r,i02.] XATL'RAL SCIENXES OF PHILADELPHIA. 203 



cycle, which is followed by another like it. This germiual 

 cycle too must have had a commencenieiit, and, as I have attemp- 

 ted to show elsewhere, the period of fertilization (conjugation 

 of ovum and spermatozoon) is to be considered the starting point. 

 In the case of change of generation for the particular objects 

 under examination, the Hydrozoa, the polyp stage is generally 

 regarded as the beginning, and the medusa as the end of the 

 cycle. This has its justification, for the medusa is more complex 

 in structure than the polyp, that is, has passed through a longer 

 ontogenetic stage. Here, too, however, the cycle really begins with 

 the fertilized ovum, and ends with the stage that produces ova, if 

 we would draAV a parallel to the gei-minal cycle of INIetazoa. In the 

 metagenetic Hydrozoa it is the medusa which produces the ova. If, 

 on the other hand, greatest complexity of stnicture should not coin- 

 cide with the end of the reproductive cycle, such complexity would 

 of course not be a criterion of this stage of the cycle, which never- 

 theless should be the one classified. Such a case would be an 

 organism with, change of generation, but with a regressive develop- 

 ment in the latter part of its ontogeny (as might be induced by 

 parasitism), so that the terminal stage would be less complex than 

 some preceding one. No such case is known, however; so that with 

 the facts at hand we are justified in concluding that greatest com- 

 plexity of structure is the test of the last stage of the cycle in alter- 

 nating generations ; and it is that last stage which is the one to be 

 classified. 



IV. The Classification of Polymorphic Individuals. 



In the preceding section have been considered such polymorphic 

 individuals as succeed each other in alternation of generations In 

 many species are found together in the same generation individuals 

 which differ from each other, as either in sexual dimorphism or in 

 polymorphism. The former shows the distinction singly of males 

 and females, and is found in dioecious species ; while the latter shows 

 the males and females in several different forms (as the polymorphic 

 Hymenoptera). Polymorphism also may show in the same species 

 a distinction of sexual from asexual individuals, as in certain corms. 

 Sexual dimorphism and polymorphism are the results of a division 

 of lal)or between the individuals of a species, as the structural 

 differences in the parts of one organization are due to a similar 



