MARSII.IA VESTITA. 1 89 



the upper is the mother cell of the antheridium (pi. in, 

 hg. 8). The size of this vegetative cell (.v) varies a 

 good deal in different cases. In the earliest accounts of 

 Marsilia this cell was overlooked, but Sadebeck * de- 

 monstrated it later. In Pilulariaf there is in addition a 

 smaller one that is frequently found. The next division 

 differs in different cases, but usually is effected by a wall 

 approximately parallel to the first one, but more or less 

 concave upward, being in fact the homologue of the first 

 funnel-shaped wall found in the antheridium of the Poly- 

 podiaceai (pi. in, fig. 9), and the lower cell, which has 

 very little granular contents, corresponds to the lower 

 ring-cell of the wall of the ordinar}^ fern antheridium. 

 Sometimes, however, the antheridium mother-cell divides 

 at once by an oblique wall into two nearly equal cells, 

 which indicate the position of the two groups of sperm- 

 cells found in the older antheridium. In no case ob- 

 served was there certain indication of the formation of a 

 perfect dome-shaped w^all in the upper cell of the anthe- 

 ridum such as occurs regularly in the homosporous lep- 

 tosporangiate ferns, and also, as a rule, in the nearl^^ re- 

 lated but less specialized Pilularia. 



The formation of this wall seems to have been partiallv 

 lost as a result of the extremely rapid development of the 

 antheridium, and the separation of the groups of sperm- 

 cells takes place by the formation of cells, cut off in 

 a more or less irregular manner from the peripherv of 

 the two cells into which the upper cell of the antheridium 

 is at first divided. The cap-cell, at the top of the anthe- 

 ridium (pi. Ill, fig. 13, d) is almost always plainly visible, 

 so that the only difference between the normal develop- 



*Schenk's Handbuch, vol. i, p. 238. 

 tCampbell, 1. c. p. 23S. 



