CYCADOFILICALES 33 



a very advanced seed condition may be attained, while the leaf and 

 vascular anatomy remain at the fern level. So the stelar structures 

 may advance while the leaf and general topography remain more or 

 less stationary. 



We cannot agree that the heterosporous ferns may be as old as the 

 homosporous; or that the seed plants may be as old as the ferns. 

 When all the geological evidence is in, it must confirm the sequence: 

 homospory, heterospory, seed. Heterospory in the Carboniferous, 

 and earlier, was attained in the same way as it is today. A section 

 of Bothrodendron from the Upper Carboniferous of Illinois, described 

 by Reed,''^' shows four megaspores and a plasmodium-like mass 

 which can hardly be anything else than the broken down remains of 

 wall cells, tapetal cells, and abortive spores (fig. 26). The resem- 

 blance to a sporangium of Selaginclla at this stage, where the origin 

 of the mass is definitely known, is striking. In all the living genera of 

 heterosporous Pteridophytes the homosporous ancestry is unmis- 

 takable. 



While the ancestors of Cycadofilicales were in the heterosporous 

 condition, but were not yet up to the seed stage, they were Pterido- 

 phytes, and this would be true, whether they were coming from some 

 known Pteridophyte by direct descent, or were coming by parallel 

 development from some Pteridophyte ancestry not yet discovered. 



When structures like the leaves are so similar in ferns and seed 

 plants that the two groups cannot be distinguished by their leaves, 

 it seems more probable that the sporangia, and also the vascular 

 system, of the higher groups have advanced from the heterosporous 

 fern condition to the seed condition, while the leaves have remained 

 stationary. It is logically possible that leaves so identical as to be 

 indistinguishable may have developed in ferns and seed-ferns from 

 entirely unrelated ancestors; but it seems more likely that the simi- 

 larity is due to genetical relationship. A Latin poet, in giving advice 

 to young play writers, once advised them not to bring a god upon 

 the stage unless the situation demanded it. In the case of ferns and 

 seed-ferns, it would seem that genetical relationship, functioning as 

 it is known to function in living plants, where the relationship is 

 fairly well known, is a sufficient factor to account for the observed 

 structures without putting upon parallel development the strain of 



