Embryo of Sciadopitys verticillata. 409 
a reticulum is really the effect of the separate threads interlacing and 
crossing with one another. This condition is shown in Fig. 17. 
During these changes in the chromatin there is a gradual enlargement 
of the nuclear cavity, presumably resulting from the increase in the amount 
of nuclear sap. As shown in Fig. 18, with the enlargement of the nuclear 
cavity the reticulated nature of the chromatin becomes completely lost, and 
the individual threads of which it is composed become sharply visible. 
A somewhat similar stage is shown in Fig. 19, but here the threads are 
isolating themselves from one another, and are spreading out into the 
enlarged nuclear cavity. It may be seen also from this figure that the 
chromatin threads are finely granular. The separation of the threads from 
one another continues until they are quite evenly distributed through the 
nuclear sap, and, as shown in Fig. 20, the threads become shorter and 
thicker, and their granular nature becomes more pronounced. This shorten¬ 
ing and thickening of the threads proceeds rapidly until we finally have the 
appearance represented in Figs. 21 and 22. A large number of preparations 
were made at this time showing all intermediate stages as well as those 
here figured. In none of them was I able to find any evidence that the 
chromatin consisted of a continuous thread. In fact, all of the evidence 
seemed to indicate the opposite, namely, that the number of chromatin 
threads in the early stages correspond with the number of chromosomes 
into which they later develop. And I am strongly of the opinion that this 
is also true for the resting stage of the nucleus. 
Up to the stage represented in Fig. 22, there was no evidence of the 
reduction in the number of threads having taken place. In fact, in Fig. 23, 
which is undoubtedly a later stage than those represented in Figs. 21 and 
22, we still find the diploid number. But at the stage shown in Fig. 23, 
there was some evidence of the actual fusion and consequent reduction of 
the chromosomes beginning to take place ; and while the actual fusion 
was not followed with certainty, the reduced number always appeared after 
this stage and never before. In Fig. 24 is represented a slightly later stage, 
which I interpret to be immediately after or during the actual fusion of the 
chromosomes in pairs. 
This stage is almost immediately followed by the disappearance of the 
nuclear membrane, the separation of the paired chromosomes from one 
another, and the formation of the reduction spindle. The spindle fibrils 
originate out of the cytoplasm after the manner which prevails throughout 
the flowering plants. An early stage of their development is shown in Fig. 25, 
with the reduced number of heterotype chromosomes lying freely in the 
cytoplasm and attached to the growing spindle fibrils. The spindle in the 
early stages is multipolar, but very soon becomes bipolar. The poles are 
at first quite broad, but later become sharply pointed (Figs. 26 and 27). 
After repeated counting the reduced number of chromosomes appeared to 
