606 F. A. POTTS. 



Also in young (tliougli even tlien sexually mature) hermit 

 crabs of I'O — I'S cm. of carapace length, the internal ramus 

 is less developed than in the adult, and everything points to 

 the conclusion that the biramous character of the abdominal 

 appendage has been acquired to provide a more secure anchor- 

 age for the developing embryos. For this purpose the basal 

 joint and the internal ramus, both in the middle and apical 

 part, are provided Avitli long, radiating bunches of hairs. 

 Of the bunch on the basal joint only is there any representa- 

 tive in the male. But though the external ramus is well 

 provided with hairs it is noticeable that eggs are never 

 attached to them, so that this ramus is left quite free to sub- 

 serve its proper function of aerating the interior of the shell, 

 for which it is admirably adapted by its oar-like shape and 

 border of stiff hairs. 



In the male the external ramus is developed as in the 

 female and is used in the same manner; but in correlation 

 with its non-participation in the nursing functions, the internal 

 ramus is rudimentary in the first four appendages. 



Account of Modification in the Male. 



The effect of the parasitism of Peltogaster on the secondary 

 sexual characters of the male E. meticulosus is to stimulate 

 the development of the rudimentary internal ramus of the 

 first three abdominal appendages. The extreme members of 

 the modified series reach a stage indistinguishable from the 

 corresponding normal female appendage. The modification is 

 rendered complete by the assumption of the typical gibbous 

 form and the bunches of egg-bearing hairs of the internal 

 ramus of the female. 



Various stages of the complete series which exists between 

 totally modified and unmodified crabs are here figured (pi. 33, 

 figs. 1 — 5). An examination of these shows that, as in the 

 development of the female, so in the modification of the male, 

 it is in the first appendage that the internal ramus earliest 

 responds to the stimulus (iig. 3). And though in some cases 



