Chap, xiv.] ALTERNATION OF GENERATIONS. 543 



have been variously regarded as grades of development 

 or oi degradation. 



We find, that is, that the medusiforni buds do not 

 always become separated from the stock that has 

 produced them ; and while in some cases (e.g. Syn- 

 coryne itself, towards the end of the breeding season, 

 or Tubularia) the buds are fully formed medusae, in 

 others, though still bearing the sexual organs, they 

 are nothing more than projections from the sides of 

 the body, in which the medusoid characters are hardly, 

 if at all, apparent (Hydractinia). These stages of 

 difference in the medusoid buds are allied, 011 the one 

 hand, to the condition which obtains in the common 

 Hydra, where ova and spermatozoa are developed in 

 one and the same individual, and in which the young 

 do not pass through any larval stage ; and on the 

 other, to what is seen in Geryonia, for example, where 

 the hydriform condition isaltogether suppressed, and the 

 larva, after a certain amount of metamorphosis, passes 

 into the medusoid condition of its parent. In both of 

 these cases there is no alternation of generations. 



A series of very interesting conditions are exhi- 

 bited by different Annelids. In Lumbriculus there 

 may be simple transverse division of the body, one 

 half of which acquires a new tail, and the other a new 

 head ; in Ctenodrilus it has been observed that the 

 anterior half of the body may again divide ; in Syllis 

 the generative products are developed in the posterior 

 half only of the body ; in Myrianida the same pro- 

 ducts are confined to the forms that arise by budding, 

 so that from a simple case of transverse division we 

 have come to a complete example of alternation of 

 generations. 



In some cases (e.g. the fresh-water Nais) there is 

 not simple transverse division, but the formation first 

 of all of a so-called " zone of gemmation ; " here the 

 zone becomes converted anteriorly into an anal zone, 



