Claire Patterson (1922–1995) was an American geochemist at Caltech ...
Prior to the 20th century, scientific estimates for the Earth's ...
Alfred Nier (1911–1994) was an American physicist who built high-pr...
To calculate the Earth's age, Arthur Holmes and Friedrich Houterman...
The strategy of the paper is as follows: **Move one** Date t...
## On Radiogenic Lead Patterson's methodology rests on one fact ...
This formula is the mathematical backbone of the lead-lead dating m...
Patterson is warning that the lead measurements are not equally cle...
Part of the criticism is that the Pb-Pb isochron may not date the o...
This point matters more for Earth than for meteorites. If Earth for...
Having pinned the age with lead, Patterson now brings in a second, ...
This is the third independent clock Patterson checks. The Rb-Sr met...
This is the fourth clock - and the one that fails. Helium dating w...
The **Canyon Diablo** iron meteorite (hit Arizona 50,000 years ago,...
Geochimica et Cosmochimica Acta, 1956, Yol. 10, pp. 236 to 23i.
P~~WIXXI Prers Ltd., London
Age of meteorites and the earth
CLAIRE PATTERSOX’
Division of Geological Sciences
California Institute of Technology, Pasadena. California
(Received 23 Janwrry 1956)
Abstract-Within experimental error,
meteorites have one age as determined by three independent
radiometric methods.
The most accurate method (Pb*07/Pb*06) gives an age of .i*BS T O,OT x. IO” yr.
Using certain assumptions which are apparently justified, one can define the isotopic evolution of lead
for any meteoritic body. It is found that earth lead meets the requirements of this definition. It is
therefore believed t,het the age for the earth is the same as for meteorites. This is the time since the earth
attained its present mass.
IT seems we now should admit that the age of the earth is known as accurately and
with about as much confidence as the ~oncentratio~l of aluIninium is known in the
Westerly, Rhode Island granite. Good estimates of the earth’s age hare been known
for some time. After the decay-constant of U235
and the isotopic compositions of
common earth-leads were determined by NIER, initial calculations. such as
GERLING’S, roughly defined the situation.
Approximately correct calculations
were made by HOLMES and by HOUTERMAWS 011 the basis of bold assumptions
concerning the genesis of lead ores.
Subsequent criticism of these calculations
created an air of doubt about anything concerning common leads and obscured the
indispensable contributio~ls which these investigators made in est~ablishiIlg the new
science of the geochemistry of lead isotopes. When the isotopic col~lpositio~l of
lead from an iron meteorite was determined, we were able to show that a much more
accurate calculation of the earth’s age could be made, but it still was impossible to
defend the computation. Now, we know the isotopic compositions of leads from
some stone meteorites and we can make an explicit and logical argument for the
computation which is valid and persuasive.
The most accurate age of meteorites is determined by- first, assuming that.
meteorites represent an array of nranium-lead systems with eert,ain properties,
and by then computing t’he age of this array from the observed lead pat,tern. The
most. accurate age of t,he earth is obtained by denlollst.ra~~i~~g that t,he earth’s
urallilln~-lead system belongs to the array of meteoritic uranium-lead syst,ems.*
The following assumptions are made concerning meteorit,es: they were formed
at, the same time: they existed as isolated and closed systems: they originally
contained lead of the same isotopic composition; they contain uranium which has
* C. PATTER~~K: S.R.C. Conference on nuclear processes in geologir ncttiitgs. I%.? Scptembcr
meeting, I’ennsylvnnia State University.
Except for minor disaprecments. this paper IS probably a
concrete expression of the attitudes of most investigators in this field. both hew and iii Europt~. The
author is grateful to his coih?agues, CLAYTON, INC~HRAX, TILTOX, \~.ws~lmc~rt:, ILHC~ \\~ETIIERILL, for
their critici&ns which helped ctarify this paper.
230
Age of meteorites and the earth
the same isotopic composition as that in the earth. On the basis of these assump-
tions various leads might be expected to evolve as a result of different original
U/Pb ,ratios in separate meteorites, and an expression* for any pair of leads derived
from such an array is:
4, - 4,
(eAIT
- 1)
R,, - RPb = k(e”zT - i)
(1)
where R, = Pb*07/Pb204 and R, = Pb206/Pb
204 for leads from different met,eorites
a and b, k = U23s/U
235 today (137.8), A1 = U23S decay-constant (9.72 x lo-lo yr-l),
A, = U238 decay-constant (1.537 x lo-lo yr-l), and T = age of the array.
The isotopic composit,ions of leads isolated from three stone and two iron meteor-
ites are listed in Table 1 (PATTERSON, 1955). Because the radiogenic and nonradio-
genie leads may occur in different mineral environments in a stone meteorite and the
sample dissolution procedures may be chemically selective, the lead ratios for the
first three meteorites in Table 1 have estimated errors from t,he absolute of about
- ,O. The lead ratios for the last two meteorites in Table 1 hare estimated errors
*, 0 :
from the absolute of about 1%.
Table 1. The isotopic compositions of lead in meteorites
/
I
Pb Composition
Meteorite
_~.
206/%04 207/204
208/‘304
Xuevo Laredo, Mexico
I 50.28
34.86 67.95
Forest City, Iowa
19.27
15.95 39.05
Modoc, Kansas
19.48
15.76 38.21
Henbury, Australia
9.55
10.38 29.54
Canyon Diablo, Arizona
9.46
10.34 29.44
These leads cover an extreme range in isotopic composition and satisfy
expression (1). yielding, within experimental error. a unique value of T. This is
illustrated in Fig. 1, where it is shown that the Pbzo6/Pbzo4 and Pbzo7/Pb204 ratios
from meteorite leads lie on a straight line whose slope corresponds to an age of
a.55 x lo9 yr. The dotted lines indicate how stone meteorite leads have evolved.
It is clear that, the assumpbions of the age method are justified by the data. Errors
in the lead data and in the decay-constants contribute about equally to the overall
error in the calculated age, which amounts to about 1 i 96. The age for the meteorite
array is calculated t,o be 4.+% _C 0.07 x lo9 yr.
The assumptions have not been shown t’o bc unique. The data can be explained
* Ai similar form of this expression was first wctl hy A.
SIER iI1 1!)3!!. F. HO~TERIANS has termed
the rspression an “isochron.”
IXeferences for the constant,s we’: (k) 31. INGHILUY; Vol. 14 Mallhattan
Project Tech. SC\-., Div. 2, Gaseous Diffusion Projrct. Chap. I-. p. 33 (l!M6);
-4. GHIORSO, nnd R. ~LTNNINGH.A51: ,%1/S. he. 88, tiJ:! (l!).i?).
(i,, 1,) E. FLEMING,
231
CLAIRE PATTERSOS
by other qualifying or even contradictory assumptions. Most of t,hese can be
excluded as improbable. One common criticism should be mentioned: the time of
a process of division or agglomeration of meteoritic material (without differen-
tiation) cannot be distinguished by this age method. It seems probable thal any
such process of division or agglomeration would be accompanied by chemical
differentiabion. Any meteorite which had a differentiat8ion historT after its initial
formation would fall off Dhe isochron. The fire meteorites in Table 1 represent a
most extreme range of differentiation which occurred during t’he initial process of
1
I
I
40- A = 4.6 x 10gyrs.
B = 4.5 x IO' yrs.
A
B
30-
x
%I
g zo-
%
n_
IO-
I I 1
0 IO
20 30
40 50
Fi:r. 1. The leatl isochron for meteorites and its estimated limits. The outline around
each point. indicates measumnent error.
formation. This criticism is not serious a.s far as meteorites are concerned. since if
it were valid the lead-lead isochron would date the occurrence of different,iation
processes :
however. it is important with respect to the age of the earth and will
be mentioned later.
At the present, t,ime, t)he nest most accurate meteorite age is determined by the
A40/Ii”o method. The argon ages of six stone meteorites, three of them determined
b- WASSERBI-RG and HATDES (1%X), and three of them clet~ermined by THOMSON
mcl MATSE (1035). are listed in Table 2.
The age of Forest City has been rrdrter-
mined without, change by RETSOLDS and LIPSOS (IS%). Two sets of ages are
calrnlated on the basis of the two reasonable limits of the c-//j- branching ratio.
The 0*0S5 branching ratio is the ra.lue obtained by studies of old pot.az;sium
minerals dated by uranium-lead techniques.
The 0.1% branching ratio is the
value obtained by counting techniques and 1,~ direct measurements of the amounts
of clecag products.
The difference between t,he two values can be accountecl for 1)~
systemat,ic loss of radiogenic argon in the old potassium minerals. If one assumes
_ige of meteorites and the earth
that a fixed amount of about LO
3 96 of radiogenic argon is lost, from all stone meteor-
ites, i.e. using a branching ratio of about O-10. then t,here is agreement of lead and
argon ages for the same st,one and an indication that the stones hare existed aa
cold and solid bodies since they were formed. Argon meteorite ages different from
the ones mentioned here have been reported by GERIXSG (10.51). and PAVLOVA
GERLIXG and RIK (1954). Since errors in the data present,ed by GEXLISG and
P_kv~ova cannot be eraluat,ed with anF certainty. we cannot be concerned by
differences bet,ween ages calculated by them and ages calculated from ot’her data.
Because of logarithmic behaviour, ralues for calculated ages of these old samples
are insensitive to changes in the e-l/?- branching ratio. For t,his reason only
disagreements of about 15 “/: between A40/I<4’J and Pb2’J’/Pb20G meteoritic ages can
be accounted for by a twofold change in the branching ratio.
Large age differences
must, therefore be reconciled on the basis of other experimental errors. Measure-
ments of the amounts of nonradiogenic argon in radiogenic and nonradiogenic argon
mixtures are subject t,o large uncertainties, and for t.he first four meteorites in
Table 2. AJo/K40 ages of meteorites
Meteorite
Age x lo-$
Investigators
(e-//3- = 0.085)
(e-/p- = 0.1%) 1
-
Beardsley, Kansas
Holbrook, Arizona
Forest City, Iowa
Akabu, Transjordan
Brenham Township,
Kansas
Monze, Sorthern
Rhodesia
4.8 4.2
4.8 4.'
4.7 4.1
4.4 3,s
4 3
1
I WASSERBURG and HAYDEN
~~.~SSERB~RG~~~~HA~DES
IKASSERBURG md HAYDEN
IREYNOLDS and LIPSOS
THonfsos and MAX-NE
THOMPSON; and ~\IAYXE
1,
Ta.ble 2, nonradiogenic argon corrections were small. For t,he last two meteorites
in Table 2, nonradiogenic argon corrections were eskemely large and t,he errors in
calculated age are excessive.
The isotope dilution det8ermination of potassium,
used by MrA4ss~~~u~~ and HAYDEN;? is nearlv an absolut,e method. while the flame-
photomet’ric determination of potassium. used by THOMSOS and K~YSE, requires a
nat,ural absolute standard which they did not use.
The age of meteorites 1ia.s been determined by bhe SrS’/Rb’J’ method. The
concentrations of rubidium and st,rontium and the isotopic compositions of
strontium have been determined in two stone meteorites by SCHUMACHER (1955).
The Rb/Sr ratio in one stone was so low that any change in isotopic composition of
st’rontium due bo radioactivity would be within experimental error. The Rb/Sr
ratio in the other stone (Forest City. Iowa) was considerably higher and sufficient
233
CLAIRE PATTERSON
to cause a loo/ difference in t,he relative abundance of Sr8i when the isotopic
compositions of strontium from both stones were compared.
The value for the decay-constant of Rbsi Is in question at the present time.
Reported values range from 4.3 to 6-i x lOlo yr for the half-life. Part of the
difficulty in the counting techniques of measuring the half-life arises from the fact
that, the frequency of 8-s at the low end of the energy spectrum increases rapidly
with no appearance of a maximum.
Measurements of decay products in terrestrial
rubidium minerals dated by uranium-lead technique inrolx-e errors of open chemical
systems.
SCHUMACHER’S experiment probably constitutes an ideal case of the
geological measurement of the half-life of Rbsi. since the ages have been det’ermined
by lead methods and the possibility of open chemical systems are remot’e. His
methods of measurement are at least as accurate as the radiometric methods.
One would t’herefore use his data to calculate the half-life of Rbs7. using the Pb-Ph
isochron age of meteorites. The half-life of Rbsi, as determined by these data. is
5-l j; lOlo yr, and is probably the most reliable value at, present. The half-life
determined by the geological method on terrestrial minerals (5.0 x lOlo yr) agrees
well with this.*
Because of t’he overwhelming abundance of nonradiogenic helium in iron
met,eorites and the large errors associated with the determinabion of the concen-
trations of uranium and thorium in iron and stone met8eorites. the age of meteorites
by the helium method is not accurate to much better than an order of magnitude
(PAKETH ef al., 1953; DALTON et al., 1953). It has been reported that iron meteorites
and the metal phases of stone meteorites were outgassed of helium as of about
5 >: lOa yr ago, while the silicate phases of stone meteorites were not, (REASBECR
and M/lsun-E, 195.5). Such an event would be highly significant and would require
detailed evolutionary theory for meteorites. Recent neutron activation (REED and
TURKEVICH) and nuclear emulsion (PICCIOTTO) analyses of iron meteorites show
that the concentrations of uranium in t,hese bodies are very low. and that’ the
uranium concent,rations used for helium a,ge calculations of iron met8eorites may be
erroneously high.
The question is unresolved at present. but it seems reasonable
to believe that, invest,igations of meteoritic helium will become vitally important to
cosmic-ray studies and may be decisive in meteorite evolution theory, but cannot be
used for accurate meteorite-age calculations at the present time.
The Canyon Diablo lead listed in Table 1 was isolated from troilite where the
U238/Pb204
ratio was shown by direct analysis to be 0.025 (PATTERSO?; ef al.. 1953).
This ratio is accurate to at least an order of magnitude. and it is so small that no
observable change in the isot,opic composition of lead could hare resulted from
radioactive decay after the meteorite was formed. Since stone meteorites were
cold and solid during their lifetime, it is unliliely that lead transport could have
occurred between iron and st,one meteoritic phases if they existed in one body.
This iron-meteorite lead is therefore primordial and represents the isotopic com-
position of primordial lead at the time meteorit’es were formed. Using the isotopic
composition of primordial lead and the age of meteorites. expressions can be
* 4 value rerommendecl by the work of the geochronology laboratory at the Dept. of Terrestrial
Magnetism, Carnegie Institute of Washington.
234
Age of meteorites and the earth
writt,en for a representative lead which is derived today from any system belonging
to the met’eoritic array:
Pb2”6/Pb2”4 = 9.50 + l-014 U238/Pb204
(4
Pb207/Pb204 = 10.36 + O-601 U23s/Pb204
(3)
If any two of the three ratios above can be independently measured in the earth’s
uranium-lead system, and they satisfy expressions (2) and (3). then this system
belongs to the meteoritic array and must have it’s age. Two of t,he ratios can be
measured in a sample of earth lead. but the problem of choosing such a sample is
complex because the ratio of uranium to lead varies widely in different rocks and
minerals whose ages are short compared to t’he age of the earth.
One approach is to partition the earth’s crust int,o separate chemical syst’ems
of uranium and lead and consider t’heir interactlions.
Such systems may range from
minerals t’o geochemical cycles.
Nearly all of the lead-isotope data concerns either
minerals in which the uranium-to-lead ratio is very high (uraninites, et,c.) or
minerals in which this ratio is essentially zero (galenas). The approsimat,e times of
formation of some galenas have been determined! and of these. two dozen or so
lately formed galenas may be used as a measure of earth lead (Suclea)
Qeobogy (lg.%). IV. Faul, Ed.). The isotopic compositions of lead in some recent
oceanic sediments have also been determined (PATTERSOK, GOLDBERG. and
IXGHRAM 1953), and these may be used as a measure of earth lead.
Any of these samples will be improper or biased if they are derived from a
system of uranium and lead which is only partially closed and is subject to slow
but appreciable transport from other systems with different I_?238/Pb204 ratios.
In this respect, the sample which may represent the system of largest mass is
probably t’he more reliable.
One sample of oceanic sediment lead probably
represents more material than a dozen galenas. The isotopic composition of t,his
sediment lead is Pb206/Pb204 = 19-O and Pb207/Pb20f = 15-5, which satisfies
expressions (2) and (3) surprisingly well. It is doubtful if these figures are grossly
biased, since a few measurements of uranium and the isotopes of lead in rocks
with widelv different
-CT238 Pb204
I
ratios indicate rather good mixing t,o be t,he
first-order effect on the isotopic composition of lead in the earth’s crust (PATTERSOS:
TILTOX, and ISGHRAM, 1955).
Independent of the absolute abundances of lead isotopes, a rough measure of
t’he rates of change of the lead-isotope abundances in the earth’s crust may be
obt,ained from the isotopic composit’ion of galenas of different ages. These rates
of change are defined by the ratios of uranium and thorium to lead in the material
from which the galenas are derir’ed. From t,he observed rate of change of Pb206,
the U23s/Pb204
ratio in the earth’s crust is found to be 10 (COLLINS, RUSSELL, and
FARQCHAR? 1953). This value satisfies expression (2) and (3) for sedimentary lead
with unexpectedly good agreement.
In Fig. 2
it is shown that oceanic sediment lead (open circle) falls on the
meteoritic lead isochron.
Most of the lately formed galenas fall within the dotted
outline, although a few are widely aberrant.
isochron is determined by t(he U23S/Pb204
The position of a lead along the
ratio in the system from which the lead
235
CLAIRE PATTERSON
ev elves .
The arrow indicates the position on the isochron which sediment lead
should occupy as predicted by the isotopic evolution of dat’ed ore leads. Inde-
pendently measured values for all tShree ratios adequa,tely sat,isfy espressions (2)
and (3), and therefore the time since the eart.h attained its present mass is
455 f 0.07 x log yr.
Fig. 2. The relationship between COM~OI~ earth lea& and the
meteoritic lead isochn3u.
If the earth is a late agglomeration without differentiation of meteoritic material
then it, can have ally age less than meteorit,ic material. Rather thaxi arguiug that
such a process would be accompanied by chemical differeilt,iatioI~ (aud a change of
the U/Pb ratio), it seems reasonable to beliere instead that such a late agglomera-
tion process would be less probable than one xvltere both n:eteorit,es and the
eart,h were formed at the same time. It is a fact that. estreme chemical differen-
t’iation occurred during the process which led to the mechauical isolat,ion of the mass
of material of which the earth is made, and since changes in this mass were accom-
panied by chemical differel~tiatioI1,
the Pb/Pb met,eorite isochron age properly
refers to the time since t.he earth attained it,s present, mass.
REFERESCES
Age of meteorites and the earth
PATTERSOX C., TILTOS, G., and IXGHRAM 11. (1955) Science 121, 69.
PICCIOTTO E. Suclear Physics Centre, Universuy of Brussells, manuscript.
REYXOLDS J. and LIPSOS J. Epipoleological Societ,y, spring 1955 meeting, U.C.L.A.
REASBECK P. and JI~YxE H. (1955) Satwe 176, 186.
REED G. and TURKEYITCH A. Inst. Kuciear Studies, Universitp of Chicago, manuscript.
SCH~-M.~CHER E. S.R.C. conference on nuclear processes in Geologic Settings (communicated b>-
H. UREY and JI. INGHRAV), 1955 Sept. meeting, Penn. State University. (Copies of his manu-
script are available).
THOMPSON S. and 3I_4YxE E. (1955) Geociiim et Cosmocllim. Acta 7, 169
VT.~SSERB~RG G. and Brly~~s R. (1955) Phys. Rev. 97, S6.
IVASSERBURG G. S.R.C. Conference on nuclear processes in Geologic Settings. 1953 Sept.
meeting, Penn. State UniversitS.
Swlecrr (:coZogy (1951) H. Faul, Ed.. J. Wiley, S.T.

Discussion

This point matters more for Earth than for meteorites. If Earth formed by a late, simple agglomeration of older meteoritic material, then the meteorite isochron would date the material, not Earth itself. Patterson therefore has to show that Earth’s lead system belongs to the same meteoritic array. If Earth lead falls on the same isochron, the age is not just the age of meteorites; it is the time since Earth reached its present mass and began evolving as its own uranium-lead system. This formula is the mathematical backbone of the lead-lead dating method. It allows scientists to calculate the age of a system ($T$) without needing to know the absolute amount of uranium originally present in the samples. The left side of the equation, $\frac{R_{1a} - R_{1b}}{R_{2a} - R_{2b}}$, calculates the slope of a line connecting the measured lead isotope ratios of two different meteorites, designated $a$ and $b$. The right side of the equation represents the theoretical slope derived purely from the physics of radioactive decay. The variables interact as follows: - The terms $(e^{\lambda_1 T} - 1)$ and $(e^{\lambda_2 T} - 1)$ define the accumulation of radiogenic $Pb^{207}$ and $Pb^{206}$ over time $T$, governed by the known decay constants of their respective uranium parents ($\lambda_1$ and $\lambda_2$). - The constant $k$ adjusts for the present-day relative abundance of $U^{238}$ compared to $U^{235}$. The brilliance of this expression is that the variable for total uranium concentration entirely cancels out. Even if separate meteorites started with vastly different amounts of uranium, as long as they formed at the exact same time from a shared primordial lead pool, their current lead ratios will predictably plot along a straight line known as an isochron. The steepness of this line is dictated exclusively by the time elapsed, $T$. This is the third independent clock Patterson checks. The Rb-Sr method uses \(^{87}Rb \to ^{87}Sr\): a meteorite with enough rubidium should accumulate extra \(^{87}Sr\) over time. One stone had too little rubidium for the change to stand out; Forest City had enough to show about a \(10\%\) shift. The weak point in 1956 was the decay rate of \(^{87}Rb\), which was still uncertain because the decay is extremely slow. Patterson therefore uses the Pb-Pb meteorite age to help calibrate it. The \(^{87}Rb\) half-life was later standardized much better; by 1977 geochronologists commonly used about \(4.88 \times 10^{10}\) years. Alfred Nier (1911–1994) was an American physicist who built high-precision mass spectrometers in the 1930s. In papers published between 1938 and 1941, he measured (1) the isotopic composition of common terrestrial lead - the “ordinary” lead found on earth, which contains both primordial lead and the small amounts of radiogenic lead that have accumulated over geologic time—and (2) the precise natural abundance ratio of the two uranium isotopes $ ^{238}\mathrm{U}/^{235}\mathrm{U} $. His work also yielded improved values for the decay constant of $ ^{235}\mathrm{U} $. This data supplied the essential starting isotopic ratios and decay rates that Gerling, Holmes, and Houtermans needed in the mid-1940s to perform the first quantitative calculations of Earth’s age from the evolution of lead isotopes in terrestrial samples. Patterson is warning that the lead measurements are not equally clean for every meteorite. In stone meteorites, radiogenic lead, made by uranium decay, and non-radiogenic or primordial lead may sit in different minerals. If the lab dissolves some minerals more easily than others, the measured lead mixture can be slightly biased. That is why the first three meteorites in Table 1 have larger estimated errors, about \(2\%\). The last two are iron meteorites, where the relevant lead was easier to isolate more consistently, so their errors are estimated at about \(1\%\). This matters because the isotope ratios are used to draw the meteorite isochron; measurement error affects how confidently the age-line can be fitted. Having pinned the age with lead, Patterson now brings in a second, independent clock as a cross-check: the potassium-argon method ($\text{K}^{40} \rightarrow \text{Ar}^{40}$). If a completely different element and decay system gives the same answer, the 4.55-billion-year result is far harder to dismiss. The complication is that K⁴⁰ is unusual - it decays two different ways. Most of it beta-decays to calcium-40, but a minority converts to argon-40 (by electron capture). Only the argon branch is useful for dating here, so you need to know what fraction of decays produce argon. That fraction is the branching ratio ($e^-/\beta^-$ in the text), and at the time it wasn't well measured - different techniques disagreed. That's why Patterson reports two sets of ages in Table 2 rather than one: he calculates each meteorite's age at both reasonable limits of the branching ratio, bracketing the uncertainty instead of pretending it away. The ratio was only firmly nailed down later, with refined measurements through the 1960s and the standardized decay constants adopted by convention in 1977. ## On Radiogenic Lead Patterson's methodology rests on one fact of nuclear physics: uranium slowly turns into lead. Two different uranium isotopes decay into two different lead isotopes, each at its own fixed rate: $$ {}^{238}\mathrm{U} \rightarrow {}^{206}\mathrm{Pb} \qquad (\text{half-life} \approx 4.5\ \text{billion yr}) $$ $$ {}^{235}\mathrm{U} \rightarrow {}^{207}\mathrm{Pb} \qquad (\text{half-life} \approx 0.70\ \text{billion yr}) $$ Each is a long cascade through unstable intermediates (radium, radon, polonium, and others), but the endpoints are stable lead, and only the start and finish matter for dating. Lead produced this way is called radiogenic; it was manufactured by radioactive decay rather than being present from the beginning. An important feature of radioactive decay is that its rate is essentially unaffected by ordinary physical conditions. Temperature, pressure, chemical state, and geological processes can move uranium and lead around, but they do not measurably alter the rate at which uranium nuclei decay. A uranium atom buried deep inside the Earth and one floating in space decay according to the same nuclear clock. This remarkable stability is what makes radioactive isotopes reliable chronometers over billions of years. The key is that the two clocks run at very different speeds. $^{235}\mathrm{U}$ decays roughly seven times faster than $^{238}\mathrm{U}$, so in any uranium-bearing system the amount of $^{207}\mathrm{Pb}$ relative to $^{206}\mathrm{Pb}$ changes steadily over time. That changing ratio is the clock, and it is why Patterson can read an age from the slope of his isochron without knowing how much uranium anything started with. To measure this, you need a fixed reference, and that is $^{204}\mathrm{Pb}$, the one lead isotope that is not produced by any uranium decay chain. Its quantity in a sample never changes. So Patterson expresses everything as ratios to $^{204}\mathrm{Pb}$ ($^{206}\mathrm{Pb}/^{204}\mathrm{Pb}$ and $^{207}\mathrm{Pb}/^{204}\mathrm{Pb}$): the denominator is a frozen yardstick, and any growth in the numerator records radiogenic lead accumulating over time. One more piece completes the picture. When the solar system formed, it already contained some primordial lead, a starting mix of all four lead isotopes inherited from earlier generations of stars. The total lead in any sample today is therefore primordial plus radiogenic. Patterson's whole argument is about separating those two contributions. The **Canyon Diablo** iron meteorite (hit Arizona 50,000 years ago, fragments discovered in 1891) gave Patterson exactly what he needed: a sample of **primordial lead**. Lead extracted from its troilite (FeS) nodules has an extremely low \(\frac{\mathrm{U}^{238}}{\mathrm{Pb}^{204}}\) ratio of only 0.025. With so little uranium present, essentially no radiogenic \(^{206}\mathrm{Pb}\) or \(^{207}\mathrm{Pb}\) has been added since the meteorite formed. This troilite lead therefore preserves the original (“initial”) isotopic composition of lead at the time the solar system formed—the crucial starting point for the isochron (Fig. 1) and the Earth-lead evolution equations later in the paper. Fragments of Canyon Diablo can still be seen at places like the American Museum of Natural History in New York and UCLA’s Meteorite Museum. ![](https://i.imgur.com/XX0ye5h.png) *Fragment of Canon Diablo Meteorite* The strategy of the paper is as follows: **Move one** Date the meteorites. Patterson doesn't date a single meteorite. He treats the meteorites as a family of independent uranium-lead clocks that all started at the same instant with the same starting lead, but with different amounts of uranium. Over billions of years, the uranium-rich ones brewed more radiogenic lead than the uranium-poor ones. Plotted against each other, their lead compositions fall on a straight line - the **isochron** - and the slope of that line depends only on elapsed time. No need to know how much uranium any individual meteorite started with; the pattern across the array gives the age. **Move two** Show the Earth is part of the family. This is the clever part. Patterson never directly dates an Earth rock - the crust's lead has been reshuffled too many times for that to work. Instead he shows that a sample of Earth's lead falls on the same isochron line as the meteorites. If it sits on the line, the Earth's uranium-lead system "belongs to the array" - it shares the meteorites' common origin and therefore their age. Membership in the family is the proof. To calculate the Earth's age, Arthur Holmes and Friedrich Houtermans assumed that the lead found in terrestrial ores evolved in a single, uniform environment. Their mathematical models relied on the premise that the source material for these ores maintained a constant $U/Pb$ ratio from the exact moment the Earth formed until the ore crystallized and locked the lead isotopes in place. Critics argued that the Earth's crust is far too dynamic for this assumption to hold true universally. Lead ores are frequently subjected to complex geological processes - such as tectonic subduction, regional melting, and fluid transport - which mix leads from different rock layers with drastically different $U/Pb$ ratios over billions of years. Because terrestrial rocks often have a "multi-stage" history, their isotopic compositions can be highly anomalous. Skeptics argued that this constant crustal recycling made it impossible to use common earth leads to calculate a definitive, globally accurate age for the planet. This skepticism temporarily discredited the whole field of lead geochronology. The brilliance of Patterson's paper lies in how it completely sidesteps this terrestrial mixing problem. By utilizing meteorites - which have existed as cold, isolated, and closed systems since the dawn of the solar system - Patterson was able to find an undisturbed, uncontaminated baseline for primordial lead, rendering the crustal mixing argument moot. Claire Patterson (1922–1995) was an American geochemist at Caltech who, with this 1956 paper, became the man who finally measured the age of the Earth - a number that has scarcely changed since. He'd been a young mass-spectrometer technician on the Manhattan Project, which is where he learned the isotope techniques he later turned on meteorites. His most important legacy, though, grew out of a problem he hit while doing this work. To measure such tiny quantities of lead, Patterson found that ordinary laboratories were hopelessly contaminated with it - lead was everywhere. Building the world's first ultra-clean lab to get honest numbers, he realized the contamination wasn't natural: it came from leaded gasoline and industrial pollution, and modern humans carried far more lead in their bodies than their ancestors. Patterson spent much of the rest of his life proving this and fighting the lead industry over it. His evidence drove the removal of lead from gasoline, food cans, paint, and water pipes - one of the great public-health victories of the 20th century. The same obsessive precision that let him date the planet also let him show that we were quietly poisoning ourselves. ![](https://i.imgur.com/z3qBq5Z.jpeg) *Claire Patterson* Part of the criticism is that the Pb-Pb isochron may not date the original formation of meteorite material. That is indeed true, technically the isochron doesn't actually time "formation" - it times the moment uranium and lead got chemically separated (a differentiation event), since that's what starts each meteorite's clock. The loophole: if meteoritic material were merely split apart or clumped together without changing its U/Pb ratios, the method would be blind to it. So strictly, the isochron dates the last differentiation, not necessarily the true age. Patterson’s reply is that this is unlikely. Real division or agglomeration of meteorite material would usually involve chemical sorting, melting, or mineral separation. That would change \(U/Pb\) ratios and push the points off the isochron. Since very different meteorites still fall on the same line, the line most likely dates their initial formation, not a later mechanical reshuffling. This is the fourth clock - and the one that fails. Helium dating works because uranium and thorium emit alpha particles (helium nuclei) as they decay, so radiogenic helium accumulates over time. Two problems wreck it for meteorites. First, iron meteorites are flooded with nonradiogenic helium - helium made not by decay but mostly by cosmic rays striking the metal in space. This swamps the small decay-produced signal you're trying to measure. Second, the method requires knowing exactly how much uranium and thorium produced that helium, and those concentrations are extremely hard to measure accurately. Together these push the error to an order of magnitude - basically useless for a precise age. The two new techniques named are ultra-sensitive ways to measure those tiny uranium amounts. - **Neutron Activation**: The meteorite sample is bombarded with neutrons in a nuclear reactor, making trace elements temporarily radioactive. Their exact concentrations can then be accurately measured via their emitted gamma rays. - **Nuclear Emulsion**: Specialized, thick photographic plates are used to capture the physical, microscopic tracks left by individual decaying alpha particles. This essentially allows researchers to count the radioactive atoms one by one. Both revealed that iron meteorites hold far less uranium than earlier helium calculations had assumed — meaning those old ages were built on inflated numbers. Patterson's verdict: helium is promising for studying cosmic rays, but can't yet date meteorites. Prior to the 20th century, scientific estimates for the Earth's age were drastically lower than modern values. In the 1860s, the Scottish physicist Lord Kelvin estimated the Earth was between 20 and 400 million years old based on the cooling rate of a presumed molten sphere. This figure famously frustrated geologists and evolutionary biologists, such as Charles Darwin, who required vastly more time to account for slow geological changes and natural selection. Kelvin's physics was careful but his premise was incomplete: he had no knowledge of radioactivity, which continuously reheats the interior. Radioactivity turned the problem into an atomic clock. Boltwood (1907) measured uranium-to-lead ratios in minerals and reached ages up to ~2.2 billion years. Arthur Holmes spent decades refining lead dating and by the 1940s had pushed the figure to ~3–4 billion years. The key conceptual step came with the Holmes–Houtermans model (1940s): rather than dating individual minerals, one could work backwards from the isotopic composition of common lead to its primordial starting point. That idea is the direct ancestor of the isochron Patterson builds his argument on here.