Pollen Mother-cells of Lilium cancidense . 197 
are in many places in close contact and seemingly fused. They now 
approximate much more closely throughout their length, twist tightly about 
each other and fuse into what appears to be a single thread. Although the 
close approximation of the threads occurs quite early in the synapsis 
period, the completion of their fusion is a matter of considerable time. 
During the process all stages of fusion may be observed in the same nucleus. 
In places the two strands are parallel but not in contact ; some portions of 
the thread show their double nature only by a lighter line in the middle ; 
in other places the two threads are twisted about each other, and in still 
others there appears a single homogeneous thread. After a considerable 
time all signs of a double nature disappear, and the synaptic mass contains 
apparently a single relatively thick thread (Fig. 19). After the fusion and 
during the continuance of synapsis, the spirem shortens and thickens to 
some extent, but the thickness of the thread at the close of synapsis, as 
may be seen by comparing Figs. 19-21 with Figs. 12 and 13, is not much 
greater than its thickness immediately after the fusion. The diameter 
of the thread, as the figures show, is at no time perfectly uniform. The 
synaptic aggregation still persists (Figs. 19), and the mass of the strands 
composing it is about the same as at its first appearance, but the cluster 
is looser, and it is easier to follow a single strand for a considerable distance. 
The looser arrangement is partly accounted for by the fusion of two threads 
into one, and the consequent lessening by one-half of the total length 
involved, partly by the shortening and thickening just noted ; but there 
is also already evident a tendency of the thread to disentangle itself and to 
become distributed more uniformly within the nuclear cavity. 
In the most favourable preparations of material in the stages just 
describedj lightly stained with iron-haematoxylin, there appears very clearly 
the differentiation into chromatin and linin already mentioned as occasionally 
visible at the stage of Fig. 8. The thread in such a preparation is seen to 
be composed of a lightly-stained ground substance in which darker bodies 
are imbedded. By careful washing in the iron alum, the ground substance 
may be almost entirely decolorized, so that the thread at first sight seems to 
consist merely of a row of granules (Figs. 15-18). In preparations stained 
with the triple stain, the chromatin bodies are blue, the linin is red. If the 
section is over-stained with safranin, the chromatin bodies are dark red, 
the linin being much lighter. 
This differentiation within the spirem thread appears most plainly 
in cells at or near the ends, especially the cut ends, of the sections, where 
the fixing fluid first came in contact with the tissues, and where, therefore, 
its effect may be supposed to have been strongest. In these regions, both 
cells and nuclei are often shrunken, and the cytoplasm has the peculiar 
granular appearance which is a characteristic effect of exposure to a too 
strong solution of osmic acid. The spirem shows little or no change. 
