See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/300527019

Potassium-Argon/Argon-Argon Dating Methods

Chapter · January 1997

DOI: 10.1007/978-1-4757-9694-0_4

CITATIONS READS

5 1,110

1 author:

Robert C. Walter
Franklin and Marshall College
84 PUBLICATIONS   4,697 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Eritrea View project

East Africa Human Origins View project

All content following this page was uploaded by Robert C. Walter on 22 April 2017.

The user has requested enhancement of the downloaded file.

NCE

Chronometric Dating
in Archaeology

Edited by
R. E. TAYLOR
University of California, Riverside
Riverside, California

and

MARTIN]. AITKEN
Oxford University
Oxford, England

Published in cooperation with the

Society for Archaeological Sciences

livery

PLENUM PRESS • NEW YORK AND LONDON

 Library of Congress Catalog1ng-in-Publ1cation Data

Chronometric dating in archaeology I edited by R.E. Taylor and Martin

J. Aitken.
p. cm. -- (Advances in archaeological and museum science ; v.
2)
Includes bibliographical references and index.
ISBN 0-306-45715-6
1. Archaeological dating.
I. Taylor, R. E. <Royal Ervinl, 1938-
~ . II. Aitken, M. J. (Martin Jiml III. Series.
I CC78.C46 1997
930.1'028'5--dc21
97-31098
CIP

Arnold
University
Bradfon

Raine·
University of Cal
Los Angeli

jonathoIJ
University of I
Irvine,

Donald
University
Seattle,

Garma1
Brookhaven N
Brookha~
ISBN 0-306-45715-6
© 1997 Plenum Press, New York
A Division of Plenum Publishing Corporation Davi
233 Spring Street, New York, N. Y. 10013 Univers
Londo
http://www. plenum.corn
10 9 8 7 6 5 4 3 2 1
W.Da·
All rights reserved Univers
No part of this book may be reproduced, stored in a retrieval system, or transmitted in any Tues!
form or by any means, electronic, mechanical, photocopying, microfilming, recording, or
otherwise, without written permission from the Publisher
B. F
Printed in the United States of America Wellingto
tE. TAYLOR

:em discussions
and Theory for
dy of the First

Interdisciplinary

1rbon dating by Chapter4

, Archaeological

g the "Piltdown

~s: strati graphic Potassium-Argon/Argon-Argon

e calibration of
e Reviews (Qua- Dating Methods
n Antiquity 43:

:ation on earth. ROBERT C. WALTER

of catalytically
·nstruments and

ccelerator mass
ABSTRACT
1984A 40 keV
ics Research B5: During the latter half of this century anthropological surveys in East Africa
have made significant contributions to understanding how the human species
ric radiocarbon has evolved. In the past two decades, particularly, discoveries of our fossil
ancestors have been made in unprecedented numbers and diversity. Detailed
.ques. In Berger,
:alifornia Press: studies of these fossils provide new insights into human evolution, such as the
origin oflocomotion and cultural activity, and the evolution of the brain, among
many other complex features that have come to define humanity. Even during
the time this manuscript was written, new hominid discoveries in Ethiopia and
Kenya were announced that trace our earliest ancestors further back into the
Pliocene. The ages assigned to these fossils have been obtained through
radiometric dating of volcanic rocks interbedded with the fossiliferous sedi-
ments. Such numerical calibrations are crucial to understanding rates and
timing of eV01Utio-rlary Change·. ~-------- -----------------------~----------------

K-Ar dating has played a key role in unraveling the temporal patterns of
hominid evolution as far back as the first significant discovery of East African
australopithecines at Olduvai Gorge in 1959. It was in large part due to the
desire to understand the age of the Olduvai hominid remains that pioneering
attempts were made to date geologically young materials using the K-Ar

ROBERT C. WALTER • Institute of Human Origins, Berkeley, California 94710, USA.

Chronometric Dating in Archaeology, edited by Taylor and Aitken.

Plenum Press, New York, 1997 ..

97
98 ROBERT C. WALTER POTASSIUM-ARGON/,

method. Yet even this seminal K-Ar dating study was plagued by the seemingly The Radiometric l'I
insurmountable problem of contamination. The principal materials for dating
East Africa hominid sites are volcanic ashes, yet many of these ashes are not In one of its mo
deposited as primary air fall tephras (Greek for ash). Rather, most are reworked lation of a daughter is
by stream action and are redeposited into the sedimentary environment. In the decay of a parent isot
process of reworking, these ashes can pick up pre-existing detrital grains that, examples of these so-
by definition, are older than the juvenile ash. If during K-Ar analyses these distinction to the rad
detrital grains are not recognized and eliminated then they can cause the the activity or quanti1
measured ages to be systematically too old. Accumulation c
Recent advances in K-Ar geochronology, specifically the 40Ar/3 9Ar variant rials, such as the age
of the K-Ar method, have helped to establish a precise and accurate temporal of daughter atoms in
framework for hominid evolution in East Africa. Single-crystal laser-fusion easier the measureme
40
Ar/3 9Ar dating has been a major factor in this success. This grain-discrete an age limit below w
method now permits precise and accurate ages to be measured on single grains small to measure ace
and, thus, contaminating grains can be eliminated. The laser-fusion 40Ar/3 9Ar determining the age c
technique has had a profound impact on geochronology by enabling reliable for the radiocarbon d
ages to be obtained where none were possible before. is to measure, yieldin,
impractical.
The radioactive
INTRODUCTION temperature, pressuH
which, simply put, th
The basis for K-Ar dating can trace its origins to von Weiszacker (1937) t, is proportional to t1
who proposed that 40Ar was generated from 4 °K by radioactive decay and that
old potassium-bearing minerals should, therefore, contain measurable quan-
tities ofradiogenic 40 Ar. Aldrich and Nier (1948) confirmed von Weiszacker's
predictions by observ~ng 40Ar concentrations (actually 40Ar/3 6Ar ratios) in the minus sign is req1
four minerals that were significantly greater than atmospheric values (von The decay constant, A,
Weiszacker 1937; Aldrich and Nier 1948). By the mid 1950s, K-Ar dating specific radioactive rn
experiments were performed on many different materials (Shaeffer and Zarin-
tion 1 yields:
ger, 1966) but the development of ultra-high vacuum mass spectrometers by
Reynolds (1956) paved the way for making precise K-Ar measurements on
geologically young samples. When the K:-~r IIl.e_thod was first applied to
young volcanic rocks it was largely in response to the need for calibrating
hominid evolution (Evernden et al. 1957; Leakey et al. 1961; Evernden and and
Curtis 1965). Since then, K-Ar geochronology has played an increasingly
important role in calibrating geological and biological events throughout
Earth's history, from the Precambrian to the late Quaternary. As a consequence,
this radiometric method is one of the most widely used of all geochronological where c is the const:
techniques. The purpose of this discussion is to address ·the conditions and atoms at t = 0, then c
problems encountered in dating geologically young deposits by the K-Ar and
40
Ar/3 9Ar methods, and to help the users of these data to understand the
geological and experimental factors that can contribute to or diminish ana-
lytical success. Solving for P gives th'
ffC. WALTER POTASSIUM-ARGON/ARGON-ARGON DATING METHODS 99

the seemingly The Radiometric Method

ials for dating
ashes are not In one of its most useful forms, radiometric dating relies on the accumu-
, are reworked lation of a daughter isotope over time in a closed system due to the radioactive
nment. In the decay of a parent isotope. The U-Th-Pb, Rb-Sr, Sm-Nd, and K-Ar methods are
al grains that, examples of these so-called accumulation clocks. These methods are in contra-
malyses these distinction to the radiocarbon (1 4 C) method, which is a decay clock based on
:an cause the the activity or quantity of 14 C remaining after a period of time.
Accumulation clocks can be used to determine the age of ancient mate-
,r/39Ar variant rials, such as the age of the earth, moon, and meteorites, because the number
rate temporal of daughter atoms increases with time. In general, the older the sample, the
1 laser-fusion easier the measurement. The corollary to this is that accumulation clocks have
~rain-discrete an age limit below which the number of radiogenic daughter isotopes is too
Lsingle grains small to measure accurately, and therefore are generally not very useful for
sion 40Ar/3 9Ar determining the age of geologically young deposits. The exact opposite is true
bling reliable for the radiocarbon decay clock, where the older the sample the less 14 C there
is to measure, yielding an older age limit beyond which 14 C measurements are
impractical.
The radioactive decay process is independent of external factors, such as
temperature, pressure, and chemical environment. It is a statistical process in
which, simply put, the number of parent atoms, P, that disintegrate over time,
acker (1937)
t, is proportional to the number of parent atoms times a constant, A:
ecay and that
:urable quan-
-dP/dt =PA (1)
Weiszacker's
Ar ratios) in
values (von the minus sign is required because the number of atoms decreases with time.
K-Ar dating The decay constant, A, is a proportionality constant that is characteristic of each
er and Zarin- specific radioactive nuclide. Establishing the differential and integrating Equa-
tion 1 yields:
trometers by
urements on
;t applied to -fdP/P = Afdt (2)
1r calibrating
vernden and and
increasingly
throughout -lnP =At+ c (3)
'1'
:onsequence,
hronological where c is the constant of integration. Defining P0 as the number of parent
nditions and atoms at t = 0, then c = -lnP0 • Substituting for c yields:
the K-Ar and
:lerstand the -lnP = At - inP 0 (4)
iminish ana-
Solving for P gives the fundamental relationship for radioactive decay:
 ROBERT C. WALTER POTASSIUM-ARGONIi
100

p = p e-At (5) THE K-Ar METHO

t 0

which states that the number of parent atoms, P, remaining after a certain Principles
period of time, t, and with a specific decay con~tant, 'A, is an exponential
Potassium (K) i
function of the number of parent atoms, P0 , present at t=O. This equation was
first deduced empirically by Rutherford and Soddy in 1902 from their experi- configuration of the n
ments on the decay of 220 Rn (Dalrymple 1991). has three natural isotc
In order for Equation 5 to be useful for radiometric dating there must be rymple and Lanphere
a way to determine the original amount of the parent, P0 • This is done. by isotopes as well; 36Ar,
summing the number of parent atoms remaining, P,, and the number of nor produced by rad
daughter atoms accumulated, D,, over time, which from mass balance consid- decay of 40 K. Actually,
erations must be equal to the number of original parent atoms, P0 , provided the common isotope c
that closed system conditions were maintained: converts to 40Ar by ca1
neutron. The atomic·
(6) remains the same (Z =
is reduced by one (N
Substituting for P0 in Equation 5 yields: abundance of 40Ar in·
von Weiszacker (193i
P, = (P 1+D 1)e-A1 (7) on the fact that the al
expected. The presen
Solving for D, expresses how the number of daughter atoms grows through which 99.60% is 40Ar;
tively. The 40Ar!3 6Ar n
time:
The K-Ar methi
D = P (eA'-1) (8) lation clocks because
t 't
to the age of the univi
little or no concern fc
Solving Equation 8 for t yields the fundamental age equation for isotope geochro-
sample. This is becau:
nology:
heated. Therefore, th~
metamorphic rocks b
t = l/'A ln(D/P, +l) (9)
heat (metamorphose
Dalrymple and Lanpl
For simple accumulation clocks the age of a rock or mineral can be
and minerals used in
determined if the amount of the parent and daughter isotopes can be accurately
Volcanic rocks
determined:--1-f-the sample contained none of-the-daughter isotope at the time
Imagine a magma c:
of formation, then the age of the sample can expressed by Equation 9. In most
systems generally ha'
systems, however, an unknown amount of initial daughter can be present in the
sition, depth and vol:
sample, which precludes the use of the simple accumulation clock presented
tens to hundreds of I
in Equation 9. If the sample retained some amount of the daughter isotope
van den Bogaard 199
when it formed, D0 , then this quantity must be subtracted from the total
rocks, comprising up
measured amount for the equation to produce the correct age. Though seem-
Although 4°K contir
ingly an insurmountable problem, correcting for initial daughter isotope con-
temperature is so hig
tents can be resolved empirically through the application of isochron and other
erupting, the magma
methods (Dalrymple 1991). In general, the initial daughter problem does not
freezing the molten li
apply to the K-Ar method, as explained below.
:::. WALTER POTASSIUM-ARGON/ARGON-ARGON DATING METHODS 101
'I I

(5) THE K-Ar METHOD

1,'

r a certain Principles
x:ponential
11ation was Potassium (K) is a common rock-forming element. It has the electronic
.eir experi.- configuration of the noble gas argon (Ar), plus a single valence electron, and it
has three natural isotopes; 39 K (93.258%), 4°K (0.012%), and 41 K (6. 730%)(Dal-
re must be rymple and Lanphere, 1969; Steiger and Jager 1977). Argon has three natural
1s done. by isotopes as well; 36 Ar, 38Ar and 40Ar. The first two are stable (neither radioactive
mmber of nor produced by radioactivity), whereas 40Ar is produced by the radioactive
tce consid- decay of 40K. Actually, only about 10.5% of 4°K decays to 40Ar; the rest decays to
,, provided the common isotope of calcium, 4°Ca. In the K-Ar decay scheme the 4°K nucleus
converts to 40Ar by capturing an orbital electron and transforming a proton to a
neutron. The atomic mass (Z = number of protons + neutrons) of the nucleus
(6) remains the same (Z = 40) but the number of protons, the atomic number (N),
is reduced by one (N := 18). One obvious outcome of this process is that the
abundance of 40Ar in the earth and atmosphere increases through time. In fact,
von Weiszacker (1937) predicted that K underwent such a decay scheme partly
(7) on the fact that the abundance of argon in the atmosphere was far greater than
expected. The present-day atmosphere, for example, contains nearly 1% Ar, of
rs through which 99.60% is 40Ar; 38Ar and 36Ar comprise only 0.063% and 0.337%, respec-
tively. The 40Ar/3 6Ar ratio in the modern atmosphere is 295.5 (Nier 1950).
The K-Ar method of dating is one of the most commonly used accumu-
(8) lation clocks because it has an effective age range from less than 100,000 years
to the age of the universe, and it is the only decay scheme that can be used with
re geochro- little or no concern for the presence of the initial daughter isotope, 40 Ar, in the
sample. This is because Ar, an inert gas, readily diffuses out of the system when
heated. Therefore, the principal materials used for K-Ar dating are igneous and
(9) metamorphic rocks because they formed from magmas or were transformed by
heat (metamorphosed) sometime after formation. The reader is referred to
ral can be Dalrymple and Lanphere (1969) for a thorough account of the common rocks
and minerals used in the K-Ar method.
accurately
H the time
Volcanic rocks provide the simplest model to explain K-Ar systematics.
9. In most Imagine a magma chamber beneath a large central volcano. Such magma
:sent in the systems generally have temperatures in excess of 800°C, depending on compo-
presented sition, depth and volatile content, and they can be long-lived, often exceeding
ter isotope tens to hundreds of thousands of years before erupting (Halliday et al. 1989;
i the total
van den Bogaard 1995). The element K is a major component of most volcanic
ugh seem- rocks, comprising up to several weight percent of the bulk magma composition.
otope con- Although 4°K continually decays to 40Ar, even in the magma system, the
1 and other
temperature is so high that the 40Ar gas readily diffuses out of the magma. After
n does not erupting, the magma cools quickly to ambient surface temperatures, essentially
freezing the molten liquid to form either rocks or pyroclastic materials depend-
 102 ROBERT C. WALTER POTASSIVM-ARGOl'

ing on the mode of eruption. At this point there is expected to be no initial by radio-frequency
radiogenic (daughter) 40Ar in the system. Any subsequently produced 40Ar will spectrometry and tl
accumulate as the system cools below the closure temperature for the retention (Dalrymple and Lar
of Ar (generally between 500°C and 150°C). The 40Ar atom is large enough that fused in a single ex
it is physically trapped within chilled, solid crystal lattices and glass networks, equivalent to hund1
where it can remain indefinitely at ambient surface temperatures. In volcanic
systems, time zero (t = 0) is the moment of eruption, which is analogous to
Assumptions
resetting a stopwatch: the subsequent accumulation of 40Ar in the rock records
the time elapsed since eruption. The primary <
Many radioactive nuclei decay by the emission of electrons from the each is representati'
unstable parent nucleus, called ~-decay. The radioactive decay of 4 °K is unusual a rock is highly het1
10
in that it occurs by two schemes, by electron capture to 40Ar (A-,= 0.581 x 10- the two aliquots mi
yr- 1) and by ~-decay to 4°Ca (A,~= 4. 962 x 10-10yr- 1). This dual scheme is referred 4
°K and 40Ar from 01
to as a branching decay. Roughly 90% of the total decay of 4°K is by ~-decay to useless as a measure
4°Ca. The half-life of the total decay (A-,+ A,~) of 4°K is 1.250 x 10 9yr. It follows sample homogeneit
that the fraction of 4°K atoms that decay to 40Ar is (A-/A-) 4°K, where A = A-, + "-w and obsidians (horn
In the K-Ar dating method the growth of 40Ar over time is expressed as follows: AK-Ar age is
crucial assumption
(10) the time of formati
radioactive decay oJ
, wh~re 40Ar* is radiogenic argon. Compare the form of this equation with 40
Ar contamination.
Equation 8. Solving Equation (10) for t yields the fundamental K-Ar age further assumed th<
equation, which the reader is encouraged to compare with Equation (9): There are sev€
the K-Ar method is
t = W. ln [(40 Ar* /4°K) (AJA-,) + 1] (11) a "closed system" 'i
system is one in wh
10 1
where tis the calculated age; 1 is the total decay constant (5.543 x 10- yr- ), size, from a single g
the proportion of 4 °K atoms that decay per unit time; A./A. = 0.1048 is the the system, usually
4
fraction of 4°K decays yielding radiogenic 40Ar, and 40Ar* and °K are the and daughter isoto]
amounts of radiogenic daughter and parent isotopes in the sample at the present leaching or enrichi1
time. The constants are those recommended by the IUGS Subcommission on its formation, are E
Geochronology (Steiger andJager 1977), based on physical measurements of change the sample ·
4°K and its decay. It is evident from Equation 11 that to determine a K-Ar age tion should be sm
the amounts of 40Ar* and 4°K must be measured. Several excellent books cover modern atmospher·
. .
e1im1nate th e atmosI
the background of the K-Ar method in more detail (Dalrymple and Lanphere
1969; York and Farquhar 1972; Faure 1986). composition of K i:
except for that char
and (4) It is assume1
Procedures
In conventional K-Ar dating, the sample is split into two aliquots; one split
Limitations
is used to measure potassium and the other to determine argon. Potassium is
usually measured by flame photometry or atomic absorption. Argon gas, which In practice, tr
is liberated by melting the sample in a high vacuum extraction system, usually the K-Ar method ar
 ROBERT C. WALTER
xpected to be no initial
ently produced 40 Ar will
1erature for the retention
.tom is large enough that
:ices and glass networks,
:mperatures. In volcanic
L, which is analogous to
f 40Ar in the rock records
in of electrons from the
ve decay of 4 °K is unusual
:o 40 Aro.. = 0.581 x 10-10
.is dual scheme is referred
ay of 4°K is by ~-decay to
. 1.250 x 10 9yr. It follows
/A,) 4°K, where A= A,+ A~.
ie is expressed as follows:
(10)
m of this equation with
le fundamental K-Ar age
e with Equation (9):
l] (ll)
nstant (5.543 x l0- 10yr- 1),
ime; A./A = 0.1048 is the
nd 40 Ar* and 4°K are the
ln the sample at the present
: IUGS Subcommission on
physical measurements of
at to determine a K-Ar age
reral excellent books cover
(Dalrymple and Lanphere
: into two aliquots; one split
ermine argon. Potassium is
Jsorption. Argon gas, which
i extraction system, usually
POTASSIUM-ARGON/ARGON-ARGON DATING METHODS IOJ
by radio-frequency induction or resistance heating, is measured by mass
spectrometry and the Ar isotopic abundances are obtained by isotope dilution
(Dalrymple and Lanphere 1969). Generally milligrams to grams of material are
fused in a single experiment, depending on age and K content, which can be
equivalent to hundreds to thousands of mineral grains.
Assumptions
The primary assumption is that the two aliquots are homogeneous, and
each is representative of the Kand Ar compositiq.n of the bulk rock. If, however,
a rock is highly heterogeneous, as are some porphyritic lavas and ignimbrites,
the two aliquots might not be chemically equivalent, causing a decoupling of
4°K and 40Ar from one split to the other. In this case the K-Ar result is generally
useless as a measure of geological "age" since formation. Therefore, to facilitate
sample homogeneity, the best materials for K-Ar dating are fine-grained lavas
and obsidians (homogenous glasses), or pure mineral separates.
AK-Ar age is a function of the accumulation of 40Ar over time. Another
crucial assumption is that there is no initial radiogenic 40Ar in the sample at
the time of formation, that all 40Ar in the sample is attributed to the in situ
radioactive decay of 4 °K (except for that which can be ascribed to atmospheric
40
Ar contamination, see below), a condition met by most volcanic rocks. It is
further assumed that 40Ar and 4 °K can be accurately measured.
There are several other equally significant conditions that must be met if
the K-Ar method is to yield valid dates: (1) The rock or mineral must remain
a "closed system" with respect to K and Ar throughout its history. A closed
system is one in which matter neither enters or leaves. A system may be of any
size, from a single grain to the size of the universe, but for radiometric dating,
the system, usually a rock or mineral grain, need only be closed to the parent
and daughter isotopes. Weathering, alteration, and thermal overprinting, by
leaching or enriching the rock in K, or by diffusion of Ar out of the rock after
its formation, are examples of open system behavior that can dramatically
change the sample from its true K-Ar age; (2) Atmospheric argon contamina-
tion should be small -and-appropriate- corrections can--be--made. Since the
modern atmosphere is 1% argon, of which 99 .6% is 40Ar, it is important to
eliminate the atmospheric 40Ar contaminant from the sample; (3) The isotopic
composition of K is assumed to be the same for all samples and unchanged
except for that change which can be accounted for by radioactive decay of 4 °K;
and (4) It is assumed that the decay constants (A, and A~) are accurately known.
Limitations
In practice, the limiting factors in dating geologically young samples by
the K-Ar method are twofold. First, severe problems can arise in attempting to
104 ROBERT C. WALTER

Vent-Wall Con
measure small volumes of 40Ar* produced by young and/or K-poor samples.
The reason is simple; the older the sample, the more 40Ar* there is to measure,
and the easier the measurement becomes. The corollary of this is that the
1
~/
younger the sample, the less 40Ar* there is to measure, and the more difficult
the measurement becomes. / /' / /
Second, errors in detecting small amounts of gas are magnified if the ! / / .
sample contains significant amounts of atmospheric argon contamination. TEPHRA FALLC
i ! ! !

Atmospheric argon is present to some degree in all volcanic rocks primarily

\ '\
due to absorption of atmospheric gases into the magma before and during \
'
""'. \" '
\
eruption, and secondarily from adsorption onto rock surfaces during even \ ,,
incipient hydration and alteration. This atmospheric gas is enriched in 40 Ar,
which is analytically indistinguishable from radiogenic ~0Ar*. However, since
the present-day atmospheric abundance ratio of 40Ar/36Ar is well known
40 6
( Ar/3 Aratmos = 295.5), then the desired radiogenic component is merely the
difference between the measured 40Ar/36Ar value and 295.5:

40 40
Ar* = Ar measured - 295.5 36Ar (12)
Detrital Conran
An additional limitation is imposed on rocks that have been physically
contaminated. This is particularly true for pyroclastic rocks that can pick up
vent-wall fragments during eruption, or detrital minerals during secondary
redeposition (Fig. 4.1). Such contamination may not be ~olumetrically signifi-
cant in older rocks, but will cause significant errors in Quaternary samples
(Curtis, 1966).

THE 40
Ar/3 9Ar METHOD

Principles, Assumptions, and Procedures

The 40Ar/3 9Ar method is based on the same fundamental principles and
assumptions as the conventional K-Ar method. The main difference is the way
in which K is measured. In this method K and Ar are measured virtually
simultaneously on the same sample split. Prior to analysis the sample is irradi-
ated with high-energy neutrons at a nuclear reactor. This process converts 39 K (a
stable isotope ofK) in the sample to 39Ar, an isotope that does not occur in nature Figure 4.1. Illustration o
and is only produced in nuclear reactors according to the following relationship: deposit. The upper figure
Here, the force of the eru]
39
rock) from the vent, whicl
Ar = 39 K 11T fcp(e) cr(e) de (13) the eruption, and-deper
ages to be systematically 1
where 39K is the number of atoms of the target isotope, 11T is the duration of can be reworked into ti
the irradiation, and cp(i::) is the neutron fluence density at energy e and cr(e) is reworking processes, the
surface. Detrital mixing i
the capture cross section for 39 K at energy e, and the integration is carried out tephras used to calibrate l
 '
! I

ROBERT C. WALTER

Vent-Wall Contamination
l/or K-poor samples.
* there is to measure,
y of this is that the
nd the more difficult

are magnified if the

rgon contamination. \
:anic rocks primarily
.a before and during \\.. \ \\,
'
.
""'
;urfaces during even "'-.. \
40
5 is enriched in Ar,
~Ar*. However, since
.
136
Ar is well known
ponent is merely the
.5:

(12)

have been physically

cks that can pick up
tls during secondary
olumetrically signifi-
Quaternary samples

1ental principles and

difference is the way
e measured virtually
; the sample is irradi-
rocess converts 39 K (a
es not occur in nature Figure 4.1. Illustration of the processes by which older grains can become incorporated in a tephra
>llowing relationship: deposit. The upper figure shows a cross section of a composite volcano that is erupting explosively.
Here, the force of the eruption is shown to rip clasts of country rock (xenoliths=strange or foreign
(13) rock) from the vent, which become part of the primary tephra fallout. Xenoliths obviously predated
the eruption, and-depending on the their age and abundance-can cause measured radiometric
ages to be systematically too old. The lower figure illustrates that an unconsolidated volcanic ash
:).T is the duration of can be reworked into the sedimentary environment by storm and river action. During the
: energy s and a(s) is reworking processes, the primary tephra can pick up older detrital grains from the preexisting
gration is carried out surface. Detrital mixing is the most common source of contamination and systematic error in
tephras used to calibrate hominid evolution in East Africa.
 106 ROBERT C. WALTER POTASSIVM-ARGOl

over the entire energy spectrum of the neutrons (Mitchell 1968). The basis for bombardment in thE
40 9
the Ar/3 Ar dating method was laid down by Merrihue (1965) and Merrihue argon correction, t
and Turner (1966).
atmospheric correc1
The irradiated sample is heated in a high vacuum extraction system, Ca-rich samples, su1
usually by radio-frequency induction or resistance heating, liberating the three ous ages if the apprc
riatural isotopes of Ar plus the new artificially produced 39Ar isotope. The 39Ar made by irradiating·
concentration becomes a proxy measurement for the concentration of K. All produced, and appl)
argon isotopes are measured by mass spectrometry, and the 40Ar* /3 9Ar ratio now
40
replaces the Ar*/4°K ratio in Equation 11. The 40Ar/3 9Ar age is calculated by
comparing the measured 40Ar*/3 9Ar ratio of the sample to the 40Ar*/3 9Ar ratio
of a standard of known age that is irradiated along with the unknown as a where:
neutron dosimeter (Merrihue 1965; Merrihue and Turner 1966; Dalrymple and
Lanphere 1971). Combining Equations 10 and 13 yield the following relation-
ships:

(14a)

and The subscript m si~

calcium and potasst

(14b)
Advantages
39
where ArK signifies the production of this isotope from 39K, and J, which The 40Ar/3 9Ar r
represents the constants ('A/A and 40K/39K) and irradiation parameters (.1 T, cp, technique, includin
and cr), is empirically determined by irradiating a rock or mineral standard of isotopic ratios in a s
known age (t,) together with the sample in question. The neutron fluence (cp) of K content; (2) sm
and the capture cross section (cr) are difficult parameters to determine uniquely
improved because P'
because the energy spectrum of the incident neutrons and the capture cross and because ratios,
39
sections of K for neutrons of varying energies are not well known. However, heated to progressiv~
by measuring the 40Ar and 39Ar isotopes of the irradiated standard, cp and cr
can be collected, me
become implicit to the standard, and] becomes:
a closed system with
should be equivalen
(15) the case, then one OJ
Such incremental h
where the subscripts refers to the standard. The 40Ar/3 9Ar age equation, then, provide valid ages fj
reduces to: and Lanphere 1971)

(16a) The Laser Revolu

The subscript u is for the unknown sample. A laser was fir~
Unfortunately, there are several unwanted reactions that produce interfer- the short burst of E
ing isotopes during irradiation (Mitchell 1968; Berger and York 1970; Brereton milligram in size) q
39
1970). Specifically, Ar and 36Ar are produced from fast neutron reactions with 40
Ar/3 9Ar ages withi1
40
Ca, and Ar is produced from slow neutron reactions with 40 K during neutron continuous laser hel1
 ROBERT C. WALTER POTASSIUM-ARGON/ARGON-ARGON DATING METHODS 107

ell 1968). The basis for
e (1965) and Merrihue
um extraction system,
mg, liberating the three
i 39 Ar isotope. The 39 Ar
:oncentration of K. All
he 40Ar* / 39Ar ratio now
Ar age is calculated by
to the 40Ar* /3 9Ar ratio
·ith the unknown as a
:r 1966; Dalrymple and
the following relation-
bombardment in the reactor. Since 36Ar is crucial in determining the atmospheric
argon correction, the extent of this interference must be known if reliable
atmospheric corrections and 40Ar/39Ar ages are to be achieved. Clearly, young,
Ca-rich samples, such as basalt, plagioclase and hornblende, could yield errone-
ous ages if the appropriate Ca corrections are not applied. These corrections are
made by irradiating pure calcium and potassium salts, measuring the Ar isotopes
produced, and applying the corrections to the unknown as follows:
where:
(16c)

(16d)
>(s)cr(s)ds (14a)
The subscript m signifies the measured quantity, and Ca and K refer to the
calcium and potassium-derived isotopes, respectively (Hall and York 1984).

(14b)
Advantages
om 39K, and ] , which The 40Ar/3 9 Ar method has several advantages over the conventional K-Ar
on parameters (~ T, cp, technique, including: (1) ages are based entirely on measurements of Ar
e>r mineral standard of isotopic ratios in a single sample, and do not require a separate measurement
le neutron fluence (cp) of K content; (2) smaller samples can be dated; (3) precision and accuracy are
to determine uniquely improved because points 1 and 2 eliminate problems of sample heterogeneity
and the capture cross and because ratios, not amounts, are measured; and (4) the sample can be
;vell known. However, heated to progressively higher temperatures and the Ar released from each step
:ed standard, cp and cr can be collected, measured, and an age calculated. If the sample has remained
a closed system with respect to Kand Ar, then the age of each temperature step
should be equivalent to each other within analytical uncertainty. If this is not
(15) the case, then one or more of the fundamental assumptions has not been met.
Such incremental heating data provide a self-checking mechanism and can
\r age equation, then, provide valid ages for samples that have not been closed systems (Dalrymple
and Lanphere 1971).

(16a) The Laser Revolution

A laser was first applied to 40Ar/39Ar dating by Megrue ( 1973), who used
that produce interfer- the short burst of energy from a pulsed laser to fuse tiny fragments (sub
d York 1970; Brereton milligram in size) of rock in order to determine the spatial distribution of
40
teutron reactions with Ar!39Ar ages within lunar breccias. York et al. (1981) demonstrated that a
th 4°K during neutron continuous laser held several distinct advantages over the pulsed laser system,
108 ROBERT C. WALTER POTASSIUM-ARGON/Al

such as permitting step heating for the acquisition of age spectra, and providing tions, such as the rare 3'
rapid and complete fusion of a variety of mineral grains. Following irradiation, are very difficult to me<
the samples are loaded into an ultra-high vacuum extraction line and fused by in K-Ar and 40Ar/3 9Ar d
the laser beam through a vacuum window (Megrue 1973; York et al. 1981). The For example, for
argon gas that is released during fusion is purified and analyzed by mass the sample and the an
spectrometry, and an age is calculated after the appropriate corrections are small. Since 36Ar conci
applied. A useful by-product of the laser-fusion method is a cleaner system. measurement is the de
The Ar background of the system is dramatically reduced because such small sample with 95% atm1
amounts of material are fused and only the material itself is heated. basalts, plagioclase, an
Early experiments with the laser were done on lunar and Precambrian 0.5% will have an appr
terrestrial materials, because the great age of these rocks made the 40Ar/39Ar age Methods of calcu
determination relatively straightforward despite small sample size. The limita- Dalrymple and Lanphe
tion of the laser-fusion method, however, is the detection limit of small amounts (1970), and Dalrymple
of 40Ar and 36Ar in very young and, now, very small samples. Primarily an ple and Duffield (1988:
instrumentation problem, this limitation was overcome in the past decade by technique is that becau
the development of sensitive, low-background mass spectrometers. For exam- simply by increasing t
ple, the pioneering work at the University of Toronto was accomplished on a study of sanidine from 1

25 year old AEI MS-10 mass spectrometer, showing that step-heating analyses of 6 grains per sampl
could be made on single grains of Precambrian biotite by sequentially increas- increasing the numbe1
ing the wattage of the continuous laser (York et al. 1981). After the acquisition precision to 0.2% or le:
of a new, sensitive mass spectrometer, the Toronto group quickly advanced the Accuracy is cont1
applicability of the laser method to much younger ages. By 1986, single grain to deviate in a consiste
ages of 35.4±0.6 Ma were achieved on microtektites from the marine record, the generic K-Ar syste1
which helped to calibrate the Eocene-Oligocene boundary (Glass et al. 1986). of systematic errors fro
By 1987 the Toronto group was routinely dating single grains of 500 ka old and erroneous constan
sanidines from volcanic ashes in France and Germany (LoBello et al. 1987; van laboratory errors can
den Bogaard et al. 1987). international standards
Perhaps even more important than demonstrating that the single-crystal are accurately known.
laser-fusion ages could be measured on Quaternary minerals, LoBello et al. by application of the g
(1987) and Bogaard et al. (1987) showed that the method provides meaningful whereas step-heating, l
eruption ages even for volcanic rocks that have been severely contaminated by be used to overcome s
older grains, because ages were obtained on individual mineral grains. The (McDougall and Harri:
grain-discrete laser-fusion method is essential for dating samples that consist
of mixed ages, and/or when large amounts of samples are unavailable. Not only
can ages be measured from single mineral grains, but Ca/K ratios can be APPLICATIONS TC
discerned via measured 37Ar/3 9Ar ratios. Such geochemical information is useful I
in helping to identify multiple age populations (Walter et al. 1991). Several exampl<
40
Ar/39 Ar dating have
comprehensive survey
PRECISION AND ACCURACY particular emphasis 011
involve human evolu
Precision is the measure of analytical uncertainty, and it is caused by pure interesting insights th<
random fluctuations in data acquisition (Bevington 1969). Very low concentra- ological studies.
ERT C. WALTER POTASSIUM-ARGON/ARGON-ARGON DATING METHODS 109

a, and providing tions, such as the rare 36Ar isotope, and 40Ar* of very young and K-poor samples,
ving irradiation, are very difficult to measure, and contribute the largest sources of random error
:ne and fused by in K-Ar and 40Ar/39Ar dating of young samples.
et al. 1981). The For example, for Quaternary samples the difference between 40Ar/36Ar in
.alyzed by mass the sample and the atmospheric 40Ar/36Ar value of 295.5 can be vanishingly
corrections are small. Since 36Ar concentrations are also typically very small, the error in its
cleaner system. measurement is the dominant source of error in the calculated age. Thus, a
ause such small sample with 95% atmospheric argon contamination (typical for Quaternary
eated. basalts, plagioclase, and hornblende) with 36Ar measured with a precision of
nd Precambrian 0.5% will have an approximate age error of 10% (Hall and York 1984).
the 40Ar/3 9Ar age Methods of calculating analytical error have been described in detail by
size. The limita- Dalrymple and Lanphere (1969) for the K-Ar method and by Berger and York
>f small amounts (1970), and Dalrymple and Lanphere (1971) for the 40Ar/39Ar method. Dalrym-
es. Primarily an ple and Duffield (1988) illustrate that yet another advantage of the laser-fusion
~past decade by technique is that because of its speed and ease of analysis precision is improved
eters. For exam- simply by increasing the number of replicate analyses. In their laser-fusion
:omplished on a study of sanidine from Oligocene rhyolite flows, the standard error of the mean
heating analyses of 6 grains per sample yielded uncertainties of about 0.5% (lcr) and that
lentially increas- increasing the number of analyzed grains by a factor of four improved the
r the acquisition precision to 0.2% or less (Dalrymple and Duffield 1988).
dy advanced the Accuracy is controlled by systematic errors that cause the measurement
~86, single grain to deviate in a consistent manner from the true response (Bevington 1969). In
~ marine record, the generic K-Ar system, alteration and contamination are the largest sources
lass et al. 1986). of systematic errors from natural causes, whereas poor calibration procedures
1s of 500 ka old and erroneous constants can lead to systematic errors in the laboratory. Most
:i et al. 1987; van laboratory errors can be assessed and minimized by analyzing well-known
international standards, with the assumption that the standard ages themselves
he single-crystal are accurately known. Systematic errors from contamination can be overcome
ls, LoBello et al. by application of the grain-discrete laser-fusion method (LoBello et al. 1987),
rides meaningful whereas step-heating, by the laser or by more traditional heating methods, can
contaminated by be used to overcome some problems caused by alteration or metamorphism
Leral grains. The (McDougall and Harrison 1988).
.ples that consist
·ailable. Not only
'K ratios can be APPLICATIONS TO HOMINID EVOLUTION
rmation is useful
1991). Several examples have been chosen that illustrate how K-Ar and
40
Ar/39Ar dating have helped to calibrate hominid evolution. This is not a
comprehensive survey, but one that illustrates a variety of applications with
particular emphasis on tl:ie East African hominid record. Not all cases directly
involve human evolution, but they are mentioned because. they provide
is caused by pure interesting insights that are potentially useful for anthropological and archae-
y low concentra- ological studies.
 110 ROBERT C. WALTER pOTASSIUM-ARG01

Olduvai Gorge: Early K-Ar Dating & Recent Laser 40 Ar/3 9Ar Revisions was shown to be a
grains. The criteria
K-Ar Dating and the Pleistocene signs of being rewo1
if different concent
A pioneering application of K-Ar dating was brought about by the discov-
markedly different
ery of hominid fossils and artifacts at Olduvai Gorge, Tanzania. In the late 1950s,
influenced the eva
jack Evernden, a geophysicist at the University of California at Berkeley; visited
obtained on tephrn
Olduvai Gorge and collected samples, principally tephras, for K-Ar dating from
Curtis 1965). Even
the lowest artifact-bearing deposits in Bed I. The discovery of the hominid fossil,
difficult to interpret
"Zinjanthropus" (Australopithecus boisei), from lower Bed I in 1959 made this
it was shown that 01
seminal K-Ar study even more timely. Seven tephra deposits from the hominid-
considered reliable
bearing portion ofBed I were initially dated, with ages ranging from 1.89 to 1.57
expressed by R.L. H
Ma, demonstrating for the first time the great antiquity of the hominid remains
depends to a consid
from Olduvai (Evernden et al. 1957; Evernden and Curtis 1965).
dates from a much l
The discovery of Zinjanthropus coincided with the advent of ultra-high
vacuum mass spectrometers for K-Ar dating, such as those developed by john
Reynolds, a physicist also at Berkeley (Reynolds 1956). Using this new tech- Laser-Fusion Met
nology, Evernden analyzed the volcanic rocks from Olduvai, in part, to test the Recent laser-J
. K-Ar method on Pleistocene materials (Evernden et al. 1957). It should be marked a great imp:
borne in mind that even by the early 1960s the age of the Pleistocene was still et al. 1991; Hay 19'
in question, and that the old K-Ar ages from Olduvai were met with skepticism. the correct stratign
Such skepticism was expressed by von Koenigswald et al. (1961), who stated two or more ages w
that "As the lower part of Bed I is said to have an age of 1. 75 Myr (in reference accounting for son
to Leakey et al. 1961), the whole Pleistocene Period would have lasted well over dates. For example
2.00 Myr. This is more than double of what we, cautiously and with good above the Naabi I/
reasons, have estimated". Straus and Hunt (1962) echoed von Koenigswald in feldspar grains an:
their general conclusion that "Having considered the many unanswered sedi- yielded ages rangi
mentological and paleontological questions which make Olduvai a difficult test outcrops of the Pai
case, we prefer ... to be cautious until more dates for the lower Pleistocene are grains yielded con
available". Ultimately, with more dates, these concerns were alleviated, provid- ment with its strati
ing anthropologists and geologists alike with the first hard evidence for an "old" al. 1991; Walter et
Pleistocene, more than doubling its time span, and for the age of associated feldspars would yi1
hominid and archaeological remains (Leakey et al. 1961; von Koenigswald et One of them
al. 1961; Curtis and Evernden 1962; Leakey et al. 1962; Straus and Hunt 1962). Gorge was the disc
In addition, this groundbreaking K-Ar study provided calibration points for its, such as the Na~
defining the Olduvai Geomagnetic Event (Gromme and Hay 1963; Gromme simply identifying
and Hay 1967; Brock and Hay 1976), the first magnetostratigraphic study of a bulk fusion K-A:
fossiliferous terrestrial sediments, paving the way for numerical and paleomag- single-grain ages tl
netic calibration of faunal sequences worldwide.
Despite these dramatic successes, the vast majority of the K-Ar dates from
Berekhat Ram a
Olduvai Gorge were considered unreliable (Evernden and Curtis 1965). The
criteria used for accepting an age at Olduvai were: (1) if several samples from the A useful illw
same tephra at different localities produced the same age; and (2) if the sample the work of Fera1
 ROBERT C. WALTER POTASSIUM-ARGON/ARGON-ARGON DATING METHODS 111

40
Ar/39Ar Revisions was shown to be a primary volcanic deposit with no contaminating mineral
grains. The criteria for rejecting an age were: (1) if the tephra showed any visible
signs of being reworked and thus subject to admixture of detrital grains; and (2)
.t about by the discov- if different concentrations of mineral separates from the same tephra yielded
mia. In the late 1950s, markedly different ages. The consistency of K-Ar ages with stratigraphy also
1ia at Berkeley; visited influenced the evaluation of each analysis. Using these criteria, of 41 ages
, for K-Ar dating from obtained on tephras from Bed I, only 15 were considered valid (Evernden and
·of the hominid fossil, Curtis 1965). Even the 15 "valid" dates showed considerable scatter and were
difficult to interpret in terms of stratigraphy (Hay 1992), and upon further study
:l I in 1959 made this
it was shown that only a fraction of the 15 dates accepted as valid were ultimately
its from the hominid-
considered reliable. The difficulty in selecting a valid K-Ar age was aptly
5ing from 1.89 to 1.57
expressed by R.L. Hay, who noted, "The implication is that accurate K-Ar dating
· the hominid remains
depends to a considerable extent of the ability to select a small number of reliable
1965).
dates from a much larger number of unreliable ones" (Hay 1992).
: advent of ultra-high
se developed by John
Using this new tech- Laser-Fusion Method Resolves Anomalies
rai, in part, to test the
Recent laser-fusion 40Ar/39Ar studies based on analyses of single crystals
1957). It should be
marked a great improvement over the conventional K-Ar dating of Bed I (Walter
: Pleistocene was still
et al. 1991; Hay 1992; Walter et al. 1992b). Ages on all the marker tuffs fall in
met with skepticism.
the correct stratigraphic order, spanning from 2.02 to 1. 75 Ma. Tephra yielding
l. (1961), who stated
two or more ages were identified by the single-crystal laser-fusion method, thus
l.75 Myr (in reference
accounting for some and perhaps most of the variability in the earlier K-Ar
have lasted well over
dates. For example, Tuff lA is a reworked tephra exposed west of the Fifth Fault
)usly and with good
above the Naabi Ignimbrite at Olduvai Gorge (Hay 1976). Seven of the ten
l von Koenigswald in
feldspar grains analyzed by the single-crystal laser-fusion 40Ar/39Ar method
.ny unanswered sedi-
yielded ages ranging from 340 to 740 Ma, no doubt derived from nearby
)lduvai a difficult test
outcrops of the Pan-African crystalline basement. Two of the remaining three
lower Pleistocene are
grains yielded concordant ages of 1.98±0.02 Ma, which is in excellent agree-
:re alleviated, provid-
ment with its stratigraphic age above and below well-dated tephras (Walter et
evidence for an "old"
al. 1991; Walter et al. 1992b). A bulk fusion K-Ar or 40Ar!39Ar analyses of these
the age of associated
feldspars would yield an erroneous age in excess of 300 Ma.
; von Koenigswald et One of the most illuminating aspects of the laser-fusion study at Olduvai
raus and Hunt 1962). Gorge was the discovery of contaminating grains in primary pyroclastic depos-
:alibration points for its, such as the Naabi Ignimbrite and pumice clasts from Tuff IE, indicating that
Hay 1963; Gromme simply identifying a unit as a primary volcanic rock is not enough to validate
tratigraphic study of a bulk fusion K-Ar or 40Ar/3 9Ar age. It is only through a population study of
1erical and paleomag- single-grain ages that primary eruptive ages can be determined with certainty.

,fthe K-Ar dates from 40

Ld Curtis 1965). The Berekhat Ram and Disturbed Ar/3 9Ar Age Spectra
eral samples from the A useful illustration of the application of 40Ar/39Ar age spectra is given by
and (2) if the sample the work of Feraud et al. (1983) at Berekhat Ram, a small crater lake in
 112
ROBERT C. WALTER
POTASSIUM-ARG01
Pleistocene basalts of the Golan Heights in Israel. In the northwest quadrant
of the crater two basalt flows are interbedded with 2 meters of reddish-brown integrated 40Ar/3 9Ar
paleosols that contain Acheulian handaxes. The upper lava yields concordant tional K-Ar age of 5
40 39
K-Ar and Ar/ Ar ages of about 230 ka, providing a minimum age for the probably formed soi
artifacts (Feraud et al. 1983). spectrum exhibits t'
The lower flow, however, yields discordant K-Ar and 40Ar/39Ar ages, as ka, and it is not pos
well as a disturbed release spectrum, with ages increasing from 205 to 780 ka likely explanation f,
(Fig. 4.2), suggesting that the lower lava was not a "closed" K-Ar system. The to alteration or the:
flow. There is no in<
important point to ·
Berekhat Ram Without contrastini
reason to question ti
800

Young Lavas witl

700
During the Pl
Q00
BRG/Bl/26 glaciers covered an~
The Nome river gla1

-
L..
=n
500
on conventional K-.
Minnie Creek, in the

-
-~
Cl)
c:n
<
400
Based on these resl
Pleistocene (Kauffrr
fusion 40Ar/39Ar me;
basalt, indicated the
300 K-Ar ages on larger
1991). The younger
200 K-Ar age of 564±64
predates the Minnie
BRG/Bl/23 Laser-fusion 4 <
100 Minnie Creek basal!
mean age of 600±341
plot (Fig. 4.3) revea
0 10 20 30 40 50 million years. This l
60 70 80 90- 100
analyses were compe
Cumulative 39Ar (~)
40 9
Figure 4.2. Ari3 Ar age spectra from Berekhat Ram upper (BRG/81/23) and lower (BRG/81/26)
~;-~~?=~~;:;~:;, l
39 An isochron pl
basalt flows. Length of bars indicate percent Ar released. Width of bars represent ±1 sigma errors.
40 39
Each bar represents an Ar/ Ar analysis at a specific temperature increment, from lowest at the
the notion that the l
left to highest (fusion) at the right. If the sample is undisturbed, that is, if it behaved as a closed from 40 Ar/39Ar data p
39
system with respect to Kand Ar, then the age measured at each temperature increment will be the Ar/4°Ar: such plots
same, within analytical uncertainty, as it is with BRG/81/23. Such ideal behavior implies that the represent portions o
measured age is the true formation age of the sample. If the sample has been disturbed, however, initial 40Ar/3 6Ar rati
then the release spectra will be discordant, as it is for BRG/81/26. In this case, no meaningful
formation age can be determined. Data and figure from Feraud et al. (1983).
conform to ideal be
presence of excess
 ROBERT C. WALTER
.e northwest quadrant
:ters of reddish-brown
ava yields concordant
minimum age for the
and 40Ar/39Ar ages, as
1g from 205 to 780 ka
ed" K-Ar system. The
) 80 90- 100
(~)
:) and lower (BRG/81/26)
represent ±1 sigma errors.
~ment, from lowest at the
, if it behaved as a closed
ure increment will be the
behavior implies that the
been disturbed, however,
this case, no meaningful
183).
POTASSIUM-ARGON/ARGON-ARGON DATING METHODS 113
integrated 40 Ar/3 9Ar age of 4 70 ± 8 ka is in good agreement with the conven-
tional K-Ar age of 510±32 ka. However, the disturbed pattern indicates the lava
probably formed somewhere between 290 and 800 ka (Feraud et al. 1983). The
spectrum exhibits two poorly defined age plateaus, at roughly 290 ka and 640
ka, and it is not possible to give a single, satisfactory interpretation. The most
likely explanation for this pattern is either argon loss at low temperatures due
to alteration or thermal overprinting, or the presence of excess argon in the
flow. There is no independent support for either hypothesis. Nevertheless, the
important point to be gleaned is that its K-Ar age of 510±32 ka is unreliable.
Without contrasting 40Ar/3 9Ar spectra (Fig. 4.2) there would be no analytical
reason to question the validity of the conventional K-Ar age of the lower basalt.
Young Lavas with Excess 40 Ar and the Initial Daughter Problem
During the Pleistocene Nome River glaciation of northwestern Alaska,
glaciers covered an area ten times more extensive than any later glacial interval.
The Nome river glaciation was initially believed to be about 800 ka old based
on conventional K-Ar ages on basaltic lavas that overlie Nome River drift at
Minnie Creek, in the central Seward Peninsula (Kauffman and Hopkins 1985).
Based on these results the Nome River glaciation was assigned to the early
Pleistocene (Kauffman and Hopkins 1986). Subsequent analysis by the laser-
fusion 40Ar/39Ar method, on milligram-size subsamples of the Minnie Creek
basalt, indicated the presence of extraneous 40Ar that rendered the conventional
K-Ar ages on larger bulk samples of the same flow too old (Kauffman et al.
1991). The younger age for the Nome River glaciation is supported by a new
K-Ar age of 564±64 ka on the Lava Creek basalt flow that stratigraphically
predates the Minnie Creek basalt.
Laser-fusion 40Ar/39Ar analyses on 14 milligram-size subsamples of the
Minnie Creek basalt produced a wide variation in ages, yielding an apparent
mean age of 600±340 ka .. Evaluating these data further using an age-frequency
plot (Fig. 4.3) revealed that 2 of the 14 analyses produced ages greater thatl
million years. This bimodal age distribution was interpreted to mean that the
analyses were composed of 2 isotopically distinct phases: a primary population
of 12 subsamples that represent the true age of the Minnie Creek basalt
(0.467±0.049 Ma), and the 2 older contaminated subsamples (1.31±0.16 Ma).
An isochron plot of this data set provides additional evidence supporting
the notion that the basalt is contaminated. Isochron diagrams are constructed
from 40Ar/3 9Ar data plotted as either 40Ar/36Ar versus 40Ar/39Ar or 36Ar/40Ar versus
39
Ar!4°Ar: such plots will yield a straight line provided that all data in the set
represent portions of the same rock that formed at the same time with the same
initial 40Ar/3 6Ar ratio. Such plots permit an evaluation of whether the data
conform to ideal behavior or whether systematic errors occur, such as the
presence of excess 40Ar (York 1969; McDougall and Harrison 1988). An
 114
ROBERT C WALTER POTASSIUM-ARGONI.A

Mean z 0.467 Ma
SD s 0.049 Ma ·3
3.40 x 10
Bin= 0.02 Mean = 1.31 Ma
n z 12, x = 3 SD z 0.16 Ma
Binz 0.02

....
<( 3.00 x 10
-3

0
~
\::
<(
co .3
M 2.26 x 10

Inter~
·1.0 ·0.6 ·0.2 0.2 0.6 1.0 1.4 1.8 (40/3E
Age (Ma) -3
MSW!
2.20 x 10 n = 14
Figure 4.3. An age-probability graph of fourteen mini-bulk analyses of the late Pleistocene Minnie
Creek basalt. The plot, based on a Gaussian probability distribution function, is a powerful way
40 9 0
to graphically depict a suite of Ar/3 Ar results from a single sample. If all the subsamples analyzed
are the same age, then the data would yield a single bell-shaped curve, where the peak of the curve
is equivalent to the mean age and the tails of the curve represent purely random analytical error
40 39
If, however, the sample is contaminated by grains significantly older than the primary age, then Figure 4.4. Inverse Ar/
multiple curves will be displayed. In the case of the Minnie Creek basalt, the histogram yields two and the Y-intercept value f<
curves, one with an mean age of 1.31±0.16 Ma (n=2), whereas the other yeilds a mean age of This so-called isochron dia.
40 9
0.47±0.05 Ma (n=l2). The horizontal scaie is in millions of years. and the vertical scale is relative conform to ideal Arl Ar
probability. Bin is the bin width (Ma) used to construct the plot, and X is a scaling factor (Deino regardless of K content be~
and Potts, 1992). that accumulates is a functi
points will fall on or stati:
!east-squares, linear regress'
"inverse" isochron plot, so-named when 40Ar is in the denominator, of all 14 square of weighted deviates
laser-fusion analyses was constructed using the least-squares regression tech- from random error, suggestj
nique for correlated errors (York 1969). In this plot, the Y-intercept is equal to is a reduced Chi-squared SI
40 meets the expected distribu
the initial Ar/36Ar ratio and the x-intercept is a function of the sample's age as
defined by the fitted line. Although the initial 40Ar/3 6Ar ratio is not significantly
different from the atmospheric, the mean square of weighted deviates (MSWD) indicates that the con
is greater than expected from pure random errors alone, suggesting that the fit dispersed in the lava,
is poor and that nonrandom errors are associated with the data. The MSWD is into the lava during e
a reduced x statistic Cx 2/n-2), which by definition equals 1 if the observed
2
itself cooled below the
population meets the expected distribution (Bevington 1969). A second iso- that the laser-fusion 1
chron was constructed excluding the 2 analyses that produced the anomalously ages in rock fragment
old dates yielding a MSWD of 1, defining an age of 466±55 ka, and an initial A similar case '
40 6
Ar/3 Ar ratio of 296±4 (Fig. 4.4). This age compares favorably with the volcanic field of New
minimum age observed in the age-frequency plot (Fig. 4.3). assumed because con
These results demonstrate that extraneous 40Ar enrichments can occur candy older than eXJ
even in basaltic lava flows that erupted at temperatures exceeding 1000°C, ubiquitous occurrenc
contrary to one of the primary assumptions in the K-Ar method. In the case of Auckland lavas sugge
the Minnie Creek basalt, the excess 40Ar observed in 2 of 14 rock fragments magmas by older com
 ROBERT C. WALTER
1= 1.31 Ma
0.16 Ma
0.02
'x .. 3
1.4 1.8
,f the late Pleistocene Minnie
function, is a powerful way
: all the subsamples analyzed
where the peak of the curve
rely random analytical error
than the primary age, then
•It, the histogram yields two
other yeilds a mean age of
I the vertical scale is relative
X is a scaling factor (Deina
:lenominator, of all 14
.uares regression tech-
Y-intercept is equal to
i of the sample's age as
1tio is not significantly
ited deviates (MSWD)
suggesting that the fit
1e data. The MSWD is
uals 1 if the observed
1969). A second iso-
.uced the anomalously
±55 ka, and <:J,n initial
~s favorably with the
f.3).
1richments can occur
~s exceeding 1000°C,
nethod. In the case of
of 14 rock fragments
POTASSIUM-ARGON/ARGON-ARGON DATING METHODS 115
.3
3.40 x 10
....
<( 3.00 x 10 -3
0
~
<(
co
(")
.3
2.26 x 10
Intercept Age .. 0.466 ± 0.055 Ma
(40/36)i .. 296.2 ± 3.8
.3
Mswo .. 1.05
2.20 x 10 n .. 12
0 0.5 1.0 1.5 2~0
39Ar/40Ar
Figure4.4. Inverse Ari3 Ar correlation dia~ram showing the X-intercept age (0.47±0.05 Ma)
40 9
and the Y-intercept value for the initial 40Ar/ 6Ar ratio (296.2±0.8) for the Minnie Creek basalt.
This so-called isochron diagram (isochron=Gr. same age) can be used to test how well these data
conform to ideal 40Ar/39 Ar behavior. It is a fundamental assumption that all phases of a rock
40 36 40
regardless of K content begin at t=O with the same initial Ar/ Ar value. The amount of Ar*
that accumulates is a function of K content and time. If these assumptions are met, then all data
points will fall on or statistically near a straight line constructed through the points using a
least-squares, linear regression technique (York 1969). For example, this plot shows that the mean
square of weighted deviates (MSWD) is one, which indicates that the scatter about the line is solely
from random error, suggesting that the Gaussian model used in Fig 4.3 is appropriate. The MSWD
is a reduced Chi-squared statistic (c 2/n-2), which by definition is one if the observed population
meets the expected distribution.
indicates that the contamination is a discrete phase that is inhomogeneously
dispersed in the lava, as if xenocrysts of basement material were incorporll,ted
into the lava during eruption and were incompletelydegassedbefore tne1ava
itself cooled below the blocking temperature for the retention of argon. It seems
that the laser-fusion 40Ar/3 9Ar method was able to measure the dispersion of
ages in rock fragments small enough to reveal this isotopic heterogeneity.
A similar case was made for young basaltic lavas from the Auckland
volcanic field of New Zealand (McDougall et al. 1969), where excess 40Ar was
assumed because conventional K-Ar ages on large bulk samples were signifi-
cantly older than expected from stratigraphic and 14 C considerations. The
ubiquitous occurrence of xenolithic quartz and pyroxene aggregates in the
Auckland lavas suggested that the excess 40Ar was from contamination of the
magmas by older country rock, probably the underlying Mesozoic graywackes
 116 POTASSIUM-ARGONIARG
ROBERT C. WALTER

and argillites (McDougall et al. 1969). In a 40Ar/3 9Ar step heating analysis of (Leakey 1973). The age
one of the same samples analyzed by McDougall et al. (1969), Hall and York divergence between thes
(1984) established the presence of excess 40Ar in the Auckland lavas by 2.61 Ma age for the KB~
obser~ing an Ar* enrichment throughout the spectrum, but particularly at
40
that an age of 2.0 Ma W(
the high temperature, low K portion of the release spectrum. This pattern is East African sites (Cook
most readily explained by the high temperature release of excess 40Ar from controversy was compou
Mesozoic pyroxene xenocrysts, as first suggested by McDougall et al. (1969). from one locality and 1.8
age was later found to b
et al. 1980). This confusi
Hominid Evolution, Time Scale Calibrations, and Laser-Fusion that yielded an age of 2.'
Koobi Fora Formation 1976). Fitch et al. (197f
data and revised their ag
In the vicinity of Lake Turkana, Kenya, there occurs a thick sequence of In the meantime, p
Plio-Pleistocene lacustrine, fluvial and deltaic sediments named the Koobi Fora the KBS Tuff was in the C
Forn:a~ion that contain abundant remains of vertebrate fossils, including Ma (Hillhouse et al. 1977
homm1ds, and stone tools. Interbedded rhyolitic tephras facilitate stratigraphic the KBS Tuff were found t
mapping, correlation and dating. Some of the tephras contain pumice clasts, the Shungura Formation i
from which anorthoclase phenocrysts were separated for K-Ar and 40Ar/39Ar comprehensive and concl
analysis. Seven ruffs were dated (McDougall 1985), ranging from 4.1 to 0.74 gall et al. (1980) and Glei:
Ma, yielding concordant ages on multiple samples from each tephra. The results analyses to suggest ages
are stratigraphically consistent, and their 40Ar/3 9Ar age spectra are nearly flat, (Gleadow 1980; McDoug:
indicating that the feldspars remained a closed system with respect to K and Ma age (Fitch and Miller
Ar isotopes since eruption. The radiometric ages established by this study perhaps due to detrital COI
helped to place the abundant hominid remains from Koo bi Fora into a precise with assiduous sample p1
temporal framework (Feibel et al. 1989). The Turkana Basin is the longest and of -1.8 Ma suggested by C
most comprehensively dated hominid sequence in the world, based on these incomplete fusion of the l
and subsequent studies (Brown et al. 1985; McDougall, 1985; McDougall et al. analyses yielded a concorc
1985; Feibel et al. 1989; McDougall et al. 1992a; Leakey et al. 1995). between K-Ar and 40Ar/3 9i
The radiometric dating story at Koobi Fora has not always been so tidy Furthermore, McDougall
It has, on occasion, been the source of heated controversy; as was the case for analytical errors may exp:
the KBS Tuff (Hay 1980). The KBS Tuff, a water lain, pumice-bearing volcanic In addition to prov
ash, is a prominent stratigraphic marker in the middle of the Koobi Fora radiometric ages were use
Formation. In 1969, primitive stone artifacts were aiScovered in tfie tu ff irself to construct a detailed rr
(Isaac et al. 1971). Feldspars from this ruff were dated by both the conventional Formation (Hillhouse et
K-Ar total fusion method and the newly developed 40Ar!39Ar step-heating the tephra ages from Ko
method (Fitch and Miller 1970). The conventional K-Ar method produced an parently at variance wit~
age of 2.37±0.01 Ma, whereas an age of 2.61±0.26 Ma was assigned to the ruff polarity time scale (GPTS
40 9
based on the Ari3 Ar age spectrum. The younger conventional K-Ar age was Dalrymple 1979), exceec
believed to be erroneous due to argon loss caused by regional hydrothermal first, this discrepancy wi
alteration. Extraneous argon contamination via detrital minerals, which could eruption and redepositioi
account for the older age, was not considered to be a serious problem. in due time it was recogni
By 1972, numerous hominids were found stratigraphically below the provided data to revise <
KBS Tuff, including individuals attributed to Australopithecus and Homo McDougall et al. 1992b).
 ROBERT c. WALTER POTASSIUM-ARGON/ARGON-ARGON DATING METHODS Il7

(Leakey 1973). The age of the KBS was critical for determining when the
r/39Ar step heating analysis ol
divergence between these two hominid groups occurred. Doubt about the
.ll et al. (1969), Hall and York
2.61 Ma age for the KBS began to surface as faunal correlations suggested
\r in the Auckland lavas by
that an age of 2.0 Ma would be more compatible with evidence from other
spectrum, but particularly at
East African sites (Cooke and Maglio 1972; White and Harris 1977). The
ease spectrum. This pattern is
controversy was compounded by new K-Ar dates of 1.60 Ma for the KBS Tuff
re release of excess 40Ar fro111
from one locality and 1.82 Ma from another (Curtis et al. 1975). The 1.6 Ma
d by McDougall et al. (1969).
age was later found to be too young due to a K measurement error (Drake
et al. 1980). This confusion was further complicated by a fission track study
[)IlS, and Laser-Fusion that yielded an age of 2.44±0.08 Ma on zircons from the tuff (Hurford et al.
1976). Fitch et al. (1976) recalculated their original 40Ar/39 Ar step-heating
data and revised their age to 2.42 Ma.
In the meantime, paleomagnetic data strongly supported the view that
.ere occurs a thick sequence ol
the KBS Tuff was in the Olduvai Event and could be no older than about 1.8
diments named the Koo bi Fora
Ma (Hillhouse et al. 1977). Furthermore, glass shards and feldspar grains from
' vertebrate fossils, including
the KBS Tuff were found to be chemically similar to the -1.8 Ma old Tuff H2 of
: tephras facilitate stratigraphic
the Shungura Formation north of Lake Turkana (Cerling et al. 1979). The most
tephras contain pumice clasts,
comprehensive and conclusive study of the KBS Tuff was provided by McDou-
Jara ted for K-Ar and 40Ar/3 9Ar
gall et al. (1980) and Gleadow (1980) who used careful K-Ar and fission track
}85), ranging from 4.1 to 0.74
analyses to suggest ages of 1.89±0.01 Ma and 1.87±0.04 Ma, respectively
es from each tephra. The results
(Gleadow 1980; McDougall et al. 1980). These studies showed that the 2.4-2.6
'9Ar age spectra are nearly flat,
Ma age (Fitch and Miller 1970; Fitch et al. 1976) for the KBS Tuff is too old
. system with respect to K and
perhaps due to detrital contamination ( Gleadow 1980), which could be avoided
lges established by this study
with assiduous sample preparation. Likewise, the slightly younger K-Ar ages
from Koobi Fora into a precise
of -1.8 Ma suggested by Curtis et al. (1975) were incorrect, most likely due to
lrkana Basin is the longest and 40 39
incomplete fusion of the feldspars during argon analysis. Subsequent Ar/ Ar
e in the world, based on these
analyses yielded a concordant age of 1.88±0.02 Ma, resolving the earlier conflict
)ougall, 1985; McDougall et al.
between K-Ar and 40Ar/3 9Ar measurements on the KBS Tuff (McDougall 1981).
~;Leakey et al. 1995).
Furthermore, McDougall (1981) suggested that unrecognized, systematic,
ra has not always been so tidy.
analytical errors may explain the erroneously old 40 Ar/39Ar results.
:ontroversy, as was the case for
In addition to providing calibration points for hominid evolution, the
: lain, pumice-bearing volcanic
radiometric ages were used in conjunction with paleomagnetic measurements
:he middle of the Koobi Fora
to construct a detailed magnetostratigraphicJram~wQ;tl< fQrt_h<eKoobi Fora
·ere discovered in the tuff itsell
Formation (Hillhouse et al. 1986). Feibel et al. (1989) noted. that although
dated by both the conventional
the tephra ages from Koobi Fora were internally consistent, they were ap- I
veloped 40 Ar/39Ar step-heating I I
parently at variance with the estimated boundary ages for .the geomagnetic
:mal K-Ar method produced an
polarity time scale (GPTS) (McDougall and Chamaluan 1966; Mankinen and
.26 Ma was assigned to the tufl
Dalrymple 1979), exceeding the boundary ages by about 100,000 years. At
.ger conventional K-Ar age was
first, this discrepancy was thought to be due to a time lag between tephra
used by regional hydrothermal
~ruption and redeposition in the sedimentary record (Feibel et al. 1989), but
L detrital minerals, which could
m due time it was recognized that the radiometric ages for Koo bi Fora actually
:o be a serious problem.
provided data to revise and recalibrate the GPTS (McDougall et al. 1992a;
nd stratigraphically below the
McDougall et al. 1992b). .
:o Australopithecus and Homo
 118
ROBERT C. WALTER POTASSIUM-ARGONIi

The notion that the GPTS needed to be revised was first proposed by
inference, the Sidi Ha
Johnson (1982) and elaborated by Shackelton et al. (1990) based on marine-
older that 3.4 Ma. Sul
astronomical calibrations using oxygen isotopes as proxy indicators of climatic
the Gulf of Aden that:
fluctuations due to variations in the earth's orbit. The initial purpose of this
Bor Tuffs (Sarna-Wojc
type of study was to evaluate the orbital theory of the Pleistocene ice ages and
tephra layer. Addition:
to develop a high resolution time scale for the past 780 ka Qohnson 1982;
Tulu Bor Tuff to 3.36±
Imbrie et al. 1984). Such orbitally tuned data have now been used to calibrate
known age (Feibel et u
an "astronomical" polarity time scale (APTS) for the past 3 to 4 Ma (Shackleton
directly or reliably dat
et al. 1990; Hilgen, 1991), indicating that the ages for the radiometrically based
In 1990, after a
GPTS were consistently too young by 5 to 7%. Similar age discrepancies were
resampled for dating.
subsequently noted radiometrically for the Olduvai Subchron (Walter et al.
technological improve
1991) and for the Brunhes/Matuyama boundary (Baksi et al. 1992). The most
in dating young, smal
likely explanation for the consistent underestimation of the widely used K-Ar
et al. 1981; LoBello et
based GPTS is due to loss of small but significant amounts of 40Ar* from
provided the first reh
basalts-until now the most commonly used material for GPTS calibra-
markers at Hadar, ind
tion-most likely from glassy, poorly crystallized, or altered phases, despite
Hadar Tuff (Walter an
careful choice of samples (McDougall et al. 1992a).
Alkali feldspars :
six small grains (ca. 0
Hadar Formation of tephra. Some grain:
with the primary teph1
The Hadar Formation of Ethiopia is one of the most prolific early hominid
in 1984 the six grains
sites in the world. Initial surveys in the early 1970s recovered more than 240
the new laser-fusion
hominid specimens including the partial skeleton nicknamed "Lucy" and the
University of Toronto,
fossil assemblage of 13 hominid individuals called the "First Family". All of
mass spectrometers. ~
these specimens are attributed to Australopithecus afarensis Qohanson et al.
(0.4 to 0.2 mm in widt
1978). Accurate age estimates for the Hadar hominid fossils, however, were
Four single grain and
elusive for many years. Initial dating attempts were hampered by either a
fuse the grains and a ~
scarcity of datable components in the tephras or by detrital contamination
of3.40±0.03 Ma (Wall
(Walter 1981; Walter and Aronson 1982) Unlike at Koobi Fora, none of the
for the lower Hadar
Hadar tephra contain pumice clasts from which primary feldspar could be
estimate for the oldest
extracted, prohibiting the selection of obvious primary phases. Attempts to date
The Kada Hadar
the altered Kadada Moumou basalt by bulk K-Ar and 49Ar/39Ar methods proved
Kada Hadar Member, t
difficult because of an apparent loss of 40Ar during alteration, coupled with
the First Family occur
variable amounts of excess 40Ar (Aronson et al. 1977; Walter and Aronson 1982;
Renne et al. 1993; Walter and Aronson 1993). Previous attempts to d
Ma on bulk fusions of[
The apparent K-Ar age of 3.6 Ma for this basalt was challenged when a
zircon indicated sign
chemical correlation was proposed between the Sidi Hakoma Tuff at Hadar and
tephras, including the l
the Tulu Bor Tuff at east Turkana (Brown and Ceding 1982). The Sidi Hakoma
single-crystal laser-fusi
Tuff defines the base of the Sidi Hakoma Member of the Hadar Formation, and
of 3.17±0.03, 3.18±0.0
is stratigraphically below the basalt. The Tulu Bor Tuff is a widespread marker
where 10 individual gi
bed in the Turkana basin of northern Kenya and southern Ethiopia, where its age
grains measured, fom
is well constrained within the Gauss normal chron (Brown et al. 1978), between
explaining the anomi
3.15 and 3.4 Ma (using boundary ages for the GPTS known at that time). By
method. Pooling the n
 ROBERT C. WALTER
1sed was first proposed by
. (1990) based on marine-
Jroxy indicators of climatic
The initial purpose of this
1e Pleistocene ice ages and
st 780 ka Qohnson 1982;
1ow been used to calibrate
past 3 to 4 Ma (Shackleton
r the radiometrically based
lar age discrepancies were
ti Subchron (Walter et al.
1ksi et al. 1992). The most
n of the widely used K-Ar
Lt amounts of 40Ar* from
aterial for GPTS calibra-
or altered phases, despite
1ost prolific early hominid
recovered more than 240
icknamed "Lucy" and the
the "First Family". All of
a
afarensis ohanson et al.
Lid fossils, however, were
re hampered by either a
'Y detrital contamination
Koobi Fora, none of the
rimary feldspar could be
y phases. Attempts to date
40
Ar/39Ar methods proved
alteration, coupled with
,\'alter and Aronson 1982;
lt was challenged when a
Iakoma Tuff at Hadar and
1982). The Sidi Hakoma
he Hadar Formation, and
ff is a widespread marker
:rn Ethiopia, where its age
own et al. 1978), between
known at that time). By
POTASSIUM-ARGON/ARGON-ARGON DATING METHODS 119
inference, the Sidi Hakoma Tuff, and the Kadada Moumou basalt, could be no
older that 3.4 Ma. Subsequently, a tephra layer was found in DSDP cores from
the Gulf of Aden that matched the glass chemistry of the Sidi Hakoma and Tulu
Bor Tuffs (Sarna-Wojcicki et al. 1985), suggesting an age of about 3.2 Ma for the
tephra layer. Additional dating studies at Turkana refined the age estimate of the
Tulu Bor Tuff to 3.36±0.02 Ma, based on stratigraphic scaling between units of
known age (Feibel et al. 1989). However, the Tulu Bor Tuff itself has never been
directly or reliably dated by any radiometric technique.
In 1990, after a 14-year hiatus in field activity, the Hadar Formation was
resampled for dating. The renewed sampling strategy was guided by recent
technological improvements in 40Ar/3 9Ar analyses that achieved a breakthrough
in dating young, small samples using a laser beam to melt the material (York
et al. 1981; LoBello et al. 1987; van den Bogaard et al. 1987). This technique
provided the first reliable radiometric ages for important tephrostratigraphic
markers at Hadar, including the Sidi Hakoma Tuff, the Triple Tuff, and the Kada
Hadar Tuff (Walter and Aronson 1993; Walter 1994).
Alkali feldspars are extremely rare in the Sidi Hakoma Tuff. In 1977,just
six small grains (ca. 0.2 mm in diameter) were recovered from roughly 50 kg
of tephra. Some grains had glass rims, suggesting that they were comagmatic
with the primary tephra. Being insufficient by far for conventional K-Ar dating,
in 1984 the six grains were consumed in a failed attempt to date them using
the new laser-fusion 40Ar!39Ar dating system then being developed at the
University of Toronto, prior to the advent of the latest generation of sensitive
mass spectrometers. After resampling this tephra in 1990, 20 feldspar grains
(0.4 to 0.2 mm in width) were recovered from approximately 200 kg of tephra.
Four single grain and four multiple grain dates were obtained, using a laser to
fuse the grains and a sensitive, low blank mass spectrometer, yielding an age
of 3.40±0.03 Ma (Walter and Aronson 1993). This result established a firm age
for the lower Hadar Formation, providing the first direct radiometric age
estimate for the oldest A. afarensis fossils at Hadar.
The Kada Hadar Tuff is a widespread marker that defines the base of the
Kada Hadar Member, the youngest member of the Hadar Formation. Lucy and
the First Family occur a few meters above and below this tephra, respectively.
Previous attempts to date the Kada Hadar Tuff yielded K-Ar dates of 4.1 to 4.6
Ma on bulk fusions of K-feldspars (Aronson et al. 1980). Fission track dates on
zircon indicated significant amounts of Miocene contamination ill Hadar
tephras, including the Kada Hadar Tuff (Aronson et al. 1980; Walter 1981). New
single-crystal laser-fusion 40Ar/39Ar analyses (Walter 1994) yielded mean dates
of 3.17±0.03, 3.18±0.02, and 3.19±0.01 Ma for three subsamples of the tephra,
where 10 individual grain ages were pooled for each subsample. Of the thirty
grains measured, four produced Miocene ages ranging from 8 to 24 Ma,
explaining the anomalously old dates obtained by the conventional K-Ar
method. Pooling the remaining 26 grains yielded a mean age of 3.18±0.01 Ma.
 120
ROBERT C. WALTER
POTASSIUM-ARGONIA

Lucy, recovered from a small sand channel 2-3 m above this tuff, is now precisely
dated to just less than 3.18 Ma, probably within the standard error of the mean. (2) smaller samples ca
Early work at Hadar and elsewhere showed morphological stasis in A. heterogeneity, precisio1
afarensis Qohanson and White 1979), but temporal controls were inadequate. obtained by increment
We are now in a position to accurately measure rates of faunal change at Hadar, whether the fundamen
and to suggest that A. afarensis existed, virtually unchanged for at least 500 ka The advent of the laser-
(Kimbel et al. 1994). The tempo of hominid evolution at Hadar is now well precise ages to be obt
established solely because the laser-fusion 40Ar/3 9Ar method was able to date mineral grains or smal
individual mineral grains. The 40Ar/3 9Ar lase
In addition to calibrating hominid evolution at Hadar, these ages provided can be used on proble1
supportive evidence that the boundary ages for the Kaena and Mammoth that have insufficient n
subchrons of the GPTS needed to be increased (Walter et al. 1992a; Renne et bulk fusion methods. 1
al. 1993; Walter and Aronson 1993; Walter 1994) similar to the conclusion provided new insights i
drawn by McDougall et al. (1992). It seems that high-precision ages on of igneous (Pringle et G
K-feldspars from rhyolitic tephra in long sedimentary sequences is an excellent Hodges 1993), impact
way to calibrate magnetostratigraphic events. et al. 1993). An excelle
been employed to calil
contamination of teph1
Recent Discoveries these ashes by bulk-fu~
In addition to the examples presented above, the laser-fusion 40Ar/39Ar The laser-fusion
method has been applied to: (1) dating the earliest evidence of Acheulian research tool for some
artifacts to 1.4 Ma at the site ofKonso-Gardula in Ethiopia (Asfaw et al. 1992); date the late Quaterna1
(2) the oldest Australopithecus afarensis fossils dated to between 3.4±0.04 Ma 1995; Chen et al. 1996:
and 3.89 ± 0.02 Ma, from Maka, Ethiopia (Hall et al. 1984; White et al. 1993); samples that overlap in
40
(3) calibrating a new hominid genus, Ardipithecus ramidus, dated to around 4.4 Ar/3 9Ar dating will be
14
( ages, particularly tb
Ma, from Aramis, Ethiopia (White et al. 1994; WoldeGabriel et al. 1994; White
et al. 1995) and (4) calibrating a new 4.0 Ma old species, Australopithecus the correction for the v
anamensis, from Kenya that appears to be a direct ancestor of Australopithecus may not be well define1
afarensis (Leakey et al. 1995). In each case, the ability to detect and eliminate be intriguing to see wb
detrital contamination by dating single grains has been the only way that method will be applied
reliable dates could have been obtained. Without the single-crystal laser-fusion
40 39
Ar/ Ar method, hominid evolution in East Africa would not be as precisely
calibrated as it is today. ACKNOWLEDGME

I extend my sine~
CONCLUSION to contribute a chapter!
gratefully acknowledge
The K-Ar method is one of the most commonly used of all radiometric an anonymous reviewe:
techniques mainly because of its versatility and its applicability to such a wide helpful suggestions. I
range of ages. Over the past thirty years, the 40Ar/39Ar techn.ique, an .elegant advisors, friends, stude
variation of the conventional K-Ar method, has become a major force m K-Ar knowledge of K-Ar and
geochronology because: (1) ages are based entirely on the measurement of Ar son, Derek York, Johr
isotopes in a single sample, and do not require a separate measurement of K: Gabriel, Bill Hackett, ,
Smith, Yanshao Chen,
 ROBERT C. WALTER.
bove this tuff, is now precisely
1e standard error of the mean.
~d morphological stasis in A.
ral controls were inadequate.
.tes of faunal change at Hadar,
unchanged for at least 500 ka
olution at Hadar is now well
19
Ar method was able to date
tat Hadar, these ages provided
>r the Kaena and Mammoth
Walter et al. 1992a; Renne et
)4) similar to the conclusion
that high-precision ages on
.tary sequences is an excellent
>ve, the laser-fusion 40Ar/3 9Ar
.rliest evidence of Acheulian
1 Ethiopia (Asfaw et al. 1992);
ated to between 3.4±0.04 Ma
t al. 1984; White et al. 1993);
; ramidus, dated to around 4.4
)ldeGabriel et al. 1994; White
old species, Australopithecus
:t ancestor of Australopithecus
1bility to detect and eliminate
has been the only way that
the single-crystal laser-fusion
tea would not be as precisely
nonly used of all radiometric
ts applicability to such a wide
A.r/39Ar technique, an elegant
become a major force in K-Ar
dy on the measurement of Ar
~ separate measurement of K:
POTASSIUM-ARGON/ARGON-ARGON DATING METHODS 121
(2) smaller samples can be dated; (3) by circumventing errors due to sample
heterogeneity, precision and accuracy are improved; and (4) age spectra can be
obtained by incrementally heating the sample, providing a means to evaluate
whether the fundamental assumption of closed system behavior has been met.
The advent of the laser-fusion 40Ar/39Ar technique over the past 20 years permits
precise ages to be obtained on very small samples, even the size of single
mineral grains or smaller. .
The 40Ar/3 9Ar laser-microprobe has revolutionize.cl K-Ar dating because it
can be used on problematic samples, such as contamin;:i.ted samples or those
that have insufficient material, which could not have been dated by traditional
bulk fusion methods. The grain-discrete ability of the laser-fusion method has
provided new insights into K-Ar systematics that can now be appl.ied to a variety
of igneous (Pringle et al. 1991; Pringle et al. 1992), metamorphic (Hames and
Hodges 1993), impact (Izett et al. 1991), and even sedimentary rocks (Smith
et al. 1993). An excellent example of its effectiveness is seen in how well it has
been employed to calibrate the hominid record in East Africa, where detrital
contamination of tephra has long plagued earlier attempts to accurately date
these ashes by bulk-fusion K-Ar and 40Ar/3 9Ar means.
The laser-fusion 40Ar/3 9Ar method will continue to be an invaluable
research tool for some time to come. Currently, the method is being used to
date the late Quaternary (Deino et al. 1994; Hu et al. 1994; van den Bogaard
1995; Chen et al. 1996), with the aim of providing precise single-grain ages on
samples that overlap in time with the 14C method. If this can be achieved, then
40
Ar/39Ar dating will be used as a means to check and possibly even calibrate
14
C ages, particularly those at the older end of the radiocarbon spectrum w~ere
the correction for the variation in cosmogenic production of 14 C through time
may not be well defined. The future of K-Ar dating lies in its versati~ity. It wi.ll
be intriguing to see where, how, and in what form the next generat10n of this
method will be applied.
ACKNOWLEDGMENTS
I extend my sincere appreciation to Professor R. E. Taylor for inviting me
to contribute a chapter to this volume. His encouragement and forbearance are
gratefully acknowledged. I wish to thank Donald Johanson, Eric Meikle, and
an anonymous reviewer for critically reading the manuscript, and for providing
helpful suggestions. I am indebted to a number of individuals (teachers,
advisors, friends, students, and colleagues) who have helped to formulate my
knowledge of K-Ar and 40Ar/3 9Ar systematics and tephrochronology: Jim Aron-
son, Derek York, John Westgate, Stan Mertzman, Bill Hart, Giday Wolde-
Gabriel, Bill Hackett, Chris Hall, Paul van den Bogaard, Norm Evensen, Pat
Smith, Yanshao Chen, Paul Manega, and Richard Hay, to name but a few. If,
 122 ROBERT C. WALTER POTASSIUM-ARGON/AR

however, this chapter suffers from any errors, .they are mine alone. I am Cooke, H.B.S., and Maglio, 1
extremely grateful for the financial support of the Institute of Human Origins, proboscidean and sui
the National Science Foundation, and the National Geographic Society for Hominoid Evolution EC
Curtis, G.H. 1966 The probl1
providing the research platform from which this manuscript was written. I rocks. In Shaeffer, O.A
extend my thanks to the Ethiopian Ministry of Culture and Information and Verlag: 151-162.
the Ethiopian Institute of Geological Surveys for their continued support of Curtis, G.H., Drake, R.E., C
our research at Hadar. I also thank my colleagues at the Institute of Human Formation, East Rudo
Origin and with the Hadar Research Project for the many shared experiences Curtis, G.H., and Evernden,j.
Dalrymple, G.B. 1991 The Ag,
that have helped to shape my appreciation for the complexities of human Dalrymple, G.B., and Duffielc
evolution. Finally, I dedicate this work to my wife, Gail, for her patience and from the Mogollon-I:
wisdom. Research Letters 15: 46
Dalrymple, G.B., and Lanphe1
Dalrymple, G.B., and Lanphe1
the conventional tech1
REFERENCES Deino, A.L., Curtis, G.H., Sou
of the Campanian Igni
Aldrich, LT., and Nier, A.O. 1948 Argon 40 in potassium minerals. Physical Review. 74: 876-877. Turrin, B. ICOG-8 Ab!
Aronson,j.L., Schmitt, T.]., Walter, R.C., Taieb, M., Tiercelin,j.-j.,johanson, D.C., Naeser, C.W., Drake, R.E., Curtis, G.H., Ce
and Nairn, A.E.M. 1977 New geochronologic and paleomagnetic data for the hominid-bear- geochronology of tuffa
ing Hadar Formation of Ethiopia. Nature 267: 323-327. Africa. Nature 283: 36:
Aronson,j.L., Walter, R.C., Taieb, M., and Naeser, C.W. 1980 New Geochronological Information Evernden,j.A., and Curtis, G
for the Hadar Formation and the adjacent Central Afar, Ethiopia. In Leakey, R.E., and Ogot, Africa and Italy. Curre1
B.A., eds., Proceedings of the 8th Pan African Congress of Prehistory and Quaternary Studies, Evernden, j.F., Curtis, G.H.,
Nairobi 5to10September1977. Nairobi, The International Louis Leakey Memorial Institute canics. Quaternaria 5:
for African Prehistory: 47-52. Faure, G. 1986 Principles of ii
Asfaw, B., Beyne, Y., Suwa, G., Walter, R.C., White, T.D., WoldeGabriel, G., and Yemane, T. 1992 Feibel, C.S., Brown, F.H., and
Konso-Gardula: the earliest Acheulian. Nature 360: 732-735. the Omo Group Depo!
Baksi, A.K., Hsu, V., McWilliams, M.O., and Farrar, E. 1992 40Ar/39Ar Dating of the Brunhes-Ma- of Physical Anthropoloi
tuyama Geomagnetic Field Reversal. Science 256: 256-257. Feraud, G., York, D., Hall, C
Berger, G.W., and York, D. 1970 Precision of the 40Ar/39Ar dating technique. Earth Planetary Science Acheulian site in Israe
Letters 9: 39-44. Fitch, F.j., Hooker, P.j., and
Bevington, P.R. 1969 Data Reduction and Error Analysis for the Physical Sciences. New York, Formation, East Rudo!
McGraw-Hill, Inc. Fitch, F.J., and Miller, J.A. l~
Brereton, N.T. 1970 Corrections for interfering isotopes in the 40Ar!39Ar dating method. Earth and Nature 226: 226-228.
Planetary Science Letters 8: 427-433. Glass, B.P., Hall, C.H., and Yo
Brock, A., and Hay, R.L. 1976 The Olduvai Event at Olduvai Gorge. Earth and Planetary Science Fragments from Barba1
Letters 29: 126-130. (Isotope Geoscience) 5\
Brown, F.H., and Cerling, I.E. 1982 Stratigraphical significance of the Tulu Bor Tuff of the Koobi Gleadow, A.j.W. 1980 Fission.
Fora Formation. Nature 299: 212-215. Kenya. Nature 284: 2~
1

Brown, F.H., McDougall, I., Davies, T., and Maier, R. 1985 An integrated Plio-Pleistocene chronol- Gromme, C.S., and Hay, R.L.
ogy of the Turkana Basin. In Delson, E., ed., Paleoanthropology: The Hard Evidence. New 560-561.
York, Alan R. Liss: 82-90. Gromme, C.S., and Hay, R.L.
Brown, F.H., Shuey, R.T., and Croes, M.K. 1978 Magnetostratigraphy of the Shungura and Usno Earth and Planetary Sc
Formations, southwestern Ethiopia: new data and comprehensive reanalysis. Geophysical Hall, C.M., Walter, R.C., W1
journal of the Royal Astromomical Society 54: 519-538. geochemistry of Cind
Cerling, I.E., Brown, F.H., Cerling, B.W., Curtis, G.H., and Drake, R.E. 1979 Preliminary correla- Awash, Ethiopia. Natu
tions between the Koobi Fora and Shungura Formations, East Africa. Nature 79: 118-121. Hall, C.M., and York, D. 1984
Chen, Y., Smith, P.E., Evensen, N.M., and York, D. 1996 The edge of time: dating young volcanic W.C., eds., Quaterna1y,
ash layers with the 40Ar- 39Ar laser probe. Science 274: 1176-1178. Halliday, A.N., Mahood, G.A
Evidence for long resi1
 ROBERT C. WALTER
r are mine alone. I am
titute of Human Origins,
Geographic Society for
muscript was written. I
ire and Information and
:ir continued support of
the Institute of Human
nany shared experiences
complexities of human
ail, for herpatience and
. Physical Review. 74: 876-877.
Johanson, D.C., Naeser, C.W.,
tetic data for the hominid-bear-
Geochronological Information
)pia. In Leakey, R.E., and Ogot,
history and Quaternary Studies,
)Uis Leakey Memorial Institute
briel, G., and Yemane, T. 1992
5.
'Ar Dating of the Brunhes-Ma-
hnique. Earth Planetary Science
Physical Sciences. New York,
;3 9 Ar dating method. Earth and
5e. Earth and Planetary Science
the Tulu Bor Tuff of the Koo bi
rated Plio-Pleistocene chronol-
ology: The Hard Evidence. New
)hy of the Shungura and Usno
b.ensive reanalysis. Geophysical
R.E. 1979 Preliminary correla-
ast Africa. Nature 79: 118-121.
of time: dating young volcanic
-1178.
POTASSIUM-ARGON/ARGON-ARGON DATING METHODS 123
Cooke, H.B.S., and Maglio, V.J. 1972 Pho-Pleistocene stratigraphy in East Africa in relation to
proboscidean and suid evolution In Bishop, WW and Miller, ].A., ed., Calibration of
Hominoid Evolution Edinburgh, Scottish Academic Press, Ltd.: 303-329.
Curtis, G.H. 1966 The problem of contamination in obtaining accurate dates of young geologic
rocks. In Shaeffer, 0.A., and Zaringer, ]., eds., Potassium argon dating. New York, Springer-
Verlag: 151-162.
Curtis, G.H., Drake, R.E., Cerling, T.E., and Hampel,]. 1975 Age of KBS ruff in Koobi Fora
Formation, East Rudolf, Kenya. Nature 258: 395-398.
Curtis, G.H., and Evemden,J.F.1962Age of basalt underlying Bed I, Olduvai. Nature 194: 610-612.
Dalrymple, G.B. 1991 The Age of the Earth. Stanford, Stanford University Press.
Dalrymple, G.B., and Duffield, 'WA. 1988 High precision 40Ar/39Ar dating of Oligocene rhyolites
from the Mogollon-Datil volcanic field using a continuous laser system. Geophyscial
Research Letters 15: 463-466.
Dalrymple, G.B., and Lanphere, M.A. 1969 Potassium-Argon Dating. San Francisco, WH. Freeman.
Dalrymple, G.B., and Lanphere, M.A. 1971 40Ari3 9Ar technique ofK-Ar dating: a comparison with
the conventional technique Earth and Planetary Science Letters 12: 300-308.
Deino, A.L., Curtis, G.H., Southon,J., Terrasi, F., and Campajola, L. 1994 14 C and 40Ari39Ar dating
of the Campanian Ignimbrite, Phelgrean Fields, Italy; Dalrymple, G.B., Lanphere, M.A., and
Turrin, B. ICOG-8 Abstract. Circular 1107 32. Berkeley; U.S. Geological Survey:
Drake, R.E., Curtis, G.H., Cerling, T.E., Cerling, B.W, and Hampel,]. 1980 KBS Tuff dating and
geochronology of tuffaceous sediments in the Koobi Fora and Shungura Formations, East
Africa. Nature 283: 368-372.
Evernden,j.A., and Curtis, G.H. 1965 The potassium-argon dating oflate Cenozoic rocks in East
Africa and Italy. Cumnt Anthropology. 6: 343-364.
Evernden, J.F., Curtis, G.H., and Kistler, R.W. 1957 Potassium-argon dating of Pleistocene vol-
canics. Quatemaria 5: 348-385.
Faure, G. 1986 Principles of isotope geology. New York, John Wiley&: Sons.
Feibel, C.S., Brown, F.H., and McDougall, I. 1989 Stratigraphic Context of Fossil Hominids From
the Omo Group Deposits: Northern Turkana Basin, Kenya and Ethiopia. American]oumal
of Physical Anthropology 78: 595-622.
Feraud, G., York, D., Hall, C.M., Goren, N., and Schwarcz, H.P. 1983 40Ar/39Ar age limit for an
Acheulian site in Israel. Nature 304: 263-265.
Fitch, F.]., Hooker, P.]., and Miller, ].A. 1976 40Ar/39Ar dating of the KBS tuff in Koobi Fora
Formation, East Rudolf, Kenya. Nature 263: 740-744.
Fitch, F.J., and Miller, ].A. 1970 Radioisotopic age determinations of Lake Rudolf artefact site.
Nature 226: 226-228.
Glass, B.P., Hall, C.H., and York, D. 1986 4l1Ar/39Ar Laser-Probe Dating of North American Tektite
Fragments from Barbados and the Age of the Eocene-Oligocene Boundar-Y. Chemical Geology
(Isotope Geoscience) 59: 181-186.
Gleadow, A.J.W. 1980 Fission track age of the KBS tuff and associated hominid remains in northern
Kenya. Nature 284: 225-230.
Gromme, C.S., and Hay, R.L. 1963 Magnetization of basalt in Bed I, Olduvai Gorge. Nature 200:
560-561.
Gromme, C.S., and Hay; R.L 1967 Geomagnetic polarity epochs: New data from Olduvai Gorge.
Earth and Planetary Science Letters 2: 111-115.
Hall, C.M., Walter, R.C., Westgate, ].A., and York, D. 1984 Geochronology, stratigraphy and
geochemistry of Cindery Tuff in the Pliocene hominid-bearing sediments of the Middle
Awash, Ethiopia. Nature 308: 26-31.
Hall, C.M., and York, D. 1984 The applicability of 40Ar/39Ar dating to young volcanics. In Mahaney;
WC., eds., Quaternary dating methods. New York, Elsevier: 67-74.
Halliday; A.N., Mahood, G.A., Holden, P., Metz, ].M.,'Dempster, T.]., and Davidson, JP 1989
Evidence for long residence times of rhyolitic magma in the Long Valley magmatic system:
124 ROBERT C. WALTER POTASSIUM-ARGON!ARG

the isotopic record in the precaldera lavas of Glass Mountain. Earth and Planetary Science Leakey, R.E.E 1973 Further ev
Letters 94: 272-290. Kenya, 1972. Nature 242
Hames, WE., and Hodges, K.V. 1993 Laser 40Ar/39 Ar evaluation of slow cooling and episodic loss LoBello, P., Feraud, G., Hall, CJ
of 40Ar from a sample of polymetamorphic muscovite. Science 261: 1721-1723. and laser fusion dating ol
Hay, R.L. 1976 Geology of the Olduvai Gorge: A Study of Sedimentation in a Semiarid Basin. Berkeley, defeat of xenocrystic cor
University of California Press. Mankinen, E.A., and Dalrymple,
Hay, R.L. 1980 The KBS controversy may be ended. Nature 284: 401. 0-5 m.y. B.P.joumal of G
Hay, R.L. 1992 Potassium-argon dating of Bed I, Olduvai Gorge, 1961-1972. Quaternary Interna- McDougall, I. 1981 40Ar/39 Ar ag
tional 13/14: 31-36. 120-124.
Hilgen, Ej. 1991 Astronomical calibration of Gauss to Matuyama sapropels in the Mediterranean McDougall, I. 1985 K-Ar and
and implications for the Geomagnetic Polarity Timescale. Earth and Planetary Science sequence at Koobi Fora
Letters 104: 226-244. Bulletin 96: 159-175.
Hillhouse, ].W., Cerling, T.E., and Brown, EH. 1986 Magnetostratigraphy of the Koobi Fora McDougall, I., Brown, EH., Cer:
Formation, Lake Turkana, Kenya. journal of Geophysical Research 91: 11581-11595. netic polarity time scale t
Hillhouse,j.W., Ndombi,].W.M., Cox, A., and Brock, A. 1977 Additional results on paleomagnetic Research Letters 19: 234~
stratigraphy of the Koobi Fora Formation, east of Lake Turkana (Lake Rudolf), Kenya. McDougall, I., Brown, EH., Cerl
Nature 265: 411-415. netic polarity time scale
Hu, Q., Smith, P.E., Evensen, N., and York, D. 1994 Lasing in the Holocene: extending the 40Ar/39 Ar 629.
laser probe method into the carbon-14 range. Earth and Plalnetary Science Letters 123: McDougall, I., and Chamalua1
331-336. 1415-1418.
Hurford, A.J., Gleadow, A.].W., and Naeser, C.W. 1976 Fission-track dating of pumice from the McDougall, I., Davies, T., Maie
KBS tuff, East Rudolf, Kenya. Nature 263: 738-744. Koobi Fora, Kenya. Nati
Imbrie, J., Hays, j.D., Martinson, D.G., Mcintyre, A., Mix, A.C., Morley,].]., Pisias, N.G., Prell, McDougall, I., and Harrison, 1
W.L., and Shackleton, N.]. 1984 The orbital theory of Pleistocene Climate: Support from a Method. New York, Oxfc
revised chronology of the marine dl 80 record Milanlwvitch and Climate. Hingham. (Massa- McDougall, I., Maier, R., Southe
chusettes), D. Reidel: 269-305. the KBS tuff, East Turka1
Isaac, G.L., Leakey, R.E.E, and Behrensmeyer, A.K. 1971 Archaeological traces of early hominid McDougall, I., Polach, H.A., and
activities, east of Lake Rudolf, Kenya. Science 173: 1129-1134. from the Auckland vole
Izett, G.A., Dalrymple, G.B., and Snee, L.W. 1991 40Ari39Ar Age of Cretaceous-Tertiary boundary
1485-1520.
tektites from Haiti. Science 252: 1539-1542.
Megrue, G.H. 1973 Spatial di:
Johanson, D.C., and White, T.D. 1979 A systematic assessment of early African hominids. Science
Geophysical Research 78:
202: 321-330.
Merrihue, C.M. 1965 Trace-elen
Johanson, D.C., White, T.D., and Coppens, Y. 1978 A new species of the genus Australopithecus
copy of neutron-irradia tE
(Primates: Hominidae) from the Pliocene of eastern Africa. Kirtlandia 28: 1-14.
Merrihue, C.M., and Turner, G
Johnson, R.G. 1982 Brunhes-Matuyama magnetic reversal dated at 790,000 yr B.P. by marine-as-
journal of Geophysical RE
tronomical correlations. Quaternary Research 17: 135-147.
Mitchell, j.G. 1968 The argon-
Kauffman, D.S., and Hopkins, D.M. 1985 Late Cenozoic radiometric dates, Seward and Baldwin
chimica et Cosmochimica
Peninsulas and adjacent continental shelf, Alaska. U.S. Geological Survey Open-File Report
Nier, A.O. 1950 A redeterminat
85-374.
Kauffman, D.S., and Hopkins, D.M. 1986 Glacial history of the Seward Peninsula. In Hamilton, T., oxygen, argon, and pota
Reed, K., and Thorson, R., eds., Glaciation in Alaska: The Geologic Record. Anchorage, Pringle, M.5., McWilliams, M., i
Alaska, Alaska Geological Survey: 51-77. dating of Quaternary fe:
Kauffman, D.S., Walter, R.C., Brigham-Grette,]., and Hopkins, D.M. 1991 Middle Pleistocene age Geology 20: 531-534.
of the Nome River glaciation, northwestern Alaska. Quaterary Research 36: 277-293. Pringle, M.S., Staudigel, H., anc
Kimbel, W.H., Johanson, D.C., and Rak, Y. 1994 The first skull and other new discoveries of Geology 19: 364-368.
Australopithecus afarensis at Hadar, Ethiopia. Nature 368: 449-451. Renne, P.R., Walter, R.C., Veros
Leakey, LS.B., Evernden, ].A., and Curtis, G.H. 1962 Age of Basalt underlying Bed I, Olduvai. Hadar Formation (Ethi1
Nature 194: 1-7. Research Letters 20: 106"
Leakey, L.S.B., Evernden, j.E, and Curtis, G.H. 1961 Age of Bed I, Olduvai Gorge, Tanganyika. Reynolds, ]. 1956 High sensiti1
Nature 191: 478-479. Instruments 27: 928-93~'
Leakey, M.G., Feibel, C.S., McDougall, I., and Walker, A. 1995 New four-million-year-old hominid Sarna-Wojcicki, A.M., Meyer, 1
species from Kanapoi and Allia Bay, Kenya. Nature 376: 565-51. African early hominid si
JBERT C. WALTER POTASSIUM-ARGON/ARGON-ARGON DATING METHODS 125

1 and Planetary Science Leakey, R.E.F. 1973 Further evidence of lower Pleistocene hominids from East Rudolf, North
Kenya, 1972. Nature 242: 170-173.
iling and episodic loss LoBello, P., Feraud, G., Hall, C.M., York, D., Lavina, P., and Bernat, M. 1987 40Ar/39Ar step-heating
1721-1723. and laser fusion dating of a Quaternary pumice from Neschers, Massif Central, France: The
miarid Basin. Berkeley, defeat of xenocrystic contamination. Chemical Geology (Isotope Geoscience) 66: 61-71.
Mankinen, E.A., and Dalrymple, G.B. 1979 Revised geomagnetic polarity time scale for the interval
0-5 m.y. B.P. Journal of Geophysical Research 84: 615-626.
2. Quaternary Interna- McDougall, I. 1981 40Art39Ar age spectra from the KBS Tuff, Koobi Fora Formation. Nature 294:
120-124.
; in the Mediterranean McDougall, I. 1985 K-Ar and 40Ar/39 Ar dating of the hominid-bearing Pliocene-Pleistocene
and Planetary Seience sequence at Koobi Fora, Lake Turkana, northern Kenya. Geological Society of American
Bulletin 96: 159-175.
1y of the Koobi Fora McDougall, I., Brown, F.H., Cerling, I.E., and Hillhouse, J.W 1992a A reappraisal of the geomag-
l: 11581-11595. netic polarity time scale to 4 Ma using data from the Turkana Basin, East Africa. Geophysical
;ults on paleomagnetic i
Research Letters 19: 2349-2352. l'1
:Lake RudolD, Kenya. McDougall, I., Brown, F.H., Cerling, I.E., and Hillhouse,j.W 1992b A reappraisal of the geomag- I

netic polarity time scale to 4 Ma using data from the Turkana Basin, East Africa. Eos 73:
9
~xtending the "'0Ar/3 Ar 629.
y Science Letters 123: McDougall, I., and Chamaluan, F.H. 1966 Geomagnetic polarity scale of time. Nature 212:
1415-1418.
1g of pumice from the McDougall, I., Davies, T., Maier, R., and Rudowski, R. 1985 Age of the Okote Tuff Complex at
Koobi Fora, Kenya. Nature 316: 792-794 .
.]., Pisias, N.G., Prell, McDougall, I., and Harrison, T.M. 1988 Geochronology and Thermochronology by the Ar/ Ar
40 39

limate: Support from a Method. New York, Oxford University Press.

lte. Hingham, (Massa- McDougall, I., Maier, R., Southerland-Hawkes, P., and Gleadow, A.J.W 1980 K-Ar age estimate for
the KBS tuff, East Turkana, Kenya. Nature 284: 230-234.
·aces of early hominid McDougall, I., Polach, H.A., and Stipp,].]. 1969 Excess pdiogenic argon in young subaerial basalts
from the Auckland volcanic field, New Zealand. Geochemica et Cosmochimica Acta 33:
ous-Tertiary boundary 1485-1520.
Megrue, G.H. 1973 Spatial distribution of 40Art39Ar ages in lunar breccia 14301. Journal of
lean hominids. Science
Geophysical Research 78: 3216-3221.
Merrihue, C.M. 1965 Trace-element determinations and potassium-argon dating by mass spectros-
enus Australopithecus
copy of neutron-irradiated samples. Transactions of the Ame1ican Geophysical Union 46: 125.
lia 28: 1-14.
Merrihue, C.M., and Turner, G. 1966 Potassium-argon dating by activation with fast neutrons.
D yr B.P. by marine-as-
Journal of Geophysical Research 71: 2852-2857.
Mitchell, JG. 1968 The argon- 40/argon- 39 method for potassium-argon age determination. Geo-
;, Seward and Baldwin
chimica et Cosmochimica Acta 32: 781-790.
urvey Open-File Report
Nier, A.O. 1950 A redetermination of the relative abundances of the isotopes oLcarb.on, nitrogen,
nsula. In Hamilton, T., oxygen, argon, and potassium. Physical Review 77: 789-793.
0 39
:ic Record. Anchorage, Pringle, M.S., McWilliams, M., Houghton, B.F., Lanphere, M.A., and Wilson, C.N.]. 1992 "' Ar/ Ar
dating of Quaternary feldspar: Examples from the Taupo Volcanic Zone, New Zealand.
vliddle Pleistocen.e age Geology 20: 531-534.
irch.36: 277-293. Pringle, M.S., Staudigel, H., and Gee, J. 1991 Jasper Seamount: Seven million years of volcanism.
1er new discoveries of Geology 19: 364-368.
Renne, P.R., Walter, R.C., Verosub, KL., Sweitzer, M., and Aronson, j.L. 1993 New data from the
rlying Bed I, Olduvai. Hadar Formation (Ethiopia) support orbitally tuned time scale to 3.3 Ma. Geophysical
Research Letters 20: 1067-1070.
ai Gorge, Tanganyika. Reynolds, J. 1956 High sensitivity mass spectrometer for noble gas analysis. Review of Scientific
Instruments 27: 928-934.
Ilion-year-old hominid Sarna-Wojcicki, A.M., Meyer, C.E., Roth, P.H., and Brown, F.H. 1985 Ages of tuff beds at East
African early hominid sites and sediments in the Gulf of Aden. Nature 313: 306-308.
 126 ROBERT C. WALTER

-Shackleton, N.j., Berger, A., and Peltier, W.R. 1990 An alternative astronomical calibration of the
lower Pleistocene timescale based on ODP Site 667. Transactions of the Royal Society of
Edinburgh 81: 251-261.
Shaeffer, 0.A., and Zaringer,j. 1966 Potassium argon dating. New York Springer-Verlag.
Smith, P.E., Evensen, N.M., and York, D. 1993 First successful 40Ar- 39Ar dating of glauconies: Argon
recoil in single grains of cryptocrystalline material. Geology 21: 41-44.
Steiger, R.H., and Jager, E. 1977 Subcommission on Geochronology: Convention on the use of
decay constants in geochronology and cosmochronology. Earth and Planetary Science Letters
36: 359-362.
Straus, W.L., and Hunt, C.B. 1962 Age of Zinjanthropus. Science, 136: 293-295.
van den Bogaard, P. 1995 40Ar/39 Ar ages of sanidine phenocrysts from the Laacher See Tephra
(12,900 yr BP): Chronostratigraphic and petrological significance. Earth and Planetary
Science Letters 133: 163-174.
Fis~
van den Bogaard, P., Hall, C.M., Schminke, H.-U., and York, D. 1987 40Ari39Ar Laser Dating of
Single Grains: Ages of Quaternary Tephra from the East Eifel Volcanic Field, FRG.
Geophysical Research Letters 14: 1211-1214. joHN WESTG
von Koenigswald, G.H.R., Gentner, W., and Lippolt, H.J. 1961 Age of the basalt flow at Olduvai,
East Africa. Nature 192: 720-721.
von Weiszacker, C.F. 193 7 Ober die Moglichkeit eines dualen-Zerfalls Kalium. Physih Zeitsch1ift
38: 623-624. ABSTRACT
Walter, R.C. 1981 The Volcanic History of the Hadar Early Man Site and the Surrounding Afar
Region of Ethiopia. Ph.D. disseration, Case Western Reserve University.
Fission tracks are zo
_ _ 1994 Age of Lucy and the First Family: Laser 40Ar/39 Ar Dating of the Denen Dora Member
of the Hadar Formation. Geology 22: 6-10. fragments through a :
Walter, R.C., and Aronson,j.L. 1982 Revisions of K/Ar ages for the Hadar hominid sit~, Ethiopia. a known rate, the ag1
Nature 296: 122-127. /, of uranium and num
Walter, R.C., and Aronson, j.L. 1993 Age and source of the Sidi Hakoma Tuff, Hadar Formation, glass are the most 5
Ethiopia.journal of Human Evolution 25: 229-240.
Quaternary deposits
Walter, R.C., Deino, A., Renne, P., and Tauxe, L. l 992a Refining the Plio-Pleistocene GPTS using
laser-fusion 40Ari39Ar tephrochronology: case studies from the East African Rift. Eos 73: as the most desirable
629. track retention proi:
Walter, R.C., Manega, P.C., and Hay, R.L. 1992b Tephrochronology of Bed I, Olduvai gorge: An procedures for parti
Application of Laser-fusion 40Ar/39Ar Dating to Calibrating Biological and Climatic Change. considerably improv1
Quaternary International 13/14: 3 7-46.
Walter, R.C., Manega, P.C., Hay, R.L., Drake, R.E., and Curtis, G.H. 1991 Laser-fusion 40Ari3°Ar
Accurate and p
dating of Bed I, Olduvai Gorge, Tanzania. Nature 354: 145-149. isothermal plateau fo
White, T.D., and Harris, J.M. 1977 Suid evolution and correlation of African hominid localities. track fading is achiev
Science 198: 13-21. for 30 days at 150°C.
White, T.D., Suwa, G., and Asfaw, B. 1994 Australopithecus ramidus, a new species of early hominid dating of fine-grainec
from Aramis, Ethiopia. Nature 371: 306-312.
White, T.D., Suwa, G., and Asfaw, B. 1995 Nature 375: 88. of detailed chronoloi
White, T.D., Suwa, G., Hart, W.K., Walter, R.C., WoldeGabriel, G., de Heinzelin, ]., Clark, j.D., from volcanic centrei
Asfaw, B., and Vrba, E. 1993 New discoveries of Australopithecus at Maka in Ethiopia. Nature Glass-ITPFT d:
366: 261-265. netostratigraphic tee:
WoldeGabriel, G., White, T.D., Suwa, G., Renne, P., de Heinzelin,J., Hart, W.K., and Heiken, G.
1994 Ecological and temporal placement of early Pliocene hominids at Aramis, Ethiopia.
JOHN WESTGATE • Ft
Nature 371: 330-333. MIC 1A4, Canada. AMA
York, D. 1969 Least squares fitting of a straight line with correlated errors. Earth and Planetary Amritsar, India. PHILIP
Science Letters 5: 320-324. New Zealand.
York, D., and Farquhar, R.M. 1972 The Earths Age and Geochronology. Oxford, Pergamon Press.
York, D., Hall, C.M., Yanese, Y., and Hanes,j.A. 1981 40Ar/39 Ar dating of terrestrial minerals with Chronometric Dating in An
a continuous laser. Geophysical Research Letters 8: 1136-1138. Plenum Press, New York,

View publication stats