BLACKMAN : THE SPERMATOGENESIS OF SCOLOPENDKA. 69 



ff.it.ed into one mass near the nucleus at the hase of the axial filament. 

 However, at ahout the time when the elongation of the cell begins (Fig. 

 92), small portions become detached from the main mass, and some of 

 these migrate around the nucleus and come to rest at a point directly 

 opposite the base of the filament (Figs. 92, 93, 94 et seq.). This archo- 

 plasmic material, which is the first indication of the acrosome, that later 

 becomes so prominent, has at this time the form of a number of small 

 spherical masses. These are rather dense and granular, and are stained 

 more strongly by the Congo red than any other structure of the cell. 

 By succeeding changes, which are very interesting, they become altered 

 from their granular consistency into a large vesicular structure which is 

 characteristic of the young spermatozoa, and later into the long filament- 

 like acrosome of the mature spermatozoon. 



Apparently the first stage in this transformation is shown in Figure 

 98. This presents an appearance quite similar to the early stages in 

 the formation of the acrosome in Amphiuma as figured by McGregor 

 (98) ; but, as we shall see, the later stages are entirely different from 

 those shown for this amphibian. The changes in Scolopendra are as 

 follows: Each of the archoplasmic bodies at the anterior pole of the 

 nucleus becomes enveloped in a mantle of clear substance (Fig. 98) 

 which has a sharply defined boundary toward the surrounding proto- 

 plasm. "With the increase in thickness of this mantle, there is a gradual 

 reduction in the size of the granular mass, until the latter has entirely 

 disappeared and there remains in its place a clear vesicle with distinct 

 smooth walls (Figs. 95, 96). Sooner or later these vesicles unite with 

 one another, forming a larger and larger vesicle (Figs. 95, 97, 99). At 

 first this is very irregular in shape (Fig. 97), and often shows evidence 

 of its origin from a number of smaller vesicles. Thus in Figure 99 the 

 fusion is still imperfect, portions of the vesicle walls still persisting as 

 partitions. However, the fusion becomes more and more complete 

 (Figs. 99-104), until, as shown in Figures 105 and 106 (Plate 7), 

 there is no evidence that the acrosome has arisen from a number of 

 vesicles. 



"When the elongation of the cell begins, changes also occur within 

 the nucleus preparatory to its condensation and lengthening. The first 

 observable alteration affects the arrangement of the chromatin. "We have 

 seen that at the time of the migration of the centrosome from the cell 

 membrane toward the nucleus, all, or nearly all, of the chromatin bodies 

 collect upon the side of the nucleus nearest the centrosome. This aggre- 

 gation of the chromatin persists until the axial filament has attained a 



