924 
subsidence, ¢.g. coral reefs and islands, are parts of the earth’s 
surface which have Jately increased rapidly in weight ; and it 
may be worthy of consideration whether coral and volcanic 
islands have contributed to deepen the bed of the ocean. 
J. F. ANDERSON 
Cauterets, Hautes Pyrenees, July 12 
Curious Rainbow 
AN unusual atmospherical effect was witnessed here to-day, 
which I had a good opportunity cf observing. The sun was 
about 8° from the horizon, shining brightly upon a heavy shower 
which had a background of dark clouds. The result was, of 
course, a double rainbow of remarkable brilliancy. In addition, 
however, to the ordinary circular and concentric bows, there 
was a third of an elliptical form, the two ends of which respec- 
tively sprang from the two ends of the inner arc, while the 
elliptical curve cut the outer are at each extremity of a chord, 
which was parallel to, and which intersected the normal radius 
at a point about two-thirds of its length above, the diameter 
that formed the common base. The top of the elliptical bow 
was thus the outermost of the three, but the space between its 
inner margin and the outer margin of the second bow, although 
quite distinct, was not large. 
The appearance of the third bow was due to light reflected 
from the sea. The sun being low, the resulting line of reflection 
was long, and it was the linear character of the source of light 
which gave the elliptical form to the bow it occasioned. 
Dunskaith, Ross-shire, July 10 GEGRGE J. ROMANES 
CHLOROPHYLL COLOURING-MATTERS * 
Il. 
I THINK there can be no doubt that the spectra 
of the various yellow substances given in PI. IL, 
Figs. 3, 4, and 6 of Dr. Kraus’s work, are due to a 
variable mixture of xanthophyll, yellow xanthophyll, 
and lichnoxanthine. These can be separated, and 
do occur in different kinds of plants, either alone or 
mixed in such variable proportions that the spectra of 
the solutions show the absorption-bands, not only in 
variable positions, but also much less distinctly in some 
cases than inothers. This difference is ascribed by the 
author, not to a variation in the relative proportion of 
two or more substances, each having definite and unvary- 
ing characters, but to the modification of one single sub- 
stance, due to some unknown cause, assigning as a reason 
for this supposition that the chemical reactions are the 
same, and that the positions of the absorption-bands vary 
so gradually from one extreme to the other that no dis- 
tinct demarcation can be detected. Now this is so very 
fundamental a question in such studies, and, according 
as it is decided, would modify the conclusions so much, 
that it is requisite to discuss it somewhat fully. No 
doubt the position of the absorption-bands seen in the 
spectra of solutions in different liquids does differ very 
considerably, but I feel persuaded that the spectrum of 
the sare chemical compound, dissolved in the same 
liquid, is the same in all cases ; and that, if there is any 
difference between the spectra of two similar solutions, 
it is due to a difference in the substances themselves. I 
would restrict the term modification to those changes 
sometimes produced by the action of weak alkalis or 
acids, or by deoxidizing reagents, which are only of a 
temporary nature, so that when the solution is restored to 
its original state, the spectrum is seen to be just as at 
first. We really do require such a term, and I have my- 
self constantly ysed it in this sense. There is, however, 
no such relation between the different colouring-matters 
belonging to what I have called the xanthophyll group ; 
and, though the presence or absence of oily substances 
may, and sometimes does, materially influence the posi- 
tion of the absorption-bands seen in the spectra of plants 
themselves, yet, when dissolved in a relatively large 
quantity of a solvent, this effect is altogether overcome, 
As | have shown in my late paper the position of the 
* Continued from p, 2g4) 
NATURE 
|Fuly 17, 18 73 
absorption-bands in the different members of the xantho- 
phyll group is very different, and yet it would be easy so 
to mix them as to have a perfect series of connecting 
links, and in my opinion the variations from what appear 
to be independent compounds may be explained in an 
extremely simple and satisfactory manner, without sup- 
posing that the optical characters are subject to any 
such variations as are ascribed to them by the author. 
Whenever I have met with these variations I have 
looked upon them as presumptive evidence of there 
being a mixture, and have always been able to prove 
the truth of this principle by subsequent conclusive 
experiments. The following example will serve very 
well to explain my views. Many yellow flowers are 
coloured by a variable mixture of what I have called 
xanthophyll, yellow xanthophyll, and lichnoxanthine. 
The former occurs separately in the Alga, Porphyra 
vulgaris, the second in such pale yellow flowers as the 
yellow Chrysanthemum, and the last in the yellow fungus, 
Clavaria fusiformis. The absorption-bands of these’two 
kinds of xanthophyll are in a very different position, and 
the lichnoxanthine gives no bands, only an uniform ab- 
sorption, extending over about one half of the spectrum 
from the blue end, The chemical reactions are also 
equally distinct. On dissolving each in absolute alcohol, 
and adding a little hydrochloric acid, the first fades 
slowly, without being first changed into another yellow 
substance, and without turning blue or green ; the second 
is first altered into another yellow substance, giving a 
spectrum with two absorption-bands in a different posi- 
tion, and then turns to a deep blue, whilst the last 
remains unchanged for a much longer time, and fades 
very slowly. Now, of course, if all these were mixed 
together in variable quantities, we should get results 
varying according to the relative amount of each. The 
absorption-bands due to the two kinds of xanthophyll 
would lie in an intermediate position, according to the — 
relative amount of each constituent, and would be more 
or less indistinct, according as there was more or less of 
the lichnoxanthine ; and on adding a little hydrochloric 
acid to the solution in alcohol the colour would turn to a 
more or less blue green, and subsequently fade to a pale 
or deeper yellow, according to the relative quantity of 
each constituent. : 
In order to make my meaning more clear, let us sup- — 
pose that we were to take a mixture of equal quantities 
of xanthophyll and yellow xanthophyll, Using the nota- 
tion I have so often explained in former papers, the 
centres of the absorption-bands of the spectra of a 
solution in bisulphide of carbon would then be— 
Xanthophylgge = 20) sl". aye OD 
‘The aboveammxture 50° Ss ls Oe nee 
Yellow xanthophyll. . . . .. 7 8 
Now on exposing solutions of xanthophyll or yellow 
xanthophyll to the sun both fade, and if examined when 
very little colour was left undecomposed, the bands would 
be seen to be in the same position as at first, the solution 
being in fact just as if a large part of the colour had been 
removed, or as if it had been much diluted. In the case 
of the mixture this would not be the case. Xanthophyll — 
is more rapidly decomposed than yellow xanthophyll, so 
that when very little colour was left the bands would be 
no longer in the original position, but in the same place 
as those of yellow xanthophyll, showing that a small 
quantity of this is left, when all the other has been 
destroyed, If some lichnoxanthine had been mixed with 
the solution, after longer exposure to the sun no absorp= 
tion-bands would be seen, only the general absorption 
due to that substance, Moreover if we took equally 
deep coloured solutions in absolute alcohol of the same 
three different specimens, and added a little hydrochlorig 
acid to each, the xanthophyll would fade till it was colours 
less, the yellow xanthophyll would turn to a fine blue, 
