338 BOTANICAL GAZETTE | NOVEMBER 
properly belongs. The others are probably caused by the action 
of the reagents. 
Mottier’s*3 figures representing synapsis in Lilium do not show 
nearly as much contraction as in Convallaria. From the descrip- 
tion it would seem also that they represent a much later stage 
in the development of the spirem than the one in which synapsis 
occurs, either in Convallaria or in Potamageton. 
FIRST NUCLEAR DIVISION OF THE MOTHER-CELL. 
The spirem stage preceding division is very well marked in 
Convallaria (fig. 0). The much coiled wire-like thread found 
immediately after synapsis gradually increases in thickness, and 
the chromatin granules become less prominent. One large 
nucleolus and usually two or three smaller ones are present at 
this period. The longitudinal splitting of the chromatin thread 
is accomplished so quickly that it was found impossible to 
observe the successive steps in the process. At the same time 
the thread becomes considerably thicker than immediately before 
the division (fig. rz). It is difficult to understand just how this 
doubling takes place, because at the very first indication of such 
a condition the threads are already separated. They lie parallel 
with each other and slightly coiled. The further development 
indicated a more or less complete subsequent fusion of the parts, 
so that before passing into the nuclear-plate stage the dual 
nature is entirely lost except for an occasional lobing at either 
end (fig.12). Not even the granules are longer visible, and the 
chromosomes are at this stage apparently homogeneous. ae 
is probably only apparent, however, and due really to the density 
of the stain and the close proximity of the parts. 
After the chromatin thread has become double, beside i 
thicker than before, it also possesses fewer coils, which 8 sr 
ably due to a longitudinal contraction of the whole spirem: 
about the time when the nuclear membrane disappears the thic 
: ; pager a: somes 
ened chromatin band segments into the individual chromo 
sides being 
3 Beitrage zur Kenntniss der Kerntheilung in den Pollenmutt 
Dikotylen und Monokotylen. Pringsh. Jahrb. f. wiss. bot. 30: 175- 1897. 
; 
q 
: 
i 
