Johnson — Forhesia Cancellata. 181 



comparable to that of birds to mammals. Zoologists see the line of 

 connexion of mammals with the Sauropsida throiigh its lowest group — the 

 Lacertilia — and not through the most higlily specialized, the Aves. 



Dichotomous branching as a primitive character was not so much 

 appreciated fifty years ago as it is now. Several of Baily's figures (1), e.g. 

 fig. 1 b and d, need redrawing, that some omitted indications of dichotomy may 

 be included. More stress may now be laid on the phylogenetic value 

 of the dichotomously veined and branclied cotyledon of ferns. At the 

 same time, the danger of attaching undue importance to external form must 

 not be overlooked. The fern cotyledon is constant in character, both in its 

 dichotomous lobation and associated dichotomous venation. Such marked, 

 constant features must have antecedents. There is no sign of them in any 

 sporophyte of moss or Liverwort; and it seems more natural to look for ancestral 

 forms in the gametophyte generation of a Bryophy te through such a connecting 

 form as Forbesia, and not to pass over the Muscinese altogether. Potonie (6), 

 e.g., seeks for a connecting link in Fucus-like Algse, and treats the fern 

 prothallus as an intercalated, secondary, aquatic adaptation. Didyota 

 dichotoma gives a hint of the line along which specialization might take place. 

 In it we have an alternation of generations in which the dichotomously divided 

 evasoular aquatic thallus is externally the same in both generations. One 

 generation is, however, asexual, and bears tetraspores only ; the other bears 

 sexual organs only. The two alternate, and yet are alike in appearance, in 

 keeping with their aquatic habitat. It would take an enormous time to 

 evolve a fern -sporophyte, as hypotheticated by Bower (7), from the simplest 

 sporogonium of a Bryophyte by sterilization of sporogenous tissue and by 

 appendicular expansions as leaves from a slowly erecting radial axis. It 

 seems simpler to accept the view that the alternation has arisen by a change 

 from an aquatic to an aerial mode of life, with a gradual production of 

 vascular tissue as a necessary accompaniment of the adoption of an aerial 

 mode of life. The converse process is well seen in aquatic flowering plants 

 which lose their xylem, and show a reduced vascular tissue. The Hepaticse 

 show, in their own gametophyte, the change from a thalloid to foliose habit, 

 and that which we have before us perpetuated in the Liverwort, may have 

 occurred in the ancestral Pteridophytes. D. H. Campbell (10) has shown by a 

 consideration of the geographical distribution of Museinese that they are an 

 ancient group, and their infrequency in the earlier rocks is explicable by the 

 perishable nature of their tissues. (Lindley, e.g., found no trace of the six 

 mosses which he left for two years, with flowering and other plants, in a tank 

 of water. Very muddy water would have been a test more in keeping 

 with the conditions of fossilizatiou). The morphological value altaolied to the 



