104 The Waterlilies. 



of slightly flattened cells, poor in contents. The inner integument has 

 both of its layers very thin, and the outer one nearly empty. 



Dahmen (1892) critically examined the funiculus of the ovules of 

 Nymphaea. Since the so-called aril originates as a swelling on the 

 funiculus before fecundation occurs, he considers that it is not an aril 

 in the usual sense, restricting the use of the term to outgrowths which 

 occur when the seed is ripening. The distinction, however, is extremely 

 unimportant, especially as the aril of Nymphaea does not reach maturity 

 until the seed is ripe, and most of its growth occurs after the seed 

 has attained its full size. In N. lotus Dahmen noted two swollen zones 

 on the funiculus near the ovule, but only one in N. rosea [quid?) and 

 zanzibariensis. The epidermis of the funiculus has slightly thickened 

 walls ; within it is a thin-walled parenchyma and a vascular bundle with 

 phloem and a little xylem. 'The bundle has no connection with the aril. 

 Small round starch grains are found temporarily in the parenchyma, 

 epidermis, and phloem. Intercellular spaces occur only in the arillar 

 swellings according to Dahmen, but I find them plentifully elsewhere in 

 N. odorata. The tissues are all filled with mucilage, which is only precip- 

 itated by strong alcohol. The protoplasm is dull yellowish. Asparagin 

 occurs in the funiculi, but no sugar was found. 



Unfortunately we cannot accurately trace all of these tissues to their 

 mature state, but the changes are not important in any part except the 

 embryo-sac. The early stages were worked out by Hofmeister (1858) in 

 Nuphar luteum and advena, which differ from Nymphaea in that the narrow 

 inner part of the embryo-sac is much longer, and extends after fecun- 

 dation to the chalaza ; an axial tube remains open even in the mature seed 

 (Weberbauer, 1894). In Nuphar there are three cells at the top of the 

 embryo-sac, two upper synergidae and an ovum below. About the time 

 of fecundation a transverse wall divides the narrow part of the embryo sac 

 from the upper wide part, and the latter is soon filled by two or three 

 endosperm cells. The fertilized ovum now divides transversely, and a 

 suspensor of one to five cells is formed above the embryo. The obser- 

 vations of Mr. Cook (1902) on N. odorata seem to be corroborated up to 

 the first cell-wall laid down in the embryo-sac. His views on the develop- 

 ment of the embryo, however, which he considers to be monocotyledonous 

 and without suspensor, are totally at variance with my own preparations. 

 I cannot imagine that his interpretations are correct. This portion of Mr. 

 Cook's paper is based chiefly on Nuphar advena, in which Caspary's 

 drawings (in Hb. Berlin) seem also to indicate the absence of a suspensor. 



