Y? 
— 
1906} SCHAFFNER—LILIUM TIGRINUM 185 
esis appears still to be an open question. Such contractions are so 
easily produced by the ordinary killing reagents used, and have been 
described for such a variety of the early stages of nuclear division, 
that it seems to me no importance is to be attached to observations 
which have been made on killed material. So far as any opinion is 
to be expressed upon the appearances in Lilium tigrinum, I am still 
convinced that the extraordinary distortions commonly figured owe 
their origin to the action of the fluids used before imbedding in par- 
affin. : 
The spirem remains simple during the entire synapsis. The linin 
thread becomes thickened and the chromatin granules are usually 
more or less elongated (figs. 4, 5). At this stage the spirem has 
already a strong tendency to be thrown into loops and coils (fig. 4). 
After the microsporocytes have become partly separated and more 
spherical, they were rarely observed to be in synapsis. The nuclear 
cavity enlarges and the chromatin ribbon becomes thicker, with the 
granules still more prominent (jigs. 6-8). This is an important point 
to consider, for we have here a clear case of sporocytes, long past the 
supposed synapsis stage, showing with remarkable clearness a simple 
continuous spirem, with a single chain of chromatin granules. 
Synapsis, therefore, can have nothing to do in this case with a sup- 
posed longitudinal conjugation of two spirems or two networks of 
chromatin before the spirem is formed. After the nuclei have passed 
on to the stage represented in fig. 6, the stages are so easily followed 
and the threads so prominent, that a longitudinal conjugation, if one 
occurred, could not escape notice. Shortly after the stage shown in 
jig. 6 the spirem begins to show double rows of elongated chromatin 
granules, but the relative quantity of chromatin ribbon present in the 
nuclear cavity is not much diminished (figs. 9, 10). If the amount 
of spirem were diminished one-half by a longitudinal conjugation the 
fact would certainly be noticeable. Often a part of the spirem appears 
still single, or it will appear with a double row of granules and gradu- 
ally change to a row apparently single (figs. 11-17). The appearance 
would be the same whether the granules were dividing or conjugating. 
The uniformity of the pairs of granules on the linin thread is remark- 
able, and the pairs themselves suggest a division. If a conjugation 
of the chromatin granules were established, it would certainly show 
