﻿134 Proceedings of the Royal Physical Society. 



Since the various lots which go to make up the total of 

 5165 specimens here considered were collected at different 

 times, by different methods and in different places, they 

 cannot be regarded as forming a homogeneous group, and 

 therefore in dealing with sizes it is best to compare the 

 percentage of each sex at each separate length. (Had the 

 total been a homogeneous group, one would have preferred 

 to form Polygons of Frequency, which would indicate the 

 frequency of occurrence of individuals of different lengths 

 for (1) males, (2) females, and (3) males and females 

 together.) From the percentages, formed in the way in- 

 dicated and given in the last column of Table I., Diagram I. 

 has been constructed. It is certain that the character of 

 the curves in this diagram would be much altered for 

 animals of from 3 to 7 or 8 cms. in length if a sufficient 

 number of small specimens had been examined. It is 

 obvious, for example, that the curves are much too far apart 

 towards the zero end of the length axis, and that we are 

 dealing with incomplete series. What is wanted is an in- 

 vestigation into the relative numbers of males and females 

 produced, as well as a record of the relative rates of growth. 

 And there are other factors to be considered. We are 

 probably entitled to assume that the sexes begin life with 

 approximately equal numbers, and that the males not only 

 grow larger, but may very well grow more rapidly. It has 

 been found ^ that in the case of Oarcinus mcenas, " until a 

 carapace-breadth of 15 mm. is reached, male and female 

 crabs exist in equal numbers. After this the females begin 

 to preponderate, until for a carapace-breadth of 36-40 mm. 

 they are more than twice as numerous as the males. AVhen 

 the carapace-breadth exceeds 50 mm. the males outnumber 

 the other sex." Mr Punnett attributes this to a greater rate 

 of mortality among the males, owing to the more frequent 

 risks they run on account of their more active habits. But 

 too little is known of the effect of seasonal changes, sexual 

 activities, and times of moulting, as well as of breeding 



^ R. G. Punnett, Proc. Oamb. Phil. Soc, vol. xii., part iv., pp. 293-296 

 (1904). 



