398 MOSSES AND FERNS chap. 



In the microsporic sorus, the apex of the placenta does not 

 develop sporangia, but remains as a sort of columella (Fig. 202, A). 



The sporocarps of Salvinia are like those of Azolla, but the 

 two layers of cells are separated by a series of longitudinal air- 

 spaces which correspond to ridges upon the surface of the sporo- 

 carp (Fig. 196, D). 



The microsporangia of Azolla have a long stalk, which 

 is composed of usually two, but sometimes three rows of 

 cells. The sixteen sporogenous cells all develop, so that 

 there are normally sixty-four microspores in each sporangium. 

 These have the exospore thin and smooth, and are included in 

 a kind of common epispore, which here too owes its origin 

 mainly to the tapetal cells. This episporic substance is divided 

 into masses (massulae), which have the foamy structure of the 

 episporic appendages of the macrospore. This appearance is 

 apparently due to the formation of vacuoles, which make these 

 massulae look as if composed of cells. The tapetal nuclei are 

 confined to the outside of the massulae, and can be detected 

 almost up to the time they are fully developed. Finally, upon 

 the outside of the massulae are formed the curious anchor-like 

 "glochidia" (Fig. ig7,gl), whose flattened form is due to their 

 formation in the narrow spaces between the massulae. 



In Salvinia the microsporangia arise as branches from 

 sporangiophores which bud out from the columella, so that their 

 number much exceeds that of the macrosporangia, or of the 

 microsporangia of Azolla. There are no separate massulae, and 

 in the macrosporangium the epispore is much less developed 

 than in Azolla. 



The MarsiliacecB 



The two genera of the Marsiliaceae, Marsilia and Pilularia, 

 are much more closely related than Salvinia and Azolla, and at 

 the same time their resemblance to the homosporous Ferns is 

 closer, and of the two genera Pilularia is evidently the nearer to 

 the latter. The development of both gametophyte and sporo- 

 phyte in the two corresponds very closely. 



The sporangia are borne in " sporocarps," which are morpho- 

 logically very different from those of the Salviniaceae, being 

 metamorphosed leaf segments enclosing several sori, and not 

 single sori enclosed simply in an indusium. The spores 

 germinate with extraordinary rapidity, especially in Marsilia, 



