﻿BOTANICAL GAZETTE 



cells have a nutritive function. Campbell has suggested that the 

 massive basal part of the outer integument in the Araceae may be 

 physiologically considered as perisperm. As both the inner walls 

 of the outer integument and the embryo sac wall, except at its 

 base, are cutinized, it is obviously impossible for food to pass into 

 the embryo sac except through the enlarged endosperm cells. By 

 the time the embryo is mature these cells have disappeared and 

 much of the food material in the basal part of the integument has 

 been absorbed. 



It is a well known fact that cells which are active in nutrition 

 possess nuclei which differ from nuclei in the ordinary resting 

 condition in their larger size, their chromatin aggregations, and 

 appearance of lack of organization. Huie (8, 9) and Rosenberg 

 (12) have described this phenomenon in the gland cells in the 

 tentacles of Drosera, while Magnus (ii) has shown that the same 

 thing occurs in the digestive cells of certain orchids having endo- 

 trophic mycorhiza. These facts seem to indicate that a nutritive 

 function is to be ascribed to these basal endosperm cells. 



Discussion 



Great variety in development of the embryo sac, even within 

 a single species, has been reported for the Araceae, and it was with 

 a view to finding out in what respects Richardia africana resembled 

 the other members of the family that this investigation was under- 

 taken. Richardia africana, or Lantedeschia aethiopica as Engler 

 (6) calls it, belongs to the Philodendroideae, to which group also 

 belong Homalonema, Philodendron, Aglaonema, and Diejfenbachia. 



According to Campbell and Gow, all these genera, with the 

 exception of certain species of Algaonema, show a normal 8-nucleate 

 embryo sac. In a series of papers on the subject, Campbell 

 (1, 2, 4) has dealt with 4 species of Aglaonema. He finds that the 

 embryo sac of A. commutatum may contain 4-12 nuclei, and shows 

 little uniformity in arrangement. This species has a group of cells 

 at the base of the endosperm bearing some resemblance to that m 

 Richardia. In 1900 Campbell suggested that these might originate 

 by division of the antipodals, but later in 1903 inclined to the view 

 that they were of endospermic origin. Gow (7), however, in 



