40 



DISCONTINUOUS VARIATION. 



sandths, and a probable error of 12.06, witli 58.55 per 

 cent, of individuals of mean breadth 654.66 and prob- 

 able error 8.41 (i. e., groups of individuals represented 

 by the two lower dotted line curves in the Figure), the 

 upper dotted line curve would be obtained. It will be 

 seen that this corresponds very closely with the ob- 



40- 



90 



SO 



GO 

 50 



40 



30-r 



-SO- 



SO 

 20- 



10- 



580 590 600 610 620 630 640 650 600 670 680 6.90 TOO 



FIG. 9. Distribution of frontal breadths of Carcinus mcenas. 



served values, and so supports Weldon's hypothesis. It 

 is somewhat curious that of all the 22 series of measure- 

 ments made by Weldon on Naples and Plymouth crabs, 

 this was the only characteristic in respect of which 

 dimorphism was exhibited. As an explanation of it, 

 Giard * has suggested that one of the two groups owed 

 its altered frontal breadth to the presence of an internal 

 parasite, Portunion mcenadis. Thus he measured five 

 specimens of C. mccnas infested by this parasite, and 

 found that their mean relative frontal breadth very 

 * Comptes Rendus, cxviii. p. 870, 1894. 



