162 BOTANICAL GAZETTE [SEPTEMBER 
greater, no sterile prothallial cells appearing, the whole struc- 
ture being an antheridium, as in the angiosperms. Our historical 
evidence accords with this progressive reduction of the male 
gametophyte, the Taxus and Cupressus lines having attained 
modern expression after the Abies line ; and back of the Abies line 
we find the Cordaites, with probably a still greater development of 
the sterile region of the male gametophyte indicated. To derive 
the Cordaites or Abies lines, with their two or three to many- 
celled sterile tissue of the male gametophyte, from such hetero- 
sporous lycopod forms as we know today, with their constantly 
more reduced male gametophytes, is not within the bounds of 
probability. Besides, the reduction of the male gametophyte 
seems to be so prompt a response to heterospory, that its par- 
tially reduced condition in certain conifers, and probably in Cor- 
daites, would seem to argue for their near derivation from some 
homosporous type. 
The development of a suspensor in the lycopod forms has 
also suggested a genetic connection with gymnosperms, in which 
the suspensor development is so conspicuous. This organ, how- 
ever, seems to have no morphological constancy. In gymno- 
sperms it may be developed from a plate of cells formed in the 
oospore, as in most conifers; or from a mass of cells formed 
basally or parietally in the oospore, as in cycads; or from free 
cells formed within the oospore, as in Ephedra; or from the 
elongation of the oospore itself, as in Gnetum; or from the 
downward elongation of the archegonium, as in Welwitschia. 
The suspensor, therefore, seems to be a temporary organ of 
the embryo, of various morphological origin, intended to relate 
the embryo properly to its food supply, and not of phylo- 
genetic significance. 
The testimony of history and morphology seem to combine 
in pointing to avery generalized Paleozoic type as the origin of 
Symnosperms. This type is characterized by its advancement 
towards seed production rather than by its habit, which must 
have been extremely varied to have given rise to such types 45 
cycads and conifers. The usually recognized Cordaites show 
