Sunteți pe pagina 1din 5

known toxic effects of salt spray on References and Notes pacted are calculated by multiplying the

the twigs that grow into the higher- 1. F. H. Borinann and G. Likens, Scienice 155, amount of Na by the K/Na, Ca/Na, and
velocity portions of the airstream flow- 424 (1967). Mg/Na ratios, 0.0435, 0.0519, and 0.139, re-
2. F. W. Clarke, U.S. Geol. Surv. Btull. 770 spectively.
ing off the ocean and over the second- (1924); B. Mason, Pritnciples of Geochemistry
14. The relation between cation exchange capacity
in milliequivalents per 100 g of soil (Y) and
ary dune (10). The growth response (Wiley, New York, 1966). organic matter expressed as percentage loss on
3. G. Likens and F. H. Bormann, in Ecosystem
of the forest trees simultaneously Structure and Function, J. A. Wiens, Ed. ignition (X) is given by Y = 1.51X-0.24
(Oregon State Univ. Press, Corvalis, 1972), (multiple correlation coefficient, .97; F value
adapts the ecosystem to the wind-salt pp. 25-67.
of the regression, 300.5; 1 and 22 degrees of
conditions and limits further biomass 4. G. M. Woodwell and R. H. Whittaker, in freedom; significant at P < 0.1 percent). The
relation between total exchangeable K + Na +
accumulation resulting from upward Symposium on Pritary Productivity and Min- Ca + Mg in milligrams per kilogram of soil
eral Cycling in Natuiral Ecosystems, H. E.
growth. However, proliferation of Young, Ed. (Univ. of Maine Press, Orono, (Y) and organic matter expressed as percentage
loss on ignition (X) is given by Y = 67.44X -
1967), p. 151.
branches at the airstream-canopy inter- 5. H. W. Art, thesis, Yale University (1971). 19.56 (multiple correlation coefficient, .88; F
valLie of the regression, 72.8; 1 and 22 degrees
face guarantees a large surface area 6. F. Cowan, personal communication. Cowan
also reported the complete absence of feldspars of freedom; significant at P < 0.1 percent).
that probably strains aerosols from and micas in a mineralogical analysis of soils 15. C. 0. Tamm, Physiol. Plant. 4, 461 (1951);
from the Sunken Forest. H. L. Mitchell, Black Rock For. Pap. 1 (No.
slower, nontoxic portions of the air- 7. H. C. Crandell, U.S. Geol. Sutrv. Water Supply 6), 30 (1936).
16. P. W. Richards, The Tropical Rain Forest
stream. Thus, there are continual ad- Pap. 1539-X (1962). (Cambridge Univ. Press, Cambridge, 1964); I.
8. W. B. Harris, Water Resour. Res. 3. 89 (1967).
justments between the restriction of the 9. R. H. Whittaker, Ecology 42, 171 (1961); ibid. Langdale-Brown, in The Soil Resouirces of
system by toxic effects of salt spray 43, 357 (1962); ibid. 44, 176 (1963); ibid. 46, Tropical Afriea, R. P. Moss, Ed. (Cambridge
365 (1965); ibid. 47, 103 (1966); and Press, Cambridge, 1968), p. 61.
and dependence of the system on salt G. M. Woodwell, Am. J. Bot. 54, 931 (1967); 17. H. W. Art and P. L. Marks, in Forest Bio-
mass Stuidies, H. E. Young, Ed. (Univ. of
spray as a source of nutrients. J. Ecol. 56, 1 (1968); ibid. 57, 155 (1968); G.
Likens and F. H. Bormann, Yale Univ. Sch. Maine Press, Orono, 1971), p. 1.
The rates of development, biomass For. Bull. 79 (1970); 18. H. C. Cowles, Bot. Gaz. 27, 95, 167, 281, 361
, N. M. Johnson, (1899); J. S. Olson, ibid. 119, 125 (1958).
accumulation, and net primary pro- R. S. Pierce, Ecology 48, 772 (1967). Cation 19. L. A. Sirkin, Bull. Torrey Bot. Cltub. 94, 131
analysis was performed by atomic absorption (1972).
duction approach the means for tenm- spectrophotometry. 20. A. T. Wilson, Natuire (Lond.) 184, 99 (1959).
10. S. G. Boyce, Ecol. Monogr. 24, 29 (1954).
perate forests (5). Dimension analyses 11. E. Eriksson, Tellus 7, 243 (1955); F. H. T. 21. D. W. Cole, S. P. Gessel, S. F. Dice, in Sym-
(17) indicated a biomass of 17,000 Juang and N. M. Johnson, J. Geophys. Res. posium on Primary Productiv ity anid Mineral
72, 5641 (1967); E. J. Conway, Proc. R. Ir. Cycling in Natural Ecosystems, H. E. Young,
g/m2, a net primary production of Acad. Sect. B 48, 119 (1942). Ed. (Univ. of Maine Press, Orono, 1967),
p. 198.
1110 g/m2, and a leaf area index of 12. R. B. Miller, N.Z. J. Sci. 6, 388 (1963). 22. Supported by the Fire Island National Sea-
13. The relation between kilogramiis Na impacted
5.9, all of which fall well within the per hectare (Y) and thousands of kilometers shore, National Park Service, Brookhaven Na-
of wind blowing onshore (between 79' and tional Laboratory, and Yale University. We
range for temperate forests. This is in 238°) at speeds of < 7 (X1) and > 7 (X,,) thank the U.S. Coast Guard, the U.S. Geo-
part because nutrients from meteoro- meter/sec is given by Y = 2.49X, + 3.57X.,- logical Survey, B. Bormann, P. Bernacki, K.
0.264 (multiple correlation coefficient, .84; F K. Turekian, W. Oakley, and T. G. Siccama
logic sources alone are about equiva- value of the regression, 39.4; 2 and 34 de- for assistance during the study, and P. L.
Marks for helpftul comments on the imianuscript.
lent to nutrients from meteorologic grees of freedom; significant at P <0.1 per-
cent). The amounts of K, Ca, and Mg im- 10 December 1973 a
plus weathering sources in a variety
of other ecosystems (Table 2).
The development of an ecosystem
may be extremely slow if low weather-
ing rates are coupled with low rates of Radiohalos in a Radiochronological and Cosmological Perspective
meteorologic input. These factors may
be important along Lake Michigan, Abstract. New photographic evidence, data on halo ring sizes, and x-ray fluo-
where fresh water dune ecosystems rescence analyses provide unambiguolus evidence that polonium halos exist as a
take thousands of years to reach the separate and distinct class apart fromn uraniumn halos. Because of the short half-
tree stage (18). Radiocarbon dates and lives of the polonium isotopes involved, it is not clear how polonium halos may
geological evidence suggest that the be explained on the basis of currenitly accepted cosmnological m1odels of Earth
Sunken Forest has reached a compara- formlation.
ble stage of development in only 200
to 300 years (19). I have examined some 105 or more halos (2) or "actinium" halos (8) to
Our results also suggest that the high radiohalos, mainly from Precambrian account for them. (Figure 1, a to d, is
levels of biomass accumulation and granites and pegmatites located in sev- a schematic comparison of U and Po
primary production found in some eral continents. In addition to U and halo types with ring radii drawn pro-
coastal regions (5, 17, 20), may be in Th halos, originally studied (1, 2) for portional to the respective ranges of
part due to high weathering inputs plus information on the constancy of the a-particles in air.) To explain Po halos,
high meteorologic inputs. a-decay energy Ea and the decay con- Henderson (7) postulated a slow ac-
HENRY W. ART stant A, I have discussed X halos (2, cumulation of Po isotopes (or their
Biology Department and Center for 3), dwarf halos (3), and giant halos respective /3-decay precursors) from U
Environmental Studies, Williamns (4), and explained how these remain daughter product activity. I demon-
College, Williamstown, prime candidates for identifying un- strated that this secondary accumula-
Massachutsetts 01267 known a-radioactivity and, not impos- tion hypothesis was untenable and
F. HERBERT BORMANN sibly, unknown elements as well. showed, using the ion microprobe (3),
GARTH K. VOIGT I have also reported (5) on a class that Po halo radiocenters (or inclu-
School of Forestry and Environmental of halos which had been tentatively sions) exhibit anomalously high 206Pb/
Studies, Yale University, New Haven, attributed (6, 7) to the a-decay of '07Pb isotope ratios which are a neces-
Connecticut 06511 210po, 214Po, and 218Po. Earlier inves- sary consequence of Po a-decay to
GEORGE M. WOODWEI.L tigators (2, 7-10), possessing only a 206Pb
Biology Department, Brookhaven sparse collection of Po halos, at times Recently, these ion microprobe re-
National Laboratory, Upton, confused them with U halos or invented sults have been questioned, Henderson's
New York 11973 spurious types suLch as "emanation" results misinterpreted, Po halos con-
sidered to be only U halos, and allu- as well as other distinguishing charac- counted for in other comparative meth-
sions made to the geological difficulties teristics need to be made abundantly ods.
that Po halos would present if they clear. I made more than 350 irradiations
were real (11) [see (12) for com- In order to ascertain the E. corre- 1 to I04 seconds in duration using 4He
ments]. sponding to a specific halo radius, I ions with energies ranging from 1 to
Admittedly, compared to ordinary have produced a new series of standard 15 Mev, on over 40 samples of biotite,
Pb types, the Pb isotope ratios of sizes against which halo radii may be fluorite, and cordierite (14). Selecting
Po halos are unusual, but new ion mi- compared withouLt relying on estimates the band sizes which correspond to the
croprobe analyses have confirmed (13) derived from ranges of a-particles in energies of the 23'8U a-emitters (see
my earlier results (3). It is also appar- air. Standard sizes may be prepared Table I ) permits a direct comparison
ent that Po halos do pose contradic- by irradiation of halo-bearing mineral with new as well as previous (1, 9,
tions to currently held views of Earth samples with 4He ions (4); the colora- 10, 15) U halo measurements in bio-
history. tion bands thuLs produced show varying tite, fluorite, and cordierite. Figure 1 e
For example, there is first the prob- sizes (as measuLred from edge to colora- shows a coloration band in biotite pro-
lem of how isotopic separation of sev- tion extinction) which are dependent duLced by 7.7-Mev 'He ions, and Fig.
eral Po isotopes [or their a3-decay pre- on energy, total dose, and dose rate, 2a shows a densitometer profile of
cursors (13)] could have occurred the latter two factors not being ac- Fig. 1e.
naturally. Second, a straightforward
explanation of `11Po halos implies that
the 1 -[cm radiocenters of very dark halos
of this type initially contained as many
as 5 x 109 atoms (a concentration of
more than 50 percent) of the isotope
'18Po (half-life, 3 minutes), a problem 4
that almost defies reason. A further W.w ...
necessary consequence, that such Po r ..E :0

halos could have formed only if the

host rocks underwent a rapid crystalli-
zation, renders exceedingly difficult, in
my estimation, the prospect of explain-
ing these halos by physical laws as
presently understood. In brief, Po halos
atre an enigma, and their ring structure

Fig. 1. The scale for all photomicrographs

is I cm - 25.0 ,m, except for (h') and
(r'), which are enlargements of (h) and
(r). (a) Schematic draw ing of "U halo f
with radii proportional to ranges of a-
particles in air. (b) Schematic of "3Po
halo. (c) Schematic of -1-Po halo. (d)
Schematic of '9Po halo. (e) Coloration
band formed in mica by 7.7-Mev 4He
ions. Arrow shows direction of beam pene-
tration. (f) A 'U halo in biotite formed
by sequential a-decay of the 'U decay p
series. (g) Embryonic 'U halo in fluorite qAM,- :..VWi:;i
w ith only two rings developed. (h) Nor- ""I.1171-".--.
mally developed 'U halo in fluorite with 10-fth, Ilk
nearly all rings visible. (h') Same halo
as in (h) but at higher magnification. (i)
Well-developed -U halo in fluorite with
slightly blurred rings. (j) Overexposed 218Po Hlo 1 Halo 238U C.- EMITTERS
"'U halo in fluorite, showing inner ring Nuclide Ea(MeV) T,,2
diminution. (k) Two overexposed "U -238U 4.19 4.5x109y
halos in fluorite showsin- inner ring dimi- 234U 4.77 2.5x105y
230Th 4.68 7.6 x 104y
nution in one halo and obliteration of "1,2 = 3m f 8226Ro 4.78 1.6x103y
inner rings in the other. (1) More over- 218Po -222 Rn 5.49 3.8d
exposed "U halo in fllorite, showing d 2t8Po 6.00 3.Om
outer ring reversal effects. (im) Second- d a Po 5.30 138 d
stage reversal in a -U halo in fluorite.
The ring sizes are unrelated to `U a-
particle ranges. (n) Three "1'Po halos of
/ .234Th_. 234po Q
light, medium, aind very dark coloration in 238 u

biotite. Note the differences in radius. (o) 2t4 Po

20Po 234u 230Th a 2 26Ro a-
222 Rn a, 21Bpo a 2i4Ppb -
Three ""'Po halos of vary,ing degrees of 210Po - -
coloration in fluiorite. (p) A "1Po halo in T:2
, 22y
2 for
O B24 St 214Po-a
214g i-. 214p a 20Pb
$ 1P

biotite. (q) Two "Po halos in biotite. (r) r,T2 = 27rn for 2 Pb P
Tso "1'Po halos in fluorite. (r') Same 214Po Halo e 210po Halo I
halo as in (r) but at higher magnification.
5 APRIL 1974 63
00O The coloration extinction boundary
is poorly defined near threshold colora-
lishing the constancy of X to within
35 percent, and under certain assump-
tion; only a few very light bands in tions U halos provide no information
00 X~* .
lv ve*; z. o - z* p~*
biotite could be reliably measured. Re-
producible measurements were obtained
at all in this respect (16).
While halos with point-like nuclei
Xlr~ in the plateau region (14), where varia- which show well-defined, normally de-
0i tions in band size are minimal. Darker
halos in biotite generally have slightly
veloped rings (as in Fig. lh) can be
used to determine the Ea's of the radio-
larger radii than lighter halos (3, 4). nuclides in the inclusion, there are pit-
Also, reversal effects in some biotites falls in ascertaining what constitutes a
oo immediately exterior to the terminus of normally developed ring. In contrast to
- 00 a halo ring cause apparent diminution the easily recognizable U halos in fluo-
of the radius. Therefore, while there rite in Fig. 1, g to i, the overex-
las 0 are differences between the sizes of
posed fluorite U halo in Fig. lj shows
$2 d
medium coloration bands (Table 1, a diminutive ghost inner ring, which
column 2) and the radii of U halos could be mistaken for an actual 238U
Z~ Z z z Z Zm in biotite (Table 1, columns 8, 9, and ring. Figure 1k shows two other par-
V-4 10) that could be interpreted in terms tially reversed U halos, one of which
of an actual change in Ea and X (16), shows the diminutive inner ring, while
such differences more likely arise from in the other all the inner rings are ob-
r.2 a combination of dose and reversal
00 literated. The U halo in Fig. 11 is even
oR effects (15, 17), producing slightly dimin- more overexposed, and encroaching re-
0u - z - ze z ished radii. Diminution of U halo versal effects have given rise to another
Z, radii may also result from attenuation ghost ring just inside the periphery.
of a-particles within the small but Figure 1 m shows a still more over-
vo c ,
relatively dense zircon radiocenters.
-ven though slight differences between
exposed U halo; in which second-stage
reversal effects have produced spurious
band sizes and U halo radii do exist ghost rings that are unrelated to the
in biotite, the idealized U halo ring terminal a-particle ranges.
structure (Fig. la) compares very well Since this association of the halos in
with an actual U halo in biotite (Fig. Fig. 1, 1 and m, with U a-decay cannot
.- .0
t--" 4
.SD 0
-4 -.
if). be easily proved by ring structure anal-
Biotite and fluorite are good halo ysis alone, I have utilized electron-
0C oo 00 0 t C
detectors, but fluorite is superior be- induced x-ray fluorescence to confirm
0 -
-' P
cause the halo rings exhibit more de- this identification. Figure 3a shows the
ea I.O
6 oo Z oq tail, often have smaller radiocenter di- prominent Ca x-ray lines of the fluorite
C 4 "I ooi
06 _;
ameters (< 1 ncm), and have almost matrix (the F lines are below detection
0) negligible size variations due to dose threshold) along with some background
0 effects in the embryonic to normal Ag and Rh lines which are not from the
o7 J-4" -.4 , 4 Cl stages of development. Figure 1g shows sample, but are produced when back-
an embryonic U halo in fluorite with scattered electrons strike a Ag-Rh alloy
only the first two rings fully developed; pole piece in the sample chamber. Fig-
the other rings are barely visible be- ure 3b, the x-ray spectrum of a halo
^ t P * XlIo cause, due to the inverse square effect, radiocenter typical of the halos in Fig.

threshold coloration has not been 1, 1 and m, clearly shows the x-ray
reached. Figure lh shows a U halo in lines due to U (as well as a small
z fluorite in the normal stage of develop- amount of Si) in addition to the matrix
(Y ,4
ment, when nearly all the rings are and background peaks. A more detailed
'.6 visible. This halo closely approximates analysis (18) reveals that the Ug line
the idealized U halo in Fig. 1 a. Under masks a small amount of Pb probably
- z- en
N high magnification even separation of generated by in situ U decay.
the 210Po and 222Rn rings may be seen. The variety of U halos shown in
Figure I i shows another U halo in Fig. 1, g to m, establishes two points:
0 0 W fluorite, with a ring structure that is (i) only a thorough search will reveal
clearly visible but not adequate for ac- the numerous variations in appearance
curate radius measurements. of U halos, and (ii) unless such a
In Table 1, columns 4, 11, and 12, search is made, the existence of halos
. . .

2 .. 2 q0 --. C.en "t4 the fluorite band sizes agree very well originating with a-emitters other than
el; "'t .e 2. .. cq eq N
m with the U halo radii measured in this 238U or 232Th could easily be over-
mineral by myself and Schilling (9). looked.
This suggests that the differences be- So far, three criteria have been used
0 tween U halo radii and band sizes in to establish the identity of U halos:
biotite are not due to a change in Ea. (i) close resemblance of actual halos
. . . .
However, experimental uncertainties in in biotite (Fig. lf) and fluorite (Fig.
:z:; z;z:z:z
measuring U halo radii preclude estab- I h) to the idealized ring structure
64 SCIENCE, VOL. 184
Fig. 2. Densitometer profiles of the photo- 1 r) to the idealized 218Po halo (Fig.
graphic negatives of (a) Fig. le, (b) Fig. 1d) showing the ring structure from the
If, (c) the light 210Po halo in Fig. ln, (d) a I f
the medium zPo halo in Fig. In, (e) the sequential a-decay of 218po, 214Po, and
210po. In biotite such halos may ap-

dark ='Po halo in Fig. In, and (f) Fig. Ip.
pear very light to very dark with radii
correspondingly slightly lower and
(Fig. la), (ii) identification of lines in higher (excluding reversal effects) than
x-ray fluorescence spectra, and (iii) those measured for medium coloration
agreement between U halo radii and bands (compare Table 1, columns 2,
equivalent band sizes (very good in b 3, 18, and 19). Cursory examination
fluorite and fair in biotite and cordi- of inferior specimens of this halo type
erite). Using the third criterion (either I. could lead to confusion with the U
band sizes or U halo radii), I can de- halo, especially in biotite, where ring
termine Ea for a normally developed sizes vary slightly because of dose and
fluorite halo ring to within + 0.1 Mev. other effects. However, good specimens
For biotite halos, U halo radii may
form a suitable standard for determin-
.1' of this type are easily distinguished
from U halos, even in biotite. In fluo-
ing Ea for rings that show reversal or rite, where the ring detail is better, a
other effects characteristic of U halos most important difference between 238U
in the same sample. If good U halos C e
and 218Po halos is delineated, that is,
are not available, and if the halos with the presence of the 222Rn ring in the U
variant sizes show well-developed rings halo (Fig. 1 a) in contrast to its ab-
without reversal effects, then the band sence in the 218Po halo (Fig. ld). For
sizes form a suitable standard for Ea example, note the slightly wider an-
determination when coloration intensi- nulus (3.9 rim) between the 210Po and
ties of variant halos and band sizes are 218Po rings of the 218Po halo compared
matched. to the equivalent annulus (3.0 [cm) in
Therefore, if halos result from the spondence of the radii with band sizes. the '-'38U halo (Fig. 1, a, d, h, h', r, and
a-decay of 210po to 206Pb, their appear- The prominent unmistakable feature of r'). This is evidence that the 21'Po halo
ance should resemble the idealized sche- the 214Po halo is the broad annulus indeed initiated with 218Po rather than
matic (Fig. ib), and the light and separating the inner and outer rings with 222Rn or any other a-decay pre-
dark halos of this type in biotite should [see the densitometer profile of Fig. Ip cursor in the U chain. As further proof,
exhibit radius variations consistent with shown in Fig. 2f and figures 7 to 9 Table 1 (columns 4, 11, 12, and 21)
the differences between lower and in (6)]. With respect to comments in shows that the 218Po halo radii agree
higher coloration band sizes (Table 1, (11) it should be noted that the 214po very well with equivalent band sizes
columns 2, 3, 6, 14, and 15). Further, halo can easily be distinguished from and U halo radii in this mineral. Addi-
such halos, whether very light or very a U halo. tional Po halo types also exist (3) but
dark, should appear without any outer The last correspondence to be estab- are quite rare. [As yet I have found
ring structure, as illustrated in Fig. ln. lished is the resemblance of the two no halos at all in meteorites or lunar
Compare also the densitometer profiles three-ring halos in biotite (Fig. lq) rocks (19)].
of the halo negatives of Fig. If (the and two similar halos in fluorite (Fig. The preceding discussion has shown
U halo) and Fig. In, shown in Fig. 2b
and Fig. 2, c to e, respectively. Fig. io
shows three similar halos in fluorite; Ca Ka
here, irrespective of coloration differ- Ca Ka 3.69 keV
ences, the halo radii are the same and 3.69 keV
correspond to the Ea of 210po (Table
1, columns 4, 6, and 20). Accordingly,
the halos in Fig. 1, n and o, are desig-
nated 210Po halos. (Actually I should
emphasize that since not all biotites Ca Ka
3.69 keV
exhibit the same coloration responses,
the radius measurements in Table 1 are U Ma A
strictly valid only for the particular 3.16 keV Rh
Ca K,B
micas I used. I did try to illustrate a /4.01 keV
range of responses by utilizing four dif-
ferent biotites for the U halo and (a) Fluorite Matrix
the three Po halo types.)
By analogy, the moderately devel-
oped biotite halo in Fig. lp shows a - CaKkeV
marked resemblance to the idealized 4.01
halo that would form from the se- (b) U Halo
quential a-decay of 214Po and 210po Fig. 3. Scanning electron microscope-x-ray fluorescence spectra of (a) the fluorite
(see Fig. lc). Table 1, columns 2, 3, (CaFg) matrix, (b) a U halo radiocenter in fluorite characteristic of Fig. 1, 1 and m,
6, 7, 16, and 17, shows the corre- and (c) a 21"Po halo radiocenter in fluorite characteristic of Fig. Ir.
5 APRIL 1974 6S
that Po halos can be positively identified one of the rings at 14.0 um and 14.4 ,um is 6 X 1012 4He ions per square centimeter. Band
a ghost ring. I also rarely observe a light sizes monotonically increased with dose to
by ring structure studies alone. That at about 16 jim, but do not presently associate about 100 CT but were reproducible in a
x-ray fluorescence analyses also provide this ring with :23Th (lo) a-decay. plateau region around 10 to 20 CT. Because
10. C. Mahadevan, Indian J. Phys. 1, 445 (1927). band sizes were unpredictable at high beam
quite convincing evidence is seen in 11. C. Moazed, R. M. Spector, R. F. Ward, intensities it was necessary to use beams of
Fig. 3c, where I show for the first time Science 180, 1272 (1973). only about 10 na/mm2.
12. Moazed et al. (11) stated that because they 15. D. E. Kerr-Lawson, Univ. Toronto Stud.
the x-ray spectra of a Po halo radio- could not find halos with dimensions matching Geol. Ser. No. 27 (1928), p. 15.
those of Henderson's type B halo (the 24Po 16. From a-decay theory, dX/X (3/2)(ZR)' 2
center (specifically, a 218Po halo). halo in my terminology) such halos do not (dRIR) + (2Z/E' 2) (dEIE), where Z is the
Comparison of Fig. 3, b and c, reveals exist; however, Henderson gave both measure- atomnic number, R is the nuclear radius in
ments and photographic evidence (6, figure 10-'3 m, and E (= E,) is the a-decay energy
that the Pb in the Po halo radiocenter 4, facing p. 242). They then inferred that in million electron volts. A particle of mass
in fluorite did not arise from in situ a different halo (a U halo) must be the m and charge z has a range r (halo radius),
equivalent of the type B halo, although the given by the expression r = constant X E2/mz2.
decay of U. [Longer runs have shown radii of the inner ring of Henderson's type Then d\/X 43(dR/R) + 46(dr/r). If the

small amounts as Se as well as U in B halo and the outer second ring of their difference between the halo radius and the
halo were significantly different (20 compared coloration band size at 4.2 Mev is real, then
some Po halo radiocenters (18).] On to 22.3 Arm). They concluded that all Po Ar = - 0.4 Am and dX/X 46(- 0.4/13) =
halos are only U halos, without having U - 1.4. Since the minimum uncertainty in
the other hand, the presence of Pb halos with normal ring structure available for making comparative range measurements is
is to be expected in a 218Po halo comparison. I showed (5) that Po halos and Ar = 0.1 MAm, it is actually impossible to estab-
U halos are distinguished by the nunmber of lish the constancy of X (for MU) from radio-
radiocenter because the decay product fossil fission tracks after etching; that is, few, halo data any better than dX/X 46(0.1/13)

is 20'6Pb. That the parent nuclide was if any, compared to a cluster of 20 to 100
tracks. I also showed that the threshold
= 0.35. Also, if dEIE =0 while dRIR #0,
then dX/X + 0. In such a case, halos furnish
218Po and not a 8-decaying isomer coloration dose is directly obtainable by con- no proof that X is constant.
precursor (13, 20) follows from half- verting a U halo fossil fission-track count (20 17. Some inner ring coloration in Fig. If results
to 100) to the number of emitted a-particles from other a-emitters in the U decay chain.
life considerations of the U halo by using the :03U branching ratio, X./Xf; this Fission track analysis shows that the dose of
U/Pb ratio (> 10); the proposed iso- contradicts the supposition that such data are a-particles from 23fU is only about 1013 per
unknown to two orders of magnitude. Ion square centimeter, about ten times less that
mer, if formed at nucleosynthesis, probe analyses of U halos show that a high the 'He ion dose for medium coloration.
U isotopic ratio can not be responsible for a 18. R. V. Gentry, in preparation.
should now be detectable in Po halo small induced fission-track count. Further- 19. , in Proceedings of the Second Luinar
radiocenters. No trace of this isomer more, contrary to a statement by Moazed et Science Conference (MIT Press, Cambridge.
al., Henderson was able to distinguish reliably 1971), vol. 1, pp. 167-168.
has yet been found, and I thus view between his type B and type C halos (6, pp. 20. , Atnn,i. Rev. Nutcl. Sci. 23, 347 (1973).
the isomer hypothesis as untenable. 246-248). 21. This work was sponsored by the Atomic
13. R. V. Gentry, S. S. Cristy, J. F. McLaughlin, Energy Commission under contract with Union
The x-ray data in Fig. 3c are unam- J. A. McHugh, Natutre (Lond.) 244, 282 (1973). Carbide Corporation, and by NSF grant GP-
biguous and should remove any doubt 14. The irradiated biotite samples were cleaved 29510 to Columbia Union College, Takoma
in about 5-Mum sections for microscopic exam- Park, Maryland.
that previously reported 21'6Pb/207Pb ination. The coloration threshold (CT) for
mass ratios (3, 13) actually are Pb iso- 30-,m biotite sections varied from 3 X 10's to 2 July 1973; revised 26 December 1973 .

tope ratios, and in fact represent a new

type of Pb derived specifically from Po
a-decay. In summary, the combined re-
sults of ring structure studies, mass V.Glutamyl Transpeptidase in Brain Capillaries:
spectrometric analyses, and electron
induced x-ray fluorescence present a Possible Site of a Blood-Brain Barrier for Amino Acids
compelling case for the independent ex-
istence of Po halos. The question is, Abstract. A fraction containing capillaries and rich in y-glutamyl transpeptidase
can they be explained by presently ac- was isolated from homogenates of bovine brain cortex by..density gradient cen-
cepted cosmological and geological con- trifugation. The enzyme was localized in the endothelial cells by a histochemical
cepts relating to the origin and develop- procedure. y-Glutantyl transpeptidase may function in the transfer of some amino
ment of Earth? acids across the blood-brain barrier.
Chemistry Division, y-Glutamyl transpeptidase catalyzes plexus, and brain capillaries (4). The
Oak Ridge National Laboratory, transfer of the y-glutamyl residue of reaction catalyzed by y-glutamyl trans-
Oak Ridge, Tennessee 37830 glutathione to amino acids according peptidase was later integrated into the
References and Notes to the following reaction (1) y-glutamyl cycle (5), proposed as a
1. G. H. Henderson, C. M. Mushkat, D. P.
system for amino acid transport. After
Crawford, Proc. R. Soc. Lond. Ser. A Math. Glutathione + amino acid I
the amino acid is converted to its -y-
Phys. Sci. 158, 199 (1934); G. H. Henderson --glutamyl amino acid + cysteinylglycine glutamyl derivative by y-glutamyl
and L. G. Turnbull, ibid. 145, 582 (1934); G.
H. Henderson and S. Bateson, ibid., p. 573. It was proposed that this reaction, transpeptidase at or near the cell sur-
2. J. Joly, ibid. 217, 51 (1917); Nature (Lond.) which makes the degradation of glu- face, a translocation step occurs bring-
109, 480 (1920). I have examined Joly's collec-
tion and found that he associated certain Po tathione dependent on amino acids, ing the y-glutamyl amino acid into the
halos with U halos and incorrectly associated
the 2lPo halo as originating with Rn a-decay. functions in amino acid transport (2). cell (5). The amino acid is then re-
R. V. Gentry, Science 173, 727 (1971).
, ibid. 169, 670 (1970).
The hypothesis is supported by the leased through the action of -y-glutamyl
5. , ibid. 160, 1228 (1968). finding that y-glutamyl transpeptidase cyclotiansferase (6) with the con-
6. G. H. Henderson and F. W. Sparks, Proc. is associated with cell membranes, es- comitant formation of L-pyrrolidone-
R. Soc. Lond. Ser. A Math. Phys. Sci. 173,
238 (1939). pecially in those cells where a high carboxylic acid (7). Conversion of
7. G. H. Henderson, ibid., p. 250. A fourth rate of amino acid transport is antici- pyrrolidonecarboxylic acid to glutamate
type attributed to 2"Ra a-decay is in error.
8. S. limori and J. Yoshimura, Sci. Pap. Inst. pated. Thus, histochemical studies have (8) and synthesis of glutathione com-
Phys. Chem. Res. Tokyo 5, 11 (1926).
9. A. Schilling, Neues Jahrb. Mineral. Abh. 53A, shown that the enzyme is heavily con- plete the cycle. Evidence for the par-
241 (1926). See translation, Oak Ridge Natl.
Lab. Rep. ORNL-tr-697. Schilling, as did Joly,
centrated in the brush border of the ticipation of the y-glutamyl cycle in
erroneously designated 21'(Po halos as emana- proximal convoluted tubules of the amino acid transport has been re-
tion halos. As for explanation of the 14.0-Mm, kidney, the apical portion of the in- viewed (5, 9-11).
14.4-Am, and 15.8-Mm rings which Schilling at-
tributed to Ul, UII, and lo, I can state that testinal epithelium (2, 3), the choroid y-Glutamyl transpeptidase was de-
66 SCIENCE, VOL. 184