X THE HOMOSPOROUS LEPTOSPORANGIAT^ 377 



enia. In the Hymenophyllacese no cap cell is formed, but as in 

 Osmunda and Gleichenia, the upper cell is divided by walls 

 running over the apex. The divisions in the central cell and 

 the structure of the spermatozoids, so far as these have been 

 studied, correspond with those of the other Leptosporangiatse. 

 A single archegonial cushion is not formed, but the arche- 

 gonia occur in small groups at different points upon the margin. 

 Goebel (10) has shown, however, that these archegonial groups 

 arise first near the growing point of the prothallial branch, and 

 that they are simply separated by the intervention of zones 

 of sterile tissue. At the point where they arise the prothaUium 

 becomes more than one cell thick, and in all cases where the 

 development could be certainly followed, the archegonium 

 arose from one of the ventral cells, and never directly from a 

 marginal cell. The details of the development have not been 



Fig. 218. — Part of the filamentous prothallium and archegoniophores of Trichomanes 

 rigidum. (After Goebel.) 



followed, and whether there is any division of the neck canal 

 cell is not known. The neck is straight, as in Osmunda and 

 Gleichenia. 



In Trichomanes the archegonial meristem (archegonio- 

 phore) may be formed as a short branch, directly upon the fila- 

 mentous prothallium. 



The lateral walls of the prothallial cells are in all the species 

 thicker than is the case in most Ferns, and there are distinct pits 

 in them. In the rhizoids a parasitic fungus is frequently 

 found. 



The embryogeny is almost unknown (Janczewski (2) ), but 

 the first divisions and the very young sporophyte correspond 



