96 Woodburn. — Spermatogenesis in B Iasi a pusilla , Z. 
fact in connexion with the study of such plants as are so noticeably affected 
by the pressure or lack of moisture. In order to draw conclusions which 
are not the result of observation of artifacts, the material should show some 
signs, at least, of decided activity. 
Cytology of Spermogenous Tissue. 
A resting condition of the antheridial cells frequently found is re- 
presented in Fig. i. The cytoplasm is finely and evenly granular. The 
nuclear cavity surrounding the nucleolus is filled with an extremely fine 
granular substance, almost homogeneous in appearance. A somewhat 
lighter space immediately surrounds the nucleolus. The latter is quite 
sharply defined, and stains very dark and homogeneously. This condition 
seems to be typical of the tissues after remaining a considerable length of 
time in a resting condition. As the active conditions arise, and the nucleus 
passes through the prophases of division, a fine network appears around the 
nucleolus, with lumps of chromatin of very irregular size, which form at 
the intersections of the threads. A definite distinction could not always be 
made between chromatin and linin, the former being drawn out gradually 
into finely granular threads, forming a network in which the chromatin 
seems embedded (Figs. 2 and 3). The nucleolus stains more lightly as the 
chromatin lumps increase in size and finally disappears, as is the general 
rule during metakinesis (Figs. 2, 3, and 5). The nucleolus is normally 
present, except immediately preceding metakinesis and during early telo- 
phases. It may reappear in the spermatid soon after the last division 
(Fig. 10), where it remains differentiated only for a short time if the sperma- 
tids are in active development. The finely granular substance of the 
nucleus passes first into a delicate network, and then the threads of the 
network seem to be absorbed by the chromatin lumps as the latter increase 
in size and irregularity (Figs. 2, 3, and 5). A definite spireme has not been 
observed, but very clearly defined chromosomes result, either from such 
a formation or from the coalescence of the chromatin lumps. Because of 
the small number of dividing nuclei observed, an exact chromosome count 
was not obtained, but Fig. 4 suggests five or six as the probable number. 
In this and similar mitotic figures no evidence of centrosome-like bodies 
could be found. 
Formation of Sperms. 
The last division, like the previous ones, evidently occurs very quickly, 
judging from the number of stages found immediately preceding and 
following metakinesis and anaphase. Although not yet observed, the 
position and form of the resulting spermatids show that the spindle of the 
last division frequently has an oblique position. Fig. 7 follows very closely 
the last division. The cytoplasm has not yet divided, as is the case in 
