150 
Charles E. Allen 
in the androcyte mother cells no longer display any trace of the activity 
which forraerly characterized them in relation to spindle formation; and 
of conrse the mere fact of a similarity in staiuing reactions of itself furnishes 
no valid reason for the identification suggested. It is also possible that 
the diminution in the amount of dark-staining kinoplasm has proceeded 
to the extent of its entire disappearance ; or, what is perhaps more likely, 
that the relatively small amount left in the cell is now less densely aggre- 
gated, so as not to be differentiated by its staining reactions from other 
cytoplasmic constituents. On the latter hypothesis, some of the formerly 
conspicuous kinoplasm may be still present in that portion of the apparently 
undifferentiated cytoplasm which, in the division of the androcyte mother 
cells as of their ancestors, usually collects rather densely in the neighbor- 
hood of the spindle poles. 
Besides the granules of irregulär occurrenee and distribution just 
referred to, the cytoplasm of each androcyte mother cell at a certain 
period of its development contains a single small, dense body (Figs. 66—68), 
for the present referred to as the central body, which is the center of 
a System of radiating fibers. This body lies in the neighborhood of the 
nucleus, sometimes (Fig. 66) apparently in contact with the nuclear mem- 
brane, commonly somewhat more remote (Fig. 67) but nearer to the 
nuclear membrane than to the plasma membrane. 
The central body is sometimes approximately spherical, sometimes 
(Fig. 67) more or less flattened in a plane parallel to the nearest portion 
of the nuclear membrane. The fibers which radiate from it are few in nuin- 
ber, well-defined though becoming slenderer and less easy to follow in 
their distal portions, straight or gently curved ; individual fibers can often 
be traced to the plasma membrane; seldom does one run directly toward 
the nucleus, but commonly, if not always, several fibers touch the nuclear 
membrane tangentially, thus forming a cone. In its general appearance, 
this astral figure is suggestive of those observed by Farmer and Reeves 
(1894) in germinating spores of Pellia. In some respects the best con- 
ception of the aster is gained from wliat may be called a polar view, 
like that represented by Figure 68, in which case the line of vision through 
the central body passes nearly through the center of the nucleus; the 
latter, lying in lower plane, is shown only in outline. The visibly un- 
differentiated cytoplasm is commonly most dense immediately about 
the central body, but there is no definitely outlined zone surrounding 
the latter. The astral rays can generally be traced to the central body, 
which is often slightly drawn out at the point of attachment, as though the 
ray were a pulled-out portion of the substance of the central body itself. 
