196 Mr. H. Thompson. On certain Changes observed in the 



every 1 mm. of growth in carapace length, or twenty-five groups 

 for the whole range of 10 15 mm. carapace length. The numbers 

 contained in the separate groups ranged from seventy-two in the 

 smallest group to 178 in the largest group. The arithmetical 

 mean and mean error in each group is set out in Table I infra. 



Similar measurements were made in the case of 1,957 young male 

 C. mcsnas from Plymouth of the year 1895. These were likewise 

 divided into groups differing by 0*2 mm. of carapace length : and 

 the numbers contained in the twenty-five groups between 10 and 

 15 mm. carapace length ranged from thirty-four in the smallest one 

 to 111 in the largest. The arithmetical means and mean errors are 

 given in Table I infra. 



On comparing the two sets of measurements (expressed in terms 

 of the carapace length which was taken as the unit) it appears, as 

 regards the " frontal breadth," that in every one of the twenty-five 

 groups without exception the average size of the frontal breadth in 

 the 1893 crabs exceeded that of the 1895 crabs of corresponding size. 

 Seeing how small the groups are the result is a striking one, and is 

 given in greater detail in the following Table : - 



C. mcenas. Frontal Breadth. 



On the other hand, if the species in 1895 has a smaller average 

 frontal breadth, it compensates for the deficiency by having a larger 

 right dentary margin. This was found to be the case in twenty- three 

 out of the twenty-five groups, the two non-conformist groups lying 

 one near each end of the range. The arithmetical means and mean 

 errors are given in Table I infra, and the results, tabulated in a 

 corresponding form to those of the frontal breadth measurements, 

 are as follows : 



