252 c. iSHiKAWA : 



and Equisetum the same author ('98) still hokis this view as to 

 the nature of the stainahle body, but is inclined to consider it 

 as representing the centrosome-like body in the sense of Ikeno 

 ('98, a). Webber ('07, c) in Zamia, Ikeno ('97, a) in Cycas 

 and HiRASE ('97, a and '08, c) in Ghikgo liave observed similar 

 modes of formation of the cilia in the spermatid cells. The 

 extremely large centrosome-like body becomes spirally twisted 

 and eventually gives rise to the 2^oi"t of attachment of the 

 cilia. But while Ikeno identifies it with the centrosome and 

 HiRASE with the attraction-splicre, Webber considers it to be 

 quite a distinct organ of the prot02'>lasm of spermatic cells, 

 having for its function the formation of the cilia and called by 

 him the '' blephaloplast " ('97, d). Shaw ('98, d) came to the 

 same conclusion in his studies on the spermatogenesis of Onoclea 

 and Marsilia, and denies the centrosomc-nature of the body in 

 question. 



Hermann ('97, h) in his recent work on the spermatogenesis 

 of Selachia and Salamandra considers the middle-piece of the 

 spermatozoa to be an extremely elongated centrosome and that 

 the tail develops out of the substance of the central 

 spindle fibres. He has thus arrived at the same conclusion as 

 I had done four years ago ('î)4, c) with respect to the formation of 

 the flagellum in Noctiluca s|)ores, and which I have again tried 

 to prove distinctly in the above description, on the ground of 

 renewed observations. There is, however, a marked difference 

 between Hermann's case and mine. For while in Noctiluca, as 

 already given, the fibres of the central spindle are directly 

 transformed into the flagellum, in Selachia and Salamandra these 

 return first to their earlier homogeneous condition, and then 

 out of this homogeneous mass the tail is developed (p. 308). 



