226 MOSSES AND FERNS chap. 



numbers from these. The inner tissue was colourless and 

 large-celled, and in the cells of the lower part of this tissue 

 were observed in all cases great numbers of irregular colourless 

 filaments that had all the appearance of an endophytic fungus. 

 Whether this is of the nature of a mycorrhiza remains to be 

 seen, but it is by no means impossible that such should be the 

 case. This lower tissue forms a distinct zone, which in section 

 appears much more opaque than the upper zone, from whose 

 outer cells the sexual organs arise. 



The sexual organs were found as a rule only upon the 

 upper side, in which respect it differs from B. lunaria, where 

 the archegonia usually are formed on the lower side. The 

 material was too old to make it possible to tell whether the 

 growth of the prothallium was from a single apical cell or 

 not. All of the archegonia found were old, and it was im- 

 possible to follow their development, and the details of the 

 structure of the ripe archegonium could not be made out. 

 A striking point of difference between it and the other forms 

 hitherto investigated is the long neck, which projects quite 

 as much above the prothallium as that of the leptosporan- 

 giate Ferns (Fig. iii, C). In general appearance it closely 

 resembles that of Osmunda, being straight and not curved 

 backward, as in most Ferns. The antheridia may either 

 occur singly, in which case they are sunk in the thallus, or 

 sometimes groups of them occur together upon short branches 

 projecting from the upper surface. In the latter condition the 

 individual antheridia often project somewhat. So far as could 

 be judged from a study of a very small number of young 

 stages, the development corresponds exactly to that in Equisetmn 

 or Marattia. The antheridium mother cell probably, as in 

 these, divides first by a wall parallel to its outer surface into 

 two cells, and the inner one divides next by alternate trans- 

 verse and vertical walls into the mass of sperm cells. In some 

 cases the outer cell divides both by vertical and transverse 

 walls, so that the outer wall of the ripe antheridium is two- 

 layered, as in B. lunaria ; but quite as often it remains but one 

 cell thick (Fig. i i i, B), in which respect it resembles Equisetum 

 or Marattia. The spermatozoids were not observed. 



In only one case was a young embryo found, and this, so 

 far as could be determined, also resembled in the arrangement 

 of its cells the similar condition in Marattia, but the prepara- 



