Eastman — Dipnoan Affinities of Arihrodires. 139 



specialization than any existing Dipnoid.* But in Arthrodires, 

 on the other hand, all available evidence goes to show that 

 the skull was constructed upon essentially the same model as 

 in Neoceratodus / consequently the latter serves as a most valu- 

 able criterion for interpreting various structural details which 

 have heretofore been misunderstood. So far as can be deter- 

 mined from the interior of the headshield in Macropetalich- 

 thys,-\ Clielyopliorus^X Homosteus% and a few other genera, 

 the chrondrocranium of Arthrodires was even less ossified than 

 in Neoceratodus. It is certain that the parasphenoid was 

 largely cartilaginous, and for all that appeal's to the contrary, 

 the palato-pterygoid elements must have remained entirely so. 

 The presence of a pineal gland, sometimes but not always 

 communicating with the external surface, is clearly indicated 

 in the Arthrodiran skull, its position being as in Neoceratodus. 

 There are also conspicuous ridges on the under side of the cra- 

 nial shield, both in Arthrodires and in Neoceratodus, which 

 extend forwards and inwards from the posterolateral angles, and 

 give off descending processes in front. These are seen in the 

 recent form to furnish support for the palato-pterygoid dental 

 plates. Herein we have an explanation for the great solidity 

 of these ridges in Dinichthys, for it can scarcely be doubted 

 that they served a corresponding function as regards the 

 powerful shear-teeth of that form. 



We have next to speak of the dermal bones forming the 

 cranial roof. Their origin is admitted to have been through 

 fusion of numerous small dermal plates ; but that which is truly 

 remarkable, and claims our closest attention, is that the pri- 

 mordial Dipterws-\ike plates should have become reduced prac- 

 tically to the same extent, and rearranged almost exactly in 

 the same fashion, both in Arthrodires and Neoceratodtis. 

 "When was this simplification brought about % Amongst Cteno- 

 dipterines, despite their specialization in other respects, we 

 know that the mosaic pattern of cranial roofing bones persisted 

 as late as the Carboniferous. Amongst Arthrodires, reduction 

 had already taken place in the lower Devonian, after which a 

 fairly uniform pattern was adhered to. Amongst Ceratodonts, 

 we have yet to learn how their cranial plates were arranged 

 anterior to the Trias. Manifestly modern Dipnoans cannot 

 be derived from both Otenodipterines and Arthrodires ; and if 

 descendants of the former, how are we to explain the extraor. 



*Traquair, E. H., On the genera Dipterns, Palfedaphus, Holodus, etc., 

 Ann. Mag. Nat. Hist., ser. 5, vol. ii (1878), p. 5. 



fCope, E. D., On the Characters of some Palaeozoic Fishes, Proc. U. S. Nat. 

 Mus., vol. xiv (1891), p. 453, pi. xxix. 



JEichwald, E. von, Letluea Rossica, vol. i (1860), p. 1529, pi. lvii, figs. 1, 2. 



§ Woodward, A. S., Note on some Dermal' Plates of Homosteus, Proc. 

 Zool. Soc, 1891, pp. 198-201. 



