120 THE EVOLUTION OF PLANTS 



Similar pollen-grains have been detected in the 

 pollen-chamber of a seed of the same general 

 type as those which are referred to the Neurop- 

 teridese among Pteridosperms. The identifica- 

 tion of the seed with Dolerophyllum is not, how- 

 ever, certain. The simple leaves and the strange 

 stamens, with pollen-sacs completely buried in 

 their tissue, separate this genus from all known 

 Cycadophytes or Pteridosperms; the structure, 

 however, both of the leaves and of the pollen- 

 grains establishes a connection with one or the 

 other group. While not forming a link in any 

 known evolutionary chain, Dolerophyllum shows 

 that at the close of the Palaeozoic age the evolu- 

 tion of the Cycadophytes was going on in more 

 than one direction. It may be added that in 

 the same Permo-carboniferous rocks, stems have 

 been found which are in some respects inter- 

 mediate between those of Lyginodendron and 

 Cycads, though not on the direct line leading 

 to any known types. 



The gap between Pteridosperms and Cycads 

 is practically certain to be filled up by future 

 research; probably the critical fossils may be 

 already known, for there are various fronds, 

 both late Palaeozoic and Mesozoic, which may 

 well have belonged to transitional forms; at 

 present, however, the data are insufficient for 

 tracing the connection in detail, though the 

 affinity of the two great groups is clear. 



