206 W. K. BROOKS. 



which ultimately spring from the progeny of a single egg is very 

 great indeed and practically unlimited, and yet when these buds 

 produce reproductive elements they are of the same sex in all the 

 members of the colony. 



This fact is very remarkable when we bear in mind the length 

 of the path from the egg to the sexual bud, and the complicated 

 character of the alternation of generations in hydroids. 



One of the most constant results of a sedentary or fixed habit 

 of life is hermaphroditism ; and its influence in this direction is "so 

 potent that it has, in the barnacles, broken down one of the most 

 ancient and persistent of all the characteristics of animals, the 

 separation of the sexes of arthropods. 



If the sessil habit of hydroids were primitive we should certainly 

 expect them to be hermaphrodites and the potential unisexuality 

 of each hydroid cormus would be unintelligible ; although it is 

 easily understood if the cormi have arisen by the asexual multipli- 

 cation of the larvae of unisexual locomotor adults, for in this case 

 each larva, before it began to form cormi, must have been potentially 

 either a male or a female, and all its progeny, produced by budding, 

 would naturally inherit th* same sex, so long as sex remained po- 

 tential and was not called into activity, for there is no reason why 

 the sedentary habit of larvse should be followed by hermaphro- 

 ditism in locomotor adults, nor is there any reason why the sex of 

 larvae should be modified by a sessil life so long as sex remains 

 latent or undeveloped in the bodies of these larvae during the sessil 

 stage. It is remarkable that those species in which the gonophores 

 are most degenerated should adhere to the same law, but this fact 

 shows the firm hold which the separation of the sexes has taken 

 upon the organization of hydroids, and is to my mind one of the 

 most conclusive proofs that the sessil cormus is secondary. 



In the third place we must remember that the craspedota with a 

 sessil hydroid cormus form only a part of the group ; that the 

 trachoraedusae in which, so far as we know, the hydra larva is 

 always a free, floating solitary animal, are numerous and diversified, 

 and that, while the life history of the various narcomedusae has 

 been greatly modified by parasitism, they may all be reduced to a 

 type in which each egg gives rise to a simple floating hydra-like 

 larva which grows up into an adult like the larva of a trachomedusa. 



