XXXIX] COMPARISON 93 



The facts relating to the sporangia together with a rather primitive 

 vascular construction have led to the segregation of L/avea, Ciyptogramnie 

 Jamesouia, and Ceratopteris as Primitive Gymnogrammoid Ferns. Of these 

 Llavea and Cryptogramme are confirmed in the position near to Plagiogyria 

 assigned to them by Diels {I.e. p. 279). With them goes also Onychmvi 

 Kaulf., which is ranked as a section of Cryptogramme, differing chiefly in 

 the formation of coenosori. The occurrence of fossils described as Onychiopsis 

 from so early an horizon as the Wealden appears to support this reference. 

 For reasons stated at length above (pp. 72-74), it appears probable that 

 these genera represent derivative lines which originated in relation to such 

 antique types as the Osmundaceae, and particularly Todea: nevertheless 

 some Schizaeoid relation for them is not to be excluded. 



It is more difficult to place Jamesouia with any confidence, for its 

 characters are so mixed. In its hairs, and the absence of scales, as well as 

 in its high spore-counts it possesses primitive features: in the Acrostichoid 

 spread of the sporangia it shows advance, while the instability of sporangial 

 structure indicates transition. Its anatomy is of a primitive type. It may 

 be held to represent an isolated genus, perhaps the sole representative of an 

 independent line in the plexus of the Gymnogrammoid Ferns. 



The same appears also for Ceratopteris. The vascular anatomy is of a 

 more advanced type than the rest, with its disintegrated leaf-trace, and in 

 large stems a medullary system within the dictyostele: these details together 

 with the presence of dermal scales place it structurally in advance of 

 Jamesouia. But the non-soral sporophylls, with their superficial thick-stalked 

 sporangia, and the wide latitude of the annular variations point to a 

 primitive and transitional state: nevertheless, the low but very variable 

 spore-output appears to suggest advance. Again the phyletic decision must 

 be left in the air. Ceratopteris may owe some of its peculiarities to its aquatic 

 habit: apart from this, which may account for its apparent isolation, it may 

 also be best regarded as the single representative of a transitional line from 

 relatively primitive Simplices. 



The salient interest of all these relatively primitive genera reflects 

 downwards rather than upwards. It would be a mistake, however, to say 

 that none of them suggest linkage with higher and more definitely Lepto- 

 sporangiate types: but the most distinctive features which they have in 

 common lie in the sporangia: these are all relatively massive- and short- 

 stalked. The numerical output of spores is low and variable, but the spores 

 themselves are large and constantly tetrahedral, while the segmentation of 

 their sporangia is that associated with transition from the more massive 

 types of the Simplices, and more particularly with those of the Schizaeoid 

 and Osmundaceous Ferns. 



The central group of the Gymnogrammoid Ferns presents less interest 



