XXXIV] COMPARISON 319 



degree of certainty been recognised as related to these living Dipteridaceae. 

 The correspondence often appears so close as to justify their being included 

 within the Family. There are hardly any Ferns in respect of which the 

 comparisons with fossils serve so effectively to consolidate the phyletic 

 position of the living types. These are generally regarded as survivals from 

 Mesozoic times, at which period their fossil congeners flourished. 



The living species may themselves be seriated according to the characters 

 of leaf-architecture and of sorus, as relatively primitive and relatively 

 advanced. D. Lobbiana with its narrow forked lamina, its sori in a simple 

 row on either side of the midrib, and with the sporangia in each of them 

 produced simultaneousl}', may be held as a relatively primitive type. 

 D. quinquefiircata with slightly broader leaf-segments, and with its sori 

 scattered over the enlarged surfaces and showing signs of fission, may take 

 an intermediate place, leading on to the broadly webbed leaves oiD. conjiigata, 

 in which the sori are found to be of the "mixed" type. These features 

 illustrate a progression from a type of sporophyll characteristic of the 

 Gleicheniaceous affinity to one with leaf-surfaces of large area, and a sorus 

 characteristic of many later Leptosporangiate Ferns. The)- suggest that 

 Dipteris is a synthetic type, linking primitive with more advanced P'erns. 

 With this conclusion the vascular construction is in accord : the solenostele 

 and integral leaf-trace of D. conjiigata readily compare with those of 

 Gleichenia pectinata or Metaxya, though they fall far short of the special 

 elaboration in the more nearly related Matonia. On the other hand 

 D. Lobbiana presents in its disintegrated leaf-trace a sign of advance which 

 does not accord with its simpler leaf and sorus. But many instances may 

 be quoted where all the features used in comparison do not progress 

 uniformly. Lastly, the stiff spinous hairs suggest a middle position. They 

 are advanced structurally as compared with the simple hairs of Matonia, 

 but they are not widened out into those broad protective scales characteristic 

 of the later Leptosporangiates. 



On the other hand the small size of the sporangia, their almost vertical 

 annulus, lateral dehiscence, and the small spore-output appear to indicate 

 an advanced state. But the complete ring of the annulus and the four-rowed 

 stalk point in the direction of the Cyatheaceae and simpler Gleicheniaceae, 

 while the form and structure of the sporangia resemble those of Gleichenia 

 lineata. Against this comparison the low spore-output appears as an obstacle. 

 It is here that the recentl}' acquired evidence from fossils comes in with 

 special cogency. 



It has long been known that certain Ferns of Mesozoic age resembled 

 the modern Dipteridaceae. Nathorst includes under his separate Family of 

 the Campteridinae the genera Dictyophylluni, Thaiunatopteris, Caniptopteris, 

 and Clathroptcris, all of Rhaetic or Jurassic age. They were all large- 



