. On Azolla and Salvinia. 237 



lous, but sometimes traces of radiating lines remain to a 

 latish period. 



About the same period I have pretty constantly observed a 

 mammilliform process in the cavity of the sac, corresponding 

 to the trifacial line (PL 19. f. 17.); it has a mucilaginous ap- 

 pearance and gradually passes off into a thin layer apparently 

 applied over the whole surface of the cavity of the sac. About 

 this period also, the outer surface of this sac will be found 

 studded with granules, by the increasing deposit of which it 

 becomes at length enclosed in an incrustation. This increases 

 in thickness, presents subsequently three lobes at the apex, in 

 the centre of which is an attaching ? process of the sac (PL 

 20. f. 13. 14.) Its next and last change is to assume a cellu- 

 lar appearance and harder consistency. The cavity of the 

 sac also generally presents appearances of cellularity, the cells 

 seeming to be very irregular in size, and I believe having 

 some connection, at least at first, with the mammilla above 

 mentioned (PL 19. /. 21.) The contents of the sac, how- 

 ever though seemingly so cellular, are scarcely appreciable; 

 grume only of excessive tenuity escaping under pressure. 

 Soon after this, the yellow sac which has increased in yellow- 

 ness and thickness, presents a viscid granular matter, not in 

 my opinion organised. 



The mature secondary capsule, which is attached to a free 

 central receptacle by means of a short, stout, compound stalk 

 (PL 20. f. 12.)* presents a cellular papillose surface of a 

 brown colour, becoming hyaline in water ; the component 

 cells are separable ; it contains a large, single, whitish body 

 of a chalky embossed aspect, without any obvious attach- 

 ment ; this is of an oblong shape, and has a rugose or 

 irregular surface (PL 20. f. \2.) The upper end, which is 



* The assumption that the more numerous pedicelled capsules are derived 

 from the development of every cell of the surface of the nucleus, while the fewer 

 oblong ones are derived from the development of several cells of the same surface, 

 will explain the difference of their stalks. 



