RESULTS OF PAL^EONTOLOGICAL RESEARCH 59 



groups of higher standing must be referred to an older 

 and older period. ' It is not long ago/ says Lotsy, 

 ' that it was thought that the seed-bearing plants 

 were of comparatively 

 recent origin and that 

 at least in the Coal 

 period they were en- 

 tirely absent/ Now 

 the Cryptogams (non- 

 seed-bearing plants) 

 are not even conceded 

 predominance in the 

 later Palaeozoic period 

 (Fig. 18). ' Gradually 

 the Ferns, one after 

 the other, showed 

 themselves to be seed- 

 bearers, and it is diffi- 

 cult to say to what 

 number relatively this 

 alteration will extend. 

 It will probably be a 

 large one/ ] 



Yet must we agree 

 with Gothan's warning against over-hasty conclusions. 2 

 The true fern nature of a leaf is certainly only 

 beyond all doubt when we find the spore heaps 

 on the under side of the foliage or elsewhere, since 

 thereby they are directly shown to be spore-bearing 



1 Biol Zentralblatt, 1905, p. 414. 2 Entwicklung der Pftanzenwelt, p. 35. 



FIG. 18. ANG, Angiospermse; CON, Coni- 

 ferse; COR, Cordaiteae; CYC, Cycado- 

 phyta ; EQ, Equisetinese ; F, Filicineae ; G, 

 Ginkgoales; LYC, Lycopodineae ; PT, 

 Pterido sperm se or Cycadofilices ; S, 

 Sphenophyllales. 

 It is seen that the Palaeozoic seed-bearing 



plants PT and COR approximately equal 



the non-seed-bearing, F, LYC, S, and EQ. 



(After Oliver.) 



