74 bulletin: museum of comparative zoology. 



side can be distinctly seen (Figs. 107, 108). It appears as though there 

 were a vacuole of less deeply stained material in this region. Occasion- 

 ally smaller vacuoles are pr*esent in all parts of the head (Figure 106). 

 Furthermore in all cells in which the decolorizing process has been 

 carried sufficiently far, a lighter area is seen at the base of the head 

 (Figs. 107, 108). 



This, from one point of view, presents the appearance shown in Fig- 

 ure 107, whereas, viewed in another direction, it appears as in Figure 

 108. Here the two lateral centrosomes and the end knob are very 

 plainly visible. In fact, they more closely resemble typical centrosomes 

 than do the same bodies in the stages immediately preceding (Figures 

 104, 105). There can be little doubt of their identity, even at stages 

 when numerous similarly staining chromatin granules are present in 

 other parts of the nucleus. In Figure 108 there can be no doubt 

 whatever as to the nature of these bodies, for none of the chromatin of 

 the nucleus takes on this form. Furthermore, each of the lateral cen- 

 trosomes is surrounded by a fairly distinct clear area and is connected 

 with the centrosorae at the base of the filament by a distinct fibre. 

 When viewed at right angles to this, the lateral centrosomes are usually 

 not visible (Fig. 107), but when they can be seen they lie at such dif- 

 ferent levels that both do not appear at one focus. 



The head of the young spermatozoon continues to elongate and soon 

 assumes the appearance represented in Figure 109. By careful decolor- 

 izing, all of the chromatin exhibits a paler grayish color, while the cen- 

 trosomes are readily distinguishable both by reason of their position and 

 by their stronger retention of the stain. At this stage the base of the 

 nucleus forms a plane perpendicular to the long axis of the nucleus, and 

 the two lateral centrosomes are so closely apposed to the membrane that 

 each is flattened out into a hemispherical body. These two bodies lie 

 side by side, quite close together, but each is distinct from the other 

 (Fig. 109). The connection between intranuclear centrosomes and the 

 one at the base of the axial filament is very evident, and presents the 

 appearance of a number of fibres forming a sort of membrane rather 

 than that of a single thread. 



At this stage (Fig. 109) a peculiar cuff like structure, apparently at- 

 tached to the posterior margin of the head, is often to be observed. 

 What its function may be I cannot say, for I have been able to find no 

 such structure in the more advanced stages. Its origin and history are 

 of interest, as at an earlier stage it is similar to structures described in 

 other myriapods (Tonniges :02) and in the guinea pig (Meves, '99). 



