Roquette P.-Contributions To The History of Number Theory in The 20th Century (2013) PDF

Peter Roquette, Oberwolfach, March 2006
Peter Roquette
Contributions
to the History of
Number Theory
in the 20th Century
Author:
Peter Roquette
Ruprecht-Karls-Universitt Heidelberg
Mathematisches Institut
Im Neuenheimer Feld 288
69120 Heidelberg
Germany
E-mail: roquette@uni-hd.de
2010 Mathematics Subject Classification (primary; secondary): 01-02, 03-03, 11-03, 12-03 , 16-03,
20-03; 01A60, 01A70, 01A75, 11E04, 11E88, 11R18 11R37, 11U10
ISBN 978-3-03719-113-2
The Swiss National Library lists this publication in The Swiss Book, the Swiss national bibliography,
and the detailed bibliographic data are available on the Internet at http://www.helveticat.ch.
This work is subject to copyright. All rights are reserved, whether the whole or part of the material is
concerned, specifically the rights of translation, reprinting, re-use of illustrations, recitation, broadcasting, reproduction on microfilms or in other ways, and storage in data banks. For any kind of use
permission of the copyright owner must be obtained.
2013 European Mathematical Society
Contact address:
European Mathematical Society Publishing House
Seminar for Applied Mathematics
ETH-Zentrum SEW A27
CH-8092 Zrich
Switzerland
Phone: +41 (0)44 632 34 36
Email: info@ems-ph.org
Homepage: www.ems-ph.org
Typeset using the authors TEX files: I. Zimmermann, Freiburg
Printing and binding: Beltz Bad Langensalza GmbH, Bad Langensalza, Germany
Printed on acid free paper
987654321
To my friend Gnther Frei who introduced me to and kindled

my interest in the history of number theory
Preface
This volume contains my articles on the history of number theory except those which
are already included in my Collected Papers. All articles have been checked and
reedited. Two articles which originally were written in German have been translated.
I would like to thank all the people who have helped me to prepare this volume,
foremost Keith Conrad and also Milena Hering who have streamlined my translation.
Particular thanks go to my wife Erika who has meticulously read and corrected
the whole manuscript.
November 2012
Peter Roquette
Contents
Preface
vii
List of figures
xi
1 The BrauerHasseNoether Theorem
2 The remarkable career of Otto Grn
77
3 At Emmy Noethers funeral
117
Emmy Noether and Hermann Weyl
129
Emmy Noether: The testimonials
163
6 Abraham Robinson and his infinitesimals
175
Cahit Arf and his invariant
189
HasseArfLanglands
223
Ernst Steinitz and abstract field theory
227
10 Heinrich-Wolfgang Leopoldt
239
11 On Hoechsmanns Theorem
245
Acknowledgements
255
Bibliography
257
Name Index
273
Subject Index
277
List of Figures
Richard Brauer, Helmut Hasse, Emmy Noether

Otto Grn
2
79
Noethers tomb
125
Emmy Noether, Hermann Weyl
146
Emmy Noether 1933
165
Abraham Robinson
176
Cahit Arf
190
Helmut Hasse, Cahit Arf, Robert Langlands
225
Ernst Steinitz
227
Heinrich-Wolfgang Leopoldt
239
Chapter 1
The BrauerHasseNoether Theorem
Revised version of:

The BrauerHasseNoether theorem in historical perspective.
Schriften der Math.-Phys. Klasse der Heidelberger Akademie der Wissenschaften Nr. 15 (2005).
1.1
1.2
1.3
1.4
1.5
1.6
1.7
1.8
1.9
Introduction
The Main Theorem: cyclic algebras
The paper: dedication to Hensel
The Local-Global Principle
From the LGP to the Main Theorem
The Brauer group and class field theory
The team: Noether, Brauer and Hasse
The American connection: Albert
Epilogue: Kte Hey
1
3
6
11
20
31
45
56
73
1.1 Introduction
The legacy of Helmut Hasse, consisting of letters, manuscripts and other papers, is
kept at the Handschriftenabteilung of the University Library at Gttingen. Hasse had
an extensive correspondence; he liked to exchange mathematical ideas, results and
methods freely with his colleagues. There are more than 8 000 documents preserved.
Although not all of them are of equal mathematical interest, searching through this
treasure can help us to assess the development of Number Theory through the 1920s
and 1930s. Unfortunately, most of the correspondence is preserved on one side only,
i.e., the letters sent to Hasse are available whereas many of the letters which had been
sent from him, often handwritten, seem to be lost. So we have to interpolate, as far as
possible, from the replies to Hasse and from other contexts, in order to find out what
he had written in his outgoing letters.1
1An exception is the correspondence between Hasse and Richard Brauer. Thanks to Prof. Fred Brauer, the
letters from Hasse to Richard Brauer are now available in Gttingen too.
The present article is largely based on the letters and other documents which I
have found concerning the
BrauerHasseNoether Theorem
in the theory of algebras; this covers the years around 1931. Besides the documents
from the Hasse and the Brauer legacy in Gttingen, I shall also use some letters from
Emmy Noether to Richard Brauer which are preserved at the Bryn Mawr College
Library (Pennsylvania, USA).
We should be aware that the BrauerHasseNoether Theorem, although to be
rated as a highlight, does not constitute the summit and end point of a development.
We have to regard it as a step, important but not final, in a development which leads
to the view of class field theory as we see it today. By concentrating on the Brauer
HasseNoether Theorem we get only what may be called a snapshot within the great
edifice of class field theory.
A snapshot is not a panoramic view. Accordingly, the reader might miss several
aspects which also could throw some light on the position of the BrauerHasse
Noether theorem, its sources and its consequences, not only within algebraic number
theory but also in other mathematical disciplines. It would have been impossible to
include all these into this paper. Thus I have decided to present it as it is now, being
aware of its shortcomings with respect to the range of topics treated, as well as the
time span taken into consideration.
R. Brauer
H. Hasse
E. Noether
A preliminary version of this article had been written in connection with my

lecture at the conference March 2224, 2001 in Stuttgart which was dedicated to
the memory of Richard Brauer on the occasion of his 100th birthday. For Brauer,
the cooperation with Noether and Hasse in this project constituted an unforgettable,
exciting experience. Let us cite from a letter he wrote many years later, on March 3,
1961, to Helmut Hasse:
ist es 35 Jahre her, da ich durch Sie mit der Klassenkrpertheorie
bekannt geworden bin. Da ich in Zusammenarbeit mit Ihnen und Emmy
Noether ein wenig dazu beitragen konnte, ist auch mir eine der schnsten
Erinnerungen, und ich werde die Aufregung der Tage, in denen die Arbeit
entstand, nie vergessen.
it is now 35 years since you introduced me to class field theory. It
belongs to my most delightful memories that I was able, in cooperation
with you and Emmy Noether, to give some little contribution, and I shall
never forget the excitement of those days when the paper took shape.
The available documents indicate that a similar feeling of excitement was present
also in the minds of the other actors in this play. Besides Hasse and Noether we have
to mention Artin and also Albert in this connection. Other names will appear in due
course.
As to A. Adrian Albert, he had an extended exchange of letters with Hasse, starting
in 1931, on the Local-Global Principle for algebras. In the paper of BrauerHasse
Noether the authors explain that and how Albert had an independent share in the
proof of the Main Theorem. Accordingly some people have suggested that perhaps
it would be justified to include his name as an author, i.e., to talk about the Albert
BrauerHasseNoether Theorem. But in this article we shall use the original name
of the theorem, i.e., without Albert, since this has become standard in the literature.
In Section 1.8 we will describe the role of Albert in the proof of the BrauerHasse
Noether theorem, based on the relevant part of the correspondence of Albert with
Hasse.
Acknowledgement. Preliminary versions had been on my homepage for some time.
I would like to express my thanks to all who cared to send me their comments each
of which I have carefully examined and taken into consideration. Moreover, I wish
to thank Falko Lorenz and Keith Conrad for their careful reading, their corrections
and valuable comments. Last but not least I would like to express my gratitude to
Mrs. Nancy Albert, daughter of A. A. Albert, for letting me share her recollections of
her father. This was particularly helpful to me while preparing Section 1.8.
1.2 The Main Theorem: cyclic algebras

On December 29, 1931 Kurt Hensel, the mathematician who had discovered padic numbers, celebrated his 70th birthday. On this occasion a special volume of
Crelles Journal was dedicated to him since he was the chief editor of Crelles Journal
at that time, and had been for almost 30 years. The dedication volume contains
the paper [BHN32], authored jointly by Richard Brauer, Helmut Hasse and Emmy
Noether, with the title:
Beweis eines Hauptsatzes in der Theorie der Algebren
Proof of a Main Theorem in the theory of algebras

The paper starts with the following sentence:
Endlich ist es unseren vereinten Bemhungen gelungen, die Richtigkeit
des folgenden Satzes zu beweisen, der fr die Strukturtheorie der Algebren ber algebraischen Zahlkrpern sowie auch darber hinaus von
grundlegender Bedeutung ist:
At last our joint endeavours have finally been successful, to prove the
following theorem which is of fundamental importance for the structure
theory of algebras over number fields, and also beyond
The theorem in question, which has become known as the BrauerHasseNoether
Theorem, reads as follows:
Hauptsatz. Jede normale Divisionsalgebra ber einem algebraischen
Zahlkrper ist zyklisch (oder, wie man man auch sagt, vom Dicksonschen
Typus).
Main Theorem. 2 Every central division algebra over a number field is
cyclic (or, as it is also said, of Dickson type).
In this connection, all algebras are assumed to be finite dimensional over a field. An
algebra A over a field K is called central if K equals the center of A. Actually, in the
original BrauerHasseNoether paper [BHN32] the word normal was used instead
of central; this had gradually come into use at that time, following the terminology of
American authors, see e.g., [Alb30]. 3 Today the more intuitive central is standard.
Cyclic algebras are defined as follows. Let LjK be a cyclic field extension, of
degree n, and let denote a generator of its Galois group G. Given any a in the
multiplicative group K , consider the K-algebra generated by L and some element
u with the defining relations:
un D a;
xu D ux
.for x 2 L/:
This is a central simple algebra of dimension n2 over K and is denoted by .LjK; ; a/.
The field L is a maximal commutative subalgebra of .LjK; ; a/. This construction
2 Falko Lorenz [Lor05] has criticized the terminology Main Theorem. Indeed, what today is seen as a Main
Theorem may in the future be looked at just as a useful lemma. So we should try to invent another name for
this theorem, perhaps Cyclicity Theorem. But for the purpose of the present article, let us keep the authors
terminology and refer to it as the Main Theorem (in capitals).
3 It seems that in 1931 the terminology normal was not yet generally accepted. For, when Hasse had sent
Noether the manuscript of their joint paper asking for her comments, she suggested that for German readers
Hasse should explain the notion of normal. (Letter of November 12, 1931.) Hasse followed her suggestion
and inserted an explanation.
of cyclic algebras had been given by Dickson; therefore they were also called of
Dickson type. 4
Thus the Main Theorem asserts that every central division algebra A over a number
field K is isomorphic to .LjK; ; a/ for a suitable cyclic extension LjK with generating automorphism , and suitable a 2 K ; equivalently, A contains a maximal
commutative subfield L which is a cyclic field extension of K.
When Artin heard of the proof of the Main Theorem he wrote to Hasse: 5
Sie knnen sich gar nicht vorstellen, wie ich mich ber den endlich
geglckten Beweis fr die cyklischen Systeme gefreut habe. Das ist
der grsste Fortschritt in der Zahlentheorie der letzten Jahre. Meinen
herzlichen Glckwunsch zu Ihrem Beweis.
You cannot imagine how ever so pleased I was about the proof, finally
successful, for the cyclic systems. This is the greatest advance in number
theory of the last years. My heartfelt congratulations for your proof.
Now, given the bare statement of the Main Theorem, Artins enthusiastic exclamation
sounds somewhat exaggerated. At first glance the theorem appears as a rather special
result. The description of central simple algebras may have been of importance, but
would it qualify for the greatest advance in number theory in the last years? It seems
that Artin had in mind not only the Main Theorem itself, but also its proof, involving
the so-called Local-Global Principle and its many consequences, in particular in class
field theory.
The authors themselves, in the first sentence of their joint paper, tell us that they
see the importance of the Main Theorem in the following two directions:
1. Structure of division algebras. The Main Theorem allows a complete classification of division algebras over a number field by means of what today are called
Hasse invariants; thereby the structure of the Brauer group of an algebraic
number field is determined. (This was elaborated in Hasses subsequent paper
[Has33a] which was dedicated to Emmy Noether on the occasion of her 50th
birthday on March 23, 1932.) The splitting fields of a division algebra can
be explicitly described by their local behavior; this is important for the representation theory of groups. (This had been the main motivation for Richard
Brauer in this project.)
2. Beyond the theory of algebras. The Main Theorem opens new vistas into one
of the most exciting areas of algebraic number theory at the time, namely
the understanding of class field theory its foundation, its structure and its
4 Dickson himself [Dic27] called these algebras of type D. Albert [Alb30] gives 1905 as the year when
Dickson had discovered this construction. Dickson did not yet use the notation .LjK; ; a/ which seems to
have been introduced by Hasse.
5 This letter from Artin to Hasse is not dated but we have reason to believe that it was written around November
11, 1931.
generalization by means of the structure of algebras. (This had been suggested

for some time by Emmy Noether. It was also Artins viewpoint when he praised
the Main Theorem in the letter we have cited above.)
We will discuss these two viewpoints in more detail in the course of this article.
1.3 The paper: dedication to Hensel

But let us first have a brief look at the dates involved. The Hensel Festband of Crelles
Journal carries the publication date of January 6, 1932. The first copy was finished
and presented to Hensel already on December 29, 1931, his birthday. 6 The Brauer
HasseNoether paper carries the date of receipt of November 11, 1931. Thus the paper
was processed and printed within less than two months. This is a remarkably short
time for processing and printing, including two times proofreading by the authors.
It seems that the authors submitted their paper in the last minute, just in time to
be included into the Hensel dedication volume. Why did the authors not submit it
earlier? After all, Hasse himself was one of the editors of Crelles Journal and so he
was informed well in advance about the plans for the Hensel dedication volume.
The answer to our question is that the authors did not find their result earlier. For we
can determine almost precisely the day when the proof of the Main Theorem had been
completed. There is a postcard from Emmy Noether to Hasse dated November 10,
1931 which starts with the following words:
Das ist schn! Und mir ganz unerwartet, so trivial der letzte Schlu ist;
der ja auch bei Brauer steht (Jede Primzahl des Index geht im Exponenten
auf.)
This is beautiful! And completely unexpected to me, notwithstanding
that the last argument, due to Brauer, is quite trivial (Every prime number
dividing the index is also a divisor of the exponent.)
This is a response to a postcard from Hasse telling her that he had found the last step
in the proof of the Main Theorem, by means of an argument which Hasse had learned
from Brauer. The theorem of Brauer which she cites in parentheses had been proved
in [Bra29b]. Of course she does not mean that Brauers theorem is trivial, but that the
application of Brauers theorem in the present situation seems trivial to her. Actually,
we shall see in Section 1.4.2 that this theorem of Brauer is not really needed but only
his Sylow argument which he had used in [Bra29b].
Only two days earlier, on November 8, 1931, Noether had sent a long letter 7 in
6 We know this because Hasse mentioned it in his laudation which he read to Hensel on the birthday reception.
See [Has32a].
7 The letter has four pages. This must be considered as long by the standard of Emmy Noether who often
scribbled her messages on postcards, using up every conceivable free space on the card.
which she congratulated Hasse for his recent proof that at least every abelian central
simple algebra A is cyclic. Here, a central simple algebra AjK is called abelian if
it admits a splitting field which is an abelian field extension of K.
But Noether did not only congratulate. In addition, she showed Hasse how to
obtain a simplification (which she called trivialization) of his proof, and at the
same time to generalize his result from abelian to solvable algebras by means of
an easy induction argument. Moreover, she gave some ideas how it may be possible
to approach the general, non-solvable case. These latter ideas were quite different
from the final solution which consisted in applying Brauers Sylow argument; this
explains her surprise which she shows in her postcard of November 10.
In those times, postal mail went quite fast. Between Marburg (where Hasse
lived) and Gttingen (Noethers place) ordinary mail was delivered the next day after
dispatch, sometimes even on the same day. 8
Thus it appears that on November 9, Hasse had received Noethers earlier letter
of November 8. While studying her proofs for the solvable case he remembered an
earlier letter of Brauer, where a Sylow argument was used to reduce the general case
to the case of a p-group which, after all, is solvable. Putting things together Hasse
saw the solution. Brauers letter had been written some days earlier, on October 29. 9
Immediately Hasse informed Emmy Noether about his finding, and so it was
possible that she received his message on November 10 and could send her reply
postcard on the same day.
Accordingly we may conclude that November 9, 1931 is to be very likely the
birthday of the BrauerHasseNoether Main Theorem, i.e., the day when the last step
in the proof had been found.
The same day Hasse informed Richard Brauer too. Just two days earlier, on
November 7, Hasse had sent a long 10 page letter to Brauer, explaining to him in
every detail his ideas for attacking the problem. He used Brauers Sylow argument
but then he said:
Leider mu ich bekennen, da ich hier am Ende meines Knnens stehe
und alle meine Hoffnungen auf Ihr Knnen setze. Es handelt sich, wie
Sie sehen um ein Faktorensystem, das zu einem nicht-galoisschen Zerfllungskrper gehrt
I have to admit that here I am at the end of my skills and I put all my hope
on yours. As you see, there is a factor system involved which belongs
to a non-galois splitting field
Since Brauer had introduced and investigated factor systems for non-galois splitting
fields [Bra26], [Bra28], it appears quite natural that Hasse turned to him for the
8 Mail
was delivered two times a day: once in the morning and a second time in the afternoon.
Brauer in his letter did not have the Main Theorem in mind but the related question whether the
index of an algebra equals its exponent, over an algebraic number field as a base.
9Actually,
solution of the problem. But two days later Hasse could send a postcard with the
following text:
Lieber Herr Brauer ! Eben bekomme ich einen Brief von Emmy, der die
ganze Frage erledigt, und zwar so, da ein Eingehen auf die Struktur
der Faktorensysteme gar nicht notwendig wird man kann nmlich
den Beweis durch schrittweise Reduktion in Primzahlschritten fhren.
Man muss nur den Abbau nicht, wie ich ungeschickt versuchte, beim
Krper unten, bei der Gruppe oben beginnen, sondern umgekehrt Ich
habe mich furchtbar geqult, und doch nicht den einfachen Gedanken
von Emmy gehabt.
Dear Mr. Brauer! Just now I receive a letter from Emmy which takes
care of the whole question, and such that it will not be necessary to
know the structure of the factor systems It is possible to get a proof
by stepwise reduction to steps of prime degree. The only thing to do is
not, as I had clumsily tried, to start the reduction process with the field
below, i.e., with the group on top, but to do it the other way I had gone
to many troubles but did not find the simple idea of Emmy.
And Hasse continued to describe Emmys idea, all on the same postcard.
Brauer lived in Knigsberg which was somewhat more distant from Marburg than
Gttingen; thus the postcard to him may have needed one day longer than that to
Noether. 10 In fact, Brauers reply to Hasse is dated November 11, one day later than
Noethers reply. He wrote:
Herzlichen Dank fr Ihren ausfhrlichen Brief und Ihre Karte, die ich
eben erhielt. Es ist sehr schn, da das Zyklizittsproblem jetzt erledigt
ist! Ich hatte Ihnen gerade heute schreiben wollen und Ihnen genau die
Methode der Emmy mitteilen wollen; allerdings mu ich offen sagen,
da ich frchtete, einen dummen Fehler dabei zu machen, weil mir
die Sache zu einfach vorkam. Ich hatte Sie gerade deswegen anfragen
wollen, was ja nun berssig geworden ist. Es war mir brigens von
vornherein klar, da durch Ihre Reduktion die wesentliche Arbeit geleistet war.
Many thanks for your detailed letter, and for your postcard which I just
received. It is very nice that the problem of cyclicity is now solved!
Just today I had meant to write you and to inform you in detail about
Emmys method; but I have to admit that I feared to make a silly mistake
because I had the feeling that the thing was too simple. I just wanted
to ask you about it, but now this is unnecessary. By the way, right from
10 From Marburg to Gttingen there are about 140 km, whereas from Marburg to Knigsberg we have counted
about 975 km.
the beginning it was clear to me that with your reduction, the essential
work had been done already.
This shows that Brauer was directly involved in finding the proof. When he
mentions Emmys method he refers to Hasses postcard where Hasse had explained
the reduction step sent to him by Emmy Noether. But in fact, Brauer had found the
same method independently and he too had realized that, if combined with his Sylow
argument, this method would give the solution. At the same time we see his modesty,
which made him claim that Hasse had done the essential work already. Two days
later when Hasse had sent him the completed manuscript, he wrote:
Heute frh erhielt ich Ihre Arbeit; ich bin ganz berrascht, da meine
doch wirklich nur geringfgige Bemerkung Sie veranlat hat, diese
besonders schne Arbeit mit unter meinem Namen zu publizieren.
Today in the morning I received your paper; I am quite surprised that my
really small remark has caused you to publish this particularly beautiful
paper jointly under my name.
Well, Brauers contribution was not confined to a small remark. On the contrary,
Hasses arguments relied heavily and substantially on Brauers general results about
division algebras and their splitting fields.
We have seen that the birthday of the Main Theorem had been November 9, but
we have also seen that the manuscript was received by the editors on November 11.
We conclude that Hasse had completed the manuscript in at most two days. Actually, it
must have been within one day because on November 11 already, Emmy Noether had
received from him the completed manuscript and wrote another letter to Hasse with
her comments. This haste is explained by the fact that the deadline for contributions
to the Hensel dedication volume had passed long ago (it was September 1, 1931) and
Hensels birthday was approaching at the end of the year already, when the volume had
to be presented to him. And Hasse was eager to put this paper, which he considered
important, into this dedication volume. Kurt Hensel had been his respected academic
teacher and now was his paternal friend (vterlicher Freund). In the introduction
of the BrauerHasseNoether paper we read:
Es ist uns eine besondere Freude, dieses Ergebnis, als einen im wesentlichen der p-adischen Methode zu dankenden Erfolg, Herrn Kurt Hensel,
dem Begrnder dieser Methode, zu seinem 70. Geburtstag vorzulegen.
It gives us particular pleasure to be able to dedicate this result, being
essentially due to the p-adic method, to the founder of this method,
Mr. Kurt Hensel, on the occasion of his 70th birthday.
Emmy Noether commented on this dedication text in her letter of November 12 to
Hasse as follows:
10
Mit der Verbeugung vor Hensel bin ich selbstverstndlich einverstanden.

Meine Methoden sind Arbeits- und Auffassungsmethoden, und daher
anonym berall eingedrungen.
Of course I agree with the bow to Hensel. My methods are workingand conceptual methods and therefore have anonymously penetrated
everywhere.
The second sentence in this comment has become famous in the Noether literature. It
puts into evidence that she was very sure about the power and success of her methods
which she describes quite to the point. But why did she write this sentence just here,
while discussing the dedication text for Hensel ? The answer which suggests itself is
that, on the one hand, Noether wishes to express to Hasse that, after all, her methods
(as distinguished from Hensels p-adic methods) were equally responsible for their
success. On the other hand she does not care whether this is publicly acknowledged
or not.
In the present context her methods means two things: First, she insists that the
classical representation theory be done within the framework of the abstract theory
of algebras (or hypercomplex systems in her terminology), instead of matrix groups
and semi-groups as Schur had started it. Second, she strongly proposes that the
non-commutative theory of algebras should be used for a better understanding of
commutative algebraic number theory, in particular class field theory.
Perhaps we may add a third aspect of her methods: the power to transmit her
ideas and concepts to the people around her. In this way she had decisively influenced
Richard Brauers and Helmut Hasses way of thinking: Brauer investigated division
algebras and Hasse did non-commutative arithmetic.
The great hurry in which the BrauerHasseNoether paper had to be written may
also account for the somewhat unconventional presentation. For, Hasse says in a
footnote that the material is presented
in der Reihenfolge ihrer Entstehung, die der systematischen Reihenfolge entgegengesetzt ist
in the order of the discovery, which is the reverse of the systematic
order
This footnote was inserted on the insistence of Noether. For, in still another letter
written 3 days later, on November 14, 1931, she had expressed her dislike of the
presentation as given by Hasse. She wrote that in this presentation the proof is difficult
to understand, and that she would have insisted on a more systematic arrangement
except that the time was too short. Therefore Hasse should at least insert a footnote to
the effect as mentioned above. And Hasse did so. He wished the paper to be included
into the Hensel volume, hence there was no time to rewrite the manuscript.
Three months later Hasse seized an opportunity to become reconciled with Emmy
Noether by dedicating a new paper [Has33a] to her, on the occasion of her 50th
11
birthday on March 23, 1932. There he deals with the same subject but written more
systematically. Those three months had seen a rapid development of the subject; in
particular Hasse was now able to give a proof of Artins Reciprocity Law of class
field theory which was based almost entirely on the theory of the Brauer group over a
number field. Thereby he could fulfill a desideratum of Emmy Noether who already
one year earlier had asked him to give a hypercomplex foundation of the reciprocity
law. In the preface to that paper Hasse bows to Emmy Noether as an invaluable
source of inspiration.
Section II.6 of that paper [Has33a] contains a new presentation of the Main Theorem. Hasse starts this section by admitting that in the earlier joint paper [BHN32]
the proof had been presented in a somewhat awkward manner, according to the order
of its discovery. Now, he says, he will give the proof (which is the same proof after
all) in a more systematic way. Clearly, this is to be viewed as a response to Noethers
criticism in her letter of November 14.
By the way, three days after her birthday Emmy Noether replied to this present:
I was terribly delighted! (Ich habe mich schrecklich gefreut! ). There follow
two pages of detailed comments to Hasses paper, showing that she had studied it
already in detail.
1.4 The Local-Global Principle

Let K be an algebraic number field of finite degree. For every prime p of K, finite or
infinite, let Kp denote the p-adic completion of K. For an algebra A over K we put
Ap D A K Kp , the completion (also called localization) of A at p. An important
step in the proof of the Main Theorem is the celebrated
Local-Global Principle for algebras. Let K be a number field and
AjK be a central simple algebra. If Ap jKp splits for every p then AjK
splits.
Here, splitting of AjK means that A is a full matrix algebra over K. Note that
the Local-Global Principle is formulated for simple algebras, not only for division
algebras as the Main Theorem had been. Quite generally, it is more convenient to work
with simple algebras and, accordingly, formulate and prove the Main Theorem for
simple algebras instead of division algebras only. By Wedderburns theorem, every
simple algebra AjK is isomorphic to a full matrix ring over a division algebra DjK,
and D is uniquely determined by A up to isomorphisms. Two central simple algebras
over K are called similar if their corresponding division algebras are isomorphic.
We shall discuss in Section 1.5 how the Local-Global Principle was used in the
proof of the Main Theorem. In the present section we review the long way which
finally led to the conception and the proof of the Local-Global Principle.
12
1.4.1 The Norm Theorem

First, consider a cyclic algebra A D .LjK; ; a/ as explained in Section 1.2. Such an
algebra splits if and only if a is a norm from the cyclic extension LjK. Accordingly,
the Local-Global Principle for cyclic algebras can be reformulated as follows:
HilbertFurtwnglerHasse Norm Theorem. Let LjK be a cyclic
extension of number fields, and let 0 a 2 K. If a is a norm in the
completion Lp jKp for every p then a is a norm in LjK. 11
This theorem does not refer to algebras, it concerns algebraic number fields only.
Now, in the case when the degree n of LjK is a prime number, the Norm Theorem
was known for a long time already, in the context of the reciprocity law of class
field theory. It had been included in Hasses class field report, Part II [Has30a]
where Hasse mentioned that it had first been proved by Furtwngler in [Fur02] and
subsequent papers. For quadratic fields .n D 2/ the Norm Theorem had been given
by Hilbert in his Zahlbericht [Hil97]. In March 1931 Hasse succeeded to generalize
this statement to arbitrary cyclic extensions LjK of number fields, not necessarily of
prime degree; see Section 1.7. He published this in [Has31a], April 1931.
Now, the Main Theorem tells us that every central simple algebra over a number
field is cyclic, so we could conclude that the Local-Global Principle holds generally,
for every central simple algebra over a number field. However, in order to prove
the Main Theorem, Hasse needed first to prove the Local-Global Principle generally,
regardless of whether the given algebra is already known to be cyclic or not. Hence
there arose quite naturally the problem how to reduce the general case of the LocalGlobal Principle to the case when the algebra is cyclic.
1.4.2 The reductions
In the BrauerHasseNoether paper [BHN32] this reduction is done in two steps:
(2) Reduction to the case when A has a solvable splitting field.
(3) Further reduction (by induction) to the case when A has a cyclic
splitting field.
Here and in the following we use the same enumeration of these reductions which
is used in the BrauerHasseNoether paper. There is another reduction, called reduction (1), which reduces the Main Theorem to the Local-Global Principle. That
we will discuss in Section 1.5 and the following sections, as the contribution of Hasse.
accordance with the definition of Ap one would define Lp D L K Kp . In general this is not a field
but the direct sum of fields LP where P ranges over the primes of L dividing p. If LjK is a Galois extension
(in particular if it is cyclic) then all these fields LP are isomorphic over Kp , and a is a norm from Lp if and
only if it is a norm from LP for some and hence all P. Usually, one chooses one prime Pjp and writes Lp for
LP (thus forgetting the former, systematic notation for Lp ). Let us do this here too.
11 In
13
The reduction (2) is due to Brauer who, in his letter to Hasse of October 29, 1931,
had provided a Sylow argument for this purpose. Reduction (3) had been provided
by Noether in her letter to Hasse of November 8, 1931.
Brauer had developed the theory of division algebras and matrix algebras in a series
of several papers in the foregoing years, starting from his 1927 Habilitationsschrift
at the University of Knigsberg [Bra28]. His main interest was in the theory of
group representations, following the ideas of his academic teacher I. Schur. It was
Emmy Noether who gradually had convinced him that the representation theory of
groups could and should be profitably discussed within the framework of algebras.
In Brauers papers, in particular in [Bra29b], we find the following theorems. Brauer
had reported on these theorems in September 1928 at the annual meeting of the DMV
(Deutsche Mathematiker Vereinigung) in Hamburg; see [Bra29a]. Although in that
report no proofs are given, we can recommend consulting it since Brauers theorems
are very clearly stated there.
Brauers theorems
(i) The similarity classes of central simple algebras over a field K form
a group with multiplication well-defined by the tensor product A K B
of two algebras. 12
Today this group is called the Brauer group of K and denoted by Br.K/. The name
Brauer group was given by Hasse in [Has33a]. The split algebras belong to the
neutral element of the Brauer group.
(ii) Every central simple algebra A over K has finite order in Br.K/.
This order is called the exponent of A. This terminology had been chosen by Brauer
because, he said, in the context of the theory of algebras the word order is used for
another concept. 13
(iii) The exponent of A divides the index m of A.
The index m of A is defined as follows: Let D be the division algebra similar to
A. The dimension of D over its center is a square m2 , and this m is the index by
definition.
(iv) Every prime number dividing the index of A also divides its exponent.
Brauer had used these theorems (i)(iv) in order to show:
12At the time of BrauerHasseNoether, the tensor product was called direct product and denoted by AB.
Brauer considered only perfect base fields K; it was Emmy Noether who in [Noe29] was able to wave the
hypothesis that K is perfect.
13 I am indebted to Falko Lorenz who pointed out to me that this theorem (ii) is contained in Schurs paper
[Sch19] already, as well as theorem (iii) if m is interpreted suitably. See [Lor98].
14
(v) Every division algebra A of index m can be decomposed as the tensor

product of division algebras Ai of prime
index pi i , according to
Q power
i
the prime power decomposition m D i pi of the index.
In Hasses first draft of the joint manuscript which he had sent to Emmy Noether,
these theorems were used. Although we do not know this first draft, we can conclude
this from the following: First, in Noethers reply postcard of November 10 (which we
have cited in Section 1.3) Brauers theorem (iv) is mentioned. Secondly, in a letter of
Hasse to Brauer dated November 11, Hasse reports that Noether had finally thrown
out the reduction (v) to prime power index, because that was superfluous. And so
Hasse continues:
Daher fand ich auch eigentlich nicht genug Gelegenheit Ihre Arbeit
aus Math. Zeitschr. zu zitieren. Es wird fast nichts daraus gebraucht,
auer den einfachsten schon vorher feststehenden Tatsachen ber Zerfllungskrper.
For that reason I did not find a suitable occasion to cite your paper of
Mathematische Zeitschrift. Almost nothing from there is needed, except
the most simple facts on splitting fields.
Here Hasse refers to Brauers paper [Bra29b]. These most simple facts which
are used in the final proof are the following:
(vi) The degree of every splitting field of A over K is divisible by the
index m of A, and there exist splitting fields of degree m.
Using this, the reduction steps (2) and (3) are quite easy if combined with the
functorial properties of the Brauer group. Let us briefly present the arguments. Our
presentation is the same as Noether had proposed it in her letter to Hasse of November 10, 1931, and which Hasse then used in his Noether dedication paper [Has33a].
If K L then we use the notation AL D A K L. If we regard A and AL in
their respective Brauer groups Br.K/ and Br.L/ then the map A 7! AL defines a
canonical group homomorphism Br.K/ ! Br.L/.
Let K be a number field and AjK a central simple algebra which splits everywhere
locally. The claim is that A splits. Suppose A does not split and let m > 1 be the
index of A. Let p be a prime number dividing m. Consider a Galois splitting field
LjK of A, so that AL splits; then p divides L W K. Let G be the Galois group of
LjK. Consider a corresponding Sylow p-group of G and let L0 L denote its fixed
field. Since the Sylow p-group is solvable there exists a chain of fields
L0 L1 Ls1 Ls D L
such that each Li jLi1 is cyclic of degree p .1 i s/. Since A splits everywhere
locally, so does every ALi . Now, ALs1 has Ls D L as a cyclic splitting field.
15
Hence the Norm Theorem implies that ALs1 splits. Therefore ALs2 has Ls1 as a
cyclic splitting field, hence again, ALs2 splits. And so on by induction. Finally we
conclude that AL0 splits. Thus A admits the splitting field L0 whose degree L0 W K
is relatively prime to p. But L0 W K is divisible by the index m which contains p
as a prime divisor. Contradiction.
Since each Li jLi1 is of degree p, it is evident that Hasses Norm Theorem has to
be used only in the case of cyclic fields of prime degree p, i.e., the original Hilbert
Furtwngler Theorem is sufficient. Hasses generalization to arbitrary cyclic fields
is not needed and is a consequence of Noethers induction argument. This had been
immediately observed by Noether (letter of November 8, 1931), and she had asked
Hasse to mention it in their joint paper (which he did). At that time this observation
indeed could be considered a simplification. But half a year later, in [Has33a], Hasse
remarked that this would not make a difference any more because in the meantime
new proofs of the Norm Theorem had been found by Chevalley and Herbrand, and
those proofs work equally well for arbitrary cyclic extensions (using the so-called
Herbrands Lemma) regardless of whether the degree is prime or not. 14
From todays viewpoint the above proof of the Local-Global Principle looks rather
trivial, once the HilbertFurtwngler Norm Theorem is accepted. In particular if the
arguments are given in the language of cohomology, as it is usually done nowadays,
we see that only the very basic properties of the cohomological restriction map are
used. This seems to justify Brauers words, cited above, that right from the beginning
it was clear to me [Brauer] that with your Hasses reduction, the essential work had
been done already. But these words are valid only if, firstly, Brauers fundamental
theorems are accepted and, secondly, there had already developed a certain routine
for using those theorems for particular problems. While the first was certainly the
case within the circle around Brauer, the second was not yet. Otherwise, the simple
proof above could well have been given much earlier.
We should not underestimate the conceptual difficulty which people had at that
time working with algebras and their splitting fields, and the notions of index and
exponent of algebras. There was no established routine to work with the functorial
properties of Brauer groups. Based on the cited work of Brauer and, in parallel, on
the monumental work of Emmy Noether [Noe29] such routine came gradually into
being.
1.4.3 Factor systems
The idea for a proof like the above, required in the first place some insight into the
relevant structures, in particular the interpretation of the Norm Theorem as a splitting
theorem for cyclic algebras. Only under this aspect it makes sense to generalize it
from the cyclic to the general case. In fact, Hasse originally did not do this step. In his
14 The HerbrandChevalley proof was included in Hasses Marburg lectures 1932 on class field theory. See
[Has33b], Satz (113).
16
class field report Part II [Has30a] he had conjectured that the Norm Theorem holds
for arbitrary abelian extension of number fields. But in [Has31a] he had to admit that
for non-cyclic extensions the Norm Theorem fails to hold.
It was Emmy Noether who then suggested to Hasse that the generalization of
the Norm Theorem would require considering algebras instead of norms, the latter
representing split cyclic algebras. This is evidenced by the following excerpt from
her letter of November 12, 1931. In that letter she wished to have some further
changes in the manuscript of the joint BrauerHasseNoether paper, for which Hasse
had composed the draft. She wrote:
Ebenso mchte ich auf S. 4, im 4.-letzten Absatz, mitgenannt sein, oder
etwa das H. Hasse durch wir ersetzt haben. Da nmlich die Fassung
mit den Faktorensystemen die richtige Verallgemeinerung ist, habe ich
Ihnen schon auf dem Hanstein-Spaziergang im Frhling gesagt, als Sie
mir die Widerlegung der Norm-Vermutung im Abelschen Fall erzhlten.
Sie haben es damals wahrscheinlich noch nicht ganz aufgefat; und es
sich spter selbst wieder berlegt. Genau genommen habe ich es Ihnen
schon in Nidden gesagt.
Similarly, I would like to be mentioned too on page 4, in the 4th paragraph
from below, or maybe the H. Hasse should be replaced by we 15 . For,
I have mentioned to you already in the spring on our Hanstein-walk 16
that the version with factor systems is the correct generalization, after
you had told me the refutation of the norm conjecture in the abelian
case. Perhaps you had not yet fully grasped it at the time, and later you
have come to the same conclusion by yourself. Strictly speaking I had
mentioned this to you already in Nidden. 17
We observe that Noether talks about factor systems and not about algebras. Factor
systems are used to construct algebras. Given any finite separable field extension LjK
let Br.LjK/ denote the kernel of the map Br.K/ ! Br.L/, consisting of those central
simple algebras over K (modulo similarity) which are split by L. Brauer had shown
that Br.LjK/ is isomorphic to the group of what he called factor systems (modulo
equivalence). A factor system consists of certain elements in the Galois closure of
LjK, and it can be used to construct a central simple algebra AjK such that the
elements of the factor system appear as factors in the defining relations of a suitable
15 In the printed version, this is the last paragraph of section 4. There indeed we find the word we as Noether
requested.
16 Hanstein is a hillside near Gttingen. It appears that in the spring of 1931, on one of the many visits of
Hasse to Gttingen, they had made a joint excursion to the Hanstein.
17 Nidden at that time was a small fishermans village in East Prussia, located on a peninsula (Kurische
Nehrung) in the Baltic sea and famed for its extended white sand dunes. In September 1930, Hasse and Noether
both attended the annual meeting of the DMV at Knigsberg in East Prussia, and after the meeting they visited
Nidden.
17
basis of the algebra. Brauers invention of factor systems was essential for the proof
of his theorems.
It is true that the appearance of factor systems had been observed earlier already by
Schur and also by Dickson. But it was Brauer who defined and used them systematically to construct algebras, thereby writing down explicitly the so-called associativity
relations.
We will not give here the explicit definition of factor systems in the sense of
Brauer. For, today one mostly uses in this context the simplified form which Noether
has given to Brauers factor systems. Noether considered Galois splitting fields LjK
only. Let G D G.LjK/ denote its Galois group. Consider the K-algebra A which is
generated by L and by elements u ( 2 G) with the defining relations
u u D u a; ;
xu D u x .for x 2 L/
where ; 2 G and a; 2 K . It is required that the factors a; satisfy the
following relations which are called associativity relations:
%
a; % a; % D a; % a;
:
Sometimes they are also called Noether equations. The algebra A thus defined is a
central simple algebra over K which has L as a maximal commutative subfield. A
is called the crossed product of L with its Galois group G, and with factor system
a D .a; /. 18 Notation: A D .LjK; a/. Every central simple algebra over K which
admits L as a splitting field can be represented, up to similarity, as a crossed product
in this sense. If G is cyclic then (by appropriate choice of the u ) we obtain the cyclic
algebras in this way.
This theory of factor systems was developed by Emmy Noether in her Gttingen
lecture 1929/30. But Noether herself never published her theory. Deuring took notes
of that lecture, and these were distributed among interested people; Brauer as well
as Hasse had obtained a copy of those notes. (The Deuring notes are now included
in Noethers Collected Papers.) The first publication of Noethers theory of crossed
products was given, with Noethers permission, in Hasses American paper [Has32c]
where a whole chapter is devoted to it. The theory was also included in the book
Algebren by Deuring [Deu35].
A factor system a; is said to split if there exist elements c 2 L such that
a; D
c c
:
c
18 The German terminology is verschrnktes Produkt. The English term crossed product had been used by
Hasse in his American paper [Has32c]. When Noether read this she wrote to Hasse: Are the crossed products
your English invention? This word is good. We do not know whether Hasse himself invented this terminology,
or perhaps it was H. T. Engstrom, the American mathematician who helped Hasse to translate his manuscript
from German into English. In any case, in the English language the terminology crossed product has been in
use since then.
18
The split factor systems are those whose crossed product algebra .LjK; a/ splits. Today we view the group of factor systems modulo split ones as the second cohomology
group of the Galois group G of LjK in the multiplicative G-module L . The notation
is H 2 .G; L / or better H 2 .LjK/.
Thus the BrauerNoether theory of crossed products yields an isomorphism
Br.LjK/ H 2 .LjK/
which has turned out to be basic for Brauers theory.
In mathematics we often observe that a particular object can be looked at from different points of view. A change of viewpoint may sometimes generate new analogies,
thereby we may see that certain methods had been successfully applied in similarly
looking situations and we try to use those methods, suitably modified, to deal also with
the problem at hand. This indeed can lead the way to new discoveries. But sometimes
it can also hamper the way because the chosen analogies create difficulties which are
inessential to the original problem.
We can observe such a situation in Hasses first attempts to deal with the LocalGlobal Principle for algebras. Instead of dealing with algebras directly he considered,
following Noethers suggestion, factor systems. Given a factor system in H 2 .LjK/
which splits locally everywhere, he tried to transform it in such a way that its global
splitting is evident. This then reduces to the solution of certain diophantine equations
in L under the hypothesis that those equations can be solved locally everywhere.
Now it is well known, and it was of course known to Hasse that the local solution
of diophantine equations does not in general imply their global solution. But several
years earlier Hasse had already proved one instance of a Local-Global Principle for
certain diophantine equations, namely quadratic equations. In case of the rational
field Q as base field this had been the subject of Hasses dissertation (Ph.D. thesis)
in 1921, and in subsequent papers [Has24b], [Has24a] he solved the same problem
for an arbitrary number field K as base field. 19
Accordingly, Hasse tried first to invoke the analogy to the theory of quadratic
forms in order to approach the Local-Global Principle for algebras. However, it
turned out that this created difficulties which only later were seen not to be inherent
to the problem.
We are able to follow Hasses ideas for these first attempts (which later were
discarded as unnecessary) since Hasse had written to some of his friends explaining
19 With this result Hasse had solved, at least partially, one of the famous Hilbert problems. The 13th Hilbert
problem calls for solving a given quadratic equation with algebraic numerical coefficients in any number of
variables by integral or fractional numbers belonging to the algebraic realm of rationality determined by the
coefficients. Hilberts wording admits two interpretations. One of them is to regard the phrase integer or
fractional numbers as denoting arbitrary numbers of the number field in question. In this interpretation Hasse
could be said to have solved the problem completely. The other interpretation is that Hilbert actually meant two
different problems: The first is to solve the quadratic equation in integers of the field, and the second is to admit
solutions with arbitrary numbers of the field. In this interpretation, which would generalize Minkowskis work
from the rationals to arbitrary number fields, Hasse would have solved only the second of the two problems. The
first problem (solution in integers) has been studied by Siegel and others.
19
these ideas, obviously in the hope that someone would be able to supply the final
clue. One of those letters, the one to Brauer dated July 27, 1931, is preserved. Hasse
writes:
Ich mchte Ihnen gerne schreiben, wie die Sachlage nun mit der einzigen noch offenen Frage nach der Zyklizitt aller normalen einfachen
Algebren steht. Ich glaube nmlich, da diese Frage jetzt angriffsreif
ist, und mchte Ihnen die mir vorschwebende Angriffslinie vorlegen.
I would like to write to you about the only question which is still open, the
question whether all central 20 simple algebras are cyclic. For I believe
that this question is now ripe and I would like to present to you the line
of attack which I have in mind.
(In this connection Hasse means algebras over an algebraic number field as base field,
although he does not explicitly mention this.)
Hasse continues that, following his line of attack, he is trying to use his LocalGlobal Principle for quadratic forms. Let wi be a basis of the given central simple
algebra A over K. The trace matrix tr.wi wk / defines a quadratic form. If A splits
locally everywhere then for every prime p there exists a basis transformation which
transforms the given basis into a system of matrix units, and this defines a certain
transformation of the quadratic form. The Local-Global Principle for quadratic forms
then yields a certain basis transformation over the global field K. Hasse asks whether
it is possible to deduce the splitting of A from the special structure of this transformed
trace form. In other words, one has to construct from it a system of matrix units over
K. But Hasse does not yet know how to do this, not even whether it is possible at all.
He writes to Brauer:
Ich mchte diese Sache zur berprfung nach diesem Gesichtspunkt in
Ihre kundigen Hnde legen.
I would like to put this problem into your hands for examination from
this viewpoint.
Several days later, on August 3, 1931 Brauer replied that at present he is not able to
say anything about Hasses problem, and that first he has to study it in detail. But
relying on Hasses own creative power he adds:
Ich hoffe wenigstens so weit zu sein, da ich alles verstehen kann, wenn
Sie selbst die Lcke ausgefllt haben werden.
I hope to be able to understand all these things at the time when you will
have filled the gap yourself.
20 Hasse writes normal instead of central. For the convenience of the reader we will replace normal in
this context by the modern central, here and also in other citations which follow.
20
Hasse had sent similar letters to Artin and Noether. These letters are not preserved but
we know the respective answers. Artin, returning from a vacation in the mountains
wrote on August 24, 1931:
Inzwischen haben Sie sicher den Satz ber Schiefkrper bewiesen. Ich
bin schon sehr gespannt darauf.
Meanwhile you will certainly have proved the theorem on division
algebras. I am looking forward to it.
Noether wrote on the same day:
Natrlich kann ich Ihre Frage auch nicht beantworten ich glaube man
mu so etwas liegen lassen bis man von anderer Seite selbst darauf
stt
Naturally, I too cannot answer your question I believe one should
leave such things alone until one meets them again from another point
of view
But she adds some remarks about the work of Levitzky (her Ph.D. student) who
provided some methods to construct bases of split algebras.
These answers do not sound as if they had been very helpful to Hasse. But he did
not give up so easily. After a while he managed to prove the Local-Global Principle
for those algebras A which admit an abelian splitting field LjK. We do not know
this proof but from Noethers reaction we can infer that indeed Hasse had explicitly
constructed, by induction, a split factor system for the algebra. We have already
mentioned in Section 1.3 (p. 7) Noethers reaction to Hasses letter; the Noether letter
was dated November 8, 1931 and gave a simplification and generalization of Hasses
result to algebras which admit a solvable splitting field, not necessary abelian.
From then on things began to develop rapidly as we have explained in Section 1.3,
and one day later the proof of the Local-Global Principle was complete.
As a side remark we mention that Hasse in his letter to Brauer of November 16,
1931 states that when Noethers postcard arrived on November 9 he had essentially
been through with his complicated proof. But, as we have seen, he immediately
threw away his complicated proof in favor of Noethers trivialization.
1.5 From the LGP to the Main Theorem

Sometimes the Local-Global Principle is considered the most important result of the
BrauerHasseNoether paper while the Main Theorem is rated as just one of the
many consequences of it. But the authors themselves present the Main Theorem as
their key result. We now discuss the step from the Local-Global Principle to the
21
Main Theorem. This is the reduction (1) in the count of the BrauerHasseNoether
paper, and it is due to Hasse.
1.5.1 The Splitting Criterion
Let A be a central simple algebra over a number field K. Then A W K is a square;
let A W K D n2 with n 2 N. It is known that n is a multiple of the index m of A. In
order to show that A is cyclic one has to construct a cyclic splitting field LjK of A
of degree L W K D n. To this end one needs a criterion for a finite extension field
L of K to be a splitting field of A.
According to the Local-Global Principle the problem can be shifted to the local
completions, namely:
A is split by L if and only if each Ap is split by LP for Pjp.
In the local case, there is a simple criterion for splitting fields:
Local Splitting Criterion. Ap is split by LP if and only if the degree
LP W Kp is divisible by the index mp of Ap .
Thus the Local-Global Principle yields:
Global Splitting Criterion. A is split by L if and only if for each prime
p of K and each P dividing p the local degree LP W Kp is divisible by
the local index mp of Ap .
If LjK is a Galois extension then for all primes P dividing p the completions LP
coincide; they may be denoted by Lp according to the notation explained in footnote 11.
The local criterion was essentially contained in Hasses seminal Annalen paper
[Has31d] on the structure of division algebras over local fields. But the criterion
was not explicitly stated there. Therefore Hasse in their joint paper [BHN32] gave a
detailed proof of the criterion, based on the main results of [Has31d]. But again it
was not explicitly stated; instead, the statement and proof was embedded in the proof
of the global criterion which was Satz 3 in the joint paper.
So the local criterion, although it is one of the basic foundations on which the
Main Theorem rests, remained somewhat hidden in the BrauerHasseNoether paper
another sign that the preparation of the manuscript was done in great haste. It was
so well hidden that even five months later Emmy Noether was not aware that its proof
was contained in the paper of which she was a co-author after all. In her letter of
April 27, 1932 she wrote, referring to a recent paper of Kthe: 21
21 Gottfried Kthe (19051989) was a young post-doc who in 1928/1929 came to Gttingen to study mainly
with Emmy Noether and van der Waerden. Later he switched to functional analysis under the influence of
Toeplitz.
22
Tatschlich zeigt Kthe mit seinem Invariantensatz ja direkt, da im

p-adischen die Gradbedingung auch hinreichend ist fr Zerfllungskrper
In fact, Kthe with his theorem on invariants shows directly that in the
p-adic case the degree condition is also sufficient for splitting fields
With invariants are meant what today are called the Hasse invariants of central
simple algebras over a local field Kp (see Section 1.6.1, p. 35 below). Kthes theorem
in [Kt33] describes the effect of a base field extension to these invariants. If the base
field Kp is extended to a finite extension Lp then, according to Kthes theorem, the
Hasse invariant of the extended algebra ALp is obtained from the Hasse invariant of
Ap by multiplication with the field degree Lp W Kp . This implies the local splitting
criterion.
One week later Noether admitted that she had overlooked Hasses proof in the
joint paper [BHN32]. Obviously responding to a reproach of Hasse she wrote:
Als ich nun Kthe in die Hand bekam, el mir als erstes auf, da jetzt
diese alte Frage ja beantwortet ist. Bei Ihnen hatte ich drber weggelesen;
oder was wahrscheinlicher ist, ich dachte an meinen alten Beweis und
habe bei Ihnen im wesentlichen berogen.
When I got Kthes paper it occurred to me that now this old question
was settled. In yours I had overlooked it; or, what is more likely, I
thought about my old proof and had only skimmed through yours. 22
Now let us return to the Global Splitting Criterion. Its degree conditions are
non-trivial only for the primes p for which the local index mp > 1. For a given
central simple algebra there are only finitely many such primes. This is by no means
trivial; it had been proved by Hasse in [Has31d] where he showed that the reduced
discriminant (Grundideal) of a maximal order of A contains p to the exponent
mp 1. We conclude that the existence of a cyclic splitting field LjK of degree n
for A is equivalent to the following general
Existence Theorem. Let K be an algebraic number field and S a
finite set of primes of K. For each p 2 S let there be given a number
mp 2 N.23 Moreover, let n 2 N be a common multiple of the mp s.
Then there exists a cyclic field extension LjK of degree n such that for
each p 2 S the local degree Lp W Kp is a multiple of mp .
22 Since Noether had wished to inform Hasse about Kthes results it seems that she did not know (or not
remember) that Kthes paper [Kt33] was written largely under the influence and the guidance of Hasse. This is
expressed by Kthe in a footnote to his paper which appeared in the Mathematische Annalen right after Hasses
[Has33a].
23 For infinite primes the usual restrictions should be observed: If p is real then m D 1 or 2; if p is complex
p
then mp D 1. This guarantees that in any case mp is the index of some central simple algebra over Kp .
23
This then settles the Main Theorem.

The Existence Theorem as such does not refer to algebras. It belongs to algebraic
number theory. We shall discuss the theorem and its history in the next sections.
1.5.2 An unproven theorem
A proof of the Existence Theorem had been outlined in a letter of Hasse to Albert
written in April 1931. This is reported in the paper [AH32]. But the proof is not
given and not even outlined in [AH32]. Hasse did not publish a proof of his existence
theorem, not in the joint paper [BHN32] and not elsewhere. Why not? After all, the
existence theorem is an indispensable link in the chain of arguments leading to the
proof of the Main Theorem. Without it, the proof of the Main Theorem would be
incomplete. Now, in a footnote in [AH32] we read:
The existence theorem is a generalization of those in Hasses papers
[Has26c], [Has26b] and a complete proof will be published elsewhere.
This remark gives us a clue why Hasse may have hesitated to publish his proof
prematurely. He regarded his existence theorem as an integral part of number theory
and was looking for the most general such theorem, independently of its application
to the proof of the Main Theorem. We shall see that Grunwald, a Ph.D. student of
Hasse, provided such a very general theorem. This then leads to the GrunwaldWang
story.
The story begins with a reference which Hasse had inserted in the BrauerHasse
Noether paper [BHN32] for a possible proof of the Existence Theorem. This reference
reads: [vgl.d.Anm.zu H,17Bb]. This somewhat cryptical reference can be decoded
as: compare the footnote in the paper H, section 17, Proof of (17.5) part B, subsection (b). The code H refers to Hasses American paper [Has32c] on cyclic algebras.
That paper had not yet appeared at the time when he wrote down the manuscript for
the BrauerHasseNoether paper, hence he could not give a page number. We have
checked that the page number is 205. But there, in the said footnote of [Has32c] it
is merely stated: The existence of such a field will be proved in another place. 24
This does not sound very helpful to the reader.
Let us check the next paper of Hasse [Has33a]. This is the one which he had
dedicated to Emmy Noether and in which, among other topics, he repeats the proof
of the Main Theorem more systematically. There he says at the corresponding spot
on page 749:
Ein solches hinreichend scharfes Existenztheorem hat inzwischen Engstrm [1] bewiesen. Auch ergibt sich ein solches, wohl in grtmglicher
Allgemeinheit, aus der krzlich erschienenen Dissertation von Grunwald
[1]; siehe Grunwald [2].
24 The footnote continues to announce that this existence theorem will be another one in a series of former
existence theorems proved by Hasse same remark as we had already seen above in the paper [AH32].
24
Such a suffiently strong existence theorem has been proved recently by

Engstrm [1]. Alternatively, it is possible to deduce such a theorem,
probably in its greatest possible generality, from the thesis of Grunwald
[1] which has recently appeared; see Grunwald [2].
Checking the bibliography of [Has33a] we find under Engstrom [1] the entry:
Publication in an American journal in preparation. However we were not able to
find, either in an American journal or elsewhere, any publication of H. T. Engstrom
where this or a similar theorem is proved.
Howard T. Engstrm was a young American postdoc from Yale who had stayed
in Gttingen for the academic year 1931. Through Emmy Noether he got in contact
with Hasse. He had helped translating Hasses American paper [Has32c] into English.
Emmy Noether wrote about him in a letter of June 2, 1931:
Engstrm war mit Ihrem Englisch, bis auf die Umstellungen, sehr zufrieden; hoffentlich werden Sie es auch mit seinem Existenztheorem sein
knnen! Er ist berhaupt sehr begeistert von allem, was er in Deutschland
gelernt hat. Ich schicke Engstrms Manuskript an Deuring, der schon
lange ungeduldig darauf ist
Engstrm was quite satisfied with your English, apart from the rearrangements; hopefully you will be satisfied with his existence theorems
too! He is really very enthusiastic about everything which he had learned
in Germany. I am sending Engstrms manuscript to Deuring who for a
long time is waiting impatiently for it
It appears that Hasse had proposed to Engstrm to write up the proof of the Existence
Theorem according to his (Hasses) outline, and that Deuring was to check Engstroms
manuscript.
But Engstrm did not complete his manuscript before he returned to Yale. We
have found a letter from Engstrm to Hasse, dated February 27, 1932 from Yale,
where he apologizes that he has not finished the manuscript on existence theorems as
yet. He concludes:
Your outline indicates to me that you have expended considerable thought
on the matter, and that it would really require not much effort on your
part to write it up for publication. If this is the case please dont hesitate
to do so .
We get the impression that Deuring had found a flaw in Engstrms manuscript
and that Hasse had given Engstrm some hints how to overcome the difficulty. But
at Yale Engstrm was absorbed by different duties and, hence, returned the subject to
Hasse.
It remains to check Grunwald, the second reference which was mentioned by
Hasse in [Has33a].
25
1.5.3 The GrunwaldWang story

Grunwald had been a Ph.D. student with Hasse at the University of Halle, and he
had followed Hasse to Marburg in 1930. The reference Grunwald [1] in Hasses
paper refers to Grunwalds thesis [Gru32] which appeared 1932 in the Mathematische
Annalen. The subject of the thesis belongs to the fundamentals of algebraic number
theory; from todays viewpoint it can be viewed as a first attempt to understand the
role of Heckes Grencharaktere in class field theory. Grunwalds thesis does not
contain the Existence Theorem, but Hasse discovered that Grunwalds methods could
be used to obtain a proof of the theorem. From the correspondence between Grunwald
and Hasse (which is preserved) we can infer that Hasse had proposed to Grunwald to
extract from his thesis a proof of the Existence Theorem and publish it in a separate
paper.
And Grunwald did so. The reference Grunwald [2] in [Has33a] refers to Grunwalds second paper, at that time still forthcoming, which appeared 1933 in Crelles
Journal [Gru33]. There Grunwald proved a general existence theorem which became
known as Grunwalds theorem. This theorem is much stronger than Hasses Existence Theorem:
Grunwalds theorem. Let K be an algebraic number field and S a
finite set of primes of K. For each p 2 S let there be given a cyclic
field extension Lp jKp . Moreover, let n 2 N be a common multiple of
the degrees Lp W Kp . Then there exists a cyclic field extension LjK
of degree n such that for each p 2 S its completion coincides with the
given fields Lp .
Whereas Hasse needed only the fact that the local degrees Lp W Kp should be
multiples of the given numbers mp (for p 2 S), Grunwalds theorem claims that even
the local fields Lp themselves can be prescribed as cyclic extensions of degree mp of
Kp (for the finitely many primes p 2 S ). This was a beautiful and strong theorem,
and clearly it settled the question. 25
The proof of Grunwalds theorem used class field theory and was considered to
be quite difficult. In 1942 a simplified proof was given by Whaples [Wha42]; it also
used class field theory but no analytic number theory which had still been necessary
at the time of Grunwald.
In the year 1948 Artin, who was at Princeton University at that time, conducted a
seminar on class field theory. One of the seminar talks was devoted to Whaples new
proof of Grunwalds theorem. Here is what happened in the seminar, told by one of
the participants, John Tate 26 :
25 Wilhelm
Grunwald (19091989) did not continue to work in Mathematics but decided to become a science
librarian. He finally advanced to the position of director of the renowned Gttingen University Library but he
always preserved his love for Mathematics, in particular Number Theory. He kept contact with Hasse throughout
his life.
26 in a personal letter to the author.
26
I had just switched from physics to math, and tried to follow it [the
seminar] as best I could. Wang also attended that seminar. In the spring
of 1948, Bill Mills, one of the students Artin had brought with him from
Indiana, talked on Grunwalds Theorem in the seminar. Some days
later I was with Artin in his office when Wang appeared. He said he
had a counterexample to a lemma which had been used in the proof.
An hour or two later, he produced a counterexample to the Theorem
itself Of course he [Artin] was astonished, as were all of us students,
that a famous theorem with two published proofs, one of which we had
all heard in the seminar without our noticing anything, could be wrong.
But it was a good lesson!
The error was not contained in Grunwalds paper [Gru33] itself but in Grunwalds
thesis [Gru32] from which the author cited a lemma. That lemma referred to a prime
number p but the author did not see that the prime p D 2 needed special care when
compared with the odd primes p > 2.
The fact that there was an error in Grunwalds (as well as in Whaples) theorem
caused a great stir among the people concerned. Would this mean that the Main
Theorem of BrauerHasseNoether was wrong too?
Fortunately, the situation was not that serious. In most cases Grunwalds theorem holds, and exceptions can only occur if n is divisible by 8. Also, Hasses Existence
Theorem is much weaker than Grunwalds and it turned out that this weaker theorem
holds in any case, including the cases where the full Grunwald theorem collapses.
This was established by Wang in his Ph. D. thesis [Wan50], where the whole situation
was investigated and precise conditions were given for the validity of Grunwalds theorem. Since then the corrected theorem is called the GrunwaldWang Theorem.27
And the Main Theorem of BrauerHasseNoether was saved.
Independently of Wang and immediately after Wangs counter example became
known, Hasse also published a paper in which he carefully analyzed the exceptions
in Grunwalds theorem [Has50a]. In particular he too established the validity of his
weaker Existence Theorem which he had used in the Main Theorem. 28
In the ArtinTate Lecture Notes on class field theory of 1951 there is a whole
chapter devoted to the GrunwaldWang theorem [AT68]. 29
27 Shianghao Wang (19151993) received his Ph.D. at Princeton University in 1949 and afterwards returned to
China. He published two more papers connected with the GrunwaldWang theorem but later turned to Computer
Science, in particular control theory. He was professor and chairman of the Math. Dept. at Jiling Universiy
since 1952; vice president 1980/81. He became a member of the Academia Sinica. It is said that Wang was a
versatile person. He was good at chess, bridge, novels, Chinese opera. I am indebted to Professors Eng Tjioe
Tan and Ming-chang Kang for information about the vita of Shianghao Wang.
28 Much later, in the year 2008, Patrick Morton discovered an error in Hasses paper which he corrected in
[Mor11].
29 The statement of theorem 6. chapter 10 in [AT68] contains an essential misprint. This has been remarked by
Geyer and Jensen in [GJ07] who also showed that this had already led to erroneous statements in the literature.
27
1.5.4 Remarks
Let us add some more remarks to this story. Even before the error in Grunwalds
theorem was found, Hasse seemed not to be satisfied with its proof. In his opinion
the proof as given by Grunwald, which used a lot of class field theory, was not adequate
as a basis for such a fundamental result like the Main Theorem. Therefore he thought
of ways to avoid the application of Grunwalds theorem in the context of the Main
Theorem, if possible.
The weak Existence Theorem. The Existence Theorem can be weakened by removing the requirement for LjK to have a fixed degree n. In this weak form one
is looking for a cyclic extension LjK whose degree is not specified, with the only
condition that its local degrees Lp W Kp should be divisible by the given numbers
mp (for p 2 S ). Already in 1932 Hasse had given a relatively elementary proof of
this weak form [Has33a]. This does not yield the full Main Theorem but only its
weak form that every central simple algebra A is similar to a cyclic algebra (which
a priori does not necessarily imply A itself to be cyclic). Quite often this weak form
of the Main Theorem turns out to be sufficient in the applications, and so Hasses
proof in [Has33a] provides a simplified access to those applications, without using
the complicated class field proof via the GrunwaldWang theorem.
Moreover, in order to satisfy this weak form of Hasses Existence Theorem, it
turns out that the required cyclic extension
LjK can be constructed as a cyclotomic
p
`
extension of K, i.e., a subfield of K. 1/ for suitable ` (which may even be chosen as
a prime number). 30 This fact became important when Hasse gave a proof of Artins
Reciprocity Law within the framework of the theory of algebras, as he did in [Has33a].
See Section 1.6.2.
Group representations. In the joint BrauerHasseNoether paper there is a section
titled Applications and it is attributed to Hasse. One of those applications concerns
representations of finite groups:
Every absolutely irreducible matrix representation of a finite group G
can be realized (up to equivalence) in the field of nh roots of unity where
n is the order of G and h is sufficiently large.
It is understood that representations are to be meant over a field of characteristic 0. 31
Two representations over the same field are equivalent if they can be transformed into
each other by a non-singular matrix; in modern terms: if they determine isomorphic
G-modules.
30 The proof needed a lemma on prime numbers satisfying certain congruence relations, of similar kind as Artin
had to use for his general reciprocity law. Simplified proofs of this lemma were later given by Hasse himself, by
Chevalley, Iyanaga and finally in greater generality by van der Waerden [vdW34].
31At that time the theory of modular representations, i.e., over fields of characteristic p > 0, had not yet
been developed. It had been systematically started by Richard Brauer in the late 1930s.
28
This theorem had been conjectured by I. Schur [Sch06] but with h D 1. In view
of this Hasse did not say that his theorem is a proof of Schurs conjecture; instead he
says that this theorem constitutes a support of Schurs
conjecture.
p
n
In order to provePthe theorem, let K D Q. 1/. Hasse considered the group
algebra KG D i Ai , decomposed into its simple components. 32 The center
of Ai is K. The assertion of the above theorem is now transformed to say that
p
h
K. n 1/ is a splitting field of each Ai , provided h is sufficiently large. Applying the
Global Splitting Criterion this means that for each prime p of K the p-local degree
p
h
of K. n 1/ over K is a multiple of the local index mi;p . Only the prime divisors p
of n are relevant and for those, mi;p is seen to divide the group order n. Thus one
p
h
has to prove the lemma that n divides the local degree K. n 1/ for each pjn and h
sufficiently large. This then is easily checked by the known decomposition behavior
of primes in cyclotomic fields. 33
When sending his draft of their joint manuscript to Brauer on November 11, 1931,
Hasse wrote:
Ich knnte mir denken, da Sie von sich aus zu dem Satz noch etwas
hinzuzufgen oder eine Verschrfung anzubringen haben. Ich habe das
nur sehr roh angepackt Habe ich I. Schur richtig und gengend zitiert ?
I could imagine that you would perhaps have a comment to this theorem
or maybe a sharper result. I have considered the question only very
roughly Have I cited I. Schur correctly?
Brauer replied on November 13:
Dieser Satz hat mir besonderen Eindruck gemacht; ich htte nicht geglaubt, da man die Methoden auf dies Problem so unmittelbar wrde
anwenden knnen. Sehr interessant wre es ja, wenn man die Zulssigkeit von h D 1 zeigen knnte. Ich bin zur Zeit auerstande, mchte
mich aber damit noch weiter beschftigen, sobald mir der SemesterAnfangstrubel wieder Ruhe dazu lt natrlich nur in dem Fall, da Sie
es nicht inzwischen selbst erledigt haben, was ich fr recht wahrscheinlich halte.
I was particularly impressed by this theorem; I did not believe that one
could apply the methods that directly to this problem. It would be very
interesting if one could prove h D 1. At present I cannot do this but I
will think about it as soon as the commotion of the semester beginning
is over of course, only in case that you have not solved the question in
the meantime which I believe is quite probable.
32
is our notation for direct sum.

Actually, the proof in [BHN32] of the last mentioned lemma contains an error. Hasse corrects this error in
[Has50b].
33
29
Well, Hasse did not solve the question but it was Brauer who many years later verified
the Schur conjecture [Bra45]. Two more years later he proved even the stronger result
that the field of d -th roots of unity suffices where d is the exponent of the group G,
[Bra47]. See also [Roq52] and [BT55].
This may be an appropriate occasion to cite a letter which Carl Ludwig Siegel
wrote to Hasse on December 7, 1931 when he had been informed about the Brauer
HasseNoether paper. It seems that in the past Siegel too had tried to prove Schurs
conjecture, but without success.
Lieber Herr Hasse! Das ist in der Tat das schnste Geburtstagsgeschenk fr Hensel, dass seiner p-adischen Methode ein solcher Triumph beschieden wurde. Ich konnte noch nicht einmal das Schursche
Problem richtig anpacken Der Pessimismus, den ich den Aussichten
der Mathematik gegenber im Allgemeinen empnde, ist wieder einmal
wankend geworden.
Dear Mr. Hasse! This is indeed the nicest birthday gift for Hensel that
his p-adic methods have been developed to such triumph. I had not even
been able to approach the Schur problem properly The pessimism
which I harbor generally towards the prospects of mathematics has again
been shaken
There had been an exchange of letters between Hasse and Siegel before this. When in
June 1931 Siegel visited Marburg, Hasse told him about his attempts, unsuccessful at
that time, to prove the Local-Global Principle for algebras. On his return to Frankfurt,
Siegel wrote a postcard to Hasse with a proof that the discriminant of any division
algebra D over a number field is of absolute value > 1; this would have settled the
problem at least if D is central over Q. But after examining Siegels proof Hasse
pointed out to him that this proof does not work, which Siegel conceded (Many
thanks for your exposition of my unsuccessful proof!).
Algebras with pure maximal subfields. In 1934 there appeared a paper by Albert
with the title Kummer fields [Alb34]. There, Albert proved the following theorem:
A central division algebra D of prime degree p over a field K of characteristic 0 is cyclic if and only if D contains an element x K such
that x p 2 K. 34
Of course, this is trivial if K contains the p-th roots of unity because then K.x/ would
be a cyclic subfield of D of degree p. If K does not contain the p-thproots of unityp
then
Albert constructs a cyclic field LjK contained in D such that L. p 1/ D K.x; p 1/;
this can be done by the classical methods of Kummer. Albert formulated this theorem
34Albert
calls K.x/ a pure extension of K since it is generated by radicals.
30
over fields of characteristic 0 only, but from its proof it was immediately clear that it
remains true over all fields of characteristic p. 35
Concerning this theorem, Hasse wrote to Albert in February 1935:
Your result seems to me of particular interest. It allows one to eliminate
Grunwalds complicated existence theorem in the proof that every central
division algebra D of prime degree p over an algebraic number field K
is cyclic.
And Hasse proceeds to explain how to derive the Main Theorem for division algebras
of degree p from Alberts result. This is easy enough. For, let S denote the set of
those primes p of K for which the local index mp of D is 1. Choose 2 K which
is a primepelement for every finite p 2 S, and < 0 for every
p infinite real p 2 S.
Then K. p / splits D by the Splitting Theorem, hence K. p / is isomorphic to a
subfield of D. Applying Alberts theorem it follows that D is cyclic. Hasse continues:
We are trying to generalize your theorem to prime power degree. This
would eliminate Grunwalds theorem altogether for the proof of the Main
Theorem.
We do not know whether Albert replied to this letter of Hasse. But three years later in
[Alb38a] he showed that Hasses idea could not be realized. He presented an example
of a non-cyclic division algebra of index 4 containing a pure subfield of degree 4.
The base field K is the rational function field in three variables over a formally real
field.
Exponent = Index. One of the important consequences of the Main Theorem is
the fact that over number fields, the exponent of a central simple algebra equals its
index. This is a very remarkable theorem. It has interesting consequences in the
representation theory of finite groups, and this was the reason why Richard Brauer
was particularly interested in it. The theorem does not hold over arbitrary fields since
Brauer [Bra33] has shown that over function fields of sufficiently many variables,
there are division algebras whose exponent e and index m are arbitrarily prescribed,
subject only to the conditions which are given in Brauers theorems which we have
cited in Section 1.4.2 (p. 13). See also Albert [Alb32b] where a similar question is
studied. 36
In the BrauerHasseNoether paper the exponent-index theorem is obtained by
using the Existence Theorem (see Section 1.5.1, p. 22). To this end the number n in
35 In Hasses mathematical diary, dated February 1935 we find the following entry: Proof of a theorem of
Albert, following E. Witt. Witts proof of Alberts theorem is particularly simple, following the style of Emmy
Noether, but the essential ingredients are the same as in Alberts proof. It seems that Witt had presented this
proof in the Hasse seminar, and that Hasse had noted it in his diary for future reference.
36 From the references in those papers it appears that Brauer and Albert did not know the results of each other
concerning this question.
31
the Existence Theorem is chosen as the least common multiple of the local indices
mp of the p-components Ap . Then L splits A by the Splitting Theorem and hence n
is a multiple of the index m of A, hence also of its exponent e. On the other hand,
from Hasses local theory [Has31d] it follows that the local index mp equals the local
exponent ep of Ap . From Ae 1 it follows Aep 1 for each p; therefore e is a
multiple of ep (for all p) and therefore also of n. It follows e D n D m.
From the above sequence of arguments it is immediate that in fact it is not necessary
to know that LjK is cyclic. But if cyclicity is not required then it is easy, by means
of the Chinese remainder theorem 37 , to construct a field extension LjK of degree
n (as above) with the given local degrees mp for finitely many primes p. This was
pointed out by Hasse in his letter to Albert of February 1935 which we had already
cited above. Hasse wrote:
In my Annalen paper [Has33a] I derived theorem (6.43) (exponent =
index) from Grunwalds existence theorem. In point of fact this deep
existence theorem is not necessary for proving index = exponent. For
one can carry through the proof with any sort of splitting field L instead
of a cyclic L. See my first existence theorem [Has26c]. 38
GrunwaldWang in the setting of valuation theory. Both proofs of the Grunwald
Wang Theorem, the proof [Wan50] by Wang himself and Hasses proof [Has50a], use
heavy machinery of class field theory. The same is true with Artins proof in the ArtinTate lecture notes [AT68] where there is a whole chapter devoted to the Grunwald
Wang theorem. But the question arises whether the GrunwaldWang theorem does
really belong to class field theory, or perhaps it is valid in a more general setting, for
arbitrary fields with valuations. If so then it is to be expected that the proof would
become simpler and more adequate. Therefore Hasse [Has50a] wondered whether it
would be possible to give an algebraic proof using Kummer theory instead of class
field theory. This is indeed possible and has been shown by the author in collaboration
with Falko Lorenz in [LR03]. See also the literature cited there, in particular the paper
[Sal82] by Saltmanwho works with generic polynomials.
1.6 The Brauer group and class field theory

The Main Theorem allows us to determine completely the structure of the Brauer
group Br.K/ of a number field K. As we have reported in Section 1.2 already,
there are also infinite primes p involved, the Chinese remainder theorem has to be interpreted such
as to include infinite primes too. In other words: this is the theorem of independence of finitely many valuations.
38 Hasses first existence theorem is stated and proved for finite primes only, i.e., prime ideals in the base field.
It seems that Hasse himself, when he cites his paper in the letter to Albert, regarded the inclusion of infinite
primes in his first existence theorem as trivial (which it is).
37 Since
32
the authors of the BrauerHasseNoether paper regard this as one of the important
applications of the Main Theorem.
Let K be a number field and p a prime of K. If we associate to every central
simple algebra A over K its completion Ap then we obtain the p-adic localization
map of Brauer groups Br.K/ ! Br.Kp /. Combining these maps for all primes p of
K we obtain the universal localization map
X
Br.Kp /
Br.K/ !
p
where the sum on the right hand side is understood to be the direct sum. (In this
context the Brauer group is written additively.) The Local-Global Principle can
be interpreted to say that this localization map is injective. Thus Br.K/ can be
viewed as a subgroup of the direct sum of the local groups Br.Kp /. Accordingly, the
determination of the structure of Br.K/ starts with the determination of the structure
of the local components Br.Kp /.
1.6.1 The local Hasse invariant
First we consider the case when p is a finite prime of K.
The description of Br.Kp / had essentially been done in a former paper by Hasse
[Has31d] in the Mathematische Annalen. We have already had occasion to mention
this paper in Section 1.5.1 when we discussed the Local Splitting Criterion. In
fact, that criterion is a consequence of the following Local Structure Theorem from
[Has31d].
We denote by Kp.n/ the unramified extension of Kp of degree n. It is cyclic, and
the Galois group is generated by the Frobenius automorphism; let us call it '. 39 In
[Has31d] we find the following
Local Structure Theorem (level n). Let Ap be any central simple algebra over Kp , of dimension n2 . Then Ap contains a maximal commutative
subfield isomorphic to Kp.n/ . Consequently Ap is cyclic and admits a
representation of the form Ap D .Kp.n/ ; '; a/ with a 2 Kp .
The remarkable fact is not only that all of those algebras Ap are cyclic, but that
each of them contains the same canonical field extension Kp.n/ as a maximal cyclic
subfield. Even more remarkable is how Hasse had derived this. Namely, he applied
the classical p-adic methods of Hensel to the non-commutative case. Let us explain
this:
.n/
we should perhaps write ' .n/ since it is an automorphism of Kp depending on n. But let
us interpret the symbol ' as the Frobenius automorphism of the maximal unramified extension of Kp ; if applied
.n/
.n/
to the elements of Kp this gives the Frobenius automorphism of Kp . This simplifies the notation somewhat.
39 More precisely,
33
The general case is readily reduced to the case when Ap D Dp is a division algebra
of dimension n2 over Kp . Now, Kp being a complete field, it carries canonically a
valuation which we denote by v. Writing this valuation additively, the axioms for the
valuation are
v.ab/ D v.a/ C v.b/;
v.a C b/ min.v.a/; v.b//:
Now, Hasses method consisted of extending this valuation to the given division
algebra Dp . It turns out that such an extension is uniquely possible; the formula for
the extended valuation is
v.x/ D
1
v.N x/
n
.x 2 Dp /;
where N denotes the reduced norm from Dp to Kp . This formula and the proof are
precisely the same as developed by Hensel for extending valuations to commutative
extensions, in particular it uses Hensels Lemma. Now Dp appears as a valued skew
field with center Kp . As such it has a ramification degree e and a residue degree f .
But unlike the commutative case it turns out that here, Kp being the center of Dp ,
we have always e D f D n. Since f D n it follows readily from Hensels Lemma
that Dp contains the unramified field extension Kp.n/ of degree n, as announced in
the theorem.
We have said above that this proof is remarkable. This does not mean that the
proof is difficult; in fact, it is straightforward for anyone who is acquainted with
Hensels method of handling valuations. The remarkable thing is that Hasse used
valuations to investigate non-commutative division algebras over local fields. 40 The
valuation ring of Dp consists of all x 2 Dp with v.x/ 0. It contains a unique
maximal ideal, which is a 2-sided ideal, consisting of all x with v.x/ > 0.
Remark. Hasse denotes this valuation prime ideal by the letter }, and this shows up
in the title of his paper [Has31d]. This somewhat strange notation is explained by
the fact that, in Hasses paper, the symbol p is used for the canonical prime ideal in
the complete field Kp , and the corresponding capital letter P was used to denote the
valuation ideal in commutative field extensions. Thus, in order to indicate that in the
non-commutative case the situation is somewhat different, Hasse proposed to use a
different symbol, and he chose } for this purpose. Formerly this symbol, known as
the Weierstrass-p, was used to denote the elliptic function }.z/ in the Weierstrass
normalization. Hasses notation for the prime ideal of a valued division algebra did
not survive, but the Weierstrass notation }.z/ is still in use today in the theory of
elliptic functions.
40After Hasse, the valuation theory of non-commutative structures developed rapidly, not only over number
fields but over arbitrary fields. We refer to the impressive report of Wadsworth [Wad02] about this development.
All this started with Hasses paper [Has31d] which is under discussion here.
34
Emmy Noether, after having read Hasses manuscript for [Has31d], recognized
immediately its strong potential. Hasse had sent this manuscript to her for publication
in the Mathematische Annalen for which Noether acted as unofficial editor. 41 On a
postcard dated June 25, 1930 she wrote to Hasse:
Lieber Herr Hasse! Ihre hyperkomplexe p-adik hat mir sehr viel Freude
gemacht.
Dear Mr. Hasse! I have found your paper on hypercomplex p-adics very
enjoyable. 42
And, as it was her custom, she immediately jotted down her comments and proposals
for further studies. About the local theory, which concerns us here, she wrote:
Aus der Klassenkrpertheorie im Kleinen folgt: ist Lp zyklisch
n-ten Grades ber einem p-adischen Grundkrper Kp , so gibt es in
Kp wenigstens ein Element a 0, derart, dass erst an Norm eines
Lp -Elements wird. Knnen Sie das direkt beweisen? Dann knnte man
aus Ihren Schiefkrperergebnissen umgekehrt die Klassenkrpertheorie
im Kleinen begrnden
From local class field theory it follows: If Lp is cyclic of degree n
over a p-adic base field Kp then there exists at least one element a 0
in Kp such that only an becomes a norm of an Lp -element. Are you
able to prove this directly? Then one could derive local class field theory
from your skew field results
In other words: Noether asks whether the Brauer group Br.Lp jKp / contains an
element of exponent n.
We do not know precisely what Hasse replied to her. But from later correspondence with Noether we can implicitly conclude that he replied something like I do
not know. It took him some time to follow her hint and to realize that he could have
said yes in view of the Local Structure Theorem above.
Let us briefly indicate the arguments which Hasse could have used. These can be
found in Hasses later papers, the BrauerHasseNoether paper [BHN32], the American paper [Has32c] and the paper [Has33a] dedicated to Noethers 50th birthday.
But in fact the arguments are essentially based on the Local Structure Theorem.
If we associate to each a 2 Kp the cyclic algebra .Kp.n/ ; '; a/ then we obtain a
homomorphism from Kp to the Brauer group Br.Kp.n/ jKp /. By the Local Structure
41 Unofficial means that her name was not mentioned officially on the title page. But people who knew sent
their paper to her if it belonged to Noethers field of interest. As a rule, Noether read the paper and, if she found
it suitable, sent it to Blumenthal who, as the managing editor, accepted it. The date of received by the editors
was set as the date when the paper was received by Emmy Noether. Thus Hasses paper [Has31d] carries the
date of June 16, 1930.
42 Emmy Noether used the symbol p (in German handwriting) since she did not know how to write } as she
admitted in a later letter. Can we conclude from this that she never had worked with elliptic functions?
35
Theorem this homomorphism is surjective. Its kernel is the group of norms from
Kp.n/ . But Kp.n/ is unramified over Kp and therefore an element a 2 Kp is a norm
if and only if its value v.a/ 0 mod n. 43 Consequently, for any a 2 Kp its value
v.a/ modulo n represents its class in the norm class group, and hence represents the
algebra .Kp.n/ ; '; a/. In particular we see that Br.Kp.n/ jKp / is isomorphic to Z=n.
By the Local Structure Theorem, Br.Kp.n/ jKp / contains all central simple algebras
Ap of index dividing n. Consequently, if Lp is an arbitrary extension of Kp of degree
Lp W Kp D n then the Brauer group
Br.Lp jKp / Br.Kp.n/ jKp / Z=n:
Hence, Noethers question is answered affirmatively if we have equality here, which
is to say that every algebra Ap in Br.Kp.n/ jKp / is split by Lp . Since the index of Ap
divides n D Lp W Kp this follows from the Local Splitting Criterion.
Thus indeed, Hasse could have answered Noethers question with yes, already in
1930 ; in fact he did so later. But Noethers conclusion that one could derive local class
field theory from this, was too optimistic. Noethers question was concerned with
cyclic extensions only, but class field theory deals with arbitrary abelian extensions.
It was only later that Chevalley [Che33] showed how to perform the transition from
cyclic to arbitrary abelian extensions.
Let us return to the Local Structure Theorem for level n. We have seen above
that this implies an isomorphism Br.Kp.n/ jKp / Z=n, via the map .Kp.n/ ; '; a/ 7!
v.a/ mod n. In other words: the residue class v.a/ mod n is an invariant of the
algebra Ap D .Kp.n/ ; '; a/. To obtain an invariant which is independent of n Hasse
divided v.a/ by n and thus defined what today is called the Hasse-invariant:
Ap
p
v.a/
mod Z I
n
this is a certain rational number which is determined modulo integers only. If m is

a multiple of n then every Ap 2 Br.Kp.n/ jKp / can also be viewed to be contained in
Br.Kp.m/ jKp / and it turns out that the Hasse invariant as defined above is the same
for m as that for n. If n ! 1 the final version of the Local Structure Theorem for
the Brauer group emerges:
Local Structure Theorem. Let p be a finite prime of K. If we associate
to every central simple algebra over Kp its Hasse invariant then we
obtain the canonical group isomorphism

invp W Br.Kp /
! Q=Z:
43 We
assume here that the valuation of Kp is normalized with value group v.Kp / D Z.
36
If p is an infinite real prime then Kp D R and there is only one non-trivial central
division algebra over R, namely the quaternions of Hamilton. If we associate to this
the Hasse-invariant 12 then we obtain
1
Z=Z if p is real,

! 2
invp W Br.Kp /
0
if p is complex.
1.6.2 Structure of the global Brauer group
Having settled the local structure theorems, Hasse turns now to the global structure,
i.e., the structure of Br.K/ for a number field K. If A is a central simple algebra over

A
K and Ap its p-adic completion then Hasse writes briefly A
instead of pp . If we
p

associate to each A its local Hasse invariants A
then we obtain the global invariant
p
map
X0
inv W Br.K/ !
Q=Z
p
where the prime on the sign of the direct sum should remind the reader that if p is
infinite then Q=Z has to be replaced by 12 Z=Z or 0 according to whether p is real
or complex. By the Local-Global Principle this global invariant map is injective. In
other words: every algebra A in Br.K/ is uniquely determined by its Hasse invariants
(up to similarity). In order to describe the structure of Br.K/ completely one has to
describe the image of the invariant map. In other words: What are the conditions that
a given system of rational numbers, rp for each prime p of K, is the system of Hasse
invariants of some algebra A 2 Br.K/ ? The following conditions are evident:
1. There are only finitely many p with rp 6 0 mod Z.
0 or 12 mod Z if p is real,
2. For an infinite p, we have rp
0
mod Z if p is complex.
Apart from this there is only one further condition, expressed in the following structure
theorem for the global Brauer group.
Global Structure Theorem. (i) For any central simple algebra A over
the number field K, the sum formula for its Hasse invariants holds:
X A
p
0 mod Z:
of rational numbers, subject to the condi(ii) If rp is an arbitrary systemP

tions 1. and 2. above, and if p rp 0 mod Z, then there is a unique

A 2 Br.K/ such that A
rp mod Z for each p.
p
37
Today we would express this theorem by saying that the sequence of canonical maps
inv
0 ! Br.K/ !
X0
add
Q=Z ! Q=Z ! 0
(1.1)
is exact. This describes the Brauer group Br.K/, if considered via the map inv as
a subgroup of the direct sum, as being the kernel of the map add which adds the
components of the direct sum.
Although the Global Structure Theorem uses the Local-Global Principle and is
built on it, it is by no means an easy consequence of it. Perhaps this is the reason why
the theorem is not treated in the BrauerHasseNoether paper [BHN32]; recall that
this paper had to be written in haste. Nevertheless, as said above already, the authors
of [BHN32] stressed the point that their Local-Global Principle is of fundamental
importance for the structure of the Brauer group.
The Global Structure Theorem, at least its first part (i), is in fact equivalent to
Artins General Reciprocity Law of class field theory. Hasses proof can be found
in [Has33a], and it is the end point of a historic string of events stretching over
several years since 1927. Let us briefly sketch chronologically the highlights in this
development.
1927. Artin succeeded in [Art27] to prove his General Reciprocity Law which he had
conjectured since 1923. Given an abelian extension LjK of number fields, Artins
theorem established an isomorphism between the group of divisor classes attached
to LjK in the sense of class field theory, and the Galois group G of LjK. This
isomorphism is obtained by associating to every prime p of K which is unramified
in L, its Frobenius automorphism 'p 2 G. This theorem has been said to be the
coronation of Takagis class field theory.
Even before the appearance of his paper [Art27], Artin informed Hasse about his
result and its proof. There followed an intense exchange of letters between Hasse and
Artin discussing the consequences of Artins Reciprocity Law. Already in his first
such letter dated July 17, 1927, Artin mentioned that probably Hilberts version of the
reciprocity law may now be proved in full generality. Later he asked Hasse whether
he could do it and Hasse agreed. Accordingly, Hasse published in the same year 1927
a supplement to Artins Reciprocity Law [Has27c] where
(among other things) the
product formula for the general m-th Hilbert symbol a;pb m was established.
Here we do not intend to describe the definition of the Hilbert symbol; let it be
sufficient to say that it serves to decide whether a given number a 2 K is an m-th
power modulo a prime p, and that the product formula includes as a special case the
Kummer m-th power reciprocity law together with its various supplementary laws.
The definition and management of the Hilbert symbol requires that the m-th roots
of unity are contained in the base field K. Now Artin, in another letter to Hasse dated
July 21, 1927, asked whether it would be possible to define some kind of Hilbert
38
symbol without assuming that the proper roots of unity are contained in K. Hasse
succeeded with this in 1929.
1929. Hasses paper [Has30d] appeared in 1930 but since it had been received by the
editors on March 7, 1929 already we count it for 1929. In this paper Hasse defined
for an arbitrary
extension LjK and a 2 K , for each prime p of K the norm
a; LjKabelian

symbol
as an element of the Galois group of LjK. More precisely, it is
p
an element of the p-adic decomposition group Gp G. This symbol assumes the
value 1 if and only if a is a norm in the local extension Lp jKp . The norm symbol
is in some sense a generalization ofthe Hilbert
symbol. If p is unramified in L and

a D a prime element for p then ; LjK
equals
the Frobenius automorphism 'p
p
appearing in the Artin map. Hasses norm symbol satisfies the product formula:
Y a; LjK
D1
p
p
where the 1 on the right hand side denotes the neutral element of the Galois group.
Hasse obtained this product formula from Artins Reciprocity Law. Conversely,
Artins Reciprocity Law may be deduced from the above product formula.
As a side remark we mention that through this definition of the norm symbol
Hasse discovered local class field theory. See [Has30c].
1931. On May 29, 1931 Hasse submitted his American paper [Has32c] to the Transactions of the AMS. In that paper he presented a comprehensive treatment of cyclic
algebras over number fields. This was prior to the discovery of the Main Theorem,
so Hasse did not yet know that every central simple algebra over a number field K
is
For a cyclic algebra
A D
Acyclic.

a;LjK
.LjK; S; a/ he compared the Hasse invariant
with the norm symbol
. It turned out that the product formula of the
p
p
(multiplicative) norm symbol provides the key for the proof of the sum formula for
the (additive) Hasse invariant as stated in part (i) of the Global Structure Theorem.
Actually, the sum formula was not yet explicitly written down in Hasses American paper; this was done in his next paper [Has33a] only. But all the necessary
ingredients and computations can be found in Hasses American paper [Has32c] already. Although that paper had not yet appeared when the BrauerHasseNoether
paper [BHN32] was written, the content of Hasses American paper was known to
Brauer and Noether too since Hasse had informed them about his results.
1932. In March 1932 Hasse sent his dedication paper [Has33a] to Emmy Noether.
This paper does not only contain a new proof arrangement for the Main Theorem, as
we had reported earlier. In addition, Hasse stated and proved explicitly the Global
Structure Theorem, not only Part (i) (which tacitly was already contained in [Has32c])
but also Part (ii). Moreover, the paper went well beyond Hasses former papers in as
much as now he did not use Artins Reciprocity Law in proving the Global Structure
Theorem.
39
We note that all three foregoing papers of Hasse, those of 1927, 1929 and 1931,
were
built
a; LjK
on Artins Reciprocity Law because the definition of the local norm symbol
depended on Artins global law. But now, in the Noether dedication paper,
p

Hasse was able to use his invariants A
for a purely local definition of the norm
p
a; LjK
symbol
. This had been suggested to him by Emmy Noether who on a postcard
p
of April 12, 1931 wrote the following. This letter was the reaction of Noether to a
report of Hasse about his results in his American paper [Has32c].
Ihre Stze habe ich mit groer Begeisterung, wie einen spannenden Roman gelesen; Sie sind wirklich weit gekommen! Jetzt wnsche ich
mir noch die Umkehrung: direkte hyperkomplexe Begrndung der Invarianten und damit hyperkomplexe Begrndung des Reziprozittsgesetzes! Aber das hat wohl noch gute Weile! Immerhin haben Sie doch,
wenn ich mich recht erinnere, in der Schiefkrper-Arbeit mit Ihren Exponenten ep den ersten Teil schon gemacht?
I have read your theorems with great enthusiasm, like a thrilling novel;
you have got really very far! Now I wish to have also the reverse:
direct hypercomplex foundation of the invariants and thus hypercomplex foundation of the reciprocity law! But this may take still some
time! Nevertheless you had done, if I remember correctly, the first step
already in your skew field paper with the exponents ep ?
The skew field paper which Noether mentions, is Hasses [Has31d] which we
had discussed above in Section 1.6.1. 44 As we have explained there, the Local Structure Theorem indeed can be used to provide a local definition of the Hasse invariant.
Thus Noether had seen clearly the potential of this for her plan to reverse the argument, so that one first proves the sum formula of the Global Structure Theorem, and

then interpret this as the product formula for Hasses norm residue symbol a;LjK
.
p
The latter is equivalent to Artins Reciprocity Law.
And Hasse followed Noethers hint and succeeded to give what Noether called a
hypercomplex proof of Artins Reciprocity Law. Thus a close connection between
the theory of algebras and class field theory became visible.
While Artins paper [Art27] with his reciprocity law had been named as the Coronation of class field theory, similarly Hasses paper [Has33a] could now be regarded
as the Coronation of the theory of algebras.
We can now understand Artins exclamation which we have cited in Section 1.2,
namely that he regards this as the greatest advance in Number Theory of the last
years When Artin wrote that letter in November 1931, Hasses paper [Has33a]
had not yet appeared. But Artin seemed to have clearly seen, as Emmy Noether had
done, the potential of the Local-Global Principle as a foundation of class field theory.
In fact, in his letter he continued:
44 See
also Section 1.6.3 below.
40
Ich lese jetzt Klassenkrpertheorie und will nchstes Semester anschlieend hyperkomplex werden.
At present I am giving a course on class field theory, and next semester
I will continue by becoming hypercomplex.
Thus Artin intended to discuss the theory of hypercomplex systems, i.e., algebras,
with the view of its application to class field theory. 45
1.6.3 Remarks
Arithmetic of algebras and Hensels methods. Twice in our discussion we had
occasion to refer to Hasses paper [Has31d] on local algebras. The first time this
was in Section 1.5.1 when we reported that the local splitting theorem was an almost
immediate consequence of the results of that paper. The second time was in Section 1.6.1 when we discussed the local structure theorems. In both situations we have
seen that Hasses paper [Has31d] contained the fundamental ingredients which led
to success.
We note that this paper [Has31d] was received by the editors on June 18, 1930
already, long before the BrauerHasseNoether paper was composed, and even before
Hasse had formulated the conjecture of the Main Theorem in a letter to Noether (see
Section 1.7.2, p. 51). In fact, the original motivation for Hasse to write this paper was
not directly connected with the Main Theorem. From the introduction of [Has31d]
we infer that Hasse regarded his paper as a new approach to understand the arithmetic
of algebras, based on the ideas of Hensel, in the same manner as he had applied those
ideas to the investigation of the arithmetic of commutative number fields.
We cannot here give a comprehensive account of the development of the arithmetic
of algebras during the 1920s; this is an exciting story but would need much more space
than is available. The following brief comments should help to put Hasses paper
[Has31d] into the right perspective.
The study of the arithmetic theory of algebras had been started systematically by
Dickson whose book [Dic23], entitled
Algebras and their Arithmetics,
had received much attention, in particular among German mathematicians. This book
contained not only a complete treatment of the Wedderburn structure theorems for
algebras, but also a systematic attempt to develop an arithmetic theory of orders of
an algebra.
45 The next semester was the summer term 1932. In that semester Artin gave a course with the title Algebra,
and he presented there the algebraic theory of hypercomplex systems, i.e., algebras. Lecture Notes for this course
had been taken down by the student Ernst August Eichelbrenner, and a copy is preserved. However from these
notes it appears that Artin covered the algebraic theory of algebras over an arbitrary field only, but not the special
situation when the base field is a number field. In particular, the connection to class field theory is not mentioned.
But it may well have been that Artin covered those more advanced topics in a special seminar parallel to this
course.
41
Let K be a number field and ZK its ring of integers. Let A be any finitedimensional algebra over K. In this situation an order R is defined to be a subring
of A (containing the unit element) which is a finite ZK -module and generates A as a
vector space over K. The arithmetic of R manifests itself in the structure of ideals
of R. The arithmetic of such order becomes particularly lucid if the order is maximal, i.e., not properly contained in a larger order of the algebra. Perhaps it is not an
exaggeration to say that the most important feature of Dicksons book was to give the
definition of maximal orders of an algebra and to point out that the arithmetic of those
maximal orders is particularly important in the same way as in the commutative
case, i.e., algebraic number fields K, where the maximal order ZK and its prime ideal
structure is the first object to study. Whereas arbitrary orders, i.e., those which are
not integrally closed, carry a more complicated ideal theory.
Now we observe that in Dicksons book, after maximal orders have been defined
and their elementary properties developed, they are in fact not treated systematically.
The discussion is largely restricted to very special cases, namely when there exists a
euclidean algorithm. One knows in algebraic number theory that such cases are rare.
The first who set out to remedy this unsatisfying situation was Andreas Speiser
in his paper [Spe26]. Also, he arranged for a German translation of Dicksons book
[Dic27] and included his paper as an additional chapter. 46
But still, Speisers treatment was only the beginning. SoonArtin published a series
of three seminal papers in which the arithmetic of maximal orders was fully developed
[Art28a, Art28c, Art28b]. The second of these papers contained the generalization
of Wedderburns structure theorems to what today are called Artinian rings, i.e.,
rings with minimum condition for ideals. 47 This is necessary if one wishes to study
the structure of the residue class rings of a maximal order with respect to arbitrary
two-sided ideals which are not necessarily prime. The third of Artins papers then
developed the ideal theory of maximal orders of a simple algebra, in complete analogy
to Noethers theory of Dedekind rings (which had just been published the year before).
The non-commutativity implies that the ideals do not necessarily form a group but
(with proper definition of multiplication) a so-called groupoid in the sense of Brandt
[Bra30a].
It was Speisers work and, in addition, this series of papers by Artin which had
inspired Hasse to write his paper [Has31d] on local division algebras. In his introduction he refers to Speiser and Artin, and says:
Ich baue den ursprnglichen Speiserschen Ansatz in demselben Sinne
aus, wie es die Henselsche Arithmetik der algebraischen Zahlkrper mit
46 The translation had been done by J. J. Burckhardt who recently had his 100th birthday in good health.
See [Fre03]. Actually, the German edition is not merely a translation of the American book. Dickson had
presented a completely reworked manuscript for translation. The book had been reviewed by Hasse [Has28] in
the Jahresbericht der DMV.
47At that time it was not yet known that the minimum condition for ideals implies the maximum condition.
Hence Artin in his paper required the validity of the minimum as well as the maximum condition.
42
dem ursprnglichen Ansatz von Kummer tut. Betrachtet man nmlich gleichzeitig die Restsysteme nach jeder noch so hohen Potenz
von p, so kommt das darauf hinaus, da man den rationalen Koefzientenkrper p-adisch erweitert. An Stelle des Speiserschen Restsystems
mod p s tritt demzufolge ein hyperkomplexes System in bezug auf
einen Krper, den p-adischen Zahlkrper, im Sinne der Wedderburnschen Theorie.
Auf dieser einfachen Grundlage gelingt es berraschend einfach den
Aufbau der hyperkomplexen Arithmetik zu vollziehen.
I am extending the original idea of Speiser in the same sense as Hensels
arithmetic of number fields had done with the original idea of Kummer. For, if one considers simultaneously the residue classes with
respect to arbitrary powers of p then this means to extend the field of
coefficients p-adically. In this way, Speisers residue classes modulo
p s are replaced by an algebra over a field, the p-adic completion, in
the sense of Wedderburn.
In this way it is possible to build the hypercomplex arithmetic in a surprisingly simple way.
And as part of this program, Hasse mentions:
Darber hinaus gelingt es mir, eine einfache bersicht ber alle berhaupt vorhandenen Schiefkrper ber einem p-adischen algebraischen
Zahlkrper als Zentrum und deren algebraische und arithmetische
Struktur zu erhalten, in Analogie zu der bekannten Tatsache, da es
ber dem reellen Zahlkrper als Zentrum nur einen Schiefkrper, den
Quaternionenkrper, gibt.
Moreover I have succeeded in giving a complete description of all existing division algebras with a p-adic algebraic number field as its center,
and of its arithmetic and algebraic structure in analogy to the wellknown fact that with the real number field as center there exists only one
skew field, the ordinary quaternion field.
Thus Hasses local structure theorems which we had cited from [Has31a], constituted
only one aspect of this paper. The other and broader one was to build non-commutative
arithmetic in maximal orders of algebras, by using Hensels ideas of localization.
In this light we can understand why Hasse in his dedication text of the Brauer
HasseNoether paper had mentioned Hensels p-adic methods as being responsible
for the success.
Class field theory and cohomology. In the section Consequences (Folgerungen) of the BrauerHasseNoether paper there is a subsection concerning
43
Verallgemeinerung von Hauptstzen der Klassenkrpertheorie auf allgemeine relativ-galoissche Zahlkrper.

Generalization of central theorems of class field theory to the case of
arbitrary Galois extensions of number fields.
The problem is the following: Ordinary class field theory in the sense of Takagi refers
to abelian extensions of number fields. The abelian extensions L of a number field K
are characterized by certain groups, called ray class groups of divisors, which are
constructed within the base field K in such a way that the decomposition type in L
of primes p of K can be read off from the behavior of the primes in the corresponding
ray class group. Question: Is a characterization of a similar kind possible for Galois
extensions of K which are not necessarily abelian? It was known since Hasses class
field report [Has26a] that this is not possible by means of ray class groups; this is the
content of what Hasse [Has33b] called Abgrenzungssatz (theorem of delimitation).
But there may be other groups or objects which are defined within K and can serve
to describe Galois extensions LjK.
Now, in the BrauerHasseNoether paper, Hasse proposes to use Brauer groups.
He shows that every Galois extension LjK of number fields is uniquely determined
by its Brauer group Br.LjK/. And the decomposition type in L of a prime p of K
can be read off from the p-adic behavior of the elements in Br.LjK/. The proof is
almost immediate using the splitting theorems of Section 1.5.1 together with well
known density theorems of algebraic number theory.
In consequence there arises the problem how to describe the Brauer groups
Br.LjK/ within K without resorting to L. This problem is not treated in the Brauer
HasseNoether paper. However it has stimulated several mathematicians, including
Artin and Noether, to look more closely into the Brauer group or, equivalently, into
the cohomology group H 2 .LjK/.
In the ArtinHasse correspondence we find 5 letters between March and May 1932
where Artin tries to give congruence criteria for the decomposition type of a prime
p of K in L by means of factor systems. However his results were disappointing to
him. He wrote to Hasse:
Im nicht abelschen Fall kommt einfach die alte Methode heraus die
Klassenkrpertheorie anzuwenden auf Unterkrper in bezug auf die
der ganze Krper cyklisch ist Dieses ist nur eine etwas verschnte
Zusammenfassung der gewhnlichen Klassenkrpertheorie Ich
habe den Eindruck, dass noch etwas ganz Neues hinzukommen muss
um zu Isomorphie und zu Existenzstzen zu kommen.
In the non-abelian case we just obtain the old method to apply class field
theory to subfields over which the whole field is cyclic This is only a
somewhat beautified combination of the ordinary class field theory I
have the impression that something completely new has to be added.
44
Hasse seems to have been more optimistic. In his talk [Has32b] at the International
Congress of Mathematicians, Zrich 1932, he said, after having reported about the
Main Theorem:
Schlielich erweist sich in Untersuchungen von Artin, E. Noether und
mir Satz 3 (und berhaupt diese Methode) als krftiges Hilfsmittel
bei der Behandlung der groen, im Mittelpunkt der modernen Zahlentheorie stehenden Frage nach dem Zerlegungsgesetz in allg. galoisschen
Zahlkrpern.
Further work of Artin, E. Noether and myself has shown that Theorem 3
(and that method in general) is a powerful tool to deal with the great
question in the center of modern number theory, the decomposition law
in general Galois number fields.
Here, Theorem 3 means the Main Theorem, and that method had been explained
in Hasses text before, namely:
Kombination der von Hensel geschaffenen arithm. Methoden, die
ich im Anschluss an Speiser in diese Theorie hereingetragen habe, mit
gewissen algebr. Methoden, die, auf frheren Untersuchungen von Speiser und I. Schur fuend, krzlich von R. Brauer und E. Noether entwickelt wurden.
a combination of the arithmetic methods of Hensel, which I have
carried into this theory following Speiser, with certain algebraic methods which, based on earlier investigations of Speiser and I. Schur, have
recently been developed by R. Brauer and E. Noether. 48
Emmy Noether seems to have steered a middle line. On the one hand she was
informed about Artins unsuccessful attempts. In her invited address [Noe32] at the
Zrich Congress she says the following, after having reported on Hasses proof of the
reciprocity law by means of algebras, and on some further developments of Chevalley
[Che33] about factor systems:
Zugleich muss ich aber doch einschrnkend bemerken, dass die Methode
der verschrnkten Produkte allein allem Anschein nach nicht die volle
Theorie des galoisschen Zahlkrpers ergibt. Das folgt aus neuen noch
unpublizierten Arbeiten von Artin, die an den obigen Beweis von Hasse
anschliessen
48 It may seem strange that Hasse did not mention Albert who also had an independent share in the proof of
the Main Theorem (see Section 1.8.) We can only speculate about the reason for this (if there was any particular
reason at all). It may have been the fact that, for one thing, Hasse had mentioned Albert earlier in his text
together with Dickson and Wedderburn, and that on the other hand the methods which he refers to in the present
connection are concerned with computations on factor systems in the realm of class field theory, which indeed
cannot be found in Alberts papers.
45
At the same time I have to qualify this by saying that the method of
crossed products does not seem to cover the full theory of Galois number
fields. This is a consequence of new, still unpublished results of Artin
which are based on Hasses above mentioned proof
On the other hand, Emmy Noether in her own work pushes the computation
with factor systems further along by what she calls the Principal Genus Theorem
(Hauptgeschlechtssatz) [Noe33a]. Seen from todays viewpoint, her work is of
cohomological nature and her Principal Genus Theorem is essentially the vanishing
of the 1-cohomology of the idele class group, or at least of some finite level of it.
In the course of later developments the idea of approaching class field theory for
Galois extensions by means of factor systems has been dropped. The new concept for
an edifice of class field theory for Galois extensions, due to Langlands, looks quite
different.
But the extensive computations with factor systems have had a significant consequence in the long term, namely the rise of algebraic cohomology and its application
in algebraic number theory. While Hasse had introduced simple algebras into class
field theory in [Has33a], these have survived in modern times as 2-cohomology
classes only. Accordingly the exact sequence which we have written down in (1.1)
immediately after the Global Structure Theorem (p. 37), is now written in the form
0 ! H 2 .K/!H 2 .IK /!Q=Z ! 0;
P
IK being the idele group (suitably topologized) and H 2 .IK / D p H 2 .Kp /:
Many years later, Artin and Tate presented in their 1952 Seminar Notes [AT68] an
axiomatic foundation of class field theory. Their axioms were given in the language of
cohomology which by then was already well developed. There are two main axioms.
Their Axiom I is essentially the cohomological version of the exactness of the above
sequence at the term H 2 .K/. And their Axiom II is essentially equivalent to the
exactness at H 2 .IK /.
We have mentioned all this in order to point out that it started with the Brauer
HasseNoether paper [BHN32].
1.7 The team: Noether, Brauer and Hasse

As seen above, there was a close collaboration between Brauer, Hasse and Noether
which finally led to the Main Theorem. We have tried to find out how this cooperation
started and developed.
46
1.7.1 Noethers error

We do not know when Emmy Noether and Richard Brauer had met for the first time.
There is a letter from Noether to Brauer dated March 28, 1927 which seems to be a
reply to a previous letter from Brauer to her. 49 The letter is published and discussed
in the beautiful book of C. Curtis Pioneers of representation theory [Cur99], p. 226;
it starts as follows:
Sehr geehrter Herr Brauer! Es freut mich sehr, da Sie jetzt auch den
Zusammenhang der Darstellungstheorie und der Theorie der nichtkommutativen Ringe, der Algebren, erkannt haben; und den Zusammenhang zwischen Schurschem Index und Divisionsalgebren.
Dear Mr. Brauer! I am very glad that now you have also recognized the
connection between representation theory and the theory of noncommutative rings, the algebras, and the connection between the Schur index
and division algebras.
The tone of the letter is somewhat like that from an instructor to a young student 50 ,
giving him good marks for success in his studies. But then she becomes serious:
In diesen Grundlagen stimmen unsere Untersuchungen natrlich berein; aber dann scheint mir ein Auseinandergehen zu sein.
In regard to these fundamentals our investigations are, of course, in
agreement; but then it seems to me there is a divergence.
And she continues to describe this divergence, followed by an essay on how she likes
to view the situation, with the unspoken invitation that Brauer too should take the
same viewpoint.
The subject is representation theory, the Schur index and splitting fields. Noether
advocates that the whole theory be subsumed under the theory of algebras. On this
project she had been working already for some time. In the winter semester 1924/25
she had given a course on this subject. 51 In September 1925 she had given a talk at
the annual meeting of the DMV at Danzig with the title Group characters and ideal
theory. 52 In the abstract of this talk [Noe25] she writes:
49 In the same year 1927 Brauer did his Habilitation at the University of Knigsberg. It seems that Brauer
had sent his thesis (Habilitationsschrift) to Noether asking for her comments, and that the above mentioned letter
is Noethers reply. Brauer submitted his paper [Bra28] to the Mathematische Zeitschrift on July 22, 1927, four
months after Noethers letter.
50 Noether was 19 years older than Brauer, who was 26.
51 I am indebted to Mrs. Mechthild Koreuber for showing me her list of the Noether lectures in Gttingen 1916
1933, copied from the Vorlesungsverzeichnis. For the winter semester 1924/25 we find the announcement of
a 4-hour lecture on group theory. This seems to be the lecture where Noether first expounded her ideas of doing
representation theory within the framework of algebras.
52 This was the same meeting where Hasse gave his famous report on class field theory. Noethers and Hasses
talks were scheduled at the same session. See [Roq01].
47
Die Frobeniussche Theorie der Gruppencharaktere also der Darstellung endlicher Gruppen wird aufgefat als Idealtheorie eines vollstndig reduziblen Ringes, des Gruppenringes.
Frobenius theory of group characters i.e., representation of finite
groups is seen as the ideal theory of a completely reducible ring, the
group ring.
She continues with the Wedderburn structure theorems for algebras and how these
are to be interpreted within representation theory. She ends up with the sentences:
Damit ist aber die Frobeniussche Theorie eingeordnet. Eine ausfhrliche
Darstellung soll in den Math. Ann. erscheinen.
Thus the theory of Frobenius is subsumed. A detailed account is to be
published in the Math. Ann.
We see that already in 1925 Noether had a clear view of what was necessary to
develop representation theory within the framework of algebras. But the promised
publication had to wait for quite a while. Noether was not a quick writer; more often
her ideas went into the papers of other people rather than forcing herself to write a
manuscript. It is not clear which paper she was announcing here; some years later
there are two publications of Noether on representation theory [Noe29] and [Noe33b]
(both in the Mathematische Zeitschrift and not in the Mathematische Annalen). The
first of these papers [Noe29] consists essentially of the notes taken by van der Waerden
of Noethers lecture in the winter semester 1927/28. These lecture notes have been
said to constitute one of the pillars of modern linear algebra. 53 The second paper
[Noe33b] is somewhat more closely related to the topics of her correspondence with
Brauer. We have the impression that in the announcement she had in mind one longer
paper but in her hands this became too long and thus was divided into two parts.
Returning to Noethers letter to Brauer on March 28, 1927: In one of the statements
in that letter she claims that each minimal splitting field of a division algebra D is
isomorphic to a maximal commutative subfield of D. But this turned out not to be
true. There do exist minimal splitting fields of D whose degree is larger than the
index; an embedding into D is possible if and only if the degree of the splitting field
is minimal, i.e., equals the index of D.
We will see that this error had important consequences, leading to the Brauer
HasseNoether theorem.
We do not know whether Brauer had replied to Noethers letter. But we do know
that both met at the next annual meeting of the DMV, on September 1824, 1927
at the spa of Kissingen. Neither Noether nor Brauer were scheduled for a report
at that meeting but certainly they talked about the topic of Noethers letter. From
the correspondence over the following weeks we can obtain a fairly good picture of
53 Cited
from [Cur99] who in turn refers to Bourbaki.
48
what they had discussed in Kissingen. Apparently Brauer knew that Noethers claim
cited above was erroneous, i.e., that there do exist minimal splitting fields whose
degree is larger than the index. And he told Noether. But then Noether asked whether
the degrees of the minimal splitting fields may be bounded. This too seemed to be
doubtful. Noether wished to check the question at the smallest example, namely the
ordinary quaternions over the rational number field Q.
In Kissingen they could not settle the question. Two weeks later, on October 4,
1927, Noether sent a postcard to Hasse (with whom she had corresponded since
1925):
Lieber Herr Hasse, Knnen Sie mir sagen, ob aus den allgemeinen Existenzstzen ber abelsche Krper direkt dieser folgt: Es gibt zu jedem n
mindestens einen (vermutlich beliebig viele) in bezug auf den Krper der
rationalen Zahlen zyklischen Krper des Grades 2n, derart da sein Unterkrper vom Grad 2n 1 reell ist, und da .1/ in ihm als Summe von
hchstens drei Quadraten darstellbar ist (Quadrate gebrochener Zahlen).
Dear Mr. Hasse! Can you tell me whether the general existence theorems
for abelian fields yield the following: For every n there exists at least
one (perhaps arbitrary many) cyclic field over the rational number field
of degree 2n such that its subfield of degree 2n1 is real, and it admits
a representation of .1/ by at most three squares (squares of fractional
numbers) 54
If such fields would exist then there would exist minimal splitting fields of arbitrary
large degree. Noether continued:
R. Brauer uerte (in Kissingen) die Vermutung der Nichtbeschrnktheit. Seine Beispiele waren aber komplizierter als Quaternionenkrper.
Es wrde folgen, da man ber diese kleinsten Krper viel weniger wei,
als ich eine Zeitlang dachte.
R. Brauer conjectured in Kissingen the non-boundedness [of the degrees
of minimal splitting fields], but his examples were more complicated
than quaternion fields. It would follow that one knows much less about
those minimal fields than I had believed for some time.
Hasse reacted immediately. Already on the next day, on October 6, 1927, he sent
to Noether a 4-page manuscript in which he gave a detailed proof that indeed, such
fields do exist. For us it is of interest that his proof was based upon the Local-Global
54 Of course, the condition that the subfield of degree 2n1
should be real, is always satisfied and hence could

be omitted. This is another instance where we can see that Noether often wrote her postcards very impulsively and
dispatched them immediately without thinking twice about the text. (Very much like some people send e-mail
messages nowadays.) If she had, she would certainly have noticed that the condition to be real is superfluous, as
she admits in her next postcard.
49
Principle: not the Local-Global Principle for algebras (this was not known yet) but
for quadratic forms, which Hasse had discovered 1922 in his famous habilitation
thesis [Has23b]. The quadratic form relevant to Noethers question is the sum of four
squares: f .x/ D x12 C x22 C x32 C x32 . The question whether f .x/ has a non-trivial
zero in a field is, of course, identical with the question whether the field is a splitting
field of the ordinary quaternions.
So we see here, in Hasses letter of October 1927, the nucleus of what in 1931
would become the Local-Global Principle for algebras in the BrauerHasseNoether
paper.
Immediately thereafter, on October 10, 1927, Noether wrote to Brauer sending
him Hasses letter which solves the question under discussion, and proposing a joint
note, to be published together with a note of Hasse. There followed a series of
letters within the triangle BrauerHasseNoether, discussing details about the planned
notes and possible generalizations. After a while Brauer succeeded to construct those
fields without using Hasses Local-Global Principle for quadratic forms. Hasse asked
Brauer to explain to him the group theoretic relevance of his example, which Brauer
did in full detail.
Finally there appeared a joint note of Brauer and Noether [BN27], and immediately
after it in the same journal a note by Hasse [Has27a]. The offprints of both papers
were bound together and distributed in this form.
Here we see the first instance where Brauer, Hasse and Noether had formed a team
towards a common goal as a consequence of Noethers error concerning minimal
splitting fields.
1.7.2 Hasses castles in the air
Thereafter Hasse became increasingly interested in the theory of algebras because
he had seen that number theory, in particular class field theory and the local p-adic
theory, could be used there profitably. Brauer became interested in class field theory
because of the same reason. And Noether, who had brought the two together in the
first place, was pleased because she observed that her methods were accepted by
both.
There resulted a regular exchange of letters and information between the three
members of the team. Brauer learned from Hasse about class field theory and Hasse
learned from Brauer about algebras and group rings.
For instance, upon a request of Hasse, Brauer wrote him on July 9, 1929 all that he
knew about group rings in the setting of algebras. On October 26, 1929 Brauer sent
to Hasse his notes which he had composed for his lecture course at Knigsberg. On
March 16, 1930 Hasse wrote to Brauer explaining in detail his main ideas and results
on skew fields over }-adic fields (they appeared later in [Has31d]). We have discussed
that paper in Section 1.6.1 in connection with the Local Structure Theorem. (See also
Section 1.6.3.) At the end of his letter Hasse observed that over a local field Kp (with
50
p finite) there exist extension fields with non-cyclic Galois group; nevertheless every
central skew field over Kp is cyclic. And then he asks Brauer:
Ich mchte nun fragen, ob Ihnen ber dem rationalen Krper als Zentrum
oder einem gewhnlichen algebraischen Zahlkrper als Zentrum ein
Schiefkrper bekannt ist, fr den es keinen Abelschen oder wenigstens
keinen zyklischen minimalen Zerfllungskrper D maximalen Teilkrper gibt. Ist etwa das von Ihnen angedeutete direkte Produkt zweier
Quaternionenalgebren in diesem Sinne nicht zylisch ?
Now I would like to ask whether you know a central skew field over the
rational or an algebraic number field which does not admit an abelian, or
at least no cyclic, minimal splitting field which is a maximal commutative
subfield. Is it true that the direct product of two quaternion algebras,
which you mentioned the other day, is non-cyclic in this sense?
We see that Hasse contemplates about whether globally every central skew field is
abelian or perhaps even cyclic, i.e., the Main Theorem. But he is not sure and wants
to know the opinion of Brauer.
In his reply on April 18, 1930 Brauer thanked Hasse for this letter, saying that he
was highly interested in Hasses beautiful results on the arithmetic of hypercomplex
numbers. Concerning Hasses question he wrote:
Die von Ihnen gestellte Frage kann ich leider zur Zeit noch nicht beantworten. Ich wei nicht einmal, ob es Schiefkrper (mit endlichem Rang
ber ihrem Zentrum) gibt, die keinen Normalkrper als maximalen
Teilkrper besitzen. Ich habe frher vergeblich versucht, die Existenz
eines solchen Normalkrpers zu beweisen. Jetzt will ich umgekehrt versuchen, ein Beispiel zu konstruieren, bei dem es keinen derartigen Normalkrper gibt.
Unfortunately I am not able yet to answer your question. I do not even
know whether there exist skew fields (of finite rank over their center)
which do not admit a normal 55 field as a maximal subfield. Formerly I
have tried without success to prove the existence of such a normal field.
Now I am trying the opposite, to construct an example which does not
admit such a normal field.
But, he adds, although he has an idea how to construct such an example, this will
probably be very sophisticated. Apparently he did not succeed. To our knowledge,
such an example was first given by Amitsur [Ami72]. But, of course, the base field
in Amitsurs construction was not an algebraic number field. 56 Brauer continues:
55A
normal field extension, in the terminology of the time, means a Galois extension.
that time there has developed an extensive literature trying to understand the construction of non-
56 Since
51
Auch die andere von Ihnen gestellte Frage, ob es Schiefkrper gibt, die
nicht vom Dicksonschen Typ sind (d.h. die keinen zyklischen maximalen
Teilkrper haben), wei ich nicht. Das von mir erwhnte Produkt der
beiden Quaternionensysteme scheidet aus, da das Zentrum dabei nicht
algebraisch ist Sobald ich genaueres wei, schreibe ich Ihnen dann
noch einmal.
Also your more specific question whether there are skew fields which
are not of Dickson type (i.e., which have no cyclic maximal subfields),
I am not able to answer. The product of two quaternion algebras which
you mentioned, is not eligible since its center is not algebraicAs soon
as I will know more about it I will write again to you.
The said product of two quaternion algebras had been treated by Brauer in his paper
[Bra30b] over the rational function field Q.u; v/ of two variables. The examples
provided by Brauer have exponent 2 and index 4 but the question whether they are
cyclic is not treated in [Bra30b]. The first who explicitly constructed a non-cyclic
division algebra was Albert [Alb32a]. 57
At the end of his letter Hasse said that he would very much like to talk personally
to Brauer about these questions. And he announced that in the fall he will be in
Knigsberg (the place where Brauer lived) for the meeting of the DMV (German
Mathematical Society). We may safely assume that Hasse and Brauer had a very
thorough discussion there, together with Emmy Noether who also participated at the
meeting. Unfortunately there is no record about their conversations. 58
In December 1930 Hasse seemed to have made up his mind and written up some
coherent conjectures about the structure of algebras over number fields. He did so in
a letter to Emmy Noether. We do not know Hasses letter but we do know the reaction
of Emmy Noether. From that we can conclude that among Hasses conjectures was
the Main Theorem and the Local-Global Principle, but also the consequence that over
number fields, the index of a central simple algebra equals its exponent. It appears
that Hasse had mentioned that his conjectures do not yet have a solid foundation. For
Noether replied on December 19, 1930 :
Ja, es ist jammerschade, da all Ihre schnen Vermutungen nur in der
Luft schweben und nicht mit festen Fen auf der Erde stehen: denn ein
crossed product division algebras. As a noteworthy example we mention the work by Brussel who uses Wangs
counterexamples to Grunwalds theorem to construct non-crossed products. See [Bru97], and also subsequent
papers of the same author.
57Albert used a construction similar to but not identical with Brauers. Note that in a footnote on the first page
of [Alb32a] it is claimed that Brauers construction was false, but at the end the author admitted in an additional
note that the difficulty was one of the interpretation of language, rather than a mathematical error.
58 Except Noethers reference to their trip to Nidden of which she reminded Hasse in her letter of November 12,
1931 which we have cited in Section 1.4.3. As a side remark we would like to draw the readers attention to
the essay [Tau81] by Olga Taussky-Todd who reported about her experiences during this Knigsberg meeting,
including her vivid recollection of how Noether and Hasse seemed to have a good time discussing her (Olgas)
results on the capitulation problem of class field theory.
52
Teil von ihnen wieviel bersehe ich nicht strzt rettungslos ab durch
Gegenbeispiele in einer ganz neuen amerikanischen Arbeit: Transactions
of the Amer. Math. Society, Bd. 32; von Albert. Daraus folgt zunchst,
da der Exponent wirklich kleiner sein kann als der Index, schon beim
rationalen Zahlkrper als Zentrum; und damit also weiter da Ihre Formentheorie sich nicht auf Formen hheren Grades bertragen lt. Ob
Ihre Vermutung mit dem zyklischen Zerfllungskrper gilt, wird zum
mindesten zweifelhaft.
Yes, it is a terrible pity that all your beautiful conjectures are floating in
the air and are not solidly fixed on the ground: for part of them how
many I do not yet see hopelessly crash through counterexamples in
a very new American paper by Albert. From that it follows, firstly,
that the exponent can indeed be smaller than the index, already with
the rational number field as center, and furthermore that your theory of
forms cannot be transferred to forms of higher degree. Whether your
conjecture concerning cyclic splitting fields holds is at least doubtful.
After this, Noether proceeds to explain to Hasse the counterexamples which she
purports to have found in Alberts paper [Alb30]. And she ends the letter by asking
Hasse to inform her if his cyclic splitting field is crashed too; this is an indication
that indeed the Main Theorem was conjectured by Hasse, as we had stated above
already.
We see that Noether did not yet believe in the validity of the Main Theorem of
which one year later she would be a co-author.
In her letter Noether had mentioned Hasses theory of forms (Formentheorie).
We do not know precisely what was meant by this. In the present context it seems
probable that forms were to be understood as norm forms from central simple
algebras, and that Hasse had the idea that some kind of Local-Global Principle should
hold for those norm forms in analogy to quadratic forms which he had treated in his
thesis and following papers. If our interpretation is correct then we can conclude that
here, in December 1930, Hasses conjectures included the Local-Global Principle for
algebras.
The vivid language Noether had used in her reply appears to be quite typical of
her style. Hasse apparently did not mind it. He seems to have checked Alberts paper
cited by Noether and found out that Noethers interpretation of Alberts result was
incorrect. And he wrote this to her. Whereupon on December 24, 1930 (Christmas
eve!) she replied:
im brigen aber ist dieser Brief ein pater peccavi. Ihre Luftschlsser
sind nmlich noch garnicht umgefallen ich habe aus der Albertschen
Frchterlichkeit ziemlich genau das Gegenteil dessen herausgelesen
was drinstand. Erst Ihr Gegenbeispiel hat mir die Sache klar gemacht.
53
This letter is a pater peccavi. For, your castles in the air are not yet
crashed I have extracted from Alberts horrible thing exactly the opposite of what was in it. 59 Only your counterexample has cleared up the
situation for me.
Noether said that she had extracted from Alberts paper the opposite of what was in
it. After longish explanations of her error she writes:
Es scheint also doch ganz wahrscheinlich, da bei algebraischen Zahlkrpern als Zentrum immer Exponent und Index bereinstimmen.
After all, it appears probable that with an algebraic number field as
center, exponent and index do always coincide.
From then on, Noether was on Hasses side and she vividly advocated his conjectures. She arranged that Hasse was invited to give a colloquium talk in the Gttingen
Mathematical Society on January 13, 1931. The title of his talk was On skew fields
(ber Schiefkrper). Hasses manuscript for this talk is preserved. This shows:
In Gttingen, on January 13, 1931 Hasse publicly announced his conjecture for the Main Theorem. 60
Although Hasse in his manuscript did not say anything about how he would try to
approach this conjecture, implicitly we can see the Local-Global Principle behind
it. For, he reports extensively on his results about local algebras [Has31d], and that
he had discovered they are always cyclic. And starting from the local results he
formulates the conjectures for the global case.
1.7.3 The Marburg skew congress
Some weeks later Hasse and Noether met again, this time in Marburg on the occasion
of a small congress, today we would say workshop, on skew fields. 61 Hasses idea
was to bring together mathematicians who were active in the theory of algebras, of
class fields and of group representations, in order to join forces with the aim of solving
his conjectures. It seems that he had discussed this plan with Emmy Noether when
he visited Gttingen for his colloquium talk in January; they had agreed to have the
workshop at the end of the winter semester, February 26 to March 1, 1931. Noether
liked to call the meeting the skew congress.
59Alberts style of writing was mostly using extended computations which Noether hated; she was advocating
the use of abstract notions and structural arguments. This explains that she called his paper horrible.
60 In the book [Cur99] it is said that the Main Theorem had been conjectured by Dickson already. But we did
not find this in Dicksons works, and after inquiring with the author he replied: The statement in my book about
Dicksons conjecture has to be withdrawn. His statement was based on an assertion in Feits obituary article on
Brauer [Fei79] which he had accepted without checking.
61 Starting with the summer semester 1930 Hasse had accepted a professorship in Marburg as the successor of
Kurt Hensel. On this occasion he had been granted some funds for inviting visitors to Marburg for colloquium
lectures etc.
54
Of course Richard Brauer was invited too. In the letter of invitation to Brauer,
dated February 3, 1931, Hasse said this will become a small, gemtlicher congress
on skew fields. 62 Hasse mentions the names of the other people who were invited:
Noether, Deuring, Koethe, Brandt and Archibald (disciple of Dickson). Moreover,
Hasse wrote, invitations will be sent to all people who are interested, e.g., Artin,
Speiser, I. Schur but the funds available to Hasse were not sufficient to cover the
expenses of all of them. From the correspondence of Hasse with Krull we know
that Krull was invited too, as well as F. K. Schmidt (both in Erlangen) but they were
unable to come because at precisely the same day they had invited a guest speaker to
the Erlangen colloquium.
Emmy Noether, in her letter to Hasse dated February 8, 1931 offered a title for
her own talk, and she forwarded already some proposals for the program of the skew
congress. She proposed the talks of R. Brauer, Noether, Deuring, Hasse to be held in
this order, so that every one of the lecturers could build on the foregoing talks. The
other lectures, she wrote, were independent. In her letter she also mentioned Fitting,
a Ph.D. student of hers. 63
Moreover, Noether strongly recommended to invite Jacques Herbrand, her Rockefeller fellow. She wrote that Herbrand had worked on Logic and Number Theory only.
Number Theory he had learned from Hasses Class Field Report and Hasses papers
on norm residues. She recommends him as a mere participant only but if he is to give
a talk too then he could report on his results about the integral representation of the
Galois group in the group of units [Her30].
In those years it was not uncommon that colloquium lectures, meetings etc. of
the foregoing year were reported in the Jahresbericht der DMV. Accordingly, in its
1932 issue we can find the Marburg 1931 skew congress program as follows (English
translation): 64
62 The following passage from Hasses letter to Brauer may also be of interest: As I told you already last
summer, we are doing representation theory this year, and we use mainly your lecture notes which you had kindly
sent me.
63 Recently some letters from Emmy Noether to Paul Alexandrov have been published by Renate Tobies
[Tob03], and there, in a letter of October 13, 1929, we read: Hasse will go to Marburg as the successor of
Hensel; I wrote to him concerning connected visiting lectures but have not yet obtained a reply. In her comments
Tobies interpreted this as Noether having written to Hasse proposing to establish in Marburg something like the
skew congress. But this interpretation remains doubtful. Among the letters from Noether to Hasse of that time
we did not find any passage of that kind. Instead, a few days earlier than her letter to Alexandrov she wrote a
letter to Hasse (October 7, 1929) where we read something else. There she first congratulates Hasse for having
received the offer from Marburg University, and then she proposed Alexandrov to be named as Hasses successor
in Halle. It is not clear how this blends with what she wrote to Alexandrov about connected visiting lectures
(zusammenhngende Gastvorlesungen). We could speculate that she hoped, if Alexandrov would have been
mentioned in the list of possible successors of Hasse in Halle, then at least she could obtain funds for inviting
him to Germany for longer periods. But this is pure speculation and so the meaning of Noethers writing to
Alexandrov remains in the dark, as for our present knowledge. In any case, as we see, once Noether knew
about Hasses idea for this skew congress, she actively stepped in and helped him in the planning.
64 We have found the reference to this program in Tobies article [Tob03].

26.2.
H. Hasse
27.2.
R. Brauer (Knigsberg)
Galois theory of skew fields.
M. Deuring (Gttingen)
Application of non-commutative algebra to norms and

norm residues.
E. Noether (Gttingen)
Hypercomplex structure theorems and number theoretic applications.
R. Archibald (Chicago)
The associativity conditions in Dicksons division algebras.
H. Fitting (Gttingen)
Hypercomplex numbers as automorphism rings of

abelian groups.
H. Brandt (Halle)
Ideal classes in the hypercomplex realm.
G. Kthe (Mnster)
Skew fields of infinite degree over the center.
28.2.
55
Dickson skew fields of prime degree.
There may have been other participants in the workshop who did not deliver a
talk. Perhaps Herbrand was one of them.
We see that Noethers proposals for the order of the talks were not realized;
perhaps the mutual dependance of the talks was not so strong that a unique order
would follow. The first day (February 26) was the day of arrival; we know that the
visitors arrived late at noon and hence Hasses talk had probably been scheduled
some time in the afternoon. We observe that in the title he used the old terminology
Dickson algebras. Hasses notes for this lecture are preserved and there, however,
he speaks of cyclic algebras. From the notes we infer that Hasse presented a
report on cyclic algebras of prime degree, based on his earlier work [Has31d] and the
HilbertFurtwngler Norm Theorem (see Section 1.4.1). At the end he presented a
number of problems, thus setting the pace for this conference and for future work in
the direction of the Local-Global Principle (see Section 1.4) and the structure of the
Brauer group (Section 1.6).
If Hasses aim in this workshop had been to get a proof of the Main Theorem then
this aim was not achieved. But there resulted a general feeling that the final solution
was near. Hasse, in particular, seems to have been encouraged and kept himself busy
working on the problem.
Already a week later, on March 6, 1931 Hasse proudly sent a circular to the
participants of the skew congress with the following telegram style message:
Liebe(r ) Herr/Frulein: Habe soeben den fraglichen Normensatz fr
zyklische Relativkrper bewiesen, und mehr braucht man fr die Theorie der zyklischen Divisionsalgebren nicht.
Dear Sir/Madam: Just now I have proved the norm theorem in question
for relatively cyclic fields, and more is not needed for the theory of cyclic
division algebras.
56
In other words: He had proved the Local-Global Principle to hold for cyclic algebras
of any index, not necessarily prime. This was one of the problems which he had
stated in his talk on February 26. Hasse published the result in [Has31a], and its
consequences for cyclic algebras were announced in [Has31c]. A complete theory of
cyclic algebras over number fields followed in his American paper [Has32c].
From here on, we have already reported the further development in the foregoing
Sections 1.41.6.
1.8 The American connection: Albert

1.8.1 The footnote
There is an extended footnote in the BrauerHasseNoether paper [BHN32] which
reads as follows (in English translation):
The idea of reduction to solvable splitting fields with the help of Sylows
group theoretical theorem has been applied earlier already by R. Brauer,
namely to show that every prime divisor of the index also appears in the
exponent [Bra28]. Recently A. A. Albert has developed simple proofs
for this idea, not dependent on representation theory, also for a number
of general theorems of the theory of R. Brauer and E. Noether ([Alb31c],
[Alb31b]; for the reduction in question see in particular Theorem 23 in
[Alb31b]).
Added in proof. Moreover A. A. Albert, after having received the news
from H. Hasse that the Main Theorem has been proved by him for abelian
algebras (see in the text below), has deduced from this, independent from
us, the following facts:
a) the Main Theorem for degrees of the form 2e ,
b) the theorem 1 below (exponent = index),
c) the basic idea of reduction 2, and also of the following reduction 3,
naturally without referring to reduction 1, and accordingly with the
result: for division algebras D of prime power degree p e over
there exists an extension field 0 of degree prime to p over , so
that D0 is cyclic.
Of course, all three results are now superseded by our proof of the
main theorem which we have obtained in the meantime. But they show
that A. A. Albert has had an independent share in the proof of the main
theorem.
57
Finally, A. A. Albert has remarked (after knowing our proof of the main
theorem) that our central theorem I follows in a few lines from the
theorems 13, 10, 9 of a paper which is currently printed ([Alb31d]). The
proof of those theorems is based essentially on the same arguments as
our reductions 2 and 3.
Here, the reductions are the steps in the proof of the Main Theorem in the Brauer
HasseNoether paper. As explained in Section 1.4.2, reduction 2 is Brauers Sylow
argument, reduction 3 is Noethers induction argument in the solvable case.
This footnote has aroused our curiosity. We wanted to know more about the
relation of Albert to Hasse, and about Alberts role in the proof of the Main Theorem.
The correspondence between Albert and Hasse during those years is preserved. Our
following report is largely based on these documents.
To have a name, we shall refer to this footnote as the Albert-footnote. For later
reference we have divided the Albert-footnote by horizontal lines into three parts. It
will turn out that these parts were added one at a time. The horizontal lines are not
contained in the original. Also, in the interest of the reader we have changed the
references to Alberts papers in the original footnote to the corresponding reference
codes for this article.
1.8.2 The first contacts
A. Adrian Albert (19051972) had been a disciple of Dickson. We have said in
Section 1.6.3 (p. 40) already that Dicksons book Algebras and their Arithmetics had
a great influence on the work of German mathematicians in the 1920s. The results
of Dickson and his disciples were noted carefully and with interest by the German
mathematicians around Noether.
Already in Section 1.7 (p. 51) we have met the name of Albert. There we reported
that Noether, after reading a recent paper by Albert, thought erroneously for a short
time that, with the help of Alberts results, she could construct a counterexample to
some of Hasses conjectures in connection with the Main Theorem.
We find the name of Albert again in the manuscript which Hasse had written for
his personal use on the occasion of the colloquium talk at the Gttingen Mathematical
Society, January 13, 1931. This was the talk where Hasse publicly announced for
the first time his conjecture for the Main Theorem; we have discussed it on p. 53.
After having announced his conjecture, Hasse, according to his manuscript, reported
what was known for division algebras D of small index n. For n D 2 and n D 3
he mentioned Wedderburn and Dickson, respectively, for the fact that D is cyclic.
For n D 4 he cited Albert for the fact that every division algebra contains a maximal
subfield of degree 4 which is abelian with Galois group non-cyclic of type .2; 2/.
But Hasse noted in his manuscript that Alberts proof cannot be valid because there
exist, as is easily seen, cyclic division algebras of index 4 which do not contain an
abelian subfield with group of type .2; 2/.
58
We do not know what Hasse had in mind when he wrote as is easily seen. His
own Local-Global Principle, which he conjectured at this colloquium, can be used
to prove that his assertion is not true over number fields. In any case, we observe
that Hasse did not write that Alberts proof is not valid but he wrote it cannot be
valid; this indicates that he had not checked Alberts proof in detail but had in mind
some construction of those algebras which would yield a counterexample to Alberts
assertion. We do not know when Hasse had discovered that his construction did not
work. Maybe this was shortly before he actually went to Gttingen, and then clearly
he would not have mentioned it in his talk. Maybe it was after his talk in the discussion
with Emmy Noether who, having had her own experience with Alberts paper (as we
have seen in the foregoing section), had now studied it in detail and could assure
Hasse that it was correct. In any case we know that Hasse, after this experience, now
wished to establish contact with Albert in order to clear up the situation.
We do not know Hasses first letter to Albert, or the precise date when it was sent.
Alberts reply is dated February 6, 1931 and so, taking into consideration the time
for overseas postal delivery 65 Hasse wrote his letter shortly before or shortly after
his Gttingen colloquium lecture on January 13. Thus started the correspondence
between Hasse and the 7 years younger Albert, which continued until 1935. There
are 15 letters preserved from Albert to Hasse and 2 letters in the other direction.
As we learn from Alberts reply, Hasse had addressed his first letter to Dickson
(with whom he had exchanged reprints in the years before) who forwarded it to Albert.
At that time Albert was at Columbia University in New York. Hasse had described
his own work in his letter; Albert replied that he was very interested in it and he
introduced himself. The next letter of Albert is dated March 23, 1931. As we have
said earlier, only the letters from Albert to Hasse are preserved while most of the
letters from Hasse to Albert seem to be lost. Accordingly, when in the following we
report about letters from Albert to Hasse we have to remember that usually between
two such letters there was at least one from Hasse to Albert. 66 Now, on March 23 there
is already some mathematical discussion in the letter. Replying to Hasses question
on the existence of non-cyclic division algebras of index 4 over a number field, Albert
wrote: 67
The question seems to be a number-theoretic one and I see no way to get
an algebraic hold on it. It seems to be a hopeless problem to me after
more than a years work on it.
We observe that this is the same question which Hasse had asked Brauer a year before
(letter of March 16, 1930) but could not get an answer either (see p. 50). It appears that
the motivation of Hasse was to secure his conjecture concerning the Main Theorem.
65 This
was about two weeks there was no air mail yet.

there were more than one, for Albert on June 22, 1932 wrote: You are, may I say it, a very
pleasing friend to write me so often without receiving any answer.
67 The letters between Albert and Hasse were written in English. We are citing directly from the original.
66 Sometimes
59
If the experts were not able to construct non-cyclic algebras over a number field then
this would add some point to the conjecture being true.
Albert also reported on the results of his new paper [Alb31c] which had appeared
in the January issue of the Transactions of the AMS. Every central division algebra of
index 6 68 over any field of characteristic 0 is cyclic provided it satisfies what he calls
a mild assumption R2 (and which he could remove in his later paper [Alb31b]). In
reply to some other related question, namely about the product of two central simple
algebras, Albert presented an erroneous answer, saying that A K A is a total matrix
algebra. But he corrected this four days later, writing that it had to be the product of
A with the reciprocal algebra A0 . 69
This and more was of course known already to the trio BrauerHasseNoether,
by Brauers theorems of 1928 which we have cited in Section 1.4.2 (p. 13), and
also by Emmy Noethers Gttingen lectures 1929/30 which however, were not yet
published. 70 In the next letter, dated May 11, 1931, Albert wrote that he had completed
a paper containing most of Brauers results but which he had obtained independently
and with his own new methods. Fortunately, he added, he had discovered Brauers
papers before it was too late and hence could give Brauer priority. This refers to
Alberts paper [Alb31b]. The methods of Albert in this paper are independent of
representation theory and, in this sense, they can be regarded as a simplification of
Brauers approach.
We see that Hasses initiative to open direct contact between German andAmerican
mathematicians working on algebras, had from the start been accepted by Albert. We
have to be aware of the fact that at the time there were not many international meetings
to establish contacts, no e-mail, and journals arrived usually much later than the time
when the results had been discovered. The letters of Albert show that he was fully
aware of the possibilities which the correspondence with Hasse opened to him: To
present his ideas and results to Hasse (and hence to the German group working on
algebras) and at the same time to learn about their methods and results (which he
regularly shared with Dickson and Wedderburn). In his letter of May 11 he wrote:
Your work on quadratic forms is not new to me. In fact I have been
reading your Crelle and Jahresbericht work ever since your first letter to
me. In this period I have also been able to apply your most fundamental
result on quadratic forms in n 5 variables, together with my above
mentioned new methods to prove the following results
It appears that Hasse had explained to Albert his idea of using the Local-Global
Principle for quadratic forms, perhaps in a similar way as he had written to Brauer
68Albert spoke of algebras of order 36, thereby defining the order as the dimension of the algebra as vector
space over the base field. We shall avoid this terminology which is in conflict with the terminology used by
Noether and Hasse.
69Albert corrected this error in the already published paper [Alb31c] by putting a page of Errata into the same
Transactions volume 33.
70 See [Noe29], [Noe33b], [Noe83], [vdW31].
60
(see p. 19). And Albert had reacted immediately, proving theorems about quaternion
algebras and algebras of 2-power index which are the obvious candidates whose norm
forms may possibly be handled by quadratic forms. Among the results which Albert
reports in this letter is his answer to Hasses former question which he (Albert) had
classified as hopeless even in his foregoing letter, namely:
Every central division algebra of index 4 over an algebraic number field
is cyclic. 71
Moreover he writes that, over a number field, the product of two quaternion algebras
is never a division algebra, and that the same is true for division algebras of 2-power
index. Finally, for division algebras of 2-power index he proved Hasses exponentindex conjecture.
Albert adds that the last of the above statements would probably remain true if
2 is replaced by any prime number p provided he could prove some kind of LocalGlobal Principle for the norm forms of division algebras of index p. But in the next
letter (June 30) he apologizes for having given the impression that he could prove the
exponent-index conjecture generally. He is still working on it. Also he writes:
I want to remark in this connection that I have proved that your results
imply that the direct product of any two central division algebras is a
division algebra if and only if the indices of the two algebras are relatively
prime (for an algebraic reference field).
And so on. The letters from Albert are full of information about his results, some of
them obviously following Hasses suggestions. Albert writes:
The work of the German mathematicians on algebras is very interesting
to me and I should like to know all of it if possible and am very pleased
and thankful for the opportunity to communicate with you and know of
your results.
In this connection we have to mention Hasses American paper [Has32c]. We
have already had several occasions to cite this paper. It contains a comprehensive
treatment of cyclic algebras over number fields. The Local-Global Principle for cyclic
algebras over number fields is cited there and used in an essential way. We have seen
in Section 1.7.3 (p. 55) that it was around March 6, 1931 when Hasse, as a follow-up
to his skew congress, had obtained the Local-Global Principle for cyclic algebras
of arbitrary index. On the other hand, Hasses American paper was received by the
71 One year later, in April 1932 Albert presented to the American Mathematical Society for publication a
construction of non-cyclic division algebras of index 4, defined over the rational function of two variables
[Alb32a]. We have already mentioned this in the foregoing section; see footnote 57). These algebras have
exponent 2. Some months later, in June 1932, he was able to give a refined construction. this time of non-cyclic
division algebras which have both index and exponent 4. Albert was 26 then, and throughout his career he
seemed to have been quite proud of this achievement.
61
editors on May 29, 1931. Thus Hasse had conceived and completed this paper in
about two months. In an introductory paragraph to this paper Hasse says:
I present this paper for publication to an American journal and in English
for the following reason:
The theory of linear algebras has been greatly extended through the
work of American mathematicians. Of late, German mathematicians
have become active in this theory. In particular, they have succeeded
in obtaining some apparently remarkable results by using the theory
of algebraic numbers, ideals, and abstract algebra, highly developed
in Germany in recent decades. These results do not seem to be as well
known in America as they should be on account of their importance. This
fact is due, perhaps, to the language difference or to the unavailability
of the widely scattered sources 72
Reading this text and knowing that it has been written in the months March to
May 1931 when the first letters Hasse-Albert were exchanged, we cannot help feeling
that to write this paper in English and to publish it in an American journal, was meant
predominantly as a source of information for Hasses correspondence partner Albert.
It seems that Hasse had observed Alberts high qualifications and great power as a
mathematician, and he knew that Albert was eager to absorb the methods and results
which had been developed in Germany. And so when Hasse wrote the paper, he had
in mind Albert as the first and foremost reader. In fact, Albert informed him (letter
of November 6, 1931) that he was fortunate to read your paper for the editors of the
Transactions. In other words: Albert had to referee the paper, and so he was the first
in America to know its contents, long before the paper finally appeared in 1932. 73
Between Alberts letter of June 30, 1931 and his next letter of November 6 there is
a gap of several months. On Hasses side, this gap can perhaps be explained because
in the summer semester of 1931 he had Harold Davenport as a visitor from England.
Hasse had invited Davenport to stay in his Marburg family home under the condition
that they would speak English only, so Hasse could refresh and upgrade his knowledge
of the English language. It is conceivable that Hasse, besides his mathematical
activities, was now absorbed in his English studies. 74 Moreover, as is well known,
the beginning friendship between Hasse and Davenport induced Hasse to become
72 The last mentioned fact seems to have been quite serious in those times. Even a big and renowned institution
like Columbia University of New York (where Albert stayed in the summer of 1931) did not have the journal
Hamburger Abhandlungen with the important papers of Artin in its library; for in his letter of May 11, 1931
Albert asks Hasse for information about what is contained in those papers because he was not able to obtain
them.
73 In [Cur99] it is said (p. 232) that the results mentioned in Hasses introductory paragraph had already been
well known to Albert. We believe that the AlbertHasse correspondence of 1931 shows that the Hasse letters were
what stimulated Albert to study eagerly in more detail the work on algebras which was conducted in Germany.
74 Perhaps Hasse had written to Albert about his English studies, for in the letter of June 30 Albert wrote: Your
English is very clear and understandable. I only wish I could write German half so well!
62
interested in Davenports work on the solution number of congruences, which later

led to Hasses proof of the analogue of the Riemann hypothesis for elliptic function
fields thereby starting a series of seminal papers of Hasse on algebraic function
fields. We know that after the end of the summer semester 1931 the Hasses together
with Davenport went on an extended tour through central Europe (in Davenports car)
which ended only in September at the DMV-meeting in Bad Elster. 75 In view of this
it is plausible that in this summer Hasse did not find the time to work intensively on
algebras and to write letters about it.
However in October 1931 Hasse seemed to have again taken up his work on
algebras. And he informed Albert about his results. Alberts reply of November 6
starts with the following text:
I received your very interesting communication this morning and was
very glad to read of such an important result. I consider it as certainly
the most important theorem yet obtained for the problem of determining
all central division algebras over an algebraic number field.
What was the result that Hasse had communicated to him, which Albert classified as
the most important result yet obtained ? Taking into account the delivery time
for overseas mail, we conclude that Hasse may have dispatched his letter around
October 20. At that time, as we know, Hasse had not yet obtained a proof of the Main
Theorem. But we remember from Section 1.3 (p. 7) that Hasse had already succeeded
with the proof that every abelian algebra is cyclic, and that he had informed Emmy
Noether about it. He had also informed Brauer. Now we see that Hasse had also
informed Albert at the same time.
We conclude that by now Hasse had accepted Albert as a correspondence partner
on the same level as he had Richard Brauer and Emmy Noether. Thus the triangle of
Brauer, Hasse and Noether had become a quadrangle by the addition of Albert, the
latter however being somewhat apart because of the longer distance, which implied
a longer time for the transmission of mutual information.
This disadvantage of longer distance became apparent soon.
1.8.3 Alberts contributions
Albert in his above mentioned letter of November 6 informed Hasse about his results
which he had obtained during the summer (when there was no exchange of letters
with Hasse), some of which he had already submitted for publication. Together with
Hasses new results on abelian algebras they would lead to interesting consequences,
Albert wrote. And he proposed a joint paper with Hasse.
But before this letter reached its destination, Hasse had found the proof of the
full Main Theorem. We know from Section 1.3 that this happened on November 9,
75 For
all this see our paper [Roq04].
63
when Hasse had received from Brauer and Noether the relevant information. The new
proof 76 turned out to be so simple (trivial as Noether had called it) that Hasses
former results and methods developed in this direction became obsolete. ThusAlberts
letter of November 6 was now superseded by the new development. Nevertheless,
Albert may be conceded an independent share in the proof of the Main Theorem,
as was expressed in the Albert-footnote. Let us describe this in some more detail.
The Albert-footnote consists of three parts. The first part was written when Hasse
composed the first draft for the BrauerHasseNoether paper; this was on November 10 as we have seen in Section 1.3 (p. 9). At this time Hasse had not yet received
Alberts letter of November 6, and so he mentioned Alberts contributions which
he knew at that time, i.e., those which were contained in the Transactions papers
[Alb31b] and [Alb31d]. These papers had been announced to him by Albert in his
letters of March 23 and May 11 respectively. In particular, theorem 23 of [Alb31d]
is mentioned in the Albert-fotnote. This theorem reads as follows:
Theorem 23. Let A be a central division algebra over K of prime
power index p s > 1, and M a maximal commutative subfield of A.
Then there exists a field extension L0 jK of degree prime to p such that
AL0 D A K L0 is a central division algebra over L0 with maximal
commutative subfield L D M K L0 , such that there is a chain of fields
L0 L1 Ls1 Ls D L
where each Li jLi1 is cyclic of degree p (1 i s). 77
Comparing this with the reductions of Brauer and Noether as presented in section 1.4.2
(p. 14) we observe that both statements are very similar to each other. Moreover,
Alberts proof of theorem 23 contained the same ingredients as the BrauerNoether
reductions 2 and 3, namely a Sylow argument (like Brauer) and some kind of
induction argument from L0 to L (like Noether). This was the reason why Hasse in
the Albert-footnote mentioned theorem 23 in connection with Brauer and Noether.
But at that time it seemed not yet to be clear whether theorem 23 indeed was
sufficient to replace completely the BrauerNoether arguments. This question was
cleared up only later when Alberts letter of November 6 had reached Hasse.
Before discussing this and the later letters of Albert let us report on the reaction of
Emmy Noether when she read Hasses Albert-footnote. We recall from Section 1.3
that Hasse had sent a draft of their joint paper to Noether, and she commented on it in
her replies. In her letter of November 12 she writes concerning the Albert-footnote
(i.e., its first part):
76 More
precisely: that part of the proof which consisted in the reductions 2 (Brauer) and 3 (Noether).
notation is ours, not Alberts. Albert formulates this theorem for base fields K of characteristic 0
only, but from the proof it is clear that it holds for any field of characteristic p.
77 The
64
Das mssen Sie aber bei Albert abschwchen; er hat, Satz 19, nur fr
den Fall zyklischer Algebren gezeigt da jeder Primteiler des Index im
Exponenten aufgeht; von der allgemeinen Brauerschen Reduktion kann
ich wenigstens nichts nden.
But you have to weaken your reference to Albert; he has (theorem 19)
shown for cyclic algebras only that every prime divisor of the index
divides the exponent, and I cannot find anything of the general Brauer
reduction.
Theorem 19 refers to the paper [Alb31c] of Albert. It seems that Hasse in his reply
protested and pointed out that Albert indeed had essentially the full theorem in question. For in Noethers next letter on November 14, obviously replying to Hasses
protest, she writes:
Ich habe Albert noch einmal eingesehen: auch in Satz 20 handelt es sich
nur um zyklische Algebren; Auch spter bleibt die Voraussetzung des
Zyklischen
I have again checked Albert; also in Theorem 20 only cyclic algebras are
investigated; And later too he keeps the assumption of cyclicity
And she proposes that Hasse should change the footnote; it should be said that Albert
did not have the full result, only in the cyclic case. But Hasse seems to have insisted
on his point of view, and to have explained the situation to Emmy Noether. For in
her next letter of November 22, she wrote:
Gut da Sie die Sache mit Albert in Ordnung gebracht haben: da die Hefte
noch ungebunden waren, dachte ich nicht daran die brigen einzusehen,
als ich die Arbeit zu haben glaubte. Er scheint mir also wirklich etwas
zu knnen! Mit der Funote bin ich ganz einverstanden.
It is good that you have settled the Albert case. Since the fascicles were
still unbounded it did not come to my mind to look at the others since I
believed I had the paper in question. It now seems to me that he really
is very able! 78 Now I quite agree with your footnote.
In other words: The volume of the journal in question (Transactions AMS, vol. 33)
came in several parts (the whole volume had 999 pages!) and those parts were still
not bound together in the Gttingen library when Noether looked for Alberts paper.
Noether had studied only one of those parts and, hence, read only one of Alberts
papers. 79 So she was not aware about all the relevant results of Albert. She had been
advised of this by Hasse and now she was happy that Hasse had settled the case in
the footnote. And Noether added the remark:
78 Certainly,
this comment from Emmy Noether means high praise for Albert.
[Alb31c] which are of relevance here, Albert had a third paper in the same
volume, namely [Alb31d].
79 Besides the two papers [Alb31b],
65
Da brigens alle Leute den Beweis nden, kommt einfach daher, da

Sie ihn gefunden haben. Denn das noch Fehlende war trivial fr jeden,
der nicht wie Sie in die Sache verbohrt war.
By the way, the fact that now all people find this proof, is a consequence
of the fact that you have found it first. What was lacking was trivial for
everybody who was not completely absorbed in the details of proof as
you have been.
Here she refers to the fact that in the case of cyclic algebras Hasse had already proved
the Local-Global Principle in [Has32c]. And the generalization to arbitrary algebras
she now considered as being trivial for everybody (although she and Brauer and
Albert had had a hard time to do it). From todays viewpoint we would agree with her.
But we have already stated earlier that we should not underestimate the difficulties
which former generations of mathematicians had to overcome before they could settle
the questions which seem to be trivial for us today.
Comparing the dates: The last mentioned Noether letter had been written on
November 22, in reply to a letter of Hasse. Albert had dispatched his letter to Hasse
on November 6. So we may assume that Hasse had received Alberts letter around
November 20, upon which he had inserted the second part (Added in proof) into
the Albert-footnote, and had Noether informed about this. Noethers letter of November 22 which we have cited above would have been her reply to this.
In this second part of the Albert-footnote Hasse listed the three results a), b),
c) of Albert, which Albert had mentioned to him in his letter of November 6 (see
page 56). However there arises some question concerning statement c). We have
checked Alberts letter and found that Hasses statement is precisely as Albert had
written to him. But this statement is not what Albert has proved in his papers and
what in later letters he referred to. The difference is that in statement c) the degree
of 0 (in Hasses notation) over the base field is required to be prime to p, whereas
Albert later in his letters and in his work 80 does not insist that this is the case. In fact,
using the above theorem 23 it is not difficult to show that the field 0 WD Ls1 has
the property that D K Ls1 is similar to a cyclic division algebra of index p. But
the degree Ls1 W K is divisible by p s1 and not prime to p.
This weak form of c) (where it is not required that the degree of 0 is prime to p)
appears as theorem 13 in Alberts [Alb31a], so let us simply call it theorem 13 in
the following, as Albert does in his letters. This theorem 13 is quite sufficient for
the proof of the Local-Global Principle in case K is an algebraic number field, as is
easily seen.
For, suppose that the non-trivial division algebra D over K splits everywhere; we
may suppose without restriction of generality that the index of A is a prime power
p s > 1. According to theorem 13 there is a finite field extension K 0 of K such that
80 Including Alberts
Colloquium Publication Structure of algebras [Alb39].
66
D 0 D D K K 0 is of index p and has a cyclic splitting field. Now, since D splits

everywhere so does D 0 . By the HilbertFurtwngler Norm Theorem it follows that
D 0 splits, i.e. has index 1. Contradiction.
We see that indeed, the Local-Global Principle follows just in a few lines from
Alberts result, as has been said in [Zel73]. However, by looking more closely into
the matter it turns out that theorem 13 is based on theorem 23 whose proof uses
the same arguments as BrauerNoether (as we have said above already). Conversely,
theorem 13 is immediate if one uses the chain of arguments given by Brauer and
Noether (p. 14). In other words: both methods, that of BrauerNoether and that of
Albert, are essentially the same, the differences concerning non-essential details only.
It is not an unusual mathematical story: A major result is lurking behind the
scenes, ready to be proved, and more than one mathematician succeeds. So this
happened here too.
1.8.4 The priority question
Let us return to November 9, 1931, the date when Hasse found the last steps in the
proof of the Main Theorem. We have seen in Section 1.3 that Hasse immediately
informed Noether and Brauer about it. But he also informed Albert. Hasses letter to
Albert was dated November 11. This letter crossed paths with Alberts of November 6.
Albert received it on November 26, and his reply is dated November 27. There
he congratulated Hasse to the remarkable theorem you have proved. But in the
meantime, he added, he had already obtained results which also could be used to
prove the Local-Global Principle; they are contained in a paper (which Albert had not
mentioned to Hasse before) in the Bulletin of the AMS [Alb31a]. Albert pointed out
that it had been submitted on September 9 and the issue of the Bulletin had already
been delivered in October. It contained (among other results) the theorem 13 which
we have just discussed. Albert writes:
As my theorems have already been printed I believe that I may perhaps
deserve some priority of your proof. I may say, however, that the remarkable part of your proof for me is the obtaining of the cyclic field. I
of course knew your theorem 3.13.
It is not clear what theorem Albert refers to when he cites 3.13. Neither the Brauer
HasseNoether paper nor Hasses American paper contains a theorem with this number. From an earlier part of Alberts letter it seems probable that theorem 3.13 may
stand for the Norm Theorem (see Section 1.4.1). But in Hasses American paper this
has the number (3.11); thus it may have been just a misprint on the side of Albert.
When Albert speaks of obtaining of the cyclic field to be the remarkable part
of Hasses proof then he refers to the Existence Theorem which we have discussed in
Section 1.5.2. This shows that Albert immediately saw the weak point of the Brauer
HasseNoether paper. For, the existence of the required cyclic field was not proved
67
in the BrauerHasseNoether paper. Albert expressed his hope that the existence of
the cyclic field
will be made clear when you publish your proof. In all my work
on division algebras the principal difficulty has been to somehow find a
cyclic splitting field. This your p-adic method accomplishes.
Probably Albert meant not just some cyclic splitting field but a splitting field of degree
equal to the index of the division algebra. As we have seen in Section 1.5.2 the story
of this Existence Theorem is not quite straightforward. But if he really wanted only to
find a cyclic splitting field of unspecified degree, then the proof of the relevant weak
existence theorem was contained in Hasses paper [Has33a], as we had mentioned
already in Section 1.5.4.
Hasse responded to Alberts wish for some priority by extending the Albertfootnote a second time, adding a third part where he stated what Albert had written
to him in his letter of November 26. But when Hasse wrote in the footnote that
Alberts paper is currently printed (im Druck befindlich), then this may have
been a misinterpretation of Alberts text in the letter, which reads as follows:
The part of the proof which you attribute to Brauer and Noether is already
in print.
Obviously, Albert meant that the paper has already appeared (namely in October) and
thus is available in printed form. The German translation of in print would be im
Druck. But in German, if it is said that some paper is im Druck then the meaning
is that the paper is in press, i.e., in the process of being printed. This may have led
Hasse to the wrong translation of Alberts text. In any case, neither Hasse nor Noether
nor Brauer had yet seen Alberts Bulletin paper. If indeed it had appeared in October
in the U.S.A. then it was not yet available in German libraries in the beginning of
November.
We conclude that Alberts Bulletin paper and the BrauerHasseNoether paper
had been written independently of each other. On the other hand, as Albert points
out correctly, his results in his Bulletin paper can be used to prove the Local-Global
Principle for algebras and hence indeed constitute an independent share in the proof
of the Main Theorem. We have discussed this in the foregoing section.
Responding to Alberts wish for priority, Hasse did two more things besides extending the Albert-footnote a second time. First, he sent Albert a copy of the proof
sheets of the BrauerHasseNoether paper, so he could check in particular the actual
text of the Albert-footnote. Secondly, Hasse suggested that they write a joint paper,
to be published in the Transactions, documenting the sequence of events which led to
the proof of the Main Theorem, on the Albert side as well as on the side of Brauer
HasseNoether. Albert should write up the article. In his letter of January 25, 1932
Albert reported:
68
I have finally found time to write up the article by both of us A determination of all normal division algebras over an algebraic number field for
the Transactions. I gave a historical sketch of the proof, my short proof
and a slight revision (to make it more suitable for American readers) of
your proof.
This 5-page paper [AH32] appeared in the same volume as Hasses American paper
[Has32c]. It documents for the mathematical community the chain of events which
we have extracted here from the AlbertHasse correspondence.
Irving Kaplansky in his memoir on Albert [Kap80] writes:
In the hunt for rational division algebras, Albert had stiff competition. Three top German algebraists (Richard Brauer, Helmut Hasse,
and Emmy Noether) were after the same big game It was an unequal
battle, and Albert was nosed out in a photo finish. In a joint paper with
Hasse published in 1932 the full history of the matter was set out, and
one can see how close Albert came to winning.
This comparison with a competitive sports event reads nicely but after studying the
AlbertHasse correspondence I have the impression that it does not quite reflect the
situation. In my opinion, it was not like a competitive game between Albert on the
one hand and the trio BrauerHasseNoether on the other. Instead it was teamwork,
first among Brauer, Hasse, Noether and then, starting March 1931, Albert joined the
team as the fourth member. Within the team, information of any result, whether small
or important, was readily exchanged with the aim to reach the envisaged common
goal. If a comparison with a sports event is to be given, then perhaps we can look at
it as a team of mountaineers who joined to reach the top. The tragedy was that one
of the team members (Albert) in the last minute lost contact with the others (because
communication was not fast enough) and so they reached the summit on different
trails in divided forces, 3 to 1.
Nevertheless, Albert in this situation was upset, which of course is quite understandable. Zelinsky in [Zel73] writes that Albert was hurt and disappointed by this
incident. I may be allowed to cite Professor Zelinskys answer when I asked him
about his memories of Albert.
Perhaps my use of the word hurt was injudicious, since besides the
mental pain that Albert must have felt, the word could connote feelings
that he had been taken advantage of, that his correspondence with Hasse
was used without due consideration. I have no evidence that he felt
that way. You are correct, he was content at last with the resolution of
the priority questions. And by the time I knew him, he had become
established as a leading mathematician in his own right, which surely
affected his attitude toward events of the previous decade.
69
We believe that, indeed, one can infer from the letters of Albert that he was content
with the solution of the priority question as offered by Hasse. The exchange of letters
with mathematical results continued in a friendly tone. We have the impression that
from now on the tone of the letters was even more open and free. At one time Albert
criticised an error which Hasse had made in a previous letter, and at another occasion
Hasse pointed out an error of Alberts in an open and friendly way. At the end of his
letter of January 25, 1932 Albert added a somewhat remarkable postscript as follows:
Permit me to say I did not believe it possible for mere correspondence to
arouse such deep feelings of friendship and comradeship as I now feel
for you. I hope you reciprocate.
We do not know Hasses answer. In Alberts next letter (April 1, 1932) he thanks
Hasse for sending photographs and books. Moreover we read in this letter:
I am very pleased to have been asked to write a report on linear algebras
for the Jahresbericht. I shall certainly accept this kind of proposition I
shall study your report and try to understand better precisely what type
of report you wish me to write.
Here, your report means Hasses class field report [Has30a].
It seems that Hasse wished to revive the long tradition of the Jahresbericht of
the DMV, to publish reports on recent developments in fields of current interest. 81
However, just at the same time there arose a stiff competition, in the concept of
Springer-Verlag to initiate a series of books called Ergebnisse der Mathematik. We
do not know whether it was this competition or there were other reasons that the
Jahresbericht report series was not continued by the DMV. In fact, Hasses report
[Has30a] turned out to be the last one in this series and the plan to publish Alberts
report in this series was not realized. 82
In some of the next letters which followed, Albert still tried to convince Hasse that
the arguments in the BrauerHasseNoether paper included unnecessary complications whereas his chain of arguments was, in his opinion, shorter and more lucid. This
is understandable because as the author he was more familiar with his own version.
But we have the impression that Hasse was not convinced, although he indeed had
high respect for Alberts achievements which had been obtained in a relatively short
81 One
of the best known such reports is Hilberts Zahlbericht 1897.
82As a side remark we may mention that already in 1931 Albert had been asked to write a survey on the theory
of algebras, just by the competitor of the Jahresbericht series, the newly inaugurated Springer series Ergebnisse
der Mathematik. But after a while this proposal was withdrawn by Otto Neugebauer, the editor of the series.
According to Albert (letter of December 9, 1931) Neugebauer wrote that it had been arranged with Deuring
to write a survey on Hypercomplex Systems, and that he (Neugebauer) had discovered just now only that
this is the same subject as Algebras. Deurings book appeared 1934 with the title Algebren. Actually, we
know that Neugebauer had first approached Emmy Noether to write such a survey but she declined and, instead,
recommended her best student Deuring for this task. As to Deurings book, see also [Roq89]. It seems that
Albert after these experiences decided to write his book on the Structure of Algebras [Alb39] independently and
publish it elsewhere.
70
time after their correspondence started. Sometimes Hasse scribbled some marginal
notes on Alberts letters as reminders for his reply, and from this we can extract at
least some marginal information about Hasses letters to Albert. On April 1, 1932
Albert wrote that he believed
your whole Transactions paper could be simplified considerably if
this reduction had been made to begin with. Of course it is a matter of
personal taste and you may even yet not agree.
Here, Albert has in mind the reduction from arbitrary algebras to those of prime
power index by means of Brauers product theorem which he, Albert, had discovered
independently.
To the second sentence we find the note yes! on the margin of the letter, written
by Hasses hand. This seems to imply that Hasses personal taste was somewhat
different. After all, as we have seen in Section 1.4.2, Hasses first draft contained
this reduction to prime power index, and it was Emmy Noether who threw this away
because it was unnecessary. The first sentence is commented by no! (class field
theory!). Here Hasse refers to the close connection of the theory of algebras to
class field theory something which was outside of the realm of interest of Albert,
except that he admitted class field theory as a means to prove theorems on algebras,
if necessary. See also Section 1.6.3.
1.8.5 Remarks
It seems that Hasse had invited Albert to visit him in Germany. In his letter of June 30,
1931 Albert wrote:
I hope that in perhaps two years I may visit Germany and there see you
and discuss our beautiful subject, linear algebras.
On November 27, 1931 the plans had become more specific:
I am very glad that you are interested in the possibility of my visiting
you. I hope that I will be able to leave Chicago on Sept 1, 1933 to return
here not later than Dec 31, 1933. I do not believe I can make the trip
before that time.
Due to the disastrous political events which took place in Germany in the year
1933 these plans could not be realized. Instead, Albert applied for and received an
appointment at the Institute for Advanced Study in Princeton for the academic year
1933/34. There are two letters from Albert in Princeton to Hasse which are preserved.
They do contain interesting material but this is not immediately connected with the
Main Theorem, hence we will not discuss it here.
While Albert was in Princeton in 1933/34 he met two-thirds of the German team,
Richard Brauer and Emmy Noether, who had been forced to leave their country.
71
On the relationship between Albert and Brauer, Mrs. Nancy Albert, the daughter of
A. A. Albert, reports: 83
I have news about Brauer from letters written to my parents. When
Brauer arrived in America from Germany, he spoke little English, and
was rather traumatized from all that he had been through. He stopped
at Princeton in 1933. My father took Brauer under his wing, made him
feel welcome, and took him to meet Wedderburn. Later, my father put
together a large mathematical conference in Chicago, where the Alberts
hosted a large dinner party, and the Brauers became good friends of my
parents. Their relationship continued the remainder of my fathers life.
Mathematically, however, Alberts and Brauers work went in somewhat different
directions. Albert continued to work on algebras, including more and more nonassociative structures. Brauer concentrated on group theory and representations.
Most of the work on finite simple groups and their classification can be traced to
his pioneering achievements, and he advanced to one of the leading figures on the
international mathematical scene (J. A. Green).
About the Princeton contact of Albert with Noether we have the following information. In an undated letter to Hasse from Princeton, probably written in January
1934, Albert writes:
I have seen R. Brauer and E. Noether. They passed through here and
stayed a short while.
And on February 6, 1934:
E. N. speaks here tomorrow on Hypercomplex numbers and Number
theory.
Emmy Noether herself, in her letters to Hasse, is somewhat more detailed about the
mathematical life in Princeton. In a letter of March 6, 1934 she writes the following
report, and we note that Albert is mentioned:
habe ich seit Februar einmal wchentlich eine Vorlesung in Princeton angefangen am Institut und nicht an der Mnner-Universitt, die
nichts Weibliches zult Ich habe mit Darstellungsmoduln, Gruppen
mit Operatoren angefangen; Princeton wird in diesem Winter zum erstenmal, aber gleich grndlich, algebraisch behandelt. Weyl liest auch
Darstellungstheorie, will allerdings zu kontinuierlichen Gruppen bergehen. Albert, in einem Leave of absence dort, hat vor Weihnachten
etwas hyperkomplex nach Dickson vorgetragen, zusammen mit seinen
Riemann matrices. Vandiver, auch Leave of absence, liest Zahlentheorie, zum ersten Mal seit Menschengedenken in Princeton. Und von
83 Personal
communication.
72
Neumann hat nach einem berblick von mir ber Klassenkrpertheorie im mathematischen Klub gleich zwlf Exemplare Chevalley als
Lehrbuch beordert (Bryn Mawr soll auch etwas davon bekommen!).
Dadurch erfuhr ich auch, da Ihre Ausarbeitung ins Englische bersetzt wird, jetzt hoffentlich in gengend vielen Exemplaren darauf
hatte ich die Leute schon gleich im Herbst gehetzt. Ich habe wesentlich
Research-fellows als Zuhrer, neben Albert und Vandiver, merke aber
da ich vorsichtig sein mu; sie sind doch wesentlich an explizites Rechnen gewhnt, und einige habe ich schon vertrieben!
I have, since February, started a lecture in Princeton once a week
at the Institute and not at the mens-university which does not admit
anything female At the beginning I have started with representation
modules, groups with operators. This winter Princeton is treated algebraically, for the first time but quite thoroughly. Weyl also lectures about
representation theory but will soon switch to continuous groups. Albert,
in a leave of absence there, has last year lectured on something hypercomplex in the style of Dickson, together with his Riemann matrices.
Vandiver, also leave of absence, lectures on number theory, the first
time in Princeton since time immemorial. And after I had given a survey on class field theory in the Mathematics Club, von Neumann has
ordered twelve copies of Chevalley as a textbook (Bryn Mawr also shall
get a copy). On this occasion I was told that your Lecture Notes will
be translated into English, now hopefully in sufficiently many copies
I had recommended this already in the fall. My audience consists essentially of research fellows, besides Albert and Vandiver, but I noticed
that I have to be careful; these people are used to explicit computations,
and some of them I have already driven away!
We can safely assume that Albert was not one of the dropouts from Noethers course.
He knew about the importance of Emmy Noethers viewpoint on algebra and on
the whole of mathematics. Noethers ideas have often been described and so we
will not repeat this here. 84 But at the time we are considering, Noethers ideas
had not yet penetrated mathematics everywhere. Albert himself had his training with
Dickson, and his papers in those first years of his mathematical activity were definitely
Dickson style. It was only gradually that Albert started to use in his papers the
Modern Algebra concepts in the sense of Emmy Noether and van der Waerden. In
1937 Albert published the book Modern higher algebra [Alb37] which was a student
textbook in the modern (at that time) way of mathematical thinking. 85
Albert explicitly stated that his textbook was meant as an introduction to the
methods which will be used in his forthcoming book on algebras. That second book
84 See,
85 The
e.g., Hermann Weyls obituary address in Bryn Mawr 1935 [Wey35].

book was refereed in the Zentralblatt by Helmut Hasse.
73
appeared in 1939 [Alb39] with the title Structure of Algebras, and it was also written
with the viewpoint of Modern Algebra. 86 It seems to us that to a large degree this
was a direct consequence of Albert being exposed at Princeton to Emmy Noethers
influence.
In the preface to his book Albert says:
The theory of linear associative algebras probably reached its zenith
when the solution was found for the problem of determining all rational
division algebras. Since that time it has been my hope that I might
develop a reasonably self-contained exposition of that solution as well
as of the theory of algebras upon which it depends and which contains
the major portion of my own discoveries.
We do not intend here to give a review of Alberts book which, after all, is well
known and has become a classic. It is our aim here to point out that to a large extent
the book is the outcome of his participation in the team together with Hasse, Brauer
and Noether notwithstanding the fact that the book contains also other aspects of
the theory of algebras, e.g., Riemann matrices and p-algebras.
But when Albert said that the theory of algebras had reached its zenith with
the Main Theorem then we cannot agree. Since then a number of highly important
results have been established, and the theory is still flourishing.
1.9 Epilogue: Kte Hey

In the history of mathematics we can observe not infrequently that after an important
result has been found, it is discovered that the very same result, in more or less explicit
form, had been discovered earlier already. This happened also to the Local-Global
Principle for algebras which is the basis for Hasses proof of the Main Theorem.
On January 26, 1933, one year after the appearance of the BrauerHasseNoether
paper, the editor of the Hamburger Abhandlungen received a manuscript of a paper
[Zor33] which begins as follows:
Die Theorie der -Funktion eines Schiefkrpers ist von Frulein K. Hey
in ihrer Dissertation (Hamburg 1929 ) eingehend entwickelt worden: ich
mchte in dieser Note auf die arithmetischen Konsequenzen, die dort aus
der analytischen Theorie gezogen werden, aufmerksam machen und sie
auf Grund einiger Korrekturen und Abrundungen als
neuen Beweis eines Hauptsatzes fr Algebren sowie
des allgemeinen quadratischen Reziprozittsgesetzes
86 This
book was refereed in the Zentralblatt by W. Franz, a former Ph. D. student of Hasse.
74
erkennen lassen. Der in Frage stehende Hauptsatz ist die Grundlage fr

die Herleitung des Reziprozittsgesetzes mit nichtkommutativen Mitteln; seine independente Begrndung ist also methodisch wichtig.
The theory of the -function of a skew field has been developed in detail
by Miss K. Hey in her dissertation (Hamburg 1929). In the present note
I would like to draw attention to the arithmetic consequences which are
derived there, so that after some correction and streamlining they are
recognized as a
new proof of a main theorem on algebras
and of the general quadratic reciprocity law.
The said main theorem on algebras is the basis for deriving the reciprocity law with non-commutative methods; therefore its independent
foundation is important for methodical reasons.
The main theorem which is meant here is not quite the Main Theorem of Brauer
HasseNoether but the Local-Global Principle as formulated in Section 1.4. The
general quadratic reciprocity law is extra mentioned by the author because it follows
directly from Kte Heys treatment in the case of quaternion algebras. In the next
sentence however, reciprocity law means Artins reciprocity law; to derive this from
the Main Theorem one had to follow Hasses method as explained in Section 1.6.
The author of this article was Max Zorn, a former Ph. D. student of Artin in
Hamburg. 87 He had been the second Ph. D. student of Artin, the first one had been
Kte Hey whose thesis he is referring to in his note. She received her degree in 1927.88
Her thesis [Hey29] had never been published in a mathematical journal but it was
printed, and was distributed among interested mathematicians. We know that Hasse
and Emmy Noether each owned a copy, perhaps Richard Brauer too. The thesis was
refereed in the Jahrbuch fr die Fortschritte der Mathematik, vol. 56.
The aim of Heys thesis was to extend the known methods of analytic number
theory to division algebras instead of number fields in particular the methods of
Hecke which lead to the functional equation of zeta functions. She defined the zeta
function D .s/ of a division algebra D whose center K is an algebraic number field.
But she considered only the finite primes p of K. If that function is supplemented
by factors corresponding to the infinite primes of K (which today is the standard
procedure) then the analytic treatment of that extended function, including its functional equation, shows that, if compared with the zeta function K .s/ of the center, it
admits two poles (if D K), which in some way correspond to primes p which are
87 Zorn
(19061993) received his Ph. D. 1930 with a paper on alternative algebras. In 1933 he was forced
by the Nazi regime to leave Germany. His name is known to the mathematical community through his Zorns
Lemma.
88 Kte Hey (19041990) left the university some time after she had obtained her degree, then she became a
teacher at a gymnasium. More biographical details can be found in [Lor05] and [Tob97].
75
ramified in D, i.e., at which Dp does not split. Indeed, the existence of such primes
is the content of the Local-Global Principle. Hey used Artins paper [Art28b] on the
arithmetic of algebras, and also the full arsenal of analytic number theory known at
that time which centered around Heckes work. Deuring says in [Deu35]:
Dieser Beweis des Hauptsatzes ber Algebren ist gleichsam die strkste Zusammenfassung der analytischen Hilfsmittel zur Erreichung des
Zieles.
In a way, Heys proof of the Main Theorem represents the strongest
concentration of analytic tools to reach the aim.
For a discussion of Heys thesis and Zorns note we refer to the recent essay
[Lor05] by Falko Lorenz.
Heys thesis is considered to be difficult to read. It seemed to be generally known
at the time, at least among the specialists, that Heys thesis contained errors. But Zorn
points out how those errors could be corrected in a quite natural and straightforward
manner.
Thus if Hasse (or Noether, or Brauer, or Albert) had known this earlier, then the
proof of the Main Theorem could have been completed earlier. It is curious that Heys
thesis had not been mentioned in the correspondence of Hasse, not with Artin, not
with Noether, Brauer or Albert. At least not before Zorns note became known. 89
Later, when Hasse and Noether discussed how much analysis should and could
be used in class field theory, Hasse wrote to her (letter of November 19, 1934):
wenn man schon einmal Analysis zur Begrndung der Klassenkrpertheorie braucht, man dann die Kanone auch gleich auf den Normensatz, das Summen-Theorem fr die Invarianten und den Satz von
den berall zerfallenden Algebren richten soll. Konsequenz: Man verwende die Heysche Kanone und dann Rckschuss auf die klassische
Klassenkrpertheorie wie zu Ihrem 50. Geburtstag.
if analysis is to be used in the foundation of class field theory then
one should aim with Heys cannon at the norm theorem, the sum formula
for the invariants of algebras, and the theorem on algebras splitting everywhere. Consequently, one should use Heys cannon and then aim
backwards to class field theory in the classical sense, like on your 50th
birthday.
89 Emmy
Noether got to know Zorns note some time in winter 1932/33. She was so impressed that she
suggested to two of her Ph.D. students to continue work in that direction. One of those students was Ernst Witt
who in his thesis transferred Heys results to the function field case. The other student was Wolfgang Wichmann
who presented a much simplified proof of Heys functional equation of the zeta function of a division algebra,
however up to a sign only.
76
Here, the reference to Noethers 50th birthday is to be read as the reference to Hasses
paper [Has33a] which he had dedicated to her on the occasion of that birthday; we
have mentioned it several times in this paper.
In his letter Hasse pointed out that no one at that time had been able to develop class
field theory without using methods from analytic number theory. His own approach
in the birthday paper [Has33a] is based on the HilbertFurtwngler Norm Theorem
which in turn was proved using analytic methods of zeta functions of number fields.
But soon after, Chevalley [Che35] succeeded to give a foundation of class field theory
free from analysis; see also [Che40].
Chapter 2
The remarkable career of Otto Grn
From FLT to finite groups. The remarkable career of Otto Grn.

Jahresbericht der Deutschen Mathematiker Vereinigung 107 (2005), 117154
(Section 2.9 has been added.)
2.1 Introduction
2.2 The first letters: FLT (1932)
2.3 From FLT to finite groups (1933)
2.4 The two classic theorems of Grn
2.5 Grn meets Hasse (1935)
2.6 The Burnside problem (1939)
2.7 Later years (after 1945)
2.8 Epilogue
2.9 Addendum
77
78
85
90
98
104
109
113
115
2.1 Introduction
Students who start to learn the theory of finite groups will soon be confronted with
the theorems of Grn, at least with Grns first and second theorem, and with
its generalizations.. These theorems found their way into group theory textbooks
immediately after their publication in the mid 1930s, with the comment that they are
of fundamental importance in connection with the classical Sylow theorems. But
little if anything is known about the mathematician whose name is connected with
those theorems.
Recently, scanning through the legacy of Helmut Hasse which is kept at the
University Library in Gttingen, I found 50 letters which were exchanged between
Hasse and Grn, from 1932 to 1972. Hasse is known to have had an extended
correspondence, freely exchanging mathematical information with his colleagues.
Thus at first sight, I was not really surprised to find the name of Otto Grn among
Hasses many correspondence partners.
78
But while reading these letters there unfolded to me much more than just mathematical information, namely the remarkable and fascinating story of a mathematician,
quite rare in our time, who was completely self-educated, without having attended
university, and nevertheless succeeded, starting at age 44, to give important contributions to mathematics, in particular to group theory.
I am writing this article in order to share this discovery with other interested
mathematicians. But I would like to make it clear that this is not meant to be a complete
biography of Otto Grn. This article comprises mainly what we can conclude from
the correspondence files of Hasse and some secondary sources, with emphasis on the
genesis of Grns main theorems. Perhaps a more detailed search of other sources
could bring to light more facets of Grns personality and work.
Acknowledgement. I had sent a former version of this article to a number of people
who (like myself) had met Grn and still remember him. I would like to thank all
colleagues for their interest and for their various comments on the work and the
personality of Grn. In particular I would like to thank B. Huppert and W. Gaschtz
for their help concerning the group theoretic part of Grns work. It seems to me that
a more detailed survey of Grns role in the development of group theory would be
interesting and worthwhile. Last but not least I would like to thank the referee for
several well founded comments.
2.2 The first letters: FLT (1932)

2.2.1 Grn and Hasse in 1932
Little is known about the early life of Grn. In his vita which he wrote in 1955 we
read:
Ich bin am 26. Juni 1888 zu Berlin geboren, besuchte das FriedrichWerdersche Gymnasium zu Berlin, das ich 1908 mit dem Reifezeugnis
verlie. Zunchst widmete ich mich dem Bankfach, nahm am ersten
Weltkriege teil, ohne Schden davonzutragen, und war weiterhin kaufmnnisch ttig.
I was born on June 26, 1888 in Berlin. I attended the Friedrichs-Werder
Gymnasium in Berlin until 1908. Then I worked in the banking business,
participated in the first world war without being injured, and afterwards
continued to work in a commercial job.
It is not known what kind of job this had been. 1 Grn continued:
1 But
see our Addendum at the end of this article.
79
Da ich stets lebhaftes Interesse fr mathematische Fragen hatte, beschftigte ich mich nebenbei wissenschaftlich und kam auf diese Art zu einem
Briefwechsel mit dem berhmten Algebraiker Helmut Hasse
All the time I had strong interest in mathematical problems, and in my
spare time I occupied myself with scientific problems. 2 In the course
of this activity there started an exchange of letters with the famous algebraist Helmut Hasse 3
The first letter of Otto Grn to Hasse is dated March 29, 1932, from Berlin. At that
time Grn was (almost) 44 years old.
Otto Grn in 1957
Helmut Hasse, 10 years younger than Grn, at that time was professor of Mathematics at the University of Marburg (since 1930) as the successor of Kurt Hensel. The
years in the late twenties and early thirties are to be regarded as the most fruitful period in Hasses mathematical life. Hasse had completed the last part of his class field
theorys report [Has30a], he had proved (with Richard Brauer and Emmy Noether) the
Local-Global Principle for simple algebras [BHN32], he had determined the structure
of cyclic algebras over a number field [Has32c], he had discovered local class field
2 Here
and in the following we use our own free translation of German text into English.
I have found the vita which starts with the cited sentences, in the archives of the University of Wrzburg
where Grn had a teaching assignment (Lehrauftrag) during the years from 1954 to 1963 (see Section 2.7.3
below). It is dated August 2, 1955. I do not know the occasion for which Grn had presented this to the university.
Probably it was connected with his teaching assignment. I am indebted in particular to Hans-Joachim Vollrath
for his help to obtain access to the Wrzburg archives.
3
80
theory [Has30c], given a new foundation of the theory of complex multiplication

[Has27b], [Has31b], and more. In Hasses bibliography we have counted more than
fifty papers in the period from 1926 to 1934. In March 1932, when he received Grns
first letter, he had just completed his seminal paper dedicated to Emmy Noether on
her 50th birthday [Has33a], where he presented a proof of Artins Reciprocity Law
in the framework of simple algebras and at the same time determined the structure
of the Brauer group over a number field. Now he was preparing his lecture course
on class field theory which he was to deliver in the summer semester of 1932, the
notes of which [Has33b] would be distributed widely and would influence the further
development of class field theory. One year later, in March 1933, Hasse would prove
the Riemann hypothesis for elliptic function fields.
It seems remarkable that in the midst of all this activity, Hasse found the time to
deal carefully with the letters of Otto Grn, whom he had never heard of before. Hasse
had the strong viewpoint that every letter from an amateur mathematician represents
an unusual interest in mathematics by the sender and, hence, has to be taken seriously.
And so he did with Grns letter, thereby discovering that the sender was not one of
the usual Fermatists but, despite his lack of formal mathematical education, was
unusually gifted and had a solid mathematical background.
2.2.2 Vandivers conjecture and more
Grns first letter begins as follows:
Sehr geehrter Herr Professor! Ich habe aus Ihren Arbeiten die Takagische
Klassenkrpertheorie kennengelernt. Ich glaube nun, auf dieser Grundlage zeigen zu knnen, da auch im irregulren Krper k./ der `-ten
Einheitswurzeln der 2-te Faktor der Klassenzahl nie durch ` teilbar sein
kann. Darf ich Ihnen vielleicht hier ganz kurz den Beweis skizzieren, zumal da ich als reiner Amateurmathematiker denselben doch nicht verffentlichen wrde.
Dear Herr Professor! From your papers I have learned Takagis class
field theory. I believe that on this basis I can show that also in the
irregular field Q./ 4 of the `-th roots of unity the 2nd factor of the class
number can never be divisible by `. May I sketch briefly the proof since
anyhow, as an amateur mathematician, I am not prepared to publish it.
By class number Grn means the number of ideal classes of the `-th cyclotomic
field Q./ where denotes a primitive `-th root of unity, ` being an odd prime. It is
well known since Kummer [Kum50] that the class number h of Q./ has a product
4 Grn writes k. / (in conformity with the older notation) where we have written Q. / (which is todays
notation). In the interest of the reader we shall freely change notations from the original, whenever it seems
appropriate for better understanding,
81
decomposition
h D h1 h2
where the second factor h2 equals the class number of the maximal real subfield
Q.C 1 /. The first factor h1 is a positive integer, called the relative class number.5
The prime `, or the field Q./, is called regular if the class number h is not divisible
by `. One of the monumental achievements of Kummer [Kum50] was the discovery
that FLT holds for a regular prime number `, which is to say that the diophantine
equation
x` C y` C z` D 0
is not solvable in integers x; y; z 0.
If ` is regular then, of course, both factors h1 and h2 are not divisible by `. If ` is
irregular then it was known to Kummer already that h1 is divisible by `, but nothing
much was known about h2 . Now Grn claimed that h2 , even in the irregular case, is
not divisible by `. This statement is today known as Vandivers conjecture, and it
is considered quite important with respect to the structure of cyclotomic fields. 6
Thus in effect Grn claimed to have proved Vandivers conjecture, although he
did not mention Vandiver in his letter. Most probably he was not aware at that time
of Vandivers work.
Hasse replied on April 1, 1932 already, three days after Grn had dispatched his
letter. We do not know the text of Hasses letter 7 but from Grns answer we can
deduce that Hasse had pointed out the proof to be erroneous. Grn wrote on June 27,
1932:
Gegen Ihre Bedenken kann ich natrlich nichts einwenden; der Beweis
ist eben in der vorliegenden Form miglckt.
Of course there cannot be any objection against your doubts. Thus my
proof has not been successful in this form.
In fact, Vandivers conjecture has not been proved or disproved until today, despite
strong efforts by many mathematicians. 8
5 The terminology first and second factor of the class number is generally used in the literature. But Hasse
in his book [Has52] says that the inverse enumeration would be more natural: h2 should be called the first and
h1 the second factor. In later letters (1957/58) Grn uses therefore this inverse terminology. Hasse himself in
[Has52] writes h for h1 and h0 for h2 . Ribenboim [Rib79] writes hC for h2 .
6 I am indebted to Franz Lemmermeyer for pointing out to me the paper [Van41] in which Vandiver expresses
his hope that h2 is always prime to `. Ribenboim [Rib79] remarks that Vandivers conjecture is already stated
in a letter of Kummer to Kronecker, dated December 28, 1849.
7 Quite generally, the letters from Grn to Hasse are preserved in the Hasse legacy, whereas many of the
letters from Hasse to Grn have to be considered as lost. Only in later years, in case the letters were written
by typewriter, Hasse used to make a carbon copy for himself and so his letters are preserved too. But this was
not the case for his early letters to Grn which were handwritten. In most cases however, by interpolating from
Grns replies we can deduce approximately what Hasse had written.
8 By now the conjecture has been verified for all odd primes ` < 12 106 (communication by Franz
Lemmermeyer).
82
But Hasse had not been content just to point out the error in Grns proof. He
had added some comments for further work. Maybe he also recommended to Grn
some of the relevant literature. For, Grn wrote in his second letter to Hasse (June
27, 1932) the following:
glaube ich aus Ihrem Hinweis eine Folgerung fr die Fermatsche
Behauptung ziehen zu knnen, die ich Ihnen gern mitteilen mchte
Wenn x ` C y ` C z ` D 0 in rationalen Zahlen x, y, z und etwa x durch
` teilbar, yz prim zu ` ist, so mu der zweite Faktor der Klassenzahl
durch ` teilbar sein.
in view of your comments I believe that I can derive the following result towards Fermats Last Theorem which I would like to communicate
to you If x ` C y ` C z ` D 0 is solvable in nonzero rational integers
x, y, z and x is divisible by ` while yz is not divisible by ` then the
second factor of the class number is divisible by `.
In dealing with the Fermat equation one usually distinguishes two cases: In the first
case one assumes that none of x, y, z is divisible by `. In the second case one of
them, say x, is divisible by ` whereas y, z are not. Thus Grns claim says in effect:
If the second class number factor h2 is not divisible by ` then the Fermat
equation is not solvable in the so-called second case.
And he sketched a proof of this result. But again, there was an error which Hasse
pointed out to him in a letter two days later.
We should keep in mind that Grn had not received any formal mathematical
education; mathematically he was completely self educated and this was the first
occasion where he could discuss his ideas with a competent mathematician. The
subject required a high level of sophistication, and after all he had no training in
presenting mathematical ideas. Thus the failure of his first attempts to produce
a consistent proof is understandable. He was lucky to have found Hasse as his
correspondence partner who, it seems, had recognized the mathematical capacity of
the author of those letters.
After some more months, on September 28, 1932 Grn wrote again. He said
that he had indeed observed the difficulty which Hasse had pointed out to him but
had erroneously assumed that this could be handled by the methods of Kummer.
Nevertheless, he now presented a correction of his result, namely with an additional
hypothesis concerning certain divisibility properties of Bernoulli numbers.
The sequence of Bernoulli numbers Bn can be defined as the coefficients appearing
in the power series expansion
X xn
x
D
Bn :
ex 1
n
n0
83
These Bn are certain rational numbers which are well known to be connected with
the structure of the cyclotomic field Q./. Kummer had used the Bn to formulate a necessary and sufficient criterion for ` to be regular. Namely, the numbers
B2 ; B4 ; : : : ; B`3 should not be divisible by `. 9 Kummer had also discussed the
irregular case to some extent, and there too he had given sufficient criteria for the
validity of FLT.
Now Grns additional hypothesis reads as follows:
Es mgen zwar beliebig viele Bernoullische Zahlen Bi mit i < ` 1
durch ` in erster Potenz teilbar sein, jedoch gelte fr keine von ihnen
B`i 0 mod `3 bei geradem i.
Arbitrary many Bernoulli numbers Bi with i < ` 1 may be divisible
by `, but for none of them we have B`i 0 mod `3 with i even.
It was well known, already to Kummer, that this hypothesis implies certain structural
properties of the group of units of Q./. Grn showed that it is sufficient (in addition
to the hypothesis that the second class number factor h2 is not divisible by `) to
deduce that the Fermat equation for exponent ` has no solution in the second case. 10
This time Hasse did not find an error in Grns proof. But he wanted to be sure
that Grns result was new. Perhaps Hasse remembered a paper by Vandiver [Van29]
which in fact contained Grns above cited result. But Grns computations yielded
at the same time a somewhat more general result than we have cited above, showing
the impossibility not only for the Fermat equation in the second case, but also for
certain other diophantine equations within the cyclotomic field Q./, going beyond
Vandivers results. Hence, even if Grns result for the Fermat equation was known,
perhaps his more general result was new?
Thus Hasse proposed that Grn should write to Vandiver at the University of
Texas who was considered to be a specialist on those problems. Grn replied that he
is afraid not to know the proper mathematical terminology in English language, and
anyhow he does not know the address of Vandiver. Upon this Hasse himself wrote to
Vandiver on behalf of Grn. Since several years Hasse had exchanged reprints with
Vandiver and, as can be seen from the correspondence between the two, the latter had
visited Hasse at least twice, once in Halle and another time in Marburg. 11
Vandiver replied in a letter of November 14, 1932:
9 If the index n > 1 is odd then B D 0. Because of this, the enumeration of the Bernoulli numbers is
n
sometimes changed, i.e., writing Bn instead of B2n for n 1. But we will keep the notation as given by the
definition above.
10 Ribenboim [Rib79] (p. 188) says erroneously that Grns result refers to the first case.
11Vandiver too, like Grn, did not have a formal mathematics education. In the biography of Vandiver
(18821973) in [Leh74] we read: This remarkable man was self-taught in his youth and must have had little
patience with secondary education since he never graduated from high school. However, already with 22 years
Vandiver wrote his first mathematical paper whereas Grn was 47 when his first paper appeared.
84
The two theorems you mention appear to be quite new. The first one
seems to be a modification and extension of the Theorem I of my paper
of the year 1929 in the Transactions A. M. S.
Here Vandiver cites his paper [Van29].
This must have been sufficient for Hasse. Perhaps he was aware of the fact that one
year earlier, in 1931, Vandiver had been awarded from the American Mathematical
Society the highly prestigious Cole Prize in number theory for his work on FLT, in
particular for his paper in the Transactions which Vandiver was citing in his letter. If
Grns result was an extension of Vandivers then certainly, it should be published.
Thus Hasse decided to accept Grns manuscript for Crelles Journal.
However, in the form as presented so far Grns manuscript seemed not publishable. Hence Hasse would first do what he always used to do as an editor of Crelles
Journal: He would study the paper carefully and on that basis give advice to the author
to produce a text which, in his opinion, meets the standards of scientific publication.12
But he needed some time for this. Grn replied in a letter of December 19, 1932:
Vielen Dank fr Ihr freundliches Schreiben von 12. Ich bin Ihnen sehr
verpichtet, wenn Sie sich dem Beweis zum Fermatproblem weiter widmen wollen und es ist selbstverstndlich, da Sie jede Frist dazu haben.
Thank you very much for your kind letter of 12.13 I would be very
obliged to you if you would continue to attend to my proof on Fermats
problem, and it is clear that there will be no time limit for this.
We should note that just in this period, the last months of 1932 and the first months
of 1933, Hasse was busy with his attempts to prove the Riemann hypothesis for curves.
We have reported in [Roq04] that in November 1932, when Hasse gave a colloquium
lecture in Hamburg, he had a conversation with Artin who pointed out to him that
his (Hasses) research project on diophantine congruences was in fact equivalent to
the proof of the Riemann hypothesis for the curves in question. This comment by
Artin had decidedly changed the viewpoint of Hasse. He went to work intensively on
this idea with the result that already in March, 1933 he arrived at his first proof for
elliptic curves. In view of this we can understand that Hasse in this period tended to
postpone, if possible at all, other obligations including the reading and correcting of
Grns manuscript. It was May 1933 until he turned to Grns manuscript again.
Grns paper [Gr34b] appeared in 1934. The date of submission is recorded as
May 17, 1933. His result in the final form reads as follows. As above, ` denotes an
irregular prime number and a primitive `-th root of unity. Let k0 D Q. C 1 /.
12 Rohrbach
[Roh98] reports: With his [Hasses] characteristic conscientiousness, he meticulously read and
checked the manuscripts word by word and formula by formula. Thus he very often was able to give all kinds
of suggestions for improvements to the authors, concerning contents as well as form. The correspondence
HasseGrn gives ample witness of this.
13 This means December 19, 1932.
85
If the second class number factor h2 is prime to ` and if none of the

Bernoulli numbers B`n 0 mod ` 3 (for n D 2; 4; : : : ; ` 3) then the
equation
". 1 /m 1` C 2` C 3` D 0
is not solvable in algebraic integers 1 ; 2 ; 3 2 k0 which are prime to
`, provided m 3` 1 and " is a unit in k0 .
This was Grns first publication. Compared with the other existing literature
on FLT it cannot be rated as exceptional. Grn followed the known footsteps in the
direction which had been pointed out by Kummer in the mid 19th century and his
result was close to that of Vandiver [Van29]. But we should keep in mind that FLT
had not yet been proved generally at that time. Hence any partial result which points
towards the validity of FLT was welcomed, even if the progress compared with former
results seemed to be small.
However, if we consider that Grn had originally not been aware of Vandivers
paper and that his result containedVandivers, then we have to rate Grns achievement
as extraordinary in particular if we remember that he had no formal mathematical
training and had reached his high status of expertise through self-education.
2.3 From FLT to finite groups (1933)

In a letter of December 6, 1932 Grn started to discuss other problems; these belong
to general class field theory and are only indirectly connected with FLT. Here we will
not go into all details but restrict our discussion to the following two topics.
2.3.1 Divisibility of class numbers: Part 1
Grn wrote to Hasse:
Ich mchte noch einen Satz beweisen, der vielleicht gelegentlich
gebraucht werden kann: Wenn K den Krper k enthlt und kein Zwischenkrper existiert, der ber k Abelsch mit der Relativdiskriminante 1
ist, so ist die absolute Klassenzahl von K durch die absolute Klassenzahl
von k teilbar.
I would like to prove yet another theorem which may be useful occasionally: If K contains the field k and there is no proper intermediate
field which is abelian over k and of relative discriminant 1 then the class
number of K is divisible by the class number of k.
This is quite interesting. We know that five years earlier Artin had observed the
same fact, and he had found it worthwhile to communicate it to Hasse. Let us cite
from Artins letter of July 26, 1927:
86
Nun etwas anderes, das mir grossen Spass bereitet hat und das ich gestern
im Heckeseminar erzhlte. Das Resultat scheint, so trivial der Beweis
ist, neu zu sein. Eine ganz kindische Vermutung jedes Anfngers ist
doch diese: Ist k Unterkrper von K, so ist die Klassenzahl von k ein
Teiler der Klassenzahl von K. Ich mchte zeigen, dass dies fast immer
richtig ist, mehr noch:
Satz: Enthlt K=k keinen in bezug auf k Abelschen und gleichzeitig unverzweigten Zwischenkrper, so besitzt die Gruppe der absoluten
Idealklassen von K eine Faktorgruppe isomorph mit der Gruppe der
absoluten Idealklassen in k.
Now something else which I had talked about yesterday in Heckes
seminar with great fun. The result seems to be new in spite of the
simplicity of proof. A very childish expectation of every beginner is the
following: If k is a subfield of K then the class number of k divides the
class number of K. Now I show that this is true almost always, and
even more:
Theorem: If Kjk does not contain any intermediate field which is
abelian and at the same time unramified then the class group of K
admits a factor group isomorphic to the class group of k.
Artin proceeds in his letter to describe a proof which, as he had said, is quite
simple. After checking we found that Grns proof was the same as Artins. The
essential fact to be used in the proof is that, under the hypothesis of the theorem, the
absolute class field of k is linearly disjoint to K. It seems that Hasse in his reply to
Grn had mentioned Artin, for Grn wrote in his next letter (December 19, 1932) that
he did not wish to claim priority. 14
Thus again, on his way teaching himself algebraic number theory, Grn had found
for himself a theorem which was familiar to the specialists, this time Artin. Note that
Artin had never published his proof.
But there had been a recent publication by Chevalley [Che31] containing the
same theorem. Certainly Hasse, who at that time was in close contact with Chevalley,
knew about Chevalleys paper, and perhaps he had pointed out that paper to Grn after
receiving Grns letter. At first sight Chevalleys proof looks somewhat different than
that of ArtinGrn but at closer inspection we find that it is essentially the same 15 .
14 Ich
wollte den Satz nicht als mein geistiges Eigentum angesehen wissen.
Let k 0 be the absolute

class field of k. Artin uses only the fact that k 0 is abelian and unramified over k, and that these properties are
preserved after extending the base field from k to K which directly implies the result. Chevalley uses the
Verschiebungssatz (shift theorem) of class field theory in order to describe Kk 0 explicitly as class field over K.
Thus he uses more machinery from class field theory than ArtinGrn. However, if one comes to think of it, the
proof of the Verschiebungssatz in this special case reduces to the argument of ArtinGrn and so, in this sense,
we may regard both proofs as essentially the same.
15 The difference between Chevalleys and Artins arguments can be described as follows:
87
Chevalley mentions that the same result would be contained in a forthcoming paper
by Herbrand which we have found to be Thorme 2 in [Her32]; there Herbrand used
it for a new foundation of Kummers theory of ideal classes in cyclotomic fields. After
checking we found that again, Herbrands proof is the same as Artins and Grns.
From all we know about Grn we have no doubt that he had found his proof
independently, i.e., independent not only of Artin but also of Chevalley and of Herbrand. Grn in his letter cites Hilbert who in his Zahlbericht [Hil97], 117, p. 378 16 ,
mentions that Kummer had stated the above theorem for the subfields K of Q./ but
that Kummers proof was incorrect.17 Of course, Kummers theorem is an immediate
consequence of the general theorem of ArtinGrn since Q./ is purely ramified.
But in Kummers case, i.e., for subfields of Q./ where is a prime power root of
unity, the theorem had been proved much earlier by Furtwngler [Fur08]. Although
in 1908 class field theory was not yet completed by the theorems of Takagi and Artin,
enough was known to prove the divisibility of class numbers which Kummer had
conjectured. It seems that neither Artin nor Grn had been aware of Furtwnglers
proof. But Hasse did know it, for in Hasses diary we have found an entry dated
October 10, 1925 with the title: The ideal class groups of relatively abelian fields.
(Generalization of a theorem of Furtwngler.)18 There, Hasse proved the ArtinGrn
theorem in the special case when Kjk is abelian. Thereby he regards K as class field
over k, thus he used still more machinery from class field theory. As it turned out in
the proof of ArtinGrn, this is not necessary. Here again, as it is the case so often in
Mathematics, the generalization (omitting the assumption that Kjk is abelian) leads
to a simplified proof. 19
2.3.2 Divisibility of class numbers: Part 2
In his letter from December 19, 1932 Grn mentions another problem concerning class
numbers. While his above mentioned result yields a lower bound of the class number
h of K (it is divisible by the class number of a subfield under certain conditions), he
now claimed to have an upper bound for h (under certain conditions it divides the
class number of a subfield times a certain factor dependent on the structure of the
Galois group). This time, however, he is not sure that his arguments are correct, and
so he writes:
Aber ich gestehe Ihnen, verehrter Herr Professor: Ich traue meinen eigenen Ergebnissen nicht; die Stze sind mir zu berraschend. Ich kann aber,
16 The page number refers to the original Zahlbericht whereas its copy in the Collected Papers of Hilbert has
different pagination.
17 The same reference to Hilberts Zahlbericht we have found in Artins letter to Hasse, cited above.
18At the end of this entry Hasse later had added a reference to Artins letter of July 26, 1927 which we have
cited above.
19 By the way, the ArtinGrn theorem with the same proof appears in [ACH65]. There, Hasse cites the letter
of Artin and also his own diary entry of October 10, 1925.
88
soviel ich mich auch bemhe, den Fehler nicht nden. Und deshalb bitte
ich Sie, mir zu sagen, ob und wo in meiner Rechnung ein Fehler steckt.
But I admit, dear Herr Professor, that I do not trust my own results: the
theorems are very surprising to me. However I cannot find the error
although I have tried to. Therefore I am asking you to tell me whether
and where there is an error in my computations.
The situation is the following: Kjk is a Galois extension of number fields 20 . Let
K 0 be the maximal subextension which is abelian over k. Grn proved:
Suppose that the class group of K is cyclic. Then the class number h
of K divides the product of the class number h0 of K 0 with the relative
degree K W K 0 .
Actually Grn wrote that he assumed the cyclicity of the class group of K for
simplicity only, and claimed that his proof could be extended to cover the case of
an arbitrary class group. However that is not the case. Hasse pointed out this fact to
Grn, and we shall see below that this led to remarkable consequences.
In the case of a cyclic class group of K, Grns proof turned out to be correct.
But it seems that Hasse was not sure whether this result was known already, since
he proposed to put this theorem as a problem in the Jahresbericht of the DMV. 21
At that time, the Jahresbericht provided a section where any member could state a
problem, and the incoming solutions were published in the next issue. Quite often
such problems were posed even if the author had already obtained a solution, but he
wished to find out whether a solution, possibly simpler, was known already.
Grn consented and Hasse submitted the theorem (for cyclic class group of K)
under the name of Grn as a problem, which appeared as no. 153 in volume 43 (1934)
of the Jahresbericht. Promptly there were two solutions received, published in volume 44, one of L. Holzer and the other of A. Scholz, both being renowned number
theorists. It turned out that both solutions were essentially the same as Grns original
proof in the letter to Hasse, and were independent of class field theory.
The proof is short and straightforward: One has to use the fact that the automorphism group of a cyclic group is abelian and, hence, the commutator group G 0 of the
Galois group G of Kjk acts trivially on the ideal class group of K. Consequently,
0
if c is any divisor class of K then the norm NKjK 0 .c/ equals c KWK and therefore,
0
0
since NKjK 0 .c/ is a divisor class of K 0 , we have that c KWK h is the principal class.
Hence the exponent of the class group of K divides K W K 0 h0 . Since the class group
of K is assumed to be cyclic the contention follows. 22
considered only the case k D Q.
= Deutsche Mathematiker Vereinigung = German Mathematical Society.
22 I am indebted to Franz Lemmermeyer for the information that Yamamura had rediscovered and used this
theorem of Grn. See [Yam97], p. 421. Lemmermeyer himself has used (and proved) this theorem in [Lem97],
Proposition 6, citing Grn.
20 Grn
21 DMV
89
But the result seems to be quite special because of the assumption that the class
group of K is cyclic. Therefore Hasse proposed to Grn to investigate the general
case, with class group of arbitrary structure. Clearly, whenever a subgroup G1 of G
can be found which acts trivially on the class group of K then a similar argument can
be applied to obtain an upper bound for the exponent of the class group of K (not
necessarily for the class number h itself), with K 0 being replaced by the fixed field
K1 of G1 .
2.3.3 Representations over finite fields
On April 19, 1933 Grn answered that his attempts to deal with non-cyclic class
groups had not been successful. However after some time, on December 5, 1933 he
wrote:
Nach langer Zeit kann ich Ihnen heute wieder etwas berichten. Ich
habe mich mit gruppentheoretischen Untersuchungen beschftigt Ich
knpfe an an meine Aufgabe No. 153 im Jahresbericht. Als ich Ihnen
damals das Resultat mitteilte, stellten Sie die Frage: Wie lautet das
genaue Analogon fr allgemeine Abelsche Klassengruppen? Um dieses
Problem handelt es sich hauptschlich.
After a long time I am able again to report something to you. I have been
busy with group theoretical questions I refer to my problem no. 153
in the Jahresbericht. When I had reported to you on that result, you
asked: What is the exact analogue for arbitrary abelian class groups?
The following is mainly concerned with this question.
And Grn continues with a description of his results. Let G be a finite group
which acts on an abelian group A of exponent p, a prime number. (We observe that
Grn discussed, as a first step, not arbitrary abelian class groups but only p-groups
of exponent p, i.e., vector spaces over Fp .) Let m be the rank of A. For any prime
` p let m` denote the order of p mod `. Grn wrote that indeed he has found
general statements about subgroups of G which act trivially on A. He proved:
If ` > mm` then the commutator group of an Sylow `-group of G acts
trivially on A. In other words: If ` > mm` then the Sylow `-groups of the
automorphism group of A are abelian.
Hasse had Grns manuscript refereed by Magnus who at that time was already
considered to belong to the leading German mathematicians in the field of group
theory. 23 Hasse asked him whether Grns result has appeared already in the literature. Magnus replied that he knew only one source, an American paper by Brahana
23 Wilhelm Magnus in Frankfurt had received his doctorate 1931 with Max Dehn as his supervisor. The
correspondence between Hasse and Magnus is preserved; it had started in 1930 when Magnus submitted his
dissertation [Mag30] for publication to Hasse as an editor of Crelles Journal.
90
[Bra34], which dealt with similar problems. Brahanas result appears as a special
case of Grns. And he added (letter of December 16, 1933):
Ich nde die Sache wirklich sehr hbsch, auch der von ihm angegebene Beweis des schon von Brahana gefundenen Spezialfalls scheint mir
etwas durchsichtiger zu sein als bei B., und wenn sich die Ergebnisse
auf Klassenkrperprobleme anwenden lassen, wre das ja besonders
erfreulich.
I regard the matter as quite nice. Also, Grns proof in the special
case which had already been found by Brahana seems to me to be
somewhat more transparent than B.s proof. And if Grns results can
be used in class field theory then this would be particularly nice.
Obviously Hasse had written to him that he expects Grns results to be applicable in
class field theory. In fact, as we have seen, Grns group theoretical problem arose
from a question about class numbers.
Grn in his letter also mentions that in addition to the above result, he has determined the structure of all Sylow groups, not only those for large `, of the automorphism group of an abelian p-group A of exponent p. Moreover, all those results
are valid for the automorphism group of any vector space of finite dimension over an
arbitrary finite field of characteristic p.
In fact, this is the content of the paper which Hasse finally accepted for Crelles
Journal, already in the same year [Gr34a].
We see that Grns main interest had by now shifted to group theory in consequence of Hasses question. The application to class field theory, he writes, will be
given later. But he never did so. It seems that from now on group theory absorbed
all his interest.
2.4 The two classic theorems of Grn

More than one year later, on March 30, 1935, Grn submitted to Crelles Journal
another manuscript on group theory. This has turned out to become a classic and
made his name widely known [Gr35].
There are two main parts of the paper. In the first part he gives a direct generalization of what we have discussed in the foregoing section (and what had already
appeared in Crelles Journal). Namely, he dropped the condition on the G-module A:
Let G be a finite group which acts on an abelian p-group A of arbitrary structure,
not necessarily of exponent p. Let m denote the rank of A. Let ` be a prime p.
Then:
91
If k is the smallest exponent with `k > mm` then the k-th commutator
group of an Sylow `-group of G acts trivially on A.
The paper contains an even further generalization, namely for an arbitrary p-group
A, not necessarily abelian, on which G acts. Then one has to consider the ascending
central series of A, and in the condition for `k the number m has to be defined as the
maximal rank of the factor groups of that series.
2.4.1 The second theorem of Grn
But the main results of this paper are to be found in the second part where we find
the two famous Theorems of Grn.
Given a finite group G and a prime number p, the problem is to describe the
structure of its maximal abelian p-factor group G=G .p/ . Here, G .p/ denotes the
p-commutator group of G. This description turns out to be particularly simple for
groups which have a property which is called p-normal. 24 This property is defined
as follows: the center C of a Sylow p-group P of G coincides with the center of any
other Sylow p-group in which C is contained. Grn proves:
If G is p-normal then the maximal abelian p-factor group G=G .p/ is
isomorphic to the maximal abelian p-factor group NC =NC.p/ , where NC
denotes the normalizer of the center C of a Sylow p-group P of G.
The idea behind this is that a Sylow p-group P , its center C and the normalizer NC
are in general much smaller and easier to handle than the whole group G. Hence this
theorem yields a criterion for a group G to have a non-trivial p-factor group, namely:
this is the case if and only if NC has this property. Note that NC contains C as an
abelian normal subgroup, thus we have the situation which Grn considered in the
first section of this paper, and that result is applicable to NC acting on C .
If in particular the Sylow p-group P of G is abelian and is contained in the center
of its normalizer then G is p-normal and it follows the isomorphism G=G .p/ P
which is a classical theorem of Burnside, and was well known also to Grn. In this
sense Grns theorem can be regarded as a generalization of Burnsides theorem
and a far reaching generalization at that.
The above theorem is usually called the second theorem of Grn although in
Grns paper it is proved first, whereas the first theorem of Grn is what Grn proves
afterwards. The switch in the enumeration is probably due to Zassenhaus25 who in his
group theory text book [Zas37] included the two theorems of Grn and introduced
the enumeration used today. This makes sense since Grns second theorem (in
Zassenhaus enumeration) can be regarded as a corollary of his first theorem.
24 This
terminology had been proposed by Hasse (letter to Grn of May 28, 1935).
Zassenhaus got his doctorate 1934 under the supervision of Artin. From 1934 to 1936 he worked at
the University of Rostock, and there he wrote his famous text book on group theory.
25 Hans
92
2.4.2 The first theorem of Grn

The first theorem of Grn gives a description of G=G .p/ for an arbitrary finite group
G, not necessarily p-normal. This is somewhat more complicated than in the case of
a p-normal group. Namely:
For an arbitrary finite group G, its maximal abelian p-factor group
G=G .p/ is isomorphic to the following abelian factor group of its Sylow
p-group P :
G=G .p/ P =P ? ;
where the normal subgroup P ? P can be described as
Y
P ? D .P \ NP0 /
.P \ 1 P 0 /:
2G
Note that here appears the normalizer NP of the whole Sylow p-group P in G (not
only NC ). As usual, P 0 denotes the commutator group of P , and similarly NP0 is the
commutator group of NP .
Admittedly, this result looks somewhat complicated because of the definition of
P ? . Nonetheless it has turned out to be quite important in group theory, in as much
as it shows that the maximal abelian p-factor group of any group G can be found as
an explicitly given factor group of the (usually much smaller) Sylow p-group P . Its
kernel P ? depends very much on how the Sylow p-group P is embedded into the
group G.
We have already said that Zassenhaus, who at that time was writing a textbook
on group theory, immediately recognized the importance of Grns theorems and
decided to include them into his book [Zas37]. 26
While reading this paper of Grn one can observe that its style is quite different
from that of his other papers. The paper is well written, careful in the use of notations,
and it contains several diagrams which nowadays are known as Hasse diagrams.
The explanation of this is that the manuscript, in the form as published, had been
entirely written by Hasse himself.
2.4.3 Hasse and the transfer
We have already said that Hasse, being an editor of Crelles Journal, used to check
every manuscript carefully before sending it to print. So he did also with Grns
manuscripts, and in particular with the manuscript under discussion. After all, Grn
as an amateur had no experience with writing a paper. The letters of the Hasse
Grn correspondence show that Hasse worked quite hard to put this paper into shape.
26 Zassenhaus, in his paper [Zas35b] on finite near-fields, had already discussed certain results centered around
the classical Burnside theorem as mentioned above. This may explain Zassenhaus great interest in results of the
kind of Grns theorems.
93
After an extended exchange of letters there were so many corrections, additions and
deletions that the original manuscript was hardly readable any more. Finally Hasse,
seemingly somewhat exasperated, proposed that he himself will now compose a new
manuscript. To which Grn replied (letter of May 18, 1935):
Ihre Mitteilung, da Sie ein neues Ms. herstellen wollen, hat mich zwar
einerseits hoch erfreut, aber darf ich denn das annehmen ? Ich wei
wirklich nicht, ob ich das zugeben darf. Wir mten natrlich auch Ihre
ttige Mitarbeit ausdrcklich vermerken. In jedem Falle : Wenn Sie
von Ihrer eigenen Zeit etwas opfern wollen, mssen Sie die betr. Sache
schon fr sehr wichtig halten. Das ist das beste Lob, das ich mir denken
kann.
On the one hand, I am very glad about your proposition that you will
compose a new manuscript but could I accept this ? Really, I do not
know whether I am allowed to give my consent. Of course, we would
have to state explicitly your extensive cooperation. In any case : If
you will spend your own time on this then you must consider it very
important. This is the best praise from you which I can imagine.
Hasse replied on May 21, 1935:
Ich halte es wirklich fr das Beste, wenn ich hier ein neues Manuskript
herstelle. Die Arbeit daran wrde mir Freude machen und Sie brauchen
sich darber keine Gedanken zu machen. Dies in der Arbeit selbst zu
erwhnen, wrde mir nicht zusagen. Sie mgen das so auffassen, dass
es zu meinen Aufgaben als Herausgeber gehrt, wenn man diese im
weiteren Sinne auslegt.
Indeed, I believe it is the best solution if I will write a new manuscript
here. It will be a pleasure to me and you do not have to worry about
it. But I would not like that this be mentioned in the paper. You might
regard it as belonging to my tasks as an editor, if one interprets them in
a wider sense.
Certainly, Hasse regarded Grns results as important and this was one of his
motivations to help Grn to put it into a form which would be appreciated by the
mathematical public. But another reason which required a complete rewriting of
the manuscript, was Hasses proposition that the transfer map (Verlagerung) should
be used as an adequate tool which provides the isomorphisms of Grns theorems.
For, in his original version Grn had not used the transfer and not obtained those
isomorphisms, but he was content with saying that if one of the two factor groups
(which we now know to be isomorphic) is non-trivial then the other is non-trivial too.
The transfer VG!H from G to a subgroup H is a homomorphism from G into the
factor commutator group H=H 0 . It can be defined as the determinant of the canonical
94
monomial representation of G modulo H with coefficients in H=H 0 . It had first been

constructed and used by Schur [Sch02]. Later in 1927 the transfer was re-discovered
by Artin and Schreier during their attempts to prove the conjectured principal ideal
theorem of class field theory. We know this from Artins letter to Hasse of August 2,
1927. See also Artins paper [Art29]. It seems that neither Artin nor Hasse were aware
of the old paper by Schur because they never mentioned it in their letters nor in their
publications. Artin was able, by means of his general reciprocity law, to reformulate
the principal ideal theorem as a purely group theoretical statement concerning the
transfer. 27 Hasse in his class field report II [Has30a], p. 170 introduced the name
Verlagerung for this group theoretical map, which then was translated into English
as transfer.
By 1935 the transfer map was a well established tool but apparently it was used
mainly in number theory in connection with the principal ideal theorem and related
questions. It seems that in abstract group theory it had not yet found many applications
(except in Schurs paper mentioned above). But this changed after Grns paper.
In Grns letter of May 18, 1935 we read:
Haben Sie vielen Dank fr Ihre Briefe und die darin enthaltenen wertvollen Anregungen. Der Gedanke, die Theorie der Verlagerung heranzuziehen, ist auerordentlich glcklich. Ich hatte ja auch bei meinem
Beweis von Satz 5 hnliche Wege eingeschlagen, ohne aber diese Theorie wirklich zu benutzen. Die Verlagerungstheorie gestattet, in einfacher
Weise die Stze 4 und 5 voll zu beweisen. Fr Satz 4 haben Sie dies ja
schon liebenswrdiger Weise so weit durchgefhrt,
Many thanks for your letters and the valuable suggestions therein. The
idea to use the transfer theory is extraordinarily fortunate. In my proof
of theorem 5 I had used similar methods but without really using that
theory. Transfer theory leads to simple complete proofs of theorems 4
and 5. In case of theorem 4 you have already kindly done it so far,
Grn proceeds to expound in detail the proofs which Hasse had indicated using
transfer theory. And later in this letter he writes:
Natrlich mu aber [in der Arbeit ] in jedem Falle darauf hingewiesen werden, da die Anwendung der Theorie der Verlagerung auf Ihre
Anregung hin erfolgt ist und ich somit diese eleganten Beweise Ihnen
verdanke.
Of course, it should be mentioned [in the paper] that the application of
transfer theory is due to your suggestion and that, hence, I owe these
elegant proofs to you.
27 One year later, in 1928, Furtwngler [Fur29] succeeded to prove this group theoretical statement. Later
there were simplifications of Furtwnglers proof, one also by Magnus [Mag34], but the most significant one by
Iyanaga [Iya34]. (By the way, Iyanaga says in the introduction that the greater part of his paper is due to Artin.)
95
But Hasse replied:

scheint es mir aus sachlichen Grnden notwendig, in einer Fussnote
zu erwhnen, dass der Gedanke, die Verlagerung bei den Beweisen von
Stzen Burnsidescher Art zu benutzen, von Herrn Ernst Witt, Gttingen,
stammt.
I find it necessary to mention in a footnote that the idea to use the
transfer in the proofs of theorems like Burnsides is due to Mr. Ernst
Witt, Gttingen.
Whereupon Grn, in a footnote to his paper [Gr35], inserted the following text:
Den Gedanken, bei diesem Beweis die ursprnglich von mir verwendeten monomialen Darstellungen durch die Verlagerung zu ersetzen, verdanke ich einer Mitteilung von H. Hasse. Dieser wurde seinerseits geleitet durch eine mndliche Mitteilung von E. Witt, wonach sich der klassische Beweis des Burnsideschen Satzes in ganz entsprechender Weise
einfacher und durchsichtiger gestalten lt.
The idea to replace the monomial representations (which I originally
used) by the transfer map, arose from a suggestion of H. Hasse. He had
been led by an oral communication of E. Witt who pointed out that the
classical proof of Burnsides theorem can similarly be simplified. 28
There is another footnote, after the statement of the first theorem of Grn, reading
as follows:
Diesem Satz und seinem Beweis hat Herr Hasse die vorliegende Form
gegeben. Ich habe mich ursprnglich darauf beschrnkt, bei den gemachten Voraussetzungen eine zyklische p-Faktorgruppe nachzuweisen.
This theorem and its proof has been put into the present form by
Mr. Hasse. Originally I had been content with showing, under the assumptions as stated, the existence of a cyclic p-factor group.
By this Grn means a non-trivial cyclic factor group of P =P ? as a necessary and
sufficient condition that G=G .p/ is non-trivial. Certainly, the idea to establish group
isomorphisms (when possible) instead of only considering the group orders, is part of
the Modern Algebra which had been propagated by Emmy Noether and had found
28 Burnsides theorem (as explained in Section 2.4.1) can be found in his book [Bur11], 243. The computations
performed there are indeed the same as computing the kernel and the image of the transfer map in the special
situation at hand. However, Burnside does not mention (nor does he care) that this is a general procedure,
referring to a generally defined map. Therefore, if it is said that the definition of the transfer map goes back
to Burnside, such statement has to be interpreted with appropriate caution. It takes some insight to realize that
Burnsides arguments indeed can be looked at as evaluating a homomorphic map. We do not know whether Witt
had known Schurs paper [Sch02] or whether he had observed this himself.
96
its expression in van der Waerdens text book [vdW31]. Hasse explained this to Grn
in his letter May 28, 1935 as follows:
Entgegen Ihren brieichen Andeutungen sehe ich allerdings doch das
Hauptgewicht Ihrer Stze in der Herleitung von notwendigen und hinreichenden Bedingungen fr die isomorphe bertragung von Untergruppen oder Faktorgruppen innerhalb P auf Faktorgruppen von G, und
nicht so sehr in der blossen Folgerung auf die Ordnungen dabei. Daher
habe ich in den Formulierungen immer nur die Isomorphiebehauptungen angefhrt und meine, man kann es ruhig dem Leser berlassen,
die daraus ohne weiteres ablesbaren Folgerungen fr die Ordnungen zu
ziehen.
Contrary to your hints in your letters I regard the main point of your
theorems to be the isomorphic transport of subgroups and factor groups
within P to factor groups of G, and not so much in the mere consequence
for the group orders. Therefore, I have formulated all the theorems as
referring to isomorphisms. In my opinion it can be left to the reader to
draw from this the consequences concerning the group orders
Finally on June 7, 1935, when the manuscript seemed to have acquired a form
satisfactory to both, Grn wrote:
Lieber Herr Professor Hasse ! Vielen Dank fr die bersendung des
Manuskriptes und Durchschlages. Jetzt ist doch wirklich etwas aus meiner ursprnglichen Arbeit geworden. Ich gestehe Ihnen, da ich erst nun
wirkliche Freude an meinem Manuskript habe.
Dear Professor Hasse ! Many thanks for the manuscript and carbon copy.
Really, now there has developed something out of my original paper. I
have to admit that only now I have real pleasure with my manuscript
But the correspondence about this continued and several points had still to be cleared.
It took until August 13, 1935, after more than eighteen letters 29 had been exchanged
between Hasse and Grn concerning this manuscript, that finally Grn could send
the corrected proof sheets to Hasse. The paper appeared in the same year 1935 in
Crelles Journal [Gr35].
We have reported about this part of the HasseGrn correspondence in a somewhat
greater detail, since it does not seem to be widely known to what extent Hasse had a
share in Grns classic paper. The title of the paper is:
Beitrge zur Gruppentheorie I.
Contributions to group theory I.
29 This
means that eighteen letters have been preserved, six of them by Hasse and twelve by Grn. Those
letters of Hasse which are preserved are written with typewriter, and Hasse had made carbon copies. Probably
another six letters by Hasse were handwritten and, hence, not preserved.
97
In one of his letters Grn had announced that there will be a second and perhaps more
parts of such contributions. But the next he submitted in 1943 only (due to the
problems in war time it appeared in 1945; see [Gr45]). Later in the course of time
Grn produced 10 such contributions, the last appearing 1964 again in Crelles
Journal, when Grn was 74. 30
2.4.4 Grn, Wielandt, Thompson
Let us jump 4 years ahead to the Gttingen group theory conference in 1939. 31 There
on June 27, 1939, Wielandt 32 delivered a talk with the title: Sylow p-groups and pfactor groups. This is precisely the topic of Grns classic paper [Gr35] which we
just have discussed. In fact, Wielandt presented (among other results) a far reaching
generalization of Grns result. The main theorem of Wielandt ist somewhat involved
and we do not reproduce it here. One of its many consequences concerns the case
when a Sylow p-group P of G is p-regular in the sense of Ph. Hall. This means that
x p y p .xy/p mod hx; yi0 p
holds for every x; y 2 P . (In other words: The operation p-th power can be
performed termwise, modulo a product of p-th powers of commutators from the
group generated by x and y.) Under this assumption it follows from Wielandts main
results that the maximal p-factor group of G is isomorphic to the maximal p-factor
group of the normalizer NP . Note that here the p-factor groups in question may be
non-abelian whereas Grns results refer to abelian p-factor groups only. Wielandt
achieves this by manipulating the monomial representation directly in a suitable way,
not only the transfer map which is the determinant of the monomial representation.
Wielandts talk was published 1940 in [Wie40]. It is evident that Wielandts
paper is directly influenced by Grns. B. Huppert has given following comment to
Wielandts paper: 33
Eines der Ziele von Wielandt wird in dieser Arbeit mit keinem Wort
erwhnt, nmlich die Nilpotenz des Frobenius-Kerns einer FrobeniusGruppe. Diese wurde zuerst von J. Thompson bewiesen. Im Sommer
1958 gab es in Tbingen eine lange Unterhaltung zwischen Wielandt
und Thompson. Unmittelbar danach sagte Wielandt zu mir: Das ist ein
sehr scharfsinniger Bursche, von dem kann man etwas lernen. Einige
Monate spter reichte Thompson seine Arbeit bei der Mathematischen
Zeitschrift zur Publikation ein. Demnach gibt es eine ganz deutliche
30 In Grns enumeration there were Contributions no. IIX and XI published, but not no. X. We do not know
his plans for no. X.
31 For more on this conference see Section 2.6.2.
32 Helmut Wielandt had studied in Berlin with I. Schur and was awarded his doctorate in 1935. In 1939, the
year of the Gttingen group conference, he held a position of assistant at the University of Tbingen.
33 Private communication.
98
mathematische Verbindungslinie von Grn ber Wielandt bis zu Thompson.

One of Wielandts motivations is not mentioned at all in this paper
[Wie40], namely to prove the nilpotency of the Frobenius kernel of a
Frobenius group. This was proved later only by J. Thompson [Tho59].
In the summer of 1958 there was a long discussion in Tbingen between
Wielandt and Thompson. Immediately thereafter Wielandt said to me:
This is a very sharp-witted guy, from him one could learn a lot. Several months later Thompson submitted his paper [Tho59] to Wielandt
for publication in the Mathematische Zeitschrift. Thus we can observe
very clearly a line of mathematical influence from Grn over Wielandt
to Thompson.
Remark. R. W. van der Waall has pointed out to me that the line of mathematical
development which started with Grns paper can be traced much further: There are
quite a number of subsequent papers continuing the ideas of Grn and supplementing
his results. Of particular interest is the following result contained in a paper by
T.Yoshida published in the Journal of Algebra 52 (1978), pp. 138. It says that the
transfer isomorphism G=G .p/ NP =NP.p/ holds quite generally, with exceptions
possible only if P admits a factor group isomorphic to the wreath product of the
cyclic group of order p with itself. Indeed this is a very strong generalization of the
first theorem of Grn. The transfer map and its dual have become standard tools in
the theory of finite groups.
2.5 Grn meets Hasse (1935)

2.5.1 Hasses questions
Grn, in his first letter to Hasse, had introduced himself as an amateur mathematician.
But it seems that Hasse, impressed by Grns achievements, had some doubts by now.
Although there had been an exchange of letters since three years, he did not know
anything definite about Grns mathematical background. So Hasse at last asked in
his letter of May 8, 1935:
Sind Sie eigentlich Mathematiker von Hauptberuf, oder treiben Sie
die Mathematik nur nebenbei als Liebhaberei ?
By the way, are you a mathematician by profession, or are you doing
Mathematics as a hobby?
To which Grn replied (letter of May 9):
99
Ich wollte, verehrter Herr Professor, ich wre Mathematiker von Hauptberuf. Leider ist das nicht der Fall, ich mu mich ohne besondere Begeisterung kaufmnnisch bettigen, um zu leben.
I would wish, dear Herr Professor, that I could be a professional mathematician. Unfortunately this is not the case; I have to work for a living
in a commercial job, though without particular enthusiasm.
But Hasse continued to inquire (letter of May 13):
Wo haben Sie sich denn Ihre mathematischen Kenntnisse erworben?
Haben Sie einen bestimmten Mathematiker zum Lehrer gehabt?
But where did you pick up your mathematical knowledge? Have you
had a teacher who was a mathematician?
Grns reply (letter of May 15):
Ob ich einen bestimmten Lehrer gehabt habe ? Ich habe meine Kenntnisse nur aus Bchern geschpft und da sind Sie selbst zu einem groen
Teil mein Lehrer gewesen. Ich bekam zufllig Ihre beiden Berichte in
die Hand und damit begann mein intensives Interesse fr Klassenkrpertheorie. Natrlich war ich mathematisch so weit vorgebildet, da ich
fhig war, die Berichte durchzuarbeiten. Die auerordentliche Klarheit
und Durchsichtigkeit Ihrer Darstellung nimmt ja dem Leser jede Arbeit ab. Bis dahin hatte ich mich eigentlich mehr fr Funktionentheorie interessiert, allerdings hatte ich wenigstens Hilberts Zahlbericht,
Dirichlet, Dedekind und die einzelnen Kummerschen Arbeiten gelesen.
Nun wurden Ihre Berichte fr mich Veranlassung, mich intensiv mit
Gruppentheorie zu befassen.
Whether I have been taught by a particular teacher? I have acquired my
knowledge from books only, and there to a large degree my teacher has
been you. Your two reports 34 came by chance into my hands, and this
started my intensive interest in class field theory. Of course I had already
acquired enough of the mathematical prerequisites which enabled me to
read your reports. After all, the wonderful clarity and transparency of
your presentation spares the reader much of the work. Until then I tended
to have more interest in the theory of complex functions, but I had already
read Hilberts Zahlbericht, Dirichlet, Dedekind and various papers by
Kummer. 35 Now your reports had induced me to look intensively into
group theory.
34 Grn refers to Hasses class field reports, the first on Takagis class field theory [Has26a], and the second on
Artins reciprocity law [Has30a].
35At that time, the Collected Papers of Dirichlet and Dedekind were available, but not yet Kummers. The
latter would be published in 1975 only, edited by Andr Weil.
100
When Grn states that his interest in group theory had been induced by Hasses
class field report, then we see that by now he had well grasped the main trend in
the then modern class field theory, as expressed in the foreword to Part II of that
report:36
Artins Reciprocity Law constitutes an advance of the utmost importance. Its importance lies not so much in the direction which might be
suggested by the name reciprocity law and its classical formulation,
but in the general class field theory. The ultimate aim of it is the coding of all arithmetical properties of a relative abelian number field in its
Galois group, similarly as the aim of Galois theory is the coding of field
theoretic properties in the Galois group.
However, Grns mathematical interests had now shifted from FLT and class field
theory almost entirely to group theory. The application to class field theory does not
appear in his further publications. In group theory Grn had found his main subject
where he would be active in the future. A majority of 21 of his total of 26 papers
from 1934 to 1964 belong to group theory.
2.5.2 Grns visit
In view of this correspondence, Hasse now wished to meet Grn personally, in particular since Grn had announced to have many more results in his files. For, in his
letter of May 9, 1935 Grn had written:
Nach der Verffentlichung meiner beiden Noten ber den Fermat und
Gruppen im Galoisfeld hat mir das Kultusministerium eine gewisse Untersttzung zuteil werden lassen, die mich in Stand setzte, mich einige
Zeit fast ausschlielich mathematischen Untersuchungen zu widmen.
Die Folge ist, da ich geradezu eine Unmenge von Notizen habe, in denen die wesentliche Vorarbeit fr eine Verffentlichung schon geleistet
ist; alle diese Arbeiten sind gruppentheoretischer Natur, natrlich mit
krpertheoretischen Anwendungen.
After publication of my two notes on Fermat and on groups in a Galois
field the ministry of education had granted me a certain stipend which
enabled me to devote almost all my time to mathematical work. As
a consequence I have a huge pile of notes which already contain the
essential ingredients of future publications. All of this work is of group
theoretical nature, of course with applications to field theory. 37
36 The following is a free translation of essential features of Hasses foreword of [Has30a]. The reader may
compare this with Hasses foreword in his book on abelian fields [Has52].
37 To this Grn added: I have to acknowledge with thanks the support which I have found with the minister
of education, for neither was I a member of the party nor have I become such. Of course, the party which
101
So Hasse wrote on May 13, 1935:

Wir haben hier in diesem Semester gerade eine kleine Arbeitsgemeinschaft ber Gruppentheorie, in der wir mit Ihren Untersuchungen sehr
verwandte Dinge betreiben, insbesondere die beiden neuen gruppentheoretischen Arbeiten von Zassenhaus studieren, die im letzten Heft der Abhandlungen des Hamburger Mathematischen Seminars erschienen sind.
In this semester we have here a small workshop on group theory, on
topics which are closely related to your investigations. In particular we
are studying the two new group theoretic papers of Zassenhaus which
have appeared in the last issue of the Hamburger Abhandlungen. 38
And Hasse continued:
Sehr gerne wrde ich Sie auffordern, doch im Monat Juni einmal hierher zu kommen und bei uns in der Arbeitsgemeinschaft ber Ihre gruppentheoretischen Studien vorzutragen, ganz zwanglos, d. h. so dass man
dazwischenfragen darf, wenn man etwas nicht versteht, und das ganze
mehr den Charakter einer gemeinsamen Erarbeitung hat.
I would like very much to invite you to visit us some time in June, and to
inform us about your group theoretic work. This should be completely
informal, so that it will be possible to put questions; the whole thing
should have the character of a common discussion.
On June 13, 1935 Grn arrived in Gttingen 39 ; his talk in the workshop was scheduled for the next day, a Friday. Hasse had offered him to lodge in the Mathematical
Institute where there was a visitors room available, and to stay over the weekend in
order to have opportunity for discussions with the people of, in Hasses words, the
small but lively group of algebraists in Gttingen. We know from other sources the
names of the members of that group, the most outstanding members besides Hasse
being Witt, Teichmller and also H. L. Schmid. 40 The latter was to play, ten years
later, an important role in Grns life.
he alludes to, was the NSDAP, the Nazi party which had come to power in Germany in January 1933. Indeed
it seems remarkable that Grn was supported in his work by the government of that time although he did not
conform to the official party line. Later in 1946 he wrote that he had to suffer severe personal repression because
he repeatedly had been urged to join the party but always refused.
38 These were the papers [Zas35a] and [Zas35b], the first one on the characterization of linear groups as
permutation groups, and the second on finite near-fields.
39 Note that in the summer of 1934 Hasse had left Marburg and accepted a position at the University of
Gttingen. Thus Hasses invitation to Grn was meant for Gttingen, not Marburg. For details of Hasses change
to Gttingen in the midst of the political upheavals of the time, we refer to [Fre85] and [Sch87].
40 This was the same Arbeitsgemeinschaft in which one year later the Witt vectors were discovered, together
with their application to cyclic extensions in characteristic p and class field theory, as well as to the structure
theory of p-adic fields. Those results are all published in one volume of Crelles Journal (vol. 176), together with
the seminal paper of Hasse who used Witt vectors for the explicit p-power reciprocity law of class field theory.
102
In a former letter Grn had asked whether his talk in the workshop could be about
p-groups, and Hasse had replied that the choice was entirely up to the speaker. And,
knowing from his correspondence that Grn may have some problems to explain
mathematical arguments in a correct form, Hasse had added the advice that Grn in
his talk should be very explicit in all details.
Perhaps it is not without interest to cite Hasses words where he tried to inform
Grn about what had been discussed in the workshop so far, i.e., what he could assume
to be known:
ber p-Gruppen haben wir auch schon gesprochen. Wir haben die klassische Theorie (Speiser) durchgenommen, ferner noch einige weitere
Stze ber die Anzahlen der Untergruppen oder Normalteiler gegebener Ordnung in einer p-Gruppe. Weiter die Theorie der Hamiltonschen
Gruppen (alle Untergruppen Normalteiler ) und der p-Gruppen, in denen es nur eine Untergruppe der Ordnung p gibt (nur fr p D 2 gibt
es nicht zyklische solche Gruppen). Ich werde morgen ber Satz 5 und
Satz 9 Ihrer Arbeit vortragen.
We have already discussed p-groups. We worked through the classical
theory (Speiser) 41 and in addition some theorems about the number of
subgroups and normal subgroups of given order in a p-group. Furthermore the theory of Hamiltonian groups (all subgroups are normal), and
the p-groups with only one subgroup of order p (only for p D 2 there
are non-cyclic groups with this property). Tomorrow I shall talk about
theorems 5 and 9 of your paper.
Theorems 5 and 9 were the second and the first theorem of Grn as discussed above.
The above lines show that in the circle around Hasse there was lively interest to
learn more about the newest results of finite groups, in particular p-groups. This
may have its explanation by the fact that during those years the theory of p-groups
had been used heavily in algebraic number theory. We only mention the work of
Arnold Scholz (who has had an extensive exchange of letters with Hasse) and who
just recently had proved the existence of number fields with a given p-group of class
two as Galois group [Sch35b]. (And one year later Scholz would prove the same for
an arbitrary finite p-group [Sch37].) This gives us perhaps another clue why Hasse
was so much interested in Grns results on p-groups.
Unfortunately we have not found any record about what Grn had actually talked
about, nor how his talk was received by his young audience. Did Grn indeed talk
about p-groups and what were his results which he presented? We can imagine that
Grn, not being used to lectures and colloquium talks, had some difficulties to address
such a group of brilliant young mathematicians who were used to high standards not
only with respect to the mathematical topics under discussion but also as to the way
41 Hasse
means Speisers monograph on group theory [Spe27].
103
of presenting new material. Doubtless Grn met high respect among these people, in
view of his outstanding results so far. But did they appreciate his talk? From other
sources (in later years) we infer that Grns talks used to be somewhat clumsy and
difficult to follow.
One week after Grns talk, Zassenhaus visited the workshop in Gttingen, on
June 21, 1935. Hasse had offered Grn to stay longer in order to meet Zassenhaus,
and Grn did so. Note that Grns paper [Gr35] had not yet appeared, and that
Zassenhaus was just working on the text of his group theory book [Zas37]. It seems
probable that Zassenhaus, when he met Grn in Gttingen, learned about Grns
theorems and realized their importance. In the foreword to his book (which appeared
in 1937) Zassenhaus says that he wished to include the new and far-reaching results
in group theory of the last 15 years; certainly Grns theorems were among those and
thus found their way into Zassenhaus book. 42
Two months after Grns visit to Gttingen he wrote to Hasse (letter of August 13,
1935):
Lassen Sie mich Ihnen nochmals danken fr die Gastfreundschaft, die
ich in Gttingen gefunden habe. Es war geradezu eine Wohltat fr mich,
einmal nur mit wissenschaftlichen Problemen beschftigt zu sein. Wenn
nicht meine wirtschaftliche Lage etwas anderes forderte, wrde ich mich
in Gttingen niederlassen und mich vllig meinen mathematischen Untersuchungen widmen.
Thank you again for the hospitality which I have found in Gttingen. It
was really a great pleasure to me to be occupied exclusively by scientific
problems. If my economic situation would have been different then I
would settle in Gttingen and would occupy myself completely with
mathematical research.
This sounds as if Grn had hoped to be offered a position at the University of Gttingen
which would enable him to exclusively follow his research work. But this was not
the case.
With the same letter Grn returned the proof sheets of his paper [Gr35]. Recall
that the title of that paper carried the label Part I which implied that there would
be more parts, at least a second part. Accordingly, Grn mentioned in his letter his
plans for Part II, and that this would include investigations on p-groups. From this
we may perhaps conclude that indeed, his talk in Gttingen was about p-groups, and
that he had been asked to send a manuscript about his talk to Crelles Journal, to be
published as Part II of his investigations.
42 Zassenhaus
book on group theory has been said to have been for decades the bible of the group theorists
(Reinhold Baer). Nowadays both Grns theorems do appear in many textbooks on group theory, for instance
in Huppert [Hup67]. Perhaps it is not without interest to note that Grns theorems have been included and
generalized in the setting of homological algebra. See, e.g., the book of CartanEilenberg [CE56] chap. XII
theorem 10.1.
104
But this Part II did not materialize in the form as planned. Several months later,
in a letter of February 7, 1936, Grn apologized to Hasse that the envisaged paper on
p-groups is not yet finished. He announced the manuscript to be finished in about
two weeks, but finally it took several years for this. And the real Part II, which we
have said appeared in 1945 only, did not deal particularly with p-groups [Gr45].
2.6 The Burnside problem (1939)

2.6.1 Dimension groups
After the appearance of Grns paper [Gr35], his exchange of letters with Hasse
slowed down in frequency and intensity. Grn had found his main interest to be
group theory. He knew that Hasses main interest was number theory, and so he may
have felt that now he could pursue his work without having to rely every time on
Hasses advice. 43
In the year 1936 there appeared the paper [Gr36] on the descending central series
of free groups. This paper is never mentioned in the HasseGrn correspondence.
Grn proves, with an unusual and somewhat peculiar argument using group representations, that the dimension groups as defined by Magnus [Mag35] do coincide
with the members of the descending central series of the given free group. This was
considered an important result.
Since Grns paper directly refers to a paper by Magnus it is not unreasonable to
assume that Grn had discussed it with Magnus before publication. Maybe it was
Magnus himself who had posed the problem to Grn. We know from several sources
that there was mathematical contact between Grn and Magnus in those years since
1935. But the correspondence GrnMagnus seems to be lost and so we do not know
the details of how strong Magnus influence had been for this paper.
In any case, one year later Magnus himself provided a simplified proof, published
in Crelles Journal [Mag37]. But Grns proof was duly registered as the first, and
was appreciated by the specialists.
2.6.2 The group theory conference in Gttingen
In June 1939 Hasse had organized a 5-day group theory conference in Gttingen.
About the preparations for this conference we read in a letter which Hasse had
sent jointly to Magnus and Zassenhaus, dated February 18, 1939:
Die Gttinger Mathematische Gesellschaft plant in der letzten Woche
des Sommersemesters 1939 eine grssere Vortragsveranstaltung ber
43 In [JL98] it is said that, according to Grn himself, it was Hasse who had advised him to switch from number
theory to group theory.
105
das Thema Gruppentheorie. Wir haben dazu Herrn P. Hall von Kings
College, Cambridge eingeladen, uns drei grssere Vortrge aus seinem
Arbeitsgebiet zu halten. Zu meiner grossen Freude hat Herr Hall sich
dazu bereit erklrt
The Mathematical Society of Gttingen is planning a conference on
Group Theory. We have invited Mr. Ph. Hall from Kings college,
Cambridge, for three lectures from his field of research. I am very glad
that he has consented 44
Hasse then explained that the lectures of Philip Hall should form the core of the
conference, but in addition he wished that a number of German mathematicians who
were working in group theory, should be given the opportunity to participate as invited
speakers. And Hasse asked Magnus and Zassenhaus to help him with their expertise
and advice to prepare this conference.
In the ensuing correspondence between Hasse, Magnus and Zassenhaus it was
decided that not too many talks should be scheduled, which meant that only those
German mathematicians should be invited as speakers whose field of research had
some connection to Halls, which is to say mainly p-groups and solvable groups and
related topics. This then would include Grn, as Hasse observed:
Wenn Grn gewonnen werden knnte, so wre das natrlich sehr schn.
Er hat doch bei allem Ungeschick seiner Darstellung die Gruppentheorie
um einige wichtige Erkenntnisse bereichert, die in engstem Zusammenhang mit den Hallschen Arbeiten stehen. Ich bitte Herrn Magnus, sich
mit ihm in Verbindung zu setzen.
If Grn could be won over then this would be very nice indeed. Notwithstanding his awkwardness in the presentation of material, he has enriched
group theory with some important discoveries which are very closely
connected with Halls papers. I am asking Mr. Magnus to get in touch
with him.
When Hasse mentioned the awkwardness in the presentation then he may have
recalled his experiences four years ago with Grns paper which he (Hasse) had to
rewrite completely. Maybe Grns talk in the Gttingen Arbeitsgemeinschaft had
also added to this impression. Nevertheless, in view of Grns achievements Hasse
did not hesitate to name him as invited speaker of the conference.
And when Hasse asked Magnus to get in touch with Grn, then this reflects the
fact that, as said above, by now the mathematical contact of Grn with Magnus had
become closer than his contact with Hasse.
44 In the end, Hall delivered four lectures. Hall was criticised for going to Germany at this difficult time but
he defended his actions saying: the German mathematicians [are] as little responsible for the present
situation (and probably enjoy it as little) as you or I do. (Cited from The MacTutor History of Mathematics
archive.)
106
The Gttingen group theory conference took place from June 26 to June 30, 1939.
The program is published in 1940 in volume 182 of Crelles Journal, together with
the papers presented at the conference. 45 Hence it will not be necessary to go into
all details here. The paper of Grn [Gr40] has the title:
Zusammenhang zwischen Potenzbildung und Kommutatorbildung.
The connection between forming powers and commutators.
The paper is motivated by and closely connected to the old
Burnside problem. Is every finitely generated group of finite exponent
necessarily finite ?
See [Bur02]. For m D 2 the problem has a positive answer, already given by Burnside.
This is so because every group of exponent 2 is commutative, as a consequence of
the formula
t 1 s 1 t s D t 2 .ts 1 t 1 /2 .ts/2
which expresses commutators as products of squares. This led Grn in his paper to
study similar formulas connecting commutators and powers.
The Burnside problem has also a
Restricted version. Are there only finitely many finite groups with a
given number r of generators and a given exponent m ?
Grns paper [Gr40] was the first in which this restricted Burnside problem was
specifically addressed, but not under that name. The term restricted Burnside problem was coined later by Magnus [Mag50].
Let Fr denote the free group with r generators, and Frm the subgroup generated
by the m-th powers. The Burnside problem asks whether the factor group Fr =Frm is
finite. In his paper Grn considers the case when m is a prime power p k ; this implies
that the group of Fr =Frm and its factor groups are p-groups.
Grn observes that the restricted Burnside problem has an affirmative answer
for the pair r, m if and only if the descending central series of Fr =Frm terminates
after finitely many steps. Note that the descending central series is defined by commutators, and so the above condition requires certain relations between powers and
commutators. In his proof Grn used his results of his former paper [Gr36], as well
as results of Magnus [Mag35], of Witt [Wit37b] and Zassenhaus [Zas39].
Grns paper was refereed in Zentralblatt by Zassenhaus, in Fortschritte der Mathematik by Speiser, and in the newly founded Mathematical Reviews by Baer. In the
review by Zassenhaus we find the statement that Grn solved the restricted Burnside
problem in the positive sense for r D 2; m D 5. Baer in his review says the author
may prove that without saying that he really had proved it. The computations
45Among
them the paper by Wielandt which we have mentioned in Section 2.4.4.
107
are quite involved and it seems that nobody had checked it. Later Kostrikin [Kos55]
claimed that he had proved the restricted Burnside problem for r D 2, m D 5 but
again, this seemed to be doubtful until Higman [Hig56] independently had settled the
question positively, for aritrary r and m D 5. 46
Although Grns paper carries the date of receipt as of August 21, 1939, Hasse
would accept it only after it had been checked carefully by Magnus. With Magnus
help the paper underwent a thorough clean up. On January 21, 1940 Magnus wrote
from Berlin 47 that he had worked the last two weekends with Grn, and that the latter
had promised to complete his manuscript until the next weekend. The final version
ready for printing arrived at Hasses office on January 31, 1940.
The attentive reader will have observed that between the dates involved, June 1939
(date of the Gttingen conference) and January 1940 (receipt of Grns paper in final
version) there was September 1, 1939, the outbreak of World War II. The publication
of the conference papers in Crelles Journal was somewhat delayed because one of
the authors, Philip Hall, was a citizen of a country which now was in state of war with
Germany. Hence it was necessary for Hasse to obtain the permission of the proper
German governmental offices to publish Halls papers in Crelles Journal. When that
permission was finally granted it turned out that only two of the four anticipated
papers by Philip Hall had arrived. Since postal service between Germany and Great
Britain had ceased there was no hope that the two missing articles would arrive by
ordinary mail, and Hasse had to find other ways to obtain those articles. This was
finally possible with the good services of Carleman at Djursholm who resided in
Sweden, a neutral country.
2.6.3 A letter of 1952
Although Grns power-commutator formulae in [Gr40] turned out to be useful in
several respects, they did not lead Grn to the general solution of Burnsides problem,
restricted or not, as he had hoped.
But Grn did not give up. Twelve years later, on June 30, 1952, after Hasse had
sent him gratulations for his 64th birthday he thanked Hasse for it and then wrote:
ich habe ein Ergebnis erhalten, das ich sehr hoch einschtze: Die absteigende Zentralreihe hat gesiegt! Die Vermutung von Burnside Setzt
man in einer aus endlich vielen Elementen erzeugten freien Gruppe F
46 For arbitrary parameters r, m the restricted Burnside problem has been finally solved in the positive sense
by E. Zelmanov who had been awarded the Fields Medal in 1998.
47 Magnus was in Berlin at that time. From the correspondence Hasse-Magnus we know that one year earlier,
at the annual DMV-meeting in Baden-Baden, he had approached Hasse and asked whether Hasse could help
him to find a new job since his position of Privatdozent at the University of Frankfurt had become unsustainable
for political reasons. Hasse was able, with the help of Wilhelm Sss who had acquired some influence in the
ministry of education, to find for Magnus a position at the University of Knigsberg. Magnus went there for the
summer semester 1939 but then accepted a job in industry with the electronic company Telefunken in Berlin.
108
alle m-ten Potenzen gleich 1, m eine beliebige natrliche Zahl, so entsteht eine endliche Gruppe ist irrig. Es gilt im Gegenteil: Fr =Frm ,
kann nur dann endlich sein, wenn entweder Fr zyklisch .r D 1/ oder
m D 2i 3k ist. In allen anderen Fllen ist Fr =Frm gewi unendlich.
I have obtained a result which I estimate quite highly: The descending
central series has won! The conjecture of Burnside, If in a finitely
generated free group F all m-th powers are put to 1 then there appears a
finite group, is not true. On the contrary: Fr =Frm can be finite only
if either Fr is cyclic (r D 1) or m D 2i 3k . In all other cases Fr =Frm is
infinite.
Hasse replied on July 15, 1952:
Was Ihr neues Resultat betrifft, so ist das ja in der Tat ganz aufregend.
Herr Witt, dem ich sofort davon Mitteilung machte, meinte, Sie htten
wohl das Resultat nicht ganz przis mitgeteilt, denn bei zwei Erzeugenden sei doch im Falle m D 5 bekannt, dass die Gruppe endlich sei.
Concerning your new result, this is indeed very exciting. I have immediately informed Mr. Witt 48 , and he thinks that you had not stated the
result in sufficiently precise form, for with two generators and m D 5 it
is known that the group is finite. 49
And Hasse asked Grn to send him the precise formulation of the result.
We do not know Grns proof but since he did not reply to Hasse and did not
publish this result there was probably an error in it. Maybe Grn had shown his proof
to Magnus who pointed out the error. Note that Magnus had published two years
earlier another paper connected with Burnsides problem [Mag50], hence he was still
interested and informed about the problem.
At the DMV-meeting 1953 in Mainz, Grn had announced a talk mentioning the
Burnside problem and the BakerHausdorff formula in the title. In the same year
Grn published a paper [Gr53] on p-groups in the Osaka Mathematical Journal in
which some connections to the Burnside problem were given. The paper was rated as
an interesting paper by Suzuki in his Zentralblatt review. But apparently nothing
decisive concerning the Burnside problem came out of these activities.
So this is another case where Grn had attempted to solve a famous great problem
but failed in the end, although he was able to contribute interesting methods, formulas
and lemmas.
48 In
1952 Hasse and Witt were colleagues at the university of Hamburg.

As far as I know the Burnside problem in the unrestricted sense
is still open in the case r D 2 and m D 5. Did Witt have a proof which he never published? Or did Witt refer to
the restricted Burnside problem? But the text of Grns letter indicates that he is concerned with the unrestricted
problem.
49 I am somewhat puzzled by Witts statement.
109
2.7 Later years (after 1945)

Grn had expressed in one of his first letters to Hasse that he did not particularly
like his commercial job, and that he wished to be free to do mathematical research
exclusively. In the year 1938 he finally had the opportunity to leave his unbeloved
commercial job (whatever it was). As he reports in his vita 50 :
Auf Bemhungen einussreicher Mathematiker wurde ich 1938 Chefmathematiker am Geophysikalischen Institut in Potsdam.
Due to the help of influential mathematicians I was appointed chief
mathematician at the Geophysics Institute in Potsdam. 51
But we are somewhat doubtful whether this new job did leave him much more time
for group theory research as did his former job. (Although, as we have seen in
Section 2.5.2, he could participate in the Gttingen group theory conference in 1939.)
In any case, during the war years until 1945, Grn was drafted to work as an expert at
the Navy Headquarters in Berlin 52 ; from this work there resulted a paper on theoretical
physics (which was published later in 1948 [Gr48b]). Again it does not seem likely
that in this period Grn had much time to spare for group theory.
After the war Grn found himself in the devastated city of Berlin without a job,
hence free to tend exclusively to his mathematical research, but also without any
income. In this situation he was picked up by Hermann Ludwig Schmid.
2.7.1 H. L. Schmid and Grn
The mathematical scene in Berlin of the immediate post-war years has been vividly
pictured by Jehne and Lamprecht [JL98]. 53 H. L. Schmid was the main figure who
took the necessary initiative and started to rebuild Mathematics at Berlin University
and at the Berlin Academy from level zero. He was successful to attract mathematicians of high standing to Berlin, like Hasse and Erhard Schmidt (and others). He built
and managed the new editorial office of the Zentralblatt der Mathematik in Berlin.
Against many obstacles he founded a new mathematical journal, the Mathematische
Nachrichten, and served as its managing editor. Using his diplomatic skills he succeeded to create a quiet atmosphere where mathematical life could prosper, protected
50 We
are referring to the same vita from which we have cited in Section 2.2.1.
do not know the identity of the influential mathematicians mentioned by Grn. It seems unlikely that
it was Hasse; the topic of Grns job in Potsdam was never mentioned in their correspondence. But see the
appendix!
52 Sachverstndiger beim Oberkommando der Marine, according to his own words in his vita. We do not
know whether it was the same military department where Hasse and a group of other mathematicians (including
Magnus) were working during the war years.
53 Klaus Krickeberg has pointed out to me that the article [JL98] describes only part of the mathematical
scene in Berlin of those years. Another part was dominated by Erhard Schmidt in the direction of analysis.
51 I
110
from an evironment full of all kinds of basic day-to-day problems. He led mathematics in Berlin to a first revival. 54 For a time it looked like Berlin could become a
leading center in Germany for Mathematical Sciences.
H. L. Schmid took Grn under his wing and was able to get him some financial
support, first in the University of Berlin 55 and since 1947 in the newly founded
Mathematics Research Institute of the Berlin Academy of Science. 56
H. L. Schmid had been assistant to Hasse in 1935, and he had met Grn when
the latter visited Gttingen (see Section 2.5.2). Since 1940 H. L. Schmid worked
in Berlin as an assistant to Geppert in the editorial office of the refereeing journals
Zentralblatt fr Mathematik and Fortschritte der Mathematik. At the same time
he was Privatdozent at Berlin University. From then on H. L. Schmid lived in the
same city as Grn and it is possible that they had met there occasionally. In any
case, H. L. Schmid knew about the mathematical background and the achievements
of Grn, and he knew what Grn needed: namely a quiet place to pursue his research
on group theory. This was what he could offer now, with remarkable consequences
for Grns output of mathematical papers in the years to follow (see Section 1.5.4).
Grns salary at the Berlin Academy was not high, in fact it was quite small and
just enough to live on. But since Grn was single, this was acceptable to him. 57
In October 1946 Grn had received an offer for a teaching position from the
University of Greifswald, as he narrates in his vita written August 2, 1955. However,
they required there that he publicly committed himself to a political party in the Sowjet
occupation zone, and this he refused. Grn was a non-conformist: in the 1930s he
had refused to join the Nazi party, and now he did the same thing with the communist
dominated parties. 58
Perhaps we are not wrong to assume that there was another reason for Grn,
conscious or unconscious, to reject this offer to Greifswald. For, he did not like
to teach. In fact, by all indications we know he was not a good lecturer. And
so he preferred to live on the small but sufficient income he got from the Berlin
Academy, free to pursue his studies on group theory without worrying about teaching
and administrative or political problems.
54 Cited
from [JL98].
a letter to Hasse dated July 1, 1946 Grn wrote: I am relatively well off considering the circumstances.
I am working at the university but as a researcher only, which after all is what I wish to do.. After the war in
1945, the Friedrichs-Wilhelm Universitt of Berlin was short named Universitt Berlin, and later in 1949 it
was renamed Humboldt Universitt zu Berlin. It was situated in the Eastern (Soviet) sector of Berlin and is to
be distinguished from the Free University which had been founded in the Western sector.
56 The documents of Grns employment at the Berlin Academy are preserved and available in the Academys
archive.
57 H. L. Schmid was able to support also a number of other young (and not so young) mathematicians who
needed help. One of them was Kurt Heegner, the man who later would be the first to solve the class number 1
problem for imaginary quadratic fields [Hee52]. (Heegners paper was formulated in too fragmentary style and
hence it was not understood properly until Deuring [Deu68] cleared up the situation.)
58 Quite generally, people who knew him tell me that Grns opinions and beliefs were remarkably independent
of the Zeitgeist.
55 In
111
Already in 1942 Hasse had written to Grn explaining what possibilities there were
for him to obtain his doctorate. But at that time nothing came out of this. Now in 1946
H. L. Schmid proposed to Grn to apply to the university for admission to promotion
for doctorate. It is reported (by hearsay) that Grn was quite hesitating because he
did not like formalities of any kind. For, there had to be an extra permission because
Grn had not been a student of Berlin University, in fact he had never attended any
university. But H. L. Schmid finally succeeded to persuade Grn. 59
Thus on April 2, 1946 Grn submitted the necessary application form to the
dean of the science faculty of Berlin University. The fields in which he asked to be
examined were Pure Mathematics, Applied Mathematics and Theoretical Physics.
He submitted the thesis Beitrge zur Gruppentheorie III (Contributions to group
theory III) which two years later was published in the first volume of the new journal
Mathematische Nachrichten (See [Gr48a]). Officially H. L. Schmid signed as the
first referee for the thesis but he mentioned in his report that Magnus, as an expert in
this field, had checked it thoroughly.
The promotion documents for Otto Grn are preserved at the archives of the
Humboldt University. The examination took place on June 20, 1947 and the final
doctors diploma is signed on September 20, 1948. At this date Grn was 60 years.
2.7.2 16 more papers
In Section 2.5.2 we have cited a letter of Grn (dated May 9, 1935) in which he
claimed to have a huge pile of notes which already contain the essential ingredients
of future publications. Some of those publications, until 1945, we have already
mentioned. But it seems there was more in Grns pile of notes. For, from 1948 to
1964 Grn published 16 more papers, 13 of them on p-groups and related topics.
(The first of those papers he had used as his doctoral thesis.) About every year he
completed a new paper. This activity seems quite remarkable, considering that Grn
in 1948 was of age 60, and he was 76 at the time when his last paper appeared.
The first few of these papers were still checked by Magnus before publication,
but later, Magnus had emigrated to USA, Grn was at last able to work on his own.
He had learned to avoid erroneous conclusions in his publications and had become a
respected colleague among group theorists. He wisely stayed away from great and
famous problems, in view of his experiences he had gone through in earlier years with
Vandivers conjecture, Burnsides problem and the conjecture of Schur. 60 His papers
constituted valuable and useful contributions for the specialists; they appeared in
good journals in Germany and elsewhere. Grn became a known expert in p-groups
and related structures, and he was consulted as a referee for doctorate theses etc.
59 It
is not unlikely that H. L. Schmid used the argument that if Grn had the title of doctor then this would
imply some increase of his (small) salary.
60 In 1938 Grn had published a paper [Gr38] in which he claimed (among other results) that every representation of a finite group of exponent m can be realized in the field of m-th roots of unity. Schur had conjectured
this in 1912 with the group order instead of exponent. However, Grns proof turned out to be erroneous.
112
Two of Grns papers from this time were on number theory: perfect numbers,
and Bernoulli numbers. But these were only small notes.
In 1958 there was an increase in exchange of letters between Hasse and Grn,
and this concerned class groups of cyclotomic fields. Thus Grn had not completely
forgotten this topic with which he had started in the 1930s. As a result of this
correspondence Grn obtained a theorem which, however, turned out to be a special
case of Leopoldts Spiegelungssatz [Leo58]. Leopoldts paper was in press but not
yet published. Hasse offered to publish Grns manuscript since, after all, it had been
obtained independently, but Grn withdrew his manuscript. Nonetheless his letters
show that Grns number theoretical interest was still alive, and his standard was
high.
2.7.3 Wrzburg (19541963)
The hope that Berlin would be able to establish itself as a center of Mathematics
in Germany dwindled soon. Around 1950 the Gleichschaltung, in the communist
sense, of academic (and other) institutions in the Soviet occupied part of Germany
was intensified. As a consequence many people tried to go to West Germany. Hasse
accepted a position in Hamburg in 1950, and a number of younger people of his
circle went with him. In 1953 H. L. Schmid changed from Berlin to the University of
Wrzburg and again, a number of people went with him there.
Otto Grn too was among those who followed H. L. Schmid to Wrzburg. The
latter had been able to find means there for the financial support of Grn. At first
Grn became a member of the research center for applied mathematics in Wrzburg
which H. L. Schmid had newly founded together with Bilharz. 61
Later, after the early death of H. L. Schmid in 1956, Grn could be supported
through a teaching job (Lehrauftrag) for group theory at the University of Wrzburg,
which he received almost regularly for several years. There are still people living
who have attended Grns lecture courses, or at least have tried to do so. The story
is that each semester Grn announced a lecture on group theory, and after 23 hours
every student had dropped out because of Grns awkwardness in the presentation
of material (which Hasse had already observed in 1939). After that, Grn was happy
to be able to turn to his research without having to worry about lectures.
Between 1954 and 1961 Grn attended every group theory meeting in Oberwolfach; these meetings were directed by Reinhold Baer, one of them by Jean Dieudonn.
Since participation in Oberwolfach meetings is possible by personal invitation only,
this shows that his results were appreciated by the international group theory community. Four times Grn presented talks at those meetings (1955, 1959, 1960, 1961).
61 Herbert Bilharz had been, like H. L. Schmid, a graduate student of Hasse. In his Gttingen thesis [Bil37] he
had solved Artins conjecture for primitive roots in the function field case assuming the Riemann hypothesis
for function fields (which was finally verified by A. Weil). Later he went to applied mathematics and worked for
a time in the aircraft industry. In Wrzburg he held a chair for applied mathematics.
113
The abstracts of those talks are still available in the Oberwolfach abstract books (Vortragsbcher), they show that Grn talked about the results which he had obtained
in his papers. But as some participants of those meetings remember, his style of
lecturing had not improved.
2.8 Epilogue
In 1955 Grn was 67 years. It became clear that something had to be done to secure for
him some retirement pension. 62 This was difficult since he never had held a regular
position in a university. In the archives of Wrzburg University I have found a number
of documents, between 1955 and 1962, written by the Mathematics Department Head,
with the intention to obtain some kind of retirement pay for Grn.
In order to back those efforts, some leading group theorists were asked to write
their opinion on Grn. Let us cite excerpts of those opinions, all dated in 1955, in
order to put into evidence that Grn was respected as a group theorist throughout the
world:
F. W. Levi, Freie Universitt Berlin: Es ist Herrn Grn gelungen, neue Methoden
fr die Erforschung der endlichen Gruppen zu entwickeln und dadurch dieses
Gebiet neu zu erschliessen. Schon seine ersten Ergebnisse haben Aufsehen
unter den Algebraikern erregt und sind schnell in die Literatur, sogar in Lehrbcher bergegangen. Seit dieser Zeit hat er unermdlich weiter gearbeitet,
wichtige Ergebnisse erzielt und dadurch anderen Mitarbeitern den Weg zu
neuer Forschung geebnet. Herr Grn ist Autodidakt, hat nie ein Lehramt
bekleidet, aber er ist ein echter Gelehrter, und zwar ein Gelehrter von groer
wissenschaftlicher Bedeutung.
Grn succeeded to develop new methods for the investigation of finite groups
and thus to open this field from a new viewpoint. Already his first results have
attracted great attention among algebraists and were quickly included into the
literature, even into textbooks. Since then he has ever continued to work, he
has obtained important results and thus opened the way for the research of other
mathematicians. Grn is self-educated, has never had a teaching position,
but he is a true scholar with great scientific standing
R. Baer, University of Illinois, Urbana: O. Grn ist unzweifelhaft einer der fhrenden Gruppentheoretiker unserer Zeit. In der fundamentalen Arbeit ber die
endlichen p-Gruppen ist es ihm gelungen, die Ph. Hallsche Theorie der regulren p-Gruppen auf beliebige p-Gruppen auszudehnen, den dabei entstehenden neuen Phnomenen Rechnung zu tragen und dadurch neues Licht auf
die Flle der Erscheinungen in diesem reichen Gebiet zu werfen.
62 In
a letter of Grn to Hasse of August 29, 1955, Grn writes that he gets only 160 DM monthly.
114
Without doubt Grn is one of the leading group theorists of our time. In the
fundamental paper on finite p-groups he succeeded to extend Ph. Halls theory
of regular p-groups to p-groups of arbitrary structure. He was able to deal
with the new phenomena which showed up in this process, and thus to throw
new light upon the many aspects of this rich mathematical discipline.
B. H. Neumann, Hull: Otto Grn muss heutzutage als einer der bekanntesten und
berhmtesten Gruppentheoretiker gelten, und zwar keineswegs nur in Deutschland, sondern berall, wo Mathematik getrieben wird In drei so verschiedenartigen Monographien wie Lehrbuch der Gruppentheorie von Zassenhaus,
Gruppi astratti von Scorza und Teoriya Grupp von Kurosch werden die
Resultate von Grn mehrfach herangezogen.
Nowadays Otto Grn has to be counted as one of the most prominent group
theorists, by no means in Germany only but wherever mathematics is present
In three quite different monographys like Lehrbuch der Gruppentheorie
by Zassenhaus, Gruppi astratti by Scorza and Teoriya Grupp by Kurosh
his results are repeatedly used.
J. Dieudonn, Evanston, Ill.: conrmer tout lestime et ladmiration que jai
pour les travaux de M. le Prof. O. Grn. Ses ides sur la thorie des groupes se
distinguent par une remarquable originalit et une profondeur peu commune
the estimation and admiration which I harbor for the works of Prof. O. Grn.
His ideas about group theory are distinguished by a remarkable originality and
a rarely found depth
W. Magnus, New York University: Grn ist ein Mathematiker von wohlbegrndetem internationalen Ansehen. Seine Arbeiten zur Gruppentheorie werden
von mathematischen Autoren aller Lnder zitiert, und einige der von Herrn
Grn gefundenen Resultate gehren zum bleibenden Bestand der Gruppentheorie, was darin zum Ausdruck kommt, dass sie in allen modernen Lehrbchern
dargestellt werden (z. Bsp. Zassenhaus, Kurosch)
Grn is a mathematician of well founded international standing. His papers are
cited by mathematical authors of all countries, and some of his results belong
to the perpetual stock of group theory, which is evidenced by the fact that they
are treated in all modern textbooks (e.g., Zassenhaus, Kurosh)
H. Zassenhaus, McGill University, Montreal: Im Bereiche der mathematischen
Forschung dieses Jahrunderts ist mir kein anderes Beispiel der Entdeckung
eines hervorragenden Mathematikers im vorgerckten Alter bekannt geworden. Im neunzehnten Jahrhundert hat es die Flle von Sophus Lie und Weierstrass gegeben Durch seine Arbeiten hat sich Otto Grn einen Namen als
ausgezeichneter tiefforschender deutscher Mathematiker gemacht, den ich in
115
England und in den Vereinigten Staaten immer wieder mit Achtung und Bewunderung habe nennen hren.
In the realm of mathematical research I do not know any other example of an
excellent mathematician who was discovered in his midlife years only. In the
19th century there were the cases of Sophus Lie and Weierstrass Through his
work Otto Grn has become a well known name as a German mathematician,
doing deep research. I have heard mention his name again and again in England
and in the United States with respect and admiration
It is not clear from the Wrzburg documents whether the initiative on behalf of
Grn was successful. I am afraid it was not.
In any case, Grn returned to (West-)Berlin, his home town, in the year 1963 when
he was 75. After that date there were still some letters exchanged between Grn and
Hasse but they were restricted mainly to birthday greetings and the like. All the time
Grn continued to respect Hasse as his teacher, the one who opened mathematics for
him, and he expressed his thanks and admiration for Hasse in his letters.
Starting from 1971 we find in Grns letterhead the title of Professor. Perhaps
we can conclude from this that he had obtained from the government this official title
and, we hope, finally some adequate retirement pension in view of his achievements.
In October 1974 Grn died at the age of 86. Among Hasses papers I found a
brief obituary, about half a page, dated October 10, 1974. But I do not know where
it had been published; perhaps it was a newspaper clip. There was no obituary in the
Jahresbericht of the DMV of which Grn was a member since 1939.
2.9 Addendum
I am indebted to Prof. Siegmund-Schultze who, after the above article had appeared,
had sent me a message dated January 9, 2006 with the information that the Bundesarchiv in Berlin contains a file on Otto Grn. 63 In the meantime I have been able to
look at it.
That file starts with a letter from Grn dated November 4, 1936 to the REM stating
that he was without any income and asking for help to find a position. The letter is
adressed to the mathematician Theodor Vahlen who, at that time, was a powerful
figure in the REM. In reply to Grns letter Vahlen asked him for a curriculum vitae
and a list of publications, and Grn supplied this eventually.
From his curriculum vitae we learn some facts about Grns former commercial
activities: Until 1925 he had been the manager of the Berlin branch of an Austrian
winery, thereafter he had been working as an independent auditor.
63 REM 2634. The abbreviation REM stands for Reichserziehungsministerium, i.e., the Ministerium for
educational policy of the German government which had been established by the Nazi government. In the
following we shall use this abbreviation for short.
116
The REM then asked Hasse, Bieberbach and Tornier for their opinion on Grn.
Hasse, as we have seen, knew Grn from earlier correspondence and from Grns
visit to Gttingen in 1935. He sent an extended opinion saying that Grn is talented,
has already obtained siginifant results, and seems to promise further success in his
research activities. But he adds that Grn, having no formal academic training, has
problems in teaching; he is not able to present his material in a clear way. Also, his
mathematical knowledge is very narrow. Therefore, Hasse continues, he recommends
to give Grn a grant for continuing his reasearch, preferably in connection with a
university. Then one could see whether he would develop to become a useful member
of the teaching staff.
Bieberbach wrote that he had not known Grn but has invited him for an interview.
From this he got the impression that Grn had acquired unusual knowledge in algebra
and number theory and was familiar with the traditional problems and methods. But
in Bieberbachs opinion it would be too early to decide whether a research grant
would be appropriate. Moreover, he added, Grns mathematical interests were quite
one-sided and so he would be of no use for, say, working on mathematical problems
of aircraft construction. He proposed to employ Grn in some administrative position
in such a way that there would be sufficient time for him to follow his mathematical
interests. Bieberbach also mentioned that his letter was written in accordance with
Tornier who had been present at the interview with Grn. 64
The next document is an entry of the REM, signed by Dr. Dames on December 14,
1936, reporting that he had met Grn and informed him that he will get some financial
support for his research work, at least for the next months. (A somewhat modest
amount of Reichsmark was mentioned.) Grn was advised for a possible job to
get into contact with Professor Bartels, the director of the Geophysical Institute in
Potsdam.
Apparently, for such a job Grn had to provide evidence for his arian descent
which was required in Nazi Germany. It took some time for him to provide that
document, and so he could start his work in Potsdam in January 1938 only. From
then on, every year in the month of March there was an appplication, sent by Grn,
for continuing his job at the Potsdam Institute, and this was supported by the director
of the Institute. According to these documents Grn did certain analytic and numeric
computations which arose in the work of the Institute, e.g., in order to describe
homogeneous magnetic fields. The last of those documents is dated March 1944, so
that apparently Grns financial situation was secured until March 1945. But we have
found no evidence that Grn had won enough freedom during that period to continue
his group theoretical research.
64 This
corroborates the fact that E. Tornier had been removed from his professorship in Gttingen in view of
his personal conduct, and that he is now at Berlin University. But since he posed as a staunt Nazi he still seems to
enjoy the protection of Bieberbach and of Vahlen. (According to the files at the archive of Humboldt-University
in Berlin, this changed during the year 1937 when it became known that Tornier was addicted to alcohol (and
drugs) and was involved in financial fraudulence; he then was removed from the university.)
Chapter 3
At Emmy Noethers funeral
Translation of the article:

Zu Emmy Noethers Geburtstag. Einige neue Noetheriana.
Mitteilungen der Deutschen Mathematiker Vereinigung 15/1 (2007), 1521.
3.1
3.2
3.3
3.4
3.5
Introduction
Funeral speech by Hermann Weyl
Grete Hermann to van der Waerden
President Park to Otto Noether
Marguerita Lehr, Professor at Bryn Mawr College
117
123
125
126
127
3.1 Introduction
The day of March 23, 2007 marks the 125th anniversary of Emmy Noethers birthday.
It may be suitable on this occasion to present in this article some new Noetheriana,
commemorating this great master of our science. I shall present some documents
which may help us to understand her life and the impact of her ideas on mathematics
and mathematicians.
First, I will show some documents from the Nachlass of Grete Hermann:
1. The text of a short speech of Hermann Weyl at Emmy Noethers funeral, delivered on April 17, 1935 in Bryn Mawr. The speech, which we present here
in English translation, was delivered in German language and was addressed
to a small circle of mourners.
2. A letter of Grete Hermann (in English translation) addressed to van der Waerden, dated January 24, 1982, in which she remembers their common days as
students of Emmy Noether. In this letter she mentions also the text of the above
mentioned speech by Hermann Weyl.
Secondly, I will show some documents from the archive of Bryn Mawr College where
Emmy Noether had found shelter after her forced emigration from Germany:
118
3. A report by Marion Park, Ph.D., at that time the president of Bryn Mawr
College, about the funeral ceremony mentioned above. The report is addressed
to Emmy Noethers cousin Otto Noether in Mannheim.
4. The text of an address of Professor Marguerita Lehr from Bryn Mawr, read
one day after the above mentioned funeral, on April 18, 1935, in the Chapel
of Bryn Mawr. Lehr reports about the reception of Emmy Noether by faculty
and students of Bryn Mawr, and her activities there.
I do not claim that these four documents are of special importance. But they
give us at least some evidence about the impact which Emmy Noether had left on
her environment. For we still have the following question unanswered: How can we
explain the fact that Emmy Noether had been able to exert, during her lifetime, such
a wide influence on the mathematical thinking of her contemporaries?
After all, she had comparatively few publications: her Collected Papers amount
to just one volume, and only less than half of it contains those papers which are
usually cited as witnesses of her fame. And even those were often written not by
herself but by her students (van der Waerden, Deuring). Also, she was not a brilliant
lecturer; all accounts of her contemporaries tell us that her lectures have to be rated as
chaotic, according the usual criteria. The lectures were comprehensible to the small
selected circle of disciples only, who were used to her style of talking. Her way
to do mathematics was not appreciated in every corner. For instance, Olga Taussky
reports that even some of Noethers Gttingen colleagues criticized the abstract form
in which she expressed her ideas. And this still happened in the early 1930s, i.e., at
a time when her name was already known worldwide and the attraction of Gttingen
as a mathematical center rested mainly upon her fame. 1 But then, what made her rise
to a unique personality among the mathematicians of the 20th century?
On first sight the answer to this question may be that she represented a mathematical trend which was to spread anyhow, namely Modern Algebra and, more
generally, the modern abstract way of reasoning. But this seems to be short of
providing a meaningful explanation of the whole phenomenon. Certainly her unique
personal character was involved.
Unfortunately we have only few direct records about her. There is no film, there
are only a few photos and no interviews. There do exist meaningful reports written
by contemporaries, but not many. Hence every new document about her, and about
her environment, may let us catch a glimpse of some characteristic trait of her, and
so may serve as a contribution in forming a valid picture.
Before presenting those documents it is perhaps not without interest to report
about how I discovered them; this happened not without some luck.
All of this started with our plan to edit the correspondence between Hasse and
Emmy Noether [LR06]. Those letters document in rare clarity the development of the
1 Compare
the testimonials cited in Chapter 5.
119
mathematical relation between the two. Hasse was not one of the so-called Noether
boys who had gathered around her in Gttingen. He had been raised in a completely
different mathematical environment, viz. in Marburg with Hensel. As we learn from
the letters it happened in 1924, when Hasse had the position as Privatdozent at the
University in Kiel, that he came into the circle of influence of Emmy Noether. He met
her at the annual meeting of the DMV (German Mathematical Society) in Innsbruck
where she presented her axiomatic description of what today are called Dedekind
rings. Hasse was impressed about the easiness and the generality of Noethers theory.
From the letters we learn that he became more and more convinced of the great
power which was inherent in Emmy Noethers conception of mathematics 2 (although
the contact with Emmy Noether never led Hasse to give up his own mathematical
individual style).
The influence of Emmy Noether, on Hasse as well as on others, was based on her
ability to formulate their problems in an abstract form which, in her opinion, clarified
the situation. She did not solve mathematical problems but she led the way to the
solution by putting them on an abstract track which, in her opinion, would lead to
the solution by simplification. 3 We observe this not only in her relation with Hasse
but also, e.g., with van der Waerden, with Alexandrov and to a certain degree with
Hermann Weyl.
Hermann Weyl, three years younger than Emmy Noether, reports:
I have a vivid recollection of her when I was in Gttingen as visiting
professor in the winter semester of 19261927, and lectured on representations of continuous groups. She was in the audience; for just at
that time the hypercomplex number systems and their representations
had caught her interest and I remember many discussions when I walked
home after the lectures, with her and von Neumann through the cold,
dirty, rain-wet streets of Gttingen. When I was called permanently to
Gttingen in 1930, I earnestly tried to obtain from the Ministerium a better position for her, because I was ashamed to occupy such a preferred
position beside her whom I knew to be my superior as a mathematician in
many respects In my Gttingen years, 19301933, she was without
doubt the strongest center of mathematical activity there, considering
both the fertility of her scientific research program and her influence
upon a large circle of pupils. 4
How can it be explained that one of the leading mathematician of the time looked
at Emmy Noether as his superior ? After all, Emmy Noether did not have a position
as professor in Gttingen but only as a lecturer with very small remuneration, and
2 See,
e.g., his invited lecture 1929 in Prague with the title Die algebraische Methode. [Has30b].
e.g., the obituary, written by van der Waerden [vdW35].
4 Cited from the published Memorial Address of Weyl for Emmy Noether, which he delivered on April 26,
1935 at Goodhart Hall in Bryn Mawr. This address is printed in the Noether biography of Auguste Dick [Dic70].
3 See,
120
this position was granted only for one year at a time and had to be applied for
each year again. There were many other brilliant professors and lecturers at the
mathematics department in Gttingen, but Weyl did not say anyting like this about
other people. From the letters between Emmy Noether and Hasse we can conclude
that an important ingredient in their relation was Emmys ability, not only to distribute
new and inspiring ideas for future development, but also to establish warm personal
contacts to her colleagues and pupils. And this applies of course not only to Hasse
but also to others, in particular to Hermann Weyl. This personal note, paired with
her insistent persuasive power, can be observed in all reports which we have about
Emmy Noether.
Emmy Noether died on April 14, 1935. There exists a letter of Hermann Weyl to
Hasse, dated April 30, 1935, in which he reports about the small funeral ceremony
which had been held on April 17, 1935. Weyl wrote that the wreath of the Gttingen
mathematicians had been placed visibly on the coffin, as Hasse had wished it to be.
Moreover, Weyl wrote:
The friends in Germany can be assured that everything was done here
to offer the deceased a dignified farewell.
Weyl added that he includes in the letter a copy of his short speech which he
delivered at the funeral. But at the end of the letter he wrote:
P.S. After all, it appears to me more prudent not to include the text of
my speech. H.W.
Weyl does not explain why he did not include that text but looking at the date we can
guess why it was the year 1935, two years after the Nazis had come to power.
We have searched for the text of Weyls speech, since we wished to present it in
our book of the HasseNoether correspondence. But we did not find it in the papers
left by Weyl, nor in those of Richard Brauer who was close to Emmy Noether in his
years in Princeton, and also not in the papers of other friends of hers as, e.g., van
der Waerden. The text of Weyls longer address which he delivered a week later at
the official memorial in Bryn Mawr has been published several times and hence is
available to the mathematical community. Therefore we did not reproduce it again
in our book. We only informed the reader in a footnote that the text of Weyls short
speech at the funeral seemed to be lost.
However, recently I found that text quite unexpectedly at a place where we had
not expected it to be. But it was not possible any more to include it in our book. That
is the reason why I am presenting it here to the mathematical public.
It was quite by accident that I was led to that text. While reading the autobiography
of the physicist Werner Heisenberg Der Teil und das Ganze [Hei69]. I found a
chapter where he reported on a seminar in Leipzig in which he discussed with other
121
colleagues the philsophical foundations of the new quantum theory, which had been
established by his own essential contributions. There he says:
A special opportunity for philosophical discussion came one or two
years later when a young lady philosopher, Grete Hermann, joined us in
Leipzig in order to dispute with the nuclear physicists their philosophical
statements.
In the following chapter Heisenberg reports on the content of those discussions with
Grete Hermann (and with Carl Friedrich von Weizscker), which took place in the
year 1934.
Upon reading the name of Grete Hermann my curiosity was aroused. Was this
possibly the same Grete Hermann who had been in 1925 the first doctoral student
of Emmy Noether, with a thesis about polynomial operations in finitely many steps?
In this connection her name used to be well known among people doing number
theory; her paper in the Mathematische Annalen is still of interest (including van der
Waerdens comments in the next volume of Mathematische Annalen). However, I
did not know anything about her life. (After all, it happens quite often that people
know the name and perhaps the most important results of mathematicians of former
generations, but much less about their personal life.)
Heisenberg had reported that the said seminar discussions took place in the town
of Leipzig. This reminded me that, at that time, van der Waerden had been professor at Leipzig University. We know from his own words that he had accepted that
position particularly since there he could establish contact with Heisenberg and the
people around him. Van der Waerdens book Die gruppentheoretische Methode in der
Quantenmechanik (Group theoretical methods in quantum mechanics) documents his
connection to the theoretical physicists in Leipzig of that time. In Gttingen van der
Waerden had been a fellow student of Grete Hermann. Was it possible that she had
come to Leipzig through her contact with van der Waerden?
My inquiries then showed definitely that the Grete Hermann who was mentioned
in Heisenbergs book was identical with the Grete Hermann who had been a student of
Emmy Noether. Here we cannot go into the interesting and quite unusual biography
of Grete Hermann, although this would be a worthwhile task, of interest not only for
mathematicians but also for physicists, philosophers, social scientists and educators.
But I learned that there exists a Nachlass of Grete Hermann in the Archiv der sozialen
Demokratie in Bonn.
I went to that archive hoping to find letters from Emmy Noether but was disappointed. In retrospect this is understandable since while in Gttingen they did not
communicate by letters but they talked with each other.
However, quite unexpectedly I found there the German text of Weyls speech.
It seems that although Weyl did not send the text to Hasse he had sent it to Grete
Hermann. Or, maybe she obtained this text at some later time, perhaps through van
der Waerden? We do not know.
122
In any case, I believe that this moving text should not be forgotten. We feel that
there was more than high esteem for Emmy Noether as a colleague. She had also
been able to establish a special personal relationship to Weyl. He was deeply moved
by her sudden, unexpected death, and this enabled him to present such a speech in a
convincing manner.
We also show a letter of Grete Hermann to van der Waerden which we found in
her Nachlass too. As we see, the speech of Weyl is mentioned there. In that Nachlass
there are also two letters of Auguste Dick (who had written the first and only Noether
biography) to Grete Hermann, but we did not include it in our article.
A first version of our article had been submitted already in September 2007 but
only with the first two documents cited at the start of this chapter. Since then I have
discovered in various archives more Noetheriana which in my opinion also would
be of interest. Those are planned to be published later, except the two which I have
found in the archive of Bryn Mawr and which refer explicitly to the funeral of Emmy
Noether.
The report of President Park, mentioned in 3. above, is part of a letter which
she wrote to Dipl. Ing. Otto Nther, a cousin of Emmy Noether, who resided in the
German town of Mannheim. The letter contains, among others, a precise medical
report about the cause of the death of Emmy Noether following a tumor operation.
It also contains details about the Nachlass of Emmy Noether. Here we have only
included that part of this letter which concerns her funeral. We read that not only
Hermann Weyl had delivered a short speech but also Richard Brauer and Olga Taussky
and, in addition, Anna Pell Wheeler, Head of the Mathematics Department in Bryn
Mawr.
At that time Richard Brauer held a position as assistant to Hermann Weyl at the
Institute for Advanced Study in Princeton. During Emmy Noethers frequent visits to
Princeton there had developed a close relationship between her and the Brauer family.
Olga Taussky had been introduced to Emmy Noether during her time in Gttingen
as a post-doc. In 1935 she worked at Bryn Mawr college as a lecturer.
Anna Wheeler had studied in Gttingen with Hilbert in the years 19061908.
There she had written her thesis on integral equations but she did not get her Ph.D.
in Gttingen. (I had some trouble with Professor Hilbert.) It was only in 1910
that she got her doctoral degree in Chicago with E. H. Moore. We know from various
sources that in Bryn Mawr a friendly relationship developed between Anna Wheeler
and Emmy Noether.
The text of the speeches of Brauer, Taussky and Wheeler seem not to have been
preserved. On the other hand, I found in the archives of Bryn Mawr a handwritten
manuscript by Marguerita Lehr, professor in the mathematics department, in which
she reports on the reception of Emmy Noether in Bryn Mawr. The text is dated:
Chapel Thursday April 18, 1935, i.e., one day after the funeral. We suppose that
it was read at a service in the chapel of Bryn Mawr on that day. The text describes,
from the point of view of her colleagues in Bryn Mawr, how Emmy Noether had been
123
received and how she had started her new academic life there. In a short time Emmy
Noether had won the hearts of her colleagues and students. I believe this is of interest
in addition to the reports which she had sent to Hasse from Bryn Mawr.
At the end of her manuscript Lehr writes that Emmy Noethers two years at Bryn
Mawr were happy ones and for this she refers to statements of friends. In connection
with this we have found a letter of Abraham Flexner, the founder and first director of
the Institute for Advanced Study in Princeton; the letter is dated April 25, 1934 and
is addressed to President Park. There we read:
it ought to make you and Mrs. Wheeler happy to know that a few
weeks ago she [Emmy Noether] remarked to Professor Veblen that the
last year and a half had been the very happiest in her whole life, for
she was appreciated in Bryn Mawr and Princeton as she had never been
appreciated in her own country.
Now, here are the announced documents: 5
3.2 Funeral speech by Hermann Weyl

The following text 6 was read out loud by Hermann Weyl on Emmy Noethers funeral
on April 17, 1935.
The hour has come, Emmy Noether, in which we must forever take our leave of
you. Many will be deeply moved by your passing, none more so than your beloved
brother Fritz, who, separated from you by half the globe, was unable to be here, and
who must speak his last farewell to you through my mouth. His are the flowers I lay
on your coffin. We bow our heads in acknowledgement of his pain, which it is not
ours to put into words.
But I consider it a duty at this hour to articulate the feelings of your German
colleagues those who are here, and those in your homeland who have held true to
our goals and to you as a person. I find it apt, too, that our native tongue be heard at
your graveside the language of your innermost sentiments and in which you thought
your thoughts and which we hold dear whatever power may reign on German soil.
Your final rest will be in foreign soil, in the soil of this great hospitable country that
offered you a place to carry on your work after your own country closed its doors on
you. We feel the urge at this time to thank America for what it has done in the last
two years of hardship for German science, and to thank especially Bryn Mawr, where
they were both happy and proud to include you amongst their teachers.
5 With friendly permission of the Archiv der sozialen Demokratie in Bonn and the archive of Bryn Mawr
College.
6 Translated from German by Ian Beaumont.
124
Justifiably proud, for you were a great woman mathematician I have no reservations in calling you the greatest that history has known. Your work has changed
the way we look at algebra, and with your many gothic letters you have left your
name written indelibly across its pages. No-one, perhaps, contributed as much as
you towards remoulding the axiomatic approach into a powerful research instrument,
instead of a mere aid in the logical elucidation of the foundations of mathematics, as
it had previously been. Amongst your predecessors in algebra and number theory it
was probably Dedekind who came closest.
When, at this hour, I think of what made you what you were, two things immediately come to mind . The first is the original, productive force of your mathematical
thinking. Like a too ripe fruit, it seemed to burst through the shell of your humanness.
You were at once instrument of and receptacle for the intellectual force that surged
forth from within you. You were not of clay, harmoniously shaped by Gods artistic
hand, but a piece of primordial human rock into which he breathed creative genius.
The force of your genius seemed to transcend the bounds of your sex and in
Gttingen we jokingly, but reverentially, spoke of you in the masculine, as den
Noether. But you were a woman, maternal, and with a childlike warmheartedness.
Not only did you give to your students intellectually fully and without reserve
they gathered round you like chicks under the wings of a mother hen; you loved them,
cared for them and lived with them in close community.
The second thing that springs to mind is that your heart knew no malice; you
did not believe in evil, indeed it never occurred to you that it could play a role in
the affairs of man. This was never brought home to me more clearly than in the
last summer we spent together in Gttingen, the stormy summer of 1933. In the
midst of the terrible struggle, destruction and upheaval that was going on around us
in all factions, in a sea of hate and violence, of fear and desperation and dejection
you went your own way, pondering the challenges of mathematics with the same
industriousness as before. When you were not allowed to use the institutes lecture
halls you gathered your students in your own home. Even those in their brown shirts
were welcome; never for a second did you doubt their integrity. Without regard for
your own fate, openhearted and without fear, always conciliatory, you went your own
way. Many of us believed that an enmity had been unleashed in which there could
be no pardon; but you remained untouched by it all. You were happy to go back to
Gttingen last summer, where, as if nothing had happened, you lived and worked
with German mathematicians striving for the same goals. You planned on doing the
same this summer.
You truly deserve the wreath that the mathematicians in Gttingen have asked me
to lay on your grave.
We do not know what death is. But is it not comforting to think that souls will
meet again after this life on Earth, and how your fathers soul will greet you? Has
any father found in his daughter a worthier successor, great in her own right?
125
You were torn from us in your creative prime; your sudden departure, like the echo
of a thunderclap, is still written on our faces. But your work and your disposition
will long keep your memory alive, in science and amongst your students, friends and
colleagues.
Farewell then, Emmy Noether, great mathematician and great woman. Though
decay will take your mortal remains, we will always cherish the legacy you left us.
Hermann Weyl
Emmy Noethers tomb in Bryn Mawr
3.3 Grete Hermann to van der Waerden

Grete Henry-Hermann
2800 Bremen 1, den 24. Januar 1982

Am Barkhof 19
Herrn
Prof. Dr. B. L. van der Waerden
Wiesliacher 5
8053 Zrich
Dear Herr van der Waerden!
Please permit me to address you in this simple manner; this reflects my thoughts
back to our common time as students! In front of me there is the invitation to
a collquium in Erlangen named after Emmy Noether. There you will talk about
126
your Gttingen years of study. This aroused so many memories: In my mind I

am seeing the lecture room No. 16 in the second floor of the Gttingen auditorium;
Emmy Noether stands at the blackboard pondering intensively about something; in
front of her only a small group of students is sitting who are heavily involved; you
and I are among them.
I shall not come to the colloquium not only because my ears which have grown
old do not any more do their full duty, but mainly because in the course of the
past decades I have lost the contact to ideal theory. Already Emmy Noether said
resentfully to me, when after my examinations I became assistant to my other teacher
in Gttingen, the philosopher Leonard Nelson: Now she has studied mathematics
for four years and suddenly she discovers her philosophical heart!
But in this year, which contains the 100. birthdays of the two most important
teachers of my student years, I remember joyfully and thankfully this woman, of
mathematical originality and human affection. She lent me her help not only in
mathematics but also with some annoying formal problems connected with examinations.
On the blackboard in the lecture room no. 16 she developed the many ideals as
mentioned in one of her obituaries: With many gothic letters you have left your
name written indelibly across the pages of mathematics. 7
In remembrance of Emmy Noether I am greeting you!
Yours
G.H.
3.4 President Park to Otto Noether

Extract from a letter dated May 18, 1935.
We arranged to hold the funeral at my own house on Wednesday, the 17th. The
service at my house took place at three oclock in the afternoon. There were, I think,
about sixty persons present, all the members of faculty of Bryn Mawr College who
had known her, and her students here, and a large number of faculty and students of
the Department of Mathematics at Princeton University. The coffin was completely
covered with beautiful flowers sent by many friends and organizations The exercises were very simple, and I thought beautiful. A trio of violin, cello and piano
music of Bach and Mozart was played for ten minutes at the beginning and again at
the end of the service. Professor Wheeler and Dr. Olga Taussky spoke briefly on
behalf of Bryn Mawr College in English. Professor Weyl and Dr. Brauer spoke on
behalf of her German colleagues and friends in German
7 Here
Grete Hermann obviously refers to the text of Weyls speech.
127
3.5 Marguerita Lehr, Professor at Bryn Mawr College

Text of the speech, handwritten, dated April 18, 1935.
When Bryn Mawr opened in 1933, President Park announced the coming of a
most distinguished foreign visitor to the Faculty, Dr. Emmy Noether. Among mathematicians that name always brings a stir of recognition; the group in this vicinity
waited with excitement and many plans for Dr. Noethers arrival. At Bryn Mawr there
was much discussion and rearrangement of schedule, so that graduate students might
be free to read and consult with Miss Noether until she was ready to offer definitely
scheduled courses. For many reasons it seemed that a slow beginning might have
to be made; the graduate students were not trained in Miss Noethers special field,
the language might prove a barrier , after the academic upheaval in Gttingen
the matter of settling into a new and puzzling environment might have to be taken
into account. When she came, all of these barriers were suddenly non-existent, swept
away by the amazing vitality of the woman whose fame as the inspiration of countless
young workers had reached America long before she did. In a few weeks the class
of four graduates was finding that Miss Noether could and would use every minute
of time and all the depth of attention that they were willing to give. In this second
year her work had become an integral part of the department; she had taken on an
honors student, her group of graduates has included three research fellows here on
scholarships or fellowships specially awarded to take full advantage of her presence,
and the first Ph. D. dissertation directed at Bryn Mawr by Miss Noether has just gone
to the Committee bearing her recommendation.
Professor Brauer in speaking yesterday of Miss Noethers powerful influence
professionally and personally among the young scholars who surrounded her in Gttingen said that they were called the Noether family, and that when she had to leave
Gttingen, she dreamed of building again somewhere what was destroyed then. We
realize now with pride and thankfulness that we saw the beginning of a new Noether
family here. To Miss Noether her work was as inevitable and natural as breathing,
a background for living taken for granted; but that work was only the core of her
relation to students. She lived with them and for them in a perfectly un-selfishness
way. She looked at the world with direct friendliness and unfeigned interest, and she
wanted them to do the same. She loved to walk, and many a Saturday with five or six
students she tramped the roads with a fine disregard for bad weather. Mathematical
meetings at the University of Pennsylvania, at Princeton, at NewYork, began to watch
for the little group, slowly growing, which always brought something of the freshness
and buoyance of its leader.
Outside of the academic circle, Miss Noether continually delighted her American
friends by the avidity with which she gathered information about the American environment. She was proud of the fact that she spoke English from the very first; she
wanted to know how things were done in America, whether it were giving a tea or
128
taking a Ph. D., and she attacked each single subject with the disarming candor and
vigorous attention which won every one who knew her.
Emmy Noether might have come to America as a bitter person, or a despondent
person. She came instead in open friendliness, pleased beyond measure to go on
working as she had, even in circumstances so different from the ones she had loved.
And our final consolation is that she made here too a place that was hers alone. We
feel not only greatly honoured that she wanted to stay and work with us; we feel
profoundly thankful for the assurance that her friends have brought to us that her
two short years at Bryn Mawr were happy ones.
Chapel Thursday April 18, 1935
Marguerita Lehr
Chapter 4
Emmy Noether and Hermann Weyl
Emmy Noether and Hermann Weyl.

In: Groups and Analysis: The legacy of Hermann Weyl, ed. by Karin Tent.
London Mathematical Society Lecture Note Series Vol. 394 (2008), 285326.
This is the somewhat extended manuscript of a talk presented at the Hermann Weyl conference
in Bielefeld, September 10, 2006.
4.1 Preface
4.2 Introduction
4.3 The first period: until 1915
4.4 The second period: 19151920
4.5 The third period: 19201932
4.6 Gttingen exodus: 1933
4.7 Bryn Mawr: 19331935
4.8 The WeylEinstein letter to the NYT
4.9 Appendix: documents
129
130
131
136
138
149
153
156
159
4.1 Preface
This is a conference in honor of Hermann Weyl and so I may be allowed, before
touching the main topic of my talk, to speak about my personal reminiscences of
him.
It was in the year 1952. I was 24 and had my first academic job at Mnchen when
I received an invitation from van der Waerden to give a colloquium talk at Zrich
University. In the audience of my talk I noted an elder gentleman, apparently quite
interested in the topic. Afterwards it turned out to be Hermann Weyl he approached
me and proposed to meet him next day at a specific point in town. There he told me
that he wished to know more about my doctoral thesis, which I had completed two
years ago already but which had not yet appeared in print. Weyl invited me to join him
on a tour on the hills around Zrich. On this tour, which turned out to last for several
hours, I had to explain to him the content of my thesis which contained a proof of the
130
4 Emmy Noether and Hermann Weyl
Riemann hypothesis for function fields over finite base fields. He was never satisfied
with sketchy explanations, his questions were always to the point and he demanded
every detail. He seemed to be well informed about recent developments.
This task was not easy for me, without paper and pencil, nor blackboard and chalk.
So I had a hard time. Moreover the pace set by Weyl was not slow and it was not
quite easy to keep up with him, in walking as well as in talking.
Much later only I became aware of the fact that this tour was a kind of examination,
Weyl wishing to find out more about that young man who was myself. It seems that
I did not too bad in this examination, for some time later he sent me an application
form for a grant-in-aid from the Institute for Advanced Study in Princeton for the
academic year 1954/55. In those years Weyl was commuting between Zrich and
Princeton on a half-year basis. In Princeton he had found, he wrote to me, that there
was a group of people who were working in a similar direction.
Hence I owe to Hermann Weyl the opportunity to study in Princeton. The two
academic years which I could work and learn there turned out to be important for
my later mathematical life. Let me express, posthumously, my deep gratitude and
appreciation for his help and concern in this matter.
The above story shows that Weyl, up to his last years, continued to be active
helping young people find their way into mathematics. He really cared. I did not
meet him again in Princeton; he died in 1955.
Let us now turn to the main topic of this talk as announced in the title.
4.2 Introduction
Both Hermann Weyl and Emmy Noether belonged to the leading group of mathematicians in the first half of 20th century, who shaped the image of mathematics as
we see it today.
Emmy Noether was born in 1882 in the university town of Erlangen, as the daughter of the renowned mathematician Max Noether. We refer to the literature for information on her life and work, foremost to the empathetic biography by Auguste Dick
[Dic70] which has appeared in 1970, the 35th year after Noethers tragic death. It
was translated into English in 1981. For more detailed information see, e.g., the very
carefully documented report by Cordula Tollmien [Tol90]. See also Kimberlings
publications on Emmy Noether, e.g., his article in [BS81].
When the Nazis had come to power in Germany in 1933, Emmy Noether was dismissed from the University of Gttingen and she emigrated to the United States. She
was invited by Bryn Mawr College as a visiting professor where, however, she stayed
and worked for 18 months only, when she died on April 14, 1935 from complications
following a tumor operation. 1
1 See
footnote 48.
131
Quite recently we have found the text, hitherto unknown, of the speech which
Hermann Weyl delivered at the funeral ceremony for Emmy Noether on April 17,
1935. 2 That moving text puts into evidence that there had evolved a close emotional
friendship between the two. There was more than a feeling of togetherness between
immigrants in a new and somewhat unfamiliar environment. And there was more than
high esteem for this women colleague who, as Weyl has expressed it 3 , was superior
to him in many respects. This motivated us to try to find out more about their mutual
relation, as it had developed through the years.
We would like to state here already that we have not found many documents for
this. We have not found letters which they may have exchanged. 4 Neither did Emmy
Noether cite Hermann Weyl in her papers nor vice versa 5 . After all, their mathematical activities were going into somewhat different directions. Emmy Noethers
creative power was directed quite generally towards the clarification of mathematical
structures and concepts through abstraction, which means leaving all unnecessary
entities and properties aside and concentrating on the essentials. Her basic work in
this direction can be subsumed under algebra, but her methods eventually penetrated
all mathematical fields, including number theory and topology.
On the other side, Hermann Weyls mathematical horizon was wide-spread, from
complex and real analysis to algebra and number theory, mathematical physics and
logic, also continuous groups, integral equations and much more. He was a mathematical generalist in a broad sense, touching also philosophy of science. His mathematical writings have a definite flair of art and poetry, with his book on symmetry as
a culmination point [Wey52].
We see that the mathematical style as well as the extent of Weyls research work
was quite different from that of Noether. And from all we know the same can be said
about their way of living. So, how did it come about that there developed a closer
friendly relationship between them? Although we cannot offer a clear cut answer to
this question, I hope that the reader may find something of interest in the following
lines.
4.3 The first period: until 1915

In the mathematical life of Emmy Noether we can distinguish four periods. 6 In her
first period she was residing in Erlangen, getting her mathematical education and
working her way into abstract algebra guided by Ernst Fischer, and only occasionally
visiting Gttingen. The second period starts in the summer of 1915 when she came
2 See
[Roq07b]. We have included in the appendix an English translation of Weyls text; see Section 4.9.2.
[Wey35].
4 With one exception; see Section 4.5.3.
5 There are exceptions; see Section 4.4.
6 Weyl [Wey35] distinguishes three epochs but they represent different time intervals than our periods.
3 See
132
to Gttingen for good, in order to work with Klein and Hilbert. This period is
counted until about 1920. Thereafter there begins her third period, when her famous
paper Idealtheorie in Ringbereichen (Ideal theory in rings) appeared, with which she
embarked on her own completely original mathematical path to cite a passage
from Alexandrovs memorial address [Ale83]. The fourth period starts from 1933
when she was forced to emigrate and went to Bryn Mawr.
4.3.1 Their mathematical background
Hermann Weyl, born in 1885, was about three years younger than Emmy Noether.
In 1905, when he was 19, he entered Gttingen University (after one semester in
Mnchen). On May 8, 1908 he obtained his doctorate with a thesis on integral
equations, supervised by Hilbert.
At about the same time (more precisely: on December 13, 1907) Emmy Noether
obtained her doctorate from the University of Erlangen, with a thesis on invariants
supervised by Gordan. Since she was older than Weyl we see that her way to Ph.D.
was longer than his. This reflects the fact that higher education, at that time, was
not as open to females as it is today; if a girl wished to study at university and get a
Ph.D. then she had to overcome quite a number of difficulties arising from tradition,
prejudice and bureaucracy. Noethers situation is well described in Tollmiens article
[Tol90]. 7
But there was another difference between the status of Emmy Noether and Hermann Weyl at the time of their getting the doctorate.
On the one side, Weyl was living and working in the unique Gttingen mathematical environment of those years. Weyls thesis belongs to the theory of integral
equations, the topic which stood in the center of Hilberts work at the time, and
which would become one of the sources of the notion of Hilbert space. And Weyls
mathematical curiosity was not restricted to integral equations. In his own words, he
was captivated by all of Hilberts mathematics. Later he wrote: 8
I resolved to study whatever this man [Hilbert] had written. At the end
of my first year I went home with the Zahlbericht under my arm,
and during the summer vacation I worked my way through it without
any previous knowledge of elementary number theory or Galois theory.
These were the happiest months of my life, whose shine, across years
burdened with our common share of doubt and failure, still comforts my
soul.
We see that Weyl in Gttingen was exposed to and responded to the new and
exciting ideas which were sprouting in the mathematical world at the time. His
mathematical education was strongly influenced by his advisor Hilbert.
7 For
additional material see also Tollmiens web page: http://www.tollmien.com/.

from the Weyl article in MacTutor History of Mathematics Archive.
8 Cited
133
On the other side, Noether lived in the small and quiet mathematical world of Erlangen. Her thesis, supervised by Paul Gordan, belongs to classical invariant theory,
in the framework of so-called symbolic computations. Certainly this did no longer
belong to the main problems which dominated mathematical research in the beginning of the 20th century. It is a well-known story that after Hilbert in 1888 had proved
the finiteness theorem of invariant theory which Gordan had unsuccessfully tried for
a long time, then Gordan did not accept Hilberts existence proof since that was not
constructive in his (Gordans) sense. He declared that Hilberts proof was theology,
not mathematics. Emmy Noethers work was fully integrated into Gordans formalism and so, in this way, she was not coming near to the new mathematical ideas of
the time. 9 In later years she described the work of her thesis as rubbish (Mist in
German 10 ). In a letter of April 14, 1932 to Hasse she wrote:
Ich habe das symbolische Rechnen mit Stumpf und Stil verlernt.
I have completely forgotten the symbolic calculus.
We do not know when Noether had first felt the desire to update her mathematical background. Maybe the discussions with her father helped to find her way;
he corresponded with Felix Klein in Gttingen and so was well informed about the
mathematical news from there. She herself reports that it was mainly Ernst Fischer
who introduced her to what was then considered modern mathematics. Fischer
came to Erlangen in 1911, as the successor of the retired Gordan. 11 In her curriculum
vitae which she submitted in 1919 to the Gttingen Faculty on the occasion of her
Habilitation, Noether wrote:
Wissenschaftliche Anregung verdanke ich wesentlich dem persnlichen
mathematischen Verkehr in Erlangen und in Gttingen. Vor allem bin
ich Herrn E. Fischer zu Dank verpichtet, der mir den entscheidenden
Ansto zu der Beschftigung mit abstrakter Algebra in arithmetischer
Auffassung gab, was fr all meine spteren Arbeiten bestimmend blieb.
I obtained scientific guidance and stimulation mainly through personal
mathematical contacts in Erlangen and in Gttingen. Above all I am
indebted to Mr. E. Fischer from whom I received the decisive impulse to
study abstract algebra from an arithmetical viewpoint, and this remained
the governing idea for all my later work.
9 Well, Noether had studied one semester in Gttingen, winter 1903/04. But she fell ill during that time and
had to return to her home in Erlangen, as Tollmien [Tol90] reports. We did not find any indication that this
particular semester has had a decisive influence on her mathematical education.
10 Cited from Auguste Dicks Noether biography [Dic70].
11 More precisely: Gordan retired in 1910 and was followed by Erhard Schmidt who, however, left Erlangen
one year later already and was followed in turn by Ernst Fischer. The name of Fischer is known from the
FischerRiesz theorem in functional analysis.
134
Thus it was Fischer under whose direction Emmy Noethers mathematical outlook
underwent the transition from Gordans formal standpoint to the Hilbert method of
approach, as Weyl stated in [Wey35].
We may assume that Emmy Noether studied, like Weyl, all of Hilberts papers, at
least those which were concerned with algebra or arithmetic. In particular she would
have read the paper [Hil90] where Hilbert proved that every ideal in a polynomial
ring is finitely generated; in her famous later paper [Noe21] she considered arbitrary
rings with this property, which today are called Noetherian rings. We may also
assume that Hilberts Zahlbericht too was among the papers which Emmy Noether
studied; it was the standard text which every young mathematician of that time read
if he/she wished to learn algebraic number theory. We know from a later statement
that she was well acquainted with it although at that later time she rated it rather
critically 12 , in contrast to Weyl who, as we have seen above, was enthusiastic about
it. But not only Hilberts papers were on her agenda; certainly she read Steinitz great
paper Algebraische Theorie der Krper (Algebraic Theory of Fields) [Ste10] which
marks the start of abstract field theory. This paper is often mentioned in her later
publications, as the basis for her abstract viewpoint of algebra.
4.3.2 Meeting in Gttingen 1913
Hermann Weyl says in [Wey35], referring to the year 1913:
She must have been to Gttingen about that time, too, but I suppose
only on a visit with her brother Fritz. At least I remember him much
better than her from my time as a Gttinger Privatdozent, 19101913.
We may conclude that he had met Emmy Noether in Gttingen about 1913, but also
that she did not leave a lasting impression on him on that occasion.
As Tollmien [Tol90] reports, it was indeed 1913 when Emmy Noether visited
Gttingen for a longer time (together with her father Max Noether). Although we
have no direct confirmation we may well assume that she met Weyl during this time.
In the summer semester 1913 Weyl gave two talks in the Gttinger Mathematische
Gesellschaft. In one session he reported on his new book Die Idee der Riemannschen
Flche (The idea of the Riemann surface) [Wey13], and in another he presented his
proof on the equidistribution of point sequences modulo 1 in arbitrary dimensions
[Wey16] both pieces of work have received the status of a classic by now. Certainly, Max Noether as a friend of Klein will have been invited to the sessions of
the Mathematische Gesellschaft, and his daughter Emmy with him. Before and after
the session people would gather for discussion, and from all we know about Emmy
Noether she would not have hesitated to participate in the discussions. From what
12 In a letter of November 17, 1926 to Hasse; see [LR06]. Olga Taussky-Todd [Tau81] reports from later time
in Bryn Mawr that once Emmy burst out against the Zahlbericht, quoting also Artin as having said that it delayed
the development of algebraic number theory by decades.
135
we have said in the foregoing section we can conclude that her mathematical status
was up-to-date and well comparable to Weyls, at least with respect to algebra and
number theory.
Unfortunately we do not know anything about the possible subjects of the discussions of Emmy Noether with Weyl. It is intriguing to think that they could have
talked about Weyls new book The idea of the Riemann surface. Weyl in his book defines a Riemann surface axiomatically by structural properties, namely as a connected
manifold X with a complex 1-dimensional structure. This was a completely new approach, a structural viewpoint. Noether in her later period used to emphasize on every
occasion the structural viewpoint. The structure in Weyls book is an analytic one,
and he constructs an algebraic structure from this, namely the field of meromorphic
funtions, using the so-called Dirichlet principle whereas Emmy Noether in her later
papers always starts from the function field as an algebraic structure. See for instance
her report [Noe19]. There she did not cite Weyls book but, of course, this does not
mean that she did not know it.
We observe that the starting idea in Weyls book was the definition and use of
an axiomatically defined topological space 13 . We wonder whether this book was the
first instance where Emmy Noether was confronted with the axioms of what later was
called a topological space. It is not without reason to speculate that her interest in
topology was inspired by Weyls book. In any case, from her later cooperation with
Paul Alexandrov we know that she was acquainted with problems of topology; her
contribution to algebraic topology was the notion of Betti group instead of the Betti
number which was used before. Let us cite Alexandrov in his autobiography [Ale80]:
In the middle of December Emmy Noether came to spend a month in
Blaricum. This was a brilliant addition to the group of mathematicians
around Brouwer. I remember a dinner at Brouwers in her honour during
which she explained the definition of the Betti groups of complexes,
which spread around quickly and completely transformed the whole of
topology.
This refers to December 1925. Blaricum was the place where L. E. J. Brouwer lived.
We have mentioned this contact of Emmy Noether to the group around Brouwer
since Weyl too did have mathematical contact with Brouwer. In fact, in his book The
idea of the Riemann surface Weyl mentioned Brouwer as a source of inspiration. He
writes:
In viel hherem Mae, als aus den Zitaten hervorgeht, bin ich dabei durch
die in den letzten Jahren erschienenen grundlegenden topologischen Untersuchungen Brouwers, deren gedankliche Schrfe und Konzentration
man bewundern mu, gefrdert worden;
13 The
Hausdorff axiom was not present in the first edition. This gap was filled in later editions.
136
I have been stimulated much more than the citations indicate by the
recent basic topological investigations of Brouwer, whose ideas have to
be admired in their sharpness and concentration;
Brouwers biographer van Dalen reports that Weyl and Brouwer met several times in
the early 1920s [vD99]. By the way, Emmy Noether, Hermann Weyl and L. E. J. Brouwer met in September 1920 in Bad Nauheim, at the meeting of the DMV. 14
Returning to the year 1913 in Gttingen: In the session of July 30, 1913 of the Gttinger Mathematische Gesellschaft, Th. v. Krmn reported on problems connected
with a recent paper on turbulence by Emmy Noethers brother Fritz. Perhaps Fritz too
was present in Gttingen on this occasion, and maybe this was the incident why Weyl
had remembered not only Emmy but also Fritz? That he remembered Fritz much
better may be explained by the topic of Fritz paper; questions of turbulence lead to
problems about partial differential equations, which was at that time more close to
Weyls interests than were algebraic problems which Emmy pursued.
4.4 The second period: 19151920

In these years Emmy Noether completed several papers which are of algebraic nature, mostly about invariants, inspired by the Gttingen mathematical atmosphere
dominated by Hilbert. She also wrote a report in the Jahresbericht der DMV on
algebraic function fields, in which Noether compares the various viewpoints of the
theory: analytic, geometric and algebraic (which she called arithmetic) and she
points out the analogies to the theory of number fields. That was quite well known
to the people working with algebraic functions, but perhaps not written up systematically as Emmy Noether did. Generally, these papers of hers can be rated as good
work, considering the state of mathematics of the time, but not as outstanding. It is
unlikely that Hermann Weyl was particularly interested in these papers; perhaps he
didnt even know about them. 15
But this would change completely with the appearance of Noethers paper on
invariant variation problems [Noe18] (Invariante Variationsprobleme). The main
result of this paper is of fundamental importance in many branches of theoretical
physics even today. It shows a connection between conservation laws in physics and
the symmetries of the theory. It is probably the most cited paper of Emmy Noether
up to the present day. In 1971 an English translation appeared [Tav71], and in 2004
a French translation with many comments [KS04].
In 1918, when the paper appeared, its main importance was seen in its applicability
in the framework of Einsteins relativity theory. Einstein wrote to Hilbert in a letter
of May 24, 1918:
14 DMV
15 In
= Deutsche Mathematiker Vereinigung = German Mathematical Society.

those years Weyl was no more in Gttingen but held a professorship at ETH in Zrich.
137
Gestern erhielt ich von Frl. Noether eine sehr interessante Arbeit ber
Invariantenbildung. Es imponiert mir, dass man diese Dinge von so allgemeinem Standpunkt bersehen kann Sie scheint ihr Handwerk zu
verstehen.
Yesterday I received from Miss Noether a very interesting paper on the
formation of invariants. I am impressed that one can handle those things
from such a general viewpoint She seems to understand her job.
Einstein had probably met Emmy Noether already in 1915 during his visit to Gttingen.
Emmy Noethers result was the fruit of a close cooperation with Hilbert and with
Klein in Gttingen during the past years. As Weyl [Wey35] reports:
Hilbert at that time was over head and ears in the general theory of
relativity, and for Klein, too, the theory of relativity brought the last
flareup of his mathematical interest and production.
Emmy Noether, although she was doubtless influenced, not only assisted them but her
work was a genuine production of her own. In particular, the connection of invariants
with the symmetry groups, with its obvious reference to Kleins Erlanger program,
caught the attention of the world of mathematicians and theoretical physicists. 16
Noethers work in this direction has been described in detail in, e.g., [Row99], [KS04],
[Wue05].
It is inconceivable that Hermann Weyl did not take notice of this important work
of Emmy Noether. At that time Weyl, who was in correspondence with Hilbert and
Einstein, was also actively interested in the theory of relativity; his famous book
Raum, Zeit, Materie (Space, Time, Matter) had just appeared. Emmy Noether had
cited Weyls book 17 , and almost certainly she had sent him a reprint of her paper. Thus,
through the medium of relativity theory there arose mathematical contact between
them. 18 Although we do not know, it is well conceivable that there was an exchange
of letters concerning the mathematical theory of relativity. From now on Weyl would
never remember her brother Fritz better than Emmy.
In 1919 Emmy Noether finally got her Habilitation. Already in 1915 Hilbert and
Klein, convinced of her outstanding qualification, had recommended her to apply for
Habilitation. She did so, but it is a sad story that it was unsuccessful because of her
gender although her scientific standing was considered sufficient. The incident is told
in detail in Tollmiens paper [Tol90]. Thus her Habilitation was delayed until 1919
after the political and social conditions had changed.
16 In [Row99] it is said that nevertheless few mathematicians and even fewer physisists ever read Noethers
original article .
17 The citation is somewhat indirect. Noether referred to the literature cited in a paper by Felix Klein [Kle18],
and there we find Weyls book mentioned. In a second paper of Noether [Noe23] Weyls book is cited directly.
18Added in proof: We read in [Row99] that there is a reference to Noether in Weyls book, tucked away in a
footnote.
138
We see again the difference between the scientific careers of Weyl and of Emmy
Noether. Weyl had his Habilitation already in 1910, and since 1913 he held a professorship in Zrich. Emmy Noethers Habilitation was possible only nine years
later than Weyls. As is well known, she never in her life got a permanent position;
although in the course of time she rose to become one of the leading mathematicians
in the world.
4.5 The third period: 19201932

The third period of Noethers mathematical life starts with the great paper Idealtheorie
in Ringbereichen (Ideal theory in rings) [Noe21]. 19 After Hilbert had shown in 1890
that in a polynomial ring (over a field as base) every ideal is finitely generated, Noether
now takes this property as an axiom and investigates the primary decomposition of
ideals in arbitrary rings satisfying this axiom. And she reformulates this axiom as an
ascending chain condition for ideals. Nowadays such rings are called Noetherian.
The paper appeared in 1921.
We note that she was nearly 40 years old at that time. The mathematical life of
Emmy Noether is one of the counterexamples to the dictum that mathematics is a
science for the young and the most creative work is done before 40. Emmy Noether
would not have been a candidate for the Fields Medal if it had already existed at that
time.
4.5.1 Innsbruck 1924 and the method of abstraction
We do not know whether and how Weyl took notice of the above-mentioned paper
of Noether [Noe21]. But her next great result, namely the follow-up paper [Noe26]
on the ideal theory of what are now called Dedekind rings, was duly appreciated by
Weyl. At the annual DMV-meeting in 1924 in Innsbruck Noether reported about it
[Noe24]. And Weyl was chairing that session; so we know that he was informed first
hand about her fundamental results.
In her talk, Emmy Noether defined Dedekind rings by axioms and showed that
every ring satisfying those axioms admits a unique factorization of ideals into prime
ideals. Well, Noether did not use the terminology Dedekind ring; this name was
coined later. Instead, she used the name 5-axioms-ring since in her enumeration
there were 5 axioms. Then she proved that the ring of integers in a number field
satisfies those axioms, and similarly in the funtion field case. This is a good example
of Noethers method of abstraction. By working solely with those axioms she first
generalized the problem, and it turned out that by working in this generalization the
19 Sometimes the earlier investigation jointly with Schmeidler [NS20] is also counted as belonging to this
period.
139
proof of prime decomposition is simplified if compared with the former proofs (two
of which had been given by Hilbert [Hil94]).
How did Weyl react to Noethers method of abstraction? At that time, this method
met sometimes with skepticism and even rejection by mathematicians. But Hilbert in
various situations had already taken first steps in this direction and so Weyl, having
been Hilberts doctorand, was not against Noethers method. After all, in his book
Space, Time, Matter Weyl had introduced vector spaces by axioms, not as n-tuples20 .
Weyls reaction can be extracted implicitly from an exchange of letters with Hasse
which happened seven years later. The letter of Weyl is dated December 8, 1931.
At that time Weyl held a professorship in Gttingen (since 1930) as the successor of
Hilbert. Thus Emmy Noether was now his colleague in Gttingen. Hasse at that time
held a professorship in Marburg (also since 1930) as the successor of Hensel. The
occasion of Weyls letter was the theorem that every simple algebra over a number field
is cyclic; this had been established some weeks ago by Brauer, Hasse and Noether,
and the latter had informed Weyl about it. So Weyl congratulated Hasse for this
splendid achievement. And he recalled the meeting in Innsbruck 1924 when he first
had met Hasse.
For us, Hasses reply to Weyls letter is of interest. 21 Hasse answered on December 15, 1931. First he thanked Weyl for his congratulations, but at the same
time pointed out that the success was very essentially due also to the elegant theory of Emmy Noether, as well as the p-adic theory of Hensel. He also mentioned
Minkowski in whose work the idea of the Local-Global Principle was brought to light
very clearly. And then Hasse continued, recalling Innsbruck:
Auch ich erinnere mich sehr gut an Ihre ersten Worte zu mir anllich
meines Vortrages ber die erste explizite Reziprozittsformel fr hheren
Exponenten in Innsbruck. Sie zweifelten damals ein wenig an der inneren
Berechtigung solcher Untersuchungen, indem Sie ins Feld fhrten, es sei
doch gerade Hilberts Verdienst, die Theorie des Reziprozittsgesetzes
von den expliziten Rechnungen frherer Forscher, insbesondere Kummers, befreit zu haben.
I too remember very well your first words to me on the occasion of my
talk in Innsbruck, about the first explicit reciprocity formula for higher
exponent. You somewhat doubted the inner justification of such investigations, by pointing out that Hilbert had freed the theory of the reciprocity law from the explicit computations of former mathematicians,
in particular Kummers.
We conclude: Hasse in Innsbruck had talked on explicit formulas and Weyl had
critized this, pointing out that Hilbert had embedded the reciprocity laws into more
20 This
21 We
has been expressly remarked by Mac Lane [ML81].

have found Hasses letter in the Weyl legacy in the archive of the ETH in Zrich.
140
structural results. Probably Weyl had in mind the product formula for the so-called
Hilbert symbol which, in a sense, comprises all explicit reciprocity formulas. 22 Usually this product formula was considered as the final word on reciprocity, and Weyl
too adhered to this opinion. But for Hasse this was only the starting point for deriving explicit, constructive reciprocity formulas, using heavily (if possible) the p-adic
methods of Hensel.
We can fairly well reconstruct the situation in Innsbruck: Emmy Noethers talk
had been very abstract, and Hasses achievement was in some sense the opposite
since he was bent on explicit formulas, and quite involved ones too. 23 Weyl had been
impressed by Emmy Noethers achievements which he considered as continuing along
the lines set by Hilberts early papers on number theory. In contrast, he considered
Hasses work as pointing not to the future but to the mathematical past.
We have mentioned here these letters WeylHasse in order to put into evidence
that already in 1924, Weyl must have had a very positive opinion on Emmy Noethers
methods, even to the point of preferring it to explicit formulas.
But as it turned out, Hasse too had been impressed by Noethers lecture. In the
course of the years after 1924, as witnessed by the HasseNoether correspondence
[LR06], Hasse became more and more convinced about Noethers abstract methods
which, in his opinion, served to clarify the situation; he used the word durchsichtig
(lucid). Hasses address at the DMV meeting in Prague 1929 [Has30b] expresses his
views very clearly. Hensels p-adic methods could also be put on an abstract base,
due to the advances in the theory of valuations. 24 But on the other hand, Hasse was
never satisfied with abstract theorems only. In his cited letter to Weyl 1931 he referred
to his (Hasses) class field report Part II [Has30a] which had appeared just one year
earlier. There, he had put Artins general reciprocity law 25 as the base, and from this
structural theorem he was able to derive all the known reciprocity formulas. Hasse
closed his letter with the following:
Ich kann aber natrlich gut verstehen, da Dinge wie diese expliziten Reziprozittsformeln einem Manne Ihrer hohen Geistes- und Geschmacksrichtung weniger zusagen, als mir, der ich durch die abstrakte
Mathematik DedekindE. Noetherscher Art nie restlos befriedigt bin,
ehe ich nicht zum mindesten auch eine explizite, formelmige konstruktive Behandlung daneben halten kann. Erst von der letzteren knnen
sich die eleganten Methoden und schnen Ideen der ersteren wirklich
vorteilhaft abheben.
22 But Hilbert was not yet able to establish his product formula in full generality. We refer to the beautiful and
complete treatment in Hasses class field report, Part 2 [Has30a] which also contains the most significant historic
references.
23 Hasses Innsbruck talk is published in [Has25]. The details are found in volume 154 of Crelles Journal
where Hasse had published five papers on explicit reciprocity laws.
24 For this see [Roq02]. See also Chapter 9, in particular page 232.
25 By the way, this was the first treatment of Artins reciprocity law in book form after Artins original paper
1927. We refer to our forthcoming book on the ArtinHasse correspondence.
141
But of course I well realize that those explicit reciprocity formulas

may be less attractive to a man like you with your high mental powers and
taste, as to myself. I am never fully satisfied by the abstract mathematics
of DedekindE. Noether type before I can also supplement it by at least
one explicit, computational and constructive treatment. It is only in
comparison with the latter that the elegant methods and beautiful ideas
of the former can be appreciated advantageously.
Here, Hasse touches a problem which always comes up when, as Emmy Noether
propagated, the abstract methods are put into the foreground. Namely, abstraction
and axiomatization is not to be considered as an end in itself but it is a method to deal
with concrete problems of substance. But Hasse was wrong when he supposed that
Weyl did not see that problem. Even in 1931, the same year as the above cited letters,
Weyl gave a talk on abstract algebra and topology as two ways of mathematical
comprehension [Wey32]. In this talk Weyl stressed the fact that axiomatization is
not only a way of securing the logical truth of mathematical results, but that it had
become a powerful tool of concrete mathematical research itself, in particular under
the influence of Emmy Noether. But he also said that abstraction and generalization
do not make sense without mathematical substance behind it. This is close to Hasses
opinion as expressed in his letter above. 26 The mathematical work of both Weyl and
Hasse puts their opinions into evidence.
At the same conference [Wey32] Weyl also said that the fertility of these abstracting methods is approaching exhaustion. This, however, met with sharp protests by
Emmy Noether, as Weyl reports in [Wey35]. In fact, today most of us would agree
with Noether. The method of abstracting and axiomatizing has become a natural and
powerful tool for the mathematician, with striking successes until today. In Weyls
letter to Hasse (which we have not cited fully) there are passages which seem to
indicate that in principle he (Weyl) too would agree with Emmy Noether. For, he
encourages Hasse to continue his work in the same fashion, and there is no mention of
an impending exhaustion. He closes his letter with the following sentence which,
in our opinion, shows his (Weyls) opinion of how to work in mathematics:
Es freut mich besonders, da bei Ihnen die in Einzelleistungen sich bewhrende wissenschaftliche Durchschlagskraft sich mit geistigem Weitblick paart, der ber das eigene Fach hinausgeht.
In particular I am glad that your scientific power, tested in various special
accomplishments, goes along with a broad view stretching beyond your
own special field.
26 Even more clearly Hasse has expressed his view in the foreword to his beautiful and significant book on
abelian fields [Has52].
142
4.5.2 Representations: 1926/27

We have made a great leap from 1924 to 1931. Now let us return and proceed along
the course of time. In the winter semester 1926/27 Hermann Weyl stayed in Gttingen
as a visiting professor, and he lectured on representations of continuous groups. In
[Wey35] he reports:
I have a vivid recollection of her [Emmy Noether] She was in the audience; for just at that time the hypercomplex number systems and their
representations had caught her interest and I remember many discussions
when I walked home after the lectures, with her and von Neumann, who
was in Gttingen as a Rockefeller Fellow, through the cold, dirty, rainwet streets of Gttingen.
This gives us information not only about the weather conditions in Gttingen in winter
time but also that a lively discussion between Weyl and Emmy Noether had developed.
We do not know precisely when Emmy Noether first had become interested in
the representation theory of groups and algebras, or hypercomplex systems in her
terminology. In any case, during the winter semester 1924/25 in Gttingen she had
given a course on the subject. And in September 1925 she had talked at the annual
meeting of the DMV in Danzig on Group characters and ideal theory. There she
advocated that the whole representation theory of groups should be subsumed under
the theory of algebras and their ideals. She showed how the Wedderburn theorems
for algebras are to be interpreted in representation theory, and that the whole theory
of Frobenius on group characters is subsumed in this way. Although she announced
a more detailed presentation in the Mathematische Annalen, the mathematical public
had to wait until 1929 for the actual publication [Noe29] 27 . Noether was not a quick
writer; she developed her ideas again and again in discussions, mostly on her walks
with students and colleagues into the woods around Gttingen, and in her lectures.
The text of her paper [Noe29] consists essentially of the notes taken by van der
Waerden at her lecture in the winter semester 1927/28. Although the main motivation
of Noether was the treatment of Frobenius theory of representations of finite groups,
it turned out that finite groups are treated on the last two pages only out of a total
of 52 pages. The main part of the paper is devoted to introducing and investigating
general abstract notions, capable of dealing not only with the classical theory of finite
group representations but with much more. Again we see the power of Noethers
abstracting methods. The paper has been said to constitute one of the pillars of
modern linear algebra. 28
We can imagine Emmy Noether in her discussions with Weyl on the cold, wet
streets in Gttingen 1926/27, explaining to him the essential ideas which were to become the foundation of her results in her forthcoming paper [Noe29]. We do not know
27 This
appeared in the Mathematische Zeitschrift and not in the Annalen as announced by Noether in Danzig.
from [Cur99] who in turn refers to Bourbaki.
28 Cited
143
to which extent these ideas entered Weyls book [Wey39] on classical groups. After
all, the classical groups which are treated in Weyls book are infinite while Noethers
theory aimed at the representation of finite groups. Accordingly, in Noethers work
there appeared a finiteness condition for the algebras considered, namely the descending chain condition for (right) ideals. If one wishes to use Noethers results for infinite
groups one first has to generalize her theory such as to remain valid in more general
cases too. Such a generalization did not appear until 1945; it was authored by Nathan
Jacobson [Jac45]. He generalized Noethers theory to simple algebras containing at
least one irreducible right ideal.
At this point let me tell a story which I witnessed in 1947. I was a young student
in Hamburg then. In one of the colloquium talks the speaker was F. K. Schmidt
who recently had returned from a visit to the USA, and he reported on a new paper
by Jacobson which he had discovered there. This was the above mentioned paper
[Jac45].29 F. K. Schmidt was a brilliant lecturer and the audience was duly impressed.
In the ensuing discussion Ernst Witt, who was in the audience, commented that all
this had essentially been known to Emmy Noether already.
Witt did not elaborate on his comment. But he had been one of the Noether
boys in 1932/33, and so he had frequently met her. There is no reason to doubt his
statement. It may well have been that she had told him, and perhaps others too, that
her theory could be generalized in the sense which later had been found by Jacobson.
Maybe she had just given a hint in this direction, without details, as was her usual
custom. In fact, reading Noethers paper [Noe29] the generalization is obvious to
any reader who is looking for it. 30 It is fascinating to think that the idea for such
a generalization arose from her discussions with Weyl in Gttingen in 1927, when
infinite groups were discussed and the need to generalize her theory became apparent.
By the way, Jacobson and Emmy Noether had met in 1934 in Princeton, when she
was running a weekly seminar. We cannot exclude the possibility that she had given
a hint to him too, either in her seminar or in personal discussions. After all, this was
her usual style, as reported by van der Waerden [vdW35].
4.5.3 A letter from N to W: 1927
As stated in the introduction we have not found letters from Emmy Noether to Weyl,
with one exception. That exception is kept in the archive of the ETH in Zrich. It is
written by Emmy Noether and dated March 12, 1927. This is shortly after the end
of the winter semester 1926/27 when Weyl had been in Gttingen as reported in the
previous section. Now Weyl was back in Zrich and they had to write letters instead
of just talking.
29 In those post-war years it was not easy to get hold of books or journals from foreign countries, and so the
1945 volume of the Transactions of AMS was not yet available at the Hamburg library.
30A particularly short and beautiful presentation is to be found in Artins article [Art50] where he refers to
Tate.
144
The letter concerns Paul Alexandrov and Heinz Hopf and their plan to visit Princeton in the academic year 1927/28.
We have already mentioned Alexandrov in Section 4.3.2 in connection with
Noethers contributions to topology. From 1924 to 1932 he spent every summer
in Gttingen, and there developed a kind of friendly relationship between him and
Emmy Noether. The relation of Noether to her Noether boys has been described by
Andr Weil as like a mother hen to her fledglings [Wei93]. Thus Paul Alexandrov was
accepted by Emmy Noether as one of her fledglings. In the summer semester 1926
Heinz Hopf arrived in Gttingen as a postdoc from Berlin and he too was accepted
as a fledgling. Both Alexandrov and Hopf became close friends and they decided to
try to go to Princeton University in the academic year 1927/28.
Perhaps Emmy Noether had suggested this; in any case she helped them to obtain
a Rockefeller grant for this purpose. It seems that Weyl also had lent a helping hand,
for in her letter to him she wrote:
Jedenfalls danke ich Ihnen sehr fr Ihre Bemhungen; auch Alexandrov und Hopf werden Ihnen sehr dankbar sein und es scheint mir sicher,
dass wenn die formalen Schwierigkeiten erst einmal berwunden sind,
Ihr Brief dann von wesentlichem Einuss sein wird.
In any case I would like to thank you for your help; Alexandrov and
Hopf too will be very grateful to you. And I am sure that if the formal
obstacles will be overcome then your letter will be of essential influence.
The formal obstacles which Noether mentioned were, firstly, the fact that originally the applicants (Alexandrov and Hopf) wished to stay for a period less than an
academic year in Princeton (which later they extended to a full academic year), and
secondly, that Hopfs knowledge of the English language seemed not to be sufficient
in the eyes of the Rockefeller Foundation (but Noether assured them that Hopf wanted
to learn English). 31 But she mentioned there had been letters sent to Lefschetz and
Birkhoff and that at least the latter had promised to approach the Rockefeller Foundation to make an exception.
Alexandrov was in Moscow and Hopf in Berlin at the time, and so the mother hen
acted as representative of her two chickens. 32
About Alexandrovs and Hopfs year in Princeton we read in the Alexandrov
article of MacTutors History of Mathematics archive:
Alexandrov and Hopf spent the academic year 192728 at Princeton in
the United States. This was an important year in the development of
31 It seems that his knowledge of English had improved in the course of time since Heinz Hopf had been elected
president of the IMU (International Mathematical Union) in 1955 till 1958.
32 In Kimberlings article [Kim81] it is reported: Handwritten letters dated 6/1/27 and 7/3/27 from Emmy
Noether to W. W. Tinsdale supporting Hopfs application are preserved in the International Educational Board
Collection at the Rockefeller Archive Center. Probably the letters from Weyl too are preserved there but we
have not checked this.
145
topology with Alexandrov and Hopf in Princeton and able to collaborate

with Lefschetz, Veblen and Alexander.
The letter from Noether to Weyl shows that both N. and W. were instrumental in
arranging this important Princeton year for Alexandrov and Hopf. Both were always
ready to help young mathematicians to find their way.
Remark. Later in 1931, when Weyl had left Zrich for Gttingen, it was Heinz Hopf
who succeeded Weyl in the ETH Zrich. At those times it was not uncommon that the
leaving professor would be asked for nominations if the faculty wished to continue
his line. We can well imagine that Weyl, who originally would have preferred Artin
33
, finally nominated Heinz Hopf for this position. If so then he would have discussed
it with Emmy Noether since she knew Hopf quite well. It may even have been that
she had taken the initiative and proposed to Weyl the nomination of Hopf. In fact, in
the case of Alexandrov she did so in a letter to Hasse dated October 7, 1929 when it
was clear that Hasse would change from Halle to Marburg. There she asked Hasse
whether he would propose the name of Alexandrov as a candidate in Halle. 34 It seems
realistic to assume that in the case of Heinz Hopf she acted similarly.
Remark. The above mentioned letter of Noether to Weyl contains a postscript which
gives us a glimpse of the mathematical discussion between the two (and it is the only
written document for this). It reads:
Die Mertens-Arbeit, von der ich Ihnen sprach, steht Monatshefte, Bd. 4.
Ich dachte an den Schluss, Seite 329. Es handelt sich hier aber doch nur
um Determinanten-Relationen, sodass es fr Sie wohl kaum in Betracht
kommt.
The Mertens paper which I mentioned to you is contained in volume 4
of the Monatshefte. I had in mind the end of the paper, page 329. But
this is concerned with determinant relations only, hence it will perhaps
not be relevant to your purpose.
That Mertens paper is [Mer93]. We have checked the cited page but did not find
any hint which would connect to Weyls work. Perhaps someone else will be able to
interpret Noethers remark.
4.5.4 Weyl in Gttingen: 19301933
Weyl in [Wey35] reports:
When I was called permanently to Gttingen in 1930, I earnestly tried to
obtain from the Ministerium a better position for her [Emmy Noether],
33 In
fact, in 1930 Artin received an offer from the ETH Zrich which, however, he finally rejected.
however, did not work out. The successor of Hasse in Halle was Heinrich Brandt, known for the
introduction of Brandts gruppoid for divisor classes in simple algebras over number fields.
34 This
146
because I was ashamed to occupy such a preferred position beside her

whom I knew my superior as a mathematician in many respects.
We see that by now, Weyl was completely convinced about the mathematical
stature of Emmy Noether. After all, Emmy Noether in 1930 was the world-wide
acknowledged leader of abstract algebra, and her presence in Gttingen was the main
attraction for young mathematicians from all over the world to visit the Mathematical
Institute and study with her.
E. Noether and H. Weyl (with hat) among a group of mathematicians
It would be interesting to try to find out which better position Weyl had in mind
in his negotiations with the Ministerium in Berlin. Maybe he wished tenure for her,
and an increase of her salary. The archives in Berlin will perhaps have the papers
and reports of Weyls negotiations. From those papers one may be able to extract
the reasons for the rejection. But the opposition against Noethers promotion did not
only come from the Ministerium in Berlin. It seems that a strong opposition came
also from among the mathematician colleagues in Gttingen, for Weyl continues with
his report as follows:
nor did an attempt [succeed] to push through her election as a member
of the Gttinger Gesellschaft der Wissenschaften. Tradition, prejudice,
external considerations, weighted the balance against her scientific merits and scientific greatness, by that time denied by no one.
I do not know whether there exist minutes of the meetings of the Gttinger Gesellschaft
der Wissenschaften in 1930. If so then it would be interesting to know the traditional, biased and external arguments which Weyl said were put forward against
147
Emmy Noether from the members of the Gesellschaft der Wissenschaften. Was it
still mainly her gender? Or was it the opposition to her abstract mathematical
methods? In any case, the decision not to admit Emmy Noether as a member of the
Gttinger Gesellschaft der Wissenschaften is to be regarded as an injustice to her and
a lack of understanding of the development of modern mathematics. After all, the
Emmy Noether of 1930 was quite different from the Emmy Noether of 1915. Now
in 1930, she had already gone a long way on her own completely original mathematical path , and her working and conceptual methods had spread everywhere.
She could muster high-ranking colleagues and students who were fascinated by her
way of mathematical thinking.
Nevertheless it seems that in Gttingen, even among mathematicians, there existed
some opposition against her abstract methods. Olga Taussky-Todd recalls in [Tau81]
her impression of the Gttingen mathematical scene:
not everybody liked her [Emmy Noether], and not everybody trusted
that her achievements were what they later were accepted to be.
One day Olga Taussky had been present when one of the senior professors talked
very roughly to Emmy Noether. (Later he apologized to her for this insult.) When
Emmy Noether had her 50th birthday in 1932 then, as Olga Taussky recalls, nobody
at Gttingen had taken notice of it, although at that time all birthdays were published
in the Jahresbericht of the DMV. 35 Reading all this, I can understand Emmy Noether
when later in 1935 she said to Veblen about her time in the USA:
The last year and a half had been the very happiest in her whole life, for
she was appreciated in Bryn Mawr and Princeton as she had never been
appreciated in her own country. 36
Thus it seems that Weyls statement that her scientific merits and scientific greatness by that time was denied by no one did not describe the situation exactly. Perhaps,
since Weyl was the premier professor of mathematics at Gttingen, and since he
was known to respect and acknowledge Noethers merits and scientific greatness,
nobody dared to tell him if he disagreed. Olga Taussky-Todd remembers that
outside of Gttingen, Emmy was greatly appreciated in her country.
We may add that this was not only so in her country but also world-wide. And
of course also in Gttingen there was an ever-growing fraction of mathematicians,
including Weyl, who held Noether in high esteem.
As to Hermann Weyl, let us cite Saunders Mac Lane who was a student at Gttingen
in the period 193133. We read in [Mac81]:
35 But Hasse in Marburg had sent her a birthday cake, together with a paper which he had dedicated to her.
The paper was [Has33a]. See [LR06].
36 Cited from a letter of Abraham Flexner to President Park of Bryn Mawr; see [Roq07b].
148
When I first came to Gttingen I spoke to Professor Weyl and expressed

my interest in logic and algebra. He immediately remarked that in algebra Gttingen was excellently represented by Professor Noether; he
recommended that I attend her courses and seminars By the time of
my arrival she was Ausserordentlicher Professor. However, it was clear
that in the view of Weyl, Hilbert, and the others, she was right on the
level of any of the full professors. Her work was much admired and her
influence was widespread.
Mac Lane sometimes joined the hiking parties (Ausflug) of Emmy Noether
and her class to the hills around Gttingen. Noether used these hiking parties to
discuss algebra, other mathematical topics and Russia. 37 It seems that Weyl too
joined those excursions occasionally. There is a nice photo of Noether with Weyl
and family, together with a group of mathematicians posing in front of the Gasthof
Vollbrecht. The photo is published in [BS81] and dated 1932. Since Artin is seen as
a member of the hiking party, it seems very probable that the photo was taken on the
occasion of Artins famous Gttingen lectures on class field theory which took place
from February 29 to March 2, 1932. 38 This was a big affair and a number of people
came from various places in order to listen to Artin lecturing on the new face of class
field theory. The lectures were organized by Emmy Noether. Since she was not a
full professor and, accordingly, had no personal funds to organize such meetings we
suppose that one of her colleagues, probably Weyl, had made available the necessary
financial means for this occasion. In any case we see that by now she was able to
get support for her activities in Gttingen, not only for the Artin lectures but also for
other speakers.
The International Congress of Mathematicians took place in September 1932
in Zrich. Emmy Noether was invited to deliver one of the main lectures there.
Usually, proposals for invited speakers at the IMU conferences were submitted by
the presidents of the national mathematical organizations which were members of
the IMU. In 1931/32 Hermann Weyl was president (Vorsitzender) of the DMV.
So it appears that Weyl had his hand in the affair when it came to proposing Emmy
Noether as a speaker from Germany. The proposal had to be accepted by the executive
committee. The nomination of Emmy Noether was accepted and this shows the great
respect and admiration which Emmy Noether enjoyed on the international scale.
Emmy Noethers Zrich lecture can be considered as the high point in her mathematical career.
37 1928/29 Emmy Noether had been in Moscow as a visiting professor, on the invitation of Alexandrov whom
she knew from Gttingen.
38 The photo is also contained in the Oberwolfach photo collection online. Perhaps the photo was taken by
Natascha Artin, the wife of Emil Artin.
149
4.6 Gttingen exodus: 1933

The year 1933 brought about the almost complete destruction of the unique mathematical scene in Gttingen. In consequence of the antisemitic political line of the
Nazi government many scientists of Jewish origin had to leave the university, as well
as those who were known to be critical towards the new government. The Gttingen
situation in 1933 has often been described, and so we can refer to the literature, e.g.,
[Sch87], [Seg03].
Emmy Noether was of Jewish origin and so she too was a victim of the new
government policy. On May 5, 1933 Emmy Noether obtained the message that she
was put temporarily on leave from lecturing at the university. When Hasse heard
this, he wrote a letter to her; we do not know the text of his letter but from her reply
we may conclude that he asked whether he could be of help. Emmy Noether replied
on May 10, 1933:
Lieber Herr Hasse! Vielen herzlichen Dank fr Ihren guten freundschaftlichen Brief! Die Sache ist aber doch fr mich sehr viel weniger
schlimm als fr sehr viele andere: rein uerlich habe ich ein kleines
Vermgen (ich hatte ja nie Pensionsberechtigung), soda ich erst einmal in Ruhe abwarten kann; im Augenblick, bis zur denitiven Entscheidung oder etwas lnger, geht auch das Gehalt noch weiter. Dann wird
wohl jetzt auch einiges von der Fakultt versucht, die Beurlaubung nicht
denitiv zu machen; der Erfolg ist natrlich im Moment recht fraglich.
Schlielich sagte Weyl mir, da er schon vor ein paar Wochen, wo
alles noch schwebte, nach Princeton geschrieben habe wo er immer
noch Beziehungen hat. Die haben zwar wegen der Dollarkrise jetzt auch
keine Entschlukraft; aber Weyl meinte doch da mit der Zeit sich etwas ergeben knne, zumal Veblen im vorigen Jahr viel daran lag, mich
mit Flexner, dem Organisator des neuen Instituts, bekannt zu machen.
Vielleicht kommt einmal eine sich eventuell wiederholende Gastvorlesung heraus, und im brigen wieder Deutschland, das wre mir natrlich das liebste. Und vielleicht kann ich Ihnen sogar auch einmal so ein
Jahr Flexner-Institut verschaffen das ist zwar Zukunftsphantasie wir
sprachen doch im Winter davon
Dear Mr. Hasse! Thanks very much for your good, friendly letter! But
for myself, the situation is much less dire than for many others: in fact I
have a small fortune (after all I was never entitled to pension) and hence
for the time being I can quietly wait and see. Also, the salary payments
continue until the final decision or even somewhat longer. Moreover
the Faculty tries to avert my suspension to become final; at the moment,
however, there is little hope for success. Finally, Weyl told me that
some weeks ago already when things were still open, he had written to
150
Princeton where he still has contacts. At the moment, however, because

of the dollar crisis they dont have much freedom there for their decisions;
but Weyl believes that in the course of time there may arise something,
in particular since Veblen last year was eager to introduce me to Flexner,
the organizer of the new Institute. Perhaps there will emerge a visiting
professorship which may be iterated, and in the meantime Germany
again, this would be the best solution for me, naturally. And maybe I
will be able to manage for you too a year in the Flexner Institute but
this is my fantasy for the future we have talked last winter about this
The first impression while reading this letter is her complete selflessness, which
is well-known from other reports on her life and which is manifest here again. She
does not complain about her own situation but only points out that for other people
things may be worse. Reading further, we see that the Faculty in Gttingen tries to
keep her; this shows that she was respected there as a scientist and teacher although
she still did not have a tenured position. Hermann Weyl was a full professor and
hence a member of the Faculty committee; we can surely assume that he was one of
the driving forces in trying to save Emmy Noether for a position in Gttingen. In
fact, in his memorial speech [Wey35] Weyl said:
It was attempted, of course, to influence the Ministerium and other responsible and irresponsible bodies so that her position might be saved.
I suppose there could hardly have been in any other case such a pile of
enthusiastic testimonials filed with the Ministerium as was sent in on
her behalf. At that time we really fought; there was still hope left that
the worst could be warded off
And finally, in the above Noether letter we read that, independent of these attempts,
Weyl had written to Princeton on her behalf. We do not know whom in Princeton
Weyl had adressed. Since Noether mentions in her letter Veblen and Flexner, it seems
probable that Weyl had written to one or both of them. Abraham Flexner was the
spiritual founder and the first director of the newly-founded Institute for Advanced
Study in Princeton. Oswald Veblen was the first permanent mathematics professor of
the IAS. Certain indications suggest that Weyl had written to Lefschetz too; see next
section. Solomon Lefschetz had the position of full professor at Princeton University.
One year earlier, in the late summer of 1932, Weyl had rejected an offer to join the
IAS as a permanent member. But now, since the political situation had deteriorated,
he inquired whether it was possible to reverse his decision. (It was.) From Noethers
letter we infer that Weyl did not only write on his own behalf but also on Noethers.
This fact alone demonstrates the very high esteem in which he held Noether as a
mathematician and as a personality. 39
39 In the course of time, Weyl used his influence in American academic circles to help many other mathematicians as well.
151
But of course, the best solution would be that Noether could stay in Gttingen.
This was what Weyl wished to achieve foremost, as we cited above. (It was in vain.)
Weyl reports in [Wey35]:
I have a particularly vivid recollection of these months. Emmy Noether,
her courage, her frankness, her unconcern about her own fate, her conciliatory spirit, were, in the middle of all the hatred and meanness, despair
and sorrow surrounding us, a moral solace.
That stormy time of struggle in the summer of 1933 in Gttingen drew them closer
together. This is also evident from the words Weyl used two years later in his speech
at her funeral: 40
You did not believe in evil, indeed it never occurred to you that it could
play a role in the affairs of man. This was never brought home to me more
clearly than in the last summer we spent together in Gttingen, the stormy
summer of 1933. In the midst of the terrible struggle, destruction and
upheaval that was going on around us in all factions, in a sea of hate and
violence, of fear and desperation and dejection you went your own way,
pondering the challenges of mathematics with the same industriousness
as before. When you were not allowed to use the institutes lecture halls
you gathered your students in your own home. Even those in their brown
shirts were welcome; never for a second did you doubt their integrity.
Without regard for your own fate, openhearted and without fear, always
conciliatory, you went your own way. Many of us believed that an
enmity had been unleashed in which there could be no pardon; but you
remained untouched by it all.
Parallel to the attempts of the Faculty to keep Noether in Gttingen, Hasse took
the initiative and collected testimonials 41 which would put into evidence that Emmy
Noether was a scientist of first rank and hence it would be advantageous for the
scientific environment of Gttingen if she did not leave. Hasse collected 14 such
testimonials. Together they were sent to the Kurator of the university who was to
forward them to the Ministerium in Berlin. Recently we have found the text of those
testimonials which are kept in the Prussian State archives in Berlin; we plan to publish
them separately. The names of the authors are:
H. Bohr, Kopenhagen
Ph. Furtwngler, Wien
G. H. Hardy, Cambridge
40 See
Section 4.9.2.
German word is Gutachten. I am not sure whether the translation into testimonial is adequate. My
dictionary offers also opinion or expertise or letter of recommendation. I have chosen testimonial since
Weyl uses this terminology.
41 The
152
H. Hasse, Marburg
O. Perron, Mnchen
T. Rella, Wien
J. A. Schouten, Delft
B. Segre, Bologna
K. Shoda, Osaka
C. Siegel, Frankfurt
A. Speiser, Zrich
T. Takagi, Tokyo
B. L. van der Waerden, Leipzig
H. Weyl, Gttingen
We see that also Hermann Weyl wrote a testimonial. We have included it in the
appendix, translated into English; see Section 4.9.1. Note that Weyl compared Emmy
Noether to Lise Meitner, the nuclear physicist. In the present situation this comparison
may have been done since Meitner, also of Jewish origin, was allowed to stay in Berlin
continuing her research with Otto Hahn in their common laboratory. After all, the
initiatives of Hasse and of Weyl were to obtain a similar status for Emmy Noether in
Gttingen.
As is well-known, this was in vain. Perhaps those testimonials were never read after the Kurator of Gttingen University wrote to the Ministerium that Emmy Noethers
political opinions were based on Marxism. 42
Let us close this section with some lines from a letter of Weyl to Heinrich Brandt
in Halle. The letter is dated December 15, 1933; at that time Weyl and Noether
were already in the USA. Brandt was known to be quite sceptical towards abstract
methods in mathematics; he did not even like Artins beautiful presentation of his
own (Brandts) discovery, namely that the ideals and ideal classes of maximal orders
in a simple algebra over a number field form a groupoid under multiplication. 43 (The
notion of groupoid is Brandts invention.) Weyls letter is a reply to one from
Brandt which, however, is not known to us. Apparently Brandt had uttered some
words against Noethers abstracting method, and Weyl replied explaining his own
viewpoint: 44
So wenig mir persnlich die abstrakte Algebra liegt, so schtze
ich doch ihre Leistungen und ihre Bedeutung offenbar wesentlich hher
42 See [Tol90]. By the way, there was another such initiative started, namely in favor of Courant who also
had been beurlaubt from Gttingen University. That was signed by 28 scientists including Hermann Weyl and
Helmut Hasse. Again this was not successful, although this time the Kurators statement was not as negative as
in Noethers case. (We have got this information from Constance Reids book on Courant [Rei76].)
43Artins paper is [Art28b].
44 I would like to thank M. Gbel for sending me copies of this letter from the Brandt archive in Halle. The
letter is published in [Jen86], together with other letters Brandt-Weyl and Brandt-Noether.
153
ein, als Sie das tun. Es imponiert mir gerade an Emmy Noether, da ihre
Probleme immer konkreter und tiefer geworden sind.
Personally, the abstract algebra doesnt particularly suit me but apparently I do estimate its achievements and importance much higher
than you are doing. I am particularly impressed that Emmy Noethers
problems have become more and more concrete and deep.
Weyl continues as follows. It is not known whether Brandt had written some comments on Noethers Jewish origin and connected this with her abstract way of thinking, or perhaps Weyls letter was triggered by the general situation in Germany and
especially in Gttingen:
Warum soll ihr, der Hebrerin, nicht zustehen, was in den Hnden des
Ariers Dedekind zu groen Ergebnissen gefhrt hat? Ich berlasse
es gern Herrn Spengler und Bieberbach, die mathematische Denkweise
nach Vlkern und Rassen zu zerteilen. Da Gttingen den Anspruch
verloren hat, mathematischer Vorort zu sein, gebe ich Ihnen gerne zu
was ist denn berhaupt von Gttingen brig geblieben? Ich hoffe und
wnsche, da es eine seiner alten Tradition wrdige Fortsetzung durch
neue Mnner nden mge; aber ich bin froh, da ich es nicht mehr gegen
einen Strom von Unsinn und Fanatismus zu sttzen brauche!
Why should she, as of Hebrew descent, not be entitled to do what had
led to such great results in the hands of Dedekind, the Arian ? I leave
it to Mr. Spengler and Mr. Bieberbach to divide the mathematical way
of thinking according to nations and races. I concede that Gttingen has
lost its role as a high-ranking mathematical place what is actually left
of Gttingen? I hope and wish that Gttingen would find a continuation
by new men, worthy of its long tradition; but I am glad that I do not have
to support it against a torrent of nonsense and fanaticism.
4.7 Bryn Mawr: 19331935

As we have seen in the foregoing section, Weyl had written to Princeton on behalf
of Emmy Noether, and this was in March or April 1933 already. Since he was going
to join the Institute for Advanced Study in Princeton, one would assume that he had
recommended accepting Emmy Noether as a visiting scientist of the Institute. We
know that some people at the Institute were interested in getting Noether to Princeton,
for at the International Zrich Congress Oswald Veblen had been eager to introduce
Emmy Noether to the Institutes director, Abraham Flexner. (See Noethers letter to
Hasse, cited in the foregoing section.)
154
But as it turned out, Emmy Noether did not receive an invitation as a visitor to the
Institute. We do not know the reason for this; perhaps the impending dollar crisis,
mentioned in Noethers letter to Hasse, forced the Institute to reduce its available
funds. Or, may there have been other reasons as well? On the other hand, from the
documents which we found in the archive of Bryn Mawr College it can be seen that
the Institute for Advanced Study contributed a substantial amount towards the salary
of Emmy Noether in Bryn Mawr.
We do not know who was the first to suggest that Bryn Mawr College could be
a suitable place for Emmy Noether. Some evidence points to the conclusion that it
was Solomon Lefschetz. In fact, we have found a letter, dated June 12, 1933 already,
adressed to the Emergency Committee in Aid of Displaced German Scholars, where
he discusses future aspects for Emmy Noether and proposes Bryn Mawr. 45 Lefschetz
had visited Gttingen two years ago and so he knew Emmy Noether personally.
Lefschetz letter is quite remarkable since, firstly, he clearly expresses that Emmy
Noether, in his opinion, was a leading figure in contemporary mathematics; secondly
we see that he had taken already practical steps to provide Bryn Mawr with at least
part of the necessary financial means in order to offer Emmy Noether a stipend. Let
us cite the relevant portions of that letter:
Dear Dr. Duggan: I am endeavoring to make connections with some
wealthy people in Pittsburgh, one of them a former Bryn Mawr student,
with a view of raising a fund to provide a research associateship at Bryn
Mawr for Miss Emmy Noether. As you may know, she is one of the most
distinguished victims of the Hitler cold pogrom and she is victimized
doubly; first for racial reasons and second, owing to her sex. It occured
to me that it would be a fine thing to have her attached to Bryn Mawr
in a position which would compete with no one and would be created
ad hoc; the most distinguished feminine mathematician connected with
the most distinguished feminine university. I have communicated with
Mrs. Wheeler, the Head of the Department at Bryn Mawr, and she is not
only sympathetic but thoroughly enthusiastic for this plan.
So far as I know, your organization is the only one which is endeavoring
to do anything systematic to relieve the situation of the stranded German
scientists. As I do not think random efforts are advisable, I wish first
of all to inform you of my plan. Moreover, if I were to succeed only
partially, would it be possible to get any aid from your organization? I
would greatly appreciate your informing me on this point at your earliest
convenience.
In the preliminary communication with my intended victims I mentioned
the following proposal: to contribute enough annually to provide Miss
45 We
have found this letter in the archives of the New York Public Library.
155
Noether with a very modest salary, say $ 2000, and a retiring allowance
of $ 1200.
Yours very sincerely, S. Lefschetz.
Already one month later the committee granted the sum of $ 2000 to Bryn Mawr for
Emmy Noether.
There arises the question from whom Lefschetz had got the information, at that
early moment already, that Emmy Noether had been suspended. 46 We are inclined
to believe that it was Hermann Weyl. I do not know whether the correspondence of
Lefschetz of those years has been preserved in some archive, and where. Perhaps it
will be possible to find those letters and check.
Emmy Noether arrived in Bryn Mawr in early November 1933. Her first letter
from Bryn Mawr to Hasse is dated March 6, 1934. She reported, among other things,
that since February she gave a lecture once a week at the Institute for Advanced Study
in Princeton. In this lecture she had started with representation modules and groups
with operators. She mentions that Weyl too is lecturing on representation theory, and
that he will switch to continuous groups later. It appears that the Gttingen situation of
1926/27 was repeating. And we imagine Hermann Weyl and Emmy Noether walking
after her lectures around the Campus of Princeton University 47 instead of Gttingens
narrow streets, vividly discussing new aspects of representation theory.
In the book [Rei76] on Courant we read:
Weyl sent happy letters from Princeton. In Fine Hall, where Flexners
group was temporarily housed, German was spoken as much as English.
He frequently saw Emmy Noether
Perhaps in the Courant legacy we can find more about Weyl and Noether in Princeton,
but we have not been able yet to check those sources.
Every week Emmy Noether visited the Brauers in Princeton; Richard Brauer was
assistant to Weyl in that year and perhaps sometimes Hermann Weyl also joined their
company. The name of Hermann Weyl appears several times in her letters to Hasse
from Bryn Mawr. In November 1934 she reports that she had studied Weyls recent
publication on Riemann matrices in the Annals of Mathematics.
Emmy Noether died on April 14, 1935. One day later Hermann Weyl cabled to
Hasse:
hasse mathematical institute gottingen emmy noether died yesterday
by sudden collapse after successful operation of tumor 48 few days
ago burial wednesday bryn mawr weyl
46 Emmy Noether had been beurlaubt, i.e., temporarily suspended from her duties, in May 1933. Observing
that mail from Europe to USA used about 23 weeks at that time, we conclude that Lefschetz must have started
working on his Noether-Bryn Mawr idea immediately after receiving the news about her suspension. Noether
was finally dismissed from university on September 9, 1933.
47 The Institutes Fuld Hall had not yet been built and the School of Mathematics of the Institute was temporarily
housed in Fine Hall on the University Campus.
156
At the burial ceremony on Wednesday Weyl spoke on behalf of her German friends
and colleagues. We have included an English translation of this moving text in the
appendix; see Section 4.9.2. One week later he delivered his memorial lecture in
the large auditorium of Bryn Mawr College. That text is published and well known
[Wey35].
4.8 The WeylEinstein letter to the NYT

On Sunday May 5, 1935 the New York Times published a Letter to the Editor,
signed by Albert Einstein and headed by the following title:
Professor Einstein Writes in Appreciation of a Fellow-Mathematician.
We have included the text of this letter in our appendix; see Section 4.9.3.
Reading this letter one is struck by the almost poetic style which elevates the text
to one of the pearls in the literature on mathematics. The text is often cited, the last
citation which I found is in the Mitteilungen der DMV 2007, where Jochen Brning
tries to connect mathematics with poetry [Br07]. But because of this character of
style it has been doubted whether the text really was composed by Einstein himself.
If not then this would not have been the first and not the last incident where Einstein
had put his name under a text which was not conceived by himself provided that
in his opinion the subject was worth-while to support. Since Weyls poetic style was
known it was not considered impossible that the text was composed by Hermann
Weyl.
Some time ago I have come across a letter signed by Dr. Ruth Stauffer-McKee. I
include a copy of that letter in the appendix; see Section 4.9.4. In particular I refer
to the last paragraph of the letter. Based on the information provided by Stauffer I
came to the conclusion that, indeed, the text was essentially written by Weyl. I have
expressed this opinion in my talk in Bielefeld and also in a Letter to the Editor of
the Mitteilungen der Deutschen Mathematiker-Vereinigung [Roq07a].
However, recently I have been informed that Einsteins draft of this letter in his
own handwriting has been found by Siegmund-Schultze 49 in the Einstein archive in
Jerusalem. The article appeared in the Mitteilungen der DMV, see [SS07]. This then
settles the question of authorship in favor of Einstein. But what had induced Ruth
Stauffer to claim that Weyl had inspired Einsteins letter?
In order to understand Stauffers letter let us explain its background.
48 President
Park of Bryn Mawr had sent a detailed report, dated May 16, 1935, to Otto Nther in Mannheim,
a cousin of Emmy Noether. A copy of that letter is preserved. There it is stated that according to the medical
diagnosis of the doctors who operated her, Emmy Noether suffered from a pelvis tumor.
49 I would like to thank R. Siegmund-Schultze for a number of interesting comments and corrections to this
article.
157
In 1972 there appeared a paper on Emmy Noether in the American Mathematical Monthly, authored by Clark Kimberling [Kim72]. Among other information the
paper contains the text of Einsteins letter to the New York Times. Kimberling had
obtained the text from an article in the Bryn Mawr Alumnae Bulletin where it had
been reprinted in 1935. Together with that text, we find in [Kim72] the following:
A note in the files of the Bryn Mawr Alumnae Bulletin reads, The above
was inspired, if not written, by Dr. Hermann Weyl, eminent German
mathematician. Mr. Einstein had never met Miss Noether.
(Here, by above was meant the text of the Einstein letter to the New York Times.)
While the first sentence of that note can be considered as an affirmation of the
guess that Weyl had conceived the text of Einsteins letter, the second sentence is
hard to believe. Emmy Noether often visited the Institute for Advanced Study in
Princeton, the same place where Einstein was, and it seems improbable that they did
not meet there. After all, Einstein was already in May 1918 well informed about
Noethers achievements, when he wrote to Hilbert praising her work [Noe18]. And
in December that year, after receiving the printed version of this work, he wrote
to Felix Klein and recommended her Habilitation. In the 1920s, Einstein had a
correspondence with Emmy Noether who acted as referee for papers which were
submitted to the Mathematische Annalen. It is hard to believe that in Princeton he
would have avoided meeting Emmy Noether, whom he esteemed so highly. Moreover,
we have already mentioned in Section 4.4 that Einstein probably had met Noether in
1915 in Gttingen. Also, on the DMV-meetings 1909 in Salzburg and 1913 in Wien
both Einstein and Emmy Noether presented talks and there was ample opportunity
for them to meet.
Thus it seemed that the note which Kimberling mentioned had been written by
someone who was not well informed about the situation in the early thirties. Actually,
that note was not printed in the Bryn Mawr Alumnae Bulletin but it was added later
by typewriter, maybe only on the copy which was sent to Kimberling. It is not known
who had been the author of that note.
In the same volume of the American Mathematical Monthly where his article
[Kim72] had appeared, Kimberling published an Addendum saying that Einsteins
former secretary, Miss Dukas, had objected to the statement that the letter written by
Einstein was inspired, if not written by Dr. Hermann Weyl. She insisted that the
letter was written by Einstein himself at the request of Weyl.
This, however, induced Ruth Stauffer to write the above mentioned letter to the
editor of the American Mathematical Monthly, which we are citing in Section 4.9.4.
Ruth Stauffer had been a Ph.D. student of Emmy Noether in Bryn Mawr and in her
letter she recalls vividly the mathematical atmosphere in Princeton at that time.
On this evidence I was led to believe that the statement of Einsteins secretary
Dukas may be due to a mix-up on her part. For, only shortly before Noethers death
Einstein had written another letter in which he recommends that Emmy Noethers
158
situation in Bryn Mawr College should be improved and put on a more solid base.
At that time President Park of Bryn Mawr had tried to obtain testimonies on Emmy
Noether, which could be used in order to get funds for a more permanent position.50
Einsteins testimony is dated January 8, 1935 and is written in German; we have
found it in the archives of the Institute for Advanced Study in Princeton. Its full text
reads:
Frulein Dr. Emmy Noether besitzt unzweifelhaft erhebliches schpferisches Talent, was jeweilen von nicht sehr vielen Mathematikern einer
Generation gesagt werden kann. Ihr die Fortsetzung der wissenschaftlichen Arbeit zu ermglichen, bedeutet nach meiner Ansicht die Erfllung einer Ehrenpicht und wirkliche Frderung wissenschaftlicher
Forschung.
Without doubt Miss Dr. Emmy Noether commands significant and creative talent; this cannot be said of many mathematicians of one generation. In my opinion it is an obligation of honor to provide her with the
means to continue her scientific work, and indeed this will be a proper
support of scientific research.
It is apparent that the style of this is quite different from the style of the letter to
the New York Times.
Although we now know that Miss Dukas was right and Einstein had composed his
NYT-letter with his own hand, there remains the question as to the basis of Stauffers
contentions.
Stauffer was a young student and what she reports is partly based on what she
heard from Mrs. Wheeler. But the latter, who was head of the mathematics department
of Bryn Mawr College at the time, had studied in Gttingen with Hilbert in the same
years as Hermann Weyl had; so they were old acquaintances and it seems probable
that Weyl himself had told her the story as it had happened. Thus it may well have
been that first Weyl had sent his obituary on Emmy Noether to the New York Times,
and that this was returned with the suggestion that Einstein should write an obituary
as Ruth Stauffer narrates. And then Einstein wrote his letter at the request of Weyl,
as Miss Dukas has claimed. Whether there was any cooperation between Einstein
and Weyl while drafting the letter is not known. But we can safely assume that
both had talked if not about the text of the letter but certainly about Emmy Noethers
personality, her work and her influence on mathematics at large. In this way Stauffers
claim may be justified that Weyl had inspired Einstein in writing his letter.
Remark. It has been pointed out to me by several people that the very last sentence
in the English version of Einsteins letter deviates in its meaning from the original
50 This was successful, but Emmy Noether died before she got to know about it. Other testimonials, by
Solomon Lefschetz, Norbert Wiener and George D. Birkhoff are published in Kimberlings article [Kim81].
159
German text wheras otherwise the translation seems to be excellent. 51 In the English
version it is said that Noethers last years in Bryn Mawr were made the happiest and
perhaps most fruitful years of her entire career, but the German text does not refer
to her entire career and only pointed out that death came to her mitten in froher und
fruchtbarer Arbeit. I do not know who had translated the German text into English.
There is a letter of Abraham Flexner, the director of the Institute for Advanced Study
in Princeton, addressed to Einstein and dated April 30, 1935, in which Flexner thanks
Einstein for the beautiful tribute to Miss Noether and continues: I shall translate
it into English and send it to the New York Times, through which it will reach, I
think, many of those who should know of her career. But it does not seem justified,
I believe, to conclude that Flexner personally did the translation job. He was quite
busy with all kinds of responsibilities and certainly he had contacts to experts who
would have been willing and competent to do it. 52
Final remark. Weyls solidarity with Emmy Noether extended to her brother and
family. Emmys brother Fritz had emigrated to Russia where he got a position at the
university in Tomsk. In 1937 he was arrested and sentenced to 25 years in prison
because of alleged espionage for Germany. In the Einstein archive in Jerusalem we
have found a letter, dated April 1938 and signed by Einstein, addressed to the Russian
minister of foreign affairs Litvinov. In this letter Einstein appeals to the minister in
favor of Fritz Noether, whom he (Einstein) is sure to be innocent. In the Einstein
archive, right after this letter, is preserved a curriculum vitae of Fritz Noether in
Weyls handwriting. Thus again it appears that Weyl has inspired Einstein to write
such a letter. 53
Among Weyls papers I found a number of letters from 1938 and the following
years, which show that he cared for the two sons of Fritz Noether, Hermann and
Gottfried, who had to leave the Soviet Union after their father had been sentenced.
Weyl saw to it that they obtained immigrant visa to the United States, and that they
got sufficient means to finance their university education. Both became respected
members of the scientific community.
4.9 Appendix: documents

4.9.1 Weyls testimony
The following text 54 is from the testimonial, signed by Hermann Weyl on July 12, 1933
and sent by Hasse to the Ministerium in Berlin together with 13 other testimonials.
51 The
German text is published in my Letter to the Editor [Roq07a].
52 Siegmund-Schultze [SS07] advocates reasons to assume that indeed, Flexner himself did the translation job.
53 The appeal of Einstein was in vain. In 1941, when German troops were approaching the town of Orjol where
Fritz was kept in prison, he was sentenced to death and immediately executed. See, e.g., [Sch91].
54 Translated from German by Ian Beaumont.
160
We have found these testimonials in the Prussian state archive Berlin.

Emmy Noether has attained a prominent position in current mathematical research
by virtue of her unusual deep-rooted prolific power, and of the central importance of
the problems she is working on together with their interrelationships. Her research and
the promising nature of the material she teaches enabled her in Gttingen to attract the
largest group of students. When I compare her with the two woman mathematicians
whose names have gone down in history, Sophie Germain and Sonja Kowalewska, she
towers over them due to the originality and intensity of her scientific achievements.
The name Emmy Noether is as important and respected in the field of mathematics
as Lise Meitner is in physics.
She represents above all Abstract Algebra. The word abstract in this context in
no way implies that this branch of mathematics is of no practical use. The prevailing
tendency is to solve problems using suitable visualizations, i.e. appropriate formation
of concepts, rather than blind calculations. Frulein Noether is in this respect the
legitimate successor of the great German number theorist R. Dedekind. In addition,
Quantum Theory has made Abstract Algebra the area of mathematics most closely
related to physics.
In this field, in which mathematics is currently experiencing its most active
progress, Emmy Noether is the recognised leader, both nationally and internationally.
Hermann Weyl
4.9.2 Weyls funeral speech
Hermann Weyl had read his funeral speech on April 18, 1935 at the funeral ceremony
in the house of President Park in Bryn Mawr. For the text see Chapter 3, Section 3.2,
page 123.
4.9.3 Letter to the New York Times
The following text was published on Sunday, May 5, 1935 by the New York Times, with
the heading: Professor EinsteinWrites in Appreciation of a Fellow-Mathematician.
To the Editor of The New York Times:
The efforts of most human-beings are consumed in the struggle for their daily
bread, but most of those who are, either through fortune or some special gift, relieved
of this struggle are largely absorbed in further improving their worldly lot. Beneath
the effort directed toward the accumulation of worldly goods lies all too frequently
the illusion that this is the most substantial and desirable end to be achieved; but there
is, fortunately, a minority composed of those who recognize early in their lives that
161
the most beautiful and satisfying experiences open to humankind are not derived from
the outside, but are bound up with the development of the individuals own feeling,
thinking and acting. The genuine artists, investigators and thinkers have always been
persons of this kind. However inconspicuously the life of these individuals runs its
course, none the less the fruits of their endeavors are the most valuable contributions
which one generation can make to its successors.
Within the past few days a distinguished mathematician, Professor Emmy Noether,
formerly connected with the University of Gttingen and for the past two years at Bryn
Mawr College, died in her fifty-third year. In the judgment of the most competent living mathematicians, Frulein Noether was the most significant creative mathematical
genius thus far produced since the higher education of women began. In the realm of
algebra, in which the most gifted mathematicians have been busy for centuries, she
discovered methods which have proved of enormous importance in the development
of the present-day younger generation of mathematicians. Pure mathematics is, in
its way, the poetry of logical ideas. One seeks the most general ideas of operation
which will bring together in simple, logical and unified form the largest possible
circle of formal relationships. In this effort toward logical beauty spiritual formulas
are discovered necessary for the deeper penetration into the laws of nature.
Born in a Jewish family distinguished for the love of learning, Emmy Noether,
who, in spite of the efforts of the great Gttingen mathematician, Hilbert, never
reached the academic standing due her in her own country, none the less surrounded
herself with a group of students and investigators at Gttingen, who have already
become distinguished as teachers and investigators. Her unselfish, significant work
over a period of many years was rewarded by the new rulers of Germany with a
dismissal, which cost her the means of maintaining her simple life and the opportunity
to carry on her mathematical studies. Farsighted friends of science in this country were
fortunately able to make such arrangements at Bryn Mawr College and at Princeton
that she found in America up to the day of her death not only colleagues who esteemed
her friendship but grateful pupils whose enthusiasm made her last years the happiest
and perhaps the most fruitful of her entire career.
Albert Einstein.
Princeton University, May 1, 1935.
4.9.4 Letter of Dr. Stauffer-McKee
The following letter was sent by Dr. Ruth Stauffer-McKee on October 17, 1972 to the
editor of the American Mathematical Monthly, Professor H. Flanders. A carbon copy
had been sent to Professor Kimberling. I am indebted to Clark Kimberling for giving
me access to his private archive. 55
55 The Kimberling archive on Emmy Noether is now contained in the Handschriftenabteilung of Gttingen
University.
162
Dear Mr. Flanders,

After reading the Addendum to Emmy Noether in the August September issue
of the American Mathematical Monthly, I was much disturbed by the apparent lack
of information concerning the thirties at Princeton! Rechecking the reference to the
original article which appeared in February 1972 I was even more disturbed to note
that the quote was attributed to a note in the files of Bryn Mawr Alumnae Bulletin.
A telephone conversation and a careful check by the Staff of the Bulletin assured me
that there was nothing in the files of the Bulletin to even imply that Mr. Einstein had
never met Miss Noether.
In respect to the thirties at Princeton, I should like to note that there was an air
of continued excitement at the Institute for Advanced Study. Solomon Lefschetz, a
guiding spirit who worked diligently to help the displayed mathematicians, Hermann
Weyl, a leading mathematician of that time who had learned to know Miss Noether
in Gttingen, and John von Neumann, then considered a brilliant young genius, were
all at the Institute when Einstein arrived in December of 1933. Mrs. Wheeler, of
Bryn Mawr, often told of the welcoming party which she and Miss Noether attended.
Mrs. Wheeler usually drove Miss Noether to Princeton for lectures and included
Miss Noethers students in the parties. We listened to talks by these men who were the
leaders in new exciting theories. It was a friendly group and after the talks everyone
gathered for more talk and coffee in a long pleasant common room. There is no doubt
that Einstein and Noether were acquainted. I saw them in the same group!
As regards the quote in the addendum to Emmy Noether inspired, if not
written by Dr. Hermann Weyl is certainly true. The writing of the obituary was a
very natural occurrence. Hermann Weyl was considered by the mathematicians as the
mathematical leader of the time and at the peak of his productivity and he had probably
the greatest knowledge and understanding of her work. Einstein had begun to slow
down and von Neumann was relatively young and still growing. It was, therefore,
obvious to all the mathematicians that Weyl should write the obituary which he did.
He, furthermore, sent it to the New York Times, the New York Times asked who is
Weyl? Have Einstein write something, he is the mathematician recognized by the
world. This is how Einsteins article appeared. It was most certainly inspired by
Weyls draft. These facts were told to me at the time by Mrs. Wheeler who was
indignant that the New York Times had not recognized the mathematical stature of
Hermann Weyl.
Very truly yours,
Ruth Stauffer McKee
Senior Mathematician
Chapter 5
Emmy Noether: The testimonials
Translation of the article:

Emmy Noether: Die Gutachten. Einige neue Noetheriana
Mitteilungen der Deutschen Mathematiker Vereinigung, Vol. 16 (2008), 118123.
5.1
5.2
5.3
5.4
5.5
Preface
The testimonials
The accompanying letters
The petition of students
The American testimonials
163
165
168
170
172
5.1 Preface
In the course of our work to edit the HasseNoether correspondence 1 we have found
some documents which perhaps deserve independent interest for the history of mathematics and of mathematicians. Some of these Noetheriana have been already
published. 2 The present article concerns the testimonials for Emmy Noether from the
years 1933/34. Recently I have discovered the originals of these testimonials.
Remark. I have chosen the English word testimonial as the translation of the German Gutachten but I am not quite sure whether this translation reflects the meaning
of the German in an unambiguous manner. In any case, the testimonials which are
discussed here are statements of mathematicians, written in the years 1933/34, assessing the importance of Emmy Noethers work and the impact of her ideas. Perhaps the
word opinion or assessment would be a little more to the point. I have decided in
favor of testimonial since Hermann Weyl had used this expression in his Memorial
Address for Emmy Noether which he delivered in Bryn Mawr on April 26, 1935.
1 This
2 See
has appeared in the meantime: [LR06].

[Roq07b, Roq08]. (Chapters 3 and 4.)
164
5 Emmy Noether: The testimonials
There he said: I suppose there could hardly have been in any other case such a pile
of enthusiastic testimonials filed with the Ministerium as was sent in her behalf.3
The existence of those testimonials was known not only from Weyls memorial
address but also from the letters of Emmy Noether to Hasse in the summer of 1933.4
Let us recall: On January 30, 1933 Hitler became Reichskanzler of Germany and
already on April 7 the so-called Beamtengesetz was decreed which, among other
consequences, prohibited people of Jewish descent or those with oppositional political
views to work at state universities. Accordingly Emmy Noether was temporarily
dismissed. She had to complete, within two weeks, a questionnaire inquiring about
her Aryan descent as well as her membership in political parties in previous years.
There was the danger that she would be permanently dismissed. But some people still
hoped that it would be possible to obtain an exceptional permit which would allow
her to stay in Gttingen and continue her work.
In this situation Helmut Hasse, at that time in Marburg, tried to solicit testimonials
from prominent mathematicians all over the world, to show that Emmy Noether was
one of the leading mathematicians and hence it would be a great loss for mathematics
in Gttingen and in Germany if she were forced to emigrate.
Until recently the text of those testimonials was unkown, and not even the names
of their senders were known. Gnther Frei reports in [Fre77] that Hasse had collected
such testimonials and sent them to the Ministerium. In the report of Schappacher
[Sch87] on the Gttingen Mathematical Institute it is only said that there were 14 such
testimonials. We read the same in the extensive and informative writings of Cordula
Tollmien on Emmy Noether [Tol90]. At first my own search for these testimonials
was also unsuccessful. The archives of Gttingen University did not contain more
information about this issue.
But in June 2006 I visited (for another project) the Geheimes Staatsarchiv
Preuischer Kulturbesitz in Berlin. On this occasion I inquired again about the possible whereabouts of the personnel records on Emmy Noether. It turned out that just
recently some material had been received including a dossier on Emmy Noether.5
Apparently these documents had been moved during the war from Berlin to some
other place and only now had it been possible to integrate it again into that archive.
This file contained, among other documents, the missing 14 testimonials. More
precisely there are only 13 testimonials since one of them is signed jointly by two
mathematicians, Harald Bohr and G. H. Hardy. I do not know whether Hasse had
written to even more mathematicians who had not answered or at least not answered
in time. In any case I find the list of names of mathematicians remarkable. It testifies
to the high esteem which Emmy Noether met within the mathematical community
of the time. I believe this will be of general interest, which is why I am writing this
article.
3 Weyls
text has been reprinted in Auguste Dicks biography of Emmy Noether [Dic70].
[LR06], in particular the letters between May 10 and Sep 13, 1933.
5 GStA PK, I. HA Rep. 76 Kultusministerium, Nr. 10081.
4 See
165
Emmy Noether 1933 at the train station in Gttingen on her departure
However, because of lacking space I cannot show here the full text of those testimonial letters. Copies of the originals (mostly in German) can be seen and downloaded
from my homepage: www.roquette.uni-hd.de.
5.2 The testimonials

The testimonials were written as letters addressed to Hasse but obviously they were
formulated in such a way that they could be presented to the official people in the
government. Here is the list of mathematicians who had sent testimonials:
1.
2.
3.
4.
5.
H. Bohr (Kopenhagen) and G. H. Hardy (Cambridge):

Ph. Furtwngler (Wien)
H. Hasse (Marburg)
O. Perron (Mnchen)
T. Rella (Wien)
166
6.
7.
8.
9.
10.
11.
12.
13.
J. A. Schouten (Delft)
B. Segre (Bologna)
K. Shoda (Osaka)
C. L. Siegel (Frankfurt)
A. Speiser (Zrich)
T. Takagi (Tokio)
B. L. van der Waerden (Leipzig)
H. Weyl (Gttingen)
In all these testimonials it is stressed that, as Perron, for instance, writes:

Emmy Noether gehrt zu den fhrenden Persnlichkeiten in der modernen Mathematik
Emmy Noether belongs to the leading figures in contemporary mathematics
It is due to her and her school that (after Bohr):
die Algebra eine neue Blte erlebt hat und in der ganzen mathematischen
Welt an fhrender Stelle steht und ihren Bereich weit ausdehnen konnte
in geometrische und andere Forschungsgebiete hinein
Algebra has seen a new blossoming and ranks in the mathematical world
as a leading force; it has widely expanded its domain into geometric and
other fields of research
Again and again her impact upon the younger generation of mathematicians is stressed;
in this respect Weyl writes:
Sie wusste in Gttingen durch ihre Forschung und durch die Suggestivitt ihrer Lehre den grssten Kreis von Schlern um sich zu versammeln
She was able to collect in Gttingen the largest circle of students, through
her research and the persuavive power of her teaching
and Furtwngler:
Sie hat auch durch ihre selbstlose und nur von idealen Zielen geleitete Lehrttigkeit einen grossen Kreis von Schlern herangebildet, die
sich heute bereits einen geachteten Namen in der mathematischen Welt
gemacht haben
Through her generous teaching which was exclusively motivated by
idealistic goals, she has educated a large circle of students who have
already achieved a respected standing in the mathematical world
167
After closer examination of the testimonials it strikes me that the illegitimacy of

the removal of Jewish scientists from German universities is never mentioned, neither
from the judicial and political point of view nor under scientific and humanitarian
aspects.An exception is the testimonial of Schouten who indeed expresses this clearly:
Es wre ein grosser Skandal wenn eine solche Kraft wegen Rassenvorurteil abgebaut wrde. Man macht sich in Deutschland anscheinend
keine Vorstellung davon wie emprt das deutschfreundliche Ausland
ber solche Sachen ist
It would be a great scandal if such a person is removed due to racial
prejudice. Apparently in Germany one has no idea how outraged people in foreign countries are about such things people who otherwise
consider themselves to be pro-German
The same point is taken up by Takagi:
Es wre schade, wenn ihr die venia legendi an der Universitt Gttingen beraubt wird, zumal wegen eines Umstandes, woran sie keine
Schuld trgt !
It would be a shame if she would be deprived of her venia legendi at
the University of Gttingen, in particular if this were because of a fact
which she cannot be blamed for !
Otherwise, while reading the testimonials one may get the impression that the
situation was not particularly upsetting. In essence these testimonials could have been
written in the same way if, for instance, Noether would have received a tempting offer
from a university in a foreign country and therefore one should try to make her stay
in Gttingen. Perhaps the referees had underestimated the seriousness of the political
situation. Or, maybe this academic reservation can be explained by the fact that one
wished to achieve something and therefore tried to avoid affronting the ruling political
forces. Certainly, this last motive is the reason why in some testimonials it is stressed
that Emmy Noethers work is important for mathematical science in Germany.
Clearly, all referees agreed that mathematics is international and cannot be reduced
to national borders. As an example we cite van der Waerden:
Aus aller Welt kamen vor ihrer Beurlaubung die Algebraiker nach Gttingen um ihre Methoden zu lernen, ihren Rat zu holen, unter ihrer
Fhrung zu arbeiten
Before Emmy Noether had been dismissed, algebraists from all over the
world came to Gttingen in order to learn her methods, to get counsel
from her and to work under her guidance
and Carl Ludwig Siegel:
168
Insbesondere hat die sog. Theorie der hyperkomplexen Systeme durch

Verffentlichungen und Vorlesungen von Frl. Noether so grosse Frderung erfahren, dass die daran anschliessenden Probleme jetzt bei den
Algebraikern der ganzen Welt im Vordergrund des Interesses stehen
In particular the so-called theory of hypercomplex systems has greatly
advanced through the publications and lectures of Miss Noether. The
subsequent problems are now of primary interest among algebraists
around the world
This statement by Siegel is remarkable indeed since he had never hidden his disgust
of the abstract methods which Noether had advocated. He regarded them as a sign of
deterioration of mathematics. But he seems to exempt the theory of hypercomplex
systems, i.e., algebras. Already on December 9, 1931 he had sent to Hasse his
compliments for the theorem on cyclic algebras over number fields which had been
jointly discovered by Hasse, Richard Brauer and Emmy Noether. At that time he had
written: The pessimistic outlook which I generally have towards mathematics has
been shaken once again . Siegel
5.3 The accompanying letters

Besides the testimonials I found also the accompanying letters from Hasse to the
curator of Gttingen University, who was named Valentiner. (His job in this case was
to forward the testimonials to the Ministerium and to add his own opinion.) These are
the following documents: 6
1. Hasse to curator on June 3, 1933. Hasse announced that after consulting
Professor Neugebauer (Gttingen) he had started to solicit testimonials for
Emmy Noether. Apparently this announcement was sent in order to forestall
final decisions about the dismissal of Emmy Noether before the testimonials
had been received. Hasse expressed his hope that Emmy Noether could remain
in Germany
durch eine Lehrbeauftragung mit Spezialvorlesungen
through a teaching appointment for specialized courses
Indeed, this covered precisely her past activities in Gttingen. Emmy Noether
had approved this wording in a letter to Hasse of June 26, 1933.
2. Hasse to curator on July 31, 1933. This letter accompanied the testimonials
which Hasse had sent the same day. Hasse emphasizes that the teaching of
Emmy Noether would be directed only
6 These letters are already cited by Tollmien since copies had also been found in the archive of Gttingen
University.
169
to a relatively small group of students most of them aspiring to an

academic position.
Perhaps this was said to point out that her presence would not come into
conflict with the student mass organizations which were indoctrinated by the
Nazi propaganda.
3. Curator to Ministry on August 7, 1933. The curator states that he is informed
about the scientific standing of Ms. Emmy Noether but adds:
To my knowledge, politically Ms. Noether had stood on Marxist
ground since the revolution of 1918 until today regardless of my
high esteem of the scientific standing of Ms. Noether I feel unable
to advocate her case.
According to Tollmien the allegation of Marxist ideology covers, following
the terminology used at the time, all leftist non-communist parties including
the Social Democrats.
Actually, neither the endorsing testimonials of the mathematicians nor the critical
statement of the curator had any visible effect. The whole action of dismissing scientists of Jewish descent, or of politically critical opinion, was done for ideological
and political reasons. The people in power who were responsible for this action had
foreseen and clearly accepted the negative consequences for science, as well as the
decline of reputation of Germany in the international world and they were determined to complete their action ruthlessly. However, some people who were directly or
indirectly affected seem to have believed that reason would prevail in due course and
objective arguments would again be taken into account. We know that, for instance
Hasse wrote to his friend Davenport on May 1933:
hope that reason will come back in due course.
and Emmy Noether to Hasse also in May 1933:
wird ja wohl ziemlich bald eine Beruhigung kommen!
probably there will pretty soon come a time of slow-down!
As a side remark let us mention that Hasse had also written to Tornier, asking to
advocate Noethers case at the government agencies. Hasse knew that Tornier had
outed himself as a member of the Nazi party after the Nazis had come to power, and
that Tornier had gained a certain influence on people in power. In any case there exists
a letter of Tornier to Hasse, dated September 23, 1933, in which he warns Hasse for
Gods sake 7 not to send the testimonials to the Ministerium, for:
7 um
Gottes willen
170
Durch solche Methoden erreichen Sie im gnstigsten Falle, dass Sie in

allen realen Dingen der Regierung gegenber absolut einusslos werden, im ungnstigsten Falle knnen die Folgen fr Sie selbst sehr, sehr
schlimm sein.
With such methods you can at best achieve that you will have no influence
whatever on the government, but in the worst case the consequences for
yourself could be very, very severe.
I do not know whether and how Hasse has reacted to this threat. Anyhow, the
testimonials had already been sent to Berlin in August, and Emmy Noether had
already lost her venia legendi 8 on September 13, 1933. Nevertheless Emmy Noether
and apparently Hasse too had some hope. This may have been the reason why Hasse
had approached Tornier. Hermann Weyl reports in his obituary on Emmy Noether:
It was attempted, of course, to influence the Ministerium and other responsible and irresponsible but powerful bodies so that her position
might be saved. At that time we really fought
It appears that Tornier belonged to the irresponsible bodies which Weyl mentioned.
In his letter Tornier had posed as a kind of a high-ranking officer in the Nazi hierarchy.9
5.4 The petition of students

There was also a petition of 12 students supporting Emmy Noether, but of course this
had no effect either. We know of this petition from Emmy Noethers letter to Hasse
of June 26, 1933. There she writes:
Wichmann hat dem Kurator noch gerade, als dieser Pngsten nach Berlin
fuhr, die Studentenunterschriften es waren wesentlich die Algebraiker
gegeben
Wichmann has given the student signatures to the curator, just before the
latter went to Berlin they were essentially from the algebraists
Wichmann had been one of her last Ph.D. students. 10 The text of this petition
has been published by Tollmien already, but the names of the undersigned were not
8 The
permission to deliver lectures at the university.
9 Reichsobmann fr Mathematik im Fhrerrat der Reichsfachschaft N-S-Hochschullehrer und Wissenschaftler.

10 The thesis of Wichmann was published 1936 in the Monatshefte fr Mathematik und Physik. It contains,
among other results, a simple proof of the functional equation of Heys zeta function of a simple algebra, by means
of reduction to the functional equation of the center (which is assumed to be a number field). It is remarkable
that in this publication, in the year 1936, Wichmann dared to thank Emmy Noether for her counsel. But in the
functional equation there remained an undetermined sign which Wichmann could not handle. Recently Falko
Lorenz has discussed this beautiful but almost forgotten result and presented it in a lucid manner. At the same
time Lorenz showed that the sign in question can be determined using Hasses sum formula for the invariants of
the algebra [Lor08b].
171
known yet. The original which I found carries the signatures. All but one of the names
could be identified. These names are as follows:
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
E. Bannow
E. Knauf
Tsen
W. Vorbeck
G. Dechamps
W. Wichmann
H. Davenport (Cambridge, Engl.)
H. Ulm
L. Schwarz
Walter Brandt (?)
D. Derry
Wei-Liang Chow
Biographical information on these mathematicians can be found in the standard

literature (some of them online) and we do not have to present them here. 11 Information concerning the unidentified no. 10 is welcomed.
Tollmien has already observed that in the petition it was attempted to help Emmy
Noether by pointing out that the undersigned students are of Aryan descent. In view
of the above mentioned names of the undersigned this appears as a tactical move since
one wanted to achieve something. It is apparent that the undersigned did not agree
with the division of mathematics into an Aryan and a non-Aryan part. But when
they say that Emmy Noether
niemals politischen Einuss auf ihre Schler ausgebt hat,
has never exerted political influence on her students,
then this cannot be regarded as a tactical move, considering all that is known about
her. Here I agree with Tollmien.
By the way: For the summer term 1933 Emmy Noether had announced a lecture
with the title: Hypercomplex methods in number theory. Most probably this meant
the proof of Artins Reciprocity Law which Hasse had given a year earlier by means
of the theory of algebras and dedicated it to her on her 50th birthday. (Hasses paper
had just appeared, in spring 1933.) This lecture could not be given anymore in the
summer term 1933 but Emmy Noether then invited the interested students into her
apartment to discuss the proof there. The above list of signatures shows that she had
students of high level. They wrote that her aim was not to teach single theorems and
results but vision and understanding of the theory. And they continue that
11 In this connection see also the list of doctoral students in mathematics 19071945 at German universities,
prepared by Renate Tobies [Tob06].
172
ihre Vorlesungen alle ihre Schler mit Begeisterung und Leidenschaft

fr die Mathematik erfllt haben
her lectures have filled all her students with enthusiasm and excitement
for mathematics
This text, I admit, has impressed me very much. This is unfeigned. We can safely
assume that also her students from previous years had felt this way. For any teacher,
can there a more convincing praise by ones students? All the testimonials on Noether
have the same assertion but here we have the statement directly from her students.
In this connection we would like to mention the text of the commemoration speech
by Professor Marguerite Lehr which had been delivered on April 18, 1935, four days
after Emmy Noethers death, at the Chapel of Bryn Mawr. This speech contained a
description of the lively contact of Emmy Noether with the students of the college. 12
5.5 The American testimonials

As said above, the testimonials collected by Hasse are impressive documents for the
esteem which Emmy Noether enjoyed among the mathematicians worldwide. But
the picture would not be complete if the testimonials written in the USA were not
mentioned. These were aimed at securing a position for Emmy Noether at Bryn Mawr
where she would be able to continue her work. I am showing here three testimonials
which I have found in American archives. Although these testimonials have been cited
already in Kimberlings article [Kim81] I believe that they are of interest also in the
present context. The authors of these testimonials are:
1. S. Lefschetz (Princeton)
2. G. D. Birkhoff (Harvard)
3. N. Wiener (M.I.T.)
It is quite remarkable that the importance of Emmy Noether had been clearly
recognized in the USA at that time already at least among the leading mathematicians. The authors vehemently support her case. Of course, those testimonials were
not meant to impress representatives of a repressive government as it was the case in
Germany, and hence one could write quite clearly. But even when considering this
fact the writings seem to me quite remarkable.
Thus Lefschetz writes: 13
12 The
text is published in [Roq07b]. (See Chapter 3.)

Committee in Aid of Displaced Foreign Scholars records. Manuscripts and Archives Division,
The New York Public Library. Astor, Lenox and Tilden Foundations.
13 Emergency
173
she is the holder of a front rank seat in every sense of the word.
As the leader of the modern algebra school, she developed in recent
Germany the only school worthy of note in the sense, not only of isolated
work, but of very distinguished group scientific work. In fact, it is no
exaggeration to say that without exception all the better young German
mathematicians are her pupils.
Birkhoff comments: 14
She is generally regarded as one of the leaders in modern Algebraic Theory. Within the last ten or fifteen years she and her students in Germany
have led the way much of the time.
And Wiener: 15
Miss Noether is a great personality; the greatest woman mathematician
who has ever lived; and the greatest woman scientist of any sort now
living Leaving all questions of sex aside, she is one of the ten or twelve
leading mathematicians of the present generation in the entire world and
has founded what is certain to be the most important close-knit group of
mathematicians in Germany the Modern School of Algebraists Of
all the cases of German refugees, whether in this country or elsewhere,
that of Miss Noether is without doubt the first to be considered.
These testimonials were written not in the year 1933 but one year later, in 1934/35
when it was time to renew the temporary position of Emmy Noether at Bryn Mawr
(or even convert it into a permanent position if possible). This was not quite trivial
since as far as undergraduate work is concerned, she will be probably of no use
at Bryn Mawr (as G. D. Birkhoff observed in one his letters). The excellence of
Emmy Noether rested in research and in high level teaching. Nevertheless, everybody
including President Park of Bryn Mawr promoted vehemently the continuation of
Emmy Noethers appointment there. Eventually it turned out that Emmy Noethers
appointment was secured for some time (not without some financial help from the
Institute for Advanced Study in Princeton) but unfortunately she did not live to hear
about this.
The efforts to get Emmy Noether to Bryn Mawr had originally started in the
summer of 1933. Reading the correspondence of people and organizations of the time
we are impressed by the quick and unprejudiced way in which help was organized for
the academics who had to emigrate from Germany. In those early letters we also find
assessments in various forms of Emmy Noethers work, but at that time there were
no testimonials in the formal sense since decisions had to be obtained quite quickly.
14Archives
of Bryn Mawr College.

Wiener Papers (MC 22). Institute Archives and Special Collections, MIT Libraries, Cambridge,
Massachusetts.
15 Norbert
174
The first document in support of Emmy Noether which we have found is a letter by
S. Lefschetz. (We found it in the NewYork Public Library.) The letter is dated June 12,
1933 and is addressed to an organization in New York (c/o Dr. S. P. Duggan) which,
Lefschetz says, is endeavoring something systematic to relieve the situation of the
stranded German scientists. At that time Lefschetz had already talked to Professor
Anna Wheeler, Head of the Mathematics Department in Bryn Mawr, with the aim of
securing a place of work for Emmy Noether, for instance a research associateship.
Lefschetz writes that Wheeler is not only sympathetic but thoroughly enthusiastic
to this plan. In addition, Lefschetz writes, he had contacted some wealthy people in
Pittsburgh with a view towards raising a fund for that purpose.
It is remarkable how quickly Lefschetz had taken the intitiative in support of
Emmy Noether at a time when people in Europe like, for instance Hasse and Weyl
(and also Emmy herself), still had hopes to be able to keep her in Gttingen. It seems
that Lefschetz and his colleagues in USA saw the situation more realistically than the
mathematicians in Germany.
Remark. At this point we would like to draw the readers attention to the archive
of Clark Kimberling in which he has collected documents of his research on Emmy
Noether. This archive is now kept by the Handschriftenabteilung of the University
Library at Gttingen and is open for historical research. The mentioned letter of
Lefschetz is available there too.
Kapitel 6
Abraham Robinson and his infinitesimals
Revised version of:

Numbers and Models, Standard and Nonstandard.
Mathematische Semesterberichte 57 (2010), 185199.
The following is a somewhat extended manuscript for a talk at the Algebra Days, May 2008,
in Antalya. I talked about my personal recollections of Abraham Robinson.
6.1
6.2
6.3
6.4
6.5
How I met Abraham Robinson

What is nonstandard analysis?
Robinsons visits
Nonstandard algebra
Nonstandard arithmetic
175
177
183
185
187
6.1 How I met Abraham Robinson

It was in the early months of 1963. I was visiting the California Institute of Technology on my sabbatical. Somehow during this visit I learned that one year ago
Wim Luxemburg had given a lecture on A. Robinsons theory of infinitesimals and
infinitely large numbers. Luxemburg was on leave but I got hold of his Lecture Notes
[Lux62]. Although the topic was somewhat distant from my own work I got interested
and, after thorough reading I wished to meet the person who had been able to put
Leibniz infinitesimals on a solid base and build the modern analysis upon it.
At that time Abraham Robinson was at the nearby University of California at Los
Angeles, and I managed to meet him there. I remember an instructive discussion
about his theory which opened my eyes for the wide range of possible applications;
he also showed me his motivations and main ideas about it.
Perhaps I am allowed to insert some personal words explaining why I had been so
excited about the new theory of infintesimals. This goes back to my school days in
Knigsberg, when I was 16. At that time the school syllabus required that we were to
176
be instructed in Calculus or, as it was called in German, in Diffentialrechnung. Our

Math teacher at that time was an elderly lady who had been retired already but was
reactivated again for school work in order to fill the vacancy of our regular teacher;
the latter had been drafted to the army. (It was war time: 1944.) I still remember the
sight of her standing in front of the blackboard where she had drawn a wonderfully
y
smooth parabola, inserting a secant and telling us that x
is its slope, until finally
dy
she convinced us that the slope of the tangent is dx where dx is infinitesimally small
and dy accordingly.
This, I admit, impressed me deeply. Until then our school Math had consisted
largely of Euclidean geometry, with so many problems of constructing triangles from
some given data. This was o.k. but in the long run that stuff did not strike me as
to be more than boring exercises. But now, with those infinitesimals, Math seemed
to have more interesting things in stock than I had met so far. And I decided that I
would study Mathematics if I survived the dangers of war which we knew we would
be exposed to very soon. After all, I wanted to find out more about these wonderfully
strange infinitesimals.
Abraham Robinson
Well, I survived. And I managed to enter University and start with Mathematics.
The first lecture I attended to was Calculus, with Professor Otto Haupt in Erlangen.
There we were told to my disappointment that my Math teacher had not been up to
date after all. We were warned to beware of infinitesimals since they do not exist, and
dy
then this does
in any case they lead to contradictions. Instead, although one writes dx
not really mean a quotient of two entities, but it should be interpreted as a symbolic
y
notation only, namely the limit of the quotients x
.
I survived this disappointment too. Later I learned that dy and dx can be interpreted, not as infinitesimals but as some entities of an abstract construction called
177
dy
would
differential module, and if that module is one-dimensional then the quotient dx
make sense and yield what we had learned anyhow. Certainly, this sounded nice but
in fact it was only an abstract frame ignoring the natural idea of infinitesimally small
numbers.
So when I learned about Robinsons infinitesimals, my early school day experiences came to my mind again and I wondered whether that lady teacher had not been
so wrong after all.
The discussion with Abraham Robinson kindled my interest and I wished to know
more about it. Some time later there arose the opportunity to invite him to visit
us in Germany where he gave lectures on his ideas, first in Tbingen and later in
Heidelberg, after I had moved there.
Before continuing with this let me briefly explain what I am talking about, i.e.,
Robinsons theory of nonstandard analysis.
6.2 What is nonstandard analysis?

6.2.1 A preliminary Axiom
Consider the hierarchy of numbers which we present to our students in their first year:
N Z Q R:
Everything starts with the natural numbers N which, due to Kronecker, are created by God (or whatever is considered to be equivalent to Him). The rest is constructed by mankind, i.e., by the minds of mathematicians. In each step, the structure
in question is enlarged such as to admit greater flexibility with respect to some operations defined in the structure. In Z the operation of subtraction is defined such that
Z becomes an additive group; in fact Z is a commutative ring without zero divisors.
In Q the operation of division is defined such that Q becomes a field. Finally, in R
every Cauchy sequence is convergent, such that R becomes a complete ordered field.
In each step we tell our students that the respective enlargement exists and we explain
how to construct it.
In order to develop what nowadays is called analysis the construction usually
stops with the real field R; this is considered to be adequate and quite sufficient as a
basis for (real) analysis. But it had not always been considered to be that way. Since
Leibniz had used the natural idea of infinitesimals to build a systematic theory with
it, many generations of mathematicians (including my lady teacher) had been taught
in the Leibniz way. Prominent people like Euler, the Bernoullis, Lagrange and even
Cauchy (to name only a few) did not hesitate to use them.
Leibniz idea was to work in a further enlargement:
R R
178
such that the following Axiom is satisfied. In order to explain the main idea I will
first state the Axiom in a preliminary form which, however, will not yet be sufficient.
Later I will give the final, more general form.
Axiom (preliminary form).
(1) R is an ordered field extension of R.
(2) R contains infinitely large elements.
An element ! 2 R is called infinitely large if j!j > n for all n 2 N. Part (2) says
that the ordering of R does not satisfy the axiom of Archimedes.
Fields with the properties (1) and (2) were known for some time but the attempts
to build analysis on this basis were not quite satisfactory. Among all such fields one
has to select those which in addition have more sophisticated properties. But for the
moment let us stay with the Axiom in this preliminary form and see what we can do
with it.
The elements of R are called standard real numbers, while the elements of R
not in R are nonstandard. This terminology is taken from model theory but I find
it not very suggestive in the present context. Sometimes the elements of R are
called hyperreal numbers. Perhaps someone some time will find a more intuitive
terminology.
The finite elements in R are those which are not infinitely large, i.e. which
satisfy Archimedes axiom: jj < n for some n 2 N (depending on ). These finite
elements form a subring E R. It contains all infinitesimal elements " which are
defined by the property that j"j < n1 for all n 2 N. It follows from the definition that
the set of infinitesimals is an ideal I E. We have:
! infinitely large
! 1 infinitesimal 0:
It is well known that this property characterizes E as a valuation ring in the sense of
Krull.
Theorem. The finite elements E form a valuation ring of R with the
infinitesimals I as its maximal ideal. The residue class field E=I D R.
Two finite elements ; are said to be infinitely close to each other if
is infinitesimal, i.e., if they belong to the same residue class modulo the ideal I of
infinitesimals. This is written as
:
The residue class of 2 E is called the monad of ; this terminology has been
introduced by Robinson in reference to Leibniz theory of monads. Every monad
contains exactly one standard number a 2 R; this is called the standard part of ,
and denoted by st./. There results the standard part map
st W E ! R
179
which in fact is nothing else than the residue class map of E modulo its maximal
ideal I .
In this situation let us consider the example of a parabola
y D x2
which, as I have narrated above, had been used by my school teacher to introduce
us to analysis. Suppose x is a standard number. If we add to x some infinitesimal
dx 0 then the ordinate of the corresponding point on the parabola will be
y C dy D .x C dx/2 D x 2 C 2xdx C .dx/2
which differs from y by
dy D 2xdx C .dx/2 D .2x C dx/dx
so that the slope of the corresponding secant is
dy
D 2x C dx 2x
dx
since dx 0 is infinitesimal. Hence:
If we step down from the hyperreal world into the real world again by using the
standard part operator, then the secant of two infinitely close
becomes the
dy points

tangent, and the slope of this tangent is the standard part: st dx D 2x.
I believe that such kind of argument had been used by my school teacher as
narrated above. As we see, this is completely legitimate.
It is apparent that in the same way one can differentiate any power x n instead of
2
x , and also polynomials and quotients of polynomials, i.e., rational functions, with
coefficients in R. All the well known algebraic rules for derivations can be obtained
in this way. However, analysis does not deal with rational functions only. What can
be done to include more functions?
6.2.2 The Axiom in its final form
As described by the preliminary Axiom, R is an ordered field. This can be expressed
by saying that R is a model of the theory of ordered fields. The theory of ordered
fields contains in its vocabulary the function symbols C for addition, and for
multiplication, as well as the relation symbol < for the ordering. The axioms of
ordered fields are formulated in this language. If we add to the vocabulary constants
for all real numbers r 2 R and to the theory all statements which are true in R then
the models of this theory are precisely the ordered field extensions of R.
If we wish to talk about functions and relations which are not expressible in this
language, then we have to use a language with a more extended vocabulary. In order
180
not to miss anything which may be of interest let us include into our language symbols
for all relations in R. 1 The theory of R consists of all statements in this language
which hold in R. Thus, if we generalize the first part of the above Axiom as

R is a model of the theory of R,
then this will allow us to talk in R about every function and relation which is defined
in R.
In order to generalize the second part of the Axiom we have to refer not only
to the relation < of the ordering, but to every relation of similar kind. More
precisely: Consider a 2-place relation '.x; y/ defined in R. Such a relation is said to
be concurrent if, given finitely many elements a1 ; : : : ; an 2 R in the left domain of
', there exists b 2 R in its right domain such that '.ai ; b/ holds for i D 1; : : : ; n. 2
Such element b may be called a bound for a1 ; : : : ; an with respect to the relation '.
Axiom (final form).
(1) R is a model of the theory of R.
(2) Every concurrent relation ' over R admits a universal bound 2 R,
i.e., such that '.a; / holds simultaneously for all a 2 R which are
contained in the left domain of '.
It is clear that this form of the Axiom is a generalization of its preliminary form,
and a far reaching generalization at that. It was Abraham Robinson who had observed
that Leibniz, when he worked with infinitesimals, seemed tacitly to use something
which is equivalent to that Axiom.
Of course, the essential point is that indeed there exists a structure R satisfying
this Axiom. This is guaranteed by general results of model theory. The most popular
construction is by means of ultrapowers.
There is some ambiguity which has to be cleared. The Axiom refers to the theory
of R, and this refers to a given language as described above, its vocabulary including
symbols for all relations over R. On first sight one would think of relations (and
functions) between individuals, i.e., elements of R. This would lead to the first
order language (Lower Predicate Calculus), where quantification is allowed over
individuals only. But in many mathematical investigations it is necessary to enlarge
the language such as to contain also symbols for sets of functions, relations between
sets of functions etc., and quantification should be allowed over entities of any given
type. For instance, if we wish to state the induction axiom for the set N of natural
numbers, we may say that
1 Functions
can be viewed as 2-place relations and thus are included. Subsets may be defined as the range of
their characteristic functions and hence are included too.
2 The relation ' may not be defined on the whole of R. The left domain of ' consists of those a 2 R for
which there exists a b 2 R such that '.a; b/ holds. The right domain is defined similarly. For the ordering
relation < the left and the right domain coincides with R.
181
Every non-empty subset of N contains a smallest element

and this statement contains a quantifier for subsets.
In order to include such statements too we have to work with the Higher Order
Language containing symbols also for higher entities, i.e., relations between sets,
functions of relations between sets, etc. Quantification is allowed over entities of any
given type. In other words:
We interpret the above Axiom as referring to the full structure over R and accordingly work with the corresponding higher order language.
This implies, among other things, that in R we have to distinguish between
internal and external entities. Here we do not wish to go into details but refer, e.g.,
to the beautiful introduction which Abraham Robinson himself has given in his book
Non-standard Analysis [Rob66]. See also the first section in [RR75].
Robinson introduced the terminology enlargement for a structure satisfying the
Axiom. As said above, such an enlargement can be obtained by ultrapower construction. It is not unique. In the following we choose one such enlargement and regard
it as a fixed mathematical universe during our discussion.
6.2.3 Some excercises
Having learned all this from Abraham Robinson, my immediate reaction was what
probably every newcomer would have done: I wished to put this method of reasoning
to a test in simple exemplary situations. I do not have time here for a long discussion
although much could be said to convince the reader of the enormous potential of the
new way of reasoning which Robinsons theory of enlargements offers to us. Here
let me be content with a few examples.
Let f be a standard function and consider an element x 2 R in its domain of
definition. According to the part (1) of the Axiom, f extends uniquely to a function
on R.
Continuity. f is continuous in x if and only if
x 0 x H) f .x 0 / f .x/:
(6.1)
Of course, it is assumed that x 0 is contained in the domain of f , so that f .x 0 / is

defined. If the domain of the original f is open then f .x 0 / is defined for every x 0 in
the monad of x.
The above statement can be used as definition of continuity of a function. Note
that the usual quantifier prefix 8 " 9 : : : is missing.
I have chosen this example because I found precisely this definition in an old
textbook. This was the German Kiepert, Differential- und Integralrechnung, of which
the first edition had appeared in 1863. It had been very popular, and it got at least
12 editions, the 12th appearing in 1912 [Kie12]. The text there reads as follows (in
English translation):
182
If some function is given by y D f .x/ then, in general, infinitely small

changes of x will give rise to infinitely small changes of y. For all values
of x where this is the case, the function is called continuous.
We see that this is precisely the definition (6.1).
Derivative. Let dx be an infinitesimal. Define dy by y C dy D f .x C dx/. Then
the derivative f 0 .x/ 2 R is defined by
f 0 .x/
dy
:
dx
(6.2)
dy
should
More precisely: it is required that this is a valid definition, i.e., the quotient dx
be finite and its monad should be independent of the choice of the infinitesimal dx. If
this is the case then f is called differentiable at x and f 0 .x/ is defined as the standard
dy
part of dx
.
I have chosen this example since it is the definition presented by my school
teacher mentioned above. It is well possible that she had been trained using Kieperts
textbook.
Integration. Suppose the function f .x/ is defined in the closed interval a; b with
a; b 2 R. Let n be a natural number and divide a; b into n subintervals xi1 ; xi of
equal length. We take n infinitely large; then the length dx D ba
of each subinterval
n
is infinitesimal. Now the integral is defined by:
Z
b
a
f .x/dx
n
X
f .xi /dx:
(6.3)
iD1
More precisely: It is required that this is a valid definition, i.e., the sum on the right
hand side should be a finite element in R and its monad should be independent of
the choice of the infinite number n. If this is the case then f is called (Riemann)
Rb
integrable over a; b and the integral a f .x/dx is defined as the standard part of
that sum.
Maybe some explanation about infinite natural numbers is in order. R is an
enlargement of R, and therefore every subset of R has an interpretation in R. So
does N. This new subset of R is denoted by N. (In fact, N is an enlargement of
N.) Using part (2) of the Axiom, it follows that there exists n 2 N which is larger
than every number in N, i.e., n is infinite. The sum on the right hand
P side of (6.3)
is to be interpreted as follows: For every finite n 2 N the sum sn D niD1 f .xi /dx
has finitely many terms, and so sn is well defined in R. The function n 7! sn from N
to R has an interpretation in the enlargement, i.e., it extends to a function from N to

R. Thus sn is defined for every n 2 N. Note that sn for infinite n is not an infinite
series in the usual sense. It is to be regarded as the nonstandard interpretation of a
sum whose number of terms is a natural number.
183
The definition (6.3) of the integral explains Leibniz idea that the integral is essentially
R a sum (up to infinitesimals). This idea had led him to introduce the integral
sign as a variant of the letter S which he used for sums (instead of which is used
today).
I have been inspired to choose example (6.3) because of its relation to Archimedes
method of measuring the area of a plane region. This method consists of cutting the
area into parallel strips of, say, length ` and infinitesimal breadth "; then ` " is the
(infinitesimal) area of the strip and the sum of all those areas will give the area of the
whole region up to infinitesimals. The Leibniz formula (6.3) does precisely this in
the case of a positive function, when the region to be measured is that between the
function graph and the x-axis.
That Archimedes had indeed used this idea (contrary to what is commonly attributed to him) is well documented by the Archimedes Codex which has been recently discovered and deciphered; see the report [NN07] about what is called the
world greatest palimpsest.
6.3 Robinsons visits

6.3.1 Tbingen
As said at the beginning I had met Abraham Robinson in Los Angeles in California
during my sabbatical. In the summer term of 1963 I was back at my university
in Tbingen. There I started a workshop where together with some students and
colleagues, we read Robinsons papers and his book on model theory [Rob63] which
had just appeared. We tried to understand his ideas for nonstandard analysis and to
apply them to various situations. His book on nonstandard analysis [Rob66] had not
yet been written.
Some time later when I had heard that Robinson was in Germany, I was able to
meet him and suggested that he spend a month or so in Tbingen as visiting professor,
for a course on a topic from nonstandard analysis. He reacted favorably and so he
visited us in Tbingen in the summer of 1966. 3
I had advertised his lecture course to students and colleagues, and so he had a full
auditorium. The aim of the course, two hours weekly, was to cover the fundamentals
of model theory with particular emphasis on the application to analysis and algebra.
This job was not easy since the students (and most colleagues) did not have a formal
training in mathematical logic; so he had to start from scratch. He was not what may
be called a brilliant lecturer who would be able to rouse a large audience regardless
3 I am relying here on the extensive Robinson biography by Dauben [Dau95] where this year is recorded for
Robinsons Tbingen visit. Unfortunately I did not save our letters or other documents from that time and so I
have to look at Daubens book for help in the matter of dates.
184
of the content of his talk. His way was quiet, with great patience when questions
came up from the students, but strong when it came to convince the students about the
impact of nonstandard applications. And this kept the attention of the large audience
throughout his lecture.
In addition Robinson was available for discussion in our workshop. Just in time
his book on nonstandard analysis [Rob66] had appeared; he presented to us some of
the more sophisticated applications.
I recall my impression that his Tbingen visit could be considered as a success,
and from what is reported in Daubens biography it appears that Robinson thought
so too.
6.3.2 Heidelberg
Next year, 1967, I moved from Tbingen to the University of Heidelberg. The general
academic conditions in Heidelberg in those years were quite favorable. So it was
not difficult to convince the faculty and the rector (president) that the visit of a
distinguished scholar like Abraham Robinson would be of enormous importance
for the development of a strong mathematics group in Heidelberg. And so in the
following year, 1968, I was able to extend a cordial invitiation to Abraham Robinson
to visit us again, this time in Heidelberg. And he came, this time not from UCLA but
from Yale where he had moved in the meantime.
Again he delivered a course on model theory and applications. To a certain extent
this job was kind of a repetition of his Tbingen lecture; again he had a large audience.
But there was a difference. For in his seminar, on a smaller scale, he found an audience
which was highly motivated. On the one hand, there was a group of gifted students
and postdocs who had also switched from Tbigen to Heidelberg and who had already
attended Robinsons Tbingen lecture. On the other hand, in Heidelberg there had
been regular courses on Mathematical Logic (by Gert Mller who held a position as
associate professor), and so there had been opportunities for the students to acquire
knowledge in this field, in particular in model theory.
But the essential new feature of Robinsons Heidelberg visit was that he talked
not only on nonstandard analysis but also on nonstandard algebra and arithmetic; in
the seminar he was able to expound his ideas in more detail. This found a respondent
audience. His impact on the work of these young people in the seminar was remarkable. And so it came about that he more or less regularly visited us in Heidelberg
during the following years, continuing his seminar talks and working with those who
responded to his ideas.
In the next two sections I will give some kind of overview on the work resulting
of his influence on the Heidelberg group, which was apparent even after his untimely
death in April 1974.
Robinsons influence was also helpful in another project. In view of Robinsons
striking applications of model theory to mathematics proper, I became convinced that
185
a chair devoted to mathematical logic could be of help to mathematicians in their

daily work, in particular if the chair was occupied by someone from model theory.
Therefore I tried to obtain help from the university administration and the ministry of
education for establishing such a chair in the mathematics faculty. I had started this
project in Tbingen already but after I moved to Heidelberg this would have to be a
chair for the Heidelberg faculty. Indeed, after some time such a chair was installed
(in those times such thing was still possible). This was in 1973. It was clear to me
that Robinsons encouragement and judgement had been of great help in this matter.
When I asked him whether he would accept an offer to Heidelberg for this chair then
he did not say no but from the way he reacted it seemed to me that he really meant
no. After all, Heidelberg seemed to be no match for Yale at that time. In any case,
in a few months after that the problem was not existent any more. But it should be
remembered that this chair, which is still in existence, had been installed with the
strong help of Abraham Robinson.
During his repeated visits to Heidelberg we came to know Abraham Robinson
not only as a mathematician and scholar but also as a friend. He lived around the
corner from our house and on his way to town he regularly stepped in for a coffee
and conversation with us. (If I say we and us in this context then I include my
wife Erika.) He was a man with a wide horizon and far reaching interests. If he
talked about Leibniz then one could feel not only his knowledge about Leibniz life
and work but also his sympathy for that remarkable man. There was only one thing
about which he strongly disagreed with Leibniz, namely Leibniz insistence that our
world is the best of all possible worlds.
Abby liked to talk to people, and sometimes we had the impression that he knew
more about our neighbors than we did. He was keenly interested in the local history.
When we took him on tour to show him the country and its places then it often turned
out that he knew more about it than we did, and he gave us a lecture on the history of
those places.
In the course of those years there developed a friendship of rare quality. Abby
belongs to the few close friends whom I have met in my life. I have learned much from
him, not only in Mathematics but also in questions of attitude towards the problems
of life.
6.4 Nonstandard algebra

Looking at the Axiom in its final form (in section 6.2.2) it is apparent that this Axiom
has little to do with the special properties of the real number field R. It makes sense
for every mathematical structure. And so there is not only nonstandard analysis, but
nonstandard mathematics at large. Abraham Robinson was well aware of this; he
has applied his method, partly in collaboration with others, to various mathemati-
186
cal problems ranging from topology, Hilbert spaces, Lie groups, complex analysis,
differential algebra, quantum theory to mathematical economics.
There were also investigations in the direction of algebra and number theory. As
said above, Abraham Robinson reported on this in his Heidelberg seminar lectures.
One of his first topics was his nonstandard interpretation of Hilberts irreducibility
theorem (jointly with Gilmore in [GR55] ). This paper of Robinson has been said
to mark a watershed in the development of model theory (in the same line with
another paper of Robinsons, of the same year 1955, on Artins solutiom of Hilberts
17th problem [Rob55] ).
Hilbert had published his irreducibility theorem in 1892 [Hil92]. Suppose that
f .X; Y / is an irreducible polynomial in 2 (or more) variables then, Hilbert showed,
there are infinitely many specializations X 7! t such that f .t; Y / remains irreducible.
The coefficients of f are taken from the rational field Q and the specialized variable
t is also assumed to be in Q. Since then there had been numerous proofs of this
theorem, also over other base fields K, e.g., number fields. Hasse had the idea to
study arbitrary fields over which Hilberts irreducibility theorem may hold, and his
Ph.D. student Wolfgang Franz started the theory of such fields which today are called
Hilbertian fields [Fra31]. This was the point where Abraham Robinson stepped in.
He stated a nonstandard characterization of Hilbert fields.
As a follow-up of our discussions with Robinson we were able to amend his result
of [GR55] by presenting a new, metamathematical proof of Hilberts irreducibility
theorem in the number field and the function field cases. It turned out that Hilberts
irreducibility is, in fact, equivalent to the well known theorem of Bertini in algebraic
geometry [Roq75]. Further investigations by R. Weissauer showed that every field
with a set of valuations satisfying the product formula is Hilbertian. This covered all
classical fields which were known to be Hilbertian. Moreover, Weissauer found quite
a number of new and interesting Hilbertian fields, e.g., formal power series fields in
more than one variable [Wei82].
Weissauers paper is a good example of the usefulness of Robinsons enlargements.
On the one hand, it can be shown that any result which has been proved using the
notion and the properties of enlargements can also be obtained without this. On the
other hand, the use of enlargements provides the mathematician with new methods and
it opens up new analogies to other problems which sometimes help to understand the
situation. Abraham Robinson used to say that his method may reduce a dynamical to
a statical situation. For instance, an infinite sequence t1 , t2 , t3 ; : : : which preserves
the irreducibility of the polynomial f .X; Y / under the specialization X 7! ti leads
to a nonstandard t which renders f .t; Y / irreducible.
For another topic of algebra: Remember that group theory had been started by
Galois in order to study the roots of algebraic equations. Today the notion of Galois
group belongs to the basics of algebra. But there arose the need to study simultaneously infinitely many algebraic equations; this led Krull in 1928 to the discovery
of the topological structure of infinite Galois groups [Kru28], and this developed
187
into the theory of profinite groups. Robinson has pointed out that profinite Galois
groups can be naturally understood within the enlargement, connected to the finite
groups in the sense that their order is an infinitely large natural number n 2 N. The
corresponding profinite groups in the standard world are obtained from these nonstandard finite groups by a similar process as the derivative f 0 .x/ is obtained from
dy
in the manner as explained above. Hence
the nonstandard differential quotient dx
again:
If we step down from the nonstandard world into the standard world again, then
Krulls Galois theory of infinite algebraic extensions appears as an immediate consequence of the Galois-Steinitz theory for finite algebraic field extensions.
There arises the interesting question which fields K are uniquely determined (up
to elementary equivalence) be their full profinite Galois groups GK . See [Pop88],
[Koe95].
The description of the structure of GQ as profinite group is at the focus of current
arithmetical research.
6.5 Nonstandard arithmetic

Remember Hensels p-adic number fields which Hensel had conceived at around
1900 and which today have become standard tools in algebraic number theory and
beyond. In the course of time it became necessary to consider all p-adic completions
at once; this has led to the introduction of adeles and ideles in the sense of Chevalley
which play a fundamental role, e.g., in class field theory. Now, Abraham Robinson
has pointed out that his notion of enlargement comprises all those constructions at the
same time. His enlargements are indeed the most universal completions in as much
as every concurrent relation admits a bound. The classical notions of p-adics, adeles
and ideles, profinite groups etc. are obtained from his enlargement by a universal
transfer principle, similar to obtaining the derivative f 0 .x/ as the standard part of the
dy
differential quotient dx
as explained above.
In the ensuing discussions with Abraham Robinson we wished to test his method
in some more situations of fundamental importance. The SiegelMahler Theorem
seemed to us a good example to begin with. Finally in November 1973 he invited
me to Yale with the aim of discussing in more detail the possibility of a nonstandard
proof of this theorem.
Let : f .x; y/ D 0 be an irreducible curve defined over a number field K of
finite degree. If is of genus g > 0 then Siegels theorem says that admits only
finitely many points whose coordinates are integers in K. Mahler had generalized
this by proving that for any finite set S of primes of K there are only finitely many
points in whose coordinates are S-integers in K. The S-integers are those numbers
in K whose denominator consists of primes in S only.
188
Nonstandard methods seem to be useful to distinguish between finite and infinite.

We work in a fixed enlargement K of K, with the properties as statetd in the Axiom.
If would admit infinitely many S-integral points in K then it would also admit a
nonstandard S -integral point in K. Such a point .x; y/ is a generic point of over
K and hence F D K.x; y/ is the function field of over K. By construction F is
embedded into K:
F K:
Now, both these fields carry a natural arithmetic structure: F as an algebraic function
field over K and K as a nonstandard model of the number field K. What is the
relation between the arithmetic in F and in K? In our joint paper [RR75] we were
able to prove the following
Theorem 1. If F is of genus g > 0 then every functional prime divisor
P of F is induced by some nonstandard prime divisor p of K.
From here it is only a small step to deduce the validity of the SiegelMahler
Theorem. Abby agreed to work out the proof of the theorem for elliptic curves, and I
was to deal with curves of higher genus. Two weeks after I had left Yale he sent me
his manuscript for the elliptic part. But he could not see any more my part for higher
genus.
Actually, there is a famous conjecture of Mordell to the effect that a curve
of genus g > 1 has only finitely many K-rational points, even without specifying
that they are S-integers. This conjecture has been proved by Faltings in 1983. In
nonstandard terms it can be formulated as follows:
Theorem 2. A function field F jK of genus g > 1 cannot be embedded
into the enlargement K.
Clearly, this contains Theorem 1 in the case g > 1, which was my own contribution
in the joint work with Robinson. But in 1973 Mordells conjecture had not yet been
proved and hence, at that time, the proof of Theorem 1 was necessary also for the
case g > 1.
In 1973 I discussed with Robinson also a possible nonstandard proof of Mordells
conjecture. We planned first to develop the tools which we believed to be necessary
for this project. However, due to Robinsons sudden death our plan could not be
realized.
In later years Kani [Kan80b], [Kan80a], [Kan82] has studied systematically function fields which are embedded into the enlargement K. In my opinion, the tools
and the results which have been obtained in his work are well capable to give a nonstandard proof of Mordells conjecture, together with Roths theorem (which is also
used in the proofs of Theorems 1 and 2). But this has not yet been worked out. It
remains an open challenge.
Chapter 7
Cahit Arf and his invariant
Cahif Arf and his invariant

by Falko Lorenz and Peter Roquette
Mitteilungen der Mathematischen Gesellschaft in Hamburg Bd. XXX (2011), 87126.
7.1
7.2
7.3
7.4
7.5
7.6
7.7
7.8
7.9
7.10
7.11
7.12
7.13
Introduction
Arfs first letter
Some personal data
Quadratic spaces
Clifford algebras
Binary quadratic space
Higher dimensional quadratic spaces
Witt equivalence
Arfs Theorems
An assessment of Arfs paper
Perfect base fields
Epilog
Appendix: Proofs
190
191
194
196
198
199
203
206
207
212
214
215
216
Originally this manuscript was prepared for my talk at the workshop on Sequences, Curves
and Codes in Antalya, 2529 September 2009. Later I had given a talk with the same title on
October 5, 2009 at the conference on Positivity, Valuations, and Quadratic Forms in Konstanz.
In the discussion after the talk I learned that there may be an error in Arfs paper and
perhaps his main theorem has to be modified. I am indebted to Karim Johannes Becher for this
comment. Indeed, after another check I found the error in the proof of one of Arfs lemmas.
Accordingly the manuscript had to be corrected, taking care of the situation and clarifying the
scope of Arfs theorems after the correction. 1
The present article is the result of unifying and re-editing our two papers mentioned above.
For the convenience of the reader we have included an appendix containing very simple proofs
of Arfs main results in the corrected form. It seems to be of interest that all the facts which we
use in these proofs can be found in Arfs paper already.
1 We are indebted to Detlev Hoffmann for providing us with the relevant information. Also we would like to
thank K. Conrad and S. Garibaldi for their help in this matter. We are indebted to the referee for his informative
comments, who called our attention to the book [EKM08], in particular 39.
190
7 Cahit Arf and his invariant
7.1 Introduction
In January 2009 I received a letter from the organizers of the workshop in Antalya
on Sequences, Curves and Codes, with a friendly invitation to participate. The letter
was accompanied by a bank note of 10 Turkish Lira. Reading the letter I found out
that this was not meant as an advance honorarium for my talk, but it was to tell me
that the note carried the portrait of the Turkish mathematician Cahit Arf (19101997).
Besides the portrait there appears some mathematical text pointing to Arfs discovery
of what today is called the Arf invariant. Accordingly the organizers in their letter
suggested that perhaps I would want to talk about the Arf invariant of quadratic forms.
Cahit Arf on Turkish bank note
Although I do not consider myself as a specialist on quadratic forms, it was my

pleasure to follow this suggestion. CahitArf had been a Ph.D. student of Helmut Hasse
in 1937/38. Arfs thesis [Arf39] has become widely known, where he had obtained
a generalization of a former theorem of Hasse about the ramification behavior of
abelian number fields; today this is known as the HasseArf theorem. 2 His next
paper, after his thesis, contains the Arf invariant which is our concern here. This
work too was inspired by a suggestion of Hasse. So this report about the Arf invariant
fits into my general project to investigate the mathematical contacts of Hasse with
various other mathematicians, including Emil Artin, Emmy Noether, Richard Brauer
and others, and now with Cahit Arf.
Much of what I know in this respect is based upon the letters between Hasse and
his correspondence partners. Those letters are kept in the Handschriftenabteilung
of the Gttingen University Library, they contain a rich source for those who are
interested in the development of algebraic number theory in the 20th century. Among
those documents there are preserved about 65 letters between Arf and Hasse from
1939 until 1975. 3 We can see from them that in the course of time there developed a
heartfelt friendship between the two.
2 This generalization had been asked for by Artin in a letter to Hasse. For details from the historic perspective
see, e.g., section 6 of [Roq00].
3 In addition there are about 90 letters between Hasse and Arfs wife Halide, mostly in Turkish language, in
which Hasse tried to practice and improve his mastery of the Turkish language.
191
7.2 Arfs first letter

The first ten letters are concerned with Arfs work on quadratic forms in characteristic 2. But where are the earlier letters, those about Arfs thesis? The answer is easy:
There were no earlier letters, for during his graduate studies while composing his
thesis, Arf worked at Gttingen University where Hasse was teaching. And people
at the same university usually do not write letters but talk to each other.
Fortunately for us, when Arf worked on quadratic forms in characteristic 2 he was
back in Istanbul, and therefore the communication with his former academic teacher
travelled by means of letters. On October 12, 1939 Arf wrote to Hasse: 4
Sehr geehrter Herr Professor,
Ich habe Ihren Brief vom 29. 9. 39 mit grosser Freude erhalten
Ich habe jetzt eine unschne Arbeit ber quadratische Formen fast fertig
geschrieben. Diese Arbeit wollte ich Ihnen vorlegen. Ich glaube aber,
dass Sie jetzt wenig Zeit haben. Es handelt sich kurz um folgendes:
Sie hatten einmal den Wunsch geuert, die Geschlechtsinvarianten einer quadratischen Form mit Hilfe der Algebrentheorie begrndet zu sehen.
Ich habe versucht dies zu tun. Da die Aufstellung dieser Invarianten fr
p 2 fast trivial ist, habe ich gedacht, dass es ntzlich sein wrde
wenn man zunchst die Theorie in Krpern von der Charakteristik 2 zu
bertragen versucht. In der genannten Arbeit bertrage ich die Ergebnisse von Witt durch passende nderungen in den Krpern von der
Charakteristik 2 und ich gebe dann die vollstndigen Invariantensysteme fr arithmetische quivalenz der ternren und quaternren Formen
in einem Potenzreihenkrper k..t//, wobei der Koefzientenkrper k
die Charakteristik 2 hat und vollkommen ist
Dear Professor,
I am very glad to have received your letter of September 29, 1939 I
have almost completed the draft of a paper on quadratic forms. I had
intended to submit it to you. But I believe that now you will not have
much time for it. In short, the situation is as follows:
You had once expressed your wish to see the genus invariants of a
quadratic form be established with the help of the theory of algebras.
This I have tried to do. Since the compilation of those invariants is almost trivial in characterisctic 2 I thought it would be useful at first
to try to transfer the theory to fields of characteristic 2. In the above
mentioned paper I transfer the results of Witt by suitable modifications
4 Observe the date of this letter. On September 1, 1939 World War II had started. Perhaps this was the reason
why Arf believed that Hasse would not be able to devote much time to deal with Arfs paper since in war time
Hasse may have been assigned to other duties.
192
to fields of characteristic 2. And then I give a complete system of invariants for the arithmetic equivalence of ternary and quaternary forms in a
power series field k..t// 5 where the field of coefficients k is perfect of
characteristic 2
From this we learn that it had been Hasse who had suggested the topic of Arfs
investigation. Hasses interest in quadratic forms stems from the time of his own
thesis, 1923/24, when he had proved the Local-Global Principle for quadratic forms
over number fields [Has23a, Has24a]. Later he had established the Local-Global
Principle for central simple algebras over number fields, in cooperation with Emmy
Noether and Richard Brauer [BHN32]. There arose the question as to the mutual
interrelation between the theory of quadratic forms and the theory of algebras. Perhaps
it would be possible to deduce the Local-Global Principle for quadratic forms from
that for algebras ?
This question (and more) had been answered beautifully in Witts seminal paper
[Wit37a]. At that time Witt held the position as assistant professor in Gttingen, and
he was the leading member of the Arbeitsgemeinschaft in cooperation with Hasse. In
his paper Witt associates to every quadratic form f a central simple algebra S.f / of
2-power index, called the Hasse algebra which, together with the dimension and the
discriminant of the form, makes a complete set of invariants at least over global and
local function fields. (Over number fields there is an additional invariant, namely the
signature of a quadratic form over the real localizations of the field.)
Witts paper represents a watershed in the theory of quadratic forms. It provided
the basis of the subsequent enormous expansion of the theory of quadratic forms. His
biographer Ina Kersten says that this paper ranks as one of his most famous works
[Ker00]. However, Witts theory covered only forms over a field of characteristic 2.
This is the point where Arfs paper comes in. He extended Witts theory to fields of
characteristic 2. In particular this applies to the case of local and global function
fields of characteristic 2.
The desire to extend Witts result to characteristic 2 had also been expressed by
A. A. Albert in a paper which had just appeared in 1938 in the Annals of Mathematics
[Alb38b]. There had been some letters exchanged between Albert and Hasse during
the years 19311935 and we know that Alberts interest in the theory of quadratic
forms over global fields had been encouraged by Hasse. 6 The above mentioned
paper by Albert shows that this interest continued. His paper is a follow-up on
Witts [Wit37a] on quadratic forms: Albert first reproves, in his own way, some of
Witts results for global function fields and then shows that these hold also in the
characteristic 2 case: namely that every quadratic form in 5 or more variables is
isotropic. And then, in a footnote, he says about Witts general theory:
5Arf writes here hti but we will use throughout our own notation and do not follow the various notations
in the original letters and papers.
6 See chapter 8 in [Roq05].
193
The results of Witt on quadratic forms on a field of characteristic not

two may probably be obtained for the characteristic two case only for
forms with cross product terms. 7 It would be very interesting to study
the analogues of Witts results for our characteristic two case but the
author has not yet done so.
We see that Arf did just what was proposed here. I do not know whether Arf knew
about Alberts paper and its footnote. In his own paper he cites Witt only and says in
the introduction:
Die Anregung zu dieser Arbeit verdanke ich H. Hasse.
I owe to H. Hasse the suggestion for this work.
As we learn from Arfs letter, there was a second part in his manuscript where he
investigates quadratic forms over rings of power series and their arithmetic invariants,
at least for quadratic forms of low dimension. Again, the motivation for this comes
from number theory. Due to his results in this second part Arf can be regarded as a
forerunner of the general theory of quadratic forms over rings, not necessarily fields.8
But in the end it turned out that Parts 1 and 2 were published separately. Part 1
appeared in Crelles Journal where Hasse was editor [Arf41]. Hasse would have
liked to get also Part 2 for Crelles Journal but it seems that there arose difficulties
with the printing due to paper shortage in war times and so the second part appeared
in the journal of the University of Istanbul. 9
Here I will discuss only the first paper of Arf [Arf41] where he introduces his Arf
invariant. It has turned out that there is an error in Arfs paper which on the one hand
reduces the scope of his main result but on the other hand has led to an interesting
development in the theory of quaternion algebras over fields of characteristic 2. We
shall discuss this in due course.
But before going into details let us familiarize a little with the people involved
and with the time of the game.
7 It appears that Albert means quadratic forms which are regular in the sense as defined below (and completely regular in Arfs terminology). See Section 7.4.
8 This is evident when we look at the book by Knus who gives a survey on quadratic forms over rings [Knu80]
where the notion of Arf invariant over rings is systematically treated. (However we do not find Arfs Part 2
[Arf43] cited in the bibliography of Knus book.) It seems that the most recent survey on the topic is contained
in the book [EKM08].
9 See [Arf43]. The paper is usually cited for the year 1943 but at the end of the paper we read that it has been
received on May 15, 1944.
194
7.3 Some personal data

Cahit Arf was born in 1910 in the town of Selanik which today is Thessaloniki. At
that time it belonged to the Ottoman empire. But in the course of the Balkan war
1912 his home town was affected and the family escaped to Istanbul, later in 1919 to
Ankara and finally moved to Izmir. 10
Cahit Arfs childhood encompassed the Balkan wars, the World War I,
the grand war at Gallipoli, the Greek invasion of westernAnatolia and the
invasion of Istanbul by the Allied Powers. When finally Turkey emerged
as a new independent parliamentary republic in 1923 Cahit Arf was 13
years old. It was the beginning of a new era. The new Republic was
hopeful, determined and full of invincible self confidence. These traits
were also deeply trenched in young Arf. This would shape his attitude
towards mathematics in the future.
In public school Arfs ease in mathematics was soon to be noticed by his teachers.
In 1926 his father sent him to France to finish his secondary education at the prestigious
St. Louis Lyce. Because of his extraordinary grades in mathematics he graduated in
two years instead of the expected three years. Then he obtained a state scholarship
to continue his studies at the cole Normale Suprieur, again in France.
After his return to Turkey in 1932 Arf taught at high school and since 1933 he
worked as instructor at Istanbul University.
It did not take him long to realize that he needed graduate study in
mathematics. In 1937 he arrived at Gttingen University to study with
Helmut Hasse.
At that timeArf was 27 years of age. I do not know why he had chosen Gttingen as
his place of graduate study. Although Gttingen used to be an excellent mathematical
center which was attractive to students throughout the world, that period had ended
in 1933 when the new Nazi government decided to discharge the Jewish and the
non-conformist professors; this had disastrous effects to the mathematical scene in
Gttingen. It seems improbable that Arf had not heard about the political situation
in Germany and its consequences for the academic life in Gttingen. Perhaps his
mathematical interests at that time leaned towards algebra and arithmetics and he had
found out (maybe someone had advised him) that in Gttingen there was Hasse who
was known as an outstanding mathematician in those fields. In fact, measured by the
high standard of Arfs thesis which he completed within one year after his appearance
in Gttingen, it seems that already in Istanbul he had acquired a profound knowledge
in the basics of modern algebra and algebraic number theory, and accordingly he may
have chosen Gttingen because of Hasses presence there.
10 The following citations are taken from the Arf biography written by Ali Sinan Sertz [Ser08]. There one
can find more interesting information about the life, the work and the personality of Cahit Arf.
195
Helmut Hasse was 38 years of age when in 1937 Arf arrived in Gttingen. Hasse
was known as a leading figure in the development of algebraic number theory, in
particular of class field theory. Just one year earlier in 1936 he had been chosen as
an invited speaker at the International Congress of Mathematicians in Oslo. There
he reported on his proof of the Riemann hypothesis for elliptic function fields over
finite fields of constants; this proof had appeared in three parts in the 1936 issue of
Crelles Journal [Has36c], [Has36b], [Has36a]. I would say that in those years Hasse
was at the height of his mathematical power (notwithstanding a certain peak of his
mathematical activities in the years after World War II). In 1934 Hasse had decided to
leave the University of Marburg and to accept an offer to Gttingen. Hasse was not a
Nazi but he described his political position as being patriotic. He strongly disagreed
with the policy of expelling so many scientists from Germany; he considered this as
a tragic loss of intellectual power in Germany. And he tried to do what he could to
counteract this. When he decided to move from Marburg to Gttingen he did this with
the expressed intention to restore, at least to a certain extent, the glory of Gttingen
as an international place for mathematics. Although he spent a lot of time and energy
on this, he could not be successful in the political situation. 11
However, on a relatively small scale Hasses activity in Gttingen had remarkable
success. He managed to attract a number of highly motivated students to his seminar
and the Arbeitsgemeinschaft. The latter was organized by Witt but Hasse participated
at the meetings and led the mathematical direction of the work.
The high scientific level of the work in the Arbeitsgemeinschaft is documented in
a number of publications in Crelles Journal and other mathematical journals. Here
we only mention volume 176 of Crelles Journal which appeared just in 1937 when
Arf came to Gttingen. A whole part of this volume 12 contains papers which arose
in the Arbeitsgemeinschaft and in the Seminar, of which Witts famous paper on the
so-called Witt vectors is to be regarded as a highlight. In the same volume (but in
another part) appeared Witts paper on quadratic forms [Wit37a] which, as said earlier
already, has decisively influenced Arfs paper on his invariant [Arf41].
Ernst Witt was 26 when he met Arf, hence one year younger. He had studied
in Gttingen since 1930. He had received the topic of his Ph.D. thesis from Emmy
Noether but since she had been dismissed she could not act as his thesis referee.13
The thesis was concerned with central simple algebras over function fields in the
course of which he proved the Riemann-Roch theorem for algebras, a ground breaking
paper [Wit34] which nowadays attracts new interest in the setting of non-commutative
algebraic geometry. 14 When Hasse came to Gttingen in 1934 he accepted Witt as his
assistant on the recommendation of Emmy Noether. Witt has not many publications
11 For more facts from Hasses biography see, e.g., Freis biography [Fre85], as well as Freis recollections
about Hasse in [FR08]. Hasses involvement with the Nazi regime is discussed, e.g., in [Seg03].
12 Each volume of Crelles Journal appeared in 4 parts (4 Hefte).
13 This was Herglotz with whom Witt maintained close relationship.
14 It is said that Witt completed the manuscript of his thesis within one week.
196
when compared to the work of other mathematicians, but each one is of high level and
witnesses a profound insight into matematical structure. We have already mentioned
his 1937 paper where he introduces Witt vectors [Wit36]. 15 Earlier the same year
there had appeared his paper on quadratic forms [Wit37a] which actually constituted
his Habilitation thesis. 16
It seems fortunate that Arf in Gttingen had the chance to join the inspiring and
motivated group of young mathematicians around Hasse, and among them Witt. There
arose a friendship between the two which lasted for many years. It is without doubt
that Arfs paper on quadratic forms in characteristic 2 has been influenced by Witts
in characteristic 2.
7.4 Quadratic spaces

Let K be a field. Classically, a quadratic form over K is given by an expression
X
q.x/ D
aij xi xj with aij 2 K:
(7.1)
1ij n
Two quadratic forms are said to be equivalent if one is obtained from the other by
a non-degenerate K-linear transformation of the variables x D .x1 ; : : : ; xn /. An
invariant is a mathematical entity attached to quadratic forms which does not change
if a form is replaced by an equivalent form.
Witt had replaced the above notion of quadratic form by the notion of quadratic
space which was adapted to the Modern Algebra of the time [Wit37a]. A quadratic
space over K is a vector space V equipped with a function q W V ! K and a bilinear
function W V
V ! K subject to the following conditions:
q. x/ D 2 q.x/
for 2 K; x; y 2 V:
(7.2)
q.x C y/ D q.x/ C q.y/ C .x; y/
We assume V to be of finite dimension n. If u1 ; : : : ; un is a basis of V then any
x 2 V may be written as x D x1 u1 C xn un with xi 2 K and then q.x/ appears
in the form (7.1) with aij D .ui ; uj / for i < j and ai i D q.ui /. In Witts setup
the notion of invariant now refers to isomorphisms of quadratic spaces instead of
equivalences of quadratic forms.
15 It should not be forgotten that Witt vectors had been discovered somewhat earlier already by H. L Schmid
[Sch35a] who also was one of Hasses assistants at that time. H. L. Schmid however worked with the main vector
components only (Hauptkomponenten) where the formulas for addition and multiplication are quite cumbersome.
It was Witt who observed that the structural operations for Witt vectors can be described quite easily in terms of
the ghost components (Nebenkomponenten). In this way he made the calculus of Witt vectors widely applicable.
16 More biographic information about Witt can be obtained from Ina Kerstens biography [Ker00] and the
articles cited there.
197
It is common to interpret q.x/ as the length of the vector x 2 V , more precisely

as the square of its length. In fact, Witt and also Arf write jxj2 instead of q.x/.
Similarly .x; y/ is interpreted as the inner product of the vectors x and y and
accordingly Arf writes x y for it. Witt however writes x y for 12 .x; y/ which is
possible in characteristic 2 and corresponds more to our geometric intuition, for
then one has x x D jxj2 . In characteristic 2 however this is not possible and so we
have x x D 2jxj2 D 0. In other words, in characteristic 2 we have to live with the
fact that every vector is orthogonal to itself.
This has consequences. The first observation is that the process of diagonalization
is not generally possible in characteristic 2. Recall that in characteristic 2 every
quadratic form admits an equivalent diagonal form:
X
q.x/ D
ai xi2 with ai 2 K:
(7.3)
1in
In Witts terminology this means that every quadratic space V admits an orthogonal
basis u1 ; : : : ; un , where q.ui / D ai . Thus V splits as an orthogonal direct sum of
one-dimensional quadratic subspaces:
X
V D?
Kui in characteristic 2.
(7.4)
1in
But in characteristic 2 this is not always possible. Arf observed that one has to admit
also two-dimensional subspaces:
X
X
V D?
.Kui C Kvi /C ?
Kwj in characteristic 2
(7.5)
1j s
1ir
where ui and vi are not orthogonal to each other, i.e., .ui ; vi / 0: After suitable
normalization we may assume that
.ui ; vi / D 1: 17
The dimension n of V is n D 2r C s. Thus in characteristic 2 any quadratic form
admits an equivalent form as follows:
X
X

q.x/ D
cj zj2
(7.6)
ai xi2 C xi yi C bi yi2 C
1j s
1ir
for
xD
.xi ui C yi vi / C
zj wj ;
1j s
1ir
where we have put

ai D q.ui /;
bi D q.vi /;
cj D q.wj /:
17Arf however does not assume this and he admits for .u ; v / any non-zero element in K. Therefore his
i
i
formulas for the Arf invariant look a little more complicated than ours.
198
Arf speaks of quasi-diagonalization since only the second sum in (7.6) is in pure
diagonal form whereas the matrix of the first sum splits into 2
2 submatrices along
the diagonal. Note that in characteristic 2 the square operator is additive; accordingly
the second sum in (7.6) is called the quasi-linear part of q. The second sum in (7.5)
consists of all z 2 V which are orthogonal to V ; therefore it is denoted by V ? .
In characteristic 2 a quadratic form (7.1) is called regular, or equivalently
non-singular, if it has no equivalent form which can be written in fewer than n
variables; this means that in the diagonal form (7.3) all the coefficients ai 0. In
characteristic 2 things are different. Arf retained the above definition of regular; in
characteristic 2 this means that the coefficients cj in the quasi-linear part of (7.6) are
linearly independent over the subfield K 2 . If there does not appear a quasi-linear part,
i.e., if s D 0, then Arf called the quadratic form completely regular (vollregulr).
Today the terminology has changed. Instead of Arfs completely regular one
says regular (or non-singular) whereasArfs regular is now referred to as semiregular. This reflects the experience that in the new terminology, the regular forms
in characteristic 2 behave in many respects similar to the regular forms in characteristic
2.
We shall use here the terminology of today, deviating from Arfs.
The notions of regular und semi-regular are invariant, and can also be applied
to a quadratic space: V is regular if V ? D 0, and V is semi-regular if V ? does
not contain a vector x 0 with q.x/ D 0, i.e., if V ? is anisotropic.
Every quadratic space .V; q/ can be scaled. If 0 c 2 K then the scaled quadratic
space with scaling factor c is defined to be .V; cq/. Notation: V .c/ . Thus in V .c/ the
length of every vector x is c q.x/ whereas in V it is q.x/.
7.5 Clifford algebras

The Clifford algebra C.V / of a quadratic space V is defined as an associative
K-algebra (not commutative in general) generated by the K-module V and with
the defining relations:
x 2 D q.x/ for x 2 V:
(7.7)
In view of (7.2) this implies
xy C yx D .x; y/
for x; y 2 V:
(7.8)
If u1 ; : : : ; un is a K-basis of V then a K-basis of C.V / is given by the products

ui1 ui2 uik with i1 < i2 < < ik and 0 k n. The K-dimension of C.V /
is 2n .
In view of its definition C.V / is an invariant of V . So is the subalgebra C0 .V /
C.V / which is generated by the products ui1 ui2 uik with an even number k of
199
factors. The invariance of C0 .V / is a consequence of the fact that the defining

relations (7.7) are of degree 2. The dimension of C0 .V / is 2n1 .
If K is of characteristic 2 then we have the following rule:
If V D V1 ? V2 then C.V / D C.V1 / C.V2 /:
(7.9)
Here I have written V1 ? V2 to indicate the orthogonal direct sum of V1 and V2 . The
tensor product is taken over K as the base field. The validity of (7.9) is immediate
if we observe that any x1 2 V1 and x2 2 V2 are orthogonal to each other, i.e.,
.x1 ; x2 / D 0. Hence from (7.8) we conclude that x1 x2 D x2 x1 D x2 x1 showing
that C.V1 / and C.V2 / commute elementwise. (At the same time we see that in
characteristic 2 this is not the case since the appearing minus sign cannot be
disregarded.)
In view of the decomposition (7.5) we obtain for characteristic 2 that the Clifford
algebra C.V / decomposes into the tensor product of r factors C.Kui C Kvi / of
dimension 4, and the factor C.V ? / of dimension 2s . The latter is the center of C.V /
and does not appear if V is regular.
7.6 Binary quadratic space

First Arf investigates the Clifford algebra of a regular quadratic space which is binary,
i.e., of dimension 2. This discussion is quite elementary but it is fundamental for all
of Arfs results.
A binary space is generated by 2 elements u; v with the relations
q.u/ D a; q.v/ D b;
.u; v/ D 1:
(7.10)
Its Clifford algebra C.V / is given by the defining relations

u2 D a;
v 2 D b;
uv C vu D 1:
(7.11)
This is a central simple algebra of dimension 4, i.e., a quaternion algebra, with the
K-basis 1, u, v, uv.
The even subalgebra C0 .V / is of dimension 2 and has the basis elements 1 and
uv with the relation:
.uv/2 C uv D uv.vu C 1/ C uv D u2 v 2 D ab:
Putting w D uv and introducing the ArtinSchreier operator }.X/ D X 2 C X we
may write this as:
C0 .V / D K.w/ with }.w/ D ab:
(7.12)
200
Suppose first that ab }.K/. Then K.w/ is a separable quadratic field extension. According to the ArtinSchreier theory this field is uniquely determined by the
residue class of ab modulo }.K/. The nontrivial automorphism of K.w/ is given by
conjugation with u:
u1 wu D vu D w C 1:
(7.13)
Since u2 D a we see that C.V / is a cyclic crossed product of the separable quadratic
field K.w/, whose factor system is determined by the element a 2 K modulo the
norm group from K.w/ .
The original quadratic form q.x; y/ can be rediscovered (up to equivalence) from
these data by the norm function N W K.w/ ! K as follows:
a N.x1 C x2 w/ D a .x1 C x2 w/.x1 C x2 .w C 1//
D ax12 C x1 .ax2 / C b.ax2 /2
D q.x1 ; ax2 /:
(7.14)
Although this is not the original quadratic form q.x1 ; x2 /, it is equivalent to it. This
formula can be interpreted as follows: We may regard K.w/ as a quadratic space with
respect to the norm function N W K.w/ ! K. Consider the scaled space K.w/.a/
which has the quadratic form aN W K.w/ ! K. Then formula (7.14) says that
K.w/.a/ as a quadratic space is isomorphic to V . This isomorphsm is given by
1 7! u, w 7! a1 v.
As a consequence we state the following fact for later use:
The image set q.V / of the quadratic space V equals the coset of the
norms from K.w/ which contains a, i.e.:
q.V / D a N.K.w//:
(7.15)
In the above discussion we had assumed that ab }.K/. The case ab 2 }.K/
is somewhat exceptional since in this case K.w/ is not a field but a commutative
separable K-algebra which decomposes into the direct product of two copies of K.
Writing ab D }.c/ with c 2 K and putting e1 D w C c and e2 D e1 C 1, we obtain
orthogonal idempotents e1 , e2 , hence K.w/ D Ke1 Ke2 . By suitable choice of the
basis u, v of V one can achieve that u2 D a 0; then u admits an inverse in A and
conjugation with u induces the automorphism of K.w/ which permutes e1 and e2 .
Hence again, C.V / is a crossed product of K.w/, determined by the element a 2 K
modulo norms from K.w/. 18 But every a 2 K is a norm from K.w/ in this case,
18 K.w/ is a commutative separable algebra over K with an automorphism group G of order 2 (the group
interchanging the two copies Ke1 and Ke2 of K). Thus K.w/ is a quadratic Galois algebra. The theory of
crossed products of Galois algebras can be developed in complete analogy to the theory of crossed products for
Galois field extensions. The first who had done this explicitly seems to be Teichmller in his paper [Tei36b]. His
terminology was Normalring for what today is called Galois algebra. Quite generally, for cyclic algebras
we refer the reader to [LR03] and to [Lor08a].
201
and therefore C.V / splits, i.e., it is a full matrix algebra over K. The formula (7.14)
is still valid. In this way the case ab 2 }.K/ appears quite analogous to the case
ab }.K/.
But there is one essential difference. If ab }.K/ then the quadratic space V is
anisotropic, i.e., q.x/ D 0 only for x D 0. This we see from (7.14) since K.w/ is
a field and therefore the norm function N.z/ 0 if z 0. But if ab 2 }.K/ then
V is isotropic since N.e1 / D N.e2 / D 0. In this case it turns out that the quadratic
form is equivalent to q.x1 ; x2 / D x1 x2 . The corresponding quadratic space is called
the hyperbolic plane and denoted by H .
In any case, the residue class of ab modulo }.K/ is an invariant of V since it is
determined by K.w/ D C0 .V /. This residue class is by definition the Arf invariant
of the binary space V : 19
Arf.V / W ab mod }.K/:
(7.16)
The above discussion shows the validity of the following

Theorem. Let V D Ku C Kv be a binary regular quadratic space, so
that (7.10) holds. The Clifford algebra C.V / is a quaternion algebra
over K. Together with the Arf invariant Arf.V / the Clifford algebra
C.V / completely determines V up to isomorphism. V is isotropic if and
only if Arf.V / 0 mod }.K/, and then V is the hyperbolic plane H .
Remark. Central simple algebras in every characteristic p > 0 with defining relations
up D a;
}.w/ D c;
uw D .w C 1/u:
(7.17)
had been systematically studied earlier, in particular by Teichmller in his paper

[Tei36a]. Such an algebra is called p-algebra. Teichmller denotes it by .a; c . 20
In the case p D 2 we obtain a quaternion algebra A. In view of the above
considerations we see that A D .a; c is the Clifford algebra C.V / of the binary
quadratic space V D Ku C Kv with (7.10) for b D a1 c. Whereas the relations
(7.11) represent the description of A as the Clifford algebra of the given quadratic
space V , the relations (7.17) put into evidence the description of A as a crossed
product of some separable quadratic extension K.w/ of K with }.w/ D c. The
connection between the two is given by the fact that c is the Arf invariant of the space
V , while the quadratic form of V is given by the norm form of K.w/ scaled by a;
see (7.14).
The paper of Teichmller mentioned above appeared in 1936, one year before
Arf came to Gttingen. In this paper Teichmller studies, among other things, the
19Arf
in his paper writes .V /.

this notation had been chosen to signalize the fact that the symbol .a; c is not symmetric in a
and c. Compare it with the notation .a; b/ for a quaternion algebra in characteristic 2, given by the defining
relations (7.19) below. That symbol is symmetric in the sense that .a; b/ D .b; a/.
20Apparently
202
conditions for two such p-algebras .a; c and .a; c 0 to be isomorphic. If p D 2 then
this result has some bearing on Arfs investigations. It would have been desirable
that Arf cites Teichmllers paper and points out the connection between his and
Teichmllers investigation. However Arf did not do this. Why not? Did he not know
Teichmllers paper?
Teichmller had been a very active member of the Gttingen Arbeitsgemeinschaft
but he had left for Berlin in early 1937. Hence Arf had probably not met Teichmller.
But certainly Teichmllers results were known and valued in Gttingen and Arf
must have heard about it. I find an explanation for Arfs silence about Teichmllers
work in a certain character trait of Arf which is mentioned in the biography of Sertz
[Ser08]: 21
Arf was in the habit of encouraging young mathematicians to discover
mathematics by themselves rather than to learn it from others. To support
his cause he would tell how in his university years, i.e., his cole Normale
years in Paris, he would never attend classes but proceed to develop
that theory himself.
It seems that during his stay in Gttingen Arf had proceeded similarly, for Sertz
reports in his biography:
Years later in Silivri, Turkey, Hasse would recall that after taking his
problem 22 Arf had disappeared from the scene for a few months only to
come back with the solution.
This suggests to me that when Arf in 1939 was back in Istanbul and worked on
quadratic forms then again he had proceeded similarly, i.e., discovering the solution
of his problem by himself and not consulting other people or papers. In fact, in his
paper [Arf41] Arf cites only one paper explicitly, namely Witts on quadratic forms
[Wit37a].
The above theorem holds in characteristic 2. Let us briefly compare it with the
similar situation in Witts paper for characteristic 2: In this case a binary quadratic
space is of the form V D hu; vi with mutually orthogonal vectors u and v and instead
of (7.10), (7.11) we have
q.u/ D a; q.v/ D b;
.u; v/ D 0:
(7.18)
The Clifford algebra C.V / is now given by the defining relations

u2 D a;
21 The
v 2 D b;
uv D vu:
(7.19)
story has been confirmed to me by several Turkish colleagues who had known Cahit Arf personally.
the problem for his Ph.D. thesis.
22 Namely
203
Again, this is a quaternion algebra. In the theory of algebras it is often denoted by

.a; b/. The even subalgebra C0 .V / is of dimension 2 and has the basis 1; uv but this
time with the relation:
.uv/2 D uv.vu/ D u2 v 2 D ab D d:
where d is the discriminant of V . Thus
p
C0 .V / D K. d /:
If d K 2 then 23 this is a quadratic field extension whose non-trivial automorphism
is given by transformation
with u. And again, we conclude that C.V / is a crossed
p
product of K. d / which splits if and only if a is a norm from L.
If d 2 K 2 then K.uv/ is not a field but the direct product of two copies of K.
In this case and only in this case the quadratic space is isotropic, and it turns out
that in this case the corrsponding quadratic form is equivalent to q.x; y/ D xy, the
hyperbolic plane.
So we see that for binary quadratic spaces Arfs situation in characteristic 2 is
quite similar to Witts situation in characteristic 2, the only difference being
p quite
natural, namely that the quadratic splitting field of C.V / is generated by ab in
the case of characteristic 2, whereas in characteristic 2 it is generated by a root of
the ArtinSchreier equation }.x/ D ab. And we see already here in the binary case:
In characteristic 2 the Arf invariant Arf.V / 2 K=}.K/ is the analogue
of the discriminant d.V / 2 K =K 2 in characteristic 2.
This was the guiding idea of Arf when he wrote his paper.
7.7 Higher dimensional quadratic spaces

Now let V be an arbitrary regular quadratic space over a field K of characteristic 2. From (7.5) we know that V decomposes into an orthogonal direct sum of
two-dimensional spaces:
X
V D?
Vi where Vi D hui ; vi i
(7.20)
1ir
and
q.ui / D ai ;
23 K
q.vi / D bi ;
.ui ; vi / D 1
.1 i r/:
denotes the multiplicative group of the field K and K 2 is the group of squares.
(7.21)
204
Definition of Arf invariant.

Arf.V /
Arf.Vi / mod }.K/:
(7.22)
1ir
Recall that by definition Arf.Vi / ai bi mod }.K/ so that this definition can also
be written as:
X
Arf.V /
q.ui /q.vi / mod }.K/:
(7.23)
1ir
This formula is printed on the Turkish 10-Lira note where, however, the underlying
field is restricted to be K D F2 , the prime field in characteristic 2. In that case
}.F2 / D 0 and hence the congruence sign in (7.23) can be replaced by equality.
If r > 1 then it is not clear a priori that Arf.V / is an invariant of V . For, the
definition (7.23) depends on how V is decomposed into orthogonal subspaces Vi
in the form (7.20). One has to show that for every two such decompositions the
corresponding sums in (7.22) are in the same class modulo }.K/. Arf does it in
his paper but the proof requires some cumbersome computations. In later years Witt
[Wit54] and Klingenberg [KW54] have given simplified descriptions of Arf.V / from
which one can see more directly its invariance. In the comments to Witts paper in
[Wit98] the editor Ina Kersten reports:
It was Witts concern in the fifties to eliminate the assumption that the
characteristic of the ground field is different from 2.
We interpret this such that Witt had carefully read Arfs paper and tried not only to
simplify Arfs proof but also to build a unified theory of quadratic forms, independent
of the characteristic. In particular Kersten mentions Witts cancellation theorem (see
Section 7.8 below) and his attempts to investigate in detail the geometric situation
which guarantees its validity.
Today we would verify the invariance of Arf.V / by investigating in more detail
the structure of the Clifford algebra C.V /. We have already said in Section 7.5
that C.V / contains a subalgebra C0 .V / which is canonically defined by V , namely:
C0 .V / is generated by the products with an even number k of factors in V . And in
Section 7.6 we have seen that in the binary case, C0 .Vi / D K.wi / is a quadratic
extension defined by the relation wi2 C wi D ai bi which shows, using ArtinSchreier
theory, that the class Arf.Vi / of ai bi is an invariant of C0 .Vi /, hence of Vi . But Arf
did not consider the subalgebra C0 .V /, probably he was not aware at that time that
C0 .V / was canonically defined by the quadratic space V . Therefore he had to use
somewhat cumbersome explicit computations.
But using the invariance of C0 .V /, the following statement immediately shows
that Arf.V / is an invariant: 24
24 I
have found it in the book by Knus [Knu80].
205
Proposition. Let V be a regular quadratic space, represented as an

orthogonal sum of two-dimensional spaces as in (7.20), (7.21).
P For
each i let C0 .Vi /PD K.wi / with }.wi / D ai bi . Put w D
i wi ,
so that }.w/ D i ai bi . Then the quadratic extension K.w/ equals
the center of C0 .V /, and hence by ArtinSchreier theory the class of
P
25
i ai bi is an invariant of V .
The essential part of the proof consists in verifying w to commute with every
element in C0 .V /. I recommend to verify this for r D 2 (and then use induction).
One has to use that
C0 .V / D C0 .V1 / C0 .V2 / C C1 .V1 / C1 .V2 /
D K.w1 ; w2 / C V1 V2
where C1 .Vi / denotes the K-space generated by all products of an odd number of
elements in Vi , hence C1 .Vi / D Vi : Show that w D w1 Cw2 commutes with w1 , with
w2 and with every product x1 x2 with xi 2 Vi . (Use the fact that wi xi D xi wi C xi ).
Let us mention that in Witts situation of characteristic 2 there arises a problem
with the Clifford algebra C.V /. For, in general this is not a central simple algebra
and it is not a product of quaternion algebras. For this reason in characteristic 2
Witt replaced the Clifford algebra C.V / by another algebra S.V / which Witt has
called Hasse algebra; this is defined as follows: First recall the notation .a; b/ for
the quaternion algebra defined by the relations (7.19). Now consider the coefficients
ai appearing in the diagonal form (7.3) and put di D a1 a2 ai . Then the Hasse
algebra is defined as the n-fold tensor product
Y
Y
S.V / D
.di ; ai /
.ai ; aj /: 26
(7.24)
1in
1ij n
This is a central simple K-algebra and plays a role in Witts theory of quadratic forms
in characteristic 2, analoguous to the Clifford algebra in characteristic 2. But its
definition (7.24) depends on the coefficients ai in the diagonal form (7.3). In order
to show that it is an invariant, it is necessary to study the transformation from one
diagonal form to an equivalent one. Witts computations for this are similar to Arfs
computations for the invariance of Arf.V / in characteristic 2. It seems to me that
Arf had modelled his invariance proof for Arf.V / after Witts invariance proof for
S.V /. 27
25 If }.w/ 0 mod }.K/ then K.w/ is not a field but the direct sum of two fields isomorphic to K. We
have discussed this situation already in the case of two-dimensional quadratic spaces.
26 Quite generally we write A
B if A; B are central simple K-algebras which determine the same element
in the Brauer group Br.K/.
27 We follow a suggestion of the referee and remark that the Hasse algebra S.V / in characteristic 2 is not
an invariant of the class of V in the Witt ring WQ.K/ contrary to the situation in characteristic 2 with the
Clifford algebra C.V / (see the next section).
206
7.8 Witt equivalence

For any type of mathematical structures, the quest for invariants is motivated by the
hope to be able to characterize the structures by their invariants (up to isomorphisms),
and thus to obtain a classification of the structures under investigation. Here we are
concerned with quadratic spaces V in characteristic 2 and in particular with regular
spaces. We now know three invariants:
1. the dimension dim.V /,
2. the Clifford algebra C.V / in the Brauer group Br.K/,
3. the Arf invariant Arf.V / in the additive group K=}.K/.
For arbitrary fields we cannot expect that these three invariants characterize V up
to isomorphisms. But Arf wished to show that for special fields K this is indeed
possible. Although, as we shall explain, his main result cannot be upheld in its full
generality, it turns out that the theorem is valid, e.g., over global and local fields
K in characteristic 2. In order to approach this problem, Arf follows Witt who had
discovered the Witt ring by introducing a certain equivalence relation.
Recall that a quadratic space V is called isotropic if there exists a non-zero vector
x 2 V with q.x/ D 0. The prototype of an isotropic regular space is the hyperbolic
plane H already introduced in Section 7.6. The Arf invariant of H is Arf.H /
0 mod }.K/, and the Clifford algebra is C.H / 1, which means that C.H / splits.
Arf proves the following
Kernel Theorem. (i) If the regular quadratic space V is isotropic then
V D H ? V 0 where V 0 is uniquely determined by V (up to isomorphisms).
(ii) Consequently, every regular quadratic space V can be decomposed
into an orthogonal sum of a number of spaces isomorphic to H and a
space V which is anisotropic, and V is uniquely determined by V (up
to isomorphisms).
The space V is called the anisotropic kernel of V . Its quadratic form is called the
kernel form 28 of V .
As a consequence of this result Arf proves the general
Cancellation Theorem. Suppose the quadratic space W is regular. If
there exist quadratic spaces V1 ; V2 such that W ? V1 W ? V2 then
V1 V2 .
In characteristic 2 this famous cancellation theorem was contained in Witts paper.
Arf has observed that it holds also in characteristic 2, but only if W is assumed to be
regular. 29
28 In
German: Grundform.
207
So Arf had obtained a new invariant of V , its anisotropic kernel V . The original
space V is obtained from V by adding an orthogonal sum of a number of hyperbolic
planes, as many as the dimension of V requires. We note that
C.V / C.V /
and Arf.V / Arf.V / mod }.K/:
(7.25)
since C.H / 1 and Arf.H / 0 mod }.K/. We conclude:

In order to classify the quadratic spaces it is sufficient to classify the
anisotropic spaces.
It is useful to work with the following
Definition of Witt equivalence. Two regular quadratic spaces V , W (or
quadratic forms) are Witt equivalent if they have isomorphic anisotropic
kernels. (Notation: V W .)
This is indeed an equivalence relation. It blends with the orthogonal sum, i.e., if
V1 W1 and V2 W2 then V1 ? V2 W1 ? W2 . 30 The Witt classes of regular
quadratic spaces with the operation ? form a group which we denote by WQ.K/.
We have V ? V 0, i.e., the elements of this group are of order 2.
7.9 Arfs Theorems

Now we are able to state the main result of Arfs paper. For an algebraic function
field or power series field K over a perfect base field of characteristic 2 he wished
to prove that the above three invariants completely characterize the regular quadratic
spaces. The relevant property of these fields was, in his opinion, the following which
concerns the Brauer group of K:
(Q): The quaternion algebras over K form a group within the Brauer
group Br.K/. In other words: If A and B are any quaternion algebras
over K then A B C where C is a quaternion algebra again.
It is not difficult to show that in characteristic 2 function fields and power series
fields over a perfect base field have the property (Q). But it seems that Hasse had not
seen it immediately and so he asked Arf about it, who replied in a letter of March 29,
1940:
29 On
the other hand, there is no such restriction necessary for V1 ; V2 . Consequently the definition below
of Witt equivalence applies to arbitrary quadratic spaces. But for simplicity we will restrict our discussion to
regular spaces.
30 If V and W are semi-regular but not regular then V ? W need not be semi-regular. Arf considers also this
situation but then the relation V ? V
0 does not hold generally.
208
Wenn A und B normale einfache Algebren vom Grade 2 sind, so ist

1
A B hchstens vom Index 4. Da aber K 2 W K D 2 so enthalten A
1
und B Teilkrper die zu K 2 isomorph sind. A und B enthalten also
Elemente u; v mit u2 D v 2 2 K die nicht zu K gehren. Es gilt daher
.u v/2 D 0
ohne, dass u v D 0
gilt.
A B enthlt also ein nilpotentes Element. Der Index von A B ist

daher hchstens 2.
If A and B are central 31 simple algebras of degree 2 then the index of
1
A B is at most 4. But since K 2 W K D 2, both A and B contain
1
subfields which are isomorphic to K 2 . Hence A and B contain elements
u and v respectively with u2 D v 2 2 K, and u, v do not belong to K.
Hence we have
.u v/2 D 0
but not
u v D 0:
Thus A B contains a nilpotent element. Therefore the index of A B

is at most 2.
This settled Hasses question but at the same time it showed that property (Q)
1
holds for all fields with K 2 W K D 2.
Arf stated his main results in the form of two theorems. As indicated earlier, there
is an error in the proof of his first theorem and in fact there do exist counterexamples.
Hence his first theorem has to be corrected. Nevertheless let us first state it as it
appears in Arfs paper.
Arfs Theorem 1 (to be corrected). 32 Assume that the field K of characteristic 2 satisfies property (Q). Then any regular quadratic space V
of dimension > 4 is isotropic. Consequently, its anisotropic kernel V
is of dimension 4. 33
Arfs Theorem 2. Assume that the field K of characteristic 2 has
the property that every regular quadratic space of dimension > 4 is
isotropic. Then every regular quadratic space over K is uniquely determined, up to isomorphism, by its dimension, its Clifford algebra and its
Arf invariant.
31Arf used the terminology normal but nowadays it is usually said central to indicate that the center of the
algebra equals the base field. The K-dimension of a central simple K-algebra A is a square n2 . The number
n is called the degree of A. The index of A is defined to be the degree of the division algebra D
A.
32 In Arfs paper [Arf41] this is Theorem 11, and our next Theorem 2 is numbered as Theorem 12 there.
33Arf also considered quadratic spaces V which are semi-regular but not regular. For those he claimed that
the regular part of V is of dimension 2.
209
Certainly, Arf regarded his second theorem as the highlight of his paper. He had
been able to accomplish his aim, namely to characterize quadratic forms by their
invariants. His first theorem was to give a sufficient criterion, in terms of the Brauer
group, for the field K which implies the characterization.
It can be easily verified that the condition (Q) is necessary for the validity of the
assertion in Arfs first theorem. But Arf was wrong to believe that it is also sufficient.
In order to find the correct condition, necessary and sufficient, we first remark that
condition (Q) is well known to be equivalent to the following condition:
(S) Any two quaternion algebras A; B over K admit a common quadratic
splitting field.
If two quaternion algebras have a common quadratic splitting field then they are
called linked. If every two quaternion algebras over K are linked, i.e., if K satisfies
condition (S), then K is called linked.
Every non-split quaternion algebra in characteristic 2 has two kinds of quadratic
splitting fields: separable and inseparable ones. If the quaternion algebras A, B have
a common inseparable quadratic splitting field then there is also a common separable
quadratic splitting field. This seems to have first been observed by Draxl [Dra75]. A
short and easy proof can be found in Lams paper [Lam02]. 34 See also Section 7.13.4
below.
But now comes the surprise: the converse does not hold. If A, B have a common separable quadratic splitting field then they do not necessarily have a common
inseparable quadratic splitting field. This has been observed by [Lam02]. 35 In view
of this the following condition appears stronger than (S):
(Sins ) Any two quaternion algebras A; B over K admit a common inseparable quadratic splitting field.
Fields K with this property may be called inseparably linked. R. Aravire and
B. Jacob [AJ95, AJ96] have shown that the iterated power series field F2 ..X//..Y //
is linked but not inseparably linked. We conclude that condition .Sins / is properly
stronger than (S), hence also properly stronger than Arfs condition (Q) which is
equivalent to (S).
It turns out that the proper correction of Arfs first theorem consists of replacing
his condition (Q) by the stronger condition .Sins /. This has been shown by Baeza
[Bae82]:
Baezas Theorem. (i) If K satisfies condition .Sins / then every regular
quadratic form over K of dimension > 4 is isotropic.
34 Perhaps it is not without interest to note that the formulas for quaternion algebras which have been used in
[Lam02] are special cases for p D 2 of formulas which have been stated 1936 by Teichmller for p-algebras in
characteristic p [Tei36a].
35 Lam cites [Tit93] and [Knu93] but his example is simpler and easier to verify.
210
(ii) Conversely, if every regular quadratic form over K of dimension > 4

is isotropic then .Sins / holds.
Quite generally, the so-called u-invariant u.K/ of a field K is defined to be
the smallest number u such that every regular quadratic form of dimension > u is
isotropic. Thus Baezas theorem can be formulated as follows:
If K is inseparably linked then u.K/ 4, and conversely.
If we observe that Arfs first (incorrect) theorem can be formulated as: If K is
linked then u.K/ 4, then we see that Arfs essential difference to Baezas theorem
consists in the absence of the inseparability condition. Apparently Arf was not aware
of the fact that there is a difference of the linkage behavior of quaternions according
to separability or inseparability. In fact, this question was first raised in 1974 only,
by Draxl [Dra75].
If K is a function field of one variable over a perfect field of constants then there
1
is only one inseparable quadratic field over K, namely K 2 . Hence K is inseparably
linked and Baezas theorem is applicable. In fact, in this case this is almost trivial;
see our appendix.
1
The classical fact that K 2 W K D 2 for a function field K over a perfact base
field, was also observed by Albert in his paper [Alb38b] which we have cited above
already. On this basis Albert had already proved that quadratic forms of 5 variables
over such function fields are isotropic, i.e., that u.K/ 4.
Apparently Arf did not know Alberts paper. When O. F. G. Schilling reviewed
Arfs paper in the Mathematical Reviews he wrote: The author is unaware of the
work of A. A. Albert. We observe that Schilling did say this as a statement, not as a
guess. Schilling had been a student of Emmy Noether in Gttingen and after Noethers
emigration got his Ph.D. with Helmut Hasse in Marburg. Later he went to the USA. 36
At the time when he wrote this review he held a position at the University of Chicago
with Albert. He had kept contact to Hasse by mail, and on these occasions he had
asked for information about the results in Hasses Gttingen mathematical circle. It
seems likely that he had been informed by Hasse or by someone else from Gttingen
about Arf and his results; this enabled him to state that Arf was not aware of Alberts
work, and that he did not add apparently or something like this. Certainly Schilling
himself knew Alberts papers.
Arfs (erroneous) proof of theorem 1 is not easy or straightforward but it is well
arranged. It seems to me that Arfs style in his paper was much influenced by the
suggestions and the advice of his academic teacher Hasse. For, I have found in Arfs
paper a footnote which Hasse, being the editor of Crelles Journal, had placed at the
end of the introduction:
36 He first stayed at the Institute for Advanced Study in Princeton where he had been accepted on the recommendation of Hasse who had written a letter to Hermann Weyl.
211
Anmerkung des Herausgebers: Im Einverstndnis mit dem Verfasser

habe ich dessen ursprngliches Ms. berarbeitet.
Note by the editor: With the consent of the author I have revised his
original manuscript.
We see that Hasse did with Arfs manuscript what he always did as an editor of Crelles
Journal, namely checking manuscripts carefully. As Rohrbach reports in [Roh98]:
With his [Hasses] characteristic conscientiousness, he meticulously
read and checked the manuscripts word by word and formula by formula. Thus he very often was able to give all kinds of suggestions to the
authors, concerning contents as well as form
So he did with Arfs paper. In the HasseArf correspondence we read several times
that Arf responds to changes suggested by Hasse, both approvingly and critically.
Finally on February 8, 1941 Arf returned the final version to Hasse and wrote:
Mit gleicher Post schicke ich Ihnen die Korrekturbogen und das Manuskript der Arbeit ber quadratische Formen zurck. Die nderungen an
drei Stellen die Sie vorgenommen haben scheinen mir unrichtig. Meine
Grnde habe ich am Rand des Manuskripts geschrieben.
At the same time I am returning the galley proofs and the manuscript
on quadratic forms. At three instances your proposed changes seem
not to be correct. I have explained my reasons at the margin of the
manuscript.37
The paper appeared in the same year 1941.
It seems curious that Hasse had not detected Arfs error although he was quite
interested in the subject and had closely examined Arfs paper. This is even more
curious since the error is of the same kind which many years ago, in 1927, Emmy
Noether had committed in a similar situation and there resulted a close correspondence
between Hasse and Emmy Noether about it. This correspondence finally led to their
renowned theorem about cyclicity of algebras over number fields. It appears that in
1940 Hasse had forgotten that incident.
The situation back in 1927 had been as follows: 38 Emmy Noether, in a letter to
Richard Brauer of March 28, 1927, wrote to him that every minimal splitting field of a
division algebra can be embedded into the algebra. Brauer knew that this was not the
case and provided her with a counterexample. But this example seemed unnecessarily
complicated to Emmy Noether; so she wrote to Hasse, in a postcard of October 4, 1927
asking whether he could construct easy counterexamples for quaternions. Hasse did
so: He constructed fields of arbitrary high degree (over Q) which were splitting fields
37 The
38 I
manuscript is not preserved. I do not know which changes Hasse had proposed.
have told this story in detail in [Roq05].
212
of the classical quaternions but no proper subfield had this splitting property. Thus
Noether as well as Hasse learned that one has to distinguish between minimal splitting
fields and splitting fields of minimal degree; the latter indeed can be embedded into
the algebra.
Now, in Arfs proof in the year 1940, a situation ocurred which was quite similar
to the earlier one in 1927. On page 164, in the second paragraph of the proof of his
Satz 11, Arf considered two quaternion algebras A1 ; A2 (i.e., Clifford algebras of
binary quadratic spaces V1 ; V2 ). He took separable quadratic extensions K1 ; K2 of K
which were splitting fields of A1 and A2 respectively, and he assumed K1 K2 . Then
A1 A2 is split by the field compositum K1 K2 which has degree 4 over K. In view
of his hypothesis (Q) Arf knew that A1 A2 is similar to some quaternion algebra,
hence it admits a splitting field of degree 2. And now he argued that necessarily
such splitting field can be found within K1 K2 . But this is not necessarily the case.
In other words: K1 K2 may be a minimal splitting field of A1 A2 although it is not
a splitting field of minimal degree.
Although Hasses example of 1927 had referred to quaternions over Q we have
here in characteristic 2 a similar situation. Why had Hasse not seen this error? We
will never know. We know that in 1940 it was wartime and Hasse had been drafted
to the Navy. He worked at a Navy research institute in Berlin and could attend to his
activities as an editor of Crelles Journal in the evenings and on weekends only. So
it seems that he did not check Arfs paper as thoroughly as he was used to in earlier
times with other papers for Crelles Journal.
7.10 An assessment of Arfs paper

Arfs paper [Arf41] is the first in which quadratic forms over arbitrary fields of characteristic 2 are systematically studied. It is true that there was already some work
before Arf which was concerned with quadratic forms in characteristic 2, e.g., Dickson in [Dic01] and Albert [Alb38b] but these discussed only special base fields:
finite fields and function fields of one variable, respectively. Due to Arfs general
Ansatz he has opened the door to an extensive expansion of the theory of quadratic
forms, not only over fields but also over arbitrary (commutative) rings.
Arf used the structural language, modern at his time, which had been introduced
by Witt into the theory of quadratic forms. Thus he spoke of quadratic spaces
instead of quadratic forms. Arf was able to extend a good part of Witts seminal
results in [Wit37a] to the case of characteristic 2. He showed the possibility of quasidiagonalization, he extended Witts important cancellation theorem (Krzungssatz) to
characteristic 2, he investigated the role of Clifford algebras for quadratic forms, and
he defined his Arf invariant in characteristic 2 as a substitute for the discriminant
in characteristic 0.
213
Arfs theorems were meant to find conditions for fields K of characteristic 2

which guarantee that all regular quadratic spaces over K are characterized (up to
isomorphism) by their three invariants:
Dimension,
Clifford algebra, Arf invariant.
His condition of linkage concerns the structure of the group of quaternion algebras
inside the Brauer group over K. Although the proof of his first theorem contained an
error and the theorem had to be modified, his paper still keeps its importance.
The error occurred due to the pathological and unforeseen behavior of quaternion
algebras in characteristic 2, which nobody of the time was aware of. In fact, it took
many years until this was discovered and cleared up, and thus Arfs first theorem
could be corrected [Bae82].
Due to Arfs number theoretical background his main interest was directed to
global and local fields K. In a letter to Hasse dated March 29, 1940 (from which we
have already cited on page 208) he had explained that any two quaternion algebras
1
over K have a common inseparable quadratic splitting field, namely K 2 . From this
he concluded that K is linked; this is true but it does not lead to the conclusion of
Arfs first theorem, namely u.K/ 4. In fact, what he did show in his letter is that
K is inseparably linked and this, due to Baezaa theorem, is sufficient for u.K/ 4.
It seems not impossible that Arfs first version of proof worked for local and global
fields of characteristic 2, and that he used correctly this inseparable linkage which he
had shown to Hasse. In any case, as will be put into evidence in our appendix, Arf
had all the ingredients of such proof at his disposal.
But since Arf did not realize the difference between linked and separably linked
quaternions, in his attempt to generalize his argument, he started with the linkage
condition (S) (or rather its equivalent condition (Q)), and since he could not prove
that this implies inseparable linkage (which we know today is not true), he tried to use
separable quadratic splitting fields. And so, since he was convinced of his theorem
due to his experience with global fields, he stumbled into his error. And even Hasse,
Witt, Albert and O. F. G. Scilling (among others) did not detect his error.
There are many examples in the history of mathematics showing that people,
even respected and competent mathematicians, who are convinced of the validity of
a theorem, are apt to accept any decent looking proof even at the cost of overlooking
some little detail which then may necessitate a correction of the theorem. 39
39 One of those examples is Grunwalds theorem in class field theory (1933), which was accepted by Artin,
R. Brauer, Hasse and Albert (among others) until Wang presented a counterexample (1948). See, e.g., section 5
of [Roq05].
214
7.11 Perfect base fields

If K is perfect of characteristic 2 then there exists only one quaternion algebra over
K (up to isomorphisms), namely the split one. The linkage conditions (S) and also
.Sins / are trivially satisfied. Therefore, in the characterization of regular quadratic
spaces the Clifford algebra can be omitted. Hence for a perfect base field, every
regular quadratic space V is characterized by its dimension and its Arf invariant.
But this result does not need Arfs two theorems for its proof. For perfect base
fields it follows almost immediately from the kernel theorem on page 206 that the
kernel V is of dimension 2. If w is the Arf invariant of V (and V ) then the
quadratic form of V is given by the norm formula (7.14) on page 200, where we can
put a D 1 since K is perfect. The original space V is obtained from V by adding a
number of copies of the hyperbolic space H , as many as is required by the dimension
of V .
In the special case when K is finite then }.K/ is an additive subgroup of index
2 in K and hence there is essentially one anisotropic quadratic space. Its quadratic
form is q.x; y/ D x 2 C xy C by 2 where b }.K/. If K D F2 then b D 1.
In the literature Arfs Theorem is often confined to this case. For instance, in
the Wikipedia 40 we read the following:
In mathematics, the Arf invariant of a nonsingular quadratic form over
the 2-element field F2 is the element of F2 which occurs most often
among the values of the form. Two nonsingular quadratic forms over
F2 are isomorphic if and only if they have the same Arf invariant. The
invariant was essentially known to Dickson (1901) and rediscovered by
Cahit Arf (1941). 41
Certainly, this is all true. But does it give an idea about the main discovery of
Arf ? In Arfs paper the field F2 is not mentioned at all. In a small remark, covering
four lines only, Arf mentions how his theory applies easily and almost trivially in
the case of perfect fields of characteristic 2. The main motivation and the main
results of the theory of Arf invariants are concerned with fields which are not perfect.
Arf studied the role of central simple algebras in the theory of quadratic forms in
characteristic 2. This aspect was not even scratched in that article of Wikipedia.
Moreover, the definition of Arf invariant as given in that article is valid only for the
base F2 and does not apply to other fields of characteristic 2. 42
The Wikipedia article seems to be written in view of the application of Arfs theory
in topology. For, several of those applications are mentioned in the article. And indeed
in topology one has to compute cohomology and other functors with coefficients
40 This refers to a former English version, August 20, 2009. A similar text, also restricted to the base field F ,
2
appears in the Encyclopedia of Mathematics.
41 It seems that the author tacitly assumes that both these quadratic forms have the same number of variables.
42 It is a good exercise to identify the defintion in that article with Arfs definition when the base field is F .
2
215
modulo 2 which means that the base field is F2 . An overview of the application of
Arf invariants in topology is given by Turgut nder in the appendix of the Collected
Papers of Cahit Arf [Arf90]. 43 But, as said above, this is not representative of Arfs
work which is meant to exhibit the role of central simple algebras in the theory of
quadratic forms in characteristic 2.
By the way, in the 1901 book of Dickson on linear groups [Dic01] which is
mentioned in the Wikipedia article, also the case of an arbitrary finite base field of
characteristic 2 is treated, not only F2 . The fact that Arf did not cite this book may have
one of two reasons: either he knew Dicksons book and found it is of no relevance
for his investigation (which would be understandable), or he did not know it (which
is more probable in view of his particular character trait which we have mentioned in
Section 7.6).
In any case, the statement that Arf has rediscovered what Dickson had known
is misleading. Arf discussed a quite different theorem, and in a very special case this
implies the statement of Dickson.
7.12 Epilog
After my conference talk I was asked aboutArfs biography for the years after his paper
on quadratic forms. I will not repeat here what is said in his biographies contained
in [Arf90, Ser08]. Let me only mention that he became a prominent member of
the Turkish scientific community (which is documented by the fact that his portrait
decorates an official banknote) but he also was a dedicated teacher. Many younger
mathematicians in Turkey had been introduced by him into mathematics, he had
encouraged them and showed them understandingly the way into our science. He is
widely remembered in the mathematical community of Turkey. Robert Langlands, in
his article about his impressions in Turkey, remembers warmly his discussions with
Arf [Lan04]. In particular Arf had directed Langlands attention to a paper by Hasse
on the local decomposition of the "-factors; these factors appear in the functional
equation of Artins L-series. As Langlands says (English translation):
I had rapid advance in my research having read Hasses paper
and
thanks to Cahit bey, I solved this problem during my stay in Ankara
and proved the existence of the local "-factor.
I was also asked to report more extensively on the correspondence between Arf
and Hasse, in particular the letters after 1941. I plan to do this some time in the
43 For the so-called ArfKervair invariant see, e.g., the article by Snaith in Morfismos Vol. 13 (2009), no. 2,
153, see also arXiv:1001.4751v1 [math.AT].
216
future. Let me only mention that these letters, although they do not discuss any
more mathematics proper, show a growing friendship between the two. Hasse visited
Turkey several times between 1957 and 1975. The last two preserved letters, dated
March 1975, concern the proposal to have an international colloquium on the structure
of absolute Galois groups. This colloquium was planned by Arf jointly with M. Ikeda
(who had earlier got a position in Turkey on the recommendation of Hasse). This
conference took place in September 1975 in Silivri, a small village on the beach of
the Marmara sea. I had the chance to participate in this conference and was able
to observe the close friendly relationship between the two mathematicians, Arf and
Hasse.
7.13 Appendix: Proofs

The aim of this section 44 is to put into evidence that the proof of Baezas theorem,
i.e., the correction of Arfs first theorem, can be done solely with the arguments which
can be found in Arfs paper. Hence Arf could well have proved Baezas theorem, thus
avoiding his error, if only he would have recognized the difference between linkage
of separable and inseparable type.
We also add a simple proof of Arfs second theorem, as well as of Draxls lemma.
We believe that our proofs, based on Arfs paper, are simpler than any of those
which can be found in the literature.
7.13.1 Baezas Theorem (i)
K denotes a field of characteristic 2. We use the following notation:
Let V be a regular quadratic space of dimension 2 over K. There is a K-basis
u; v of V with
q.u/ D a;
q.v/ D b;
.u; v/ D 1
.a; b 2 K/:
(7.26)
Here, q W V ! K denotes the quadratic form of V and is the corresponding bilinear

form. Let A D C.V / be the Clifford algebra of V . This is a quaternion algebra over
K with basis 1; u; v; w and the relations
u2 D a;
v 2 D b;
uv C vu D 1;
w D uv:
(7.27)
We identify the quadratic space V with the subspace of A generated by u and v, and
then q.x/ D x 2 for all x 2 V . If x 2 K 2 then K.x/ is an inseparable quadratic
subfield of A.
Lemma 1. Let y 2 A. Then y 2 2 K if and only if y 2 V C K.
44 This
section has been written jointly with Falko Lorenz.
217
If A does not split then this lemma gives a complete description of the inseparable
quadratic subfields L D K.y/ A : then y D x C c with suitable x 2 V , c 2 K.
The statement of Lemma 1 can be found in Arfs paper [Arf41] on page 161.
Arf does not formulate the statement in the form of a lemma, he just performs the
computations which we give in the proof below and uses them in his text. 45
Proof of Lemma 1. We represent y 2 A in the form
y D x C z with x 2 V;
z D c0 C c1 w 2 K.w/; c0 ; c1 2 K;
(7.28)
and compute
y 2 D x 2 C z 2 C xz C zx
D x 2 C z 2 C c1 .xw C wx/
D x 2 C z 2 C c1 x
(7.29)
where we have used that xw C wx D x for x 2 V , which is a consequence of

the relations (7.27). Now, x 2 D q.x/ 2 K, z 2 2 K.w/ and c1 x 2 V . Since
V \ K.w/ D 0 we conclude that y 2 2 K if and only if c1 x D 0 and z 2 2 K, hence
c1 D 0 since K.w/jK is separable.
In the following we regard K as a 1-dimensional quadratic space with the quadratic
form q.c/ D c 2 for c 2 K. The sum V C K A is direct and can be regarded as
the orthogonal sum V ? K of quadratic spaces.
Lemma 2. Let V; V 0 be regular quadratic spaces of dimension 2 and let A; A0 be their
Clifford algebras. Assume that A and A0 do not split. Then A; A0 have a common
inseparable quadratic splitting field if and only if V ? V 0 ? K is isotropic.
Proof. Let K.y/ A and K.y 0 / A0 be isomorphic inseparable quadratic subfields
of A and A0 respectively. We choose the generators y; y 0 in such a way that they
correspond to each other in the isomorphism K.y/ K.y 0 /, so that y 2 D y 0 2 . By
Lemma 1 we have y D x C c with 0 x 2 V and c 2 K. Similarly y 0 D x 0 C c 0
with 0 x 0 2 V 0 and c 0 2 K. We conclude
x2 C c2 D x0 2 C c0 2:
Putting d D c C c 0 we obtain
x2 C x0 2 C d 2 D 0
which shows that the quadratic space V ? V 0 ? K is isotropic.
45Arfs
notation is different from ours.
(7.30)
218
Conversely, assume V ? V 0 ? K is isotropic. There is a nontrivial relation of

the form (7.30) with x 2 V , x 0 2 V 0 , d 2 K. It follows
x 2 D .x 0 C d /2 :
(7.31)
Let, say, x 0. Since A does not split we have x 2 K 2 . Hence K.x/ is a

quadratic inseparable subfield of A. From (7.31) we conclude that K.x 0 C c/ A0
is isomorphic to K.x/.
Proof of Baezas theorem, part (i) (see page 209). We assume that every two nonsplit quaternion algebras over K have a common inseparable quadratic splitting field.
We claim that every regular quadratic space of dimension > 4 is isotropic. Write this
space in the form
V ?V0 ?W
where V and V 0 are of dimension 2 and W of dimension > 0.
Let y 2 W and assume first that q.y/ D 1. Then the subspace Ky W is
isomorphic to K as a quadratic space. Since the Clifford algebras C.V / and C.V 0 /
have a common inseparable quadratic splitting field (by assumption) we infer from
Lemma 2 that V ? V 0 ? Ky is isotropic. Hence V ? V 0 ? W is isotropic too.
But W may not contain a vector y with q.y/ D 1. In this case we use the
method of scaling (see page 198). If 0 c 2 K then the scaled quadratic space
V .c/ ? V 0 .c/ ? W .c/ is isotropic if and only if V ? V 0 ? W is isotropic. Now
choose y 2 W with q.y/ 0 and take the scaling factor c D q.y/1 . Then
c q.y/ D 1 which means that the scaled space .Ky/.c/ is isomorpic to K. From
what has been shown above it follows that V .c/ ? V 0 .c/ ? W .c/ is isotropic, hence
so is V ? V 0 ? W .
7.13.2 Arfs second theorem
We assume that every regular quadratic space of dimension > 4 over K is isotropic.
We claim that every regular quadratic space over K is uniquely determined (up to
isomorphism) by its dimension, the Brauer class of its Clifford algebra and its Arf
invariant.
Proof. Recall that WQ.K/ denotes the (additive) group of Witt classes of regular
quadratic spaces over K; see Section 7.8. The Witt class of the space V is represented
by its anisotropic kernel V . The whole space V arises from V by adding a number
of hyperbolic planes H , as many as the dimension of V requires.
Hence our claim reduces to the claim that the Witt class of V , i.e., the anisotropic
space V , is uniquely determined by its Brauer class of C.V / and the Arf invariant
Arf.V /.
We recall that
C.V / C.V /
and Arf.V / Arf.V / mod }.K/:
219
The Clifford algebra yields a homomorphism V 7! C.V / of WQ.K/ into the Brauer
group Br.K/. And the Arf invariant yields a homomorphism V 7! Arf.V / of WQ.K/
into the }-factor group K=}.K/. We have to show:
If C.V / 1 in Br.K/ and Arf.V / 0 mod }.K/ then V 0 in WQ.K/, which
is to say that the anisotropic kernel V of V vanishes.
We have dim V 4 by hypothesis. Since V is regular we have dim V D 0; 2
or 4. The case dim V D 2 is not possible since Arf.V / 0 mod }.K/ implies
V D H , the hyperbolic plane, hence V would not be anisotropic.
Suppose dim V D 4 and write V D V1 ? V2 as the orthogonal sum of two
binary spaces. Since C.V / 1 we have C.V1 / C.V2 /. Since both algebras have
the same dimension it follows that they are isomorphic: C.V1 / D C.V2 /. Also, since
Arf.V / 0 mod }.K/ we have Arf.V1 / Arf.V2 / mod }.K/. We have seen in
Section 7.6 that a binary regular space is uniquely determined by its Clifford algebra
and its Arf invariant. It follows V1 D V2 hence V D V1 ? V1 0, so again V
would not be anisotropic.
7.13.3 Baezas theorem (ii)
The following lemma is the separable analogue to Lemma 1. The situation is the
same as in Lemma 1.
Lemma 3. Let y 2 A. Then }.y/ 2 K if and only if y 2 V C K C w.
If A does not split then this lemma gives a complete description of the separable
quadratic subfields L D K.y/ A : then y D x C c C w with suitable x 2 V ,
c 2 K.
The statement of Lemma 3 can also be found in Arfs paper [Arf41] on page 165.
Arf does not formulate this statement in the form of a lemma, he just performs the
computations which we give in the proof below and uses them in his text. 46
Proof of Lemma 3. We represent y 2 A in the form as before in (7.28) and compute:
}.y/ D y 2 y D x 2 x C z 2 z C c1 x
D x 2 C .c1 1/x C }.z/
(7.32)
where we have used that xz C zx D c1 x. We conclude that }.y/ 2 K if and only if

c1 D 1 and so z D c0 C w, hence y 2 V C K C w.
Changing notation, we write c D c0 and hence y D x C c C w. Observing that
}.c C w/ D }.c/ C }.w/ we have shown that
}.y/ D }.x C c C w/ x 2 C }.w/ mod }.K/:
We will have occasion to use this formula later.
46 Our
notation differs from Arfs notation.
(7.33)
220
In the following lemma we consider K.w C w 0 / A A0 as a quadratic space

with respect to its norm function N W K.w C w 0 / ! K. Explicitly we have
N.c0 C c1 .w C w 0 // D c02 C c0 c1 C c12 }.w C w 0 /:
(7.34)
The sum V C V 0 C K.w C w 0 / A A0 is direct and can be regarded as the

orthogonal sum V ? V 0 ? K.w C w 0 / of quadratic spaces.
Lemma 4. Let V , V 0 be regular quadratic spaces of dimension 2 and let A, A0 be
their Clifford algebras. Assume that A and A0 do not split. Then A, A0 have a common
separable quadratic splitting field if and only if V ? V 0 ? K.w C w 0 / is isotropic.
Proof. Let L A and L0 A0 be isomorphic separable quadratic subfields of
A and A0 respectively. We may write L D K.y/ and L0 D K.y 0 / where y, y 0
correspond under the isomorphism L L0 and hence }.y/ D }.y 0 / 2 K. By
Lemma 3 we have y D x C c C w and y 0 D x 0 C c 0 C w 0 with suitable x 2 V ,
x 0 2 V 0 , c; c 0 2 K. From (7.33) we obtain:
x 2 C }.w/ x 0 2 C }.w 0 /
x Cx
02
x Cx
02
2
2
hence
mod }.K/;
0
}.w/ C }.w /
0
D }.d / C }.w C w /
mod }.K/;
(7.35)
with d 2 K:
(7.36)
x 2 C x 0 2 C }.d C w C w 0 / D 0:
Here, }.d C w C w 0 / D N.d C w C w 0 /; see (7.34). It follows that the quadratic

space V ? V 0 ? K.w C w 0 / is isotropic.
Conversely, assume that V ? V 0 ? K.w C w 0 / is isotropic. We have to show that
there exists a common separable quadratic splitting field of A and A0 . There exists a
nontrivial relation of the form
x 2 C x 0 2 C N.z/ D 0
with x 2 V; x 0 2 V 0 ; z 2 K.w C w 0 /:
We write z D c0 C c1 .w C w 0 / and use formula (7.34) for the norm. It follows

x 2 C x 0 2 C c02 C c0 c1 C c12 }.w C w 0 / D 0
(7.37)
Suppose first that c1 0. After dividing by c12 on both sides in (7.37) and
changing notation we may assume c1 D 1. Hence
x 2 C x 0 2 C }.c0 C w C w 0 / D 0
which gives
x 2 C }.w/ x 0 2 C }.w 0 / mod }.K/
221
and using (7.33):
}.x C w/ }.x 0 C w 0 / mod }.K/:
It now follows that the quadratic ArtinSchreier extensions K.x C w/ A and

K.x 0 C w 0 / A0 are isomorphic.
If c1 D 0 then (7.37) shows that V ? V 0 ? K is isotropic. From Lemma 2
we infer that A and A0 have a common inseparable splitting field. We have already
mentioned Draxls lemma which says that then there is also a common separable
splitting field. For a simple proof of Draxls lemma see the next Section 7.13.4.
Proof of Baezas theorem part (ii) (see page 209). We assume that every regular
quadratic space of dimension > 4 is isotropic. We claim that every two non-split
quaternions A, A0 over K have a common inseparable quadratic splitting field.
We write A D C.V / as the Clifford algebra of a 2-dimensional regular quadratic
space V as in (7.26), (7.27), and similarly A0 D C.V 0 /. As above we consider
the separable quadratic extension K.w C w 0 / A A0 as a quadratic space with
respect to the norm. The 6-dimensional space V ? V 0 ? K.w C w 0 / is isotropic
(by assumption) and hence Lemma 4 shows that A and A0 have a common separable
quadratic splitting field L. But we are looking for a common inseparable quadratic
splitting field; this will be established as follows.
The common separable quadratic splitting field L can be embedded into A and into
A0 ; this yields isomorphic separable quadratic subfields in A and in A0 . After changing
notation we now identify these two fields, so that A, A0 appear as crossed products
of the same separable quadratic field L. Writing L D K.w/ with }.w/ D c 2 K
we have A D .a; c and A0 D .a0 ; c with certain a; a0 2 K which represent the
factor systems of LjK defining A and A0 respectively. We refer to our discussion
on page 201. As explained there, A, A0 appear now as the Clifford algebras of the
0
quadratic spaces L.a/ and L.a / respectively, which are scaled quadratic spaces of the
space L with respect to the norm.
Now, by hypothesis the 6-dimensional space
0
L.a/ ? L.a / ? L
is isotropic. Thus there is a nontrivial relation of the form
aN.z/ C a0 N.z 0 / C N.y/ D 0
with
z; z 0 ; y 2 L:
(7.38)
If y 0 then after dividing by N.y/ and changing notation we obtain

aN.z/ C a0 N.z 0 / C 1 D 0
0
with z; z 0 2 K.w/.
This shows that L.a/ ? L.a / ? K is isotropic. Applying Lemma 2 we see that there
exists a common inseparable quadratic splitting field of A and A0 .
0
If y D 0 then from (7.38) we infer that L.a/ ? L.a / is isotropic, hence L.a/ ?
0
L.a / ? K is isotropic too and again Lemma 2 applies.
222
7.13.4 Draxls Lemma

Lemma 5 (Draxl). Let A and A0 be two nonsplitting quaternion algebras over K.
If A; A0 have a common inseparable quadratic splitting field then they also have a
common separable quadratic splitting field.
Proof. We represent A in the form
A D V C K.w/ with V D Ku C Kv, w D uv
(7.39)
as explained in Section 7.6, and similarly A0 . Let L be a common inseparable

quadratic splitting field of A and A0 . Embedding L A we have L D K.x C c/
with x 2 V , c 2 K, according to Lemma 1 on page 216. We may assume
that v is linearly independent of x C c (otherwise interchange v and u). Writing
uQ D x C c, vQ D v, w
z D uQ vQ we have another representation of the same kind as
(7.39): Az D Vz C K.w/
z : : : Changing notation, we omit the tilde and now have the
situation (7.39) with L D K.u/.
Similarly, we take a subfield L0 A0 isomorphic to L and we adapt the representation of A0 such that
A0 D V 0 C K.w 0 /
with V 0 D Ku0 C Kv 0 , w 0 D u0 v 0
and L0 D K.u0 / where u and u0 corespond to each other under the isomorphism
L L0 , hence
u2 D a D u0 2 :
We have to find isomorphic separable quadratic subfields of K.y/ A and
K.y 0 / A0 . To this end we put
y D cu C v C w
and y 0 D cu0 C w 0
where the coefficient c 2 K will be determined below. Using (7.33) we compute

(congruences are mod }.K/):
}.y/ .cu C v/2 C }.w/
c 2 a C b C c C }.w/ I
}.y 0 / c 2 a C }.w 0 /:
We see that
}.y/ }.y 0 / c b C }.w/ C }.w 0 /:
If c is chosen that way then the separable quadratic ArtinSchreier fields K.y/ A
and K.y 0 / A0 are isomorphic.
Chapter 8
HasseArfLanglands
Introduction of Robert Langlands at the Arf lecture, November 11, 2004.

Middle East Technical University, Ankara.
In those years, each year there was scheduled a special colloquium lecture at the Middle East
Technical University in Ankara, commemorating Cahit Arf who had served as a professor for
many years at METU. In the year 2004 the invited speaker was Robert Langlands, and I had
been asked to chair his lecture. The following text was read to introduce the speaker. I tried to
point out a historic thread from the work of Helmut Hasse to Robert Langlands, via Cahit Arf.
It is my pleasure and a great honor to me to introduce Robert Langlands as the

speaker of the Arf Lecture of this year 2004, here at the Middle East Technical
University.
Professor Langlands is a permanent member of the Institute for Advanced Study
in Princeton, which is one of the most prestigious places for mathematical research
studies. He has received a number of highly valued awards, among them the Cole Prize
in Number Theory from the American Mathematical Society in 1982, the National
Academy of Sciences reward in Mathematics in 1988, and the Wolf prize from the
Wolf foundation in Israel 1996. He is elected fellow of the Royal Society of Canada
and the Royal Society of London, and he has received honorary doctorates from at
least 7 universities (to my knowledge).
In the laudation for the Wolf prize it was said that he received the prize for his
path-blazing work and extraordinary insights in the fields of number
theory, automorphic forms and group representations.
Now, while I was preparing the text for this introduction it soon became clear to
me that in this short time I would not be able to describe, not even approximately,
Langlands rich work, its underlying ideas, its enormous impact on the present mathematical research world wide, and its consequence for the future picture of Mathematics. After all, our speaker today with his own words will be the best man for
explaining the main lines of his vision.
Instead, please allow me to add some comments of a more personal nature.
Perhaps I should explain that my own mathematical interests at present are predominantly of historical nature. I am interested in particular in the transmission of
224
8 HasseArfLanglands
ideas from one mathematician to another, and from one generation to the next. Mathematics does not develop just by itself, it is shaped by people. In the process of
history of science the communication between people plays a decisive role, and I find
it fascinating to follow the lines of communication along which mathematical ideas
have developed in the past.
Under this aspect I recently have found a very interesting statement by Professor
Langlands himself. I have found it in an article in which he recollects his impressions
during his first visit to Ankara in the academic year 196768. The article is written
in Turkish language but a friend kindly translated it for me into English. From this
I learned that Langlands year at Ankara was quite decisive for his own research, in
particular through his contact with Cahit Arf. Langlands writes that Arf had pointed
out to him a paper by Helmut Hasse who had proved the first results in the direction
which he, Langlands, was exploring just then. Citing Langlands own words:
Thanks to Cahit bey, I solved this problem during my stay in Ankara and
proved the existence of the local "-factor. Perhaps this theorem would
never have been proved if I had been somewhere else that year
Indeed, I find this is a very interesting statement.
As a young student, Cahit Arf had worked with Hasse in Gttingen to obtain his
doctorate. This was in the years 193738. In this formative period of his life, Arf
became familiar with the mathematical ideas of that time in the circle around Hasse.
Many of those ideas were directed towards the generalization of class field theory,
from abelian to general Galois field extensions.
Perhaps I may be allowed here to use some technical mathematical expressions
whose precise meaning I cannot define here. I hope that in any case I am able to convey
to you that in the 1930s there were strong attempts concerned with a theory, called
class field theory, which had been inherited from many generations of mathematicians
and now was aimed at generalization to a much wider scope than it was known at the
time. In fact, the topic of Arfs doctoral thesis was a direct fall-out of those attempts.
The main result of his thesis, nowadays known as the HasseArf Theorem (more
precisely: Arfs part of the HasseArf theorem), had been envisaged already in 1930
by Artin in a letter to Hasse. Thus Arfs thesis was fulfilling eight years later what
has been called Artins dream.
Helmut Hasse and Emil Artin belonged to the group of leading figures at the
time of the said activities; other names were Emmy Noether, Claude Chevalley and
Shokichi Iyanaga, Andr Weil. As I have mentioned, Cahit Arf was also working on
those projects.
However, despite strong and penetrating attempts from the side of those people
and other mathematicians, the desired general results for Galois extensions could not
be reached in the 1930s, and class field theory remained to be essentially an affair of
abelian extensions only. In a letter from Artin to Hasse dated March 9, 1932 Artin
said:
225
8 HasseArfLanglands
I have the impression that there has to be a completely new idea in order
to obtain the desired general results.
A similar mood can be read from Emmy Noethers letters to Hasse, but she was
somewhat more specific already. On April 26, 1934 she wrote:
It seems to me that first we have to clear up the situation with respect to
the irreducible representations, which means the Galois modules, and
at the same time to Artins conductors.
Even one week before Noethers death in 1935 she wrote that she was still thinking
hard about this problem.
Certainly, Arf had become familiar with this situation when in 1937 he studied
with Hasse; this can be concluded from his letters which are preserved in the library of
Gttingen University. Also, Arf knew that a certain problem connected with the theory
of Artins L-functions, namely the existence of the local "-factor, was considered to
be a step towards further development. But in the 1930s, this and the big problem
of Galois class field theory remained unsolved, with little essential progress in the
following decades.
Finally, 30 years later in 1967 Langlands, after discussing with Arf and after
studying Hasses work, succeeded to have the completely new idea which Artin
had wishfully foreseen in his letter to Hasse. And it was connected with Artins
conductors and with irreducible Galois representations, like Noether had envisaged.
This was a step forward in the development of what is now called the Langlands
program which implies a new form of class field theory valid for Galois extensions,
and which Langlands had outlined in his famous letter to Andr Weil 1967.
H. Hasse
C. Arf
R. Langlands
Let me repeat: The historical moment when the first inspiration for the "-factor
came, was in 1967/68 and, according to Langlands himself, it happened here at the
Middle East Technical University, through the conveyance of Arf who had kept close
226
8 HasseArfLanglands
and friendly relationship to his former teacher Hasse all through his life time, and who
thus was informed about Hasses ideas and work. In this way Langlands has become
part of a long historical line of mathematical development, through Arf directly to
Artin, Hasse and Noether, and from there on far back through Hecke, Takagi and
Hilbert to Abel.
Certainly this will not be the only historical line which can be found leading to
Langlands widespread work. But today, at this occasion, this line through Arf seems
to be of particular significance.
On this background I feel that this years Arf lecture, here at the Middle East
Technical University, will become something very special.
Professor Langlands, on behalf of the Mathematics Department at METU I have
the honor to invite you to start with your address.
Chapter 9
Ernst Steinitz and abstract field theory
In Memoriam Ernst Steinitz.

Journal fr die reine und angewandte Mathematik (Crelles Journal) 658 (2010), 111.
Ernst Steinitz
In the year 1910, in volume 137 of Crelles Journal there appeared a paper with the
title
Algebraische Theorie der Krper (Algebraic Theory of Fields).
The author was Ernst Steinitz. Let us use this occasion of a centenary to recall the
impact which Steinitzs paper had upon the mathematicians of the time, and its role
in the development of todays algebra. Bourbaki [Bou60] states that this paper has
been
228
9 Ernst Steinitz and abstract field theory
un travail fondamental qui peut tre considr comme ayant donn

naissance la conception actuelle de lAlgbre.
a fundamental work which may be considered as the origin of todays
concept of algebra.
One of the first eager readers of Steinitzs paper was Emmy Noether. At the time
when the paper appeared she was still living in her hometown Erlangen, during what
may be called the period of her apprenticeship, studying the highlights of contemporary mathematics of the time. Her guide and mentor in this period was Ernst Fischer.
We can almost be sure that Steinitzs paper was the object of extensive discussions
between Fischer and Emmy Noether. 1
Steinitzs ideas contributed essentially to the shaping of Emmy Noethers concept
of mathematics and in particular of algebra. During the next decade every one of
Noethers papers (except those on mathematical physics) contains a reference to
Steinitz [Ste10].
Later in Gttingen, when she had started her own completely original mathematical path 2 she took Steinitzs results as well known and she used them as the basis
for her work on fields and rings in particular in her project to reformulate classical
algebraic geometry in terms of abstract commutative algebra. She urged her students
and her fellow mathematicians to study the classical papers which she considered
to be the roots of abstract algebra among them was invariably the 1910 paper by
Steinitz. Van der Waerden reports in [vdW75]:
When I came to Gttingen in 1924, a new world opened up before me. I
learned from Emmy Noether that the tools by which my questions could
be handled had already been developed by Dedekind and Weber, by
Hilbert, Lasker and Macaulay, by Steinitz and by Emmy Noether herself.
She told me that I had to study the fundamental paper of E. Steinitz
Van der Waerdens textbook Moderne Algebra [vdW30], which was based on
lectures by Emmy Noether and Artin, contains in his first volume a whole chapter
about Steinitzs theory. Van der Waerden says in [vdW75]:
In earlier treatises, number fields, and fields of algebraic functions were
usually treated in separate chapters, and finite fields in still another chapter. The first to give a unified treatment, starting with an abstract definition of field, was E. Steinitz in his 1910 paper. In my Chapter 5,
called Krpertheorie, I essentially followed Steinitz
Van der Waerdens Moderne Algebra was widely read and translated into many
languages; in this way Steinitzs ideas became known worldwide as part of the basics
1 The mathematical letters between Fischer and Noether are preserved in the archive of the University of
Erlangen. Fischers name is still remembered today from the RieszFischer Theorem in the theory of Hilbert
spaces.
2 Quoted from Alexandrovs obituary for Emmy Noether [Ale83].
229
of contemporary algebra, and they found their way into the syllabus of beginner
courses. Almost all the notions and facts about fields which we teach our students in
such a course, are contained in Steinitzs paper.
But what are those notions and facts? Let us point out first that the main point of
Steinitzs paper was his abstract approach. As Purkert says in his essay on the genesis
of abstract field theory [Pur73]:
Hier wurde erstmalig eine abstrakte algebraische Struktur auf der Grundlage ihres Axiomensystems zum Gegenstand der Untersuchung gemacht.
Dieses formalalgebraische Denken einerseits, die Verbindung mit der
Mengenlehre andererseits das sind die Charakterzge der modernen
strukturellen Algebra.
This was the first time that an abstract algebraic structure was studied on
the basis of its system of axioms. On the one hand the formal algebraic
thinking, on the other hand the connection to set theory these are the
characteristics of the modern algebra of structures.
Hence, when we now describe the content of Steinitzs paper we have to keep
in mind that not only are those definitions and theorems important, but also the fact
that they were obtained in the abstract axiomatic setting, notwithstanding the fact that
some of them had already appeared in earlier treatises for fields of special kinds (fields
of numbers, of functions, and finite fields). Steinitzs paper was the first systematic
investigation of the structure of abstract fields. 3
In his preface Steinitz states that he wishes
eine bersicht ber alle mglichen Krpertypen zu gewinnen und ihre
Beziehungen untereinander in ihren Grundzgen festzustellen.
to obtain a general overview of all possible types of fields and to determine their relations with each other.
In any abstract field, Steinitz showed that there is a unique smallest subfield which
he called the prime field; this is either infinite (and then isomorphic to the rationals) or
its cardinality is a prime number p (and then it is isomorphic to the integers modulo p).
Accordingly he defined the characteristic of a field to be either 0 or p respectively.
Any integral domain determines a unique field of quotients; this is a frequently used
method to construct fields. For any field K and an irreducible polynomial f .X/ 2
KX there is an extension field L containing a root # of f .X/. If L is minimal
with this property then L is uniquely determined up to K-isomorphism. Steinitz
discovered that in prime characteristic an irreducible polynomial may have multiple
3Actually, an earlier article by Heinrich Weber [Web93] also works in the framework of abstract field theory.
Steinitz mentions that in his article. But he points out that Webers investigation is directed to Galois theory only
while his (Steinitz) goal is the systematic investigation of the structure of all fields. As said in [Kle99]: While
Weber defined fields abstractly, Steinitz studied them abstractly.
230
roots, and so one has to distinguish between separable and inseparable algebraic
extensions. 4 Galois theory in this abstract setting holds for separable extensions
only. Steinitz defined the notion of transcendence degree and he showed that every
field can be obtained as an algebraic extension of a purely transcendental field, i.e.,
a field of rational functions over the prime field. Finally, he constructed for every
field an algebraic closure and showed that it is unique up to isomorphism this is
probably Steinitzs most important result. This theorem on the algebraic closure of
any field is sometimes considered to be the proper Fundamental Theorem of Algebra.
All the above has been included by van der Waerden in the first volume of his
textbook Moderne Algebra. This appeared in 1930. The second volume, appearing
one year later, contains the beginning of modern algebraic geometry in the framework
of commutative algebra. In his Heidelberg lecture [vdW97] van der Waerden gives
a lively account of how he discovered the algebraic definition of generic point and
dimension of an algebraic variety and in this connection he again refers to Steinitzs
paper. He says about it:
Die Wichtigkeit dieser Arbeit kann man garnicht berschtzen, das Erscheinen dieser Arbeit war ein Wendepunkt in der Geschichte der Algebra des 20. Jahrhunderts.
One cannot overestimate the importance of this paper. The appearance
of this paper marks a turning point in the history of algebra of the 20th
century.
Immediately after the publication of van der Waerdens book it was rated, by the
referee in Jahrbuch fr die Fortschritte der Mathematik, to be the
Standardwerk der modernen Algebra in der ganzen mathematischen
Welt.
standard treatise of modern algebra in the whole mathematical world.
Van der Waerdens was not the first textbook in which Steinitzs theory was incorporated. Two such textbooks had appeared earlier. One of them was Haupts
two-volume Algebra which appeared in 1929 [Hau29]. This book too had been
written under the influence of Emmy Noether. Otto Haupt held a professorship in
Erlangen since 1921. In his reminiscences [Hau88] 5 he wrote:
Ich kam nach Erlangen als klassisch gebildeter Mathematiker, noch
vllig unberhrt von den damals aufkommenden, als modern bezeichneten neuen Ideen in der Mathematik. In dieser Verfassung machte ich
4 Steinitz speaks of exensions of the first kind and second kind respectively. The terminology separable and
inseparable appears in van der Waerdens Moderne Algebra. It seems probable that this terminology had been
created by Artin. For, van der Waerden reports in [vdW75] that he took lecture notes of Artins course on algebra
in the summer of 1926. He says: In the theory of fields Artin mainly followed Steinitz, and I just worked out
my notes the presentation given in my book is Artins.
5 written when Haupt was 100 years old.
231
die Bekanntschaft der eifrigen Propagandistin der modernen Algebra

Emmy Noether. Auf gemeinsamen Spaziergngen erzhlte E. N. uns von
ihren algebraischen Arbeiten. Ich verstand nicht viel von ihren Erzhlungen und fragte E. N. wie ich zu einem besseren Verstndnis kommen
knne. Sie verwies mich als beste Einfhrung auf die 1910 erschienene
Crellearbeit von Steinitz.
I arrived in Erlangen as a classically educated mathematician, still
untouched by the so-called modern ideas which emerged at that time.
In these circumstances I got to know Emmy Noether 6 , the eager
propagandist for modern algebra. On joint walks she told us about her
algebraic work. I did not comprehend much of what she told us, and I
asked her how to get to a better understanding. She recommended the
Crelle paper of Steinitz which had appeared in 1910.
We can picture the situation: Emmy Noether being an ardent walker, striding
speedily up the Rathausberg near Erlangen and fervently persuading Haupt, who
tries to keep pace with her, not only to understand modern algebra but also to write a
textbook on it. Which he did.
For one year, 19291930, Haupts book was the only source for outsiders to learn
about the modern ideas of algebra. Even in far awayYale, a young student, Saunders
Mac Lane, was told by his teacher Oystein Ore to
read the monograph by Steinitz and the textbook on algebra by Otto
Haupt
As Mac Lane reports [ML81], this happened in the year 1929, one year before
van der Waerdens Moderne Algebra appeared. Today Haupts book is almost forgotten, although it contains some more material of Steinitzs field theory than van der
Waerdens, in particular concerning inseparability phenomena. The fact that van der
Waerdens book was more popular than Haupts in terms of both its number of editions and translations is probably due to the style of writing. Although sometimes it is
pretended that in mathematics only the facts are important, it is a common experience
that even in mathematics the style of writing counts.
Let us turn back to the year 1910 when Steinitzs paper [Ste10] had appeared.
Besides Emmy Noether there was another mathematician who was keenly interested
in this paper, namely Kurt Hensel in Marburg. For Steinitz had mentioned Hensels
p-adic fields in a footnote of the introduction:
Zu diesen allgemeinen Untersuchungen wurde ich besonders durch Hensels Theorie der algebraischen Zahlen angeregt, in welcher der Krper
der p-adischen Zahlen den Ausgangspunkt bildet, ein Krper, der weder
6At that time Emmy Noether was living in Gttingen but she used to visit her home town Erlangen from time
to time.
232
den Funktionenkrpern noch den Zahlkrpern im gewhnlichen Sinne

beizurechnen ist.
I was inspired to these general investigations by Hensels Theory of
Algebraic Numbers, in which the field of p-adic numbers is used as the
starting point. Such a field cannot be counted among the function fields
or number fields in the ordinary sense.
Here, Steinitz refers to Hensels book [Hen08] which had appeared in 1908.
In the year 1912 Hensel attended the 5th International Congress of Mathematicians
(ICM) in Cambridge, England. There he met the Hungarian mathematician Josef
Krschak who gave a talk with the title:
ber Limesbildung und allgemeine Krpertheorie.
On the concept of limit and general field theory.
Krschak worked in the framework of Steinitzs abstract field theory. He defined
what today is called a valuation of a field K, as a map a 7! jaj from K to the
real numbers, with the standard properties. He showed that Cantors method, which
Cantor had used for the construction of the reals by means of Cauchy sequences of
rationals, works for any valuation of an abstract field in the sense of Steinitz. In this
way he constructed the completion of a valued field. As an application this gives a
construction of Hensels p-adic field as the completion of the p-adic valuation of the
rationals. This method is standard today.
Moreover, Krschak showed that the algebraic closure (in the sense of Steinitz)
of a complete field carries a unique valuation extending the valuation of the base
field. In other words, he proved that a complete field is Henselian in contemporary
terminology. In this abstract setting Krschak used and proved what is now called
Hensels Lemma. It seems remarkable that his proof works at the same time for both
archimedean and nonarchimedean valuations.
Kurt Hensel was impressed by this and he took Krschaks paper for publication
in Crelles Journal of which he (Hensel) was the chief editor. The paper [Kr13]
appeared in volume 142 as kind of follow-up to Steinitzs paper [Ste10] which had
appeared in volume 137.
Krschak did not publish any further paper on valuation theory. But there was
a young mathematician in Marburg with Hensel who was eager to take over and to
extend, generalize and simplify the SteinitzKrschak theory. This was Alexander
Ostrowski. 7
Ostrowski was a brilliant young mathematician. He had started his studies in Kiev
but was sent by his academic teacher Grave to Germany since, being of Jewish origin,
Ostrowski seemed not to have much chance in the academic world in Russia at that
7 Hensel himself, although interested in Krschaks work, stayed on the traditional side. He continued to use
his own construction of p-adic fields, based on p-adic power series.
233
time. Ostrowski arrived in Marburg in 1911 when he was 18 years old. He soon was
engrossed in valuation theory in the sense of SteinitzKrschak. His contributions
during the years 19131918 shaped valuation theory into essentially the form we use
today. One of the interested readers of Ostrowskis papers was Emmy Noether, who
in the year 1916 started a correspondence with him. 8
As a somewhat bizarre story let us mention that Ostrowski, being a citizen of
Russia, had been interned in Marburg during World War I when Germany was at
war with Russia. But Hensel could persuade the authorities to permit Ostrowski to
use the University Library for his studies during the day. Hence for more than three
years Ostrowski sat daily in the reading room of Marburg University Library. It was
there where he wrote his seminal papers on valuation theory which were published
in Crelles Journal, Acta Mathematica and Mathematische Annalen. He also wrote
his big monograph on valuation theory [Ost34] which was completed in 1916 but
appeared only in 1934 not as a book but in three parts in the journal Mathematische
Zeitschrift. Van der Waerden says in [vdW97]:
Ostrowski setzte der Bewertungstheorie mit seinen groen Abhandlungen in der Mathematischen Zeitschrift, Band 39, die Krone auf. Die
Darstellung dieser Theorie im zweiten Band meiner Algebra beruht ganz
auf dem Werk von Ostrowski.
Ostrowski crowned valuation theory with his great papers in the Mathematische Zeitschrift, volume 39. The presentation of this theory in the
second volume of my Algebra is completely based on Ostrowski. 9
After the war, in the year 1918, Ostrowski left Marburg for Gttingen. He did not
publish any other paper on valuation theory. (But in his paper [Ost33] on Dirichlet
series he used the results of SteinitzKrschak and himself.)
It was some years later, in 1921, that the young student Helmut Hasse arrived
in Marburg from Gttingen in order to study p-adic fields. There Hasse learned
about the work of SteinitzKrschakOstrowski. In the course of many years Hasse
became, in the words of van der Waerden [vdW75],
Hensels best and a great propagandist of p-adic methods.
In the fall of 1922 Hasse went to the University of Kiel as Privatdozent; he stayed
there until 1925. Since 1920 Steinitz had held a professorship in Kiel; thus Hasse and
Steinitz were at the same university during those years. I did not find any information
about whether Hasse and Steinitz had mathematical discussions or joint work during
8 For
more details, see [Roq02].

the same paragraph in [vdW97] van der Waerden says that valuation theory was started by Rychlik. But
there seems to have been some mix-up. As said above, valuation theory had been inititiated by Krschak in
1912 on the basis of Steinitzs paper. Rychlik has given some contributions to it, beginning in 1919, and he cites
Krschak and Ostrowski.
9 In
234
this time. Steinitz was known as der groe Schweiger 10 (the great silent man)
which meant that it was not easy to come into close contact with him. But certainly
both had met. Frei [Fre77] reports that there was only one office in the mathematics
department, and this had to be shared by all staff : by the two professors Steinitz and
Toeplitz and the two Privatdozenten Hasse and Robert Schmidt.
In the academic year 1924/25 Hasse gave a course on Hhere Algebra. In the
second part he presented field theory in the form of Steinitzs paper [Ste10]. Hasses
notes from this course became the basis for his two-volume textbook Hhere Algebra
(Higher Algebra). The book appeared in the Gschen textbook series, which at that
time was well known among the German-speaking mathematicians. Its second part
appeared 1927. (This was two years before the publication of Haupts book mentioned
above.) The referee (R. Brauer) of this part stated:
Der zweite Band behandelt die Theorie der Gleichungen hheren Grades,
er schliet sich an die Arbeit von Steinitz (1910) an
The second volume covers the theory of equations of higher degree. Its
concept follows that of Steinitzs paper (1910)
In the correspondence between Hasse and Emmy Noether [LR06] one can see that
the latter sent Hasse some advice for his algebra book. But this concerned certain
details of proof only, e.g., the primitive element theorem. As said above, Hasse
devised the concept of the book during his years in Kiel, independent of Emmy
Noether.
Hasses estimate of Steinitzs paper can be seen from a footnote in the second
volume of his algebra textbook where he said:
In diese grundlegende Originalarbeit zur Krpertheorie sollte jeder Algebraiker einmal hineingesehen haben.
Every algebraist should have read at least once this basic original paper
on field theory.
In order to facilitate this, Hasse had Steinitzs paper reprinted in book form,
together with comments and an appendix by Baer [Ste30]. This happened in the
years 1928/29 when Hasse held a professorship at the University of Halle. Reinhold
Baer was Privatdozent there. In the preface of the book the editors Baer and Hasse
praise the text as a
klassisch schne, formvollendete und in allen Einzelheiten durchgefhrte Darstellung
classically beautiful, perfectly structured exposition taking care of every
detail
10 Quoted
from Haupts reminiscences [Hau88].
235
And they continue:

auch heute noch ist die Steinitzsche Arbeit eine vortrefiche, ja geradezu unentbehrliche Einfhrung fr jeden, der sich auf dem Gebiet der
neueren Algebra eingehenden Studien hingeben will.
still today Steinitzs paper is an excellent and in fact indispensable
introduction for everybody who wishes to study modern algebra more
extensively.
The appendix by Baer contains a detailed presentation of Galois theory, which
was not explicitly covered by Steinitz.
The comments by the two editors concern those proofs in Steinitzs paper which
use Zermelos well-ordering theorem and the principle of transfinite induction. These
were necessary to treat infinite algebraic extensions, or fields with infinite degree of
transcendency over their prime field. Steinitz was well aware that this depends on
the axiom of choice which at that time was not generally accepted. He said:
Das Auswahlprinzip erscheint auch unvermeidlich, wenn man den Beweis der Existenz einer algebraisch abgeschlossenen Erweiterung fr
jeden beliebigen Krper fhren will Noch stehen viele Mathematiker
dem Auswahlprinzip ablehnened gegenber. Mit der zunehmenenden
Erkenntnis, dass es Fragen der Mathematik gibt, die ohne dieses Prinzip
nicht entschieden werden knnen, drfte der Widerstand gegen dasselbe
mehr und mehr schwinden
Also, the axiom of choice seems to be unavoidable if one wishes to
prove the existence of an algebraically closed extension of an arbitrary
field 11 Many mathematicians still object to the use of the axiom of
choice. This resistance against using the axiom of choice will dwindle
with the realization that there are mathematical problems which cannot
be decided without this axiom
But those proofs in Steinitzs paper which use the principle of transfinite induction
were somewhat long-winded and therefore the comments of Baer and Hasse try to
simplify and streamline Steinitzs arguments. Today we would prefer to use Zorns
Lemma instead; this would lead to a still greater simplification combined with a
considerable shortening of Steinitzs paper (which originally had 134 pages). But
Zorns Lemma was not yet formulated in 1929.
The HasseBaer edition of Steinitzs paper was reprinted 1950 by the Chelsea
Publishing Company in New York, in its series reprinting classical treatises. (In 1997
the American Mathematical Society acquired Chelsea. This title is not listed anymore
11 But note that Banaschewski proved in 1992 that the existence and uniqueness of the algebraic closure can
be derived from the Boolean Ultrafilter Theorem already, which is weaker than the axiom of choice [Ban92].
236
as being available.) I had bought a copy myself in the 1950s and followed Hasses
advice to read this classic work.
Let us return to the year 1924. In that year the Hamburger Abhandlungen published the paper by Artin and Schreier: Algebraische Konstruktion reeller Krper
(Algebraic construction of real fields). This paper is in some sense a companion to
Krschaks: whereas Krschak developed the notion of valued field on the basis of
Steinitzs abstract field theory, Artin and Schreier do the same for the notion of ordered
field. They construct the real closure of an ordered field with the help of Zermelos
well ordering theorem similar to what Steinitz had done for the construction of the
algebraic closure. In fact, they cite Steinitz in connection with some details of this
proof. And in the introduction the authors say:
E. Steinitz hat durch seine Algebraische Theorie der Krper weite
Teile der Algebra einer abstrakten Behandlungsweise erschlossen; seiner
bahnbrechenden Untersuchung ist zum groen Teil die starke Entwicklung zu danken, die seither die moderne Algebra genommen hat
E. Steinitz, through his Algebraic Theory of Fields, has opened up
large parts of algebra to an abstract treatment; since then, thanks to his
groundbreaking work, modern algebra has seen a strong revival
Nowadays we do not often find such enthusiastic references to Steinitz [Ste10] in
current mathematical papers. The reason for this is that the main ideas and results of
Steinitz have become a matter of course, not the least through the early textbooks by
Hasse, Haupt and van der Waerden mentioned above.
At the end of his introduction Steinitz says in his paper of 1910:
Der vorliegende Aufsatz behandelt nur die Grundzge einer allgemeinen
Krpertheorie. Weitergehende Untersuchungen sowie Anwendungen auf
Geometrie, Zahlen- und Funktionentheorie beabsichtige ich, in einigen
weiteren Anhandlungen folgen zu lassen.
The present article covers the foundations of a general field theory only.
I am planning to follow-up with more advanced investigations, and with
applications to Geometry, Number Theory and the Theory of Functions.
But Steinitz did not publish anything in this direction, and also in his literary estate
nothing of this kind was found. We do not know why he did not later write what he had
announced. In any case, we observe that his 1910 paper has exerted a great influence
upon the delopment of modern algebra during the next decades. As evidence for this
we may cite the following papers which are to be considered as immediate follow-ups
to [Ste10], and have each opened up a long line of development:
Analysis, including p-adic analysis: KrschakOstrowski on valuation theory [Kr13], [Ost34];
237
Number Theory: Emmy Noether on Dedekind domains in abstract fields

[Noe26];
Algebraic Geometry: Emmy Noether on primary decomposition of ideals in a
Noetherian ring [Noe21], and van der Waerden on the foundations of algebraic
geometry [vdW75];
Real Algebra: Artin and Schreier on real fields [AS27a], [AS27b].
This, of course, is not a complete list. 12 For a biography of Steinitz we refer to
the Dictionary of Scientific Biography.
Ackowledgment. I am indebted to Keith Conrad for helpful critical comments, and
for streamlining my English.
12 Compare
with [Pur73] and [Kle99].
Chapter 10
Heinrich-Wolfgang Leopoldt
Obituary.
Journal of Number Theory 132 (2012), 16411644.
H. W. Leopoldt
Heinrich-Wolfgang Leopoldt, member of the editorial board of Journal of Number

Theory from 1969 to 1987, has passed away on July 28, 2011 at the age of almost
84 years. His name will be remembered as the discoverer of the p-adic L-functions
( jointly with T. Kubota) and as the author of the Leopoldt conjecture.
Leopoldt was born and raised in the small German town of Schwerin at the Baltic
sea. Still a school boy he was drafted in war time to various military services. After
the war he started an apprenticeship in view of the uncertain prospects of the future.
However there was a former mathematics teacher with whom Leopoldt played music,
and who introduced him also to astronomy and its mathematical fundamentals. He
suggested to Leopoldt to finish his Gymnasium (high school) in order to be able to
attend university. Leopoldt did so, and in the academic year 1947/48 he started his
mathematical studies at the Humboldt University in Berlin.
240
One of the first lecture courses he attended was an introduction to number theory
by Helmut Hasse. In his Erinnerungen (memoirs) Leopoldt reports that this lecture
had left a deep impression upon him, in particular Hasses remarks on the relationship
between beauty and truth in mathematics, combined with various instances of parallels
between number theory and music. This lecture, Leopoldt recalls, led him to study
number theory with highest priority. We can observe this dedication to number theory
throughout his mathematical work.
In the year 1950 Hasse left Berlin for Hamburg and his student Leopoldt followed
him there. 1954 he obtained his Ph.D. at Hamburg University. Thereafter he got a
position as assistant professor in Erlangen until 1962, interrupted for two years at
the Institute for Advanced Study in Princeton. 1959 he made his Habilitation, i.e.,
his qualifying examination as a university teacher. 19621964 he taught at Tbingen
University, again interrupted by a visiting professorship at Johns Hopkins University
in Baltimore. In the year 1964 he obtained an offer for a permanent position at Johns
Hopkins and another one at Karlsruhe University. He accepted the latter.
Leopoldts mathematical work is governed by the attempt to systematically open
up the structure of abelian fields such that one can work in them freely in the same
way as one can work in quadratic fields. This challenge had been formulated by
Hasse in a monograph 1952, and his student Leopoldt set about to perform this task
step by step. Here I cannot give an assessment of Leopoldts complete work. But
while I am writing this obituary I remember particularly three of his most brilliant
achievements which had impressed me at the time and I would like to share my
recollections here.
First I remember his paper on the structure of the ring of integers in an abelian field
K which was his Habilitationsschrift at Erlangen [Leo58]. If K is tamely ramified
then Emmy Noether had shown that there exists a integral normal base, i.e., the ring
of integers I of K is G-module isomorphic to the group ring QG, where G denotes
the Galois group of K. However, if the ramification is not tame then there does not
exist a normal integral base and it was not known how Noethers theorem should
be be modified in order to describe the G-structure of the integers. But Leopoldts
paper solves this problem, first in an abstract setting showing that I is G-moduleisomorphic to a certain order of QG, and then interpreting this in a very explicit and
concrete manner by constructing explitly a normal basis with the help of the Gaussian
sums .
/ belonging the characters
of G.
Secondly I recall his brilliant paper, composed in Princeton, on the Spiegelungssatz in an absolute Galois number field K [Leo58]. This concerns the structure of the
class group C of K, more precisely: the structure of its p-part Cp where p is a fixed
prime. Assume that the degree of K is prime to p and that the p-th roots of unity are
in K. The action of the Galois group G of K induces a direct decomposition of Cp
into subgroups C' where ' ranges over the p-adically irreducible characters of G.
Let 0 denote the character belonging to the representation of G on the p-th roots of
241
unity. Then there is what Leopoldt calls the reflection map ' ! 'x with
'.
x /D
0 ./'.
1
.for 2 G/:
This construction of 'x had arisen in various situations before but Leopoldt had discovered a relationship between the '-part C' and the '-part
x
C'x . In some sense both
are of about the same size. More precisely, if e' , e'x denote the G-ranks of C' , C'x
respectively then
'x e'x e' ' :
where the bounds ' , 'x depend in a well defined manner on the p-adic Galois
structure of the group of units E of K. Hence the study of the Galois structure of the
units of K leads to information about the deviation of the rank of C' from that of its
mirror image C'x .
Leopoldts proof of this theorem uses class field theory; it is quite lucid and
straightforward. It did not use any newly developed method and could have been
proved much earlier. The essential new ingredient is the way of looking at the various
objects from the structural point of view. Instead of dealing with class numbers
Leopoldt deals with the class groups and their Galois structure. This point of view
had been advocated by Hasse in his monograph cited above, and Leopoldt now uses
it competently to advance a big step towards the understanding of the class groups of
number fields.
Nowadays this point of view, i.e., the investigation of class groups by means of Galois action, has become standard in algebraic number theory, as it is manifested, e.g.,
in what is called the Iwasawa theory. At the time of publication the Spiegelungssatz
won widespread interest among number theorists not only because it offered a common background for classically known results about divisibility properties of class
numbers, but also since it admitted to obtain much more information of this kind.
Whereas classically (since Kummer) the properties of class numbers are connected
with the Bernoulli numbers B n , Leopoldts new results referred to his generalized
Bernoulli numbers B
n belonging to Dirichlet characters
. They are defined recursively in a similar way as are the ordinary Bernoulli numbers. Leopoldt showed that
his B
m , which are contained in the cyclotomic field Q.
/, occur as the values of the
Dirichlet L-functions L.s;
/ at the odd negative integers:
L.1 m;
/ D
B
m
:
m
These relations lead to the third oustanding result of Leopoldt which I would
like to recall, namely the discovery (jointly with Kubota) of the p-adic L-functions
[KL64]. They belong to the standard tools of todays algebraic number theory but
perhaps it is good to recall that they are due to Leopoldt. The story is as follows.
For any prime p his generalized Bernoulli numbers B
m satisfy certain p-adic
congruences, the so-called Kummer congruences. Leopoldt observed that those conBm
gruences can be interpreted such that the m are in a sense p-adically continuous
242
functions of m. More precisely: There is one and only one p-adically continuous
function Lp .s;
/ defined on Zp such that
Lp .1 m;
/ D
B
m
m
.1
.p/p m1 /
for the negative integers 1 m with m 0 mod p 1. 1 These numbers 1 m are

dense in Zp . But it turns out that Lp .s;
/ is holomorphic in a region which is larger
than Zp , at least if
1 whereas for
D 1 there is one pole for s D 1.
Based on these p-adic L-functions Leopoldt considered for any abelian number
field K the corresponding p-adic zeta function K; p .s/ as the product of the Lp .s;
/
for the characters
of K. He arrives at p-adic class number formulas which are the
analogue of the ordinary class number formulas for abelian fields K, and they look
quite similar but the terms have to be interpreted in the p-adic sense.
However there was one obstacle to overcome, namely the proof of non-vanishing
of the so-called p-adic regulator of an abelian field which appears in those formulas.
This regulator RK; p is obtained if one replaces the ordinary logarithms in the classical
regulator by the p-adic logarithms. The non-vanishing of RK; p means that the p-adic
rank of the group of units of K equals its ordinary rank, i.e., r1 C r2 1 where r1 , r2
are the numbers of real or complex infnite primes respectively.
Leopoldt had tried hard to prove this conjecture. He delayed the publication of the
second part of his L-series paper [Leo75] since he still hoped to be able to include a
proof of it. Nevertheless he presented an exposition of his theory when he lectured at
Johns Hopkins in the year 1964. In this form the theory somehow made it to Princeton where Iwasawa included it in his Lectures on p-adic L-functions. This made
Leopoldts theory widely known and started an extended research in consequence
of which p-adic L-series became indispensable tools of number theorists. Brumer
succeeded 1967 to prove Leopoldts conjecture for arbitrary abelian fields. 2
A characteristic feature of Leopoldts work is that he aims at concrete results given
by explicit and effective formulas. He competently uses and investigates abstract
structures, but such considerations serve him as motivation only, as guideline towards
the goal of explicit algorithms. To a large degree his results were obtained by extended
numerical computations.
This attitude led him quite early to develop computer programs for the use in
algebraic number theory. His team in Karlsruhe was one of the first in Germany which
systematically developed the necessary algorithms for this project, also in cooperation
with Hans Zassenhaus. A number of Leopoldts students are now working in scientific
computing and so follow up his ideas.
Leopoldts personality can be characterized as quiet, unassuming, always willing
to hold back his own in favor of supporting the cause. His firm and objective counsel
1 For
simplicity we assume here that

is an even character and that p > 2.
Mihailescu has announced a proof of Leopoldts conjecture for all number fields.
2 Recently
243
in scientific matters, always to the point, was valued by all who had to deal with him.
His lectures stood out by their clarity and intensity. He was known as a master in
exposition. The list of his publications is not large but his work belongs to the pearls
of mathematical research in the last century.
Leopoldt was married and has five children. After his retirement from the University of Karlsruhe he had moved to a small village in Northern Germany where he
dedicated himself to his beloved piano music.
Chapter 11
On Hoechsmanns Theorem
On the local-global principle for embedding problems over global fields.

Comments to an old paper of Klaus Hoechsmann.
Israel Journal of Mathematics 141 (2004), 369379.
11.1
11.2
11.3
11.4
11.5
Introduction
Statement of the result
The setting
Hoechsmanns theorem
Construction of counter examples
245
245
247
249
252
11.1 Introduction
The origin of this note was the attempt to answer a question of Moshe Jarden who
had asked me:
Hat jeder Zahlkrper ein endliches Einbettungsproblem, das lokal berall lsbar aber global nicht lsbar ist?
Does every number field admit a finite embedding problem which is
locally solvable everywhere but not globally solvable?
As a first reaction I referred him to an old paper of Hoechsmann [Hoe67] on the
embedding problem. But after a second reading of Hoechsmanns paper I found that
the answer to Moshes question which is affirmative is not explicitly stated there.
Although the answer can be readily derived using Hoechsmanns ideas, it is perhaps
not without interest to do this explicitly. This is what I propose in this note.
11.2 Statement of the result

Let K be a global field and GK its absolute Galois group. Let A be a finite GK module. The action of GK on A factors through a finite factor group. Let G be such
246
a factor group, i.e. G D GK =U where U is an open normal subgroup of GK which

acts trivially on A. We consider embedding problems of the form
GK
/A
/E
}|
(11.1)
'
/G
/1
1
where ' is the natural projection, and where E is a group extension of A with G.
Such group extensions correspond to the cohomology classes " 2 H 2 .G; A/. We
are looking for solutions of the embedding problem. We do not require that be
surjective. But note that for a global field, it is well known that the existence of any
solution, surjective or not, implies the existence of a surjective solution. (A proof can
be found in Hoechsmanns paper.)
If p is a prime of K then Kp denotes its completion. The absolute Galois group
GKp is considered as a subgroup of GK , viz. the decomposition group of an extension
of p to the algebraic closure (unique up to conjugation). Let Gp D '.GKp / denote
the decomposition group of p in G.
Given an embedding problem (11.1) its localization at p is
GKp
/A
/ Ep
}{
(11.2)

/ Gp
/1
1
where Ep is the inverse image of Gp under the map E ! G. The factor system of
this localization is the restriction ResGp ."/ of the factor system " of (11.1).
The solvability of (11.1) implies the solvability of (11.2) for each p. The LocalGlobal Principle LGP.A; K/ asserts that conversely, if an embedding problem (11.1)
is locally solvable for each p then it is globally solvable. For a given GK -module A
the Local-Global Principle may hold or may not hold. Our result in this note is
Theorem 6. For every global field K of characteristic 2 and any given m 3 there
exists a GK -module A of order 2m such that the Local-Global Principle LGP.A; K/
does not hold.
Remark. The modules A to be constructed will be cyclic groups of order 2m with
the action of GK defined suitably. If 2 is replaced by a prime number p > 2 then
247
the situation is completely different. For, if GK acts on a cyclic group A of order p m

with p > 2 then the Local-Global Principle LGP.A; K/ does hold (irrespective of
the characteristic of the field K). This is a consequence of Gudrun Beyers theorem.
(See Corollary 11 below.) The exceptional role of the prime 2 in this context is a
consequence of the difference in the structure of the automorphism group of cyclic
groups of p-power order p m . If p > 2 then the automorphism group is cyclic whereas
if p D 2 this is not the case for m 3. In this respect the situation here is similar to
the situation of the GrunwaldWang theorem. (See [LR03].)
Concerning the characteristic hypothesis in Theorem 6, this is necessary if one
wishes to construct counter examples to the Local-Global Principle by means of
cyclic groups A, as we do in this paper. If K is of characteristic 2 and A is a cyclic
group of 2-power order with any action of GK then the LGP.A; K/ holds. This
is a consequence of Witts theorem that for a global field K of characteristic 2 the
maximal pro-2-factor group of GK is free in characteristic 2 (and similarly for any
non-zero characteristic). I do not know whether non-cyclic groups A can serve as
counter examples to the Local-Global Principle.
11.3 The setting

Let me first recall some of the results in Hoechsmanns paper.
The solvable embedding problems (11.1) form a subgroup of H 2 .G; A/, and this
is precisely the kernel of the inflation map
inf W H 2 .G; A/ ! H 2 .GK ; A/:
(11.3)
(Note that the inflation map is well defined since the kernel of GK ! G acts trivially
on A.) This holds for any base field K, hence also for the localizations. Now, every
element in H 2 .GK ; A/ is the inflation of some element in H 2 .G; A/ for a suitable
finite factor group G. We conclude:
Proposition 7. The Local-Global Principle LGP.A; K/ holds if and only if the map
H 2 .GK ; A/
Q
p
H 2 .GKp ; A/
(11.4)
is injective.
At this point Hoechsmann cites the duality theorem of TatePoitou for global
fields. That duality theorem holds if the order of A is relatively prime to the characteristic of K (including the case of characteristic 0) which we assume henceforth.
Let Ay denote the dual GK -module of A. It consists of the characters
of A, i.e.,
the homomorphisms of A into the multiplicative group of the algebraic closure of K.
248
The action of GK on Ay is given by

1
.a/ D
.a /
.a 2 A; 2 GK /:
(11.5)
Note that in this formula acts twofold: First 1 acts on A since A is a GK -module.
Secondly, acts on the character values since is an automorphism of the algebraic
y
closure of K. In Hasses terminology, this is a crossed action of GK on A.
Now, the TatePoitou duality theorem asserts that for a global field K, the map h
in (11.4) is dual to the following map:
y
H 1 .GK ; A/
Q
/
y
H 1 .GKp ; A/:
(11.6)
In particular, h is injective if and only if j is injective. We obtain:

Corollary 8. The Local-Global Principle LGP.A; K/ holds if and only if the map j
in (11.6) is injective.
By this result, the problem is transferred from cohomological dimension 2 to
dimension 1. This is the starting point of Hoechsmann. First he reduces the problem
to a finite factor group of GK .
y i.e., the factor group
Proposition 9. Let G be the action group of the GK -module A,
y
of GK modulo the normal subgroup which fixes A elementwise. Then LGP.A; K/
holds if and only if the map
y
H 1 .G; A/
jG
Q
p
y
H 1 .Gp ; A/
(11.7)
is injective. 1
Here, Gp denotes the decomposition group of p in G, i.e., the image of GKp in G.
y In
Proof. (i) First we consider the case when G D 1, i.e. GK acts trivially on A.
this case it is asserted that the LGP.A; K/ holds, i.e., that the map j in (11.6) is
y D Hom.GK ; A/.
y Every
injective. Now, in case of trivial action we have H 1 .GK ; A/
y
x of GK .
homomorphism f W GK ! A factors through a finite, abelian factor group G
x
Let N 2 G. Using Chebotarevs density theorem we conclude that there exists a
prime p of K whose decomposition group contains .
N Hence, if f vanishes on all
decomposition groups then f .N / D 0. Since this holds for all N we conclude f D 0.
(ii) Now consider the general case. Let L be the finite Galois extension of K
corresponding to G, so that G is the Galois group of LjK. Consider the commutative
1 This
proposition and the following corollaries remain valid for any finite factor group G of GK modulo a
y
normal subgroup which acts trivially on A.
249
diagram
/ H 1 .G; A/
y
Q
p
/ H 1 .G ; A/
y
K
inf
jG
y
H 1 .Gp ; A/
Q
p
/ H 1 .G ; A/
y
L
Res
y
H 1 .Gp ; A/
Res
Q
p
jL
(11.8)
y
H 1 .GL;p ; A/
with self-explaining notations. The rows are exact. The vertical arrow jL on the right
y Consequently, if the arrow jG
hand side is injective by (i), for GL acts trivially on A.
on the left hand side is injective then j in the middle is injective too, and conversely.
y If the
Corollary 10. As in Proposition 9 let G denote the action group of GK on A.
group indices G W Gp of the decomposition groups have greatest common divisor 1
then LG.A; K/ holds.
y If c vanishes at p, i.e., if ResGp .c/ D 0 then it follows
For, let c 2 H 1 .G; A/.
G W Gp c D 0. If this holds for all p then c D 0, provided the indices G W Gp
have greatest common divisor 1.
Corollary 11. If the action group G of GK on Ay is cyclic then LGP.A; K/ holds.
For, if G is cyclic then by Chebotarevs density theorem there exists p with
Gp D G.
Corollary 11 is the theorem of Gudrun Beyer. It is remarkable that the validity
y not on A itself. This has
of LGP.A; K/ depends on the action of GK on the dual A,
been discovered by Gudrun Beyer. For Corollary 10 Hoechsmann cites Demukin
and afarevic.
11.4 Hoechsmanns theorem

From now on we assume that A is a cyclic group. After decomposing A into its Sylow
components we may assume that the order of A is a prime power, jAj D p m . Its dual
y D p m too. If p > 2 then the automorphism group of
Ay is also a cyclic group and jAj
Ay is cyclic and it follows that G is cyclic, hence LGP.A; K/ holds by Gudrun Beyers
theorem (Corollary 11).
Consequently, in looking for a counter example to LGP.A; K/ we have to take
p D 2. (This implies that K is of characteristic 2 since the order of A is supposed
to be relatively prime to the characteristic of K.) The GK -module A should be a cyclic
group such that the action group G on Ay is non-cyclic. In particular m 3. If there
exists a prime p of K with Gp D G then by corollary 10 we have that LGP.A; K/
holds. We conclude:
250
Let A be a GK -module which is a cyclic group of prime power order p m . If the

Local-Global Principle LG.A; K/ does not hold then the following conditions are
satisfied:
1. p D 2.
2. The action group G of GK on Ay is non-cyclic, hence m 3.
3. For every prime p of K, the decomposition group Gp is a proper subgroup
of G.
Now we can formulate Hoechsmanns theorem:
Theorem 12. The conditions 13 above are not only necessary but also sufficient for
A to be a counter example to LGP.A; K/.
In view of Proposition 9 this is an immediate consequence of the following group
y
theoretical observation. For simplicity we write X instead of A.
Lemma 13. Let X be a cyclic group of order 2 m .m 3/ and G a non-cyclic group
of automorphisms of X. Then there exists 0 c 2 H 1 .G; X/ such that its restriction
ResH .c/ vanishes for every maximal subgroup H G.
Proof. We identify X D Z=2 m (additively) and G with a group of units in .Z=2 m / .
The action of G on X is given by multiplication. Any element in H 1 .G; X/ can be
represented by a crossed homomorphism f W G ! X. The functional equation of a
crossed homomorphism is
f . / D f ./ C f ./
for ; 2 G:
(11.9)
In particular, for D we note that

f . 2 / D . C 1/f ./:
(11.10)
We shall prove the lemma by explicitly exhibiting a crossed homomorphism f representing c.

The non-cyclic group G is a direct product
G D h1i
hui
where u 1 is a certain unit of X which can be assumed to be u 1 mod 4. (If
this should not be the case then we replace u by u.) Let k be the exact exponent by
which 2 appears in u 1, so that
u 1 D 2 k
where is not divisible by 2, hence a unit in X. We have
2 k m 1:
251
(If k would be m then u 1 mod 2m , contradicting the fact that u 1 as operator

on X .) The group theoretical meaning of k is the following:
The group 2mk X consists precisely of those elements of X which are fixed by u.
For, the relation ux x mod 2m is equivalent to .u 1/x 0 mod 2m which,
by definition of k, means x 0 mod 2mk .
Every crossed homomorphism f W G ! X is already determined by its values
on the generators 1 and u of G. We claim that there is a crossed homomorphism f
with the values
(11.11)
f .1/ D 2mk ; f .u/ D 0
and that its class c 2 H 2 .G; X/ satisfies the requirements of the lemma.
First we consider the subgroup h1i of G of order two. Consider the function
f0 W h1i ! X given by the values f0 .1/ D 2mk ; f0 .1/ D 0. This is a crossed
homomorphism. To verify this one has to check the validity of (11.10) for D 1
only. Indeed, we have
f0 ..1/2 / D .1 C 1/2mk D 0 D f0 .1/:
We have the exact sequence
1 ! hui ! G ! h1i ! 1:
As observed above, the value f0 .1/ D 2mk is fixed by u. Hence we may extend
f0 W h1i ! X by inflation to a crossed homomorphism f W G ! X such that
its values f ./ depend on the residue class of modulo hui only. This crossed
homomorphism satisfies (11.11).
Let c 2 H 1 .G; X/ denote the class of f . We claim that the restriction of c to
every maximal subgroup of G vanishes. There are three maximal subgroups of G,
namely the two cyclic groups hui and hui, and the group h1; u2 i which in general
is not cyclic except if u2 D 1 (which means that k D m 1).
The restriction of c to hui vanishes since f .u/ D 0 by (11.11).
As to the restriction of c to hui we first note that f .u/ D f .1/ D 2mk
does not vanish. But consider a crossed homomorphism g W G ! X belonging to the
same class c as f , which means that
g. / D f . / C . 1/x
. 2 G/
(11.12)
for some x 2 X. Can we choose x 2 X such that g.u/ D 0 ? This means

f .u/ D 2mk D .u 1/x D .u C 1/x:
Since u 1 mod 4 we have u C 1 2 mod 4 hence u C 1 D 2 with a unit in
X . Hence by choosing x D 1 2mk1 we indeed have g.u/ D 0.
252
Can we choose x such that g vanishes on the third maximal group h1; u2 i ? This
means, firstly, g.1/ D 0 and thus
f .1/ D 2mk D .1 1/x D 2x
(11.13)
and so we take x D 2mk1 . Secondly, the condition g.u2 / D 0 requires that

f .u2 / D 0 D .u2 1/x D .u 1/.u C 1/x D 2kC1 x:
The same x D 2mk1 as above satisfies this condition since 2m x D 0.
We have now shown that c vanishes if restricted to any of the three maximal
subgroups of G. It remains to verify that c 0 in H 1 .G; X/. In other words: It
is not possible to choose x 2 X such that g.1/ D g.u/ D 0. Now the condition
g.1/ D 0 implies by (11.13) that x is precisely divisible by 2mk1 (and not by a
higher power of 2). On the other hand, the condition g.u/ D 0 requires that
f .u/ D 0 D .u 1/x D 2k x
and hence x should be divisible by 2mk . Both these conditions are not compatible,
and so c 0.
11.5 Construction of counter examples

In the following we let A be a cyclic group of order 2m with m 3. We try to define
a non-cyclic action of GK on A such that condition 3 of Theorem 12 is satisfied. This
will give a counter example to LGP.A; K/. The main tool for this is the following
Lemma 14. For any global field K there exists an abelian extension LjK of prescribed 2-power degree 2rC1 whose Galois group G D Gal.LjK/ has the structure
G Z=2
Z=2r ;
and such that for every prime p of K its decomposition group Gp is a proper subgroup
of G.
There are many possibilities to construct such a field extension. First assume that
K is a number field. Consider the field K .2/ of 2-power roots p
of unity over K. Its
Galois group is either a free cyclic pro-2-group (for instance if 1 2 K) or else it
is the direct product of such a group with a group of order 2. In any case the Galois
group of K .2/ jK contains finite cyclic factor groups of arbitrary large 2-power order.
Accordingly let L0 jK be a cyclic extension of degree 2r which is contained in K .2/ .
We observe that the only primes p of K which are ramified in L0 (if there are any)
are divisors of 2. This follows from the fact that 2 is the only prime number in Q
which is ramified in Q.2/ .
253
Now we take a rational prime number p > 2 such that

p 1 mod 2N
(11.14)
for sufficiently large N and put

p
L D L0 . p/:
If N and hence p is sufficiently large then p is unramified in K, i.e., every prime
p
divisor pjp appears in p with the exponent 1. We conclude that p K, and that
p
p
p is ramified in the quadratic extension K. p/. Therefore K. p/ is not contained
p
in L0 , and K. p/ is linearly disjoint to L0 over K. The Galois group G of LjK is
p
the direct product of Gal.L0 jK/ (which is cyclic of order 2r ), with Gal.K. p/jK/
(which is of order 2).
Let p be a prime of K and Gp its decomposition group in G. If p is unramified
in L (including the case when p is an infinite prime) then its decomposition group
is cyclic and hence Gp is a proper subgroup of G. If p is ramified in L then either
p
pj2 or pjp. In the first case, pj2, if N 3 then (11.14) implies p 2 Q2 , hence
p
p 2 L0;p , thus Lp D L0;p is of degree L0 W K D 2r over Kp . Hence its
Galois group Gp is of order 2r and thus a proper subgroup of G. In the second
case, pjp, let N be large enough such that L0 is contained in the field of 2N -th roots
of unity over K. The condition (11.14) implies that Qp contains the 2N -th roots of
p
unity, thus L0 Qp Kp and consequently Lp D Kp . p/ is of degree 2.
Now assume that K is a function field of characteristic 0. Let k be its field
of constants, and consider the unique extension k0 of degree 2r over k. We put
L0 D Kk0 ; this is the constant field extension of K of degree 2r . It is cyclic and
unramified over K. Now let t 2 K be a separating variable. Consider a prime
polynomial p.t/ 2 kt with the condition that its residue field contains k0 . This
condition is the analogue to condition (11.14) in the number field case. Since there
are infinitely such polynomials we may assume that p.t/ is not ramified in K.
If the characteristic of K is 2 then we put again
p
L D L0 . p.t/ /:
Quite analogous to the number field case it is seen that L satisfies the requirements
of the lemma. The situation here is even easier since L0 jK is unramified, hence it is
not necessary here to discuss the prime divisors which are ramified in L0 , as we had
to do in the number field case. The only primes p of K which are ramified in L are
the prime divisors of p.t/. For any such p its residue field contains k0 and hence its
completionpKp too contains k0 . It follows that Kp contains Kk0 D L0 and therefore
Lp D Kp . p.t// is of degree 2.
p
If the characteristic of K is 2 then K. p.t// is inseparabel and useless for our
construction. Instead of a square root we have to use a root of the appropriate Artin
254
Schreier equation:
L D L0 ./;
2 D
1
:
p.t/
Again, the only primes of K which are ramified in L are the prime divisors of p.t/
and the discussion now proceeds as in the case of characteristic 2.
Lemma 14 is proved. In that lemma we have not excluded the case of characteristic
2 because it is not necessary. However, in the following proof we have to assume that
char.K/ 2 in order to be able to apply Hoechsmanns theorem which is based on
the TatePoitou duality theorem.
Proof of Theorem 6. Let us put X D Z=2m . The automorphism group Aut.X/
consists of the units in Z=2m which act by multiplication. Aut.X/ is non-cyclic and
has the structure
Aut.X/ Z=2 m2
Z=2:
We see that Aut.X/ is isomorphic to the Galois group G D Gal.LjK/ of the field
extension of Lemma 14 if in that Lemma we take r D m 2.
Let is fix an isomorphism G Aut.X/. In this way X becomes a G-module. X
appears as a GK -module via the projection GK ! G. The action group of GK on X
is G.
y Then A is a GK -module of the same order 2m as X. We
Now we take A D X.
y
y D X. Thus the action group of GK on Ay is G. The conditions 13
have Ay D X
of Theorem 12 are satisfied in view of Lemma 14. We conclude that A is a counter
example to LGP.A; K/.
Problem. Prove Hoechsmanns theorem directly, without reference to the Tate
Poitou duality theorem. It seems that the reciprocity law for global fields will be
sufficient.
Acknowledgements
The author and the publisher wish thank Elsevier, Springer Science + BusinessMedia,
Verlag Vieweg + Teubner, Walter de Gruyter, the DMV, the LMS, and the Managing
Editor of the Israel Journal of Mathematics for granting permission to republish the
articles in this collection in their original or a revised and extended form. Thanks
also go to the Mathematisches Forschungsinstitut Oberwolfach for providing a high
resolution copy of the photo showing Otto Grn on page 79.
Bibliography
[ACH65] N. C. Ankeny, S. Chowla, and H. Hasse. On the class-number of the maximal real
subfield of a cyclotomic field. J. Reine Angew. Math., 217:217220, 1965.
[AH32]
A. A. Albert and H. Hasse. A determination of all normal division algebras over an

algebraic number field. Trans. Amer. Math. Soc., 34:722726, 1932.
[AJ95]
B. Aravire and B. Jacob. p-Algebras over maximally complete fields. Proc. Symp.
Pure Math., 58(2):2749, 1995.
[AJ96]
R. Aravire and B. Jacob. Versions of Springers theorem for quadratic forms in characteristic 2. Amer. J. Math., 118:235261, 1996.
[Alb30]
A. A. Albert. New results in the theory of normal division algebras. Trans. Amer.
Math. Soc., 32:171195, 1930.
[Alb31a] A. A. Albert. Division algebras over algebraic fields. Bull. Amer. Math. Soc., 37:777
784, 1931.
[Alb31b] A. A. Albert. On direct products. Trans. Amer. Math. Soc., 33:690711, 1931.
[Alb31c] A. A. Albert. On direct products, cyclic division algebras, and pure Riemann matrices.
Trans. Amer. Math. Soc., 33:219234, Correction p.999, 1931. Remark: This paper
has not been included into the Collected Papers of A. A. Albert.
[Alb31d] A. A. Albert. On normal division algebras of type R in thirty-six units. Trans. Amer.
Math. Soc., 33:235243, 1931.
[Alb32a] A. A. Albert. A construction of non-cyclic normal divisrion algebras. Bull. Amer.
Math. Soc., 38:449456, 1932.
[Alb32b] A. A. Albert. On the construction of cyclic algebras with a given exponent. Amer. J.
Math., 54:113, 1932.
[Alb34]
A. A. Albert. Qn normal Kummer fields over a non-modular field. Trans. Amer. Math.
Soc., 36:885892, 1934.
[Alb37]
A. A. Albert. Modern higher algebra. Univ. of Chicago Press, Chicago, 1937. XIV.
319 pp.
[Alb38a] A. A. Albert. Non-cyclic algebras with pure maximal subfields. Bull. Amer. Math.
Soc., 44:576579, 1938.
[Alb38b] A. A. Albert. Quadratic null formes over a function field. Ann. Math. (2), 39:494505,
1938.
[Alb39]
A. A. Albert. Structure of Algebras. Amer. Math. Soc. Coll. Publ, 1939. XI, 210 pp.
[Ale80]
P. Alexandroff. Pages from an autobiography. Russian Math. Surveys, 35:315358,

1980.
[Ale83]
P. Alexandrov. In Memory of Emmy Noether. In Emmy Noether, Collected Papers.

Edited by N Jacobson., pages 111. Springer, 1983. VIII, 777 pp.
[Ami72] S. A. Amitsur. On central division algebras. Israel J. Math., 12:408420, 1972.

[Arf39]
C. Arf. Untersuchungen ber reinverzweigte Erweiterungen diskret bewerteter perfekter Krper. J. Reine Angew. Math., 181:144, 1939.
258
Bibliography
[Arf41]
C. Arf. Untersuchungen ber quadratische Formen in Krpern der Charakteristik 2.

I. J. Reine Angew. Math., 183:148167, 1941.
[Arf43]
C. Arf. Untersuchungen ber quadratische Formen in Krpern der Charakteristik 2.

II. Re. Fac. Sci. Univ. Istanbul (A), 8:297327, 1943.
[Arf90]
C. Arf. The Collected Papers. Turkish Math. Soc., 1990. 422 pp.
[Art27]
E. Artin. Beweis des allgemeinen Reziprozittsgesetzes. Abh. Math. Semin. Univ.

Hamb., 5:353363, 1927.
[Art28a] E. Artin. ber einen Satz von Herrn J.H. Maclagan Wedderburn. Abh. Math. Semin.
Univ. Hamb., 5:245250, 1928.
[Art28b] E. Artin. Zur Arithmetik hyperkomplexer Zahlen. Abh. Math. Semin. Univ. Hamb.,
5:261289, 1928.
[Art28c] E. Artin. Zur Theorie der hyperkomplexen Zahlen. Abh. Math. Semin. Univ. Hamb.,
5:251260, 1928.
[Art29]
E. Artin. Idealklassen in Oberkrpern und allgemeines Reziprozittsgesetz. Abh.

Math. Semin. Univ. Hamb., 7:4651, 1929.
[Art50]
E. Artin. The influence of J. H. Wedderburn on the development of modern algebra.

Bull. Amer. Math. Soc., 56:6572, 1950.
[AS27a]
E.Artin and O. Schreier.Algebraische Konstruktion reeller Krper. Abh. Math. Semin.

Univ. Hamb., 5:8599, 1927.
[AS27b] E. Artin and O. Schreier. Eine Kennzeichnung der reell abgeschlossenen Krper. Abh.
Math. Semin. Univ. Hamb., 5:225231, 1927.
[AT68]
E. Artin and J. Tate. Class field theory. W. A. Benjamin, Inc., New York-Amsterdam,
1968. XXVI, 259 pp. Nachdruck der Ausarbeitung eines Seminars im Jahre 1951 an
der Universitt Princeton.
[Bae82]
R. Baeza. Comparing u-invariants of fields in characteristic 2. Bol. Soc. Bras. Mat.,

13(1):105114, 1982.
[Ban92]
B. Banaschewski. Algebraic closure without choice. Z. Math. Logik, 38(4):383385,

1992.
[BHN32] R. Brauer, H. Hasse, and E. Noether. Beweis eines Hauptsatzes in der Theorie der
Algebren. J. Reine Angew. Math., 167:399404, 1932.
[Bil37]
H. Bilharz. Primdivisoren mit vorgegebener Primitivwurzel. Math. Ann., 114:476

492, 1937.
[BN27]
R. Brauer and E. Noether. ber minimale Zerfllungskrper irreduzibler Darstellungen. Sitzungsberichte Akad. Berlin, 1927:221228, 1927.
[Bou60]
N. Bourbaki. lments dhistoire des mathmatiques. Histoire de la Pens. Hermann,

1960.
[Bra26]
R. Brauer. ber Zusammenhnge zwischen arithmetischen und invariantentheoretischen Eigenschaften von Gruppen linearer Substitutionen. Sitzungsber. Preuss. Akad.
Wiss., 1926:410416, 1926.
[Bra28]
R. Brauer. Untersuchungen ber die arithmetischen Eigenschaften von Gruppen linearer Substitutionen. Math. Z., 28:677696, 1928.
Bibliography
259
[Bra29a] R. Brauer. ber Systeme hyperkomplexer Gren. Jber. Deutsch. Math. Verein., 28(2.
Abteilung):4748, 1929. kursiv.
[Bra29b] R. Brauer. ber Systeme hyperkomplexer Zahlen. Math. Z., 30:79107, 1929.
[Bra30a] H. Brandt. Zur Idealtheorie Dedekindscher Algebren. Comment. Math. Helv., 2:13
17, 1930.
[Bra30b] R. Brauer. Untersuchungen ber die arithmetischen Eigenschaften von Gruppen linearer Substitutionen. II. Math. Z., 31:733747, 1930.
[Bra33]
R. Brauer. ber den Index und den Exponenten von Divisionsalgebren. Thoku Math.
J., 37:7787, 1933.
[Bra34]
H. R. Brahana. Prime power abelian groups generated by a set of conjugates under a

special automorphism. Amer. Journ. Math., 55:553584, 1934.
[Bra45]
R. Brauer. On the representations of a group of order g in the field of g-th roots of

unity. Amer. J. Math., 67:461471, 1945.
[Bra47]
R. Brauer. Applications of induced characters. Amer. J. Math., 69:709716, 1947.
[Bru97]
E. Brussel. Wang counterexamples lead to noncrossed products. Proc. Amer. Math.

Soc., 125:21992206, 1997.
[Br07]
J. Brning. Die Flche des Poeten. Mittt. Deutsche Math. Vereinigung, 15(1):3336,
2007.
[BS81]
J. Brewer and M. Smith, editors. Emmy Noether. A tribute to her life and work. Marcel
Dekker, New York, 1981. X + 180 pp.
[BT55]
R. Brauer and J. Tate. On the characters of finite groups. Annals of Math., 62:17,
1955.
[Bur02]
W. Burnside. On an unsettled question in the theory of discontinuous groups. Quart.

Journ. Math., 33:230238, 1902.
[Bur11]
W. Burnside. Theory of groups of finite order. Cambridge Univ. Press, 2 edition, 1911.
XXIV + 512 pp.
[CE56]
H. Cartan and S. Eilenberg. Homological Algebra. Princeton Univ. Press, 2 edition,

1956. XII + 390 pp.
[Che31]
C. Chevalley. Relation entre le nombre de classes dun sous-corps et celui dun surcorps. C. R. Acad. Sci., Paris, 192:257258, 1931.
[Che33]
C. Chevalley. La thorie du symbole de restes normiques. J. Reine Angew. Math.,

169:140157, 1933.
[Che35]
C. Chevalley. Sur la thorie du corps de classes. C. R. Acad. Sci., Paris, 201:632634,

1935.
[Che40]
C. Chevalley. La thorie du corps de classes. Ann. Math. (2), 41:394418, 1940.
[Cur99]
C. Curtis. Pioneers of representation theory: Frobenius, Burnside, Schur and Brauer.

History of Mathematics. Amer. Math. Soc., Providence, R.I., 1999. XVI, 287 pp.
[Dau95]
J. W. Dauben. Abraham Robinson. The creation of nonstandard analysis. A personal

and mathematical odyssey. Princeton University Press, Princeton, NY, 1995. XIX,
559 pp.
[Deu35]
M. Deuring. Algebren. Erg. d. Math. u. ihrer Grenzgebiete. Julius Springer, Berlin,

1935. 143 pp.
260
Bibliography
[Deu68]
M. Deuring. Imaginre quadratische Zahlkrper mit der Klassenzahl 1. Invent. Math.,

5:169179, 1968.
[Dic01]
L. E. Dickson. Linear groups: With an exposition of the Galois field theory. Dover,
New York, 1901. XVI, 312 pp. Reprint from the original edition 1901.
[Dic23]
L. E. Dickson. Algebras and Their Arithmetics. Univ. of Chicago Press., Chicago,

1923. XII, 241 pp.
[Dic27]
L. E. Dickson. Algebren und ihre Zahlentheorie. Mit einem Kapitel ber Idealtheorie
von A. Speiser. Orell Fssli (Verffentlichungen der Schweizer Math. Ges. Bd. 4),
Zrich, 1927. bersetzt von J.J. Burckhardt und E. Schubarth.
[Dic70]
A. Dick. Emmy Noether 18821935. Beiheft No. 13 zur Zeitschrift Elemente der
Mathematik. BirkhuserVerlag, 1970. English translation 1981 by H.I. Blocher.
[Dra75]
P.K. Draxl. ber gemeinsame quadratische Zerfllungskrper von Quaternionenalgebren. Nachr. Akad. Wiss. Gttingen. Math. Phys. Kl. 3.F., 16:251259, 1975.
[EKM08] R. Elman, N. Karpenko, and A. Merkurjev. The algebraic and geometric theory of
quadratic forms., volume 56 of Colloquium Publications. American Mathematical
Society, Providence, RI, USA, 2008. 435 pp.
[Fei79]
W. Feit. Richard Brauer. Bull. Amer. Math. Soc. New Ser., 1:120, 1979.
[FR08]
G. Frei and P. Roquette, editors. Emil Artin and Helmut Hasse. Their correspondence 1923-1934. With contributions of Franz Lemmermeyer and an introduction in
English. UniversittsVerlag, Gttingen, 2008. 497 pp.
[Fra31]
W. Franz. Untersuchungen zum Hilbertschen Irreduzibilittssatz. Math. Zeitschr.,

33:275293, 1931.
[Fre77]
G. Frei. Leben und Werk von Helmut Hasse 1. Teil: Der Lebensgang., volume 37
of Collection Mathmatique, Srie: Mathmatiques pures et appliques. Universit
Laval, Qubec, 1977. 59 pp.
[Fre85]
G. Frei. Helmut Hasse (1898-1979). Expositiones Math., 3:5569, 1985.
[Fre03]
G. Frei. Johann Jakob Burckhardt zum 100. Geburtstag am 13. juli 2003. Elem. Math.,
58:134146, 2003.
[Fur02]
Ph. Furtwngler. ber das Reziprozittsgesetz der l-ten Potenzreste in algebraischen

Zahlkrpern, wenn l eine ungerade Primzahl bedeutet. Gttinger Abhandlungen, 2:3
82, 1902.
[Fur08]
Ph. Furtwngler. ber die Klassenzahlen Abelscher Zahlkrper. J. Reine Angew.

Math., 134:9194, 1908.
[Fur29]
Ph. Furtwngler. Beweis des Hauptidealsatzes fr die Klassenkrper algebraischer

Zahlkrper. Abh. Math. Semin. Univ. Hamb., 7:1436, 1929.
[GJ07]
W.-D. Geyer and C. U. Jensen. Embeddability of quadratic extensions in cyclic extensions. Forum. Math., 19:707725, 2007.
[GR55]
P.C. Gilmore and A. Robinson. Metamathematical considerations on the relative

irreducibility of polynomials. Can. J. Math., 7:483489, 1955.
[Gru32]
W. Grunwald. Charakterisierung des Normenrestsymbols durch die }-Stetigkeit, den

vorderen Zerlegungssatz und die Produktformel. Math. Ann., 107:145164, 1932.
[Gru33]
W. Grunwald. Ein allgemeines Existenztheorem fr algebraische Zahlkrper. J. Reine

Angew. Math., 169:103107, 1933.
Bibliography
261
[Gr34a] O. Grn. ber Substitutionsgruppen im Galoisfeld. J. Reine Angew. Math., 170:170

172, 1934.
[Gr34b] O. Grn. Zur Fermatschen Vermutung. J. Reine Angew. Math., 170:231234, 1934.
[Gr35]
O. Grn. Beitrge zur Gruppentheorie I. J. Reine Angew. Math., 174:114, 1935.
[Gr36]
O. Grn. ber eine Faktorgruppe freier Gruppen. Deutsche Math., 1:772782, 1936.
[Gr38]
O. Grn. Gruppentheoretische Untersuchungen. Deutsche Math., 3:547555, 1938.
[Gr40]
O. Grn. Zusammenhang zwischen Potenzbildung und Kommutatorbildung. J. Reine

Angew. Math., 182:158177, 1940.
[Gr45]
O. Grn. Beitrge zur Gruppentheorie II. J. Reine Angew. Math., 188:165169, 1945.
[Gr48a] O. Grn. Beitrge zur Gruppentheorie III. Math. Nachr., 1:124, 1948.
[Gr48b] O. Grn. Berechnung des elektrischen Feldes bei einer gewissen Materialverteilung.
Math. Nachr., 4:419433, 1948.
[Gr53]
O. Grn. Beitrge zur Gruppentheorie. V. ber endliche p-Gruppen. Osaka Math.

J., 5:117146, 1953.
[Has23a] H. Hasse. ber die quivalenz quadratischer Formen im Krper der rationalen
Zahlen. J. Reine Angew. Math., 152:205224, 1923.
[Has23b] H. Hasse. ber die Darstellbarkeit von Zahlen durch quadratische Formen im Krper
der rationalen Zahlen. J. Reine Angew. Math., 152:129148, 1923.
[Has24a] H. Hasse. quivalenz quadratischer Formen in einem beliebigen algebraischen
Zahlkrper. J. Reine Angew. Math., 153:158162, 1924.
[Has24b] H. Hasse. Darstellbarkeit von Zahlen durch quadratische Formen in einem beliebigen
algebraischen Zahlkrper. J. Reine Angew. Math., 153:113130, 1924.
[Has25]
H. Hasse. ber das allgemeine Reziprozittsgesetz in algebraischen Zahlkrpern.

Jahresber. Dtsch. Math.-Ver., 33, 2.Abteilung:97101, 1925.
[Has26a] H. Hasse. Bericht ber neuere Untersuchungen und Probleme aus der Theorie der algebraischen Zahlkrper. I: Klassenkrpertheorie. Jahresber. Dtsch. Math.-Ver., 35:1
55, 1926.
[Has26b] H. Hasse. Ein weiteres Existenztheorem in der Theorie der algebraischen Zahlkrper.
Math. Zeitschr., 24:14916, 1926. 0.
[Has26c] H. Hasse. Zwei Existenztheoreme ber algebraische Zahlkrper. Math. Annalen,
95:229238, 1926.
[Has27a] H. Hasse. Existenz gewisser algebraischer Zahlkrper. Sitzungsberichte Akad. Berlin,
1927:229234, 1927.
[Has27b] H. Hasse. Neue Begrndung der komplexen Multiplikation I: Einordnung in die
allgemeine Klassenkrpertheorie. J. Reine Angew. Math., 157:115139, 1927.
[Has27c] H. Hasse. ber das Reziprozittsgesetz der m-ten Potenzreste. J. Reine Angew. Math.,
158:228259, 1927.
[Has28]
H. Hasse. Book Review: L. E. Dickson, Algebren und ihre Zahlentheorie. Jahresber.

Dtsch. Math.-Ver., 37:9097 (kursiv), 1928.
[Has30a] H. Hasse. Bericht ber neuere Untersuchungen und Probleme aus der Theorie
der algebraischen Zahlkrper. II: Reziprozittsgesetz. Jahresber. Dtsch. Math.-Ver.,
6(Ergnzungsband), 1930. IV + 204 pp.
262
Bibliography
[Has30b] H. Hasse. Die moderne algebraische Methode. Jahresber. Dtsch. Math. Ver., 31:22
34, 1930. Reprinted in English translation in the Mathematical Intelligencer, vol. 8,
1986.
[Has30c] H. Hasse. Die Normenresttheorie relativAbelscher Zahlkrper als Klassenkrpertheorie im Kleinen. J. Reine Angew. Math., 162:145154, 1930.
[Has30d] H. Hasse. Neue Begrndung und Verallgemeinerung der Theorie des Normenrestsymbols. J. Reine Angew. Math., 162:134144, 1930.
[Has31a] H. Hasse. Beweis eines Satzes und Widerlegung einerVermutung ber das allgemeine
Normenrestsymbol. Nachr. Ges. Wiss. Gttingen, Math.Phys. Kl. I, pages 6469,
1931.
[Has31b] H. Hasse. Neue Begrndung der komplexen Multiplikation. II. Aufbau ohne Benutzung der allgemeinen Klassenkrpertheorie. J. Reine Angew. Math., 165:6488,
1931.
[Has31c] H. Hasse. Theorie der zyklischen Algebren ber einem algebraischen Zahlkrper.
Nachr. Ges. Wiss. Gttingen, Math.Phys. Kl. I, pages 7079, 1931.
[Has31d] H. Hasse. ber }-adische Schiefkrper und ihre Bedeutung fr die Arithmetik hyperkomplexer Zahlsysteme. Math. Ann., 104:495534, 1931.
[Has32a] H. Hasse. Ansprache zum 70. Geburtstag des Geh. Regierungsrats Prof . Dr. Kurt
Hensel am 29. Dezember 1931. Mitteilungen Universittsbund Marburg, 1932(1):2
6, 1932.
[Has32b] H. Hasse. Strukturtheorie der halbeinfachen Algebren ber algebraischen Zahlkrpern. In Verhandlungen Kongre Zrich 1932., volume 2, pages 1819. IMU, 1932.
[Has32c] H. Hasse. Theory of cyclic algebras over an algebraic number field. Trans. Amer.
Math. Soc., 34:171214, 1932.
[Has33a] H. Hasse. Die Struktur der R. Brauerschen Algebrenklassengruppe ber einem algebraischen Zahlkrper. Insbesondere Begrndung der Theorie des Normenrestsymbols
und Herleitung des Reziprozittsgesetzes mit nichtkommutativen Hilfsmitteln. Math.
Ann., 107:731760, 1933.
[Has33b] H. Hasse. Vorlesungen ber Klassenkrpertheorie. Preprint, Marburg. [Later published in book form by Physica Verlag Wrzburg (1967)], 1933.
[Has36a] H. Hasse. Zur Theorie der abstrakten elliptischen Funktionenkrper. III. die Struktur
des Meromorphismenrings. J. Reine Angew. Math., 175:193207, 1936.
[Has36b] H. Hasse. Zur Theorie der abstrakten elliptischen Funktionenkrper. II. Automorphismen und Meromorphismen. Das Additionstheorem. J. Reine Angew. Math., 175:69
88, 1936.
[Has36c] H. Hasse. Zur Theorie der abstrakten elliptischen Funktionenkrper. I. die Struktur
der Divisorenklassen endlicher Ordnung. J. Reine Angew. Math., 175:5562, 1936.
[Has50a] H. Hasse. Zum Existenzsatz von Grunwald in der Klassenkrpertheorie. J. Reine
Angew. Math., 188:4064, 1950.
[Has50b] H. Hasse. Zur Frage der Zerfllungskrper des Gruppenrings einer endlichen Gruppe.
Math. Nachr., 3:46, 1950.
[Has52]
H. Hasse. ber die Klassenzahl abelscher Zahlkrper. AkademieVerlag, Berlin,

1952. Reprint 1985 with an introduction of J. Martinet.
Bibliography
263
[Hau29]
O. Haupt. Einfhrung in die Algebra I, II. B. G. Teubner, 1929.
[Hau88]
O. Haupt. Erinnerungen des Mathematikers Otto Haupt. Unpublished. Archive of

Erlangen University, 1988.
[Hee52]
K. Heegner. Diophantische Analysis und Modulfunktionen. Math. Z., 56:227253,

1952.
[Hei69]
W. Heisenberg. Der Teil und das Ganze. Piper, Leipzig, 1969. XI 349 pp.
[Hen08]
K. Hensel. Theorie der algebraischen Zahlen. I. Teubner, Leipzig, 1908. XI 349 pp.
[Her30]
J. Herbrand. Nouvelle dmonstration et gnralisation dun thorme de Minkowski.

C. R. Acad. Sci., Paris, 191:1282, 1930.
[Her32]
J. Herbrand. Sur les classes des corps circulaires. Journ. de Math. (9), 11:417441,
1932.
[Hey29]
K. Hey. Analytische Zahlentheorie in Systemen hyperkomplexer Zahlen. Dissertation,

Hamburg, 1929. 49 p.
[Hig56]
G. Higman. On finite groups with exponent five. Proc. Cambridge Philos. Soc.,
52:381390, 1956.
[Hil90]
D. Hilbert. ber die Theorie der algebraischen Formen. Math. Ann., 36:473534,
1890.
[Hil92]
D. Hilbert. ber die Irreducibilitt ganzer rationaler Functionen mit ganzzahligen

Coefficienten. J. Reine Angew. Math., 110:104129, 1892.
[Hil94]
D. Hilbert. ber die Zerlegung der Ideale eines Zahlkrpers in Primideale. Math.
Ann., 44:18, 1894.
[Hil97]
D. Hilbert. Die Theorie der algebraischen Zahlkrper. Jahresber. Dtsch. Math. Ver.,
4:IXVIII u. 175546, 1897. Englische bersetzung: The Theory of Algebraic Number Fields. Springer, Heidelberg, 1998.
[Hoe67]
K. Hoechsmann. Zum Einbettungsproblem. J. Reine Angew. Math., 229:81106,

1967.
[Hup67]
B. Huppert. Endliche Gruppen I. Springer, Berlin, 1967. XII, 793 pp.
[Iya34]
S. Iyanaga. Zum Beweise des Hauptidealsatzes. Abh. Math. Semin. Univ. Hamb.,
10:349357, 1934.
[Jac45]
N. Jacobson. Structure of simple rings without finiteness assumptions. Trans. Amer.

Math. Soc., 57:228245, 1945.
[Jen86]
W. Jentsch. Auszge aus einer unverffentlichten Korrespondenz von Emmy Noether

und Hermann Weyl mit Heinrich Brandt. Historia Math., 13:512, 1986.
[JL98]
W. Jehne and E. Lamprecht. Helmut Hasse, Hermann Ludwig Schmid and their
students in Berlin. In H. B. W. Begehr et al., editor, Mathematics in Berlin., pages
143149, Berlin, 1998. Berliner Mathematische Gesellschaft., Birkhuser.
[Kan80a] E. Kani. Eine Verallgemeinerung des Satzes von Castelnuovo-Severi. J. Reine Angew.
Math., 318:178220, 1980.
[Kan80b] E. Kani. Nonstandard diophantine geometry. In Proc. Queens Number Theory Conf.
1979., volume 54 of Queens Pap. Pure Appl. Math., pages 129172, Kingston, 1980.
264
Bibliography
[Kan82]
E. Kani. Nonstandard methods in diophantine geometry. In Journees arithmetiques,

Exeter 1980, volume 56 of Lond. Math. Soc. Lect. Note Ser., pages 322342, London,
1982.
[Kap80]
I. Kaplansky. Abraham Adrian Albert. November 9, 1905 - June 6, 1972. Biographical

Mem. Nat. Acad. Sci., 51:322, 1980.
[Ker00]
I. Kersten. Biography of Ernst Witt. Contemp. Math., 272:155171, 2000.
[Kie12]
L. Kiepert. Grundri der Differential- und Integralrechnung. I Teil: Differentialrechnung. Helwingsche Verlagsbuchhandlung, Hannover, 12 edition, 1912. XX, 863
pp.
[Kim72] C. Kimberling. Emmy Noether. Amer. Math. Monthly, 79:136149, 1972. An addendum to this article appeared on page 755 of the same volume.
[Kim81] C. Kimberling. Emmy Noether and her influence. In James W. Brewer and Martha K.
Smith, editors, Emmy Noether. A tribute to her life and work., pages 361. Marcel
Dekker, 1981.
[KL64]
T. Kubota and H.W. Leopoldt. Eine p-adische Theorie der Zetawerte. I: Einfhrung
der p-adischen Dirichletschen L-Funktionen. J. Reine Angew. Math., 214/215:328
339, 1964.
[Kle18]
F. Klein. ber die Differentialgesetze fr die Erhaltung von Impuls und Energie in
der Einsteinschen Gravitationstheorie. Gtt. Nachr., 1918:171189, 1918.
[Kle99]
I. Kleiner. Field Theory: From Equations to Axiomatization. Part II. Amer. Math.
Monthly, 106(9):859863, 1999.
[Knu80]
M.-A. Knus. Quadratic and Hermitian Forms over Rings. Grundlehren der mathematischen Wissenschaften. Springer-Verlag, Berlin, Heidelberg, New York, 1980.
VII, 520 pp.
[Knu93]
M.-A. Knus. Sur la forme dAlbert et le produit tensoriel de deux algbres de quaternions. Bull. Soc. Math. Belg., 45:333337, 1993.
[Koe95]
J. Koenigsmann. From p-rigid elements to valuations (with a Galois-characterization

of p-adic fields). J. Reine Angew. Math., 465:165182, 1995.
[Kos55]
A. Kostrikin. Solution of a weakened problem of Burnside for exponent 5. (Russian).

Izv. Akad. Nauk SSSR. Ser. Mat., 19:233244, 1955.
[Kt33]
G. Kthe. Erweiterung des Zentrums einfacher Algebren. Math. Ann., 107:761766,

1933.
[Kru28]
W. Krull. Galoissche Theorie der unendlichen algebraischen Erweiterungen. Math.

Ann., 100:687698, 1928.
[KS04]
Y. Kosmann-Schwarzbach. Les thormes de Noether. Invariance et lois de conservation au XXe sicle. cole polytechnique, Palaiseaux, 2004. 173 pp.
[Kum50] E. Kummer. Allgemeiner Beweis des Satzes, dass die Gleichung x C y D z

durch ganze Zahlen unlsbar ist, fr alle diejenigen Primzahl-exponenten , welche
ungerade Primzahlen sind und in den Zhlern der ersten 12 . 3/ Bernoullischen
Zahlen nicht vorkommen. J. Reine Angew. Math., 40:130138, 1850.
[Kr13]
J. Krschk. ber Limesbildung und allgemeine Krpertheorie. J. Reine Angew.

Math., 142:211253, 1913.
Bibliography
[KW54]
265
W. Klingenberg and E. Witt. ber die Arfsche Invariante quadratischer Formen mod
2-. J. Reine Angew. Math., 193:121122, 1954.
[Lam02] T.Y. Lam. On the linkage of quaternion algebras. Bull. Belg. Math. Soc., 9:415418,
2002.
[Lan04]
R. Langlands. Benim tanidigim Cahit Arf (Recollections of a year in Turkey with

Cahit Arf) (turkish). Matematik Dnyasi, 2004(winter number), 2004.
[Leh74]
D. H. Lehmer. Harry Schultz Vandiver, 1882-1973. Bull. American Math. Soc.,

80:817818, 1974.
[Lem97] F. Lemmermeyer. On 2-class field towers of some imaginary quadratic number fields.
Abh. Math. Semin. Univ. Hamburg, 67:205214, 1997.
[Leo58]
H.-W. Leopoldt. Zur Struktur der `-Klassengruppe galoisscher Zahlkrper. J. Reine

Angew. Math., 199:165174, 1958.
[Leo75]
Heinrich-Wolfgang Leopoldt. Eine p-adische Theorie der Zetawerte. II: Die padische - Transformation. J. Reine Angew. Math., 274-275:224239, 1975.
[Lor98]
F. Lorenz. Ein Scholion zum Satz 90 von Hilbert. Abh. Math. Seminar Hamburg.
Universitt, 68:347362, 1998.
[Lor05]
F. Lorenz. Kte Hey and the Main Theorem in the theory of algebras. In W. Wiesaw,
editor, European Mathematics in the last centuries., pages 5776, Wrocaw, 2005.
Stefan Banach International Mathematical Center, Institute of Mathematics Wrocaw
University.
[Lor08a] F. Lorenz. Algebra. Volume II: Fields with structure, algebras and advanced topics.
Transl. from the German by Silvio Levy. With the collaboration of the translator.
Universitext. Springer, New York, 2008. 336 pp.
[Lor08b] F. Lorenz. Zum Beweis der Funkionalgleichung der Heyschen Zetafunktion in der
Dissertation von Wolfgang Wichmann, Emmy Noethers letztem Gttinger Doktoranden. Mitt. Math. Ges. Hamburg, 27:167183, 2008.
[LR03]
F. Lorenz and P. Roquette. The theory of Grunwald-Wang in the setting of valuation

theory. In Franz-Viktor et al. Kuhlmann, editor, Valuation theory and its applications, vol II. Proceedings of the international conference and workshop, University
of Saskatchewan, Saskatoon, Canada, July 28August 11, 1999., volume 33 of Fields
Inst. Commun., pages 175212, Providence, RI, 2003. American Mathematical Society.
[LR06]
F. Lemmermeyer and P. Roquette, editors. Helmut Hasse and Emmy Noether. Their
correspondence 1925-1935. With an introduction in English. UniversittsVerlag,
Gttingen, 2006. 303 pp.
[Lux62]
W.A. J. Luxemburg. Non-standard analysis. A.Robinsons theory of infinitesimals and

infinitely large numbers. Math. Dept. California Institute of Technology, Pasadena,
CA, 1962. 150 pp.
[Mac81] S. MacLane. Mathematics at the University of Gtingen 19311933. In W. Brewer

and M. Smith, editors, Emmy Noether. A tribute to her life and work., pages 6578,
New York, 1981.
[Mag30] W. Magnus. ber diskontinuierliche Gruppen mit einer definierenden Relation. (Der
Freiheitssatz.). J. Reine Angew. Math., 163:141165, 1930.
266
Bibliography
[Mag34] W. Magnus. ber den Beweis des Hauptidealsatzes. J. Reine Angew. Math., 170:235
240, 1934.
[Mag35] W. Magnus. Beziehungen zwischen Gruppen und Idealen in einem speziellen Ring.
Math. Ann., 111:259280, 1935.
[Mag37] W. Magnus. ber Beziehungen zwischen hheren Kommutatoren. J. Reine Angew.
Math., 177:105115, 1937.
[Mag50] W. Magnus. A connection between the Baker-Hausdorff formula and the problem of
Burnside. Ann. Math. (2), 52:111126, 1950.
[Mer93]
F. Mertens. Ueber ganze Functionen von m Systemen von je n Unbestimmten.

Monatsh. f. Math., 4:297329, 1893.
[ML81]
S. Mac Lane. History of abstract algebra: origin, rise, and decline of a movement. In
American mathematical heritage: algebra and applied mathematics. El Paso, Tex.,
1975/Arlington, Tex., 1976, volume 13 of Math. Ser., pages 335, Lubbock, Tex.,
1981. Texas Tech Univ.
[Mor11]
P. Morton. A correction to Hasses version of the Grunwald-Hasse-Wang theorem. J.

Reine Angew. Math., 659:169174, 2011.
[NN07]
R. Netz and W. Noel. The Archimedes Codex. Revealing the secrets of the worlds
greatest palimpsest. Weidenfels & Nicolson, The Orion Publishing Group., London,
2007.
[Noe18]
E. Noether. Invariante Variationsprobleme. Nachr. Ges. Wiss. Gttingen, 1918:235

257, 1918.
[Noe19]
E. Noether. Die arithmetische Theorie der algebraischen Funktionen einer Vernderlichen in ihrer Beziehung zu den brigen Theorien und zu der Zahlkrpertheorie.
Jahresber. Dtsch. Math.-Ver., 38:192203, 1919.
[Noe21]
E. Noether. Idealtheorie in Ringbereichen. Math. Ann., 83:2466, 1921.
[Noe23]
E. Noether. Algebraische und Differentialinvarianten. Jahresber. Dtsch. Math.-Ver.,

32:177184, 1923.
[Noe24]
E. Noether. Abstrakter Aufbau der Idealtheorie im algebraischen Zahlkrper. Jahresber. Dtsch. Math.-Ver., 33:102, 1924. 2. Abteilung.
[Noe25]
E. Noether. Gruppencharaktere und Idealtheorie. Jahresber. Dtsch. Math.-Ver.,

34:144, 1925. 2. Abteilung.
[Noe26]
E. Noether. Abstrakter Aufbau der Idealtheorie in algebraischen Zahl- und Funktionenkrpern. Math. Ann., 96:2661, 1926.
[Noe29]
E. Noether. Hyperkomplexe Grssen und Darstellungstheorie. Math. Z., 30:641692,

1929.
[Noe32]
E. Noether. Hyperkomplexe Systeme in ihren Beziehungen zur kommutativen Algebra und Zahlentheorie. Verhandl. Intern. Math. Kongre Zrich, 1:189194, 1932.
[Noe33a] E. Noether. Der Hauptgeschlechtssatz fr relativgaloissche Zahlkrper. Math. Ann.,

108:411419, 1933.
[Noe33b] E. Noether. Nichtkommutative Algebra. Math. Z., 37:514541, 1933. Nachdruck in
Ges. Abh. 40, pp. 642669.
Bibliography
267
[Noe83]
E. Noether. Algebra der hyperkomplexen Gren. Vorlesung Wintersemester

1929/30, ausgearbeitet von M. Deuring. In N. Jacobson, editor, Emmy Noether, Collected Papers, pages 711763. Springer, 1983. VIII, 777 pp.
[NS20]
E. Noether and W. Schmeidler. Moduln in nichtkommutativen Bereichen, insbesondere aus Differential- und Differenzenenausdrcken. Math. Z., 8:135, 1920.
[Ost33]
A. Ostrowski. ber algebraische Funktionen von Dirichletschen Reihen. Math. Z.,

37:98133, 1933.
[Ost34]
A. Ostrowski. Untersuchungen zur arithmetischen Theorie der Krper. Die Theorie

der Teilbarkeit in allgemeinen Krpern. Math. Z., 39:269404, 1934.
[Pop88]
F Pop. Galoissche Kennzeichnung p-adisch abgeschlossener Krper. J. Reine Angew.

Math., 392:145175, 1988.
[Pur73]
W. Purkert. Zur Genesis des abstrakten Krpernegriffs. NTM, Schriftenr. Gesch.

Naturwiss. Techn. Med., 10(2):820, 1973.
[Rei76]
C. Reid. Courant in Gttingen and New York. Springer, New York, 1976. 314 pp.
[Rib79]
P. Ribenboim. 13 Lectures on Fermats Last Theorem. Springer, New York, Heidelberg, Berlin, 1979. xvi, 302 pp.
[Rob55]
A. Robinson. On ordered fields and definite functions. Math. Annalen, 130:287271,

1955.
[Rob63]
A. Robinson. Introduction to model theory and to the metamathematics of algebra.

Studies in Logic. North-Holland, Amsterdam, 1963. ix, 284 p.
[Rob66]
A. Robinson. Non-standard analysis. Studies in Logic. North-Holland, Amsterdam,

1966. xi, 293 pp.
[Rob67]
A. Robinson. Nonstandard arithmetic. Invited address. Bull. Amer. Math. Soc.,

73:818843, 1967.
[Rob69]
A. Robinson. Topics in Nonstandard Algebraic Number Theory. In Application of

Model Theory to Algebra, Analysis and Probability. Proc. internat. Sympos. Calif.
Inst. Technol. 1967., pages 117. Holt, Rinehart and Winston, New York, 1969.
[Rob72]
A. Robinson. Algebraic function fields and non-standard arithmetic. In Contributions

to non-standard analysis (Sympos. Oberwolfach, 1970)., volume 69 of Studies in
Logic and Found. Math., pages 117. North-Holland, Amsterdam, 1972.
[Rob73]
A. Robinson. Nonstandard points on algebraic curves. J. Number Theory, 7:301327,

1973.
[Roh98]
H. Rohrbach. Helmut Hasse and Crelles Journal. J. Reine Angew. Math., 500:513,
1998.
[Roq52]
P. Roquette. Arithmetische Untersuchung des Charakterringes einer endlichen

Gruppe. J. Reine Angew. Math., 190:148168, 1952.
[Roq75]
P. Roquette. Nonstandard aspects of Hilberts irreducibility theorem. In D. H. Saracino and B. Weispfenning, editors, Model Theor. Algebra, Mem. Tribute Abraham
Robinson, volume 498 of Lect. Notes Math., pages 231275. Springer, Heidelberg,
1975.
[Roq89]
P. Roquette. ber die algebraisch-zahlentheoretischen Arbeiten von Max Deuring.

Jahresber. Dtsch. Math.-Ver., 91:109125, 1989.
268
Bibliography
[Roq00]
P. Roquette. On the history of Artins L-functions and conductors. Seven letters from
Artin to Hasse in the year 1930. Mitt. Math. Ges. Hamburg, 19*:550, 2000.
[Roq01]
P. Roquette. Class field theory in characteristic p, its origin and development. In

Katsuya Miyake, editor, Class field theory its centenary and prospect. Proceedings
of the 7th MSJ International Research Institute of the Mathematical Society of Japan,
Tokyo, Japan, June 312, 1998, volume 30 of Adv. Stud. Pure Math., pages 549631,
Tokyo, 2001. Mathematical Society of Japan.
[Roq02]
P. Roquette. History of valuation theory. Part 1. In F. V. Kuhlmann et al., editor, Valuation theory and its applications, vol.I., volume 32 of Fields Institute Communications,
pages 291355, Providence, RI, 2002. American Mathematical Society.
[Roq04]
P. Roquette. The Riemann hypothesis in characteristic p, its origin and development.

Part 2. The first steps by Davenport and Hasse. Mitt. Math. Ges. Hamburg, 22:169,
2004.
[Roq05]
P. Roquette. The Brauer-Hasse-Noether Theorem in historical perspective., volume 15 of Schriftenreihe der Heidelberger Akademie der Wissenschaften. Springer
Verlag, Berlin, Heidelberg, New York, 2005. I, 77 pp.
[Roq07a] P. Roquette. Brief an die Herausgeber. Mitt. Dtsch. Math.-Ver., 15(2):8687, 2007.
[Roq07b] P. Roquette. Zu Emmy Noethers Geburtstag. Einige neue Noetheriana. Mitt. Dtsch.
Math.-Ver., 15(1):1521, 2007.
[Roq08]
P. Roquette. Emmy Noether and Hermann Weyl. In Katrin Tent, editor, Groups and
analysis. The legacy of Hermann Weyl., pages 285326. Cambridge University Press,
2008.
[Row99] D. Rowe. The Gttingen response to relativity and Emmy Noethers theorems. In
J. Gray, editor, The symbolic universe. Geometry and physics 1890-1930. Selected
papers at a conference., pages 189233, Oxford Univ. Press, 1999.
[RR75]
A. Robinson and P. Roquette. On the finiteness theorem of Siegel and Mahler concerning diophantine equations. J. Number Theory, 7:121176, 1975.
[Sal82]
D. J. Saltman. Generic Galois extensions and problems in field theory. Adv. Math.,
43:250283, 1982.
[Sch02]
I. Schur. Neuer Beweis eines Satzes ber endliche Gruppen. Sitz. Ber. Preuss. Akad.
Wiss. Berlin, 1902:10131019, 1902.
[Sch06]
I. Schur. Arithmetische Untersuchungen ber endliche Gruppen linearer Substitutionen. Sitz. Ber. Preuss. Akad. Wiss. Berlin, 1906:164184, 1906.
[Sch19]
I. Schur. Einige Bemerkungen zu der vorstehenden Arbeit von herrn A. Speiser. Math.
Zeitschr., 5:710, 1919.
[Sch35a] H. L. Schmid. Zyklische algebraische Funktionenkrper vom Grade p n ber

endlichem Konstantenkrper der Charakteristik p. J. reine angew. Math., 175:108
123, 1935.
[Sch35b] A. Scholz. Die Kreisklassenkrper von Primzahlpotenzgrad und die Konstruktion
von Krpern mit vorgegebener zweistufiger Gruppe II. Math. Ann., 110:633649,
1935.
[Sch37]
A. Scholz. Konstruktion algebraischer Zahlkrper mit beliebiger Gruppe von

Primzahlpotenzordnung. I. Math. Z., 42:161188, 1937.
Bibliography
269
[Sch87]
N. Schappacher. Das mathematische Institut der Universitt Gttingen 19291950.

In Heinrich Becker and andere, editors, Die Universitt Gttingen unter dem Nationalsozialismus., pages 345373. K. G. Saur, 1987.
[Sch91]
K. Schlote. Noether, F. Opfer zweier Diktaturen. NTM-Schriftenreihe, 28:3341,

1991.
[Seg03]
S. L. Segal. Mathematicians under the Nazis. Princeton University Press, Princeton,

NJ, 2003. xxii, 530 pp.
[Ser08]
A.S. Sertz. A Scientific Biography of Cahit Arf (1910-1997). unpublished

manuscript, 2008.
[Spe26]
A. Speiser. Allgemeine Zahlentheorie. Vierteljahrsschrift der Naturforschenden

Gesellschaft in Zrich, 71:848, 1926.
[Spe27]
A. Speiser. Die Theorie der Gruppen von endlicher Ordnung. Mit Anwendungen auf algebraische Zahlen und Gleichungen sowie auf die Kristallographie. Die
Grundlehren der mathematischen Wissenschaften mit besonderer Bercksichtigung
ihrer Anwendungsgebiete Bd. 5. J. Springer, Berlin, second edition, 1927. IX + 251
pp. mit 38 Abb.
[SS07]
R. Siegmund-Schultze. Einsteins Nachruf auf Emmy Noether in der NewYork Times

1935. Mitt. Dtsch. Math.-Ver., 15(4):7p., 2007.
[Ste10]
E. Steinitz. Algebraische Theorie der Krper. J. Reine Angew. Math., 137:167309,

1910.
[Ste30]
E. Steinitz. Algebraische Theorie der Krper. Neu herausgegeben, mit Erluterungen

und einem Anhang: Abri der Galoisschen Theorie versehenen von R. Baer und H.
Hasse. de GruyterVerlag, Berlin, 1930. 177 pp.
[Tau81]
O. Taussky. My personal recollections of Emmy Noether. In J. W. Brewer and M. K.

Smith, editors, Emmy Noether. A tribute to her life and work., pages 7992. M.
Dekker, New York, 1981.
[Tav71]
N. A. Tavel. Milestones in mathematical physics. Transport Theory and Statistical

Physics, 1:183207, 1971.
[Tei36a]
O. Teichmller. p-Algebren. Deutsche Math., 1:362388, 1936.
[Tei36b] O. Teichmller. Verschrnkte Produkte mit Normalringen. Deutsche Math., 1:92

102, 1936.
[Tho59]
J. Thompson. Normal p-complements for finite groups. Math. Z., 72:332354, 1959.
[Tit93]
J. Tits. Sur les produits de deux algbres de quaternions. Bull. Soc. Math. Belg.,
45:329331, 1993.
[Tob97]
R. Tobies. Die Hamburger Doktorin der Mathematik Kte Hey (19041990). In Promotionen von Frauen in Mathematik ausgewhlte Aspekte einer historiographischen Untersuchung., volume 16, pages 3963. Mitt. Math. Ges. Hamburg, 1997.
[Tob03]
R. Tobies. Briefe Emmy Noethers an P.S. Alexandroff. NTM N.S., 11(2):100115,

2003.
[Tob06]
R. Tobies. Biographisches Lexikon in Mathematik promovierter Personen an

deutschen Universitten und Technischen Hochschulen WS 1907/08 bis WS 1944/45.,
volume 58 of Algorismus. Dr. Erwin Rauner Verlag, Augsburg, 2006. 403 pp.
270
Bibliography
[Tol90]
C. Tollmien. Sind wir doch der Meinung, da ein weiblicher Kopf nur ganz ausnahmsweise in der Mathematik schpferisch ttig sein kann. - Emmy Noether 1882-1935.
In Gttinger Jahrbuch., volume 38, pages 153219. Erich Goltze, Gttingen, 1990.
[Van29]
H. S. Vandiver. On Fermats Last Theorem. Transactions Amer. Math. Soc., 31:613

642, 1929.
[Van41]
H. S. Vandiver. On improperly irregular cyclotomic fields. Proc. Nat. Acad. Sci.

U.S.A., 27:7783, 1941.
[vD99]
van Dalen. Mystic, Geometer, and Intuitionist. The Life of L. E. J. Brouwer., volume I.
Clarendon Press, Oxford, 1999. XV + 440 pp.
[vdW30] B. L. van der Waerden. Moderne Algebra. Unter Benutzung von Vorlesungen von E.
Artin und E. Noether. Bd. I. Die Grundlehren der mathematischen Wissenschaften in
Einzeldarstellungen mit besonderer Bercksichtigung der Anwendungsgebiete Bd.
23. Springer, Berlin, 1930. VIII + 243 pp.
[vdW31] B. L. van der Waerden. Moderne Algebra. Unter Benutzung von Vorlesungen von E.
Artin und E. Noether. Bd. II. Die Grundlehren der mathematischen Wissenschaften in
Einzeldarstellungen mit besonderer Bercksichtigung der Anwendungsgebiete Bd.
24. Springer, Berlin, 1931. VII + 216 pp.
[vdW34] B. L. van der Waerden. Elementarer Beweis eines zahlentheoretischen Existenztheorems. J. Reine Angew. Math., 171:13, 1934.
[vdW35] B. L. van der Waerden. Nachruf auf Emmy Noether. Math. Ann., 111:469476, 1935.
[vdW75] B. L. van der Waerden. On the sources of my book Moderne Algebra. Historia Math.,
2:1140, 1975.
[vdW97] B. L. van der Waerden. Meine Gttinger Lehrjahre. Mitt. Dtsch. Math.-Ver.,
1997(2):2027, 1997.
[Wad02] A. W. Wadsworth. Valuation theory on finite dimensional algebras. In F. V. Kuhlmann
et al., editor, Valuation theory and its applications, vol.I., volume 32 of Fields Institute
Communications, pages 385449, Providence, RI, 2002. American Mathematical
Society.
[Wan50] S. Wang. On Grunwalds theorem. Ann. Math., 51:471484, 1950.
[Web93] H. Weber. Die allgemeinen Grundlagen der Galoisschen Gleichungstheorie. Math.
Ann., 43:521549, 1893.
[Wei82]
R. Weissauer. Der Hilbertsche Irreduzibilitaetssatz. J. Reine Angew. Math., 334:203

220, 1982.
[Wei93]
A. Weil. Lehr- und Wanderjahre eines Mathematikers. Aus dem Franzsischen bersetzt von Theresia belhr. Birkhuser, Basel, 1993. 212 pp.
[Wey13] H. Weyl. Die Idee der Riemannschen Flche. Teubner, Leipzig, 1913. 183 pp.
[Wey16] H. Weyl. ber die Gleichverteilung von Zahlen modulo Eins. Math. Ann., 77:313
352, 1916.
[Wey32] H. Weyl. Topologie und abstrakte Algebra als zwei Wege mathematischen Verstndnisses. Unterrichtsbltter, 39:177188, 1932. English translation in Amer. Math.
Monthly 102 (1995) 453-460, 646-651.
[Wey35] H. Weyl. Emmy Noether. Scripta math., 3:201220, 1935. Reprinted in the Noether
biography of Auguste Dick 1970.
Bibliography
271
[Wey39] H. Weyl. The classical groups, their invariants and representations. Princeton Univ.
Press, Princeton, 1939. 302 p.
[Wey52] H. Weyl. Symmetry. Princeton Univ. Press, Princeton, 1952. 168 pp.
[Wha42] G. Whaples. Non-analytic class field theory and Gruenwalds theorem. Duke Math.
J., 9:455473, 1942.
[Wie40]
H. Wielandt. p-Sylowgruppen und p-Faktorgruppen. J. Reine Angew. Math., 182:180

193, 1940.
[Wit34]
E. Witt. RiemannRochscher Satz und -Funktion im Hyperkomplexen. Math. Ann.,

110:1228, 1934.
[Wit36]
E. Witt. Zyklische Krper und Algebren der Charakteristik p vom Grad p n . Struktur diskret bewerteter perfekter Krper mit vollkommenem Restklassenkrper der
Charakteristik p. J. Reine Angew. Math., 176:126140, 1936.
[Wit37a] E. Witt. Theorie der quadratischen Formen in beliebigen Krpern. J. Reine Angew.
Math., 176:3144, 1937.
[Wit37b] E. Witt. Treue Darstellung Liescher Ringe. J. Reine Angew. Math., 177:105115,
1937.
[Wit54]
E. Witt. ber eine Invariante quadratischer Formen modulo 2. J. Reine Angew. Math.,
193:119120, 1954.
[Wit98]
E. Witt. Collected papers Gesammelte Abhandlungen. Ed. by Ina Kersten. With an

essay by Gnter Harder on Witt vectors. Springer, Berlin, 1998. xvi, 420 pp.
[Wue05] D. Wuensch. Zwei wirkliche Kerle. Neues zur Entdeckung der Gravitationsgleichungen der Allgemeinen Relativittstheorie durch Albert Einstein und David
Hilbert. Termessos, Gttingen, 2005. 126 pp.
[Yam97] K. Yamamura. Maximal unramified extensions of imaginary quadratic number fields
of small conductors. J. Thorie des Nombres de Bordeaux, 9:405448, 1997.
[Zas35a] H. Zassenhaus. Kennzeichnung endlicher linearer Gruppen als Permutationsgruppen.
Abh. Math. Semin. Univ. Hamb., 11:187220, 1935.
[Zas35b] H. Zassenhaus. ber endliche Fastkrper. Abh. Math. Semin. Univ. Hamb., 11:187
220, 1935.
[Zas37]
H. Zassenhaus. Lehrbuch der Gruppentheorie. Bd. 1., volume 21 of Hamburg. Math.

Einzelschriften. B. G. Teubner, Leipzig, Berlin, 1937. VI, 152 pp.
[Zas39]
H. Zassenhaus. ber Liesche Ringe mit Primzahlcharakteristik. Abh. Math. Semin.

Univ. Hamb., 13:1100, 1939.
[Zel73]
D. Zelinsky. A. A. Albert. Amer. Math. Monthly, 80:661665, 1973.
[Zor33]
M. Zorn. Note zur analytischen hyperkomplexen Zahlentheorie. Abh. Math. Semin.

Univ. Hamb., 9:197201, 1933.
Name Index
Abel, 226
Albert, 3, 29, 30, 44, 5153, 5673, 75,
192, 193, 210, 212
Albert, Nancy, 3, 71
Alexander, 145
Alexandrov, 54, 119, 135, 144, 145,
148, 228
Amitsur, 50
Aravire, 209
Archibald, 54, 55
Archimedes, 178, 183
Arf, 189226
Artin, 3, 5, 20, 25, 27, 31, 39, 41,
4345, 54, 61, 74, 75, 8587,
91, 94, 112, 134, 143, 145,
148, 152, 186, 190, 215,
224226, 228, 230
Bach, 126
Baer, 103, 106, 112, 113, 234, 235
Baeza, 209, 210, 216, 221
Banaschewski, 235
Bannow, 171
Bartels, 116
Beaumont, 123, 159
Becher, 189
Bernoulli, 177
Bertini, 186
Beyer, 247, 249
Bieberbach, 116, 153
Bilharz, 112
Birkhoff, 144, 158, 172, 173
Blumenthal, 34
Bohr, 151, 164166
Bourbaki, 47, 142, 227
Brahana, 89
Brandt, 41, 54, 55, 145, 152, 171
Brauer, 176, 79, 120, 122, 127, 155,
190, 192, 211, 234
Brouwer, 135, 136

Brning, 156
Brussel, 50
Burckhardt, 41
Burnside, 91, 92, 95, 106
Cantor, 232
Carleman, 107
Cartan, 103
Cauchy, 177
Chevalley, 15, 27, 35, 44, 72, 76, 86,
87, 187, 224
Chow, 171
Conrad, v, 3, 189, 237
Courant, 152, 155
Curtis, 46
Dauben, 183, 184
Davenport, 61, 171
Dechamps, 171
Dedekind, 99, 124, 141, 153, 228
Dehn, 89
Demukin, 249
Derry, 171
Deuring, 17, 24, 54, 55, 69, 75, 110,
118
Dick, 122, 130, 133
Dickson, 4, 5, 17, 40, 41, 51, 5759,
71, 72, 212, 215
Dieudonn, 112, 114
Dirichlet, 99
Draxl, 210, 216, 221
Duggan, 174
Dukas, 157, 158
Eichelbrenner, 40
Eilenberg, 103
Einstein, 136, 137, 156160, 162
Eng Tjioe Tan, 26
274
Name Index
Engstrm, 24
Euler, 177
Faltings, 188
Feit, 53
Fermat, 82
Fischer, 133, 228
Fitting, 54, 55
Flanders, 161, 162
Flexner, 123, 147, 150, 153, 155, 159
Franz, 73, 186
Frei, 164, 195, 234
Frobenius, 47, 142
Furtwngler, 12, 55, 87, 94, 151, 165,
166
Garibaldi, 189
Gaschtz, 78
Geyer, 26
Gilmore, 186
Gbel, 152
Gordan, 132134
Grave, 232
Grn, 77116
Grunwald, 23, 25
Hahn, O., 152
Hall, Ph., 105, 107, 113
Hardy, 151, 164, 165
Hasse, 1226, 233236, 240, 241, 248
Haupt, 176, 230, 231, 234, 236
Hausdorff, 135
Hecke, 74, 86, 226
Heegner, 110
Heisenberg, 120, 121
Hensel, 3, 6, 9, 10, 29, 33, 40, 42, 44,
53, 79, 119, 139, 140, 187,
231233
Herbrand, 15, 54, 55, 87
Herglotz, 195
Hering, v
Hermann, 117, 121, 122, 125, 126
Hey, 7376, 170
Higman, 107
Hilbert, 12, 69, 87, 99, 122, 132134,
136140, 148, 157, 158, 186,
226, 228
Hoechsmann, 245254
Hoffmann, 189
Holzer, 88
Hopf, H., 144, 145
Huppert, 78, 97, 103
Ikeda, 216
Iyanaga, 27, 94, 224
Jacob, 209
Jacobson, 143
Jarden, 245
Jehne, 109
Jensen, 26
Kani, 188
Kaplansky, 68
Kersten, 192, 196, 204
Kiepert, 181, 182
Kimberling, 130, 144, 157, 161, 172,
174
Klein, 133, 134, 137, 157
Klingenberg, 204
Knauf, 171
Knus, 193, 204
Kthe, 21, 22, 54, 55
Koreuber, 46
Kostrikin, 107
Krickeberg, 109
Kronecker, 81, 177
Krull, 54, 178, 186, 187
Kummer, 29, 31, 42, 8083, 85, 87, 99,
139
Kurosch, 114
Krschak, 232, 233
Lagrange, 177
Lamprecht, 109
Langlands, 215, 223226
Name Index
Lasker, 228
Lefschetz, 144, 145, 150, 154, 155,
158, 162, 172, 174
Lehr, 118, 122, 123, 127, 172
Leibniz, 175, 177, 178, 183, 185
Lemmermeyer, 81, 88
Leopoldt, 112, 239243
Levi, 113
Levitzky, 20
Lie, 114
Litvinov, 159
Lorenz, 3, 4, 13, 31, 75, 170, 189, 216
Luxemburg, 175
Mac Lane, 139, 147, 148, 231
Macaulay, 228
Magnus, 89, 94, 104108, 111, 114
Mahler, 187
Meitner, Lise, 152
Mertens, 145
Mills, 26
Ming-chang Kang, 26
Minkowski, 18, 139
Moore, 122
Mordell, 188
Morton, 26
Mozart, 126
Mller, G., 184
Nelson, 126
Neugebauer, 69, 168
Neumann, B. H., 114
Noether
Emmy, 176, 79, 80, 95, 117174,
190, 192, 195, 210, 211,
224226, 228, 230, 231, 233,
234
Fritz, 123, 134, 136, 159
Gottfried, 159
Hermann, 159
Max, 124, 130, 134
Otto, 118, 122, 126, 156
275
nder, 214
Ore, 231
Ostrowski, 232, 233
Park, 118, 122, 123, 126, 127, 147,
156, 158, 173
Perron, 152, 165, 166
Purkert, 229
Rella, 152, 165
Ribenboim, 81, 83
Robinson, 175188
Rohrbach, 84, 211
Roth, 188
Rychlik, 233
Saltman, 31
Schappacher, 164
Schilling, 210
Schmeidler, 138
Schmid, H. L., 101, 109112, 196
Schmidt, Erhard, 109, 133
Schmidt, F. K., 54, 143
Schmidt, Robert, 234
Scholz, A., 88, 102
Schouten, 152, 166, 167
Schreier, 94
Schur, 13, 17, 28, 44, 54, 94, 97, 111
Schwarz, 171
Scorza, 114
Segre, 152, 166
Sertz, 194, 202
afarevic, 249
Shoda, 152, 166
Siegel, 18, 29, 152, 166168, 187
Siegmund-Schultze, 115, 156, 159
Snail, 215
Speiser, 41, 42, 44, 54, 102, 106, 152,
166
Spengler, 153
Stauffer, 156158, 161
Steinitz, 134, 187, 227237
276
Name Index
Sss, 107
Suzuki, 108
Sylow, 77
Takagi, 43, 80, 87, 152, 166, 167, 226
Tate, 25, 31, 45, 143
Taussky, 51, 118, 122, 126, 134, 147
Teichmller, 101
Teichmller, 201, 202
Thompson, 97
Tinsdale, 144
Tobies, 54, 171
Toeplitz, 21, 234
Tollmien, 130, 132134, 137, 164,
168171
Tornier, 116, 169, 170
Tsen, 171
Ulm, 171
Vahlen, 115
Valentiner, 168
van Dalen, 136
van der Waall, 98
van der Waerden, 21, 27, 47, 72, 96,
117122, 125, 129, 142, 143,
152, 166, 167, 228, 230, 231,
233, 236
Vandiver, 71, 81, 8385
Veblen, 123, 145, 147, 150, 153
von Krmn, 136
von Neumann, 72, 119, 162

Vorbeck, 171
Wadsworth, 33
Wang, 26, 50
Weber, 228, 229
Wedderburn, 40, 42, 47, 57, 59
Weierstrass, 33, 114
Weil, 112, 144, 224, 225
Weissauer, 186
Weizscker, 121
Weyl, 71, 72, 117, 119123, 127,
129162, 166, 170, 174
Whaples, 25
Wheeler, 122, 123, 126, 154, 158, 162,
174
Wichmann, 75, 170, 171
Wielandt, 97, 106
Wiener, 158, 172, 173
Witt, 30, 75, 95, 101, 106, 108, 143,
191193, 195197, 202,
204206, 212
Yamamura, 88
Yoshida, 98
Zassenhaus, 91, 92, 101, 103106, 114
Zelinsky, 68
Zelmanov, 107
Zermelo, 235
Zorn, 74, 75
Subject Index
Albert-footnote, 56, 57, 63, 65, 67
anisotropic, 201
Arf invariant, 189222
ArfKervair invariant, 215
Artins Reciprocity Law, 11, 27, 74,
80, 100, 140, 171
Artinian ring, 41
ascending chain condition, 138
Bernoulli number, 82, 83, 112
Betti group, 135
Brauer group, 5, 11, 32, 35, 36, 43, 55,
80, 206
Brauers theorems, 6, 13
BrauerHasseNoether Theorem, see
Main Theorem
Bryn Mawr, 2, 72, 117120, 122, 123,
126128, 130, 132, 134, 147,
154158, 172174
Burnside problem, 106, 108
differential module, 177

Dirichlet principle, 135
elliptic curve, 84, 188
elliptic function field, 62, 80, 195
embedding problem, 245247
enlargement, 181
Erlanger program, 137
exponent, 13, 51, 53
exponent-index theorem, 30, 60
factor system, 16
Frobenius automorphism, 32, 38
function field, 62, 75, 112, 130, 135,
136, 188, 192, 210, 212
generic point, 230
groupoid, 41, 152
Grns theorems
first, 92
second, 102
Grunwalds theorem, 25, 26
GrunwaldWang Theorem, 26, 31, 247
cancellation theorem, 206

central algebra, 4
class field theory, 5, 1012, 25, 26, 45,
70, 76, 79, 80, 87, 90, 94,
101, 148, 187, 195, 224
class number, 80, 85, 86, 88
second factor, 81, 82, 85
Clifford algebra, 198, 199, 201, 202,
204206, 212
cohomology group, 18, 43, 45
complex multiplication, 80
concurrent relation, 180
continuity, 181
crossed product, 17, 201
cyclic algebra, 4, 55, 56, 60, 79
Hasse algebra, 192, 205

Hasse diagram, 92
Hasse invariant, 5, 22, 35
Hasses diary, 30, 87
Hasses norm symbol, 38
HasseArf Theorem, 190, 224
Hensels Lemma, 33, 232
Hilbert space, 132
Hilberts 13th problem, 18
Hilberts 17th problem, 186
Hilbertian field, 186
hyperbolic plane, 201
hyperreal, 178, 179
Dedekind ring, 41, 119, 138

derivative, 182
index, 13, 46, 51, 53

infinitesimal, 175179, 182, 183
278
Subject Index
integration, 182
isotropic, 192, 206
quadratic space, 196

quaternion algebra, 50, 193, 203, 205
Langlands program, 225

linkage, 209, 210, 213
local class field theory, 38, 80
Local-Global Principle, 3, 11, 32, 49,
51, 52, 66, 67, 74, 75, 79,
139, 192, 246248
regular prime, 81, 83

regular quadratic form, 198
relativity theory, 136, 137
representation theory, 46, 47, 142, 155
Riemann matrices, 73, 155
Main Theorem, 2, 4, 5153, 62, 75

maximal order, 41
minimal splitting field, 47, 211
model theory, 178, 180, 183, 184, 186
Modern Algebra, 118
modular representations, 27
monad, 178
Noether equations, 17
Noetherian ring, 134, 138
nonstandard, 177, 178, 181186, 188
Norm Theorem, 12, 15, 16, 55, 66, 75,
76
p-adic number, 187, 232
p-algebra, 73, 201
Principal Genus Theorem, 45
profinite group, 187
quadratic form, 191, 192, 195, 196
Schur index, see index275

Siegels pessimism, 29, 168
SiegelMahler Theorem, 187, 188
skew congress, 53, 54
splitting algebra, 11
TatePoitou duality, 247, 248
transfer, 93, 94
u-invariant, 210
valuation, 31, 232
valuation ring, 178
Vandivers conjecture, 81, 111
Wedderburns Theorem, 11, 142
Witt equivalence, 207
Witt ring, 206
Witt vectors, 101, 195, 196
Zorns Lemma, 74, 235

Roquette P.-Contributions To The History of Number Theory in The 20th Century (2013) PDF

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Roquette P.-Contributions To The History of Number Theory in The 20th Century (2013) PDF

Uploaded by

Copyright:

Available Formats

Peter Roquette, Oberwolfach, March 2006

To my friend Gnther Frei who introduced me to and kindled

1 The BrauerHasseNoether Theorem

2 The remarkable career of Otto Grn

3 At Emmy Noethers funeral

Emmy Noether and Hermann Weyl

Emmy Noether: The testimonials

6 Abraham Robinson and his infinitesimals

Cahit Arf and his invariant

Ernst Steinitz and abstract field theory

Richard Brauer, Helmut Hasse, Emmy Noether

Emmy Noether, Hermann Weyl

Emmy Noether 1933

Helmut Hasse, Cahit Arf, Robert Langlands

The BrauerHasseNoether Theorem

Revised version of:

1 The BrauerHasseNoether Theorem

A preliminary version of this article had been written in connection with my

1 The BrauerHasseNoether Theorem

1.2 The Main Theorem: cyclic algebras

1 The BrauerHasseNoether Theorem

Proof of a Main Theorem in the theory of algebras

1 The BrauerHasseNoether Theorem

1 The BrauerHasseNoether Theorem

generalization by means of the structure of algebras. (This had been suggested

1.3 The paper: dedication to Hensel

1 The BrauerHasseNoether Theorem

1 The BrauerHasseNoether Theorem

1 The BrauerHasseNoether Theorem

1 The BrauerHasseNoether Theorem

Mit der Verbeugung vor Hensel bin ich selbstverstndlich einverstanden.

1 The BrauerHasseNoether Theorem

1.4 The Local-Global Principle

1 The BrauerHasseNoether Theorem

1.4.1 The Norm Theorem

1 The BrauerHasseNoether Theorem

1 The BrauerHasseNoether Theorem

(v) Every division algebra A of index m can be decomposed as the tensor

1 The BrauerHasseNoether Theorem

1 The BrauerHasseNoether Theorem

1 The BrauerHasseNoether Theorem

1 The BrauerHasseNoether Theorem

1 The BrauerHasseNoether Theorem

1 The BrauerHasseNoether Theorem

1.5 From the LGP to the Main Theorem

1 The BrauerHasseNoether Theorem

1 The BrauerHasseNoether Theorem

Tatschlich zeigt Kthe mit seinem Invariantensatz ja direkt, da im

1 The BrauerHasseNoether Theorem

This then settles the Main Theorem.

1 The BrauerHasseNoether Theorem

Such a suffiently strong existence theorem has been proved recently by

1 The BrauerHasseNoether Theorem

1.5.3 The GrunwaldWang story

1 The BrauerHasseNoether Theorem

1 The BrauerHasseNoether Theorem

1 The BrauerHasseNoether Theorem

is our notation for direct sum.

1 The BrauerHasseNoether Theorem

calls K.x/ a pure extension of K since it is generated by radicals.

1 The BrauerHasseNoether Theorem

1 The BrauerHasseNoether Theorem

Q=Z ! Q=Z ! 0