MARSILIA VP^STITA. 20^ 



in that fainilv. This is also true of the division of the 

 sperm-cells into two distinct groups, not nearly so promi- 

 nent in Pilularia, and merely indicated by the first division 

 of the central cell of the antheridium in the Polypodiacea^. 

 So, too. the female prothallium is less reduced than in 

 Marsilia, but in the latter its limits are alreadv seen in 

 the ungerminated spore. 



The peculiar quadrifoliate leaf of Marsilia seems to be 

 also a form not derived directlv from the lower ferns. 

 In Pilularia the leaf is perfectlv simple, and this is the 

 form of the first leaf in Marsilia: and it is not until sev- 

 eral leaves have been developed that the characteristic 

 four-parted leaf is met with. 



Without going further into detail, we are prett^' safe in 

 assuming, as the wa'iter has alread}' done * that the Mar- 

 siliaceie represent the end terms of a series of forms 

 whose lower members are found among the leptospo- 

 rangiate ferns, and probably the PolypodiacCct ; that of 

 the two genera, Marsilia is the more specialized, and 

 stands at the top, with Pilularia between it and and its 

 homosporous relations ; and we are not therefore to look 

 for any connection with forms higher up, but conclude 

 that this special line of development ends with Marsilia. 



*Campbell. The systematic position of the Rhizocarpea^, Bull, of the Torrey 

 Botanical Cluh, Oct. 1888. 



