224 Papers from the Marine Biological Laboratory at Tortugas. 



been observed in the case of Stromhus do not support this view. In the first 

 place, there is not the slightest morphological evidence that the proximal 

 centrioles form rings; they clearly form nodes just beneath the membrane 

 and each one lies at the base of a flagellum. Then again, the flagella have 

 begun to shorten before there has been any appreciable lengthening of the 

 spermatosome in either direction; if any has occurred, it is certainly not 

 sufficient to account for the ver^^ noticeable shortening of the flagella that 

 has taken place. 



It remains to be pointed out that there exists a striking similarity 

 between the apyrene spermatosome (as regards its centrosomal structure 

 at the time of the early growth of the axial fibers) and the well-known 

 ciliated epithelial cells of the Mollusca. In figure 35, for example, the 

 proximal centrioles can well be compared with the so-called basal granules 

 of the ciliated epithelial cells and the distal centrioles with the inner granules. 

 The slender fibers growing out from the distal centrioles are analogous to 

 the fibers which extend across the epithelial cell. With the progressive 

 growth of the axial fibers across the spermatosome the resemblance dis- 

 appears to a great extent. While a strict homology between the two 

 structures can not be maintained, we have here a strong, if indirect, support 

 for the Lenhossek-Henneguy theory of the centrosomal origin of the basal 

 granules found in a ciliated cell. 



Several changes occur in the spermatosome about the time that the distal 

 plate of centrioles reaches the cell-membrane. At this time or perhaps a 

 little earlier, the chromatic vesicles begin gradually to lose their sap, so 

 that they decrease in size and the chromatic granules come closer together 

 (figs. 43-45). As the process continues the chromatin gathers on one side 

 of the vesicle which is now becoming more and more reduced in size (figs. 

 46, 47, 48, and 49). Soon the membrane disappears, either by being dis- 

 solved or by the total loss of the sap, so that the whole structure shrinks into 

 a solid mass. In either case, what were formerly active chromatic vesicles 

 are now reduced to degenerating lumps of chromatin (fig. 63) ; eventually, 

 all of these are dissolved in the cytoplasm and disappear, leaving no traces 

 in the adult spermatozoon. The whole process of the degeneration of the 

 vesicles can be followed through in the living cells (figs. 1-6). 



The direct eff^ect of the degeneration of the chromatic vesicles is found 

 in the appearance at this time of the secreted albuminous bodies. Just as 

 in some epithelial cells where the formation of the secreted granules is 

 accompanied by the degeneration of the nucleus, so here the processes 

 leading to the eventual disappearance of the chromatic vesicles go hand 

 in hand with the secreh'on of the albuminous bodies. A similar corre- 

 lation between the degeneration of the nucleus and the secretion of 

 the albuminous bodies has been described for the so-called nurse-cells of 

 Littorina (Reinke, '12). In Strombus the albuminous bodies appear first 

 in the anterior portion of the spermatosome; a few can always be found 

 around the advancing end of the bundle of axial fibers by the time it has 



