34 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY. 



the same character as at other times. The most reasonable explanation 

 of the cytoplasmic changes is the one I have already suggested. In 

 the early prophase the archoplasm is dissolved in the hyaloplasm, and 

 in this condition is present in all parts of the cytosome. During the 

 rest of the prophase it remains in this condition, the only evidence of 

 its presence being the deeper stain which the inter-reticular substance 

 at this time assumes. However, at the time of the establishment of 

 the division figure this substance again appears, now in its kinetic, 

 fibrillar form, and continues to become more marked, until, in the 

 period of greatest cellular activity, during the anaphase and telophase, 

 every part of the cell is penetrated by these archoplasmic fibres. In 

 the metaphase and succeeding stages these threads are much more 

 numerous in the region of the centrosomes than elsewhere in the cell 

 (Figs. 32 et seq.). Many of them at these points are apparently short, 

 extending only a little distance into the cytoplasm. Indeed at this 

 period these small fibres form the outer zone of the centrosphere 

 (Plate 4, Figs. 49, 50). 



Of the behavior of the centrosome during these stages I will speak in 

 detail later, but would like to describe at this time its general appearance. 

 During the metaphase and early anaphase the centrosome is a bipartite 

 body of an irregular dumbbell shape contained in, or suiTounded by, a 

 centrosphere composed of two layers or zones (Figs. 50, 51). The inner 

 of these zones stains in a fairly definite manner, but much less deeply than 

 the centrosome itself. It is of a homogeneous or very finely granular 

 consistency and of an oval, or later of a lobate form. The outer zone, 

 as I have said, is formed by the closely apposed bases of the astral rays, 

 and has no distinct outline. In preparations stained in the ordinary 

 manner with iron-haematoxylin, it is darker in color than the inner 

 zone, but upon further decolorizing, this dark stain is lost and it appears 

 lighter. In the deeply stained preparations the spaces between the fibres 

 are filled with masses of the stain, but in the lighter-colored sections these 

 deposits are removed and the fibrillar nature of this zone is disclosed. 



The longitudinal axis of centrosomes and centrosphere bears no con- 

 stant relation to the direction of the spindle during the metaphase and 

 anaphase, but during the telophase is always parallel to the plane of 

 cell cleavage. 



The spindle as first formed is of symmetrical shape (Plate 8, Fig. 145), 

 but very soon changes begin which destroy this symmetry (Plate 3, 

 Fig. 33). When it first appears it extends at right angles to the length 

 of the cell, but the poles of the spindle soon show a tendency to move 



