eTHENOGENESIS IN THALICTRUM PURPURASCENS 365 
eveloped in these bags from the time the pistils 
e determined by means of the microscope until the seeds 
Each week a head was removed, the flowers being 
as above for future study. 
ss that survived under all the conditions, in the 
in the garden, or as rootstocks, produced abun- 
d fully developed seeds. Of course in such a case it 
mikely that the treatment with the solutions produced 
ct. Therefore the problem resolved itself into an inves- 
of the embryos, to determine whether or not there was 
C nesis, the embryo developing from the unfertilized 
egetative apogamy. Abundant and good material had 
sen ed, which it was hoped would show all stages 
O answer this question. The material after fixing and 
brought gradually into 70 per cent. alcohol and there 
used. The xylol-paraffin method was used entirely, © 
were cut with a microtome from 5-15 thick as the : : 
Sections were stained with Delafield’s haematox- Q 
with Flemming’s safranin, gentian-violet, and orange 
: drawings were made with a Zeiss Camere. A and = 
: Lomb oil immersion. ae 
i 
RESULTS. S 
c irpose of this paper to describe the even 
megaspore in detail, for it differs. in no way if 
: ong angiosperms, In the very young ale the 
ell is distinguishable (jig. 7, 2), later enlarging 
‘dividing unequally to form the so-called tapetal 
and the larger primary sporogenous cell (fg. ae 5). 
ell may or may not divide further. Fig. 4 shows 4 
Primary sporogenous cell gives rise to the usual 
megaspores (fig. 5), the innermost nar of the : 
| ~ 5, fm 
pore germinates in the usual way, 
h surrounding cells until it occupies 4 ‘large P 
@ Sy 2 and the nuclear divisions resulting it in 
