part 2] PREGLACIAL FLORAS FROM CASTLE EDEN. 127 



•Rosacea (Pomotdeje) genus (?) sp. (PL VIII, figs. 29a-29c.) 



A small three-carpelled ('?), synearpous fruit (or possibly when 

 ripe breaking into cocci), funnel-shaped, coriaceous, rough with 

 irregular longitudinal rugosities ; calyx five-toothed, adherent 

 almost to the apex ; carpels semiovate, facetted, two-seeded ; seeds 

 fiattish, rough, attached by the inner edge. 



Length = 1 mm.; breadth — -{y mm. 



The fruit broke while it was being examined, and part of it 

 crumbled ; but the base and one of the cocci were preserved. The 

 coccus is broken dorsally, and shows two minute seeds, of which 

 one later became loose. 



The fruit suggests Kosacese of the sub-order Pomoidese, but I 

 have been unable to identify it. 



Potejntilla akgextea Linnams. (PL VIII, fig. 30.) 



Length = '75 mm. ; breadth = - 6 mm. 



A great number of minute carpels of Potent Ma agree in every 

 respect with those of P. argentea. At Tegelen a single carpel was 

 found ; at Castle Eden it is the most abundant fossil. Both the 

 smooth and rugose types occur. 



It inhabits the northern temperate regions of both worlds, its 

 habitat being hedge-banks and dry gravelly places. 



Poxentilla pliocenica, sp. uov. [ = -?■ sp. Bidart, -pi. vii, 

 fig. 9.] (PL VIII, figs. 31 & 33.) 



Carpella ovata sive reniformia, hevia, tenuiter foveata. 



Carpels subovate, inflated ; dorsal margin rounded, scarcely 

 keeled ; ventral margin convex below, concave above ; apex hooked, 

 and turned towards the ventral margin ; surface-sculpture con- 

 sisting in very minute regular pits over the whole surface : one 

 specimen shows faintly the elongate, irregular, curved ribs found in 

 most species of Potentilla, though by no means in all individuals 

 ■of these species. 



Length = l - 1 mm. ; breadth = - 7 mm. 



This species is not British. It is identical in every respect with 

 •a species described from Bidart near Biarritz, 1 the deposit being 

 correlated by us with the Reuverian (PL VIII, figs. 32 & 34). 



Both at Bidart and at Castle Eden two forms occur, the hooked 

 form described above, and a rounded form. The sculpture in both 

 these forms, at both places, is the same. Whether it is a 

 species which developed achenes of two shapes, or whether they 

 represent two species both of which occur at Bidart and Castle 

 Eden, I am uncertain, but think the former the more probable 

 •explanation, though I do not know any recent species that has 

 two forms of achenes. 



1 Jules Welch, ' Les Vallees Pliocenes avec Lignite de Bidart, &c.' Bull. 

 Soc. Geol. France, ser. 4, vol. xv (1915) pi. vii, fig. 9. 



