REVISION OF ENDOGONEAE. 335 



radially, Figures 89-90, shows a rather clearly defined distinction 

 between an outer thicker, and an inner thinner dense layer and a 

 looser broader sporogenous region which lies between them, but is not 

 separated by any clean cut line of demarkation. 



Although the fungus is usually described as immature or sterile, I 

 have myself seen no specimens which are not fertile and but two that 

 were even moderately young. Of these the youngest was found in 

 1891, near Kingston, Jamaica, partly buried in a very soft rotten log. 

 The sporocarp was completely enveloped by a thin white separable 

 universal membrane within which it was already free and perf9rate 

 below. The hyphae composing this membrane are slender, thick- 

 walled and septate. In this unexpanded condition, which was thick 

 and somewhat flattened, the spores, although for the most part 

 nearly mature, Figure 90, were crowded in an irregularly double layer, 

 and the plant was set aside as a species of Endogone. 



As the sporocarp matures and enlarges this crowded condition dis- 

 appears, and the outer and inner layers, especially, evidently take an 

 important part in the tangential increase, so that the crowded spores 

 become more and more discrete as the wall stretches itself, so to speak, 

 and eventually assume a more or less definite arrangement in a single 

 layer. The hyphal elements which take part in this process are thick- 

 walled with rather frequent septa, densely compacted in the two walls 

 and somewhat looser in the middle region (Fig. 94) . As the sporocarp 

 increases in size the septa become very numerous and the filaments 

 greatly enlarged, so that the inner and especially the outer layers ap- 

 pear to be made up of dense thin-walled parenchyma which gradually 

 passes into the looser lacunose middle region, the hyphae of which are 

 also thin-walled for the most part and submoniliform. Figure 93, 

 with large interhyphal spaces. 



Unfortunately no information is available as to the earlier stages in 

 the formation of the sporocarp, or the initial processes of spore- 

 formation. The sporogenous zone is evidently traversed at intervals 

 by radial wefts of finer, closely septate hyphae, which are dense and 

 have a dark granular appearance (Figs. 90-91). These wefts are the 

 sporogenous centres above referred to, and conceal the spore-origins 

 which traverse them. The spore-initials emerge from them as shown 

 in Figure 91, a tangential section, in the form of long clavate structures, 

 terminating apparently aseptate filaments which, although it has not 

 been possible to trace them to' their origin, appear to grow radially 

 from the inner margin of the inner dense layer of the sporocarp. The 

 tips of these clavate structures become greatly enlarged and the walls 



