) LIFE HISTORY OF THE PONTEDERIACE 331 
with the transparency of the stains. Delafield’s haematoxylin 
ad erythrosin, and, in ovaries isolated from the perianth, anilin- 
siranin and gentian-violet gave the best results. Numerous 
other combinations were tried, but all were more or less unsatis- 
factory. 
_ Atypical spindle of the first karyokinesis of the elongated 
_ tegaspore mother cell is represented in fig. 20. The large fibers 
' tfached to the chromosomes are plainly seen to be composed 
olmany delicate threads. On the other hand the fibrillar struc- 
_ ie of the central spindle is less clearly perceptible. 
Cytoplasmic radiations about the poles occur in all good prep- 
_ Nations and frequently the individual rays appear to terminate 
“ternally in large granules which stain like nucleoli. The 
chromosomes, 14 to 16 in number, were never seen to assume a 
‘shaped outline at any stage of the nuclear division. This state- 
— Ment holds good also for Pontederia, where by the use of iron- 
dum-haematoxy lin a sharper definition of the chromosomes was 
— Secured, 
All the preceding description of the spindle will apply equally. 
_ Wthe divisions of the pollen mother cell. 
The nucleus of the elongated sporogenous cell passes through 
_-*Ynapsis Phase. If such a phase precedes the first division 
| — that occurs early (as in fig. 72) I found no evidence of it. 
| Sesto Special significance can be assigned to the case shown 
MG. 18 (see also figs. 22 and 23), in which the two sporog- 
: oh Whose nuclei are in synapsis have arisen apparently 
© early division of the primary sporogenous cell. But 
daa been suggested by some cytologists, synapsis occurs 
» 49 those cells in which a reduction of the number of chro- 
aa iS going on, it would follow that when the primary 
8€ous cell divides early without elongating, the chromo- 
| ae takes place in each of the paren ee 
- tty, the sat Ep eniary sporogenous cell elongates be 
= Uction is completed in the first nucleus. 
