Infinite Series by Bromwich

cr>
CM
CM
BH
EnHI
UNIV.OF
TORONTO
BINDING
AX [NTRODUCTION To TIIK THEORY OF

INFIMTK SKIMES
MACMI-LLAN AND
LONDON
CO.,
LIMITED
BOMBAY CALCUTTA
MELBOURNE
THE MACMILLAN COMPANY

NEW YORK
BOSTON
CHICAGO
SAN FRANCISCO
ATLANTA
THE MACMILLAN
CO.
OF CANADA,
TORONTO
LTD.
AN INTRODUCTION
TO TIIK TIIKoRY OF
INFINITE SERI1>
T.
.1.
LA.
BROMWH'H.
M.A..
F.fc/8
KKI.I
FORMKRLY PROFESSOR OF M
.,
LIMITED
M ACM ILLAN AN]
ST MAR UN'S STREET, LONDON
1908
($>/?
GLASGOW: PRINTED AT THE UNIVERSITY PR1

hV KOHKHT MA< 1.KH08K AND CO. LTD.
PREFACE,
THIS
hook
on
based
is
courses
of
l<
(^m-i-n's Coll.-^.-.
Analysis given
ialway. dm
Hut additions have naturally
LOUS I!MiM!M)7.
Mi"
;it
preparing the manuvript fot press: in parti

r
hapter X I. ami th
part of
added,
in
tinbeen
r,
selecting
subja
n.
ln-,-n
in
!ia\-.-
t> include proofs
>!'
Irrationalzahlen
th-
////'/
Chapter I.
limit ami
a
ol'
in
attempt
Priu^lM.-iin'-
wnendlicher
I\
tli.-oi-rins
oi'
-\<-rpti<.n
In
tliron-n,
all
itli
tVad
n-latin^- to coiitinu'l
preliminary account
COTr
is
i.
the
1'
n-.
n-
ha\e
not
in
this
clia-
the fun-i
to supply arithmetic proof- ,,f
thecrein^ (-..nctTnin-- the existence of limits, l.m bave allow-'-l
ath-iiipt.-.l
their truth
to
the hope that

live
to
rest on an appeal to the reader's intuition

the diseuîon may thus lr mai-i,
Ih-înners.
An
arithnietie treatment
will
!>
f"iin<l
in
where Dedekiml's definition ..i irratinnal num

is
the
adapted as fundamental: this metliod lea- is made ,,f the n
and principles of the >itl-ivntial and Integral Calculus: I
for BOme time l>em c. mvince.l that b
uld not attempt
.1
to
study
Intinite
Series
in
any
detail
1'1>.
17 ;in.l
after
until
H.I.
they
I..
H'Jl).
t Not only
litTf.
>e
Imt
by a
in ni:tny <>th-:
systt nur
i.uin
have been
and ininiinuin limits.
PKEFACE.
vi
mastered the differentiation and integration
the
of
simpler
and the geometrical meaning of these operations.

The use of the Calculus has enabled me to shorten and
functions,
simplify the discussion of various theorems (for instance, Arts.

11, 61, 62), and to include other theorems which must have
been omitted otherwise (for instance, Arts. 45, 46, and the latter
part of 83).
It will be noticed that
made
from Art. 11 onwards,
free
use
is
of the equation
although the limit of (1 + 1/V)" (from which this equation is

commonly deduced) is not obtained until Art. 57. To avoid
the appearance of reasoning in a circle, I have given in
Appendix II. a treatment of the theory of the logarithm of
a real number, starting from the equation
The use of this definition of a logarithm goes back to Napier,

but in modern teaching its advantages have been overlooked
until comparatively recently.
An arithmetic proof that the
integral represents a definite number will be found in Art. 163,
although this fact would naturally be treated as axiomatic
when the subject is approached for the first time.
In Chapter V. will be found an account of Pringsheim's theory

which has not been easily accessible to English
of double series,
readers hitherto.
The notion
uniform convergence usually presents diffifor this reason it has been explained at
some length, and the definition has been illustrated by Osgood's
The use of Abel's and Dirichlet's names
graphical method.
for the tests given in Art. 44 is not
strictly historical, but i^
of
culties to beginners;
intended
to emphasise the similarity between the tests for

uniform convergence and for simple convergence (Arts. 19, 20).
In obtaining the fundamental power-series and products
constant reference
is
made
to the
principle of
uniform con-
vergence, and particularly to Tannery's theorems (Art. 49)

the proofs are thus simplified and made more uniform than is
;
otherwise possible.
Considerable use
is
also
made
of
Abel's
ri;i:i
(Arta, 50
the,,ivm which,
88) on
'-1
\ii
<-.ntinuity of power-series, a
tip-
spit.importance has u-'ially not been

diseased
in
text-book-.
Mately
--ral.lv
contaii
('hapt.-r XI
complete account of thof
not,
nt
ami asymptotic
iitly de\,-l,,p,'d theoriefl
its
nl'
iii
iiP-nt
Beri<
applicatiMiis tM
iM,,k.
tip-
with
inv.-Nt i-at SMUS
tip-
,,f
was
I'l'Min
which
tlp-n-fMiv
tin-
lrl
L21,
t'ir
th--
M!'
and pa
XI. imply an
,,p,
th-
no
pi-intini:
out
writ*-
CMiiM
App.-iplix
inaiiii-
was
ln>ok
Kiiuli^h
h-< IP-IMS
ar.juaiii'
wlp-n th-
ut
in!
Qe0688ary
ti
sr
IL':;.
intinit.-
repared
a\ailaM.-
th.-
systematic
('haj.t.T
COnveTJ
antlim-
nutsid'-
IM -MUM- exteosiODfl
l.-.l
Instance, Arta L18
For
Tin-
exp
has
r.-sults
tli-
t,,
\H'\wf f
JMii-tlh-Mi-y
IH-
mi-'lit
Afl
knnwn
tip-
M!'
'-ontim-d
li
i'linct
qnot
III.,
an
îvin^r
intnuliictiun to tip- throry ..!' int-_i:raN: IP-:

HMH
is (lir ct. ,1 to the
ln-tv.
of
of
and
difference
points
similarity
this tlp-ory ami that
To emphasise tip- siinila961166,
!'
the tests of convrr-viic,- ami of unit'Mnn <-Mii\vrirence (Ai

171. 17'J. are c-alli-d l>y tip- saiip- nanp-s as in the case ol
ami
Second Theorem of M.-an Valui-i-plat-i-il

l,y iip-<|iialitirs (Art. Uiii) which are IMMI-.- obvioc
Connected with Alu-l's Lcnnna (Art
To illustrate the
ral tlu'Mi-y. a ^h<.rt discu^siMii nt' IHrichlet'- inteî-aN and
tip-
traditional 1'nnn of the
i^
'
i.
the (ianmia integrals i^

given; it i- hoped that these pmnt's
le t'niind Imth simple and ri-orou-.
"t'
will
ol
The examples (.,!' which there are over li(M>) include a nuinl.rr
tlp-Mivms which CMiild not
inserted
n the fcert, and in
!
sm-h caaee referencefl are î\-en

ThrMU-'liMUt
the
l.M,,k
in
analysis.
),,,p,.
theinsel\.
I
nl'
that
which present
9,
Settled
'
\VhiK-
\UiK-li
my
-'ml
l>ook has
l.t-fii
in
tin-
l.y
it
the
iisin-
press, three
tn
keep
in
]ml.l.
connexion with
-.
and
BO I'Mrth.
the results -m-,,

books have appeared,
h.
<
MB
Integral* (ch. iv.),
ntctiuns </
aim
dniiMe-limit
in
nl'
my
tlp-Mreins to ev-
most
r<iU;/.
dillt-rentiatiMii
witliMUt ditliculty
roiitums
I'urther inl'm-inatiMn.
have made
view the practical applicationw..rk
SOD
t<>
and Pierpont'a Theory
PREFACE.
viii
Mr. G. H. Hardy, M.A., Fellow and
Lecturer of Trinity
great help during the preparation of
the book he has read all the proofs, and also the manuscript
of Chapter XL and the Appendices.
I am deeply conscious
College, has given
me
much increased by Mr.

and
Hardy's
suggestions
by his assistance in the
selection and manufacture of examples.
The proofs have also been read by Mr. J. E. Bowen, B.A.,
Senior Scholar of Queen's College, Galway, 1906-1907 and in
part by Mr. J. E. Wright, M.A., Fellow of Trinity College, and
Professor at Bryn Mawr College, Pennsylvania. The examples
have been verified by Mr. G. N. Watson, B.A., Scholar of Trinity
College, who also read the proofs of Chapter XL and Appendix
III.
To these three gentlemen my best thanks are due for
that the value of the book has been
valuable
their careful work.

T.
CAMBRIDGE, December,
J.
I'A.
BROMWICH.
1907.
The following list comprises those books of

use in arranging the material
which
have made most
Chrystal, Algebra, vol.
2.
Hobson, Trigonometry.
Osgood, Infinite Series.
De la Vallee Poussin, Cours d* Analyse Infinitesimal?.
Goursat, Cours d' Analyse Mathdmatique.
Tannery, The'orie des Fonctions d une Variable, t.

Cesaro, Lehrbuch der Algebraischen Analyti*.
Pringsheim, Mathematische Annalen, Bd. 35, pp.
y
1.
297-3!)
Reference has also been made to works on Analysis and Theorv .if
Functions by Baire, Borel, Dini, Harkness and Morley, Hohs.m, Jordan,
Lebesgue, Nielsen, Osgood, Picard, Runge, Schlomilch, Stolz, Vivanti, and
various other authors, in addition to the sources iiu-nUnm-d alu.vf and
in
Chapter XI.
CONTENTS
(
'HAITI-:!!
I.
M"!l<'l"!
l'i
|'i"-i
i!i''i]>:<
ami
L.iu.-r
Maximum and Minim am Limit*
Lii:
Intiuir.
17
CHAPTEB
SI:KIK>
01
I'OMTIVK TKK.MS.
'>
;.
II.
'.iiiil.-ii-
;.
('"iii|
Mrin.tk-i:
Logarithm:
Aii"tli-:
EXAMPl
rilAITKK
III.
i\\..
and
Hati"
l>irirhi.
Lfiiiin;t.
Xwtl
Abel
rniatiun.
60
HAl'TKi;
t.
i-:.\
IMP
Ajj)li.-ati"ii>.
i'
IV
CONTENTS.
CHAPTER
V.
PAGES
DOUBLE SERIES, Sum of a Double
Series.
Repeated Summation.
72-90
Series of
Tests for Convergence.

Absolute ConMultiplication of Series. Mertens' and Prings-
Positive Terms.
vergence.
heim's Theorems.
Substitution of a Power-Series in another
Non-Absolute Convergence.
Power-Series.
90
EXAMPLES,
Theta-Series (16)
other Elliptic Function-Series (19-24).
CHAPTER
VI.
95-103
INFINITE PRODUCTS,
Weierstrass's
either Sign.
Positive
Inequalities.
Absolute Convergence.
Terms
Uamma-Product.
Terms.
of
103
EXAMPLES,
Theta-Products (14-20).
CHAPTER
SERIES OF VARIABLE TERMS, Uniform Convergence of Sequences.
of
Series.
Weierstrass's,
Abel's,
VII.
-
Continuity, Integration, and Differentiation

Products. Tannery's Theorems.
of
Series.
125
EXAMPLES,
Bendixson's Test of Uniform Convergence
CHAPTER
POWER
108-125
Uniform Convergence
and Dirichlet's Tests.
SERIES,
(14).
VIII.
Intervals of Convergence.
Abel's Theorem.
Differentiation, Integration.
Theorem
128-142
Continuity,
of Identical Equality
between two Power-Series.

Reversion of Series.
POWER
SPECIAL
SERIES,
Series.
Multiplication and Division.

Lagrange's Series.
-143-160
Sine and Cosine Series.
Exponential
Series.
Logarithmic Series,
Binomial
arc sin and arc tan
Series.
161
EXAMPLES,
Series for
&
u-s
TT
(A. 48)
(B. 30, 31)
Abel's
Theorem
(B. 13-21)
Lagrange's
Differential Equations (B. 34-36).
\i
.II:NTS.
EAFTEE
<
IX.
TRIGONOMETRICAL INVK>HSin.-
IT,
'
Miiltip!-
188
CHAPTEB
<
\M>
'o.MI'I.KX
X.
-
I'!;
le Moj\|.
Complex Numbers.
Ooir
in.
Alisolutr convergence
T--'
llnti.i
l'i-i[i_'-li.-inr>
lencec,
IV
ami
-el's
Tr>t
hii-i.-hk-t's
\\
Oonvergenoe.
'iivk-
1'.'
rnifoiin
<
!'
l-;\lt.iH-nti,i
Sim- ami
Cnsiii.-
S.
Log U
Binomial Srrirs.
Bine
imnM'tri.Mi sM-ir>.
-
gent
Kult-rs
!!-.
it
tnd
In:
Differentiatioo "
and Conne ProdacU,
I'.rrnonlli's
NuinliT>.
an
(A.
The Ootan-
15fi ii'ulli;i!i
Fiiu.-ti.
BZAMPl
-41
Tri
('.
9-11):
An.irli-
\\Vi.-i >u-ass's
12
I)..ul.k-
Tli-.i
Series Tlu-.i VMI
CHAlTKIi
.-(
'uNVKK'iKNT AM'
BORBL'8
AsY.MI'T TI
Historical
I'lililiu^rapliy.
MKMiMi'
Alôluti- Siiiiiinalulity.
-nt iati"ii,
OTHKI;
32).
..
(n-m-ial
Cn-
Addition of
<-i
Suiniiiati
<
Multiplication.
'oininuit y. Di:
Ai.
--in.
OF SCMMATIMN.
and Le
Examples
(I
SKIM
and Integration.
MKTII->I.S
Boivl's
27.
15.
XI.
Intr>iln.-ti<.n.
I'liiulitimi
SuiuiMaM-
lo a
<
UMATION,
Borel's Integral.
MIS
'
vm
of
Kulei-s
Kvil.
heorem.
'
othT
I{.y's
ition.
Minit
Series
Multiph'
ion>.
an
Q
1
I'.oi
of
,-al.
\.iuiplea
CONTENTS.
xii
PAfiE.S
ASYMPTOTIC SERIES,
Remainder
Eider's use of Asymptotic Series.
Formula.
322-346
in Euler's
Fresnel's Integrals.
Integral.
Poincare's Theory of Asymptotic Series.
Logarithmic
Stirling's Series.
Summation of Asymptotic
Stokes' Asymptotic Formula.
Series.
Applications to Differential Equations.
347
EXAMPLES,
Fejer's
and de
la
Vallee Poussin's Theorems
APPENDIX
(5-7).
I.
ARITHMETIC THEORY OF IRRATIONAL NUMBERS AND LIMITS, 357-389

Infinite Decimals.
ations with
Extreme Limits
Lemma.
of-
Algebraic Oper-
Numbers.
of a Sequence.
Limits
vergence.
Dedekind's Definition.
Irrational
Monotonic Sequences.
General Principle of ConExtension
Quotients.
of
Abel's
Theorems on Limits.
390
EXAMPLES,
Infinite Sets (15-17)
Goursat's
Lemma (18-20)
Continuous
Functions (21-23).
APPENDIX
DEFINITIONS
OF
THE
LOGARITHMIC
II.
AND
EXPONENTIAL
-
FUNCTIONS,
Definition
and Fundamental Properties
Exponential
Function.
Exponential Series.
396-410
of the Logarithm.
Scale of Infinity.
Logarithmic
Arithmetic Definition of an Integral
(Single and Double).
410
KXAMIM.KS.
APPENDIX
III.
SOMK TllKOUKMS ON INFINITE INTEGRALS AND G. \.MMAFUNCTIONS,
414-466
Iviinitiim U
Convergence, Divergence, and Oscillation.
Limit of a Sum. Tests of Con \ri-rii.v. Analogue of Abt-l's
Lemma
(or
S>n<l Theorem
of
the
Mean).
Absolute
FAOE*
lMri.-lil.-t
.ui.l
Int.
nif.-iin
>iri<-hl<-t'>
hil'
T--t-
M. ,in.l
,.->
.f
T.--I-
'
--1'g
Intr-iati.
.n.
Liinit-
Sj)-.-i;il
hit.-Jl
Illtr<_ri:i
Repeated
Tli--
lnt-::i;i!-.
<
;:iiniii;i
ii
hunma Panction
<
ili.t
f-
I:
Inti-u'r.-il.
Fi
.--nil.
Stii
mn
i-'-i
Camma
Fun. -ti.
.us (35-47,
55-57).
EASI MISCELLANEOUS EXAMPLE

EXAMPLES,
MISCELLANEOUS EXAMPLI
i-506
EXAMPLES,
iiann's
(19,20):
)isi-.intiinnus
(-FnnctioD
.)U..nu-tri.-al
85);
iNhKX
S.-i-i.-s
SPECIAL
Kunrtimi (32-36)
(44-49);
Hifinann's
(64-66);
J.icolii's
Fiinrtii.ns \vith..,it
Kumm.
OF
(18);
i.
\V.'i.-r>tras>'> \(,ii-|)illrivnti;il.lf
Miami's
(70-77):
St-rirs
Analyti.-al
;Mc ?ow{
LNTBGEALS,
'1
Then-i-ius
li*-ta-
Kun.-i
n
:
'..ntinuati..ns (78-
PBODUCT8
87-100).
AND
7-508
I;AL
INII
509-Mi
ADDENDA AND CORRIGENDA.

p.
16,
The
l."17.
1
series is here
-2 + 3-4+...
supposed to oscillate
finitely
such a series as
excluded.
is
p.
29,
Art. 11, and p. 80, 1. 2.

The integral test is commonly attributed to
Cauchy it occurs in Maclaurin's Fluxions, 1742, Art. 350.
p.
97.
Ex.1.
p.
101.
Art. 41.
A proof of the
Art. 77.
p.
122.
The
Thevalueof the second product
thus,
(sinh7r)/27r= 1-845... (Art. 91.)
can be somewhat shortened by the use of extreme

if
is chosen so that
discussion
limits
is
part can also be given on the lines of
first
we
get at once
l
For methods
determining the region of convergence of Lagrange's

series, see Goursat,. Cours d' Analyse Math., t. 2, p. 131, and
Schlomilch, Kompendium der hoheren Analysis, Bd. 2, p. 100 ;
the relation between these methods will be seen from theorems
due to Macdonald, Proc. Lond. Math. Soc., vol. 29, p. 576.
p.
141.
p.
146, Ex.
p.
164.
Ex. 25,
p.
190.
Ex.
of
The numerical results should

23-14... instead of 23 '00.
18.
pp. 212, 213.
This
is
= v(u-
omitted from
is
{(1
and
instead of 4 '80,
-xy)(l -xly)}~
1
.
taken from the Mathematical Tripos Papers, 1890.
assumed that
It is
v(w)
- 1
The index
4.
1.
be 4-8105...
v is
a function of w w ith a period

r
2?r,
so that
2?r).
p.
226.
Professor Dixon's own version of his proof has just been published in
the Quarterly Journal of Mathematics (vol. 39, p. 94, Oct. 1907).
p.
227.
1.
8.
The reference should be
p. 323.
1.
3.
The
410,
1.
4.
p..
to Arts. 44, 45
(1),
instead of Art. 49.
should be 8 instead of
the following four figures are 6060, according to Gauss.
last figure in Euler's constant
5,
and
The proof that 2 > 8 n and a < S n can be made purely arithmetical
2', <r', the sums obtained by superposing the two modes of
by using
We
division.
2, and a'
similarly
is
<r
have then 2'rxr', while 2' is less than both

2'
Thus 2
a'
greater than both s n and <r.
>
< a' < 2' < Sn
> >
N,,
*
and
ami
Similar integrals occur in Electron-Theory

Soinmerfeld, (iitttimji n Xachrichten, 11MI4, p. 117).
p.
471.
Exs.
p.
490.
Other examples of a simple character

Non-differentiable functions.
have I.e. M -ixen recently by H. von Km-h. A<-tn Matin ninfifi. \'\.
/hnt*>-l, n Math.
30, 1!K)7, p. 14",; and by Faber, Jahr, *l rirhf <1>
20, 21.
Verdi,,
p.
495,
Ex. 51.
!',<!.
The
Hi,
1JI07,
function /(a,
p.
;/i
(compare
:>3S.
is
Mippo.-etl monntniik-
with respect to
n.
AN INTRODUCTION TO THE THEORY OF

INFINITE SERIES
CHAPTEB
I.
SEQUENCES AND
LIMITS.
convergence and divergence.

we have agreed upon -MIL- rule, OT rul<
which we can BSSOCiate a definite numher ,/, \\itli any .i-H^ned
Infinite sequences
1.
SiipjM.s.-
that
positive inte-vr
,,
then the
of numl..-!^
set
'
"i-
BO as to
arranged
-
1.
'!.
c<
will
tin-
notation (am) to
the
set
i-allr.l
an
to
IM-
shall
\V.-
siiujily
!!-
H-IV>JM m<
/'
of
]nsit:
infill
timl
l'r.-i|ucnt ly
it
convenienl
this
'I'h'lisequence,
the word infinite simply means that every t.'i-ni iii the m-iju.
is followed
hy anothei- term.
The
ome
rule
detinin--
loi-iiiuln
the >e|n,-iice
formulae)
1.
Ex.
2.
Ex.3.
l>e
expl
_d\ in-- a
from the preceding
be determined, either directly or
Ex.
either
may
BS an explicit fui;
verbal Statement which indicates IK,\
(oi-
me
i--]ir-s.'nt
-Id
I'
Tlir
ni'_C
imnil.rrs
1.
li;u iM"iii<
I'M.'
rm
t.-nns.
thr
't'
rati.-iial
positi
arm
of
to
ami
80
this
nu-tlr
.f
UK!
-\\
ill'-
ini>
>.!
h'T'-
aiiuvm.Mit
ami
.1-.--
I--
'-an
n..t
'
!"
It-ail
anv OOITeBpOod
the
..'to
si-<|in-ii.
it
mil-
<>f
llx-ii
"hrir inn.
SEQUENCES AND LIMITS.
H.
[(
I,
The most important sequences in the applications of analysis

are those which tend to a limit,
The
be
limit of a sequence (an ) is
found
to
correspond
small, such that
provided
said
< an m.
only that
more convenient
It is generally
to be
to every positive
contract these
to
two
in-
equalities into the single one
\l-a n \<6,
where the symbol
used to denote the numerical value
is
\x\
of x.
The following notations will be convenient abbreviations for

the above property
:
I = Km a n
or
= lim an
or
a n -+l
the two latter being only used when there

to what variable tends to infinity.
Amongst sequences having no limit

those with an infinite limit.
A
the
sequence (a n ) has an
number
may
be,
n>
it is
useful to distinguish
can
be
found such
that
>N,
771.
This property is expressed by the equations

or a.n ->>oc
lim a n = oo
or lim a n = oo
;
no doubt as
is
no matter how large
infinite limit, if,
an index
an
provided only that
>-oo
><x>
In like manner, we interpret the equations

oo
oo
a w ->
lim an =
lim a n =
,
oo
n >oo
In case the sequence (a n ) has a

convergent and
sequence has an
Ex.
'oo
Ex.
1 big.
is
said
to
converge
finite
to
infinite limit, it is called
With a w = 2w
(the sequence
of
limit
I,
it
is
as a limit',
called
if
the
divergent*
odd integers) we Imvj
a diverge/it sequence.
ftfa.
With a n =l/n
(the harmonic sequence)
we have
-*<);
convergent sequence.
*
but it
<lir<i-<j< at us equivalent to non-conn r<j< nt
distinguish l/<-t\veen sequences which tend to infinity as a limit and
We shall fall the latter sequences <>.<<,-i//<tturit (Ai
those which oscillate.
Some
writers regard
.oii\ cnit-nt to
1]
Ex. 3
til
It
ler
liiHttrl
ln>
h<
liu
in-
in
tin*
s.Mjii.-li'
tli.-i.-
-.
will
,i!.v;i\-
!:
unlimited nun.
\V- shall tind
MM
;iiul
convenient some! im
it
an OrdlE
iiid;.
:>hirally,
'
abfiG
<-<|iial
tin-
pirtun-d hy joining
In
liin-.
tli'
|U.-n<-.-
'
points with a l.ml
OOni
\\holly within a liuri/mital .strip <>!' \\i<lt!

rtain valu-: if tin- sequence is divergent, fch<
has passed
al')\,wholly
(or below) a certain
i-taiii
valur.
lii-
'I'in-
It
^rajhical i-rjirr^-ntat
alr.-aly considered
i-
will
srrn
1)0
tin-
iliM-T.-ini
tin-
s.-cMinl
>n
that
tin-
tirst
converge, and
is
9
makt
'//
Smdi
a definition
may
aa
n f 4)
taking
as .small
M W
and
tin-
thr--
tennfl r<-j'r-s,-ntrl
is
in
111
oft.-n ]oo>,-ly
th
/'//M'/
id-
'
which
third to <>scillat- (see Art
<!
Hd-.-. l.y
tmns
/.
I-
Been tY-mi th
l>-
tin- t'.-w
s.-.junin-
tin-
Tin- definition nf a limit
initial
given
a ^lann- that
at
sln-w
tin-
!'
i-
'/'/"
is
may
li.-
points
in
sr
the
'
if
pos^jl.i;
'
'
h
is
,'\
en.
large enough,
i-ut
r(,/ -f
tin-
tin-
we can
tin-l
berm
whirh
I.
[CH.
(2) Infinity.
remembered that the symbol oo and the terms

infinity, infinitely great, etc., have purely conventional
It is to be
infinite,
in fact, anticipating the

that infinity must be regarded
q# an upper limit which cannot be attained. The statement
that a set contains an infinite number of objects may be under-
meanings in the present theory;
definitions of Art. 4,
we may say
stood as implying that no
number
count the
suffices to
set.
Similarly, an equation such as lima n =oo is merely a conventional abbreviation for the definition on p. 2.
In speaking of a divergent sequence (aw ), some writers use
The numbers an become infinitely great, when
phrases such as
n increases ivithout limit. Of course this phrase is used as an
:
equivalent for tend to infinity
we
but
shall avoid the practice
in the sequel.
(3) It
is
evident that the alteration of a
finite
number
of
terms in a sequence will not alter the limit.

Ex,
4.
The two sequences
have the same limit
Further,
terms from
the limit;
1,
2,
3,
4,
1,
i,
i, J,
J,
i i
i,...
*, fc..
zero.
evident that the omission of any number of

a convergent or divergent sequence does not affect
it
is
but such omission
may
the character of o
change
oscillatory sequence.
Ex.
Thus the sequences
5.
1, 5, 9,
limits as those considered in Exs.
1,
13,
2.
... and 1,
|, 1, ... have the same
But the omission of the alternate
;\,
t'ini> in the sequence

1, fc I, i,
changes
it
!,*,*,...
into a convergent sequence.
(4) In a convergent sequence, "//,
or none of the terms
may
/////"//////.
be equal to the
limit.
Kxainplrs of these four possibilities (in order) aiv

1;
the limits of which an-,
= -sm(iw7r);
in
"N
=
W
m-drr, 1,0,0,0.
-^ +
ji i/ifc
îvm ly
a-^J
mi mix
1,
IfONOTONK
2|
We
ive
'
shall
si
hOW
rid
>//"'//.
A'
l.ut
II. \\.-\rJ-
choic.-
denot,-
an
iii'i
<!'
'lie. -s
n>
^>
,>
N.
j,
limit
\}'
can
mal.-
can
w.-
than
that
hy
>ull
tin-
hypoth.
can
I"-
lini",,
<
such that
\.
rarily small
arl-it
which cout ralicts
this, supposi-
then
wer
ii'
ol'
we could
we two
can
f,,,
".
-|
emphasise
ri-rtain
K'-r
/
i"
mal.-
leSfl
than
ch..,
ditr.-rciM
j.i-ovidcd
///
hat
wln-n-
Special '\aui]lc
hy takin-
denote an arl.itrarily
to
hail
c< Hi \
s.-ijii,
th-
positiv,- nunili-r.
any
ha]iM-iis
than
thru intVr that /_'.

not ma'',
98 than
A-
(or
xnliu-t ilil.-s
is
it
tliau
'')
avoids roMt'usioM
it
M.'ii
lesa
!)
/,
whir!
861
are usually added
fco
with
'/
variahl'
In-
ily
irord
,.-
b<
Inn
unii
is
!!.
al
numher:
tive
that
..!'
l.ut
shall
<!
i\
\\.-
to
usually employ
iiuiiil..-!'It,
added
-i
that
//>///.
Thus
lini
/>
only: hcnc.- we must hav<- lim/;

It
should
Bequencee
U rh
that
..... l).s,Tv.-l
that
,,.
K..r
the
Thus
imit.
hif'|Ualit\
<
easily ha].]M-n
that
lim
j.osi
'-.uiclusion
th-
\'v\\,
lim
2.
ii'
Ithou-'h constantly
.lit!
'
lin
i(
may
it
lim "
Monotonic
sequences
<i,,
:_
lim
and conditions
for
their
con-
vergence.
iu
an
yah,
and
which a
tor all
and
calle
vulu-
all
similar.;
1
ied
in
ailed
The
first
[CH.
may now
general theorem on convergence
I.
be stated
monotonic sequence has always a limit, either finite or

the sequence is convergent provided that \an is less
than a number A independent of n; otherwise the sequence
infinite;
diverges.
For the sake

that a n
is
detiniteness, suppose that

tt+1
constantly less than the fixed number A.
of
however small the
positive
fraction
may
and
Then,
be
will
it
be,
such that a n <â m -\-e, if n^>m',

for, if not, it would be possible to select an unlimited sequence
of indices p, q, r, s, ...
such that ap >a^ + e, a q
p + e,
possible to find an index
>a
ar
>a +
g
e,
as ^>ar -\-,
and consequently,
etc.;
after going far

should arrive at an
in the sequence p, q, r, s, ... we

A contrary to hypothesis.
index v such that a v
if
we
the
Thus,
graphical representation described in
employ
the last article, we see that all the points to the right of the
line x =
will be within a strip of breadth e; and that the
enough*
>
breadth of
by going
the
representation
enough
it
approaches some
made
be
can
strip
far
the
to
as
small
From
right.
we
as
the
please
graphical
appears intuitively obvious that the sequence

limit, which cannot exceed A (but may
But inasmuch as intuition has

this value).
in mathematical reasoning,
led
to
serious
blunders
occasionally
it is desirable to give a proof depending entirely on arithmetical
be
equal
to
grounds; such a proof will be found in the Appendix, Art.

Ex,
1.
which
is
As an example
14!>.
consider the increasing sequence
8.
i>
>>
represented by the diagram below.

i
1234567
FIG.
In this case we
may
take
4.
A = \, and
there
is
no diHirulty in
of oourse \u- might
less than .1.
but
that the limit of the sequence is equal to .1
have taki-n .1=2, in which case the limit would In;
*The nmnher
of
terms to he taken
in
the
seûmce
equal to the integer next greater than (A -i)/c.
/i.
</,
/,
v.
..
would he
KONOTON*
2|
Ex.
2.
si-
that
iv
the sequence
Ait.
||i-iuli\.
in
I'.ut
">s
ease
then
;1
If
0<r<l,
j..Mti\.1
('niisiih-r
3.
> _ o.
/
the
and
!''y
/'
l ^
'/.
tin-
s.Mjufnce
m+j
in niodifyini:
case
ie
tt
0,,
a sequence which
n-
= ^.
ulily decreases but tin- ti-rm-
se<|
approadirs a
can lind
wt-
+00,
^m.
no ditliculty
tr)
so that,
inl,
iinl
tin-
if
.1.
r..nsfi|iuMitly
'I'hu-
that
plain
iva.lcr will havt-
BO tliat
Ex.
lit,
i.s
is
it
th-
is
jn-.,ji.-i
numbrr such
always an
may In-, thnv
>
Thwork
ni..ii"t,,ni,-
n<>
1
> .1.
ainl tin-
th-
!'
obtained
limit
formal pr.mf
tli
that
\
(1
//<
^<r"<^ +
J<r" + <r(/ +
t,
limit
such
that
that
s.i
n>r/i.
if
<r(/ +
II.
dftinitt*
:".nd to
'
);
itly
/(I
is
-jiiality
\Vhi-ii
Thus
/>
1,
it
follows
>
\v- find /"
iaMi>h-d
MI
inn-,
fi'.'in
-;)</.
however
Miiall
U-,
ty "f tin-
Mquenoe
An
1.
(-l)".|r|"
tin-
sr.ju.'iH-e
can be drtfrniim
btained.
"]>,
= 0.
:it
t'
It
havi-
we can
It
Sniinuinjr
\vt-
t'lMin tin- iimiiMtMnii- projii-i
K.\.
may
th- last result that
''"it
-Kr<l,
r"-*0;
1
"i
by
oo,
r2n+
r=-l,
^=
r-" +1
Take next
1,
[CH.
find
I.
=-l.
o n =r/w!.
p be its integral part. Then the sequence (a,,) decreases

and since
is positive it follows
steadily, after n exceeds the value p
that a n -+l ^0.
If r is positive, let
()
Now
n+ I
an
Thus we can
find
..
n+2
2n
a-2n
3,
so that
and
Hence, as in Ex.
Zn
we
>TO
<a
obtain
l<^(l + t) or
It follows that
When
= 0.
r is negative,
we
obtain the same result by writing
a
Thiis for all values of r
= (-l)". |r|"/!.
we have
lira
n\
3.
=0
General principle of convergence.
If a sequence is not monotonic, the condition that a

constantly less than a fixed number is by no means sufficient
to ensure convergence
this may be seen at once from the
|
sequence given in
Example
The necessary and

it
may
positive
e,
Art.
1,
for
which
0<a n <l.
sufficient condition for convergence
is tin it
find an index m, corresponding
to any-
be possible to
number
3,
such that
n greater than m.
Interpreted graphically, this implies that
sequence which are to the right of x = m,
for all values of
of breadth
2e.
The statement
is
all
lie
points of the
within a strip
then almost intuitive, since
the breadth of the strip can be made as small as we pithy going far enough to the right; an arithmetical, proof will
be found in the Appendix, Art. 151.
GENERAL PRINCIPLE
2,3]
OONVJ
"I
the
Ex.
J>
012345678
Pic
hieh
aiid
is
it
>.-t-n
easily
that tin- limit
is
Tin- diairiain
1.
CAT
above
must
der
test of
;i
(even
wunu-d
is
certainly
f<>
^,II>IH,S,j
-l'-l
arbitral
then
/')=0,
B6qU61
\J
be
l."7.
no
ditliculty
ndea for calculating
!//,
6ll)=limall
Iim(all
*///
>'//
liui l..Lr(l
-r/*
W MO
tin-
I'.ut
have
will
ivjulrr
//.
Appendix, Ait.
lnit
/////
~
,
a/rbtirary function
suppose that
|=
conv*
f<rr
necessary,
//-/'/A
/'///////'///
HIM
seen t'mni
tin-
>.
This condition
/////.
as equivalent t<> a condition sometii

in hooks which an- i;vnT;illy
nann-ly:
<-oii\ .-I'Li-'iicf
condition
may
i-
will
it
in
that
prnviii^-
ll>
limits a
liin6
linn
]>r<\ idi-d
that
]T"\-id-d
tli
tin-
nt.
sequencer
liu
And
and that
^-n. -rally
liin/(" a
where
,.)=/(HmamJ Iim6w>
/
.
demurs
1:
that
any
n>iiil linat
in
of
th-
lim
l'<>ur
-l-iM-ji-
operations, MiKj.vt to conditions similar to those ah-.-ady
sj,,.ritii-d.
10
[CH.
I.
If the functional symbol contains other

operations (such as
extraction of roots), the equation above may be taken as a
definition of the right-hand side, assuming that the left-hand
side
is
indices
found to converge. On this basis the theory of irrational

and logarithms can be satisfactorily constructed.*
remembered that the
It is to be
limits
on the
left
may be
perfectly
To
illustrate
definite without implying the existence of liman and lini b n

= (-!)", b = (- l)"-^! !/).
this possibility, take
tl
a n + b n = (-
Then
n~l
l)
a n .b n =-(l
and
/n
(a n
+ l/n)
+ 6,,)-+0,
(.&)-> -1,
a n /b H =-n/(n
so that these three limits are quite definite, in spite of the non-existence of
Km an and lini bn
.
6 n ->0, we cannot infer that n /& n ôo

that
a n /b n has a fixed sign.
proving
If a n ->0, and 6 n ->0, the quotient ajb n may or may not have a
If a n is convergent
without
and
first
limit (see Appendix, Art. 152).
Thus with aw = l,
= (-l) n /n, we
bn
so a n /b n oscillates between
Again, with a
lt
= l/n,
bH
and
oo
=(
see that b n-*0,
+ oo
but a n /b n = (-l)"n and
1)"/X the value of a n /b n oscillates between
and +1.
When one of the sequences diverges (say n -*oo ) and the

other converges (say to a positive limit) it is easy to see that
<
(a&n)^c
a n .b n -+x>
n /6 n ->oo
6 n /a n
when
-0
and then the

and (a n/b n ) need special discussion.
both a n and b n diverge to oo we have
the only case of exception arising
6 n ->0,
.sequences (an .b n )
Again,
if
but both (a n
bn )
and (a n /b n ) have to be examined specially
Appendix, Art. 152).

If a w ->oo and fr w ->oo
there are three distinct alternatives with
rc-pect to the sequence (a n/b n ),

(i)
aw /6 n ^0;
(ii)
*<nin,n/ tlmt
ajbn ->k>0:
TlniH \vc can define A-V^ a8 lima?"", where

n \\hcn
iunl /
ate botli nionotonic,
(
;.
.\|.|i.-ii(lix,
Art.
l.VJ,
(see
Kx.
4).
(a,,)
(iii)
= l,
it
a.
tlu- srijiirn,
J,
/>
H.
-*\-.cd not con-
hi
UM
\M- LOWER
rri-l;
3, 4j
11
in
than
////
eon\ enient
In
>'<.
eastr
(ii)
<'
i'
use the notation
to
wbn
iven
Kill.
Mm
mhiat ion of
the
in
Appendix
a number of
in
<
Ait.
the d
in.
practical work.
in
Upper and lower
4.
limits of a sequence.
rm
tin-
for
\:,'2\
cases which an-
.Mid
upp
l.-a^t
l.rm
I.
r.ut
quence
term,
may
such
imlici-.s
he. th--rc
I'ui-th.'i-. tln-rt-
tli
t.tln-r\vis.'
!:
similarly.
\\
th- lower limit.
callcl
is
it
this
//.
is
i'oll.\\-
it
always
lar^-r
an
is
tliat
inimhcr of such
in
fcerm
ii
Iliake
definite
n.
in-1-
the
tile
Sllpp'sr j, I,,
index sati^l'yine- the required condition.
Hence t)i- terms
the sequence which fall het ween am and Op are all
J)
\\
'
than
a,,
and
('lioose
now
Succession
"^""r
such that
/>
">_-
Pt>j
P*>Pl>
/'!>'-
and
values of
<>f
for simplicity den..'
Then WG
-truct.-d
moiiotonic sequence blt

and this se.jUence
limit (Art. 2). either a finite numl.er // or ao.
If lim
eaii
how
find
small
///
BO that
is
e<-/
/
//'
ll
bm
///.
<
,
f
if
>//>{
lim
h,,\\
lar-e
and
A',
A'
if
of
fi-
can find
provided that
ma;
//
\\-
lin.
>
no mailer
prn\ ided that
//,
'Hi/ iiff.iinn/
Similarly,
hetweeii
lies
and consc,|Uently we ha
he:
may
//
II
>
,i>
so that
//.
has a
no m.r
//.
12
[CH.
I.
If the upper limit of the sequence is H, whether attained

or not, the sequence has the two following properties:
No
term of the sequence is greater than H.

least one term of the sequence is greater than
(ii)
however small e may be*
But if the upper limit of the sequence is oo, it has the
(i)
He,
At
property
An
how
infinity of terms of the sequence exceed
great
may
modify these
It is easy to
oo
and
definitions
refer to the lower limit (h or
no matter
be.
).
FIG.
results so as to
1O
11
12
7.
The diagram gives an indication of the mode of selecting the

and h\ these are represented by dotted lines.
sub-sequences for
Ex.
1.
Ex.
&
and
3.
0,
which
(Art. 1)
oc
A = l, because
is
an =
is
l/it.
the greatest number in the sequence and A is

not actually attained by any number of the sequence.
-i,
is
1,
Here the sequence (&)

is
is
in the sequence.
(Art. 1)
J5T=1, because this
seen to be
so the upper limit
2.
Here
Ex.
a w = 2?i-l.
(Art. 1)
Here we have
the least number
1,
is
|,
found from the sequence
indicated in Fig. 3 of Art.
f,
,
f,
|,
f,
f,
...
and gives
^, J, |, ^,
f,
....
to be 0.
...
by the dotted
H=l
and similarly
These sequences are
lines.
Maximum and minimum limiting values of a sequence.

We have seen in the last article that any infinite sequence
5.
lias
upper and lower
Consider successively the sequeiuvs
limits.
a lt az a s a4> a
,
":;
fi
"
";
".,-
",.
and so
*
If //"is not attained,
,...,
"r
>
<V".
>
on.
there will be ;m infinite niimlirr of surh terms.
4,
MAXIM
5|
\M. MINIMI M
ami I0W6T
i|>jM-r
//,,
II
Thru
//
>
:
trnii in
in-
the sequence
a
Jl
i
which 0886
in
//
jnonoh.nir
i-
>
Thus
//
.t
<-d
l.y
w-
miM
",
^liall lia\ e //
//
//
//
//
givee a limit
.-UP!
th--
!>
or
/:
Similarly
limit
in
//.
<
limit- 1- d>-n<
iae
//.
Mil v
BO OIL
sequ<
cam
M all
<Ar
A,
//,.
may ha\r //,-/,.
\\<-
an.
'
:
1.
OftSe
//
//!
= h-> = h i=
...
may
It
gTOr-foo,
be DOticed
... = +
//
//
ran only
ran
mily be
tliat
and
3c:
.'/
-f
00,
'/:
ii'
is
It
that
notice
t<
imp.M-tant
/-
CCW1
./
//"
14m
rithrr
//
//..
,<//
...
thr sub-sequence
t> tllr >.M|llr]ir
Ic'loll--
for
i-
<>
II
//
thru
tVi.ni
//,
is
in,
lim//
small
<,['
if
con
is:
Bub-Sequel
BO that
///
An
(i.
have
G
///"//
//
drtinition of
tin-
ly
and
',.
</"< n
tha/n
greater
ran find
luit.
...
//.
#ub-&
///
//'
is
n'l,ii-/i
= ^r we
lt
''
which
//
<
II
a c-n'tain
suh-sr.jii.-nc,- d.'tin.-
ajiplirs to
arûiwnt
o
thr limit
("). so that this samr
similar
it
Itself
'I
with
rr.incidmt
!'
//.
&+,
no
\v-
have
//,,.
matt.-i-
//
so
hat
Thus,
//_
if
(j -f-e,
///.
thr limit of any num-r^-rnt

8ub-8equenC
nm-t ha\r /i.^'-j-e: and.
rlutrarily small.
this
\\c
In
likr
mann.-r
0.
requiree
prove that m can
found to makr a,
if
and
drdnrr tlia
//-e,
ii'
is
Thr two properties

'
in
rMaMislird justify us
just
ant/
Thr symhol
Mm"
is
<
'
.
<
1 i
ralli-
th.-
/< lim a <
/'
maximum
impli.-s tliat
/'.
hapjirns that ( = x \\r ha\.th<Tr must hr an intinity of terms

it
in
n^-d to denote rithrr
limit: thus an
inequality
/<</ and d
If
MM
//<>'// /'///
symbols we
.
minimum
llf
;/
,'
//
s
irr
an-;
than an\
..-ntly
14
number N, however great;

an
!)
terms
infinity of
when y =
similarly
than
less
[CH.
On
N.
there must
the other hand,
Inn a n =
I.
if
oc
co
it is easy to see that
and similarly
=4-oo we must have limaw =+oo.
From what has been explained it is clear that every sequence
has a maximum and a minimum limit; and these limits
liin
lim/< u
if
are equal, if, and only

It is convenient to
the sequence converges.
if,
call
maximum and minimum
sequences
oscillatory
We
are
limits
when
the
shall
call
unequal.
these limits the extreme limits of the sequence, in case
wish to refer to both maximum and minimum limits.
It will
maximum
be evident that the
we
limit coincides with
upper limit, except ^vhen the latter is actually attained by

one or more terms of the sequence; and similarly for the
minimum and lower limits.
the
123456789
FIG.
1O
8.
The diagram gives an indication of the process. The points

and h n are marked with O and are joined by dotted lines.
Hn
Ex.
In the sequence (Ex.
1.
we have //"=!, A B =
A. 3.
i t, i i
# = 1, g = Q.
so that
Art. 1)
3,
Here it is plain that convergent sub-sequences can be selected

any limit between the extreme limits. Thus
i
and
With
8,
//
1
=/t.
With
3.
1,
-2,
...gives the limit
fa ...gives the limit -L.
v*
=
J
//
#=
ErS,
!
J,
:5
it
limits
will
<.tli-i-
(-])-i(i+l)
//,
= !!,...
=-^
A,,
QO
</= -oc.
he seen that
tin-
o-(-l)
<:,...
5,
/(
than
//
ll
-1.
i.
= 00,
6' = 00,
and HO
//,
|,
-if, /<3
3,
//
tiinl
,,
fc.-fc...
//,
Ex.
In
2,
6 = l,
80 that
-f,l,
J71 =
get
ami
\vc
f, ft,
Ex.2.
we
f>
to give
IH>
suit -M'i|inMH''s
cxticnir limits.
can
In-
t'ouiul
to
5,
6|
OF AN
MINITI: SERIES
reme
der will timl some ralee
Exa
limiis in
6.
Sum
we
that
17
the
'-n-l
of an infinite series
lY-ni
il'-tlm-.-
;iMlt
tliis
thi- cha]
..i
addition of two series.
a second
.s.-.|M-iire
'hat
x
Thrn
= "l + ":i +
l=
and
th,
/'/'
A">
/A*
lim
tl
i
"+
+-- =
'
*!//,/
//,.
'
It
||iiw.-\
is,
that
is
"
!ii'l;nii-nt;il
unl acconlin-'ly care must
'in it:
nme without jrnt'

rily tnir
aiv
Similarly, if
that
-f
nsr',ll,ih
to
/'-
1'"
'//'///
tinitr
or oedtto
div(
/,/
1.
oi
propi-rtit's
reap
-'
/)
1.
\vh.-n
i,t
/=!,
,,
and
=w.
\\h.
-1 <><!,
|r|
\\hrr.'
i>
|rj"<
I
iy tin- liiniiinial tln-oi
(1
+ </),
(1
cm.
Thus
liinr"=0,
\vhifh was
ill
it
is
lim*
nli\ j.ius
*oo
indf]'-nd-ntly in
ulitaiiii-d
II.
sd that
Tln-n
jM.sitivf.
that
\,
'.
Kr<l.
-_
/-,
lim
and KOOOrdil
/
ll;
When
is
1.
1.
and
\vr
(
II-
have
-r)">
f 6),
d.l.
H<
iniii'l
tak'-n DOl
IM-
sud
t1><
/
Ex.
familial-
limits
i'or
in
iiii])(.i-t;inc.-
~2.
MUMS
16
Thus
lim
arid so the series oscillates

If
r=
= l,
sn = 0,
and
We
=-*
lim
between
sn
1,
Thus the
sn
oo
s,,
= + cc
and +
if
ig
odd,
if
is
even.
oo
[CH.
series again oscillates.
have now
justified all the statements of the enunciation.
It follows at once
from the
results stated in Art. 3 that if
=a + a + 3 +... to oo
T=& + &2 + & 8 +... to GO,
a.
and
then
fifT=(a1
51 )+(a2
The
rule for multiplication of S,

readily (see Art. 34).
62)+(a8
should be observed that the insertion
It
series
equivalent to the
is
68)+....
does not follow quite so

of
brackets in a
a sub-sequence from
selection of
the sequence (s n )', and since an oscillatory sequence always

contains at least two convergent sub-sequences (those giving
the extreme limits), it is evident that an oscillatory series can
always
made
be
to
converge
by
grouping
the
brackets; and, conversely, the removal of brackets

a convergent series to oscillate.
Ex.
The
2,
and
306...
to the
1-306....
It is
sum
series
1-306...;
'306...,
[-306. ..
any
l-J + -f + i-f + --. oscillates between

series (1 - |) + (|-f) + (f-f) + ...
while
-(i
-)-(-)-
number
...
in
cause
the values
converges to
converges
the
sum
see Arts. 21, 24, 63.]
when we
are only concerned with deterconvergent or not, we may neglect

of terms of the series; this is often con-
evident that
finite
may
but the
= l-log2 = - + -+...,
mining whether a
terms
series is
venient in order to avoid some irregularity of the terms, at

the beginning of the series.
In particular, it is clear that the series
The sum of the latter

ore simultaneously convergent.
terms of the former.
the remainder after
is
often
17
1.
1
Arts.
1.
It'
3.
+ /'.
i
3.
here
inn. us
'.
<
fin.
:.!
if
.!,
if
Ex.
4.
I'
5.
t" tin- limit (a
+ /.
If
6.
prove that
..
(<>,
/,<,)
an
!'
(!+/).
In
tl.-
'
tllf
and
::iiiniii
in.-
similarly
val
tin-
iken,
/
7.
If
:i"ii
limit.
If
:IM"M linn:
!>n
limit
is
(sinli
8.
tli--
;n nlniM-tic
ami
li;irm..n:
'nil- ami
9.
tlif
If
".,- Ac
aiithiiM-r
ii.'ii,-t,,ni.-
ami
and
i
la
TT
La
rl
hen
inmoii limit
ri"
tip-
are
18
[CH.
10. In general, if a H+ i=f(a ), and if (a n ) converges to a limit I, then I

must be a root of the equation x = f(x).
But there are two limits which can be derived from (a n ) by means of
sub-sequences, they must satisfy the equations x = f(y\ y =f(x).
For illustrations, see Ex. 11.
ll
a n+ i = &",
If
11.
>0, a number
the condition in the form
loga,, +1
curve y = log.r and the line y = \x

(i)
if
is
A>l/e, (a n )
we can write
of alternatives arise;
= Aa,,,
A = log&.
if
we can
By means
of the
easily prove that
a divergent monotonic sequence
the equation A# = log x has two real roots a, /? (say

<A<
that a < ft)
then the sequence (a,,) is monotonic, and a n -* a
(ii) if
/e,
!</?; but
if
When A
we may
y = log (log-
\
)
write log (log

>
and the
the sequences (a2n ),
A<-<9,
(iv) if
if
Here
w, v
a-j
>a
line
(a-2n+i)
-e<A<0,
(iii) if
><x).
()
the sequence
difficulty
!>/?, a n
if
negative, the equation log#=Ao; has one real root (a); but
will be seen to be no longer monotonic.
To meet this
is
= Iog(
A) + Aa,,_i and use the curve
n + l/
y = log(- A) + \x.
It can then be proved that
are separately monotonic,
and
a w ->a.
+i-*,
and
are such that
a*,,
= a,
a:
a.> n
û,
a.,)+1
^v,
a.
and ^ M = v,
u<a<v
but
a 1 <a',
if
-*<>,
if
^*
= w.
This problem was discussed in the special case a^ = k by Seidel (Abhandlungen der k. Akad. der Wissensch. zil Miinchen, Bd. 11, 1870), who was the
to point out the possibility of oscillation, in case (iv).
Previously,
Eisenstein (CrelUs Journal fur Math., Bd. 28, 1844, p. 49) had obtained the
this series is the same as
root a as a series proceeding in powers of A
first
the one given in Art. 56, Ex.
4,
below.
Arts.
-
12.
limits,
Th<
5.
4,
The reader may find it instructive to determine the upper and lower
and also the extreme limiting values of the following sequences.
relations of the terms a n to the limits should also be considered.

(1) a M
a,,
(3)
= (-l)"n/(2w+l).
= + (-l) n (2H + l).
7i
(2)
(4)
CT,,
= (-l)"( + l)/(2w + l).

= 2w + l+(-l) M w.
13. In an oscillatory sequence there may be a finite numlx-r of limits

derived from sub-sequences, all, sinnc, or none of the limits bring attained,
as may bo seen by considering
:
(1)
tl
= Hin(ttir),
1
(2)a,,=
all
M4-
luit
which consists of
tlir
sevrn numbers
0,
-ti,
as
in
repeated infinitely often.

jsin(jb<7r)
only
tin*
has
value n
the
is
same
attained.
seven
limits
ca
i x/3,
.'
J.I
ba
Tli.
\V!IM!.
:i
limits
:n
nil
tin-
limits (>
..f
may
l.x.
1.
lin.
rith
tilt-
"t
.m-t
ail.
tli.-|vt'..iv
f.,MM
raiiUMt
''
14.
It'
and
/
liiii-/,,
/,
lim<>,, = A'
liin
/,
lim
/'
= A,
/>
+ /^lim(^, + /O = A'+A
K-l
/'-/.
a n.l
iiiiilti|ilii-atinn
rt-siilts
tin-
may
and A
A"
>M|)|M.>.-
I"
<></
Mil
\vt-
tplej
|fi
I= A'A;
it'
<></<
A'
and /<i)<r
A,
w.-
/<0<A,
\v-
1:
,-,-A-A
if
/<>< A"
and
A
A
win-re
and
tlu>
is
_lim ((//>)__//,
nuin.-rically
nlvalsii in the liniitiui;

r
is
It
also
tln-sr
t,,
/-A
and
In-
tliat
tin-
ruli-s <-an
present,
M ,,ti .....
that
in. -.jualit
than
-iscs /---A*,
iriv,-
D"
iv
nnK'h
//
in
-:ist-s
when
\\id.-r
lii
',,6 W )
jiarii.-nlaiis
multiplication ly
If
,.f
A'^
",
inf.>rinati..n
'
(1)
//
i-..Mv-ri:'-i
ma\
i'-s
in
15.
A'/.
A"/..
!-tandinu
that
i:
tin-
usually
rM-i|ii-,,,-al.
!we
an-1
lind
(</1
/.
and
l.ut
if
in
i'iu-1 -i
/<()
all;
A'.
tin-
:n
values of
thosi/,
I-
Jjive
no
20
16.
To
(1)
(O=
1,
=
(2)(a,,)
(6,,)=
i,
-f,
17.
bility,
(3)
is
1,
(&)=!,
There
i,
^,
i,
-2, +3, -4, +5,
(O=l,
no
},
1,
14,
15,
the reader
may
consider
(&)=-> +
(3)
of Exs.
illustrate the results
the following cases
[CH.
2,
i,
J,
2,
3, i,
4,
3,
...
&,,-* 0,
an /b n
-0.^
...
lim
(&/<>=-<*, Hm" (6n /a w ) = x
...
lim
(&,,)=-!.
..
..
a n b n -*
lira
1.
(a n /b n ) = 0,
Era"
(/&) = x
difficulty in constructing further examples.
Verify the following table, and construct examples of each possiwhere (1) denotes convergence to a limit not zero, (2) to zero.
divergence
to
+x,
(5) infinite oscillation.

CL n
(3')
divergence to
-x,
(4)
finite
oscillation,
LMPLE8.
I.]
22.
1-f
s.uin-
':
lines ax tin-
'lu-t lira
'_'
6.
'.
prOgTM0i<
Miscellaneous.
23.
If
the
-rue of
sr|M.-ii.-.
/'ili
h.it
tin-.-
tern
Sequence* vary
tip
iii
:ise.
-...+/;),
24.
sequence (l-a")/n
.i
<a<l,
>1,
ami examine th- eztenaioi]
25.
frni
hftlur,-
it
limit
f(a\
26.
.'in.!
tin-
t'
tln-^-
iiuMjnaliii-
iii<-([iialiti'
i-.-iuai'
>
deduce
that
that
,
"" th-
I"
1
-f-. *
nattiiv nf thIf
nf tin-
i'".,t-
,,iv
ti,
|iiali"n
i|ii:ili-:.-
l.v
ili-iiMtnl
\vi-
a, /3,
can
J.K.V.-
-a,
Himtly,
r
in
i..iial
that
if
/j
i'
case
;al
ft\
j.i-oviilnl
that
\vh.
t..
til--
'
thr
\\h.
f..nil
-i
lent
timl
that
|
ami
t-asily
expressed
will
that
jirin.i:
i" >!"'
ial
CHAPTER
II.
SEEIES OF POSITIVE TERMS.

a2 a3
If all the terms (a lt
7.
of the series are positive, the
...)
sequence (sn ) steadily increases and so (by Art. 2) the series 2 n

must be either convergent or divergent; that is, oscillation is
It is therefore clear (from the same article) that
impossible.
;
The
series converges if sn is less

(1)
all
values
for
of n.
The
(2)
than N, no matter how large
sn
Compare
= i + J_ + J_ + JL
i
and
found
so that
is.
with the sum

!
<Tn=l
is
be
Consider the series given by a n = 1/nl, so that
1.
It
a value of n can
series diverges if
sn is greater
Ex.
than some fixed number
clear that
so on,
+ oî'
+ + 02 + 03+
,
= 3.2>2
n\ = n. (n
1)
2
3!
4!
...
= 4.3.2>2 3
3.
2>2"~
1
.
Thus, from the third term onwards, every term in <r n is greater than the
and the first and second terms in the sums
corresponding term in s n
;
are equal.
Thus
cr,,>s,,.
so that
<
x,
'"usequently the series

1'
By
In-
Mini
direct
and
is
2a n
denoted by
calculation
to
<
<r n
< 2.
its sum cannot exceed
we ran prove similarly that
is
convergent and
1,
as usual,
decimals
we
find
that
l-l-*;
lies
and that l/{7(7!)} is less than 00003, 10

between 2'7182 and 2'7183. Further calculations ha\e shewn
1^71*-'*
-2-7182818!
-2.
between
that
that
lies
81
Ex.
2.
POeiTn
tin-
<'<>n-i(!.-i
i:
EB
On^ln). (r
liaiiiK.nir serieH
\O
li
i-oiiii;iiT
.<
the
\\itli
is
-r<>u|>
a power of
-*(^-4
-DHi
-fi
as
>;tni-
Imt
tli-
last
t.-rm
tlu-
!'
i]
cai'li
'_MiMij
in
(after tlf !ii>t)
<r n
.-.|ii:il
i>
t-ijiin
t.
-
'
Tim-
L'.V-l
all
),
>)t(nt + \).
and
<-<>n-i'.|mMit ly
//-
t0TM
ten
tin-
wmiM
lifl
in
t.idt-r
Hate,
iinilai-lv
It"
.f...
*G
it
\N'-
tl.
tion.
Sn
l''"i-
in
tl,'
n.iita,1
all
u i"iij.
Tlirn
Hut
Sow
ainl
-2.
sum
i:i
DOt
that
"
5
+f+
t. a
million trniis
is
less
than
i'
<l'"r<)8.
SERIES OF POSITIVE TERMS.
24
The method used here can
But the
[CH.
II.
be applied to discuss the two series
easily
and have the advantage
discussions in Art. 11 are as easy
of
being more easily remembered.
The method given
Ex. 2 can be put in the following
in
rule (often called Cauchy's test of condensation):

The series 2a n converges or diverges with ^Na N
if
N=2 n
and an ^ an+l and it is easy to extend the proof given above

so as to shew that we may take
as the integral part of kn
where k is any number greater than 1.
(3) It is clear also, from the results of Art. 2, that if we
can find n v so that s ni s n ^>h (where, h is a fixed positive
constant), no matter how large n may be, then the series must
;
>
be divergent.
For we can then
n,
...
select a succession of values
n l} n z n s>
,
such that
-Sh
h
Thus, on adding,
and therefore
nf
sufficiently great;
we
etc.
find that
can be made arbitrarily large by taking r
and so the
series diverges in virtue of (2)
above.
As an example,
consider Ex. 2 above

s
ni
because
so,
-s n >(n l -n)ln^
?i
= 2w, we
get
*Jn
(4) If
ami
sum
- *n >
;ies;
all
+ 1)
to
/^
ami
of a convergent series of positive terms,

8 the value of sn cannot
n ) increases to the limit
<i
for
!/(??
fortiori, cannot exceed S.

the sum of any number of terms,
includes
On
the
is
the sequence
than
we have then
~s n contains (ni-n) terms ranging from

ni
by taking
reach,
can be chosen
lai-^v
Thus S must be îvntrr
taken arbitrarily, in
enough to ensure th.-r
tlu>
these terms.
number smaller than N. (say N e),

we can liml terms in the 861168 whose
the other hand, any
has the property that

sum exceeds N e.
7,
81
now
i-
It
cl.-ai-
tli.-i
mini
an
it'
IM-
\,-.l.
8.
l!'
l.rack.-tx,
iii
ijM-il
-i
pro\,-i|
will still
it
fco
wh.-n
Comparison test for convergence

til.'
t-
f,
ami
Mt
IP-
that all tfk
/,,
'';;
anot
if
"
series
li.-r
of
valut-s
c>f
positive
-f
l
than >>MM-
ber
lias
<i.,-\-<t.,-\- ...
ti\i'ii
ent,
wli.-n
>
a'
we
1'
4-...-f",
if
is
'/'
the sun.
Thi
-
[&BB
than a fixed number
establishes the convergence

In cast- the ine<|itality holds
s
',,
<C
TinIn-
mlititm
<///
tliat
for
sum cannot
the
BO that
tin- ti-i-ins
independent of
valm-
all
t-xc.M-d
must
It.-
,,
wi
}.
we
hav.-
T.
jHîtivi' in
^1"
and
'
in>
at
In-
ii.it
ill
nun
lM"_rimiinur
tin-
atl'i-.'tfil
liy
tli--
.mi--i..ii
,.f
tinit--
tli--
tak-
tlr
+ i_
with
in
tin-
-(. inline
"-
>,.,-,,
tri-ni
Hut
tin-
s-.th.-u
In
ml m
in
'
.uid tin-
in
thi-
timl
SERIES OF POSITIVE TEEMS.
26
9.
The comparison
which
II.
test leads at once to the following form,
work with
often easier to
is
[CH.
ân
Let the series 2(l/C n ) be convergent; then

r
will converge,
provided that
not infinite, both series containing only positive terms.

when this condition is satisfied, we can find a constant
is
For,
independent of n, such that
Hence a n
convergent
is less
than G/Cn which
is
the general term of a
series.
It is useful to
remark that
there is
existence of the limit lim (an Cn )
no need
this is seen
to
assume
the
by considering
the convergent
series
o
which a n Cn
is alternately equal to 1 and 2.

the
test is sufficient only and is not necessary
Further,
for
Cn = nl and
= l/2 n ~
as
then anGn >n/2

by taking
But Ea n converges (see Ex. 1, Art. 6).
so that lim(a (7n ) = oo.
The corresponding test for divergence runs
we may
see
a.n
ri
Let the series 2(l/Dn ) be divergent, then
2aw
itfiM
diverge,
provided that
lira( n Z)n )>0,
/>o///
series
containing only positive terms.

is practically identical with the previous investiga-
The proof
when the signs of inequality are reversed. We note also

that the limit lim(a n Dn ) need not exist; and that the test is
not necessary.
tion,
follows immediately that

necessary but not sufficient
It
the
folln-! ,/</
rttmlitnmx are
For convergence,
for divergence,
lint,
in
^vn.-ral
tlicrc
of (a n Cn ) to <-xist: and
Dn ) =
Mm
(a n
Mm
(",/')
is
////
'
no need for
condition,
given as necessary for convergence,
tin-
limits
<>!'
(<>
1>
[im(anDn )=Q, sometimes
is incorrect,
9,
10
BS]
Ex.
If
\M-
tilnl
i
liiu
iiat
liiu
lie
seen
will
(a
Art.
in
1.
tliat
eaay bo see
m lim (/'">
is
It
/'
cmnlitinn
il
is
.....
MK
if
///-
t>
<>
liin
we ran
rergent,
-li-
-...+"<.
\M\V
i-acll
!'
tllrsr
t.TlllS
a/
)'',.<
r.m
it'
//
is
<
/"/,
-2t\
if
>
//
an-
r.
liin
Tliat tins ciiiiilition
>
>
/'
it'
e,
//'.
iiat
//"'<
6.
'.
Tims
</
BO tliat
1,
choose
iii
>
i,,,t
is n.
.t
which -ivrv
"li(liti"M
li'ini
cnn.s.-|U-ntly
nOn)-
snllii-irnt
f..ll<.\vs f i.n.
di\
\it. 11)
liiu (/"',,)
.iltlmuirh
such as lim(. //>..) (l

ilv than N
and r\ani]>
DO ili-liiiitr limit will le found in
= 0.
\\hi.-h
("
L'I
iiiLTshrin
M-t inijih
HIM--
/>).
10.
It'
C(,lii],;l!V.l
iiil'.-r
\\ill)
wliicli
Cauchj
til.
i-
tlj.M.n-tit-ally
of
__
If
It
tin-
it
liin
"">!.
inijM.rlaiK-,-
rat;
limit in^
:-i
\a
J.
ti"
n-nn-inl-i-r that, in contrast
\vitii
28
II.
[CH.
!_
n
Further, to ensure divergence, it is not necessary that a n
should be ultimately greater than unity, in spite of what is
i
sometimes stated in text-books; and if a nn oscillates between

limits which include unity, the series diverges*
To prove
these rules, suppose
that
first
j_
lima n n = Gf
Take any number p between
_i_
an
and
6'
<p<
<
1.
then we can find
n>m.
if
Hence, after the rath term, the terms of

series S/o
convergent
after p terms
is less
that
2a n
is,
is
rn so that
ro
are less than those of the
And
convergent.
than p p /(lp) provided that
the remainder
p>m.
But
(say
lima n w >l, there
if
n2 n3
n-t,
be an
will
sequence of values of
infinite
...),
i_
a n n >],
if
n = np
a w >l,
if
n = np
and therefore
Thus the sum 2a, taken from 1 to np must be

may be taken as large as we please, so that 2 n
greater than
We know from Art.
54 that lim a n n
if
limO',,
") >>
But on the other hand,

between the extreme
and
if
\im(a n+l /a.n )
<
1,
This shews that d'Alembert's
1.
from Cauchy's.
test (Art. 12) is a deduction
falls
diverges.
between the extreme
lies
limits of (a n+1 /a n ); thus the series converges
and diverges
n,
such that
we only know
since
limits of (an+l /an ),
we cannot deduce Cauchy's
in
test
its
that linuf,,"
clear that
is
it
full generality
from
d'Alembert's.
If
we
consider a power-series 26 n aj n (in which b n and x are
supposed positive), Cauchy's test will give
#<!,
where
./
This
.
gives an exact boundary between

series,
rni^lit
SIM-HI
'Jncl
edition.
implying the
.,!'
for divergence,
Z=lim/v.
= ]/
Tims
and divergent
toi in
and se>l,
for convergence,
to
insiiHii-ii-in-y
In- t.-st
is
îvi-n
at
In-
<-li.
iy
of
(
ariaiic-
\\\l.
Jj
tin-
">,
I):
\\itli
Imt
ns.-d
nictliod
'hrystal in
his
slat i-im-nl
niad-
ivinaik must
in
that
com vrgrnt
from
to be diflrrcnt
supposing
In-
atticlc.
of tin- sain.- c!ia].l.T.
/m
\>\
and
('lirvMal
umli-rst
cod as
Tin- roniplrt
.-
I\TI:;I;
10, ll
li
it
all
= 0,
'
It
ive
p-.-ii
the
condition
aln
x-
tin-
rge
\;dil>
r.ui
if
\\.-apply d'Alembert's
inf.-r
mly
MM
\ i.
;id
t;
(as they
no
tl,,.
'/.
information
as
t,,
'
Em(6
gr-Umi
(i are
when
unequal
'iat
I'htain
>
./
and
\vli.
to the pov.
test
may
IM-!I;I\ i,,ur
ra-dly
of tl
.</<-'<
In
this
th.-
of
it
this
its
t.
th-
simp!
(with
t.-^t
in
d'Ali-nilf-rt^
'on.
ti
aMish
\\id.-ly
ordinary \\ork.
Second test for convergence; the logarithmic scale.

:nv ,-in-anî-il
Suppose that tli.- tennfl of a p,
11.
of ma-iiitudf. so thai
ord.-i-
>
that th- function
may happm
is
also di-linilr
tha-
and
for
\ai
which
ever ina
//
-f
I,
it
i->
aiv
Th.-n.
if
not
./
plain that
th,.
d,-tinitioii
of an
int-'-ral.
r-fi
J n
<>r
+r
=|
.ml
we
find
Add'
11.
!'
.(iialit
ii-s
we
h:.
>().
we
md
ii
n,. s
>o.
Thus. iVoin
in
O.
ha\--
beti
30
Further
(s n+l
- 7n+1 ) -
(.s
- 7J =
n+l
- fn+lf(x)dx ^ 0,
Jn
and therefore the sequence whose ??th term is s n

increases', and since its terms are contained between
the sequence must have a limit (Art. 2) and
a,
IF.
[CH.
In never
and a l>
Slim (-/.) SO.
Thus, the series Sctn converges or diverges with the integral*
f(x)dx ; if convergent, the sum of the series differs from the

I"
integral by less than a l ; if divergent the limit of (s n In )
,-
and
nevertheless exists
Ex.
a n = l/n(n
If
lies
/(#) = !/#(# +
+ l),
and a r
between
and f*/{*)drlog.8.
!),
3C
And 2
ci n
in
= l, which
contained between the values log 2 and ^ + log2,
is
agreement with the general
result.
A large number of very important special series are easily

tested by this rule
:
(1)
-+
Consider
if
Here,
is
where
^4-+...,
ci
= n'P.
positive, the rule applies at once,

^.
p
thus the integral to
oo is
and gives
(î-p-1);
convergent only
if
>
1.
Thus
the
given series converges only if p^>l: and the sum is tlifi)

contained between the values l/(p 1) and p/(p l).
If p = 1 the integral is equal to log x, and shews that the
,
harmonic
series is divergent (see Act. 7)
we
infer also that
n->cc
.xists
and
lies
between
and
1.
This
is
Euler's or
M*rln ',<>,/
/".s-
constant.
The convergence
L
can
*
in
The
of the st-rifs used in
+ gf+3
^,
Art. 9
+ ..^-^ +
^-1-...
inferred.
integral converges or <livrrm:s with the scqurnrr (/): for further details
iix
III.
LOG
\i:i
an
HMi'
\u.
9<
in-i'i.i.-.l
'
'+V-16*
4* 9*
leM than
toi sum
ii
'
lonmder
\\
\Qogi)
liidi
./,
'.'
and
()
f-aTHl
II.
Of
Thus
1
s|.\\.
tinit
th-
MUM
in
nuinlx-r
aiv
.t'
..I'
I).-
jMi.sitiv.-.
ami
//=
1.
th-
>mit
\\v
it'
nisun- llmt
t>
'^
'^'
li\.
!.-;
similarly that
pi-ov.-il
rarly t-nns
tin-
if
t.-nns
lillinn
/'=!
/'>! an
that
n..t.-<l
if
com
seriee
îvi-n
-lioiiM
]<Mr(lo^r
<!-/'>
all
tin-
a
1
suti
'ârit
Inns
that
tli-
:iv.-
lnit
it'
n n =(
i'
or
if
>
I.
F'I
int.-î-als
i'
slog
tWO
<1.
=!
[/'
nnl
/'
/'
<li\<
deci
|>r..\
il-l
tl
a limit, tin-
Ifl
>am-
]sitivi- anl BO
will
IM-
we can
tn;
l.-ilnr.- tli.-
r.-nlt
1.
This
1.
'hat
iv.suit
binued
work. th- two types
i.
;n
th.
\vith<-ut
are
For
nli
in
32
The following
Let
(Mn )
if
For,
see
that
substantially equivalent to (4)
an increasing sequence such
denote
are divergent
the
value
is
Mn+1
as
limMn = cc
is
then
convergent if
n ranges from q
M )/Mr+l
than ~Z(Mn+l -
greater
that
p~
^ (Mn+l
n )/'M H
sum of (Mn+l -Mn)/Mn
while
series,
we take
its
which are independent of the Calculus, have a
results,
field of application
II.
[CH.
p>\.
to
r,
we
(Mr+l - Mg)/Mr+l
because in the summation J/n
make J/r+l ^2Mg

be.
may
large q
and so
Thus the
Similarly, 2(J/n+1
If
= 2,
/?
vergent
series,
sum
We
series diverges.
- M n )/Mn+l
p>
if
2,
can choose r large enough to
greater than
is
is
24, Ch.
17
'
an ^ so
con "
the terms are less than those of a convergent

is
given by
 ( 1 - a n )/n, if m< n.
I.,
( 1
an
Write now
and then we have

where k is a proper
To apply
:=
^(jr--*/
so the only case left for discussion
From Ex.
no matter how
^,
(Art. 7 (3).)
divergent.
the third series reduces to
thus
and
<J/r+1
this
m/n = k,
c,
c*>(l-c)
fraction.
lemma, write
this
and then we get
Mn+l
^
-M
< -L- [l -
(^
1/1
)" ']
From this it is plain that the given series has its terms less than those
of a convergent series.
Ratio-tests for convergence.
Rummer's test for the series 2a n runs thus:
12.
//
SAr
is
a divergent
series,
t/
(C) convergent,
in
positive
tin-
first
numlxT
case,
less
i!'
lim (l) n
\
if
I,
For
if
//
"
''
is
L> /J+1
)>0,
X
"'"
lini(/>,,
>
is
tin-
'
'
>/
'
^n)
i>
''
AH-I)<O.
/
minimum
than //.an intr^vr
-n-fl
2aM
then
'*'
//,
can
limit
IK-
;m<l
found
//
is
siu-h
any
tl-.at
11, 12
i;
P8
Tl,
W(
Mini:.
+<..+an <amDm /h
II-
-inn
nn
</<><>
ri-'lit
lli'-
SO
ft
tllfl.
lit.
v,.r..nl
tin-
Iii
case
'
/>
;
<<>,
if
I'n-i-
if
II-
th.-
th'
term
aft-r
861166
diveruynt
hat
if
ami BO
of
and
number,
fixed
5]
,/<,
>
than
tin-
!'
Phllfl
<
tl
also
i-
_--nt.
Special cases of importance are

(
</'A/< n>l
tlu-n
1.
it
If
th
..
>!'lli<-<l
i
.,
<
i,
t.t
rhf >-rirs
>]
,1.
'lit-n
thrll
tll-
hn.,
EX.
2.
If
w.-
-...
l,ut
we
wh.-n
minlitinlis
-!)]>!;
lini
(OM
'.,
1.
;,]v
IS
(D
-!
tak.-
x
^)
.1
and
if
'
th.-
with
s.
ft>a+
1.
see
the test jfives no result
ol.vimi>ly
!.!
/'.
)<1.
In.
(D)
e-ninpaivd with Caiu-hy's test of Art. 10.
In-
Ex.
condition^ are
tin-
i>
that
/'
It"
o'
if
SEEIES OF POSITIVE TERMS.
34
[CH.
we must
(3) If the limits used in (2) are both equal to 1,

delicate tests, found by writing
II.
use
more
Dn = nlogn, n log n log (log n),

These functions are of the form
and so
where f(x)
f(ri),
on.
continuous
is
and f"(x) tends to zero as x tends to infinity. Then Rummer's

test becomes
(C) iim^ w >0; (D) lhn~/c n <0,
^=
where
For
f(n + 1) -f(n) -f(n} =
Now we
[f(n + #) -f(n}}dx =
can find v so that. \f"(x) < e,

be less than ê, if n >
if
easily seen to
is
v.
ffaf*f(n
x > v, and
Thus
so the last integral
f(n + I)-f(n)-f(n)-*0, as 91^00.

Writing f(n + l) and f(ri) for Z) n+ i and Z)M in Rummer's
at once to the form given above.
In
f' (x)
1/X',
/0n
"
where
Their further
are led
tests,
>l;
(D)
KS /0n <l,
= 1 + - + ~^
given by f(x) = x log x log (log a;),
etc.,
are of
practical importance.
(4) It is
sometimes more convenient to replace the
by the following:
(C) lim
<r n
>
(D)
!W <
7l
last test
1,
w_î-=-
where
After a certain stage, we have
<an /a n +i
also
thus
we
f(x)
(C) lim
little
test,
x log x, we find /'(#) = log ar+ 1/

thus we find de Morgan's and Bertrand's first test,
if
particular
+ t)dt.
<^-lo g (l+^)=[o r^^<^,

we
(5)
see that
<p,,-(r,
< 2 (log %)/,
The most important
the quotient
ajan+l
<*
yU
remains
is
less
in
being expressed
an+1
where
cases
and
=!
if
so p n -o-, -*0.
t
practical
the form
work admit
in
'"
ff
,
?'
a constant, X an indrx ^r'at-r than

than a fixed nnmlMT .1 I'oi- all values of
1.
//.
and
of
13
12,
\Tl'
i:
'
osion
It'
T.
//<!.
tl"-
lim
Thus
//
his
UY
<.,>
log
= 0,
.=0.
I..--
~A
<
<
\\V
the
tin-
th-i
may
rule (^s.-ntially
working
mi
ins
in
t-
tin-
apply
ivMilt> in
th.-.si-
up
I.
//.
.i
ami lim
.1.
ha\v
ease
limit
tin-
<
Hut, Since
in
diaciiflfl
ider
\\-
tort's
i-
lu-
r-
!'
Qfl
t<>
Hypergeometric
//'
tlu>
quoh
>
>
^_
i'
fl
//
>
"
tl"
we
It'
ivsiilt-
tin-
a). ply
Ex.
(/.( (/
ojy
M.y*"
tfing
<
tin-
'_
di\-iu
if
/t
+ l)^ (/j +
+ l).
aj|..-ai
/^,
-^(^
+ 1X7+2)
snifIf
that
Ait.
it
.".<
>
that
.-oiivcr^s
i-
^f-ii
family
t<>
converge
!,
+ i-a-0)-
si-rii-
the
als..
It
coir,
M-!
fop
if
./=-],
-1<
if
'
tn
I.,-
imti-.l
that
n
lliat
'/'.I
will n>t
In-
^.-1>A>^
:
llivi-i
[
with
'
"
tin-
=o
-.3 7(7
Notes on the ratio-tests.
13
In
so that
<-'>ntra-t
/^.
from
1
tliis
J?>1,
it'
+ l)(y + /Q_~
= 7 + 1 -a
Ait. ^
t" divei;ur <'
=a+
will
tin
S.-i
Hy])frê')iiH-ti-i'-
d'Aleml
fiMiM
It
and
i\r^
'--
'i'-nt
thai
condition lim(/' n
(>t
\\hifh
|ir..\-.-
tl"
form' above,
1.^. ;<y
and
1,
t>
Milieu It
sufficient far
seriee in ^-iR-ral, AIM.
I'm-
('uiisid.-i-
3.
the
/'//
1.
Iv-
"nd
-suits
imt
i>
it
Art. 39,
"i'
jM.sMhlr to find a nuinlH-r

f,
36
Secondly,
Sa n
For
it
definite limit.
will be seen in Art. 26 that the order of the
the convergence of a series of positive terms

the order may affect the value of lima n /a n+l
Ex.
The
1.
II.
not necessary for the convergence of the series
it is
an /an+l should have a
that
[CH.
a+
series
but
affect
change in
+ a + a + a + a + ...
3
terms does not
of course a
a rearrangement of the
is
3
geometric series 1 +a + cr + a +... and so is convergent if 0<a<l.
in this series the quotient an /a n+l is alternately a and I/a3
,
But
Theseries
Ex.2.
is
convergent
In this series
+ a + /3 2 + a 3 + /3 4 + a5 + /36 +...
<a<@<I
if
as
is
But even when
llm
n
/3
/a
n+1
= oo
the terms are
series does not
Ex.3.
by comparison with
we have
lim a n /ft n+1 = 0,
of the
plain
arranged in order of magnitude, the convergence

imply the existence of the limit.
l+-ia + f a+|a 2 + ia 2 + a 3 + â3 +...
Theseries
< a < 1 and

terms arranged in order of magnitude, if
2
3
2
3
convergent, by comparison with l + a + a + a + a + u, + a +...
has
its
it is
then
lim(a n /a n+1 ) = l/a,
But yet
Km (an /an + = 1.
:)
Thirdly, if the quotient an /an+l has maximum and minimum

limits which include unity, the whole scale of ratio-tests will
fail.
For,
if
\im(a n /a n+l
)O>
>g = \im(an/an+l we
If
can take
A",
k such that
and then a n /a n+l

while
),
for an infinite sequence of values of

is greater than
than k for a second infinite sequence of values.
n belongs to the first set of values, we shall have
n,
it is less
n(a n /a n+l -l)>n(K-l);
but
if
it
belongs to the second set,
limw(a n /a n + - 1)=
Hence
and therefore Raabe's

extends to
If
all
lim n(a,,/a n+1
test fails entirely.
It
is
- 1)= -
co
easy to see that the failure
the following tests.
we apply Raabe's
lira
and passing
+ 00,
test to
Ex. 3 above, we iind
?*(/+ ,-!)+ oo,
to the next stage
imi (log n)[n(a n /a n +i
we
Hn7i(a M /a M+1 -l)=l
get
-!)-!]= + 00,
=
lim(log70[/<(",, ",,-n-l)- 1] 0,
so that the ratio-tests can give no information.
BUM LKO1
13, 14]
he se,-n
will
It
from
t:
i\
mparat
due--
at
ln-iii
all.
and
Kuniiner him-
convergence, where
to
siil.jrrt
pro\vd to
1..-
in
used
Ermakoff s
The
Mffiee
Art.
I.
his
ion
to
=(}.
LOna ha\
^f{n\
which f(n)
in
i-
suhject
convei^vnt
<
Inn
if
tii
/>
4"
s"
'>!.
'
any nuiul.rr
is
i'
ran tind
X>
Jug
tin-
!..-(
wmi
Th
in.ixiiiiuin
tli.-
limit
^ 1;it
vari.-il.l.-
ind.-prndt-nt
1,
<>.
-V
'
li
to the conditions
<
lim
if
divergent
if
\v-
to
hy I'rin^^heim.
'
wt-
tests.!
(ii)
tin-
was
whicli
where the i
l'.
'
in
the form
aUo was
l)ini
I-'"]-,
in
t.-st
.rm îven above,

test both for c-.
tin-
l!
is
(i
and unity,
and
liii;.
|)ini.
hy
Kin-tli-
14.
of
r ->t n'et
siijH-riliious
expresai
/.
an arhitrary Bequenoe
is
</><//>
tin-
ohtain the condition
di\
u,at. in
is
they apply) than
ratio'
Historical Note.
t'oi-
inl
086 the
numler.s.
to this
much simpler than
'i
<|iiotien!
t<>
<
answer
Tin-
In-
ii"t
ran^-
should w- tr.,uhl- bo
with the n,
why
-I.
oompaiiBon-teftfl
easier
limited
fly
may
it
to
'
in
>$
tin-
l.-ft-liand
int-gral,
jr
(l- r
I
ha\v
v.-iriatic;
Ix-i-n
Lriven
liin'> an- uixloul.u-.lly tin- UK-:

-
bo
tlu-
aiva of
tin-
cui\.
u-e
iu
\vlu-n
p-.Miu-triv.-al
f.>il-
represeiitatiou of f/(x)<fe as
SEKIES OF POSITIVE TERMS.
38
Or, again, since the last term in the bracket
greater than X,
'
this inequality is true for
that the integral
any value
of
f(x)dx must converge
the series 2/(w).

In the second case,
As
x is
positive, because e
we have
"
As
is
II.
[CH.
X greater
than
it is
clear
and, therefore, so also does
can be found so that
above, this gives
7*
or
Jx
f(x)dx
f(x)dx
how
is
X may be,
great
number X' = e x
greater than a certain constant
Thus the
These
is
/oo
if
f(x)dx,
divergent, because, no matter
can be found such that
f(x}dx
compare the argument of Art. 7
(3).
series 2/(w) is divergent.
tests
include the whole of the logarithmic
scale.
For example, consider

p
f(x)=l/{x.\ogx. [log(log#)]
e
ex
Thus
and so
f(e
,/'
^.
oo
= oo
or
That
It
is
),
= e /{e .x. [log x]?}.

f(e )/f(x) = [log(log x)y
lim e xf(e x )/f(x) = Q,
if
then
is,
the series 2/(7i) converges
if
if
1.
p>-l and
if
easy to see that if </>(x) is a function which steadily

with x> in such a way that (/>(x)>x, the proof
increases
above
diverges
may
be generalised to give ErmakofFs tests:

.
(i)
(ii)
convergence,
divergence,
if
p
^
lim 4/(x)f(A(x))
7x T
<1,
J\ x )
'
x->
if
lim
xi^;
*'<)/(())> L
./('')
LOG \i;n H.MK
15
14,
Another sequence of
15.
in
tests.
Although the follow

or
it
tlin.-iry wurk,
('and,
Ait.
seen
_T.-S
if
"iily
iii
39
;--s^
I'".
An.
p>l,
that
ii,
17-"M
if
>o
!
th-
lirst
û.-h
tli.it
:aJ
<l
log[/X tt)j
-r
that
slu-ws that
;
S|i.-.-ia
/'() = /j
(1)
l'<r,
we
can fiiul/>>l and an
>-,
>
(
(/_
tlii>
0.
f
.-<.ni|>:iri80n.
Pgea, ly thr pnnriple
~u-h that
the second cast.
in
-/"(/')[^( /)]",
>h,.\\>
wh.-i.
II .ut tends steadily to
CM6, as on previous occasions,
logK
This
divergent,
'8t
-!,,.:[/
Inn
di verges if
is
;"
11
continuation of
,i
and
<*-tecU>
imp..:
iliv.Tges.
.f this test arc-
i:i\''!i ly
and the function to examine
is
logO
logn
mction
is
log(l/wa,)
log (logn)'
;iicti"H
and
>n.
D
in
\\hi.-li
It'
is
the
I,,-
i.-lati"ii
tnuuformed into another

t>
\v write A = log(l
,/.
'a'h
).
-liape,
fii-st
given by Jamet,
h\
it
is
ea>y
t-.
>ee tliat
it
Thi;
A>
iii
i,
;^=[
provided that lim (A /<)-<>: and,
if
i^
(1
"
this Condition
is
net satisfied, Cauchy's

the test
tses of practical interest,
Will
l.e
40
Similarly
it
II.
can be proved that test (2) can be replaced by
[n(1
- a "" } " log n
log flog *)
This form proves for example that the series

but converges for x>\.
0=^=1
[CH.
^> 1
'
--lognj
diverges
if
>
General notes on series of positive terms.

Although the rules which we have established are sufficient
to test the convergence of all series which present themselves
naturally in elementary analysis, yet it is impossible to frame
16.
any
rule
which
give a decisive test for an artificially
will
In other words, whatever rule is given, a

series can be invented for which the rule fails to give a decisive
constructed series.
result.
The following notes
(l)-(3)
and
(8)
show how
certain rules
which appear plausible at
first sight have been proved to

be either incorrect or insufficient.
Notes (4)-(7) shew that
however slowly a series may diverge (or converge) we can
always construct series which diverge (or converge) still more

slowly and thus no test of comparison can be sufficient for all
;
series.
Other interesting questions in this connexion have been conby Hadamard (Acta Mathematica, t. 18, 1894, p. 319,
sidered
and
t.
27,
1903,
p.
177).
has pointed out that there cannot be a positive

function <j>(n) such that the two conditions
(1) Abel,
(i)
lim </>(n).an = Q,
lim
(ii)
</>()}).
an
>
are sufficient, the first for the convergence, the second for the
divergence of any series Zf,,.
For,
l
if so,
2J[<(wj)j~
would diverge
and
Mn = wo)]-* + wa)]- +
therefore,
. . .
if
+ ooor,
the sequence
n would be an increasing sequence tending to oo
Hence the series 2 (Mn
n -\ )IMn would diverge also (Art. 11); but
.
lim <l>(n)(MH -
so that
contradicting the
(2)
first
-I/,,
=0,
condition.
Pringsheim has proved that there cannot be a positive

c/r
such that the condition
<f>(n) tending to
function
16
15,
function
I"
n-ar
any com
for
'id
an
BeriOfl
'
fiat
lin,
'
'
J0
'
II
Mth. A
See
\>.
lias
ini
lin.
can
ami
r/>(//
th-r.-
>()
1"
DQ8 of tin- 861
BO arrani:v'l that
I)--
i'l.-<l
that
t.Tin-
tin-
Set-
!.
.
.
A1..-1
In
;
"//// di\<
lim
a po-
i>e
condition
tin-
dhvrirence of
tin-
I'm-
that
v.-l
pi-..
uch that
function
function
:ni.
8ft,
to zero.
il'
which
ii<l
nl
the
p,
that
r.-iuarkt-.l
=0,
!'
also <Mv<T---iit
is
l>ut
BOCfa
lim (6^
i/ M = a +
>
write
MriM
(5) dii
second
-'
11
.!/,
J/ M ->x
Beriee
-a,
and
^<>
it'
1-
can
!/>
n.
Tound
i^
r ,n v.-r-vnt.
,
which
has
tliat
is
7
1.
it
Sti.'ltjrs
.-\
_ 1+ ,,, i+ ._,4-
11).
->/.
tliat
that
il
as
can If I'mind so that
>
idrnt
sh.-w.-d
ndin^
,/
.<
<
jM.sitivi-
J/,..,
<-..nsMjiu-iitly tin-
6M =
limit,
...
i>
d*
di
1'
L/m,; then
M + -M
n
"~F^T
^fw-w..-*-!
1
<i n
and
MI.M
BO that
shewed that
rty
1'iovidfd
l,,
An.
IH-VIIIOII.!
1'iuis
OOnvergeni
Then
..+//,
I.v
1
''..
Mi/~"'
seriea
thi-
42
proved that
(7) Stieltjes also
v lt v 2 v 3
if
[CH.
an increasing
... is
sequence, tending to infinity as a limit, a convergent series 2c n

can be found so that Zv n cn is divergent.
For, write cM = l/vM -l/v,l+1 which makes 2c n a convergent series; then
= (vn+ i-v n )/vn+ i, so that 2vn cn is divergent.
,
i'
ll
cn
(8) Finally, even when the terms of the series 2a n steadily

decrease, the following results have been found by Pringsheim
However
the series Sc" 1
fast
However slowly
<j>(n)
may
always convergent series Ean

(although Iim7i.aw =
by Art.
converge, yet there are
that limcw a n =
2an such
series
always divergent
may
increase to
which Kmti.
for
0.
with n, there are
oo
<j)(n)
an =
<x>
9).
See Math. Annalen, Bd. 35, pp. 347, 356.
EXAMPLES.
1.
2# M
Test the convergence of the series
following expressions
where a n
is
given by the
1
"'
a 1 /"Prove that
2.
if
(log)
10
n>
10
[log(logw)]
M'
(Art. 15)
1.
(Ex. 25, Ch.
6-l>>0,
I.,
and Art.
11)
the series
a(a+l)(q+2)
converges to the
Shew
sum
also that the
(6
sum
1)/(6
a-l)(6-a-2),
[If the first series is
1).
of
if
&-
denoted by w
?i
+ M 2 +..., we
get
\\hidi
Hence (6-al.\
The
l)(x
(l
- V ) = a</,, -(< + /<)", hy
addition.
But
Hence lim* w can be found.

>ince the terms st.-adily decrease.
second series can be expressed as the difference between two series
Ait.
!,
of the fust
ii.
kiFLEfi
I'iv.- that
3.
tl
(20 +
if
\xv*
4.
6>f>0
and dm-i-en
that
-Ties
|'|..\.-
tin-
if
+ \S74j
if
and
f>>-2,
I'l'-vi-
if
that,
tliat
)3
that
I'rnviI'
O<Y<!,
-n
pla<r
al>u that
/>
.!
in
tli.'ir
is
i..-.-illati-
>/3>l,
that
lint
and
Between
tin-
10).
d-<-reaaee t
</>(/0
n-r i-(Hi\-.T^fn<-.-
an
is
no advantage
-
-.
+ i.
Kunn
whfiv
COOV-!
oonvi
i,,
Sliew that
..f
|i'-itiv-
'-1
~
+"| +
.i
V2.4.6
thrrwi-.- di\v:
..nsrcutivr t-nn>
ratio
7.
3 5
ulso that
6.
irs.4.6
5.
'
'
1X30+2)
arliitrary
in
convergent
making
jx-xitivr
,-;IM
writt-
fun-ti<.n
;.
Imt
\>\
this .-liaiige.
in
aft
if,
\vt-
stage,
[Or
-
ir
8.
degree
term
lii;h-st
rges
9.
10.
in
X-,
th-
whose
jii"tirnt
"f
hiirln-t
POT
'/
tli.-
sum
^\'(
in
tt-rins
of tin- inr
'^
s
and drdiu-f
Siniilarly
d
1'"
[I'i
Kind limits fm
that
|x.lyn..iuials in
div.-i^.-s
^=1 and J>1.
if
two
in<-iva--
in\
if
if
the
^>1,
us.]
44
by means
of the integral
C
Jo
and deduce that
cr,,
Prove that
11.
->
if
ndx _
[i
[CH.
dt
x2 + ri2 ~Jol +
'
t'
|TT.
approaches zero through positive values,
and that
where
[Di RICH LET.]
Euler's constant.
is
[To prove the latter part, note that
is positive but less than

can be put in the form
f(v)
if
The desired
l/v.
limit is that of
/"(I),
which
From this expression

equal to that of
More
12.
and
n is
and
If
lim
we can
^2
if
generally,
never
easily see that the desired limit
= an + bn
(where
and
finite.
~~
is less
|
exists
is
n,
and
l{
p-+Q \
or =(1
- l/o)/logc,
or =0,
Ex.
13.
= n*,
Utilise
II.
2w
/( ,
of Art.
= (#/")
= S,
that
"
w,,
J/N+1 /J/,,-*oc.
37.]
152 (Appendix) to shew that
^l--log?ij
2X, ('nipniv Art.
is
if
(</)
necessary (Art. 9) for the
by writing
,,
If
Mn+l /M -+c>l,
if
[PRINGSHEIM, Math. Annalen, Bd.
Tln-i.ivm
convergence of
if
2", n!
decreases steadily, the condition lim(?m,,) =
of
than a fixed value,
[DIRICHLET.]
;J/~I+P
then
bn
zero),
tends steadily to infinity with
must be
/,.
r/
l
-,,,i
,.
"
";~
"- 1
"n
prove that lim (//) =

!.").
6', (
_1
[CESA.:,..]
!,
and deduce the divergence
14.
,o also is
dm-ip- and
!">(/,
2(a N 6 N ),iustrat-d
\.-t
+ + -'
::
fnn
aft
hand,
if
be seen
iv
<
implii
15.
16.
[i'l::
Use
th.-
rhat
')
=
N,
tli",l
1
18.
if
If
that
17.
l.ut
S!H-\V
that,
).
,,f
-l)]-'
= 21og2-l, 1
\\itli
tln
8-1).
n..tati..n
:;
-+;>
:,:'^l)='
Deduce that
__=-:;
<j/,,//,.
r,-;
i>(
19.
l'i"\r -iniilarlv
1905.]
"
.'2v+l
tliat
20.
that
1
and
Slu-xv
tljat
j
]
~-
21.
Kxamin-
if
<M n )=l
if
jr<
22.
+ + +
2
3
Shrw
;tlltl
OODV]
'i.-tilar,
--
prove
that
1
th.-
'
-l~2'
2/7-7-7
v
If
,i
'
'
'
n!
\
-
tliat
tin- > -rifs
converges
CHAPTEK
III.
SERIES IN GENERAL.
The only general
test of convergence is simply a transformation of the condition for convergence of the sequence sn
(Art. 3); namely, that we must be able to find m, so that
17.
sm
\s n
"
It
must
be possible to
positive fraction
e,
no matter how
It is
n>m.
|<e, provided only that
condition in terms of the series
we
If
get
we
express this
find m, corresponding
to the
arbitrary
so that
a m+1 + a m+z +
large
p may
+ a m+p |<
e,
be.
an obvious consequence that in every convergent series*

lim an = 0,
n
-oo
lim (an+1 + a n+2 +
n >co
+ a n+p = 0.
)
But these conditions are not
sufficient unless p is allowed to

forms of variation with n; and so they are
not practically useful. However, it is sometimes possible to
take
all possible
non-convergence by using a special form for

shewing that then the limit is not zero (as in Art. 7
infer
We
therefore
are
suffice to
to
obliged
employ
shew that a large number
special
tests,
p and
(3)).
which
of interesting series
are
convergent.
*
It is clear
from the examples
not exclude divergent series

perhaps might bo expect. .1.
where lim
/,
= 0, iima n = l, and
in
Ua
\<t
'liaptcr II.
it
does not
In-
tn-ms
<
that the condition lima,, = does

-\vn exclude oscillatory series, as
ti-inl
to /.cm.
17, 18
LB8OL1
OON1
II
17
El
18. A convergent series of positive terms remains convergent when each term an is multiplied by a factor vu
whose numerical value does not exceed a constant
can
F..I- since 1
chow
nvergent, th- indr.x
/..
///
>/</,
that
Imw.-v.T small
may
be,
M+
"+/>
1 ^.
But
UJ
and therefore
TV.
1
fche
series
values
|a n
2a n
'
if
series
tl
/A*///
/'
////'
"
'<
;(
1.
(l
/ ,
///.
///,><,////,-/_//
<<,,','<>n<l'
i'
,j
/.
Tli-
'...uld
we cannot apply
that
observe
infinity
An
iii< li
is
cxanij.!.'
Ex.
it'
1.
f>
thf
atl'
\\<-
It'
-rdcd
K\.
l.y
tak-
we
'
b,
11
tlia'
.verges
_1+1_1
4
4P 5P
+
that tin-
tir>t
.-f
thr-i- i61
'i'g*
-1.
EX.
that
Kii"-
Ait-. 21,
\\ill a:
tli-
im-tlnxl
'
It
this
kn<.\\
t-nai'
tlii-Mi-mi
luit
i-
lu-!..\v.
L'
Th'
1.
oj
series
.1
positive
,>>.
is convergent.
|
deserve
th.-.,i-rin
is cm\>->.r
For hciv r/ H = |a n and

Such serifs are called
'/<
ooni
ia
cases <f this
.d
<.
s*
Thua
2.
sum
'
".
!+*-*+'
tin-
:t,'8
fa.-i
- -I.
-Jl.'2
1
+ -
'
'
+ --!- '""
tO
so
48
SERIES IN GENERAL.
The sum
of the first
terms
2?z
III.
[CH.
is
Thus, by Art. 11, Iims 2n =oo.

But *2n-]>2n, and so also Kms2 -i = oo.
Thus the new series is divergent.
The reason for the failure of the theorem
is that the original series

contains an infinity of negative terms and that the series ceases to converge when these terms are made positive (Art. 11).
;
It is
easy to see that the foregoing theorem can also be
stated in the form
An
remains convergent if each

a
whose
numerical value does
multiplied by
factor
not exceed a constant k.
term
absolutely convergent series
is
19. It is often convenient, however, to infer the convergence

of a series from one which is not absolutely convergent.
For
this
the following theorem may be used
purpose
convergent series
remains convergent if
un
is
2an (which need not converge
absolutely)
terms are each multiplied by a factor

the
that
provided
sequence (un ) is monotonic, and that u n
less than a constant k.
(Abel's test.)
its
Under
vn = u
when (un )
is
it
is
decreasing.
never increases
so that
un converges to a limit u and write

an increasing sequence, but v n = u n u
these conditions
u n when (un )
and
Then
will suffice to prove the
to infer the convergence of

Art. 23 below)
is
But by Abel's lemma
P V ni+\
greater than any of the sums
m-ni
|
-j
Now
convergence of 2ft n vn in order
2a n un
m+1
where p
clear that the sequence (v n )
it is
converges to zero as a limit.
'"
~~~H
-'
01
18, 19, 20]
Ai'.l
\M>
.1.
p=se
no matt.-r ho\\ small f
'\.
I'llMdll
can be chosen BO that
^''
fchtU
18]
'
!|
ll
f.
and
c..ns,.ijufiitly
at
thai
conditions as
If
1.
coiidil
uv takr
MriM
th.-
and
ol.tain
\\liich
\
'
oiiic
rnij.li.y
0,
w-
the factors are sul
lit
Ex.
mgent
ill.-
+\- +
}
...
Oiln-ady used
n-tors
sequ
tin-
niii>t
th.-r.-foi
.-
'!"
---ni.
that
\-iify
this
tin-
OM6,
hat
,i-l
_/l_l\_/l_2\
1
""
Thus
li
in ,<,._!
t-xi-ts (liy
= liin.N-
liin*. _,
F1
Ex.
Ex.
in
any
a.
iin-nntMiiic
Art.
2,
'"
1_
Ait.
1)
That
and
(1)),
the
is,
sine.-
=%,_i -
.<...
series
converge; the sequence of factors employed
18,
./
sMju-nc<-).
tin-
^'-,|u-ntly
is
e.
tin-
than
r.-iiiaind.-r
less
nunirrirally
choice of ni, \vhicli
;
/'//-/-
This ].r.']Tty
o/OJj
may
In-
after in terms in
than i>('\+
niak.-s
so lon^ as
.-xpv-s.-d
*-,,,
witf
r^ -f
1
>y
this
/^
and
):
ivmaindrr
-prcial ca.M- of thi>,
ran
20.
1-y
takin.i:
l.S.
whi--h
wu-
th.-
l.
I'i
//
minimwm
".,--
lvm,\
is tinit,-.
âyin-'
that
41.
below.)
it
important intViviic-r is that it' tluu depend

the conditinn of t'ori.
ay on a varial >lr
<suljrct
i
\s
(/<+i
B61
easy to sec why, fn.ni our present point of
It is
2.
In
,1
3-
i'-
origin*]
^1.
SERIES IN GENERAL.
50
[CH.
III.
terms are multiplied by a decreasing sequence (v n ) which tends

as a limit. (Dirichlet's test.) *
to zero
lemma
Abel's
gives the inequality
m+p
^^
/
(K
^^
1)
nt)
771+ 1
where p
is
any number not
less
than the greatest of the sums
a m +i>
It is sufficient to suppose p not less
than each of the differences

>
|
Now,
if
S m+p
- Sm
the extreme limits of sn are both
.
\
finite,
we can
some constant! I, such that sn is not greater than I,

any value of n. Thus sn sm = 21, and we may take p =
We can now choose so that vm+l e/Z, and then
find
for
21.
<
m+p
|2wJ<2e,
m+1
proving that the series 2a ?l f n converges.
The
Ex.
Of
series 2-yn cos nO, *2v n sin
nB converge
if
not
is
or a multiple
27T.
For
l>;
m+l
cos?i0 = sin(te9) cos

.
[m + ^(p + 1)]<9.
cosec
m+p
2 an0=sm(J0).8m[m+i(p+l)]0.eoeecJ0;
and
m+l
could here take p = cosec J0|.

the first series may be convergent or divergent according to
but the second series, being 0+0 + + ..., converges to the
the form of vn
so that
we
When = 0,
sum
0.
special case of the last article is the result:

terms of a series '2( I) n 1 v n are alternately positive
If
and negative, and never increase in numerical value, the series
21.
the
will converge, provided that the terms tend to zero as

*
limit.
Abel knew of this test the history is sketched

by Pringsheim .(Math. Annul ,-n, I'.d. 25, p. 423, footnote). But to distinguish it clearly from the test of Art. 19, it seems better to use Dirichlet's
name, following Jordan (Cours d' Analyse, t. 1, 299).
It is practically certain that
briefly
fThis constant I will be either_the greatest value of

greatest value) the greater of Ilim,*,,! and |lim.<*,,|.
|,
or
(if
there
is
no
20, 21
if
|-'<>i
of
\n: 9ER]
AI.I I;I;N
1
the
we take
la>t
-f
\v need only
article,
.".I
as the oscillat*
...
that p
.suppose
is
1;
and
til'
<<|Uently
v 1 ->'.>+>',-v4 + v & -Vt+...

it.
account of the
<)n
now
use of
frequent
give another proof of
is
It
it.
theon-m.
this
\v.
that
plain
%i-(i-^)-K^-4)+...f(^|. 1 -i
and each of
th-s.- l.i-ack.-ts
so that, as n increases,
Also
,1
]><it ive (or at least
i>
th.- s.Mjuciice
-(,> 2 -i; 3 )-
and so the sequence
>
of terin-
<>
-...-(t)
>
n<-'
ive),
never decreases.
^K
Qever increases.
..,,.;>
=--
Further
and
Hence, by Art. L the sequence (> v lias a limit not greater
than >\ and (fi^+i) Has a limit not less tlian 0. But these two
)
limits
must be equal
Hence the
Ex.
is
n->\\
1.
The
series
lini
c.,
n ..^
>een to converge, i>n>\idfd that
is
p=l we
get
easily seen t> le cvjual
'i
=0, so that
converges to a sum lying between

already mentioned
M-rii-s
In the special case
\\hirh
since
jo
+.r
t<
in
Art.
K.\.
1,
0<jol-l.
tin-
l>^-2.
F.i
1 H-
.'
f1
.'-'"
,.
1! " 1
J,','i
,,,,1,,,
1*,
if-*L(i-J+J-J+.
and v r
SERIES IN GENERAL.
52
III.
[CH.
The diagram indicates the first eight terms in the sequence (sn ) obtained
from this series by addition
the dotted lines indicate the monotonic
convergence of the two sequences
;
FIG.
9.
It is obvious that if the sequence (v n ) never increases, but

n ~ lv
n
approaches a limit I, not equal to zero, the series 2( l)
will oscillate between two values whose difference is equal to I;

in fact we have by the previous argument lim s2n+1 = lim s 2n
1.
case
If v n /v n+1
can
interest
given by the following test

special
which is similar to that of Art. 12
of
is
J^l +
the series 2(
For
if /x
\)
>
0,
in the form
be expressed
n ~ lv
n is
2+
FX>1
^5,
{j
convergent if p
after a certain stage
>n
n *^~ n*x~
>
we
oscillatory if
shall have
0,
/m
= 0.
'
and further (by Art. 39, Ex. 3) limv n = 0.

so that v w
n+1
But, on the other hand, if /x = 0, it is clear (from Art. 39) that
>v
limt> n
/JL
<
0,
is
not zero, and so the series must
after a certain stage
we
shall
oscillate.
have vn
<v
limi; w cannot be zero, leading to oscillation again.

Ex.
2.
Take the
a.
ft
l.y
series
1.2.y( 7 -fl)
So the series converges if

The same condition .ipplies
and
1.2.3.7(7 + 1X7 + 2)
to the series (Art. 20)
to the corresponding series of sines.
And,
if
n+l) so that
\Ui:i;\
22
21,
hmiM
i.
9EB]
I.
THIN
observed
l>e
that
+ i^|4 + jî
:i-
L>
-ily
therefore leu than
-f
t-6.-
(Art
-2
ntly th.
-- +
lint
1/11
-iK+5 +
tlll!
tlie first
p-
5i
>
sum
the
of the first n negative
1\
'
series tends
2n terms of the given
first
limit.
its
A ruriou^ thc(,rcm,
22.
jî
1111
l+
S
-OD as
sequences
:.
M divergent
rooogDied
than
Ii-ss
to
nag
nor
1/
ill
snch as
ii
>M<1
{>
',-,!/. 1
reason In supjMi^- that tin- tln-nivm i- -till m-r.-ssarily

fact r
CODStrUCJ '\ani]>l<-s nf tin- failmv,
tli.
fcmej ami
aii-1
tin-
if
ihr
in
of th- laM.
i^
'In-'
tn
some
bo
L-xtt-nt
of con
kin<l
'rsaro:
If a
farms
<K,
tude,th(
n a
//
it
of P*
ng order of
ca/nnoi
'/,.
//<//////
mi
=
fence th-
>f
..f
th-
if
it'
/,
<
/,
\v-
:-''
i-vj+dp,-^
-i-v
mis to a po-
if
>/,,
//
;ch
iml<
>
that
//'.
[n^
sin.
<>/
tiTin-
can tinl an
n
series.
fh<>
î/
-
But,
p-^sililr. that
Hence
terms
t<Tiii
tir-t
^+(/',
tlu-n.
jli-xt
jM.sitivf
fi n
'-h.-.-hy
Suppose now,
/;
th-
/'
sum
(Pi-9
limit
other Ivmti flam

tl>'
"
owy
>r/,
i/
<> arranged
>+'"'
given series
;
ami
i->
>
not
,C. + 'Wi +
---
alsolut<-l\
ntly
>'
tlu-
ônt.
m
+ >\
...
SEKIES IN GENERAL.
54
[CH.
III.
can be made greater that NJ^ by taking n greater .than (say)

n
n can be
Hence, no matter how large
is, a value
t
found so that
//;
if
hence
sn
must tend
if
lim (p n
n=pn+qn, and so
This proof
qn)/n cannot approach a negative
The
Ex.
As
if
series 1
a verification,
value must be
q n )/n exists its
lim(p n /q n ) exists, its
Now
0.
value must be
1.
same as one given by Bagnera.*
substantially the
is
to oo with n, contrary to hypothesis.
It follows similarly that (p n
limit; so that
?i>7i
+ ^- + i + - + ^ + - + ...
we
note that the
sum
of
cannot converge.
3n terms is certainly greater
than
so that the series
23.
If
Abel's
is
divergent.
Lemma.
the sequence (v n ) of positive terms never increases, the
2 anV
n ties between
Hv and
hv : where H, h are
the
sum
upper and
lower limits of the sums
For, with the usual notation,
al = s 1
a 2 = sz
sl
Thus
Now
v 2 ), (v2
the factors (fj
v s ),
...
(vp . l
vp ), vp are never
negative, and consequently

8 i( v i
- v z> < H v - v 8*( V - vs) <H(v - vs

-V9 )<H(vp . ^V9 8f Vp <H>
8p . (v9
(
Bagnera,
Rend. (4), t.
Hull.
Sci.
4, p.
\\\'.\.
Mnth.
(2),
t.
s).
12,
p.
),
)t
227:
Ces&ro,
Rom. Ace.
Lincei,
22, 23, 24
Hence
Ui.MM
kBEL'fi
<
Xa
or
In
manner
likr
lli-
sum
<//'-,.
r
[fl
ilian
l't
<
",,'',,
//
where H'
lil
"
it
the greater of
IB
!
:
It
Mi
<
17 and |&|; that

1
Bometimefl <l'îrallc to obtain
se
that
tin-
mote
//
ntli.-r
//
is,
i-
//. A
aiv
.!/
i^
an
iiu-n-aî
Transformation of slowly convergent
Let us writ
and
</
aw +2olH l -hh^=
.
Thru
(
and
l(
series.
tc.
\\v lia\r
rous,',,urntly
^J",/"
whI.'
r/
|"-atin^ this op.-rat
<<
="
lt
'/-f
//
= .'(
>n.
we
/
.
=o y-KjRo )y+...4
l
-f.
+.'>
tind
Th. -n exactly
for
limits of 8mt
lo\v.-r
tl
whm
hand,
limitfl
u]|nT and
th<-
find
24.
the upper
-.., |*p|.
is
while
w.-
SERIES IN GENERAL.
56
can be proved* that in all cases

converges, the remainder term
It
tends to zero as
when
increases to infinity, at
III.
the original series
when x
least
is
Consequently,
positive.
The
[CH.
cases of chief interest arise
when x = l, and
are alternately positive and negative.
a l =+v 1
a 2 =-v 2
aB =+v 3
the terms a
Write then
a= -v,
...,
and we have
where we write
Dvl = vl
We
v2
L^v l = Dv l
Dv 2 = v l
2v 2 + 1> 3
etc.
can write down a simple expression for the remainder,

is a function such that/* (as) has a fixed
v n =f(n), where f(x)
if
sign
for
all
values of
positive
x,
and steadily decreases in
numerical value as x increases.
Dv n =
For we have
dx l
so
I
f'(x l
Jo
Jo
Jo
( Xl
and generally
ri
J)Pv n
I)P
Thus the
ri
Jo
Jo
series
dx
dx l
D pv
D pv + D pv
2
...
consists of a succession
Its sum is therefore

of decreasing terms, of alternate signs.
p
in
numerical
less than
value; and consequently
v^
where
*For the
p. 188.
cose
x=l,
see L. D.
Ames, Annals of Mathematics,
series
'2,
\.>1.
ION
24
1
This ivsult
lea
of th- type
Here
d
*>
II
Teases as
tin-
\\ni-k, in
tin- ac--
arithln.-tiral
tion twice,
starting
tin'
iv.Milts
anl
ti,,
inn
tii^t
r;i]riil;iti.ilis.
aay
at
tO
and
th.-
apply
s,-<-..inlly
and
in
1.
\\v shall
tin-
r=\ if \v- work

a=l -70711
apply tin- UMH
tal'lf
l).-l.>\v
r
~^\\
i
arc suhstaiitially the sain.- \v-
T;ikf
!';i
at j
be âti-li.-d that
may
the \\orl
Ex.
li\
\ve
I
get
-',
58
SERIES IN GENERAL.
Now
'40825 --19568 = -21 257, so that
s'
certainly
[CH.
is
III.
contained between
0-21256 and 0-21258.
But s = 0-81746-5', so that $ = 0'6049 to four decimal places. If we used

the original series, it would need over a hundred million terms to get this
result.
Ex. 2. Similarly we may sum the series lTo 6 decimals, the first 8 terms give 0'634524 and from the next
terms we get the table
:
V.
24
l.i.l:
l.i
reference
n.
terms v
series of
-?.,
for
We
has always the same sign.
IP-v*
ha\.
if
Dh n
is
because the series
we ha\-
positive
in tin-
l-ia.-l.--t
we have
small
.-ry
+ ...).
-/
/>>,
/>/-,
that
\\hi.-i
tin -n
/>,',
Now
hew
will
t<
;i
59
iti.l
lim
-o.
/>-
êt
Dvj.
if
Similarly
tlu*
>*> ^(r, -f Dv^.
ir,
series ca
ii_:li
proof of
tin- tirst
Mai'kotl'
lia\.-
t-\ain]il-
il
.en
to
<lu.-
i>
degree of
by K
Kuininer
Ponoelet
transl'nnnatioiis
oth.-r
!'.,r
the
includes Kult-r's aa a
nn-ili<.l
Mark.tf>.
<i'
above WBS
acniracy
i'<un.l
th- latt.-r>
an
have
trans!'.. rniatidi <l.-rriU-<l
'I'll.-
and
ami
//-'
may
\\
-2)1
sum mnvrtly
Lnvr the
TM
ajj)ly Killer's
method
-'
\\\->\
;
<-f
|.].ly
term>
trn
Kulei's
V& -U
the
next
n<
4 3 *'")
the series in th-
..f
m.-th.-d
<1-
example the ivader may
t> this
l
to 20
>i.\.
\\
the remainder: thus
.e -9011165, and if
for the value
'0004262
get
^057 to six
'I-ofi.
4
+
j'
The >um
,,f
the
mrtli.nl -
Thus
1'
tiist
'1
t.-rms in the
teli
f-r the
1=2 (0-82 2469)=
'
l^v.
p. 178.
F.-i-oth.:
lira.-k.-;
reiiiaiii-.
j..
.-MiiMilt
i:u
/'
l'rin::vli.-ii!.
i./Jg,
1896),
SERIES IN GENERAL.
60
[CH.
EXAMPLES.
Prove that the
1.
series
i_J_ J__L.
x+\
converges for any value of

but that both the series
x which makes none
J__L_
__
~~
1_
denominators zero
J_
1,
x x+l
,
of the
are divergent.
2.
Prove that
3.
If
the
increases to
if
series
co
^na n
2a w
is convergent and
then (see Art. 23)
lim (a^J/i 4- 2 ^2 +
+ an
with
n,
Prove that the
4.
convergent, so also
is
the
is
sequence
0.
steadily
(J/,,)
[KRONECKER.]
series
.
but can be made to converge to either of
On
inserting brackets.
by
n.
Mn }IMn =
a - a? + a? - a? + a* - a* +
oscillates,
is
the other
hand the
its
two extreme
limits
series
convergent.
Shew that if a series converges, it

number of brackets are inserted, grouping
5.
the converse
is
for
x=
the
sum
7.
t\y>
for
Shew
And shew
also that
true, if all the terms in the brackets are positive.
sum
Calculate, correctly to 20 decimals, the
6.
convergent when any
still
is
the terms.
*&
x=
How many
of the series
terms would have to be taken, to calculate
fy to 3 decimals?
that the series
-a 2 + a 3 -a 4 + ...
diverges
if
an =
C_iy'-i
-j-+
n 1
]
a=l/[-x/w + (-l)
although the terms are alternately positive
tend
to zero as a limit.
and
and negative
or
if
8.
If
|#|>1, prove that
converges to the
9.
If
a,,
sum
A,
converges to the
*2 +l
A-*+l
-.
-> a and b n ->
converges to the sum
__L_ +
_2
'
.r+1
>S,
(01
6,
verify that
if
the series
then the series
- a 2 ) + 6 2 ('/ 2 -
sum S-ab.
3)
+ 6 3 (r 3 /
</
4)
-f ...
LMFLE8
111.
1
10.
the
if
-!
>...
.
ei jênt
al
.so
+^(
rt
+ ^3) + ^ (
(/
converse always
i-se
11.
certainly tin.-
is
Ni.sruss
that
ti
cS
_I_ + _L_- +
./-Cj
4-...
+_L_rl +
C,
tending to
ijucin-r
12.
h-n
\\
oo
Verifv that
/
?)
ru'i-iit
<
if
to
x and x
,
is
not equal
vak
I the
i
'
If
is
al.-olutely
13.
She*
convergent
r ia tin-
inti'êr
and
intîa! part "f
/.
_!_ = _ A
frmii
fad
if
that
in
In
verify that
fixed,
'
th-
in.
1'
-in-
th..'
>>e
the -unnnatioM.
tin-
>UIM ran
'>
writti-n
M-\ +
*fl-
_,)
14.
With the
>anie notation as
\-
~
i //<
if
)i<
Find an
SERIES IN GENERAL.
62
Discuss the convergence of the series whose general term
15.
and
[CH.
with cosw$ in place of sinnB.
also that of the series
III.
is
[Art. 20.]
[Math. Trip., 1899.}
Apply Art. 20 to the series whose sums

2
#, and deduce that
16.
n terms are sin(w
to
cos(w + !)
^^sin nd
are convergent
Shew
if
cos
n2 0,
2v n
sin
nO
sin
n2
vn steadily -O.
[HARDY.]
vn tends steadily to zero, in such a

not convergent, then the series
17.
that
if
way
that 2v n
is
r=D
converges
if
(and only
If the sequence
18.
if)
a l + a 2 + ...+ak = 0.
convergent, prove that liniw(aM+1
n) is
must
either oscillate or converge to zero.

If
19.
2an
convergent.
converges, and a n steadily decreases to

If,
in addition, aw
0,
2?i(an -an+ i) is
2an+ i + a n+2 >0, prove that
n 2 (a n - n+ i>->0.
Euler's transformation to
20.
Apply
21.
Utilise the result of Ex. 3, p. 58, to
1
T
~*\
.v
~T~ T
'.
[HARDY.]
shew that
shew that the sum
of the series
x3
*j
".
-j
~T~
tends to the limit \ as x-*\.

It
the
sum
easy to see that

lies
between ^v
(if
<#<!), Dvn
=J and
i(î
is
positive
+ î)= ^j
4
and deceases
thus.
"1
x-
By taking ?> ;l = log( + H), shew, as on p. 56, that

steadily decreases; deduce that
22.
nl
is
D p vn
is
negative
CHAITKK
IV.
ABSOLUTE CONVERGENT
25.
value.
is
matter how
projn-rty,
series
as
;
however,
an
value N
|i'itivf
so
srrics
ii\
tin-
termfl
v trnns
t
is
c-on\ ci'-vnt
^rt-attT
that
:
than
each
ha- the
BUID
MUM
(Ait.
'
.-r-ally
K\.
t-nii
is
and has a
L),
u^ arrange
L-t
.1.
positive
tin-
i'ol]o\\
th- 96068 th-n beOOl
.
.'-V
4^:;
1
1
-
i;
>_
K>
i-j
lim/.
..,
liav-
i_Ui_i
4
^^
2 -ri;
Unit--
ly no means uni\
is
-J
Now we
a
tli,-
.|'
illustration. consider th-
which we know
this
that
fact
i'ainiliar
Il
IK.
I.
J_
'
-2
.,
Thus
and
the
it
is
sum
easily
oi'
>een that
the series
i-
lim/.
..,
lim
t-niis of
ABSOLUTE CONVERGENCE.
64
[CH. IV.
Consequently, this derangement of the terms in the series
sum
alters the
of the series.
In view of the foregoing example we naturally ask under

what conditions may we derange the terms of a series without
altering its value? It is to be observed that in the derangement we make a one-to-one correspondence between the terms
of two series; so that every term in the first series occupies
a perfectly definite place in the second series, and conversely.
Thus, corresponding to any number (n) of terms in the first

series, we can find a number (n) in the second series, such that
the n terms contain all the n terms (and some others); and
conversely.
For instance,
terms of
the derangement considered above, the
in
are all contained in the
3p terms of
are
first (871
contained in the
all
first
+ 1)
terms of
4p terms of
t\
first
(2?
and the
first
s.
series of positive terms, if convergent, has a sum

independent of the order of its terms; but if divergent it
26.
remains divergent, however
its
terms are deranged.
As above, denote the original series by s and the deranged

series by t
and suppose first that s converges to the sum S.
Then we can choose n, so that the sum sn exceeds S e, however
;
small
Now, t contains all the terms of s (and if any

term happens to be repeated in s, t contains it equally often);
we can therefore find an index p such that tp contains all the
terms sn
Thus we have found p so that tp exceeds 8 e,
e
be.
may
the terms in ip s n are positive or zero. Now t

contains no terms which are not present in s, so that however
great r may be, t r cannot exceed 8} and, combining these two
because
all
conclusions,
we
get
S^t r >S-e,
Consequently the series
Secondly,
if s is
If
we attempt
Art. 25,
we
tf
>S
is
terms
r^p.
sum
cannot converge
converges,
S.
for the foregoing

must also converge.
;
divergent.
to apply this ai ûnicnt
..1-1+I-.J+...,
find that the
that
if
if
converges to the
divergent,
argument shews that

Consequently
in
,,
the
two
series considered in
1-1-1-1 + 1-1-1+,..,
arc partly
and as a matter
t<>
of fact
////<////<.
we
sec
Thus we cannot
frmn
Art.
:.'">
pi
that
this
25,
M:I; \\'.I.M
26
65
>iu.il;ul\,
that
is
It
-1
(r
now easy
convergent,
term
I.
TIM
in
a; BO, since
not -renter than 2.1,
now
SUJ.JM.S.'
al'ti-r
any
Tli.-n
is
<
sum
th.-
that N.
.1
are
/;
Ex.
absolutely
th.-
B
B = S+A.
three series
for each ser
saim-
tlie
hy what has been proved A'
ntain only
positive
and
tern
ll-A
N'=/;'-.|
}T..\ in-'
and so
of a;
th.- MIIII^ ot'
siij.|ns,.,l
hut
/'
II,:
is
the corresponding
_jes to a sum
twic,.-;
convergent,
\vl
= S'+A';
to
-;
than
.l.-rani;-i-im-nt
H'
"
a series
if
fails
it"
'2|aJ, ami th.-n introduce the new

contains no negative terms,
The aeriea
i-
in
,.-
not altered by derangement.
n |].
:-j
no
I,,-
(,,
prove that
is
0Zaw
write
ami
to
sum X
its
argument ueed above

ti
h.-i-r>\.
til--
although thi> happ.-n-
tin-
thf
ii-iil-r
1.
>.
1-.
convergent
.trly
rgent, ;nnl lias
tla-
An.
l.y
11;
same sum
;m.l
after
From our present
point
betw<
ot'
,.1.1
l.y
The
is
__L_JL.^
12
1Q2
11).
any deran^fin.-nt.
-f
'
-.ulain.-d
series
2.
ih-
t<>
t!
Ex.
therefore
th.-
tact
wriefl a,
that
Hew, WQ observe
that
tlu- iiu-jiiality
.^-i+j-j_^ + ...
these serifs are not absolutely OO1

li\.-i^.-iit, and ..f couise we havt-
o'
to infer thaTli.-
affairs
last
explained
.it
should
n-sult
article,
U-
we
lind
)-N,
lim
thrii-
lim
(.,'
thf
-tate
of
uîno- th- nota'
./
).
-f
,'/
\A
limy -
.<
-8+A
^-11
).
;
dill'-!
sum. whose value depends en
<
(8+A
whereas there the se|Uencex

êj.arately. althouo-h
lini(r,-//,
so that
with
rastt-d
c. .nt
at tin- h.---innino- of Art. l^s
th-
>
>
-.verges
to a
66
[CH. IV.
Applications of absolute convergence.

Consider first the multiplication of two absolutely convergent
series A =2a n> B = ^b n
Write the terms of the product so as to
27.
form a table of double entry
a^
afiz
a-J) 3
a^
...
/ t / t / t /
a6
ajbâj)} a b
/ / / t
aj)âj)âj} a
t
/ /
2
3 4
...
Z>
a4 b l -+a4 bz-+a 4 b 3 -+a 4 b
It is easy to
prove that
AB
is
the
...
...
4:
sum
of the series
where the order of the terms is the same as is indicated by the

table.
For the sum to n terms of this series (1)
arrows in the
is
A nBn
if
Now A = 'Z
an and B' = ^
Thus the
are convergent
bn
by hypothesis.
series
(2)
obtained
by removing the brackets from
convergent, because the
number
sum
of the
(1),
absolute
is absolutely
values of any
of terms in (2) cannot exceed A'B'.

Accordingly, (2)
sum
as the series (1).
Since (2) is absolutely
AB
has the same

convergent,
we can arrange
without changing the sum.
in any order (by Art. 26)

it
Thus we may replace (2) by
a 1 6 1 +a2 6 1 +a 1 6 2 4-a36 1 + a 2 & 2 +"
(3)
/>,
+ "/> + ...,
1
following the order of the diagonals indicated in the diagram.

Hence we find, on inserting brackets in (3),
= c + c 2 +c8 +... to
C2 = a2 6 + a 6 2
c3 = a 8 6
-A.S
where
cl
= a b 1)
1
oc,
H-a2 6 2 +a 1 6 3
and
For other results on the multiplication of series thr
should refer to Arts. 34, 35.
iv;ul<-r
AI'I'LH
27
MM.!
wh.-iv
;/
,<=
theorem of Alt. 26
tin.-
in
arrang
in
./
say
//
have !<
to
sutli-i.-nt
'>
-f
>
r/:
...
lowers
Art.
1<>.
w.-
-f l*^.
\vliere
this
966 that
,;<
\-
if
A,
iv.|iiires
= liin
condition iv.juiivs that j8d <X, and that

than BOOI6 ti\-d value; and thm the n.-ct-ssary
1668
certainly not alter th- sum

most of the ordinary 68864
will
In
shall
last
'I'ln-
19
u-A+ftf+.-.+Af, A.-IU f-||:
0.1,
and from
of
polynomial
hriv
is
It
is
step
i<
application of
us-|'ul
to justify thf
the condition
is
th-
BOTJ
\=l. and
is
//
of th-
1'orm
th.-n
In particular,
uri,
,,,,.,.
sutlii-ient
.^__2.
satisfied
ainly
be
if
,,
lut
it
to tak-
is
inaih
pusitivr
AB an
has ih.
ird
l.y
jurii'-iilar liy
in
in
/<>f/
>f
tliis
1
[1
/
x 2
1.
which
iav ). t.-mpt.',! t,. think th;it the coiulitiuii \y\<\ w..uUl

tliis is m.t the case.
K"i- we have to ensure the
illustration
>um
<
\vln-n
i>
is
-(ir-.r-')]-
any vahux }, bed
i"
get
-X*
and
V
\vhi--h
j.oint, r..n>i.l'-r
= (!-.)
>f
\viitt-n cut at length,
form.
-'.
\V!MMI
1)
\toe-a?\<l. This conditi

iad
lyin^r Between 1 -r s ;
i'.nt
if
th-
68
[CH. IV.
This equation is true if both series converge although the proof does not
follow from our present line of argument. It may be
guessed that, in
general, the condition found for f in the text is unnecessarily narrow
;
and
we
we wish
ever,
is
How-
number
this is certainly the case in a
of special applications.
are not here concerned with finding the widest limits for x
to
shew
is
that the transformation
is
certainly legitimate
what
when x
;
properly restricted.
In view of Riemann's theorem (Art. 28) it

may seem surprising that
the condition of absolute convergence
gives an unnecessarily small value
for f.
However, a little consideration will shew that Riemann's theorem
does not imply that any derangement of a non
series
-absolutely convergent
will alter its
sum
but that such a
be made to have any value

which may easily be of a far more
n
sweeping character than the derangement implied in arranging â^
according to powers of x.
by means
series can
of a special derangement,
Riemann's Theorem.
a
// series converges, but not absolutely, its sum can be made
to have any
arbitrary value by a suitable derangement of the
it
can
also be made divergent or oscillatory.
series;
Let xp denote the sum of the first p positive terms and
yn
28.
the
sum
of the first
negative terms
where p,
tend to
oo
oo
according to some definite relation.
lim xp =
p
then we are given that
lim (xp + y n ) =
=
lim(a?p-2/ n ) s,
oo
lim y n =
oc
Hence
n >oo
>oo
Suppose now that the sum of the series is to be made equal

o-; since xp -*vo we can choose p l so that x
er, and so that
Pi
p l is the smallest index which satisfies this condition. Similarly
we can find n^ so that 2/ %1 2^ 0", and again suppose that 71
is the least index consistent with the
inequality.
>
to
>
Then, in the deranged series, we place first a group of p l

positive terms, second a group of n^ negative terms, keeping
the terms in each group in their original order. Thus, if >S, is
the
sum
f
<S
of
O,
terms,
if
is
v<r,
if
p^ v
//
,/,,.
<r.
>
pi)
//,^-ho-;
terms, where
n8
ifl
IMKMANN
27, 28]
Tin-
and
indefinitely,
is
\vhil.-
if
<!
,-f-
i>,
tin-
if
</',
>' = />,.+
/> r
//,-.
t'-riu:
/', -)ili
/*,.+
/*,
is
N,
>^=p
.>^
/>,
N.-T
i'<.r
'!
on
*
evidently be carried
oannoi exceed th- (/>,+
I'M:
positive,
can
d.-ar that
is
it
BEORE14
construction
of
im-thon!
^
r
t-rm of
jMiMtiv.-,
MC
tin-
does not
Imt
these
.-it
ehangefl sign
terms.
Thus, since
increases,
ti-rm^ "f
tin-
we ha\-
th'-
limN
mu-t
tferiea
t.-nl to
zero as v
=&
easy to mo. lily tin- t'oivôin^ m.-thoil so as to get a

<li\fi-L;Tiit or o.scillat'ry 901166, by starting from a sequ<-:
is
It
which
is
ritht-r
turn to be
in
<li\
tin-
vruvnt or oscillatory and takin.
which satisfy thr
inlic-s
tir-i
>^-î s>y
--<rii
;ml so
howe\
work
and
of fact,
fur pra'-tical
\vr
tnn> prooeM i- 'juite out of the

have to adopt an entirely dift'-
>t.-adiiy dr<T.-asing to zero as
!i-ti..n.
ami
-i)"-y(/0, \vhii-h converges,

\-
that, in
tla-
tei-ni
jiusiihv
dfraniTfd
f<.ll.i\vrd
is
tin-
s.-ri,->,
negative terma (so that
+ r = r).
-In
Then the sum
and
Suj.jiosr
that
first
/'(-")
''-
inction
<>f
hange
in
sum
tlu-
\Vr ha\- thus
tcnd>
/-/'(/')
iM-tw.-iMi
th,-
and
*r-
''
?'*
integer neireet
s..
and suppose
to n
ji--iliv,-
between
lios
with
l.t-.-au-.-
then
n,
llOOSe v to
th-n
be
=/.
*M^/'
.r
of these r tern:
Tin;-
taking the series ^(-ly-'/r*,
integt
-r r
intinity
/(*+?),
/,
rr-ult
and
/-
increases
r-
of th.
!ii->t
tt-nn
t-nns i-,ntain
Art. 21.
i>\
itiv.-
first
win-re the second l>rark.-t contains r trims,
tlu-
> xP -
on.
As a matter
tin-
.....
int-'jualii
we
see
again with 1( -i)--
1 !
that v
^,
may be
may
bi-
!ogn.
i.
-J-J,
j..
1
to xitisiy tin- i-.iulitiiin.
Hut
tl
to take
always the nearest

.:--nl.
order
70
Next,
if
value of
c,
lim nf(n)
is finite,
however
small, a value
[CH.
say equal to g, it follows that, for any positive

nQ can be found such that
Let v be chosen so that k = \im(n + v)/n.

It is easy to see,
by an argument similar
Hence the
alteration
to that of Art. 11, that

rn+v
dx
contained between the two values
is
(#c)log.
Thus, since
e is
arbitrarily small,
we must have
Hence, if lim nf(n) =g, and if k is the limit of the ratio of the number of
positive to the number of negative terms, the alteration I is given by l=%glogk.
In particular, since 1 -* + ... log 2 (Art. 21), we see that when
this series is
terms
its
arranged so that
sum
is
log 2-f
log
k=%
terms to each negative term,
While,
if
positive terms correspond to n negative

and so, if there are two positive
log 4
Ten
we
get
there are two negative terms to each positive term,
we have
a result which has been proved already (Art. 25).

Finally, if there are four negative terms to each positive term,
we
find
To save space, we refer to V. of Pringsheim's paper for the discussion

of the rather more difficult case when liranf(n)=Q.
EXAMPLES.
1.
Criticise the following
paradox
+ 1+*+*+*+*+...
2
+J
+ 1 + ...)
= 0.
2.
If a transformation similar to that of
Ex.
1 is
applied to the series
1-i+l.lo.I.
2* 3* 4
6
Hhew that
(if
p<\) we
obtain the paradoxical result that the sum of the

1, the result obtained is correct and rx presses
p>
series is negative,
But, if
the sum of the given series in
terms.
terms of the corresponding series of positive
Tl
MI'LES.
iv.)
3.
=i/(n), shew that
wt writ- /(.)-!/*, and
If
+ ...H'-!* 1
t/W
")
[It
is
proved
An.
in
4.
I'K.VC that
5.
Apply
that i
71
a transfoi mation similar
tliat
of Ex.
to the series
i-i+j-H~,
ami piovr that
tin-
----.
which converges to
>!
suit iiiur
-um ItH
ies is
that of
tlian
th'-
-iven series.
convergent series may be converted into an

wriM ly tin- insertion i.f brackets. [See Art. 5.]
Any oscillating s.-rii-s may W converted into a convergent sei
and the bracket* may be arranged so that the
in>.-rti..n ..f l.ra.-ki-ts
Any
6.
n>n-alis..lutrly
absolutely coiiveru'ent
la7.
In
sum
Mi-iK-i-
that
unaltered
it-main
sutli'-it-nt
valu.-
tin-
Prove that
p.-rfe.-tly
Shew
9.
Here
//
=0
that
is
./
Il(lt
(2),
t.
is
dfierminate numl>er,
14,
luit
>nii
lim
./
tend to x
in
siidi
that
if
.<, 1
way
that lim
(</;<)
sei
1-..-
= ^ + (/,+a s +...-l-a (1
may
tei
1890, p.
that
supposed not to U> an integer, and
to le
Find the product of the two
Shew
V"tL
+...+-K., and
10.
convergent series
the order of the
displacement <>f the //th term by the

to zero as n increases to oo.
'
definite.
wheie p and
in
implies that
n.n-al>>"lutt-ly
may tend
[Iiui:Ki..
is
that the product of the
greatest subsequent term
8.
.f
a <-.-rtain chain:'-
aftt-r
30
limits of
any of the
to
i|ual
+
,
'-. ..+(-1)-
then
.
1*7.]
CHAPTER
DOUBLE
V.
SERIES.
29. Suppose an infinite number of terms arranged so as to

form a network (or lattice) which is bounded on the left and
above, but extends .to infinity to the right and below, as
indicated in the diagram
The first suffix refers to the row, the second suffix to the
column in which the term stands.
Suppose next that a rectangle is drawn across the network
so as to include the first in
rows and the
first
n columns
of
the array of terms; and denote the sum of the terms contained within this rectangle by the symbol sm |M
If # /M>M
s
t
a
limit
as
n
and
tend
to
definite
approaches
infinity
.
the
same time
<
(but independently), then
<
s is called the sun*
the double series represented by the array.*

In more precise form, this statement requires that it shall be
possible
to
an index
find
positive fraction
e,
//,
corresponding to an arbitrary
such that
!-V
-*!<*>
if
>,n>p.
the last inequality is implied that m, n are subject to no

other restriction than the condition of being greater than /z.
By
*ThiH definition
is
framed
in
arrordaixT with the om- adopted
(Miinchener Sitzunytberichte, Bd. 27, 1897,
p.
101
l.y
see particularly pp. K:i.
UK
OF
Sl'.M
29]
BLE 3EBU
DO1
th- e
pn.p.-rty
HIM
s
tii
s-.
-y
lini
,r
|iiati<
->
tinis
symhol
eoii\ .-ni.-nt
t.)
mode
u.x,.
it
of MI n
ia
<e,
I'm-
method
from
:;1
it
whidi
i,
tl>
eOUIM
of
iliis
'I'll--
\\-IK-H
n.
arily
t-nd t"
//
filiations
lini
.s-
imply that, îvm any

can tind //. Midi that
Berieci 18
similarly divergence to
It
aNi
is
th.-r.'
jM.îl,!,.
establish
-/:
that
v:
lio\v-\
fr.
.-r
'
\V<-
d(uM-
may
lii"
d. tint-
and
m.
Beriefl
modifying the ni'-thod of Art.
any
l'oi-
these
l.i
>f*:
'"
tli.
t-nd
will
fi
tin-
in
littlf ditliculty
is
it'
then Baid
than
ber
that
so
/z
/"/'///.
nuinlH-r
jH.vitiv.-
find
;i
or lim x iHi?J
y:
8m,n>G,
d<>uM-
//
ini])ly
I
tin-
and
in
/"*///
n.-t
doefl
//',
we can
converges,
>
that
pruvidi-d
t> /..T., wh.-n
sr.ju.
.!
l+*m-\,
that
t'..lln\v.s
and
unless sum.- indi'
ii
i<
''
|tt w>H
to
nthei-
give values dilli-ivm
not MiMid-iit
ifl
as to tlif
.-iM.-<|
d'Mot'd
I).-
d..ull<-
l.y
and lim
8m
(m, w)
-imply that
i-
between two
numerically less than e. ii
of
t.Tin-
tli.-
wh.-iv of
dition
d.-not,'
is
(-"iii-s,-
//'.
r.-ctan^lfs
the value of
//
/<
and
than
/*
l,y
tli,-
replaced ly
\alue of
^, such
"The
Wl,
approaches
that
;;
when
,n
i>
|<
if
full
l>y
rnnir-i
m- in
Tl
e.
it
tli-
must be
i>
>
yi.-ld-
At.
and so we can
limil
IS.^K^
in
</>
'/
(so that th- rectangle
Then OUT condition

!o-9 -<r M
Heiu-e T
>-
depend on
will
.,nd to see that
/*.
if
lt>
1900,
DOUBLE
74
Now
so, if
if
p,q>n>[jL 2
and n > // 3 we
// 2
',
the greater of /^ and
is
// 3
<K
<r
p,,-
S
\
P,- \<
S
Ex.
1.
Convergence:
Ex.
2.
Divergence: If s w> M
Ex.
3.
Oscillation
30.
[CH. V.
the general condition gives also

|
and
SERIES.
If
find
=0 and /x^2/c.
= m + w, the condition of divergence is
= ( l) m+n the extreme limits are -1
m>w =l/m+l/,
If *,,
^ 2>^-
if
>
satisfied.
and +1.
series.
Repeated
In addition to the mode of summation just defined it is often

necessary to use the method of repeated summation; then we
form the sum of a row of terms in the diagram, and
first
oo
obtain b m =
oo
a miM
after
which we sum
n=l
we
This process gives a value which
(JH)
this
6m
denote by
(n)
sum by rows of the double series.

manner we define the repeated sum
called the
is
In like
\ ^J
n=i x m=l
j
^m
>
2 2
or
which is called the sum by columns of the double series.

Each of these sums may also be defined as a repeated
limit, thus":
2 2a m M =
,
(m) (n)
lim (lim
m->m n-+<x>
s m>n ) or
lim s m<ll
(m)(n)
with a similar interpretation for the second repeated sum.

In dealing with a finite number of terms it is obvious that
M / N
\TM x^
N M
VM
v*
\
\
*-SSv^Hi;gg<v.;-
But
lias
it is
the
a)
sum
by no means
s in
necessarily true that
if
a double series
the sense of Art. 29 then also
s=i
m-
indeed the single series formed by the rows and columns of the
louble series need not converge
<>//, I ml may oscilla
/
REPEA1 ED
29, 30)
That
th.
lini
has
quatiow
I'aet
<
Vn
lim
9eri& is conr*
rows
/;/' //'.
ê,
if
//<>//;
hypothesis, this limit
dim
lim
'
u->
like
/Ae
f
</,, >i/,/,
we have
l>y
o&ot*
MI that
In
ii
proved. 1.
<i,l if th,
.//.
lim
all.
I'l-iii^h.-im
since,
1-0
hi.'h
In
he example
oolomnt need not oonrergc
'A
exists at
MM Ml-
81
>(/1
/t
= x.
n-*>oo
manner we can prove the other
\\'hen the douh''
half of equation
convergent, the equation
not
i^
<
).
g(fr*.)-|(!
i^
mt
win-never the two repeated 961166
valid
n.M-r.ârily
convergent
1
h.
no reason \\hau-ver for assuming that the equation

lim (lim *,.)
= lim
(lim
v)
the repeated limits

/>),
lim
Knm
not
(lim
x fllill
= n,
1'rin^shrim's theorem
(mvrê
tiiul
is
it
.<,
with
<
.....
lim
'2
is
valid:
(lim
.,
+ /o-, we
,,)
= = lim
.lnMe
the
series
ûpposed con-
the truth
[
(;/j
tin-
ln-in--
Init
tli-
= IHI<
= !.
clear that
(the rows and columns
vergent) unless e(|uation

no reason for avximiin^
.
we
lim (lim
oi
douMe
find
(lim -V,
Mi
lint yi-t tht- d..ul.le >eiies ca;
FT
true,
some purp.êx it
without tnmMin^
is
to
nvrt'ul
to
conîd.-r
know
th-
thai
w,i =
equation c2)
is
DOUBLE
76
SERIES.
[CH.
v.
convergence of the double series. In such cases, conditions

may be used which will be found in the Proceedings of the
London Mathematical Society,*

would go somewhat beyond our
A
may
the discussion of them here

limits.
further example, due to Arnclt, of the possible failure of equation

be added
(2)
Thus
lira
(lim
m*-x> n
but
s m>
= - J,
lim (liin* mt w)=+i-
Other examples of points in the general theory will be found at the end
(See Exs. 1-6 and
of the chapter.
31.
9.)
Double series of positive terms.
been proved in Art. 26, we may

a series of positive terms converges to the sum
in any way, it will have the same sum if summed in
any
In view of what has
anticipate that
s
other
if
way which
includes
all
the terms.
For,
however many
terms are taken, we cannot get a larger sum than s, but we can
get as near to s as we please, by taking a sufficient number of
We
shall now apply this general principle to the most

useful special cases.
(1) It is sufficient to consider squares only in testing a
terms.
double series of positive terms for convergence.

= n; then
Write for brevity s mn = o-n when
must converge
to the limit
s,
if
sw
does
iH
converges to a limit s, so also will s m<n

so that <TM lies between s and s
e;
/j.
.
greater than
/z,
so.
plainly
Further,
if
crn
<r n
For then we can find

but if
and n are
we have
so that
= ,>*
e.
Hence s m<n converges to the limit 8.

The reader will find no difficulty in extending the argument
to cases of divergence.
*Bromwich, Proc. Lond. Math.
Soc., seri-
-J,
\..l.
I.
I'.MH, p.
ITti
SERIES
30,31]
If
mem
for
puri
ion,
we
iccession of curves*
plainly,
tained
two
'
aqua
when
sum
p,
tin-
mi therefore
V particular CUTVe
of tin- square-, and that tl
thus
if
is
to
tt-iid
sum
the
in (1)
greater than
i>
//
1)V
Thus, since
\vt-
>
<
lint
lik-
nianii.-i-.
cm".
-N-
and SO
tin-
can
that
tti<
,-ive,
can
>
e,
it'
fi
>
;|
tin-
th,-
<>|'
//
N aft
tin-
if
h.-iv!', ,r,.
particular da-s
<>t'
ly cnclo-in-'
>lu-\v
M|U;UVN (and
drawing
-jiiares)
to be con-
it
tn iiitinity in all <lin---
taking
rr
In
th-<'
^crâ.
l-'urtli.-r.
makr
for
'>nd
s.juar.-
ln-t\vM-n
two
!'
the
cur
rect
boo, in
curv.-s
diagonals, equally indiiu-d to

the nrt\v<>rk as indicated in
u^-d
tin.-
in
-J
virtue of (1)
i-
t'orim-d
ly
Imri/nntal and v.-rn'cal
tin-
riidit-hand
10.
Tin-
summation ly
he noticed
>.|Uar-s
is
indicated on the
Thus, hy sjuaivs.
r(.2l
we
are suniiiiin^ the
+ ",, + ",, +
)
-f-
\vhlly IT
that each curvr enclose
in
l-l't.
It
should
DOUBLE
78
SERIES.
and by the diagonals we get

a n + (a a + a 12 ) + (a 3l +
22
[CH. V.
+a
13 ) -f-
. .
Of course the equality between these two series is now seen

to be a consequence of Art. 26
but we could not, without
further proof, infer theorem (1) from that article since Art. 26
;
refers only to single and not to double series.

By combining Art. 26 with (1) above, it will be seen that
(3)
can
No derangement
alter the
terms of a double series

sum, nor change divergence into convergence.
of the (positive)
important to note that:

the terms of the double series are positive, its convergence implies the convergence of all the rows and columns,
and its sum is equal to the sums of the two repeated series.
For, when the double series has the sum s, it is clear that
It is also
(4)
8 m> n
When
cannot exceed
and consequently the sum
of
any number
of terms in a single row cannot be greater than s.

Also, for
lims
any fixed value of m,
m>n exists and is not greater than s.
(n)
Now we
can find
than
Consequently
JUL.
JUL
so that s mt n
>
silimsTO|W >s-e,
(n)
(HI)
In a similar way,
we
see that each
m>
m, n are greater
//.
column converges and that
to (4), we have
The terms being always positive, if either repeated
:
is convergent, so also is the other
and
if
(n)
As a converse
(5)
if
e,
= s,
]
Km [lim sm
Hence
the three
sums are
and
the same.
For, suppose that

lim [lim*,,,
Hence by
series
also the double series;
= *,
Art. 2 the sequence (a-m) converges to a limit

then
it
follows from (4) above that x = rr. and that
other repeated series has the s.-mir sum.
;ui<l
y.j
tin-
ivnd'-r will
'I'll.-
in
and
(4)
find
"N\ BRGEN4
FOB
PS
31, 32]
littlr
in
ditlirnlty
79
ni'.difyi:
so as to cover the case of di\.
(."o
-i
Tests for convergence of a double series of positive
32.
terms.
we compare An.
It'
<
If
>
t;
s
,
with
<
mother doub
///,;,
see
series wi less il>i<
<l<>i>/<
known
;*
/,
we
the last article,
.f
>
rms of a
to converge, th*
'
Similarly for
Tin-
in.ist
d;
wln-iv ^^',"
that
tliis
seriee
.inn!.],-
ntain-l in a
s.|iiar:
and therefore T
Converges, ly
(
>n
ly
=(y/>
-in^lf
c2)
In-
i>
convergent we note
.i'
sj.1.-
limit.
last
i,
-i\-n by
to
i^
convergenl singl
n
recognise
of tin-
in
is
plai
series'
.-r^'-nt
i^
-\\u\
plainly r.jUa
'
r<>ii^.,|u,-iitly
tin-
<l>ul.-.
articl.-.
th- chirf ivp--
\vli<-i-.-
t>
as
diagonals,
ill-'
in
greater
+ ô- +
;/
aeries;
of tin-
liand.
tlirr
tin-
ha-
'
\y\- d' c)ii\
iiiipnrtaiil
amtH = (<
with
,n
tin-
last
,i
i'<
and
divergence'
"
is
a
^/>
1
iliv.-r^.-ncf.
In
artirlr.
take
thi^
way
Beriea
which
aeon
i-
divci--vnt.
coin paring
.n
'+..
Ex.
Ex.
?.)n-*n-fi coir
1.
2.
l'(//j
+ /)~
ft
a>l, /^>1.
di\
divei
^v=rK
Ex.
3.
- A
,
opavergea
if
X>1,
div-i
-
\\h--iv
i>
Thus thf
and
th-
L:I>
roiidit in->
-J.
iiu.
uid
.1.'
"t'
b\.
i
dix'i
it
di
tin-
with
Bum
DOUBLE
80
SERIES.
[CH. V.
The reader will have no difficulty in seeing that the following

generalisation of Cauchy's test (Art. 11) is correct:
(2) If the function f(x, y) is positive and steadily decreases
and y
as x
increase to infinity* then the double series

S/(m, n) converges or diverges with the double integral
to zero
f(x,y)dxdy.
However, nearly all cases of interest which come under the

can be as easily tested by the following method, which
depends only on a single integral
(3) // the positive function f(x, y) has a lower limit g()
and an upper limit G() when y =
x and x varies from
to
and if G(), g() tend steadily to zero as ->x
test (2)
then
the
double
series
22/(m, n) converges
//>
&()$;
if
the
integral
but the series diverges if the integral
converges;
00
0()^
diverges^
For then the sum of the terms on the diagonal x-{-y = n

lies between (n-l)g(n) and (n
l)6r(?i); thus the series con30
verges with Z(?i
1) G(n), that
is,
with the integral
but the series diverges with 2(?i
that
l)</(?i),
G()d\
with the
is,
y ()(d(.
integral
Ex. 4. A particular case of (3) which has some interest is given by

double series f(am* + 2bmn+o*,*), where /(#) is a function which
2
steadily decreases as its argument increases, and am* + %bnin+cn is subject
to the same conditions as in Ex. 3 above.
If A is the greatest of a, |6|, c, it is evident that
the
less
is
than A[a? + 2.r(
the same
2
/?
[{(
that
way
But if b
is
if
-.r)
+( -x)
=A
</,
we can put
negative,
When
2
.
b, c,
is
=(Ap)
*That
is,
and
we suppose
/(f,
f
Tin- us.-
Ml'
(</.M.
singlr int.'-ral
H'o-Ar,
lS7'i,
Hut-wit/ (Math. Ann'i/m,
is
we
see in
uivutrr than
the expression in the form

,
if
4Jfl=ac-&
</tf)
if
Cii-m.-mii
positive,
the expression
+ c-26).r + (6-c)^} 2 -f(ac-^)P]/(a + c-26)^^ 2
Hence
the least of
is
|i.
I',. I.
i?)
'
|',.r
p.
//).
=='!/
testing mull ipl-
l-VJi;
II.
=/(.''.
""''
-{
an
s:ti.
s.-rirs
'
all
I'liat
Tin-
.-ilmv.-
in\
t-st
tonn
srrniS to be
i-at ion
s.-cins
is
due to
given by
to lr imvi-1.
32, 33]
/.
il.trlv
seen to diverge
tl
and
al.o\e;
:i
..
tin-
>eries
converges;
\\
>0)
.'i
the other hand,
..ii
that
-li.-wa also
it
'
t
tin-
181161
if
divergei
p)
Absolutely convergent double series.
33.
as in tin- tlifry
lust
!
iit<l ;l
Tlit-
that
in. -ill.
all
tin-
I'..sitivr
ten
ivMilts
In
by Hardy
(1)
Ar
(Proc, A'-"/.
sc-n
Consider
liy
ll'-i,I'
tli--
plheim'l
any
aK^>lutely convergent
reader \vlm has advanced
slmuld
sul.j.-.-t
voL
l.
t-misult
I'.'o:;.
pajM-r
p.
t\v..
examples:
siiupk-
+ 1 + 1 + 1 + ...
+ 1-1-1-1-...
first
jit
+0 + 0+ ...
in tin- tirst
tt>\, sn that
l-linition.
Hut if
if
l..uM.-
-r
not necessjirily trm- for n>n-al.>'.lutely i-<nivergent
taking
tTiu-
^,, = 2
!'.
once to shew
,-it
applii-.l
:il
tlie
N''. c2)
.I/"'/'.
+1
\vli.Tc all
for
tin-
be
Art.
in
true
of
.l.-in.'iits
In-
the series
call
,n\ .-r^-nt.
,-,
oonnezioD
tliN
w-
I,
,,
pn.v.-il
Tliat tlu-se results are
may
sing
still
tin-
l-.-ycnd
<>!'
if
in
u^.-.l
.,1
aeriea
il..nl.lc
tin-
?/J,
.iniuiiij,'
tlie
rows and
\.\\
s.-ri-<
it
has
tin-
c.lui.
>um
i\
a--.-..
riling
the d<uil>l- series into a

Alt.
(1:
:il
),
\v,-
get
1+2 + 1+0 + + .. .=4.

nlivii.iisly,
too,
tip-
convergence
--f
tlii-
series does not
tWO fowl ami columns (compare

(2)
'
the (loul.K- series suggested
1,1
2~
:i
+ 21-,
:i
1,1
7
1,1
73
L6
153
imjily the con-
Art. 31).
by Cesaro
^8~16 + l6~-
16*
l.s,
(.})
16
15s
+ r6'
DOUBLE
82
Here the sums
of the
rows
SERIES.
[CH. V.
in order are
L I
2'
22
23
'
I
24
'
*"
'
'
and so the sum of all the rows is 1.

But the sums of the columns are
+ 1,
-1, +1, -1, +1. -1, +1,
proving that (5) of Art. 31 does not apply.

This result is specially striking because each
terms being
(the
formed by the sums

,
row converges absolutely
t+i+i+J+J-K..)j and
than
less
...,
rows
of the
is
*2i
+ 53
~*~
secondly, the series
which
>
2i
also converges
absolutely.
But the justification for applying (5) of Art. 31 is that the double
series still converges when all the terms are made positive, which is not
the case here ; since the sum of the first n columns then becomes equal to n.
The
fact that the
sum
of a non-absolutely convergent double
may have different values according to the mode of

summation has led Jordan* to frame a definition which admits
Such restriction seems, however,
only absolute convergence.
series
unnecessary, provided that, when a non-absolutely convergent

we do not attempt to employ theorems (1) to (5)
series is used,
of Art. 31 without special justification.
For example
in
Lord Kelvin's discussion
of
two
the force between
electrified spheres in contact, the repeated series
"
m+n
r*
m=iUt'i
is
used.t
mn~\
(-l)
(m + nY J
(m+nf
This series has the sum
(log2-), and
it
has the same value
with respect to m. However, Pringsheim's sum does not

exist but oscillates between limits^ J(log2-f) and
(log2 + |); while the
if
we sum
first
diagonal series oscillates between
oo
and
+ QO
A special example of deranging a double series is gi\vn

rule for multiplying single series given in Art. 27
the
by
above.
34.
Suppose we take the two single

construct from
*
Contra cTAiiaJye,
H.-.lrick), vol.
t-K.-lvin,
:|.
(see
them the double
/.'
Bromvvich
9).
1,
i>.
1,
p.
;>!>_';
= 2ttn 5 = 26n
,
P = 2a m bn
and
<-..mp;nv <."iii>ufs Aiiu/i/xi* (translation
l>y
:r,7.
/*///// >,f
;iiul
t.
series J.
series
MiTfricn/
Hardy, /Vof.
MM.
l.,i<l.
Mnf/i.
So,-.,
sri i-s
-J,
>1.
'_',
l!Ol,
p.
lt!l
MILTI1M.H \TI"\ OF
33, 34]
It
ha:
/'
ii.it
/>
.1
ponvi
*m.n
BO that
liin
= ("i + ",+
J
lYini^h.-im
--use, provided
lu
ire
i-
>,
in
eonvi-r;r,.s
-I.I:
+ ",></>,+/>,+ ... +
'
I!.
-*
lint
praeiiral work in analysis it

ilouhle aeriea /' into a single series,
tli'-
f"i-
usually ehoflen
This
It
follows
l>y
diagonals (see (2) A
-'. where
i^>
Art.
tVoin
Once
at
sum
tin-
hein;:
senee
siii-'l,.
that
>>
their
Jut,
,!},',,},
^_
und.-r
cnu>'<
if
cinMiinstanr.-s
tht's.-
//
//'
m,
product
<t'
tin-
we
write
;/>
both of
or
we have
/->.
tin-
of
8wn
An =n +
Is
'i.
f<>
/<ii
not converge
f/lt ,f
the aeries
/>,<///</ A.B.*
ti n
-}-...^-'i nt
should
/V<',V,>,/
pi'oduct
l>i
'.>+.
..
For
+ !>,
\ve
=alBn +atBn . l +...+amBv
Ol +Ct+.,.+On -A lB1>+AJBn . l +...+A
Hence
(App., Art.
and again
<AJ.I...
lim
<'
1.1
= AH.
ha>
II
jual
'
that
i.y
t->
...
r.
,.
have also
'
-*-.!.//
./.'
pr.ivi-tl.
+ (',+
t54
(For an
oonvt
provitlrtl
((',
An.
L
Hence
when \\m(' n =
r>4>.
lim
.-onvTL.'-.
A,
Abel's theorem:
cl +ct +...+cn
Now
it
one
however,
t<>
lciil)l-
if
eqwil
,'!<' i'f.
terms from ~\c n cannot exceed
absolutely,
serif*
ivo
is
and 2|cn converges because the sum
'.(S^,,!);
If.
!' -au.se 2|am |.|6.| converges to

oonvergentj
alôluti-ly
<-
ssary
the one
to convert
"
r
4-
...
+ 4 HI )=
A
,/y.
alt.'i-nativ.- |'ioof,
;i
thi?
siinil.u
sum
.!
nn-th<Kl.
tin-
.1 />.
see
Art 54 be
diagonal series, v
\
i-oluinii.
tl
tind
DOUBLE
84
SERIES.
[CH. V.
2c n cannot diverge (if

are convergent), although it may oscillate. For, if
= oo, and therefore also
divergent, we should have lim Cn
It should be observed that the series
2an 26W
,
2c n
is
IV
Art.
by
154;
whereas
limit
this
must be equal
AB.
to
If
2c n oscillates, it is clear from the article quoted that AB lies

between the extreme limits of 2c n that in some cases 2c n does
oo and -foe)
oscillate (and that its extreme limits may be
is evident from Ex. 3 below; but in all cases the oscillation
is of such a character* that
;
IV
Ex.
1.
Undoubtedly the
of power-series.
Thus,
if
cases of chief interest arise in the multiplication

two series
the
are both absolutely convergent for \x

g iven bv
+ +
<r
(see Art. 50), their
is
product
0^ 02^ + ...,
which
is
|#|<r; where we have written

+ î&n-i + + an bQ
also absolutely convergent for

c
=a
cn
=a
bn
. . .
the notation being slightly changed from that used in the text.
Ex.
2.
If
we have no
we apply
reason (so far) to anticipate a convergent series
find the series
in
the rule to square the series
-Q +
which the general term
is (
but we
i
1
l)"" ;^,
where
so that
Ces&ro (to whom this result is due) calls such serii-s xhnply ind>(> miniate
the degree of Indeterminacy being measured l>\ the number of means \vhioh
have to be taken before a definite value is obtained. (See Art. 122 below.
;
Mi l.THM.K LTIOH OF SEE]
35
34,
li.nfu
LI),
logn)='
9
liin "
lini
w-f
!>gn)
1
,,'-i)-;r7T
theorem,
-S('+4
10-1
this
with
agrees
in^lx-ini's
l'i
general
th.-c.r.-i:
[MEx.
Hut
3.
if
-qnare the
w.-
1900.]
s.
î-I^....,
"litain
l'(
)"
'". where
-r[2(/l
if
'.at
[./'(
-.r)]"
">/''
and
it'
iniilri
/>
b, th<> .<'r/f.<
ml.-
?J{
de for
hand,
if
>
/'
A,
I'l-in^h-
M.-rt.-n-
35.
lia^
'h<it
.! /;.
Suppose that
|.i..\,-d
tit.
and thru writr
t,
/.
r^^-f-r.-f
,
4- />+...
(/
...-f-,-...
...
/?=/
it
('.'nsr.juently
tliat
one
.-|a|,
M
-(~ 0"~ ljr n
(>n th- I'tln-r

tlu-
= </,/* + /,/;..
= (//_r,,)+
/;
#
so that
//
,-f...-
,(^-r,)
.-i
|,|i
In.
/;
f->r In-
/!,
,/
.1
t'
-f-.-.-fc
...,-i-H. ..4-
DOUBLE
86
Now,
can find
a limit
so that
>
it
/o
[CH. v.
B is convergent, the sequence p 1? p2 Pa* tends

has therefore a finite upper limit H. Also we
since the aeries
to zero as
SERIES.
m pm+ i,
/o
m+2
are
...
all less
than
thus,
if
p=n
m,
we have
Take the
limit of the last inequality as
because
so that
we must have
tends to
with
oo
HnT|fln
that
Now
n.
= 0,
lim
Note
(6), p.
and we
and
find
is fixed,
5),
= 0.
B-Cn = 0,
MmCn =AB.
Hence
lirn(A n
or
must
/f n
oo
arbitrarily small
is
(see
is,
It
n tends to
be supposed that the condition of Mertens
not, however,
is
necessary for the convergence of 2cw ; in fact Pringsheim has established

a large number of results on the multiplication of two series neither of
The simplest
which converges absolutely.

of interest) is as follows
of these (including
most cases
7=2(
product
is
If
their
)**"%,
n~ l
F=2(
l)
vn are convergent in virtue
-!)""%,
given by 2(
of the conditions
where
wn = u v n + u2 vn _ + ...+ u nv
1
provided that the series 2(wn y n ) is convergent.

Since lim^w = 0, we can write
2
where
Now, by
7= Ul - (u 2 - UJ + (u 3 - u 2 ) - (HI -
8l
=u
S2
l ,
hypothesis, 8 2
= w2 ~ w n
83 ,
6"
...
^3
are
= W3~ W25
all
negative, and accordingly
convergent, and therefore the series 8 1 -8 2 + 8 3 B 4 + ...

we
can
therefore
Mertens'
and
theorem
absolutely convergent
apply
obtain
Hence, 2|8|
is
is
where
Th UH
7i
= Vi = î
= &ii>n + 82*^-
2 U F= w - (w2 l
?j) -f (w3
- w2) ~ ( W4 ~
and therefore we have

/'!'-=
provided that lim
^',,
= 0.
Wj
- w, +
<r
:,
\v,
M:I:
35,36]
that
ii
limy,,
<>,
i.
Mi
iir*-ut
(in
virtu.-
\\-
n.-.-d
..|,1\
that
..f
M.-rteiw' theorem),
for
'
i
and
that
>iiji|.o--
i>
vx
index
:i".v
th:in/>; as
less
ha\e
\M-
--cause
in
n-v^r
7+1
ti.
1
lolll
to
have
\\e
,/,
,',.
+
Thus we
as
is
"/!-$)
find
If,
f,
8iippc-
tind 7 so
is
tl.
leas
than Ac; 7 having been fixed, take the limit of the last inequality as n
tend.- to infinity, and we find lini
Thus, since t is ail'itraiily
"_.
-0, and it follows that lim "-<.

other n-sults are due to Voss and C'ajoii, in addition to those found
of 1'ringsheim's aiti'le in the
K"i
iirjsheim.
small,
lini
'/
'i
two
I.;
vol.
/,
36.
Ikl.
nii-re
i-.-.-mt
IJHM,
-2,
]>;i])ers
and
-J.-)
j,j,.
will
found
l.e
in
4<l.
Substitution of a power-series in another power-series.

operation -ivr^ anotlicr t-xaniplc of deranging a double
and
-r
4-..
oaider tin- aeries
ThU
t**f(
,j
l,
t}
l>
i'
thru
Beries
arises
in
\vln-tluT
tlif
if
-ijual
rule
tin-
to
///
./'
tin-
at
convergent
<
9&y (see Art.

of û >^t
ivMilt
all.
th-y
Tin- <|U-ti..n
50)
nt n -
th-
that
for
thr
powers
of
//
lr
i-an
and tln-n
multiplication of series,
fOWS of the doulde s
tin-
l>;i
"
:
from
calculat.-d
-f-M
-f...
4-..
BeOOnd
ami arranîno; in jo\vrrs of X is ev.

It
valuBO, t'>r what
apj-.-ar-
8WM
+ "/'
+ "A
con\.
tir^l
usino;
if
<N.
//
and
gent,
4- ...:
/'.,./-
al.solut.-ly for
/._,//
:;
4-...
DOUBLE
88
double series
If this
we
are
equal
made
summing
SERIES.
arranged according to powers of
is
by columns
it
[CH. V.
these
x,
two sums are certainly
still converges, after every term

31
positive (Art.
(5) and Art. 33).
the double series
if
Write
|a|
and then the new
= an
series is
\b n
=/3 n
x\
= g,
not greater than
Now
summed by
this series,
is
--(2)
rows, gives
which converges, provided that

Take now any positive number less than r, say p, then
the series 2/3 m /o m is convergent, and consequently the terms
m have a finite
Thus our condition is
/3m p
upper limit M.
satisfied if
or
if
Hence
if
<s, and
/3
g<( s p
)p
/(M+s-/3o), the
series
2)
of positive terms will converge.

Consequently the derangement
of the series (1) will not alter its sum.
Thus the trans-
formation
permissible if the two conditions
is
(i)
|6
|<,
(ii)
|*|<(8-|6
|) /
,/(af+-|6
|)
are satisfied (where p<^r,

b n p n ^M).
In particular,
= 0, the conditions may be replaced by the one
if
>
n
// the series
converges for all values of y it is
evident that the condition x
is sufficient to justify fit r
y
derangement.
The
case 6
is
of special interest in practice;
the coefficient of x n in the final series

series,
is
not
itself
and then
an infinite
but terminates; a few of the coefficients are
and generally,
if
n>
2,
cn will
contain the terms
36, 37
N"\ LBSOL1
Ex.
.1
Ti.
iation
tin-
verges for
a
in
polynomial
Lining that
//,
th-
caeOfi
t.-nninat.-s
neiiefl
cH
II.
we
"
ifl
2,
...,
;|
ami
/'-I, l--ans- in
'
Non-absolutely convergent double series.
37.
AliiK.st
dily Lrnirral tyj ..... f su.-h
tlic
r.-mitly.
tiv.-ly
in
1,
degree
i-.-a-hinj
l.rf<.i-'-
the l>in'inial Bttiei (Ait-.
'il.tain
hi_;li'--t
+.r) (see
= 0,
/x
of
fj.(fj.
/'
an.l
t>/
term
tin-
= /zlog(l
vanish for
A ill
that
sn<-h
= e* and
'>'-',,
all'.waMf, pn.vid'-d that |#|<1, since z con- !

I/1.
Tin- i- -ult is .lvi..M-l\
./;", where
is
vain-
all
hyH
y=i"
tlu-n
<
Art*
Hardy;*
it
Th.- th.-.-n-ni
ia
l.y
i:. L ".
series
"//
//// '/.<
tin-
t..
th- aztenrioD
l..--n
iriv.-n.
(/<>
.<//;/i
niii.-h..
f
?
m>
.,,ni|,aiM-
discussed
t\|i-
'
^, n
''
has
>.-ri.^
n.n-,-sji,,n,ls
/"///"'/- '"^//
'<<
'
'
con>'^
.-'.*!.,
;<
'
<,
just
as in
th> proof of Abel's leninia (Art.
tin-
Lrivt'ii
i-onditions
win re
//
f"r
//
rithiM-
or
//
LB
I.
//
-''.
1
S2-Z
11112
N'that
(V
whit h .an
lM ad<-
!.,
AII apjilii-atinn
iv-n
o,
./
2-1-2
as small as w.-
is
and otherwise
A
//
.v
.i/
2-f2
i-
and
th.-
Iiy
'y
'-.',
v V'v;
and P^ | both tend to
\vhi-rr
i*:'.
\vc
- | fci-fc*-l-*-H'H
11
it'
''
''"
that
^^*
an upprr limit to
is
or n
ta.-t.
22<Vt>
s..
as either
'''"'^ ô -2^0
iiinh-i
In
/-
,,+i
,.
'/*'//
that
|>l,>as.-
>..
+, V(1 ),
th'
l.y
]I-MJ.T
.-h'-i.-.-
.f
//.
r,
because
diMilil.-
''ges.
wc/)),
<,..
an
p
'.
.l/'/'/K
SOC., srri.-s
_'.
vol.
1.
UNI:?.
},.
I-.M
roL
J.
l'.U.
]..
I'.MI.
DOUBLE
90
For then
SERIES.
[CH.
-.'
\s mttl
and
\<4\cosec$0cosec$<f>\
while lim v m>n =0.
EXAMPLES.
1.
As examples
(1) *m
in
equal to
take the following
here the double and repeated limits exist and are
I--;
all
0.
(2) s m n
,
single
we
of double sequences
m n
=(-l) +
here the double limit
(!+!);
\m n)
and repeated
do not
limits
lim /lim
m
(3) Sm,w
(4) s m n
,
= wm/(w 3 +
= m/(m + n)
>-oo
\n
-ac
again
0,
but the
although we have
exist,
= =
s m<n \
is
lim / lim sm
n >-3o \ m wo
here the double and repeated limits are again all 0.

here the double limit does not exist, but we have
<*,<!, and limsm n =0, lim *,. = !,
tt
because, however large p may be, we can find values of m, n greater than
such that s m<n
and other values of m, n for which * TO ,n>l e. But
/x,
<
the repeated limits exist and are such that

lim
lim
MO \n
>cr>
s m> ,A
J
=0,
lim
n
>-oo
lim
\n
s m<n \
= 1.
>
Similar features present themselves in the sequences

sm .n = mn/(m* + n?) and s m n = l/[1 +(m-n)*].
m 2 3
s
G
=
sm n
we have
(-l) m n /(m + n
,
(5) If
),
lim /limsw
(m)
bat yet
us
may
\=0=lim /lims m
J
(n)
limsmtn =-ao,
m=
be seen by taking
ri
lim sm n =
,
Here
it
n\
\ (m)
()
+ oc
should be noticed that the
limit of the single sequence given by putting w- = n exists and is equal.

;
although the double limit does not exist.
[PRINGSHEIM.]
to
The double
2.
series given
by
+ bn + (, - 6 ) + 2 + f a4 -f ...
(-a +!) )+(-a -!> )-a 2 -a3 -a 4 -...
b.
+0 + 0+0+...
2
+0 + 0+0+...
A,
(a
()
</.,
!>.,
/>.,
.îves
the
sum
in Pringlheim'f sense,
whatever may be the values
.f
But the sum by rows is only convergent if 2a,, converges; and

the sum by columns converges only if 17>,, is convergent. The sum by
diagonal is lim (n + />), if this limit exists; and is otherwise oscillatory.
",
/>.
lt
IfFLES
v.|
In
3.
tin-
(1.
.ul. I.-
1+2+4+*-i-1-2-4-i-J-1-2-
neriea
-I
ci.luiiin
j
l'i
iii^.-ht-iin's
The
4.
t<>
converges
sum cannot
series
and
sum
tin-
-1+0+1+0 + 0+..
+ 0-1+0+1 +
+ + 0-1+0+..
..
ha> the
<>
sum
l.v
rows; -1
doul>
tluif
5.
'I
l.v
and
is
-1
if
given by
m>n,
or
sum -
tin-
ln-twet-n
-1
The
6.
<
is
divergent.
In
fact,
m<n.
if
-2 + 1+0 + + 0+...
+ 1-2 + 1+0 + +
1
scillates
!_"-.
Pringsheim's sense.
+1
0+1-2+1+0 +
+ 0+1-2 + +
+ + 0+1-2 +
has
li\.
diagonals
columns; u ly diagonals; and naturally
'iiverge in
H<
l.v
0+1+0+0+0+...
l.v
gran
row
Inn
o,
t-xi-t
..
..
..
..
and th- diagOIM]

ly r.\\s and columns
0.
There is no sum in I'rin^hfiin's
m=/>, and is ..tlu-rwise 0.
Imth
-2 and
if
2 + 0-1+0 + + 0+..
0+2 + 0-1+0 + 0+..
-1+0 + 2 + 0-1+0+..
dnul.l.
1+0+2+0-1 + ...
+ 0-1+0 + 2 + 0+..
0+0 + 0-1+0 + 2 +
..
ha>
tin-
the
sum
sum
l.v
l.v
K.US
diaLTonals
+ +0 + 0+ ... =2, and

is
2 r<> + + + ... = 2.
1
samr sum l.v -..lui.

Thus t/iee three sums are
tin-
<>
<
/we, since
I'l-nvt- that tinmultiplication rule
summiiiL: the dmil.
7.
l.y
",/',.
+ </,'
first
liy
ru\\s
and
secoiully l>y
fr
+ 'ij>.*i-+ ...
+...
</ w
= 2.
DOUBLE
92
8.
Discuss the following paradox
If
we sum the double
first
SERIES.
[CH.
1111
series of positive terras
by rows and secondly by columns, we obtain
+ 1=5
or
1=0, where
[J.
BERNOULLI.]
If the double series 22a m>n is

9.
convergent in Pringsheim's sense, it
does not follow (in contract to the case of single series) that a constant C
can be found such that sm
C for all values of m, n this is seen by
,
<
and supposing 2a,,, 26 n to be divergent.

In like manner we cannot infer s m> n < C from the convergence of the
sum by columns or by rows (see for instance Ex. 3).
considering the series of Ex.
2,
m+n
in which a min
(
/mn does not converge
l)
sums by rows, columns and diagonals are equal to
one another and to Pringsheim's sum. The common value is, in fact, (log 2) 2
m + nu v
Exactly similar results apply to the series in which a mtn = (-l)
m n
where the sequences (u m \ (vn ) steadily decrease to zero.
10.
The double
absolutely
series
but yet
its
11.
Consider the double series in which
and
am
Here we
= 0.
2 2 = -|2<
IS = + P(W.
find
<)
12.
[M^P-
[HARDY.]
Prove that
V
m?-, n?
tends to
/.<!'
when
v tends to oo
pnvilnl
tliat
all
are omitted from the summation.

13.
and
then
m-
If
am
= 2~ m
Yi *)--
m=0 \i-0
trrms for which m = n

Moth. Trip. 1895.]
= - 2~",
If
w=0
m=0
= 0,
!.\ \.MTi
V.|
14.
It
.u.
,(-ir*"m/(>wm
1ml
-inn
tin-
sum
ti,,-
[F.-i
and
oscillates l.rt\\.-rn
,'rt
'rove that the product
<>f
two
the
.r
to
"
2
(3!) .6!
It
<
,. V
)=i+ y (.,.+
.i/^) + ^
1
and
)4....
iy:<i,
J?)-^*"*/(*/^^)-25ffi
th.-n
and Hardy, Proc.
series
x*
16.
and
17.-,.]
j,.
'
,il
'
Iin.inwi.-h
M.I,
and
+ jV
I!
15.
-x
lirtwi-rn
illatt->
i.
v=-i,
.? ).5r(y ,9
-Axing)-
2
),
.</)-."('.
Verify that
17.
L!(n-l)!+J
!':</'
-2).
K*-)!
./-- 2
4?+8
use Arts. 33, 57 to infer I'rym's identity,
1.1
'
./
x(x 4-
!)".'(.' +!)(.'
8h6H
18.
that
_.
+ 2)
+< "
Lr
1!
.r+1
3!
2! .r-f-2
_._
....
j
into
;i
dulle
M-rirs.
and transform
it
..
.
,
Taki- /r
19.
lit
Ckm^
and 8O
calcnlatt-
I'l
/,-
t> 7 d.-rimal
places.
_. + _^
into a dnul'le Beriee, and drdu.r that
it
i>
c<jual
t<>
[STII.
DOUBLE
94
Shew
20.
that
(if
and deduce that
SERIES.
|#|<1) Lambert's
X1
3?
X
r^"*"!^? 4"!-
[CH. V.
series
this series is equal to Clausen's series,
l+x
Hence evaluate Lambert's
Shew
the coefficient of
xn being the number
From Ex.
21.
series to five decimal places, for

is also equal to
that each of these series
19, or directly,
3?
3?
of divisors of
prove that,
'^l+^6 ^l+^ 10 ^"*~l-^2
Shew
that, if
(1
and n
included).
x <1,
&
l-j^^l-^ 10
~\+x
22.
&
if
+a
\x\<l]
2x*
3.r*
(l+xf
"*
between
[For the connexion

functions, see Jacobi,
If \x\
23.
where
that,
<^>
if
series
Exs.
in
19-22
and
elliptic
n).
Deduce
40.]
<1, shew that
the
.is
<^_ 1
the
Fundamenta Nova,
sum
= 0=<#>
n (including
of the divisors of
and
[Math. Trip. 1899.]

24.
If \x
<1, prove
that
where 0(w) denotes the sum

1 and n).
X/(rf)
for
all
'
the divisors of n (including

6
In
[LAGUEHKK.
25.
Shew
that in the special series
converge to the saim- sum as tinmay oscillate, for instance u = /? =

series
<>f
(loulili-
sfiies;
but the diagonal series
= <=7r, p = l,
_ 3 + _ + ....
l,
Art. 37, the repeated series also
gives for the diagonal

[H,\i:i.v.]
CHAPTER
VI.
INTIMTK PRODU(
38.
Weierstrass
In
this artirlf the nunilx-rs
and
])(>N]'ti\c
inequalities.
than
leee
(!+",)(
..
ej,
an- suppn
this In-in^r the case,
we
+ ",)= f(ol -fo1)+olat > +(",+

1
Eeuoe
(
+0,
+",>(
)(
ami
(-(.ntiiiuin-
(l)
(1 4-0^(1
In
+('/! +",)](
this proc,
>
+";;>
+<"!
+ ",+
tliat
see
H-a^lât)...(l+aw)>l-h(a1 -fa1 +a1 -f... 4
nianinT
liU(
+" >[
ha\.-
\vc
-OiX -",)= -("

1
Thus, since
i^
Oj
-h",,)
|<sitiv-.
+ " ",>l
-('/!
\\r
liave
(l-al)(l-o^(l-at)>[l-(al 4-a1)](l-at)>l-(al -|-a1 4-fl

ami so
\\v
nt-rally,
>>\-> +, +
l
tip;
/,
aid nf c2).
(
siiu'i
it
tin-
1+,/jH
...-
and
,+
+</.,+
1.
\v-
have,
>y
r.-Milt
1
_+Oj+...would tluMi
+ ./.)...< 4-OnXP -(",+">-...^".)]1
_!t-utrr tluiii
1.
tin.-
iiu-|U,ilit
1
.
INFINITE PRODUCTS.
96
Similarly,
we
[CH. VI.
find
(l_a )(l-a2 )...(l-an )<[l+(a 1 + a 2 +... + a n )]By combining these four inequalities, we find the results
1
(4)
(l-2a)-
(5)
(l+2a)-
(6)
where
a denote numbers between
all the letters
2a
is less
39. If
a lt a z
that
than
1.
a3
...
>n(l+a)>l + 2a,
>n(l-a)>l-2a,
and
are numbers between
1,
and
1,
such
the
convergence of the series 2a n is necessary and sufficient for

the convergence of the products
n Qn to positive limits
P,
as
n increases
Pn
For clearly
Now,
if
2an
that
3.
Then by the
is
to co,
increases as
increases,
we can
convergent,
and Q n decreases.
find a
= am+i + an+s + am+s + ...
to oo
number
and
...
+aB ) ^ P m
we have
'
-a
Hence, provided that
is
greater than m,
Pn Q.(l-<r).
Thus, by Art.
such that
2,
PH and Q H approach
-P^Pm /(l-<r),
if
Sa M
is
divergent,
we can
definite finite limits P, Q,
QSQ.(l-<r).
find in so that
N may be.
no matter how large

Hence, by the same
inequalities,
P n >l + N, QW <1/(1+JV)
=
lim Q n = 0.
consequently lim P n
I
and
if
w >m,
should be observed that if a product tends to zero

^mi^, without any of its factors being zero, the product /*
It
such
<L
inequalities (5), (6) of the last article,
But,
where
POSH
38, 39|
with
rallrlism
Ex.
will
_'>,
1.
approach a
97
tli. -MI
tli.-
ili'-
I'i'xl
limit
because
thr |.i"diu-t
I)(M
+ 1)
+l
so
lllu '/
tll.it
Similarly
+^
\(
+--
<
)(
It
hough iU value
is
'ulaU-il so iradily.
Ex.
2.
+ *)(i -
(i
\\ill
(livt-rirf
,,
:-i)(i-S)0-i)...
M+l
liin
Ex.
3.
n+
3
=
T-' V "- ^3'4 l
-1
'
/'==
-f.
,
lim
MI
iuulfi tin- u'ivfii
V'.
l>,,>lb>0
if
= 0.
I.,
if. nil,
f,
timl
iiii(/,,
an
= 0.
K-h that
n^tn.
have
and
:.:-
D-
M-
lim
so that
is
It
lim
</
().
easy to 896 that the ur^mm-nt "f Ai
>rove that
<</' //
x,
"
= (!+.,,
+-,.(!+.
iiti-,l
INFINITE PRODUCTS.
98
[CH. VI.
For if, in any new arrangement (represented by accents), we

have to take p factors to include the first n factors of P and Q,
then
Now n
to
Pn
can be taken large enough to bring
and
Q, respectively, as
lim
r >oo
In like manner,
diverges to
0,
close
Q r = Q.
'
r ><*>
if
so does
and Q n as
please.
P/ = P,
lim
Consequently
we
diverges to
oo
so does P'
and
if
Q'.
taking logarithms we can see at once that the present

theorem is deducible from the theorem of Art. 26.
By
40.
Convergence of
the numbers
an
infinite
u lt uz UB) ... may have

may be supposed
than
product
both signs.
that they are

for there can only be a finite
loss of generality it
cally less
products in general.
infinite
It is quite possible that in
them greater than
But without
all
numeri-
number
of
(otherwise the product would certainly

or
and
the corresponding factors can be
oscillate),
diverge
omitted without affecting the convergence.
Now we
have*
M>-1.
is positive,
the lower limit of the numbers

1,
1+Ui,
+ u.
...,
l+u nt
...,
we have
fu
For
log(l
"
Hence,
but,
if
if
is
is
positive
negative
\
.'ii
,/..
)=Jo|
p~ -log[(l+tWi)(l+*Wi)---(l+*O]
can
mad-'
!>
how
bet
large W
is.
ij
!"-
//
-4-
..r
so
2u w
if
diverges to
kites if
h;i\.-
\vi//
to oo
/<*
'/"
i/ 2u,,
in*
small by properly choosing m, no

Tim- we have tin- theorem
arhitrarily
>
in'
>
4-
//
>
'
if
"
if
0> > -1,
i^>
poeitn
/
tliat
..+
...
//
!+",.,). ..(!+"
>-ln.i;
>
L
wluT'I
is
H-...-H4) A
the upper limit of
1,
(1 +%),(!+
9O fl"lf
jrli
is
not
+x)
//////'
1"
'
/'
"//'
///'
/'>,
d-h^xi + ^jd + r
to
d"
o.
vcd/UA
.nly cases not covrr-l ly

in wliich Su^1
diverges and î//
Tin-
It
is.
v
perhaps, a littl.Loth dix.TLiv
Converge; hut
Fr. Id
typ.\
MI-.
pOSJ
and
limit (in
its inaxiiiiiiiii
,i^
it
and
s.
-''/.-l
abso,
or has
si^-lit
\vh-n
that
which
of
Mth
th<
hy
,=!+/.,
//
J^'/
(),
the pi-odiu-t
(by
)
us page.
M not
tirM
at
).
the product
i'nrni
Then, provided that lim

and BO olvimisly C
if
-illatioii
>t'
diverge
-f
divergee
<.v
prrph-xin^
t
tenii^ aiv given
-J/'th
!'
th- jn-mlurt may n.-\-.-rth-l,this

juit
t-asy to const rurt a pi-nluct
s of
a COnvei
a
li\
is
^'-,
i-as.-
in.-thod are those
in--
tl;
.-itln-r
=-
Ar;
:'.'.'
i.
-f "
>-
'
isdivergeni
ll<
+Cm )
1"
Kurthi-r
and
will
INFINITE PRODUCTS.
LOO
[CH. VI.
2unz must also diverge.* This condition can be satisfied in

many ways; one simple method is to take c n = d^, and then
we must suppose 2dn ^Zdn2 both divergent, and 2c n3 convergent;
for instance, we may take d n = n~ p where %^.p>\.
The
u2n = nproduct is then given by u 2n . = n~i
,
Ex.
Since the series
1.
1111
are both convergent, the two infinite products
0+(l-i)(l+i)(l-*)... and (lIn fact the
also.
converge
first is
obviously equal to
and the second
to}.
Ex.
Since the series ^ + |- + ^ + -^+... diverges in virtue of

it follows that the infinite products
2.
Cesaro's theorem (Art. 22)
a+jxi+i)a-}xi+ixi+#i-..*^
are both divergent. In fact the first diverges to
for they are equivalent to the products
Ex.
oo
and the second
to
Since the series
3.
J__JL _,_!__!,
is
convergent,
V2 v/3
but ^ + ^ + 4 + 5 + --.
v/4
is
V&
divergent,
it
is
clear that the
two
products
and
hoth diverge to the value
In fact
1+
j_)[,
<>.
-^3-1 -^^^[i-^-.j,
is always less than 1, and can be put in the form

Further lirn(waw ) = l, so that our two products diverge to
by
so that this product
(l-a
w ).
Art. 39.
Ex.
4.
diverges
If M n = (-l)"i, it
thus the product
must diverge
*If
to 0,
Su n2 were
the product.
is
evident that
which may be
verified
convergent, the divergence of
1>\
2?*,,
oscillates,
while
2?/,,
inspection.
2n n would imply the divergence of
40, 41
&B8OL1 TE
"N
<
101
Absolute convergence of an infinite product.
41.
<//s,,/,//,/
y convert ! ui,
tl>
II*
/"
,-./>
Writ.last
then -"
//// /'//
pendent of
converges; and, as explained
<
\v can suppose a w
articl.-.
tlso
WMW
'>T8.
"'<",.
that
BO
convergent (Art. 8), thai
--"ii
in virtin-
that
Su|ij...v,. n.-xt
a tli-!vni
>f
861166 Stt,
tin-
...
(14-14
'
I'
Up/Vn
it
B.-(i+6i)a+^)...(i+W.
(>
//
>
BO that t/p contains th-
made
Â p 'B
\Up/Vn -l
p.
.sit:
-l,
Fn ^//
that
l^-^l^.l,
N"\v. as rxplaim-.l in Art.

s,M|u-ntly
/',,
can
lim
/^,,
lim
=P
.-!
say.
>
if
//
\UV -V <
if
//>
E7^
UU
//
li
/,
Hn,
is
and
&n>i
V=U.
'hiite
TIu-
in ay
l>y
IllX'MIIH'nt
little
be
us.
'.I
I..
r-t.tl)lish
change to shew that
made
to converge to
altering the ..nl.T
-f
tin-
th.-i
<
Iff-r.Ke,
luit
Con-
-/y,,<
approaches a limit ^7 as y trnds t<> y

//>//> /^ Wr ha\V
9
is the
Thus, if
greater .if n and /^,. \v.Hut
=A
>A9 >Bn >A-ie lfjp>>

î p
and
:>!>.
-/;
BO -T--at that
t'..uinl
!>
ii'-i-
that
whole
nmtly
p contains the whole of B n ); on
is
CMiitains every
}
vi.l.-m that A p /B n
hut with the si^ns
\.
H.-nce
if
lr-vity
'
and
s.i
l-ranged SO as to b"
.4
(and
multiplying out
l.Tiu in
last arti<
'
chosen
j)
Zt/,,
-MO
11(1
r.-a+vâ+tg-O
'...0
Then suppase
i-
writ.- fur
IK!
in tin-
proved
II-
verges; and
.ind th-ivf(.nit
i:
Uii'lii;iMir-
nn-al>solutely
any
fi.-tors.
product by
tie
..
to
averment
ititinit>
jmdiverge, or to oscillate,
INFINITE PRODUCTS.
102
[CH. vi.
Perhaps the case of chief interest is that afforded by the

n~1
is positive and
a M ], where
\im(na n )=g.
infinite
product
Suppose that
the value of the product is P when the positive and negative terms occur
when the limit of the ratio of the
and let its value be
alternately
number of positive to the number of negative terms is k.
II[l+(-l)
X/P= nKm (1 +,n+1 )(l + a an+8 ) ... (1 +

\im(v/n) = k.
Then
2v-i),
-oo
where
Now
it
is
it is
article,
plain that
clear that
2a M 2
convergent, and therefore, by the last
is
can be made arbitrarily small by taking n large enough.

Pringsheim's method (Art. 28), it is clear that
lim (
2w+1
+ a2n+3 +
Further, by
+ a-i) = \g log k,
and therefore
k9
or
X/P=Jk
It
is
The Gamma-product.
42.
evident from the foregoing articles (39, 40) that the product
is
=
divergent except for # 0.
But we have
I
.
so that the expression
increases with n.
Also,
where A
1, if
as.
in Art. 40,
is
positive, or
we have
[Art11
Art.
2,
is
either
Sn approaches
Now
(Art. 11)
a definite limit
l+#,
negative.
as n increases to oo
if
is
(1)] >
Hence, by
l+||4-...+-logJi
approaches a definite limit
(7,
and therefore
P =
x log n log n log (n*IPn \

Now
so that n*IPn has also a definite limit; this limit
Gauss's notation.
is
denoted by II (.r)
n *-"
THUS
which, again, can be written in Weierstrass's form
= e cx - s =e cx
lim e~ Sn = e
H
When x
is
*-ao
'
U
r=l
a positive integer, Gauss's form gives
(l+)(2 +
)..
.(*+)
II (.r) =.<!,
n/\
because
in
G IMMA-PEODl OT,
41, 42]
Although we have
convenient
"t necetiftai
Mill
prod
an intrjjer.*
It
rasy to
i-
if
Converge
\.-rifv
rentrict (1
and
.>
it
to be positive,
is
integration ly parts that
liy
-f .f)
easy to see th,r

has any n-u'ativr valm- wh
nvergem-e
Kuler's integral
rd
has the property
,.f
,.
1,,-in^
tin-
mtii-ipatr
j,,
t<
./
when x
is
an integer; and we
(|uati.!i
r(i+./-)^n.
It
be proved to be correct
will
\\lii.-h
ebl
\\i-
-f
in
Art
in
definition
IK-
<>f
I7fi
!!(.'),
of
bhe
wr
tind
'
w*"^
"
+
",!'-
= (.r+l)II
4-l)
It
and
f,,ll,,ws
1=
I'.y
is
hi/
(I"
iMjuation in the
last
the notation t-xjilained in Art. 3.

reversing the foregoing argument
II
we
see that
form
tfo function
-4-l)
= (r4-l)Il(.'-)
II(w+./-)ru/<'i;
<
AMPLES.
Discuss the convergent
/"(/()
2.
,.
hni
often convenient to write th-
1.
+*)/!!(*),
f.iiM.-juentlv tin- definition leads to the e<juatim
or
It
(1+*X* +*)...(*+*)- H(
that
is
a polynomial
Piovt- that
absolutely
integer
II
f,.i-
[(
in
/*.
(w
any valur
ami that
of
/
if
./
<
2, 3, ...)
provided th;r
.r,
\i-i
^"VfTflJ
j converm-nt
= l,
1.
Tlu- ronvfi'iifiu't- j)Tsists also lur i-oinpU-x
values of
(set-
I'.
INFINITE PEODUCTS.
104
*
then
Wn ~
if
Hwn
If 7n
diverges to
= l, Hwn
4.
m>1
if
diverges to
a = c.
if
[CH.
if
to
oo
m<l.
if
a<c, and
to
oo
if
a>c; and
converges
[STIRLING.]
If
M 2 = 0,
(),
= P>k
u 2n -i=-n~p u 2n = n~p + n~ 2p
are both divergent, but 11(1
,
then 2w n 2wn
$
Verify that the same is true
2
where
+u n )
is
n>l
and
convergent.
if
Ma-i = - W--P, w 2n = w~* +
?i~ 2p
+ w- 3^,
(n
>1
).
[Jfo^. rnjo. 1906.]

5.
Verify the identity
-~~
~~
.,of
Shew that as n tends to infinity, the product

x except 0, but the series converges, provided
6.
it
7.
Prove that (1+^)(1+^ 2 )(1+^4 )(1+^8 )..
.*.*,
-n-
and that
8.
tan|
Shew
frm.
The
10.
+i
cos -^
tan
cos
x
3
...
= !/(! -x\
Gamma
-^-,
...
= i-cot^?.
[EULER.]
F(l+c) and T(l+x + c).
that
product
Prove
is
ir(^+2w) r(w) n
f f^TT(x + n) F(27i) J
that, if
is
an integer,
lim
verify that
and that
Shew
wa M -(w + l)a ll+ i = to + (<- l)a,, +1 w6 M _, -7ia ( + l)6 n = < 2 [(2- l)a,, +1 -f n6J.
ll
(w
l)
that limtta n = 0, Iimw 3 6 n = 0,
Hence prove that
6,,
if
= 2 ('It - 1 )
values
= sn x
J + gi tan^ +
functions
all
\x\<\.
if
Determine the value of
in terms of the
9.
cos -
Verify that
diverges for
that #>0.
^>^, and deduce
or,,
3/*.
that
LMPLE8
vi.
I
12.
tin-
L05
hree hyperge>i
'
êneral
erief
fry,
111
y>u +
\vlii.-h
/'<-
i'>
i),
/;f.
(y
i-
)/.,
(/m,,)=0.
Deduceth.u
/;-( y
_/^r, (y- tt )(.l -/*)=.,
J= (y-)(y-
and that
K\.
Kr.-in
--_
1)
:)
and
'-fty +
'
that
-)(</-&) = /:?'/-, + ("

liin
13.
-l,Ar.
Then prove
slu-\v
that
'/^7 +
i-
t-xprrssimi tt-nls t> the limit
last
tin-
(X
r(y-a-/J)-
as H-*-o(
'-
duce
.14.
,/= IK
IK.
if
-,/->,
?1
+v),
y;i
!-!^F=n(i+ n
?
=n
-,/-
(i
the four iii-udiirU air als..lut-ly (-..nverp-nt

9ctf s
= II(l
y,Y.'/ 3
15.
.
It
1"
.-iivri ur,. Ilt
i>
f,,,.
c..nvei-.;,. Ml
al>s<liitfly
;u ,y
1.
=l-
convergent,
vain.- of
./
and
tht>
it
]i<>duct
can
)....
11(14-.-
expanded
:t
+ r ./-+r^+...,
ui.c,,.
an
r,
aU,, that
11(1 +.*//)(! +/'
r)=
'+
r i(-''+
K.16.
I.
iiavc at
.nee (!--/
and
the
l.y
la>t
example
...
Thus we find
</+:
which give
=
-^)(i-
<i-f)<i-y',
and
('
u'enerally
/'
/'
[K
.lately
in
.,
Shew
<
-7^(1 -7-')...] = (l+y)(l+v-)(l+y
[<1- 7 )(1
-y"), yi7 2 =n(l-i-
and
Thu>
if
al>s<>lutel\
INFINITE PRODUCTS.
106
17.
we
[CH.
If
see,
from Ex.
we can write
15, that
But qxF(q*x) = F(x), and
find
F2 = V^, F3 =
yielding
Thus
we
so
^)=F [l+ ? Cr+l/^) +
To determine
we find
F we may
2
pn 7 -l<
Thus
use the results of Exs. 15, 16, from which
F" =
Pn+r/Pn >qQ, P >q

< q2n/q (l - q
I<?I<1 and
because
Pn V
Hence
-1
Thus, using the notation of Ex.

/(#, q)
"),
lim(Pn F ) = l, or
so that
= U(l -
where g
+...,
<Io
<Io
16,
= l/?o-
Ch. V.,
we have
= q U(l+ f-\v}(l + q^lx\
from which a number of interesting results follow.

18.
From
Ex. 17
/(I,
we
with the notation of Ex.
find,
= <M 22
?)
/(-I, ?)=<M 3
Or, writing these equations at length,
where the indices in the third
/(?, y)
we have
+<?/#) as
<?)/(!
so that
17,
tfo/?!
f(Jq,
so that
[Compare Art.
= 1 - 2? 2 4- 2? 8
2?i 2 .
n
Jq) = 2H(l-q )
11(1
W?3
the indices being of the form $n(w+l).

Similarly,
/(-<?*,
or
46.]
we get
the indices being of the form
Also
n(n+l).
# approaches -g, we have
the indices being the same as in the third series.
Again, from Ex.
14,
series are of the type
Again, by taking the limit of /(a?,
19.
[.TACOBI.]
^)=n(i- ?
no -f- w -9 '- )=n(i -r/)

l
").
M3 = l-(? + 9 )+(7
the indices being alternately
[GAUSS.]
12
16
+9 )-(9 + 9 )+-->
tt(3?il).
[EULEK.]
LMFLEa
vi.]
ml
Writ-
20.
-that
Thus.
<>n
in
j.la-
!!
iii
N 'y
tli.-
ire
'.
|>'it
Thru
107
-*),
2
:i(i-9 -',
we
niultiplirati.'ii,
tind
<.//
/>]-'
Wi*
ami
si
we have the identity
'/,'!
In
/'I"'.
we
if
particular.
\\in.-
\v-
I.
find tin- intf-n-stin.-; ivsiilt

8
7i
whii-h leads aâin
the identity
t<
(l+iy +
V + *V+---)
-(l -2y
+ 2y 4
:.'</'
"+.-.)
\\!i-iT
21.
the
sei'ir>
,,,,..
if
-..ii
verge
the product
In
if
tin-
an-
tin-
in
\-]\.
!s
-.-<-.
= / and
in
-ivt-n
l\/
4
,
,-,1,,
[.TACOBI.]
:;;;>U,e
product,,
thcs- d.nditiona are
\\"li-n
i!/>.
-K..)
8ati>ti.-.l,
-.\|
form
prove thai
parlii-ular.
(/t+a + 6)
= r(l+a)r(l4-6)
*?(+)(.
22.
I'lovr that
Tak- the trims
23.
in
.l-nntes
Thru wr
24.
is
It
\
In
if
and use the
pairs
,.ij
easy to drdiu-r
di-r
that
fi-.-m
I'./
the pr.-diK't
..n
th-
and
(>-,
_e.
n]
of
thr..i-y
!<
also di\rii:r
pi..ilu.-t> will lie
found
infinite
OODVWge
Il.<
II
..nst
Otlirr o\ani]tlr>
can write
'
||
hi),
example.]
m !,,_] __J-- s-1--... --?-Y

T 1 + r 2-fjw+jr/
,,_,V
tind
11.
r(#)/T.'),
last
at
..
01
..
f
product* Abel's
diverge togct
wilidl d:\n-gea to
re a m >0.]
the emls of Chapters
CHAPTER
VII.
SERIES OF VARIABLE TERMS.

Uniform convergence of a sequence.
may happen that the terms of a sequence depend on some
variable x in addition to the index n; and this is indicated
by using the notation Sn (x). We assume that the sequence is
convergent for all values of x within a certain interval (a, 6),
and then the limit
ii m 8n (x)
43.
It
n->ao
defines a certain function of x, say F(x), in the interval (a, b).

The condition of convergence (Art. 1) implies that, given an
arbitrarily small positive
such that
8
|
number
_F
Obviously the definition of
make
it
precise
by
<
m
agreeing to
e,
we can determine &n

if 71
e>
is
> m.
not yet precise, but
always
integer
we can
select the least integer
the prescribed inequality. When this is done,

it is natural to expect that the value of in will depend on x,
and so we are led to consider a new function ra(#), which
which
in
satisfies
e and on the nature of the sequence.

note incidentally that, regarded as a function of e, in(x)
monotonic since (for any assigned value of x) it cannot
depends on
We
is
decrease as
Ex.
1.
If
diminishes.
Sn (x) = \j(x + n\ where .rÔ, we have
= 0.
F(x)= nlim,,(.r)
+ao
Then the condition
so that
or
of convergence gives
*+*>!/>
= the integral part of (l/c)-.r, when .<!/,
m(x) = 0, when xzZlfc.
m(.r)
[FORM
43
1
Ex.
>.
It
2.
<
'
v*
<
'
>
1.
MI
kb
Ex.
It
3.
values
all
,,f
tak--
-rÔ, we have
uli.-re
-n
It
nm-t
\v,-
R,
ami
easily >-rn tint
is
il
part
"I
(>,
'=0.
and
Ex.
d+n** 1
[f
4.
j?
),
/'(./)-
bt-i
liui
..f
if0,
(/.
-//;
the
alway-
i^
>
in
1-
consideratioD
Bquenoes, which
s.-ijUcl.
98
'
me
ln-i
and
it
n luces a
u].j.-r limit
lixt-il
th- interval (a,

Ex,
ami
ni
sin
OQD
:i"t
/x
we can
th.-n
t,
in
may vary with
e.
6,
canmH t-ml
tak--
t> intinii
Uniterm t'..r
Hut ii.
e.
tin-
all
}>-
/<(e)=l
nnil'nrm in an interval
uniform
i>
/>).
si.
,-
tl>
variea tV.m a to
iii\
Thus, a-
in
it
Thix
>.
cnn\.
//"'/
1..
values of
./
uniformly in
/').
J
any point
Thus in
<
///(
ami
'
vay that
/'<"/
nni^t
distinction.
snl.tl.-
\V- ^hall
ink
,!'
as .f-*l
/:
that
sul<li\ ision
i'urth.-r
will pr-
l.nt
+(1 - 4
the functi'in
Kx. 1
in
lut in K\.
lining
flu-
[1
= 0.
ltlu'ii.
UK. iv
tin-
It
!..<
-0.
.'
-to Exs.
Tin- .i.nditiMii
ire
ted,
>'
interval
-aching
(0.
o)t
tin-
u]>
it'
'><'<!.
110
[CH. VII.
Hence in Ex. 2, x = l cannot be included in any interval of

uniform convergence such a point will be called a point <>f
:
non-uniform convergence. Similarly in Exs. 3, 4 the point

x=
must be excluded to ensure uniform convergence.
This distinction may be made more tangible by means of a
graphical method suggested by Osgood.* The curves y = Sn(x)
are drawn for a succession of values of n in the same diagram
;
done in Figs. 11-14 for the sequences of Exs. 1-4. Then,

Sn(x)-*F(x) uniformly in the interval (a, 6), the whole of
this is
if
F(f)=f
Ffo)=o
FIG. 14.
FIG. 13.
the curves for which
by y = F(x) + e.
?,>//()
will lie in the strip
bounded
glance at Fig. 11 will shew that this does

occur in Ex. 1. But in Ex. 2, as we see from Fig. 12, every
curve y = Sn (x) finally rises above 2/ = e; and the larger n is
I
thus x
taken, the nearer to x = 1 is the point of crossing
;
In the same way,

a point of non-uniform convergence.
=
()
is
of
non-uniform cona
14
shew
#
that
Ki-^s. 13,
point
is
vergence for each of the sequences in Exs.

i
'-.tin
of the American Math. Society
(_>),
vol.
:\,
3,
4.
1S'.)7,
p.
:,!).
43
OOin
\ll'o|;M
In
due.-
d.-linition
Involve the actual
in>t
of
niiinatic.il
l.-t.
uniform G
..i'
to
follov.
tin-
C0ll\
ordei
Ki:
An
of
that
>rm con/respond n>>!
number
tile
e.
and
independent of x,
#m-h
/>
fiil am
It
and for
///.
seen on coni|>ari.u with Art.

h- iinl.-jM-n<lcnt >f
///
is t<
will l>e
..ii<litinn
.i'
/A"//
vafa
that
jx
M'<|Ui-nc- an- functions of

That tlic cnmlitimi
to
F(x\
''//-
\Sn
the
II. -nc.-
an<1
convrr-v
is
al-
****
limit.
lim
we bave
|
for
if
\\
that
closed:
*/<.</'.
that .r =
<i.
it
is.
is
it
leads to
satMi.-d.
virtue of
in
ê,
if
[Art.
>
//'.
to
'i^tied.
if
converge uniformly
discontinuities oi
unifnnly
âason
found so that
A U.i.
tind
,(^)|<
a
si]
if
<<6 and
>
continuous function
90 that
l^'
and
where
je.
<J
-a<4
tf
may defend
on
//
are not
l#,
tin-
/(
will R!BO
>>
:.e).
IJL(
N,(r) t.-n-U
so that
N (.0=/
F(. O-N,,U)
noticed that
lie
if
<-r^vnc-
and BO the condition of uniform convergence

iould
\vh-r.-a-> tin- t.-nn>
,/,
n>m-l.
c >n\
\F(a-)-S n '.'-) <2e.
llenc-.-
tl-
that the nnly
')
''
///
points
id.-nt, for
///
uniform
for
some
t,,
\\-i-it.j
<H
>
i-oiMliti.ui
The condition
must
^- can
all
x.
tin-
>'
is
<,
')
,///
m, which
>n<!ex
//"'/
\\.
1!
f.-r
112
Sn (a)-Sm (a)\<e,
Hence
Sn (a)
so that
=
converges, and x a can therefore be included in
the interval of uniform convergence.

Ex. 5. The sequence
because lim*S'n (.r) =
and
interval
with
[CH. vn.
Sn (x)=x"(l
Sn (x)<.l/n,
Similarly for x
O^r-r^l, converges
#),
since the
maximum
b.
unifor/it!//,
S n (x)
of
in the
given by # = w/(w + l). The reader should contrast this result

should draw the curves y = S n (x] for a few values
2, and
is
Ex.
of n.
Uniform convergence of a
44.
series.
If, in Art. 43, we suppose the sequence to be derived from

a series of variable terms
Sn (x)=fo(x')+fl (x)+ -"+fn(x),
by writing
we
obtain the
test
for uniform convergence of a series in
(a, b) in the form:

It must be possible to find
an
interval
a number
independent of
x,
so as to satisfy the condition
\fn+i(x)+f+*(x)+...+fm+P (x)\<e,
at all points of the interval (a,
Each
series
43 can be used to construct a
by writing
but a more natural type of non-uniform convergence
*
Here we
x2
S n (x) =
find
F(x]
/X0) = 0.
is
a point of non-uniform convergence at
But, just as the general

replaced by narrower
convenient
replace
:
.r
= 0,
as
the reader
by considering the condition
will see
more
the following
and
There
is
x"-
H^ + (TT^
so that
tests
2, 3, ...
l,
b).
the examples given in Art.
of
where p =
the
test
in
tests
test
by
practice,
one
is
convergence
(compare Chap.
ordinary
above
for
of
so
the
here
usually
,nv
\\hicli
III.)
we
three
usually
following
43,
M'.KI. -
ND
NVI
44
i:t
Weierstrass's M-test for uniform convergence.
(1)
The majontv
to
proved
uniformly
de--rrih"d
has
,)
'!
i>r<>/>''rty
~M
series
tin-
and
-,,,/,,
.!/
Tin n
vergent.
t~<
if
XH/>]HM
2/n (a;)
is
interval (a
vn
f/
at
once
from
Art.
convergence, it is only
for any integral \alue <f p,
niher that
<///'/
series
fl"'
uniform
the
realisr
independent of x:
con\ci-ênce f(llows
ahsolute
Tlie
.!/-
fl"
fa
'
'irf constant,
?>
means
as the
l.rietly
Mn
re
elementary analysis can

of a test due to
in
\>y
/"
for nil
flu if
/Hist-
with
861166 im-t
<.f
<
and
Ifl8
to
I:>TS.
18;
'
to
^r^ if
IS
in
lss than
1M/,,
cli"i.
of
[g
uniform
and
Phys.
(2)
choos,-
\\v
if
('nsfijii-ntly,
e.
\\\(
nlivinllsly
!'>/
\>l.
rst
'(./)
is
i.
also
Less
th'
tlian
rt'inainder
and
e:
[CoiMpan- Stokes,
-2sl.]
Abel's test for uniform convergence.

ilrlicjititi-M
for uniform
converg
moiv
to
and has
A1..-1.
this
of X, 80 that the
satistinl.
],.
make
to
Jis
IcjiriK lent
i-
i-onvrr^.-nci-
90
///
1.,.,-n
due.
aln-ady nu-nt ion-l
in Art.
in
l!:
xeries *2a n vn (x)
and
>;,i
'//
in
K"i-.
8O
tliat,
,!
tlif
>>
virtue of
whatc\-r
+!
fl"
lemma
i-y
'
fined
we can
]>ositi\v inte-vr
may
.-.-
!-!-
that
<AT.
hyjM.tht11
find
l>e,
B H2-|-----f
Then,
6
L8.
//////
vnti /''//.
nuniei-icallx
all
'i'ir
COBV<
tli--
", .!+",
fl'
nt;
in
"+|.
\irtue
of
in.
114
[CH. VII.
Further, since a n and K are independent of x, the choice of

is independent of x also
and therefore 2a n v n (x) converges
uniformly in the interval.
.
terms a n
It is obvious that the
of
that
provided
x,
2a n
may
be themselves functions
uniformly convergent in
is
tl
interval.
uniform convergence.
(3) Dirichlet's test for
This
is
also
more
delicate than the I/- test.
(Compare Exs.
4 below.)
1,
The
(a,
b),
series *2a n v n (x) is
uniformly convergent in an interval
provided that 2a n
oscillates
between finite limits; that
any particular value of x in the interval vn (x) is positive

and never increases with n ; and that, as n tends to oo vn (x)
for
tends uniformly to zero for all values of x in the interval.

For then the expressions
+ am+2
a m +i
\am+l
are less than a fixed
,
-..,
|,
number
such that
K\
Kn+i + <Vf2+ ...-ftf m +j,|

and we can find an index
v m (x)</K
for all values of
x in the
Thus, using Abel's
interval.
lemma
as before,
we
see that
m+l
for all points in the interval.
Again, the terms a n may be changed to functions of x,

provided that the maximum limit of 2a w remains less than a
|
number throughout the
fixed
Ex.
interval.
Weierstrass's M-test.
1.
~=i
2 ^7^ 2
Consider
(fl);
these converge uniformly for all real values of
Sinn* ., 1
np
np
Ex.
yl&eJ'*
2.
Consider
the
OOHW
If
Abel's
3.
2an
is
np
and
.r,
because then
(1/ ,(P) is
con
nt
^^.
case
vn (.r)=l/7i x
uniformly in the interval
Ex.
<
np
(0,
1)
if
(OÎ./'^l);
~c,,
then
c-onverges.
^(ajn*)
converges
[DlRIOHLKT.]
test.
convergent,
converge uniformly
in
the interval
(0,
1).
[HARDY.]
Ml\l MY
45
44,
Ex.
i:,
4
.
(p> 0)
2r,
nf
'
then WTii
we
see
that
series
lioth
\vhc:
posit
i\
nnit'orinlv
converge
in
an
in
~a
Fundamental properties of uniformly convergent
45.
Caiichy and the
anirlr.
i-
earlier analysts
lined that the continuity of
\\ith the
<
/'(.'')
series.
exception of Abel)
lim >'(./) could
1>.-
deduced
n-*-<n
fr.im that
that this assumption is no-f Art. 4:i.

Kurtln-r. tin
immediately from K
that
discontinuity in
/'(./)
imj.li.-s
,ples
point
of
non-
uniform convergence; although K\. -l. An. 4:1. indicai

DOn-Unifomi convergence does not nee, ssarily invohe the
continuity of
w- wish
if
\i;ain,
dis-
/';
to integrate
r|lini
the
/'(./).
>'(./ )|</./-= lin,
"N,ir,,/./'
,tf
n..t
is
on
tni- <-ithT. as will be seen from th- .-xanipl.'
n.-ci'ssarily
US Ix-low.
}.
.,
IS47 Stok.-s* published his
In
'ivm (1)
iv.junvd
in
(-xc-pt
tl,>
',
(,t.
in
cinifiniiiin.fi
/,).
tin
can
///
ti^Lx
= ln,
pulilislird
auarr
of
tin-
in
yrar
riu>
(see
si
i>7
chosen
such
made
alsu
i^
int.--
1^7<>
'Hivergei'
\f
each
90
'/"
th>
of
9Wm
//"
''-
independently
way
of
/'
mverênce.
X (provided
c<
tln
>nlv
that
l.y \\Yi<-r>t ra>-
and
s,
,,),
],, lt
xtoljgg'g
paper
IK
some the years at I'-.i-t bd

''s's
paper WAS
dear from paper- in tin- tii-t \olunieof \V ci erst Pass's \\
distinction
pulilished.
pp.
amd
iiif'Tcnf,
ilix-civrry \\as
\\.t-
until
virtue of the definition of uniform
in
nuiiilu-r
that
l.'t-tur.-s)
,,t, ,-,-,,/
Kr.
Ifinrntai-y
fl,r
//
Tln-nn-m c2\ on
analysis.
\\Yii-rM ra-,. and smns not to ha\'-
in
due to
is
distinction
tin-
and non-uniform OOnveTgenoe, and

below, which <-stallishcs tin- continuity of
uniform
In-tw.-.-n
Beriefl
lisciv-ry of
\vljiL-h
ivinaiiifd
iui])ul>lisiuHl
for
aU.ut
.Vi
\ears.
116
no matter how small the constant
and
it'
may
be.
Now
write
then clear that
it is
where
[CH. VII.
(a^x^b),
\F(x)-Sm (x)\<&,
must be remembered that 771,
e are quite
independent
of x.
Thus
c is
if
any value of x within the
interval,
we have
\F(c)-F(x)\<$e + \Sm (c)-Sm (x)\.
so that
Now
77i
being
8m (x) is a
tind a value
fixed,
and therefore we can
continuous function of x,
S, such that
\8m (c)-Sm (x)\<fr,
if
\c-
\F(c)-F(x)\<e,
if
|c-o;|
Hence
which proves the continuity of F(x) within the interval

It
and
(a, b).
not unusual for beginners to miss the point of the foregoing proof
therefore advisable to show how the argument fails when applied
is
it is
to such a series as
(l-x) + (.v-j(?) + (x*-a*)+...,
when we take
Here
and
(Ex.
2,
Art. 43)
= l.
/m(#)+/m+i (#)+...
to
/(l)+/+i(l)+-
to oo
co
=xm
if
0<.r<],
=0.
Thus, if we wish to make both these remainders less than

choose m, if we can, so that
vm ]--^,
or
and the two
e,
inequalities (A)
................................. (B)
and (B) are mutually contradictory (supposing
that
<1).
Consequently the two steps used
patible here
and the reason
in the general
aimiment are incom-
for this difference lies in the fact that the
inequality (A) does not lead to a determination of

when x can approach as near to 1 as we please.
independent of
The assumption
that
the series converges uniformly enables us to avoid the difficulty involved

in such a condition as (A).
(2)
//
in
the
the series
F(x) = 'Zfn (x)
i#
uniformly nmrergent in (/'

continuous
fl/f
fu m-titm* /(./)
(a, b), and if each of
we
write
interval,
may
interval
/'x
ENTEOBATK
45
1
virtu.-
in
K<-r,
can
til
id
uniform
tin-
ol'
117
c>n\ .-r^-nce
l/(*)+/. + ,(.'-.)+...-f/;jr) <*,

however
i"
vndept
-mall
I-
may
!>
ami
p>m
if
value
tin-
<>f
will
be
we
see
of x, as before.
at
Hence We
ami since
we
of 2/n(#),
BO that
///
li
this
no matter how larê
true,
is
that
may
/>
be,
OOnV<
that
:il
=e'--
i|
m
J'i
At
th.-
saim- timr \\v have
).L()|;
|ji(
ami oonaeqnently
l
c,
e,
last
i^
m (a>)da>
F()da>-\"s
J
c,
Cl
inequality can be written
':
so that, coiiihinini: the
two
ine.|U;.lities,
we
find
\vhTf
Since
may
fc-
is
///
inent
-i\-en
a^ small as \\c plra^*-.
In-
no l<m-vr juvsmt
in
N,
p,
:,.
l>y
]iri|nT choice
the let't-haml
<>n
"!'
//-.
>ile, the
shews that
"
Jc
This
o].el-;iti..ll
ix
.t'tell
\vill
,,f
n>(,
hnwrviT.
i-onvi-rirt-nt
n-sults.
tilt-
|,i-,,l,;il,ly
.
onditioi)
tin.l
tlial
as
'
!.->>
Ml
/,/',/,-/,//-/,///
ilitVu-Mlty
>houl(l
here
give a really simple example

\\hi.-li irrni-l>\ -term
inte^rat i.-n
in
Tin- f'-llMU-ini:.
ahh
in
rea!
.!"
indrprlldrlil
to
r;isy
>erirs
lrsf|'il
\<A
.f
l-.ul>
./'.
It
n<>M-uir.:
!
nroneous
as good as any.
118
Sn (x) = nxe-*\
F(x) = lim Sn (x) = 0,
Let
so that
x > 0,
*co
S n (0)=0,
and obviously
if
[CH. VII.
so that F(0) = 0.
Hence
But on the other hand
2^
so that
which
is
fn (.<v)dx= lim
w >
obviously not equal to
(1 -<?-')
F(x)dx.
= and gives some

figure shews the non-uniform convergence at .iindication as to the reason why the area under y = Sn (x) does not tend
The
to zero.
FIG. 15.
Of course the argument above assumes that the range of

the conditions under which an infinite
is finite
to oo say, belong more properly
series can be integrated from
to the Integral Calculus; but some special cases are given in
integration
Art. 176 of the Appendix.

46. Differentiation of
If
we
we
an
infinite series.
consider Ex. 4 of Art. 43, for which
see that
z->0
Thus
lim
z-j
although F () =
/
M'(0)== oo,
*Sf
n->co
It follows that the equation
''
[lim
/(
,(*)]- lim
-i
*.
Sn'W
I'll!
45,46]
II!)
uln-ii
y tru.-
Hut
!'.
-hniiM
it
n.n-.
may
:.H
hri-
I/
,,t
///'////'//
i:
is
the era*
^-n.
r*mly
/,
(,/.
/''
of
f/
/'//
',
,,t,
-..
\/(#) oscillates.
\
,y//,//
'
/-
h,
Pi
,/./-..
<///
tin-
rential
,im
is
unitnrmK
Bfl
sert<
fi,,-
tin-
t'n.in
nuisidi-r similarly
nee pres
,-,
it
\\hirh
he apparent
as will
rradcr
Th<-
that
the
of
i'ailurr.
!!<>n-iinii'<.rin
untie. -<1
IM-
///////
/,,
//^'
F'(x),
/"'
ri-<if*
Write
\v.-
">.
have
i
'I'hus.
tin-
ly
t'uii.lani'-ntal
ju'ujn-rly
an
<t'
iiit-_;ral.
\\-
hav-
("
F'(x)=G(x)
'
i1
easier
than
tlms,-
an<l
hut
which
J.PM.I'S
Calculus.
on,-
only
.'-i
is IK
rasy witlnmt
the following infthn.!
ihcoi-ciii
l'(prr-(.in^
thr Intr^ral Calculus;
nf
)itlrn-ntial
tin-
ol'
jirnoi'
BOme usr
it
depend
ivsult
<.n
Ml,tain-d
is
_:!-alitu.
in
IM-
chosen so as to
<,
()
Thus
\^
i
|>r..\id.-d
that
<.,
K(x)d*
hrl,,,,-
,-.,
make
fco
U^X;
<((
th.-
int.-rval
(.
/>
>.
This
^
and
sn
\^
i/i
iufrr the
this
An
al(Hti<>nal
assumption
\\lurli
\\
.).
n..t
I]
wou.
tin-
ly
120
we have
Or, changing the notation,
^1
?x {/ to
-
[CH. VI L
-/(*)}
'ê,
provided that both x and x + h belong to the interval
Now we
=x,.
We
(a, 6).
have
have
proved
h
t
+ Fro, say.
=^
that
and
in
virtue
of
our
original choice of m,
But we have
identically
_
h
and so we
^(05
77i
J
J
find
+ ^-^(05)
Since
-Q(x)
has been fixed without reference to
to tend to zero, without changing
/ij
we can allow
m;
the right-hand side of

the last inequality then approaches the limit 2e, because each
term under the summation sign tends to zero. Consequently,
the
maximum
limiting value of the left-hand side
and since e is arbitrarily small,

must be zero (see Note (6), p. 5).
than 2e
this
is
not greater
maximum
limit
Thus we have
Jim
F(x+h)-F(x)
-G(x) = 0,
or
which
in
the required theorem.
important to bear in mind that the condition of

is
merely Niijlirirnf for the truth of tlu
theorems in Arts. 45, 46 but it is by no means a necessary
condition.
In other words, this condition is too narrow but in
spite of this, no other condition of equal simplieity has been
47.
It is
uniform
com rr^vnr.-
PRODUCTS,
46,47,48]
ami
discover
Mii
not
shall
\\v
into
t'urtln-r
tin-
ra
That
til-
unit''.
t\Vii
mi
not
following
<lis.-,,iitinuity.
.,-+...
,-'
,1
+ .r),
(0
<.<!).
also equal to the SIM
is
\<^-2
-f
:ind
do*
Don-uniform convergence
hat
(i)
imply
D6CC
,.\;i||,|.
i-t+J-J+found hy integrating tenn-Ky-t'-mi.

Nevertheless r=l is a point of non-uniform ...UMT^-IH-O of the series
in
./
l>e-au-f
to a
l.M.ls
indudnl
is
Tin-
OO**Oi
wlii.-li
/,,
an
])rnlucl
If far
"//
l>ut
values
<>/
im
sntlicient
In-
at
jrim-ijl.}>r<
interval (0,6) the fwn*
'!>'
has th /"/>< rty \fn (x)\^M,,.

slant (independent <>f x), tl
>/ *!>
i>r<xl
tin-
,M])]lic;itins
/,,,<
>)
flu
(v.
./
infinite product.
following theorem will
tin-
of
ind.-jx.nd.-nt
rlcinrntary analysis, and for our
Irss I'rtMjuriitly in
purpos.-
the condition ^jr"<
convergence can be extonl-l
unii'orm
oi'
infinit.-
intt-rval
tin-
l'tinition
I.,
occur
in
:.l
of
Uniform convergence of an
48.
one.'
ivmaindi-!
tin-
(It-t.-nnination
aeries
isapoeMix
-M
>
convenj
is
art
/v-)=fi +./;,(y)][i + /;(.')]! 1+u.i-)]...

continuous fumctio^ of x im th*
>l.
continuous in
the fun
l" lf
' !
For writ-
I
[
LI
then
1+ /-,(./)]. ..[H-/
-K, (,0][
Ml
+/(')][
ii:
]...
'
i
><n
<
=P
tooo-<
..,(r)]
B1
I-/-
^Hil-l/J}; fl;i+/, <1

+ !/.+,(*) +
t
-/'
(I-/-
'
'
where
/,'
...
Q.(x)-l|< RJ(l-R m ).
in.ulr U>
s.-1-ii-s
'J.
to
Thus
vol.
1.
!'.(! I.
|..
int.-uiation
t-nn-l>y-t*Tin
i
in
IS7]
li.i\,-
;
:
;.
l...n
thr Apjx nli\. Ai
tin-
bg
I..HM-I
P.I
..
jivBO
tt,
author (JVofc
-'
IT-''.
bj
/. ..... /.
.V'C
122
R m < Mm + Mm+l + Mm+ + ...
Now
to oo
where
[CH. VII.
,
Hence ?/i can be

are independent of #.
chosen (independently of x) to make 72 TO <-J-e, for all values of
# within the interval (a, b).
all
the
JJ/'s
Hence, assuming
than
less
we have
ifaô^fc,
<e/(4-e)<ie,
\Q m (x)-l
where
be
e to
(as already explained)
is
chosen quite independently
of x.
Thus, if c is
the inequalities
any value
x within the
of
and
i-i^Q^
Or
Q(x)
ê
l+êQ
interval,
we have
m (c)
1-fr P..(aO^P(s)
'
Now,
since
is
fixed and independent .of x, and since
are continuous functions of
thus
we can
x,
so also
the product
is
/',(,'):
find 3 so that
.,
provided that \x
*IMM
< -P-() <
c\<^8.
T
"^ n/^v "\
71
\9
'
and so*
provided that
|a?
c|<[<5.
of x within the interval (a,
If the function /(;/)
<md
Hence
P()
is
///*
//
derivate, snch
if
*For
so that
a continuous function
b).
(l
(l-)(H-iO<(l-iO s
In
it
\
fn '(x)
< M,,,
49
48,
\\.\l.
>
l;\
III
///,/,
//'
,i
tieee
-|
Pi
i-"iiditi<iis
ire
''
<-
'
!+./
An.
ihai
v,,
ran
h;
!..
and
a|.].li.-,l;
\v.-
ac -,,i.lin-ly.
tin<l.
1
49.
i.s.
!elv connected with
tin- I'dllou ini:
OpO84
ilnit
theory of unit'iTn,
tin-
theorem 4 whirl
u-- in sob
.jiu-iit
mi
"
in
\^vl (n)+vt (n)^...^vp
and
we wcwt
tlntt
/////
to
/,,
./;/_//
//////
vided
//"'/
/'(//)=
',(/')
Kirst
<!'
//
i.
th-
h-t
that
note
a= IT.
to 00
.U,
independent of
choose a nuinlnT
'/
.l/-t-st
alnmst thr
is
th<-
for
bhe
sain-
irMl'. tOO,
(which
.f
due
t>
th.-nn-m
,i
he taken
.-
lari;.-
+ 'Vfi-f -.. +
+ ...
sain.-.
DOOT86
is
x <e,
make />></:
to
r
rm.u_u h to
'|.
=^/ + J/,
1/
.;.
+ ...
iii.l.-]--inl,-nt
in tlu-
tot of Art,
then
x<e
to
e ot
is
\\'.
Mich that
M
and
'/'A',
/*.
At
'I'll''
).
ii
(r lu-in^ fixed),
<'/"/'
.I/,.
will
reader
Milistantially
(Art.
with
+ <'-+ ^''o-f ...
*/
00TM8 l.U,
'rh.-
l-\,<\.
/*//
lini
///r
it
iiijin
limvr(n)wf
////
lim
fl- limit
fond
1.
124
\w q + w q+l + w q+2 +... to
Also
<|(% + î+--. + ^-i)--( w

/
= M + Mq+l +...
to
<z>\
\F(n)-(wQ + w l + w 2 +...
Thus
[CH. VII,
to
+w
oo
<
oo)|
1
+...
+ wq-i)\ + 2e
>
to be remembered that so far n has only been

by the condition p^>q.
Now, since q is fixed and independent of n, we can allow
n to tend to infinity in the last inequality, and then we find
and
is
it
restricted
because
lirn t;r (?i)

n><x,
Hence, since
= wr
e is arbitrarily small,
HmFn=W=W+w
or
we
on
find, as
+ w+...
p. 120,
to oo
The following example will serve to shew the danger of trying to use
the foregoing theorem when the J/-test does not apply.
Consider the sum
=
tv()=logf 1 + ^J and p n.
so that
Then obviously
and
so the
But
sum
vr (n) lies
wr = }imv r (n) =
n
of the series
*oo
w + 20+w + ...
is 0.
between r/n2 and rj(n^+n\ and
2> =
so that
lim F(n) = \.
and hence
Another theorem of importance

analogous result for products
Suppose that
where
in
this
connexion
tends steadily to infinity with n.
Then if \imv r (n) = wr
n->oo
independent of
n and I,Mr
the
is
and
if
\vr
(n)\^Mr where
is convergent,
we have
...
the
to oo
is
equation
MI'LES.
49]
Tin-
Mini
terms Of
I'm-
'-inj.!'.;
tin]
to 00
"'.,.,)...
J/
+ J/,. + ...
condition >n<-h as
Die
ii
foregoing,
limits
timl
t-V<l4
+"V< +
be
li:
bo
n-ni;mil.-r
tin-
b6W the
tllf
in
ditlicnlt
littl<-
tin-
Art-.
(,!'
in
on
this tln-.TiMii
ivsnlts
ill.-
liavr
sl,,,uM
iv.-ul.-r
<>f
proof
to oo
tin- .l/-t-
X.-MMjilr
which
= /, = ... = !/,
j-,
that
I'.Ml
...=0.
tllf
IJui
,-,,M;itinl|
niM-i-ssarily
value of liiu(yy
/<),
In
true.
tin-
fa<-t
)'
=1
valm-
the limit
<.f
because
<
.,
/'
I.X \M!'I.I>.
1.
Shi-w
that
if
^.(i) = .i"
OOnvei
(\
r=l
+.'"),
limit.
it>
>
is
raw graphs
of
j)int
of
noii-u;
&(.) and
lim >
*
1
2.
all
Sh-\v that
,)=^--
tlir
.'ii
If /'(.) is nr\.
3.
i-ontinunus fuiirtioii f
uniformly
in
.,-
iii
th- interval
(",
interval,
shew
of the
int,-ival.
w.-
/o.
in
dnt'.'i .-nt iat i.-n.
and
that tin-
if
Lfl
any
jM.int
0^
\K
continuous,
>.,(.r)-5
r-c| <8.
= F(.r)-Sm (.r)<
Senoe
.f
can timl
//?
siu-h
(C)<j.
:
pi-vidcd that
]>>ints
t.-ini-l.y-t.
same
r in tin-
l-'uithd.
all
Ity
uniformly o.iivrn:.
th- int.-r\al.
at
is
t
the inteival (c-8, C + S).
we can
find &
iMxi.]
12()
Sn (x)
since
Now,
[CIL
never decreases (as n increases), we have

(.*)
Thus
^Sm
(.r),
if
n (.v)<,
if
F(x)-S
n>m.
n>m,
at all points of the interval (c-8, c + S).

Consequently there can be no
point of non-uniform convergence in the neighbourhood of c and therefore
;
no such point in the whole interval (a, b).

The reader should observe that this argument fails in cases such as
those illustrated in Figs. 14, 15, because then Sn (x) may be further from
F(x) than Sm(x) is. In the cases considered here, no two of the approximation
there
is
curves can intersect.]
Shew
4.
that the series
continuous for
is
all
values of
.i\
and deduce from Ex. 3 that it converges uniformly.

= is the only possible point of discontinuity,
[It is easy to prove that ^
by means of the J/-test. Now if we take v as equal to the greatest integer
contained in
l/.#
we have
r
=< r'
since
the limit of which as x-*0

5.
of
Shew
<.r(l +log
we
2<l/#v,
f),
--
that f(x) = 2^
L
~p
0<#<1. Hence
filld
zero, so that the series is
is
if
v +i
continuous at #=0.]
for
% 2 ~f~ 71 r~x converges uniformly
examine whether f(0) can be found by term-by-term

1
Shew
6.
that
2-1
n\
converge uniformly
for
all
values
i?/'"*
differentiation.
//"
_^_
l+x*a 2n
values of
all
and that
<1
if
and
they are respectively equal to the series

e~ a
and
obtained by expanding each fraction in powers of

If
7.
then
x.
[PRINGSHEIM.]
/(*)=*''(! -.r);
we have
2/n (#)=#/(l -.r2 ),

2/', (l) = 0, although
but
if
Shew
8.
that
it
9.
that the series
cannot be integrated from

t>
If \fn
(x)\<.Mn lx ,
converges and jo>l,
Apply
to Ex.
2 l/(w + ^)
8.
where
then,
this to the series
if
J/,,
converges uniformly
to oc
is
;i
if
.vÔ; but
[OsoooD.]
positive mnstant such th,U
I'.l/,,
a>0,
2 !/(+#)*
and shew that
it
does not apply
[OsaooD.]
L27
10.
If
fmit.-U
oscillates
1'",,
the series 2(ajn*)
!!
is
[IMKK MLKT.]
.,,,..., ,,
11.
|.J.
I
Km-
.'-"
12.
It'
fcbsolntely)
Shew
that
tin-
in
uniformly
;md
my
(l>ut
int--r\al.
1",,- '-ii\ orgec,
tinili-
'
intei val.
produ.-ts
II[l+(-
II|.l+(-l.
D".'-/'J,
in
OOQVerge unifm-mly
linitc
any
int.-rval,
'han a
lix.-tl
ii-n
II.-..>.
)] t
and
nuinlM-r
tliinl
tin-
tliat
at
;ill
for
int.-rval
all
it
/),
(./.
[For, divide
win -re
6<^,
'hat
at
001
tli-
-s at
,nifniily.
into \- snl.-intfrvals
interval
any assiîu-d small
lriii'4
tip.-
it'
t-nds of
f"ii\erges
j).in;
-o
and
>ll-
illts.
;u id
erges absolutely
uniformly
13.
00]
Itely
11(1+
;ill
poin
[HKM-IX>N.]
,-sich
>f
Imu'th
N-xt
nunil-r.
jM.sitivc
tinJ
suit-interval
ra--li
(^-1,
<t>(>>'r)="'fn(-'-r\
2,
^ ...)
,,,-t-l
is
less
numerically
than
h of thr-M- points,
Thi-
o.
./////
i-
further distant than
A/
le.-;m>e th
Now
-fl).
of tin- interval the nt-aiest
any point
possil.le
//';/
end of
if
a sul>-intt-rval
not
hem-.-
\<P(*)\<
and so the
15.
te>t
|<K*)|<I<M.V)| + S<2S<,
of uniform
<
Apply li'ndi\*Mifs
ti-st
^16.
with
their
and
It
to th.-
OODVergM M
-'
hi.-h
t.-nls steadily to
Deduce
'
n,
it,
do,-s
tin-
//
OOHYergM
in
-.!!, Tal,
not
if
OODVergQ
nil
:
<
val
<>f
that
'-.^0.
OOOFM
it
n.
'
>(*
is
|M>ssil>le
that the
values of
Shew
unit'..rni!y
if
>oiue nuinl.er ;: sii-h thai
17.
'led.]
.'
l.i
exai.
[CA
CHAPTER
POWER
50.
SERIES.
The power-series 2a wo; n

of
types
We
one of the most important
is
series.
uniformly convergent
recall the result-
VIII.
proved in Art.
10, that if
Inn
>w
when x
the power-series converges absolutely

series
cannot converge
if
>
I /I, for

and
so there will be an infinity of terms in the series whose absolute

values are greater than 1.
Thus any power-series has an interval ( 1/, +!/) within

which it converges absolutely, and outside which convergence
is impossible.
By writing x in place of Ix, we can reduce this
interval to the special one (
1, +1): and we shall suppose
this done in what follows (we exclude for the moment the
cases
or
oo
).
Thus suppose that we have a power-series which is absolutely

1 and + 1
so that if
convergent for values of x between
n
/
is
1
number
between
and
the
series
any
^|^,,|& is con:
vergent.
series
Then, by Weierstrass's
Jf-test,
2a n#n converges uniformly
because in that interval

|
n -suit that
a power
series
a nxn
it
is
clear
the interval
in
an kn
that the
k,
+/>),
Hence we have the

converges uniformly in an intercut
|
which falls entirely within its interval of absolute converge

It sometimes happens lint further tests (such as those given
in Art. 11) shew that the scries is .il>solm<-ly convergent for
= l; and then the interval of uniform <-<>n\ i-r^viice extends
|oj|
.
50]
vi
U)
-f
:npare
ami apply
But
\\
al>
may
it
\rt.
lerau*e
11),
tli,;'
happ.-n
nt:
ith
tin-
,in.
tli.-
inn-eases with //.ami
ca -
this
in
variahle
leqUi
is
than
(if
Oû
Con-
1
i.
rmly
nei-tl
IP
an interval which
in
as far as ./=
The
1.
l"t\v-cii
and
<>
ft
in
uniformly
lnit
tlit-
Similarly
).
it'
./
2a,,a;
~(
(hut
1.
(-,
int-rval
the-
includes
uniform C0nv<
nt ili- interval nf
Ex.
not
apply Abel's test

factors X n D<
jH.ints
-4-1
1,
+/)? where
is
any
niunl.-i-
l-l.nig to the region of
n,,t
.1,,
uniform
Ex.
Tin-
2.
i-.ifi's
Ex.
+...
The
3.
the interval (-1, +1).
in
uniformly
ft
8
<-"ii\erL. es
f
uniformly
li.-tv.
\\V
luit
the p-'int -
ivturn
HM\\
hitherto
side.
tin-
1"
seriea
the
(-.1.
-f-^1).
Tims the
Where
or
()
oo
which wv havr
"=0
converge ai.sMhurly tor any value of x\

uniform convi
may be taken as
will
SB"
interval
happen^ that
it
It'
O,
and
+1). whore / is any urn

the region of uniform
(-,
il>es n>t lifhuig to
cases
th.-
to
one
.n
k-t'l
-i
the interval
in
>i'
.1
can
)>
arhitrai-ily
la:
si-i
1-f
.y foi-1,
On
+.1
any value
the other luunl.it'
lini
\
t'or
An example
* ami
is
a n |" =
of
any value of x other

of
uniformly omverênt
ii
).
tli
led 1-y
i:
tha.
the
Sei
"
eannot
POWEE
130
There
is
SERIES.
[CH. VIII.
one important distinction between the intervals of
absolute and of uniform convergence; the interval of uniform

convergence must include its end-points, but the interval of
convergence need not.

terminology, the former interval
absolute
use
to
Or,
is
closed',
the
convenient
latter
may
be unclosed.
That the interval of absolute convergence of a power-series

is evident from Ex. 1 above, in which the
series is absolutely convergent for any value of x numerically
less than 1, but the series diverges for x = \ and oscillates for
1
x=
On the other hand, Ex. 2 gives an illustration of a
need not be closed
closed interval of absolute convergence.

But we proved (at the end of Art. 43) that the interval of
uniform convergence must be closed, whenever the function
Sn(x) is a continuous function of x. Now for

2aw # n we have
Sn (x) = a + a :x + a^x? + -f anxn
a power-series
which is obviously continuous for all values of x. Consequently

the interval of uniform convergence of a power-series is certainly
This fact
closed.
p.
129,
is
or Art.
top,
not deducible from Abel's theorem (see

51), for it does not appear impossible
a priori that a power-series might diverge

1.
be uniformly convergent for x
|
theorem
51. Abel's
is
x=l
at
<
and yet
expressed by the equation
Km
provided that 2a n is convergent; this of course follows from
the fact (pointed out in Art. 50) that x = l belongs to the
region of uniform convergence and from the theorem (1) in
Art. 45.
Abel
also
appeal to uniform convergence

theorems.
A =a
Write
Then
since
to + oo
then 2a n # w also
This theorem cannot be proved by any
but the following method applies to both
shewed that when 2a M diverges, say
tends to +oc as x approaches
1,
form of Abel's
x,
#2
1.
;
A^ciQ+a^
,
...
Lemma
is
...,
J,,
rt
+ a 2 +... +
a decreasing sequence,
(Art. 23)
'
we have by
the second
50, 51
ABEL'fl THfiOBl
\\hnr
//
//,
'
ar-
//
tin-
ami
IIJIJMT
limit-
1" wt-r
are those
limit- an-
ill,-,-
Sin..-.-
.1,
.1,,,
...... m _i, and

l
(1)
ind-|M-nIi-nt
<!'
y,
we have
-O-H /'.-''" =Sa^^ //
'
//,,*".
*MWI
I
*,
we can
then
h:il
however small
be.
may
ami
indi-x
the
Thus
Hut
Me frin
\vt-
]..
only on the series
-ntls
2# n and
is
therefore
5) both th.-
maximum
that
(1),
and so
arbitrarily small,
is
di
///
oi
independent
ami minimum limiting
Note
(see
(6), p.
valu.
lim (l'/,,r M ) =
Hence
to
lal
s.
=-
x-*l
lar.u't-
may !.
.V
lim
x
ility
\\hnv
/,
way
with
c.uin.M-trd
that
A'
as
In
//. n
.V.
ust-d
we prove
alx.ve,
that
that
results
is
the theorem of com-
series.
are
t\\"
ilm-r^-m
from
//
that
by the same argument
inur
ami
/..
s<>
>//
seiief,
ive
tl.-
an
(1),
Z?,,
=^
-i-î
+ -" + ^.
IM(>
and
Thus, provided that

(2)
_:
see
f..icur
tin-
I'/-,
I'./,,,
similar t>
timl
J n |fjniafrr n
two divergent
parison for
Suppose
I'./,, ./"
we can
// So,, otcillatesi
lim
ly
can
*!
an aixumrnt similar t> that
l.y
/-'////,
we
oo,
Thus
so that from (1)
Thus,
to
say
h" \\.-v.-r
//,
/>,
//v
".
tind
yc<
applications of the theorem
v=
it
is
.")+*-
-"
generally the case that a*
is
always
POWEK
132
.
lim
Obviously
</>
Xf-l
(# ) = km jHm ,
SERIES.
and lim
[CH. VIII.
= Km/h m
3> (x)
X>I
Bn /A n approaches a limit
then possible to choose m, so that
and now suppose that
however small
lim
X>I
Hence from
we
(2)
Then we have
be.
may
<f>
(.r)
^l
if
have,
as
n increases
is
it
at once
and lim
<I?
X --l
(#)
/(^)
X-+\
X-+1
Repeating the former argument, we find that

!
In a similar way we can prove that

that
Bn IA n
if
oscillates,
f(x)
may
Bn /A n-*&,
if
then f(x) -+cc

and
between limits which cannot
oscillate
be wider than those of B n /A n

Thus we have the theorem, due to Cesaro
+ b n ) / ( + !+...+ a M ) approaches a
If BJA n = (6 + 6 X +
.
. . .
Jmite or
definite
limit,
then
infinite,
lira [(26 B ay)/(2o H

ac^l
)]
= lim Bn /A n
>oc
a result which can be obtained also from Art. 153 of the Appendix.
It should be noticed that if b n /a n approaches a limit, Bn/A n approaches the
same limit (Appendix, Art. 152) thus a particular case of the theorem is
:
If b n /a n approaches
definite limit, finite or infinite, then
lim (bn /a n ).
Ex.
It
1.
form
possible to obtain the first
is
of Abel's
theorem from
the last theorem, by comparing the two series
Ex.
2.
by comparing the
Similarly,
4;r +
we
see that
.J
+ .4 + .4# + ... and
lim -(A
if
series
+A
then
lim(2,,.r
[FROBENIUS.]
Z.
x >!
Ex. 3. Again, if the limit in Ex. 2

a further mean. Suppose that
,.
tlien
+ (2^ + ^
n.
.t<-
we may
consider
the series
)^+(3/l
and prove that

\\'-
not definite,
we can compare
y|
is
tliat eacli
+^J
lim
-i-.l
i)^-f-...
(!'</,.../")
of tin- <-.\ainik's
1,
=
2,
and
[C'ompare Art. 123.]
3 includes the preceding one.
CONTIM
51, 52]
Ex.
AH
4.
ITY.
n-a<ler
tin-
<'
'hat
in,.
+ *'+...) =
(ii)
+ 3'-V+...)=I
'.r
*!
In rase
the
(i),
+ ...)=
'-Kr' -.!>
'
sei
14.
^4.... gi vt-
while the series for (l-.r)~* gives

'ml
M0
DM
\ve
(iii)
.l
we
A(
It'
r,
'I'ln-
liin
[
LT'-ii'-i'al
rl
<)...2.
-f 1).
roA.
great generality on
quote th- f-ll-.v.
ire
examj.],-
th.-n
<.
A(./-) .*]
-.r)] for values
(1
thcoiviii^
to a ptwvr-srrii^.
iiu-nts
(1)
,,f,
<>f
2A'(n)
neaf t<
./
1.
Properties of a power-series.
52.
of
that
approximately ly A[l
the,, reins of
pi-..\,-l
A> an
.l
-2
log(l-x)
!'
+ ... -f
.-!,
1)
tin-
J,,+
liav.-
r=l.
funetiin
i.-
than
+
\\hile
tli;c
li:ive
1'rin.u'slHMui*
\vhieh diverge at
= 3.
ZfH
cx>logtt/log,
th- fa-t
Kinally, in case (iv)
sl,,wly
f|
40 of course apply
in Arts. 43,
proved
that
BO
can inakr thr
\v.-
i'ullnwin--
state-
2an#* is u
power-series
rval
'/
"/////'//
/'/>
<;,,ili
n>iu.-< fu
region
<Val ii'itliin
//'
is
r>
!'",>
(<-;
')"
uithin
any
f/
+1
*(Saj> xfc-2
fl
<>
any
<
-'-r
<>f
+t+I
).
noe
reg
:
!/,
\\'r
nutf
that
p^
interval
tin-
nit
thr
I'rnin
80 that
lit,
alt.'ivd
i'act
"
,Mf
i(
M,
ditrcrcntiatiun
l,y
that
lini
//
= lini 1^1" = lim
f:
convei
altsolut,-
<t'
-.
linn.
>f
1,
p.
i.
\vi,,:i-
full
will he found.
i
>r
lim(n +!)/!=
1.
so that liui/c
1-y
Art.
>
l.Vt in
.i\.
references
POWER
134
interval
the
[CH. VIII.
Abel's theorem (Art. 51) to the integrated series

(2) the point c z may be taken at the boundary
By applying
we see that in
of
SERIES.
of absolute
that
convergence, provided
integrated series converges there,

series does so or not.
An
example of
(4)
If a power-series f(x) =
no matter whether
the
the original
this has occurred already in Art. 47.
â x
converges
an
within
interval
k,
an
there is
-f-/c),
interval within which f(x) =

t
has no root except, perhaps, x = 0.

For suppose that a m
then the series
is
the
first coefficient in
f(x) which does not vanish
a continuous function of x, in virtue of (1) above, and fm (x) = Q for #=0.

find c, so that \fm (x}
is less than ^ a m
if
x
c
k
consequently in the interval ( c, c),
is
Thus we can
Hence /(#) =
and
is
if
interval
c,
has no root other than # =
different
from
zero (so that ra
= 0),
oo
a =b
two
/(#)
c)
has no root in the
(4) that:
If two power-
g(x)
n
^b nx are both convergent in the
be,
we can
which
satisfies
may
S),
= &!,
we hardly
known
x |< &
is
|
series are identical.
In practice,
f(x)
that
< <
oo
= ânxn
then
the
in the interval (-c,
interval ( k, +/c), and if, however small 8

find a non-zero value x l in the interval ( 8,
and
c).
an immediate deduction from
(5) It is
series f(x)
ever need this theorem except wlu.-n

be
to
equal to g(x} for all values of x, such
We
have hitherto discussed the continuity of the po\\< Tfrom the point of view of the variable x\ but it sometimes happens that we wish to discuss a series 2/n (y) xn
53.
series
regarded
as
function of
the variable
y.
The following
theorem (due to Pringsheim) throws some light on this question:*

*
Further results have been established by Haitn^s (Math. Annalen,

9), Using more elalmiatr aiiaK
1006, p.
K<1.
CrJ,
PEINGSHEIM'e THEOREM,
52, 53)
caTi be
\'
and
are fixed
A,
trl,f>re
2/,<
la
th.it
<
./
',
if
8u</< ////
interval
we
theorem,
tli-
;>
,
(-pJ-J
which
in-.-.l
is ino!
compare the
<.nly
ami
of y
-p.-ii<l.-nt
i>
ainl
is
a continuous
tlu-r.-1'ore
function
series
OQQVefg
Weierstrass's
(1
th-
'),
<
continuous for all finite mines ofn, and
when |05|<A': tlms the s.-rics ^/' (^).a5* (by

cm vi-r^-s uniformly with ivûnl to // in
;
Then
any value.
f/tc
,i
A'.
To prov
with -/l?i
?i /ias
'//'/
>ntlnuou8 fuiirfiun
found
of
int.-rval
//
in
that
Interval.
was erroneously ûppo^nl
It
1/
'/
rriii^shrini
that this c.nlitimi
The
>
sullici-nt
< <A
con-
(a-<umin^ /',(//)
an .-xanipl.- s
./
not sutlicient (see
is
the convergence of
to ciivun- th-
that
lias cotistruct-<|
rxainpk-s are due to AK-1,
follnwitiLT
(1) Tlu-
was
But
tinuous).
Aln-1
l>)
tin-
of
continuity
li\-
intn-val (a,
n
^/ ,(iy)a; for
in
A'
l'jc+-2 '.c-+...
Example
\\itli
+ ,<\c" + ...
(5)
the i-xccptimi of (5):
(\jc
<1)
represents a continuous function of y.
ffiiny+i
MtiiiMini.-
COnvergea
'
it"
miniums
\,
it
i-
(,
I)
Thr
-.
./
<1
ntiiiu.us at
<.f
>i
if
v=0,
thus
Art
(see
tin-
4;r, ...
2?r,
X\<\\ and
+ ...
Imt altln-uirh
liin
tlu-
tii>t
...nv.TLrt-iu-c
."'
"'-'
tan"
t.-nn y(l-f//-).
Thus
/'(y)
u)
of tlu-
sn
Spim
//
it
st-rit-s
still
(see Art. 65).
'
from
tiian
ili>
fniH-ti.>ii
the 9eries
'->in ^// -.',' sin 3//
\vlii-n
th.it
= U.
Hut
if
POWER
136
does not follow from that of
SERIES.
2/M (#)#
[CH. VIII.
for all values of
Thus the
y>0.
series
converges
if
Bin2y
sin2y-
n2^
#<1, when y>0; but
the series
x>\.
if
diverges
8in.y
(5) Pringsheim's
Example
Let Jfn tend steadily to oo with n in such a way that Km

and let J/ = 0. [For example J/" = 0, J/ =wn .] Then write
J/'n+1 /J/ B
7l
and
of
it
is
y and
evident that the series
any value*
for
tinuous for
discontinuous at
values of
real
all
of x.
2/n (y)#2n converges for all real values

Further, the functions / /j, ... are con,
But
?/.
#^1,
if
the
2/n (y)^r2n
series
is
,y
For
and
so
if
|y|>0,
o
Now
the terms
(y) are positive, so that
rt
î
From
these facts
Of course
if
it is
and thus the Weierstrass-test

54.
^1
clear that the series
<J, the series

than 1, and so
\x
positive but less
if
is
is
discontinuous at # = 0, if x^\.
= 0, because /,(y) is;/
continuous at
applies.
Multiplication and division of power-series.
As regards multiplication of two

shew that if both series
power-series, the
results
of Art. 34
converge absolutely in the interval!

is given by
9
/,
+/), their product
* Because
ami
^x2n /M n
Of course
converges for any value of x, since lim Afn+1 / J/ M = ao
to be zero; if y = 0, all the terms of the series are zero,
.
we have taken y not

and
S/M (0).x2 = 0.
two
f If the
UK-
-///'///
/'
series
of tin- two.
have different regions of convergence,
this
iuUTval
\\ill
be
MUI/ni'l.K \||M\
53,54]
which OOnTflVgei
the same
ill
al.-olut<-ly
.\Nh I.IVI>ION.
apply Abel'd theorem (An.
\v-
\\
H'/V
Co
If
int.-r\al.
to the equate
~>\)
**-(So,
we can deduce
pn>\ id-d that
oner
at
all
we
</;/;>;<;,/,
hi-
(Art.
tln-nivin
assuni,-
that
tirst
from zero; and for simplicity
ditr.-ivnt
>nsilrr
CAE,
that
:{-!
convei
Beriefi
tin---"-
tin-
term
constant
\v- tak.-
08
it
i>
Thus
1.
tir-t
where
//
Now
and
Art.
liy
pi-i>vil-l
.')U.
nuiiiln-r
any
and
<!+//)-
This
./:,
arran-v,l in
1..-
pw.-rs
th.-
than
Irss
upper
radius of coir.
th.-
limit of \b n \p n (of cour>-
h.'iv
1=1-^., + ,^ -:_/,
seriefl
and we
may
th.-n
ol.tain
l>e
,.,-:
_,/,,.-:_ O/^/,^.
multiplird
-
..v
by any other
j.
for tlie (jiioti.-nt
^r" r ''-h" ,'-+...)(l-f^ r ''-f^

:
in.,
tli.-
.f
may
.lu.'tinit
Th
wben
( ,
= 0,
th,.
initial
still
l^=(l
l.c
1'iit
trim-
in tin- ilriic-minat
found as
tliat
/-_,
is
powei
n..t
/
/>',
Tlu-n, as
al,.
:u l
'',
Thus
<>:
16
ohtain
\\'-
in
may
that
In-ill^',,.'-",
th
BWO)
th.-n
w-
h.,
of
>=!.
POWER
138
In practice
mined
it is
[CH. vin.
often better to use the
we
thus
coefficients
SERIES.
method
of undeter-
should write
multiply up, and obtain, in virtue of Art. 52
(5),
from which we get successively

^0
A
is
'
2'
'
more exact determination
* ' '
'
of the interval of convergence
given in Art. 84 below.

55.
Reversion of a power-series.
Suppose that the
series
y = a^x + a 2x
+a x
converges absolutely in the interval

required to express x,
Let us try to solve
-i- ...
+ k),
k,
and that
it is
possible, as a power-series in y.
if
the equation, formally, *by inserting a
power-series
If
2b n y n
is
convergent for any value of y other than zero
(by Art. 36), the resulting series may certainly be re-arranged

in powers of y without altering its value, at any rate for some
values of y
leaving the question of what these values are
;
for subsequent examination,
y=(
A) y +
0,^ = 1,
or
and so
Thus
A+
have, in a certain interval,
+ OA + 2 & A + 3 V)2/ 3 +
= 0, a^3 + 2a b b + a^ = 0,
V)2/
a 1 6 2 4-a 2 6 1 2
we
2
on, in virtue of Art. 52 (5).
we can
determine,
step-by-step,
the
succession
of
coefficients,
6X
= l/a v
It is evident
different
from
*For the
K.xs.
case
0,
b.
= -a
z /<**,
bs
= 2a22/a 6 - aja^,
1
from these results

or the assumed solution
when
....
that a t must be supposed
aj
and a a
is
B, 31-33, at the end of the chapter.
will certainly fail.*
not zero, the reader
may
refer
to
REVERSION OF
54, 55]
\\'e
may
without
loss of generality,
with a slight
BO,
//
Then the equations
change of notation, we can start from
= + ",.''' + 'V
:1
.,-
i'or
l>
+....
.-.juations sh.-w tliat \b n \^/3n ,
])l'o\
Now
where
given
that
id.-d
the (Mjuations
ol)taine(l
in
for the
he writ!
equation ran
and
then
9E1
for
the
are
,-J's
those
which
wouM
in.srrtin^ the series
l.y
the equation
Hut
any positive nuniher less than /. the aeriefl 1

iver-vnt, and BO we can timl* a nuiuher .17 such that
it'
ft
is
/'
fol all
values of
//.
Thus we ran put
_,_
/.'/
this equation
-'i
Of
iiiii.itic.n
below.
i.f
M.
Mfl
l.e
POWEE
140
SERIES.
[CH. VIII.
the negative sign being taken for the square-root, because

and tj vanish together.
But
(p
where
+ tf-^M+p^X-Mft-n),
X = 2M+ p + 2 [M(M+ p)]
*,
Thus we have
and
can be expanded* in a
thus, since X>yu, the value of
convergent series of powers of rj, provided that 0<i;</x. But
this series is clearly the
if
0<>7</u.
as 2/3n >7 n
which therefore conw is
so
that
finally 26 n i/
j8nS|&|<
same
Now
verges
absolutely convergent in the interval ( /x, +/x). Consequently
the formal solution proves to be a real one, in the sense that
it
certainly convergent for sufficiently small values of y.
is
It is perhaps advisable to point out that the interval (

/*,
+//-) has
not been proved to be the extreme range of convergence of the series 26,,y
we only know that the region of convergence is not less than the interval
;
For
instance, with the series

$
X1
^ = r + 2! + 3!
'
we
could take as the comparison-series
This
is
found to give
X=
and
the method above gives an interval only slightly greater than

+). But actually (see Arts. 58, 62 below)
so
(_^
and the
series for
y = e*-l, so that ar = log(l+^),

^ converges absolutely in the interval
1,
+1).
Lagrange's Series.
In books on the Differential Calculus, an investigationt is commonly given
for the expansion of x in powers of y, when an equation holds of tin.- form
56.
We
fSee
=
anticipate here the binomial expansion of Art. 61, for the case v lt.
-ntinl
7:
Khvanl>,
for instance \Villiain.-on,
('d/nt/it*,
/>////
chap.
Differential Colai'm, dia|). 18.
I.
55, 56)
process 'gives an
an
LOB
-t
tin-
jH.îl.K-
in a
l'-d
""
ivality,
,-,
d-
tin-
mis ^grange's prol>h-m
iw.-i >i<.n
tli-
t..
t-<|iiivalrnt
Iâgrange's
tin-
nf
n>w seen
i-
the power-series
.f
-)i,-\\>
iii\-r>tiâii"ii
M
in
<
''"""
MIC -"it
in
expansi>:
nvergent powerpOMible to write the
tli.-ii
"-"
th<-
hi-h
inf-.i
-\|.i:
'ally u.
f-r the coefficient* in
ratioo
anal
gives no
it
\\'-i
the
in
â,^,
ti
to be,
to the
i.il
'
"!',
[/'(-')]"
ng
r\|aii-
\v.-
inert-
tin-
cstalilislj
ur 'n-ral
form
Lagrange's series in
i.f
00
\vhirh ;H
knew
\\'.-
p'\\-
small values of \x\ and
wlu-n-
/,----<,
f,
anil
"
in
iiidr<l
tlie
^_
in fa.
|.y|,
(lV"in
we can
otln-r
Art-.
and
''><>
another
is
that
."."))
ntly
\\
have
c..i-Hi.'ii-nts
still
The
to be found.
convergence cannot In- f.iuinl. ly flrnit-ntary ni'tli"ls, until

the coftli.-ifiits havr IMM-II (l.-trnniiifd.
inti-rval
..f
Mtiaic tin
.'1
we
find
-|(*.;
hividc in.w
liv
r
i/
Suppose both
tln-ii.
uiv \vliolr nuinlier,
\\1,
thifl
on the
riu'lit.
t<>
i-|iiatiin
thriv
is
and wr get
-\pandi-d in asn-iulin^
IK-
..nly mi.- t.-nn+
containing *
'
and
this
= r.
is
It
n,\v
,-1,-ar
th.-
tli
'.lie
\cept for
7i
r,,-Mi.-i-ut
..f
./--
the rxpansion of
in
finl of tlu- cha]
= r,
1.,
= Jo*-"' +
hut
in
this
sen
t-\i
if
is
because x
On
tlu-
loss
than
;
othfcr h:inl.
if
r, \vr
nt
'
>
r,
,.f
any pow-r
1,
g+^
= + J5,^
term
in
POWER
142
r
in ascending
g'(x}jy
formula, although
It
[CH. VIII.
and this is equivalent to Lagrange's

powers of x
put in the above form by Jacobi.*
;
first
to be observed that
is
SERIES.
if
the equation in #,
solved by some algebraic (or other) process, there will

additional solutions as well as the series found by reversion.
is
usually
be
This series
gives the solution which tends to zero with y.
of
y=x-ax?, and g(x}=x, we have
Ex.
1.
x~ l
in the expansion of
If
(x-ax*)-
Thus we get
and
=x- r (\-axY r
to
find the
coefficient
rdr =
x =y + a# 2 + 2a 2y 3 + 5a3?/ 4 +
so
which converges
Of course this
if
|
ay
. . .
< J.
series gives only one root of the quadratic in x,
Ex.
2.
In like manner,
Ex.
3.
The reader
if
y=x-axm+l we
,
will find similarly that
namely
find for one root
if
y=.r(l+#)
37i37i+l
n
,
then
+ ~,
method of expanding g(x\ we take the followbx

ax
e
in powers of y = xe
To
expand
ing example
r (a ~ rb)x
r
and so the coefficient of x~ is easily seen to be
Here g'(x)ly =ax~ e
Ex.
4.
To
illustrate the
a(a-rby-
f(r~l)l.
Thus we have
S* = l
V + -a(a-36) y +
+ ay+ a(a-26)
2
.
s
.,
which converges if |y|<l/e|6|.

In particular, with a = l, 6=-l, ^ = e*, we obtain Eisenstein's solution
of the equation log=,y (see Ex. 11, p. 18), in the form of the series
Qes.
Werke, Bd.
6, p.
37.
56, 57
KXroNKNTI
SPECIAL
CI:I;TAI.\
57.
\|.
The exponential
SKHIER
i'<>\\ 1:1;
SERE
limit.*
\\v pn.r.-.-d tu prove that
win-!-.-
05
= liliH.
-*<>
Consider
tin- sin-rial
first
intr-Tal values
Thru we
//.
tiiiil.'-
and
6886 wh.-n
t-nds to intinity th;
n*=*X.
writ.-
>n .-xpaiKlii
"
i-
-3(i-J)...(i-^
Now.
can
uî-
ezpressioi] satisti-s tin- cnmlitions
this
as
ni'
Art. 4!>:
tin- roinj.arisoii-series,
-,,..-
.,..
where
F<
the <nvatrst value} of |X| for any value
is
A',,
ol
we have
>r
wln-n-
A',,
is
(,!'
Also
coarse iinl<'|enlem
Hill
>'
=J
n
i
use
lini
-)=
n->.\
-|iial t<> n.
and
Finallv the
liinA'=.'-.
and so of com
infinity.
|iml+*l+s+.l_^+..
Thus
"Tin-
in-i
Mitik'ndnl
;i\
to
II.
x.
before proo
further.
tFor
tht-
That
L'<
-iitTiil ti-rni
tluTi-
-lu-s
tht-
is
;i
limit
in the liiiunnial
jji-i-ati-.st
x,
us
H
expansion of (1+^)
val
in.n
;,-iit.
,ist-s
t<>
livcause it
intinity.
is
is
supposed
SPECIAL POWER SERIES.
144
[CH. vill.
now
If
v tends to infinity in
any other way,
be
contained between the two integers
stage,
say; and of course n will tend to infinity as
(!
)"
vg will
and
since (!
follows that
does.
n+1
)" is
Thus
and
n >oo
contained between (1
2
we
M
)
see that
and
(l
n+1 it
,
write for brevity
(*+)"will
any
= Km (n+l)=x.
Thus, from what has already been proved,
it
(n + I)
between* (l + ) n and (l +
be contained between ng and (n + l)g, so that
lim
we
at
will be contained
n>x(ng)
If
v will,
n and
2/r(*,*),
r=0
be seen that we have used the theorem

lira 2/r(*,
n >-oor=0
n) =
r=0
r (#,
[lim/
>)
)].
That is, we have replaced a single limit by a double (repeated) limit and
of course such a step needs justification (see the examples in Art. 49).
;
Special cases.
If
= l/i/, we
have the equation

l
The value
+ d!
i-,+
+ l+i
z!
...
tooo=&
of e has been calculated in Art. 7
and proved to be
2-7182....
If
=l(i/-l), we have
lim (1
so that
l/v)~
e.
v->ao
These two results
where X approaches
*It
is
be combined into the single equation
from either
On
tin- ih-tinitioii
side.
understood that the positive value of (!+)" is taken; and

H+1 if " is
obviously fnntaiiu-d between (!+{)" and (l+f)
a
is
if
is
the
statement
other
h.md,
the
irrational,
consequence of
<>f un irrational ]>o\ver.
of course
then this value

rational.
may
is
>
58
57,
58.
\V.
CES,
The exponential function.

may denote hy tli.- synil."!
'
'I'd-
-ii
x*
/'<
a*
ill.-
tliat
..f
and that
X,
term
ditl.-ivntial
its
xjMuiential series
a*
(1
t'uneti..n
B)
K(x)
a contin
is
coefficient
i^
-i\-)i
by
that
ditl'-ivntiation, go
i)-i++lJ+jP+...-jrM
.=
we
see
Art.
!'r.ni
TT| .
(/--I).
rl
">7
that
"x
This rrsult
I-'A
i-an
also
.lut.-ly
..|uati.iis
dir.-ctly
prn\-r(l
(An
rule
tin-
>
values
Sequence
l'
Hut
B,
ot'
nuinlx-i-^
i-ati"iial
is
it
!'
step
In
'I'
valu- nf
t\.r
the
sometimes clearer
w,-
wli..x,.
limit
is
./,
we
write
,-
mplicate
ezpx, as
the c-hapt.T.
in
A'.
p-nerally
e\]M.ii-iit
>r
h;.
'
shall
to use
i'
.ts
it'
alid IMMMU^-
I'utuiv.
when
;/-
relation must also be tru-
tl..-
i'or
(-
lim //(/ i=
hut
oonvergeni serie& Tli.- i-.-ult

which
re positive int.-êrs)
in-atinnal
E(x)\
for
rios
inultiplyin^
leads din-ctly tc
(Art 46)
>
(in
rational valu.- of
]:ist
tin-
+-[(iyf-E(n)-[E(
Thus W B6 that
the
I'rmn
l.
instra
i>n
POWER
SPECIAL
146
The reader
[CH. VIII.
an independent proof (depending on
find
will
SERIES.
the Integral Calculus) of the equation
in the
Appendix
As
Ex.
(Art. 162).
may shew
a numerical example, the reader
that
e^=
4-800...,
6^ = 23-0 ____
and hence that
may be made to the example of Art. 59 for some of the

needed in the calculation.]
[Reference
results
The
59.
sine
and cosine power-series.

5
[3
Hn+gr-"*-
and
Then it is plain
functions of x and
as
may
that
Sn (x), Cn (x)
both continuous
are
be seen by differentiation.
Now C
(x)
= cosx
but
negative
negative when x
Thus -r-[Cl (x)]=
is
and consequently
-i-
SQ (x)
with x and consequently
(x) vanishes
is
is
1 is negative,
positive.*
(x) is positive
GI(X) vanishes with x, and therefore
when x
C^x)
is
is
positive; but
positive
when x
is positive.
Hence
-T-[Sl (x)]
=C
S^x) vanishes with
x,
(x) is positive
so that
when x
is
S^x) must be
positive;
positive
and
when x
is positive.
That
and
is,
-r-[C2 (x)]=
therefore, since
when x
*We
is
make
S^xy
C2 (x)
is
negative
vanishes with
when x
is
positive;
C2 (x)
is
negative
x,
positive.
use throughout
this
increasing with x and is zero for

for positive values of x.
article
of
x=Q, then y
the
(if
obvious fact that
if
is
continuous) must be positive
SINK AM'
58, 59]
\Ve can cnnt iniie
us INK SKKIKS.
his ai'innm-nt
and
l.y
W6
doing SO
(7 (ar),
Cz (x), C4 (x), CQ (x), ...
fi
17
find that
are negativ.
\\
h -n
...
(aj),
J
Dispositive,
while the expressions with odd Mitlix.- are positive,
This .sh.-ws that sin,/' lirs between th- tw<.
+ i>:
1
'
.v
+(
-
3-2/1+3
-I
-...+<-
,.,,1
'
(2iT~
Hence, since
sin2
we uhave
In like
manner we
(see Ex. 4, p. 8),
2n+3)l
Xs X6 X 7
KT+T? &-:+
t
to
ao,
find
...
7.6
>.4
These results have been established only for positive values

of x; but it is evident that since and its series both change sign
while cos./' and its series do not change sign, so that
with
./',
the results are valid for negative values of x also.

The figure lelw will s.-rve to sli'W the relation between
and the
first
two or three terms
Ku:.
sin./-
the infinite ser
If,.
Lot us calculate c<.s(irr) and
Ex.
\\Y
in
li,t\,- .r
This
= ^7r=r.")7<s
vci v
'
-0
&**=
07W,
=
L-S546
Also
= -00003.
-0209,
i-j:>;
0,
tlu-
wror
r-1, the error
1-oing less
IKMH.I:
leea
than -00003.
than -00003.
SPECIAL POWEK SERIES.
148
60.
[CH. VIII.
Other methods of establishing the sine and cosine
power-series.
(1)
is
to
Probably the most rapid method of recalling the series to memory

assume that sin x and cos x may be represented by power-series.
Thus
sinx=
if
we have
and
cos x
-j-
+ 3a 3# 2 +
(sin x)
. . .
so
= 0,
Further, a
because sin.r
Hence we get
and
is
2 a2 =
-!=-],
3.4.a-4 =
-a2 = 0,
and
1,
for
#=0.
=0,
2. 3. o 3 =
4.5.<x 5
COS.T is
4.
=-a
3,
or
or && =
=-
so on.
But of course we have no a priori reason for supposing that sin x and
cos# can be expressed as power-series and therefore this method is not
;
logically complete.
(2)
We may
start
from the
series,
and
call
them,
'say,
S(x\
C(x).
Then
multiplication of the series gives
C(x) C(y)
Hence
in particular
[C
S(*c) = ZS(x)C(x\
and
From
these formulae
we
C(Zx) = \C(xff-\S(xff.
can shew that S(x) and C(x) satisfy the ordinary
results of elementary trigonometry.
Further,
that C(x) =
and S(x)
is
it
can be seen without
much
in that interval.
liccaiiHt-
and
2.
and
2,
because
is
negative,
so
But
always positive! for any value of x between
C(x) can have only one root between
trouble that (7(2)
has at least one root between
it
and
2.
Thus
steadily decreases
\\i'
[ES
~
'all
this r,,ot
thru
and ao
l>e
positive.
Hence
ami
W
MI
<
>
these facts
in
<
nut
is
[t
IX.
'hap.
tin- \vh..lr
of Analytical Tri.ur "ii"iiirtry
to piove,
ditliciilt
C(X+T)--C
imliK-tioii
liy
or
l.y
can be based.
methods given
the
that
l>rlo\v,
n(n-l)
2!
\vhrre
= t&n0, and
both
aeries trnninatr after An or 4(
i(/>
that rinpl..yed
f..r
n>
To theae azpnemona
u>inu as
r
tin-
-an
i:
apply the
the.-:
coinjtarison-series
...
<>r
# = .r,
Thus we have, on putting
+ l)
;!
=|tano:|^|f |.
\\itli
exponential
limit
r+
Ait.
in
ami
1\/
iLM
.~7
\ve
get
2\^*
1
"*)!
*-\
-]1
-f-
= COS -,
^
we have
lim n /=0,
that
an.l
thus
li,n
eoe*-
lim
(H
M=l
(see Art. 58),
the
150
Another instructive method
(4)
is
[CH. VIII.
to apply the process of integration
by
parts to the two equations

c
sin# =
rx
co.s#-l=-/
cos(x-t)dt,
Jo
If
we
integrate twice
by
parts,
we obtain
t
2
~\
and so on.
Thus we
sin(#-) +
co$(x-t)
r-
cos#=l-|
x 2
(x-t)--2jsin(x-t)\
and
s\u(x-t)dt.
Jo
f-2
cos(#-)J
f-x
fi
-^m(x-
find that, in the notation of Art. 59,
Hence we
find
and
61. The binomial series.
Let us examine the series
-l)-j
+ K^-l)(^-2)+...
to oo.
We know
from elementary algebra that if v is a positive

now
this
series
terminates and represents (l+x) v
integer
for
values
to
examine
the
theorem
other
proceed
corresponding
We
of
v.
By
by
Art.
x
|
<
vergent in
Now
the
12,
region
and so
any
f( a; ) =
(Art.
interval (
^l +
= vg(x)
where g(x)
differs
( l/
absolute
of
50)
A:,
the
/.,),
convergence
series
is
given
uniformly con-
< <
where
is
-l)a;4.( l;-l)(^2)^+... to
x]
say,
from f(x) by having
(v
in place of
v.
BINOMIAL
60, 61 J
Also
(1
\+( v -l)x + (i>-
-f
1)(.
x
-2)jj+...
BO that
Hence we
see that
or
/(aj)=.-l(l
where
a constant in<l. JM n<l. lit of

v\<- choose the
po*>'
consequently,
.4
But/(0)=l: and
x.
is
have
.1
that
for (l+x)",
>lue
it'
18,
This result has, of course, been proved only for the im

'.\trnl-l t. inrhnl.(-/, + /.); K't u- now see if it can
!
th.-
points
ly the
series
,/
>th
//
and so the
for
an.l
l'h-
ni">t
tile
for
assuming a
iy
On
.r=+l
'..,
the
f,,r
sum
r.-calling
is
'2'
if
i'-fl
1
is
the series to m-
18
0,
if
positive.
+a r v + <i.,.r-+
mu>t
+ e
..nU-i
t..
o>ni)>l-t-
|i"..t".
in
r,
Thus by Abel's
iC
lind that
\vr
-'
tin-
l,
).
f(.r)=\
tliis
th--
M|uatinii
sei
sul.stitMti..ii.
in
1 if v is positive (Art.
positive (Art. '2\

of thf Sel'leS for
is
sillll
(1
I
/>>i' +
if
1).?:,
nictli-ul
rapitl
th- ditlrivntial
i^
(??
positive: and
is
i-
(Art.
term
/>th
is
(r+1)
if
">
the
<>t'
<jU<'ti.-nt
series con\< r-vs for ic=
./=+!
thenl-elll
Th.-
+1.
1,
ami
whii-h
r.)
multijilv l'_:-!li.[
and
\<
\it.
\rt.
.iiimt
must
t\\,,
L'T
s;i.
writ
that
\vitli
tlu-ir
tlitlVrmt values of f (say
j.l.xlu.t
as was
lu-
tlu-
rth
|H,
\\n
,,f
its
valur
f.-i
= l.
SPECIAL POWEE SERIES.
152
(3)
We
[CH. vin.
have
where we obtain the
we
last line
by integrating by
parts.
Continuing thus,
get
Now,
if
is
provided that
On
positive,
(l+x-t)
between
lies
and (l+#),
so that
> v - 1.
the other hand,
if
is
we can only say that
negative,
x}
IJ
-,
if
r>v-l.
Thus, in either case, the difference
tends to zero as r increases to infinity.

If
x=
1 it is
,
interesting to note that we can sum the binomial

number of terms. Thus we have
series to a finite
and so on.
Hence the sum
Therefore
is positive,
62.
it is
to (r-fl) terms
clear
to infinity
is
from Art. 39 that

if v is
The logarithmic
this
take as our definition (see Appendix

logarithm the equation
to
if v
series.
We
Because the series
sum tends
negative.*
+i+ +
...
to GO
is
divergent.
II.) of
the natural
GABITHM*
I'
81, 62]
\vln-ii
<
./'
'i
wriit-
ml-t+t
(
...tOQO,
from
tin-
Art.
l.y
log(l
llo\\.-v.-r.
"1
(oiivrr^rnrc
Can
Thus
+ .!) = ./-
ry
<>nl-r
in
in-
to
tends
to
//M
unit''.
mi
t.-i-m-ly-i -Tin
i*
cl.-arly
l.-ss
than
+1
/'
as // tends to infinity.
Ini/m-it/iiHtc series if >"//''/*
when
the otht-r han<l.
)n
tin-
/,-ro
Thus
1.
of
inlt-ji-al.-
P<
<s=
+.
= ,--^,^-f!.' -...-H(-l
i-f.'-)
which
=/.
positive, tin- la^t intf--ral
is
to
\\rib-
oF+
It'
make use
i"
861160
tin-
:!
A./--+^./- -J.i-
not
i^
it
uniformly
ha\- th- -\|>ai,
\\-
./
is
i>n.vMMl
even for
-,-111
that
x\.
only say that
jf;
than
98
anl from
+#
Jo
Ssion
tl.
be '-xrlnd.-.l iVoin
it
woull
!.
of
ami, as a matter of fact,
9;
been
|'ro\
tin-
to
.<!
li\
th<-
loirarili
96li6fl
-2
(1-f J-X = tf
i.l-
tin-
./=
cxi..-ct-il that
'
..
-i
Tli.it
.-re
is,
of
Art. 47 UM-1 Ar-
unitnim
OCH
be
ext-
tegrated series (compare
154
[CH. vni.
It is now necessary to consider whether the first of these series can

be re-arranged in powers of y without changing its value.
By Art. 26,
this derangement will be permissible if the series
is
if
convergent, where f = |# |, ^ = \y\.

But the last series is the expanded form of (1 -)~ T!, which is convergent
<1 that is, if |#|<1. Thus the derangement will not alter the sum.
;
Hence we get (from the
coefficients of
and y 2 ) the equations
?/
to oo,
...
^(l + HHi)+-..
Similar (but less
of log (1+0?).
63.
is
simple) series
be deduced for
may
[Compare ChrystaPs Algebra, Ch. XXVIII.,
to oo.
higher powers
9.]
For purposes of numerical computation of logarithms
it
better to use the series
x
which can be found from the previous series by writing
2
x ).
for x and then subtracting; or directly, by integrating 1/(1
In either
than
way
-I
the remainder after
* ~T~
~*
J-
~1
terms
is
seen to be less
iX.
The
details of calculation for log f
I
5
= -20 00
-^3=
-J3-=
-^r=
may
00 00
be arranged as follows
-20
800000 +3
12 80
+7
51
+9
00 00 00
2666
32000 +5
67
6400
83
06
20 27 32 56
2
Tin-
error involvrd
-i4L_-.
result
is
rff))
in
= -40
log f
54 65
terms
l.cvoiul
m^li-ct in^
which cannot
affect the eighth
correct to the sixth decimal.
tln>
tifth
<lt<imal.
is
loss
than
Hence the
\M
>l\
LBH
62, 63, 64]
AJEf
get
Also
log}
th.-><-
In
ivMilts \\-
log:.
nmiilM-r
l"ic
log 2, logî.
lt;
64.
II'
and
>f
.")
terms
..f
illu>t iat<-.
\\-f\-rr.
it
i^
-on verge
r
\\ritiii'_ ./--!
31,
and
,n
b;i\
reversion of
principle of
coetlici.-nts in this
<
+ ir* + ...).
series for arc sin z
.>'.
\vlii.-h
be easily calculated to a
the reader may find formulae for
to
the three series obtained by
can expreefl th- Dumerically

in
l"g).
= 1-09861:.'.
febove hav- I>MI f"iin<l.
tin-
--
tlit-
r'i"in
simOarlj
=
log 10 ^:
T<.
placrs.
in
l>y
Iog3
-698147,
'"
t'--iind
1..-
and so enable the logaritlnns
The power
>in ^. we
BO,
whi.-h .-an
7,
han
rapitlly,
'KXXXKX)3]
havr
\vi-
particular,
!.
ri^hth
the natural logarithms of all integer-
tlnil
Iog2=
more
tin-
decim
t" -ix
In (with tin- rM-i'ptinn of
t..
unit in
;i
= :.'[-imilll+-OOCn.=a'2L'.'Ji
th;ui
l-ss
again
tin-
t\v>
first
ii"'
t.-nn*
obtain
tx>
i-asy
manner: hut
Beriefl
value of
b-ast
\v<-
(Art
we
~>~>),
as a
_ ent
r
\\bicb arc
<!'
tin-
^'iirral
me
can
la\\"
th>-
<!'
tlic
ditliculty
by nsin^ the Calcn!

\V.- have in i'act
i
^cîl-^r
wlr
to
i-
The
+ :r +
.:;
between
lie
i
(
:>. 4''
,g
'
^.l.,;
-ATT and -fl-. sn that
positive
r-
B6VJ
nhtain.-d
i^
lV..in
binomial
the
Beri
I'mit
will th.
-f.r:
--^. ami
unil'nrmly in any interval (/,-. +/,) if ()</,<!.
We may int.- urrat.- fcenn-hy-term and so obt-iin
:
writing
./
= ,- + ,
i
arcrina
i.
156
which
1,
converges
+1), as
writing x =
may
[CH. VIIL
absolutely and uniformly in the interval

be seen from the test of Art. 12 (5). Thus,
we have a series for |TT.

Although we have not found* a series for tana?, we can
we have
For, writing x = tan
easily find one for arc tan x.
l,
<j>,
dx
or
d>
'o
where
The
is
</>
supposed to
lie
between
|TT
integral last written is less, in
and
+ JTT.
numerical value, than
2n+l
aj
/2n,7/__!
;
lo
and this tends to as n tends

Hence we ha vet
to cc
provided that x
to
...
where
In particular
==+!,
= 1.
'oo
jTr^
we have
Of course this series converges very slowly, but by the aid

method given in Art. 24, the reader will find no great
difficulty in calculating JTT to five decimals, from the first 13
or 14 terms. The result is JTT = '78540.
of the
Art.
We
can, of course, form such a series
54),
but there
is
by dividing
no simple general law for the
sin.c
by cos a- (compare
coefficients.
The
first
three terms are
more coefficients are given in Chrystal's Alyebra, vol. II., p. 344.

The method used in Art. 54 shews that this series will certainly be convergent
XZ yA y&
< 1 &"il a short calculation will shew
any interval for which ol + T-, + ^j +
anil
in
that this is satisfied in the interval (-1*3, -f I'.'i); but by means of a tlu-ornn
given in Art. 84, we can shew that the region of convergence is ( îr, + ^TT).
fOf course the term-by-term integration could also have been justified by
making use
of the
uniform
coim-riM-in-i-
<>f
the series
-x'2
+ x*-
...
ntlGONOMETRK
65
64,
t"
rfoai
kii'.un
uvergence of the series
,'\,
tan
7r )
.tli. -r
= 2.H
I*-.
1.
Various trigonometrical power-series.
65.
clear
is
Art.
iVi'in
'21
that
tin-
BZpCU
i
l-2rco80+r*)- =l+(2rco80-r )+
without
may In- arraii-vd in ]M\V.-I^
|mvi<lrl that*
<r^-|.
f+... to
..!'
Tin- srijih-nre of coeflirimts
fraction
tli.-
t'..r
[en
Il
i\.
Tl.
I
\i.
.1
.1
(l-rcoa0)
....
.1
is.
alt-riiii:
h..\v\
er,
its
\alu\
MHI-- -aîly l--ti-niiiinMl
(12
are of cuur^.' functiona
<>i'
H.
have thf Mi-ntity

1
+J
-rooed- L-h^r
:i
4-
-l
/:;
-^00802^^00802
ami
h.-ncr
we
Art.
u-in--
get,
+...
80+...
52
-I
-0-1
39,
.1
,
and BO <>n.
Thus w,
ha\.-
It'
\\v
tli.-
-l+rco80-Hrt coe20-frI coe30-h...
r_
sul.tract
and
ilivi.lt-
l.y
r,
8
an.l whn. r |
..atfl+r^âiri+r
a
<l,as..
Thiii|2roM0
:
to
\\
J^+y-co-
Jr|+^ = f|<l.
oo
;f9-f ...
tn
POWER
SPECIAL
158
Combining
1
these,
we
SERIES.
[CH. VIII.
get also
r2
...
An
to
inspection
we
see
that
r + r 2 + r3 +...
led to
exactly similar argument will give the result

to oo
By
oo
or
these series converge
all
when
when
Thus we are
0<r<l.
converges,
enquire whether the equations are not also true for
the interval
(0,
Now we
1).
find, identically,
Hence, as for the geometrical progression (Art. 6), we see

that lim.R n = 0, if
and accordingly the first equation
holds for the interval (0, 1). And the other equations can be
0<r<l,
extended similarly.
Again we have
= - 2(cos 0+r cos 20+r2 cos 30+
.)
by what has been proved.

Hence, integrating,*
2r cos + r2 ) =
log(l
we have
2(r cos
+ Jr
no constant being needed because both

Also we have
dt
>
Thus we
cos 20
+ ^r3 cos 30 +...),
sides are zero for r =
dt
(*'
==
~1
$7\
0.
5 ^ r^ ^ an
find
sin
<ft
Jo
= r sin + ^r2 8^20 + 1^ sin 30+..

*
Term-by-terni integration is purinissihU- hrcause, if 0^r^fc< 1, tlie series

be compared with 1 + k + k- + k?+ ..., ami Wt-ierstrass's J^-test can be u]])lied.
may
65
TBIGONOMETRK
\\
li.-iv.-
only
.-stal.lMi.-d
</<!:
ill--
-,,
SEED
\i.
nations above on
we knew from Art 20 thd
hut
tin-
ih-
1,
two
series
>-f ...
rin0+j MB
tin-
lepi
20+in30+-= or
i'or
Thus, hy AbeTfl theorem (Art
C080+
!ir>t
~>i
we have
>.
Mm "J
''+..--
I"-'*
-l-Vn.sfl + r2 )
-!Tjlog(4Bin
and similarly,
cos0-4cos20+Jcos:W-...=
a
in
In
likf
1'iv^li
lit-
in
tioo
driluo-il
inaniii-r
\v.-
iV.iiu
i^
tin-
-r2 )
li
the la-t
unnecessary,
]>ivcc(lin--
1,\-
clian-iiiL:-
l'r
have
hisin30-|-...=limarctaD
v
\ii\\
is
tin-
t'roin
tli.-
anu.-lc
tan
r/>,
runctioii)
tiûn-
wliich
must
-vi<l-iit
is
it
(acc..nlin^
l>-t\\v<'ii
lit-
Î
that
tin-
the
to
J.
TT
an^l.- in
drtinitioii
ijii-
of
and +J7r; so
th--
that
17.
Thu>
I
Jut
t.Tiii
Bm0+ism20+i8m30-|v .. = J(ir-0)
= 0, nr "2~. th- valu.- of th- Beriefl i>
if
in
it
vanishes: thu^
tin-
0<d<
if
>
(>
l-rc-au-
.>ntinu':
'
and
If
lies
Ix'Uveen
'2L--
aegatiye,
w.-
and 2 (Jb 4-1) IT,

.
MM these are points
wli.-iv
/,-
is
an
iir
ha\-
= j[--
of n.-n un.
series (A:
N].
POWER
SPECIAL
160
It
SERIES.
[CH.
not difficult to discuss the series
is
by a
direct method.
/(0),
we
have,
by
In
fact, if
Sn (0)
sum
the
is
of the first n terms in
differentiating,
Thus
n(#)
Now, by
Art.. 174, Ex. 2 (Appendix), the limit of this integral
provided that ^0
between
lies
and
TT
/(0) = i(7T-0),
if
is
TT,
and consequently
0<0<27T.
/(0) = 0=/(27r).
But
Thus the curve y=f(6) consists of a line making an angle arc tan ^ with
the horizontal and two points on the horizontal axis.
glance at Figs. 12 and 13 of Art. 43 suggests the conjecture that
the limiting form of the curve y = Sn (0) consists of the slanting line and
two
vertical lines, joining the slanting line to the axis (see the figure
But as a matter of 'fact this is not quite correct, and the vertical
below).
lines really project
For
above and below the slanting

$_ w^ ? ,_ g
clearly the point
line.
(X.lri)
belongs to the curve y = Sn (B), whatever the positive number

Now, as tt-*oo, this point approaches the limiting position
/i
A sin
in virtue of Ex.
3,
Art. 174.
A may
be.
Similarly the point
u = Sir.
/"*
y=
sin
dt,
Jo
belongs to the limiting form of the curve.
Now
the integral
(sin t/t) dt
can have any value from
.'o
to its
maximum
TTX (1-1790), which occurs for A = TT so that in the limiting form

of the curve y = Sn (B), the two vertical lines have lengths T85194, instead
1'851 94 =
of
^7r=r57080, as conjectured.
85
--
-J7T
-1-85-
FIG. 18.
Some
of
of the approximation curves
y = S n(0) are drawn
,sv-/vV, etc., p.
/'////.
Mag.
49
(5),
and the curve w = 80
vol.
45, ls)s,
I'].
is
for various values
and in arslau s /',.,///>/.<

given by Michelson and Stratton,
in Byerly's Fourier Series, etc., p. 63 (No. II.),
XII., Fig.
5.
<
vin.
A.
161
and Integration.
Differentiation
1.
.Ju-tify
tin-
i-iju.-itinii
5-a4s + J^-J:i/- + -~A'n:? rf
<a>
'
Thus the
De<1 ......
series can
be found
"""
Unit.
in
4>0
>-
..nal.
[GAUSS.]
+ -=
1111
ll^'-sba+-=
-Jl
[Math. Trip. 1896.]
that
.'\v
Utilise
3.
if
and
are positive integers
i-
Ex. 2 to prove that
x*
(l-.r)
r^73 + 273T4 + 3T4T5 + --^~2^ + ^^~
<<
and
find a
fonnula
fr
the
sum
of the
111
Obtain the former result also by

4.
i:
the logarithmic
2r4-4-rr6 + 6TY78--=i('r
Shew
5.
that
with
restri-tions
rrt:iin
and dedu
Ft
\:>,.
on
a,
j8,
- 3 )-
se:
.i''
y,
rh.
r (y)r(y-q-)
7-)r<y-!sy
*In
number
of these examples, the
me
are implied.
I.s.
is
used aa equiva
cases the words /or sufficiently small values of
x
^
POWER
162
6.
The complete
7.
Prove that
8.
From Ex.
SERIES.
[CH.
elliptic integrals are
Ch. VI., or Ex. 5 above, prove that
13,
r(2)
-4
,2.4.6.8
[Write
9.
a=
J,
/3=-, 7 = !.]
1
Multiply the expansions of (1-x)' and log(l-.r), and deduce by
integration that
[Compare Art.
10.
If
62.]
n
y = (l+^)- log(l+^), shew that
Deduce the expansion
11.
Prove that
[By direct multiplication, or by expanding the
(l-^)
12.
Prove that
[Use
Shew
13.
the equation
similarly that
Prove
that,
if
\x\<\,
Is the result true for
x=\
differential coefficient
LMPLE8
vin.
I
14.
Sln-w that,
if
\jc.
A.
< 1,
i
:j
:>,
log[i{lW(l-*)}]=15.
Prove that
K^,,^.,^.'^,
[It
is
easy to see
th.it
(1-
+ J* + J** + ...)
,(.
+ (l+*%r-J^ + J*-.
= 2{*-(J
,116.
tf
j/=(seco:-ftanar)
m=
17.
prove that
-IK-i + jrfa-SK.s -... =
Of
I6C,
[Hen
2a^,
(\
.1896.]
[.V
//.]
Verify that
J(arctan.r).log(l-
where
and dciluoe that
J log 2
[Here
(1
=
.1
"
!>-....
+)$-<
= S...
For the second
]>art,
[Math.T
.S)^+(*6
fi
use Al.rls thftOT
Derangement of Expansions.
18.
/
Prove that the
iy .w(m
+ l)...(iM+M-l)
,.r
'
71
whei
Sh.-w that
it
f .;-"
I'lu-fli.-ii-iit
in th.-
><(w
+ /0
i-\ji:uiin
/>
>
//(/<-!)(/
4!
2!
-,
is
inultiplr
.f
the o..tlicient <f
in
--7V
- -- --
(1+-.
(1-
the expansion of
i+qt
3-;;
(1-
POWER
164
Expand exp (arc tan .r) up
19.
SERIES.
[CH.
term which contains
to the
x>.
[Math. Trip. 1899.]
(l+^2 )=y, and
[This function satisfies
^ = 1,
/nl ) then
This gives
=1,
=-l,
=-7,
possibility of the expansion follows
20.
Shew
if
find that
if
an+l = a n -n(n-l)a n -i.
The
that,
we
so
....
from Art.
36.]
|#|<1,
>. 1902.]
[This
is
If
21.
the expansion of (2+y)a where
#=2y+# 2
y=a
+ a 1 ^+a2 -^2 + .-.
obtain the
>
first
use Lagrange's series.]
and second terms
in the
power-series for
1
yApply the
22.
and
a^x
I
Ids
-----
=-.
example to prove that
last
x dy
y-a
if
- [(x - a) cos
f(x) = (sin x sin a)~
a]"
!
j- [/(a)]
Prove that the
23.
n\
number
the
- lim/(#) = | sec 3 a -
coefficient of
r
L
/"(a)
= lim/(.r),
A sec
a.
â
[
Math. Trip. 1896.]
in the expansion of (1
^-1)^-2)^-3)
7i(7i-l)
"*"
2271-1
2(271-1)
2.42w-l2rc-3
-1
"'J'
of terms being ^(^ + 1) or
Determine the
24.
J
then
first
three terms in the expansion of
2
2
1
25. If [(l-#3/)(l-tf,y )(l-#/.y)(l -tf/^ )]" is expanded in powers of x,
the part of the expansion which is independent of y is equal to
s
- *2 2
(1 +s )/(l -^)(1
)
[Math. Trip, 1903.]
.
we expand {(\-xy)(l
[If
It
is
26.
up
[The
obtain
then easy to pick out the specified terms in the form
Expand
to
-x/y)},
we
(l+.r)* = exp(l
and including the term

first
three terms are e(l
follows from Art. 36.]
2
3
-^x + ^x - J.r -}-...)
in x*.
-J^ + JJ^
2
)
the possibility of the expansion
165
VIM.]
th-
27.
powers of
in
that
28.
\\li-
she win* that
*,
tlu-
1.3.5
(T^?
3!
1.3
th- ooefl
MI
iii->t
x*
(1
ia
..-"
to
,-ial
N '(
and hence find an
-./-'),
HARDY.]
29.
in
!p'-
ami extend to any nunil>-r

Shf\v
30.
-<*<!.
""-rr?
that
indices of 8iiiuniati"n.
-f
tlu
tliat
IH-\\
and by
-(inverg/nt,
are as
gh
^TT
-M7
:K)3.]
[.I/"'
following series are alislutely
tin.-
summing with
SB
,.
r.r
_j_
X
X
-
[STERN.*]
Special Series.
l-.-^T^'^-
i-tsii*
sln-w that
also
th-
y+1
that
V>0,
[If
'(1
-tyjt
and
If
wt'
x-i-i.-s
ran
this .-an
the
32.
tli.-
j.iov,.
I..-
\aiial.lr
i-
|t..>itivi-:
Art.
l,\-
to
\~>
that
'(
(1-r
/-,
ise
\-f
if
1).
extended
from
and
njual to
i>
in
th-
iri-t
y)
is
equal to
0>y> -1,
by
~'ilt.]
.+ ftr +
s
-(i-ô+^y1900.]
*See
Dirichlet's
.-,te
InttgraU
(c
( i
POWER
166
33,
Shew
power-series
that the
sum
SERIES.
of the squares of the coefficients in the binomial
is
>-*-
[Put a = /3=-i/, y = l in Ex.

34.
35.
[Math. Trip,
Ch. VI., or in Ex. 5 above.]
13,
Prove that
x=\.
Discuss the validity of this equation for
[CH.
Shew
times
its
that at r per cent., compound interest, a capital will increase to

original value in n years, approximately, where
In particular, the capital will be doubled in the time given by the
approximation
n=
-69.3
r
36.
h'35.
If
prove that/,(#)
is
a polynomial of degree
in
x which
satisfies
**
x ~tt
the equation
l
/2 =(.r + a) 2 +#, /3 = (# + a) 3 + 3#(. + )+#, and that if

a is positive all the roots of /,(#) = are real and negative, and that they are
=
[HARDY and Math. Trip. 1902.]
separated by the roots of /,_i(#) 0.
Shew
that /j =x + a,
37.
If
â nxn = (l-x)-%, prove
that
and deduce that

1
n-
n-...
[For the latter part, use Ex. B,
[Math. Trip. 1890.]
2.]
Trigonometrical Series.
38.
By
writing
r=-#sin0
in the power-series for
arc tan {r sin @/(l
r cos
#)},
shew that
arc tan (a -f x)
- arc tan a = (x sin
0) sin
6-
(.r
sin 0)2 sin
20 + $(x sin
3
0) sin
36
...
where
39.
is
=cot#.
Prove that the
eerie s
convergent and is equal to $ir-6, when 6 lies between

-eot0 in lv\. :is
or
[Put r = cos0 in Art. (;:
:
[EULER.]
and
TT.
167
vin.
I
Sh-\\
40.
that
li.-iiiflrr
tin
i,.-i_-lit
If
41.
of a regular h-jtagon inscribed in a circle IB nearly

il
triangle whose Bide is equa.
le
equal
qual to
<
tan./'
-A
42.
<l,tain the f..ll.\vin;
rals
d.-li
Iog(l-2rcos0 + r-').
z)*'
'(i
Vl -rc<-
oosne
in
ItiOD
1
43.
an
sin>^si
/"'
'-f'-
l-2rcosi
Jo
'.
<;.')
intt-_'
Pv int"jratinu thr
and tTiii-ly-terni integration (Art.

and 1.
|Mati'ii
(Art.
and
./
lies
between
<>
and
C080 + ;,
nn
:J0+...
1,
= (Awe can
infer the results
')7T
2
,
r3
30+...=
cos^-f^cos
I.M.]
[!'.
[\.t- that
tin-
.iixtrmts of
\aiiiil-.
intrifiati..!!
i->
can
!<
f,,i
all
values of
0.
Th>;
d-l.-i inin.-d
unifcim
'
tinnuus.
-f ...
J,
-JA**,
I
thus:
anil thei
= ATT*, we
ur et
whi,
45).
'
*-...
~>
tind th.
The
POWER
168
Shew
44.
that,
if
SERIES.
[CH.
|r|<l,
arc tan
Prove that
45.
6+ cos 3(9 + 1 cos 50+ ... = J log cot 2 J0,

0+sin30 + sin50+...= +7r, (0<0<7r)
- 1 sin 30 + ^ sin 50 - ... = log (sec 0+tan 0) 2
sin
cos
sin
The second
series
these values of
0=^5
= 7r, STT, ..., and

changes sign at
while the fourth changes sign and
is
is
discontinuous for
discontinuous at
IT*
cos
Prove that
0cosa + ^cos2#cos2a + cos3#cos3a+...= - J log [4 (cos 0- cos a) 2 ],
cos
0cosa-cos20cos2a-f ^cos30cos3a-
46.
20 cos 2a + sin 30 cos 3a +

sin a + J cos 20 sin 2a + J cos 30 sin 3a +
cos a +
sin
cos
sin
<a<
where, assuming
Prove
47.
and
find the
if
that,
sums
...
. . .
. . .
= Jlog[4(cos + cosa)2],
=/(0),
=g(0\
TT,
i(ir-a),
ifO<0<a,
-â,
ifa<0<7r.
0^0^7r,
of
" cos
710
sin
?i
'
48.
Series for
Sin ce
TT.
- ?r = arc tan - + arc tan 5

o
2>
4
2 2
,2.4/2\
both of these results are due to Euler.
3F
,2.4/1
LMPLEfi
vni.|
'1
'.sit
6 fl-i*
1+
also
-L
YJ. la. "flJ.
+
OV!oJ f -J 4TooL +
l
'
/'
TT
a.
'
/ l Vj.
5(100;
+-}
gave the result
-^^
lead
-7r
2 4
3 io
\\lii.-ll
169
i;
Hn
which leads to
4"fi>L
'$
:<
11=5,
COB
lu'lilv
*-*
-|
l++
7r,*4f.
2.4/144V
+-
In writing these serif* il"\\n, \\in
tin-
n..ti-
that
..
2,
below
l>e
'--in
put
f'-rni
r,
\
1
EXAMPLES
-J-
1',
Euler s Transformation.
1.
Sl,.'\\
hy the
saint-
nirtli.xl as
...
il
N'(
-/r,.
'hat
in
!+//-)[
...1
01
Similarly. jr'V- that

;
^-
.r'+...^
If
>'/
2.
By takin_
',
= J,
logty
cr
=l,
...
in
+...],
[KrLKR.]
K\.
1.
ihew that
if
:/
< 1,
'
K\
Dedu*-c that
J\J*\_
,2.41
2.4.6
v:
V3 log V~7r/~~
h"
=1
3.
Shr\\
in the first
ftlM that
'
V:
Ml
[Ei
if
4-...=
vV)-(Mr,,),-^ + (/>-,,)^
,
^;'
POWER
170
In particular,
4.
a =0,
Thus from Ex.
if
,,
=w
we
DaQ =-l,
SERIES.
find
^^
3,
2 ^ #" = (a? -f 7#2 + 6-r3 + .z4 ) e*.
Similarly,
oo
If
5.
is
[CH.
,,r
*=2|L,
an integral multiple of
and
e,
in particular
[WOLSTENHOLME.]
6.
By
integrating the formulae of Ex. 2 above, prove that
if
|y|<l,
24
l|+|:
2 w4
2 y4
Are these equations

7.
From Ex.
valid for
y = l?
4 above, prove that
oo
n
i
xn
n
[J/aÂ.
^n>. 1897 and 1905.]
if
^(-l)
n- l
= 0.
Sn
71
Shew
and that
2 4
[Math. Trip. 1904.]
similarly that
Obtain these results also by differentiating the exponential
series
and the
series for (1 -4?)~*.

8.
Apply
Euler's
method
to prove that
*
m m+l m + 2 f+3
__(_*\+
"*"+ UH-W
9.
Apply
Euler's
method
to
F(a, ft 7, ^) =
where
/(,
=
ft 7 ,)
,
2
/"
Vi
"
+ l)(m + 2)\l+#/
"(7tt
prove that
(T^-a^(
!
'
x+
7- A
7'^)
x* +
...
[GAUSS.]
\MI'I. l>
171
I',
Miscellaneous.
10.
If
//
[#=;
11.
ail.
(!+.'-)
l...th
that
tha:
less
If
',-_'!'
'
Also
1.4
12.
From
the fxpan-i-.n
drrinial i-lan-s.
(>l>tain
[Til
the
in
.f
ih-t-i niin.-
(1-^.,)"
tli--
Uirr.ilL'.-JTS.]
saint-
\\-iv
tin-
of
root
t-iiln-
from
tin-
-.\]>ansi.>!
:
KK -l
Theorems of Abel and Frobenius.

13.
With
thr notation of Kx.
!>.
sh-\v
tliat
tf
14.
If
n.i
if
then
OUL]
that
apply AWl's
15.
If
<
tlit-..|
ivea
-iii.
]
ITS and
PJ
pr'\c (hat
if
I'/
^.'iil
lin,
that
if
/''
.1
oi-
di\
.-1
L.
51 to
liui t
'-nt
I
tenda to
J-^,
lim{2
X *1
Hi;.'
ntiv
and
that
.1
tendl to
POWER
172
SEEIES.
[CH,
16. If the coefficients

b n satisfy the conditions of Ex. 15, and fn (x)
n
decreases as n increases (but is always positive), prove that ^b n fn (x) will
converge provided that 2a,,/n (.r) does, if lim A n fn = Q. Deduce that when
,
/n (l)=0,
and lim
2, ^(.r) = rr>0,
[Apply the lemma

17.
If
of Art. 153.]
= 0,
ra
ll
fn(x) = &?vn
A 2v
then the series

are equal.
-f
in
Ex.
shew that
16,
2 A 2 ?> 2 + 3A 2 y 4 +
. .
If further v n has the limit
series has the limit | as
Use the method

\imn&v n = and limv
x tends
of Ex.
18.
= l,
rt
17 to
x *!
[For another method see Ex.
From Ex.
lim
x-+i
to
3,
v
1
if
2,v l
lim(?iv n )
y n is positive,
+ 3v 2 - 4y 3 +
x tends
as
. .
to
1,
shew that
and
if
converge and
then the last
1.
shew that
then lim(i'
x vl
20.
\i
Ayn = y -v n+ i, & 2 v = v n - 2v n+ i + v n +2, and
Writing
19.
then
&2 v n
if
is
positive,
and
if
-^4- y 2 -y 3 +...) = |.
Art. 24.]
prove that
15,
= 1 log 2 = lim |(- 1

IX-l)- -^^-u
-f>.
)-^^
1-.^'
*_i
n(l+x ) 2
1
Establish the asymptotic formulae (as^-*!),

,
and $n(-x)
[The difference between the two sides of the

\-x and use Ex. 18.]
~--
first is
less
than
in
the second, multiply by

21.
On
the lines
formulae (as /r-*l)
of
Exs.
17-20 establish the following asymptotic
Tf
--'~4irp
where
is
Euler's constant.
[CEsXno.]
Lagrange's Series.
22.
(t
Shew
that,
if
+ a) n = n + wi(t +
where w r
is
a positive integer,
~
n~l
+ 4- n r a(a - rh) r
l>)
is
the ordinary
. . .
+ //>)" ~ r + ... + a(a
nb)"-
binomial coefficient
[ABEL.]
vni.
B.
3
I
multiply
i.il
.". ;ind
I'.v
authors have con-id- n-d th- \aliditv
i.f
tin-
coetl
equate
equation, also due to
A1...I.
4(t + a)=<t>(t) +
h.-ii
i.
--ults raniK't
if
</>(f)
I..-
is
'((
..'ivi-ii
is
\\'.
IM-M-.
M.-iy
remark, ho\v-v.-r. tha'

Tin-
1:<1.
Prove that the
<!>"((+.
Ini:
23.
-* b)
+ b) + a(a
in
i-ot-Hiri,
28,
tin-
").]
-1)]"
BZpl
\\
l'i"\,- tliat tin- roctVirirnt
'f.,"
'
in tin- <-\j>ansi<.n
(l+.r)
2 "- 1
(2
<.f
+
[Mtl,.
[Tse Lagrange's series
(1) for
y = e*-
and
(L
y
24.
and
t*
Kxjtaml
tin-
drtri-iiiint-
[Write
l>u
in
>
is
interval
= (a-
pc \vt-r-st-rics
J.a^ran.u'*
in
./,
tlie
n,
-e.]
\v
/('')]", in a>.-,-nding
for tlie f.-llowini;
coefficient
n.r;
'-
wh.
the r.\jiaii.in <>f [1

the '"tli. -it-lit ..f .'" in th- Sip
iiiiin.-
1906.]
in
.'
Trij..
-j/), wli-
)"
(*-^l+y and apply
If /'(.')
25.
']
r
t-f*-t-
and
recent
iiio.-t
that the
powers
of
:-e to
f>
(3) log(l
MK.]
[\\'<
[The ivsnlts
:
0)'"" ^
(8)0or(-l^
;;':;
Oor(-l
<4)
Tin- values
26.
occur
\vh-
(6)0
\
.-n.]
that
.i
N
27.
where
Lagrange's
s.
Use Lagrange's
(3)1/( W -1):
series to establish the eqi:
/=./(! -.r) and
on.
POWER
174
28.
[CH.
Prove that
6.7# 5
+x2)
and
where
29.
SERIES.
t=x(\
Shew
8.9.10*7
2
1
1
< fa.
that
on-l
,
nt *n-S
tt(M+l)
u-3
*(*+ 1)
(2tt
- 2)
[MÂ.
is
[This is the coefficient of l/# in the expansion of

therefore equal to na n if
2o^" =
TT
log(l
2.9?),
2a ny n = - J log (1 - 4;y),
Hence
\,v(l
-z)\-
n
(l
TVzp. 1903.]
-Zx)-\ and
where y = x{ 1 - .r).
or a n = 4"- 1 /??.]
An alternative way of stating the result is to say that the sum of the
terms in the binomial series for (1 -|)~ M is equal to the remainder.
30.
Extend the method
of Art. 55 to
y = a*>x*'+ a$y? +
there are two expansions for
Shew
where
31.
also that
n is
the coefficient of
y^x^+x}-,
l/.r
in the expansion of
+ ^2 = a,y + (ab + c)?y 2 +
we
find
It
if
form
. . .
is
easy to write
in the following cases
down
if
# + ff2 = 2a,y,
1
a 1 = m, a 2 = wl (2^-l)(2m-2)/3
where
33.
. . .
a particular case of the. last example, shew that
.?!
If
if
then
32.
<7Oi
W
As
of the
prove that
,r
first
l)
!,
- 2) (3m - 3) (3m - 4)
(3 w
etc.
the general forms for the expansion of
Differential Equations.
34.
If
are two power-series which converge for |.rj</i', prove that the differential
equation
,/--,,
has solutions of the type y A + Ai. + J.y-+...

\x\<R. Here A and A\ are arbitnuy, while
conil>iri;itions of J
and J,.
(
whirh also converge for

.!._,,
.1..,
^J 4
,...
are linear
LMPLES
vni.|
17
r,
the
.'e
coii
,,-t
-f ...
we
*B
i,/;
r<R
where
\\,
and
that
/;
.i,,j,
..v+c*-;
u mob
that
if
tmvergM
Witli
35.
.!:
and
J^
/;,,|<J//-",
lin:
th-'
.1,
i-t-J,//
/;
//,,
tind
ti,,.,,
Thn-
#,=
if
+ .!,'/.
-r.
</; and
.-
lind that
\v.-
1.
.verges
til-
differential
<juati(n
tli.-
type
-...),
t(t-l) = tpQ + q
ie
inultiilr- ot
/'
[If
and
other
the
arl-itiai-y,
erti'-i-nt8
th.-
..f'
we
<|ua<lrati<-,
lind
-{(*-l+t)pl +qJA n-l +...+(tpn +qn)A

wliidi
!*
may
being
thr nth-r rM.t
is
is
J,,
with
i-mnjiai -d
K-l+T)^+...+(l-fT)^},
re S
we
If
= \t-t'\,
tak.-
/>
r = \t\,
\p,,
= |ôl
that
il
that
1'.!
.-
;us
<
as in
/,',
that
shall
have
tht-
-MI
"(uadiat
\\i!l
f-und
].
in
!'
'hat
y=y,,+
an integer. wh-n d(
(+t-t') vanishes
|*-blâ, \y-y \^b,

and
J0^8, we
=r.
COnV<
i|tinn of tin- i-haiiL'es
Sup|iose (hat
l.-n-r
nicditicai
in-tli-..i
tip
36.
A,
lin,
6
with
Bp =
u>8.
Fur-
o.
= 1.
\\
POWER
176
Then y n approaches
SERIES.
as its limit a continuous function of x, say
dv
satisfies
dx=f(x
\f\<M,
which
>
[PICARD.]
\f(x,y')-f(x,y)\<A\y'-y\,
and so in particular f(x, y)
M and A are constants
77,
for sufficiently small values of \x-x \.

[It is assumed that for these values of x,
where
[CH. VIII.]
is
a continuous
function of y.
Then
-y |< M\x xQ
and so, if c is the smaller of a and b/M, the

\,
within the prescribed limits, provided that X-XQ <c.
Suppose now, for brevity of statement, that x>x& then
|yw
values of
yn
fall
ly-n-yl = r\f(x,yn )-f(x,y n-i}\dx<

Jxo
and
I
Thus, by integration
and generally
yn
#1
we
- .yo
77,
< M(x ~ ô) < Me.
find
yn ^ \<^MA
Thus, by Weierstrass's rule, y

their limits
n~l
(x
- x^ n <
and (similarly)
dx
g = lim^
Also,
because f(x, y)
is
continuous as regards
yJ]
-^
MA n ~ cn
1
converge uniformly to
CHAPTBB
IX.
VKSTH IATK
Tl;I<;<>N<>.MKTl;irAL IN
66.
and
Expressions for
<-in /<M sin
tf>
)XS.
as polynomials
in c
\\V have seen (Art. 65) that
+ } r^cos 20 + ^cos 30 +
cos
Bui
f/ .-) = _[(
wln-n-
y = 2cos0; ami,
arr.in-vd in
j.o\vi-rs
of
It
is
y _r-)+^/-y-
'
<
,<'
the latter series
t'urtlu-r.
r,
2
/>2
without alteration of
7-^)3 +...]
may
be
re-
valu.-. jn-ovided
that
()</-:^|.
coefficient of
in
/'
(/-y-'-'
/ /
becan^'
)"
+ 7;-^T 7(
"/.
i"
+2
,
n />2
...
+ ---+<''y-
contain no terms in
Thus
"
,
//I
//
(f/
(71-8).
innnlxT of tt-nns
1
lciiiLT
ritlu-r
i(//
*!
Of
<
/'
1.
= y"
4-
+
l.s.
..
2
/?
y""
is
r\j session
tin-
evident that -cos//#
therefore
2
*
r*.
the
TRIGONOMETRICAL INVESTIGATIONS.
178
we have
Similarly,
[CH. IX,
seen that
rsinO
= rsin#+r sin
2
1-ZrcosO+r2
Hence we deduce that
the coefficient of rn
is
7r
~l
in the
sin $
series
+ (ry - r + (ry - r + (ry - r +

2
2 3
2 2
)
Thus
where the number
of terms
is
either
$n
or $(n
that this formula can be 'deduced from the last
+ l). We
by
note
differentiation.
both cosnO and sin7i$/sin$ are

in
of
cos$,
polynomials
degrees n and (n 1) respectively. But
It is therefore evident that
some purposes
for
of
nO
more useful
is
it
in terms of sin
0.
This
we
to express the functions

do in the following
shall
article.
we
Before leaving the formulae above, it is worth while to notice that if

write y = t + Ift, instead of 2cos0, then 1 -ry + r2 = (l -rt)(l -r/t).
Hence
and
log (1
ry + r
2
)
= log (1 - rt) + log (1 - rjt) = -
rn (t n
+ t~ n \
from the foregoing argument, we get the algebraic identity
so,
+ t~ n = y n ny n z + ...
as above.
Similai ., y ,
and
we
so
67.
find
~r. =
n~ l
Further forms for
-(n-2
cosTifl
n -3
+ ...
as above.
and
In the formulae of the last article change
y = 2 sin 6, and we find

- 1 )m 2 cos 2ra0 = y 2m - 2my'
(
2 "'
=
in case
is
even and equal to 2m.
--+... to
to (J^r
(m + 1 )
0)
terms,
then
Bir
-1
njiial in
M.l.l aii-1
iNm<2,,,+l)0 = y
<
M(2m + 1 )d
(-l)
c<
llo\vr\.-r.
arm
it
AND 006INE8 OF Mil.Tiru:
BINES
66,67]
tli.-sr
lnit
l.i-.iically.
->
+... to
(m+1)
terms.
more ele<:
pe \vli-n
nding powers of sin0; of course
tn
litlicult
iln-
rearrange
instnirt i\v to obtain
is
it
>
nnul;u' take a
f.
sji.M-ially
-(2m-l)y
asy
I7:
+ l)y 2m -' + ...to(m+ l)terms,
wording
not
i-
-2n> -f
<
+ -<:>/,<
2 'M
\\<;i,E8.
ill.-
ezpreeeiona al.
n-ults in another
way,
in
\vrit-
that
966
<(
//H,
nîlT
/'//*
tin-
//
//
polynomial of degree
ha\-
\v-
anl w- have
sin H.
.'
\\-.-
it'
or cos
//rt
//
fxjiri-svioii
'
in X,
îvrii alx>ve for cos
2m6,
can be expressed as a
containing only even powers; thu^
in \vritr
cos
ronxtaiit t.Tin
tin-
\\.'
Il'
?>
=1+J
l.'iii;_;-
vul.stitutr this
I.
,.'-'
-1
h.-caiisc
expression
* 4 4-
4-
A nx n
0=0 gives CO8 1101, 050.

in
th<*
diti'.-r.-ntial
equation,
tin.l
ire
.2.
.l..-f:{.
4.
/--'+.
..+<,/-
and
so
H
\\ht-n
L__
.1,
ii!!<.
//
oo6ft0>l~
r-
42 )
k
.
t
i'
Siinilarl;.
of degree
__ZV
/'.
tin<l
tliat
sin
nO
which contain^ milyodd powers of

sin 710=
4-...-I-J
.r
is
:
a polynomial
tl.
180
the
first coefficient
being determined by considering that for
we
Hence, on substitution,
+n
Thus
A x+5
3
(nx + A 3x
3.2.^ 3 +
dx
[CH. IX.
+A
= 0,
find
.
x5 +
-l)?i = 0,
x*
+n(n- l)A nxn
+ A nxn ) = 0.
5.4.^ 5
giving
to J(7i
?i
being
+ l) terms,
oc?<i.
To verify the algebraic identity between these results and those of

Art. 66, consider the case n = Q. Then Art. 66 gives
or
cos
Change from 9
cos
60 = 32 cos 6 0- 48 cos 4 0+ 18 cos 2 -
to (TT
1.
and we get
0),
66= 1 - 18 sin 2 + 48 sin 4 6- 32 sin 6

6 2 2 -2 2 )(6 2 -4 2 )
62
6 2 (6 2 -2 2 )
2
=
^sm 4 -- (6
sm ^
l-^sm +
~6!
.
-^
in
agreement with the formula for cosnO.

Again, consider the case n
=l
we
have, from Art. 66,
cos70 = 64cos 7 0- 112 cos 6
or
Hence, changing
to
TT
+ 56cos 3 0- 7 cos d.
we have
0,
sin7<9=7sm0-56sin 0+112sm 6 0-64sin 7

3
-l
v
- 7(7
2
2
)
7(7
-l
6'
)(7
-3 2 )
differentiating the formulae just obtained for

cosnO, we find
By
to A(?i
when n
is odd',
and
sinnO and
+ l) terms,
--
AND 008INE8 OF
SINES
67,68]
Ml
AV-LE8.
I.Tll'l.i:
1-1
BID
3T
to JTI term-.
when n
is
iv;ulrl-
Til'-
that
>1|. \V
in;iy
+ 80 sin 4 0-64 sin 6
co70/coB0^
an.lth.it
i(>0/co80
::i'>iii
;unl '"iiiuif tli.-sr formula--
68.
til"-,-
conditions on
ran
J,
tin-
i-,.ino\-<-i
//-T-:^>in 5 ^f
::
the last article are resti
tonn of
in
An
,,f
tone obtained in
'I'd'-
c.-i-tain
ly
\\ith
any
L.-t
//.
us no\v
st-e
il'
these
\vuy.
the &
.iniplr
which was proved to terminate and to represent cos??$, win -n

//
is
,-.
\\
/'
ifl
au odd integer, or
ninatc.
and
is
It
is
natural
not an integer, the series does

to consider whether it is con-
to investigate its >inn.

The test (5) of Art. 12 ahewa at once that the seri
.Tges
X =1; and so, as we have explained in
al>solut'-ly when
,
Art.
the
-")(),
it'
so,
series
l'llnws that
It
as on
17!,
j>.
From
it
converges
is
it
6+
7>
for
nO=l
//
is
U/nifan
values of
?j2
6,
and
# = 0-
of the
form
sin
= 0.
for
cos
and
all real
follows^ that
= 0, we have y = l,
Thus .1 = 1, 7^ = 0, and so
w
! >/
the equation
+
~7^
J.cos
\
nt>
continuous for
satisfies
this equation
,i1,..,/
-,')(
-**+%!^s*-...to
oo
any valu
This
ol)\
constants,
Equations, Art. 8).
:ul)itr;uy
.juiitiiMi
no other
solution
;m.l
can
since
l>e
tin-
found
solution contains tiro
(Forsyth,
Differential
182
[CH. IX.
Similarly, we find that all the other formulae of Art. 67 are

valid for any value of n, provided that x
1 and that the
series are continued to infinity.
\
<
however be noticed that the third and fourth series

are not convergent when x = 1 but they converge absolutely
for x
x ^k
1.
1, and uniformly for
It should
69.
<
<
Various deductions from Art.
We know
that
sin
nO _
we can
67.
write
(n odd)
sin0
= ^ + ^2 sin2 + "- +n-
sinflcos
where the
coefficients
sin"- 0,
worked out in
form at present.
are the same as those
Art. 67, but are not needed in
an
explicit
the left-hand side vanishes for
Now
(n even)
0=Tr/n, + 27T/71,
37i>,
...
so that the right-hand side (regarded as a polynomial in sin 0)

must have roots
sin
When n
are
is
sin
must be (n 1) of these roots which

and these can be taken as
TT
if
is
m.
sm0=+sin-,
But
+ sin (Zir/n), + sin (3-Tr/n), ____
odd, there
all different;
.
(TT/TI),
2-7T
even, there are only

TT
^
sm0=sm-,
.
,...
2) different roots,
(?i
2?r
,...,
Thus we have the formulae

(n odd)
sin 710
sin
sin 2 0\/
sinVV
sin 2
sin 2 2a/
'" [
X^
(n even)
sin 710
~n/
sin 2 0\/
sinVV
where, in each case, a = 7r/7i.
sin 2
2
sin 2a/
"' [
sin 2
namely
SINKS
68, 69]
It
\\|,
<>siNi:>
compare these with the
\ve
ii
.r
\I
I.TITI.
I'-TIM*
explicit
^T
Ja
In
8m-.
cos nfl
,,v
8ini(n-2)a'
way w- pn.v-
-imilar
-^y
sin
= /.1
l
identities
tin-
0\/ ~ sin 2 \
sm-/3/\
P = w/2n and
ih tin-
iI^8-
<"/'/
Again,
if
7,
/8
appear.
see that
'
sinV-2))8'
ri
\\v cMHi>itl-r tin-
we
-+
f
;
---
în 2
l--^-o7
sm-(/'L
multiples of
fonn>
^3^ +
",iii(n-
f^
...
an*3/8/
only th-
-ii.-'
I
where
"si,Â
/.
^m'0\
,>nO = (\- sm-0\/^
B (l--v-.7~b
4n<
we can
+-
<'>7,
8in"(-l)^
I'm-muK-u- of
Art.
()(i,
it
is
that (cos 710 c-.s ,,\ may

-x]>r^sed as a polynomial of degree
;/
in cos^, tin- term of hi^hrst
tindegree beinî
\
cos
iv
if
zero
(cos ?i0
!
,,
-2-//\, 47T/'\,....
tin-
factors of
wli.-rr
It
is
the polynomial
of theXprefisiODfl
<lit}-r.'iit
,.,
as befon-.
.-asily
s.-.-n
H_,.,, S
i-
u^--.l
in
will
(jurstion
(\2a), C080
cos(X4a
.....
f<>r
that the n ditirn-nt factors ean
>.<'-
ooe(X+2a
1>-
tak-n as
''
cosO-cos(\-f -2,
because ooe (X 2
H.-ii,-.- \v,- ha\- the
(A+2(n
i'l.-ntity
n-l
cos/-
x(X+2
.:
r=0
If
\\r
\\i-itf
in
this
2
be
form
-X]IV>-MIM wt- ha\r

II
-'.
184
[CH. IX
with a change of notation,
or,
sin
2^11
nO=
+ TV*).
sin (0
r-o
< <
But the sign must be

a, all the factors;
because, if
are positive, and it is easily seen that both sides change sign
together (as
passes through any multiple of a).
70.
We
as an infinite product.
Expression of sin
have seen in the last article that, if n is an odd integer,
sin
71
\
/
where
a = 7r/n.
Thus
if
To
we
this equation
we
T<7
= 6 we
n<j)
write
have
r=l
we apply
the general theorem of Art. 49
have, in fact,*
sin 2 (0/7i)
_02_
^4r
sin 2 (r7r/7i)
because rTr/n
than
less
is
Now
JTT.
'
this expression is inde-
00
pendent of n, and the
series
^0
r=l
/4<r
is
convergent; consequently
the theorem applies. But we have

lim n sin(0/7i) = 0,
r
hm
'
.
sin
Consequently,
The
special value
= rhm
=
-Ô = %TT
II (I
leads at once to Wallis's
Theorem t
7T
TT
We
from
see,
224466
by differentiation or otherwise, that
to ^ TT
sin
xfx decreases as x increases,
thus
1
Consequently,
> (sin x)fx >
2/Tr,
if
sin (rw/n) "> 2r/n,
if
<x<
< i n.
jr.
Also |sin0/n|< |0/w|, for any value of 6; and so the inequality follows.
THE SINE
69, 70, 71]
Art.
l'"nnula
th-
combining
I'.y
PRO]
MM
i'.r
witb
fJ
the results of
ee that
''--""'
,
ifffr
to
\\hile
\v..i-tli
:,.-ijis
an
Kin-fly to
in-
ompl.-tf
M
]
given in Bonn
for
'
0=nr,
1,
ur^-d that
is
it
>ii\
fj
the
-*0
i\v
th-
hut .-\artly th-
* /
only suggest* that

:iii"t
iian
the
th.-
is
pr..(bi,-t
more
q negative factors),
Art.
fsee
L
is
pp'iif
an.l
41,
when
A(/<--l),
= sin</>,
'f
jipplyinj;
h>kl ôod.
the thi-Mi-em of
An
additional
follows that the values of sin(rr

tlmse Nvhen r ranges
it
ti.
but
thf iwris" "idt-r.
-'r
t:
sin^^y
SLâ
pa
\v-
of fin.
>
may be m
8in 2
if
the (lanî-r
dees n..t
..ut
J/-t<-.-t
'/'-f-1)
in
M<>il.-y
1^).
tin-
his risk
.>)
by Haiku.-- and
Lri\-i!
I
to cases
some knowledge
thf..- lin.-s inv..lv-s
i. it
j.i-.M.f
_u">"il
have Already piint-tl
\\
and
tl.
0^=-.]
i
t<
--ry
ff) /**]?).
Vf/ \ ^ /
~i
A sat
ii"t.-l
is (say f>
be of the form a 6 sin 0/0;
{jossibly
--
l-^-j
n-L
thi-
/M \
In this (.(niH-xioi). it luav be

/
6\ /
f) \
),
and th
into factors (l
(l-\
pi>
separate
so that
number,
ival
(l-^-rJ may
11
//-
we
would apply equally to
s.r
'
applv the theorem
h'
:.
\\-
appeal to get
whi.-h eonti;idi-t> thf
..btaim-d
n-siilt
before.
<
'
i-
that
thf i!if<|uality
be greater than A-
__"
'_,
and
it
i-.
<
in fact,
is
n,,
imp.
lon^.-r
-ible t"
true,
since
nr/n
may
<>nstnHt a n divergent
oomp
71.
Expressions for cosO,
cotfl.
Tin- reader ^b<.uhl tin<l n<> dilliculty

illtillitr
\\'r
pl-odlK't,
hive
in
l'nlln\\ ilio-
fed
<
ami
lilli-s
nl'
in I'xjT.-îng
thi'
Art.
cos
wb.
tb<'
*= "ill.- .T
i^
"'/'/.
cosO as an
p!-fCC(li]i_
TRIGONOMETRICAL INVESTIGATIONS
186
Here the comparison-series
is
20 2 /r2 and
,
0=
Alternative methods are to write cos

to the product of Art. 70
rearrange the factors.
To express
cot$,
n~
so
we take
if
1
4.0
-=cot
1_
^
V
is
an<^ appeal directly
/~S~?
in that product
1}
.,,
we have
sin(0/n)cos6n
n smo, r7rn /
.
not a multiple of
si
TT.
To
this identity we apply the theorem of Art. 49,

first step is to obtain a
r
comparison-series
^M
Now
and
(n being odd)
\\,
sm 2 (r7r/^)J'
logarithms and then differentiate,
the identity
assuming that
S1 "
sin 2 (0/7i)-]
is
that
0^
sm = n sin- II r,1
n r=
and
the result
or to write $7r-@ for
we know
[CH. IX.
and our
(see footnote, Art. 70, p. 184)
sm2 (T7r/n) > 4r2 /^2

.also
provided that r
sin (0/7i)
sin 2 (r7r/%)
Thus,
<
/7i
- sin2 (0/7i) >
for all values of r such that
n>
< \n
2r
(4r
>
sin(0/7i)cos(fl/?i)
sin 2 (r7r/7i)
sin 2 ($/?i)
)/
and consequently
if
Since 6 is fixed, there may be a fixed number of terms (i.e.

a number independent of n) at the beginning of the expression
for cot 6 for which the last inequality is not valid.
But for
the rest of the terms
which
To
is
we can
use the comparison-series
obviously convergent and independent of n.

we can apply the theorem of Art. 49
these terms
the fixed
number
of terms preceding
them no
for
special test is
necessary.*
*No test of this character is requisite for any fixed number of terms; the
object of Art. 49 is to enable us to handle a sum the numher of whose terms
tends to oo with n.
71]
<'T\v.i:vr SEB1
in:
"
'
li,,,
'
I!,,
ocri
i
.
187
'-/tan
J)-J
n/ 9
n\
8in(0/7i)co8(0/n)
sinVir/n)-8m*(0/
'I'
-0i
This ivsult can also
derived
l.e
01106 lYoin the
at
product for
u>e the
Logarithmically; we ^hould then
>in(>.
ty differentiating
theorem at the end <!' Art. 4s
In exactly the saint-
way th-
to
jn^til'y
identitir- (p.
ÎJ'S"^^^
and
the operation.
1*3),
Cod
dd> n
(r
-a-^s="S
2
nsin 2 (r-
w*
apply intr An.
.11
I'.',
two
the
l_y_l_
r
-f-
a~TrV
or
s<-:
4.
and
>-..=
f+:
uis- tlin-c
i
is
no
'
2~7'
dilliciilty in d-lucin.i;
+ ... =
traiisf.,nnati.ns
i'h.
IV.)
r];il,o|-at<-
(altliouM-h
,,r
being
('liry.stal's
greal
Algebra,
lar^c minihcr
ju<tili.-il
4*
interest)
ol'
vol.
l.v
dflails
1'urtluT
content
fr'iii tlu
-p+-
-J)(i'
th.-
the second of th--c
use
w.ull
oursi'lvi'S
2.
us.-l'ul
ch.
Ait.
of
analytical
l-al UN too i'ar
with making
XXX..
in
and interesting
which
ident;
\\'.-
ati.-ll.
<
i
will
l.c-
to
i'ouml
TRIGONOMETEICAL INVESTIGATIONS.
188
[CH,
EXAMPLES.
From
1.
the formula (of Art. 68)
sin(mO)=msin 9
obtain a power-series for (arc sin xf
==
ai
6T
2.
'
From
2^\l
the formula for sin
n9
to
oo
namely
3 2 /2. 4"5"^\1 2 "^32
+ ...
-sin 3
o2 /
given at the
2.4. 6
end
of Art. 69, prove that r
cosec 2 6 + cosec 2 (0 + a) +
. . .
+ cosec 2
+ (n - 1) a = n 2 cosec 2 nO.
}
[Math. Trip. 1900.]

3.
If
odd and equal to 2m + 1, shew that
is
tan 4 (r7r/7i) = ^(7i-l)(7i2 +^-3).
[Math. Trip. 1903.]
4.
If
odd, shew that
is
'^cosec^nr/Ti) = J(%
r=l
n = abc...k, where
If
If
- 1).
extended only
to
a, 6, c, ..., & are primes, shew that the above sum,

values of r which are prime to n, is equal to
2
(a -l)(6 -l)(c
5.
Shew
-l)...(F-l).
#2 + if =
that the roots of the equation 3?
AV7 = 2sin(^r),
2sin(f7r),
2 sin
[Math. Trip. 1902.]

are given
[Write ^ = 2sin0 in the formula of Art. 67 for sin
/=
-l)
To distinguish between the two
2sin(f7r) are both greater than
6.
Shew
or
7(3/
Then we get
70.
^=^7(^-1).
sets of three roots note
1,
while 2sinf;r
lies
that 2 sin (f TT),
between -1 and
that the roots of the equation
are
2cos(y7r),
2cos(f?r),
2 cos
(fir).
[The formula of Art. 66 for sin 70 shews that
f - bf + 6y -
are the roots of
The
substitution
7.
Shew
and deduce
y=x+Z
that,
if
that, if
= 0.
leads at once to the required result.]
\x\ <
y=2
1,
cos 20,
by
(fir).
0.]
iMl'LKS.
IX.]
the series
.in
\vhi--h
Shew
is
;i
that
0<*<*,
if
all
,'
the greatest
int.
_'
IN
-r
tl
to
').
Sum
9.
terra* for
series
tin-
fn.iii
.f
multiplr
RISKS.-
the series
-)+^( Kin
-i(
!si
'
[~'l!u-
in
MUD
is
smalK-i-
tin-
anl
>f
<>__.,<
if
~'^
7T.J
it
"'(- + /',)(/Mtl
II. -li..
sin
(qâin (a*r) ... sin (a^)^
'~sin(/>J 7r)8in(^7r)...8in(
It'
S6-2a^
>h-\v that,
tin-
if
:
.
:lu-r, if
lim(// r)
= /\
;io-rnt
= li m
hl ,t
_x
/,,
wh.-r.-
~i
-J
12.
/'.
caniict tin- first ]n-olu.
i>
rxcliuled
and KM
-C
**
Why
II
\\- hav.-
n. ,.,..
that
iinj.lirs
l.r
8 " iL
=1 -
[K
8
-fi?^-
written without an i-xpciu-ntial
fa
'
Sh.-w that
(l-.r)(l
tin-
terms
-fr)...
= co8(|-
aj]>l\
MX. Jl.
in
j.air-
ami
'h.
VI.. ..Staining
r(-)
4.1
\viit.'
13.
OU<
tin-
Prove that
i>r..lm-t
f.-rm
n fi
L
_r
-(
-.r)}
-**- 1= (nTT+jr^J
sin(|7r).]
:
[Ki
190
14.
Determine the limit
where
15.
/x,
tend to
Shew
of the product
in such a
oo
sin (x
= cos x,
TT)
and deduce the periodic properties

Deduce the
16.
17.
and
of the sine
that
sin (TTX)
cosine.
from the equation
product for sin.r
infinite
fi*
oos(2f)c&
Jo
of the series for cos (2xt) in
Shew
x and cos x that

- sin #,
sin
cos (x + ^TT) =
ran (&)*=* 2#
by means
Km (/x/v) = k.
that
way
from the products for
directly
"
irx II
_ QQ
powers of
7r[cot( 7r^)-cot(v 7ra)]=

^J
cot
(TTJ?)
cosec 2 (7r^) =
oo
^^) - ?o cosec
In the
18.
first
(TT^)"]
_l
last
L\(
%/
lim
limf 2 2
L m= _v = - ^
and
[We
in the double
Shew
=2
-j
ri
3
W)
(.1?
oo^^-'
n=-V X
v _ô
19.
= -+ 2'
-n a-n
[EULER.]
^.4y
is
excluded,
example that
I/->ao
and
(Art. 68).
and second equations the accent implies that n =
Deduce from the
where
=2
---
-co
sin
F/ - #\
- )e*/ n H
1
-
7T
[CH,
summation
7
(a?
all
ll
- m)(x - n),1=-^
J
,
values
m = n are
excluded.
that
sec ^
r^2-CTZ^ + yjl a*"'"'
[EULER.]
can get the expansion for cosec# from the identity

cosec x = cot
( Ja?)
- cot #.
The second series can be derived from the series (equivalent to the
by writing x+\ir
for
x and grouping the trims
differently.]
first)
iMFLES
i\.
I
Shew
20.
that
-lfc+A+A+.~
I
Deduce the value
of
l)-'
n=0
being excluded from
21.
l'lo\e
Miiimiation.
tin-
?("
If
22.
term
the general
in
1896.]
[J/.-
tli:it
<
-6-7^7 + "can be divided into partial
form
tlic
1\1=0,
2"
tli. -11
1.1-'->t(</7r),
-B
all
\\li.-ic
23.
,i
the ntunbers
a are supposed
Find the sum of the
ditl.-r-ni
from
xero.
sci
-0
being
57i-2
[\
H<
L
iv\
'hided.
1\
"m/+
6^^57?00
Find
24.
vain-
tlu-
w -
-1A_
3
-I)
;tnd dftt-nniiif 5,
25.
We
C so
that
have seen <rh.
(i'/'
V!..
(271)3
may be
Kx. 11) that
quadratic
if
rr(nr(QT + rr^noT rr(3)r(/n-'

f/)J
Lr(i+oJ"*"Lr(3+o'J
th-n
Now
M.I
^l=
if^>A.
J{L
=
--
l.v
Ait. 71,
We
lind
^ = ^(^-9), ^
^fcW+fcW^-S
i:
CHAPTER
X.
COMPLEX SERIES AND PRODUCTS.

The algebra
of complex numbers.
assume that the reader 'has already become acquainted
with the leading features of the algebra of complex numbers.
The fundamental laws of operation are
72
We
x = g-\-irj,
If
x + x' =
then
x'
xx
x'
+' +
= g' +
i(>i
irj',
(addition)
*)'),
=
j/),
g'-\-i(r)
= gg' w 4- i (gri + g'rj),
.
En
(subtraction)
(multiplication)
'ti
laws include those of real numbers

and that these four operations can be carried
out without exception (excluding division by zero). Further,
these laws are consistent with the relations with which we
It is easily seen that these
as a special case
are familiar for real numbers, such as
= (xy)z.
Thus any of the ordinary algebraic identities, which are
established in the first instance for real numbers, are still true
if the letters are supposed to represent complex numbers.
It
natural to ask whether other assumptions might not be made

Thus the analogy for addition might
is
which would be equally satisfactory.

suggest for multiplication
But
this
then give
whereas
is
inconsistent
##' = {f' + i7;i/'.
with
tin-
2.r=2
x+ x= + +
relation
+ i(0) = 2f,
t'
x+x = 2x,
since #' = 2
would
[72, 73
Mill
H.sXUIIIpl
MI
laws an
tive
tin-
to fix the
Uw of ranlttp
iuulti|>li<-Hti'.ii,
tliil'
MIlllllMT*.
be shewn
r.l.
(t*
II..
'
rail
It
*-\-2) that
\>\>.
mad*-
jitjons
a8iii:
by
we
i-lse
we
are forced
'he produithout
\\
t-it
ht-r
./-,
Argand s diagram.
73.
Tin- reader
douhti*
is
I'amiliar
with the usual represen-
tati"ii d' tin-
;nt
wit):
(^,
f
19.
r.ut
may
it
i-nnvi-ni-iit
!>
inv.-
t<
ûinmary of
ln-i.-t'
th.-
tod
If
we intnxlucf polar
coordin.-,
./
shall call r
\\.-
l,y tli.-
it
symbol
with the notation

her
\V"
+./
call
From
tlu-
or
H the
/((.-<
^-f
\v.-
can
writ.-
si:
/'
>/
<>l'
mo
times also callcfl the
denote
</*w of a*):
\x\.
'I'his.
i-
33
(it
is
some-
and wo
shall
(juitc ccnî^tcnt
previously: for if x is real, \x\ will

according as x is pov-itiv.- or negat
n^,-,!
./.
wrgwneni
<>f
diagram the nieani
./
it
is
snm- times
:-
:.
called
the
ident.
COMPLEX SEKIES AND PKODUCTS.
194
we draw through A, AB, CA equal and

C are respectively x+x' and x x.
If
then B,
x +x
A'B
= x'+ x
is
parallel to
OB<OA + AB
Since
we have
the relation
|
and
x +x'
x
similarly,
x'
\
Again, supposing
# 4-^
x
also
or
OB < OA + OA,
x
|
x
|
>
x'
+
+
x'
1
33).
|,
x'
1
I.
.
\
we have
OB>OA-OA'.
or
x'
\
x'
^>\x
It is easy to
(Euclid
<
<
OB + AB>OA
and so
OA
OA < OA,
Thus,
OA
',
parallel to
The fact that
by the geometrical theorem that
represented
equal and
is
prove similarly that
<E
2x
and generally, that
.
\
These facts can be proved algebraically, thus, consider the
and write
R = \x + af\,
J
Hence
this
is
so that IP = (f +
= r2 H-r + 2(^
2
2 /2 = 2 (r/ (r + /)
Then we have
and
[CH. X.
ft
certainly positive
s positive,
/2
if
f ) + (^ + T?') 2
first
inequality
^' + t]tj'
is
or negative.
zero
But
if
we have
so that
the sign of equality only occurring
Thus
if
in all cases
or
and the sign

which
74.
If
is
'
and
+ >;>/> r
represented geometrically by supposing that 0^1' falls along OA.
of equality can only appear
if
>?'-'?7
Multiplication; de Moivre's theorem.
we
multiply together the two numbers
x = r (cos
the product
is
x'
= r'(cos 0' + i sin 0'),
found to be
xx' = rr'(cos 6 cos
= rr
+ i sin 0),
& - sin 6 sin 0') + irr'(cos
[cos (0 + 0') + i sin (0 + 0')].
sin 0'
+ sin
cos 0')
Dl KOIVBE'S THEOREM,
73, 74]
Thus
tlio
absolute valu- of
/'
/T';
I-I*M*'
and
tli-
In pariicular,
to
of
ai'innm-nt
/'=!,
if
O + tf,
is
././'
or
i.
sum
tli-
product ,/<>->
tin-
195
the arguments
of
f)
4-
-in H
is
equal
h* sin (0+0')],
r[oo
o
therefore clear that
is
It
20+ % six
ami so
if
=cas0 + -isin 0,
/- l
hisin30;
on, as indicat.-d
to
1-
\^
pial
t<>
cos(-0)+z sin(-0);
th-
in
22.
if
i^
(cos 6
To
\vli<.l.-
any
intrrj.r.-t
''
wr
(f
thus,
I.
is
>?0
//
")"
arm
f
r n. -Dative. \\v
<
sin
/?
0.
wliol,-
is
still
have
nun
true.
./<-^''sin</>),
p (cos ??<+/
/',
and
if
whciv
= T = cos
///
indi<-.->
)"
\vh.-iv
agree that th- law of
Qenoe
si
nuinl>T. ]M>siti\-
any whole
sin
)r-r..x //-f)+ 'sin

^
,><;>
and ty
//<t)4--2/"7r.
nunil>.T.
'rh;i-
-cos
|
thus
I.
has
Th'
/,
casr
j(m0+2fcr)l
litlrn-nt
/
()
+ifflu
values, uiv.-n
constitutes d-
[^(m0+l
l.y
taking fc0,
I.
196
large
number
Morley, some
to the
of elegant geometrical applications have been given by

will be found in Harkness and Morley's Introduction
which
of
Theory of Analytic Functions, chapter
The
1.
[CH. X.
and
triangles x, y, z
x', y',
z'
x', y',
z'
1,
1,
II.
We
note a few samples
are directly similar
if
If the triangles x, y, z and x', y', z are directly similar, any three
points dividing xx', yy', zz in the same ratio form a third similar triangle.
2.
The conditions
3.
=
1,
(where
a, ft,
y are
and
zx-\-yw,
a,
1,
real
1,1
numbers) are respectively the conditions that

x, y, z, w should be concyclic.
x, y, z
should be collinear, and that
General principle of convergence for complex sequences,
75.
Sn = X n +iYnt we say that the sequence (Sn ) converges,

when both (Xn ) and (Fn ) are convergent; and if Xn -*X, and
Fn HF, we write Sn -*X + iY.
The necessary and sufficient test for the convergence of the
If
sequence (Sn )
e,
however
is that,
small,
corresponding to any real positive number

find an index
such that
we can
|#n-m|<e,
To prove
\Sn -Sm \^\X,n
so that
Sn -Sm \^\Xn -Xm
and
Now,
find
we observe
this statement,
if
+ \Yn -Ym
-S
m \^\Yn -Ym
\Sn
\,
\.
are also convergent, so that
we can
to satisfy
and consequently
so that the condition
On
that
-Xm
(Xn ) and (Yn )
(Sn ) converges,
>m.
if
the other hand,
is
Sn - S,n
<
if
e,
>m
necessary.
when
we have
this inequality is satisfied,
Xn -Xm \<,
\Yn -Ym \<,
if
and therefore the sequences (Xn ), ( F,,) are convergent.

Thus (S) converges and therefore the condition is
;
also
n>m,
sufficient.
application of this principle we consider the interpretai sin 6 and a is irrational.

note
tion of t* where t = cos
As an
We
as a preliminary that
|
(cos
$ + i sin
$)
(cos
tjs
4- i sin
\js)
=2
sin
|(0
75
74,
<
OF
i:
OMPL1
<
numl.-i^ which
can lind
lias
so
in
-,,,|<e/|tf
>m.
11
'I'llUS
9)+* 00(0*0)}
0!<e,
<l -
|{cos(an 0)-f
>
//<
and so the sequence
gent
of course to be
limlcosajtf +
/"
<-on-
9)-j-isini
1
thT
'iiral
rfiii.-nilM-i-.--l
that
IN
It
<///
as lini/'
/"
but
-;
limits*
tin-
+ :>/,-)]
*)
l.'
to delin.-
(/.'=+!,
2,
3,
to
Thus
equally well In- regarded a> incIulMl in tin- symbol
tak.-n
-I
th.m.-anincr
must
be
be
attach.
care
to
to
^{..-cit'y
special
may
to<*; for most purposes

r.-diir, - U)
\vln-ii
t)
it
i->
>utlici'-iit
h;it
r.-tain tin- vulin-
valur î\'-n
tin-
l>y
which
Convergence and divergence of a series of complex terms.

"n
It'
il'
w' bave
....
Th. u
iti n
fj
.,
-f
* and
to
sums
tn tin-
nt
s,
OOIH
tli-
sum
It
^L"
is
easy
<li\
COSfflOI] l'r
inn
an
infinite
\\e
n.-.-.l
in
*
:ui
All
iitt<
-|-/y 2
4-
to
...
are separately
...
to oo
div.Tês (or oscillates),
we say
o^-illates).
i^r.-at
\\v
j>
and
tin-
l'>n-î.in_
///.
LB Dimply th
can tind /// such that
manner
like
<m this drtinitinn
no mattci- lm\v
In
+ '/.,+
Z// (1
BtMjurncrs that
\vt-
e,
//
resptrtivi'ly, \vc say tliat
N-(-/V.
"/
that
may
..Main
-ji.mdin^
tin- n-al
be.
th.-
<lrtiniti.>n
l,
.!'
n\
<
rgence of
in
ordinary \\
only absolutely convergent }>roducts, which we shall
oniplr\
]>r<><i
Art.
77.
these values are uiuMjujil. bOMMi

(
must not be allowed
makes la equal to
t
;h
198
[CH. X.
Absolute convergence of a series of complex terms.

a n = gn -\-iri n and if 2|a n is convergent, we shall say that
2an is absolutely convergent. It is evident in this case, from
Art. 18, that the separate series 2 n 2^ are both absolutely
76.
If
convergent, because
|fi.|^
and Ifcl^KI.
2an
It follows therefore that
convergent in virtue of our
is
definition (Art. 75); and by Art. 26 the sums 2 n 2^/ n are

independent of the order of the terms. Hence also, the sum of
,
an
absolutely convergent series is independent of the order of
arrangement.
It
is
perhaps worth while to give a graphical illustration of the method
2a n from
of inferriDg the convergence of
= a + a 2 +.
ovHail +
Let
sn
and represent the numbers
in Argand's diagram, as
ff
0*
ff,
<7fa
below
O-
FIG. 23.
By
definition
we have
Os 1 = 0(T li
Thus
*n*p
But we have
<r H <rp
=
=
S 1 S2
= <T (T2
1
S2S3
= 0-2 0-3,
etc
by elementary geometry.
0"n<7>
p>n,
if
o- no-,
represents the sum 2 an = lima w

Now, by the definition of convergence, we can choose n to correspond
to
so that
<r Mor = cr-<r w <e.
where
<r
Consequently, n can be found, so that
**,<,
centre is s n
shews
(see Art. 75) that

f,
\\hi--li
(li;r_iT;un
An
can
Since
may
In-
!<
tin-
indicates the
alti-rnaiivr
ffiii
A.*ndU nf Mathematics
p-tints
ciivlc
tin-
<>
>f
(2),
if
must
p>n
within a circle of radiiis

irJnw
taken as small as we please, the last result
or, geometrically, all the points 8P
riivli-s
diagram
\.,1.
:.,
*,,
lie
must duster alumt
Imt cannot be outside
with centres
lias
2,
3,
s.unc
limiting
it.
.-.,.
been proposed by G. H.
AlisoLI
76.77|
is
It
what has been said as to series

is not
oonvei
necessary
obviMiis n-Miu
CMIU-S,-
ni'
of ival
199
(X)NVJD
II.
-hut
alisMluh-
series whirl
i>l.-x
M<
although
-*+.
For both
of absolute values
s.-ries
which
and imaginary parts cunvi-r^, by Art. 21; but
ival
by
div.-rjres
Art.
evident I'mm chap.
may
oonvergeni series
is
K\. >)
<
I\'.
altrr.-d
Ai-t.
11.
sum
the
that
M!'
a non-absolut.-ly
derangement.
l.y
Absolute convergence of an infinite product of complex
77.
factors.
Thr
intinitr proiluct
th.-
ii'
at
11(1+
j.rMluct
to be absolutely conIt foil

convergent
"ail
i-
i-
that
onc-f
lid
11(1 -f"
'v
convergent,
F.-r
know
\\.-
that
S|a n converges: and
so
t\\-M
so
also
is
Hut
have
\\v
pi-Miiu.-ts)
and
wmpU
^
tonv*
to
still
am
deck
^11 + a.l
"
m+p
'
(l-
II
th..-
</,<!. we have
if
1and SO
Art. 39
tt
<
\\\
by
11(1
|a |) are convergent.
sht-w that (in the case of complex
Jl(l4-|a n |)
pi-M-lucts
(1+rO
II
IR-fl
since
II(l-f a
that
follows
it
converges,
>
and thnvfoiv
No
1( I
from Art. 39
a n \)
|
gent
thus we can tind
<
"
//>.
>>l-e
s. that
and
"ll'\l
",
)<l+e,
m+l
\
11.
ei Mnall
may
be.
MII tind
11
hat
L+a
>
<
1+6,
-,,4-1
-mall
may
le.
and however larê
/'
may
be.
200
That
is,
II(l-fan )
[CH. X.
the general condition for con-
satisfies
vergence; and so the product is convergent.

It is easy to modify the proof of Art. 41 to shew that if
*an is absolutely convergent the value o/II(lH-an ) is unaltered
by changing the order of the factors.

Pringsheim's tests for absolute convergence of a com-
78.
plex series.*
Of course the conditions
of chap. II. can be applied at once

but since
to the series of positive terms 2 a n
1
= V(&2 + ?n2
|on|
it is
)..
On = fn + ^n,
if
evident that the square -root complicates some of the tests.

tests of Art. 9 can be at once changed to
Of course the
lim
(7n
,\a n
limDn
an
2
|
<
(convergence)
oo,
>0,
(divergence);
but the same transformation cannot, as a
rule,
be applied to
the ratio-tests of Art. 12.
Thus, the condition
n+l
by no means sufficient to ensure the convergence of 2 an ;

because whenever lim Dn oo (which is usually the case), the
above condition does not exclude the possibility
is
n+l
which may occur with a divergent

For instance, with l/a n = n\ogn and
2
series.
Dn = n
(n>2),
because
Thus,
but yet
*
2|a|
diverges.
Archivfiir Math.
u.
[See Art. 11
Phys.
(3),
Bd.
(2).]
4,
1902, pp.
1-19.
we have
TESTS FOB
77,78,79]
It
LB8OL1
can be
therefor*
conditions
tin-
ii/
li.ul />
,.ee)
\\ni\D
\\-\i\
t<i
uliMantially <-<juival.-nt
the
if
condition
first
is
-2s-
and 9st, say
the
if
th.-u
/;
:nid
/i
..l'
/"
Art.
"2.
"'
in all cases of ju-n-ti.-al intt-rost,*
r'nditi.ni "f Art.
But
"
--s;
"
limit to
conditions
lind
*!;!'
Now
s;it
tin-
it
l) n
possible to assign an
is
J?S-
-//
and so the
;(!.
sfc..n<l
C0
hld>. we
.-..nditinn
tind
si.
that
if
that
tli-
79.
It
/'
lir-t
fa'-l.r
must
nr-m iv.
Applications.
i^
/(
t<>
easy
/'
UMii^l'orii!
in Ail.
lim/v-,,>
\-2
(3),
th.-n
tind
\vc
'
win -re
fash point
only
p.
in
rnnlitinns ly writing
ri-iiiîi.-iin'.s
and
to
notice
i-
that
ii,J""'i-' =li ,,,
*-'
\irtur
-o.
>!'
and
jr-*
L'l I -
j.ital's
nili-
|/<
In particular, In
(Art.
l.">2):
no\v
w-
also tends to /.era
us take
-n,
ailiuitx slightly
m.-n-
(8)/(n)nlog
-itions.
that
202
We
after
obtain,
few
[CH. X.
the
transformations,
following
conditions:*
(1)
lim
[
(2)
lim
'n+l
an
2,
(convergence)
2,
(divergence).
7i
> 2,
'
2,
(3)
''n+i
The most interesting case in practice

(5) of Art. 12, where we can write
(4)
^n+l
'""
It is easy to see that -if
(I
is
(convergence)
(divergence).
one corresponding to
^=a
in
(4),
i2
(5)
n+l
from (2) we see that the series 2|a n converges if

and diverges if a
in case a = l, we apply (3), and
1
find divergence.
Thus we have the rule:t
and
so
.|
<
>1
When
a.
the series
2|aw converges if the

and otherwise diverges.
|
than
On
1,
real part
account of the importance of the series
of
jm
is
greater
2an 03 M when
(4)
we
some further results.

An
shews that the series converges
if
absolutely
|os|<l, diverges if \x >1; and we have just
examined the case when \x =1 and a> 1, proving that then
2an # w is absolutely convergent.
But we do not know yet whether the series may converge
proceed to consider
application of (1) above
holds,
=l
(although not absolutely) even when a^l.

will increase with n,
negative, the sequence \a n
ut least after a certain sta^v, in virtue of (o) above; and
for
|fl5|
If
is
*The
inequalities
are here used as equivalent to the two

t Weierstrass, Gea. Werke, Bd.
1,
p.
HmP>l, HtnP<l.
185.
79]
consequent
termfl
Kurth.-r.
to
2(t n x*
of
nee
-i
if
i-
it
0,
On
the other haml.
if
0>0,
to /..TO as a
ti'iiU
argument îmilar
1
linite
that |a n
th.-
sin,-,-
impassible,
see (by an
bo
OftSJ
is
ineivases.
/<
BO
Kx.
Ar
follows lY"m
it
limit
:;.
so that convergence
that
limit,
may
or.
< a =1,
and
tradition (4) holds
prove next that
\\'e
2
In the
first
that
evident
is
ii
place
vali;
|rt n
jiience
I'.
J0n_
(<
finally
!/lk +
*n+l
h-
.f
uly.
we bave, from
Kurth.-r.
obviously convergent
|q-a,.
im-n-ases:
//
istfl
is
it
after a cei-tain
least
decreases as
the
and
(at
1_I^ +
^n
|J|
/'-
"A
(5),
Îk-H,
Thus
a.
-n+l
and
follows, as
it
Art.
in
the series 2|
that
IN.
a w+1
is
|
convergent
v,
\\v
have identically.
ami
ami eon-e.juently,
if
|a?|=
ami X
litlen-nt
\^
t'n
-m
1.
the
00
'
-"
conxerês ami linn/
fl
= 0.
may aU
This result
\\'e
.leri\e.l
t'rm
it.
i-
>
<
Lilt
the
proof
is
little
tedious,
to
tlr.~
ami we simply
limit
proving
l>irirhl>
have "Mtaine.l no information as to the K-haviour nf 1

as a mat
a :_,
(
li
should
unless
= 0;
:lty
in
COMPLEX SERIES AND PKODUCTS.
204
[CH. X.
refer to Pringsheim for the details.*

Of course we have already
proved that the series diverges in the corresponding case when
an
real (see Art. 12).

final conclusion is thus
is
The
_^n_
Tf
aH+1
= l_l_/
>!)
(\\<A)
^_
VS*
<
then the series â nx n converges
absolutely for x
cannot converge for as
1.
a + i^, there are three cases:
If \x\ = \ and
|
>
1,
and
fji
>
(i)
when
(ii)
when
0<a =
l,
except for x =
\.
(iii)
when
Ex.
As
1,
the series converges

absolutely.
the series converges (but not absolutely),
^ a, the series does not converge.

we may
a special example
take the
(Weierstrass, Lc.)
hypergeometric series
which
(Art. 12), in
an
a n+l
so that
Then we
if
(i)
=1
see that, for
the real part of
series converges absolutely

if
(ii)
this real part is

for .r=l
is
(y-a-fi)
between -
positive,
the hypergeometric
and
0,
the series converges except
(iii) if
the real part
is
or less than
- 1, the
series does not converge.
Special cases of this have been found already in Arts. 12
Further tests for convergence.

will find no difficulty
of Art. 18 to establish the theorem:
and
21.
80.
The reader
(1)
that
|
// 2a n
is absolutely convergent, so also is
v n never exceeds a fixed

|
L.c. pp.
modifying the proof
in
number
2anvn provided
,
k.
13-17; the proof shews that (except for /&=!, when the series
is the same as that of
is
easily seen to diverge) the behaviour of 2a,,
1-A*
"I"
And
the
sum
of
?i
(1-AQ(2-/Q
1.2
terms of this series
(1- M )(2-/0(8-AO
1.2.3
is
L^\
1.2. 3. ,.(7i-l)
?i,
(iL +
)'"\
Lz/A
n-l)
It can easily be seen that the absolute value of this expression tends to oo with
and that its argument has no definite limit.
IBEL'8 TES1
79, 80]
It
ho\\,
is,
alt t-rat
ion
prove
ti,
i.re\
/////
vn+l
-f
t' n
+j
little
shall
\\
complex.
I.-ii.
Ahel'fl
-um
the
i',j+ 2
>at
need a
V -
i-
note
vn
'.ill
which
in
modilied form of
nsider the
for
<
Ai
that
to include cases
9,
Writ--
<>i;
of
tl
to
-f
l'n+i.
\Vc
1,
>+...+([T;
SO thai
N.\v (see Art.
Vw+1
2|vn
<
Til) tli'
SO that
lias
>
//<.
And
>-}-...4-('\
'
//
it'
-v
series -('..
COM v !;] it.
is
1
/'
:,
limit
<lctiiiit'-
G-
MS
;/
R-QC
=Hm|t; n
//
,)='',-'
'
and
n-
h-t
\vrit.-
\v
:n.
-'
.
iiat
if
we take
the limit of this
Q
if
Zn
Thus,
we
l(>\-
''.,)+
// is not
if
Ie86
I'
than
.
'..+i|
^'e tind. as in Art.
+",''
".,''.,+
\v- tind
)+...+
:>:>.
S,VW
than the upper limit of
-^
|,
|*t
|,
that
...
|w|,
tind
is
the form of
Similarly,
ti
n-...+(F-i-F
Lemma for complex
ITRFj-F^-f
which
-"j' !-f
al\\
if
[.[
\'
\^
we write }' = (!+!'
It
i- aNo obi lOOfl that the
se.jUriir,
injual to
Beqne&Oe, and that !'- I'
Thus,
ine.jiialit
/,/
:iul
1,
if
ami
//
i/
is
A.bel'fl
m-t !*< than
tin-
than
.
i_
,;( I",
r,,,)-f
>/
term-.
..f
* r -cr|, as r ranges
ujijwr limit as r ranges fron.
\ijp-r limit
tin-
//r
from
206
the sequence
If
G = \imv n
is real,
positive and decreasing, we find that
so that these inequalities are almost the same as
(#)
Vn = vn
and
[CH. X.
those of Art. 23.
Now
It
apply the
Lemma
to the
then evident that this
is
H'Vm+i, where H'
Now
e,
however small
a m+l v m+1 +
which proves that the

This
may
in
less,
absolute value, than
the upper limit of
is
since 2a,t converges,
make H' <
is
sum
can be chosen so large as to

may be and then we have
;
+ a m+p vm+p
2a Bv n
series
be called Abel's test
|< e Vm+1 < e Vv
is
for
convergent.
complex series.
a function of #, Abel's test also enables us to ensure the

uniform convergence of 2a M v n at all points x within an area for which
1^1 and C/i remain less than a fixed value. For
If vn is
F = +*7
Similarly,
we can
0<
arid
vj
tfi.
establish Dirichlet's test for complex series.
0, then Ztt n t> n will

Vn+i\ converges and limi> n
limits.
between
oscillates
2a
n
finite
converge if
In fact
is here zero, and (just as in Art. 20) we can find
(3)
If 2
vn
a constant
(independent of
m)
so that
H'
< 2.
Thus
m+l
which can be made
less
than
by proper choice
of
because
r,H.o.
in Art. 79.
special case of this test has been already used
Uniform convergence.
After what has been explained in Art. 43, there will be nouniform convergence for
difficulty in appreciating the idea of
a series 2/n (se), when x is complex the only essential point of
81.
novelty being that the region of uniform convergence now

= +irj)
q) plane (if x
usually consists of an area in the (
It is also
instead of an interval (or segment) of the real axis.
sometimes convenient to use the idea of uniform, convergence

along a line, which should present no fresh difficulty to the
reader.
80,81]
\inl;M
unaltered,
is
/,,
'
ret;i
tl,
-f+o
when
l.i
M-test for uniform convergence can be
Weierstrass's
almost
ONI
<
//--
/</"/"
irifliin
'/"
.1,
t/nif
/'///
Constant,
<><'/"
wren
oerttwn
>'
///'/
ift)
^.M
co
'
i*
a/nd uniformly
''.^nltitiln
.1.
Abel's and Dirichlet's tests for uniform convergence (for a
Once from
are <>htained at
plex
The proof
I/ i') u
of Art.
(1) can he easily
4~>
IWU8 fnixt'"!'
lis<MiwiiiM
Tli.-
tn tin-
lnt
vn
\arial)l<-
out
not
is
it
n-'ithlii
tin-
to sh.-\v that
f}
,-'
><ma
9epa/ra&
f/
(liU'.-ivntiatiun
<.f
"/
and
int.-î-ation witli respect
falls Ix-yoiid tin-
im-utiou
t<>
<>i'
tliat
BOqpe of this IxH.k:

(wlu-n tin- fun-la-
ini'iital
plannotions ha\;' hccn made clear) the results of Arts.
remain
|>raet it-ally
It
Art.
1^
i-
-till
correct,
when X N
variahle.
c<.mj.le\
Tliere
no ditliculty
i-
remain
lit
valid,
in
wlien
to
often important
is
It
4.">-47
unchanged.
evident a No that
is
com-
art;
last
iiioliti.-d
./
fluif
contfawHU
tin-
seeing that the two theor.M,

re
the functionc<m}>lex.
integrate a complex function with
rasped to a real variahle: in particular it i- usrful to consider

the mean value of a continuous function /(./) along a circle |.r = p,
which
defined hy the equation
is
where
'
,
The existence
fr-'in
and
mean value
of a definite
i-
inferred at
the o.ntinuity of /(
JUM afl in Art. lti-"> of the Appendix
the following condu>i -n^ are immediate consequences of
'
the definition
i
(iii
'
ti.
if
<
(iiii
because
\\ri<>
.ii-tant,
.1.
\Vi'
if
'+
Jo
sin /HI
00,
<J
f(
(>.
if
i.
on the
and
!
eircle.
an int.-vr (not
=(.
/.ero),
208
[CH. X.
Further, from Art. (45 2), we deduce that if

converges uniformly on the circle \x\ = p, then
We can define the mean value without using the Integral Calculus, by
supposing the circumference divided into v equal parts at #t #2 ...av, and
writing
,
This method leads to the same results as those just proved

and thus
can
be
established
without
the
Calculus.
(p. 209)
;
Cauchy's inequalities
82. Circle of
From
vergent
and the
Art. 10
n
convergence of a power-series â nx
it is
evident that the series
if
series certainly
absolutely con-
.x-i
ni
is
cannot converge
if
because then a nx n cannot tend to zero as a limit.

Hence, if we write (as in Art. 50)
(where, of course,
verges absolutely
if
is
and positive), the power-series con1/L

l/; and cannot converge if \x
real
<
>
To interpret this geometrically, let a circle of radius \jl be

drawn in Argand's diagram then the series is absolutely con;
vergent at
any point within the
circle,
and cannot converge at
Fio. 24.
any point outside the circle. The circle is called the

and it will be seen that, if a n is real, the
convergence
of convergence obtained in Art. 50
is
circle
of
interval
a diameter of the
circle.
81, 82
Of
U;i
<
|.|
with respect
i,-i
n
case to which
MI
<
'h-
<!!<
wh
in
we can
\\VI.TM
[nation b
Art. 7!.
in
nniM
It
nut.
li
raae'e
npjx.s.-il
ilitifs
OO
*in
tli.-
in<l<-c<|.
circle;
ml-
diacu
that
tin-
has
I'rui^h'-iin*
n hidi convei
three
th.-
<>n
<
circle.
l.ut
not ahsolutcly.
Tinan.
ice,
-
i.
jx ti H
of
,//,,,/,'
i.l.-nt
th.- cirrli-
x =1
n-oiii
.17
6 c;in
<
and
Art-
\.-i-lal
\\V -hall,
I.
fully
Tin* iva.lrr will tind
of
"//
h<
;ilt-rati(inx
any
!'..i-
>!'
lit
60
ti
=/,-,
\vhn'int
which'
if
<
j>
that the theorems
difficulty in
1.-
for
cninplcx
|><
>\Vfr-series,
Certain
Mnall
every
ciivlr
made.
h.-in^
de hy
.
-;i\
cirri.
h-
,,,
the ciro\
only
|.IVS,-MI
rgent
Art.
in
iiinr.-
ami
th.-
at
iinii'onnly within
between
hen - "
\\
/.
we ran
intc-'i-atiiiL:-
\,
\^
uniformly
<>n
ivnlily ohtain
.-i-jn-1
>y-t
its
mean-value
crm.
,-
tli.i!
kl = /.i/0
nf
tto2 /" '/x >;ihi
Similarlv.
w-
Thus,
is
//..
lind
8[/<
we
if
.!/
-u.;
nn the
the niaxiinuin \:ilu-
circle
./
ergence,
we may
=/>.
lia\.
<
"
.V.
<M
.-ii.l
\!l
of
belongs
t<>
>
La
210
[CH. X.
Again, since the series
converges uniformly on the circle
w^'
O
we
find
a/
Similarly,
03
1
We
= p, we
find
shall now consider the question

Can a power-series be determined so
as to have given values along a definite circle, say \x\-\ 1
Let us write the coefficients a n in the form a n + ifi n where a n /3 H are real;
:
then write 2o n # n =/1 (#), îftn xn =f2 (x\ so that f(x) = 2an xn =fl (x)+f2 (x).
Now suppose that when x = coa0 + i sin 6, we have f^x) = u l -\- iv lt f2(x) =u2 + iv2
= u + w where u lt v^ etc., are all functions of 6 such that U = U I + H.,,
a,TLdf(x)
t
v = v l + v2
if |c|<l, we find as above (assuming the uniform conn

=
vergence of 2a n # on the circle |#j l)
Then,
In the second integral put \Jx for x

so we have
far
and
= ^-1
(u l
27T-/0
If
we
this will
change u l + iv 1 to
iv^
u-^
-w C-X de.
)
l/
1
subtract the last result from the formula for
/i(c),
we
obtain
/ ( C)=J^
STT^'O
t
Similarly,
by addition we get
In the same way we can find integrals for
/2 (c)
in
terms of
argument being that when
w.2 ,
v2
the
changed to .',
u 2 + iv 2 becomes -u 2 + iv 2
This, however, does not alter the final formulae
/v.sand so by addition we see that these formulae remain true when the x
are omitted throughout.
Thus /(c) is completely determined (save for a
constant) by a knowledge of either u or v. But, given an arbitrary continuous function for u (or v\ we do not yet know that it is actually possible
to determine /(c) so that its real (or imaginary) part does assume the given
values on the circle; this problem will be discussed in Art. 83.
nly essential change in the
.*
is
/////'.
83. Abel's
theorem and
Suppose that |#|

tliat
2a
rt
is
known
=1
is
allied theorems.
the circle of convergence for
to converge, altli<>u-li
imt
Sa,,o;" ant
absolutely.
we take vn = j- n
It'
THEOREM
ABKl.s
82,831
An
in
'2\\
tind that
= lim|v n = 0, and
'/
V = 2\vn -vn+l = \l-
hat
Thus,
sine.in
unit'nniily
the series Sxt^" will converge
'it,
any an
which we
for
I-'
h,<
Ml-M),
any assigned mimher greater than
wh.
\x\<
'I'-,
i.
(See &beJ
ted OB
Thus
_-_)\ /J
n
In
this minlitinn.
./
__
|
r.
It
is
.u-'hly
(see
(l--2/,'-( s,/>-h
f
t-
liuiunni
to the circle |#|
~,
and the equation
(-^7T<0-X),
'2~>
oo8^-
2(X
tiûn-
between
;i
be NM
li.-s
of
may
(X-p)^g>
easy to see that the an-
nearly
Thus
in
32>X
.;,
arc
it
/* (CMS 0-f
-i)/,
-l)p-8
innrr
tin-
Bfl
observe thai
or
p<\d-/->
and of course
p.
interpret this iin-,,uality wre
where
1,
I'm-
nt'
1,
(witli
tin-
tintin-
case X
n>(l-
lima;.m
at
/>
-i.
ajpn);ii'lu'v
thr
taken.
lai
/A
//>/,
Il'any rcûlai- c-urvc i> drawn fmin a point inside the circle
1, then, provided the curve ruts the circle at a
tc tlie ]>'im
./
/'unbioncnthtori
i
wliii-h
in
our
n<>i.itii>n
wouUl be representetl by p=2oos<f>
use a
212
finite angle,
curve
[CH. X.
we can draw
that
a limaôn to enclose the whole of the

the series will converge uniformly along the
is,
Hence lim 2a n xn = 2a n where x approaches 1 along

any regular curve which cuts the circle at a finite angle*
curve.
This
is
the extension of Abel's theorem to complex variables.

in Art. 51 relating to the divergence of 2an
The theorems
cannot be extended so as to hold for complex variables quite

easily, because the lemma of Art. 80 gives less precise
so
lemma
n
series 2a nx
and it is necessary to
assume that the
possesses some further property in
addition to the divergence of 2aw
For as a matter of fact,
even if an is real and positive, Pringsheim has shewn that the
information than the
of Art. 23,
divergence of
2an
does not ensure!

lim
2anxn = oo
|
for all paths defined as above.
The condition introduced by Pringsheim
is
that of
uniform
divergence, which implies
where an
>
and the point x lies within the limacon.

from Art. 51 that lim *La n xn = oo
The reader will find no great difficulty in modifying the
It is then obvious
proofs given in Art. 51 so as to apply for complex variables

when Pringsheim 's condition is satisfied.
We proceed now to find
the limiting values of the integrals, given at the
end of Art. 82.

For example, suppose that v is an arbitrary real continuous function
then the second formula gives a value for /(c) which can be expanded as
;
a power-series in c, convergent if |c|<l. We shall now prove that if this

function is denoted by U+i V, then V tends to v as c moves up to any point
on the circle so that we have determined a power-series whose imaginary
;
= l.
part has an assigned continuous value v along the circle |c|
Clearly it is sufficient to establish the result for any point on the circle
we
so
*
shall calculate the limit of
Picard, Traiti d'Analyte,
t For consider
There
is
no
Af
7
t.
2,
p.
20
is
difficulty in seeing that,
negative,
moves up
and so lim E\ p->o
to
1.
73.
= A'P-^ (cos 20 -
the coefficients a n are positive and that

cos
as c
sin 20)
J,
where E(x) =
*.
when this series is put in the form

2a n diverges. However, if $* <0 <
-,
L(l -#)
J|=0.
îr,
UL
POI
83]
It
be seen that
will
j-
where now
+ c _ 1 - 1*+ 2ir sin (0 - o>)
.r=co8<o + t sin
and
+
and we note further
ti
f>r
'
>
1
a eontinii"u
is
Th*
>
<>,
we
the
positive,
ranê must be
////
than
less
1.
find
)(l-rV<o
-2r 008(0-0))+^'
(,'-
and sinee
x//<.
is
int.-jrati..n
(km over any
value of the
Min<-tiou of
we can determine a
u>,
so that
Iff*.
27T\Jo
= 2a to w = 2rr-2a; here,
next to consider the integral from
//
that
is not
i.>)
greater than cos a, and so
j
.I
+ /- ^1 - iV e, ,s
L-fJ <2(l-r>
while
;
>)
-f
/-'
2
jT sin a
we have
he upper limit to the values of \V\ on the eiivle,
(v-v )(l|l-2rcos(0n
'juently
.-^[(1-r
^ftr-fc
27rj2a
It
th.-i -i-t".
i->
'
-in*a
p(.--iMe
find
t<
|V-P |<,
a,
if
<-..
so that
tf'<u, and
r<8j
to sa\
that
is
still
-;i\rs
'0
tO and
a
id
moves up towards
tegral
tlie
did as c
\\
preceding
T-> deal with

appioa-he-, any point on the eiivle exOSpl 1.
<>
ha> the limit /. when
that
tlii'-'iiirh p.iti\e \a!
1,
neâl
i\
V=V
it
to
i-
it"
the
wluMe
<:>
limit
-1
o>
= 0,
if
we
assign
.=0.
Furth-r.
\\
becomes continuous at
evident fioin
it
in ^.
We
suppose
tin-
.*
I"
rej.ie-ent- the
same angle as
is
indicated in the diagram on
p.
:M1.
COMPLEX SERIES AND PKODUCTS.
214
Now
from the
case considered
first
lim V'
so that
we have
limv'
(w )
(r, e)
lim
V
7T
(r, 6)
where
is
the limiting value of
as
6) approaches 1.
the particular case when v is given in the form 2an sin
and then the result is
</>
[CH. X.
<f>
(r,
In
have m=-l,
sin
Tito,
we
shall
nO=
2a n cannot be convergent
were convergent we should have
lim 2 nrw (cos n + i sin n 0) = 2an
(M)
in virtue of the extension of Abel's theorem (at the beginning of this
It will be noted that in this case the series
for
if it
that
article);
is,
lim
2a^sin*0 = 0,
(r, 0)
which
not the case.
is
Taylor's theorem for a power-series.

have seen that a power-series â nx represents a continuous function of x, say f(x\ within its circle of convergence
84.
We
=R
x
|
in h.
11
let
us
Draw
now attempt to express f(x + h) as a power-series

the circle of convergence, and mark a point x
FIG. 26.
such that |oj| = r; draw a second circle (of radius

with
centre x, to touch the first, and mark a point x + h
r\
within the second circle.
We shall now see that f(x + h) can be expressed as a power-
inside
it,
series in h.
In fact f(x + h)
is
the sum,
by columns,
+a x+ a^ +
+ aj
2
of the double series
ag? +..,
TAYLOR'S THEOREM.
83, 84]
Hut this series
term by
\vh.-iv
liv
absolutely convergent,
absolute value, we ^.
ii'
r.-pla<-<-
tl.
\**\+\*i\(r+p)+\**\(r+pP+\**\(r+i
=
'Muse
Now tlii
i>
fi
|.
mnstructinii
tli.-
ln-raus.-.
is
its
215
Tliat
lately.
willmut
ami
\alu-
altri
H.'licr
<loull<- series
'In-
ti
can SIMM
\vr
i-,
loiible
tin-
converges
series
l.y
rows,
Art.
+ /,)=/!
/(/
\\h
')
ft
l
+ '*i.,.rh3.4a4 o^+.,
f...,
so that these series
applying
att.-nti..n
in
Series
It
in.
iy
ln>
may
/<
usrful
th- nifanin- of the process.
t<
may
//
t..
!><
rallfl
.-iivl.-
is
<-a>y
that
t-
it'
..f
.-ir.-N/,'
./
\\
.-..nvergence for th
that it
\v kiinw
that
f\-idcii.-f
it
>^8.
'/''//''
ivuiaik that thr
h-
l.y
LOU, will
l'>nnal nil.- for SO
tin-
any
inmi
be obtained
it
may
not rxtcixl further.
writ.-
...,
-M
th,n
mgea
J
|A|<|1
it'
\/t\<
Thus the Taylor's
series
_.!!<..
w.- !M\.'
in tin-
tli-;-
>)mded area
ii
>wei wi
i
i<
rh
\M>
II
inn.
.n
ii.
/(
in
IM
.l.l.i
,nt
..
now
(Ii,'
/,,
'/'//<-,<///
rxlrndrd
ol.t.nin .in
,>'
Mir
Mint,
Ilifiv
(If
"I
MI
mil
l"il
KlIlirMnllM,
nil
l-'niK'tioii*
.\inlliiti,'
../
Tlirii, if
that |u3|"r,
<>f
Mint
/(.'
wi^
|A|M|
':.
r. ,n
n|'
Known
in
I'tlillirl
(ill
I'M
IP
1.
/(,/
i.-,
..I'
<
be witlmi
nil
inn
pninln
nil
in
<n
>t
for nil
It)
nr.j u.dil
1<
to
I'ol'
roiix rn;r;,
inn xiiiiiini nf
t.ln'
'nnrli\
l.ldill',
|aj|
.Hid
srric
//
is
/!/
1'nnn
id
-..
Mint
but.
tlir
/',
Hiicli
!!M
dnfinil
nriînn.1
province
rin\ cri-rnrr
ill'
cirrlc
|/(.|BBM.
iln
Wrmi-MtMWl'H
111
.nil
.i.|r
Suppose
known,
nnt
Mir riirlr
tlir
.Mil
Illtrixilli'd'nii
S'J).
i.s
he area of
I.
niM-.il.il
|.
(In
\\V rnn
Ait.
IM
Morlry'H
.'Hid
I.
I..-\MIM|
<
CTS.
I'liohl
points
in
by applying Oaucliy
lui.\'(^
to tlir /^-series
r.|n.dity
,M_ M
HnN
\ppl\in
"'
ll
Mir
.
:
(m
,;imr
n)(m
mi
'/'////
Until
,!,,<{
tO
Unit
tin
l.i'
II.
I.
i]/,,."'
<>/'
IM
M.
.n
lift
oj'n
N|IC\\
rcMiilt.
of
i<
(i
......
Mill
..I
'I
.111.1
M,.|
III
\\
II
ill
-'
.
\l
mid
ll.n.l
|H
oL
n,
-;,
UK
UK
/,
<-/
siinir
:<T<>
"/'
il
Ii
oil
il
.in\
i,
n.
I.-
m.
I..
/.'
in.
A',
HIM
tin n
if
.....
,|
//,'
(/
I.
'
VlirillllloM, IIIUI <>f
.....
M|||
,,
186
i-
UK' orK/iii.
/<>
,,u\ t'l'^c
Mild
<-ir</<
/In'
Ili.ii
Is ritlirr
t'H
Ii. -\\
Inn
<K'
Intxt
(if
fioir,
Taylor'', Minirrin
/>
iH'itrrxt
is
il
//
ix
////'/r
MlJlt-
<-on i',T</f
of
(,(>
\.ih
Lond M">
IK
'
\\
lew
ia
I/.
this
ir.rd
irliirli
/'.-
which
II
,.
i,f
In-
'
oi-K/tiKil
Iffifl
I
cii'i'lr
'
term
or/-//
/(
Mir nrrdd,,, iirlio.>d
ftlsO
///.
iiiiinmmii
in
of UK
../'
,/,,,
In
.HI.
<(
,;>i,r, I</<-IK-<
to tlir Mn'O|VHl
li.i
iifi
i'</<
|(^
.-,-
,17
"
|< (/ ,
|,/.
.it
point
'
i'i
N)
[mini on UK'
con
I/
f-N)'
(r
/;,<
(f-f>
liifiilur
is,
I),
<>/'
UK- r,l</ills of
tli.it
lr;ld
'I'll!
srrirs for
.'
Un'i'ifon*
(r
I'olloU',
It
!)...(///
;/
y I" Mi"
inr<|iiiilit
.Ml
IM
to
l.-Mil.
.1
..
,lll\
II
Mill
I..
84, 85|
IK
r,
'
rf //
//
<
,
O6M,
I"
i.nlniH in NIMMI
ill-
I"-
IK.
ill. 111
\sln-M-
/<
The exponential power-series.
86.
'I'lp
ilillimlt
MiMilir\ in"
in
In
.1
|
In.. (I
)'
J?
îin.
iniill
l'.\
th.M
A in.
Oi
ii
Li
ijilir.it
An
.11
i.
IT
liii'l
hiii'i.nii. ni
In
A kind of
.1
equal
.1
li
irith
pM
the
//
"""
1,'iiuly
ln.i
'"lulu KII
in
\*
ll
require
1.
IM.||,
I,,
<
,,,//(,
in'.
/,
equal
ion
bo
ho
i!
.'I"'
tin
218
then
[CH. X.
/()x/(0)=/(a)
or
Hence a n = l, and
so
Again,
and
we take
if
the limit of both sides as y tends to zero,
we
get at once
Or, applying Art. 52
and
f'(x)
(3),
since this equation
interval
We
we have
must hold
for all values of
in the
^R, +%R\ we must have
a 2 = a:2
That
and so
= a f(x).
a2 = ax2
is,
f(x) = 1
do not
a3 = a 1 a 2
a3
a-f,
it
= a^
xz
this
conditions of the problem
power- series,
+o
know from
a4 = a 1 a3
....
See Art. 52
a n = a^\
we
(5).
X3
argument that f(x)
but
satisfies all
the
see that if there is such a
can be no other than
E(alx).
Now
E(a^x) does
satisfy the relation
E(a,x) x
E(a1 y) = E [^ (x + y)]
any real or complex values of x, y.

Consequently our problem has been solved;* and
for
where a l
the coefficient of x in the power-series for f(x).

and in many respects convenient, to write e* for
even
when
x is complex. But it must be remembered
E(x)
that this is merely a convention and that in such an equation
is
It is usual,
as
e^ = i
(see below, Art. 86)
the index does not denote an
ordinary power.
*
no other function can satisfy the

xf(y)=f(x + y), because we have assumed f(x) to be a power-series.
we assume that f'(x) is continuous, there is no difficulty in shewing that
It does not follow from the foregoing that
relation f(x)
But, if
f(x) has the exponential form.
\M'
SINK
85,86]
between
Connexion
86.
219
<
the
and
exponential
circular
functions.
in tinntial B6f
n.mplex \arial>le
variable
on
a
mil
/, we may differentiate term-by t-rm
depends
\\ itli
respect to /, and obtain the same formula as if x w<
1
tin-
1'
./'
-i
"rr->jM. mlin^ in a
that
''
.lian.u'*'
in
',
<
hangea
A.,,1
..],
ls.
r,
M 8<-*0, &r/& approaches the lim

be la-t iiM-|iialit v
foil..
!jA',,
In pai-ticular.
or,
wli.-n-
//
is
.siip|.ns,.
ival
t!
puiv iina-iiiary and equal to
thru \\c
.<laoit
;-,
li,n
/'//.
if
/:<
'-i
w;)r(oo80-f
we have
euj
Thus and ft q are

and so r=l, = o.
/
r=
Hi-lie.- L^-M.
-rally
and SO
which
-*-L
tn|
.,1'
ind.-j.endent
1.
continued
A'"V
Another
l.y tlu-
m.-tlidd
l.ut
t'.r
'sin
;y.
ivinark that
/
-;;/)
'
/
of establishing
the
last
)=^(0)=1.
r.-sult
>l.scr\in^ that
J
N"\v write
and we
see that
sin
;;
//
;/,
j-
is
//
= (cos^>-|-/
cos^-f isin#=
lim
it'
-f
mra ;/,
i-
îven ly
220
because
lim (n sin <) =
r\
- cos 0) =
r\
and
lim n(l
Hence
by
cos
r\
+ i sin
[CH. X.
lim (sin </<) = q,
- cos
= 0.
w
= lim (1 + n ) = E(irj)
lim
[(1
</>)/</>]
/c
r\
Art. 85.
Still
another method, analogous to that of Art. 59, can be used.
write
Then
But,
if
{yn _^
an d Ji
In
fact,
_ i( GOS + i sin 77).

77
y = r(cos
we have
^| =
or
Lrii
+ r'(^)
^ ~
drj
Hence
'
drj
and
3/0,
#1,
...
are all zero for
= 0.
77
Hence we
find, if
77
is
positive, the
sequence of equations
<
Thus
<
lim
71
If
and
we
so
?/
M =0.
><K
substitute i^ in the exponential series,
we have now a new method
we
find
of finding the sine
cosine power-series (Art. 59).

If we write r\ = \ir and TT,
we get the equations

E(Tri) = i, E(7ri)=-l.
Using the notation explained

cos
and changing the sign of

cos
thus
cos
f]
r\
+ i sin = e
rj,
r\
in Art. 85,
r\
we
find
i sin q
= (e^+e~ '*),
we may
ir*
sin
= e~
r\
ll>
;
-.
(e^
-e~ ^\
write
and
WX3 MMTIIMir
86, 87]
have
\\'e
at
and
;l
tin-
<-O80;
<-,,n\ .-nieiil
est ahl ish.-
already
JM.V.
'2-21
d-finiti
IK.
pivsriit
VI
I'l
when X
e.juai
I
COS
''
and
-in
'
wh'-n
to detin,- tli.-m
is
Then
real.
by
the
('"
.,
are
tni.-
comj.lrx valurs of X as well as
I'm-
also
inll(,\\s
It
that
any
ri-<
depend only on tin- addition-theorema

complex \ariahh-s: thus in particular
main
vmte x
re
will
it
///.
cosh
--in
fcoah
,'
cosh/; = A(^4-r
wh\\'-
functions:
V instance,
;/
Ix-
seen that
sin
^sinh
,h
-f
;y
sinh
l),
//
shall not elaborate tin- <l-tails
cosh
inl
is
\(<
<!'
results can
th.-
unalt<-r-il
formula*-
sin
;y.
"-,
\\n> analysis of tin- sinh

he found in many t.-xt
>
35
1
1,
.sii.l,
= x
we have
|rfn|<|
Similarly,
we have
^ eoab r|;
and.
if
87.
\\'.'
\x\
<
1,
we
find
The logarithm.
have already seen that
/.'i
'/
Henee
if
is
/////// '.<<
|cosic|<l
1,
|#|<
//"/
\\V can. liowever, replace them hy "there,
vulii/.
so, if
Arts.
;y.
Algebra, ch. XMX.i.

n-hen x is complex, fl"-
|<
as
and
of
C'lirystal's
to be noticed that
are no longer
tin.-
which
i'<ii-iiiul;n-
r.-inain
tru-.
COS X = COS
-in
i-.-al
un nn-t rical
any
inte^.-r
if
//
008 If -f
(
j.o.sit
a real
is
'
i\
s i"
AVJ, --;>=!,
ai
;
/-
..]
n.-âtive),
222
and
since E(g+ir})
the equation
other than
=0,
It follows
ri
that
= e(cosr]+i8inr)) there
E((+irf) = \
= 2mr.
[CH. X.
are no solutions of
we wish
if
E(y) = x,
to solve the equation
so as to obtain the function inverse to the exponential function,

the value obtained is not single- valued, but is of the form
y=y
where y
is
(n = 0,
+ 2mri,
1,
2,...),
one solution of the equation.
Q*
FIG. 28.
If
we
represent x geometrically in Argand's diagram,
x = r (cos
we have
+ i sin 0) = rE(i6).
logr is the logarithm of the real number r, defined

as in Art. 157 of the Appendix, we have
r = E(log r),
x = E(logr+iO).
and consequently
Thus we can take y = logr+ i0, and then the general solution
is
2,...).
(n = 0, + 1,
y = \ogx = \ogr+i(0 + 2n7r),
But
if
We
define the logarithmic function as consisting of all the
and we can specify a

exponential function
one-valued branch of the logarithm by supposing a cut made
inverses of the
along the negative part of the real axis, and regarding x as

prevented from crossing the cut. Then we shall have
=
log x log r + i9,
With
this determination,
where
logx
TT
is real
<C Q^TT-
when x
generally the most convenient assumption.

be observed that then such formulae as
is
\og(xx')
is real,
Hut
it
which
should
= log x + log x'
can only be employed with caution, since it may easily happen

that (0 + 9') is greater than TT, in which case we ought to write
= log x -f- log x' 2?ri.
log (.''./'')
The reader
will note that for
two points such as P, Q
diagram (Q being the reflexion of

of the real axis),
1
i 1 1
(log x p
log X Q )
in
= 2-Tri.
in the
the negative
half
LOG \UTH.Mh
87, 88]
oh\
rut.
continuous over
jollsly
88.
the
at
repi
ranch
I.
ill'-
whole
tlir
sejrcted
The logarithmic power-series.

know 1'n.ni
.ind lii', that
v
\\'r
(1)
= .i--
//
represents the function
In other words.
(
*J
must le
real
is
and
\X
<
A.i-'+'.r'-...
function
series (1) in the
tin-
powers of X, the result*

Hut thi- ran-fonnation is merely alêl>r;i
equally true whether X \^ real or complex.
1-j-.'.
and. as such,
log X IB
inverse I" the exponential
we substitute
if
.and then arrange
if
>,
for
of X.
plillie
to
;icc. .I'din-t
is
Since the series c2\ c]\
values of y.
all
f >r
i.solutely
the d'-ranêim-nt implied in this transformation is legitimate (see

is
Art. :>ii). provided that the series
al>olute]y COB
1
Hence,
ipiat
i-
'ii
value of
|o^|<[l, equation (1) îves one
if
<
'J
Thus, usine
ami
hranrh
the
from
furthei'.
(1).
the
of
//
defined
Inârithm
satisiyinu
>/
when
real
i-
is
./
in
real.
the
we have
artic!--.
+.,)
= .,_
^+., .:;_...
(if
|a.|
<
i).
p>
-1
From
the ti-ure.
WhThis result can he
where we proved
evident
is
it
at
that
If
we
\\
me
^'
< <
(if
)'
^=
'
./
-...-,
-f
Ml
is
a gooii
e\>
,-sin
r(oOfl H-f:;
this equation
-7T <</>= +~
once continued
<
arctan
that
sin
f))
in
Al
reference t>
l,y
'2tf+
}.
'
Ar
H-
^ tM
--j\es
Jr'cosSH-
r-sin ^(9+^/ J sin
:;>
the poi
^
/v/'sin^-^. /-sin
+ |rcos:{H-...
l^+^sin.Stf-.
i.y nuiiu-i -ir.-il
i-Diuputation
up
to,
say, a*.
224
and obviously
Thus our
p
tan
results
we now
except that
[OH. X.
= 1 4- '2r cos 6 + r
= r sin 0/(l + r cos 0).
2
are in agreement with those of Art. 65,

see that
actually lies between
<
+ JTT (because r < 1) instead of TT and

We shall obtain an independent proof of
(if
log(I+x) = x-x + x*-...
and
\-K
TT.
the equation
a?|<l)
in the following article.
The
and
series for arc since
Again, by
arc tan
a?.
Art. 64, the series

1
y =x +2
,
xs
3 x5
+ 2T4y + -3"
,
represents the function inverse to the sine-function (Art. 59).
<
1.
for real values of x, y, such that x
for
all
values
of
y,
|
is
Since the series (4)
and the
aj|<l, the algebraic relation between these series

to persist for complex values of x, and we can
series (3) for

is
now
seen
accordingly
write
1
xs
3 x5
..
since the series (4) is taken as defining the sine for complex
values of the variable (Art. 86).
Similarly the pair of functions
y=x
(5)
are inverse to one another for real values of
and we may therefore write

arc tan
a?
a?
\x*
+ \x
x,
such that x
|
<
complex values of x
for
b
...
(if
|#|<1).
In these equations the values of the inverse functions are

determined by the conditions
BINOMIAL SERIES.
88, 89]
225
The binomial power-series.
89.
/(, x).
\vheiv hoth
;nil
in.'iy
C+
r,
.11
>1*
.,
/(,.
we
hav,-
r>x/<.',*)=/(, + ,>)
pick nut
&
of
meHiri.-nt
tin-
2)+
,.
Th- -rieS
when |#|<1, ami thus we
.lit
i!'
+ ,* + ,< -
tin-
is
absolr/
il.-ntity
(|*|<
in
th- product
/<**)x/(,',a>>
rule (Art. 54) to be a polynomial
of .r r \\\
/'; thus tin- co-tlici-nt
seen by the ordinary
is
it
lth
of degree r in
and
,X)-f(+l
-o a
when
Hut.
,.
wh.'ii
and
j'
same de^
of the
(my two
are
M186,
polynomial
cniis ..ju,.ntly ~\
must be identically
any aî-nr<l
is
values of
infinity of <liti'<Mvnt
because
int.---.-^
,.
int-^-r.
(namely.
i-
I.
.V,.
'2.
is
8,...
zero for an
to oo
).
Thus, identically,
x)xf(v',x) =/(r+i
From
this
c2) to
til
I'.ut
i/
is
as a
1-y
= (!
(|
*|<1).
method indicated
In the
ditl'erently.
power-eeriee
.f
the
in
columns
1-f.
-ho'
for
doul.l,-
complex values
<.f j. w.proceed
can be expressed
x) can be regarded as the
lirst
t'd.
place /(v,
a;)
961166
fj
''
,
-f
...
11
<
-f ....
ua
in
a rational number.
to deal with the case of
somewhat
the
prove that
,.<)
when
'..'),
relatin we can apply
226
Now this double series is absolutely convergent because, if

= and |a;| =
the sum of the absolute values of the terms
a;
|i/|
[CH. X.
+ l)th
in the (p
column
is
~~x/,
which
is
the (p + I)kh term of the series

if
x <^
/(i/
a?
);
and
/(i/
o?
Art. 12.
by
converges
The double series being absolutely convergent its sum is not
altered (see Art. 33) by changing the mode of summation to
rows, which gives
f(v, x)
where
= l + vX
= x ^x
Thus, since /(j/, #)x/(V, x)=f(vd-v,

above, and deduce that*
/(,, x)
In order to determine
Thus
x
new
is
is real,
let
lt
we can apply
Art. 85
us write
i/
= l,
which gives
a value of log(l+x); and since

the value defined in Art. 87.
it is
We
is
real
when
have thus a
investigation of the logarithmic series.
X = log p +
Hence
(see
= E( vX
x),
fig.
i(f>
29, Art. 88), so that
sn
v = a + i/3,
where
is due to Abel.
The investigation above is based
Goursat
the
on
(Cours d' Analyse Mathtfmatique,
proof given by
a result which
275).
The method given in the example of Art. 36 (p. 89) applies

and the following method was suggested in 1903 by
indices
JDixon
The
complex
relation
gives at once
where n
*
to
Prof. A. C.
is
Of course
/(i>, .r)
f(v,
x/(i/, x) =f(v + v, x)
*?-[/(i
*)J-0 +f )"
Sa7
a positive integer.
v
corresponds here to
jc
of that article
and
A'j
corresponds to a^
BIKOM1
89]
\i.
>(-a(-*)*
('
-]
-=)('
the aeries in square bracket* has each of
tllf
lll.lli
'.n-t^|)Ml|ili||^
tf|
terms
its
less, in
absolute
+
;i
convergent
series,
independent of n and consequently th- limit

found ly taking tin- limit of each term
;
to oo can be
Hence
But
)-v(*-
(Art. 86)
lim(l
Thus,
+ f)" = /;(./'),
/(is #)
if
.r'
= lim(l +)" =
= lim ().
[
log(l
+.i-)].
Tin- di>eu-M<n -i\,-n alM>\i- appli--. only to points within tin*

o5|
ire
= l.
To
c-an
exaniiiif
i-.-i'.-r
convergence at points on tinbut on
tin-
back to
WetarafcasB'fl nilf (Art. 79);
tinimportance of the binomial

an
indt-pciHh-nt treatment of case (ii).
give
\\\ have at once
tint
>t'
\\<-
s.-i-irs,
shall
<-
\vhere
i<
= u4-
Thus the
a
is positive
aeries
;
|o)|
I~
-.
i
T~~9
.1.
the circumi
converges absolutely
'Ilm-.
[(*-o?+/fi
<
and SO 2a
imference.
nt
and
-;
^-*
-"I
up
aince
ff[vlog(l+flj)]
to
i-
and including the

also oontinaoi
that
,if
C)
>0)
th- cii-cuinl'crrnce |a;|=l.

points
(>n th.thT hand, ii' a+l=0, after a certain stago the
absolute \alues of the terms of the series never decrease, and
at
till
so
fd
///
Lverge \vhen
1
liut
tlu-r.-
liin
For the absolute value
is
no
|a;|
= l.
ilitliculty in MTJIIJ;
tli.U
A'[
of this exponential is
p*~^, where
(it
228
-1
if
Again,
lim|an =0;
< a^
also
n->*>
0, we see from Ex.

we have /(I, x) = I+x, so
(l+aO/fo x)=f( v +l,If
we
[CH. X.
Art.
3,
'
first
that
x).
take only the terms up to xn in this identity,
where Sn Sn are the sums of the

and /(i/H-1, x) respectively.
39,
that
we
(n+I) terms
see that
in f(v, x)
Now
the real part of j/H-1 is a + 1, and is accordingly positive

'
follows from the previous argument that Sn tends to
a definite limit as n increases to oo
and so
it
And
lim
since
n\
that lim(l+oj)
is
definite
%->oo
thus, unless
x=
1,
lim a n
= 0,
it
follows
and equal to E[(v+I)\og(I+x)]
we have
f(v,
x)= lim 8n = E[vlog( ! + #)],

n-><*>
and the series converges on the circle, except at x

For points on the circumference, it is evident that
,0
= ^0,
= 200810,
1.
(-7T<0<7r)
and so we have
f(v,
a?)
= (2 cos J0) ae-^[cos{ Ja0 + /31og(2 cos
>
1.
provided that a
For the special value
It follows, as in Arts.
but that
1,
we have
the identity (see
p.
152)
42 and 61, that
Sn ->0,
ifa>0,
Sn oscillates if a = 0.
|^n
Koo,
ifa<0,
Differentiation of Trigonometrical Series.

In some cases of interest, it is found that although the series
90.
f(x) = 2tt ne
is
inx
= 2an (cos nx + i sin nx)
uniformly convergent, yet the series of differential
coefficients
If this occurs, it is often possible to obtain

ceases to converge.
the value oif'(x) by differentiating the series
MITI:I;I:M
89,90]
\Tiov
This gives
the
pn>\ ided that
^/H/
s.-ries
fc
= 0, 2-7T
are excluded
Lence, or in
if
n(o
int.-rval
any
the 6886 (in
linly
/,.
>
interval
any
is
-...],
c.,iver-
\vhirh
1'rnin
a r.-al positive
an
"
l/<
ii'
+8(a
^ uniformly
")'
is
|
decreasing
convergent (see
Arts,
the value of f(x) as a
\ve substitute
It'
re liml
'.>'
-...].
seri-
th,-
eqaation
,,!-<
In
1.%^
will usually
it
].i-actic.-
for f(x)
-(",-found best to
and obtain thf /"/////
first,
the
litl
(Mjuatinii
and
4
l>y (1
as
if the differentiated
of
JMW.-I^
rrangemeni accinlin^
This rule will be seen in ('haj.ter XI.
Be were convergent
which
is
to
l>y
int.Tj.i-rtcd
multiplication
>
md
be more than an accidental coincidence [Ar
to'
UO(2)J
is
It
if
easy to see that exactly the same method can be
necessary, to .-staMi-di a similar rule to interpret the 91

oiwhere
+
+'
ly IIMH- the 1'artoi-'
>
<
i.
is
;iu
integ
Another process, which in practice U almost the xiine as the

L
1,
foregoing, is due to Stok.-x <.!/.////. i///// /'////>-. /'"
pp.
t
D);
the
tirst
case of Stokes's rule
may
rcdue,-d
the form
-1>
where
.1
determined
Constant
i-
<
}u-:icti.-c
8,
;unl
I'.'M;.
in
p,
(S),
^7).
principle.
t.
!_'.
H, lt
-i.
.-
liy t-.|uival-nt
-
liy
the
e<iidition
and another case can be
i- e,,i,
lv:.. p.
t
tliis
:>:.!
that
writt.-n
-/>'
B]<
luive
Item
pven by Lerch (Ann.
dt,
,m.l r.n-nk
,,g
process seems simpler, Uotli
in
230
[CH. X.
a constant which makes the series on the right convergent. These results of course depend on the fact that
if
is
in virtue of Art. 65 (see also Ex.

Ex.
ix
Consider
1.
below).
1,
f(x) = e +
This gives the formal equation
/()-t'(*
which leads to the
real equation
(\-<&)f(x) = i(**
or
/(#)=-(*+ cot
Thus
But
f(x)
for X=TT,
find
f(x) =
so that
in
we
Ex.
2.
Similarly
ix
sin J#)
. . .
==
that
J log ( tan i^) + i?,
< <
^P
TT,
ete
and
These give,
if
cos
x + } cos 5^ + ^ cos 9^4-
sin
# + J5
sin
5# +
sin
9^ +
3.
Again consider
=|TT
. . .
- j log(tan
Jo?),
= |TT + J log(sec # + tan #).
. . .
oo
Ex.
< x < 27T,
+ %i(ir-x\
65.
we can prove
+ ^tx + fypx +
0<.r<27r.
/(7r)=-log2,
- log (2
agreement with Art.
#),
- log (sin J#) - \ix + const.
--
ginx
2i
f( x ) = _,t
7i
Here we get the formal equation

f'(x)-itf(x}=-iê**,
00
which gives the
real equation
[/(^)-i(/(.r)](l-6^) = 0.
f(x) = Ae
where A does not involve

we
obtain the result
#=7T
Putting
Thus
or
itx
ê
ip
=a
__. + _._..
#,
and x ranges from
i=s7r cosec(7rO (see
Art.
to
!:>).
2?r.
\M>
BINE
90,91, 92]
Th
\ilim:
--
The
The identitie- "!'

and we deduce, as
Art.
Now.
and, since
complex values of
05,
sin 2 (a/n) 1
-an
KT
always
t-nsuiv tliat
and so Art. 86 gives
\x\ t
<
and rosa.
.^'i
to t.-ml
greater than
n(t-n)x
i
-
ivinain true for
(i!
a;
sn
BU
<
Art. 70,
in
sin
'
sin
products for
infinite
?i
-coeff-iOf
~ 2<
>
7
_
'
91.
06IKE PBQ01
we get
"',
l-y
'
sin(r7r/?i)> 2r/n
(see footnote, p. 184>.

: n 2/ r
n
,.
I iC/ TV
tUU
II
J
^N^
Hence
sin
l/r
*/ I*
'
L
r2
~
-2.")
'
and consequently we can
in
thron-in
tin-
.1'
Art.
Hence, as
P.'.
we
Art. 70,
in
h'ml
r =i
In
tli.-
ânif
\\-
way
C( )S
92.
The
find
The investigation u v
-\t. uded \sith'Ut
difficulty
'
t'nl
-Br=
series of fractions for cot

r
the
11
./'
i"
Art.
to a
71
tan
-/
,
for
./
,
real
cosec
'.
angles, can
)>
complex argument. ly making
lowing nnditicati"!
\Ve ha\
e,
of ooune, the idi-ntity

.
n)
lie
identity
merely an alêl.raieal deduction i'n>m
is
trigonometrical addit ion-t

.nients are real
li-
.\liieh
ai
\vhether the
232
Now
here, as in Art. 91,

sin (r7r/ri)
assuming that
Also
we have
> 2r/n,
n> x
[CH. X.
sin (x/ri)
\<%\x\/n,
we have
(Art. 86)
Hence
2n sin
1
and
ri
2
1
(x/ri) cos (x/ri)
sin 2 (r7r/ri)
Thus, provided that 5r
- siu
>3
(x/ri)
x
|
\,
<
> 4r
we can
and so the theorem
30|*|
of Art. 49 can be applied just as in Art. 71;
X
T
hm /n sin - \
'
n/
exactly as
if a;
lim
lim
TI,
is
- cot - = )
n/
Vn
sin 2 (rTr/n)
sin 2 (x/ri)
%x
rV
for all values of x, real or
Now
^
T
# = - + Zv -Q2
X
X
now
2x
complex (except
/
= -1 + ^v
7
~
o
2
ri 7T
VaJ
IN
)>
TlTT
7?7r/
the variable of summation, instead of
the following identities hold
tan x = cot x
2 cot 2#,
cosec x = cot J#
cot x.
identities are familiar results
when
follow from the formulae obtained in Art. 86.
r.
:t
These statements follow at once from the power-series
fThe
'
x2
TT),
cot
where
*(/n)<x*(x/n)
Thus we have,
multiples of
X
V
1
lim / cos \ = 1,
\
n/
= x,
were real instead of complex.*
Hence
and
30|a?
use the comparison-series
^
further,
-f
we have
2 sin (x/ri) cos
Consequently
"--
.i
is
for the sine
and
cosine.
real; for other values, they
92, 93J
Hence we
find,
OTANQENT
ill.
on
su)itra<-ii..n.
_*
&c
^2n-l)
811
y2i0,
93.
to
2!"*"3!
con<liti<'ii is
i-r'-vidi-d
that
+ "*= 1
l>y
s<-
<
p,
\\\.
taking
At
if
||<1-L>.
n'nd
'
\vr
'
,./'+...,
l-'- =1+ 2'
sulttract th.- lir^t of these ezpansioiifl
obtain
.*
'
we
function x
ii'.
It
p=
\\
=-,
-j
certainly satisti-l
l.y tai
can
th<- iv<-iprocal
\\v
4-1
-j
,'-"
for x
'
Thus
'
Beri<
OK
-r
and consequently (as in Art. :>4)

can
expanded in powen "i'
This last
+e
^tl.iVi-i v
The power-series
The exponential
.e
<
^-o^'
\v- M'- tliat
'
5^"i
2x
= cos* =
'
ly writing
8KB]
tin-
idmtity
r=
^3=-
-l-.
(>
bom
the aa
''-
>
234
[CH. X.
Consequently we can write
*where
5 5
JBj,
2,
3,
...
_L R
^2+J
?J_7?^_
+
7?
are Bernoulli's numbers.
It is easy to verify
A=
""
i>
by
direct division that
= *V>
#4 =
= A,
#>=F
but the higher numbers become very complicated.*

Again, from the last Article we see that
Now
x <C
if
2-7T,
each fraction can be expanded in powers of
giving
x,
ZnW
And
the resulting double series
the series of absolute values
convergent series
It
e*
of
which
2laj|
is
now
seen to be valid for x

|
By
absolutely convergent, since

obtained by expanding the
therefore permissible to arrange the double series in

x and then we obtain
is
powers
is
is
<
2-jr.J
comparison with the former expression,

T-k
11
V~>
J-
QJ X~>
T-
1
A
T-
we
J,^
*"
see that
K^>
1
-1-
and generally
*The numbers
<;i;ii.sher
by
(as decimals) and tlu-ir logarithms have been tabulated by

to fiw are givi-n
(Trans. Camb. Phil. Soc., vol. 12, p. 384); and
l
Adams
Paperx, vol.
Ch. XXVIII.
(Scientific
1,
pp. 453
and
455).
(For more details, see
0.)
Qhryitel'i .\/,,<bra,
tThat 'lir is the radius of conver^riirf may be smi tVoiu Kaker's theorem
x
(Art. 84); for the zeros of e -\ are y = 2niri, and the least distance of any of
these from the origin
is
L'TT.
I:\MII.I.I -
l:l
93, 94]
Wo
MI;|.;RS.
ol.tain thus tin- n-siilts
S-i
Alt
\^
(l)
,
-^v V
instructive to notic,. that.
is
^ V
i
'
Z^^c
It
\i
\\
hen x
is real,
1 -
^~9450*
we
liav-
value of
any
**
^+
'
(2.2
4nVJ o
.
(r
tin-
l.y
+ 3)
ail.litiMH.
terms of
following
For instance,
for
see that
srrirs
tin-
of
ti-rin
we
i-
c^ ~~
n-].ivs-nt.-l
\>y
th-
with an error which
is
l-ss
than
tli-
any
ival poîtiv.- \alu-- of
,/\
W6
h.-.
720'
an.l so on.
By HUM
Ex.
identity
WO
ca
liat
md
94.
Bernoullian functions.
r.-'i-nonllian
function of
in
th-
D of
'',-",'
whicli.
/
<
lv the
in
powers of
the
236
[CH. X.
Thus we have
so that
.
rn
P
~2~!
where the polynomial terminates with either x or x 2
From
the
this formula, or
first six
direct multiplication,
by
polynomials are
~n-4
*** x
41
we
find that
= yz,
y = x(x
where
Again,
n (o?
+ l)
(j>
\)
n (x)
is
the
\-.T-'
coefficient
of
/n\
in
the
expansion of
^ n (33 + 1 )
so that
If
= 1,
we write 03
we see that,
results,
2,
3, ...
if x is
n (x)
= nx n ~ 1
in the last equation
any positive
and add the
integer,
n
which gives one application of the polynomials.
Further, by differentiation we see that n'(x)
(j>
of
jn\ in the expansion of

t*e
_,t(e*<-\)
e'-l~
Hence we
find
is
the coefficient
94]
BERNO1
and generally
0',m(z)
LI.
\\
= 2m^,,
\< TI<
II
-I.
+ <-l
1
Change x
to
a;
ami
$*aroot of

)--=<>.
<//>u
njuatinn that B
t'nllnws 1'p.ni tin- last
it
>and thai
ia
ti\- i'uncti"!
tin-
-,> = <-!
Sim-.' 0,
hat
='*,':;
Thna
1..
to -/.
f/
root of
v lraK to
and
<j,
"x
Suppose
valu-
tliis
then
to.
up
lias
rouj.-cturr
IM-.-M
i-
a glance at
ronj.-i-nnv that
tin-
no root
that
'/"'
""///
^,n (^)
ôo<
i-sialilisli-l
-J-.
1'.
since
<-c),
:(#)
en
is
<>
innnri-ically
and
at
./-=A,
and cannot
\a:
1.
Consequently ^
('2/jL
ran
:\
Imn- ^
between
havr no xrro lM-t\v'-n X -<> and '..and thcr.-1'mv non<- l>etween
I
Mid
Thus
I.
a-fl:
and
tlu'on-ni
tinit
has
indiratrs th- ivlations

illustrates tin- general
rxti-n<h-d
IMM-II
always
iin^ly
ll
tru-.
,.
\alu-
to tin-
Tin-
'
diagram
and so
ar^p.n
^j''
It
(in
>
will
1).
lr
s.M-n
l'";in#e
thai
<.
m (.'')=
<f>
,
/x
// M/
Oy
"/"/ t/tt
tin'
i).
)-'
238
summation formula.
95. Euler's
We
[CH. X.
have seen in Art. 94 that
if
x and
are positive integers,
li
the polynomial containing 1(71 + 2) or J(ti + 3) terms.

It is obvious that if f(x) is a polynomial in x, we can obtain
the value of the sum
/(l)+/(2)+. ..+/(*-!)
multiples of the Bernoullian
functions of proper degrees.
But to obtain a single formula,
we must utilise the Calculus and so we observe that we
by
the
addition
of
suitable
can write the foregoing -expression in the form
Hence when f(x)
is
a polynomial, we have Euler's formula*
/(l)+/(2)+. ..+/(*-!)
where there is no term on the right-hand side (in its final form)
which is not divisible by x.
However, the most interesting applications of this formula
arise
when
f(x)
is
rational,
algebraic,
or
transcendental
function, and then of course the foregoing method of proof

cannot be used; and the right-hand side becomes an infinite
series which may not converge.
in these cases, we shall apply the method of
Consider in fact the integral
To obtain the formula

integration
where
<
by
parts.*
denotes the Bernoullian polynomial.

that </*_.(/) vanishes for /
KViniMiilmriiig
this integral
is
equal to
*Seliwanoff, J}i/ereuzenrechnun</,
38, 39.
<)
and
/-I,
\vi
tincl
that
M.MATH'N
'
95
|,,|;MI
|. \
ly parts again, w-
Int.-.natiiix
mil.-u-ly
+0[^-a(0 + (-l)"^Hence
if
we
v,
.V
we have the
-i=(-l)'
Let us now examine the case n =
A"
Thus, rinoa
(:
')],
we have
tli -n
,//.
<MO = '*-'> we
have
-l
Th.r
:r'
Thus
*f
r'(.r)
//(o^4-.;L/'o--Hi)-r
and
!)-/]-
fl-
uke
a succepi:itii.u>.
Tlifii
1 )
\\i-
liavr.
if
\\
M->J. at
+/(&) - ( /(')<ft +
/*-!,
A'_.
and add the
f>r instance,
* L'X'O +/"(")]
240
[CH.
And obviously we can introduce as many terras as we please on the

right-hand side.
For a further discussion of Euler's formula, see Chapter XI. Art. 131.
Ex.
If tyn (x) is the coefficient of
1.
\n\ in the expansion of
prove that
and
if
a positive integer,
is
^n(x) = l n - 2 n + 3" -
- -^n(O)
Deduce that
if
f(x)
. . .
a polynomial in x,
is
/(I)
Êx.
As
2.
particular cases of Ex.
These give, when x
is
1,
we
...
+ const.
find
a positive integer,
1-2 +3 -4 +...+(_i)*-i# =-U

2
2
2
1 - 2 + 3 - 4 + ... + ( - \- a? = (- \
or
and
In the
so on.
according as x
Ex.
3.
It
is
is
and third
first
cases the alternatives are to be chosen
even or odd.
easy to see that
From
the foregoing equations and from those of Ex. 3 prove that
Shew
also that
is
(x-%)
a factor of the odd polynomials, and
of the even polynomials.
Ex.
4.
Prove that
if
is
odd and k
is
an integer,
r=0
and obtain the corresponding
result
when n
is
even.
x(x\)
LMPLBfi
x.|
A.
\.
Complex Numbers.
inn
th.-
It
1.
an-
l".tli
ii.
od n"i
li,
;m
..ii
is
al.
;liat
!>--=a',
_rer.
2.
y are complex, prove

f
l*+yl +
'
thi^
pivt
-|uati>n
l*+yl +
in
that
Ai-iranil's ilia-^raiii.
'hat
!-'
[HAKKXKS.S and
3.
It'
are the points
/!
I,
uid
that
slu-w
.iindittMii
tin-
4.
It'
t"
.m
"I'iirin,
I.
wh(
'=0,
OA
.:
1901.]
0.]
inh //=
0=
condi
tlif
[.!/<
\\hi.-l,
:i
in
t.
.\<>i:.
Ml.
we write
\\li.-i
whirh represei.
.N-nt
..f
.1-
Q08J
OA ./:.
o.i-
[Tian-fT
.1
in
"<
mid point
if
and
if
I/--M-..
|&+]/&:= > /[:i!,-,,>i>(

\\ln-i
;-.ints
tl.
ri|iii;u
-I,
when x
d cases
>h-w
'mpl.-x,
tli.-v
sin
and that
sii
mi the roots of
nation \\, :
'
th.-
-.|
(*+l
)\
and prove
tln-\
tliat
6.
It"
the
|iiati"ii
ha-
(t \s<'
lii-nrj
real
and
t \\
-"ni|>l.
s
,-as = 0.
If
7.
/]
a,
t!
tin-
point
't.
5+y
a rouii- m- straight
lim-.
\vln-n
takes
all
n-al
8.
t
-Miplrx
\arir-. tin- point
iinniUtM
"
1.
LB,
\a!
[STU./ und
onnliti'!is for tin-u.
ti
<
IW
tliat
at + h
t-c
when it
,'ht
lin.-.
as
242
If
9.
varies so that |j
= l,
in general describes an ellipse
whose
= |6|,
which
If
10.
is
whose axes are
|a]
+ |6| and
|a|-
6|,
and
varies so that
shew that the point
given by x = ab.
prove that the point x traces out the portion of a straight
terminated by the two points x2 =ab.
foci are
If |o|
line
[CH.
|"|
= 1,
prove that the point
in general describes a limagon, whose focus is c

b 2/a.
Find the node ;
b , shew that the limagon reduces to a cardioid.
and if a
[MORLEY.]
If
11.
varies so that
|tf|
= l, shew
a
that the point

b
*-iZI + *Fl+
and find its asymptotes. Under what
the origin (1) the centre, (2) a focus of the curve ?
in general describes a hyperbola,
conditions
Prove
is
also that the point
x=
-,
(t
--
z^ H
*
+c
^-
describes a parabola.
Constructions for trisecting an angle.
12.
a = cosa + *sin
a,
of the equation in
t,
If
To
effect
the circle
1
1
the determination of ga
t
is
equivalent to the solution
= a.
this geometrically we use the intersections of a conic with

I
the form of the conic is largely arbitrary, but we shall
give three typical constructions, the first and second of which, at any rate,
were known to the later Greek geometers (e.g. Pappus).
(1)
If
rectangular hyperbola.
we write our equation
in the
form
*-o/4
= -;, we find
l/t
t
that the points trisecting the
+ irj,
2
are
+ r/ 2 =l of
given by three of the intersections with the circle
angle
the two rectangular hyperbolas
and then put
-?/
-(cosa + 77sin
a)
Of course the fourth intersection

is
not on the
= 0, 2^-^sina + ?/cosa = 0.
of the hyperbolas
is
the origin and so
circle.
Either of these hyperbolas solves the problem, but the second
is
the
easier to construct; its asymptotes are parallel to the axes (tin- one axis
being an arm of the angle to be trisected), its centre is the point
(-Jcosa, Jsina), and
it passes through the centre of the circle (that is,

the vertex of the angle to be trisected). Since a hyperbola is determined
by its asymptotes and a point on the curve, we can now construct the
hyperbola.
(2)
The
hyperbola of eccentricity 2.
hyperbola in (1) cuts the circle
first
as the hyperbola f 1 - <ty* - 2( 00*0+17 sin
-fr/
a)+ 10.
=l
in
the same points
LMPLE8
\.|
Tin-
li\
I,,,
pei
baa
la
\.
ecr.
>
),
oth.-r
I.
ran. h
tin-
tin-
,,f
pres.-n;
naturally tli.m th'.-e îven in

ad
that
in
same point-
tin-
line
u f
and
.V
2
*+/ = l
the circle
point (sec./.
x /7,
,11.
V
Oftfll
through the points
ti
...
13
(iii)
tin-
shew
kin^
hut the
1=0.
at
-----)
l'.'
=
= 11
(i)
sin u -
//
n -
to ?;=0, passes
irallel
If
13.
(1) cute
,/sin
tin-
a*
Thi- paral>ola has
ami touches
(1
tii'
I'.v
iir-t
tin-
and
vertex on the
the
IT-|M.M.|
'i:'..
A'
at
s
-f -'
N \ li'
-./O-K'-'-v
easily
that
|.i-.,v,-,l
,V-f.S"=-l,
.)-=n.
Thus
.v
a root of
is
,S-'
+ + = o,
.S'
It
is
easily
proved
the >i-n
In
like inann.'i
In
CM6
or
.V
\\\^
;.
Le
niii-t
:.
7)-
tor
tli >t
\\hi.-h
:J
X*
7)
un(&r
oompa
,th
,.
sini
|..
ease
irr
7)
188,
(ii).
now
(iii)
and
=2.
s=.,
*
II.
1.
ig
::
13.
i>
iilivimis
ly
in
\\hi.-h
the
than the
first
+ 3* + J>
l,,,t
eonsid.-i iu^
That
.s
(and the!
.mst
l.e
|.
;.
4-cos(67r
14.
With
,,
(in
tin-
term
nunn-rieal
same
i>
,nd
jh-
in
the
last
= 4 gives A'=(l
'= 8
V
(l
(iii)
[In the
In
tlu-
a- 18
...
we ha
second case, wfirst
case
In the third QMO,
thai
vain-
).]
example, she\\
that
lew
244
[CH.
a-l
15.
more general sum
The value
of the
to a, can
now be
prime
bn
y=^x
=o
where
~,
if
= 1,
6 = 3,
6 = 1,
6 = 2,
6 = 1,
if
if
fc
if
if
if
if
(-l+;)2v/2,
if
= 2,
= i,
6 = 3,
(-l-z)2 v/2,
if
(1-1)2^2,
if
b is
We
inferred in these special cases.
an integer
find, in fact,
2,
4,
5,
6.
3,
4,
5, 9,
6,
7,
8,
3, 4,
9,
10,
12,
5,
7,
8,
11.
6,
10,
= 5,
= 7.
from a consideration of the special cases discussed

ar n * may, or may not, be equivalent
n
set
x
In
a
is of the form 4/- + 1, where k is an
to the
the former case,
and
the
of
k
sum
consists
8
integer
pairs of terms, whose indices are
complementary (that is, of the form v, a v). On multiplying SS' out, it
is easily seen to be the same as k(S+S').
or JT 2 = 4& + l=a.
Thus we find
S'2 + S - k =
Similarly, if a is of the form 4& + 3, we find that the terms x~
belong
16.
It will be seen
in Exs. 13, 14, that the set of values

*.
n<1
to
',
and then
Thus
we find 88' = (2k+\) + k(S+S') = k+\.

S'2 + S+k + I=0 or X 2 = -(4 + 3)= -a.
[Mnth. Trip. 1895.]

(and indeed a complete
general determination of the sign of
discussion of the distribution of indices between S and *S") belongs to
the problem of quadratic residues in the Theory of Numbers.*
17.
When a
is
an even integer a = Zk, where k
affi+W^-aF*, so that
When
a=
is
odd,
we note
that
identically zero.
4&, the results of Ex. 14 suggest that JT=(l+) N /, but
of this requires some further discussion. t
is
complete proof
18.
Deduce from Ex. 13 that

tan
r- 1895.]
[In fact,
since
#" = !.
Thus, in the notation of Ex.

i
13,
= S-S'-2(A-
tan
*For example, see Gauss, Disq. Arithm., Art. 356; Wcrke, Bd. 1, p. 441;
G. B. Mathews, Theory of Numbers, pt. 1, pp. 200 iil-_>:
Werke, Bd. 2, p. 11
111-117.
H. Weber, Algebra, Bd. 1, 179 Dirichlet, Zahlenlheorie,
;
t Gauss, Werke, Bd.
2, pp. IM
15,
LMFLE8
l.\
x.|
Tin? ivsnli
19.
tin-
r.Mjular
in lead
heptagon
to
.in
easy geometrical
.;
gee at ODCe
'
roots of th-
:io
s7 + l)-l
tint
s<
If
),
.i--
\\nt--
\\>
x4 are
-,
'
root-
+ //. l/<"4
,\\hirh
ular
represents
f..!i>tru-ti..n
tin-
bj rith.-r
tin-
".
l.-liici-
.litl.-i-riitiati.'ii
l.y
(r=0,
duce
that,
if
rw) cosec
u = J
/N 7
that
1
'
'
that
If
an.l
.-c
/r
ticea
tlir-niî,
insnil.nl in
:{
20.
/N '7-l=0,
IIVJH-I-IM-:
n-îilai- lu'|it;i.u iMi

tlii-r
>
[J/
/'
an
is
int- ur,. r an ,l ^
...1.1
an y
intt'irn-
1901.]
jirinn- t.
/,,
shew that
||f-ll
^) i^-A,
2ain(S
6 = irqlp, ami A
wlu-re
is
sum
the \aii ..... f the
MiiiM-
inn-iM-r
any
\\hfii
fr.ni
t.
inch:.;
(l...th
than
/-.
[K:
I
fn.m the th-,rv
Writ., r-.-'". thni
;,
limit
tin*
-f
In.th
:^)7.]
..f
fra.-ti.-ns
partial
K4
sides as J>--l,
and
\v
get
-A-2<
1
-se
-A
22.
similarly that, \\ith
I'ri'vi'
A</>,
;.]
same notation
tin-
as in the last example,
(a + H#)
f,
\vhnv
[\Yriu-
is
mid and
A=
i(X-
+ ]),
n.-i
1,
= e-' '*
2
.r
in
tin-
l>ut
net-d
not be odd.
partial fra.-tions nsed in
Kx. ^l.J
246
[CH.
Convergence of complex sequences.

If (a n )
23.
and
limit,
if
a sequence of complex numbers, which converges to a as a

b is another complex number, shew that values of b n can
is
be selected so as to form a convergent sequence, whose limit

the values of ba
If
24.
is real,
and
tends to
l
prove that any value of x oscillates
drawn
^|^-
])
in the plane of the
finitely
both as x
on one side
on the other side of the
complex variable
#,
*(*-!)(*- 2).
3!
2!
of the line
and
its
modulus tends to
line.
infinity
[Math. Trip. 1905.]
If an infinite set of
points is taken within a square, the set has at
one limiting point (that is, a point in whose neighbourhood there
26.
least
is
one of
to co.
25.
straight line can be
so that the series
converges to
is
an infinity of points of the
set).
[For if the square is subdivided into four by bisecting the sides, at

least one of the four contains an
repeating
infinity of points of the set
this argument, there is an
infinity within at least one square whose
side is a/2 n where a is the side of the original square, and n is any integer.
It is then not difficult to see that we can select a sequence of squares,
;
each within the preceding, and each containing an infinity of points of

the set
the centres of these squares then define a sequence of points
which can be proved to have a limiting point. Finally, we can shew that
within any square whose centre is at this limiting point, there is an
;
infinity of points of the set.]
Sn (x)=f (x)+fl (x) + f2 (.v) + ...+fn (x), and let the roots
be marked in Argand's diagram for all values of n
if these
roots have x = a as a limiting point, the series
Suppose that
27.
of
(#)
has # = a as a zero, provided that the series converges uniformly within

an area including ,v = a.
[HURWITZ.]
EXAMPLES
B.
Power-series.
If H,
1.
then
K are
the radii of coin erênre of
^/,,.r"
and
-!>.<" respectively,
(1) HI!'
is
the radius of convergence of ^Jt .'".

n
is tin- radius of
/.'
/.'
convergence of 2(a n + bn )x ;
11= It' the radius is at least equal to A' and may l>e greater.
(2) If 11 is less than
but
if
[Apply the method
of
Art. 82.]
If a power-series is /.en. at all points of a set \\hirli lias the origin

Art. .">_'.
.as a limiting point, then the series is
|< '<>in|>aie
identically /.en.
2.
\
power-series cannot
within
jM.ints
la-t
tin-
NV- have seen that
4.
Kal'iv
that
if
ha-
this
in
many
ra-.-s
pr..\ed
\\ln-i.
lal:tni;irl.
A.
t..
e.|iial
ml
t.i
Convergence of
th<-n
pp.
=A
2ar*
!:
a singular point "f the

A does not exist, we can deter-
wh.-n thr limit
limits
drtinit'-
/>,
\B
as 7i-*oo.
>/
-jt.r-qx = Q.
[For details, si-i- Van-
tin-
nioilulus of the least root
An
5.
M-itL Soc^ roL
Vl.-.-k.
illustration of tin- last
and
1.
by the
293.]
1<>00, p.
series for
**+i*-K..>
(;j
IMOCiatod
tin-
afforded
is
example
log[(H
which
in
.f
radi
tin-
radiu-
such as
'
ThiMi
is
(te
Si.nirtiin.-s
n-luti.>n
ill.-
limit
!-.
{.WIT ->.-i
all
'in.
lun(o
and
ary) at
purely real
example.)
can be determined
of
IMPLEfi
x.|
|iiadrati.-
is
4-^)'/,,+-j=
Similarly for the series
f^ + |^+5.r+...
oUaincd from
-l<.u (l
(/
and
tht-
l.-it
.litl'en-nt
+ ar<-
If
-"
'
all
converges within
value
If.
-'/
is
the
=0,
tin-
circle
.r|
= fl(>0), shew
value of
the radiu> of
,n
a circle
p.-int
within
l'a,,r"
any
auchy's ine<|ualit
com
within
ie-
ti,
"i".
]
I'
tlien (>e,
/.'.
>educe Caudiy's
that
examine the relation between the regions
.1
rgence
be expected
and 1
maximum
ilian
iy
Sliew tiMtlier that
8.
have the same radiu-
grejiter than the valu-
[Use
-f l)// M +
=0.
>>'
vvln-r.-
get
+ (/
ciil.
'
Com
7.
we
tan r,
singular |.int>. th
under Van VK-ck's rule.
:IH-
6.
if
.'-)
+3K,
associated
:ally.
ineijiialit;
"JV ",
wh.
.<
th-
/,
is
also
248
[For we have
and
an
if
'
is
<Sfor"f(x)
the conjugate to
n n
'2a n'r* /x
where
to /(#).
If /(a?)
9.
= ant
=f (r2lx) is the conjugate

f(xW = f(x}fi(r*lx}
= 2<C^n ^#- n/(#)
2R|/(tf)|
Thus
and
[CH.
D is
= 2a H#n
the
converges for #
|
maximum
of !/(#)-/(
< R,
(Art. 82)
then
-#)| on the
circle
\x\=r<R.
[LANDAU and TOEPLITZ.]

.a l
[In fact,
= morif(x\ -a^m.v- f(-x\

l
2a l =m[x-i{f(x)-f(-x)}~},
so that
which gives the desired
result.]
n
the series
the radius of convergence of 2an#
^dnX"* will converge absolutely, provided that the argument of x is greater
Shew
10.
that
if
is
than log (!//>).

Obtain from the binomial
11.
where
series, or otherwise,
and greater than
v is real
1.
What
the equation
restrictions are required as
to the value of 61
Shew
that the equation ceases to be true for
i>
= 7r, and
explain why.
Find the sum of
12.
cos 30+...
and
of
[Apply Art.
13.
If
is
89,
putting
positive,
15.
[In
.'\
of the result
of
Examine the convergence
[Apply Weierstraas's
i9
shew that
and examine the special form

[Take x = i in the expansion
14.
x-e
v=-J,
(l+#)
when
w = l/10.
m
.J
of the power-series
rule, Art. 79.]
Discuss the convergence of the power-series
Lin-
third,
tin-
...rllirim!
,,f ./"
steadily decreases; see Ex.
-2,
Art.
.]
16.
d of
f/0-
Determine
17.
tip-
-*
expan>i..n of
in po\\er-
it)
;,,/
Jo
Shew
tliat
Jo
-1
the exponential series.]
in
[I'u;
18.
lv\.
(compare
17)
...},
\\here there are
Determine
(/<
cos
+ -2)
b(n
!)
tei-ms in
a siini!
Sh.-w that
19.
if
0=1 -6
the luackets.
/./-.
0|<l/e,
3s
22
Bin 6-*
'/*
cos
40-
33
3!
_!
[JA/M.
[Write
<t
=b=i
the formula
in
K.\.
.f
complex numlters in the place of i-eal oii.-s may

of the same type as that used in Art. 88.]
20.
It'
'
'
'.
sheii
that
ami that the

'
It"
/'**
(1
and
SF/
+ ft)"t
ted
fiom
1891.]
hy an
ariju:
< 2.
0<r<l.
shew
.-...
^H
.\IM-I,
that
-,i.?-i*nf-'
lie
justified
Lagrang*
ay
OM^H
1
lie
r/-/
intr.nlneti,.;
H+^l
21.
The
Ait. 56.
I.
c,,etlicieiits
theorem
that,
if
iii
the liinoinial series.
>o,
+--3><
.neiMr).
[Note that the ai'mimen;
-1) when ril
tirst
summation
me
is
valid
jiproaches the limit 4-Jnir

itive,
and
if
rather IIUMV ditlicult
L-- when
did tli- M-.-..ml
\\
if
'ive.]
>
lj
lut
th-
'250
.r = cos# +
isin#,
[CH.
22. If
is convergent, prove
n -*0 and 2|an -a n+1
Dirichlet's test (Art. 80) that 2a,Xl is convergent if 0<0<2?r.
If 26,Xl is convergent, under similar conditions, then shew that
cn
=a
bn
+ ^ &_, + ...+
by
2^ n Z?n converges, where

4 n = \a n -a n+1 + \an+ i-a n+3 + ...,
provided that
Bn = \b n -
b n+l
6 M+1
6 n+2
+ ...
[Note that the sequences (A n \ (B n ) are decreasing sequences and that

A n \bn
On
Then apply the method given in Art. 35
n (as in Art. 80).
for establishing Pringsheim's theorem on the
multiplication of series.]
|
<
23.
\, and/ = l,
the series
convergent, but not absolutely, while 2 v n -v n+ i\ is convergent.

that the signs of v n are the same in groups of 1, 2, 4, 8, ... terms.
is
Note
[For we have
/5</4<4/2 or /4+/5</2
and
/7</6<i/s or /6+/7</3Thus /2+/3</4+/6 +/o+/7 and so on for each group of 2* terms with
the same sign. The convergence then follows from Art. 21.
2 v n - v n+l =
+
Again,
2) +
)+
:
so that 2|v M -v n+1

24.
(t\
+/
(/2 +/4
(/4 +/8 )
[PRINGSHEIM.]
converges.]
from the previous example that the
It follows
. . .
series
=
converges, but not absolutely, at every point of the circle
|.r|
= l.
[The convergence of 2|-y n -v,l+I |, combined with limv n = 0, establishes the

n
and the last example
convergence of ^v nx at all points except x=\
;
enables us to include x=\.
As regards
that
divergent]
2|tn
25.
*|
If H,
,|2/n
2|t>1
is
= b + b + ... + b n and
l
n rant^s from
to ?>i-l,
if
and from
the absolute convergence,
A", A',,,
we note
[PRINGSHEIM.]
are the upper limits of 1.5,, as
it follows from
to GO, respectively,
Arts. 81, 83 that
'J'li
us,
if
*2,a n
is
a series of positive terms satisfying the condition of
uniform divergence (Art.
83),
we
h;ivr
0,
x
I
In-
if
li
path of approach lying within the limaôn of Art. 83.

^ciM-rallv, for such path-,
llrii'-<:
the right-haml limit Iming supposed to
e.\i>t.
[PRINGSHKIM.]
LMPLEfi
x.|
to
variaMe
\\
part
i-nl.i
theorem
-h
note Blao
in
;in<l
tO
apply
\\ithin
lyinu'
ilt
foil..
tin-
-_
It'
r\ten
to tin- coinpN-.x.
In
tin-
.r-l,
-IM-H. ;is
-\
al"
I:'!',
|.
liin(l
Mt- to
ial
nuinl't-r
finite
;i
\\itlmut
two
In
siiiijilr
\Vln-n the coeffici'
(1)
t<
ii
liiri.-iit^.
\Vlu-ii lini /"/,
Ol
quoted on
is
Converse of Abel's theorem.
27.
If
'-xtelision
t'i!ili.-i
<>,
all
and
.>
fui'tlit-r
.1.
in
ni.'ikr
l.y
any path within
tlii->
it
'--t
i'-t
.n
iotj
[\
JM.^
aj.pr-.a.-li.-
tin-
Ar
[T.\-
[Sincr in ras
of
can apjirnafli
that !'/ cannot
1
(I)./-
lini !'</. /"
llrncf, in case (1)
î-f,,
IM! valurs.
l>y
livrr^.-
convives, and
can
\\e
t'ui lli.-r, ^.n,,
then-fore eijual to J,
is
fi.m
infi-r
tin-
cann-.i
Abel's
l>y
theorem.
In
<
;
it,,
i,
,. then we
,f
r[
r)|
1(1
the ilppel- limit
then
ha\e
th.-n.
./
point
Ait. M3,
and 80
f.
|^
We
1.
path Mich that
iriven
on th-
\\
i.
//,.
t<-nd- t>
liirht
<Mt)
lt
\\
i,,
,i
^
./--
|l-j:|<A
T|^
-x,'a<'h of the
to
<m
as a
in
= |l+^ +
find
lini
z
\\
can infer
nee of
th.
within the lima
and from
tlie
Condition
lini
Thoe
coiiditinns are
taken together, they are
[Tau
(i
'2'i.
i
4-
&
/
;:
-f
Loth neecssary fop the convergence of
2^,
vutl,
.
!'.d.
^.
i;d.
:n,
UMI.
and,
252
Applications of Art.
29.
it
is
easily verified that,
and so
on.
The two
|
we
Thus,
[CH.
84.
with the notation of Art.
84,
obtain the transformation
on the right-hand are both convergent
series
#!-#!< 1+#|, and

1
which are outside the
the latter condition
satisfied for
if
.v < 1
and
some points x^
|
= 1 we have thus obtained a continuation

#j
Repeating the process, we obtain
circle
of the binomial series.
is
where we assume that the broken
line
\xr+l -xr \<\l+xr
from x to xn
= 0,
(r
I,...,
is
drawn
so that
n-1).
For example, by taking
so that
where
Thus we are led

But it should be noticed that
if we take a broken line passing below

the real axis, we find /(-2) = e^ 7r [cos(a7r)- <'sin(a7r)]
we thus obtain two
different values for/(~2) by approaching -2 along different jf>aths.
This
;
indicates (what
(3
and a
30.
for
is
we know
method
which we
to be the case) that f(x)
is
manyvfelued
unit's
an integer.
wimilai
to the last
example can be applied to
find
and
tl
<(-2) = Z7r.
<
same
And
get <^>(-2)= -/V.
|Hiints
with
.1-,
.r,,
./'
lnkni
.1-.,,
lim-
.r
us in the
last
example, we get
passing below the
ival
axis,
\\<-
x
I
If
31.
tin-
aiv
OCX
that
tli.
0<r</0
than
A'
i-oint
VIVANTI
if
possible,
--on-
l!
1111111!..
f>
nin-t
It
\\ln-n
OOnvergen(
>iimiiiMl
!" tin-
000
ill
In-
>inîilai-
as
.-i
I'//.,
{/-
i-inal hyjM.tli-
point.]
Weierstrass's double-series theorem.*
32.
thr
that
1
an-
ai.d
Su|,|,(^.-,
pi
can
kin
\\
si
(fl
a d<>ul>;
That
and
positivr
HIT;-
\v
all
BJ
84)
thm
will
,,f
the
M
it
has a
\i>*
l:..d
(t'..i
radr.
iln-
that
-i
tin-
<_'.,
all
'/////// a
OODY<
f.-i-
OOnvergenl
1>
in
L:
A',
-.)
and
that th-
nnlli.'i
r\.-ry circle
--ss
\\1.
than
A',
'i
90
(1) thr
fin-
r\.-ry
-I
valur
..f
<lt.
n=0
tn=0
tin-
\..w on this
niran IMMII^ taken aloii^ an

is
-ti!ivri'_u'-nt
in,
if
and
//
uniformly
r.jual
i-mi\ rru'nit.
and
so
th-
-t
!-
any integer and
is
'
f '.,
ha\r similarly
\\.
/;
if
M^
is
Eenoe,
thr niaxinnini
,.d
ol
/;
=r,.
ii
N=0
and
1.
LIT.
Q(
<M^
so that
-r.
li.l.
11..
p.
206.
be
254
Now, -we have
identically
F(x)-G(x) =
2 (^
-#,)'"
n=0
number
because this equation contains only a finite
we
[CH.
find
n=0
< ^(Sr, - r)/(r
- r),
of series: am'
(/z)
= r.
if
|
we can make ^fl

n
Thus, since F(x) and ~2A nx are
But, since F(x) converges uniformly on the circle #|=r,
we
as small as
by proper choice of
we must have F(x) =
please
independent of
/x,
p.
oo
71
n
Â
nx
=
EXAMPLES
J\
C.
Miscellaneous Series.
the series
where f
F(x, n) converges absolutely, provided that 2 a n \n ^does so r

n
the real part of x. Thus, in particular, if â nx has a radius
*La n
is
convergence greater than 1, "Za n F(x, n) is absolutely convergent for

But if the radius of
values of x (other than real negative integers).
convergence is less than 1, ^La n F(x,ri) cannot converge.
of
all
Finally,
where A >
if
the radius of convergence
and
|
-a.
a> n
<A
[Note that
2.
then
[KLUYVER,
is
equal to
F(.r,
n)cuT(x)n~
Shew that, with the notation of the

2 a n F(x, n) converge for the same
/
l/vn
3.
Gammafunktion,
1)3, J.)4.]
x
.']
example, the two series

[LANDAU.]
the series 2|i',,-v n+ i| and
last
values of x.
= n*F(x, n)
(2), taking v n
then
are
n
easily proved to be convergent.]
\jv fi\
[Apply Art. 80
suppose that
absolutely convergent if
is
F(x, n)
see also NIELSEN,
1,
Shew
that in the notation of Ex.
1,
n)-F(x-\, n + l) = F(*, n),

and deduce that S/'X-r, n) converges only when f >1 that is, -F(.i;
Shew also that 2( l) n /X#, n) convi
can only converge absolutely.
/'(.''-I,
if
^>0;
4.
The
and apply Ex.

series (see
Ex.
19,
to
Ch.
deduce
.r/(l -.r),
corresponding results
''
if
|.r|<l, and
1 /(I
-.<), if
|.r|>l.
[J.
5.
TANNKUV.]
Slievv that tlu; scries
are botli convergent for all values of ./, except

Kor applications, sec N IKI.SKN,
|
for
1.)
1_^41_.^
represents the function
the
u)
0,
-1, -2, -3,....
ti'-iiniii<ifnnktion<
.^33.
LMPLE8
x.|
6.
of
:i-e
to
iily
that
>0, >ln-u
within
uniformly
th-
dly
'
isilly
tin-
7.
.\
"f
vrlappmu'
a constant
and an-
lin-.
undt-r thrsr
the circle
l.y
an-
\\liich
enough to
:ll
implith-d in CftM th- p"
that
>M.-h
- CH
^.k>
'
circiii!
to
|Tx|,,-.-t
nfi
v
A-jain.
8.
= r,
.-vampi.
iionur a .-traiu'ht
\Vltll
muni"-!-
ill--
/,',
'I
Cw
.p
liy
.-it
any
th.it
f..r all
erge* abftolut.-ly
within
-n.-h
nuiiilM-rs
comph-\
ho series
if
linr.
tin-
tliat
although not
-v,
pointt\\n
in*
of
alon- a
distril.ut.-d
tli. -in
can
lilar statfini'iits
mad.- with
i-
n(iA
fuii'
ions, the
of thi- \\.
set
-iin|)l.
r-
j-oints
notr
nnm- than on-
imt
that
we
(ni
than
niol'i-
^,,t
li-
.-all
]n>int^
liMv
Eence 2
;
.
ron\-fl\
a st-ijiu-ncr of
I'
9.
and
)/,
witliin
fall
JM.;I
elliptic
jiaia!.
<ir.r.
:n.t
th- two Bqnarea (m -
>f
bheory
n<-t\\i'ik of
tl,
In-ill'.;
if
is
conijih-x
i.-al
\vhich
numli.-rs
numl"
i>
part
'
jtoîtiM-,
tin-
B81
.:/
lit.
--+''>! an<l
<.!/
.)/+,/
=A
(I ,
tin-
'" /f
)/(l
,,f
if
MC
if
tin-
othrr hand,
if
that
^3.
//
A
A,,)-
\vhi-h l.-ads to tin- n-ult
-,-n.-ral
th,.
-A
and BO
<>n
ratio
/,'
'i
use
2
1
In either rase tlu-iv
'
i>
tinit.-
i-onvrrirt^ beOtjUae the eoni]
'
up|).-r
limit
'
//
.--nt.]
t-
256
If
10.
of
2a M
convergent, the series
is
n 3/f7* is
[CH.
if
convergent
the real part
M^
is
Thus, in general the region of convergence of *2a n
positive.
bounded by a line parallel to the imaginary axis.
Further, in case 2a n is convergent, the series converges uniformly in a
is
bounded by the
sector of the plane
lines rj=
-i
K, where
is
any assigned
number.
[CAHEN.]
[For then we can use Abel's theorem (Arts. 80, 81), merely noting that
we may write Vn = HM~_ in virtue of the last example.]
It
11.
is
not hard to modify the result of Ex. 27 (p. 251) so as to apply

where f(n} is a positive function which
to series of the type ^Zan e~ xf(n \

steadily increases to infinity (this
the same type as in the last example)

by any path between the two lines ?;= K (where
K is an assigned number), the series has a limit /I, and if lim {a n f(n)/f'(n)} = Q,
then 2a w converges to the sum A.
If as
x tends
[Landau, Monatshefte /. Math.

12.
Prove that
cosx
From Ex.
cos 3^ cos 5x
+ )
^
o
_ sin x
-X(TT-X)TT
and obtain the sum
o3
)=cos.r +
if
3#
cos
shew that
13,
if
+ ...
O^^^TT,
sin
bx
+...,
^3
[EULER.]
-^=v=%Tr,
#2 \_
5x
R2O
of the series for all values of x.
7r.r/7r
8"VT~3";~
# = j7r) shew that
and (by taking
cos 3.r
sin
^p
14.
18, 1907, p. 19.]
\
...
Art. 90, deduce that,
2,
From Ex.
Bd.
u. Phys.,
O<#<^TT,
if
13.
is
to
siii3#
sin 5^7
~ar
~5*~
llTT 4
1536
15.
Deduce from Ex.

cos
3x
1. 3.
and
find the
16.
Shew
sum
v/2
2,
cos
Art. 90, that,

5.r
cos
if
*ix
-\TT^X^.
_TT
= CC
5~3. 5.75.7.9"
~3
of the series for all values of x.
that
where n=0 is excluded from the summation and

Deduce that
0<y<l.
etc -
[Put
= 0/27Ti,
^=27r//
iii
Ivx.
:i.
Art.
!>n.
[L'A.MiK and ScHlAMlLcil.]

and apply Ait. !U.|
17.
integt
\\n
i:\
x.|
i-duce that
r'n.i.1
it'
two even
between th
li
f
aild rXHUlillf the
[Writ,- *
Shrw
18.
and divide
r.i
(A
8r)r,
-f
-.]
that
this r<|iiati>n
ival
int..
and imaginary
|.
_.;-<' 0'r
Sh.-w that
19.
_.
\\li.-r-
//
is
-d
.'
If
20.
"
ditr.-ivn.T
tli,-
tin-
iii
\vhri'
1"'th
and
tin-
m* - ri*
y an-
.-,
lirt
th<-
\v-rn
<>
summation.
and
j.,
3.
[Kx.
w- ha\-
tlic
Art. 90.]
following
r-
-7r(./--i),
--('
or
-.{),
or
i'.-v)
,-'-.,-
By writing
+J
is
if
-r>v,
if
.'
if
= 7r[</>,(,-) +
or
22.
'
~~n
4n*
i-xdudt-d from
00
<.V,o>
v.
and ..W-rve that
90,
><./<!, we
<
to
int-^-!
eerifl
'
,~i
If
rintar
between
an int-grr and
is
V'
21.
-L'>/)
T-'[</,,(,-)
+ // -
J],
if
.r>y,
+ </>,(,/) 4- .i-H,
if
.r<y,
</,,( / /)
the IVrnoullian function of order
2.
Mt. 90, or otht-i-wise. sh-w that
sllll:
Use
Stok<-s'>
mil- (Art.
0<<>r.
Sllll;
!i)
t.
dcduc> cadi
of
MI
23.
i-t
77
Slllll
|c=
l.v
rxainjilf.
C08ha.r_l
= - -f- ^"
\\'t~
a
_sini
and obtain
the otln-r
[JVM.
ditlrivntiatin.i:.
that
2
-
--
>
1
-I)"" '**"
+(-!)"
7T
'
<..<...,
'/'
l>y
258
Shew
24.
that the product
converges absolutely except

[If the general term is
when x
is
a positive integer.
lim(iAi) *
!-,
#2 [(H-.r>4) and
Evaluate
25.
vol.
[Glaisher, Proc. ZOT^. 3/a*A. Soc. (1),
26.
From Ex.
27.
Prove that
10,
7,
p.
23.]
Ch. IX., find the values of
where
log H c
any determination
of logc being taken.
Shew
[If
= J(-l+iV3),
that
[HARDY.]
n^l=|
+
28.
[CH.
so that
*i
= l, we
[GRAM.]
</J
have, as in Ex. 21, Ch. VI.,
Now
29.
write
Shew
^=1, and observe
that
"1
that
TT
sinh (?r.2\/ 2 ) + sin (
=
4
J_Y_1-- i),
[Math. Trip. 1888.]
[We
L
where
_^
have
^
w^-r
$f4?\i+to
Thus, as in Ex. 22, Ch. IX., the given sum
and
1.
is
this gives the required result.]
30.
Apply a method
similar to Ex. 29 to find
259
\MI-LI-
x.|
Prove
31.
ti
that
lr;i-t
til.-
Vtlllf
/j_v
'
[.I//'/.
*)+/
and so the
%U+(* + i>)~n+<*-iy)r
sun.
Shew
32.
7 rip. 1892.]
that
-1
4/1-
-l
..
,7T V
7J\/3
êch-^-.
1898.]
[M,,-
+
,T^ 2^^"2M^rV'
\\li.-iv
<,>
= ^(-l4-'x
ii
whirh
is
.iriv-n
A7r{cosec(i7rto)
easily reduced
Sh.-w
33.
seriei
tli*
tin-
t<
-jivi-n
fm-ni.]
tliat
aivtan,
-arc tan.,
.,
.\
tan
......
...
=arc tan
1891.]
[.!/
is
[It
and
it
<>asy
t<>
prove that
aivtan-5
"-
-.'-
follows that
th.-
- an- tan f
..I -fare tan
to
il
îvt-n
\
3/
..I,
x
\,
\vh-i.
34.
:;.
iil
.\i'pl_\
the îvcii foiniula.]
Slu-w that
'=
"
L
In partii-ulai- \ve
tintl
\\
\n + .r/
ith
H3
v/
1
l<s
i
0-1=
T
--
+ are
tan
rsinh(Tr.r)~l
.
-.')
260
[CH. X.]
[We have
and
tan(^-)
log
In each of these, change x tox + iy, and equate the imaginary parts on
the two sides.]
- p,
35. The points P, Q have coordinates (p, q\ (

the point with coordinates (na, 0). Shew that
then
[If
q) respectively
is
if
[Math. Trip. 1894.]

tan0=tan(7rp/a)coth(7r?/a).
it will be found that
we write p + iq=x, -p + iq=y,
id = log sin (n-xja} - log sin (iry /a).]

36.
Verify that,
if
where we are to put EP

[Apply Art. 94.]
37.
Shew
a positive integer,
is
= -B BP+ = Q
l
after expansion.
[Math. Trip. 1897.]
that
fcotf-1-^fj-B^-.B.f;-...,
and that
x
g 2 sin
_Bx^ + B^x*_ + BS^ +

4 4!
2~2!
(Jar)""
38. In virtue of Baker's theorem (Art.

powers of x if |#|<JTT, and if we write
the coefficients
En
6 6!
are called Euler's numbers.
# = 1, ^ = 5,
^ = 61,
111
l-3^
2
we can expand sec#
84),
Prove that
=1385,
+ 5^-7^+... = 2^,
[See also Chrystal's Algebra, Chap.

39.
Shew from Ex.
in
XXX.,
TT''"
+I
^ )f
3.]
37 that
cosh x
cos x
iog--^-
-^ n+lcos
,<,_
,,
Bn
.'-"
^-)^(2^r
[Math. Trip. 1890.]
40.
Assuming
Stirling's
formula (Art. 179), shew that

J
when n
is large.
CHAPTER
\
XI.
CONVERGENT AND ASYMPTOTIC
SKIM
Short bibliography.
96.
follow in-
Tin-
are
th--
chapter has been d'Ti\<-d

r.'i;K!.,
/.-
BOOTOeS
principal
which this
l'r<.ni
is,
1901.
2,
Normal
1896, p. 103.
10,
1899, p.
n. ISDO,
<
H.MM'Y,
II.
;.
'.<> i.
,,.
^ //,///.-////
LI.
I-:.
BtoT,
ij'.iT
'am
'///
1.
IN.
]>.
!.
H',
(published 1902).
./" '/>////
nf
M,iti.
01.36,1908,1).!
I'.HH.
p.
317.
1
II.
l^"'
..
I'.
..
<
97.
tin-
and
iiu'tlii..ls
<!'
in jiarticular
l>y
Alirl
in u-in-- in ii-r
*T
art-
trnii.s
n..w rallc.l
ln---in
If
before
and
tin-
tln-ory of
j..
1906.
put
on
convergence had
t
ln-
rs
and
sum
part of
ti
and
tli.-y
r.-ach
s<
minimum, afterwards
to a stage at which the terms
on these serifs use
viotis
little
n-tical
naturally used i.nly

asymptotic (Alt, l.'JO below); in such
tak.- the
sinv
'audiy. niatlicinat iciaiis liad
th<
in
listi;
IM-.-II
n\ -i-r^-nl aeriea in Imtli
to decrease,
we
had
Analysis
investigations,
In nunirrical Work, however,
incr
l!n;.,
i'/.ig,
Historical introduction.
footing,
which
Yolk.
r.rfniv
tin-
I'.
\.-\\
trgtnt aa including
quotations.
a<i
NON-CONVERGENT AND ASYMPTOTIC
262
SERIES. [CH.
XI.
we may hope to obtain an approximation with a degree of accuracy represented by the last term
retained; and it can be proved that this is the case with
many series which are convenient for numerical calculations
(see Art. 130 for examples).
are sufficiently small,
An
important class of such series consists of the series used
by astronomers to calculate the planetary positions it has

been proved by Poincare * that these series do not converge,
but yet the results of the calculations are confirmed by observation.
The explanation of this "fact may be inferred from
:
Poincare's theory of asymptotic series (Art. 133).

But mathematicians have often been led to employ series of a
different character, in
which the terms never
decrease,
and
Typical examples of such series are
increase to infinity.
i_i + i_i + i_i + ...

1-2 + 3-4 + 5i^2 + 2 -2 3 +
(1)
(2)
may
(3)
1-2! +3! -4! +5! -6! + ....

"
sum " of a non-convergent
(4)
Euler considered the
series as the
numerical value of the arithmetical expression from the

expansion of which the series was derived. Thus he defined
finite
the
"
sums
"
of the series (l)-(3) as follows
J__l.
~1 + 1~2'
and
^_
_1.
,^_
-(1 + 1)2-4'
his discussion of the series (4) will be
of Art. 98 (see
p.
_L
-1 + 2-3'
found at the end
267).
In principle, Kuler's definition depends on the inversion of two

limits, which, taken in one order, give a definite value, and taken
in the reverse order give a non-convergent series.
Thus series
lim l-limoj + limic 2 -lim^ 3 +...
(!) is
as x tends to 1
So, generally,
" sum " as lim
Euler's definition replaces this
by
2/w (c) is not convergent, Euler defines the

2/n (#), when this limit is definite a definition
if
X-+-C
which should be compared with the principle

the beginning of Art.
*Acta Mathematica,
t.
!M).
13,
1890; in particular
i:t.
L'ormuluti'd
.it
fflSTOR*
97|
can
definition
tlii-
'I'--
aUo
raided the
('all.-t
ohtained
IM-
[KTEODUi
\i.
l,y writ
in;.;
in
the series
tliat
>lj-cti>n
<
i<
he series
+*-+...,
(5)
,
whereas the left-hand side th.-n

This ol.j.-ct ion ..!' 'all.-t 's \\as
istead of $.
n,,-i
vrho suggeated that
ami
that th.-n
-h.ul<l
ti
..tiM
tin- <l-ri\-.-.l
rk of Lagrange's,
written as
,\
!>
1+0-1+1+0-1+140-1+.
Tin- la-;
!.<>.
value
imfl
|(i
....
î\,. n
aeriee
).
<
that
BO
.
},y
|,
1(
1.
may
with
ftgrOe
had
n:athMiiaticians
i]l.1
is
.f
.">
th-
In
i-
nriînal
of which the
o, ...,
KulT and
l>y
Dnn-l
M-i-ii-^
usin^
the
in
sutlicient ly
if
other mat heinat

which do not converge,
:
statement
that
ollcr
the
-vil-
experimental
they \veiv cnnverevnt led to
^(.o<l
when they presented
,/,/////,/////.
rxumplr
atl"nl.-il
i.
i,
the majority of easefl
in
results t
COrrecl
themselves
^\>\<-
o.
sum.
the use of such series as
that
which
the fact that
t-
made numerous discoveries

\\e
i,
<>.
Kulci-'s
u-'ith
cin^r
lla\inu rc-'ard
'-
hand
.ft
the aroma are
to
tin- "
1.
p.
1>\-
Bi t"
tli-
.f
pa-au
'
in
rier
..litaiii
is
\vliat
ol)t;iiniiiu
\v<-
should
call
IK>\V
'
K"iui.
inctk>O
f\
ih
and
finds that tlu-
la
*Thi
tluMMvm
of
in
Fiolx-nius c'
Ait.
more satisfactory
l;,l.
x'.i,
ls>u,
],.
basis
_'.,
iiat
."!
liln
*-!
:iu
'
,(
lim
-f-
1)
J.
with a few
MB
|>t
al-.
sum
264
coefficient of sin x appears as 1-1+1-1 +

be expected to be
if we adopt Euler's principle.
As a matter of fact this is correct, since
Tims the
...,
SERIES. [CH. XL
and may therefore
sinh
x sin x dx = | (cosh x sin x - sinh x cos x \

2 f
so that
f(x)*v&xdx=^.
Abel and Cauchy, however, pointed out that the use of non-
convergent series had sometimes led to gross errors; and, in

place mathematical analysis on the firmest
their anxiety to
foundations, they felt obliged to banish non-convergent series

from their work. But this was not done without hesitation
;
thus
Abel
writes
to .his
former
teacher
Holmboe
in
1826
(Oeuvres d'Abel, 2me. ed. t. 2, p. 256): "Les series divergentes

sont, en general, quelque chose de bien fatal, et c'est une honte
qu'on ose y fonder aucune demonstration ... la partie la plus
Pour la
des Mathematiques est sans fondement.
il est vrai, mais
les
resultats
sont
justes,
plus grande partie,
c'est la une chose bien etrange.
Je m'occupe a en chercher
essentielle
la raison,
probleme tres interessant."
And Cauchy,
in
the
preface
of
his
Analyse Algebrique
(1821), writes: "J'ai ete force' d'admettre diverses propositions
qui paraitront peut-etre
un peu dures
par exemple, qu'une
"
serie divergente n'a pas de somnie."
the asymptotic property of Stirling's

132 below), by means of a method which can
be applied to a large class of power-series. But the possibility
of obtaining other useful asymptotic series was generally over-
Cauchy
established
series (see Art.
looked by later analysts and after the time of Cauchy, workers

the regions of analysis for the most part abandoned all
;
in
In England,
attempts at utilising non-convergent series.
however, Stokes published three remarkable papers! (dated
1850, 1857, 1868), in which Cauchy's method for dealing with
Stirling's series was applied to a number of other problems,
such as the calculation of Bessel's functions for large values
of the variable.
*Of course no one would now regard

made it.
('auehy's statement as unusual,
the sense in which he
fSee the references
of Arts.
1.'!::,
i:i.~>
Ix-low.
M.I:
97, 98|
no
Hut
-eii.-ral
ON8IDBB
\l.
tin
when papers
I'oinc.i
written hy Stiel'
of researches has.- heeii puMi-hed
until
OOming
ISM!,
.1
aO8 that time a nun
the
1'nuvl
1>\
in
we
in- articles
t'oll.\\
our exposil ion
tin-
we
mod
the
.r
r.
<!> not
General considerations on non convergent

\ii-\v
ol.taiiK-d
r.-sults
th'-
I'.-rlVrtly
j.ivcis,-
may
uiili-r |M
which
niranii
1-
M---I
.r
!'<
Tim- we attempt
jtrictioDB,
not).
tvergenl
as-ociatf
t<
function
dclinitc
jMTfcctly
the -.sum" of the Beriea
(letinition cho-.-n
he such that
M'
use of
calculation.
to a
is
\\ith
it
numhcr. which
and which
olvins
of course
i-
It
nl
l;i
the n -suiting laws of
COUT
arl.itrary;
hut
tliat
it
the
>lnull
c-alciilatioii
aeriea
not
i-
that the
i'leiit
to
BUBD
with
'i
conf-'unded
!,-
a noii-
with the sum of a
in the sense of
\rt.iii:
hut it will
Convergent
confusion if the deiinit i<ui is such that tin- same operation,
.
applied
yields th-
Qvergenl
.'In
i..
I..-
of the di>t inrti'in
|i..ii,
tliat
..in
t.-il
ln-t \\t-f
"
hi.-,
Tliu>
Of I
hat tin-
Kuii-r.
sum
lu-
"
't'
says
;it
nmi
;.n\
los-l
(jj<
ii>u.-ly raiiiu't
rate,
mi
1
.'.
tli.-
when
ordin
wai perfectly
I-:
>nlin:r
in tin-
th- larp-i-
in
1;
sum
a'
.nvcr'_:-ii;
i>,
tin- iinira
doesS.
i-th tlu-
lull
i-:
....
with those of t-onver-eiii &
convergenl
d.\
to formui.-i'
call
th-
can attach
n a
cnal'i'
lie,
\ve
ivergenl
of formal
juirpo-.^
)i
series.
l.y
]>ast
prnhaM.- that
ins
sud
th-
in
ay> adopt hisim-th
al
In
of
e..ntine
par!
non
98.
non
^hall
i)l'
theory
MMI. although
theory. 9OQ16 of which
tli.-
Ait
in
<|Uot-d
In
nil
account of this
IM-.T;.
rory see
Ar
-w).
266
SERIES. [CH.
XI.
And he adds, after referring to various difficulties, that contradictions can be

avoided by attributing a somewhat different meaning to the word sum. "Let
us say, therefore, that the sum of any infinite series
the expansion of which the series
infinite series 1
-x+x
-x* + ...
is
will be
--
is
the finite expression, by
In this sense the sum of the
generated.
because the series arises from
the expansion of the fraction, whatever number is put in place of x. If

this is agreed, the new definition of the word sum coincides with the ordinary meaning when a series converges and since divergent series have no
;
proper sense of the word, no inconvenience can arise from this new
terminology. Finally, by means of this definition, we can preserve the
utility of divergent series and defend their use from all objections."
sum,
in the
In writing to N. Bernoulli (L. Euleri Opera Posthuma, t. 1, p. 536), Euler

adds that he had had grave doubts as to the use of divergent series, but that
he had never been led into error by using his definition of "sum." To this
Bernoulli replies that the saine series might arise from the expansion of
"
if so, the
sum " would not be definite
more than one expression, and that
to which Euler rejoins that he does not believe that any example of this
However, Pringsheim (Encyklopddie, Bd. I., A. 3, 39) has

examples to shew that Euler fell into error here but in
his definition almost exclusively in the form
Euler
used
practice
-could be given.
given a
and
if
number
of
restricted to this case, Euler's statement is correct.

from these passages that Euler had views
which do not
bottom, from those held by modern workers on this subject
although of course his attempted definition leaves something to be desired,
It will be seen
differ greatly, at
modern
in the light of
analysis.
borne in mind that the legitimate use
It is to be carefully
always symbolic the operations being

convenient abbreviations of more complicated trans-
of non-convergent series
merely
is
"
shorthand
Naturally this
background.
labour
us
to
the
of
not
enable
avoid
representation" does
justifying the various steps employed but when general rules
formations
in
have been
laid
the
down and
them with confidence

It
the
may
work
in
firmly
established
we may apply
particular case.
that we might just as well writr
be
urged
very likely
in full, and so avoid all risk of misinterpretation.
any
But experience shews that the use of the series frequently

suggests profitable transformations which otherwise might never
be thought
An
of.
c.\ainj)k;
of
Nicholas JWimiilli
this
(//.
may be taUm fr>m

Euleri ()//'! /W/tuma,
KuK-r's
t.
1,
ronvspmuli'iu-.' with
p. f>47)
uhrir
ilu> ivul
WTBORA1
BBL'fl
98,99]
Caning
olij.
ll.-
th- M
that
lir-t
prOTM
'!'
which
nlitain-
In-
tin-
ti-
intrirral
if
|-;h''
"
(l
"th.T hand,
tin-
in
liv
n>inu' tin- rul-->
which
li:nl
In-
olitain-d
f-r tin-
transformation
aiid
it
j?
1+
i-h
1+
1:
&r
2.r
rr
l.v
|.r..v.-(l
N"\v this n-latiun
l.t-.>
n.it
1+
1+
1
was rntiirly guided
Writing
.r-
1,
1+
1-f
naturally \vithut
ill
iUggi
alit-ady
\v..rk
3oi-t-
I.
__-__
"~Ldt =
1
1+AY
1+
3.r
TT~ r+~
nnai -ki-d,
it
is
r\ idnit
tlie
tli..'
thr aim of evaluating the
1>\
Kult-r cl-tains
th<- c'.ntinu.-d
fi-.-ni
fia.-tion tlic
nuivfrgente
1>
and
Ity
usini;
tli-
l-'Uh
and
1-lth
In-
int"-i-
tin-
numerical
value
Hr then
sulti-a-ts this drrimal fi'dii
and
infd-s that th-- Value
-f
tin-
99. Borel's integral;

Tl).-
summable
general considerations
Specialised ly adMjitin-"
as Insum of a n
aome
<>!'
th.-
series.
la->t
m-ticl.-
lta\.-
mivrntinnal im-jmin^
n\\ to be
In-
tak.-n
Hardy,
in
tli-
inllowin-- jn-incijilr
definib ordt
vdku
A'.
Imt.
[.MJM-I-S
:
//'
i|ii"ti-.i
rniulatt'd
tlu-
268
Y,
meaningless expression
we may agree
SERIES. [CH.
XI.
as
to interpret
X.
meaning
For example, the equation

~=o
f(x,
n)dx =
[lim/(#, n)]
JO
dx
n-><x>
But
true under certain restrictions* on the function f(x, n).
definite
may happen that the right-hand side is perfectly
is
it
while the left-hand oscillates between certain values (which may

oo and + x ).
In such cases, Hardy's principle is that we
be
may
take the right-hand integral as defining the left-hand
limit.
sum
In particular, suppose that f(x, n) is the

of a series of functions of x; say that
to
n+1
terms
and that
The equation then becomes

o
is
<&(x)dx
con-
Jco
o
vergent,
when
integral ivhich
convergent
the series on the left
may
special type of function

interesting results is
<t>
un
not
if
"
so,
we have an
sum "
of a non-
series.
The
where
is
be taken as defining the
is
n (x)
independent of
<j>
n (x)
which has led to the most
= e-*un x n/nl,
x.
n (a?)
dx = un
(see Art. 178)
too
o
arid
$(x) = e~*(u Q + up + u 2 |j + u 3 |j +...)
= e~ x u(x)
us say. We must assume t that the coefficients

that the seriefl u(x) converges for all values of x.
let
*A
fAt
un
few simple cases will be found in the Appendix, Art. 172.

least for the present; but see Art. 130 below.
are such
99, 100J
Then provided
we may
tl>
i*
I'll
>\
//</
>l
//
integral
agree to associ
series
tin-
d.-n,t,-d
!>
tin-
l.y
called
IK- rall'-d
i;ty
sum may
Tin-
ma\
this int.-^ral
*grie*
/A*-
//-/'/A
<
/'-
conve,
MMIMIU1
syml)ol
Borel, who deduced it. 1,

definition u-ivni hrlnNV ill Al' III: liUt BUICe the
This definition
tlir
lias
is
due to
more serviceable
].ro\.-.i
rd
100.
oltvinus that
tin-
i'uinlaiinMital
in îvm
<l-finit
Û
lead to diilirulti.'s- unl.-ss tin- .sum

\vli-in-vt-r
in tin-
d.-tinit
it
.n.
i'l'
Fr.
is
consistency
St.
tin-
s.
rics
\J
//
arliitrarily
-e~ x
fi.
converges uniformly for
''
within an interval
,/
',
.,
df
that
if
e,
/j,
and consequently
satistinl.
<
\alucs
sum
ffrgent,VfG rnn tind a nunil).-:
Then
will
artirl'-
tin-
imw pm\
shall
last
a^r.--s with
"
oditiOD
all
in\-<-sti-j-ati<Hi-.
suls-.jurni
tin-
Condition of consistency.
is
It
in
as
tlii>
i'r..m
(0.
,\
>.
\\ln-iv
may
be
1,-r
S^
f\o
|*A
''~ z
,c?a;
Jo
by A
rx
(
fA
where
A_
te(_iT+7(n+!
V(// + D:
,\
New,
since
aence; ami
Alu-l's
is
,\
jm.siti\r.
L':!
;
'
tin-
t> tin-
r..nM-.|in-ntIy \vc
Icimna (Ait.
A(\,-\
.+...toa
sequence (X,) ia a
can apply tin- rxtriidt-d form of
-um
li
\,,.
<//(\,
and we sec that
-,\,
/(
)-f //
270
SERIES.
[CH.
XL
where H, h are the upper and lower limits of UQ+U^ ... + ur as

r varies from
1 and
to
m hm are the limits as r varies
from in to
Now
ûn
if
hm = s
to oo
oo
converges to the sum

s + e, however small
Hm =
X and X m tend to the limit
choose
so that
maybe; and as X tends

because
is fixed.*
Thus
e
1,
[h(\ Q \ m + h m \ m = h m ^s-e
lim
and
we can
s,
- X m ) + H m \ m = Hm ^
lim [#(X
A-*- oo
Hence we
see that
00
A->oo
But
arbitrarily small, and so (see

limits of 2un\ n
is
maximum and minimum

to
s,
Note (6), p. 5) the

must both be equal
so that
00
lim
A-^co
what
or,
But
if
2u
n\i
=8
_ -^v
the same thing,
is
2un
however great
>
we can
diverges (say to +00),
may
find in so that
be.
Thus, since
- \ m ) + h m \ m = h m > N,
lim [h(\
X->co
00
we
lim
see that
A->->3
Since
is
arbitrarily
is
divergent,
with
this
inequality
That
X.
v
when 2*** ^
is
great,
tends to infinity
n
^ un^n = ^
is,
divergent.
implies
that
the integral for
Of course
this
only as
it
should be
otherwise there might be a risk of
inconsistencies arising.
*For
=l-e-A,
Xm = l -
expressions tends to the limit
1.
+ i\2 +
...
m
;X
and each
)e-*>
of
these
ADDITION 0V TERMS TO
100,101]
MMAHLE SEWER
SI
00
Relation between
101.
and
//
'/
>
not follow immediately tY'-m their definitions
.loes
It
suma
8= <*UH and
,
<J/
are related
such a
in
existence
tin-
thai
way
the
<>f
the
tliat
tin-
The
ther.
existence
either
>!'
definition
ini]
relation
tin-
9lJo
r-'y/t. /)'/.',
.s'=| e-*u'(i''i
Jo
////
a;
1186
'
/.''
are -m-h that u(x)
the coefficient
any valn-
I'oi-
+";:.).+
of
./
an-l
o MISC. jn.-nt ly
Art.
")0
<
can be
OSed to justify trnu-liy-tenii <litr.'r-ntiation of u(x).

(
)n
int.'-i-at
iii'4"
tin-
for
expression
relation
ny
8*
tind
\>
tin.:
x'^
+x.
g-w(a;)
I
This shews that when
l>oth
>'
>,
lini
//
also:
fcfl
exist, tlie limit
(./)]
><*>
and the \aine of
this
cnld
otlierwise the inte-'ral for
must
limit
not conver^.-.
he
zero,
(S.-c
Miir-
Art
Appendix.)
Hein-i-.
i>i-<,
j)i'n|ierti'>.
Beiiea
>>,/></
s'=/i u
/,,,,-,.
<
M'
flnif
integrals st
Jx>t/i
thus the t\\n
aeriei
respect, ana In--, .us to

course any failure of the analogy
in
this
and w.u
the definition which we have
.lvious ditlicnltie^
bere would
reduce the
ly
the
sel<
"
li
utility
"
sum
of
sunuuahl'
I'.nt
it
must be rememhnvd
that
the oonv<
of
Jo
'o
l>y
no mean- enSUTQfi
ti;
lini
delinit- value
X >ao
as
is
cli-ai-
<>/ ./////
fn>m Art.
definite
valw
Dili
!'
the
Apii.-nilix.
It follows that
272
An
paper),
example of
by taking
been given by Hardy
this has
SERIES. [CH.
XI.
30 of his second
(p.
Then
but the integral for
s'
is
which oscillates.
Thus 5' does not converge, although
The reader
does converge.
will find little difficulty in seeing that here
the algebraical transformation will be found in Hardy's paper.

00
Although the convergence of

that
yet
Jo
e"*u(a/)daj does not ensure

Jo
definite limit as x increases to oo
u(x) tends to a
we can infer
lim e~ x u(x) = 0,
that
e~ x ii(x)dx is convergent;
when
and consequently
the
integral
that the integral
e~*u(x)dx converges
also.
For, write
e~ xu(x)da
y= Jo
/
Then
-*
d?y _ _ x
,,
and consequently
The
last expression
and then
(TJ
T;')
tends to a finite limit
tends to the limit

e
If
these
YI
;
0.
?,
as
increases to
Thus we can choose
oc
A",
so write
so that
if
has an unlimited number of extreme values, //' vanishes at each of

Thus
hence at any extreme value beyond A', we have rj <
.
|T?|<
or
Iini7;
if
r;
= 0, and
if*>.\',
hence Iiui7/ = 0.
has not an unlimited number of extreme values,
must approach a definite

77
approaches the limit -a. Now, from Art.
finally of constant sign,
say a; thus
r)'
and so
?/
must be
limit (Art. 2),

152, it follows
ADDITION OF
101, 102|
that
if
limit
tin-
:ij,|,r..:i.-li.--
-/
liiu
and
ii
\]
\ i.
tli.-n
-o,
'.r
//
approaches the same
limit.
hav
Tim* Wl
l.ut
si
hu\v
,,
Inn
",
if
be zero.
tli--
Thus we must always
liiu
?)
= 0,
lim
;/'
= 0.
X-ôo
X->oo
or
-a,
(,/ .r)
"ttto+l e-*u'(x)dx t
Jo
Jo
Mm
and
'-'//(. /')
= 0,
x->>
]>r<\ idcd
that
tin-
integral
Accordingly W6 have
vmmdbUj
//"
//
n-sult
tin-
00 "/>"
vergent.
Jo
//' f/te
the series
..
//'
r' I
1.
ion between their
mode
?////>
of stating these
i-.-sults
/>c
///<"/
and
series
is
the following:
prefixed to
remain swwmdbl*
//" .sums will be related as if (/' tm> series were convery
Hut fhi- removal of even
6
singl* term from ///
tin'
of tl
////// destroy
j>n>/r(;/ of swmrm
,
v.
/"
oortefl
"'iff
Some obvious theorems.
102.
ao
follows immediately tY<m the d-tinitin> that
It
oo
'
are BUmm&ble, tli-n
suniiuahl.-.
Kurtht-r.
''
is
any
/\n n +''n) and
1'act.u-
independent <f
o
l.s.
(*u n
and
oo
and
ii'
if
/'.
are also
274
We
SERIES. [CH. XL
should note further that the addition of a constant to
any term of a summable

same constant
the theorem on
are
series is equivalent to adding the

this of course is a special case of
the addition of summable series. For the series
sum
to its
summable and equal
all
to
k,
CO
Thus
o
although, as we have just proved (Art. 101),

is not necessarily summable when
+ u^ + u^ + us +
103.
Examples of summable
is
so.
series.
i_i + i_i + i_....
Ex.l.
u(x) = \-x +
Here
x2
x3
^-^+
...
=e~ x
/oo
e~ x u(x)dx=
foo
o
Assuming
we can
e~ 2x
dx = L
Jo
this series to be
[Compare
Art. 97.1
summable, and to have a sum
evaluate s by prefixing a term
s,
at the beginning
of the series (Art. 101): this process leads to
s-l=- + 1-1 + ...= _ s

1
so that
(by Art. 102)
= J.
l-t + t--t*+....
(t>I)
tx
=
we
find
1,
u(x) e~ and then we
Ex.2.
Here, just as in Ex.
get
Jo
Here again, assuming the series summable, we obtain

1 at the beginning of ts, giving
value s by prefixing
}
= !/(!+/).
1-2 + 3-4 + 5-....
or
Ex.3.
^(oO=l -2u; + 3-4+... =er'\l -
Here
and so
Jo
e-*u(x)dx=\
Jo
e'^(\ -.e)rfa? =
'>
i-J = J.
its
LMPLES OF 9U1CB LT
102, 103]
Assuming
!UI ,1
!!
siimiiialiilit y.
tli.-
us write
9.1-2+8-4+5-64
0+0-0+1-2+8-4+J
so
Adding
til,-
corresponding
MI. 102).
i.-rm-.
1-14-114-1-1 + ...-}
<
Ex.
(by Kx. l)
J.
(7=l+co80+ cos20+cos30 + ...,
4.
0+8H10+ sin 20+n30+....

Thus
C+i/?=l-;
ami the associatol
i'unrtion
i-
)=1+^,
Of
.)
Hi-jir,-.
tind
pi-.
= er(co+i8in)
that
id.-d
is
not zero or a
6U= T
Jn
inulti}
iiiin^-
nlitaiiu-d
l>y
nn-aiis
the
of
= s(H-
-n
COSO
Of coon
In-
1)
Mini
thi-
,\
+...=t'X 1 )(^)
2\
S1110
suimuablr property,
sann- drvici- as
tin-
in
tin-
MHH can
Kx.
'2.
which
[nation
e(('+;N)+l=
and there!
(7-fifl
Henl
Thus wr
with
tli-
l/(l-<
fmind.
rrsnlt just
havt-
1+costf + n.
()
+ sinH + siii2rt + .sii,:^+...,
.
and
In
likt-
nianin-r
w'+
find
t!
')(-
msf) + rMs:>f) + (.'0830+...=

sin
I'.y
0+ sin 26
rhan^-in^ I'nnu
'
>+...-Jc
t-
re
H.
.=
l'tain
.
,
rin0-am20+ain30-...
-..)
,__
276
C=cos0 + cos30 + cos50+...

$=sin0+sin3$ + sin 50+...
Ex.5. If
and
(7+ iS = e
we have
using Ex.
(l
+e +e
4ie
XI.
+...) = "/(!
-e
<')>
4.
< <
Hence
6
(7=0,
TT).
(0
=Jcosec0
be noticed that these two values are the same as
will
It
ie
SERIES. [CH.
are given (Ex. 4)
by summing
cos
and
sin
but there
no general theorem which would justify our
is
new series
to
By changing from
equating these
cos0
G and
to
0,
JTT
cos30 + cos50
S.
we
obtain
â^\
...=
-sm30 + sin50-...=.
Ex.
to
6.
It is easy to
+ 0, we
deduce from Ex. 4 that
is
if
not equal
have
cos
cos
sin
sin
cos
sin
+ cos 20 cos 20+.-. =

+ sin 20 sin 20+ =0,
+ cos 20 sin 20 + = ^sin 0/(cos
-g-,
...
cos 0).
Absolutely summable series.

Following Borel, we say that a series
104.
is absolutely
summable
if the integrals
/
Joo
o
e~ x u(x)dx
and
\
e~ x u^(x)dx
Jo
where \ represents any index of differentiation, are

lutely convergent.
It is at once evident that
if
all abso-
the series i^ + 1^ + 1& 2 +
. . .
is
absolutely summable, so also is the series obtained by removing

any number of terms from the beginning of the series for
the integral function associated with the series
:
by hypothesis
Jn
e~ x u'(x)dx is absolutely convergent.
OL1
103, 104|
It
\irtu-
in
follows
MM .Mill.
-I
what was proved
of
lim
Art.
in
1"1
liine-*wA (a;) = 0.
-0,
X->!
'/
nimbly.
f...
is
al*l tii'i >j swm/mable,
<//,i/
sums
their
'/
Bumrndble, BO also
'I'h.-it
is
can
\vt-
is,
90 also art
series
/A
as '///"'// were convert)'

n -f "._,+ " 3 + ... '-s
<n->
And, conversely,
"/ ;
M'rii'*
//"'
'
'
series
th<>
formed by prefixing u
th- convergence
ini'.-i-
<>!'
tli.-
e-*\u(x)\dx
Jo
'0
IV.
.111
that of
I
"
o'cte.
Jn
'0
T.
this,
irvr
h/t
us writ,'
Jn
Thus
BO
')^
li
\U(.'
Id -nc.-
''
,
//(./)
urrtainly OOD
J./'
&(x)d&
Jo
Jo
I'.ut
(Miivrr^vnt.
the r..n\
inl'.-nvd lV"in that of
.!--.
-nee of tin/ /.
l>y
last
iiit'_Ljral
is
can be
the ar^tinu-nt of Art.
HM
Jo
Tin- ]r<t', Imwrxcr, can

of tlu- inonotniiic nature of
above.
1..-
simplitird
h.-rc.
in
For we
liave tin-
fJT
o
identity
/'- -' -*A(Z)-f

Jo
"
and consequently, because
\
o
.1
Jo
/>(
and
ive,
-I
Jo
\'irtue
278
x
Hence, since the integrand e~ <^(x)
SERIES. [CH.
XI.
the integral
is positive,
Art. 166 of the Appendix.

It will be noted that this argument cannot be reversed
is
convergent,
by
that the convergence of
e~ x
(j>'(x)dx is
and
not deducible from the
Jo
Jao
o
e- x $(x)dx.
e-*<$>(x)dx
x
convergent, but lime- (f>(x) does not
is
exist,
may
be obtained by taking
so that
It follows
from a general result due
convergent (see Ex.
is
3,
to
du Bois Reymond that
2
- 3
exp ( x sin .^) dx
Art. 166).
in a power-series for
The function $(x) can obviously be expanded

value of x, although the law of the coefficients
is
any
not simple.
absolutely summable series, we may refer

to those given already in Art. 103.
Thus, for instance, take
Ex. 2; here it will be seen that
As examples
of
and
so that
dx
dx*
Similarly
the
obviously convergent, since 1
is
other
cases
are
to
proved
be
>
0.
absolutely
summable.
105.
Absolutely convergent series are absolutely summable.
For we have
so that
|
Now,
u(x)
^ v(x),
since the series
5> n
Jo
is
if
vn =
un
\.
convergent, the integral
e~*v(x)dx
converges (Art. 100), and thrn-l'mv the integral
e-*u(x)dx
{oo
is
absolutely convergent.
104, 105|
Further.
tin-
tli.-
Mill.!
^_it A+1 H-v/ A ^_.-i-
.same
tin-
'I
that
-279
this
to
argument applied
absolutely con-
is
...
series
shews
int.
\dx
Jo
That
dec absolutely convergent

absolutely
1'iit
Minmiablr,
n id.
const
This
in
is
-.1
is
îvrn hy
-I/'
K,
(in
llar.ly
lias
r-,
/-
and
\\hich
intn -\
:
i-
and
that
i-
'*
<
\
per.
"
\\ li'-ii
pp.
-~>
2&).
K r,
lid
M]
s
\\ln-i.- K
any int-u
i-
integer contain. -d
.itmle as the
.id.i
saint-
u(x)\>
L.-inu'
that
.a
]
lai-<->t
ill.-
'lie
in
proved by an example
that
and
all
utely
b.-.-n
takin.u'
\\li.-n
i-
Hanly then shews
<-<>n\
any
not correct. as
l.y
fxaiiijil.-
\\liii-h
that
in. -lit.
I!..)-.
an absolutely n.h
is,
suiiiinable.
is
It
-i.
a constant, throughout
is then
easy to prove
^'-nt.
is
It
easy to construct simple series which aiv suiimial>l<\

an example is î\-n by the series
hut not absolutely:
For here
and so
')=
x ['
(c.s./-+
|*\
i
Jo
!],
_ ^ ==
,
fl
[Art. 173 (DJ.
"i
ucc
SO
./)
But
sin
c-'-f/sinaj
-
v ,
f*cosic
(j-)<Lr=\
,.-*
U(x)\dx^\Jo
.
lit.
{
In iVLranl
this srrii-s.
\v-
m>t<-
that
-O/^
\vhirli
tends
t<
B6TO, a
'
"'
I
ges.
in
thu>. as
in
Art.
1"1,
we
B66 that
280
NON -CONVERGENT AND ASYMPTOTIC
But the convergence
of this integral
x
so that
\er
is
SERIES. [CH.
XI.
not absolute, because
if
u'(x)\^j2lx-2la?,
/oo
Thus the
integral
Similarly,
we
e~ x u'(x)
JQ
dx
is
divergent.
Vac
establish the convergence of
e~xv>(x)dx,
A.
being any
/oo
index of differentiation, and the divergence of
Thus
the
series
Â +
%+ + %+
i
+ "-
e~ x \u x (x)\dx.
summable for
all
values of
A.,
although not absolutely.
106. Multiplication of absolutely

two series
summable
series.
If the
are both absolutely summable, and
wn = u
if
then the series

is
also absolutely
summable, and
W = UV;
For then we have
I
Jo
e- x u(x)clx\
= lim
\\
A->oo J J
where the
e~
Jo
- (x+v)
u(x)v(y)dxdy,
latter double integral is
taken over the area of a
square of side X.
Now, in case u(x) and v(y) are both positive, the integral
taken over a triangle such as OA'C' lies between the integrals
taken over the squares OABC and OA'B'C'] so that, since the
latter integrals both approach the limit uv when OA tends
to oo, the integral over OA'C' will also approach the limit uv.
15nt if u(x) and v(y) are not
always positive, the difference
between
tin-
\TIN
Mi I.TIN. h
105, 106|
tin-
OA BO and
ini
n.\ /;'
I.
integral
'"
"'WaOI.
'
taken over the area
.I//'
''//.!
this
an<l
the
is
dil
the integrals
f
A
"
'(
'
'i)\dy
Jo
and
'
'|tt(j")|
tLc
which teniU
BerO, since the
to
e"V\v(
Jo
ii,
30
and
e~*
-'
<
Jo
Jo
are in.th convergept

we can writlim
when-
tin-
now
is
integral
triangle such as O.1V",

Del
intr-Tal
tli-
I'.-)-
tak-n
sil-77
"\-.-r
thai-ea of
_\.
i-
= ^,
Ix-conif-
l-iy.].
.-(
Jo
Î'-.'nt
t.
wl.^-
Jn
Owino-td the
<lxdy,
=
.,-+,/
us write
and th-n
'
>/:Jj
fact
that tin- scries
any \ah
l'>r
\\
//
./
).
-(//)
aiv
al>s,.
lately
can obtain the product
fti-)]-*(dr)
the
hy
rule
t'or
product-Seriee
resped
inultijdicat
will
The general term
and
['
tl.al
I
.11
obvi
iliat
i.
of
be
we may then
in
the
Beriefi
(Ari
uniformly
the
uid
convergent
with
integrate terin-hy-terni.
|>r
,,
282
Thus the
integral over the triangle
SERIES. [CH.
XI.
is
where
we have
-and accordingly
the equation
uv =
'0
But we have not yet completed our

the last integral
and
is
is
the
sum
(Art. 101) it does not
summable.
It
investigation, because
of the series
immediately follow that the series
we can
should be observed, however, that
infer the absolute convergence of
because
|TP()|
Hence,
if
we
Jo
u {(l
*i)} I-
reverse the argument used above,
e~%
W(g)\dg<\Jo e~
Joo
o
we
find
r\
rx
JAo
v (&)\ d>j.
\u(x)\dx.
Jo
e~y\v(y)\dy
e-*\u(x)\dx.\ e~y\v(y)\dy.
Jo
Thus, since the integrand is positive, the integral
is
convergent (see Art. 166).
To complete the demonstration of the theorem,

the two absolutely summable series
Their product
let
us multiply
given (in virtue of the foregoing) by the

absolutely convergent integral
where
is
Ml UII'LK
106|
\TI<
'I'll.'!'
)="',,+
...-,/'
"-,V
"|
ah
so
anil
--'Itrtf)
JO
ami therefore
W((
<
Jo
ix
gent
ahsnint.-iy
No*
HT()
-,4-^;-
i*()
w=
whicli
is
-r
<<>...
with the scries
t'unction assu-iatrl
tin-
i^
//-
4- /r t 4- /'._>+...
therefore suniinablr: ami
sum
its
is
r|ual to that of
0+ "'+"'!+ "',4-...
If
we multiply
similarly the ahx.lutely
we can
'out
<
(Art
//
;{
+"
+...
'
'i
Mimmable
101).
ser;
'-+'"'
establish th- oonvergenoe of
inning the
VT6
].r..cess.
can
pn\v
that
\;
an-1
Jo
|
e.n
for
By
ue
int.
any
!'<)hltely
.,f
aceuplin^ly
the
slJininahle.
conihinin- the foregoing results with the obvious
nn the addition
>lutely
the
suniinal-i
t'nllnwili.
//'
/'
''
ng
fl
'/'//
p
Ill
IIKllll'
si
/
i4
un#
6c
284
SERIES. [CH.
XI.
Examples of multiplication of absolutely summable
107.
series.
Ex.
(1-1 + 1-1 + .. .)2 = i
1.
Of course these
results agree
Ex.2. Write
+ cos 20 + cos30+...
=0 + sin + sin 20 + sin 30+
with the examples of Art. 103.
(7=l+cos
...
C7+i#=l+e + e K * + e s ''+....
'
then
Hence (0+ iS) 2 = 1 + 2e ie + 3e

2
2
or
(7 ->Sf = l + 2c
By
using the values of
+ 4e ^ +
C and S
found in Art. 103,
it is
easy
to deduce that
+ cos 0+2 cos 20 + 3 cos 30+ = + sin 0+2 sin 20 + 3 sin 30+.. = 0,
.
..
cosec 2 (|0),
agreeing with the results found in Art. 110 below.

Ex.
3.
Another exercise in multiplication
C and S
directly
is
given by squaring
this yields
= Jcosec0(sin0+ sin20+ sin 30+...)

+ JO + 2 cos + 3 cos 20 + ...),
=
S
+ sin 20+ sin 30 + .)
| cosec 0(siri
-1(1 +2 cos + 8 cos 20+...).
2
S 2 found
obviously agree with the value of C
2
above.
These
2
We can also find C + S* by multiplying (C+iS)x(C-iS);
and we can find CS directly.
108. Multiplication of non-absolutely summable series.
Hardy (p. 43 of his second paper) has given two theorems, \vhirh are the
extensions to summable series of the theorems of Mertens (Art. 35) and
The second of these extensions will be found in Art. 111.

theorem is that it is sufficient to suppose one of the series u v
then w will be summable and its sum will be equal
absolutely summable
We suppose that u is absolutely summable, while v is summabK'
to uv.
Abel (Art.
The
34).
first
but not absolutely.
On
(luubl
reference to Art. 106
it
integrals over the areas
is
plain
that the ditfeience between
OA<\ <>M>C
is
tin-
107, 108
Mi LTIPLU
\i
i"V
writ,-
ami
th. MI
/'
'
tin-
finite
;
limit
upper
c~ y
integral
<(>/}
<(>/
tt
roppoted
values
f<T all
//
is
* ami A
<.f
.n
<
vergent,
^(
further we can timl
hat
X-jrg
if
arbitrarily small.
divide tin- integral (1) int>
may
Now
\-fji to A;
exceed
in
t'oimei
tin-
Tims, sim-o
//.
>>(
t\v.
</>(./)
from
parts,
than
l-xs
is
t,
we
to
the
in
A-// and
fim
],v
timl
\D\<
,(
.'0
/"*
Now
so
<:
incijuality as A trmls t> infinity.
last
i<-
lim
we take
if
the limit
timl
\v.-
l>
A->
Hut
i-
IT trm- Miih->- lim
ami
arl'itrarily small,
Thus lim
0; and
I)
/>
p.
.")
tin- la-l
in.Mjuality
i>
intr^ral OTef
tin-
tin-
<>.(!',<'
may again
A-+-OO
!ila--l ly th- intr-ial ..VIM- tin\\,m.i\ \\ nt-|iu;.

Q Art.
'.!
II'
lim
'
.
/
|
A-*
the integral l.ein- taken over the triangle
After tin- Stage

Thus tli>- M /' \ "-,,-f
r
tlint
It
OAO.
\.i.-lly
the same lines
"', -f
one
can
)>e
shewn
that
any
s-
+ ...
is
summahh' umler
the above hyp.ththe end <f Art.
M at
that
"a
-4-
"'.,,
-i
"a
.-
r-
...
may
l.i-
H.'I
is
summal>le
an example of the possibility

'hout
t
286
SERIES. [CH.
XI.
and integration of a sum-
109. Continuity, differentiation

series.
mable
Suppose that the terms of the summable

of a variable
we may
so that
a,
a\ =
u (x
series are functions
write
û
( a \?L_.
then, in agreement with the previous use of the
word uniform,
oo
we
shall say that
Qûn (a)
Tlie series
is
uniformly summable
with respect
x
e~ u(x,
to
a in
an
interval
(/3,
y),
provided that the integral
a)dx converges uniformly in that
interval.
Thus, in particular (Art. 171, Appendix), the series is certainly

uniformly summable, if we can find a positive function M(x} >
independent of
a,
such that
\u(x,a)\<M(x\
j.90
while
e~ x M(x)dx is convergent.
Then the following theorems

(i)
(/3â^y)
If
n (a) is
o
terms
the
all
un (a)
are true
are
continuous and the series
n
uniformly summable, and ^J un (a)x /n\
is
uni-
formly convergent for any
finite value of x,
when
a lies in
00
the interval
(/3,
sum oûn (a)
y), the
is
of a in the same interval.

For then, if a is any particular value of a in the interval
r
oo
Km (?un (a) = lim

a >a
e~ x u(x,
a >a Jo
|<
a)dx= ^ e^'lim
this transformation being justified by the

of the integral (Appendix, Art. 172).
00
Next lim
[u(x, a)]
aôo
because
u n (a)
71
>
o).
a continuous function of
oo
a,
and the
series
45).
lim Qâ
JO
which sliews that the sum
any
IT,
converges uniformly (Art.
Thus
uniform convergence
2 ^() x = 2 <(o) x~ = u x
= lim
a-^-oo
is
[u(x, a)]dx,
->o
point of the interval
(/3,
is
a continuous function of u
y).
.it
di
/',/./, r
IMA n
hii ii.
109, iio|
TI\ \M>
raoB
i\
9am
tht
<* in
"
T[
(i
Jft
Jft
be cor/
it-Hi
This
f..ll..ws l,y
tiii)
//
//"
an arûinenl
similar to that of
conditions
saint
of
"/"'/'/////'///
(i).
*/////''
00
'*
This thenivm
is
also
easy deductiOD from A
aii
\~~2.
Applications of Art. 109.
110.
00
Ex.
we consider
If
1.
where x
tl.
is
& complex
number
of absolute value not
than
less
we
1.
find
that
j
f
where
me
that
may be observed
It
.
and
e.'nse.jurntly.
but positive. th-
135*
Jn
the ival pari

that |ex '| = >p
if
f
,
wlieiv
lesa
ft
\\i:
...
fi
for
inti-î-al
will
.'
than
i.
the ival part

arbitrarily small
is
conv'i-^f
unifoi
oo
with rrsprct
is
/5
t..
fuin-ti..n
r..ntiniion.s
1
Ex.
<i,
2.
as
If
nhvinus
is
in
./-,
we
.f
f|-,,m
x within
lind
P6gi
tin-
that
and 80
Thus, the integral functi..n
00
bed
,/;,
tl,.-
sum
^xn
tin-
thr valu- of
writ.-
thus
with
ti.-l
1,\-
288
because
we have
SERIES. [CH.
XI.
the algebraical identity
7^ = 7i(?i- l)...(?i-_p + l)+^l 1 ?iO- I)... (n-p + ty + ^. + Ap^n,

where A 1} A 2 ... A p . are certain coefficients which depend on
p but not on n (compare Exs. 4, 5, p. 170).
,
Hence
\up (xt)\ê*
and therefore the
(t*
+ \A
\tP-
+...
+ \A p _
1 \t),
integral
t
re~
up (xt)dt
converges uniformly
Thus we may
00
cos
if
=1
a.
differentiate the series
found in Art. 103 for
'
00
often as
<25^ sin 710, as
we
please, provided that
is
not a multiple of
Hence we
2?r.
find, for instance,
J^n* cos 7i0 = 0,
J?VS+1 sin n6 = 0.
Taking O = TT in the
we find the results:
1* _ 2* + 3* -
first
equation and J?r in the second,
= 0,
26 +1
'
- 3* +1 + 5 2s+1 -
= 0.
Again,
provided that
Thus, since
we
is
not an odd multiple of
find
TT.
_
<>>+i)
92+2
[Art. 126, Ex.
Thus
l_2 + 3-. =
..
*The
5].
in particular*
]
(Art. 103),
second and third of these may also be obtained by means of devices

See also Art. 126, Exs. 2-4.
similar to those used in Art. 103.
Di;
110]
In
lik.-
manmT
to
289
O-CQB 30+cos 50- ... =
sec
n-iilt
tli.'
where E
\\i,
sn
ih,-
COB
l.-a.U
NATION
3+5-... = i(-l)'#.,
number (Ch, V Ex, C
Buler'fl
[Ci
Hi-nee in particular
I
+ ;,-_...= _
._:;
Ex.3.
m we may tak-
As examples
the series
-n
whi<-li (as
in
are unilMi-inly suininalil- in an interval
-2)
when- <><o<
Thus \\v
ATT.
fi& s
Now
f^flq-MC*
6
2n+l
Jii
and
mi thri^ht
scries
iherel'on' tin-
2cos(2?i + l)^/(2?i + l)=
ami
<-..ii
28inc2/<
verges uniformly
J,
is
1'
l)c)
convergent;
also,
(see Ex.
4, p.
,;),
+ 1)
(2//
\V.-i.-rstraas's test.
l.y
"
--',,+r-
,,-
ami therefore
o
<-')'
./ft-.
-
1
In like
manner we
f
J
ami
pr-.ve that
dOOtdcW=2
^f *Hsin
i
hei
I *fl,.
,t
V/ _1 V82<-i
d,/rt=9
is
g
-:;io
2
2.
290)
290
Ex.
As other examples
4.
integrate the
first series in
where the resulting
XI.
of integration, we note that we may

to JTT, and so obtain
Ex. 3 from
0+icos 30+icos 50+
cos
SERIES. [CH.
= Jlog cot J0,
. . .
series converges.
(0
< <
This series
is
TT)
summed
2, and Ex. A, 43, Ch. VIII.

series
the
from
Similarly
cos0 + cos20 + cos30+... = -J,
independently in Art. 90, Ex.
+ sin 20 + sin 30+
sin
we
find,
integration from
by
sin
to
TT,
+ 1 sin 20 + J sin 30+
...
=J
the results
= JO - 0)
and _(i+cos0) + JrO-cos20)-J(l+cos30)+...
(0<0<27r)
= log sin 1 0.
These
form
series converge,
and the second one leads
+ i cos 20 + J cos 30 +
cos
...
to the simpler
= - Jlog(4 sin 2 i #).
All these series agree with the results of Art. 65.

establish the results
Similarly
we can
and
sin
0- J sin 30 + i sin 50-
sin
+ J sin 30 + isin 50+
Ex.
we
5.
T
If
^=
...
= ilog(sec + tan 0)
=i?r
(0
< <
TT).
<S cos x (a 4- nx)
get
Thus
The values
Ch. X.
(p.
of
for
cc
= 0,
259); from these

6
For other
=2
are found in Ex. C, 31 r
follows that
- -
cosh bx sinh
--26
--- 8mn
iTol
o
'cos 2a
\cosh26
1 /
u=T
cc
it
applications, leading to the values of
some
interest-
ing definite integrals, and for the investigation of conditions

under which the equation
we must
refer to the latter part of

the
references
given in Art. 96).
paper (see
is
true,
Hardy's
first
110, ill
111.
AUK i.- THEOREM,
(An. 5L)
Analogue of Abel's theorem.
<*Un
that
is
summahlf, ami consider
tin-
possihilitv
oo
of
summing
//,/
\v
n-al
i.
ami
lies
between
ami
1.
to
xt
\\V have th.-n,
li-tinit
l,y
ion.
<&u n =\Jo e-*u(xt)dx,

t
ami
(
ehaninn^
BO,
=//, sa\
tin-
iml'-]>'ml-nt
variable
from
fold
Thus,
17 = 1+0,
it'
su that
is
n-al
ami \>
ire
have
Jo
hyp'ttlM-si^. the integral
Now, by
e|
Jo
mvergent; and consequently (from Art. 171.

Aj|M-mlix) the integral
f
Jo
Y.\.
~2.
>t
e-9< l +*>u(y)dy
converges unit'nnnly with n-spect to

win-re a i^ any positive lumber.
in
6,
any interval
(0
oo
Ih-nce
^*u
o
the interval
Thus
in
t"
<>,
can le
summed uniformly with
resp.
1).
partieiilar
w-
havi- the result
')='
Further,
pi-Mvi.lr.l
ummaMe
that
the series
for every integral value of X, the series
(û n
n
t
is
also absolutely summable

for tlu'ii we have
limi
f->r all
positi\.-
\aluesof
= 0,
(XÔ).
)]
less
than
292
Thus we can
Now
SERIES. [CH.
XI.
find a constant A, such that
= P-
[*e-*^[u(xt)]dx
\e
so that
Thus, since the integrand

to
must converge
oo
(Art.
taken up
positive, the integral
is
and so the conditions of
166);
00
absolute summability are satisfied
by the
series
o
Thus we have the theorem

n
oû
o
//
is
*
:
summable, then
oûn n is
t
uniformly summable
for all values of
in the interval
and
1);
(0,
if,
further,
oo
is
summable for
all positive integral values of A, then
o
oo
(ûn n
t
is absolutely
summable for any value of
between
and
1.
An
immediate consequence from this result

theorem on multiplication of series
is
Hardy's second
// neither
uv = w
that
u nor
is still correct
is
summable
the
equation
(using the notation of Art. 106),
provided
is
absolutely
summable.
00
00
For then the three
Q7*u n t
series
n
,
CO
<^vntn
Q^wn n
t
are
absolutely summable, and therefore (Art. 106)
000
Hence, taking the limit as
(0
approaches
< <
t
1,
1).
we have
/TV
*Due
Hardy
in part to Phragmen (Complex

44 of his second paper).
(p.
Rendw,
t.
132,
1901, p.
1396)
and
to
!'<>V,
111,112]
An important
112.
Suppose that
where /()
int.-Tal
sueh
i.s
tliat
aUo absolutely
is
//
Then, appl\
i-
f(g)\dg
j'
I'm-
summable power-series.
class of
A. Art.
convergent, so that
>n\ ,-rênt.
<
have
\ve
17").
beam-.'
and
tli.-
HÎ
converges uniformly with respect to
aeri
is
'/;-
convergent
Jo
Thus
int.-ral
r.oivl's
"..//
'
I'm-
Fr<.in
imt
with
Jo
**/()<*
order of integration can

integral, provided that the real part oi
!'<>r
then the
than
a fixed nuniher / i<l);
^reati-r
is
seen to be absolutely C(n\ rr^vnt ly coniparini: it
Art.
be inverted
i-^
c-'cfl
Jo
Jo
177.
it
that
t'>ll>ws
tin*
this
in
tht-
"
Thus
IJiiri-l's
Jo
integral
reduces to
C
(
I'
J
is
It
all
-vident
values of x
thus the
that
t
V'/^
(real
part of
x^k <
.''(
the
except
last
real
integral
will
he con\ -i-iêut for
values which are greater than
int'-rai
gives a larger region nl' suiiinialility

\\V cnuld, of COUTSe, have adopted this
than IJmvl'.x integral.
"
as a definition
nl'
the
sum
1"
making another
l.y
last
,|'
\rt. 99 above); but it

applicatinn of Hardy's prineip.
would not have been evident to what extent the new definition
Could he deduce.
in
tVoin
Corel's.
.!'
applicati< ms ni' this integral
the paper ijunted above (Art. !ti). and the method has 1"
nded )>y Hardy.^ ly th.- aid of contour-integrals,
L.-
*/'/
liny has -iv.-n a nuiul.er
.oi.
;?,
ini:,.
p.
294
Ex.
We have
1.
*^L
/"V-V- dt= f f "-'[loga/ft?- d%,
n
J
Jo
+ 2* + 3* + 4P+
and
its
sum
L
2.
Again,
^-ji+
=
1
r
/
^2
expressed by the integral
is
>
Ex.
(p>0),
series
3,
summed
and accordingly the
can be
SERIES. [CH. XL
lo/-
00
e -* sin t
dt =
JQ
rff
r1
n~ l
[HADAMARD.]
sin [log ( 1 /)]
d,
-'o
so that the series

i_
I_l-l2
is
summed by means
Ex.
J.1.92'
**%'
of the integral
b>a>0, we have
a(a + l)...(q + 7i-l)
T(b)
[
6(6 + 1) ...(6 + -l)~r(a)r(6-a) J
If
3.
so that the series
can be
summed by
the integral
1
d-ft*-'- ^.
In particular,
if
[HADAMARD.]
6=1, the integral can be evaluated by writing
which gives
x is complex the last transformation requires a

beyond our range, as it depends on the theory
If
is
Ex.
where
which
of contour integration.
Consider similarly the general binomial series
4.
contained between two integers

and m + 1 then, if wÔ, we
..f Ex. 3 to the series obtained
by omitting the first
terms, and so pn>\r th.it tla- original series is summable by applying
is
can apply the nirihixl
(m + l)
little justification,
Art. 101.
112, 113|
113.
Borels
of
Application
!'"\N
power-series
in
the real part of
than
1.
to
process
general.
a
example, ennsider the series
first
1+r + .r- + .r +
;
m.
by Bond's
Jind
\\'e
^x
?
so that
<
this
less
is
!'
=\
t'~
('<(((=:-
;
'
pn, \id--d that
integral convei--.-^.
than
Jo
and
hod
-i
1.
Similarly, ly considering the 861166
we
find
tin-
Cc
HI.
j.i-n\
id.-.l
Tin-
that
h
'
'
a?
a;
<
.,
are
n-al parts of x/a, x/b, x/c
tin-
all less
examples are as
In-low: thr n-îon to be excluded is shaded and the
i.îoiis
,,r
suininaljility
in
thrv,-
t\v
in<licat- the t-ircles of coin'tT^t'iict-
of
th<-
Similarly, the -jM-cial 861168 e\amin-d in the

in the region imlicated in li-airr
power-series.
last
article are
snmmahle
I'ln
the
r.-njeeture
t!
m/mabili
,-<></
of
to
1,-ad
Tin
from
th<
origin
to
i/
il'|>"M- ;iny
/' tak.-ii
jMMut
in tln> rr_i:i"ii
-iniMil.ii 'it;.
-iivunr
It
hi.
}i
".
drawn on
tlu-i.
\m within or
le
i> tli.-j.
endooee
/'
and
>till
.,11-iti--.-.
-n
tlu-
of radius
296
Let
is
/(.r)
= 2an#n
SERIES. [CH.
it follows from Art. 82 that

any point within the second circle,
be the series; then
the mid-point of OP, and
where the mean value
is
taken round the
is
Thus
and so we get
circle
k\
XI.
if
= r.
an
where the interchange of summation and integration is permissible because

the series converges uniformly at all points on the circle since
Hence the
integral
f
\
f
u*(x, t)\e~ dt converges, provided that the real
on the circle; now

positive at all points
at P, because OP subtends an acute angle 6 at
all points
of the circle, and so (f #)/*= pe**, where p is real and positive.
Thus the series is absolutely summable at any point within the polygon
part of
(-*)/
this is satisfied
x/(
when x
is
is
specified.
By the aid of complex integrals and a slight modification of the method

of Art. Ill, it may be proved that if a power-series is absolutely summable
at P, it can have no singularity within the circle described on
as
diameter.* This is the converse of the last theorem.
OP
It will be evident that, whenever the number of singular

points is finite, the method of summation enables the value
of the power-series to be found at points belonging to certain
and we have thus
regions outside the circle of convergence
;
a process for finding an analytical continuation for the powerseries.
from Art. Ill that the power-series is uniformly

the polygon of suinmdbility.
This property completes the analogy with the circle of conIt follows
summable within any area inside
vergence^
Borel has also proved in a later paper t that if F(xt) is the
integral function associated with the given power-series, then
the straight lines (drawn as above) determine on each radius
through 0, the limit of the points x for which
Borel, Le$om p. 108.

t Borel, Math. Annaltn, Bd. 55, 1902, p. 74.
t
B0i:i:i.>
113,_114]
OWI
OF BUMMATK
OTHKi: MI:TIIOI
114.
IMTIMV
i-i.i
Borels primary definition of the sum of an oscillatory
series.
Writ.-
= "+"! +
*,
and thru consider
tin-
".,+ ..+U,,
.
6X]
-*u^=e-u'(X
<
Hence
(1)
Thus, vimv the
''//</),/./
<
inte^n-ul
is
associated with the
Jo
series
we
).
see that
the
if
<Ûn
s.-ri-s
summahle, then the
is
limit
--
)-
;
:
ami
is
sum
t" tin-
eijuul
But we have proved that
(*U
if
90
can be summed, so also can

00
(û
KM
(see Art.
wnmable,
fl
and since
):
i<><ition
rue; andccn
www nw
bath
For
existence
tin-
{and
<>f
V-V
when
th-
'"'.)]
./
/*
û
^s
'/
and &s
o
limfx-^xL
<> ,-'
when
ire see that
quati&n
limit
this a^-iin
thr
implies
e(.n\-r--.-nce
t> t-.juatin
l.-ad-
of
Th.- limit just nl.tain-'d
l>y IJoivl:*
.'iirsr
hut, as pointed
tliis
r.m
inctlioil
least
;it
!,
to thrrc
has
pr.uvss to ol.tain
l.y
]<|>|iliril
<i-i-inal d.-tinitinn
Hardy,
tin-
nuni.noal
to
integral
d.-tin:
-ii*
calctllni
the
.i
iiiiiui-ri.-al
it
is
that
this
1896,
p,
119
expression gives
possible to use th-
loddenbl
j.
BOggeeted
'
i-aliMilju.
.li-.-iinals
th.-
,,ut
11.
54
1
was
i-Ji.
298
SERIES. [CH.
XI.
(deduced by Borel from this limit) can be applied to cases in

which the limiting process gives no definite value (see for
instance the example of Art. 101).
It is almost evident from Art. 100 that the limit
the sum,
is
proof
if
easy.
<jhoose
the original series
For
if
convergent;
converges to a
but
sum
equal to
a direct
we can
s,
so that
S
Hence we
and
is
is
>
find
therefore
m-l
where
and
the upper limit of
is
we
so
But
is
is
arbitrarily small,
limit must be zero,
divergent,
lim
Ex.
ranges from
to
and so (see Note

and therefore
(6), p.
5) this
easy to modify the argument so as to prove that
It is
un
see that
lim
maximum
s as
\s n
when
-- =00.
If
we have
Thus
and
giving
The reader
will
find
eries given in Art.
it
103.
exercise to discuss similarly the other
115.
BO1
i.i:
114, 115, 116J
DBJ
IM
299
'
Le Roy's extension of Borel's

Roy has r\\ 'n an extension
definition.
>
I.--
modifying
tin-
seri.-s
in
tin-
by
integral
form
t^+0+0+... +04^+0+0+. ..+0+14+0+.,

where (p 1) zero terms anThe analytical formula i'nr
un and un+l
]!.
tin-
Mini
is
tht-n
'|
W =l
where
'I'll.-
ol.j.-et
iisin-- this .U-tinitinii is
nt'
asrs in wliich
th.-
cases are illustrated
Of course
()
Beriee
by
Ilon-I's int
oever coim-r^-:
-nul.l
such
tli.-
value
is
equal t<> that of
nsiii nlivinusly
condition of consist. -ncy (in virtu.- of Art. 100).
iimditird
tinth<-
116.
Beriee
if
is
!>/,
Nn\v.
*/.,+
assuming
bsolutely
taken
h.-r....
is
that
definition.
for
sum
tlir
of
an
oscillatory
the limit
the
series
u^-d
this
in
definition
is
convergent, we can write
T(nt + l)
('
rt^nj
Jo
rh^lI
its
independent
expression
"+// +
convergent,
and BO L-
s.-ri.-s.
Le Roy
'I'll.-
an.
id
thri-rt'..!.-
tin-
1'
that
-',':
tin-
justify tinin\ri>i.'i
n.
:,':
Jo
iivrr^nit. -
j>|.i-mlix, can be applied to

ml summation.
300
This integral
SERIES. [CH.
XI.
the same as
is
00
Jo
and so Le Roy's
definition
e~ x u(x t )dx
becomes
1 f
lim-
x )u(x)dx
ll/t
t->i
x l+a )u(x) dx.
____ Joo
If Borel's limit exists, this definition gives the
same value
for the function
x a Qxp(x
x l+a )
>
>
decreases* as x increases, if x
1 and a
0, and is equal to 1
for x = 1
and we can therefore apply Abel's test for uniform
;
convergence given in Art. 171 of the Appendix, which gives

x
r*
x l+a ).e~ x u(x)dx=
x a exp(x
i
Ji
e~ x u(x)dx,
because the latter integral converges.

Further,
lim
a^-OJO
# a exp(#
x l+a ). e~ x u(x)dx =
because the range of integration

t
Hence Le Roy's
the latter
is
e~ x u(x)dx,
JQ
is
finite.
definition coincides
with Borel's, whenever
convergent.
117. Borel's third
This method of
method.
summing
differs
from the
first
method
in
the fact that the terms are arranged in groups of k before

applying the method. It is however of less importance than
the other methods, at any rate from the arithmetical point
of view.
CO
Thus we obtain from
ûn
the integral
x
Ie~
o
Because the logarithmic derivate
is
plainly
is
a1)
l-(l+a).e + a/a:= - (x
and
when
a>0.
#>1
negative
a
\vlii--h
uk (x)dx,
AN
116, 117, 118]

\\
Of
Ml IMI
HI.
[<
here
')
-0
If
<
2u n
sun,,
its
is convcr^ nt, tliis ],r. .cuss of

grouping eann.t alter
and e.>ns,',|U<-nt ly (Art. M'l) this integral will also
verge to the sum.

Hut in -vii.-i-al tin- \aluc obtained dep.-inls on /
for example,
- 4- - -f ... -iv-s <) + <) + <)+...(=<)) if / is 2 or any
-"ii
number, lut
-f
...(
As another example, take

foregoing methods.
if
-2 + 3-4
=2 we get - 1 - 1 - 1 - 1 - ...
-'-5 + 8- ..., \vhi< h is the same as
With
118.
-H...
-2 + 3 -4 + .)-(!-! + 1-1
..
as
to
./'
is
<j)(x)
tends
integrals
''=
in
30,
has
\\lii.li
tl
divi-r^s tu
-y
- r.
the
k=3
while
~i = t-
+...) = !
which steadily
function,
pnsiti\t<
odd.
method of summation.
further extension of the

that
-s
hi.-h
\\
3(1
is
such a
way
that
all
tin-
po
1
V'
Jo
ar.-
f
Jo
it
Then
convergent
is
easily
we
if
#(*)(S;
N o
cn
s.-.-n
to be
consid.-r th-
=
'
correct
when
absolutely
nt.
K'-r
f Ui
then
because
</>(/)
and
./"
an- jm>itiv.-
Aj.jirndix, can be aj)]ili-d

and intr uqati.n. Tims
if
1"
is
\.,
1
u
\- o
\,
,i.-,-,,nliiii:ly
..f
the test
tlu-
')**'dx)
cn
I
onK-r
.f
sunnn
oonvergent within the interval
we ha\-
ag
and
justify tlu- iiivi-i>in
-^).'/
.
un
1.
-f-1),
302
Now, because <t>(x/t) steadily decreases

we have, by applying Abel's theorem for
SERIES. [CH. XL
as
definite
increases,
integrals
(Appendix, Art. 171),
provided that the latter integral converges.

Thus, provided that ûn t n converges absolutely within the
interval (
1, +1), we have
Km 2
f <j>{x)( V^
x n] dx,
c
Jo
provided that the
latter integral is convergent.

Thus, if for two different functions 0, \}s we can prove that
the corresponding integrals are convergent, we can infer that
their values are equal.
In particular, if we can shew that BoreUs
integral is convergent, we can obtain its value (when

convenient} by means -of any integral of the form
n
^~^
,0
rn
more
rlv
'
which also converges.

Ex.
Consider the series of Ex.
We know
real part of
that this series

t
is less
than
is
3,
Art. 112 (with 6 =
summable by means
so to find its value
1).
of Borel's integral,
we
if
the
could take
Then
119. Euler's
method
for
summing
oscillatory series.
Euler (Inst. Gale. Diff., Pars II. cap. I.) employed his
transformation (already given in Art. 24) for summing oscillaThis method is in many cases the most rapid
tory series.
in practice ; and
Art. 103, before
we
apply it to some of the examples of

examining its theoretical foundation.
shall
Ex.
TRANSFORMATION.
u:i;>
i:i
118,119,120)
1.
Ki"iu the coefficients
we
l.
l.
I,
get the aeries of differences
I,...
0, 0, 0,
simply
-L.
Ex.
<...
-1, -1, -1, -1,

0,
0,
the
5,
4,
3,
2,
1,
get the differences
and
ran>f<.i mati"ii
...
0,
gj
ami
."3.J
[Ex.
l-2 2 +3 2 -4 2 + 5 2 -
3.
11. l.-
...
i-J = i
Ex.
Art
4,
1-2+3-4
2.
in tin-
we
-,
[Kxs.
the
tin-
-i.-mYi-nt8 are
4,
1,
are
il
16,
9,
-'
-9,
-3,
...
0,
Eenoe the sum
Ex
0,
i-| + i=0.
is
4. As easy examples for

Kuler f>r tin.- ni">t pail
\v-
pra-ti.-.-
Ait.
2.
lio.J
give the following, taken
may
=
1-3 + 5-7 + 9-..
1-3 + 6-10+15-.. =
.
l-2 4 + 3 4 -4 + 5 -...=
4
It
L
[Ex.
O/
will be seen that in all thrse cases tin- n-sults
with
integral
This
transformation.
and we
120.
can
ivlation
ral
shall
fact
obtained by
those
suu-.-sts
ohtaincil
Art.
2,
fmnul
usino-
lM-t\v- -n
t
Ku
that
tin-
th- t\vn
methods;
investiâtr this point in the following art;
Connexion between Euler's transformation and Borel's
integral.
\^
Supposr
ival and
that
[688
>i
~^,<.t
than
1.
= ".-,
wh'Tt- a
i^
real
Tln-n. as in Art.
and
'24
I'
h-
-''
rr::
/^''M.
etc.
positive, whih\\<-
introduc.
304
From
these equations,
we
XI.
find at once
a3 = 6 -36 1 + 36 2 -63 = D3 6
etc.
we have
Thus, in Borel's integral
or, expressed in
SERIES. [CH.
terms of 6
b lt 6 2
...,
u(x)=
bjxt
We
can obtain the value of u(x) by summing this series

according to columns, provided that the series converges
absolutely.
Thus, provided that the series
converges,
we
find the equation
Hence, Borel's integral
Jo
If
we
e~ x u(x)dx=
is
e- x(l
- t)
[b
Jo
-b
(xt)+...]dx.
integrate this series term-by-term,
we
obtain Euler's
series
(l-O4
a43
equivalent to Borel's integral although of course

to consider the validity of the transformations.
still
we have
i.i
120]
\M- BOREU6
EUE8
Now.
Kult*r 8
\n ion (with
\i.
tip-
argument
changes) to prove that
ivioufl
;i
BOB
can repeat
series oonvei
il'
IN
tl
-bpt+l
is
C0nv<
ahsnlut.dy
to
MUM
til--
Kiders
//
l/
intend
BoreTfl
series is cow<
2ant*
xeries
'!
real
is
;/'
is
have obtain'l
///////
BorePs
that
ttd
.'
//////
1.
th-
summable
amd l"
is
"il*
natural
is
It
\ah;
iii.-th<><l.
to a
h.-ive
~
(
we
that
irb>t
^=r
(T=Tr^
lY M A
juirrs tin- conv.-rua-nro
diif't
Aôc.
as
fl
>f ^/-,,
ll
Now
complex
!'
hy
for
tl
coiisidrr the continuity of
Tin.- kMiiin;i !-<
lls
the same
jM.svil,!,. to apply
Irimna
Art.
the
NO leads
S.M-HI
here.
litliculty
t..
not
fact
tin-
th
it'
iKjuiiv
<1<>^
it
to
\vin_n'
Thus we should
and
t.
lut
and
^; A,,-A n + ij; no
)f
.,
integration.
if
= r + is,
= (\
= e-Mi-r) and
2An -Aw+1 = e-Mi-'0fA i-r _]
|c-A(i-0|
thus
8-|l-< -(!-/)
-
1 A,,- A M + 1
continuity
On
j-*\
<-an
\\hcn
K-a-t
A.
hand,
apply
te>t
it
if
\\
I',
should
l>e
m>t.
/)
ITti.
(/
IK-HILT
tin-
Appendix.
sei
real part of
Km- since 2 lt n p*
.I
th.it
if
-D
on
is
uile.
s.
the
lenini,.
t'nllows that
'
and consequently the sum
assume that the
of Art.
2(
It
nut wro,
thfl
p = \t\/(l
\vh.-i-e
;,
i-
.<
the other
convergent,
We
\vitli
ita
306
SERIES. [CH.
XI.
converges absolutely and uniformly with respect to x in any

finite interval.
We can now infer from our test that
sum
converges to the same
as
bn t n
2(
1)"
.v
(l
We
n+1
t)
have in fact
n\
Jo
n\
Jo
n
last expression reduces to \b n \p /(l -r).
and the
Thus the
series
o
n
convergent, because E \b n p converges.
article quoted, we can write
is
Jo
From
p<^l,
Art. 50,
say
\t\<^l(l
is
the line
thus
that
r),
eccentricity
we
I,
Consequently, in virtue of the
see that the series
Borel's
is
and
bn
n
\
sums
Euler's
if
converges
are
equal if
points* within a conic of
to say at all
is at the origin
whose focus
and whose
directrix
rl.
Euler's series will converge in an area
which
bounded by the
circle
<
the auxiliary circle of the conic; if I

1,
|,
|tf|
the area is within the circle because t =
is inside the circle
1
is
if
If
>1, the area is

=
1
1, the boundary
to the left
outside
the
circle
because
is
so.
a straight line (V = J) and the area is

of the line in the ordinary form of diagram and then
of course the conic
is
is
a parabola.
By appealing to the Theory of Functions, we can now see

that Borel' 8 integral and Euler's series must be equal of allpoints where both converge. But there is no obvious means
of determining the region of convergence of one from that
of the other, as will appear from the two simple examples
given below.
*
In case/>l, the conic is a hyperbola, and uc must take only points within
one branch; that is, the branch for which r< 1, and this is tin- /iff-lnim/ brunch
to the right of the origin).
1
in the ordinary way of drawing tin- diagram (with
/
KM
120
1
Ex.
1.
Tfckeo,
this gives /fo N
-a N -a H + 1
M/ =J, ZPoo^t,
I.
etc.,
dor's series is
r
_ oa if
|*|<2|l-f;
tli.it
In
wh.-n
tin- int.-^ral
diagram,
tin-
tin-
the region outside the circle
Ml.
linr
in-tli.(l
îvcii
in
ai--a
Tli.al..iv.-
i-
tin-
<>f
tli-
n-.-il
convergence of
"t
if_ri"ii
rdej and that

the
in
tw.i p..int-
Borel's integral gives
lint
ami
i>,
tin-
I'.nn-l's
\\liidi
ar-a
integral
Killer's series
is
the space
they air pruvt-d

tin-
t-
l-ft
-f
tli;m 2.
I.--.*
tin-
i.
].
t<
is
\\.-
the left of
t-.jnal
}>y
the
hyj,.-;
-r).
(Tea in
the
l'-ft
the
in>
if
which the twn
All and
ly
..nt-idr
o|ual
In
is
the area which
this case
one or
'
1"-
Tlie "i -i-inal
i<
applied
thin the
=
_'.
'2.
\\hi.-li
drawn.
Ex.
2.
Consider
n.-\t
tli-
Th.-n
case
hl
a*-L>X
:.
Thus
I
**
_L.
\-r(\
which
.-..
diameter
Hut
and
si.
nverês
is
tin-
if
,'
liir
i-
in
the aiea in-ide the circle
the points
Ak ,
-1.
ihat
whose
intcuM-a;
the integral oonverges wli
>
than
J.
308
SERIES. [CH.
XI.
In the diagram, the interior of the circle represents the area of conand Borel's integral converges in the space
vergence for Euler's series
to the left of AB.
The two expressions have been proved to be equal in
;
FIG. 36.
the area within the ellipse
(\t\
<(! -r))
but they are actually equal
at all points within the circle.
The
original series converges within the circle
Ex.
3.
The reader may examine
where k
|=J, which
is
not drawn.
similarly any example such as
a positive constant.
if a n = I or n, Euler's series terminates for any value of
Borel's integral converges only if the real part of t is less than 1.
is
In particular,
t
Numerical examples of Euler's transformation.
121.
transformation lends
Euler's
itself
very
numerical evaluation of non-convergent

Ex.
As an example
1.
-(* +
the
series.
us take the series
let
+ ^ 3 +...)
naturally to
[which
is
equal to log(l-),
if
|*|<1]
and write t= -2.
We obtain
2 2 -f J 2 3 -
24 +
. . .
and we
thus we sum
already made in Art. 24
which give - 19*314286, to six decimals.
;
the
shall utilise the calculations
first
eight terms separately,
The remainder can be put
in
the form
where 6 6 n b 2l ... are the differences given in Ex. 2, Art. 24.

But we can obtain a general formula for these differences, and so establish
,
the convergence of the transformed series.
For
D an =Dan -Da n
1.2
and so on
leading to
120, 121)
<
'I.
Mi:i:i<
tin.
-aily
series certain
With
tin-
\l.
I.
Mince |tf|/|l
ditl.-i'-ne.-s
than
less
ft
expct
1\
LEB
\ \\IIM.I
\\e
and so the tranBformed
1,
= |.
f.iiin.!
the values
liiul
7407
898
81
II
\\e
It
of the last
.-'
foinmla given
tli-
;i|)|ly
between J4 and
t<
-i
r>
-t
4-.
Now
^'(-1 !!); .7)
.un.-.l,
..
L129
I'.Kil
:i
lies
we can take
= -119607.
should be
an<l BO tin- Mliefl
i;n-llL'!>,
that the remainder

that
1-0986,
whii-h agrees with log 3 to the last figure.
Ex.
.*
2.
As a second example,
1ft
US
au'ain
Tin-
-HIM
write
nf
lirst
if
IKII
tin.-.-
^-iiiis
i>
1*5
and we
shall
apply the
following it-mis.
tin-
liave
.3...(2n-3) 1.3...(2-1)_
~2. 4. ..(2 + i>)~
2.4...2n
and
numerical
-2.
tin-
transformation to
\\ '.-
Borel's
\\itli
comp;iri-"ii
nrl'l-o-^
'-i'-o'
ami
for
priM'frdini: thus
we
3)
2.4.
find
1
2.4.6...(2wH-2jD)
= 3,
Thus, jnitting
.10...(2;^
:.iws that
/
.
decreases
asp
increase^ so that Kuler's
and fonsdjiu-nt ly the integral of \\\- '- niftli

mention because Boivl himself does not
t
\Ve have,
in pi-oviiiL: this dir.-.-tly
the convergence
An.
i),
2,
2,
ivn;.
p,
ll'i.
erges;
\M-a-nt
-\
s.-.-m
to
this
have succeeded
d ready established
310
In our
and
case,
SERIES. [CH.
XI.
we have
so on.
Then the series is l'5 + 2 3 (c + C! + c2 + ...).

The numerical values given by these formulae have been checked by
we have then
direct calculation of 3 a 4
6 ... and their differences
.
= '062500
c,
c2 =
15625
c3 =
5208
2025
C4 =
c5 =
868
c6 =
398
c 7 = -000192
96
c8 =
50
c9 =
26
c 10 =
14
cu =
-000378
086624
378
087002
rough estimate
of the
remainder gives
.
that
c n , so
we get
= '087021,
the approximation being probably in excess of the true value.

Thus we find for the sum of the original series
+ '232056 = 1-732056.
^3 = 1-7320508...,
1-5
Now
it is closer than Borel's,

so that our approximation is a very good one
although he works to 7 figures and uses 34 terms of the series.
;
Ex.
3.
Euler calculates in this way the value of the series

3 + log 10 4 - ...
210
Iog
Iog 10
He
starting to take differences at Iog 10 10.
obtains 0'0980601.
(Compare
Ex. 10, p. 351.)

Euler also attempts to evaluate 1 - 2 + 3 - 4 +
by this method,
and he obtains '4008..., but although the first and second figures agree
with those found in Arts. 98 and 132 (1), yet it does not appear that
his method rests on a satisfactory basis here.
!
122. Cesaro's
method
of summation.
has been proved (Art. 34) that when two convergent series
are multiplied together, the product-series is at most simphi
It
indeterminate.
That
the limit
(1)
Km
is, if
denotes the
sum
to
(?i
+ l) terms,
121,122]
and
is
Unit.-.
is
the theorem
in
have al-o
\\'<-
thU nn-an-value
that
^.,-11
l-'rohenius
.!'
mi
\.\\.
lim (i' x*).

lt
Cesi\ro has proposed to adopt the limit

d'-tinition nf the sum of a non-conv
generally, he defines
tin-
exists, as the
-ries
or,
more
^\\m as
|P
Hi"
{.S.
W },
'M.
n->-oo
wh<
8
an.l
<
.!
course the limit (1)
>i
putting
liy
With
tln-s.-
drtinitions,
NMW
(1
and hy
Thus
-x)- (r+l} = (l+x + a*+...)il-x)
juatin^ Cddlic:
'
tlu-
^,-e
drnoniinator
is
the ininicrat.r of N,
=
>'j
">',,
we
see that
Furtlu
they are
and
all
equal to the
,,'
>'.,
and
to
e<jual
BO,
that
sum
in
of the coeffic
particular,
if
'
/(
Sn /A n
(r)
(r}
.
'"'
we
that*
:=(\-X)-
1^
an.l
case of (2) which
special
cvid.-nt
is
it
V,
in
tin-
is
= !.
= (l+-'.-+
-x)- (r+11 <:
h;
c<.nstM|U.-ntlv
J" +l
.+(H
(l-,)-'=l + r, +
iiiiiltiplii-il
l.\
^"
:
tion.
the
coeffit
is
312
Thus we can write
From
SERIES. [CH.
XI.
also
the last expression
it is evident that the effect of

r
is
to
more
increasing
give
weight to the comparatively early
terms of the series 2uw it is therefore obvious that by increasing
r we may hope to counteract the oscillatory character of the
;
2un and
series
so to replace the series
Now,
similarly,
n
equating coefficients of x
By
if
limit.
2Sn(r+1}xn = (l+x+x
we have
and
by a convergent
since
we have
Consequently, from Stolz's theorem (Appendix, Art. 152, II.),

lim [Sn(r) /A n(r)] exists, so also does lim [Sn(r+l]/A n(r+1) ]
and
;
n >oo
jj,_^.oo
the second limit

exists
of
for
any
is equal to the first.

Thus if the limit (2)
value of r, it exists also for any higher value
r.
If
Cesaro
is
calls
the least value of r for which the limit (2) exists,

the series "Zu n k-ply indeterminate.
Extension of Frobenius's theorem.

We have tacitly assumed that the power-series used in the
last article would converge
This is, however,
absolutely.
(r)
tends to a finite
capable of easy proof whenever 8n /A n
limit I', for then we can fix an upper limit C to the absolute
value of this quotient. Thus 2$n(r) #n will converge absolutely
when 2M n(r) #n = (l x)~ (r+1) does so; that is to say, for values
of x such that \x
1.
Since 2u n x n and 2sn #n can be derived
from 2Sn(r) xn by multiplying this series by (1 x) r+l and
r
n
n
x) respectively, we infer that I,u n x and 2sn x are also
(1
1.
convergent for x
123.
(r]
<
<
122, 123, 124
in,
Ml.
1
ly Art.
-"'I
25;
=/
=shm
11111
-.
=
so that
818
Al.
B66 that
\\-
AN
'
pl
liia
\\hirh
is
tin
i.-sult
'tension of Frobe //' //>'* //,//,/,

.r thi> theorem are air<>r.l,.oo
/.
>aiii.-,
and
|ti'<.
tli,-
that
|,i-..vr,l
7'
lin:
x
stains
It
:.
1;
in. -ans
Kn'])p* has
!"
II.'ld.T has
"-'4-ri '^-f...
found ly
.tlmd and ly
<
this lias lu-.-n j.n.vi-d uj.
t..
vrd that \vh-:u?\vr H"ld<-r's
Ht.lder's process must

l-'or hiur h-i vallMI
J.
m hod
t
gives a limit, Cesaro's
will jîve the saint- limit.

It
i>
usually l.cttn- to apply Ceskro's
a given series on
124.
!'
it>
-_
it
,
Thus
\^
/'-ply
can
\v-
iiil<'trniiinat'
EolloWS t'n.m thr
method
.l.'tinition
lias
tlic
/,
ii,
-^.i
Benoe, remembering
tli.-
i.l.-ntiti--
of Art.
identity
,1-0
that
=0.
lim
(1)
sum
that
-h
i,
'
definition.
ami
\\hri
(2)
H.'.ld-r'>
writ,-
SflJ
tin-
tlian
ri-at-r >implirity.
Inferences from Cesaro
^LH
It'
an-oiiiit
'(i-x
rj-j.
we have
n->\v
NON-CONVEKGENT AND ASYMPTOTIC
314
SERIES. [CH.
XI.
where the
coefficients in p(x) are subject to condition (1) given

And, conversely, if (2) holds, the series
n is at 'most
Su
&bove.
sum
r-ply indeterminate and has the

It
is
to be noticed that (2) does not allow us to write
where
1)
lim{o n /-d^~
<r n =
For we have
and
so
I.
-p n _i and A
/_!
if
|
we can
/o n
< eljjtj
only deduce that
|er n
or
cr n
<e{
*~*
/A
= 0.
= A * - A^,
*~ l}
and
< *A ^\
(
/>
A^ + A^}
)<{!+ (2n/r)
}.
Thus, owing to the presence of n in the factor multiplying

r ~ l)
be inferred that
limJ
IA^
\=Q
if
we know nothing more than
is
simple example
given by taking
Then
<r
(l-x)-
(3)
(x)
satisfied.
and +2.
we have an
if
(r+l)
is
identity
2unx n = l
where P(x) is a polynomial not divisible by 1 x, the

2un is at most r-ply indeterminate, and has the sum I.
For
if
there are
p terms
in
values of the coefficients of
P(x)(\-x)-
and
is less
cannot
r=2 and
=I-
between
so that (Tn/A^ oscillates
In particular,
that the condition (1)
it
e,
is the greatest of the absolute

P(x\ and if
P(x\ it is clear that the coefficient of .v" in
than
the quotient of this
series
MpA
by A n
(r
~ l)
is
Mpr/(r + n\
which tends
to zero as
n tends
to GO
In practice the most common form of identity

- x)- (r + ûnxn = 1(1- x)~ (r+l} + P^x) (1 (1
(4)
l
is
r
)-
from which we can deduce the same results as from
theorem on the multiplication of

The following lemma will be needed:
125. Cesaro's
IfZu n
2t> n
are
wo
upper
limit, while v n /A n
(u v n + u.Vn^ +
tends to zero.
may
It
be zero here.
may
(3).
series.
such that \u n \/A ? has a finite

tends to zero, then the quotient
(
series
(
+ unv )/A ; +8+1}

(
be noticed that either (or both) of
?,
!.l
124, 125|
ll'l.h
[(
we can
\
Um
>m.
l^l-
|2*,.,*|<r
;.:,*"
|2*l<tt
ami
Now
ami
10
.ay-
A',,
/;
is
equal bo
vK^l'
and, since
can
-*^
tak
1>
proceed
HoW
-",
v \/A (r+*+"~
2u
lira)
,
a> small a> \\- please,
--i
Suppose
-..~
==0.
are /'-ply and *-ply

V BO that
I'.
^''
aii'l
-^- /)
.1
= ^/r,,.'-
p|
,1
' +1)
HIM
i:
iliate
we can write
o-(
that
.cli
f=0,
-I
[ft
Ti,
follows that
to the pFOOf of C'
that 1"..
liin
it
" +1>
respectively, and that their MUMS are
wliei-.-
= (l
'x(l
i> th,- ,.,,,-
Consequent lv,
\\'e
H)
rl
Hence
[o-,,/^!^}
>x(2
wl"
,.,-}-...e
To each
-.c)-
the
of
n alo\./
r<
i',
BO
.r
rn
(l-)
where R(x)
i->
aini
Viilu.-s
II
ami
Then
find
(r+
'
V-'')+ I'd-.'')
terms
instance,
in
/,'</>
in
/"(
\\
can
,-
if
/'(1-r)-"
we have
Similarly for the
SI
A ?'
(
lin. ;A',,
two
..th,-r
+l}
}=().
term-
Ri
we have
lin. ;/,'
in
;,
J
.1
;
=0.
apply the
lemma
316
Thus we
from
see
(1)
and
sum
its
XI.
(2) of the last article that the
2wn is
UV*
degree of indeterminacy of
and that
SERIES. [CH.
not higher than (r+s +
1),
equal to
is
The product of a series (whose

degree of indeterminacy
by a series (whose degree is s)
can be formed as if the series were convergent. The degree of
indeterminacy of the resulting series cannot exceed (r
This leads to the theorem
is r)
Examples of Cesaro's method.

is
1-1 + 1-1 + 1-1 +
126.
Ex. 1. The series

has the sum \.
Ex.
2.
has the
if
we
The
series
sum
..
1-2 + 3-4 + 5-6 +
For here we
find
=1
..
>
is
simply indeterminate and
doubly indeterminate and
i=-l> S 2 = 2 *3= 2, ...

we have to evaluate
>
and
write r = 2 in the formula of Art. 122,
(n+ 1)50 + 715! = 1, (i-l)*2 + (ii-2)*3 = 2,

=
Thus, when n is even ( 2m), we have
Now
etc.
+ ns + ...+= 1 + 2 + 3 + ... +m + (in + l) = $(m + I)(m + 2).

When n is odd ( = 2m + l), we have again
(n + I)8 + ns + ...+*=! + 2 + 3 +
(n + I)s
..
Thus the
fraction has the limiting value
It will be observed that
- 1 + 1 - 1 + ...) 2
and, according to Cesaro's theorem of the last article, (1
should be at most trebly indeterminate ; and this of course is verified here.
Ex.
3.
The
series 1
Hence
(1
which obviously
Ex.
4.
2 2 + 32
-4 2 +...
is trebly
indeterminate, for
we have
-2 2^ + 3% 2 -4 2.r +. ..) = (! -#V,

}
x}~*(\
satisfies
the condition (3) of Art. 124 and gives

l-2 2 + 3 2 -4 2 + ...=0.
In like manner, we find
Now flz^L
has the value
-i
for
# = 1,
so that
we have
_
:
where P(x)
is
a polynomial of degree
3.
*
The reader will probably find it instructive to use this method to establish
the theorem for r = 0, = 0, already established in Art. 34. Cesaro's treatment of
the general case is on the same lines as the proof given in Art. 34.
EXAMPLES OF
125, 126, 127|
we
Efi
5.
.d<
being q
diili'-nlt
t*
minate.
MM- thai
t<>
the series
il
1-ir
i'ly
BOD,
... + i)=/'(.r)(l--*V,
r...= -J,
ries
Ml
find
(1-
Ex.
\l:-
*-...
indeterminate, because
:;
we
<"+[(l_,
-..)-']=
find
WOO--'
a polynomial of degree (r-l); and
is
-ftGI
if
-0,
Bei
Ex.
of th- Beriei
lie >MIII
The reader
6.
(1-1
i*
//(A/.
>'
have
will
F.r in-tan.
inultij.licati'.n.
i-
-ARO.]
little
we
dihVulty
vnifyin 4 K
in
lind
+...)(! -i + 3 -4 + ...)= 1-3 + 6 -10+...

1-3 + 6-10 + ..
+ ...--i(l- 3 + 0-10+.. .)-(!- 2 + 3 -4 + ...),
+1-1
or
;-'
and so
Ex.
|=0.
The
7.
follou
!+
''
sin
cos ^ + cos
-.-
3^ + cos
~>0
n-ûlts agree with th.s- f..und liy
evaluated
127.
+ ...=J,
ôot^,
+ sin 2^+sin3^+...
(y
d'.\
l.y
1.
-ml
M-I
ii-in-
tin-
ni'-an-
. . .
= 0,
ni.-th..d
I'-uirl's
valu- jn
>'-o.>).
'
(Ait. 103).
tputcula
Limitations of Cesaro's method.
Tlu- iiiran-vulu.'
6
which can
l.r
this,
process
let
Miiiiinc.1
,iinc
in
^^
so that
ratln-r
li;i>
i,
this
way.
a-
narrow limits of
\\".-
(i
01
type of
tli.
ha
sc;
that
,n
which givee
3
--+l>N:,"
+ "'t
-'terms.
318
Now
value
approaches a definite limit /, we can find a

such that when i/>> r, all the terms
/A ? lie
if
S^V-d-iT
between
e)
(l
and
(l
+ e).
Thus
The
SERIES. [CH. XL
coefficient of
in this expression
is
easily seen to be zero
in virtue of the fact that
so that
Thus
\
That
and
< e.2
un
is,
r+1
(n >p),
limuA =
(
so
But
().
^;
p
Iim4 n
so that
il-D
we have
Thus,
/7i
=]/r!.
the result
which gives a necessary condition that

r-j9^2/
the series
2^n
??ia2/
te
indeterminate.
There are many series surnmable by the former methods

which do not satisfy this condition in fact the simplest of all,
;
does not satisfy
it,
and
is
therefore not
summable
in Cesaro's
sense.
128. Borel's original
We may
s
s v 82
...
method contrasted with Cesaro's.

mean amongst
regard the most general form of

as given
by
wh<
never
II.
we may
-I-.-
'-itln-r
bend to Borne definite

alternative
may
tin-
function
(iii
BUM \M
HO|;I:I.>
127, 128, 129]
<h
\\\
<>\
limit.
nil.
h<-
sei
every
*w
<
r all,
\ariall-
are
which
./
or
made to
!'..
Dili
and
ad
Cesaro's thM ami second mean
dou Mr indeterminacy come ondei
i
xn ^
or
= ./+ -
#,
wlii-iv
0,
m;i|.-
al't.-r\\ai-U
is
t.
t.-n.l
BU
by taki
O.
ii
1'.
Ami
i)
ii'
of
it'
i<
if
/'
>
>
mvMi^h
<thcr im-ans are given 1'V -imilar, lut

In these rases thr osoilla
t'oriiiulju-.
valu-o.
tin-
'
M-i|ll.-nre
>'
>
rharartrr of
rnMVerteil
into
th.
he t'Tins with hi-'h
that
little
,y
th--
wei
-'n-.l
which an- the
inilirrs.
ms that
are of importancr in studying tin- Beriea

To meet this <litlicu!ty. IJorel chooses the factors
increase
ami
,-n
firsi
ai'trrwanls
maximum
maximnm (whose
to
steadily
recedefl as
./
pi
Thr mo-t natural type ..f Function

which inciva.se> so Ion--
deCP
This
is
-:+..
r/
tinally
n+->
the definition
;.
M 4.
Connexion between Borels sum and Cesaros mean
/.
t!
/'<''. and
2TT^r
for a given series 1

\Ve have pro\rd
Art.
limit
-i'yin^
givefl
129.
the
terms
ry rapidly.
ihr m.-an
afl
.'-
which
<1'
tin-
determii:
in
afl
so as to
\\itl:
|K.-itin
which ensure^ that

part
/(
ion \ari--
The
increases,
with hi^h indice^ play an important

mean.
ilitions
-sit
'
L*:J)
tliat
i:
thru
liiii-1
-/.
320
Further,
we have proved
SERIES. [CH.
XI.
(Art. Ill) that

*
lim (&u,nt*) =
t^>i
e~ x u(x)
Jo
doc,
provided that the latter converges. Thus, when Cesaro's mean

exists, it gives the value of BoreUs integral, provided that the
latter is convergent; the integral may, however, oscillate
between extreme limits which include Cesaro's mean.* Obviously

the same result is true for any mode of summation given
by the
integral of Art. 118,
Hardy has proved in his second paper (see also the small
type below) that if r=l, and if a further condition is satisfied,
the
of
convergence
Borel's
from
follow
will
integral
the
In general, with higher values

Cesaro's mean.
seems likely that the same condition will suffice to
deduce the convergence of the integral from the existence
of the mean.
However, the algebraical difficulties involved
existence
for
of
it
r,
seem at present too formidable to make
worth while to
it
write out a rigorous proof, except for the case r =
We
l.
have seen (Art. 114) that the existence of the limit
implies the existence of Borel's integral, and that the two expressions are
equal and we shall prove now that if
;
lim
n
(*
then
. . .
(^sn -\\
nl/J
Borel's integral
Now
condition.
must
suppose that
//,
n ranges from
It
It
Art.
is
to
127).
the
l,
then obvious from Art. 114 that
is
oo
= (*0 + *1 +
+ *)/( W + ] )
while //,,
mean may
/<,
we
of course understood that Cesaio
in general,
h are the upper and lower limits of
then, from Art. 153 of the Appendix,
Z,
also be equal to L
"n
series
+ s n )l(n + 1 ) =
\\n\e-*
L
V
ôo
subject to Hardy's
as
+s +
*-oo
>
oscillate
are those of
<r
<rm+1
cr m+1
...
see that
mean
exists.
If
we
are dealing with
though the integral converges
(see
BOBEU8 BOM AND
129)
'
Letween
//
,<//
in(1
whenmi.
th.-
limit
and
./'/:
rr
tii^
tin-
i-
<,
and
'
Lrral
LN,
//
+ (//.-*....
)//!..
maximum
.n
f.-im.
making n tend to
t
t
oo,
be equal
+ *, + ... +
!%,
to
baa
and so we get
',
(/.
and
No\\
>*.-.-.[(//-//
M.inlinx t"
any
valiu- of
*>/-..
wh.-r- <,.,->0 as
'-re.,
a constant
i>
.1
>urh that
//-//
find
\\c
k
.
///.
tlu-iv
/,,-/,--.!,
tlm>
].
we ran write
//
And
increases.
+ (//,-/
^+'-'D'(7,;^
we use
if
Stirlin.:
formula
Ap^-ndix. Ait. 179),
'
we
see that
''
wln-n-
/'
tt
-(/'-
as
-nds to
Now
l'Mr-'
/ Klo g ^
<
\vhi-i-
1.
'
log [ e
A"
lnit
d'-in-t.-
ahvay>
Similarly,
"(7^):]-
liuiiil'.-i
<-.,ntain-d
wo can
n..;
1..
ju-ovr that
if
rily tin- saint- in all ini'iua!

itain ti\-<l limits, such as "0001 and 10000.
a> the intrîal part
^-W<
ll>
i-
th- integral
j.ait
!'
n
lilll <e
!"
^p,
322
Hardy remarks
that the condition lim
for the truth of the theorem
convergent series (which has

virtue of Art. 100).
Let
[en. XI.
by no means necessary
by the following
therefore summable in
is
in fact, it is not satisfied
sum
the
wN/m) =
SERIES,
*,,
and
is
= Ji-,
-a
and then
)/(
n+ l)
Thus by Art
11,
so that
we
see that
lim
nx/w)
= oo
although lime n = 0.
But Hardy shews by another example
lini(e,iv/tt)
condition
is
is
broken, the series 2w,
(I.e.
may
p.
41) that
if
not be summable
the condition
;
so that the
not merely a consequence of the special presentation of the
argument.
ASYMPTOTIC
SERIES.
use of asymptotic series.

most instructive examples of the application of
non-convergent series was given by Euler in using his formula
130. Euler's
One
of the
summation
of
(Art.
for the calculation
95)
of
certain
finite
sums.*
Thus, taking f(x) = !/#, a =

1
+v+ +
Now
because
..
n, Euler finds
l,
+ i=log7i + -^+ -j^

i-^L+.-.+const.
ATI
2
4??
o??
/
^ti
7?
this series, if continued to infinity, doe^ not converge,
we have
(Art. 92)
2r(2r
1)
^n?''
~'
Z
but,
if
r>3,
2_<llBr
Llliit
(see Art. 7),
B,^
and 2
150-
hence the terms in the series steadily increase in numerical

value after a certain value of r (depending on n). It does not
appear whether Euler realised that the series could never converge
*
but he was certainly aware of the fact that
Innt. Cole. Difl.,
17V> (Pars Posterior), cap.
vi.
it
does not
r:i
129, 130|
6
ralnilatr
\\
liidi
h.-
for
He
1.
/<
L8YMPTO]
i-
the
rinploy-<i
~1<
series
to
tli.
Hoarded m
th- get
V
The reason why
hat
ih
ran
In-
used, although
MaAntd
///.
is
6ft
tip-
U''HTal
Tin- trutli
thiM.r.-iu
fact
i-nalil-N
are correct, after
Ex.
1.
making
:
UK-HI
ti
h,-l<,\v
(see
Art.
Kul<-!-'>
-i-
f.-w
.17.
/'
'
of Kul-r'>
uuimpoi-taut changes.
...
1M1)
in
once from
>
all
con-
/'//
t>
follows at
gn and lias tinus to see at on<-'- that
thai
This
<>f
|ir<vi-|
thorn
u.,t
324
Ex.
Taking f(x) =1/^
4.
2
,
XI.
prove similarly that
SERIES. [CH.
2+... to
00=
-0951663357,
^- = 1-6449340668.
6
and deduce that

Ex.
Shew
5.
similarly that
l+^s
+ Js + p+... = 1-2020569032.
Euler obtained in this manner the numerical values of 2l/wr from r = 2 to

16, each calculated to 18 decimals (I.e. p. 456); Stieltjes has carried on the
from r 2 to 70 (Acta Math., Bd. 10, p. 299).

to 10 decimals (for r = 2 to 9) are given in Chrystal's Algebra,
vol. 2, p. 367.
calculations to 32 decimals
The values
Ex.
If /(.r)=l/(
6.
+ .r 2
),
prove that
sin 2(9
2U
P + ri2 /
Z(e
^-l)
2
+ -r
4
sin 4 0sin4#
:
^.
where tan0=/w; the constant is determined by allowing n

and using the series found at the end of Art. 92.
Ex.
In particular, by writing
7.
, =4
By
.(-,l T+
=n
_L
(in
Ex.
^
=
writing n 5, Euler calculates the value of
Ex.
8.
which
is
131.
If /(#)
= log x, we
found differently
The remainder
6),
we
find
4-
'
TT
to tend to GO
to 15 decimals.
obtain Stirling's series
in Art.
132 below.
in Euler's formula.
In virtue of the results of Art. 95 (small type), we can write

(1)
<
where
''
>
-*
i:\A\l!
130.131,132]
/,
in virtu.-
tin-
..!'
,',.
re
<
"/<
of comst
awl
-urne
'/..
////
>/
i;i/n,
.///
!'
IM
]ushr.l
cannot:
ilfci-.-.-îiiL:
in
good an approximation as
132.
1
conv.-r-vnt
of
Further examples of asymptotic
same
prop-
seriea
can be
while <l>
usuallx
necessary for
i-
the
luivt- altt-rnate
appnxiin.-itioii,
(1)
s.-rirs
tin'
pranio-
trnn> which
tin-
however,
an arbitrary degree
Init
/rm, a-n^ /ta

l,a^ th,- same
//*./:/
ol)tain.-il
Theoretically,
in th- U
'>"th
Thus if we 088"
th.-ii
(/,.>'
thr
and
..-mid in that artid.-:
pol\ nomials
he
/,'
as
[iiitc
ordinary cak-ulati
series.
~<
Th,- intr-ral*
<U
dfllot.Ml
oftrll
In
many
ini.'-ral
for
/i
it
///
li(
important to calculatr tin- value of

vahn-s of x.
Nan \vritet
is
<if
<
'H/
thr M'lllhol
l.y
proUK-ms.
(1
-'
<if
<
57-J,
/-J,,
dt
\vh.-i-,.
Baler's constani
i-
siiil.ol
i.
-nt
"
on
tin- priii.
"
li
-i
(see
pal value of tlu- integral is to
i'
unt!ii'_'
Appendix. Art.
//
',
;/
<
',
.in.l
then
tin-
be taken (see Art
meaning
li(y)=|
of
we
inus;
ti.i
'logy
du/logu.
utegrals are convergent

\vhicli
178).
except
of
326
SERIES. [CH.
XI.
But
this expansion, although convergent for all values of x,

unsuitable for calculation when \x\ is large, just as the
exponential series is not convenient for calculating high powers
is
of
To meet
e.
If
is
positive,
c x e -t
Jx
this difficulty
dt = e-
we
write
l+U
we
proceed as follows:
x(l-\-u), and then we have
du
Rn |< n
where
|
x~ (n+l)
This result can also be found by integration by parts.

When x is large, the terms of this series at first decrease
very rapidly. Thus, up to a certain degree of accuracy, this

series is very convenient for numerical work when x is large;
but we cannot get beyond a certain approximation, because the
terms finally increase beyond all limits.
For example, with # = 10, the estimated limits for jR9 R 10 are equal and
are less than any other remainder. And the ratio of their common value
,
term in the
accuracy from the first
series
# = 20, the
the
to the first
ratio of
RIQ to
series is
about
To get
2500.
we should need 35
first
term
is less
than
this degree of
Again, with
but 80 terms of
terms.
:
K> G
the ascending series do not suffice to obtain this degree of approximation.
When
then
we
is
negative,
x^
we
and
write
t
u) = (u
= x(l
1);
find
/-
1-^ du
-
Q
where
Art. 164).
where
')/
n p~ u
n4.<u,+v?+...+u*- )e-**dM+P\
Jo
-du
~\
\,
denotes the principal value of the integral (Appendix,
Thus
R* = P
du
AST MI-HU
132|
i<
l.-n-thy to be
approximation
taking
-y
of
tli.-n
inteie-4
AiP'tlHT
\\rite
ire
<.f
j;
tli.-
r.juat*-
\a!u.
tli.-
giv.-s
utilised
"
tin-
Miinniat imi
"
by
of
>eii.-s
MI,
pp.
we have
(1.6.
p.
o-;j)65996 as
round
value of
tlie
"sum"
U-2!+8!-4! + ...-'4
n->;ik fund fn-in Killer's coiitiii'
witli tin.'1
method
the
('
,n.....
brackets, which yields
ill
to Lacroix
..)-,
i:
agree in.i;
2'7r/)*.
--
-C-M
Lacroix
too
km-un
a-
('
OPd
tin-
two expansions were
nnl in
#=1 and
is
get the best

part of
apji!:
taking thr value

If
Kul.-r's
which
we
also
tin-
Dotetha!
it.
se
t. liiul
to
.|iial
valur
\\ithoiit
:i't
li--us-i>n.
;ise
tin-
tliat
aii-l
an elaborate
i.y
'-
tin-
.u'ivi-s
aiii.tlp-i-
ajipi "\iinate
.if
cal'-ulatinii
quadrature
i/y
'
'f
'.
tliis
98.
><!
..-..illati.i-y
to evaluate tip-
integral
rf^d
...),
'
\vlii
\vitli"
l.ut
r.
and he
appi-.'ximate
<|ii:,
tlie
also suggests the applica
integral
dv
1
-log.
whirh
is
f..und
'
ly writing
no nun.
'/>.
two
.-il-r tin-
/-I
rr
\vhich anin
tin-
iii.-t
x
witli in tin-
tlirory of
liysi-al
Optii-s,
\\;r-
ISM;,
illy
propel ties of the
with
lh-ory of
l.M-p-wat-r
*A*x
f* 8"3
)
this prol.l.
the
i.
p.
-.1
-Jul.
have oc
Hit
s
paper).
and also
328
We
U+iV= f
have
SERIES.
XL
[CH.
^j-dt,
Jx \/t
so that
if t
= x(l + u) we
du =
by applying the process

thus,
have
du
of integration
by
parts.
Continuing
we have
~~'
1
1.3.5
l.S.i
'
1.3.5.
ixu
7 f" e
du
'
and the process can be continued as far as we please. A

moment's consideration shews that the remainder integral at
any stage is less in absolute value than the last term of the
series and thus for any value of x we can determine the stage
at which it is best to stop in numerical work.
;
We
are thus led to the asymptotic equation (see Art. 133)
say.
*>-^(X-Y),
The
series for
be
iY
summed
in Borel's
X - i Y= -%my,
and
y_
"
U=
e~
we
way,
"f
v
dv
observing that
00
j7T
V
applying Art. 118,
so,
"summed" by
can be
1.3. 5. ..(271-1)L
=
Thus
and
e~
v n -*dv,
Jo
see that if the series
its
sum
is
.-(-Xsinx+Ycosx),
iY
can
(see Art. 136).
_
v~
given by
00
e-"v
dv
V=-j- (^Tcosa;-h F since).
>iii:u\
132]
01
these expressions
the
,..,-,.
qua] to
.j>lete
tl
(lill'.-lrntl.l!
Th
Lav
Id
/}')!= _^1.
Benoe we
r)H
</}'
1-
It
It
for
integralB
)'
.\\
th.-sr
that
;m.l
:.!
man:
du-
l.\
-t.-il
t>
and
'aiK-liv,
<-.iuliti..n
tli-
i'm.1
)-
"<1
that
to
---IHI
tl.-
f\
an<l
I'
)'
.\',
'I'lif
tin-
..
-JT--1.
l.v
tin-
1,0
an- actually
int
i.sson (see
asympt,,;
l.i-
i.--j.r.-tiv.-iy
as
'
to
be
i.lv
It
that
is
\\ni-tli
}M-rh;i}
relations
tlu-
hy th- qse 6f
[ing's
in
!><
A'.
tin-
I'
aii-i
mark
additi
tinI'
-t
naturally
asymptotic
*
(Appendix, Art
j.rovt-d
wliere
Now
between
niakr
to
wliil.-
\fs(.r
= (.'+
.1
1<>^.
1*0) that
./-
we have
aivtaiK
X ) = (v
>'/
-...
.>)-
+ (-1.
,/_,(''>
whew
(
wh.-iv
the
961
//
Appendix, Art.
ix
numerically
ITi'-.
less
Bx.
:>>'.
than the
We
li;i
tirst
term omitted
330
SERIES.
[CH. XI.
If we take the quotient of two consecutive terms and remark

that (compare Art. 130)
where Q
is
a factor slightly less than
integral part of TTX;

of accuracy which is
(4)
of (1)
we
see that the least
is
The reader may discuss, by methods analogous

and (2) above, the following integrals:
7
'
the
1,
given by taking n equal to the

but the first two terms give a degree
ample for ordinary calculations.*
value for the remainder
first of
which
is
related
OJ.AJ.
i/
to those
~r
the error-function, while the
to
second and third are the cosine- and sine-integrals.

133. Poincare's
theory of asymptotic series.

All the investigations of Arts. 130-132 resemble one another
to the following extent:
Starting from some function J(x),
we
develop
it
formally
in a series
a,
ct f
is not convergent, but yet the sum of the first

terms
{tt+1)
gives an approximation to J(x) which differs from
n+l where
than
n /x
n depends only on n and
J(x) by less
not on x.
This series
Thus,
In
i
In
if
Sn
denotes the
such cases,
sum
of the first (n
+ 1)
terms,
we have
we say
that the series is asymptotic to

function', and the relation may be denoted by the symbol
Such
all
series
*An
when x
were called aemiconvergent by the older writers.
elementary treatment of this approximation will be found (for tin ra.Mis an integer) in a paper by the author (MeHsenyer of Moths., vol. .'>(>,
J90C, p. 81).
132,133]
ho\
to be noticed.
is
It
II
more
t.,
thai
.'xampl.
lin.
vuinr
tin-
s.-rir.s
\\ ill
from
fi.ll..\\s
It
-0,
ivpi
tin-
definition
tl
..
lit.
tlf
JC6
the
that
nil.-
)^" +
(l
where
still
ril-
^/'o +
XX
1
-i
.Illill^
then
\V.-
appli.-s.
"
V
X
liml.
if
'
.'
formal product
tin-
Illptotic
pr-'dlict
Art.
of
ll(.')^'',,+
= ",/' + "/'
',
-f
J-f
+"
,-h--
r/jT,
\\liii-li
by definition 1
;!
ins
\\lu-iv /'
'
"
in
is
..
<-Min,-i(lr->
thui
polynomial
in
tip- pi
\\ it li
mlurt
>'
7'.,
lii-li--
mi
ami
iip-lu<lin;_'
'
MI
whose
...
t<>
nj
'/'
fi
.''""
if
t..
1).
us
'
t-nl t-
[ /( /v -, "][
-
/1
1-2.+'.
.'
p-0,
<r->0,
ami accordin.
lini
Thus
l.V
tin-
'=0.
.*"[./(,
OB
].rMiuct
JT
/{
\.
**>
A'
In
II'
in
iid
p-v.
us
Lft
a-ymptotir
imw
conoid.-!-
tin-
particular W6 CU ohtain any

the ordinary ru
,y
p-.ssiliility
./
in
c\id-ntly writ.- /{
th<
-f'V 7
*Of
'
>t
II"
In
may
asyiujit.
,,t'
tlir
sul'stitutin^
first
ûl^titut<-
p
-',,-f ./
an
332
SERIES.
XL
[CH.
and rearrange in powers of J 1 provided that |a is less than

radius of convergence (Art. 84); because lim /1 = 0, and we
can therefore take x large enough to satisfy the restriction that
,
the
tt
l
ol~H</il is to be less
than the radius of convergence.
This having been done, we

the asymptotic series
for
/,_
consider the substitution of
may
in the series
*Vi) = t' + C.J,

Let us make a formal substitution, as if
were convergent; then we obtain some new
*-*4+^ +
where
Let us denote by
Sn and 2 n
the
sum
of the
the series for
series
:::.
terms up to \jxn in J^ and
respectively.
Now,
if
2,/
2n and 2 n
'
polynomial
= C + C.Sn + C,Sn* +
in
l/.r,
. . .
+ CJSS,
and consequently 2 W
l/#
n2
n+1 to
thus
(l/^)
ranging from terms in (l/#)
agree up to terms
in
'
is
(1)
Next,
we
rn = CQ
if
have, since
Sn
represents J^ asymptotically,
therefore
(2)
^- Tn = Cn+
Finally,
thus, since ^("/i)
where Jf
is
is
convergent,
a constant.
Hence, we
find
(3)
because
By combining
(1), (2)
and
(3),
lim
we
see
now
that
\imxn (J1 r
Snr )=0, and
Thus
3BR1
;\
the
sequent
!;.
pou
kan
t /,,
Further, a
cor
tin-
in
rearrange
in
"i-dei-
in
his
in.-ijualitv
.'./
i^
<
/'-'/'
iii.-,,uality
wo
;iti(l
./,.
i'
j
tin-
places <>niy.
secondly
-1
and thru we must
r-^tridinn
"//
in
th--
il'
whose
that
first
tl.
>UJJM,>.-
'
ill.-
tir-t
estal
ord<
mtir.-ly M\.il
bo
M-l
bo /'(/,):
in
tl
Mini
-/
l
/(./
term
is
..fitufii-
zero
mo
-i'/f
a,,
nt.
An
t'nnm-r n^nlt
th<-
.f
application
(assuming that OQ
-".(14wht-re
A'-x,
Th.n
aini
J-i
-i
,,
sain*-
rule
t.
rstal)li>li
Fr
\\-
an asvni])t"t
as
it'
ic
Beriefl
th-
'i
= 0).
j&l
\\,-
J-8, <f
>.:.
././..]
./ '/./
is
..".
if
x>
<
if
-,{
.r
ij"
l.-t
which
\\
divide
Finally,
can
A').
Thus, applying the rule for multiplication,

,,/,,
tli-
i_A'4_A-
thus r.instnirt
.-xacily tin-
liy
"'
is
x-rn).
\\<>(
i->
asympintically
l.y
w.-iv '}\-
we
see
334
an asymptotic
But, on the other hand,
SERIES. [CH. XL
series
cannot safely
differentiated without additional investigation, for the

existence of an asymptotic series for J(x) does not imply
the existence of one for J'(x).
be
Thus
e~*sin(e*) has an asymptotic series
0+2++.x a?
But
its differential coefficient is
tends to GO
and consequently the
+ cos (e*),
which
oscillates as x-
differential coefficient has
no asymptotic
sin (e?)
expansion.
On
the other hand, if we know that J'(x) has an asymptotic

it must be the series obtained
by the ordinary rule
expansion,
for term-by-term differentiation.

This follows by applying the theorem on integration to J'(x)
We
quite as simple, and perhaps more instructive.

theorem that if <f>(x) has a definite finite limit as x tends
either oscillates or tends to zero as a limit*
is
proof
+ ajx + a2 Ar2 4-
Thus the
differential coefficient
^+
but a direct
use of the
to QC
then
<'(- r )
we have
If J(x) <x
. . .
make
/'(^)-^'(
[t
has a definite limit, must tend to zero. But xn [J(x)-Sn (x)] does tend
to zero, so that \imxn+l [J'(x)-Sn'($y], if it exists, is zero.
That is, if /'(#) nas an asymptotic series, it is
if
it
It is
instructive to contrast the rules for transforming
and
combining asymptotic series with those previously established

for convergent series.
Thus any two asymptotic series can be
whereas
the product of two convergent
multiplied together
series need not give a convergent series (see Arts. 34, 35).
:
any asymptotic series may be integrated term-by-tern i,

not
although
every convergent series can be integrated (Art. 45).
Similarly
111
fact
if <f>(x)
tends to a definite limit
we can
find
"
Thus, since
we
find
So
= 0'(a
>X
where x > * >
x
if
*'() </(*-*())
so that
.
a-
'
but the last

cannot approach any definite limit other than zero
may not take a// values greater
inequality does not exclude oxcillatioH, since
than .r as x tends to 00,
<f>'(x)
i.
Til
133)
On
ivasonin
'nptotir
apply
wi-
that
th--
l'-i
bo*
l'-l
convei
parameter with
tin-
as sm-pi
to
ivsp.-et
\\
ilitl'.-n-n!
and
no
in
l.-jH
on
-niU
an eveo
in
an
l'tinilam-ntal
jM-rt'rrt ly
tin-
the
pies,
Tin- contract
i>arani.
iiioi-i-
of
d-tinition
asyni]'-
'
tin-
may drlin.- a
aM in])!<>!ic Berii
in'" th-
enters
way
Ml
in
<-on\
i.-ii-
contrasts,
a
-in
tli-
an
li;i^
In
dealing with a COir
in
li,
iliH'eivni
in<l-ri\
-mlin^
in
aerii
the
we
unless
depend'-nt
to
'
other hand
I!M-
ditl'eivnt
<r kSYMPTl
l:Y
i:
may
l>c
ilii.
"-llicii-ir
way.
asymptotic
coinpl.-t.'(l whole
lnl-i
in
tin-
>au
ies.
It
that
when
./
convrni.-nt
sniii.-tiiiifs
i-
i-
represented asymptotically
l-linitin and -ay
our
to .-xt.-ml
tli.
l>y
''
'I*
is
represented by
two suitahly chosen functions

Thus, for example,
asymptotic formula
f
"+ X
!+..., where
l'
we can deduce from
n~.
Stirlii,_
the
i/V^...),
/!
r1-
<!.
'
12'
-2
r =-B
11
-2
Hitherto
coiiij.lex
i
i,
./
\aluex
l.ut
and tends
to
-)
sn]]ios,-d
the
/:
,-
dued analyi
theory
in
any
to
ten<l
remains
..tln-r
to
-s.
through
unaltei
definite
lin-rtion.
l;
Mt
336
In
fact, if
we can determine
r_ a
"
SERIES. [CH.
XI.
constants M, R, such that
_i <M_
x
^l*!
2'
HI C|>
!->/?
follows from elementary theorems in the theory of functions

is a regular function of l/x, and consequently the
it
that J(x)
asymptotic series
is
convergent.
For different ranges of variation of the argument of x, we may

have different asymptotic representations of the same function
which between them give complete information as to its nature.
illustration of this
good
phenomenon
is
afforded
by the
P>essel
functions which have been discussed at length by Stokes.*
Applications of Poincare's theory.
134.
An
important application of Poincare's theory is

solution of differential equations.!
The method may be
in
the
up
following steps
to
the
summed
1st.
formal solution
is
obtained by means of a non-con-
series.
vergent
2nd.
It is shewn,
by independent reasoning, that a
solution
capable of asymptotic representation. Thus we

has
been done in Arts. 131, 132 above, deduce a
as
may either,
definite ^.epral from the series first calculated; or we may
exists
which
is
find a solution as a definite integral directly,
with the
it
series.
The region
3rd.
and then identify
is
determined in
which the asymptotic
representation is valid.
Poincare has in fact proved { that every linear differential
equation which has polynomial coefficients may be solved by
asymptotic series; but his work is restricted to the case in

which the independent variable tends to oo along n specified
*Camb.
41-J;
p.
also
Phil.
Mti/h.
1857, p. 105, and vol. 11,

See
350, vol. 4, pp. 77. -Js'i.
and Papers, vol. 5, p. 283.
Stokes
Tram., vol. 9, 1850, vol. 10,

and Phy*. Papers, vol. 2, p.
Ada
Mathe.matica, vol. 26, 1902, p. :>!>:},

that in the asymptotic series examined by him, the change in representation occurs at a value of the argument, which gives the same sign to all the
r.
marks
terms of the divergent series.

Some interesting remarks on the sense in which an asymptotic series gives a
Million of a differential equation, have been made by Stokes (Papers, vol. 2,
I
:W7).
]..
Acta Mathematica,
t.
8,
1S86, p. 808,
and
tilled
tli.-
iiiiin
.lacnl.Mhnl in
l.y
applieat ion- "f
ami Manly
tin-
met
BEE!
(
i
paper
in-
This gap has

cases;* and
|ii.t'-l
l.elnw
in
made
theory have been

asy mptot ie re]
a|..j.tMl
I'.ani.^
l.y
IM-VMIK!
i^>
MM th- tlnM.ry
lr|M-nls
th-
in
which can
l'nll\\ in--
l.c
\\^-<\
in
<l-aliirj;
wln-rr
>
îmjil.-
with c-rtain
and
"
is
_p
= \, and
integer (a
elaborate methods);
because logi
u log
log Z
typ--
consider
or
This
ivefl
See a series
'.
and then
A'
th-
term
-.
ua
i-p.
IMis
form
renmvrd
/"
whnv
usin^
the greatest term
l.y
an<
A
,
c\p<
lM iir
/-J. u/
\uf>*/t).
the
ot
Is,
'!! in
HIM
ins, in;.
c<ut>. rhit.
and
\iisohnitt \'l
:.
:.
of the
is
= M /-Aul.-(^/)-\
r+;(
-
fa
not of lii^hcr order than v //.
n-Mrirtion which can
an
nioiv
give,
ar+D^,
(see Art. ls<>. Appendix),

is ronvrni.-nt to
It
suppose that X
is
\\'.-
due to Stol
tind.
\\'.-
iiH-th'.,l
inei
book, a
this
>
-/'
I--=/UL,
luro',
is
and
ival
is
oi
-rat inn.
in!
v !'(/'+', + !). ..rfri +

-'r ( , l+ 6 1 -fi)...r(7i-f
X
{he
asymptotic expression for the series
135. Stokes'
\Vi-itf
Hai
l.y
liinit^
tin-
c< >nt< .111-
<>!'
art id.-
l>y
nmnary
ly power-Beriee
referencea i^ ui\--ii
i'u',1
very
ial
tin-
given
with
we
different
,\
Poincare?fl
.nst
iii
-I
-perial
diM-uî,
detailed
tO
iii
PTl
determined.
not
.).-
Horn
hy
ma;.
equation
AlTLh LTK
133, 134, 135|
fi
see also
//,
I.
vol.
338
where q = e~^
/t
of Art. 51, Ex. 3,
Making use
SERIES. [CH.
we
q approaches the limit 1) the series in brackets

approximately by 7r*(l q)~*, or by (2-jr//tiJ*.
Thus the asymptotic expression

e'H-*
T
is
see that (since

is
represented
is
i
rr,
where
M*(arO**"
Hardy* has proved, somewhat
f(x)
XI.
= x-.
in the
same way, that
if
Aex/(27rx)\ where
represented asymptotically by
= e~*.
and
+...
SUMMATION OF ASYMPTOTIC
SERIES.
Extension of Borel's definition of summable series.

In the foregoing work (Arts. 130-135) we have shewn how
to obtain asymptotic expansions of certain given functions, and
we have established rules of calculation for these expansions.
136.
We
now
are
led to consider the converse
problem of summing
method is to apply
But on trial, it appears that Borel's method
powerful to sum even the simplest asymptotic
a given asymptotic series

Borel's integral.
arid the natural
not sufficiently
series, such as (Arts. 98, 132)
is
we apply
If
Borel's
method
of
summation
to this series
we
are led to use the associated function
which converges only if t<x. Thus the results of Arts. 99113 no longer apply, but we shall now proceed to modify the
definition so as to obtain corresponding results for such series.
We
and
have
u(t, x)
= x/(x + 1),
this defines a function
to oo
which
Thus we can agree
*Proc. Lwul. Mntli. sV.
(-2),
vol.
i>,
if
is
<
x,
regular, as
to take the integral
191)4, p.
339.
ranges from
1M \Tl'\ OF
135, 136]
as
Hi.
d-tiniii^
writ-
Tin- reatlrr uiii

\\a-
\vlii.-h
The
'_
But we can also

and so obtain a second definition
intfi-al,
integral agrees with
i,
in
i\'-n
339
i<
the series above.
this
in
U3YMPTO]
t\vc integrals are
vah-nt so long an
e.\
it'
nut
fiinrtimi
hat
ditli.'iilt
in
tin-
part
ny
two
iy
tin-
negative real
paiall-l
lilish thi>
plain-
that
can
an analytic
ir-'in the
in'
th-
Thi- p;-
/.
juiin-il
as in
tl
that
te
<4<.
that
tlu-
test
..f
Ait.
174
tin-
s.-mml integral can be proved to define an analytic

J is greater than
\
part of the plain' for which tin- r-al part
this n-îitii is l.oumlc.l l.y a line parallel to the iinaîn.i!
function in
;
and
i->
./
!>
In likr niannri
\vhi<-h
int.'-ial
th.-
Aj)pcti(li.\)
for
numl..-!-
p..-iti\.lint-s
I'MMIM-I-
tin-
"-
and
p<>
is
i>
It
and
real
is
c,mple\ th,- f,,nu.-r i- convergent e\,-,-pt \\ln-n

negative. whereas the latter converges only if tin- ival part <>f
l>ut
I.
Art. 08.
tin-
in this
nirtlio,!
example suggests a general process
an
for Minmiin.o;
that
is
an asymptotic 96riee Mich that the associated

r
r-
'.-
seri-'-
+ 'V + ", L + "::i: +);
OOB
i'roni
tl.-tiiuMl
=
\vt-
certain jiositivr valm-s

by the series i'r these
oi'
'-.and tliat thr I'unction
\aln-s of
is
i-ontinuous
to
di'tiiu'
tlu>
sum
..t'
tin-
'.
|
srrirs
l>y
tin-
integral
340
SEEIES. [CH.
XI.
To make any practical use of this definition we must, howassume further that a positive number I can be found
ever,
lime- lv fn (v) = 0,
such that
v-><n
where
any index of
is
differentiation.
evident that then the integral defines a function which

within the region indicated in the figure.
It
is
137.
is
analytic
Rules for calculation.
It is easy to see that the series of the last article represents

the integral asymptotically, x being real and positive.
For,
if
we
by parts we have
integrate
= aU +-l
/
because /(0) = a
Now, by
We
e-'f(-\dt,
\
/
I .
continue this process, and
hypothesis,
we can
w+1
so that
|/
'
and therefore
we
obtain
find the positive constants
(^)l<-^^
"
'
Consequently,
which establishes the asymptotic property.

It will be remembered (see Art. 133) that the differentiation
of an asymptotic series requires special consideration; but
this
special
any number
type of asymptotic series
integral
x>l), because
t
\te-
and the
integral
te~
Jo
be
differentiated
For
of times.
the differentiation under the

(if
may
f'(t/x)\<Ate-
~
(x l)t
dt
is
(x
sign being permissible

- l}t
,
convergent.
SUM.M VTION
136, 137]
Now, within
so that
int.-r\al
tli.-
/'(g) is
which
is
tin-
ptotically
c.
.intril
tin-
the series
l>y
for /(#) liy
orij
ant
lie
ici
with
tli.-
pat
from
.--I
of times.
oliMiients <f funrt ion-theory will
th-
l.y
.iixidn-
I'-Miicaiv
we have
>;)
pruvin^ that th-se n-snlts arc valid within the part of
indii-atfil
IH
L--t
i'mm
/<
...),
/
or
could
as
litliriilty in
|ilarn-
"i"
opn-atioM may be ivp-at-d any number
:'-al-i- \vl.
IK-
a;
d.-ri\rd
Beriefl
diir.-iviitiation.
rly tin-
lia\r
I'
OQHV6fgenoe
!'
-'
x\x
formal
AS1 KPIU1
n|
i:i;.
M.-xt
th.'
it
operations;
algebraic
follows
theory that two asymptnt

may be added
if
Uut w- must aUo pn>\v that
convergent,
ie Beriefl
and multiplied as
the product can
considered above.
!.-
ivpi-.-^.-nt-'d
an
l>y
integral
of the
type
\Ve have proved that
is
tin-
-"nvspondin^ to
asympt'itie Beriee
Now
oonverge und.-r the con. lit ion ||

such that
|a w |r"<J/./!
= r,
Thus
:J
Cn r-|<3f [/.:-h(//-i):
<
.!/-(
and oonsequently 1
Art.
product,
it'
AO-Xa
both
if
tin-
we can
find a constant
< M.
6n |r"
---j.:.-j:-h...-f-'
will
oonveige
it'
<
/.
Next, as in
106 above, we have
'"
,
..i,
ai^K'
342
and consequently by
Thus
we
differentiation
SEEIES. [CH.
XI.
find
Borel's associated function for the product-series is

V
F(v) = atf(v)+
and
But,
I
\F(v)\\a
\
Jo
f(v-t)g(t)dt
\.\g(v)\+\"\f(v-t)\.\g(t)\dt.
Jo
by hypothesis, we can
find the positive constants
and
such that
\f(t)\<A#,
\g(t)\<A#.
F(v}\< a \Ae
Hence
lv
+A
ve lv
and consequently we find

F(v)]
= Q.
lim [e-* vF*(v)]
= 0.
2lv
lim[e-
v>ao
Similarly,
we can prove
that
v-><
Thus F(v)
the necessary conditions.

these
we have the following rule
results,
By combining
a certain number
contains
polynomial expression
satisfies all
If
of
asymptotic series* (of the present type) and their derivates, the
value of the polynomial is expressible asymptotically by a
same
type), obtained
by applying the ordinary

rules of calculation, as if the series were all convergent.
And the result cannot be identically zero, unless the terms of
the resulting series are zero.
series (of the
The reader who

difficulty in
is familiar with elementary function-theory will have no

extending these results to the complex variable, in a suitably
restricted area.
*0f course some

in
I/a;.
of the functions
may have
ordinary convergent expansions
SUMMATION OF AflYMFIOTH
137,138]
Another integral for an asymptotic
138.
numlM-r of
Tin-re are a
come
ini_
i:
nn-thod,
asily ly niaki:
we suppose
IT
Art.
MS
true
I'm-
int.
/(a
In
ha-
t!
ii'
can tind a
\\c
(-D
we
shall
For
have
this
i-
tin-
t-aîly
take this integral as
DOH-ZerO ra<liuMI-
-ati-li.-<l
tOCe
i'un-ti<.n
ni'
as to
such that
\!s(g)
Jo
|
-jnation
}>roved to be true when |v|</>, ami we can

tin- d.-tinition of
f(r) for larger values
.
!/()!<
Then,
-)<f>'
Jo
an,l
That
tin-
intt-^ral
series follows at
}
fi
Me)
Jo
r
!
in
i-
one.-
l.y
fact
r-])i-i-^-uted
-...+(-1 r'+...+2
<
win
asymptotically by
Ml.s.-rvin^ that
/>
particular,
f.
anl that crrtain n-M ri-t

of increase of |/"(v)| with
06
tin- <l-tinition.
in
provid.-il that
which do not
but can be
a function Mich as is
specified in
0(.r) to be
that
th.seen
results
are still
pn-vious
rea.lily
i-
it
tin-
series.
\ampl.-s
'"
+,,.
"'
|.
+...+
tin-
NON-CONVEEGENT AND ASYMPTOTIC
344
Again*
SERIES. [CH.
XI.
J'(x)=
and, as previously, this can be shewn to give the asymptotic

series obtained by the ordinary rule for differentiation.
special type is given
and then
Thus our
cn
definition
by taking
= T(n-i-\).
becomes
=
It
is
easy to
repeat
the
previous
argument (with small
changes) to establish the corresponding theorems for operating

with this integral.
As an example of thewe could infer
Art. 132,
method, we may point out that in Ex. 2,

the differential equations for X,
by simply
last
operating on the asymptotic series.

Another example is given by the series
which leads
to f(v)
= I/(l+v),
so that the series represents asymptotically
the integral
139. Differential equations.
We
shall conclude
by giving a few examples
of the
way
in
which asymptotic
series present themselves in the solution of

and we shall illustrate the methods of
differential equations
;
Arts.
1.
136-138 by summing these series.

Let us try to solve the differential equationf
dy _a
dx~~ x
*
and
Differentiation under the integral sign
Jo
t<f>(t)dt
is justified,
because
we suppose
converges.
Thus the test of Art. 172 (3) applies.

fThis is the simplest case of a general type of equations examined by Borel
(Annalea de VEcole Normale Suptrienre (3), t. 16, 1899, p. 95).
IHKl
138, 139]
meanfl
l>y
On
'I'll
!'
.-MI
\i.
VI i"N'S.
BQ1
asyinp'
we
substitution.
is
1.1:1
find
g\:
= 0, A = -
-4
..=
-I
"
1.2a
~~
-2A,_
:'
i*
34 8 _1.2.3.a
'
Thus we
/,'
I,
I,
the formal solution
lino!
'/=-/'
ox
and by
Art.
1-,
)'+...].
J
the equivalent
bhif
l.'lti
-.
ox
-"
-,
I
and
tin-
it
7-T-|
diivetly that this inte-cal

e.jiiation, as of course must be the
bo vei'ify
Lriven
Art.
Of
integral
137,
n>i<l-r
tin-
iiK'tlitied
Bessel's
matiin
-a
Writ.- 'n
If
ire
ami then
sul.st.tui,7;
\vi-
tind
=i
ire
...
+ ...
+
Hence we
...)=o.
find
--(fife*
and
so
s45
1*.3.5 2
T3>
NON-CONVEEGENT AND ASYMPTOTIC
346
Thus
>/
is
SERIES.
[CH.
represented by the series

1
.3 2
__
.3 2
.5'2
2.4.6
That
and
so
if
---l
*=(- D"
is,
l}
we put
2.4.6...2w
2"'
cn
r(f)
v^ vn =
we have
This leads to the integral (Art. 138)
A~i
s>&
TT
(/
Hence
y=
,,00
/
-y-
ptfjf
C
Ct'O
which can be proved directly to be a solution

Similarly,
we
of
the given equation.
can obtain the solution
for the equation
d2y
du
J + 5s
3.
In like manner the differential equation
*s+<. + /~*>i-<*-*
is satisfied formally
by the asymptotic power-series
.IT,
y ~x\}~
a(l-^)
l.x
a(q+l)(l-^)(2-^)
1.2.^2
which leads to the integral
This equation has been very fully discussed by

Annalen, Bd. 56, 1903, p. 129), to whose paper we
W. Jacobsthal
may refer for
(Math.
details
as to the various solutions.
We
have now completed a tolerably full account of the

theory of non-convergent series, so far as it has been yet
developed on the "arithmetic" side. Its applications to the
Theory of Functions lie beyond the scope of this book, and
we have made no attempt to give more than a few theorems
whose proofs can be readily supplied by any reader interested
in such developments.
XI.
I
LMPLE&
Integration of
Wi- have seen
1.
Summable
bhatth<
Series.
'he series
.-
series
to be
*a).
This suggests*
\vhi.-h
is
_ ___
a-
and so
(0<a<7r),
COS a
OS 4?
b
tity
,-^ n(l +
.>
{>>
ai
MM(i
cos ./--rns a
.)u-t
f'
vi-i ili'-d
-a>ily
.os nx - cos na
2.
c 87U'-' /
in
lli-
\v-
integratiiii:
tind
log
the constant being foond
bj ufling the value
of the
sej
-.65).
'
log(f"-
An'th-r
fi.rin
tht-
..f
first
intrirral
is
-Trlogma -^)}.
e
that
,-
if
i.
0..+HX
3.
it
i>
Sinn-
\\r
sii^^f-t-tl
':
* that
if
is
//
This
i-i-siilt
is
easily
*Thr n-M
\,
nor
a multii>
HI
vt-rit'u-tl
=ir.
indr|MMid-Mtly.
intririiit
>vi-r
ti
.e
x=
NON-CONVERGENT AND ASYMFTOTIC
348
From
4.
we
Art. 103(4)
SERIES.
[CH.
see that
Thus
and
Hence, integrating with respect to
2J
The
series
2(
o
l) /(x
+ n)
0,
we
find
sin
is
sometimes denoted by
/?(#),
expressed by means of the ^-function (see Exs. 42-44,
p.
and can be
475), since
Thus, ^y x is rational, the series can be summed in finite terms the case
has occurred in Arts. 65, 90, so take now x = % as a further example.
;
=\
We get
< 6<
where
TT.
If 0/7T is rational the series
may be
expressed by means of "^-functions
and
so the integrals are then expressible in the same form.

If we allow
to tend to
in the sine-series, we get
[For
by Weierstrass's
Jlf-test,
and
^
I ?^(^) =
(.
_ ^) COS
^_J
S in
^6 log(4
si
Cesaro's Mean-value Process.
A specially
5.
by
Feje>,
who
interesting example of Ceskro's process was given recently

it to the Fourier-series for a function f(x) which
applied
does not satisfy Dirichlet's conditions (Art.

series need not converge.
In
fact, if
a n = - { f(0)<x*n$d6.
TTjO
and
and we write
6,,
174,
App.
III.), so
= - (*f(0)fa.nO dO.
'
TTJO
that the
JfPLEfi
xi.
I
\\-
find
tli.it
"
and
that
aritlinn-t
tin-
im-an
i-
"t"
.<,
x,
...
wh
ial
(A|.|.
caws
to
n.ntinuitiea present.-
u> writ.-
L.-t
is
it
that
o- w _i
=a
that (|-ayin-
us.-d
iiM-an
in
the
"lit iniK-us
la-t
th.-
.nvrr--
to
r
fl
to
L':
'
t. tin- il-tiniti-
l/l'-)J-^--
^+
^+/v)(i-^)<^ /;V
nuinl.ri
any
/-
i-
>ingle
^l
at ti-nt i'ii
''
tli-
und*-i-
L.-vnud our r;m_
i-
-^,1-''
pi.-.f
(l
rfl
tin-
in
Since
tind
mind"
Unit*-
arithiiM-ti'-
'
1~ :>'
fnin-ti..!!
l.ut
inti'^r.-ilili-
tin-
f.,r
,r,,
(litli.-ulty
\\f
'
I'm
'
limit
it>
fn-li
ni!i>t
Men
ra.-ily
baa a
\\h-
IK*
6.
W6
tWO,
iln-
The extension
thru
int..
K->s
than
,<
and
BO,
taking
tin-
limit a> N
tends
to oo, \vc find
=}
I
r=l
l"'[
becanae
Thti-
:it
(Alt.
7).
li.M./, l= o.
and
wr may
a].jly
1*^1
Tannri
y's
thf-i'-ni
(Ait.
to
1'.')
./
\vhirh
givi
1
ivsnlt
[Thi.->
IM
dt-
la
Vall'--
Ch,
N.\
MT
al>. a jnji-r
~MiVr thLft,
}<
\.
I'mis-in
+ V).
lv ilnrwit/
Wr
7.
due
is
{Mutt-
>;-'
- ,
T6
and
MT
'
I
n
;,
I'.Hlti.
j,j,.
(_>-
'
that
1<.
.'!
Lebesgue, Strie*
350
Hence the
convergent
and
sum
its
a n sin nx -f bn ( 1 - cos nx) }
this is equal to
lim
dx=
<r n
n-^Jo
<rn
sum found by taking
therefore equal to the
is
But
the arithmetic mean.
because
[CH.
series
is
SEEIES.
{ f(x)dx,
-
converges uniformly to the value f(x).

is also due to de la Vallee
Poussin.]
[This result
8.
Consider the application of Frobenius's theorem to the series
where
and x
t is
is
number
a complex
of absolute value
1,
and generally
S2
sl
sn
but
is
not equal to
1,
proved that
It is easily
real.
S3
_l-t'
~"
~
i~ So~
= {l-(-t)}/(I+t),
if
S8
'
(v-l)
_l+t
i>
*>
+
=
lim (1 -^ + .^ -.^ +. ..) !/(! +
Hence the arithmetic mean

and thus
is
found to be
1 /(!
),
x>-l
lim (1
or
- .r cos + #4 cos 20 -.r9 cos 3# +...) = 55
>1
lim (x sin
6-x* sin W+x
sin
* >1
30 -...) = i tan
[See also Appendix, Art. 155.]

9.
Apply Cesaro's mean-value process

+
/I
(/I
+/*)
(/I
+/2 +/S>
to the aeries
(/I
+/2 +/3 +/4 ) +
where fn is positive and decreases steadily to

The sum is
Kni [/i
a:
zero,
but 2fn diverges.
- (/i +/2)-^ + (/i +/2 +/3 )^ 2 --].
[HARDY.]
have, in fact,
[We
*1
=/!
*2
~/2>
*2n-l
and so the arithmetic nu-an
Apply
Stolz's
arithmetic
mean
of
.s-,,
*2 ,
theorem (Art. 152,

is
...,
II.),
%,
is
and we
lim^.,,.
that lim(^,, + i/w) = lim/o, (+1 = 0,

-as that of
s,,
found
pi <.<luct (1
tends to the same limit

l.y
applying Art. 124
find that the limit of the
Again, from Stolz's theorem we see

and so the arithmetic mean of 1} *a ..., %,+!
equal to
(<
tin-
.,,
,.., *_,.
The result can also be

+ 1 - ...)(/ -/i+/ - ...).J
2
unple are L'iv-n
'
Illn
10.
liv
t.ti
:.,-M
(I
r>
iU"
wllieh
11.
seeond
tin-
[in
base
tli
10,
Killer's .-alt-illation <|U..ted
V.-lili'--
to 6 d
\rt.
il
1 -J 1
to
ices,
.
that
II
= -098060
log ($TT)
O"
and
decreases,
Thus the
series
-n]
2(-l)r-[(n
nvergent and, fi.>m
C.
nu isseen to be
tli-
iC-filog
value nf
Tlif
rms
can
->uin
tin-
f-i)
i.l.-n
th-
in^ that
sum
t<>
-...+('0*,;:,
(.
Asymptotic
12.
l>y
is
Ksialilish tlu- asvni]titir

in./-
Vn=
furnmla
(-l)'T,1
**-
tin-
-...
(WTT)
i-
'
...<
I.
nunu-ri<-al
0787,
-- -
v ,+
(/'-)-
-;
JM
whi.-h It-ads very easily t>
Series.
-M;:ii,
<
and so on.
Art.
[C(.in[.air
13.
132.]
An.
^>
t<>
that
sli.-w
/'T
")
l0 ^
-/ 1
-" 'M
-/,^}
olitaiiu-il liy
t'Tiu in the
[Km- the
.t.-iijtin^
at
any >tage being
sn
first
less
than the
v
tin-
in;.
1
- 1
f.-lb
and iMr.im:
identity
/-
,~
'
K.T the MOOOd, integral.- l.y jur

..Cgested by the series and tan le established dirertly.]
integral*
352
SERIES.
[CH.
the error being again less than the following term.

[Apply the same methods again.]
By
f" e~ tz dt=
writing
and integrating the
16.
Generally,
if
latter
e~ t2 dt-
Jo
J^k
by
parts,
0<s<l, we
we
find that the first integral is equal to
find
r<
And
17.
similar expressions can be given for
Apply Art. 131
function /(#) = #-(*+*),
to the
A>0, and
deduce
in particular the formula
'"
72xl03
Hence evaluate the sum
l+i+i+^
2* 31
to five decimal places.
18.
/oo2i.a2
If
l-a:2
JO
^-^.
shew that
and that
ô
Hence prove that
r^dt-'
1.3. ..(271-3)
J*
and that the value

In
particular,
of the
for
value of the integral
remainder
is
approximately (a
a = 4, by taking 16 terms we can
L-fd*
lies
- 71
infer
tluit
the
between 1149400-6 and 1149400-8.

[STIKLTJES,
A eta Math., Bd.
9, p. 167.]
XI.)
<>!'
19.
^1(
r
and
tin-
i:Jl
asymptotic expansion
6!
d<
approaches
writing
1, l>y
-\!ii.. ii.
20.
From
= log(l
Com}
1..
p.
238.]
tin-
<(rb)"
-1
1.2...(r-l)
r(w+l)(^4-2)...(n + r)J
Thai
//-'
21.
n+1
Similarly |ii>\^ that
-H[See
Si
-ii I..".MI i.i -ii,
'
'7-100;
1
22.
I'lovr
icuiiiu'
al>..
1.,
pp.
I'.L'
that
numerators
I>MIILT
-'.
-I.
>
!:>.]
354
23.
More
SERIES.
[CH,
generally prove that
where
t(t-
1)
...
(t-
and
24.
It
and 2a n #n is conn
that the two functions 2a #n ?b n x have the
must not be assumed that
lim (/&)=!
if
vergent for all values of x,

same asymptotic representations.
rt
For example, consider the two series

x2 x* 3*
^ + __ = ^ +T j-_î_ g
.-11111
co
tAS
A.
*^
"^
_^
*^/T
+ ....
+ _j-_^_-j
*
Miscellaneous Applications.
25.
Shew
that
we
if
.attempt to find a Fourier sine-series for
from x = Q to ^ = TT, we obtain the
series
2 (sin 2x + sin 4# -f sin Qx -f
. .
),
which can be proved to have the sum cot# by using either Borel's integral
or Cesaro's
mean value
(Arts. 103, 126).
Hardy has extended the result of the last example as follows

If f(x) satisfies the conditions of Fourier's expansion from #=0 to
26.
x=a, f(x}- i_#

-7â + 7
except that near

conditions
(x
In particular,
at
if
+ '~ + ?
a)
expanded as
then f(x) can be
summable except
~~\2
\x
^
a)
satisfies these
Fourier series which
2?r,
is
x=a.
7n
= l, we
have the simple result

00
f(x) =
+ <2^ ( a
cos
nx + ^n sin w#)j
where
= ^-P r/(x)dx,
ZTT
Jo
bn
= -P
7T
f
JO
a n = -P f f(a;)co&nxdx
TT
Jo
f(x) sin nx dx
and
this is easily extended to the case in which /(.r) has any finite numln-r
of such singular points.
[HARDY, Messenger of Mat/is., vol. 33, 1903, p. 137.]
27.
Applying a method similar to that of Ex.

1.2
J_
A
or
while
2,
p. 42,
^ "T"
2wi
27n(2//<-l)
\
m+
(2w + l)!-(2)!.l! + (2w-l)!.2!-...=0.
prove that
BXAMPLE&
xi.
I
tin-
li.-it
355
;IM mptotK- series

'
J+p
is
,..
+ ->
'
by
represented
As
a verification,
thru trim-l.y trim
->|M-.-ti\rlv,
two
,/,-
denoted
(compare Ait. 137)
series are
ion gives
(li tl'rivii t i;it
--
a,
a.,.1
s,,
\vliich
agrees with the result deduced above fr'in
Le Roy's method desril..(l
28.
Art.
in
direi.-t
multij*!.
115 can be app!
ding's
&
-
taking
;>
=2
L\V, lind
series
Le Roy's formula.
in
in
fact
which converges absolutely
if
||<2ir|jr|.
Also (Art. 176, Ex. 2)
.-*+1)J
-
^T"
tli.-
17''',
cxplain.-il
tli.-
of
intfivluinge
A,
in
suinniatiiiii
long as
Art.
i:i;,
we get the
it
is
this last
to define
i-
as always
*}.
rain
iiiv-rsi..ii
^()
p.-nui>Ml.l.- to invt-rt tin- Drd
bare
Jo
l>ut
int^ratioii being justifiable

Thus, in a-,vrni.-nt with tin- j.rir.
we can agree
Th.-n it"
LIttegni
with rt-sjK-ct t> ^ ai
la-t
*'**
Md
and
||<:
n,.t
<r"
p n-tirularly easy to justify.]
givm
l.y
l.y
-gra-
356
SERIES. [CH.
XI.]
29. As examples of series which can be readily summed by means of

various integrals similar to those of Art. 136, we give the following simple
cases
30.
Determine a formal solution
of the equation
form
in the
and express the
[The
integral
result as a definite integral.

is
^/
V'(l +;)"'-<*,
if
> 0.]
Obtain a formal solution of the differential equation
31.
in the
_n(n-l)
*-**Vfl
4^2
L
form
and express the sum
n(n-l)(n-2)(n-8)_
4 8#*
.
1
J
of this series as a definite integral.
References.
For Exs.
Hardy, Proc. Lond. Math. Soc. (1), vol. 34, p. 55, and
For a justification of the integration used in Ex. 1, see
Hardy, Trans. Camb. Phil. Soc., 1904, vol. 19, p. 310.
For Exs. 14-16, see Diriehlet's Bestimmte Integrate (ed. Arendt), pp. 208-215
Nielsen, Mathematische Annale?i, Bd. 59, p. 94; Hardy, Proc. Lond. Math.
Soc. (2), vol. 3, p. 453.
The equality of the integrals of Ex. 14 was proved
vol.
35,
p.
1,
2, see
81.
by Cauchy, (Euvres,
t.
1,
p.
377.
In Exs. 20-23 the series are not asymptotic but convergent ; they may be used
instead of some of the asymptotic series found in Arts. 130, 132.
But the
law of the coefficients is much simpler in the asymptotic series and for this
reason they are usually preferable.
;
AITKNMX
AUTIIMKTic
[RRATIONAL ffUMBE
Tlll-:ni;v <1-
AND
I.
LIMIT-.
140. Infinite decimals.

\v-
II'
apply tin- Mi-ilinary prO060a

fraction a/6 to a <lrriinal. it
rath.nal
convert a
t>
livisi..n
!'
thai
.-\il.-ni
ix
tin-
.-ith.-r
process must terminate >r else tin- <|Utii-ntx imM n-eiir

(//-I) divi-ion^ at most; t'T in li\ilinu l.y l>, tln-n- are not
more than (/>-!) ditl'.-ivnt iviiiaiinL-rs pussiM.-. nann-ly
.
L'
3,. ../>-!>.
aft.-i
nninatin.ir
71
!:
-.
tin.-.-
= 7-
di\
frcurriiiur after s.-\cii divisions.
-.
And
i))
If tin-
drcimal part
t".
''
and
BO
/'
ll"t
\t
<li\'isil
1>V ritllfl'
)!
d--iiua!
formula
tlio
f..r
Tli';
'').
ii;il
'In-
'
<livi-.
pur-ly prrimlir and .-..ntainPm /''l"-"-l), ly im-ans of
is
ran
sunn
1) divi-
'2
"I'
<
.").
'"HV'M
it
t""ll"\vs
fr-'in
!i-i.n
K".
di\i-iM-
1'v
di\Mlh-
liv/-: thus a I
,-itlM-r
if
"i-
the drrimal pai
decimal
hrt\\
is
theorem that
can be found BO
index
/>
f..rm
tin-
/'
In'
1)
and
\\li.-ii
t'
so can be expand- d
the pel
in
liirurcs
e,,ntainin_
contain eith.
i"dic
and
Tic-
tie
the
pi..\.'d
Mi..t
.itlier pi
in
th,-
/>
it
.if
i> nii\.-.l.
Miu-t
n\MH'
tin-
an
.").
is
decimal with
i..-ridi.I'.ut
-1
-f
l.e
j-
because
relation
discussed so simply.
'I'!,,
USB, Ditq.
that
lt
if
= V&r**"t* ....
where
r,
is
prin
t,
r, ...
are
a factor of
It'
IKINV
we
intinitr <l.'cinial
inated (say
I
2t
will
).
-\-l'}\
.-vidt-nt bhal
fracfa
.rdin.K to tlie n.
it
is
more
o.nvi-niriit
;inl
\\
shall .i.llu-re to
'l
l>y
an
IRRATIONAL NUMBERS.
358
But we can
exhaust
easily
see
[AP.
I.
that the rational numbers do not
all infinite decimals.
Thus consider the decimal

1010010001000010...,
consists of unities separated by zeros, the number of zeros
increasing by one at each stage. Clearly this decimal neither
terminates nor recurs: and it is therefore not rational.
which
we may
Similarly,
take a decimal
11101010001010001...
formed by writing unity when the order of the decimal place is

prime (1, 2, 3, 5, 7, ...), and zero when it is composite (4, 6, 8, 9,
This cannot be rational, since the primes do not form a
10, ...).
sequence which recurs (in rank), and their number is infinite, as
appears from the Theory of Numbers.
If the
primes recurred in rank after a certain stage,
to find the integers a, b} such that all the
it
would be possible
numbers
would be prime. Now this is impossible, since a + ab is divisible by a;

and therefore the primes do not recur.
If the primes were finite in number, we could denote them by p lt p 2
...,>; and then the number
jt? 3
,
would not be
of primes
divisible
is infinite
by any prime, which is absurd. Thus the number

theorem and proof are Euclid's (Bk. ix, Prop. 20).
this
As an example of a different type, consider the infinite decimal

obtained by applying the regular arithmetic process for extracting the square-root to a non-square such as 2 this process
gives a sequence of digits*
;
1-414213562373....
This decimal has the property that,

of the first n digits after the point,
which gives
*
h.
rapid
VIII.
way
2-^ n
< 3/10",
of finding the decimal
since
is
if
An
denotes the value
2^1 n +l/10 w
<
3.
to use the series of Euler, Ex. B. 12,
140, 141
To
IM
l\l
this decimal
see that
only to pn.\r that there is
DEI IM
i.
cannoJ
UA
terminate
<>r
recur,
j.i-iM.f thug
Suppose, if
are .positive integers which
Now
at least one of them is odd.
\ecangivpoesilil*
tli;it
(<//6)
are mutually
= 2 we may
;
ami
j.iim.-.
<-an!i"t
a = 2c, we get 2c2 = 68
;'
'
U-
must
SO that
/*
we have
>on whose aqua/re is
i"
.ann.t
l.-
I..-
..tl.l
lint
<,<1(1.
if
we write
thu> arriv.- at a
w.-
contradiction.
The order of the system of

jM.ssiM... ;iinl in many way-
141.
It
up
is
\vhnlr
tin-
tli.
.,1V
shall
thfii
m.aninto the
that
lefl
tin-
of
mark
is
.1
BO that
/>'.
1'.
smaller.
iiiMjualit i-s
tin-
regard
tin-
ni'
ri^lit
shall
now pn.vr
,,n
ly best, to l)iiill
tin-
basis of order,
t<
we
.!>/). ]{>(.' simply as
of
th.- ri-'ht
falls
/;
u-ual. \v- ]>lac- tin- lai

aln^ a slrai^lit lin',
between
an<l the
and
mark
<'
<\
/;
\\'.'
.list in<-t
marks distinguishing certain objects
arran-v<l in a <l-tinit- nnli-r.

tin-
i>
rational nuinlH-rs
"!'
be nuinlMTs as
numlM-rs to
infinite decimals.
it
that KM cam
l>f<n<>
t lie
same
inent
I,, I
that
and that
tinin'
say that
we
w<-
I,,,
an-
tin-
same up
to a certain sta_
tind
.1.-...,
with
directly.
fiml
+;,
...
nin-fspoinlin^ interpretation for
/>',.
Then
\ve
1.
> 1/10".
me,
may
be negative, but that <r,
'i...
.1
would
in-
..it
,i.d
lew
tba;
a,,
IRRATIONAL NUMBERS.
Also lO n (B
virtue of the
Bn
[AP.
a rational number, and
is
method
of finding
B B.
while
is less
I.
Thus, in order to determine the relative position of two

infinite decimals (derived from rational fractions), we need
only compare their digits, until we arrive at a stage where the
corresponding digits are different; the relative value of these

digits determines the relative position of the two decimals.
By extending this rule to all infinite decimals (whether
derived from rational numbers or not) we can assign a perfectly
for example, the decimal
definite order to the whole system
1010010001 ... given in Art. 140 would be placed between the
:
two decimals
1010000
and
also
and
so on.
...
(zeros)
and '1011000
...
(zeros),
between
101001000... (zeros) and '101001100
we may shew
Similarly,
...
that the infinite decimal derived from extracting
the square root of 2 must be placed between f f and ff
we
ff = 1-411...,
find
so that, in
(zeros),
agreement with the

ff
ff = 1-4146 ...
For,
by
division,
rule,
< T41421
...<{f.
It must not be forgotten that at present the new infinite

decimals are purely formal expressions, although, as we have
explained, they fall in perfectly definite order into the scheme

of infinite decimals derived from rational fractions.
142.
Additional arithmetical examples of infinite decimals
which are not

Consider
If
is
first
any
rational.
the sequence of fractions (an ), where
integer,
and n
1
> m,
we
1
find
.1
IM Mil.
141, 142)
which
DEI IM
is
l.-.ss
LLa
than
'
'I-in
-'
nt'.l
//
Tim
l<
>//<
the
!,
for
iliH-iniiil
'
V(iV
ra+1/
Ix-tw.-.-n
lies
'
tin-
by
?lll
= <'.+
+l
hl)l
In this
w,.
\\.i\
tiiut
tin-
su.Ti->siv-ly
and
1708
Mid
L71888
and
hi-arly i-i|ual:
\v- an-
tin-
1...1
-1
IM-
<>nr
which
c<
n^t
<
as.Mjuival.-nt
dfcinial corresponding
to
than th-
lM-iinal
+
-2
:{
the last ex;
I'I-MMI
+ " + 2.3W - 2
'
-2
:>-
'
i-
SW
the drcimal
il.-riv.-d
+
'2
J_
wl
4!
31
MUMih'T.
i-atinnal
any
t<>
111
matt.-]- IIMW
n<>
)rr^j)<)ii'U
21
Bence,
become nmn- and m<nv

rud an infinite decimal
<l.-<-imaU
Inis
pi-MYrd that this intinitr di-ciinal can'
IIMW
tin-
+ -+' +
I-'MI-
Ami
Which T
(1)
will
two
increases, these
(] '71 8281828... X
It
a.,,1
'H.
AJB
with
w=3
and TT
TIHC, and
:!;;
and
lii:.
d.'i'i\.'d
i'r..in
'-8-!(-^>
many
th.-ir
t.-nn^
we take from
sinn will
l>c
less
'lian
21^0-H
1_
1
ir^t.-fix
>
]
.
+ ...
than th- d.-cimal
75,
I'.ui
^,-f
1.
IRRATIONAL NUMBERS.
362
so
-and
Thus the decimal
on.
[AP.
representing any
number
I.
of
terms from
must be
than
less
"I
\/ ^
O\
\ /
O\
'75.
that (1) could lead to an infinite

decimal agreeing with the decimal derived from a/c, where
a and c are positive integers. Multiply by c\, and (1) becomes
Suppose now,
The terms
if possible,
some integer
in brackets give
/, say,
and so we
iind that
that
is,
which
same
an integer equal
absurd.
is
decimal which
to a
Thus no
less
is
decimal as (1) does.

Consider next the continued fractions
infinite
&
Here,
we
=!+
lies
Thus we
between
b nl+1
=p/q,
terms
6 TO
= r/s,
= l,
\p8
qr\
and
b m+l if
?i>w + l.
find for the successive values of b n ,
1,2,
so,
recall the facts that if
then
while b n
to
jx"'
bm
converting to decimals,
than
'75,
fraction such as a/c can give the
|,
we
I,
...,
see that 6n lies
2
1-5
1-67
1-6
1-625
1-615
1-6191
between
142, 143
OBOME1
.\l.
l:i<
\\iri.l.lliniv
6
-ijiiaI.
an
i"
lead
.-mil
decimal
iiitinih-
he considered IB equivalent
bo th.-
caimoi
decimal
infinite
fraction
'
were we -houM
it
ii'
_"
<
(pcaq)
this
is
'
/.'_'
~ a<J\ <
an integer,
IB
nlivinusly alsurd.
rth cniiN.T^.nt
cnit inin-l
Similarly,
'
la-i
tin-
lit-
'I
'
'2+
an inlinitc decimal which
to
1,-ad
t<>
]u-nvrl
>
il'
lVac-tin
2+
IM-
fainilar with tinth>
rt-.-nicr \\lii. is
utiiiin-d
that the aqtuire of (3) con vergea to the value 2

in
..iiMfxii.n
A-
hut
easy to 866 that the
is
\\viv .-<pial
it
with
,-nd
&
l<>'
it'
0.
J]
...
"
C,
wh<
w- should
endfl
with
is
impossible,
that 3,5,6,7, 11.... cannot
\v
have nr
Inn-.
I-YMIM
fche .-.\aiii]>lfs
system
fulfil
143.
Geometrical examples.
143.
tin-
must
10"
Thus
Similarly.
loâi'.it
it
fi
leted
of Art.
-\aiiij>lc
.-xani].!.-.
rational.
whil<
-30H
For
rational.
or
nt
is
dfcimal
i
cann..t
all
-hew
L'-t
(M
I)
to
tin-
SMiM.-what ditl'nvnt
intinih-
r,
\vith
gives r>>-:
of
litimi
iiiinator
14-L-L.L
can
.,!,.
1.-
than
tin-
rational
'/
\PC
>inc..
lV.m
have the inequality
'/
SO that
Inn
inued n
d"-n\.-d
which can
10...,
intii.
L+ITiriTTTThis
DXM
;in.l
in
of
the n-. ds
that
,!'
in
Ar
nuinhers
alêhra.
is
\\'e
14:2
l,y
shall
it
i-
evident that
no means sullicient to
now
jrive
an ezao
does not Mitlice for geometry.
it
^ht
Euclid,
--ivcn
rational
line
Bonk
.!/;
II.
be divid.-.l at
.prop.
r ~Y~b
11).
r*
in
so that
7"
"i
"golden section"
AC \CB = AB
AC.
IRRATIONAL NUMBERS.
364
AC
then easy to see that
It is
less
Cut
than twice CB.*
AC
once that
is
divided at
off
[AP.
I.
must be greater than CB, but
AD
in the
equal to
same
CB;
ratio as
it
follows at
AB is
divided
For we have
at C.
AC:AD = AC :CB = AB:AC,

and consequently
AD:DC=AD:(AC-AD) = AC:(AB-AC) = AC:CB.
AD
Also
is less
than half of AB.
Thus
we
AN, which is
repeat this process 2n times, we arrive at a line

n
less than the 2 th part of AB.
suppose, if possible, that AC/AB can be expressed as a
if
Now
rational
DC/AB
fraction
is
then
r/s;
(2r-s)/s.
Hence
AD/AB = CBIAB is
AE/AB is (2r-s)/s
(s
and
r)/s,
and
ED/AB
Continuing this argument, we see that AN/AB

must be some multiple of 1/s and so cannot be less than 1/s.
is
(2s
3r)/s.
But we have seen than
than l/2 so that we

are led to a contradiction, because we can choose n so that 2
exceeds s. Thus the ratio AC AB cannot be rational.
AN/AB
re
is less
71
prove similarly that the ratio of the

side to the diagonal of a square is not expressible as a rational
In fact, let ABC in the figure represent half a square
fraction.
It is
not
which
of
difficult to
AB is a diagonal it is at once evident that AB is

AC and less than 2 AC. Cut off BD = BC, and erect
;
greater than
B
FIG. 38.
=
and
perpendicular to AB at D; then we have ED DA,
=
for
the
line
of symmetry
EC ED, because BE is a
quadrilateral
BCED. Thus EC=DA. If we repeat the same construction
on the triangle ADE, we see in the same way that
AF=FG =
DE
*The
less
first
follows from the definition
than twice AC.
than twice CB.
Now,
since AC'.
and so we see that
CB=AB :AC,
it
AB = AC+Cli is
AC is less
follows that
143, 144]
Tim-
D(
.1
the
Tim-.
AN
ANP,*\u-\i that
lo
-imilarly,
/'
ii
we
continuing the eond
i.y
an isosceh
at
AO\ and
than half
than hair Ah.

arrive
365
\l.
than
is less
L'
Ar
\\
A/I isa rational fraction
AF/ABis 8r 2a)/a;
AN/AJi is in lea than
that
.1
than
4 then
-1 1>
.!/>'
is(*-r)/,
ami continuing the process, we see
hat
r,
\\hieh
Tin-
ivadn
"Ex..
of tin
cle
144.
I,
m;i\
sli.'\\
My
oonrerget
ratiM
in
to tin-
di.i-"ii;il
tli.-
ad
\B
th.- coiitiiiMfd
tli.it
.f
40
.\'
less
as !!'
n.
'
to tin-
Of that
leads to a emit radict
the golden iection

side of a 8quai*e.
whfla
special classification of rational numbers.
imli-at.- tin- iif.-l I'mM>

Th<- rxain]l-v ,.f Art-.
I'.ut before
smut- thirrational numbers.
proceeding to
-'i
f'Tinal
which
l-liniti<n.
shall jrive
some
will
t<
1'4I42I
Tin- intinit-- d-ciinal
in
...
)cd. -kind's
(A) Thi
/"////
(Mich as
jM
than or
l.-.-s
infinite .i.-cimals
fractions
rational
all
smm- tmii
to
i-ijual
a
we
Art. 140 mal>l-s
in
disj-ii-M-d
which contains
r//xx,
arti-l-,
dctinition.
divid- "// rational nmnl'.Ts into tw<
'
n-xt
tin-
which -h-w how
cnnsil-iMti<.ns
which do not ivcur h-ad un
found
!><
of
tin-
sequence of t-niiinat<'d decimals

1-41.
1-4.
1
(I
It
,)
is
(i)
upper do88
T/
M6
we have
Any
numl>-r
Thnv
tin-
i-
least
truth
in
tin-
tin-
It'
and,
fraci
tin-
s-.ju,
>f
<!.. n.l
-t :it.-iiM-iit
L'-/-'
t'^r tli.-
t(S
iiiiin
sam.-
-Train- than every
m.iv
\v.-
lf a iMtii>nal MUMilx-r of
similar aix'uin.-nt that
i-
the upper da--.
'1'liriv is tlii-n-fiiiv n..
MOW w. suppose
tlian /.
in
/-/--i'(i
tin- l(.\v-r rlasx.
l.v a
rational
dasa
nuinl-cr in the
~t
numlier
..f
n]>p.-i-
the lower
in
latiiinal
bean;
etc,
all
(-uch as ;;> ^-i-.-atcr than every trrni

drar that
no
'!'
1-414-J.
ih'-n
numltcr
(ii)
1-414,
\\hich contains
that
\\ill
we have
/(,
^ to
.mlirr in tin- l.\v-r class.

tin- u|>p<-r class,
we prove
-
less
IRRATIONAL NUMBERS.
366
[AP.
Ex.
1.
Prove similarly that
Ex.
2.
Establish inequalities similar to those of Ex.
Ex.
3.
The formula, corresponding
Ex.
4.
Utilise the last
if
then l>k,
F</V,
<N,
I.
where
l=(Na + bk)l(b + ak) and b*> No*.
two may be taken
The
as
1,
2,
taking
for the nth root of
example to find approximations to 2*
the
N
first
f.
of rational
classification
to Ex.
1,
numbers which has been just
described can, however, be obtained by a different process.

the arithmetical process of extracting the square-root of
From
2, it is
evident that
2
(1-4)
(1-41)*,
(1-414)
2
,
(1-4142)
,...
but the sequence contains numbers which

are as close to 2 as we please. Thus the lower class contains
are
all less
than 2
every positive rational number whose square is less than 2 and

also contains all negative rational numbers.
Since the two
;
it
classes together
the upper class
contain all rational numbers, it follows that

must contain every positive rational number
whose square is greater than 2.

Thus the same classification is made by putting,
(A) In the lower class, all negative numbers and all positive
numbers whose square is less than 2.
In
the upper class, all positive numbers whose square
(B)
is
greater than
2.
145. Dedekind's definition of irrational numbers.

Suppose that some rule has been chosen which separates all
rational numbers into two classes, such that any number in
the upper class is greater than every number in the lower class.
Thus, if a number k belongs to the upper class, so also does
every rational number greater than /,.
There are then three mutually exclusive possibilities
(1) There may be ;i number (/ in the lower class which is
greater than every other number in that class.
(2) There may be a number I in the upper class which is
less than every other number in that class.
(3) Neither (j nor I may exist.
:
hl.M.KINh- M.I IMTi>
144, 145]
'I'll.-
i>\
nai niiiui.>
repreeenl
/'
represent thunit of !!
r.
upper class
on
tin-
lint'
and
lower class consists of v and
tin-
Similarly, \v- .MM illustrate case

thai
i/
rational points to the right of
.11
c<
and /m:
>uslv; l.ut this
b
that
falls
-ial
is
.-.vim
/;
to
and
this
the
in
-\aiujil' îvt-ii
last
<j
nor/ can
wln-n-
articl.-.
th<T' could lc no greftiesi nunilx-r in

no least nuniluT in the u]]H-r class.
that
\vlii.-h
i.-d i,\-
at tiist
sight
the hypothesis
be classed. Now $(g~

would .-...ajM- .'la ilirat imi.
TliMt thri-c an- cast's in whic-li neither
lh-
rational p-.inta to the
nuiiil"
betWi
1'nnii
all
mi^ht be thouglit
It
(2).
the
tin-
>nal
and
.'\i
\va- prov.-.l
it
lo\\vr class,
\ani|lr, h-t us illustrate mi a straiirlit liiu- tin- ajijro\iniati-.iidt-ri\'d from th;

te to the rontintifd fraction
and
ar.-
onvergenta of
tin-
+ _L
^2+
LJ_ "
'1+ -2+
lower class are Bf

'
1-
while
an- tho>- of
tin-
uiM-r
'
It
niav
..... l.srrvyil
that
if
rant
of
.-itlu-r
__ _
con\rru ciit of the same

r
i-
clttM,
the-
next
'
',
\vliil-
>>
the intermedia!.- OOnVBTgei
the other class
hewn
Tlien
ii_:uie
falls
in
tl
Je reproduction of the segment of
!..-
l.et\\.
_J
ti.
L'
II
HI
It
J_ttl
nnu
ti
!ii.-h
IKRATIONAL NUMBERS.
368
[AP.
I.
It is clear that in case (3) the rule gives a cleavage or section

numbers and to fill up the gap so caused in our
in the rational
number-system, we
agree to regard every such section as

number. This constitutes Dedekinds definition
defining a new
of irrational numbers*
On
so no
For
it
is
from what has been
clear
new numbers cannot be
said that these
rational.
the other hand, in cases (1), (2) there

new number is introduced.
146. Definitions of equal, greater, less
is
no
section,
and
deductions.
For the present we use the following notation

An irrational number is denoted by a Greek letter, such as
the numbers of the corresponding lower class by small
ft
:
a,
as
italics,
A,
a,
The
B.
those of the upper class
classes
themselves
by
capital italics, as
be denoted
may
by adding
brackets, as (a), (A).

These definitions may be indicated graphically thus
i
a
It is
<, >,
an obvious extension of the ordinary use of the symbols

to write
a<a<A, b<@<R
In particular, we say that a is positive wT hen

belongs to
(a)', a is negative when
belongs to (A).
Two irrational numbers are equal, if their classes are the
same; in symbols we write a j3 if (a) = (6) and (A) = (B).

The reader who is acquainted with Euclid's theory of ratio will recognise
that this definition of equality is exactly the same as that which he adopts
Euclid in fact says that A:B=C:D, provided that
his Elements.
in
the inequalities mA^-nB are accompanied by mC^-nD, for any values of

m, n whatever. In Dedekind's theory, the inequality mA > nB implies that
thus Euclid's definition implies
njm is in the lower class defining A B
that A B and C D have the same lower class and the same upper class.
:
the other hand, the number a is less than the number /3,
when part of the upper class (A) belongs to the lower class
{&), so that at least one rational number r belongs both to (A)
On
and
to
(b).
This definition of inequality also coincides with Euclid's.

*
Other definitions have been framed by Meray, Weierstrass and G. Cantor.
M.l'i
145, 146, 147
follou
It
<
[QNa
im the definition thai
">f
<'<i<ft,
and
.1
sii'-li
</3and $<y,
it
< y.
then u
tli;it
/i
</<,*.
we ran
that
hrloi,
and
there
NOTR
DO
i-
li-.nal
than
iste
numher
an
find
{lumber.*
one rational num:
I'-.-i^t
nuiiiln-i
tin-
in
>
Thus
/'.
a<r<<
Thru
that
any two
iv
if
thr-r rational nunihrrs
all
intr^ri-s.
j.i.sitivr
w-
hi

is
Thru consider thr
to
<
|ji.siti\r
lower rla*s (a) and
tilt'
>'
,-hi^T,
that
//
,
cla-N
u]>}M-i-
solllr illtr-vi'
fl
inning this j.rorrw.
\\
to class
dfl
30
ari'i\r
t'racti-
"o+iV "M+
t
in
!>'
/',,+
rational
......
tin
llr^ltixr) sllch
U'
MIJ
i.
that
\\v
the result
at
"
-
If
we
rail
these
i'racti<.n
merely
is
two decimal tVacti.n- " .1
plain that
made [688 than any proi-rihed rational
.
l.y
taking
We (-..IliinUe the p
numher detined l.y the
The ariji.ment
l.e
sutlicieiitly
\\
indefinitely
see
i;
great;
that
((
a:
is
the
infinite decim;il
JflO
fix
t,,
that
t'
f/,,,f
A
\\
i^
u-et'ul
;tll\
i..
not,-
further that
pl-rviTiled ilUmher
r/'
<
>f
can
l>e
the loWel"
chosen so as to
el,:
IRRATIONAL NUMBERS.
370
[AP.
a" be another number of the lower class which

a''; and then choose r so that 10 r
a').
l/(a"
But 4 r >a",
ar
<a"
so that a r
a',
>
ar
or
a'> A r
greater than
is
>
Ar
I.
Then
a".
a'.
Modified form of Dedekind's definition.
Suppose that a
classification of the rational
following properties
a belongs to the lower

number less than a;
(1) if
(2) if
numbers has the
belongs to the upper
class
so does every rational
class, so
does every rational
number greater than A

every number a is less than any number A
numbers A, a can be found in the two classes such
:
(3)
(4)
than an arbitrary rational fraction.

classification defines a single number, rational or
that
Such a
is less
irrational.
For any rational number r which does not belong to either

must lie between the two classes, since any number
less than a number of the lower class must also
belong to
the lower class
and therefore r must exceed every number
of the lower class similarly, r must be less than every number
of the upper class.
Hence, if a, A are any two numbers of
r
A
the two classes a
class
< <
Suppose now
a second rational number which belongs

to neither class; then a<^s<Â.
Hence \r s\ must be less
than A a; but this is impossible, since by hypothesis A, a can
be chosen so that A a is less than any assigned rational fraction.
that
s is
Consequently, not more than one rational number can escape

if there is one such number, the classification
may
classification
be regarded as defining that number; but
number which
section,
escapes classification,
is
no rational
we have obtained
a Dedekind
and have therefore defined an
if
there
irrational
number.
Algebraic operations with irrational numbers.

The negative of an irrational number a is defined by means
of the lower class
A and the upper class a; and it is denoted
-a.
by
148.
Tin reciprocal of
1
easily
by
an
irrational
restricting the classes at
number
first
is
defined most
to contain only terms
ami
.1
'
ive,
he
-i\'-n
li.-n
>-,il
class
upp-r
compl.-t.-
Thus
ion
who),-
framed
i-
l'ir
when
anl
ive
i^
t>
i-.jual
(l.-nt-l
is
it
class,
188 will
nurru
as
<.r
by
is
tl
al
lower
t!
only
i\-
will
th-
in
by putting
with all negative numbers, while
definition
.1
number
th--
if
th.-
,,f
of
by the lower
L/a
ft.
s|MM-ili,-;it
tin-
ih.-r
contain
In-
tli.-
OPEB kTIO
\n
\l.<;i:i;i:
H8|
147,
always
i^
ti
],..siiiv-
or
|o|.
'/X.
B|
ii
< <
< <
o'
the up],
Art.
To prove
17.
l.'tii
II
/>.
and then
BO &8 to
+/>')
.1
thr
at
represented ly
That
and
(lie
is.
MUM
tin-
DO thai
i'ollo\
It
is
.il.-d
tin-
the only rational

numlier defined l.y
jual
't
l-'or
i/
(4)
ami 8O
/;_
Art.
A'
//
II.
t.
<-ach
Mice
can tind
\\v
147.
than
l.-ss
our
rational or irrational
I..-
.',e:
classifica:
this
a)=0; for here the lower dan

J.and th.- upper das- i,y A
o+(
typ.-
'
<t.
"fall negative rational numbers

positiv rational numl.ers; he;
all
t. tlie suni
not
and
class,-.!,
tln-r-
iticatioii.
virtue of the relation
,ndit i,,us
<t+/3.
numher
is
In
than
may
<-,
lo\v-i- rlass oonsistfl
upper class of
tin-
and
i/
Lefifl
tion dctin.-s a iinnilxT \\hich
nunil
u i_ (M _ H(
mak- J
is
cl.i
that
iM'înnin^ of
-f/o
(r/
,..
that
nuin
ronditinn
s.-u i.sii.-s
not.-
lo\v.-r
iou.sly satisti.-s
it
w-
14. />>>_(, /
ami, as .'XplaiiH-d
and
that
nuiiil.-r.
///'
nl-\
nil.-
rational
tin-
/>'.
so
irrationals,
given
Tlh-n dassily
a
IIH-HI!MT
<f tin+/*
typical
b
.1,
making
o!'
two
the
be
fco
,-i
,+(
,'*
/8)
= 0, w- may
detine ,<->'
ft).
>!><</
simplicity of statement,
We
<>mit
the
numl..-rs
from the lou.-r olasses; and then \v,- detin.- the jro(lu'.-t a/3 by
usin- the tyj>e ,if, for tin- lower elass and A Ii for the &
spondine- upper claSS,
let
t-
To prove
that this
.1.
-tines
-.in-l,.
denote an arbitrary positive rational fraction
l.-sx
nun
tha
IRRATIONAL NUMBERS.
372
[AP.
1.
then choose any rational number P which is greater than a+/3+l.

Next find numbers A, a and B, b such that A a<^6 l ,B b<i l
where
el
The determination
Art. 147.
Now we
= e/P.
of A, a and B, b
have
}
is
possible in virtue of
or
That
AB-ab<6.
is,
Thus the classification by means of ab and AB defines a

single number which may be rational or irrational and this
number is called a/3.
;
In particular, if /3=l/a, the product is equal to 1 for the

lower class is represented by a/ A and the upper class by A/a.That is, the lower class contains all rational numbers less than 1
;
and the upper
class all rational
sequently the product
which escapes
is
numbers greater than
equal to
1.
the single rational
1,
Con-
number
classification.
Multiplication of negative irrational numbers is reduced

at once to that of positive numbers by agreeing to accept the
"rule of signs" as established for rational numbers.
Division.
In consequence of the relation (l//3)x/3 = l, we may define

the quotient a//3 as equal to. the product ax(l//3).
It is at once evident that any of the fundamental laws of
algebra which have been established for rational numbers
remain true for irrational numbers.

Thus, we have the following laws
For example,
By
definition
let
us prove the theorem a + )8 =
we have
and
But a + b = b + a and A
+n = fl+Aj
so that a + j3
the same two classes and are accordingly
r<|ii;il.
and
/2
+a
are dHinrd ly
KONOION*
148, 149]
NUM
t"
in a îmii
intial iiili<--s
and
of |M-^;H
if
Inn-,, ..n tli-
t.
th.-ii
li.-
i-itli.
149. The principle of convergence for monotonic sequences

whose terms may be either rational or irrational.
A monotonic sequenc
'
-main
tin-
HUM
an
i^>
nni
...<A.
term
eitJ
occur;
in
tin-
two
sequence
ry term
alt
will
(a,
IK-
than
\\\> d.-linc tin-
/,-.
\vliich
the
Beqv
it'T
than
rl<-ar
i^
It
.1.
are
\..\\
ujijier
tin-
rational
all
class
,mliti.n.
<-.
Aether
/-
i-a
\V.-
may
<-.
.ntain
which
call
/,'
ra'
///
dctin-
t'othc
\\]
t<>
rational nnndn-i
:id
than any term
any
-',
:
:
.i.
,.
and these
also
th'.
<
tequentlj
rcecd
the other liand. every rational nuinhcr lc>s than

helon-- to the lower
Now if t i- any
<
)n
tiiei-e
will
l.e
rational iiuinl-
than
these nunil"
i.
That
the
there will he rational nuinl>er-
than
!>'
And
ter than any term .'
nnmlter
in-atimal
y îvater than ft
ry
tru-
i--
as
tt)
Ty]ical
which
ju-ctivdy. defined ly the se.jiie;

i"ital numlter -Teater than >1 I'clon-'s
oil
same
ter
tin*
rational nunilM-rs
th-r-
clfl
every
nf
<-la>s
tin-
,iiliti<n.
96COnd
lln-
tin-
nuinl.cr ^. rational or irratinnal.
and lower
-.
\\liilc (II) contains

.-vry i-ational nuinh.-r
and possiMy sonic rational nuinhrrs ].-ss tlian J.
that
and
afl
tir-t
In-
>atis|'y
the C
!'
HUIII|MT>.
sati>,t'y
\\hirh
all
<>t'
t-ijual
will lê
!'
//
th--
tl.
lixed
numl.er, one of
Ol
[066
than
le8G
;lt
i.-il
to
the ^fiju-ncf
that
defiuiteneSfl
Suppose I'min which

;
in
-n
niimi
iiinnal
i>,
f/,*
.-},
they nr.
e.
o'
ft.
inu^t
IRRATIONAL NUMBERS.
374
@ = a n = /3
we have
That
I.
^ 7i>m,
if n > m.
a n = am,
Hence, since
[AP.
e,
lim a n = ft.
is,
A good example of
such a sequence is afforded by the terminated decimals

derived from an infinite decimal
and it will be seen at once that the
section described here is an obvious extension of the method used in
;
Art. 144 above.
Suppose next that the terms of the sequence, while still

It
increasing, do not remain less than any fixed number A.
is then evident that if a m ^> A, we have a n
A if 7i>m.
>
lim a n = oo
Thus
n >oo
Exactly similar arguments can be applied to a decreasing

sequence.
As an example we shall give a proof of the theorem that any continuous

monotonic function attains just once every value between its greatest and least
values.
Suppose that /(#) steadily increases from x=a to x=c, so that
b<d,
Then
is
also
if
f(d) = b and/(c) = rf.
if I is any number between b and

between b and d suppose that
we
d,
consider /{^(
found to be
this is
+ e)},
less
which
than
write then
= ( +
= c,
Cl
<*!
On
the other hand,
<!),
b 1 =f(a l )<l,
d =f(c )>l.
l
when /{J(a + c)}
greater than
is
I,
we
write
d =f(c )>l.
1
Continuing the process we construct two sequences (), (c,,), the first
M
never decreasing and the second never increasing; and c -a n = (c-a)/2
Also by the method of conso that (an ), (CM ) have a common limit k.
unless it happens that at
struction it is evident that f(an )<l<f(cn )
some stage we find /() = I, in which case the theorem requires no
;l
further discussion.
Now
since f{x)
f(c n )-f(a n
and /(CM ).
)<,
if
is
continuous
n>v,
Thus we can
we can
it is
in Art.
clear that there
is
that
f(
,)
find v so that
\f(k)-l\<
and therefore, as
an integer v such
are contained between
find
and both f(k) and
if
n> v
=
f(k) l. From the method of construction
and this is also evident
only one value such as /
1
(6),
from the monotonic nature of
#.
LIMITS
149, 150]
Maximum and minimum
150.
limiting values of a sequence
of rational or irrational terms.

all
tli.it
than SOULinn
..mil
tlian
is
it
icfa
'I'll. -ii
p,
we
i!'
iiiiinl
that
an
an
smaller
p be a
l.-t
intiiiity <f
intinit
beta
/'
id
'
the
in
M.
ater
tennfl
evideni
lall
are less
tli.-
il
y or
a finite
number
r
tl,
iii
case
we
\N
latr-
\Y<
bains
an infinity "f term
gee are
the
.-at
in tin-
diagram.
PJ
Then
the
sequence />. /',.

than /': and so
lees
'/
ID
limit
i!'
i-
article
la>t
tin-
a.
which may
ff,
sequence
Thru the
-'
irratinnal,
the
increasea
numli.T.
]Miti\e
sequ<
* n
p*(Pp)
/'
t'-nniiK'S
because
an arhitrarily small
Conse.|uent ly am
anl
1-
COUTSC K- cithfi rational
<!'
i.
/'
never
/>....
/>.,.
th.-
we can
>aiin'!'
tin.l
//
G'-fe.
vn
"i"
Bub-aeqtience 1'p'iii (" ..) which has G

a certain
stage after which all
are
less
I'f + e:
than
thus no
>e,juenn.
Thus we can .letennine

ami we can
its limit
:
the
bermfl
the
<>f
find
s/H-tr
have
can
sab-sequence
convergent
tlmt
;>
Inn
thai.
ber
//
A qut
It'
DO
such
ar-'uinent.
than
at
flu a
JO.
.V.
afl
/'
On
the other hand,

\\ill
can
IK-
only
be
80
if
l'"Un<l
the
of
;iiml)er8
naMe numher.
found, there
than
numl.ei-
th.-r.
sequence
ater
('
no such numl>er as
rm^
however
lim "
in
consequeni
can be
_^r.
IRRATIONAL NUMBERS.
376
[AP.
I.
For the sake of uniformity we may say even then that the
maximum limit, which is, of course, oo
sequence has a
foregoing discussion can be at once modified

cc ).
establish the existence of a minimum limit (g or
All the
151. The general principle of convergence stated in Art. 3
to
both
is
the terms being rational or irrational.

In the first place, the condition is obviously necessary for if
lim an = I, we know that an index in can be found to correspond
necessary and sufficient
n >oo
to
e,
in such a
way
that
a n < Je, if n^m.

-am |<e, if n>m.
-a
l-a
n \+
m |=
|an
l
Thus
In the second place, the condition
is
sufficient
for let
be
fixed so that
or
Then
has a
it
finite
between
index
is
Je,
if
follows from the last article that the sequence (a n )

maximum limit G so that an infinity of terms fall
;
and
Choose one of these, say ap whose

than
m.
Thus we have
greater
Je
ap
Also
Thus
and since
it
<
n > m,
am -Je<a n <am + Je, if n > m.
a n -a m
6r-fje.
Je
<
<x
< ap +
- f e < aw <
Je,
+f
a m \e<â n <
+ Je,
G e < an <C (r +
a,,,
follows that
since
>
>
w..
e,
e,
if ?i
> m,
if TI
>>
7?i.
Thus a n -+G and consequently the sequence is convergent.

Of course in this case g G, the extreme limits being equal
;
in a convergent sequence.
Various proofs of this general theorem have been published, sonic icing
much shorter than the foregoing series of articles. But on
1
apparently
of the shorter investigations it will be seen that.

in all cases the apparent brevity is obtained by avoiding the definition of an
This virtually implies a shirking of the whole difficulty
irrational number.
for this difficulty consists essentially* in proving that (under the condition
examining the foundations
of Art. 3) a sequence
may
be used to define a "number."
I. A. 3, 14) says:
"As the truth of this theorem
and exclusively on an exact definition of imitituml inn.',
naturally the first accurate proofs are connected with the arithmetical theories of
irrational numbers, and with the associated revision and improvement of the
Pringsheiin (Encyklopiidie,
rests essentially
older yeometrical views."
150, 151, 152
LIMIT-
MI MIII
"I
Theorems on limits of quotients.

and lii
// Mm",,
152.
<>
I.
steadily decreases,
,.
Inn
<',,
tirM
SUJ.J.M-.-
th>-
,.
--lim
second
."
</
that tin- limit
ami
linit.-
i>
-'jUal
arl.itrarily small j.o.xim,. Eractl
:i
i-
AD
</+*
---
to /: th.-n
foiiinl
so
if
/t
inoe
(''
'',.
i'-e><^-^..,<'',. -''-,<<' + *
Chaii^tllfll
\Vr
to
/'
/'
!.
/*
+'2. .../'+/'
takf the limit of this
()
Hi-nci-.
sine
or
w-
if
tin*
how-\,-i-
lar^r
re
/>
..
~ s.
w-
p ).
ol.tain
h.-r.
if
e,
'')
limit
>
before,
tin-
ly hypoth-
liini",,
the oth.-r hainl,
as
n-.Niilt
positive,
-/
On
ali
<(/ + f
(/-cHl*,,-!*, .,,)<",<-".,,
N>w
ni'l
tillil
//
if
.V.
..
/.
i^
>
\\-
can timl
>
so that
ctly the same aixr um-nt as
!i
timl
..+*).
which
1.
or
A'.
Thus
linn,/,,
is
in*
ditliciilty in
D
"11
if
AJ-
/>
>
m,
'\ti'inliii.
that, \\ith tht-
llir
iii
t>n
'
;"-""-'.
IRRATIONAL NUMBERS.
378
[AP.
I.
This theorem should be compared with the theorem (L'Hospital's) of the Differential Calculus, that:
lim0(X) = 0,
//
limbec) = 0,
lim 0(aj)/^r(aj)=lim
then
X>x>
<f>'(x)/\{s'(x),
X->oo
provided that the second limit exists and that \^'(x) has a
constant sign for values of x greater than some, fixed value.
II. If b n steadily increases to oo
then
,
n+l~n
provided that the second limit exists*

For if the second limit is finite and equal to I, as in I. above,
we see by a similar argument that it is possible to choose
so that
bm )
(l-e)(b n
Thus, since b n
Now,
< an - am <
is
since b n -+ x
and
lim
(I
+ e) (b n - b m
if
),
n>m.
positive,
we have
(i
+ e)
= +
e.
n^-
And
we
so
find
But these extreme limits are independent of m, and therefore

and so the inequalities can only be true if each of
also of e
the extreme limits is equal to I. Hence
;
lim (a n /b n) =
Similarly,
if
the limit
a n -a m
however great
Thus
*
is
oo
>N(b n -b m
may
1.
we can
),
if
find
m, so that
n > m,
be.
Extended by Stolz from a theorem given by Cauchy
for the case b n =n.
LIMITS
152
1
an.l
limit
tin-
Jim
liat
and
of
t!
on
lh-
is
lli.
-I---'
Thus
Inn
imilarU
'
| *
Cf M
with
tli:it
"~~
'/
'
IL should be compared with
IVIM
LI
>ill'T.
-Altai's) of tinincreases steadily
I'
tin-
(.It.
-o w
-o*
N, OS
nn\v this miiiiiiniiii limi?
'
I'nllov
tin-
<>rem
f<>
s_
tt-ifl,
/A
./,
'^1:
/,</
It
tin
it
d,,>
n-t
|tr<'l>,illv
second
..ut
i.f
//'///
nl mi
|il,
tliis
iiiijxirtant
theorem,
>of .*
he extemU'tl
and
>>(
')-'>(,,)
a so
-lin,.se
f"i-in
is
Ban-value
j.ositiv.-.
tln-'-i-.-ni
Tin:
\\-.-
adfl
tl
')
Ami
fr'lii
Similarly,
hn
.....
\vai(U
tin-
;n -iiniriit
])i-
that
if
>()],
</>(a-)-</>(a)
Ex.1.
,-r+...+
If
^=
(^+ i ) p
+i_ n p+i-
A
iiin.laiiirntal
H.'
ami 80
limit
..f
tl
ilus.
JL,
lin.
liiu"
6.
have
to a limit
IRRATIONAL NUMBERS.
380
Ex.
If
2.
a n = lo7?
n + i-
we have
so that
lim
if
Similarly,
a n = (logw) 2
^
which tends to
(Compare Art.
b n = n, we
=0.
by the previous
we can prove
Similarly
that
= Q.
lim [(log #)*/] =0.
a n =p n
we have
bn
by using L'Hospital's theorem.
= n,
-a n =p (p~l\ & JI+1 -6,, = 1,

-a w ) = 0, if/>ll,
= x if p > 1
or
lim(p'-0, ifjo^l,
n
a n+l
and hence
lini(a n+1
Thus
or
Of course Ex. 3
Ex.
not
is
Even when
4.
In this
exist.
An
exist.
ifjo>l,
and (6 W ) are both monotonic, lim n /6 n need not

theorem shews that lim (a n+l - a n )/(bn+l - b n ) does
is
given by
i-a=/ [^+p(-l) ll+1 -g'-( -1)"]

fl
and so
steadily increases
Then we have
if
^.iiz.
o
n+ i-o w
q>(p + l)'/(p-
1).
=
j-l.
-b
bn+1
jo-1
lim(aw /& w ) = 2-
Since
5.
find
o w +i-a n = 3 + 2(-l) ll+1
(a /l+ i-a, )/(6 H+ i-6

1
if
Theorem
II.
ow
If
see that
Again,
we
p-l
1,
these results agree with the extended form of
Thus
lim (/&) = ?+ 1.
+
(jo
l)/(p-l)>l,
while
Ex.
(Art. 161).
2.
(a n )
example
o
=ao,
only another form of Ex.
case, the
Here
we
+ !),
result.
also verify this result
If
3.
< | log (
+1)
\irn[(\ogn)*/n]
The reader may
161.)
find
= [log n + log ( + 1)] log

(l
Thus
Ex.
I.
=n
bn+ i-b n
[AP.
/(
oscillates
between
and
5,
although
152, 153
!--
\l-.l
1. 1.
MM
\.
l/2(-f 1); and 00

I,,,,;
tt-s
mill,:-
a drlimtr limit.
Ex.
6.
!">
It'
steadily in.n-aae, the Theon
ii. .t
II
neces-
.saiilx
Tim*
aw =
&.-<
ii
ntlv
luit
.-
In
same M
lie
.inn
may
= 1.
\\ln-n \fr'(x)
fail
chang*
intiiiitrU
'riiu>
"ii-id.-i
/+! +>
^('
for
\v,-
\\hi.-li
and
Ex.
and
tin.
-(),
\vliii-
Ms.-illat.-s
bet
;i>
7.
tin-
'"ii-i<tfi-
<
Mm
|>i"\f that
= l,
Mm.
wl.iK-
153.
An
.,-)=!+<,.
extension of Abel
Lemma.
'
y,,
,/, /,
ran '*"
'
.\
,,
rn
/,
\\i'itf
.1,,
A =
'',,,
ami
/'
/'
.
,
/>',
'',,
!> {t
r -{-,.,
-f- //
''.,4-
'V
~^~
.-1
''j
/>'
-}-/',.
-\-
((
Ii
,
"
~^~
'
= '/,,-f-Cj
= + +
-f~
~f
4-
/'
'>!
'
Then
\
^(Vo-^) 4-4
/*,(
'-,
'",
4-
-...
'>-',)-
-f />,
.i(v n .i~
+ ^^
%
1
('
H-i-v n )+^
re all
//'
]n.>iti\ 6,
We
'MII
//
and
//.I
nl.tain
in
.1
in
;in
n]|n-r
lin,
jla--
..f
/;
0,
IRRATIONAL NUMBERS.
382
Hm
where H,
are the upper limits of
Bn
(B
l-2-> -j-,
\-L
Îj
l
vn ) and f
(f ri-i
\-A-m
are
Bn\
"'
...,
i,
-"-?n+i
This follows because the factors
respectively.
.
orf
/Bm Bin+l
ana
^
->, -pJi
+ n/
I.
[AP.
V^
(VQ
v 2 ),
(t^
all positive.
Thus
we
we
replace J. r by its value in terms of

obtain,* since
m by definition,
and,
if
H^H
+a +
1
...
+a r>
"
/
i^ 'i
we prove
In like manner
~T
that
h m are the corresponding lower limits of Br/A r

Secondly, suppose that the sequence (v n ) steadily increases.
In the numerator of
v x ), (v x f 2 ),
n the factors (V Q
where
A,
( v n-i~ vn) are a ^ negative, while v n is positive, so that the

value of the numerator is increased by writing
and
/i
hA r
in place of
w -4r
in place of
while in the last term
Br
Br
= 0, 1, ... m 1)
= m,m + l, ...,n
(r
(r
we must put
Hm A n
in place
l),
of
Bn
This changes the numerator to
= k m (a
+ .+a n v n + (Hm - h m )A n v n
+ (h - h)(A Vm - " r - - am vm
tn
and, since h m
the
is positive,
is
less
(Hw - h m } A n vn 4- (h m - h) A
in
actually
av
is
),
this again will not be decreased by omitting

terms in the last bracket.
Hence, since the
*The numerator
which
^h,
negative
denominator
llt
+' 0,1?!
+ ... + a m r m - (a
than the value given
-f-
^+
..
+ a m )v
v in
Till
153,154)
w-
Similarly,
1.
Mil-
timl
n _
II,,,
1'
m )A n vn
+(H-H m )A Mvm
.(
'/'A*
xuppn.sr that the trim Vf
We may
sequence, and
ilir
than
is
tli-
/-.
-iv a t.-t in
the
ll.-i-,-
K-tors
while
vw are
'')
greater
/'| .!.,(
Thu^
Xn
numerator of
iln-
is
tlian
'-',)+...-!-.
,)+...
'
+/',,,M.
+ //
= y/
+4i
^4)+
-...+"
-...-"
/'
we deduce,
Hriir-
<
V
Similarly,
we
x.i,,.'-,,.-'/.
an ar--um-nt similar to
l-y
Ex.
find
...
154.
Other theorems on
follows at onre
->
Thus
ha-
v l +...+a n v n
/
divi-r.Lr-
be
-. ....
provided that
It
iha;
th<-
//
av +a
l'i-..v
'-...-
writing
as
ni
8m
if,
=a
+ 6,,
[.Ins-
limits.
Theoivm
limit.
II.
of
+Oj+...+a
fh>-
6
/
Art.
to
fcendfl
that
1">2.
the same
that
\f a
if
limit.
fi>
lim
(r/
-|-
/.,+ ...-!*"
I' 1
t!
troul)K' to i-xjiiniiu-
;ivj,
it
lirr-.
monotonic
l\it it
saves
IRRATIONAL NUMBERS.
384
Of course the second

with
..._!
= 1,
a-j,,
= 0,
limit
may
[AP.
when the
exist,
the second limit becomes
does not
first
I.
thus
although the sequence
{a n ) oscillates.
THEOREM
III.
If all the terms of a sequence (an ) are positive,
lim
if
(an+l /a n ) is definite, so also is lim an ; and the two
limits are the same.
[CAUCHY.]
and
lim (a?l+1 /a w )
if
For,
and equal
finite
is
to
(not zero),
we
can write a n = c n l n so that

,
lim(c n+1 /c n )=l,
Thus we can
find
< cn+1 n <
By
m, so that
/c
multiplication,
we
Hence
<
<c m (l + e)
cn
m (l-e)<
Now
<c, n
cw
=l
limcm
1
lim c n
(l
tt
.
+ e).
(see Ex. 4, p. 17),
n >oo
so that
n ^ m.
if
e,
obtain
c w (l-e) w
so that
^ lim cn ^ 1 +
e.
Now
these extreme limits are independent of

and therefore
e is arbitrarily small, so that the inequalities can
only be true if each of these limits is equal to 1, or if
of e;
and
lim<v = l.
Thus
But
lim
if
lim (a n+l /an ) =

<-i>
however great
= I = lim (an+l /an
n
oo
we can
n +il<*>n>N
may
a> n (f
minimum
is
limit
so that
*
is
w /JT]i.
=1
lim a n
so that
and
>l
lim K,/JV']
This
in>
ai>^[
But
n^m,
if
).
m,
be.
Hence, as above,
or
find
Lim On*
p.
17),
J\r.
independent of
arbitrarily great, so that
(Ex. 4,
m and therefore
wr must have
a oo.
oi
154]
Mm
h.-n
LIMIT-
HEOREMfi
tli.-
can be reduced to
|,rt-\
:n _"!in.-Mt
ioni
to
th.
proTi
r:il
lim
.ilnl
');
/.,".
alsi.
',+iKX-,
;
tod
'
it
Ex.
reaaes to
1.
lit.:
ft
n*,
,:,t
'
Thus
lit.,
result can
fonnuhi
f(.|'
To
Ex.2.
ami
(SIT
rig's
\.
HIM
liiul
[(//<
l)(// <
+ 2)...(
!*
tixed,
tin,,
/'+
tin-
i.-u
Ex.
3.
limit
ajruin
Similarly
\\t-
is
find that
l
HIM
[(/<
-<,
I\".
O.
]
90
Thm
tin-
-iv.'ii
ezpreaaion tuk.^
"^^'/i + .-.-f
Now, My
li"
the r.-ult at
Pi+
ill.-
tin- iM-îniiinir
that
p n ->0,
'/
f..nn
-f...
"'
+r
'lii^ artid.-. \\,- SIM-
/,-fV,, -f ...-I-.
that
IRRATIONAL NUMBERS.
386
and consequently we have only
prove that
to
lim -(Piq+piq*-i +
tv
Now,
which
JUL
since
pn
q n both tend to
and
greater than \p r
is
I.
[AP.
+Mi) =
0,
we can
\q r \;
and
find a constant
further,
we can
find
so that
\p r
Then,
if
>
2/z,
we have
\p r qn-r + i\<A\q n - r+l
and
\p rq n - r +i
ifr>//.
qr\<e/A,
\<e/A,
\<e,
n-r^$n> p,
if
r^$n,
if
r^$n >
\<A\pr \<e,
p..
Consequently
-|(Mn+Mn-l+-..+Ml)l< e
if
>
>2/*,
IV
and so tends to
n tends
zero, as
to
oo
thus the theorem
is
established.
155.
THEOREM
If 26 n
V.
2c
ri
are two divergent series of
positive terms, then
_
- 11111
i
Co + ^4-.. .+C
provided that the second limit
exists,
and
that either (1) c n /b n
the
steadily decreases, or (2) c n /6n steadily increases subject to
condition*
is fixed.
where
Let us write for brevity
Let us also write
c n /bn
= vn
',
then
gn = (^QO)Ô +
Cn
The second part
1907, p. 269;
(2),
that
t.
13,
the
1889,
p.
c n /b n increases.
61.
+...+b n v n
theorem is due to Hardy, Quarterly Journal, vol. 38,

given by Cesaro, Bulletin des Sciences Matlnmatiques
Of course A'>1 in case (2), in virtue of the t';u-t
of the
first is
b v
154, 155
we can apply
and
-(hm
whenf
//,
lini
and
si
an-
/,
:md
It'
the
in
tind.
ire
(1)
to
BEOREM8 OH LDA
//
-h&<$<Hm +(H-B
iipp.-r
&,-, We
I',,
and lower limit-.
are
'
.
Pm/Bm
!'
can tind
1'n -in
^n
BO that
01
.1
x
/
A,,,
<:
II
+ e.
>
<
i{
'
tin-
we have
An
tin-
CM6, when
tie-'
(/-,)-,/-/,,'/;<;;<(/+.)+(//-/.
BO
tak-
u.-
it'
th-
last
linii'
tind
\v.
in-.|uality,
tli.-it
/- t
I!*
in
-ii
In likr
""'
!)
.Mjual
manner,
>
l n
it'
"""
'
'.'
*'
Sine.-
"
-s_
\v-
t<
/-he.
I...,;!"
;|
each
arliitrarily small,
HUM
inn/;
^n
<t'
tin- .-xt r-ni<-
limits
inn(^C'n ) = /.
x \v<> can
H,,
,
tind
so that
\'.
>.V-(.V-A
tind
the last in-i|uality tliat
i'ruin
\in(QJCn )^
and
<
linKy,,
In tin- s-o>ml case, wln-n c n
<^<*.+
li\-
A"
hvpoth
is
we have
( fi
< A'
',"
/.
0,,
\vhnv
in-retises,
-,//-/,
//,-,//
N<.\v.
!>
QODSU
a
//
A ',//
fcei
-/,
than unity.
II.-nc.-
< /;'</',-f A'i//

then,
jn.st
as c2)
is
d.-dur,-d
/-C-Vv'-lk
I'rniii
which the same
result
fmn,
in,,/;'
i'.
.ll-.vv-s
,.
we
-A
tind
/-frJA'as
A'
from
-/-^
<
II
>
that
IRRATIONAL NUMBERS.
388
[AP.
I.
in the first case the series 2c n

diverges more slowly than 26 W while in the second case 2c M
diverges more rapidly than S6n but the final condition excludes
It is instructive to note that
which diverge
series
too fast.
It should be noticed that if sn tends to a definite limit, Theorem V. is an

immediate corollary from Theorem II. of Art. 152 for then both fractions
have the same limit as sn
;
The applications
most interest
of
'
arise
when
= \=...=b n = l,
and then we have the result:

// 2c n is a divergent series of positive terms, then
Co
n+ 1
+ Ci+.-.+c*
second limit exists
and
either (1) that c n

c
or
that
increases,
decreases,
n steadily
subject to
(2)
steadily
the restriction
provided that
the
nc n
where
is
<K(c + c
+...+c n ),
a fixed number.
Ex. 1. A specially interesting application

theorem of Frobenius (Art. 51) to the series
form an increasing
c
c lt c2
...
satisfying the condition just given.
where
Here
it is
from applying the
sequence of positive integers,
arises
evident that the series must be written in the form
= ao>
-4q>-i
ô = o> î = o>
A =a -}-a A +i = aQ + a ..., and so on.
A v =a + a + ...-}- ant
Generally we have
c + c + ... + c, _i^v<c + c + ...+c n
we have
Thus, if 8n = aQ +a + ...+a
A + A! + ...+A V =s co+ 5 c + + *n-icw_i + s(v - c - c - ...-c,,_
so that
>
Co
Co
if
1l1
--
and therefore Frobenius's mean,
is
if it exists,
given by
which we have proved to be the same as

provided that the last limit
Ex.
2.
Interesting special cases of Ex.

6-0
for
which
exists.
+ 0! +
..
may be taken
c,,
as
= (w + l)
2,
are given by taking
;2
(w
+ 1) 3
3 respectively.
etc.,
155
THEOBEMfi "\ LIMITS
Th us we have the
re.--.
an-
Ex.
Hut
3.
\e c
Thai
ae
"*"**
tr* + 0^r
I'
*'j*
wt \M
if
= 2,
our
i>,
r +...) =
+*
*+*> +
<*+...) = K">
lin.
x
=-
"ixlit ion is l.ioki-n. so
that we have no right
<
lint
'
t.-iul^
to
Ex.
+ .+ *)
We
""I
...
(oompti
4.
!>.
can n>-
taking x,,=
liv
(*o
SMJ.
particular
tlii>
Then,
1.
matter of
fact
tO 8
rt
=jy,
\\\-.
oonvergea,
[H.\
notation
tli-
'1
;ls
roved to oscillate as
j.
489).
tin-., i, -in
in
can
,,f
tl.
.res.
i
Thn-
:niot
Ex.5,
aj)]ioa<-h
/;(''
\6
an-i
Irt
A,,
/;..
/',.
th. MI
anv
limit
"tln-i
tlian xero.
;^')=/
i*i/
...
].!-o\,-
+ /;
fl
that
)'
CMC
in
i,
ly applying Art.
*1:
pi
worth whiK-
bo
i:;:;i
t This
is
in
the
tlic
'
;i
ce
1.
loth
the
fa.-t
Iii
/'
that
does actually incriMse
x, lut that
an<l
11
v.
'"Is in
the
first
sequence, and
"n.l.
only point in the proof at \\ln.h care
is
necessary.
IKRATIONAL NUMBERS.
390
[AP.
EXAMPLES.
Irrational Numbers.
1.
(1) If
is
a rational
(a + 1)
number lying between a 2 and

A-
2
,
prove that
(2) If
where
a,
p n qn
,
are rational numbers, prove that
and that
Thus
p,
-Aqn2 = (a2 -A)
n
.
an approximation to *JA, p n/q n is a closer approximation.

[The approximation p s /q s is the same as that used by Dedekind (see
if
is
Art. 145).]
2.
(1) If a, b, x,
prove that either

rational numbers.
numbers such that
are rational
?/
x = a and
(i)
?/
= &,
or
(ii)
<J(l-ab) and ^/(l-.r?/) are

[Math. Trip. 1903.]
(2) If the equations in x, y^
ax + %bxy + cy = 1 = la
have only rational solutions, then
2
J[(b-m) -(a-l)(c-n)] and ^(<^
2
are both rational.

3.
If a
is
[Math. Trip. 1899.]
irrational
and
a,
b,
c,
are rational (but such that
ad
is
not equal to bc\ then
aa + b and
numbers, except when a = in the former.

irrational number a can be expressed in the form
-are irrational
4.
Any
any assigned positive integer and c 1? c 2 c 3 ... are positive

integers less than a. Thus, in the scale of notation to base a, we may
write a as a decimal
where a
is
For example, with a = 2, that
in the
is,
binary
scale,
we
find
^2 = 1-01 10101000001....
If %, a 2 a 3
is an infinite sequence of positive integers siuli
...
7
n can be found to make (a^^... a,,) divisible by JV whatever the
integer N may be, then any number a ran bo e\|n vssrd in the form
5.
that
''
!
+...,
cH
<a
1.
1
ordri
Iii
Th.-
(hat
.-
rettliotiOO
mu>t
th
hi-
find
thii-.
and
so
.i
thr
sum
to
intini!
~1
t9J-^\
--fl -r 1 -J
)
->~>T^7
'
dtln.u^h
'
e|ual
we can determine a
If
6.
d;
linn
"
is
/,
thr
of
cast-
sjM-cial
^0,
to
int.-i;,.|
K.\.
S)
to
th-'ii
'/..a.
,i.
niu-t
irrational.
!<
nun:
[
\ mcb
intrî-i
Fr
tin-
first
part.
\v-
n..i-
.V-
thai
.)/
ii
intt-^ral
aiv irrational
numl.r!
.I/
j^
il
any assigned
.,
tlu-
v.li.-i,-
pl.-as.-.
that
Aj.ply
thai \\hi-n
an tind an
such that
pait of ô,
'
'
<
For th- sfi-ond part, express a as a conti:

The sums of
7.
!
\\lifr--
/-.
t'or
an- any po-hmtin- |i odn< 'II
[For simi
If
anil
[Ki-
-.\t-nsions, ser (Jlaisln-r,
.1
i-
>
\
/'
degree
ifl
iln-
wln-i.
.in
1."-.
n.-an-st
!_
~K^'
Thai
if
can
find
pio\ r that
that
integer to
eu/ M
n\bcr
//-
i:
""
->\.
-urh that
podtive integera.
s-i|i;.
when
of
illation
th-:
oi
/'
1SU.J
p.
8.
inti-_rcrs,
'/
"'
than
It'.
ly taking
it-ncc
9,[
Li
IRRATIONAL NUMBERS.
392
9.
Suppose that a
[AP.
an irrational number which
is
is
converted into a
continued fraction
and p n /q n
the convergent which precedes the quotient a n
is
write further
Qn+ i=g n A n + q*-i,
A n = an +
where
Then shew that
sin
|
if
man
qn
and also that

where en tends
sin q n air
to zero as
a n+ i
\
. . .
a, i+2
> Kj Qn +\
<m<q
n +i;
= sin (ir/Qn+i) = (1 + e n )7r/$, i+
n tends
i ,
to oc.
[HARDY, Proc. Lond. Math.
Soc. (2), vol. 3, p. 444.]
Monotonic Sequences.
lies between the two preceding
monotonic sequences.
two
compounded
(2) If a sequence of positive numbers (a n ) is monotonic, prove that the
sequence (bn ) of its geometric means is also monotonic, where
10.
terms, shew that
(3) If
<?!,
c2 ,
it
...,
of
is
cp
are real positive numbers, and
prove that the sequence

the same
11.
term
(1) If in a sequence (an ) each
is
(ju,
n +i//^n)
steadily
if
increases;
and deduce that
true of //,.
Sn =
If
'
is positive and independent of ^, shew that if 2a r is convergent,

gives the value of lim n (see Art. 49).
Conversely, if lim n exists, shew that 2a r converges, and that its sum
where ar
its
is
sum
equal to the limit of Sn .

to Ex. 12, taking k = l
Apply
12.
and
Apply Cauchy's theorem (Art.

1 [~n
n- 1 n-2
,
where C is Euler's constant (Art.

Prove also that for all values of
to Ex.
6,
Ch.
XL,
taking & = 2.
152, II.) to prove that

1
,
11).
n,
the expression
lies
between
and
1.
[Maf/t. Tri[>. 1!H>7.]

13.
Apply
Stolz's
Inn
then
theorem (Art. 152,

(
+!-
II.) to
+ A -"J=,
prove that
if
where A.>-1,
-
-T-.
+A
IIPLE8
1.
1
/'
[If
th.-n
th.-
\\here
(]
expression can
L'iven
/'..
Th. -ii
!"
\\
increases with
/-,
1-fA
if
//,
i>
p..
-in
-li.-u
!L'
which we
in
iven
ret
Quasi-monotonic Sequences.
<iiven a seq'
N.-t
tin-
are all less than

nil.-
-:'M
/,
ami dm--
iiii>niit<iiiii-
m.t
t"
t.-inl
tii.n
14.
it
OOHV61
iiiuit
if
great i-r than any
ami
in<Tf;i>in'_r
\
which
inl'ir
f"i-
th'-
decrwuing case.
-I'lasi-moii., toni.-
for
m-ti>n of
nut
is
index
^reateM
itli.-i
divf-ri^ to oo or
than any trrni nf th- >c|uence.
f
diverged t" / or converges to a limit
(in.i
si-ijin-iin-
wliirli
si-(|MMK
t.-rni <.f \}\<
tln-n tin- s<-ou<'n<--
is
quasi-m
,-an
ny .on\n-^rin
liaving a comnion limit.
IK-
sul i-dividt-d into
two
.jua-i-nioiiotonic
s'i|iiciic.'-
Illustrations of i|iiasi-iiionotoni<- seqaenoet

4 (-!)"/>,
in
\vh.-n-.
tin-
a n = n 2 + (-l)";^,
two.
first
ly
-{2 + (-
]>ur|>"-,,^
analysi-
,t'
("H \i:i.v.)
Numbers.
need to
\\.-
<<>
irlti<-li
,,
=!
,nt.
Infinite Sets of
For some
15.
't'tcli
Beqilrli'
i-all-
Prove that th
and
nund><
r.-al
>11
countable,
[For
-a-h
if
t.-i'in
/,,
Of,
as an
'
...
intinit.-
i--
an\
write
B6C]
decimal
'0*
Tlicn considci- the
infinit--
iff*.
<'..
d-'cimal
7=
in
\\
It
i<
hich
-lit i,.n
clear th.r
exhaust
all
the u-al nuinl<-
,t"
to tin-
.nd
1.]
IRRATIONAL NUMBERS.
394
[AP.
Given any infinite set of numbers (&) we can construct a Dedekind

by placing in the upper class all rational numbers greater than
any number &, and in the lower class all rational numbers less than some
16.
section
number
k.
This section defines the upper limit of the set

prove that this upper
limit has the properties stated on p. 12 for the upper limit of a sequence.
;
Frame also a corresponding definition for the lower limit of the set k
and define both upper and lower limits by using the method of continued
;
bisection (as in Art. 150).

17. The limiting values of an infinite set of numbers consist of numbers A
such that an infinity of terms of the set fall between A-c and A + e,
however small e may be.
Given an
section
than
infinite
by placing
but a
all
finite
numbers
rational
This section
maximum
numbers
of
set
in the
number
defines
infinite
maximum
the
we can
terms
of the
than an
less
(k)
class all rational
upper
k,
number
limit
construct a Dedekind
numbers which are greater
and in the lower class all
of terms
of the
set
k.
prove that the
a limiting value of the set, in accordance with the

definition given above
and further that no limiting value of the set
can exceed the maximum limit (compare Art. 5, p. 13). Frame a correlimit
is
sponding definition for
the
minimum
limit
Goursat's
and
state the analogous properties.
Lemma.
18. Suppose that an interval has the property that round every point
of the interval we can mark off a sub-interval such that a certain inequality
Then
is satisfied for every point Q of the sub-interval.
divide the whole interval into a finite number of parts, such that
each part contains at least one point (P) for which the inequality \Q', P} is
denoted by {$, P}
we can
lies.
every point Q' of the part in which
not
if
either
half
does
the
interval
satisfy the condition,
original
[Bisect
If we continue the process, one of two
and so on.
bisect it again
satisfied at
alternatives
must occur
we
either
shall obtain a set of sub-divisions
which
satisfy the condition, or else, however small the divisions may be, there
is always at least one part which does not satisfy the condition.
in the latter, we have an
In the former case, the lemma is proved
;
infinite
sequence of intervals
b H+ i-a n +i
and the condition
(, &),
each half the preceding, so that
= $(b H -a H ) and
aw
â
ll+1
<6,1+ i^6 n
not satisfied in any interval (a,,, 6,,).

Now the sequences (), (&) have a common limit / mark off round /
Then we
the interval (/, g) within which the inequality {Q, 1} holds.
is
have an
= = &n
Thus
an ^ we can choose n so that bn - d
">/+(& -0=/
so that the interval (a n , />)

tlii-refore
the inequality {^,
l\
is
is
lt
<l-f=g-l.
and l>n<g + (l-<*)^ff,
contained within the interval
(/,
satisfied at all points of (<(, /;).
g)
and
That
is.
'I
,ii..n
ai.
led
to
.11
111.
ih-
..t
di\
id.-
l.y
an
in:
-MII
mark
,ny two
nat
/.
-ati-lied in (/./*,.):
-o
we
'
ue.]*
'
point
Miml.er of
(I,
-mis
mci|iialit.
.a
and
asumii
parts
\\liieli
li
t.
I'loVe a
IS.]
and prove
Slate
20.
plane
..r
K\-v
I..
Is.
\\>
ie-ion>
to
j,,
in
dOB.
Continuous Functions.
'
21.
:nuoii>
in
the interval
the intei-val,
)
the tipper and lower limits

(Apj.lv the method
In
i<
i:et
the upper limit of
<-ommon
limit
of continued
an
It)
intervals
/(/)<//, choose 8
that
8.
Th-n choose n so
at
all
p"int>
22.
then
that
I,
\\i>-
s to
A.,
~a n =&; and we
'ii
tind
tl.
+S( f )}
to hypothesis.]
contrary
prove that
divided into a
in
it'
int.-i\al '-an lie
;idin-
that //
\>-\\
the inte:
in
if
of /(*) in the interval.
1'iseetion.
infinite sM|iienec of
/'(/')
Then
/.
and
//
finite
numliei of
an
int.
mn
j>art- (the
that
>ni-h
two points
iny
23.
fun<-tion
finite
upper
i-
said t"
limit
_'
finite
in
the interval:
in
that
and
in
the same subdivision.
INK.]
it-
absolute vahu-
the interval.
-
if
a!s..
that
in
an
intei \al.
it
that
l/(f
^
Tin-
form
'in-sat
suitat.l.-
will he
|s,
found
\\)
in
his C<
M-ein to ha\
f* -
<&
in
proof
is of
APPENDIX
II.
DEFINITIONS OF THE LOGARITHMIC

FUNCTIONS.
156. In the text
it
AND EXPONENTIAL
has been assumed that
and a number of
used.
to
It
is
allied properties of the logarithm have been

customary in English books on the Calculus
deduce the differential
ponential limit (Art.
57)
coefficient of log x from the exor else from the exponential series
would, therefore, seem illogical to assume these

of
properties
logarithms in the earlier part of the theory;
although, no doubt, we could have obtained these limits quite
It
(Art. 58).
book.
But from the point of view
seemed more natural to place all special limits
after the general theorems on convergence.
It is, therefore,
desirable to indicate an independent treatment of the logarithmicfunction and it seems desirable to use this way of introducing
the function in a first course on the Calculus.*
at the beginning of the
it
adopted
157. Definition of the logarithmic function.

There appears to be no real need for the logarithm at the
beginning of the Differential Calculus, but we require it in
the Integral Calculus as soon as fractions have to be integral' M!.
At
first it is
probably best to denote \dx/x by L(x), and post-
pone the discussion of the nature of the function

*See a paper by Bradshaw (Annal* of Matkematic*
sunl
Osgood's Lehrbitch der Funktionentheorie, Bd.
1,
(-2),
vol.
pp. 487-500.
4,
/,(.')
190,'i,
until
j>.
.~>l>
i\i
in-:i
157|
[0
ball
the
between
an-a
tin05
= a,
denote
It
AH
in
or.
area
tin-
A(./-i
l>y
ordinair
\B
will
h.-
..hxioiis
fi-
-\
,iii
be drawn,
th.-n
if
hound. -d
axis of
th<-
raun throii-h
two rectang
w<-
>
rhan--.-
slight
./>
(3)
to
7'
tin-
has
/.(./)>(./-!
of
At
nutation.
+.'>>.>
must
In-
u'-').
-f.o.
it
in
siij.p
h:,
/-I
itive,
is
and
//
and the Other nitiivly within AHl'M.
().
Thus
with
tiûr.-.
HTM
or.
variable
tin-
drtinition that
tin-
paralh-i
Mill
l.y
x and
w-
th-
1/j/x,
from
id<-nt
/.,!,
Furthrr.
will
notation of the Calculus,
th<-
/.</>=
as
HTM
tin-
(fl5=l),
= l/x
hyp.-i-l.ola
wht-rr.
nates
tln-Mivm
t!i.-
Ai
ordinal^ I'M
(
nli
Of
below
!i\<-d
presents
curve, the;
lan-ular
tli-
that
x=l
shall
th-
present,
t.-ikm
all
tlu-
t"
ni'ti.-.-
...ld
tliat
in-
the nuiurin-al \alnr.s the inniualit
g,,Kl
v.
///
-/-'
ten than
\vmiUl havr
easy to ahew
\\lim
1.
in
.<!.
i.J),
..i
[iialitii-s
irv
is
//////.
IM-
r.-\cj-sed.
or
LOGARITHMIC FUNCTION.
398
For instance, we get from
But
(3),
<
in numerical value
we take an
if
Again,
[AI>. N.
Z(|)|<1.
ordinate
QN, such that
ON=2.0M,
we have
L('2x) -L(x) =
PMNQ > rect. MN
and
<rect.
That
is,
and so
x(^)>L(2a)-L(x)>x(^)
we
Thus,
Now,
if
or
since
Z(2)>J,
by addition that
X>X
(n +
Hence
if
NQ,
on.
It follows
2 n+1
1>
get
MN.MP.
>a;>2
with x (Art.
1,
evident from the figure that
it is
+1
1)> Z(2
and
Note
> L(x) > Z(2) > K
evident that L(x) tends to infinity

or
it is
2),
(5)
Again,
if
we
write x =
l/t,
we have
cZa?_
~^~
so that
Z(^) =
a;c
dt
"T
=-
- ^(
or
(6)
Hence, as a? approaches zero, since Ifx tends to

L(x) tends towards negative infinity, or
infinity,
limZ(aj)=-oo.
(7)
a;->0
Again, the function L(x) is continuous for all povifiw

values of x.
For we see at once that \L(x + h)-L(x)\ lies
between two rectangles,' one of which is equal to \h\/x and
the other to \h\/(x + h).
Thus
\L(x+h)-L(x)\<c,
\\
if
Inch proves the

continuity of L(x).
S'lTION
157
1
folloWB lYnin
It
tinn
;iii'l
between d>
t'mi'lani'-nlai
tin*
int'-jiMt inn
tliat
.A,-''-'
without
hut
in
aj.jM'alin:i;it
this
/.(./
|
and
I/a;
l/(cr
L(. '))/'
nhtain
rniitain.-d
is
1,,-lv.
Thus
A).
cia;
ire
WC can
fact
+ //)
= ii..i'[(+A) T i()].i,
^_,
two
onli
that
,ch
f>
> >
'/
\v-
><(6-a
l>y
exactly
same arûin'iit
(.'>>.
we have
as
tin-
usnl
\vc
or
Tims
or
Consf(|u<-ntly.
the functior(
n.-arly olivioiis
cw
a;
.!/>'
-///'/
/'.-.
increases.
i
Tin-
.\aiii])|f, tin-
timuv
ln-lnw
ifadri
will
may jn'Vr
tin-
41.
dotted
liiu-8
rcprt'si-nt
(lu-
la^t
n.-i-al
the logarithmic I'unctinn.
'I'lic
as x
L(x)/(xl)
which cm-ivspniuls to the

that //" mean ordinaic
'.NVX;
rurves
y = (x-l)-i(a-
result
id. -a
<!'
ly
ditlci-fiitiatioii.
tin-
course of
400
158.
[AP.
Fundamental properties of the logarithmic

u
f dx
II.
function.
In the formula
change the independent variable from x to g by

x = g/v; we find then
writing
going back to the definition,
or,
= L(uv)-L(v).
Thus
(1)
From
equation (1)
follows at once that
it
n
L(u ) = nL(u),
(2)
where n
is
rational number.*
any
Suppose now
that
is
number such that L(e) = I;
existence of e follows from the fact that L(x)

x
function, which steadily increases from
varies from
to oo (see Art. 149).
Then equation
is
the
a continuous
to
+x
as
(2) gives, for rational values of n,
L(e") = n,
(3)
which proves that L(x) must agree with the logarithm

base
a?
e,
as
we
ordinarily defined;
shall,
the future logx in place of L(x).

We can obtain an approximation
observing that,
writing x =
when n
is
to
therefore,
to
write for
the value of
by
positive, (3) of Art. 157 gives, on
+ l/n,
(4)
n+1
n/
l
+
(lV
1
so,
since the logarithmic function
must tend
to
is
tends to
as its limit;
and
continuous and monotonic,
e.
Equation (2) maybe used (see Bradshaw's paper, 4) to establish the existence of roots which are not evident on geometrical grounds for example, th;
Of course, from the point of view adopted in this book, it is rnoiv

natural to establish the existence of such roots by using Dedekind's section.
lifth
root.
M>
li
158|
WG
Similarly.
\Ml-.N
PROPER]
\l.
1.
Thus, we lind
lim( 1-
'
".
)
easy to give a dn
is
It
limit
./
in
I'-i
hat
the
bi
a definite
<
that (logjr)/(jr-i) decreases aa
increases.
Thu-
I'.ut
is
that
-i-e
th-ief-'!-i-
i-
does
/)"
In
so.
MC
lo_ir(l
tlie
saint-
it)"
\\ay
prov.
that
(1
than (1-1
less
that
ur
iiHTrast-x
we
~*
1
/<)-",
and
-2.
tl.
Thus
(1
//)"<4, and c(nse|iiently
(1
-fl/0" converges to a definite
limit
A-
matlrr
In
>\\ <-\
-r,
tlu-su
limits
ar-
imt
v.-rv
Their ^Mnn-tri<
.inj.utatinn.
Letter approximation: for it will be seen that
II
fact.
!'
i'r
nuiiit-rical
c<
nice
Tlius
Wi-
it
1
(
\
tin-
]
-
will
I..-
V = 2-7048,
third of whi-h
is
found tint
(
V
only WTOB
\n -\
:nit
in
V =2
the last place.
It
i>
-alcu{>erhaps .if hist..rical int.-r.-^t to note that N
lated the tir^t tal>le <-f h i-ai'it lim^ l.y mean->
an appn -ximate
!'
formnla aili-l
to
i.
nam,/-
La
402
[AP.
ii.
The error in the approximation (8) can easily be shewn to

be about b[(a,b)/af, and in Napier's work (a b)/a does not
so that Napier's approximation is right to the
exceed* 5/10 5
13th decimal place.
;
Napier's definition of a logarithm is exactly equivalent to

the definite integral which we have employed he supposes that
the velocity of a moving point
is proportional to the distance
;
from a fixed point 0, and is directed towards 0; so that

of
the time represents the logarithm of the distance OP, if the
initial distance OP is taken as unit.
159. The exponential function.
Since the logarithmic function logy steadily increases as y
to + oo it follows from Art. 149 that, correincreases from
,
sponding to any assigned real value of x
there
is
a real
positive solution of the equation
We
y the exponential function when x is the indeand write y = exp x

the graph of the
pendent
function can be obtained by turning over Fig. 41, p. 399 and
interchanging x and y. The figure obtained is shewn below
call
variable
-2-1
01
Dotted curve, y =
FIG. 42.
It is evident that the exponential

*
function
is single-vain
Generally, the function inverse to a given function is single-valued in any

which the given function steadily increases (or steadily decreases).
interval for
I:\I-M\I.\
158, 159J
\i
mot
value of X in
OCM
.dlled
Suppose now
80
tlial
that
e-have
(1)
+*)<*<*/*
/<//</</'<//
H-m-,-.
ion
.'-}-/',
Arl
fi-MMi
'
'
<
//
Thus.
fmicti,,|l
+ /).
Mini
///,
if
|A|</(yH
|A-|<e, and
see that
>
//<
expont'iit'nil /"i><-fion
ilUOUS.
\Ve have proved in Art. i:,s (3) thai. \vli-n x is

iiu*- of e*.
-\]Mii-ntial t'unrtii.n i*> thin-atinnal, <lrtmr.l
ra
l>y tli-
u]j>-r and
ln\v.-r
= t'.\jM/
< cxp <
'\p
t'unctinn
kuse tin- cxpoin-iitial

.-\j..r
ami
i-iMtinumis.
fa
<-'>i!MM|ii,-ntly
d.'tin.-d
///i
log 1
Sine,-
.i-y,/,
= 0,
th- continuity of
can
(rompaiv
D.-drkind's ilrtinition
/V//,
"
'
'
cla^s.-s
tin-
hy
./:
it
,///'.//
tin-
expos
Thus rxpr
).
is
that
is
tin-
irrational;
/'x
now
a;
is
o7<
fil.,,rt',u,,
i'..Il..ws
=eA
./.
mad.- a^ small as
wh-n X
''.
If
we
increases with
In-
(e
(e*),
o!'
/M
p.
.1
ro/^'
///r
Al
we
sinM,-
please;
number
coincil->
with
and so for
y;ox/'/;/V
>"=!. and 80 w.
"//
/v////*
/g*.
hav,-.
iV..m
exponential function.
Hn,
,-->
= ,"=
1.
.r->0
ii.
because
Logy+logy'log(yy^ we ha\r
l
et+.>
Of course
(8) and (4) agree witli tin- ordinary laws of inaM^ln-d for rational nundtcrs in honks ..n ah'vhra.
Ki-.-ni the dctinition^ of the logarithm and
exp)iiential functions
it
follows at once that

i
/if
</>/
'/
//
Thus
that
flu-
'i
wop
derivuf
is.
j
This result can also
deduced
at
<>n
'(self,
404
The following
[AP.
II.
inequalities are often useful:*

for any value of x.
e*>l+#,
(6)
ifOO<l.
e*<l/(l-aj),
These follow from (3a) of Art. 157.
(7)
Some miscellaneous inequalities.

when a;> 1, we have from (3)
160.
Since,
\OgX<^X
(1)
it
follows that
if
is
positive,
x/(n -f x)
we
Thus,
<05,
positive index,
any
<xn
\vgx
Again, from the same
(2)
of Art. 157,
< x /n.
n
log x
or
we
article,
see that if
x and n are
< log (1 + x/ri) < x/n.
find
'
(3)
lim=#, and
Since
since the
exponential
n->w
function,
a continuous
follows from (3) that
it
(4)
+ |)".
lim(l
>t X
't''
71
we can prove
Similarly,
that
is
if
n > x > 0,
_
When n
is
we have
a positive integer,
K)"='+<'('-:>and
since all the terms are positive, this gives,
and consequently, by taking the
e*^l+a: + ia
(5)
Similarly,
we can prove
2
,
*The
geometrical meaning of
any
(3),
limit as ?i->oo,
if
>
0.
that
if
entirely above
from
(6)
of its tangents.
is
./>().
simply that the exponential cutvo
IMJ.l ALII
159,160,161]
Some
161.
limits
II.
article tliat
last
th.-
..!'
<./"
s>
if
//,
and BO
I.
LIMITS.
the logarithmic scale of infinity.
\\V have seen in c2)
AM*
>
1,
>0,
-IlCe
limllo^.r
fl
x *co
mply
1
lii.
>
if
<!<>-./
>0.
//
X'-ÔO
Since
loga=
last
tin-
log (I/a:),
(3) of th- last
SCt-
tli.-
same as
>0.
it'
we
article
is
equation
= 0,
liui(.'-"loga:)
that
(x/n)<e*>
<
or
U,
/j-n^
and therefore
or
>0, /?>0.
if
,.
,,-aj^
liin(./-'v---
lim(./"v-')
Of course w
= 0,
can also write

lim
(4)
we have
writing x in place of 2x,
1)\
= o,
if
form
in the
(.'*)
if
(/*.,
>0.
//
x ><a
Tin- results (3)

is
and
course true
//_(); lut no proof
if
thru
We
Art.
can also obtain
l.'n'
;>
(I
HUM
and
l'\
-hanging
or
we
to o^
Th<-
write
limits
</
OD
m-l
(!) (4)
It
than
get
?*
\ve
**
L'Hospital's
rule,
(1).
Inn
JT
0^*
"
>0.
>oo
raise the last to the 7ith
form
tin-
of
'.
On
th- other hand,
tends to
oo
^mall
its
li\v.-\vr
we
get
"*^
l^-./-
on.
we
power,
basifl
follows from (1) thai
any positive power
ami so
to
,^=0,
<
Si,,,.:
It
aii)ealing
\>y
;ilx)ve.
see
still
(4).
CO/
more slowly
index
may
more slowly,
from (4) that
e*
tends
406
[AP.
II.
faster than any power of x, however large its index may

and hence a fortiori e? tends to oo still faster, and so on.
Thus we can construct a succession of functions, all tending
to
be
oo
to oo
say,
. . .
< log (log x) < log x < x < < ef <

ex
and each function tends

the
preceding function,
of the following function.
to
but
. .
oo faster than any

power of
more slowly than any power
It is easy to see, however, that this scale by no means exhausts

types of increase to infinity. Thus, for instance, the function
all
e (loga:)2_ ^loga;
tends to
oo
more slowly than
x
,
but more rapidly than any
(fixed)
power
of x.
a* = e* l **
Similarly,
tends to infinity more rapidly than e*, but more slowly than e? or than ef.
Other examples will be found at the end of this Appendix (Exs. 11, 12,
p. 412).
162.
If
The exponential
we
series.
write
we have
Thus
and
so,
.(l-e-*X n ) = e-*(Xn -Xn _,) = e-*

since 1
Xn
e~ x
is
zero for
cc
= 0, we
have
a result which can also be easily obtained by applying the
method
of repeated integration
by
parts to the integral
i>Multiplying the last equation by
ex ,
we
find
ONEN
161, 162, 163]
I.
But when X
and 8O
in
i>
losa
than
(^+i?
-'
(=
integral
3.11+1
"J
i~
last
th<-
'
Iori
\\'li.
407
II \l.
less
is
say),
ir
IJo
than
e*,
and so
'
>
w.- have bherel
l-f-^-fC-r.-.-h'^-rA',.
A'.i
In
lini
t-i
<
/.'
r^-r.,
<)
(Art.
The existence
,ual.
arithin.'t
)>
found
sincr
lisciission
ic
^>
0,
wh-n
./
<
0.
Kx.
...
and so
4),
to
of an area for the rectangular hyperbola.
\Yc inv.- here a proof thai
area which can
./
-2,
.,.
163.
when
l>y
l'c\v
of
tin-
angular hypi'rl>ola has an
ivrt
ddinitf limiting process;
Ku^lish
aiva of a
tin-
this
hooks î\- any ad'juate

curxv.
The method
applies at once to any curve which can he divided into a //

numhcr of parts, in each of which the ordinate steadily increases
or sti-adily deereasee; although the at-tual proof refers only to
curve lik-' the rectangular liyp-rl)ola in wliicli the ordinate
constantly
<
Increases.
408
[AP.
II.
with this strip an outer rectangle PMNP' and an

inner rectangle Q'MNQ\ now bisect
at
and draw the
l
ordinate R^M^ This gives two outer and two inner rectangles
outside the curve, R^, QM1 inside.
But
namely,
ly
the sum
R
is
less
than
the
outer
+
1
1
obviously
original
is greater than the
rectangle; and R 1
1
original inner
associate
MN
PM R^
N
PM
M+QM
rectangle.
If we again bisect
four inner rectangles
MM
;
and M^N, we obtain four outer and

l
and the sum of the outer rectangles has
again been diminished while the

increased by the bisection.
of the inner has been
divided into 2 n equal parts, let us denote the

of the outer rectangles by Sn and the sum of the inner
When
sum
MN
sum
rectangles
by
is
sn
Then
>S >S >...>Sn >...,

< < <...< <....
S
and
sl
82
8n >Sn
AISO
(71
= 0,1,2,3,...).
Now, from the figure, we see that the difference S1

the sum of the two rectangles PR l} R : Q, which is equal
sl
is
to
Similarly,
and generally
Sn - sn = J (Sn _ x - sn . x ) =
It is therefore clear
common
limit as
area of the figure
from Art. 149 that
increases
this limit,
'Sn
^(S
and
say A,
-s
sn
is
).
approach a
called the
PMNQ.
But it is essential to prove that we find the same area A

in whatever way the base
is supposed divided to form
MN
the rectangles.
i denote
I. .-i
ili\ id.'d
<!'
the
turn
shew
will
tliat
.-1
the SHIMS
if
j
/!=T. w-
Hiii }L
=J =
^ô
//
w-
is.
A'
/"
//'/>
.main.-.l
_!
in
the
olitain
ilividol.
is
that
LCb
since
Thus.
.17
n.
-QjV),
wil-st
2-(r<e,
That
"!'
lini
MII ohOQf
tln-r't'. .!-.
lias,-
glanoo
1.
^l. rj.
llr:
ami
<T_ J.
tin-
>!'
Th-n
is
denote
independent
v/M/
linsultli
tin-
is
,-)'
<
2=^,
1.
llut
/r
w.-
Nn a
lini
l.-t
!'<
uite
'i'liu
MN
wh.-n
nd
oorresponding inner
wli.
\\hnv
4<M
I.
Vt <>r irr.
S>
We
i:\
<>\'
any
in-
li-iin-
li.-
in
up
sum
tin-
at
OF
Al;i: L
163|
S<
liin T.
^^.0
aiva
*
!>,<;,/.,/
whut-\vr way
in
.1.
f/,,,f
tin-
//,/
/-a
:'
Extensions of the definition of integration.

If
taking
(./,
//,-.
an.l
r,
function
tin-
int<-i\al
\vlin-.-
lr
t<>
/'
upj-r ami
tin-
is
i':i,
i.f
..l.tainril
iy
divi.linur (".
have gone
Supjxisi' that in
\v.-
tin-
at
prrst-nt
//// sMl-iiivi-j,,n
hut
th-
int-'i\;il
iiion.t,ni.-,
ami
BO liavr
we m\v
// ...
then
it
nn;
il.lr
n-a
iict-d
and
aU
modes
of <livisic.ii
..t'
th.-
//
t. tind
c>
liin.-
a-
W r <,
that
int.
//,.
thai
luivo
latin
hciiiu'
for
IT
<lttel
an.
th.
only
iiulio;it-.l
the
,j
b-w.-r
intr<du.-f
Qto -ul-int<-i-\
n thf iippn-
function in th- intn \al
ill
tin-
in
fum-timi
limit>
denoi
leea
in
ii;
easy to see that

limits as i< trmls t> intinit
is
it
tinitf
::
p.
limits
l<>\v.-i
mOOOtonic, hut
ii"t
i>
K\.
s.-.-
$ n ^//-
Writing
y,.
iutr'_r i"itr.l
A) (f>r (Ictinitinii,
rc.-t.ui-;.
f|ual.
410
Under
[AP.
we have obviously
this condition
(-)
=
liin
Km Sn = lira sn = A
so that
(),
sa}*.
Then, just as above, we prove that for any mode of division
and by Riemann's
condition,
S-o- <e,
It
Ex.
is
It
two
<8.
<
<,
because
^r) r
= b-a.
also easy to extend the definition of integration to a function of

let us consider the meaning of
variables, say .r, y
is
///(.r,
where x ranges from a to

If
we
yj r
for (see
easy to shew that a continuous function is integrable
- a), if r <S. Thus we find
I.) we can find 8 so that <o r
c/(b
rj
App.
22,
2i7 r a> r
if
Km2 = J=limcr.
Thus
we
divide
(a, 6)
y)dxdy,
and y from
6,
a' to
into 2 m equal parts
and
6'.
into 2 n equal parts,
(a', '&')
obtain two sums

*^m, n
where
and A M
== 2jtl
IL, vifj., v i
&m, n
^'^fJ.,
/u,
v>
are the upper and lower limits in a sub-rectangle

or n is
Then, just as above, we see that Sm n decreases if either
'TV,*.
and s m>n have each a limit when
thus
increased, while s m<n increases
,,,,
H^
wi, T&
tend to infinity in any manner (see Ch. V., Art. 31).

if
f(x, y) is continuous we prove as above that
Further,
Km S
nlt
Now we
= lira s Mi n
F, say.
have, from the definition of single integrals,

/
Km
lira
\ M >-oo
*oo
Cb
Sm>n )=
>
dx
\mâo
fb'
\
Km ( Km SW) B ) =
and
(*
''
dy
f(x,
y)dy
(* /(a?,
'
so that these two repeated integrals are each equal to F,
and
therefore to
>each other.
EXAMPLES.
1.
Prove directly from the integral for log# that
,[For
If
we have
we take
log(2) =
J*^,
these integrals from
to 1J,
from \\ to
1J, etc.,
we
find that
111
Il.'l
2.
th.-ii
t.-h
If
3.
>i, t
tends to
-*l as
gra]
prove tint
'.
Dcdll.T
th;,t
l
...+o
\\hi.'h
,f
tin-
<>na
t\v.
1>r
(*'
'
l.-j.-n
ami
ithin-
ii"t.-
that
181.]
5.
and
If
is
/'
nmnrrirallv
than a lixed number A,
l-ss
iiuU'|H-i.
if
.og
(l
then
<
lim.r
log(\ 1
+ i + _L_
/'
//
l"u
/'
/'
/'
log?i
=
Art
6.
Tsc the
1-2
(1).]
t-xain.!!-
last
t.
>h.-\v
that
if
'|
tormi
iiat
tin-
1" converges
,
if
I-
/x>l,and
..thorwisr diverges.
_] )-<"+!)
is
div.-i^riit.
7.
If
>
Art
Compare
".
that .-
pfov
IS,
-(l+./-)!lg(l
+
1906.]
[Write log (!+./) =

8.
Provt- that
as
'2,
and
UM
th.-
rank's fr..m
-1
log(l-fj-)
iitinuous
'-e'f
to
x,
tlu-
> 2^
funrti..ii
_1
-
and steadily decreases
fr'in
to 0.
if
is
412
[From the
continuity
is
[AP.
dislast example, we see that the derivate is negative

only possible at # = 0, and when \x\ is small we find that
;
_J_ -l_l-.a
x
log(l-Kr)
the series converging
Shew how
9.
if
12"*"
|.r|<l.]
to determine
so that
e*>Mx y
\tx>X,
where M,
are any assigned large numbers.
[We have to make
x > log M+ Nlog x>
which
But
(since
log x < &Jx) can be
satisfied
as a rule these determinations of
10.
The logarithmic function logx
[Apply Art.
11.
is
by taking #>21ogJ/' andlG.V-.

are unnecessarily large.]
not a rational function of x.
161.]
What
is the largest number which can be

expressed algebraically
of three 9's ? Estimate the number of digits in this number
written in the ordinary system of numeration.
by means
when
12. Arrange the following functions in the order

which they tend to infinity with x
of the rapidity with
# l0 **,
x*,
the
Indicate
members
13.
If
a?)*,
position
(logo?)'
each
of
we assume
to be
that,
Shew
if
is
(log log .r)^
functions
these
of
10
**,
that
we
find
series.
(-1W
product on the
(nl)
= 'Za nxn /nl, we
1\*
a/ ill
left is called
v,
we
e*-l
^- V = -^- =
v
x
|
positive,
dv
Taking
10 * 10
in
**.
between
the
the binomial series for any integral exponent, and
[If the
(logtf)
an integer greater than
and so obtain the exponential

14.
10 **) 8
of the standard logarithmic scale.
suppose n
Deduce
(log
get
a?
+ 2! + 3! +
,
get at once, since o
= 0,
LMPLE8
ii.
I
vre find
whi.-h
15.
idmtit
tin-
i-
rly.J
Sh.-u
that.
.iH
*-*<),
G^-r!
16.
If
-,
-'""
(l
'
-
11).
|.|-..\i-
<
<x
[It
will
_ [
Xwhirli give*
17.
'-
be seen that
tin-
good
X-J
_Mc_
0/4 ( W
\
}_
-.'
result.]
t. the
a|)ii-.'.\iin:iti.n
fuii<-ti.. n
K\.
..f
16
taking
l>y
the
ermr
[Aj.j.ly
18.
in
wliieli
is
tlie
<.f
Kuler's
I'I-MV.
and
[Dill'eiviitiutr,
Shew
20.
t<
that the
infinity
I'mve
[A.
(/<
l...tli
I=^Y^_
J
!\fl
):
-id.- irdu-,- t.
tliat
n-f2
T 8!
Mqn<
3
= e'\
nmrr rapidly than any
<-
ni.-nil.er of
ial
tli-
tl
2(loj
convergee
it'
L"i><;K.]
:)).]
= e-,
Ifl
(I""
that
19.
'il.-i
.
-i
or
it"
ami
and
<.tli-i\\i>r
scale.
APPENDIX
III.
SOME THEOREMS ON INFINITE INTEGRALS AND

GAMMA-FUNCTIONS.
Infinite integrals: definitions.
164.
range is infinite or the subject of integration

tends to infinity at some point of the range, an integral maybe conveniently called infinite* as differing from an ordinary
If either the
very much in the same way as an infinite series

from a finite series.
In the case of an infinite integral, the method commonly
used to establish the existence of a finite integral will not apply,
integral
differs
ff
if we attempt to modify the proof

must accordingly frame a new definition
as will be seen
We
of Art. 163.
f(x)dx
{00
f(x)dx when this limit
..
exists.
Secondly, if the integrand tends to infinity at either limit
fb f(x)dx
a
as equal to the limit

lim
f(x)dx
6->0 J (t+8
when
(S
> 0)
this limit exists.
Following German writers (who use uneiyentlich), some English authors have
used the adjective improper to distinguish such integrals as we propose to call
The term used here was introduced by Hardy (Proc. Lond. Math. ^
infinite.
(ser.
1),
which
is
34, p. 16, footnote), and has .several advantages, not the least of
the implied analogy with the theory of injinite series.
vol.
DEF1NITIO]
[164]
tin-
If
ran-.-
ami
.-uM
tli-
liml
,-"'.
tinitr
\vliil.-
intinitf.
- ral
liml
n-rtain proLL-niv
in
ln.th
il,,.
tin-
i^ral
la*-:
int.'-raU tends to
limit
tinit-
>
two
tlit
ly
limit^ in tin-
\\
liin
li\il.-
il-tim- tin- int
i-
within
point
to
But
intii.
in
oi'
into fcwo,
t"
tendfl
int'"_;!-an'l
\v.-
then
tin/
r\ /,-)^=ii.It
is
one,-
at
.-viilmt
'
:
-i-
['/(./) //*].
bend
tliat
temuj
tin-
OA
aii-1
u^.-
.!'
definitiona
Exs.
1
>
%-
ftr
^=
JO
IA
liiM
;='
A-.
n/.
di-
/"fr
"/./
,^
-T+lim
"
pi
K)-'-5
i-*o 2
nJ-a X
5_o\.
=
\\hfic
shnuhl
in
tin-
la>t
i\\..
ik.-d
i("
(s
int.
that
in
>u])jisf
tin-
la-t
and
/<
CMQ ^- ihoold
8
ml to a
Maik,
<
tin-
apply t> these
tlffuiiu-
1.
1.
j..
limit
-.Ml.
BI
doea
so.
416
INFINITE INTEGEALS.
Exs.
[AP. in.
divergence and oscillation).
(of
dx
diverges
l
dx
x
,.
sm.vdx
oscillates,
and
diverges
/*
and
am
Jo x*
\xj
oscillates.
must not be supposed that the two types
It
An
integrals are fundamentally different.
of
infinite
infinite integral of
one type can always be transformed so as to belong to the

other type; thus, if /(#)-> oo as x->b, but
where in the interval (a, b), we can write
a
x
=7b
and the integrand

1
Ex
Jo
in
is
a-\-b
-
or
is
continuous
else-
finite.'
everywhere
-*-/-
d-^
2
)
-^
reversing this transformation it may happen that an

integral to oo can be expressed as a finite integral.
By
rao
When
Ex.
if
5^2
= !/,
r\
ar'd-x
becomes
J^
(both integrals
still
converge
which
%'- d%,
is
a finite integral
2>s>l).
if
It is also possible in many cases to express a convergent

infinite integral of the second type as a finite integral by a
change of variable. Thus we have
x 2 y*
o(l
by writing x = sin
finite in
0,
the interval
and the
latter integral is finite if f(x) is

(0, 1) (for definition see Ex. 23, p. 395).
Kronecker in his lectures on

a transformation is always
theoretically true,
*
it
is
definite integrals states that such
possible, but although this is

not effectively practicable! in all cases.
Care must be taken in applying this kind of transformation when the infinity
is inxide the range of integration.
Here it is usually safer to divide
of f(x)
the integral into two, as already explained.

t If
new
/()-*
variable.
oo
as
-a,
we can
write
J'(x)dx = i-
Jm
Similarly in other cases
express a divergent integral in the form
and
/
in the
</;.
and introduce
as a
same sense we can always
m-:n \rrioNa
164,165]
Special case of monotonic functions.

Although. as we have pointed out in tin- last
165.
art
definition of a definite integral iv-jinn^ in gen.-ral a

for
an
case of
ilif
definition
direet
infinite
can
integral.
obtain
when
as tin- limit of a sum,
tin-
,,f
integrand steadily increases or steadily 'Increases.
f(x)dx
cMnsi.lri-
tli.-n
tin-
only
i'alls
.
un.l.-r
tin-
'I'ln-n
rul.-s.
ordinary
be a seju-!
w-
c;
'Because the integral
iiit.-^ral
function f(x)
x greater than
lrrivasr> to o for values of
steadily
tin-
in-
fcluefl
to
let
/:
.'',,(='),
liave,
1.
(*+,-'
/ -'
>'
I
>
Jx*
Thus,
tin-
ii'
integral
th-
)f
vii-tut-
>f
two
MM-ii-s
tin-
COnVi
the
contains
conver-v,
it'
every
Hw.
-\-i-r.
in
the
only
of
-J-"
./
and
<
/()=]
l>
lli
have
l.e
well as of
(as
i--
'
./-.
it
io
to A, say:
si
this
.r n
,-,
easily enûn- the

|.n>vr that
i^
limit
must
made
to
he
(1).
a^
then
chosen function of
is
independent of
= r+nlt and we have
-A[/(c)+/(+A)4
i
not
sufficiently rapid:
will l>e found that
we can
common
virtue of the int-.|ualiti'^

Miple.
limit
to
need
tii-st
(1); and we can also
common
}.i-operly
,<
ZrO ly varying//:
eijiial
The
tei-ms.
j>ositi\e
taking Xn to
l.y
\\
in
"
the
tend t"
/.
certainly converges, in
integral and the fact
srconil
ti
some parameter
Convergence
tin-
(I),
tlm rate of increase of
instance, \\ith
t'-nn
in
/.
-'
:,)Î^"
<
the value
to
/<.')</./
>
+ ...].
//
and
418
INFINITE INTEGRALS.
[AP.
between the two sums is

and their difference tends
It follows that the difference
so that both are convergent,
with h
hf(c),
to
hence
In like manner, if xn+1 /xn is independent of

= cq n we have
say, so that xn
q,
III.
n and
equal to
and
Thus we can again infer the convergence of the

from that of the second, and we see that
first
series
/-co
Consider
Ex.
1.
We
have then
xe~ x dx, with
xn =nh.
7=
h X)
a value which can be verified by integration by parts.
/oo
Here write xn =cq
by applying one
Ex.
3.
and we get
of the fundamental limits of the differential calculus.
It can
be proved
if f(x) steadily decreases to

r
(where s>l).
x~*dx,
by rather more elaborate reasoning that

x tends to oo then
as
sin
<t
r<*>
oo
xf(x) dx= lim h ^f(nh} sin nh,
cos xf(x)dx =
Let us consider the simple example

r
),
the
sum
is
then
//
sin
**
+ (sin h + $ sin 2A + ...)
liiu
h ^f(nh) cos nh.
DEFINITIO
165
1
Sin,
is
I,---,
,
I
UTT
Ex.
-an
Ar
limit
tin-
givei
i.il
be verified
th.it
5.
liy
means
lim/, v
v-o
..f
(<--*
tlii-
correct
th--
iîvi-
Art.
I.
<
verify in th- SHIM-
tin-
+ 2*-
./'
int.-î;il
that
w- ran
''/r,
e- * + 3*-'e<
\\.-iy
the
-?
Jo
Ex.
tic-
-27T),
v.
l.y
The reader may
4.
than
tin-
-H'.i
;i
+ ...) =
|r-.vi-
that
"./-'
r
whi.-li
.It
In
lik-
has already been found

if
iiiaiiin-i-.
oe
/'(./)
Art.
in
ly
Inu
-<>.
flB
"1
an-.th.-i
inrtho.!.
st-alilv decreases
ft
as
varies
a;
erges,
I
f(
fr.in
to
we can
6,
ji-<vt:
that
\vli.-n
f(x)dx
we have
i
JO
Ex.
Take [\ogrdx', we have

lim l>(\ -
6.
t..
find
= 6 (lo- />)-/>,
a-
int-'î.tt inn.
v.
In
tin-
infinite
\vln-n
limit
a
l.y
piwimis
i-anu'.
of n /in
iff
that
series;
single limit.
nt -^ra
-
nl
tlie
t<>
/'.
Then
to that
of Art.
II
l<
to
{/("
can replace a
\\e
is.
nt
<
an<l
'
from "
and
how
evaluate an
loulle limit
(See ah
Thus, suppose that in

i
have SMMI
\\.>
ly calculating the limit of an infinite 9BTJ

is tinit-- WG can also "l.tain
the result as the
intr^Ta!
the
\\-Mi-k
\\i-it.
will
-f
/'
>>
shew
>i
that
.- ////.
that
inte--ral
/'(./)
f(x)d&
steadily
tlie
and an argument similar

i/./-
lies
)+/> + 1'A )+ ... +/</>-/'
)]
liet\\-en
the
420
INFINITE INTEGEALS.
Now,
[AP.
III.
as
7^0, the integral tends to a definite limit and the

between the two sums is Ji{f(a + h) /(&)], which
tends to zero with h in virtue of the monotonic
property of
f(x) (see pp. 423, 424 below). That is,
;
difference
P
Ja
f(x)dx = limh[f(a+h)+f(a+2k)+...+f(b)],
7t->0
which gives the value of the integral as a

single
limit.
ft
Ex.
Consider
7.
Write A = -, and we have
(by Ex.
Ex.
0<s<l.
where
J^x-'dx,
to find
Art. 152, above).
1,
In the same way
8.
\Qgxdx
liml 2log(-) = limlog(

6
found as
is
YT =-1
(Ex.
],
Art. 154).
Ex. 9. If we divide the last equation of Art. 69 by sin

tend to zero, we find, if a = 7r/ n,
0,
and
let
n = 2"- 1 sin a sin 2a
Now
. . .
sin (n
change from n to 2n and write h for a
Thus, extracting the square root,

sin A sin 2A... sin
and from
For
this
we can
find
(?i
-!)/$
we
-",
(if
get, pairing the terms,
A = 7r/2w),
"log sin #.
this integral is equal to
lim h [log sin h + log sin
Ji!." 27i
=Tests
(2/<)
. . .
M)
166.
- 1 ) a.
TT
f* log
log
x ) io
2]
2.
of convergence
positive integrand.
If the function f(x)
n ~( n ~
+ log sin (nh)]
for
is positive,
infinite
integrals
with a
at least for sufficiently large

rA
values of
increases
x,
it
is
with X;
clear
that the integral
f(x)dx steadily
thus in virtue of the monotonic test
lor
165, 166]
t
integral
;,,/(
.1
\.
In
usual
tin-
practic.-
ni.-t!i."l
nppi-al to tin!'
positive
which
tlii-n
rat--
any
also
/(./)</./
of
valu.-s
for
WM
positive function
ii'
't(x)dx COIIN.T^,--,
tixrd
in the case of
is
principle of
terms; in i'a-t.
ted
thlfl
applying
!'
<
converges
than
iC
if
M>me
nuiiili'-!
(Le
I
<[
ami
tliis
la.st
Thus, suppose
rxpn-ssion
>n that
eu
\v.-
'.\\hereaispoArt.
fr.iin
i-=.rê"
-,
1.
<l,
ar
<e~i a
Ait. i;i)
".
If
-',
if
./->c,
if
r>c.
e~^*dx
(a
th
n '/./-
i<
scale of
tc.
;u
l.y
th.-
;ii<l
of
rU
Ii
inlii,:
Tim-
",/
"/'
\t.\'.
.-(.nvcr^nit.
be niultijilifd
one
convergent, and conse-
tliat
fin.l
\-fA'-(
fluif
is
in write A'=e*, we
iiinic
./ê'l^-Ô a JT-ÔO,
'
A-**
.ê~
:il'.v-. \v.- lind that
I'll
'
I'-t.-riniiH'
iind tlion
of \.
in<l-j.'nl.-nt
we ooond
_Mii\.-
[">;'
is
a <ni>/ a
///
'l
,<>o
->,
/
fi'dl
J(
J
Tin- coinparisoii
h'
80 also
Iwaya
I'm-
I'un-
<l
.:nl
as follov
iln-n
<;<
at
is
j
valur
t.-st
any
rate
at't.-r
crrtain
422
INFINITE INTEGRALS.
We have proved
found so that
type above) that
(see the small
whatever the index
/?
if
is
III.
positive, c can
be
be.
may
Now
[AP.
Jc
and
this expression tends to oo
Thus
y&e^dx
I
If
a=0,
it is
By changing
with
A, so that the integral to oo is divergent.
also diverges.
easily seen that this integral diverges if

we deduce as before that
fi=
I.
the variable,
diverges under the same conditions.
f(x)dx diverges if we can find
Joe
a
an index
aîO, such that one of the conditions

- a)
(1
a
(i) f(x)
J5(loga;y a-
>
(ii)
/()>
>
BxPer",
or a
= 0, /3^-l,
is satisfied.
These conditions are analogous to those of Art. 11 for testing

the convergence of a series of positive terms; and, as there
remarked, closer tests can be obtained by making use of other
terms in the logarithmic scale (although such conditions are
not of importance for our present purpose). But one striking
feature presents itself in the theory of infinite integrals which
has no counterpart in the theory of series.
integral
An
f(x)dx
limit
may
zero.
converge even though f(x) does not tend to the

Naturally,
function, for lim /(&)
convergence; but
()
we must then have an

is
obviously necessary in
oscillatory
cases of
all
To see, in a
we may even have lim/(#) = oo
is possible, we may use a graphical
.
general way, that this
method.
1
l-i...
IS,
Consider a curve which has an infinite series of
/>'/,
x,
of
to suppose
steadily increasing height; then, it is quite possible
that their widths are correspondingly decreased in such a AV.-IV
TESTS OF
166
1
that
areas of the peak-
the
(|ii,
ONI
'
Bl
t'.nn
f(x)dx may con
ntly
in
pai ti'-ular tin- fun.-tKui
-r* (l+.r-sin 2 .r),

1
(a>/3>0).
itli
to
series;
oonvergeni
fr
carve //
i^, at every point
In tin- intnval fi<
tin-
.c
~ a
ft
It'll
\\hi<-h
is
./
its
Xiaj.l
multiple
.-i
..f
r,
we have
(n+1)
thai
f
'
this
is
it
evident that
/'(./)</./
/> 2(/?+l)
';th;r
And
êiu-rally,
if
</>(#),
( A% )
^rs
if l\/i(/< -f 1)
or
to
steadily
.,r
.'..iivrrL't-s
diverges with the
a^2(/?+l).
oc
with
*, tin-
integral
44>
l
+v
[VK*)P converges and divrr^-s
if
-00'
diverges.
o^dor
/"*
Ex.
converges
1.
-/0
a>(j+l
Ex.
With
2.
"i
divergee according as
l+j<r\*inx\
or
a^j8+l.
[HAUI.Y.*]
.nrergM with
/"</,(
-^|^^]
-''
and diverges
[IIAK..Y.]
irff)
Ex.
with
3.
ih
/"<(*)-Jo
M/
I'
'
spit.-
!'
tin-
Accessary for
rrsuh,
last
86) that
''I'/
t/<>
.Messenger of Math>
mm iX
\v,
can
decrease*,
<jence of
April
1JX-.
i'1-ov.-
tl>
diverges
()
:r
111
jfefi
^
[1M-
Hois RBTMOSI..]
(a.s
in
Art.
i
111.
l)
INFINITE INTEGRALS.
424
[AP.
III.
For here we have
thus for convergence it is necessary to be able to find A so that

than e for any value of p greater than A.
(/x
- A)/(/x)
is less
Hence lim #/(#)=() is necessary for convergence. But even so, no such
condition as lim {x log x)f(x) =
is necessary in general (compare Art. 9);
but it is easy to shew that if (for instance) xf(x) is monotonic, then
x\ogxf(x) must tend to 0. More generally, if <f>(x) tends steadily to x
and f(x}l<f>(x) is monotonic, then f(x)<j>(x)l<f>(x') must tend to zero thismay be proved by changing the variable from x to <(#).
[PRINGSHEIM.]
;
It is perfectly easy to
as to apply to integrals in
* so
modify all the foregoing work
which the integrand tends to infinity,
say at x = 0.
The
than
results are
1),
where
167.
To
(i)
the other
(i)
The integral
Jo
f(x)dx converges (if b
is less
provided that we can satisfy one of the conditions
where either
On
a<0
a>0
or
hand,
or
(ii)
a=
0,
/3<
1.
the integral diverges
(ii)
(),
when
/3^-l.
Examples.
illustrate the last article,
we
consider
two simple
cases.
It is easy to see that the integral converges, so far as

concerns the upper limit, by applying the tests of the last
article.
There is an apparent difficulty at the lower limit,
because of the factor
the difficulty
*
is
I/a);
but since
apparent only.
Or we may obtain the
results directly by writing
l/.r
for
.r
in the integral.
166, 167
425
intt-'n-al
tin-
dx
x
-<
'
liui
i.s
(<
l-*J
"
J
nr
Benoe
int.-
<-**
Inn
CM
But
e~*
between
lirs
"
and ly
'
l>y
'"'
tend to lo^
wliicli
ul'
H.
\alu.-s
tin-
tli->,.
g-'lno-/
both
{"
J, 5
the
ri
l>y
nid
i'..unl
valn-s an-
and
r-j,i,i
as o kendfl to
^- z
1(.
e~ tx )
1.
liin
1\-
0.
->oJ
and aceordin.
(e~
iider
/.
//,
(Ae-
J*<-
^ + (v--
///*
,
are positive and
A
It
may
sh.-wn as above
tliat
tin-
integral converges
cnnditions aiv
NOW
J3
fi
In \irtu--
!'
the r(,nditi<'n
OJ*
VJl
1J./=O,
it
is
now
B
f
(l
*
'
:=
Thu>
=
"'J 2(^
,1.1.
|
and BO
Jo
1.1-
.;
-rJ6
,-\
idrnt
that
INFINITE INTEGRALS.
426
3.
The reader can prove
2Â ra r = 0.
where
similarly that,
(k = Q,
'
1, 2,
...,
[A P.
if
III.
p > n,
1)
168. Analogue of Abel's Lemma.

If the function f(x) steadily decreases, but is always positive,
in an interval (a, b), and if \<f>(x)\ is less than a fixed number
A in the interval, then *
}
f(x)<p(x)dx<Hf(a),
i
where H, h are
the
upper and lower limits of
the integral
r*
ranges from a
as
to
b.
For, assuming that f(x)
is
differentiate,
we have
J=
since /(&) is positive and f'(x) is everywhere negative,
obtain a value greater than J by replacing x(&) an d x^)
the last expression by H, and a value less than J by
Now,
we
in
them by
Thus we find
replacing
h.
-h\J af(x)dx <J<Hf(b)-H\Jaf(x)dx

or
Similarly, if
interval (a, c)
hf(a)<J<Hf(a).
Hv /^ are
and H h
2,
the limits of the integral x()

in the interval (c, b), we find
2
n the
- H, ( f(x)da> - H2
Ja
[f(<e)da>
Jc
or
*If f(x) should be discontinuous at x = a, /(a) denotes the limit of/(#) as r
tends to a through larger values.
AN
167, 168]
\\
\l.(M,i
if
ranges from
as
\Vln-u
/'./)
of
1.
oiu-
<t
-I.
p.
To
on
i^<;
p.
is
\alue
Itet
leads
ine,,uality
writing
liy
t'oini
(gee
results
an-
to
the
\(^>
is
iMjui\'alent
that
first
not
u.-.-j.-.l
Second Theorem
and 80
ri.htiniK.us.
ween A, // at least oner

t<. Bonnet's theorem
^/('Oxtfo)' ul "' iv
the
/'
= =
'
in
the
''
du Bois Reymond's theorem, which
this
fco
ronvs|,oiilin-
th.-sr
note
see thi>,
,i--i;n
hut
+ >).>f(c).
formula.ol>tainr<l
()":
ivsjM-rtm-ly. th.-n
>>.
Thus the
''o.
to
/'('')-/('>
/yj
complex,
ineijuality
:V.t~>
p.
<
Uf()
li>07,
Value.
inter\al (o,
and from
80 can be
j)i-f>i-nt
Mean
Of
to
is
Art.
i,
iiinii
,/
|7
is
nunilM-r.
an- th- upjM-r
IMA
Mil
unction of a real variable, it

n-al <T minpl.-x. air,
li
that
foi-
OF
i;
!/('')- ff(b)
i>
monotonic
true f..ran\
^hu-
f"
we
find
comnK.nly quoted
<}>(
>(l>)l^
BOM
value of
Although the restriction

importance here.
\.-t
it
i-
,,
cannot
le determined,
ditl'erential.le
l.e
t!
h---i
.-t
ic.illy
Mee
de.-iral.le
to establish
i>
equa
of
little
-udi results
\Ve shall tln-ivfore

ni; with the greatest generality possible.
a second proof, based on one due to Pringsheim*, in whuh we
me nothing aln.ut the existei
as th
Divide the interxal

.
and wri
in;
|]
part-
l.y
inserting
j...
then we ha\e
=
where
'<K),
p.
'209;
this
paper contains a
inure general form of th.
equality on
vol.
j>.
.SO,
4-t'..
HHMi,
Another
i>.
1".
}>roof
baa been given
i.\
H.ir.iy,
MtMtny
INFINITE INTEGRALS.
428
Hence,*
if
/r+1 =/(.rr+]
III.
[AP.
),
Jr-r
(1)
but in virtue of the decreasing property of f(x\ [/(#)-/r +i]

the last integral and is less than (fr fr+i), so that
is
positive in
<(fr-fr+i)A(b-a)/n,
(2)
<j>(x)\<A and xr +i-xr =(b-a)ln.

the equations (1), bearing in mind the inequality
because
By adding up
(2),
we
see that
rxr+l
n _l
$(x)dx = Rn
J-Zff+il
J XT
r=0
(3)
Rn < 4 (6 -
where
because
fn =f(b)
is
)/
positive.
now we apply
If
-/) < ~ (b -
)(/
Abel's
Lemma
(Art. 23) to the
sum
r+l
r=0
we
obtain the limits
and
hf-
.fi^ for
and the sequence /i, /2 ...,/

Thus, from (3), we find
,
because
f
is
decreasing.
A/i-^-)/P<^<^/i+^-)/f
(4)
where
/! =/[
now we
If
it,
T^r+l
+ (5 -
)/]
take the limit of (4) as n tends to infinity,
we
obtain the
desired result, t
In exactly the same way we can make the further inference that if
between a,
lies
b,
and
ranges from a
as
In case
/(a;)
to
if
c,
H^
A x are the upper and lower limits of I
while ZT2
h 2 are those as f ranges from
should be discontinuous at xr +\,
we
define
of f(x) when x approaches ar r+ i through smaller values of

exist in virtue of the monotonic property of /(.<)
will be seen that the condition
fit
|0(#)|<-4
is
fr +\
^ (.r )././
c to 6, then
as the limit
this limit will
by no means
essential,
and
t't>
that
verges
/*
may
it
;
for
be broken at an infinity of points, provided that
we can then make a
\<j>(x)\dx con-
division into sub-intervals, for each of whirh
-.
than any assigned number. But Pringsheiin has proved

/
that it is only necessary to assume that (f>(x) and f(x) x (x) are integrablc in
the interval (a, l>)
compare Proc. Loud. Math. Soc. vol. 6, 1907, p. tii?.
\<f>(x)\dx is less
429
168, 169
Tests of convergence in general.
169.
in- the
genera]
AppK
fral
be
and
<,////
tl
for
we oam find
-!n,t
/i
Convergence
<>f
tin-
nieh
j
T
/in
"//'/ e 18 ni'ftifi'"
'//.
HoW( vet
'
.
in
usually r-|laci-l
which can
tin- t'ollo\\
l>c
more
a]>plic<l
.--ral
inlinit.
I'm-
ly BOm
The thn-.-
iff
j.i'act
<|uickly.
test for con-
DMTO
chief
are
t'-sts
iii
Absolute convergence.*
1.
The
will
literal
certainly converge
it'
I/O/
'
.'
converges,
<!>'
.<'
Jf
i'.iit
naturally
analogy h.-twccn Midi
tin-
not
<|uitc
convergeni aeriea
the values of a function.
is
lut.-ly
intcî-als
c(,ni])l-tr.
since
and absois no
th.-r.-
in
In particular,
if
|/(#)|<
//'(.').
:icivascs. tin- integral
Readily decreases
wli--
/'(')'/./
will coii\>
2.
Abel's test.
Infinite
integral wh
and
Suppose
m,t absolutely)
'
//N>-
that
,/,(.
>e).
/)'/./
;id
that
inono-
function, such
tonic
\lr(.i'}
''hinnjii
that
*fr(x)
<.l:
it
i>
th-
tit
that
tt-nds t. sonic limit
Thus,
i-/
we
if
if
wri
-.-.
tnd
thai
ahsohiti-
'I'lit
pointed out
l.y B(
-.
ami
/'A
is
clearly
INFINITE INTEGRALS.
430
decreases to
and
it
[AP.
obviously sufficient
is
to
III.
prove the
f(x)<j>(x)dx.
Joo
a
Now,
by the analogue of Abel's
is
(Art. 168),
we
see that
< Hf(() < Hf(a),
IJ1
where
Lemma
the upper limit to

x
<j>(x)dx
when
of
X ranges from
<j)(x)dx,
to
Now,
'.
we can determine
in virtue of the convergence

so as to
make
and then
f/<
so that the integral
Hence
also
f(x)<f)(x)dx converges.
f
I
Ja
3.
\}r(x)<t>(x)dx converges.
Dirichlet's test.
An
infinite integral which oscillates finitely becomes convergent after the insertion of a monotonic factor which tends
to zero as a limit.
Here again we have

I
and
f^'
J
f(x)$(x)dx
H will be less than some fixed
thus, since f(x)-> 0,
we can
< Hf(),
constant independent of ^; *
so that
e, and con-
#/()<
/QO
sequently the integral

Although the
find
Ja
f(x)(h(x)dx
tests (2), (3) are
is
convergent.
almost immediately suggested by the tests
of Arts. 19, 20, yet it is not clear that they were ever
given, in a complete
form, until recently. Stokes (in 1847) was certainly aware of the theorem
(3) in the case <(#) = sin .r (Math, and Phys. Papers, vol. 1, p. 275), but he
makes no reference to any extension, nor does he indicate his method of
The first general statements and proofs seem to be due to Hardy

proof.
his argument is somewhat
(Messenger of Maths., vol. 30, 1901, p. 187)
diflerent from the foregoing, and is on the lines of the following treatment
of the special case <(.?;) = sin ./.
;
Since the integral
number G
<f>(x)dx oscillates finitely it
for all values of
thus
we have
remains
less
than some fixed
MI;K
\\i.
169
III.L;
In tin-
i>.
--sci
i.ly
Hates between the two

the figure.
in
almost
-\
intuitively
decrease in value, and have
that
id.-nt
tin-
aivas
tin-
.,f
\\aves steadily
in
alt-
and
.sim-e
-in
fl
TT)
positive in tin- integral,
so that
to
U-t
oo
Further,
obviously negative and nuinerically less than the area of the
i>
wave.
luiis
vdx--Jf(x+**+lir)KD
which
lies
tliis
tends to zero as
it
It
f*
follows that
sin
is
c.n\ t-rgent,
by applying
th- th.-MUMii of Art. 21.
In general,
method.
If
if
usiiiL'
tin-
changes sign infinitely often we can apply a similar

t->t
(An. ^>) to establisli th.- ronvergence of
<f>(.r)
i>iriehlet'>
intro-ranil
tends
convergence and
lor
Th>
tin'/
'!/
tin-
JO,
tr-.t
.!'
-ay as .r-*a, the general t-M.

als(lut- convergence run as
././
flu
/'"/
8UJj
is
<'<>n>; r< jence

if
fix
><!,
fl.-
Ib
ivherc
///.-
any
;>ox//?>v
*/
I
,ii
/'.>
|/
/A.
'//.
It
in
is
i>M>vil,l,.
to
write
and Dirichlet's tests ;

practice and an- (oetl
Altai's
nut
con-.^jioinliii^-
l>ut
such
t.-sts
to tin-
nioliticat inns
are not
ingenuity
of
ci
<>!'
tlu
reader.
432
INFINITE INTEGRALS.
Ex.
1.
.-dx
__ctr,
converge absolutely
^<foif 0<<1
1(
(
Jo
/oo
Ex.
2.
-^^ d# con verge

</>(#)
sin .rcr,
converge in virtue of Dirichlet's

Cb
For
Ex.
sin x dx
cos
cos b
if
test,
and so does
-"*, (logj?)-*
and generally
cos tfcfo?
tends steadily to zero.

I
cos x
dx
sin b
test are given
= (sina?)-*,
<^(d7)
- sin
|
2.
Ja
Further examples of Dirichlet's

/(*) =
0<p<l
<j>(x)
fb
/
2,
|
3.
if
</>(#)
7a
p>
/*oo
*-^.(lx,
if
because |sin#|^l, |cos#|^l.
x^
III.
[AP.
by
[HARDY,
log(4cos%).
I.e.]
170. Frullani's integrals.
As a simple and
interesting example
of the
tests
of the
last article, let us consider the value of
where
<f>(x) is
such that
pco
<j>(x)dx oscillates
between
finite limits
{or converges).
Then, by applying Dirichlet's or Abel's
test,
we
see that
x
is
convergent and
equal to
is
THUS
and
if
'
if
a "5
>-
X
<f>(x)
tends to a definite finite limit
the last integral has a finite range and a

thus we have
In the same
way we
can prove that
if
For examples, take 0(.) = coso5 or
sina;.
as x tends to
finite
0,
integrand;
170, 171
Tin-
foniH-r
in!'
<'\aliiat'-<|
IT
tends to
/
>
,j,(
M x tend
is
means
by
Tli.'
of
\\hi<-h
and
Hai'
a U..
i<l
>n8
l.y
i.r
th.-
l-Yullani's
.f
ill
|ii7
i>
particular case of
Uniform convergence of an
171.
be
may
integral
thai
<>f
\.-ilu.-
AM.
in
.iiml
f.
integral
formula
03
J.
infinite integral
i
tin-
lue
wliidi
lor
!'
^-\
^^
tin-
in.-(|iiality
<6,
(\>$),
JA
Art.
witli
function
Lfi
we Bay
!:).
interval (a, 3),
than
pendent of
tin-
all
vahu-s
A' if),
if
T.iit.
iliat
fr
i'linction
//.
which
interval
it'
as
w.-ll
.f
intf^r.al
.f
wliicli
//
converges uniformly in
in tlir intn-val
d.-pt-nds
on
but
r-iii,
is
inde-
t!
]>art
vahif //,
icuK-n-
to be a point of non-uniform convergence

Ex.
-nee
In
e.
tin-
tin
:.
int-^ral.
1.
(if
-"
tin-
kc
unit'
i-
intcî-.-il
-\'u
t.
Ex.
0,
^=
II flirt'
is
>/
=0.
.rnJ.
/via/,
v,.l.
:;:?.
I'.nu.
,,.
11:;.
3e
ilM
*.
l.s.
t.
Papers, vd.
I.
j..
INFINITE INTEGRALS.
434
[AP.
III.
we have usually in practice

uniform convergence which is similar to
the general test for convergence (Art. 169): The necessary and
sufficient condition for the uniform convergence of the integral
But, just as in the case of series,
to introduce a test for
r
I
Ja
f(x, y)dx,
a value of
in an interval of valuer of
independent of
y,
f(x,
where
'
we can find
y, is that
such that
y)dx
e,
has any value greater than f
and
is arbitrarily
small.
The only fresh point introduced is seen to be the fact that

must be independent of y. The proof that this condition is
both necessary and sufficient follows precisely on the lines of
Art. 43, with mere verbal alterations.
But in practical work we need more special tests which can be
applied more quickly the three most useful of these tests are
:
1.
Weierstrass's test.
Sup2)ose that for all values of y in the interval
function
the
(a, /3),
f(x, y) satisfies the condition
where M(x)
the integral
is
I
Ja
\f(x,y)\<M(x),
a positive function, independent
M(x)dx
converges, the integral
of y.
I
Ja
Then, if
f(x,
y)dx
is
absolutely and uniformly convergent for all values of y

the interval (a, /3).
For then we can choose

is less
M(x)dx
than
independently of
y,
so
that
and therefore
e;
Joe
J\
f(x,
y)dx
< M (x)dx <

Jf
Jf
Thus the
integral converges uniformly;

absolutely in virtue of Art. 169, (1).
2.
and
e.
it
con\
Abel's test.*
The integral
interval
*
<
M(x)dx
(a,
/ft),
Ja
uniformly convergent in an
f(x, y)</>(x)dx is
provided that
Hromwich, Pro?. Loud. Math. SW.
p/
<j>(x)dx
converges,
,<
(:>),
vol.
1,
1903,
j>.
-JO
I.
and
thxf,
171
1
as x increas'
<es
.'/)
then, in virtue
where
!*/(.
//
j.
.,i
the analogue to Abel's
<
y)#()*
the upper
Lemma, we
<///'.
liinii
r)dx
eonv<
ia
we can
independently of
timl
iven
i'leut
that
that
y,
so
that
<
//
int.
when
e J\
and
:y.
liy 0<.''.
//
),
pi-'A i<le,l
r^rni in the interval
<
i
(a, ft).
-'
Dirichlet's test.
Ja
i*val (a, ft)
and
if
and
then
wh'-iv
then
//
is
tix
/V.vy/rr/
Ex.
.l'
//.
//.
remain
'<**
3.
Ex.
ft).
< lff((, y),
to sati^l
provide*! that
.'it
the
<
e //.
uie limits
ti
iut-
can
of
,
,-?).
/ef.
uniformly tlii'-u^hout any
inU-rvjil
><,
of variation
4.
;iiiiormly
in
any interval (O^yl^J), because the integrals
raf*
in
virtin-
"f
Ait.
1-
f(
convi-r::*'
creases as
:ileut
0(. /)'/./
>
fj/'it
;/
91
indepen.leutly
contain
Again, ^1 c) may
I
/,,
/./
tli.iii
8tt
*
'
[ess
I,.
consta/T(
h-jit
'I'! f/i
'/'//<///
'
F..r
of
iltiffs
$(.'
fl"
nil
tO Z(
INFINITE INTEGRALS.
Generally
e~ xy $(x)dx converges uniformly in
[AP.
any similar
ill.
interval, pro-
&(x)dx converges.
Ex.
5.
Dirichlefs
test.
COB *
r
Jl
dx,
(.^+y2)i-
sin *
h (rf+y^^x
converge uniformly throughout any interval of variation of
p 1+X
cos
And
("iZMd*
fîf)^,
X
1+Z
(f)cfo.
2
Jl
in
converge uniformly
Of course the
any
y.
Jl
Jl
interval which does not include
definition of,
and the
tests for,
y = Q.
uniform con-
vergence can be modified at once so as to refer to the second
type of infinite integrals.

Applications of uniform convergence.
172.
f(x,
a
Joo
y)dx which converges uniformly in an
has properties strictly analogous to those of

and the proofs
uniformly convergent series (Arts. 45, 46)
can be carried out on exactly the same lines. Thus we find
interval (a,
/3)
1.
If f(x, y)
(a, /3),
that
it
is
a continuous function of y in the interval

a continuous function of y, provided
the integral is also
converges uniformly in the interval
(a, /?).*
alterations are needed in the proof of the correfor series (see Art. 45).
theorem
sponding
Only verbal
Ex.
1.
Thus
(see Ex.
i
3,
Art. 171)
f !^><*,, f ^>^ f "gg*,,
are continuous functions of
in
any
interval.
/oo
Ex.
2.
If
3.
But we
<f>(x)dx is convergent,
then (see Ex.
4,
Art. 171)
[DIMCHI.KT.]
Ex.
and
it
is
Stokes,
must not anticipate the continuity at
y=
of
not hard to see that they are actually discontinuous.

(See Ex. 5, Art. 171, and Ex. 6, Art.
I.e.,
\>.
173.)
171, 172]
fin/'
L'.
n-itl,
fin
it
tin-
t<>
respect
-\iruKM CONVKl:cu
vn
<
<jrate
//,
//
th>
prov*
ft).
A -Min,
tli.-
proof
f.,r Beiiefl
M ..... Is
only
v.-rhal
changes.
^
-j
equation
'
'provided tlmf
'I/id,
uniformly <nt
write
ami
,/".-.
fl
infri/îf
on
tl<
n,,
////*/////
tin'
Irft
right
/'x
converges
convert)'
/= f [/(. y +*)-/(.
y)_
A
JaL
ay
;
us liml
Irt
\vli<
for
;-
be
.ill
cni-iv.-t
if
f/nif
tin
all
that
BO
of
indeprii<l-iit
values
of
//
in
y,
the
and
the
interval
in.Mjuality
(a,
Then,
/3).
X>(,
\
the
^
value
the
of
doiihle
inteiri-al
of
function, taken over a fin He area, is inl.-j)en<lent of the order

of integration (see Art. lti:i. j.. H()I.
Thus.
</
{"',[/<'.
+ /<)-/<.'-.'/>]'.'
<
virtue of our eh<
in
.M-Y>(/0 =
80 that
|0(A'.
'1'hr
the case
hist
<>
condiiion
i.s
|>
p.
119 for
INFINITE INTEGKALS.
438
The
and
values of h under a certain limit.

|
to oo
we
greater than
tend
we make
If
obtain
is independent
Since our choice of
allow h to tend to zero without changing
we have
thus
last inequality holds for all values of
all
HI
[AP.
we can now
of
h,
and,
by
definition,
lim| ^>(oo, 7t)|^2e.
Since e is arbitrarily small, and 0(oo

this inequality can only be true if
is
h)
independent of
e,
lim
/t->0
or
Another theorem may be mentioned

little beyond our scope.
here, although the ideas involved
are a
,-oo
//",
ê
integral F(z)
Ja
f(x, z)dx, the function f(x,
an analytic
is
z)
function of the complex variable z at all points of a certain region T of the

z-plane, then F(z) is analytic within T, provided that a real positive function
M(x) can
be
found which makes

I
5.
<
M(x) at
To shew the need
convergence,
we may
the
integral
condition
satisfies the
Ex.
'TV
all points
M(x)dx
y under the
JO
integral sign,
6.
On
we
find
the other hand, the equations
f*
x&i
Jo
are quite correct.

4.
(See Exs.
1, 6,
Art. 173.)
The analogue of Tannery's theorem {Art.
{An
49).
r*>
f(x,
n)dx= J a g(x)dx
\
if
this
is
differ-
r<a
which does not converge.

Ex.
and
as that of uniform
sm ^' dx
convergent
of T.
some condition such
for
consider the integral
entiated with respect to
Ja
cos(jcy)dx,
172]
uniformly in
<
.i/>
of n,
...
'*
for a
c),
be chosen so
let
is
large
<//!/-."
will
-</
Ex.
/""
.'u
--5
;in
'
ml.-.
necessary, consider
\vitliMtit
TT'
-f
tin-
t-
//
(;is
B*
T<>
7.
integral
tip-
uniformly
*)=o,
;i
liml
i-
tin-
pp!y
limit
.T
limit
:rul :i]iji-
imt
TT.
The second type of

Th-
i.
;!<![
littU- ditlimlty in stating and

above, fn- tin- second tyj..
l)-( 4 )
Tln-rc
is
infinite integrals.
find
slu.ulil
only one case
some difficulty
tlu> ia
'
<
in
\vlii.-li
j/,
tin-
upper
alt In MIL'
limit
van.-.s
with
y,
= b(y)>
and
i>
.uity
p-ii
7^f
/"fr
ix
unif..indy
vab;
!,i,-h
//(.</)
fniind
\V.
that
In-
the problem of differentiating an integral of
F(.
for
may
\\liu-h
ivmains
lin.l
a rmiMant
tlian
1,-ss
\ve
ry siinplr rxainplo
is
^=
ire
and
)='.
lues of y.
constai
<!?!
!/(-..")-'('-
-^'
gi\
'j{x(y-z)}dx,
>ay,
y>0.
have
all
and
440
we
INFINITE INTEGRALS.
have,
and
by the ordinary theorem
for differentiating
using the inequalities which define
so,
[AP. HI.
8,
we
an integral,
find
Also,
so that
and
using a double integral as on
so,
In the
tend to
last inequality, let
we
see that
p.
437,
0,
and <(, y} then tends
to
F(y\
so that
Thus we
see that
<2e +
Take the
hand tends
A tends to zero then the rightindependent of h thus we find
limit of the last inequality as

to 2c, because 8
is
and
by the same argument
so
as before,
we have
173. Applications of Art. 172.
Ex.
1.
where
a,
Consider
may
first
the integral
be complex, provided that they have their real pairs
positive or zero.
Then
Now
is
uniformly convergent for all positive or zero values of

with respect to y. We obtain
IY
hv\
and
this integral converges uniformly so long as y

therefore equal to dJjdy^ in virtue of Art. 172 (3).
*
y*
differentiate
It is understood that neither
a nor
b is zero.
= l>0.
Its value
is
8FE4 EAL
172, 173]
vi
111,!./
'
.'"'-I.
IT:.',
lind
w.-
/r
> ,~-o'' X
H."./
int.-ial '-an
la-t
Tli.-
v;iliiri|
same
\\\-
being of
!'
n.'itui.-
t'--i
lli-
..!'
,-*>
ii
!...;(/.
has the advantage
.-
l.-xMi-itlnn
i'f
to tin-
iiuml>-r,
.-Minj.l.A
.I
:IJ|M-;I|
\v-
if
.i-tii-iilar.
\\i-itr
''.
\\-
^.r^-^
-CB.,)^.^
Ex.
\\liri.-
2.
I'.-l
Ex.
ami
By
3.
iiit-.i,
in
iraj
that
"
parts of
real
tli,>
diivrt
can prove
\\r
(Jfiirrally,
i-ati.ii
.r zero.
rninltim-il
with (1)
it
will
!>
found that
L+(a+c)
tin-t
l<i:a!-itliiii
is
dcti-riniiifd a-
>
are
in --at
!'"!
-\aiiipl-,
if
we
taki-
a=
get
l,
6=
^"[r-{l+(H-t>}-^
1.
fr)-cosa:]^=|-].
As
annth.-i-
illustr.v
6=
Thr,,
w-
.(,'[("
Ex.
4.
1,
=-0/-fi).
tilld
')
i>
It
.-.tsy
ii-ly
that
l.o,
I'.l,
i.
1\ !._,(>, and
tin-
real
parts of a,
/<.
or
442
INFINITE INTEGRALS.
Ex.
By
5.
differentiating
J=
^/^[fa
[Q
[AP.
twice,
=,
we
ill.
find that
ify>0;
the last result following from (1) above.
Hence, since linit/=0, and
remains
finite as
y-M>
Art. 172),
we
find
y tends
to oo (see Ex.
1,
J=$ir(l-e-).
we
Thus, on differentiating,
find, if
is
positive,
When y is negative we find J=\Tt(ey - 1), and so the other integrals

become \TT&, -\ictf, respectively.
Thus </and the cosine integral are continuous at y = 0. But the third
integral
is
discontinuous there (see Exs.
to the sign of
Ex.
in
which
|
f(x)
has the value
^TT,
according
and vanishes for y=0.
?/,
As an example
6.
sin (xy)
Art. 172).
1, 3,
,"<x>
In like manner, the întegral
is
of Tannery's theorem,
supposed
less
than the constant
we
take the integral
H in the interval
(0, b).
Here
so that
^-l
.v
from
this result following
can take M(x)
any fixed interval for

x tends
Ex.
then
7.
x.
to
It follows at once
linit/n
n
In applying Tannery's theorem we

to the limit /(O) uniformly
understood here that /(O) denotes the
(1) above.
H(^v[^x)lx
limit of f(x) as
--
r sin 2 .rT
f sin2a;, = TT
dx -,
dx=\L --x Jo +/Jo
x
2
Jo
in
00
rsin 2 .r,
For
and f(x/n) tends

It
is
through positive values.
from
(6) that if
ao
The reader should prove that
if
&
= TT,
this result
must be replaced by
The integral Jn is interesting on account of an application to Fourier

Series given by Fej6r (see Ex. 5, Ch. XI.).
174.
1.1
L74|
Some
further theorems on integrals containing another
variable.
Tan
as
in
related
x incf
ia
hen
(A,,
a
/
that
.1
thru
liin \
/.,i'
"/'
</(c
the upper limit
is
"
//
</<''
" )//
< -I//,
as
to
'
from
[I
ft]
can find
infinity, w.-
0)
Art.
\-
i
//
nilin-s
,i/l
(-
run verges when extended to

mak.- ,l//<e: and tli^n
//
i
j
< J // <
e,
because, as
,r
increase
we
Coiisetjlieniiy
and does
nc'
tind
<ê4
/r|.
PO 1-y
fixed tlic limit
is
Art.
lin.
*
<>t'
.iiid
tin- last
sc
we
int.-Tal as
lia\-
>
It follows that
this
maximum
limit
fp<
I
it
or
,,
tmd-
444
INFINITE INTEGRALS.
[AP.
ill.
Dirichlet's first integral.
As an
application of this theorem, consider the integral

sin
If
we change
nx
'0
the variable of integration to x/n,

w6
sin
/./#\
JTn =\f /(-)-
a?
we have
dx.
Hence, if f(x) is positive and never increases, our theorem

can be applied, because*
/(O)
and f(x/n) tends to the
Hence
lim
wôo
a/(a/) >0,
limit /(O) uniformly in
Jn =/(0)
in virtue of Art. 173, Ex.
^HL?
Joo
35
any
fixed interval.
above.
1,
remove the conditions from the

easy
function f(x) of being positive and never increasing.
Suppose,
for example, that f(x) first decreases in the interval (0, c), and
It
is,
to
however,
afterwards increases in the interval

functions F(x), G(x) defined
and
(c,
b).
Now
consider the
by
F(x)=
0(x)=f(c)+A-f(x),
A f(b)
is a constant such that f(c) -f A and f(c)
where
Then the conditions of being positive and
are both positive.
never increasing are satisfied by both F(x) and G(x)} so that
hm
n-*-oo
Jo
i-
rr/ xSin?i#
^^
F(x)- ^ -dx = ^F(0),
TT
But F(x)
with a similar equation for 0(x).

b
T
hm
n--ao
JO
S/
^^
Si n
f(x)-
G(x)=f(x\
so
th.it
^/rk\
-dx= T9-f(Q).
7
can be at once extended

where f(x) has a limited number of maxima
It is easy to see that this result
case
any
and minima and no
to
infinities between
Here we use /(O) to denote the
positive values
b.
x approaches
through
monotonic property of /i
limit of /(a?) as
this limit exists in virtue of the
and
DIRICB
174
Dirichlet's second integral.
Coiisnl.-r
(0<6 <1P)
'*'
JO
"li'-
!/=:>//"
-/()
'
Sill
x/aiux steadily increases and
interval (0.
forth
that
/<.'
/).
in
follows
it
'
lias
no iniinity
in
Hm
Bence
wil
</>(./)
)iru-hl-t's
first
inu-^ral.
provided
= \TT^
I\ H
o).
II
may
however, the ran^\\
hit
tin-
rite
<>t'
in
f^ral
extends up to
int. .-^ration
tin-
TT,
form
/f*"
(I
and
tlion
the
ves
sl
Ex.
1.
variall'
"
~"
I"
the
in
s.-c.uid
^-"
"sin-
lim
\Slllltf5
f''
+1
f,
clian^v
Bence
part to
>;
'
)J
i\,- >,/.,--
As
n.it
h...-1-Sooi
(*"
i-r
Jo
.
/I
whirh
a.u'ivr
\\itli
t!
we can
tlu'ii
\\
')
while
is
tlif
th
ât isfiee
tliat
with
lfaliii--
are positiv.- ;uul
IM
in
ti
;i
j
OK
we
INFINITE INTEGRALS.
446
Ex.
It
2.
is
III.
instructive to investigate the value of the integral

'-dx,
sin.v
by means
[AP.
of the curve
?/
= {sin(2?i + l).#}/sin#.
This curve
is
of the
same
general type as the one given in Fig. 46, Art. 169 except that the initial
ordinate is y = 2^ + l and that the points of crossing the axis are
;
Then, using the argument given there, we see that the value of the
integral
Kn
where k
is
is
expressed by a
an integer such that
Hence
differs
any integer
_
lies
between (k-l)ir and
than k\ the value of the integral
less
#/i/,
and
where
= 2n+l.
the integral
Kn
r is
^"n lies between
to
oo
we
find that the limit of
between*
lies
and
x
where
Kn
.to,
we make w tend
If
and
+ ( _ iy-1^
Thus, changing the variable to
for,
> vk > 0.
-2
^+^ _
VQ
than v r
less
by
r is
(if
from
+ l)
(2?i
>V >V >
the form
finite series of
any positive
Thus
integer.
00
,.
sin^7
Ex.
3.
It
is
TT
-^=2'
JOB'.-/,,
[DlRICHLET.]
easy to see (as in Ex. 2 or otherwise) that

,.
lim
bn
sin(2w
+ l)^
'dx= r^sin^,
,
sm.r
Jo
x dx,
where A = lim (2w6 M ). The maxima of this integral are given by A = TT,
... and the minima by A = 27r, 4?r, 677,
37r,
STT,
Glaisher (Phil. Trans.,

numerical values for the
vol.
160, 1870,
p.
387) has given the following
maxima and minima, where

sin x
sin x
- dx,
=
/r
dx = ~- /"'"
2
Jo
Jm X
71= -0-28114,
73 = -0-10397,
76 = -0'06317,
72 = +0-15264,
74 = +0-07864,
7 = +0'05276.
7
/*
fl
Thus the greatest value of the integral is \TT 7j = 1*85194, and the
least (if A>TT) is \TT- /2 =1 '41816.
(See also Ex. 12, p. 351.)
the inequalities proved in Art. 59 and in the footnote p.
*From
see that (since x\v
I*.
Thug
and so
sin
a;
184,
we
< i?r),
sin
a;
v sin (xfv)
\x*
x\
|sinx|
a;
Zv
sin (x/v)
this difference tends to zero uniformly in the interval
0^.r^(r+ l)?r
174]
Ex.
decreases in such a wa>
4.
wr
i.iiv. -r^'rnt,
and
--an
pro
...].
In
>0), the
nl.M,
|,.irti
Ex. 5. Sh-\v that

and <)<< 1, tli.-n
first
linii'.
ronditi'
liin
Jo
w-*
'
lim/,'/
n+x
J*
&
'
til."
..f
Jordan's extension of Dirichlet's integral.
n
Sup|<sL- that
never
decreases
.-=
!IT
'''.
\\
(().
i>i
poeitiv*
Jo
in
tin-
interval
whilr
rties
<A
(i)
n^-^6;
\\}
(ii)
lini
'<'-
if
Jo
\vh.-iv
ae
.1
and
a COnfltanl
i-
is
taking th- limit
in
ti\.-d
ii
= f^6,
rded
arl>i'
I'ndi-r
).
lii..
\vh.-iv
d.-nntrs
/'(())
Km-
\vr
liavi-
tli<-
"n
liinii
from Art.
his.
< <
r
j|'
j,
444.
/>,
(/.-/OlAOj-MOl + AMOX /,<(//-//.

where
//.
an-
tli.-
u})]n-
and
lo\\cr limits of
)]+//'I
Jo
to
bom
(i)
bhal
if
tj
is
A-A' <*A,
//-//
|/-V(0)'<2.l|.'
tin'
irn-att-r
//'
and
A'.
<iM.
''
v
6.
>
**
INFINITE INTEGRALS.
448
Now
choose
c so
c,
make n tend
it
follows that
that 2J.[/(0)~ /(c)]

Since lim
to infinity.
< e;
r\
n >oo
[AP.
III.
and having fixed

by condition (ii),
lim/n =
Hence
This result can be at once extended to any function f(x) of the

type considered in dealing with Dirichlet's integral (see p. 444).
Integration of series, when infinities of the intein the range.

occur
grand
It is obvious that infinite integrals are excluded from the
175.
discussion of Art. 45
one case of practical importance presents

the terms of the series are of the form (p(x)fn (x),
where <j>(x)-*oc at, say, the upper limit b. Then we can easily
:
when
itself
establish the following result
A. If
2/w (o;)
converges uniformly in the interval
(a, b)
and
b
|
0(05)
dx
convergent (and has the value J), then
is
-2
n (x)dx.
</>(x)f
J\a
For then we can find m, independently

it
m /()]<*
b
Thus
t
m
oo
It follows that
At
<t>(x)fn (x)dx
/
Jo,
the same time
< J^
a
p>m.
(#)
dx=*J.
d>(x}fn (x)dx converges, and that
we have
-"2
Q
f
Ja
*</>(')/(<)<**
find
-I
and, since
so that
Cb
I
so that
Tims we
of #,
t/ is fixed,
-'a
we can determine
to
make
27
as small
as
\\-
174, 175]
th'M.-fc.iv
!"
iiiu-t
the h-ft-hand >ide
is
independent of m,
^)i/.w*r-SVw
Thai
Ex.
440
in the last im-ijuality
but
and
This caso
1.
illustrated
is
l,y
.SC-lJr-^^k
Here
th-
-r
uniformly from
roiiv.-rurrs
log(l+./v
to
but
-oo as
On
other liand.
tin-
y also
to be uniformly convergent) as
se
\\
CM n
the
th.
often justify term-by-term
oivm
/us
r
flu>
wln-n
integration
tliis
by
may
liappens,
means of
Suppose that
B.
Unit
t.-nd to oo (or it
-l>\
<
</>(x)
fn (x} "'' r
>i>tfral
is positive in
''H
positive,
interval
////
/o.
(<>.
'
convergence of
fb
'-'',
.
n
or
21
II
<
)'/.'',
J
///
..lvinis that
|in.\vd
i-
Imth
that
<llhr of
-..nditions are
th-m
In
necessary: the only point to be
i-
W,
tli.-n,
/' /'//'-''<
|/
-it
<tses
iM-.u'.'itiv.-,
as S t.-inU
:iat
if
riih-r
liiu
a-*o
.f
-'
and
/.-n.
t<.
lim
thr
..tlirr.
the
lim
t"
tin-
fuin-tii'M
Thus, as
infinity.
iiit.-i
li.n
m-*
and
\al
|lim/\(S,
U-*o
tht-ir
values are equal.
(a, />-6),
so that
..(1)
lil
lm-H
[.8,
Ait. 31(5),
m)\
5_M)
/"
in
the iT]K';itfd limits
In.
I
and
//<
450
INFINITE INTEGRALS.
Similarly the other repeated limit
is
established
for
III.
seen to be equal to the series
is
(*)/()<** ......
Hence the theorem
[AP.
if
............................ (2)
either the integral (1) or the

and their values are equal.
series (2) is convergent, so also is the other,
Ex.
An
2.
application of
Plog(l-*)
V(l-*)
is
is
rJW(l-*)~
easily verified
given by the equation
4/2
2.4 2.4.6
+ +
+
+
*dx
gfi
Jo
This result
Theorem
3\
directly
is
found to be -4, and the sum of the
Ex.
2,
p.
where
3.
The
and
if
C.
Another example
B
p=0
and Theorem
special case
p= -\
When
is
expanding
#) instead of log(l -#).
!/>/(!
given by
Here we use Theorem

to include x=l.
jt?+l is positive.
interval
parts, the
series in brackets is 3
42).
We get another illustration by

Ex.
"7'
by integration by
integral
(see
57779
to include
x=Q
in the
gives
the integral can also be evaluated in finite terms.
the terms
we can apply
are not all positive,

case either the integral
fn (x)
a similar argument in
b
Ja
\<l>(x)\{2\fn(x)\}dx or the series S'P|

Ja
.
|
|/
Here we write
converges*
and then, under either of the given conditions, Theorem B can

be applied to each term on the right-hand side.
Of course
these conditions are easily seen to be sufficient only and not
necessary here.
Thus, for example,

n (x)
*
vol.
= (-I)
n-1
xn
if
a = 0,
ft=l,
vol. 35, 1905, p. 126; Bromwich, ibid.,

It should be noticed that the argument fails if we only
Hardy, Messenger of Mathematics,

36, 1906, p.
know
that
1.
fb
/
<f>(x)[Zfn (r)]dx is absolutely convergent.
175]
we
VI
iJ-i; VI
!-.!+..= r
i-'.-f
althnu-1,
Ex.
1,,-re
To
4.
2|/(aj)|=
H -
''
And
'/./
-r
diverges.
Jo
Theorem
il
(see Arte
.I'-~r
.'
('.
-..
-iy-'.'-'
Here we
diti"iis
ik-
.(.r)|
(1
But Theorems B, C
which
caaefl
d<>
in
thi'iii^-lvi-^
|ii-.-<mt
t>
Now
an =
].V
'I'll'
"i-
BO
In
as
-1
}
:
(\ -.<)-
+ 1>0.
(i-.'-y-',
is
.Ms,,
in
put
1
and Theorem
j.la<-t-
of
<\+, -)-^,,
8.
(1 -fa:)
.f
./
,
on
tlir
i-iu'i
tlii>
it'
is
/)
n,
th.-n
tin-
s-rii-<
the \alu-
n-dmvs
..f
the
to
-log:>,
intr^ral
and
tluis
it
mu
To
illustrate the result given
\\
p.
H>0);
this
is
and
tind
thai
6.
<
,--./..)<";
Continued.
Ex.
the
in
'.
];iiti'-i:l:ir,
verified
if
MI.1A--1.
.-...ili.-i.-nt
'I'll.'
iiinl
integral diverges
'Ill
pn.c.-.-d
/'
tin-
^]
}>ow.-i--
:i-n useful;
n
\>t
_
n \x
/(
''
If
with
.li-alin^
Oonrider the integral
because
and do not
5.
gives
number of comparat
nut sutlicr in a
any really --.-iifral tln-nn-m. In

BerieS, the n-mark made at the top of p.
and in some cases we can apply The. -ivm ('
lim
and tln-n
n '/.,).
taking X to !
Ex.
COWK' und'-r
following example
aid th.-n the con-
n particular, />
simple
=1
Caresati
..I
valid,
beoauae the resulting sen
rges.
is
easily
INFINITE INTEGEALS.
452
Ex.
are given
\_
2^-1
Further applications of Theorems B,
7.
P(lpg^)^
1 +
l'
From
we can
Jo
-C
the last integral
J^ +
By changing
8.
by the equations
obtain the results
or =
Ex.
III.
^"^
-)-
yr
or = 0,
we
[AP.
- 7T 2
if
# = 0,
if
< 0.
the variable in Ex. 6 from
to
where
.r
= e~
find
cosh (bt)
r=
I
Jo
,.
irrv\^--r
2a
\2a
cosh(atf)
Similarly, starting
from the equation
which can be established by Theorem B, we find

/TT&\
rsinh(&0v/#= TT
_tan(
),
2a
\2a/
smh(aO
v
Jo
176.
Integration
of
an
series
infinite
over
an
infinite
interval.
The method of proof employed in Art. 45 does not justify

the deduction of the equation
r *<*
Ja
from the knowledge that 2/

all values of x greater than
7i
(#) is
uniformly convergent for
fao
o
|^(0)'|(2(B
is
con-
175 can
vergent, the method used to prove Theorem A of Art.
be at once modified to establish the desired result.
But
it is
often necessary to justify the equation

is
r\<j>(x)\dx
divergent or
2 fn (x)
when
either
can only be proved to con-
175, 176
INTEGRATION
tli'-n
new
sunn-
test
be
Thus
t'--I-
\ai,i|!<
;(.<)
..
(.r)=0;
.-f
maximum
tin-
limit
its
and
i:*
nil'onnly
iut r-ihn-M|.
must
and
Ol
in
-rês uniformly to
interval
any
S.
t'oi-
= 1,
'
BO tint
lim
$,(*)<&?-!,
-'JO
and
tin-
same a
tin-
DOl
ifl
[lim
'
H-KB
This
ilia
,,
<^>(^)
tin- Mtlin-
ari>-s in tin-
liili<-iilty
dor
|
diverges (because <(#)=!)
such as the expOOentuU
ijiti-irrati"!'
whirli oOOVergee nnifirmly in any ti\.-d int.-i-val (wliidi

r
((.nv.-i'p- uniformly in an infinite interval.
arbitrarily L ivat l.ir
)
B.
Many
follow in--
cases
<>f
//'
impoi-tance are covcn-.l
j.r.^-tical
brndy
/'//
cvnyfia
is con!' u'linut for all
II-/K /<
if (f)(x)
x
tl
ly
it-
'
Of
= b,
''
f\.
X,
</>(x)fn (x)dx;
I
provided flif
'
'
./
/-/'////
/////////
^///
/'/X-^)
"/
}</.'.
//"'
Ja
)
!/(#
identity
r
$/.H<+WH/.+l/.U,in
on
at
In this
-a>c
this
|}-l/|.
th
iiu-
i-asi-
in
which
</>
and
f"
r,
are
the function
eases as
raaM
';
intcr\al m.iy
The
i"-
il!u-t i.ited
l<T!i:-
nil"
inn
the
-..f.-nnly
hut m.iy
all positive
l>y
repeat
;--ntly
mil
to /cro in
;
the
<.<
t\
^hitrurily great
values
ai.
any
>'.,(.
l'.r-Mi).
ii
the latter converges unifonnly to zero for
INFINITE INTEGEALS.
454
arguments given in Art. 31
prove that
(5) to
if
[AP.
lim (lim
A.
X*>
fJ.
F(k
//,))
ill.
exists,
>ao
so also does the other repeated limit, and the two limits are equal.
But, in virtue of the uniform convergence of 2/n(#)j we have
so that
lim {lim F(\,
ja)}
=
a
A->oo fjiHB
The other repeated
and
so the test
Ex.1.
limit
seen in the same
is
Jo
is
positive,
|
to be
established.
is
Consider
where a
way
and
sin(fe^)
b=p + iq,
= [sinh
ax
where
q\
= s<a;
since
2
(
,.00
and the integral
Jo
convergent, it follows from Theorem B that term-by-term integration

permissible,* because the terms in the series
is
are
Thus we have
all positive.
in(fo)
a
In the case
and
2.
*-l
when a = 2?r,
so in general
Ex.
it is
this expression is equal (by Art. 93) to
equal to
In like manner
we prove
/sintfcc) tf. 'Jo
2Lb
that
6
*"+!
= in_
by Art.
is
_jr
__
"1
a sinh (irb/a)J
92.
Note that exactly the same argument enables us to include
integration, although the series diverges there.
iu the
range of
l.vi
176]
Ex.
BGRATION
Taking the case a = 2jr, expand !.<>th ikUi

powers
the theorem depends upon the integral
CMS the application
3.
In this
,-HMh_{|
e*-\
irh .-..nverges if
Thus we
|6|<27T.
find
[*a*-*dx_B r
Jo
d compare Art.
see A.
Ex.
4.
Similarly,
l.y
K\.
17.'.,
7.
in
-xpuiidiii-_:
powers of
we
.r,
find that
it'
>S+-)-F
And
\vitln ut
.
iv>triction
r e-^cofi
on
6,
the valuo of
l'(.l),
r(j?), ...,
6*
VT
/ ^\
^/i
^v
+ 5i~2---- \ = iT7~
^/l
2a ex P V(- /
2\
1
N^
we have frm
1
However, Theorem
For example,
C.
IM
int.
does not cover
cases which are
all
not hard to see that the
is
it
iv|uired.
seri
-rated term-by-term between the limits 1 and

This case and others are
although the test given above fails.*

cover. <! l>y the following test:
flmt
the
series
o
/''.s
series
vn
uniformly
2gn (x)
converge*
>////>///////
(1)
//'.'"/
<<n;i
interval (a,
6),
wltl>
vn
*val
/ M HfoJ
)]././
001
r,il
///
n,s
of(
CI&
[DIM.]
ly Ait.
And. since
wt-
-i'i.
X</.<--)
>
have
uiuf..nnly r..Mvi-ri;Mit, wt- have (Art.
Inn
/" [-/(,-)],/,-- j-*r
[!',/.,(.''
See the second paper quoted on
'
^-.
p. 450.
L_.
.>
INFINITE INTEGRALS.
456
[AP.
ill.
The inversion of a repeated infinite integral.

by no means easy to determine fairly general conditions
177.
It is
under which the equation*

/30
/GO
pCO
dx\Jb f(x,y)dy=\Jb dy\Ja f(x, y)dx

fOO
a
is correct.
Here we
f(x, y)
is
is
shall
by
replaced
when either
when
f(x, y)
converge
simply consider the easiest
positive or else the integrals
still
case,
\f(x, y)\.
Let us write
f(x, y)dx,
this equation being valid (see p. 410) if, as we suppose, f(x, y)
is continuous for all finite values of x, y (or at least for all
such as come under consideration).

0(a?,
p)= Jb
\f(x, y)dy
^r(aj)
be subdivided by
each of length
I,
and
in the rth interval
Now
this
Km
lim
n>x>
lim
/A
Now we
>
F(\
decrease as
n and
(5), Art. 31,
Ui r (fi)}
=lim
n
//)
= lim
shall prove
Jb
f(x, y)dy,
Let the interval
n sub-intervals,
minimum of <J>(x, /z)
><x>
we have
//
lim
p >oo
n-->oo
lkr
if
/v
tend to infinity;!
which gives
I1ir(n)},
1
provided that one of these limits converges.

(3)
last integral.
let hr (m) denote the

then, as in Art. 163,
sum cannot
H><*>
/*)
continued bisection into
and so we may use theorem

(2)
= lim 0(0,
fj.>os
assuming the convergence of the

(a, X)
Further, write
Thus
= lim h r (fi).
/a
>>
below (see the small type,
p.
457) that
For wider conditions, see a paper in the Proc. Lond. Math. Soc. (2), vol. 1,
and other papers quoted there. Reference may also be made to
Gibson's Calculus, Ch. XXI. (2nd ed.), and Jorclm's -our* ,rAnnli/e t. 2,
71, 7'J.
t As regards n, see the argument of Art. 163; and 0(x', /t) increases \vith
(because f(x, y) is not negative), so that the same is true of
1903, p. 187,
REPEATED
177
1
minimum
the
i-
/,.
.!'
NTK<
;it,\LS.
]!]
>//(/)
//.
in
tin-
inter
/'th
'
inn
;
:
*
iiii
->
v.
fA
//',.=
the integral in
>p08ed c-onv.-r^.-nt (other
tin- equation (4)
n (I) would be obviously meaningless
that tin- ri-ht-hand limit in ("2) exists; and BO
Sine.-
-1
).
made above
1'Yum
justified.
the
'
(illations
(3)
hat
'
Jb
:.
dy\
/(.',
thai
and similarly, we
to tip
in
'
la
\'.iii.
.-
that
>f
th'-
/'( \,
//)
s.roml intf^ral in (1) is

taken in the reverse or
.">
above is correct, provided thai both
>
and
are
be seen that
tiff
\\
>till
'
'
!:
all
remain
to |.|-HVf
/(
either
(lr.
nd the theorem to cases

the integrals
M|iial
1
{,-
ill
fJL
decrease, as X and /JL increase, so that

of Art. 31; and we obt
a}nly theorem
foussin'a theorem
i
li
I
\>V
timl
limit
bed
,lim
/<
:!imt
we
y)dx=\J dx\Jb
JO
dr
is
'
and
B,
<>f
tin-
'+ j
fin
reds
rep
in place
\\ht-n /'clia
nt
\v-
of/)
!'-il
when f
put in
is
that
pi,
tll.'ll
//J
l-'i-uii
and
BO,
th,-
OD
of
ilriini;
iinl
tnal
hapiK'Hs that
hat
<!>(/>, />}
^(p)î'
also;
uiiniinu:
\//-(./-)
in
tin-
Ln(
Art. 17_
iii-
.'
that/
is
convergence, aa in
458
INFINITE INTEGEALS.
But
if
^}>k^.h(^\
(f>(p,
it
[AP. HI.
follows from Ex. 21, p. 395, that the equation

let
M denote the
<(#, p) = k has at least one root in the interval (p, q)

least root.*
so that
i>^ju,,
and
therefore tends to a limit
<X">AO = <K",
Again
so that, on
making
As an
where the
application
real part of
and
have seen (Ex.
(5)
we
is
we
as v tends to infinity.
= *,
we have
find that
^ ()=,
1,
is
therefore
shall establish the equations
and the arc tan function
positive
is
determined
Art. 176) that
therefore
fY-(_L
Ve'-l
Jo
Now
the last integral

00
-vu =2 r^& ri)&.

e*-\
2/
is
e^-i
Thus
f"
Jo
dt
I
dx <
e'* =e~
and
Jo
Jo
absolutely convergent, since
sin (xt)
^ô?^
^Zl = 24'
fl e2| " -!
35
Jo
which proves the absolute convergence
if
we
write
If the equation has
is
'
we can
therefore invert the order
and we then
find
xdx
^-\-i^ in the last equation,
under the integral sign, between
reader
'
2/
an
infinite set of roots,
is continuous)
are also roots (because
which is therefore the least root.
fThe
3 Art
"
Ve'-l
Now,
^ < -V
122'
r w_j__ 1,1^=2
Jo
Jo
respect to
Ex
if
n (**)l
of integration without altering the value of the integral,
which
'
o as to vanish with x.
We
")
v tend to infinity,
Thus Vr(f)^|, and so from

again the minimum of ^-(#).
Ex.
v>p, we ha vet
if
Then,
and
we can
oo
integrate with
for
the limiting values of the set

set attains its lower limit,
and so the
advised to use a figure in following the argument here.
THE GAMMA INTEGRAL,
177,178)
this
80 that
tli.ul.l.'
Similarly we tincl
aU..lnt-ly convergent.
onvergent, since (a'+Jf'liSft so that
is
intr.u'i-al
<
xdx
Thus
\\r tiinl tin- furtli-
-<-
JT<
where y =<)+"/
The Gamma-integral.
178.
In
We
we have
42
Art.
now
shall
integral \vh-n
seen
th.-it
express this function by means of an infinite

n-al ]uirt of l+ic, is positive.
tln
a.
Write
/,
= P y x+s (l - y) n ~' dy,

Jo
we
thru, using the
method
and so
I /In = n\/{(l+x)(2+x)...(n+x)}.
by
parts,
find that
/n = l/(n + l+aj),
But
so that
r(H-.'')=lini //'''/o,
varial>li-
tin-
chanîn^
or,
of integration
l>y
writing
////,
wu
hav.-
'<//.
r(l+a;)=lim f'Yl--Y/
n->cc Jo >
We
can apply Tannery's tlu-nn-m (Art. 172(4)) to the last
integral;
for
we have*
l
\e~
and
t'-(\
</ 1+
/cJ
p-îtiv') the intr^raiid converges to th' limit

in
unit'ormly
any _//./W interval I'm- /. Further, \v- ha
so (since a
is
ênt, because a
is
too
o
Actually
-'-(l-*/w)
- e'(l -
is
=
-j
positive
and
e'M -
less
^dv,
than ^/(-'O-
so
that,
when
positive.
is
po
INFINITE INTEGRALS.
460
Thus
the conditions are satisfied, and so
all
r(l+a)=fJo
somewhat
[AP.
we
III.
find
e-'&dt.
similar integral can be found for Euler's constant
we have
seen (Art. 11) that
lim
n
But
>vo
l++... +
Thus we
on writing x=l-t/n,
find,
And
C=
hence
and,
by
1,
[Ml-fl*)"}f Jif(l-^
n)
V
nj
V
lim
n-wo L/o
method
the same
t -\
we obtain
as before,
as the limit
Since
we
see that
lim
Another form
is
(
[^ j
f*J
easily obtained
- log(l +
8)J
by changing the variable from
to
.00
f
in the integral
number
e~ dtjt
this gives
of definite integrals for
can be obtained from the expression
-lim Hog 8+ Te-'t

g-ôL
Amongst them
It
is
are the following
J$
easy to see from Art. 180 below that
Useful properties of the Gamma-function.

1.
When x
is
a positive integer,
we can
write
STIRLING
178, 179]
result whit -h
is
also easily obtain-.
of int."_ri;iti.>M
ni.-tliM(l
POBM1 LA
n-.m the definite integral, using the
'""7,1,
-x)(2-*)...<n-*)
!
>'
)(2-.r )...(n*-^)
)!'
,.
Writing # = $
:*.
(An.
.i.)
the last result, \\-
in
ti:
T(i)
i*
}><
^-*
fir).
;i) =lim
Thus
I (-'.'+ 1)
this last expression does not contain

a result
tlii-
putting '-();
by
ajijiealin.u'
179.
we can
find its value
by
which can also be obtained
directly t> the detiniti>n.
asymptotic formula for the Gamma-function
Stirling's
when x
,:
is real,
large and positive.
In tin- integral
F(l+ .!)=[J
niMxiiiiuin
tin-
so
it'
<>1'
tin-
in!
.-rand
is
6~ maP and occurs for
we write
"*",
tin-
range of valnefl
to thr range (0,
Thus
for y will
oo,0, 4-oc
oo ) for
r(l+)
.'
Now, taking
lnr;iritlnus.
\v.-
li,
t/0)
so that
i-nnvspmul
= J\
INFINITE INTEGRALS.
462
But the properties

y lies between*
of
[AP.
III.
the logarithmic function shew that
x-
2
,
and
Thus, since y has the same sign as
we
x,
see that
lies
t/(t
x)
between
t&
/y,
\AJ
(\2/
2)
Thus, (o5/2)*(l/y)
And
therefore, since
must
lie
Hence
/i
I
(I/-
/
V
_.
^y
a;)
/(
<a
/
/i /
l> I \l/
Tifi
<Aillv4-
\fiJl
= 1 + x/(t
0"
W
i
!
we
a;),
see that
between
-7-
ay
-;
lies
between
and
Accordingly,
where
|
/r
\AJ
between
lies
tAs
ArA^/v
/ tA/ \
-i
and
<2
we have
Nowt
|y|e-^dy =
ir*
[
J -c
J -GO
and accordingly
-1
(2vttV
Hence
or,
as
we may
We
have,
e-
it,
.(t-x)lx = r,
if
which obviously
write
lies
y^x
between ixra and $#T2/(1 +r) 2

(
e- x x
(Art. 178 (3).)
OTHEB POBM1
179, 180]
Again,
we
)] < \c2irx) -2],
^
(./+i) log a-a;+ Uog(2?r),
log T(l +*)
-2
so thai
usmI!'
1,,
ml.ol
-\
we -uUrart
log
wr oUain
co
!'(.!)
<
-0+ilog<2T)
(.1--
nirtlio.l
(\.
t*60
tin-
"ii-id'-i
IJ.u\
to
.hi.-
is
Journal
Ex.
sense explains! in Art. 133.
in tin- rxt.-nd.-.l
logOJ,
Eon
'I'll.-
his
'\/
1.
wlio gave
ill.-,
L846,
it
ill
*64)
p.
value of
*<-r >="
r
(
,')
r+
D - p (* +2 5 1)/r6l( *
ve integer.
change X to r+1,
<M.<-+l)<M-'0 =
= ^(.r+l) = ^(jr + 2)=... = ^(ar+).

</>(jr)
Hei
n
or,
n"[.'{.'-
\ve see that
//
tlir
u-iiiir
is
large,
r(y + a)~
asymptotic formula
r(j/).?/'
ab--.
Hence, as y tends to intinity, </>(//) tends

WC hav- alrradv j.mvril that /.(.'-) = ^(r +
great positive
The
180.
\Y.
>]>ecial
iiiteur T,
case
>> that
n=2
has
\vi-
IM-.-M
Hi -nr.
t>
(lisni,-,l
17:>)
the limit (27r)^ (n
where
),
must have
Integrals for 1- Til

have j.rnvnl (K\. 1. Art.
a, b an-
x. that
f/>(.r)
in
= (27r)^ (n ~
Art.
that
is
if
178
~ 1)
//
rwl
/"'/V
"f
+./
la
(4).
the real parts of
w-
positive,
II
'0
log Til +.'

f./-)
and
positiye,
If
n*,
an arbitrarily
]'
hav.-
464
INFINITE INTEGRALS.
Thus we are
S(x, n)
III.
[AP.
led to consider the function
=x log 7i -f V log f--^\r-{-x
i
60
(e
-._ e
/7/
-*)9+2 JO (1 _ e
"i
Now
7i),
It is to be observed that
say.
both in F(x) and in G(x,n) the
integrands are finite at t = 0.

For if t
2-7T, we can write (Art. 93)
<
so that
and
similarly for the other integrand.
Thus,
when t<l,
cannot exceed some fixed
'-
value, independent of
than
|oj|H
6 ~~
is positive).
t,
+1
z-,
but
because \e~
xt
J.
if
<V
>
this expression
(since the real part of
Xe- M dt
Then
or
<X/n,
lim G(x, n) = 0.
n >oo
Hence
less
l+o;
Thus we can determine a value X, independent
such that
so that
is
logr(l+a) = lim S(x n) =

t
of
180
MM
This
in
in'
divided
I..-
logr(l+)
.....
ii,
and we
parts,
,/,/.:-.',.
wH
'
expressi(,n following lV-.ni the

this runs)'. rinat i< n is
last
tin-
The advantage of
example of Art.
due to two fa<
the value of 0(o?) CftD be i'-Miinl in termfl of

functions; ami secondly that \//<r) tends to zero
that
to
find
such
in
/:
the course of the
in
<1
:-
The
is
the real part of
limit
also 0,
is
when
/;
tend
^(.r).
Thus we
see that,
--./
To
-s.
we
posit
we have
A:
:i.
to
^ being kept
-'"')
^[(,--D^'+()-;
Kx.
478).
p,
irds
l.y
tends
we proved that
Thus when 5 tends
,/.
el.-im-ntary
if |a;|
also tends to oo
of Art. 177,
example
B)
where
the real part of
that
way
177.
detei mi
j'
|f
J=
if
+ A.
.
7s.
which
-'ives
-log rt- r-hl) = Hog7r.

)
logr(aJ+i)+l<
Thus
\ve ha\
lim
n iiis.-i-tinth,- value of
whi<
lin,
1.
+,'
I,
-I
-J
.)
-^...oj...^,,.,,
or*
"
Tlu'
I.S.
valr.
478).
lllllla
INFINITE INTEGRALS.
466
[AP.
Thus we can write

log T(l
+ x) = (x + 1) log x - x + \ log
i
where
2
^\ = o
and
arc tan (Wo;)
2-Tr
+ \/s(x).
|Vr(aO|<l/12f
have a formula
It is often convenient to
and varies
is real
if
Thus,
where
we
neighbourhood
x v _j_ ttj
large and a may be large, but is of the order

obtain the asymptotic expression*
v is
we
most,
in the
write
for r(l-fo5), when x

of a fixed large value.
where the error
is
of order
\]v at
a/v.
Hence
In several books on analysis, the integrals for logF(l+#) are found
different method due to Dirichlet.
In outline, this proof is as follows
by a somewhat
rr
:
(1 )
Differentiate the Gamma-integral,
e-
and we
00
logtdt = jQ e-
(2)
Invert the order of integration, and
(3)
We
find
r
dtjo
we
//i
(e
--e-<*).
obtain
must next prove that

lini
and then we have
(4) Finally, if we integrate the last equation, we arrive at the same

integral as before for log F(l +./).
The reader will find it a good exercise in the use of Arts. 160, 72,
1
177, to
shew that the steps
in Jordan's Cours
*We
have
so that here
we
get
d Analyse
(l)-(4) are legitimate.

(t. 2,
2me
6d., pp.
Proofs will be found
176-182).
4J7
in.
I
i.\
LMPLE8
Tests of Convergence.
1.
Mci mine
the valu.^
î-'cosjfdSr,
/'
.f
whi,-},
fol
O,
tl,,.
integrals
(2)jV'sin*.
are convergent.
iMscuss
2.
continuity of
tlio
tin-
/"*
intBi
l-2ff
'led as a
function
,,f
Sketch
y.
hisousB the convergence
3.
the
.f
its
graph.
[JA/'//.
EM
./;.
0<K;_i
If
4.
both the series and
2
*
liv.-i-gent
and
if
tin-
intrgral
/'
+"
,^
>\\\ i'H'
.r"
although both converge
a>0,
int
if
5.
*,
1'
tin-
inf.-r
'(.')
is
convergence
of
171).
[HARDY.]
prove from Art. 1W5 that
n-.-m
and
'
that of
w,-
/"*
j
Apply
tin-
if
(")
dedii'-dl
lie
ii
6.
the
/(.)/.;
provided
in<>n<t"nic.
Similarly. sh-\v that
ll.-
When f>^,
a=0.
l><>th
convergent.
Reconcile these results with Dirichlet's testI
is
iM<iiii.t,.|iic
from that
sequence, th- OODTergOl
of
thai,
ni'-t'
if
u, /?,
y are p<-
integral
*
if
-'.
-;
Dedu
and di\.-rges
-l>y>0
and
/^ + yi
+ y> -.
if
^3
egral
[<
ic
Shew
that
no
that
test
l,r
1,..
in the la-t
te^rand tends steadily to zei

of the Logarithmic scale suilices to establish the
convergence
of the int'
State and
p:
[H
\.
INFINITE INTEGRALS.
468
Shew
7.
that the series
[AP.
n a-l
if a is positive and A is rational (in contrast to the corresponding

But if A is the root of an algebraic equation of
in
Art. 166).
integral
2m.
1, the series converges if /3>a
However, irrational
degree
diverges
m>
values of
/3
A can be constructed
for
which the
series will diverge,
[Compare a paper by Hardy, Proc. Lond. Math.
Shew
8.
whatever
be.
may
Soc. (2), vol. 3, pp. 444-9.]
that the integrals
/cos {/(*)}<**,
pm{/G*)}db
are convergent, provided that f(x) tends steadily to infinity with x.
Prove also that
f'(x)sin{e
f'(x) tends to infinity.

[In the first case it
as
we may
f(x)
convergent no matter
is
}dx
how
rapidly
not sufficient that f(x) tends steadily to infinity,
is
by taking /(^) = x]
see
/-oo
/ cos(x )dx
and
2
s'm(x )dx are convergent,
2
2
prove that 2cos(ft 0) and 2sin(% 0) cannot converge
(see Exs. A', 13-17, Ch. XL, and Ex. 10, p. 485).
if
B/TT
is
rational
Change of Variables.
10.
If
intervals
x,
g()
is
an odd function of
(0, JTT),
TT-X, x
ir,
ZTT
TT),
(Jir,
...
X,
(TT,
|TT),
prove by dividing the range into

and introducing the new variables
,
...
respectively, that
provided that both integrals converge.
Deduce that
tan~ 1 (asin.r)
H
/
11.
of f,
(log
COS-.-')
Apply the same method

and O^K:-^
= ^7rsinh~
dx
7T
to prove that,
log
2.
if
/()
e--
cos
a>0
(!)
/(
sin,)$
>i"
=2
sn A>
V( si
(Ex. 15.)
a,
[WOLSTENHOLMK.]
is
an even function
ill.]
<>
-2
in
i.-,
[It
in
lv\.
17,
the intrgi
all
th;it
:n.
p.
Illustrations
12.
,f
cos 2nx] dx
oowe**-
p)a(
understood
'*"*
+ 2 cosh(Ka) 2
[*V(sin .r) [e
869,
|>.
rge
K.\.
.-iii.l
the series used are
:..
the last exaui]>l<-
(1)
(logC08^)~=
-7T,
(logcos .r)(logHii
1).
^
ami
similar formula rout, lining log
and log A(l -
sin'-'.'
Ol'>g(l
and log(l
similar formula <-ontainin_

-
but
this case then-
in
is
no
Differentiation
Cali-iil. ite
the
tli-
j)i
and HARDY.]
and Integration.
inteL:
o\- that
i>
10g
linn
lu-in^
>iu'n
ambigUOCU
limit
with l>^-\
[I
,=jf
and
-e'**).
lin^ foriuula
.
13.
1)
+*-
not the
same as
i-
tin-
the sign of
'lain
this
why
integral
[STOLZ.]
14.
l'i'..\-
I'.v
di"
n,
or
l,y
.xj.aiidniL:
'
liy
:lls
jwwers of
l)S
jt
in
iv
writing
dillrn-ntiati.-n
and
/b\
whi-r-
a,
that
!
o'
i>
real;
and
INFINITE INTEGRALS.
470
15.
[AP.
Prove similarly that

/
-f
log (1
and
j^log
(1
a sin x)
=
S1H 30
[TT
sin" 1 a
- (sin"
By
a)
+ a sin%) 4^=* [v/(l + a) - 1].
Obtain four other integrals by writing a = i/3.

16.
[WOLSTEN HOLME.]
integrating the equation
fSrSS*-!^ where a>0

with respect to
y,
>
prove that
and so on, the terms introduced on the left being those of the sine and
cosine power-series and the terms on the right being those of the exponential
series.
17.
[Math. Trip. 1902.]

Justify differentiating the integral
f^tan- 1 (a2 tan%)^,
(a
>0)
under the integral sign, and so prove that its value is 7rtan~ 1 {a/(a + v/2)}.
2
2
Change the variable to 0, where a tan ^=cot 0, and deduce that if
we put V< = 2a/(a2 -l),
e)
an
+ ta n 0J
cases x = 2 (Wolstenholme)
(1
Examine the
special
and * = 8 (Oxford Senior
Scholarship).
18.
By
differentiation or otherwise, prove that
jftan-'(a*)tan-'(&*) J-Ja- [olog (l
19.
By
if a,
+^)
differentiation or otherwise, prove that,
6 are positive,
+ 6 log (l + |
if
is
positive,
osin
7V/;>.
1892.]
171
ill.
I
20.
>li.-\\
are positive, a being the greatest of the three,
},..
<io:
= 4^,
COB &r COB or
>>6 + c,
if
ITT,
Deduct-
a<6-fc.
integration that
l.y
e08ar-^
= i7r&,
i7r(
if
+ &-c),
if
a>6 + c,
a <& +
<?,
'
> 6-fc,
/<& +
In
particular,
10,
or
Prove that,
21.
if
t>
"2
II
The
22.
:
*TT,
<><f<2,
'.
î = i-
...aft^r.
:MER.]
integration
l.y
by
irivt-s
v.,,
f/*"-
,s,
^.l'/-'
.0,
(^^l^S^^jp
,=(-!)"
(1
!'.!:>,
1'-
log^p
" .....
I'.l"'
I\la2"=0.
similar formulae f-T intr^rals which contain
'lish
<?.
+ ... + |a*|,
rili(
if
Exs. 20, 21 can also be found

at once
results of
this iiirtln.ll
Jo
if
sums
of sines;
and prove that
--...}
tin-
Miuiili*
in
tin-
1.
racket being i
or i
Dirichlet's Integrals.
23.
Apply the thfoifin
of Art.
liu. \' >(r

c-M)^0
and deduce
17^ (4) to justify the equation
+ ,-t}e-dt=^
that
'(y)-<*-H*dy =
>OC J_ao
'-
li,
[_;(,-
+ o)+
r\.r- 8)].
8 *0
[\VKlKi:sTRA88.]
24.
Apply
Alel's test of
4ence to prove that
in
inonotonic (at least after a certain itage)
-'(OX
according
c,
zero, or positive.
i/(ox
if
/(/) is
INFINITE INTEGRALS.
472
/ f(i)dt
is
[AP.
convergent, then
/ "cos (xv)dv j
and the same
result
true
is
the cosines are both replaced
if
by
sines.
[FOURIER.]
25.
taking /(#) = e~
By
ax
(a>0), deduce from the
example that
last
/""acosO^y)
7
2 *-
TT
-i<" =1
-W^
Consider similarly the integrals given by taking f(x)
from # =
d/(#)=0 from 1 to oo.

26. From the integrals
to
1,
[FOURIER.]
,ech*=S/>iS**,
Jo cosh Trt
sech%=.
sinh
irt
prove by the method of Ex.
Art. 174, that
4,
* cos
(2^7r.r/(o)
TT
o)
cosh(?i7r /w)'
2?r
(o
e -(x+no,)2
n cos (2w7T^/w)
~&
sinh (w7T2 /o>)
-OD
).
O)
_a
[SCHLOMILCH.]
Integration of Series.
27.
Prove that
(see Ex. 42,
p. 167), if a, b are positive,
a cos% + 6 2 sin*#)<far = TT log
and verify that these results remain correct when

Deduce that, if r2 < 1 and p, q are positive,
/" log (a
,
r
9sn.r
Jo
Compare Ex.
28.
2, p.
cos^cty
Jo
H-2^cos0 + i
and verify
<t
= 0.
_7T
7,
Ch. IX., prove that
= 2_tanjr^
"
(1 + ^)'
/*"
Jo
f'tan-
this result
and when
**py
f/<
Jo
347.
Using the series of Ex.
Deduce that
l-t4 /unf9
by expanding
in
= 8tan-/. tanlrV,
pmvrrs of
/.
[HAHHY.]
LMFLEa
in.)
29.
Ki-.-in
:1)
I
a!-,
Sli.-\v
7r
f*
30.
1^7+r
Bn
I'n.ve th.v
(1)
suin
'o
A\
(I
\7r' )
175>
!
Euler*! nuinl.*-:
ii
lly
(\ -L
:.vcos*+r*
r\j;uitliiiu
powers of
in
u,
-i
=
31.
Kroiii
tli.
Jo
logO,
-('-i'
--)
*10j
^rr'.
;
(V..ni|.;ir.-
tt-mliii^ tin-
32.
I'ul.-
t"<>!
'
IT"'
t--i
l"
in -l.y-t.-i in
iu-iit'\
tin-
Mt
th.-
..nly
n|i
[W..LSTKNII
ditli-Milties
arise
in
to the limits.]
following traiisfunnat
-J
= 1 _i
[LKGKNDRE.]
33.
shrw that
\it. 176),
ami verify
34.
[\V,
It'
tin-
Mtih
equation
1>\
mak
30.
that
474
INFINITE INTEGRALS.
Gamma
Use the method
35.
[AP.
Functions, etc.
shew that
of Art. 106 to
if
the real parts of
r, s
are positive,
and deduce that
e~ xt tn
If
36.
Cf=Jo
-l
3_
3f~
and hence prove that

using Ex.
By
r<a
co$t.tn
Jo
#=+iV? and
>0, shew that
'
U=T(n)/x
in
'dU_
<sj"!"-T*r
n
if
Art. 172, deduce that
2,
-l
where
dt,
n>0.
if
/-
dt = T(n)cos(%n7r\
and verify that the
smt.
JQ
last result is correct if
l<n<l.
Obtain the corresponding formulae for

/
coa(a?)dx
If the real part of
37.
and
lies
si
between
[CAUCHY.]
-k and -
where k
is
positive integer, prove that

.
[Apply the process of integration by parts
Shew
38.
that
if a,
ft
to the integral for
[CAUCHY.]
r(
are real,
# = 117, shew that
If
dx
Tf
]
express
A B
t
in
terms of Gamma-functions, and prove that
Assuming the value
of F(f) given in Ex. 41,

;4
40.
= 1-311029,
deduce that
=0-599070.
[GAUSS.]
Similarly express the integrals
dx
in terms of T(^),
xdx
and so obtain numerical values
for them.
[GAUSS.]
EXAMPLES,
in.
I
Deduce
41.
fr-.m
(In-
K'r<i+*)-jiog
^1
Cs- J
<7&
formula for !'(!+./) that
j,r..du.-t
-C
s
(77
= $i>-* =
also
!.
F-WTMl,
found
fi-'iu
this
#=10 +
writing
42.
If
Shew
2 7t- =
2232,
I..g 10
s.-i
i-s
ami
r(|)=!
that
if
1'
+ i) = re*, then
^=-'301'
'.802-(0-15495)i
Stirling-
>-i ies
(Art. 132),
l-o
is
/.
that
win-re
t-alculated to 7 decimals
ilt
N- r =0-0011928,
fc-U -
log, /-= 1-7 17:11
Tho.-
that
iv
logl(l
will
73-
nuim-ii'
It
[jr|<2,
-jiogQ^+c,*-^-^-...,
:;.
if
.,l
Euler's constant.
xKO-
that
Deduce from Arts.
17*,
that
17!)
)-i
\js(x)+C= f
>lt
Jo
Obtain the particular
where
that,
f(t)=
0)=
<>f
is
|...>itive).
if
^'(4)=i
p, q are posit i\e
-*)} + (! -f)log(l
if
part
results,
fa)-|(iw>-w
Shew
the real
tliat
alflO
43.
(if
11
(-TT/flJl
~0+ 2
w~' r log(l -
w r O>
INFINITE INTEGRALS.
476
Deduce from
[AP.
and the corresponding formula with q
this
in place of
p, that
Obtain the particular
results,
lK) + C= ~ I log 3 + ITT v/3,
- 3 log 2 + Jir.
<Kf ) + (7=
[GAUSS.]
30
44.
Similar results can be obtained for the function J3(x)=
2
o
thus,
IV*
-\-
shew that
/?(#)
+ 0(1 + x) = 1/x,
lim
($(x)
+ 0(1 - #) = TT cosec (wtf),
O(^) - log x\ = 0,
>oo
a;
Tl
($(x)=
lim $(x) = 0,
a:
>
fx-1
(if
^r-.dt
the real part of
is positive).
In particular, prove that
45.
If
/(a)
prove from Art. 172 (3) that

provided that a is positive.
we may
differentiate
under the integral sign,
Hence
and
46.
Shew from Ex. 45
that
rt
I
= log 2 - 1,
sin
log sin
x dx
''sin
(log sin #)*.<&?
f V(sii? ^) log sin
a?
etc
= (log 2 -
2
)
Jo
I
log sin
a;
^)
dx
2
.
dx
= - JTT log 2,
= tir
-f 1
- V,
III.
\ll-l.
l>
47.
.tiati..n
tl
.hi.-tify
th-
..f
..,,.
that
log cos x
ô
/"i*
/
sin
o:
log sin
.</.<
= ATT [( log 2)1 - A
log cos #
71
"]'
dx = 2 - log 2 - ITT*.
Miscellaneous.
Frnin the series
48.
that
it'
the real part
i>
[See Ex. 42 and use Art. 52

49.
and
From the
last
uu-att-r than
n't
UIV.H.T than
(3).]
<K-du.-r that,
example
-I,
if
ih- KM! part "f n
i>
I,
r* sinh
our dor
Ki-.)*,
and hence,
if
rove that
- log 00th JAr,
[JA/M.
r/-
1889.]
si
50.
Ki<m
nd n't
K\.
^n-alei-
B,
Ait.
than
[HA
nhlATT).
cos.
IT."-,
if
thr ival part of a
7fi
that
if
LIP
JTo
15,
[.I/
-HI
p-
=ilog,
51.
is
A,
4.8
1)
obtain the latter sn
<l->,)'-*l
-;d
.r(.r
>]~-(\-ij)'
+...
and a are p
1895.]
INFINITE INTEGRALS.
478
52.
Obtain the
x = 2 (er* - e~ Sx + e~^ -
. .
.)
Frullani's integral to the separate terms.
Obtain similarly the following integrals
where the
III.]
integral of Ex. 49 from the series
first
sech
by applying
[AP.
real part of
is
[HARDY.]
positive.
53. Write down the form of Frullani's integral when

and deduce that when p is positive,
/"*/
Jo
/
sirihpx
W^_ii
~
^
cosh px + cos rx) x
sinhp#
V cosh px
+ cos qx
[Math. Trip. 1890.]

54.
The
following integrals are allied to Frullani's integral
r^dx
(sin
Evaluate the
55.
mx
sin
=%TT
the variable from
Shew from Ex. 3, Art. 173, that of

Iog2-l and the second to log(7r).
56.
Prove that
if
m-n
these integrals
first of
By changing
nxf
is
when m, n have
to 2
opposite signs.
in Art. 180, prove that
these integrals the first
is
equal to
(See Ex. 52.)
positive, the function
V-l
to oo (see p. 234).
steadily decreases as t increases from
By applying
Art. 168, deduce that, if #= + /?/ in the formulae of p. 465,
57.
but
rj
Deduce from Ex. 56 that
if
#+tfy
where
is
positive
and
tends to infinity,
where r = |a?|.
(Compare Ex.
38, p. 474.)
[PINCHEHU:.]
BABY MISCELLANEOUS EXAMPLES,

Sh,w that
1.
can be MH.UM..I to
series
the form
in
I>i>ouss the
2.
"
t!)- ]^*
v|_
10 -
tern
71
may
express the general
<
'{./--
/.*-.]
convergence of the series

1890.]
and
,l++...+-
If
3.
d-'dur.-
Kulfr's i-i.nstant
that
is
{<>
rju;il
"+
1'iove
4.
that,
it'
(w
that
5.
[<
V1
to the limit
.
1111
//
'
'-' }
log{(l +/);(! -X-)}, wluMi

tend-
//,
tend
t..
intinity in
such
//.
-hat
j.).l>,
1
;
.'th.-r
[K<>i
mini:
i!
and
8,
6.
If
dniotes
2~"
rvalisatin::
i'p.
tln-sr
-riea
l.
into
hat
+ * + ... = l,
J^ + ii3 + l*l+- = l'.,
*,-f*4
-|-*fl
+...
= i,
'
'
[>
,12,
L906,
i'.
477.]
EASY MISCELLANEOUS EXAMPLES.
480
7.
If
is
positive, prove that
9
[Write x=e~^ and convert into a double
8.
Shew
[Math. Trip. 1894.]
series.]
that the series
x\.
does not converge uniformly in any interval including

9.
(1) If
1
/(#) = way"- -(% + !)#", prove that
and deduce that

(2)
is
S/nfaO <&?=!> while
Prove that the
series obtained
same true
10.
[Math. Trip. 1901.]
by
differentiating
real values of #, including
all
#=0.
Is the
of the given series?
Prove that
11
1.3
2
,
1, 2,
^)]
= ^(1,
tf
|, 3, #),
[^(1,
f, 3,
'
^)P = ^(2,
I, 5, ^).
but the law of

[All these can be obtained by direct multiplication
the coefficients is more quickly determined by differentiation or some
;
other special device.]

11.
a,
6,
Shew how
c,
d, e
to calculate log
2,
log 3, log
log 7 from the five series
5,
given by writing
^=A> %V> T&TJ sV> sir

.^)}
respectively in the series for log{(l +#)/(!
and prove that
[For results to 260 decimals, see ADAMS, Math. Papers,

12.
Prove that as x tends to
13.
sin
sin 2.^
log (1 -.,).
to 2?r of the functions
bx + - sin
7;
>
si
1,
p.
459.]
1,
.ru
Sketch the graphs from
vol.
lO.r
+ ~ sin
bx +
. . .
+ - sin Kb + ...
(}.*
[CESARO.]
I.
14.
AM
gift l.l.l.\\
It
'li.it
\\lit-re
i)+...+/{.r
a-2ir
nmlti|ihir>ult
111
i*4, an.l
P,
r- i-
If
+ (*-l)a}],
-.
[This
15.
4M
AMI'LKS.
nimifiiral
tli.-
apji
see
</-.<.
\-\
I.
1.
tl
[Ml
384).
s,,
IB.
in
jp|
|Takf
,vha,
ivin^j
./
i-!,,i
valiu-s
>|-.-ial
l'i"\- that
17.
s-i
tin-
it'
i-
-'\
in
I'.ni.J
j,.
;_,..
;,,,;,
tl.-
17,
190.
\>.
rn
^/*+l
rfl
17,
,!,,)
= tan
i.'
-i""l_\
[M-
Slu-Nv
18.
that
th-
remainder aftrr
\
/<
tnni>
in
th-
1907.]
lii-i
,'\
'-'
iiotcs
th-
r.frn.'iilli,,!,
fuiH-ti.'ii
(h-iin.-d
in
An.
|".1A/^.
7'
Sh.-w that
19.
[\\
H'sf Arts.
Discuss
20.
!C,
l!>
ami
th,
;.]
th-
'inj.lex.
tllC
f>r all
vahu-s
La
..f
th.-
..miil.-x
)H'i.l
vaHabl
--'n
[.I/.-
l.ME.]
EASY MISCELLANEOUS EXAMPLES.
482
21.
Prove that
if
S = l!-2!+3!-4! + ...,
l(l!)-2(2!) + 3(3!)-4(4!)+... = l
l 2 (l!)-2 2 (2!) + 3 2 (3!)-4 2 (4!)+...=5
then
and generally 2(-
\)
~*nk (n
of
the form
!- 3
1.3.5
is
!)
(Arts. 98, 132)
-2S
-2,
aS + /?, where
a,
/;?
are
integers (positive or negative).

22.
Shew
that
l
1.3.5.7
[For an Application to an important physical problem, see Lovi:

Trans., A, vol. 207, 1907, pp. 195-197.]
23.
X*=F<r*.dx,
n
If
Jo
prove that
lim (lim
X
24.
From
>ao
n)
>oc
shew that
for [log (4 sinH#)] 2
-7r<#<-,
if
of Ex. 31, p. 473.
un = - 2n
If
f "cos
u n - un+l = (
prove that
%nx log (2 cos \x) dx,

-1
+ 1 ),
)/(2w
^ = ^-(l - i + i - ..3^)
and that
26.
>an
a:
also that
Deduce the integrals

25.
X )=\.
(lim
>-oo
the power-series for 4[log(l+#)] 2
and obtain a corresponding formula
Shew
= 0, Km
Prove that
rir
if
/:
is
-r2 )cos7i.r
5^ = 0,
-2rcos#-H*2
(1
Jo
if
= 7rr",
Deduce that
k = aK and
if
[CATALAN.]
an integer and |r|<l,
= aA,
n
if
not divisible by
is
tt = v.
where
*,
are co-prime integers,
f,
Jo
27.
(I~
Prove that
If also
ac-6 2
if
is
is
positive
positive
and y 2 = cu?2 +2&.r-f c,
and equal
to
jo
prove that
where the angle
lies
between
and
TT.
[For a discussion of these and other similar cases, see BROMWICH, M<
of Maths.,
vol.
35,
1900,
]..
13I.|
Prove
28.
BLLANEO1
\n-<
RAfiTS
th;i
'gfiKJAWfOll-
all
rmta
I- -a!
"I"
th:it
(0
'Vt
"--6*>0),
/'
when
-s
form, \\hrn
(,'
\*
the second result in a real
ii'-ât
/*4w
Provt-
29.
that
tli. MI
IT.
th-
if
"iivergent
and equal
tli.-
fM
that
Converges, |.r'\il-.l
h.-d
/.'!".
.nvergen
2
log (4 co8 .r)<^)(.r )</./,
Provr that
30.
t-tnU >tcadily t>
</>(.)
l..r(
the sense defined
in
HIM
'"/).'-
/'
HXRDT.]
l'i
l.y
in-shriin (Ch.
</'-':l.<-dy
(r,8>0)
A, >X-*r.'"
-crfc- r(r)r() an J(r +)*.

-
also that
0<a<;r,
if
lim
/'%'W^coea-h^
A,
HAKDT, Messenger of
Provr that
31.
tin-
if
ivul
H^Jo
Piovr that
32.
tlian
for all
C01l\ri ur,. n t
in
if
j:irt
:--J,
r.'W.
:>2,
j,j,.
159.]
]-ositive,
^ 1^
n!
l+^~2r^-fl)...
tin-
intnval
mil
val
intr U'I-al
''
if
fl
to
/,,
nd-
to
its
fl'olii
(./.
tin*
/')
the
function
/(.,,
n)\
ift
less
ml has a
fum-tiiMi
tlli'll
Urn
pi-o\ id.-d
that
ao< .niifiin
limit
uniformly
De<lu'- that
0,
r
lim
in
any interval whi-'h
[See &\>
'.
[/>
+ (l
HARDER MISCELLANEOUS EXAMPLES,

WITH REFERENCES.
Prove that the function which
1.
and
when x
to
repeated limits
is
irrational,
is equal to 1 when x is rational,

can be defined by either of the following
(1)
lim [lim {cos(m!7T^)} 2w ].
>oc
>
[PRINGSHEIM and LEBESGUE.]

(2)
where
lim [sgn{sin 2 (ra!7r#)}],
sgn
>-oo
2
X=lim - arctan(w^T).
[PIERPOINT.]
[This function cannot be represented as a single limit of a continuous

function of x, n
BAIRE, Lepons sur les Fonctions Discontinues, pp. 75, S3.]
:
2.
to 0,
when x
interval, however small,

3.
where
which
when x
If f(x) is equal to q,
and
pjq,
is
is
irrational,
although f(x}
represented by the rational fraction

prove that f(x} is not finite in <m//
is
finite for
every value of
Discuss the solution of an equation
f,
/(*)<*) = *,
by forming the sequence
/(an+1 ) = /<XO.
are real monotonic functions,
<
[See SOMMERFELD, Gottingen Nachrichten, 1898, p. 360; and

Exs. 10, 11, p. 18. An extension to complex variables has been
FATOU, Comptes Rendus,

4.
Prove that
t.
To prove
the
in
compare
made by
-()'. 0<a<,
if
2>n <?,
then
()
143, 1906, p. 546.]
x.
first result
no,
*&
I,\Vn-vn+l
\<l+^ + ^f
take an integer r so that ra.^

r
oo
<(;-+l)a
then
2^,<-.
2><-,
a r +i
a
i
To prove the
1
v
Z
Bothat
second, take the integer
\Vn-Vn+1\ = V\
,
Vn
Vt
so that
5a^7r<(+l)a;
then
II
AIM'I
5.
\vlii.
condition
t.-iuU -h-idily
-li
\\ht-n
that,
tin-
.s.- (
infinity with
t..
"
|i].-iir.-
or
(so that
iiumotonv-
li,,,
-*
iv
fr
ntoutary
r..ii\,-i^.-M.'-.
with
ph-,
u monotonio,
[('..in
6.
"f
'
intimtv,
t<>
(1)
7.
<KIM.;
pair Arts. 9 and 166.]
th.-
tlii-n
huth
It'
as n tends
.verge and diverge
tdily
l.
JftO
tliat
tends stradih
\\hi.-h
,,
t.i
infinity with
with
inlinitv
rapidity
n..t
pn>ve that
l-.s>
OJ
with
stradily t" infinity
trn.l
l>Mtli
uith
a
M"t
rapidity
[K>r
h-ss
and
tli.--.-
.f
pi-Mil'-
than that of
as
11.
1906,
/'..
/;,.
;?-.,
arr
...
i,.,,
Plj,
i" :
and
whi.h
l'."_',
''.
p.
:V.H
ti-nd-
that
10.
I'rovr that
if
irrational
wr can
find a SftjueiK-f
>in
ntunlu-i- l.rtw.-,-n
any
i>
12.]
p.
-uch that
'
i-
it"
fy
t"
HIM
-1
very rapidly, thu-
"lit.
Sln-w (MC K\.
jrhen
that
IM-.
t!
'.MKKI..
incn-asr
...
//,.
that
9.
_J
intru'tTs, siu-h
least
tin-
between
1'
ognlog(log
>i
nuinlnTs
see
Bfying all the logarithmic
v
~
i* ~*log
tli.-ui-eni8,
29.]
foil.
v'.V_J_
wlii-r-
similar
..f
p.
Bejmond hai oonstmcted
l>u U"i^
8.
criteria
a nuinln-it.
rin(n^)
.-Mrfiiif
tint
and
limit.-
H!>(ft+l)W
<,
then
I
feduce
I'.y
:r
tljat
l'>in(//-Vyi caDl
1IM7,
ii
= (2A+
llatei
)-'/'
tliat
p|
r-:5).
when
is
//
an-
divergent
U
fail
hisruss similarly the

[Littl<-
A and
a niulti!
2i
apjirai- t-
irrational.]
^).
kno\\n a- to
th-
of tl>ee series
when
0/ir
HARDER EXAMPLES.
486
Prove in the same way that 2 sin

t
where A and /x are integers.
11.
= A/2'
$/7r
converge
only
if
Shew
12.
can
(2 #)
that 2sin(tt!7r.r) converges for
rational values of
all
and
also for
#=(2ra-fl)e, 2m/e, sinl, cos
For instance,
if
x=e, we have
1.
sin (n \ire) = (- I)"-
sin (/Or),
[RIEMANN.]
where
We
We
are
can now apply Art. 21.

note for comparison with Exs. 10-12 that the integrals
all
convergent (see Ex.
8, p.
oo
Prove that
the two series
13.
if
the series f(x}= 2 a n xn converges absolutely for x <1,

\
both converge and are equal if |#|<1, |c|<l, |y|<l.

[Compare Ex. 6, p. 126.]
n
[The series obtained by writing a n =(- l) /nl, y = c in Ex. 13 have been
considered in Dedekind's edition of Dirichlet's Zahlentheorie
and those
found by writing #=c, an =(-l) n /nl or 1/w!, have been discussed by
Pringsheim, Chicago Math. Congress Papers, p. 288.] ..
;
14.
If the
two
example are not equal, but are con-
series of the last
vergent, prove that
and deduce that
if
x=rx-x
[That the series need not be equal

15.
If
we put y = \
then
when c>l
is
we must omit
in Ex. 13,
and write
clear
the
from Ex.
first
term
28.]
in
</>(.)
,.
fcW-Sfr^J
then prove that
Deduce that
(f> l
if
(x)-(f> J (xc)=f(x)-a
r
Xi(- )
then
16.
where
'
Deduce from Exs.
m=n
IH
.
()
= f (')-</>! (''),
=
20, 21, p. 251, that if
0=y<x = b
omitted from the series; hut that the sum is zero if .r=#.
when
..f suiuniat ion
is immaterial except
Shew further (hat the culrr

#=0, # = (nee Ex. 11, p. 92).
[HARDY.]
HAEDEB
ohtain
17.
sum
tin-
i.
sen
of tin- i-j>eated
v-xpii';r/(i/M-
\MI
+,...-,;
1,
fi-'"
[Kk'.M,
li.it
ii.it
.vergent, sec
!4
_rer
to
A!
\\hi.-li
-f \
Tim.-.
Ol
1C
tends t<.
through >maller values,
lar'_ 'T \alues, \'q\ tends to -^; hut {^}
ncare*
definition.
.is
l'i.ve that
II
the -ymhol
Let
18.
"
tin-
m>
lia-^
to
||
M
/<
tends to
.'.
through
di><ntint;
-.th.-r
tl
^^...
is
uniformly
ntimiity
t>
|i"int
ry
Thus the
in.
seri'->
i-
valuer of
all
./--/-
hut that
./
;
\\li--ii-
-2n>,
the series has a

,,dd
i-
j,
and jiim-
diaCOntinuOQfl intinii
[1;
19.
the lines
Ail.
i.f
'>')
the equation
that
y=
ilnuhlr ]M\V-l
-all-tied hy a ]>.\\ er >ei ies
\\hei
\ff\
hi-
-an
li
.-. >l\
i.solnlfly
-rr^'Ilt
for
y- I
\\hieh
ei-rtaiidy
Shew
20.
.-..nverges
i>im
<4aM
in
is
iii
an <>Uain an
Art.
.\\li-:
<
//(.'
>
I*rove also that
intr^rr.
{...sitive
lain
.I/
'.6).
and
where
-Ll/ 4- '//).
the
f>r a I'uin-tii.n //(')
.',
\x
tl,
in
in
if
hy the tenn>
dete)Iliine<l
i-oli-taill
\\e
if
\\iite
4(4?)
tht-n
i't'
jm\\fi-
OOefficienl
Ajijily
21.
is
/-/A,
./
of
the eneilieient of
.<
tin- d'-i'-rmim'> all
.,
in
in
the r\jaiision
the
/!
which
i-
the expansion of
the metliod
taking
If
.'
(1)
in
e..etVieiei.'
.-
l"" vt
!-
th;lt
<M
^
l-fr)}.
the
HAKDER EXAMPLES.
488
Deduce the value of <()

x2 + x = l, then
(see also Ex. 32, p. 473),
and that
if
# = iU'o -
1),
so that
<*>(*)
[BERTRAND,
= AT* - (log x}\

270
Calc. Int.,
4>(a*)
= J.TT 2 - (log x}\
ROGERS, Proc. Lond. Math. Soc.
(2), vol.
4,
1906, p. 169.]
22.
(1)
Prove that the
differential equation
can be solved by a series
where
Shew
=A
that
|
Prove that
(2)
cos # + Z? sin #,
if
is
v(x)
un (x} =
w n <| x\
2n+l
tun _i(i)&\i\(x-i)dt.
/n\.
continuous in an interval
(a, 6),
the equation
has a solution y = û n in the same interval, where
Shew
that
if ^8,
y are certain constants depending on the nature

|
un
< (3yn/(2n)
[A
1.
C.
of
v(.'') 9
DIXON.]
Results allied to Abel's theorem.

<j>(x) tends steadily to zero as x increases, but the series
diverges, shew that as a tends to zero (from the positive side),
If
23.
/CO
An
ci0.
<$>(x)e-xdx,
illustration has already occurred in Ex. 5, p. 419
[LE ROY, Bulletin des
Extend the
24.
results of the last
in particular that if
ivTi
.'o
math.,
another
t.
is
24, 1900, p.
series,
and
(^8
î
cos*^-o>'
.r,
I,
pn>\.-
[HAI:I.V.|
Abel's theorem (Arts. 51, 83) can be extended as follows:
Siij)|)..-.-
sum
245.]
?/-fl),
where rcosw = l-.r, rsinw = l-y/, & = 2-2s.

25.
given by
do
rot)/-*"
-V^
example to double
0<s<l,
r*
sci.
^,<f>(n)
that
and that
^H H
vn
is
lim
is
/'-ply
indeterminate
a function of
nX-0,
x which
2w r A- +1
|
/-, (
(in
('
-sard's sense)
satisfies the
< A',
and
li;i>
the
following conditions
SARD!
\\hile
then
/=/.
inn !/.,
rwi(
ii.
havr
.'led
Keen found
l.\
aN. Moo
1907.
Vol.
26.
'
and
asymptotie formulae
? + *TT-
"*X'2
that
seen./
all
27.
will
..-ills
iH-xt
Mob?
trim
in
tin-
l.i-I
asymptotic fonnul;i
fo mill f>
!.
j>
:iji|>roxiin:itioii
nirthotl
thi-
Apply
r...<^4[(7r/.r)-l].
tO cMiitinue the
;itt.-ni|.t
tin-
in ol.tain tin-
I.'JI)
,..^..~''
Us-
./-,
See also
'
1+.I-U.
in
tin-
1907,
-1.]
formula (Ait.
Ajtj.i.
under
II
of
Art.
a-
BOTO: and a- a iiiattM- of
(l-'x^-**!*.
!:*:>
t:
t<>
Y'
and
the
in
28.
that
'
/"(
i-
I.
sain.-
from
l)uliii-if
that
ju-ovc
way
lv\.
.").
p.
tin-
:U^.
ami
tin-
pow.-r
i-ontimioiis function in tin- interval
any
(<>.
_'-). a
j>.lvii":
\jP+..
can
!"
ilctfj inini'il
^.
>
that
.
29.
,|,aiv
I',
Shew
that
;.
l.y
|.i..|..'i
Ch. 1\..
the
the CO1TW] ..... linur value of t, can l,,.th

>-ducc that, as
t.-nds
1, the
I
I.
In>.
<
V.]
maximum
l.roii^ht as n.-ai
value of
t..
us
we
please.
'
-f...
o-'-illato
If
30.
f
'_
et
MC limits
write
,/
(\
and
I.
1.
:V(-V
MI'rovr that
if
/'( r)
'(
:hen
and
3.
w,
HARDER EXAMPLES.
490
Verify also that, to
five places of decimals,
^(1/^/2) = -00275,
and that
Shew
if
F(l/6*)= -13185,
further that, as x tends to
= 2,
c = 6.
if c
g(x) tends to
1,
and deduce that
f(x) oscillates between limits at least as wide as

if c=2,
50275 and -49725,
or
-63185 and '36815,
if c = Q.
we
If
31.
write
,,
the series
Prove that
and
= ( - l) n /nl and
=
Ex.
15,
we
get
F(x)=f(x)-g(x\ then
if
verify that
Shew
-logo; for x in
if c
^(1/^2) =1'1960,
also that g(x) tends
as
to
x tends
= 2.
to
and deduce Cesaro's
1,
result that f{x) has the asymptotic representation
- log Hogi
Vlog c.
[For a treatment of the series
8-13 of a paper by
integrals, see
p.
269
some similar
results
[See Ex.
133.]
4, p.
Exs. 30, 31 by means of contour
in
HARDY, Quarterly Journal,
vol. 38, 1907,
have been obtained for the series
V u n a-,-A,,*
i
where X w tends steadily to

Sup.
(3),
11, 1894, p.
t.
infinity,
by CAHEN, Annales de VEcole Normal?
75.]
Functions without a derivate.
Shew
x=a, an
that
32.
at
{F(x2)-F(x
such
that
if
interval (a
l
)}/(jc2
-x
a-8<x
In like manner
if
1)
8,
lies
F(x) has a finite derivate b

a + 8) can be found such that the quotient
between b-e and
<a, and #2
F(x) has an
is
+ t when
33.
and
may
Write F(x} = ^Lrn cos(s nTrx\ where 0<r<l and

$m (x) denote the sum of the first m terms in
if
a=jt?/
m where
p
,
prove that,
if
is
is
* is
an odd
F(.r),
then
an integer,
p+1
or
+ 3, and
F(ft-F(a) = ()"["/ "

(
^^L
where
x = a, prove
greater than
is
.V.
be.
let
while
any point
2 <a-f8.
a<#
infinite derivate (positive) at
that the interval can be chosen so that the quotient
however large
#, is
any point such that
i\
/?
= <//s m
I/I
^r^^^l
|-tf|<3ir/(r-l).
then
int-_-
slu-\\
thai
II
Deduce that ii
and ran In- iii.id.
>l+f*>(]
/,
that
-he\\
".
idered in
and
{fi'
h. \\c\t-r
-V
lai'iri-
has nowhere a
tin-
a,,
',),
found
a. fall
the type
the intei
i.f
in
wliil.-
/Mr), although cmtinin>u> f-r
K\.
h-,,i,,
i<>n
\\ fn-i-strass's
i'..r
/.<
f:
".]
tluM
:{:;
y=*
than
less
is
/'(/;)
!>>th
^(a) and /-Vy)
and
minimum in the interval (a, y). 11

of maxima and minima in any interval,
has an
\VKIKI:STRASS.]
fuin-tin leads to the
Thus /-~\ would require
-^ ;;-.
instead of
a=(?-i)/r, f^=qlf
/'(./)
all
(
_-.
prove that
int.-i
In-
d.-n
d.-tinit<-
deduce that
that
-h
nnnrcfssarily naii"\\ cnndit
\vlit-rc
/, can
ii
[Thr in\r-tiâti<.n usuallx î\cn
Sh.-u
/;?,,
=7
may
thr function
that
<
in
fall
n.,,
,,,,
example, and su
la-t
tin-
take r=^, <
pl-,
joints
1)*,
sufficiently.
UK! an arbitrarily small
containing
35.
ing
are ea
M (-
the quotient ha* the nine iign
r),
î
ail.itiaiily
conditions mi
['I'll.-
34.
Al.'l'l
infinity
nail.
has sometime-*
[It
Imt
this
stated
ln-.-n
cannot
l.e
tl
ju-oved
maximum,
at
rate
any
l>y
/^
minimum
of
th<
argument.]
36.
a
(1)
"f
similai
odd
\Veieist pass's function,
in
leads to the condition
.-|iondiliu c.ises when sim-> take the

Trove that with the notation
>
in
-hew
jii
the
>
4ifi.
|
ami dediu-e that

[The
last
iction has nowln-i-e a definite/''/-/.'' dei
\\
ciinditio',
iher similai
fancttons,
M0
I>IM.
Complex
37.
Deduce from Kx.
HI.
p.
X.)
Series.
that
-Jll.
it'
/-
i-*
prime nu
~\p/
where
and
')
|^"]
if
/<
.'
is,
is
is
if
if
4-1
a
i>
a -piadiat ic
multip
\ve
ran lind
.'ill
i:.
'"*
is
& multiple
that
HARDER EXAMPLES.
41)2
In particular, with
the residues less than 7 are
1,
1
+!_i+i_
_!+ +!_J_,_L__L_! 4
^2 3U 5 G^SÔ KT11 12 IS"
1
[DIRICHLET
38.
the
see also
is
the circle j#| =
mean value being taken along

if
the real part of q
so
_JL
vol.
31, p. 98.]
prove that (see Art.
positive,
and
4,
^7'
GLAISHER, Messenger of Maths.,
If the real part of
Deduce that
"
2,
1,
89),
also positive,
is
and hence that
^
--provided that the real parts of m,
n-m
r(2w+i)
are positive.
[CAUCHY.]
subsequence selected from

(a n ) has zero as one limiting value, prove that (a n ) tends to zero as a limit.
to a (a>0),
(2) If for every value of x from
39.
(1) If (an ) is a sequence such that every
prove that
tends to zero.
(cn )
(3) If for all values of
x from a
to
ft
(both
real),
lim (a n cos nx + ft,, sin nx) = 0,
prove that
(a,,),
(ft,,)
both tend to zero.

[G. CANTOR, Math. Ann., Bd.
4,
p.
141.]
2a n # (a; = + ^)
M
40. If the imaginary part of

converges at all points of
an arc of a regular curve (which is not a radius through the origin), prove
w
(from Ex. 39) that 2an # has a radius of convergence not less than p, if
is the least distance from the
f)
origin to the given arc.
It is not
enough
to suppose the convergence given at all points of a

thus the imaginary part of 2.a? n! converges for ani/
dense set along the arc

value of r, if .r = r(cos
+ i'sin 6) and
O/TT is rational.
[KALMAN, Math. Annalen, Bd.

41.
63, 1907, p.
3-J-2.
|
For any series of the type
v
"
are two constants A,
/x
such that
the series diverges
converges if A</?(.r)</x, and converges absolutely

constants air mniM-h-d l,v tin- ..... litions
if
/?(a
<
A = lTm
l
{log
A n |}/log w,
/x
it
A'
<
= lira (log ^H )/logw,
)>/x.
These
BARD!
is
It
assumed
ih,.
ban a radius of
./"
anil
t<>
igence equal
theorems can be
eHtal.li-1,.
<l
tlie
r'..i
series
(*-l)...(*
[LAM.
Sln-w that
42.
tli.-
infinity,
.\
-//
th.
it
!-<!.
..
numheis tends
i,-al
.mil
>-.',.
..iily
I/
\\here
-
A M =a + d.2 +. + ./.
ur c
!"-. tin- oondif
it
it'
he necessary
"t
[ton, pair Kxs.

43.
Consider
192-208.]
to
.<t*;uli'fy
if,
assumed that the
2o
series
does
l.ut
..-nt.
ooi
f..i-
K>.
!>,
is
It
..
36, 1906, pp.
-2^>.\
p.
V
'TrA/
\\ln-i
and
latiniial.
l'r-v.-
that
'
has a tinitr radius

(3) div.-ru'o
I!
valut- OJ
(1)
nd.
\i;i
for every
i/Otk,
..nvrr ;M
aon-zero \alu-
..f
vol.
3,
MX).'),
<.r
p.
441.]
Riemann's t-function.
44.
tliat
l'i-i'V-
the
sei
l+2-'-H3-' + 4-' +
nd
\Vhrn
term
in
this
...-
.,f
-lily
c'c.ndit i"ii
>
ii"t
i>
g) is givat.-r
satislifd.
dftiiu-d
lh- a-ynipt'ti.- --xpan-i..
,:il
a>
than
tin-
1.
OOJ
\+-2- + Z-'+...+
Thus
if
<
/{(.<)
1,
we have
r...-f-
f()-Km F(H
fl
and
sn
on.
function {(t)
It
is
IL
* M
-an
nimv
l.y
rlalM.iatr
thai
inrth.-ds
= 1
only one singularity, a poK- at
a singularity fnll<>\\ 11. (h. II.. which gives
analytic ami
lias
HARDER EXAMPLES.
494
(2)
The following are easy examples on the {-function assuming R(s) > 1
n(l l/p*)i where p is any prime number.
[EULER.]
2 =
[f( 5 )]
2A.(w)/M*, where X(n) is the number of divisors of n.
(3)
From
45.
(1) [f(*)]~ l==
(1)
deduce that
where n is of the form a a b^...kK when expressed in terms

factors, and ag = q(q + l)...(q + a- l)/a!.
,
(4)
From
(5)
The
(2), (3)
we
of its
prime
sum
of the
obtain the familiar result
coefficient of l/n* in the
product {(s)f(s-l)
is
the
divisors of n.
1
(6) [^(s)]~
= 2/x(/i)/w*,
where
number
(7)
and
<f>(n)
denote the number of numbers
d denote any
(8)
divisor of n,
less than and prime to n

and write ^(w) = 2<(d). We have then
ft
f (2*)/f ()=2(
(9)
46.
(all
n has any repeated factor, +1

- 1 if n has an odd
different), and
if
of factors (all different).
Let
let
is
fj,(n)
n has an even number of factors
if
Shew
+/'= H(l + 1/p')-
1
-
[CAHEN.]
that the series

(*)
converges
- l) a+/3+
and only
if,
if,
= l-2- + 3--4-+...
R(s)
is
positive,
and prove that then
&)-<! -S*~)(('}
Shew
also that,
if
0</2 (*)<!,
Deduce that
and
so prove that
47.
If
0< /(*)<!,
21
-T ()
s
() cos ( Jw) = 7r f( 1
shew that the
series
r;(*)=l-3-' + 5-'-7-'+...
is
-.nvergent,
48.
and that
Prove from the definition of the {-function that
= ~ J, f (2m) = 2*"
=
f(-2m) 0, f(27 + l) = K
C(0)
where m is any positive integer.

Deduce that the relation between f() and (!-*) given in Ex. IG is
true when
is any
that it is true for all values of * is proved
integer
;
by
Hi-in;iiin
(/
Ml'LES.
Tin- snirs, in
49.
\vhi.-h
t.ik.
/.
dues,
ha-
It
|)i-M.t'
.?,.,!
in.
I-
M"t
i-
\.-l
L+I+J
that
sn
tin-
nje
<*<!,
if
luit
h-
OOmpll
there.]
Infinite Integrals.
IVnVr that
50.
if
*(,
in
A).
('/,
su.-h
in
i'Ut
way
that
tin-
t"
Mjnal
liiu
the iiit-r\a
51.
tin-
Apply
Art.
177
ly rout inm-d
iilxlividcd
<>f
>t'
example and
la-t
tli-
that
pi..vr
in
((>(.,-)
lii^-<-t
52.
ion
into n equal
th'
tlif..ivm
pr^v.-d in
th--
>inall
tyj.-
true
th.-
in
iiitt-i
v;il
any put of the
P>y
writing
-"sV,
./
.!,,
-/^.înh'/,
I.I-MV.-
that
^
(dW+P/4^-lJ
j["(V-4>/jj
iind
iW + b*l**)e-"
/-
l/4a)e--*,
53.
ly
tin-
sain.-
.iiisf..uu;it
in
as
in
nd
tin-
?>
\vlu-!
}>rinripal
an
ci-en
valtu-s arc
t>
fim<-ii..n
!>
\\~.,-,\
if
on
IHM-C-
the *econd equat
HARDER EXAMPLES.
496
Similarly,
if
#()
an odd function,
is
Special examples of these are given by taking /() = cos

[Compare HARDY, Quarterly Journal, vol. 32, p. 374.]
54.
an integral
If
f(x)dx
J^
lira
exists.
We
we may
oscillates,
call it
summable,
if
tin
{" e-t*f(x}dx
may, when convenient, indicate this limit by writing G befon

Deduce from Ex. 2, Art. 172, that the "conditu
the sign of integration.
of consistency" (Art.
100)
Prove that (compare Ex.
is satisfied.
aw
lain (*)/'
=Qt
GJ\inxdx=\.
[HARDY, Quarterly Journal,

55.
Use the
converges for
result of
all
36, p. 74)
Ex. 54 to prove that
if
vol. 35, 1903, p. 22.
the series
F(x}=^,a^
values of x, then
provided that both series on the right are convergent.
Deduce that
T
-?dx=*J
cos mo?
cos
4m H-sin 4m
),
where
56.
Following the lines of Art. 109, frame a definition of uniformly

/.on
summable integrals
and prove that
if
J
in
is
an interval,
its
value
is
f(x, a)dx
a con tinuous- function of a
is
uniformly sunnnable
and that integration
permissible under the sign of integration.

57.
Discuss
parameter.
shew that
also
the
In particular
question
of
differentiation
if
"-=-
'
with
respect
to
HAIJM
hi
in
u = l[-<r\i(e-*) + e- a \i<
samr way
th.-
he ex pan.-.
58.
ta.,1,
ln-\v
-...,
tliat
and drduo- that
*-l)-
ta.
that
I'l'-v-
59.
it
'n
ilisli
lh-
-imilaily
interchani
Deduct' Sti-k-
'
u= ^
ccs
Ol'tain
60.
tin-
that
results
and
witli
j yw
>
= 0,
ri
sin*-*
,,lx,
s.ltitiui
-0,
^c:"
intinity
tli-u
if
(.,-<-.,-;/
a.-yni|t..ti<-
l.y \viitin.
(..
limit-.
udini:
_!f +
tln-n
liniti-
and
<>f
A//i'
th- ditlnvnt
ial
i-jM;iti.n
(comi-aiv Kx. 58)
equation n
]r..vt- tliat tin-
/
0,
\vhirh
thf >>lutiui
.L'ivi->
f*e*4
8 {
[ST--KKS,
Trigonometrical Series.
>hc\v that,
61.
êi's
i:
<
(<j
wluTf
pt
<<.
= -/
and
iu
the sunn.
'akes
all
]
multi].
[Wi-itt-
La
spect to
-i
by
|>.
190.]
HARDER EXAMPLES.
498
62.
Shew
that
Deduce that
with certain restrictions on

63.
If
is
and
any integer and q
E
\p/
[H. N. DAVIS.]
<f>.
an integer not divisible by p, prove that
is
= _ +
2
p
sin
2p n=1
(2^0) cot (nO\
where E(x) denotes the integral part of x and #=7rr/p,
being any
integer prime to p.
r
Writing
= 2,
deduce that
if
p, q are odd,
where a = 7r/p.
[See Ex. 21,
64.
Shew
[EISENSTEIX.]
245, for the first part.]
p.
that
if
(an )
and
(6 n )
2 F(x) = $a x
is

Prove that
all
tend to
2 -^ (
0,
the series
cos ?w; + b n sin w^)
values of x.
Deduce from Ex. 4 and Art. 80 that

lim
-- { F(x + 2<x) + ^(o; - 2a) - 2^) } = a + 2 (
a-*0 4tt
for all values of
65.
cos
?i,r
+ 6 n sin nx)
for
which the right-hand side can be proved to convex

[RIEMANN, Ges. Werke, p. 232.]
Prove by means of Ex. 4 and Weierstrass's
J/-test that if
(A n ) tends
to zero, the series
converges uniformly for all values of a.

Hence shew that with the notation of the last example,
lim
[F(x
a *0 ^ a
provided
that
(an )
+ 2a) + F(x - 2a) - %F(jc) ] = 0,
and (&) tend
to zero,
converge or not.
no mattei* whether the
,>
[RIEMANN,
/.,-.]
\I;M
II
I>-du
66.
tin-
fi..i.-i
<-'|Mal
/.!<
I..
at
the series
if
00+2(0,, COB
Imt a timt<- inn
all
th-
in
'
ii
the coefficient*
<
[<;.
eh.
method of representing
Mittag-Leffler's
67.
/'(
It'
in such a
t.-nds to infinity at
p)
uav
<-a<-h
win
t'linrtion
I
It
and IMM!S
Shew
68.
-liosen so
if
d'-\-"i.l
tliat
with
to inlinity
that
(1)
tin-
amnr(l
i-
an infinite
>.1
ufiice (a*)
it.-
integer
'
plane.
and
\\
iiiiil
!'
that
lim
we ran
tli'-n
point
functions.
-I,",,
Indicate the
i.
..f
th--
th.-
tin*
|
-.
-i
ios conver-_
Unit.-
a,,
j.a;
never decreases,
n,
mains
than a
h->s
Ix-tw.-cn
\:
ti.\.-d
Mitta^-Li-tHi-r's represei
fractions f.r roo<
lim
Prove that
in
moduli
niu-xinn
seriea "f partial
Ha to inak-
.-insularity
(.'
//)!'(.')
and deduce that
= ( -1)"
x tends
as
',
to
'
Shew
also that
)=
e~'t
/
.'i
<:j)
Indicate the relation of Mittag-Letfler's theorem
r
I
where
(1)
'
l-)(2-q)
J+2 +
Art. :M, that
Shew, as
Q(f)
0,
if
o<y<-2-,
l.ut
Indi.-ate the n-latinn of this result

" tliat
if
is
to
flf(*)
--,.
,f
Mi-
poril
;''
where
equations
~T2
69.
the
to
[('-l-lo
g (^-
27r<y<
HARDER EXAMPLES.
500
Jacobi's Theta-functions.
70.
we put x=
If
zi *
in the functions of Ex. 16,
obtained are called theta-f unctions

a-j
(<) =g (
- e^,
=/(
[There
is
71.
e**, q)
- e2 ^
=
and Molk
q)
= 2q* sin
<f>
Chap. V., the functions
thus
2q$ s i n
3^ +
2^ sin
5<
. . .
= l + 2? cos 2< + 2? 4 cos 4< + 2?9 cos 6</> +

= 1 - 2 cos
some divergence
as to notation
we adopt
. .
that of Tannery
so far as the suffixes are concerned.]
From Ex.
Ch. V., prove that
16,
where a = ^(^) + ^),
/2
= ^ (<-!/')
the products 2 (a)$ 2 (/3),

Deduce that
and obtain corresponding formulae for
etc.
and obtain other similar formulae.

72.
From
73.
if
Exs. 18, 20, Ch. VI., prove that
^)=^^- n
[(i
prove that
where n may be any
Shew
also that, as
integer, positive or negative.

n
to l/^ ,
x tends
F(x)l(l-a?q) = (-
lim
l^V/fi
where (as in Ex. 14, p. 105), q n denotes 11(1 -q Zn ).

Deduce by Mittag-Lettler's method that, "if \p\ lies between
[It
may be proved by more advanced methods
Fonctions Elliptiquen,
74.
Assuming that
""'"
Deduce that
t.
in
(see
3,
Ex.
Art. 479) that O(x)
is
(see
and
Tannery and Molk,
zero hnv.
the formula of the last example (?(.)
-*'^
18, p. 106),
is
H
75.
Induce
\i;i.-
between
and
|1
,_!___
i-it+i^p+ri^q*
t\p)F(x)
#=
Write
and
r-r*-?*"
and deduce
'
1-
l.y
!+;>'-),
=1+HV'
.
-^'tin-
76.
l.i\v
Kx.
r'l-.m
(1
imliff-
tin-
i.f
pn.vf
>
every p-itivf intf <:!
it
\Vrit inur
.'-
sum
\pressed as the
In-
Bd
K\. 71. prove
in
<-an
A.
[.I
77.
thf
..n
i.
pp.
oi
L':i:.
'
'
'
,
M.H
any prime
\vav onlv) as the

can lie MI -\pi.
sum
4/-+1
f.u-m
;
l>ut
n<>
can
l>e
prime
expressed
<>f
tin-
(i
form
-1/-
+3
[JACUHI, /..]
Series defining functions

78.
tin-
..f
twu squares
>f
which have no analytical continuation.

is
1'
p..sitiv.-
ami <></<!.
th.-n
lim s '(l
r-*l
when
that
I-diii--
tin-
appi'>a'h
" ami
|).'int
are positive integers ami * is all.>\<,

tin- unit .-irele
along the radius,
I
2y'~/' s
where
/'
[The
r=|x|,
(see E.\
1.
if
= 4/-+l
im-lh..d faik
\\li.-n
/'=+/,
a
\-
"f th-
we get
ami
79.
= 4* + 3;
if
=4/-.
'-l)V*
ml> tn
tie
Kr'in
larity in
thf
tin-
iplf,
limit
pmve
^f unit-ri!
that
ti
1
\\i:\t
the function
'nued bey<'
:
[WK1KRSTRA88, (?e. Werte, Bd.
2, p.
228.]
HARDER EXAMPLES.
502
Shew
80.
that the functions
tend to infinity as x approaches the points

ex p C2,7rib/am \ exp (îri
respectively along the radii.
Deduce that these functions cannot be continued beyond the unit-circle.
Apply Ex.
81.
31, p. 253, to
prove that the function
f(x) = ^n-
n an
has a singularity at every point of the type exp(27ri'6/a w ). The function

therefore cannot be continued beyond the unit-circle, although f(x\ /&),
fi( x \
'
(Art. 84) all converge absolutely at every point of the circum-
ference.
82.
[PRINGSHEIM.]
Assuming
Borel's
theorem
(see Ex.
83), prove that the function
cannot be continued beyond the unit-circle, although <(#), <$>i(x\

2 (#), ...
converge absolutely at every point of the circumference. [FREDHOLM.]
<
all
83.
Although
it is
beyond the range
of the
methods given
in this book,
we state the following theorem, which includes Exs. 78-82 as special cases
The function 2cw^ cannot be continued beyond the unit-circle, provided
:
that the integers a n increase fast enough to satisfy the condition

lira
and that |#
=1
(a n+1
- an )IJa n >
0,
the circle of convergence.

narrower form of the second condition is often convenient in practice,
is
namely
n
n+l \} =
lim \og{\c \/\c
Q
-
n+l
[BOREL, Liouville's Journal de Math.

84.
(5),
t.
1896, p. 441.]
2,
Prove (by means of Ex. 79 or Ex. 83) that the function

1
has the unit-circle as a natural boundary.
Prove
then
also that
M<1
if
/(
and
>
>
J?*
= -fg > i
1
and so shew that the function inverse to f(x)

Extend the argument to the function
is
single-valued.
where 2cn is a convergent series of positive terms whose sum

and the indices satisfy Borel's condition (Ex. 83).
is less
than
It is evident that the series given in Exs. 78-84 have all large
the sequence of indices; but this is not essential to secure that the
function cannot be continued beyond the unit-circle.
85.
in
.ii'Lix
\i;i'i
11
fcfon
which has mi
in
-japs
tin-
.,f
sr.|ii.-nr,-
indices.
1
Pi:
[Compare
.-,
Kummer's
\NV
86.
I.
1.
pp.
Series for logr(l-i
-I-.,
find
the
if
that,
real
part
,
;
of
./
lies
]r
and
L-fcr-2
log
Now
is
t.
t-asv
/.'
"-
constant.
lo
f^)'
iFl
(Art.
Kumi
obtain
Power
if
thr
ju..\r that
it
s.
i-
177)
for log !"(/), nat
(A-r)(C'-|-log2) + (l-.r)]og7r-41oK
87.
be
W^-^T-J
ler's
we
summation can
vcrifv that
and
so that
to)
tind
r^l-
it
orcfer
'<?
\\<-
Series in
Two
Variables.
y=y<M
al-
til
valur-
..-}
that
[AHKI..]
I'l'-'V.-
surh
that
it
unifol mlv
ais.
i-
tliat
OOIlV6fgen1
f"i
all
\a!';.--
'
l*l^l*ol-i lyl^lyol-,
?
.
is
any
posi-
A.
to justify this strj
[\N
!',
luit a
in
'
the analysis appears to be rather tedious.
pear to
io obtain*
bt-
KIKRSTRASS.]
HARDER EXAMPLES.
504
88.
Suppose that
r is
/, such that the series
This
\y\=r'.
is
maximum
given
then there will be a
maximum
value of
=
absolutely convergent within the circles |#| r,
value of r' may be regarded as a function of r,
given by a relation
O = 0;
<(r,
and
r,
r'
are then called a pair of associated radii of convergence.
Let
where
p + q=n.
Then prove that
~'
1
are a pair of associated radii.

r
and
r'
'
Eliminating
k,
[LEMAIRE, ZtoW. des

89.
is
shew that the
relation
Prove that the
convergent
For the
22-
series
if
\x
and deduce that /
set.
math.,
relation
is
t.
20, 1896, p. 286.]
-Vy*
<1, |y|<2,
defined as a function of r
is
series
shew that the
by the equations
^T
r+/ = l.
Let
90. Evidently / = <(?) is a non-increasing function of r
the upper limit of the values of r for which <(?')>0, so that
Shew
between
is
that for the series which
we have ^=3, and

/ = 2, (r^J);
r'
is
denote
the expansion of
= l/
[For an application of this theory to the differential equation
see
GOURSAT, J5^.
91.
where
o?e
â ôc. wa^/i.
The function / = $(r)
0<r <r <r

1
satisfies
c?e
France,
t.
35, 1907, p. 81.]
the inequality
1,
logn, log </>(?
1,
log?-2
1,
logrs
</!.
[FABRY, Comptes Rmdus,
t.
137
HARTOGS, Math. Annalen, Bd.
62.]
ttABDEB
From
92.
Kx. !!
\MN.
i:\
it
limits excluded.
uhid,
for
here /?=
Shew
93.
a
\\i'
constant
In-
<.nly
l.e-miiim:
int.-i-val
as
..h,
lv\.
!>J
D an
(and not
sh-w>, not neceasa
including r=0).
[For solution.
The
94.
ll
oi
..]
may
st-i
r..iiv,-r;:-Mt.
Thus, M
;i
series,
power
#=1, y
95.
kind
'I'his
-,
/'
ll<ith.
can
funveriri-nft-
c.f
fin,-..
v,,l.
only
JjniJ,
i',
for
[11
96.
.....
'I'll.-
wh.-n
prove that
it
th.-
may converge
scii.-s
siinimrd
st-ri.-s
<
-iu
|y|<l,
the regions
1!><>4.
iM.]
1.
1
wid-r region than that of absolute

r columns, or
diagonals.
in
oonverg
Ke]ae>
particular,
by rows,
'I
for !/(]-.<--//) converges absolutely
or by .-olumi!
161.]
j,.
of \alu-'> "f
|,ails
-\ith
th- azpl
i>
but for absolute convergence
tMWIOH AM- HAIM.V.
isolat'-d
al>-. lu-
alisnlute con-
>f
the scheme
l.y
whii-h
(1rges for
nt
though
tside the reîmi
sete of
ami
if
|.r|
+ |y|<l
it
|j?|<|lin
oi
a dia_-
09
97.
I.
the
double power series

-
...,
'
-f ...
converges absolutely
I

\y
<
if
* +|y|<l, by dia-oi,al>
it"
r+^;<l, by rows
[HARTOOS.]
HARDER EXAMPLES.
506
98.
If the series
22ev
satisfies
the condition
m=0 =0
for all values of
/A,
and
v,
if
any one
(w,)
is
(m)
<K
of the three series
(m)
(n)
(w)
convergent, the corresponding one of the three limits
lim f(x,
y),
(*,)
lira f(x, y),

(*)W
lira f(x,
W(a)
y)
In these limits #, y
is determinate and equal to the sum of the series.
are supposed to approach the common limit 1, simultaneously or successively.
[BROMWICH and HARDY,
I.e.,
p.
168.]
99. Extend the last example to cases in which the series 2aw>n can only
be summed by a mean -value process similar to Cesaro's.

100.
An
example of the theorems of Ex. 98

which
is
I.e.,
p.
173.]
given by the coefficients
of Ex. 13, p. 92, for
2am n xm y n =(x- #)/(2 -x-y\

,
An example
of the
mean-value method
is
given by writing
I.e.,
pp. 169, 175.]
[For various extensions of the results given in Exs. 98-100, see BROMWICH,
Proc. Lond. Math. Soc., vol.
6,
1907, pp. 67, 74.]
INDKX OF SPECIAL [NTEGRALS, PRODUCTS AND

SERB
(The numbers
series.
Asymptotic
Kul..
Stirling's
serie-
155;
logarithmic
Integral,
Elliptic function series.
TheM .funelioiK.
172;
!W,
li>7.
luti,
r>0o
:;:,M.
various series of
fr
I.
180.
172;
Integrals.
.
171.
Dot-el's,
268
et set],
-ion).
171.
Klliptir, 17,
Krn.r fun.-tion.
:iVJ
Killer's constant,
182,
K\I...H, ini.il or Logjirithinie, 825,
-. UK.
l-J.
KourierV,
17'-'.
FrulhiniV, 182, L7&
17.
210,
Sii..
_'!_'.
.
Non-convergent
.1.
IK)
Hi.
series.
,i".
J^^. 817
:;:.]
See also
Numerical
ui.
series.
'
L87,
1
Donl
rsfor T,
-""'fnn-fcn
1
).
Ui^.
114,
SPECIAL INDEX.
508
Numerical values.
n
2( - l) /n, 58, 155 2( - l)"/(2n+ 1),
21/?i!, 22 ; 21/7iand Euler's constant, 323
2
B
156; Z(-l) /(an + 6), 161, 481; 2(-l)/v/n, 57; 21/w 59, 71, 93, 187, 324;
2
3
2r
234, 324, 481.
2(-l)"/(2w + l) 479; 21/?i 59, 324; 21/?i
log 2, Iog3, ... 154, 480; ^2, $2, 171; sin (71-), COS(^TT), e**, e* 146, 147;
;
475.
e),
Power-series.
Binomial, 89, 150, 225
(Ex. 21).
sum
of squares of coefficients in, 166 (Ex. 33), 249
Exponential, 143, 217, 406.

Geometric, 15.
Hypergeometric, 35, 52; limit as x-*l, 42 (Ex.
2),
105 (Ex.
13),
161 (Ex.
5),
171 (Ex. 13).

Inverse sine
and tangent, 155, 169, 170, 224.

Lagrange's, 140, 174 (Exs. 30, 31), 487 (Exs. 19, 20).
Logarithmic, 152, 162, 223.
Sine and cosine, 146, 221.
Theta, 93, 106, 107, 350, 500.

2
n
2
ZnP~ l xn , 2z" 2(-l) n #" 2x
,
2*'",
133,389,489,501.
*/(*-!), 234, (e*-l)/(e*-l), 235.
Products.
H(lq n nKl + ^-^Kl+g "- /*)},

2
),
Gamma-product, 102, 461, 463
sine
105,106.
and
cosine, 184, 231.
Wallis's product, 184.
Trigonometrical series.
Convergence of 2vn cos w0,
2>
sin nO,
50
sin (w2 0),
sin (2n 0),
sin (n! d),
485, 486.
),
?^,
n
v
rM
COS
(
Vi0,
?ism
157.
114, 115, 167, 168, 256, 257, 480, 498
cot ( Ta .) =
25
7rcosec(7rx)
= +
^coanx
^,?isinwa;
OCT.
v wo v/*-r <*/., 29 Q
See also under Fourier series
Zeta
series, 493-495.
2(
\ na
2_
Zl/(a?-n)', 190,475,476.
in general Index.
_^\
s
2
}
484.
187-190, 231-233.
GEN K UAL INDKX.

(The numbers refer
Lemma,
AW1'>
to pages, not tu
--ries);
Cesan/
on
I.TJ
<:::pari8on
rals).
314
207
11:;.
i);
(<>n
multiplication of series).
Complex MI
(unit*
t,
I
Abel's
continuity of
s tests);
204-206 (Abel's
Dirichlet's tests).
11")
Continuity,
iin.maMe wriM
imi (for double sen
convergence, 47 (of real

81 (of double series): 196
429 (of incomplex series)
isolate
series);
(of
tegrals).
Convergence, absolute,
convergence.
Convergence, circle
of,
real
(for
4i Hi.
i
86!
A<\:nptOtic
met!
Absolute
see
of, 8
196
(for
sequences)
:>76 (proof).
complex sequences
284,
27<'.
of a curve, arithmetic treatment,
HC.
functions);
Convergence, general principle
Al. solute sunnnability,
Area
and
serie.-
(of
summableseri'
AS. 98-
197
'rings-
Killer's
:gence,
of,
uniform,
1'J^.
see
'inform
;:;7(I'oinc;t!
'
theory).
Decimals, imiuitr.endixscin's test for
uniform
Brrnoulli's
>c<U-kiul's
of
(Ictinitinn
\\
nun
187.
numbers, 234
functions,
Petiniti
limit of a
iir
sum,
i<>:
of
:I
t.'i.
s< :.
ami
:nami's
series,
rrm.:-
!4&
stence
theor,
L';uu'l>-
11,-.
17
.*
and
17.
118
C;uid,
[d
(of
.riii.
o's
mean-value.
int.:
ils).
to
usually .|ii't-l in
Index.
tli-
text
uti.i. r
;\n
.uit:
authors
name
in the
GENERAL INDEX.
510
Dirichlet's
Dirichlet's integrals, 444, 445, 471.

test for convergence, 49,
206 (of series); 114, 207 (uniform);
430, 435 (of integrals).
Rummer's
Rummer's
Double integrals, 410 (arithmetic defini-
Le Roy's methods
tion)
73-89 (convergence) ;
series,
503-506
76-79 (of positive terms)
for
theorems).
Limits, maximum and minimum (or
extreme), 12, 375 (of sequences) ; 394
tests of convergence, 37.
Euler's constant, 30, 323, 460.

summation formula, 238, 322.
transformation, 55, 169, 302.
(of infinite sets).
11
(of
Limits, upper and lower,
sequences) ; 394 (of infinite sets).
Logarithmic function, 221 (complex
variable) ; 396-402 (real variable).
Exponential function, 145, 402 (real

variable) 217 (complex variable).
;
series, 143, 217, 406.
Logarithmic scale of
theorem on Fourier
series, 348.
series, 167, 168, 189, 230, 256,
472, 480-482, 498, 503;

non-convergent, 2/5, 276, 284, 287-
257,
infinity, 405.
Maclaurin's theorem* connecting the

convergence of series and integrals,
29 ; for double series, 80.
Mertens' theorem, 85; 284 (for non-
Fourier integrals, 447, 471, 472.
Fourier
of
(power-series).
Fejer's
summation
Exponential
series, 140, 172, 249, 487.
non-convergent series, 299.

9 (rules of comLimits, 2 (definition)
bination) ; 246 (of point-sets) 377 (of
383-389 (miscellaneous
quotients)
Double
Ermakoffs
tests for convergence, 32.
Lagrange's
456 (inversion of repeated).
series, 503.
263,
convergent series).
Mittag Leffler's theorem, 499.
Monotonic sequences, 5 373 (proof of
393 (quasi-monotonic
convergence)
290, 347-350, 354.

Fresnel's integrals, 327.
Frobenius's
theorem,
313
132,
(extension).
sequences).
490
395
Functions,
(continuous) ;
501 (without
(without derivates)
For other special
continuation).
280,
functions,
Gamma,
see
284,
vergent
under Exponential,
Logarithmic, Zeta,
of
Multiplication
etc.
314,
series,
331,
82-86;
(non-con-
66,
341
series).
Non-convergent series, 261-267 (general

remarks) see also under Summable
Series, Asymptotic Series.
Numbers, irrational, 358 (as decimals)
;
Gamma
function, 102 (product)
459
(integral) ; 461 (Stirling's formula)

463 (formulae for logarithm); 461,
474, 478 (miscellaneous properties)
503 (Rummer's series).
366-370 (Dedekiud's definition).
Goursat's
Lemma,
Huygens' zones
Poincare's theory of asymptotic series,

330-337.
394.
Poisson's integral, 210, 212.

Power-series, 128-142 (real)
in Optics, 58.
(complex)
integral, 414 (limit of an
integral) ; 417, 495 (limit of a sum) ;
420, 429 (tests for convergence).
Infinite
Integration, 116, 448, 452 (of series)
286
(of
summable
series)
437
Irrational numbers, 358 (as decimals)
197-
199 (complex).
integrals, inversion of order
of integration, 410, 456.
Repeated
fit'ii
latin's
theorem on derangement
of series, 6s.
Kit matin's theorems
498 (Exs.
on Fourier
series,
(>4-<;(>).
Riemann's Zeta function, 493-495.
grals, 44-7.
'Commonly
of series, 86.
Jucobi's theta-functions, 500.

.Ionian's extension of Dirichlet's inte-
IV.Hluets, infinite, 96-102 (real);
366-370 (Dedekind's definition).
202-216
293-296 (summable).
Pringsheim's tests of convergence, 200.

Pringsheim's theorem on multiplication
(of
integrals).
Inversion of repeated integrals, 410,456.
attiilHitfd to (tauchy
it
occurs in Maclaurin's Fluxions,
17-1-, Art.
X>
RAL
Second theorem of mean va
\\\>\.\.
Tests for convergence,
series,
Stirling's series, I
Stirling's
nee
t.
of pro-
daol
-
!'.
>.
Suinni.-ihlf-
''-gral);
COSH* and sinnff);
(for
linir
-"I
i,'ral)
181
i<*s
met
17:
ul. i.-,
TrigoiK
for
cosec'
cotB,
Syinl.i.ls,
'i;<;
w,
-/
complex)
;,:>;
'_'
liin,
lim,
|cosx|<2,
11,
loots);
43S
(integrals);
\'.\
!!_'.
-'"7
Taylor's theor.-m,
_'!
-JIM;
(senes);
tests):
Emu).
l-JI
113,
(sequence)
114,
127,
(products):
sr- unth-r .\\-l.

1
Brmlixxm, (\iuchy, Dirii-liNKuinmer, Shiclani in, 1'
-trass's
nuikot!'.
hcini, Stirling.
I'nifo:
iss.
11.'{
General
'.
!_'_'.
tests
iiinifi-i-iin
L'U
for
t
convei
(power-series);
nils).
-trass's
theorem on double
series,
167,
of series,
Zeta function, (Riemann's),
49.S
PRRB8 HV ROBERT MACLBBOAE AND CO. LTD.
Br
295
is
John
I'
An
Physical
&
Applied
PLEASE DO NOT REMOVE

SUPS FROM THIS POCKET
UNIVERSITY OF TORONTO
LIBRARY
KHDHD

Infinite Series by Bromwich

Uploaded by

Copyright:

Available Formats

You might also like

Infinite Series by Bromwich

Uploaded by

Document Information

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

Infinite Series by Bromwich

Uploaded by

Copyright:

Available Formats

cr>

AX [NTRODUCTION To TIIK THEORY OF

THE MACMILLAN COMPANY

GLASGOW: PRINTED AT THE UNIVERSITY PR1

preparing the manuvript fot press: in parti

t> include proofs

the hope that

rest on an appeal to the reader's intuition

where Dedekiml's definition ..i irratinnal num

.-taMish the ^-m-ral principle of

mastered the differentiation and integration

and the geometrical meaning of these operations.

simplify the discussion of various theorems (for instance, Arts.

from Art. 11 onwards,

although the limit of (1 + 1/V)" (from which this equation is

The use of this definition of a logarithm goes back to Napier,

In Chapter V. will be found an account of Pringsheim's theory

to emphasise the similarity between the tests for

vergence, and particularly to Tannery's theorems (Art. 49)

spit.importance has u-'ially not been

inv.-Nt i-at SMUS

Instance, Arta L18

intnuliictiun to tip- throry ..!' int-_i:raN: IP-:

the tests of convrr-viic,- ami of unit'Mnn <-Mii\vrirence (Ai

Second Theorem of M.-an Valui-i-plat-i-il

traditional 1'nnn of the

the (ianmia integrals i^

sm-h caaee referencefl are ^i\-en

the results -m-,,

view the practical applicationw..rk

and Pierpont'a Theory

Mr. G. H. Hardy, M.A., Fellow and

much increased by Mr.

that the value of the book has been

their careful work.

The following list comprises those books of

have made most

Chrystal, Algebra, vol.

Tannery, The'orie des Fonctions d une Variable, t.

Maximum and Minim am Limit*

DOUBLE SERIES, Sum of a Double

Tests for Convergence.

Substitution of a Power-Series in another

other Elliptic Function-Series (19-24).

Continuity, Integration, and Differentiation

between two Power-Series.

Multiplication and Division.

Sine and Cosine Series.

arc sin and arc tan

(B. 30, 31)

Differential Equations (B. 34-36).

and Conne ProdacU,

Series Tlu-.i VMI

Eider's use of Asymptotic Series.

Vallee Poussin's Theorems

ARITHMETIC THEORY OF IRRATIONAL NUMBERS AND LIMITS, 357-389

and Fundamental Properties

(Single and Double).

SOMK TllKOUKMS ON INFINITE INTEGRALS AND G. \.MMAFUNCTIONS,