256 BOTANICAL GAZETTE [OCTOBER 
egg. This last position may have been taken because of a 
movement of the nuclei in the direction of least resistance, 
which would be towards the place lately occupied by th 
synergids. 
The fertilized egg usually completes its first division shortly 
after the second division of the endosperm, the first wall being 
transverse. The further development of the embryo does not 
differ from that of the other Compositae which have been 
investigated. 
SILPHIUM. 
Silphium integrifolium Michx., S. terebinthinaceum L., and 5. 
lacimatum LL. were studied. The last named species received 
more attention than the others on account of the large size of 
the ovules, and the ease with which they could be oriented for 
sectioning. Merrill’s recent paper on the life history of be 
various species is so complete, except in regard to fertilization, 
that we wiil pass over all other stages. a 
As in Erigeron, the polar nuclei fuse long before fertiliza 
tion, and the resulting nucleus comes to rest near the egg. 
careful study of the microspores of S. daciniatum revealed bad 
presence of long male cells, lying side by side and reaching 
almost around the interior of the spore. The pole 
usually enters at one side of the synergids. In one preparaliol 
it was observed to turn almost at right angles, pass between sé 
nucellar cap and the synergid, and then continue its gi 
towards the egg. The synergid against which the pa t 
lies soon begins to show signs of disintegration, the ae | 
usually remaining intact until the fertilized egg begins t0 — 
In all preparations examined the pollen.tube was much expan™ 
above the nucellar cap, and its walls stained intensely. 4 
fig. 7 shows two small bodies (+) similar to those a ells 
Erigeron, which remain in the pollen tube after the eee ue 
have been discharged. This figure also shows the coiled ea 
cell (sp,) resting against the egg nucleus (@)- bee a 
male cell (5f.) is lying against the endosperm eee 
The path by which it reached its destination can be 
