330 Scientific Proceedings, Royal Dublin Society. 



This identification was supported by comparison of the restored epidermis 

 with that of Ash. The subsequent discovery of the quinately divided leaf of 

 Detvcdc[nea at 903 feet, with typical epidermal structure, seemed to indicate 

 the necessity of a revision of the first view. In both Dewalquea and Fraxinus 

 the venation, epidermis, stomata, and peltate scales (glands) all agree 

 sufficiently to allow approximation ; and we are of opinion that Dcivalquea 

 is more naturally associated with the Oleaceae than with the Eanunculaceae. 

 There is, however, another line of affinity which appeals more to us. 



PI. XI, fig. 9, shows a well-preserved fossil leaf or leaflet, 2x4 cm., 

 approximately, coriaceous, broadly oval-lanceolate, margin entire below, but 

 sinuous- dentate in its upper two-thirds, midrib pronounced, thinning out 

 apically ; secondary veins fairly numerous but delicate, sub-opposite ; angle 

 of divergence, 70°-76° ; adjoining secondary veins united towards the leaf 

 edge by loops or their own forkings, and forming a marginal network, from 

 which veins pass into the teeth ; the whole system camptodrome. There are 

 no pronounced cross-anastomoses, but there is a fairly well-developed 

 vascular network forming polygonal areas between the secondaries. 

 Shortened secondary veins occur. Externally this leaf shows great resem- 

 blance to Ilex celastrina Sap. (13), fronr the Upper Oligocene of Saint Jean 

 de Garquier and Armissan in the south-east of France. As the name 

 suggests, Saporta saw affinity in his specimen to Gelastrus also, and admits, 

 in naming it, that it is fairly remote in its characters from the Ilex of 

 to-day. 



If we were confined to a comparison of the external characters only, we 

 should feel compelled to iiame our specimen Ilex celastrina. Eestoration of 

 the leaf tissue, however, alters our attitude. Markedly wavy walls are not 

 found in Ilex epidermis, nor are peltate scales found in any of the Aquifoliacea^, 

 according to Solereder (14). Of the many other possible connexions, the one 

 combining most of the external and anatomical features is the genus Carya, 

 e.g., the modern Carya lccciniosa,in which, however, tufts of hair, not known in 

 our fossil, occur. Our specimen shows, however, such general agreement with 

 Beivalquea in its external features and minute structure that we feel compelled 

 to refer it to this genus and not to Carya, and to name it Dewalquea denticulata. 

 We are of opinion that the leaf fragment named by Heer, Pterocarya 

 denticidata ? (15) {Jit.ylans denticulata 0. Weber) from Bovey Tracey {op. cit., 

 PI. LXX, fig. 5) is near to, if not identical with, D. denticulaia; and that the 

 Carya hilinica Ettings (16) is also probably nearly allied to it, though 

 differing in form. A re-examination of the collections of fossils in our 

 museums, after restoration by maceration, would reveal many hidden 

 affinities. 



