JANUARY 24, 1907] 
MATGERE. 
205 
before its radio-activity dies out. We would expect that 
the more impervious mineral substances would show the 
least amount. Quartz is without radium, as the Hon. 
R. J. Strutt shows by his determinations. 
There appears to be no improbability that matter in 
minute quantities might reach us from the sun. Here we 
are observing the most minute traces. If the observations 
are correct as to the velocities of solar ejections, it would 
take but a few days to bring solar matter into the orbit of 
the earth. The sun-spot-weather connection may possibly 
be involved, as well as the phenomena of atmospheric 
radio-activity, although doubtless emanation from radium 
already accumulated on the surface and in the soils is 
mainly responsible for these latter effects. Whatever may 
be said as to the value of such subsidiary evidence, it 
appears as if only by looking to such an extra-terrestrial 
origin of radium can we evade the difficulties connected 
with the associated uranium. We are evidently not com- 
pelled to assume that the radium received upon the earth 
carries with it the equilibrium amount of uranium, although 
doubtless we may expect that some uranium is also 
received. 
It is possible to arrive at a rough estimate of the amount 
of radium reaching the earth if we assume the annual 
oceanic supply of radium is mainly extra-terrestrial, and 
that a state of radio-active equilibrium with the average 
rate of supply has been attained. In effecting this estimate 
we deduct the annual river supply from the supply required 
for maintenance of the radium concentration of the ocean. 
The result is an annual supply to the ocean of 1-77 x 10° 
grams. If this supply is received uniformly over the 
oceanic area, about 12-5 milligrams enter over each square 
mile per annum. This result is probably excessive, as it 
assumes no uranium to be in the ocean or received from 
extra-terrestrial source. 
If it is permissible to apply to the land area the mean 
figure deduced for the ocean, we can ascertain the depth 
to which the observed radium content of the sedimentary 
rocks would be maintained at its present value. It is but 
trifling—about ten metres. As I have indicated above, 
however, the received radium will be washed from the 
surface soils and carried into the denser and more retentive 
rocks. The due proportion is doubtless carried to the sea 
on the break-up of the rocks or by percolating waters. 
We are no longer in difficulties on this score. Jen onx. 
Trinity College, Dublin, January 6. 
P.S.—A small quantity of radium will almost certainly be 
carried to the land along with wind-borne sea-salts. From 
Pierre’s measurements of the latter on a coastal rainfall 
of 60 cm., I find that about three thousandths of a milli- 
gram per square mile (1-2x10-*° grams per sq. cm.) per 
annum will in this way reach coastal countries. This 
small quantity will not complicate the problem unless there 
is associated uranium, and unless, further, this latter sub- 
stance accumulates in the rocks. In this case the rocks 
of coastal countries might in course of time come to have 
a higher content of radium than interior continental rocks. 
But, I repeat, the supposition that uranium will con- 
tinually gather in the rocks and never follow the usual 
channels of escape seems very improbable. 
Much remains for investigation, naturally arising out of 
Mr. Strutt’s fertile work. Rain-water should be system- 
atically examined, due allowance being made for wind- 
borne radium carried from the sea. I have begun such 
observations, but they will necessarily demand time and 
care if they are to be of value. In the hope of getting 
further light on some of the points at issue, I have the 
rocks of the Simplon Tunnel and certain of the deep-sea 
oozes under examination. J. Jory. 
January 16. 
Green Sunset Colours. 
Tue green sky described by Mr. Collins in Nature of 
January 3 (p. 224) was evidently an unusually brilliant 
example of the green tints often seen in a sunset sky. As 
I have not seen any explanation of the phenomenon, it 
may possibly be of some interest to give the following 
attempt at one. 
The colour of the sky at any time is made up of two 
NO. 1943, VOL. 75] 
components: A, the light from the upper regions; B, that 
reflected from the small particles in the lower air. The 
A component is always blue, and its spectrum shows a 
deficiency in red and yellow rays. Its light passes between 
the particles, and therefore forms a background upon 
which they are projected. The spectrum of the B com- 
ponent is. variable. When the sun is well above the 
horizon the light is white, and the variations in the deep- 
ness of the blue of a clear sky are due to differences in 
the relative proportions of A and B. As the sun nears 
the horizon the B light begins to lose its more refrangible 
rays, and the absorption extends towards the green and 
yellow as the sun goes down. 
Now if we take two equally brilliant spectra, cut out 
the red, orange, and yellow from one, and the violet and 
blue from the other, and then mix the residues, we shall 
obviously have all the colours necessary to make white 
light with a double allowance of green. An eye receiving 
the whole will see pale green. This, I take it, is the 
origin of the green colours of the sky. The A component 
is deficient in the less refrangible rays, which are sup- 
plied by the B component, and the two spectra overlap 
in the green, showing an excess of that colour. 
Occasionally, but rarely, the two are exactly comple- 
mentary over a limited stretch of sky, and then white 
patches are seen amid the colours of the sunset which 
are easily distinguished from clouds. They shade off on 
one side into tints of green where the spectra overlap, 
into yellow where the B component is in excess, and into 
blue where the A light preponderates. 
When the sky is clear it is no uncommon thing to see 
a considerable expanse of green, shading on the one side 
into pale lemon-yellow where the overlap of the spectra is 
considerable, while on the other side it shades through 
a narrow border of silvery tint where the balance is exact 
into a delicate rosy hue where there is a general deficiency 
in the central rays. 
Green tints are by no means always to be seen, and I 
think the foregoing explanation shows why—their pro- 
duction depends upon such an adjustment between the 
brightness of the two components that they shall be 
approximately equal. The white~ patches are rarer still, 
as they require exact equality in brightness and correct 
apportionment of colour. ArtHuR W. CLAYDEN. 
5 The Crescent, Exeter. 
Ultra-violet Fluorescence of Benzene. 
From my observations on the emission of light by canal 
rays (Ann. d. Phys., 21, 401, 1906; Physik. Zeitschr., 
7, 355, 1906), I have concluded that absorption of light in 
a band spectrum (running towards the red) produces fluor- 
escence. Hartley and others have stated that benzene has 
a banded absorption spectrum in the ultra-violet ; adopting 
that principle, I conjectured that benzene had an ultra- 
violet fluorescence. I have confirmed this by the follow- 
ing method :—The ultra-violet rays from a mercury lamp 
by W. C. Heraeus (Hanau, Germany) fell vertically upon 
a diluted solution (0-25 per cent.) of benzene in alcohol, 
and the fluorescence light emitted by it horizontally was 
analysed by-a quartz spectrograph. In the spectrograms 
obtained there appeared, besides the mercury lines, a group 
of four strong continuous bands situated in the ultra- 
violet between AA 271 and 314 “mu; these bands run towards 
the red end of the spectrum, and the heads have the wave- 
lengths 272, 280, 283, 292 MH. 
As Hartley has shown, the absorption spectra of the 
benzene derivatives are produced by an alteration (both in 
intensity and spectral position) of the simple benzene 
spectrum; such an alteration is produced by condensation 
and substitution. According to our principle, the same is 
true for the fluorescence spectra of the benzene derivatives. 
The question as to the relations between chemical con- 
stitution and fluorescence is thus reduced to the question 
of the relations between chemical constitution and absorp- 
tion, and banded absorption must be explained with regard 
to the fact that it is coupled with fluorescence. There may 
also be drawn from the above result some conclusions 
about the constitution of the benzene ring. 
J. Srark. 
