1899] DEVELOPMENT OF THE MICROSPORANGIUM 349 
itis not possible to count more than seven or eight chromo- 
somes. 
Whether a fusion takes place here one cannot determine, 
since the small size renders it impossible to follow the process 
accurately. The exact manner of segmentation upon the spindle 
is also still in doubt. Seven chromosomes are found in the 
daughter nuclei, which leads us to infer that each of the origi- 
ual masses splits into two, but the preparations show no indica- 
tions of any such division farther than that in many cases 
under high magnification, it seemed as if the chromosomes pos- 
“ssed a +-like structure similar to that in Convallaria, but 
the figures were not distinct enough to allow of any definite 
conclusions, 
Inregard to the formation of the spindle a few notes may 
be given, although the results do not differ essentially from 
those obtained by Mottier in Lilium. The kinoplasm is at first 
ited to a thin felt-like coat surrounding the nucleus (fig. 32). 
On account of the very large space occupied by the nuclear 
| QD, it is easy to observe the entrance of the kinoplasm into the 
; tuclear cavity. This takes place apparently before the entire 
tisappearance of the membrane. The latter sometimes is still 
Mable after the nuclear cavity is nearly filled with kinoplasm. 
a first thought it seems impossible to conceive of a substance 
_ Pissing through the nuclear membrane in this way. But Mot- 
- has shown that the membrane itself is probably nothing 
Be close weft of kinoplasm. We have then merely 
~ “sume that the inner threads of this weft separate from the 
_* traverse the nuclear cavity instead. Finally the whole 
.." membrane. They are from the first a part of it. At 
7% € spindle is multipolar (figs. 33,34)» but the poles are 
i. amber. and very soon disappear, thus giving ae 
|. Normal bipolar type. In its mature condition the spindle 
. a = and the poles are very acute (jig. 35). The fibers 
: Of. cit, Pringsh. Jahrb. f. wiss. bot. 30: 176. 1897. 
