GENERATIVE OKGANS. 611 



(2) The generative products of both sexes originate in the ento- 

 derm (hypoblast): Plumularia and Sertularel la, amongst the Hydroids, 

 and the whole of the Acraspeda and Actinozoa. 



(3) The male cells are formed in the ectoderm, and the female 

 in the entoderm : Gonothyra?a, Campanularia, Hydractinia, Clava. 



In view of the somewhat surprising results to which the re- 

 searches on the origin of the genital products amongst the Coelen- 

 terata have led, it would seem to be necessary either to hold that 

 there is no definite homology between the germinal layers in the 

 different forms of Coelenterata, or to offer some satisfactory explana- 

 tion of the behaviour of the genital products, which would not 

 involve the acceptance of the first alternative. 



Though it can hardly be said that such an explanation has yet 

 been offered, some observations of Kleinenberg (No. 557) undoubt- 

 edly point to such an explanation being possible. 



Kleinenberg has shewn that in Eudendrium the ova migrate freely( 

 from the ectoderm into the endoderm. and vice versa ; but he has 1 

 given strong grounds for thinking that they originate in the ectoderm. 

 He has further shewn that the migration in this type is by no means' 

 an isolated phenomenon. 



Since it is usually only possible to recognise generative elements 

 after they have advanced considerably in development, the mei'e 

 position of a generative cell, when first observed, can afford, after what 

 Kleinenberg has shev/n, no absolute proof of its origin. Thus it is 

 quite possible that there is really only one type of origin for the 

 generative cells in the Coelenterata. 



Kleinenberg has given reasons for thinking that the migration of the 

 ova into the entoderm may have a nutritive oViject. If this be so, and there 

 are numerous facts which shew that the position of generative cells is often 

 largely influenced by their nutritive requirements, it seems not impossible 

 that the endodermal position of the genei'ative organs in the Actinozoa 

 and acraspedote Medusae may have arisen by a continuously earlier migra- 

 tion of the generative cells from the ectoderm into the endoderm ; and 

 that the migration may now take place at so early a period of the develop- 

 ment, that we should be justified in formally holding the generative pro- 

 ducts to be endodermal in origin. 



We might perhaps, on this view, formulate the origin of the generative 

 products in the Coelenterata in the following way : — • 



Both ova and spermatozoa primitively oi'iginated in the ectoderm, but/7 

 in order to secure a more complete nutrition the cells which give rise to y 

 them exhibit in certain groups a tendency to migrate into the endoderm. 

 This migration, which may concern the generative cells of one or of both 

 the sexes, takes place in some cases after the generative cells have be- 

 come recognisable as such, and very probably in other cases at so early a 

 period that it is impossible to distinguish the generative cells from in- 

 diiferent embryonic cells. 



Very little is known with reference to the origin of the generative 

 cells in the triploblastic Invertebrata. 



39—2 



