89 



less than that in G. -pectinata, but much greater than in either 

 G.flabellata or G.rapestris. In the size of the antheridium and 

 in the number of spermatozoids it is inferior to G. pectinata. 

 Rauwenhoff does not state the number of spermatozoids in G. 

 flahellata or G. rupestrls, but to judge from his figures it is 

 much less than in any of the species studied by me. 



G. dichotoma agrees closely with G. pectinata both in the 

 position and structure of the antheridium (PI. XI, Figs. 82--88), 

 but the latter is usually slightly smaller than that of G. pectinata. 



As we have already seen G. laevigata differs much from the other 

 species in the position of the very large antheridia which are even 

 larger than those of G. pectinata (PI. XI and XII, Figs. 89—93). 

 They occur often in crowded groups near the base of the lateral 

 leaf- like lobes of the prothallium. Less frequently they are found 

 upon the upper surface of the midrib. While the arrangement 

 of the cells of the wall is much like that in G. pectinata the 

 number of cells is much larger and sometimes the walls of the 

 antheridium are in places more than one cell thick (PI. XTI, 

 Fig. 91r/). Owing to the crowding together it is not uncommon to 

 meet with antheridia that are much smaller than the normal ones 

 (PI. XII, Figs. 92 — 93). Some remain rudimentary and show an 

 elongated form resulting from a definite apical growth. These 

 dwarf-antheridia seem to be formed as buds upon the larger ones. 

 In several cases what seem to be rudimentary antheridia were 

 seen to be growing from the wall cells of the larger antheridia. 

 Corresponding to the large size of the antheridium the number 

 of spermatozoids is also greater than in G. pectinata. 



The Archegonium. 



The first archegonia are formed in a cushion just back of 

 the growing point of the prothallium, very much as in ordinary 

 ferns. As the prothallium grows new archegonia arise more 

 abundantly at the sides of the growing point and thus are 

 shifted to the flanks of the massive midrib, being much less 

 numerous among the rhizoids. In the large prothallia of G. 

 dichotoma it was sometimes found that the portion of the 



