Johnson & Gilmoke — Dewalquea in Coal-bore at Washing Bai/. 329 



Eanunculaceae, consisting almost entirely of herbaceous plants, would not 

 lend themselves to fossilization, and their past history is almost a blank in 

 the rocks, though, as members of the Polycarpicae, they must have been one 

 of the earliest families of the Dicotyledons to appear. Hellebore, with its 

 coriaceous leaves in some species, would be the one most likely to be found 

 in the fossil state. It would be distinctly satisfying if Deivalquea could be 

 accepted as the ancestral form of the family. Its peltate scales, combined 

 with other features, militate, however, against this view. 



The Araceae have been suggested as a possible line of affinity with 

 Dewalquea. This is admittedly one of the most primitive families of the 

 Monocotyledons, and in, e.g. Anthurium, there are several species with pedalo- 

 digitate leaves. Their venation differs, especially in the possession of a 

 marginal vein not found in Dcvxdquea. Their epidermis and stomata also 

 differ. Peltate scales are said to occur, but they are simply saucer-like 

 depressions, in which the secretion of the glandular epidermal cells collects. 



A third line ofafhnity is that indicated by Debey's name of Araliophi/Ilum. 

 The Araliaceae were rejected by Saporta and Marion on the expert advice of 

 Decaisne that leaves showing such segmentation were not then (as they are 

 now) known in the group. Schimper, as already mentioned, accepted the 

 generic name, and described the fossils as Araliaceous. Peltate hairs occur 

 in Hedem helix and in Oreopanax, but they are stellate, and do not suggest 

 the disc-like scale-hairs of Deivalquea. 



P. Principi (8), in his recently published work on the fossil Dicotyledons 

 of the Italian Oligocene, records the Belgian species, Devxdquca gclindcncnsis, 

 and a new species, D. yrandifolia. He follows Schimper and Zeiller (9) in 

 referring the genus to the Araliaceae. D. yrandifolia shows 5-7 leaflets, 

 each on its own stalk, i.e. a compound leaf with 5-7 pinnately arranged 

 terminal leaflets. 



The "Washing Bay flora is an early one from the Dicotyledonous point of 

 view, and one would hardly expect to find highly specialized families such as 

 the Oleaceae to be well represented in it. Thus in Fraximis excelsior the 

 corolla has passed through the stage of gamopetaly to suppression. Fraxinus 

 first appeared in the Arctic regions, and gradually spread southwards in the 

 Old and New Worlds. Heer (10) records F.praecox from the Upper Cretaceous 

 beds of Patoot in Greenland ; and Lesquereux (11) describes F. cocenica from 

 the Eocene of Golden, Colorado; while Ettingshausen and Gardner (12) hst 

 F. jovis and F. prae-savinensis from Alum Bay. The seventeen European 

 species recorded are nearly all referred to the Miocene. One of our speci- 

 mens (PL XI, fig. 6), found at 909 feet, had every appearance of being the 

 distal end of an Ash leaf, such as Fraxinus ixnnsylvatica var. lanceolata. 



