BLACKMAN : THE SPERMATOGENESIS OF SCOLOPENDRA. 39 



and is also frequently more or less curved, forming structures which 

 resemble quite closely the centrosomes figured by Boveri ( :Ol). At 

 about this stage (late metaphase) the centrosphere reaches the height 

 of its development. It is usually at this time larger than at any other 

 period of cell division and is quite definitely marked off from the other 

 elements of the astral system. In many of the cells of this stage, as in 

 those preceding, some at least of the astral rays can be seen to pass 

 through this area and come into contact with the centrosome. This 

 can be seen only in those cells which have been decolorized more than 

 usual, but it probably exists in all well fixed material of this species. In 

 the late metaphase the centrosphere, which up to this time has been of 

 an oval shape (Fig. 50 a), also shows evidences of the approaching 

 division. There appeal's in its middle region a constriction which pro- 

 ceeds until the structure in question assumes a distinctly bilobed form 

 (Plate 4, Fig. 51 ; Plate 8, Fig. 146; Plate 9, Fig. 148). This change 

 is followed by the division of the centrosome and the further constriction 

 of the centrosphere. At this period we have at the centre of the aster 

 two masses of archoplasm (centrospheres) which are nearly perfectly 

 spherical in shape, but still adhere to each other upon one side. In the 

 centre of each of these spheres is a small darkly stained granule of 

 irregular shape, the centrosome. 



Thus in the anaphase is completed the division of the centrosome, the 

 first indication of which was made apparent in the mid prophase. From 

 the length of time required to accomplish this division, as well as from 

 the great variety of stages exhibited, we should seem to be justified in 

 concluding that it is of importance, and that the division is an accurate 

 one as regards bulk at least. Such, however, does not always appear to 

 be the case. In Figure 50 b is shown a stage of centrosome division in 

 which the two lobes are apparently very unequal in size. "Whether such 

 a figure would eventually result in an unequal distribution of the cen- 

 troplasm or not, I cannot say, but I have seen no completely divided 

 centrosome in which the two parts are markedly unequal. It is possible 

 that the substance is later rearranged so as to result in an equal division. 



Another poiut to be noted in considering the centrosome is the irreg- 

 ular form which it exhibits at various stages. This would seem to indi- 

 cate that this body is not a simple granule, as it is often considered, but 

 that it is composed of an aggregation of granules. This would, I believe, 

 serve to distinguish it further from the centriole as reported in many 

 other kinds of material, as this structure is, I believe, almost universally 

 said to consist of either one or two simple granules. 



