1913] GOW—MORPHQLOGY OF AROIDS 141 
dehiscing anthers in which it was found that this division had not 
yet taken place. Grains were sprouted in weak solutions of glucose 
and lactose, and examined unstained, or stained in a weak aqueous 
solution of hematoxylin. One of these is indicated in fig. 47. 
Summary 
1. The development of the archesporium in aroids follows no 
uniform rule. In one species studied,’ the primary archesporial 
cell produces two megaspores, the outer one of which functions; 
in two cases there are four megaspores, the outermost one of which 
functions; in one case the primary cell produces two cells, one 
of which may be a tapetal cell; one species develops four megaspores 
in the same horizontal plane, and the functioning one then develops 
a tapetal cell at its upper end; one species develops four megaspores, 
the innermost of which functions; and in one case the primary 
sporogenous cell develops directly into an embryo sac. 
2. The number of antipodals varies from 3 to 11, four species 
having regularly more than 3. This is a relatively primitive 
feature, and occurs in other primitive orders. 
3. Previous to the work here reported, embryos of the spherical 
type lacking a suspensor had been reported by CAMPBELL in 
Pistia, Dieffenbachia, and Lysichiton. The writer has confirmed 
CAMPBELL’s observation with regard to Dieffenbachia, and has 
discovered the same sort of embryo in Nephthytis, Philodendron. 
Arum, Aglaonema, Anthurium, Arisaema, Symplocarpus, and 
Richardia. It seems clear that this type of embryo is characteristic 
of this family. 
4. Richardia as here cultivated in the greenhouse has usually 
a sterile nucellus and does not develop an embryo sac. The dis- 
integrating nucellar cells act as conducting tissue, and the pollen 
.tube enters sterile nucelli, although gametic attraction cannot 
exist. The direction of the tube is conditioned on the presence 
of nutritive material. 
5. Reduction of parts is indicated by the presence of staminodia 
in several of the species studied. 
6. In an abnormal specimen supposed to belong to the genus 
4 Gow, James EL.is, Studies in Araceae. Bor. Gaz. 46:35-42. pis. 4-6. 1908. 
