1899] EVIDENCE OF NATURAL SELECTION 43 



captivity. Some of the specimens grew at nearly the same rate, that 

 is to say, the increase in size was nearly the same after the same 

 number of moults. But in other specimens this was not the case. 



In August 1883 he obtained five specimens in the Megalopa stage. 

 He gives the dates of the moults of two of these, and the sizes after 

 each moult. These specimens he distinguishes as A and B. The 

 latter, B, was later in moulting than A ; its fifth moult (after it had 

 attained to the crab-form) occurred on April 23, 1884, while A 

 moulted for the fifth time on March 8. The crab, after the fifth 

 moult, was in the case of B 5*7 mm. long, in the case of A 4 - 44 mm. 

 long, a difference of 1*3 mm. In February both of these crabs had 

 moulted four times, and there was a difference of *34 mm. between 

 them. This proves that crabs of different sizes on a given date in 

 some cases have moulted the same number of times. Supposing the 

 change in relative frontal breadth to be constant, or to be less variable 

 than the increase in size, then these two crabs of different sizes would 

 have nearly the same relative frontal breadth. To take another case 

 from the same paper. A crab N was 12-5 mm. long in July 1882. 

 After five moults, in the following May, ten months, it was 37 mm. 

 long. Another crab Y was 12 mm. long in September 1882; after 

 five moults, in the following June it was 45 mm. long, the moults 

 having only occupied nine months. Here we have a difference of no 

 less than 8 mm. in the carapace lengths of two crabs after the same 

 number of moults, although their original difference was "5 mm. In 

 one case the increase was 24'5 mm., in the other 33 mm., a difference of 

 8 '5 mm. in the increment. This shows how great may be the difference 

 in the amount of growth after a given number of ecdyses. Supposing 

 the larger crab did not cast its shell again for a month, we have the 

 above difference of growth after not only the same number of moults, 

 but after the same interval of time. If we look at the measurements 

 recorded, we find that after four moults the crab Y was nearly of the 

 same size as the crab N after five moults. Supposing then that the 

 change in frontal breadth at each moult is nearly constant, is not 

 proportional to the change in size, then the crab Y, though of the same 

 size as N, was one moult behind it, and therefore of distinctly greater 

 frontal breadth, since the relative frontal breadth decreases at each 

 moult. Thus, if the crabs in 1893 in Plymouth Sound grew as much 

 in a given moult as Y, while in 1895 they grew at the rate of X, then 

 the crabs of the same size in the two years would represent different 

 numbers of moults. Unfortunately Mr. Brook did not measure the 

 frontal breadths, and therefore I cannot carry the argument further, but 

 it seems to me extremely probable that the change in the frontal 

 breadth of crabs which reached a given size in four moults, would not be 

 the same as in those which reached it in five moults. If that is correct, 

 all Professor Weldon's evidence proves is that the crabs grew more at 

 each moult in 1893 than in 1895, and the evidence he has produced 



