30 
NATURE 
[ NoveMBER 12, 1896 
stated. In fact, Prof. Weldon’s original interpretation has not 
yet been, and never may, be disproved, and, for all we know to 
the contrary, selection may still be acting in the direction indi- 
cated by his measurements. Mr. Bather will see, on reference to 
my letter, that I advanced the idea ‘‘ for whatever that idea may 
be worth.” It appears necessary to point out that the ‘‘ facts” 
are before us in this case; it is a question of interpretation. The 
difficulty raised by your correspondent is one of his own creation, 
and not one of mine. Where have I stated, or even suggested, 
that the appearance of a character which formerly was of selec- 
tion value, and which now appears in the later stage of indi- 
vidual development, is a ‘‘ universal principle” ? If the sugges- 
tion which I made with reference to the interpretation of Prof. 
Weldon’s results is sound in principle, it is obvious that it refers 
to a (comparatively) recent cessation of selection, and so far Mr. 
Bather has correctly grasped my meaning. When, however, he 
lays it down, on his own responsibility, that the conditions 
then ‘‘can hardly have been very different to now,” he is laying 
claim to a knowledge of the past and present conditions of life 
and toa familiarity with the time necessary to bring about the 
modification of species, at which I can only stand aside with 
envy and admiration. If Mr. Bather really must have some 
“known fact ” after this distinct dissociation of myself from his 
methods, I will refer him to those cases of mimetic butterflies 
which have no models. There is good reason for believing 
that certain butterflies which are undoubted mimics, since they 
depart from their allies in type of pattern, have for some un- 
known reason survived, while the species which they imitated 
have, also for unknown reasons, become extinct. The mimics 
still retain their mimetic pattern and colouring, but the selective 
process which produced this type can no longer (in the absence 
of the model) be regarded as in active operation. Neverthe- 
less, the mimetic disguise is retained by virtue of heredity, and 
appears in the last stage of the ontogeny. Your correspondent 
will, I hope, pardon any apparent discourtesy if I state that the 
correspondence on this subject is, so far as I am concerned, now 
closed. I can assure him that he shall hear more about the 
results of Prof. Weldon’s measurements at no very distant 
future. Rk. MELDOLA, 
Measurements of Crabs. 
Mr. CUNNINGHAM, in NATURE of October 29, raises a doubt 
as to the trustworthiness of the results of my measurements of 
crabs. on the ground that the specimens of the year 1893 had 
been longer in spirit than those of 1895. 
The value of measurements made on animals preserved in 
spirit naturally depends on the animal so preserved ; and to any 
one who is acquainted with the rigid nature of the calcareous 
carapace of Carctnus muenas, it seems difficult to admit the pos- 
sibility of distortion from such a cause. All specimens that were 
not quite rigid (a condition due to the animal having moulted 
just Defore capture) were rejected as a matter of course. 
The fact that a deficiency in one dimension was ‘‘com- 
pensated” by an excess in the other dimension, so far from 
suggesting to Mr. Cunningham a suspicion of the specimens 
having ‘* undergone an artificial change of shape,” would, I 
think, have seemed not only natural, but a feature to be 
expected, had he realised that every crab during its growth 
passes through a similar change of shape, in which the relative 
Increase in size of the one dimension (the dentary margin) is 
accompanied by a corresponding relative diminution in size of 
the other dimension (frontal breadth). The table appended to 
the paper in question shows that, while the dentary margin in 
the youngest crabs measured increased on an average from 414 
thousandths of the carapace length to 498 thousandths in the 
adult, the frontal breadth, on the other hand, diminished in 
relative size from 813 thousandths in the young to 595 thou- 
sandths in the adult; so that a compensatory relation between 
these two parts of the hard skeleton is a normal phenomenon in 
these animals. 
I would venture to add that observations in recent years, on the 
variability of both animals and plants, tend to show that species 
are much more unstable than was once supposed. The changes 
recorded in these crabs in the space of two years are very minute, 
but they are persistently in one and the same direction in all the 
twenty-six stages of growth measured. If such a rate of change 
exceed what we should have expected, it may be because hitherto 
we have had no collections of comparative facts on which to base 
any reasonable expectation. H. THomrson. 
October 30. 
NO. I411, VOL. 55] 
I Am able to offer Mr. Cunningham direct experimental 
evidence that the hard carapace of a young shore-crab is not 
sensibly distorted by immersion in spirit for six months. 
I have lately had occasion to compare the mean frontal 
breadth in three samples of young female crabs, gathered from 
the same locality in Plymouth Sound. The range of size was 
the same in all the samples. The samples were (1) a large 
number of individuals, collected in 1892 and 1893, measured 
after immersion in spirit during a period varying between six 
months and a year: the mean frontal breadth in these crabs was 
compared (2) with that of 569 individuals, collected during the 
summer of 1895, and measured fresh while still wet with sea- 
water ; and (3) with that of 595 crabs, gathered also during the 
summer of 1895, but preserved in spirit for about six months 
before measurement. 
The mean difference between the frontal breadth in the spirit 
specimens of 1892-3 and that of the fresh specimens of 1895, was 
1°47276 of Mr. Thompson’s units; the difference between the 
crabs of 1892-3 and the spirit specimens of 1895 was 1°58992 units. 
The difference between the two results obtained from crabs of 
1895 was therefore only 0°11716 unit—a difference so small, com- 
pared with those observed by Mr. Thompson, that it may safely 
be neglected, even if it be assumed to be due entirely to the spirit, 
and not ,to be within the probable error of the determination. 
As for the distribution of deviations from the mean in the two 
samples, it would take too much space to compare them in detail 
here; I hope to do so before very long in connection with 
another problem, and therefore I will only now say that the 
“*standard deviation ”’ (= error of mean square) of the series of 
fresh shells was in Mr. Thompson’s units 11°7955, that of the 
spirit specimens being 1179628 of the same units. 
These figures seem to me sufficient to justify Mr. Thompson’s 
conclusion ; and it must therefore be held that a change, whether 
of oscillation or of evolution, is going on at a measurable rate 
among Carcénuzs menas in Plymouth Sound. 
The very great importance of this result to all students of 
animal evolution is evident; for the careful study of cases in 
which change is actually going on so rapidly that it can be 
watched and measured in a reasonably short space of time will 
assuredly be found to give the best, if not the only clue to the 
process of evolution in general. W. F. R. WELDON. 
University College, London, October 31. 
The X-Rays produced by a Wimshurst Machine. 
IN my’two papers of June 4 and June IS respectively, I proved 
the existence of non-homogeneity in the X-rays, and gave a 
simple method of strengthening and maintaining the discharge 
of these rays from an ordinary focus tube by means of a con- 
ductor wrapped round the part of the tube level with and 
behind the kathode, and separated by a small sparking gap 
from either an earthed wire or the kathode’s external wire loop, 
or the wire leading from the induction coil to the kathode itself. 
Since then my attention has been chiefly turned to the most 
remarkable results which can be attained with a Wimshurst 
machine, with which this paper is concerned. My Wimshurst 
has two 15-inch plates, and, not being of the latest type, has 
the old-fashioned metallic sectors and buttons, instead of the 
plain varnished glass disc, though I do not imagine that the 
results will be found materially different with the simpler, and 
probably better, form, weer 
At first I drove the machine by hand, and hand driving is 
sufficient for the results mentioned in this paper—a most im- 
portant fact, seeing that it renders unnecessary any other form of 
engine than the hand, and brings the copious production of any 
kind of X-rays within the reach of those who have neither 
battery nor dynamo—afterwards, for securing greater personal 
freedom and uniformity of turning force, ¥ drove the Wimshurst 
by a small motor, an easy and most convenient plan, using a 
platinum electrode acidulated water resistance by which the rate 
of rotation is easily governed. 
To me, accustomed only to the effects given by a 44” spark 
induction coil, the brilliancy of the shadows given by simply 
connecting my tube’s electrodes to the brass knob terminals of 
the Wimshurst used without condensers was surprising, and the 
steadiness of the image a most grateful rest to the eyes. The 
bones seemed more transparent than I had ever seen them, and 
though at first I thought this might be only the effect of the 
lessening contrast brought about by increased general illumina- 
tion, this does not prove to be the case. The two faults to be 
