DEcEMBER 3, 1896] 
NATURE 
IOI 
from physical data, and that we, the geologists and zoologists, 
** are free to proceed, and to look for the conclusions warranted 
‘by our own evidence.” 
Lord Kelvin’s experiments on thermal conductivity of rock 
at various temperatures are of the highest interest. I did not 
allude to them because it seemed to me unnecessary to point 
out that, until we know the nature of the material which forms 
the deeper parts of the earth, any attempt to generalise from 
the results of experiments on the material of the surface must 
be inconclusive. 
Mr. Chree thinks that I might have made something of the 
uncertainty as to the true mean temperature gradient at the 
surface. The reports of the British Association on underground 
temperature afford abundant support for a temperature gradient 
in the northern hemisphere, which cannot be very different from 
that selected by Lord Kelvin. 
I fail to grasp the object of Mr. Chree’s letter, unless it be 
to proclaim that he never accepted Lord Kelvin’s conclusions, 
and I cannot see that any great object is gained by even this 
Statement. 
As I am writing on the subject, I should wish to point out that 
the address, as printed in the Report of the Liverpool meeting, 
_ will be slightly different from that which has appeared in the 
columns of NATurE. Lord Kelvin kindly drew my attention 
to one or two errors which will be corrected in the Report. 
Oxford, November 20. Epwarp B. PouLtTon. 
Measurements of Crabs. 
I AM much obliged to Prof. Weldon and Mr. H. Thompson 
for the careful consideration they have given to the doubt which 
I raised concerning the validity of the comparison made by the 
latter of measurements of crabs collected from the same locality 
in different years. Prof. Weldon offers some evidence to show 
that immersion in spirit does not affect the relative dimensions 
of parts of the carapace. The evidence is not direct, nor 
perhaps is it complete. It refers to female crabs, and not to 
the male specimens with which Mr. Thompson was dealing. 
But I notice that according to Prof. Weldon’s measurements 
the spirit specimens of 1895 differed more than the fresh speci- 
mens of 1895, from the spirit specimens of 1893, whereas if the 
spirit were the cause, the difference would be less between spirit 
Specimens and spirit specimens than between spirit specimens 
and fresh specimens. I admit then that there is little possibility 
of the observed difference being due to the action of spirit. 
But Mr. Thompson’s letter suggests other reflections. He 
draws my attention to the fact that the difference which he 
observed between crabs of the one year and those of the other, 
is of precisely the same kind as the difference between an older 
rab and a younger crab, or rather between a larger crab and a 
smaller crab. As the male crab increases in size, its frontal 
breadth is continually becoming less in proportion to its cara- 
pace length, while its dentary margin is becoming greater. 
‘What Mr. Thompson found, therefore, was simply that in one 
sample the individuals of a given size were more advanced in 
<dlevelopment than those of the same size in the other sample. 
The development or law of growth remaining the same, the size 
of one sample had been reduced in comparison with that of the 
other. But this is not, I think, correctly described as a change 
in the character of the species. I find that the crabs of the 
same stage from the two samples differed in average length by 
about °5 mm., whichis a very small difference. Such a difference 
might well be caused in the young crabs of two different seasons 
by a difference in abundance of food, due to meteorological 
differences ; the crop of young crabs was finer in one season 
than in the other. I believe that if the lambs of two seasons 
Were compared in the same way, differences in the rate of 
growth, or in the size of the individuals which had reached the 
same stage of development, would be found. We may say, 
indeed, that such differences are known to be of general 
occurrence. Mr. Thompson’s paper suggests that a consider- 
able change in the proportions of parts characteristic of the 
spectes had been observed, whereas, in point of fact, no new 
proportions were observed at all, but the old were found to be 
present in individuals of different sizes. J. T. CUNNINGHAM. 
College of Surgeons, November 20. 
Suggested Reef Boring at the Bermudas—and elsewhere. 
Mr. W. K. Morrison’s suggestion (NATURE, November 5) 
of the Bermudas as a site for renewed reef-boring experi- 
ments, and for the establishment of -a permanent biological 
NO. 1414, VOL. 55] 
observatory is well worthy of consideration. It is, at the 
same time, desirable to remark that the Bermuda reefs scarcely 
appear to possess the most favourable conditions for boring 
operations. As long since recorded by Dana (‘‘ Corals and 
Coral Islands,” p. 361), the Bermuda coral rock abounds with 
caverns and fissures, and there would consequently be an 
imminent risk of negative results being obtained there, as has 
happened at Funafuti, through the uncontrollable infiltration 
of sea-water. The circumstance, also attested to by Dana, that 
the reef-making species of corals at the Bermudas are but few 
in number, and are constantly submerged to a depth of from, at 
least, one to four fathoms, places this island group, as a station 
for the especial investigation of coral-growths, at a disadvantage 
in comparison with many others that might be mentioned. 
Much might be written in favour of the many locations on 
the Australian Great Barrier Reef that would be particu- 
larly suited for the prosecution of the borings and biological 
investigations proposed. My later sphere of investigation has 
brought within my notice another area which, in certain re- 
spects, offers advantages and facilities possessed by no other 
spot throughout the coral regions with which I am acquainted. 
I refer to Houtman’s Abrolhos, lying thirty miles to the west- 
ward of the port of Geraldton, within a day’s journey from 
Perth, on the coast of Western Australia. A paper indicating 
certain of the more remarkable marine biological features of these 
islands was communicated by me to last year’s Ipswich meeting 
of the British Association, and the subject is dealt with in 
further detail in a chapter of my book, ‘‘ The Naturalist in 
Australia,” now on the eve of publication. The data recorded 
that are most pertinent to the present issue are as follows. 
Notwithstanding the position of the Abrolhos islands so far 
south, 28° to 29°, as to be outside the limits of the tropics, the 
marine fauna—owing to a warm southerly drifting equatorial 
current—corresponds more nearly in character with that of 
Torres Straits than with that of the adjacent mainland coast- 
line as far north at least as King’s Sound, in latitudes 16° to 17°. 
The western is a perfect archipelago of coral islands, including 
lagoon, barrier, and fringing reefs, and abounds with readily 
accessible coral growths in infinite variety. 
An abundant illustration of what Houtman’s Abrolhos can 
produce is afforded among the series of coral specimens 
from the Western Australian reefs, recently contributed by me to 
the British Museum coral galleries. 
Any practical scheme initiated in this country having as its 
object experimental reef-boring, or the establishment of a 
marine biological observatory on Houtman’s Abrolhos, would 
undoubtedly receive substantial sympathy and support at the 
hands of the Western Australian Government collectively, and 
yet more definitively through the medium of its far-seeing and 
scientific-minded Premier, Sir John Forrest. 
Any assistance towards the furtherance of this suggested 
scheme, should it merit consideration, that may lay within my 
power, would be most willingly contributed. 
London, November 9. W. SAVILLE-KENT. 
The Structure of Nerve Cells. 
THE Spanish histologist, Dr. S. R. Cajal, on p. 9 of his book 
“Les nouvelles idées sur la structure du Systeme nerveux,” 
translated into French by Dr. L. Azoulay, makes the following 
assertion :—‘‘ Les cellules nerveuses sont des unités indépen- 
dantes, ne s’anastomosant jazazs ni par leurs rameaux proto- 
plasmiques, ni par leurs expansions nerveuses ou cylindres axes.” 
(The italics are mine.) I venture to bring this statement forward 
because recently I have discovered that it is not universally 
correct ; in sections of the medulla oblongata of a young snake, 
Tropidonotus natrix, prepared according to Cox’s modification of 
Golgi’s corrosive sublimate process, I have found a pair of cells 
on the ventral edge distinctly united together by a protoplasmic 
process, or, as I would propose to term ee) dendrite. Cox's 
mercurial method is so far better than the chrom-osmium-silver 
method, inasmuch as the preparations made by it keep much 
longer. ALFRED SANDERS. 
The Hawthorns, Caterham Valley, November 26. 
Snow Buntings. 
Iv may interest some of your readers to learn that on 
November 17 I saw near the top of Snowdon a flock of snow 
buntings. The mountain was snow-clad, and had been so for 
several weeks. J. R. Dakyns. 
Penn-y-ywvyl, November 23. 
