292 bulletin: museum of comparative zoology. 



ceding the formation of the spermatozoa are always indirect, and this 

 conclusion is confirmed, for sponges, by Fiedler ('88) ; but it was not un- 

 til very recently that Aders (: 03) was able to demonstrate that, after the 

 formation of the last generation of spermatogonia, the number of such 

 divisions is invariable, there being formed here, as in other groups, never 

 more nor less than two generations of spermatocytes, and an ultimate 

 generation, the spermatids. 



The earliest detailed observations on the metamorphosis of the sper- 

 matids are those of Pictet ('91), who describes the process in the case of 

 various siphonophores. He observed the formation of a " Nebenkern " from 

 a condensation of cytoplasmic granules, and the change in chemical 

 composition of the nucleus. 



Our present knowledge of the anatomy of the adult spermatozoa of 

 coelenterates is largely due to the researches of E. Ballowitz ('94), who 

 describes them, in medusae and actinians, as being of the ordinary fla- 

 gellate type. But, in the actinian Tealia, he found an interesting devia- 

 tion, in that the adult spermatozoa retain much of the appearance 

 of their parent spermatids, thus recalling the conditions described by 

 Pictet ('91) for Halistemma. The observations of Ballowitz have been 

 extended by Retzius (=04, -.05), who has described the spermatozoa of 

 many other coelenterates. He has, however, made no contributions to 

 their histogenesis, and it remained for Gorich (: 03 a , : 03 b , : 04) to study 

 this process, in Sycandra and Aurelia. Gorich's most important ob- 

 servations relate to the role of the centrosome in the metamorphosis of 

 the spermatid, and will be fully discussed later. 



Of special importance in this connection are the researches of Guenther 

 ( :03 a , : 04) and Downing ( : 00, : 05) on the nuclear changes in the sperma- 

 togenesis of Hydra. Guenther has described a " synapsis " phase in the 

 primary spermatocytes, and has devoted much study to the nucleolus in 

 Hydra viridis, while Downing has given a painstaking account of the 

 entire course of spermatogenesis in the closely allied species H. fusca. 

 To both of these researches I shall have occasion to refer so repeatedly 

 that there is no necessity of summarizing them here. 



3. The Oocyte Nucleus. 



Kleinenberg ('72), in his classic studies on Hydra, laid the founda- 

 tions for all more recent investigations of the oocyte nucleus by the 

 discovery that its ground substance is not homogeneous, as early in- 

 vestigators had supposed, but consists of two distinct materials. This 



