68 THE ARCHAEOPTERIS FLORA fcH. 



States, the latter having something in common — though perhaps ' 

 remotely, with the Crossothecas of the Coal Measures, are at 

 present wholly obscure. 



The absence of seeds associated with the undoubted Devonian 

 floras is very remarkable. A revision of this flora has not pro- 

 duced a single undoubted specimen of a seed. The three supposed 

 examples attributed to the genus Carpolithes by Dawson in 1863 

 and derived from the Devonian of America, have no claim 

 whatever to be regarded as seeds i. One of these more recently 

 figured by White is a small object of a doubtful nature which 

 he thinks may be merely a scale, possibly comparable to 

 those of Barinophyton^- 



The present author has seen from the KUtorkan beds of South 

 Ireland, one or two small bodies which bear some slight resem- 

 blance to seeds, but it is quite possible that they may be capable 

 of an entirely dift'erent explanation. If seeds do occur at 

 Kiltorkan, they are undoubtedly exceedingly rare, and since the 

 number of species of all groups there represented is very small, 

 probably not more than four, it is unlikely that Archaeopteris, 

 which is there by far the commonest type represented, -is a seed- 

 bearing plant. This is a point to which we shall return later 

 (p. 81). For the present it is best to assume that undoubted seed 

 impressions are unknown from Devonian rocks. 



Genera of very Doubtful Occurrence in 

 Devonian Rocks. 



Lepidodendron, Sternberg, 1820. There is still no really con- 

 vincing evidence of the occurrence of this well-known genus in 

 Devonian rocks, from which, however, it has been reported from 

 many parts of the world. That is to say, no examples have been 

 figured which show a typical Lepidodendroid leaf-scar and its 

 prints. If the genus does occur, the specimens so far known 

 appear to be all more or less decorticated. This is obviously the 

 case in some examples figured from Australia, Canada, the 

 Arctic Regions, Russia, etc. It seems probable, however, that 

 some of the best preserved of these fossils belong to distinct 

 genera such as Archaeosigillaria and Leptophloewn . The type 



1 Cf. White (1905), p 78. ^ White (1905), p. 78, EI. IV, Jig. 11. 



