earth to geological problems, which occupy most of the
ensuing chapters, are based on values somewhat below
the means of all published results.
The basic rocks of the Deccan, which are of enormous
extent, have yielded radium at every point at which
they have been examined. A series of tunnel and
surface specimens, kindly sent by the Director- General
of the Indian Geological Survey Sir Henry Holland
42 RADIOACTIVITY AND GEOLOGY.
enabled estimates to be made at widely separated
points. The specimens examined up to the present are
the following :
Bedded basalt, Ghats near Thuria, . * . 27
Pavagad Hill, Panch Mahals, . 8'2
Basalt dyke, near Palasi, Dhar State (2), . . 4'9
Bedded basalt, tunnel near Palasdari, . . . 3'8
,, ,, ,, between Keloli and Vani, . 2 '4
,, ,, ,, between Kasara and Igatpuri, 3'6
Mean, . . .4'3
The convention of omitting the qualifying expression
10~ 12 will be continued, as above, throughout the rest of
this book unless occasion arises to specially refer to this
factor. It will be, therefore, always understood that
the radioactivity expressed by an unqualified number
is the quantity of radium in billionths of a gram per
gram of material under discussion. A number given in
brackets after the designation of a substance refers to
the number of tests, generally on independent samples
of the substance : the radioactivity in such cases being
the mean of the results.
The basalt of the Giant's Causeway, which is part of
the great lava flow of the Inner Hebrides, has been
treated in five experiments, all, however, on adjoining
fragments of the rock. The results agree very closely.
In all there are fourteen results on basic rocks :
Deccan basalts and traps (6), . ;','. 4'3
Giant's Causeway basalt (5), . . . .5*8
Basalt, Mellenifjord, Greenland, . . .. 5'0
" Whinstone," Balfour Bore, Fifeshire (from a
depth of 3000 feet), 3'8
RADIUM IN THE EARTH'S SURFACE-MATERIALS. 43
On granites there are six results :
Leineter granite, Bally knock en (2), . . .5*4
Mourne granite, Newry, Co. Down, . . . 3'7
Aberdeen granite (3) (from depths of 80 to 270 ft.
in Kemnary Quarry), 3*2
Mean, . . . 4'1
Syenite, Plauen, . 6'8
And the following from the very ancient rocks of
Sutherlandshire, Scotland :
Lewisian mica schist, . . . . . '.6*3
gneiss, . . . 3'8
,, hornblende schist, . . . .8*0
Mean, . . 5 '7
Twelve specimens of recent volcanic racks :
Lava, Crater, Kilauea, 1875, . . .6*2
Scoria, Stromboli, . . . . . . .6*0
Augite andesite, Val di Lachi, Lipari, . .2*4
Trachyte, Monte Olibano, Campi Phlegraei (2), . 10 '6
Obsidian, Ascension Island, . . . .2*5
Lava, Rocci Rossi, Lipari (2), . . . . 8'0
Bomb, Vulcano (2), 7 '6
Olivine trachyte, Ischia (2), . . . . 9'4
Pumice, Chimborazo, near Quito, . . . 4'8
Lava, Vesuvius, 1855 (2), . ; . . . 19 '2
Bomb, Martinique, 1902 (weathered), . . .2*0
Pumice, Krakatoa, 6 '9
Mean, . . . 7'1
To these must be added a large number of rocks
which enter into the composition of the Simplon
massif, and which are believed to be of igneous origin ;
and others again which enter into the St. Gothard and
the Finsteraarhorn massifs. These are in every case
44 RADIOACTIVITY AND GEOLOGY.
tunnel specimens, some of those from the Simplon
coming from depths beneath the surface of 2135 metres
[Reserving particulars till we come to consider
specially the results obtained in the case of these
two great tunnels, it is sufficient to say here that in
the case of the Simplon some difficulty was experienced
from what may be called sporadic radium. Possibly
the cause of this has to do with the facts that not only
are many of these rocks rather above the average in
radioactivity, but they have been exposed to extreme
metamorphic actions which may have led to a segre-
gation of the uranium. Some, bearing evident uranium
minerals, were met with, so that a low and a high mean,
the latter embracing a few very exceptional results,
appear to arise out of the experiments. Further
experiments are in progress to determine, if possible,
the most probable average. Such extreme variations
were not noticed in the St. Gothard rocks. The
Finsteraarhorn granite is included in the mean of
the St. Gothard igneous rocks.
Simplon (32) ; schists and gneisses, mean, low, . 7*6
(32); high,. 9-1
St. Gothard (32) ; granites, gneisses, and schists,
The mean of all the igneous rocks examined by the
author (92 distinct specimens), taking the low mean of
the Simplon rocks, is 6'2. This may be somewhat
unduly raised owing to the seemingly rather abnormal
results obtained with the Simplon materials. Setting
aside for the present the entire Simplon series of igneous
RADIUM IN THE EARTH'S SURFACE-MATERIALS. 45
rocks, the mean of 60 rocks is 5'5. These 60 results,
taken along with those of Strutt and Eve on similar
materials, give a general mean of 4-2 obtained on 92
specimens of various igneous rocks.
It would be impossible to predict with any certainty
the relative radioactive states of the sediments and
the parent igneous rocks. So far as the purely detrital
rocks are concerned, the question would turn, of course,
on how far the uranium is retained in the debris of
the parent rock, and on the weathering qualities of
certain minerals which appear to be very often
specially radioactive ; for instance, zircon. There is
little doubt that high results occasionally met in
rocks of the detrital class are due to the preservation
of this mineral. Evidently we could not safely
predict the result where such uncertain factors enter ;
but, reasoning from the behaviour of other heavy metallic
substances, the probability would incline in favour of
the uranium being in considerable part removed in
the general disintegration of the rock. Thus we should
expect a lowered radioactivity in the case of such rocks.
The precipitated sediments present a very different
problem, and one which experiment alone could solve.
Here the radioactivity must depend on the abstraction
of uranium along with the rock-materials in the
process of deposition, and clearly the radioactivity
of the ocean is concerned in the genesis of such
materials. The uranium in the ocean and its sedi-
ments is indeed a factor of such considerable impor-
tance in the consideration of the whole question of
radioactive denudation, seeing that the whole or the
46 RADIOACTIVITY AND GEOLOGY.
greater part of it must have come from the land, that
we will state what is known on the subject before
passing on to the sediments.
The waters of the ocean cover five-sevenths of the
earth's surface to a mean depth of about 3 '8 kilometres,
and represent, of course, far the most abundant of the
surface-materials open to our investigations. A con-
siderable number of results have accordingly been
obtained in order to arrive at some conclusion as to the
probable mean radioactivity. Quantities of the water
from 1500 to 3000 c.cs., in most cases concentrated to
about 'half bulk by evaporation in a laboratory into
which radioactive preparations had never come, were
used for testing. The results obtained are for the greater
number published with fuller particulars elsewhere. 1
Radium per c.c.
Coast round Ireland (5), 0'034
Atlantic Ocean, Madeira to Bay of Biscay (5), 0'017
Atlantic, New York to Ireland, Arctic current, 0*016
,, ,, ,, ,, Gulf Stream, . 0'014
,, Mid Atlantic, . O'Oll
,, 260 miles W. of
Ireland, . O'OOS
lat. 00', long. 3126'W., . . . 0*038
Atlantic, South ; lat. 16 54' S., long. 3749' W., 0'007
Mediterranean (2), O'OOS
Black Sea, 0'007
Arabian Sea, lat. 1040'N., long. 580'E., . 0'027
Indian Ocean to Mediterranean
Sandheads to Madras, 0'004
Off Colombo, . . 0-007
,, ,, Minicory to Sokstoa, 0*004
Red Sea, off Jiddah, 0'009
1 Joly, Phil Mag., March, 1908, p. 385.
RADIUM IN THE EARTH'S SURFACE-MATERIALS. 47
The mean of the twenty-four determinations, many
of which are individually the results of several experi-
ments, is 0'017 x 10~ 12 grams per cubic centimetre.
These results are obtained by modes of treatment
varied in such a manner as would best guard against
any systematic experimental error. Sea-waters are
only safely dealt with on addition of pure redistilled
hydrochloric acid, which lessens greatly the risk of a
precipitate forming in the process of concentration.
The variability of the results is considerable. To what
is this due ? It would seem unlikely that retention
of the emanation can account for the lower results,
for most of them were confirmed by a second boiling,
and in some cases on three or four repetitions of the
experiment. The second boiling generally gives a some-
what increased yield of emanation, but there is seldom
any subsequent rise. Again, the matter was inves-
tigated by adding to a sea-water a certain known
small quantity of radium, and seeking if the emanation
corresponding to it could be completely withdrawn.
The result was quite satisfactory. This experiment
does not indeed prove conclusively that a residuum of
emanation might not remain over in all sea-waters
after ebullition ; but if so, we would expect this residuum
to be fairly constant in amount, and the differences
from experiment to experiment have still to be
explained. Nothing is more astonishing than the fact
that these small amounts of emanation can, in any
case, be withdrawn. The volume of the emanation per
gram of radium is but 0*585 cub. mms. 1 Thus there are
1 Rutherford and Geiger, Proe. It. ., 81 A., 173.
48 RADIOACTIVITY AND GEOLOGY.
but 9 billionths of a cubic millimetre in a litre of
average sea-water !
It is, perhaps, premature to say much as to the
possible cause of the variations. The most plausible
explanation seems to be the removal of radioactive
matter from the surface-waters by precipitation under
the action of decaying organic life and sulphur
bacteria; so that according to the local organic con-
ditions and it is known that these vary considerably
under the influences of surface-currents, etc. a large
variation of the radium near the surface might arise.
The investigation of water from the deeper parts of
the ocean has yet to be carried out, and will be of
There is evidently a large quantity of radium in the
ocean. Taking the mass of the ocean as 1*452 x 10 18
tonnes, there must be about 20 x 10 9 grams, or about
20,000 tons of radium, in its waters.
It is of interest to note in connexion with the radio-
activity of sea-water that Eve 1 has found that the
ionization of the atmosphere over the central parts of
the Atlantic is comparable with that over the land,
even in the calmest weather, when it cannot be
supposed that the emanation, which, doubtless, is
responsible, could have been blown from the con-
tinents. This fact appears to be traceable to the
radioactivity of the water, and the facility with which
emanation may be supposed to escape from it. The
bubbling of air through liquids containing a trace of
radium results in loading the air with the emanation.
i Eve, Phil. Mag., 6, xiii., p. 248.
RADIUM IN THE EARTH'S SURFACE -MATERIALS. 49
Hence every bubble rising through the water, or
mingled with it in rough weather, brings emanation to
the surface. Again, it is certain that the emanation
diffuses out at the surface ; and as this is continually
being renewed by currents and disturbances of every
kind, the effect must be maintained continually at its
maximum. It must result from these causes that
the ocean, although not nearly so radioactive as
the land, must give up its emanation more freely,
and hence the observed ionization appears accounted
The condition of the deep-sea deposits fully supports
the observations on the radioactivity of the ocean.
From materials supplied to the author, mainly by
Sir John Murray, seventeen various samples of oceanic
deposits have been examined. 1
Blue Mud. "Challenger." Off E. coast of N. America.
Lat. 38 34' N., long. 72 10' W. 1240 fathoms.
Terrigenous Mud. From " mud volcano," coast of Arakan,
Bay of Bengal. Surface. Radium 2 '9.
Green Sand. Ss. "Albatross." Locality unknown. Radium
Globigerina Ooze. Off W. coast of Ireland. 570 fathoms.
Radium 6 "6.
Globigevina Ooze. Central N. Atlantic. Lat. 48 17' N.,
long. 39 49' W. 2493 fathoms. Radium 7*0.
Globigerina Ooze. "Challenger." Middle of S. Atlantic.
Lat. 21 13' S., long. 14 2' W. 1990 fathoms. Radium
1 Joly, Phil. Mag., July, 1908, p. 190.
50 RADIOACTIVITY AND GEOLOGY.
Globigerina Ooze. " Challenger." Pacific Ocean, W. of S.
America. Lat. 38 6' S., long. 82 2' W. 1825 fathoms.
Radium 7 '4.
Globigerina Ooze. "Albatross." Central Pacific. Lat. 22
11' S., long. 133 21' W. 2042 fathoms. Radium 8'0.
Calcareous Mud. "Albatross." E. of Society Islands. Lat.
21 4' 6" S., long. 133 1' 2" W. 2225 fathoms. Radium
Red Clay. " Albatross." N. Pacific, W. of Central America.
Lat. 10 38' N., long. 105 47' 6 W. 1955 fathoms.
Radium 13 '0.
Red Clay. "Challenger." N. Atlantic, off coast of Africa.
Lat. 24 20' N., long. 24 28' W. 2740 fathoms. Radium
Red Clay. "Challenger." Central Pacific, near region of
Radiolarian Ooze (as under). Lat. 13 28' S., long.
149 13' W. 2350 fathoms. Radium 52 '6.
Radiolarian Ooze. "Challenger." Central Pacific. Lat.
3 48' S., long. 152 56' W. 2600 fathoms. Radium
(From this sample the magnetic particles had been
Radiolarian Ooze. "Challenger." Central Pacific. Lat.
7 25 S., long. 152 n 15' W. 2750 fathoms. Radium
Manganese Nodule. Locality same as that of last ooze.
Radium 24 '0.
Manganese Nodule. "Albatross." Pacific, off W. coast of
S. America. Lat. 8 29 '5 S., long. 85 35' '6 W. 2370
fathoms. Radium 21 '0.
The diatom oozes have not yet been examined.
The great quantity of radium spread over the ocean
floor is well shown when we enter the above results
RADIUM IN THE EARTH'S SURFACE-MATERIALS. 51
against the extension of the several sediments as
determined by Sir John Murray :
TJ j- Extension : millions
Radium. of square miles.
Globigerina Ooze, . . 7*2 . . 49 '5
Radiolarian Ooze, . .367 . . 2 '5
Red Clay, .... 27'0 . . 51 '5
It is established on satisfactory evidence that the
more slowly collecting deposits are those of non-
calcareous character the red clay and the radiolarian
ooze ; the red clay in some cases being almost devoid
of calcareous matter, and collecting at an extremely
slow rate. The evidence for this is the amount of
unburied and even fossil remains which are brought
up with such deposits. The foregoing results show
that these deposits are richer in radium than the more
rapidly growing calcareous sediments. This would
follow if we supposed the rain of calcareous and
siliceous tests acted as a dilutant. It is of interest
to note in this connexion that the process of extraction
of lime salts from the ocean by organic actions must,
to a considerable extent, exclude the uranium in the
waters, as indeed might be anticipated. The curious
calculation of Bischof, 1 that an oyster requires the lime
from some 27,000 to 76,000 times its weight of sea-
water, would establish this fact; for if the uranium
present in this bulk of water was taken into the shelly
limestones, their radioactivity would be much greater
than experiment shows.
1 Bischof, Elements of Chemical and Physical Geoloyy, vol. i., p. 180,
52 RADIOACTIVITY AND GEOLOGY.
The following table shows that radioactivity and
the percentage of calcareous matter in these deposits
stand in an inverse ratio :
Globigerina Ooze. "Chall."338, . 92'24 . 6'7
296, . 64-34 . 7'4
Red Clay, . 5, . 12'00 . 15'4
. 276, . 28-28 . 52'6
Radiolarian Ooze, . 272, . 10 '19 . 22 -8
274, . 3-89 . 50-3
The percentages of calcium carbonate are from the
report of the " Challenger " Expedition. The red clay
fourth from the top in the table might appear ex-
ceptional to the others, having too high a percentage
of calcium carbonate for the radium content. It is
probable that this is a case of recent change in the
character of the deposit, the antiquity being shown by
the number of sharks' teeth and manganese nodules
brought up. Thus the specimen is a mixed one, so to
speak, there being a lower part rich in calcareous
matter, which acts to dilute the red clay more recently
forming above. Several such cases of recent change
are recorded by Murray and Renard. 1
The view that the radium on the earth's surface
might be of extra-terrestrial origin 2 suggested the
desirability of examining the cosmic particles long ago
identified by Sir John Murray as widely distributed
over the ocean floor, and, as it were, concentrated
in the deposits of very slow growth. If the theory
1 Report Scientific Results, " Challenger " Expedition.
2 Joly, Nature, January 24, 1907, p. 294.
RADIUM IN THE EARTH'S SURFACE-MATERIALS. 53
has any reality, we should expect that such bodies
would reveal exceptional richness in uranium, and, of
course, in radium. A tube of magnetic particles,
among which the cosmic particles would be included,
received from Sir John Murray, was accordingly
tested for radium. These magnetic particles had been
withdrawn from material raised from a depth of
2300 fathoms in the North Pacific. The quantity
available was only 1-1 gram. The result was a radium
content, with difficulty read, of 0*6. Had there been
present even one rich uranium-bearing particle, so
small a result would not have been obtained. The
theory referred to therefore lacks any support which
this experiment might afford.
"We have seen in an earlier chapter that it is to the
genetic connexion between uranium and radium that
we look for the maintenance and renewal of the latter.
It is thus of importance, in relation to oceanic radium,
to establish, where practicable, the existence of the
parent substance, and thus obtain assurance that the
radium is not exotic in character, and the radioactivity
a passing property. In the majority of rock-investi-
gations this is impossible, owing to the minuteness of
the quantities of uranium involved, and the lack of
character which the uranium spectrum exhibits. But
in these cases the conditions assure us that, unless there
be some other parent to radium, the uranium must be
there ; for how else has the radioactivity been main-
tained through past ages ? The oceanic conditions
lead us to a similar conclusion, so far as the waters
of the ocean are concerned. In the great volume of
54 EADIOACTIVITY AND GEOLOGY.
the ocean we have estimated not less than 20,000
tons of radium. This cannot have been supplied in a
period so brief as to be included in historical times.
Yet, unless the uranium is present, the whole amount
of radium must be renewed in a few thousand years
by the rivers. In order to accomplish this, however,
the rivers must be much richer in radium than the
ocean. We can acquire an idea of what their radium
content must be. The waters of the ocean possess
a total mass of 14 x 10 18 tonnes. The amount of
water annually discharged by the rivers of the world
has been estimated by Murray at 27,191 cubic kilo-
metres, or about 27 x 10 18 tonnes. Now a-sVo part of
the radium is transformed each year, so that if the rivers
are effective in keeping up the supply, they must in
one year send into the ocean an amount of radium
equal to ^ of what is present in its waters. But
the quantities of ocean-water and river-discharge per
annum are as 50,000 to 1, or, in other words, the rivers
must take 50,000 years to replenish the ocean. They
should, therefore, be richer in radium in the ratio of
50,000 to 2540, or as, say, 20 to 1. There is no evidence
of river- water possessing such richness. An examination
of Nile water afforded a radioactivity of but 0*0042,
which is about one-fourth the oceanic, and water from
the Kio de la Plata showed 0-0052.
Again, we are obviously in the difficulty of accounting
for the uranium corresponding to the radioactivity of
the ocean if not conveyed to the sea. For in this case it
must be in the detrital sediments, and these should show
the concentrated radioactivity of ages of denudation.
RADIUM IN THE EARTH'S SURFACE-MATERIALS. 55
This they do not, but in point of fact are less radioactive
than the igneous rocks a state which is in harmony
with one conclusion only : that the uranium released
from the rocks by denudation is carried into the
While this reasoning seems unassailable as regards
the ocean, we cannot apply it rigorously to the sediments
accumulating beneath. It might be that a certain pre-
cipitation of radium would give to the surface parts of
these deposits a temporary high radioactivity.
Against this view we have the fact that such ancient
deep-sea deposits as are open to our examination show
considerable radioactivity. The chalk and greensand
are instances. The former must have been a relatively
fast-formed deposit composed almost entirely of sub-
stances extracted from solution ; its radioactivity com-
pares well with that of Globigerina Ooze. A specimen
of green sandstone showed a somewhat higher radio-
activity than a similar sediment recently dredged.
Plainly in these cases the uranium must have been
precipitated along with the other materials.
As further evidence on this point, in the case of a
red clay showing the high radioactivity of 544, Professor
Emil Werner has recently estimated the uranium, using
a colorimetric method. In 8*42 grams of the clay sub-
mitted to Werner, the quantity of uranium should be
1*3 milligrams. The estimate afforded from 0'6 to 0*7
milligrams. But the separation of the uranium is
probably incomplete. 1
If we take it that the richness in radium of the
1 Joly, Phil. Mag., July, 1908, p. 196.
56 RADIOACTIVITY AND GEOLOGY.
oceanic sediments is on the average no greater than
that which we observe in the Globigerina Oozes and
this is probably a low estimate the quantity of radium
involved is enormous. We can calculate with consider-
able security the aggregate mass of such oceanic deposits
as are derived from dissolved materials; i.e. the cal-
careous oozes and muds, the diatomaceous oozes, the
radiolarian oozes, and much of the mass of the red
The elements of this seemingly insoluble problem are
really simple. We know that the sediments in the
ocean once in solution are all ultimately derived from
the igneous rocks ; we also know, as the average result
of numerous analyses of both igneous and detrital sedi-
mentary rocks, that about 33 per cent, of the former
pass into solution in the process of their conversion into
the latter. Now of these substances going into solution,
most are again precipitated from the ocean, or extracted
by organic agency more or less rapidly, forming, in fact,
the sub-oceanic sediments. The salts of sodium alone
are neither precipitated nor extracted. 1 This is owing
to their very high solubility. The recognition of this
fact gives us a clue to many great generalizations in
the history of solvent denudation. Among other facts,
it serves to tell us what total amount of original igneous
rock must have been broken up by denudation during
the course of geological time. 2 This knowledge is
obtained directly from the chemical analyses showing
the average loss of sodium from igneous rocks when
1 Joly, Trans. Royal Dublin Society, vol. vii., ser. ii., 1899, p. 23,
et seq. 2 Joly, ibid., p. 46.
RADIUM IN THE EARTH'S SURFACE-MATERIALS. 57
becoming sedimentary rocks, and the chemical analyses
which give us the quantity of sodium in the sea.
Thus we find that the average igneous rock contains
3*39 per cent, of soda and the average sedimentary rock
1-30 per cent. 1 The loss has, therefore, been
3-39 - 1-30 x ~- = 2-52 percent.
in the process of conversion. Now the sodium in the
ocean when converted into the equivalent oxide
amounts to 21 x 10 15 tonnes. The mass of the parent
igneous rock is therefore :
91 x 1 15
" .-0 X 10 = 83 ' 3 X 1Q16 t0nneS '
About 33 per cent, of this parent igneous rock
passes into solution ; and this amount would represent
all the substances now in the sediments and still in
solution in the ocean, but for the fact that some part of
what has been precipitated in the ocean in past times
is no longer to be reckoned among the deep-sea deposits,
having been, in fact, elevated into dry land. Such are
the limestones and some other deposits which, com-
paratively, are insignificant in amount. Van Hise 2