i9oS] 



COW—EMBRVOGENY OF ARISAEMA 



45 



appears that endosperm formation does not begin until 



tube has reached the embrv'o sac (and 



iperm 



and the second 



fused), it 



would seem that the fertihzation of the oosphere is followed by a 

 period of rest before the formation of the embr}^o begins. The first 

 two divisions are transverse^ followed by a longitudinal division (figs, 

 20-22). Beyond this point no definite histor>^ of the embrj^o can Ije 



Longitudinal sect 

 e. at somewliat lat* 



1 



23 



endosperm. X770 



X770- 



given until the stage illustrated in fig. 23 is reached. The embno 

 here shown is six weeks or two months old and has not yet reached 

 its maximum development; fig. 24 shoe's a somewhat older embn-o. 



cTood sections of the embr}-o were obtained. 



Beyond this stage no j^ 



SUMMARY 



dioecious character of Arisaema 



ce amon^ the aroid genera, other character! 



bear 



and orthotropous ovule, 



.. « tm*^rseei^to"indkate it as a primitive form, but the fact that 

 there are four o;-ules (typically) in each ovary, and that their basal 

 origin is apparent onlv, since they are in reality lateral outgrowths 

 of the suppressed axis, seems to prove it a more highly speciah^ed 

 type than Aglaonema, Nephthj-tis, and Spathicarpa, in ^ 



.rpel contains a single basal o\Tile 



This view is confirmed 



consider the variation^ in the number and ^f^f ^f *,^°^f ^ 

 lei in Nephthytis and Aglaonema (Campbell^ 



imitiTC character 



the embr\'0 sac 



Arisaema corresponds to the usnal tj-pe found in other '--^'>^V^' 



and 



ir' 





ffinities with lower terms. 



