SAPINDACEAE. 



GENUS? 

 PI. XI, fig. 17. 

 The fruit shown in fig. 1 7 belongs to one of the Sapindaceae with 3 carpels strongly 

 winged, probably in the section Paullinieae. The fruit is oblong, and coarsely reticulate 

 on the sides ; the wing is only partly preserved, but shows a strong dorsal vein and 

 coarse reticulation. We cannot identify the genus, and the specimen is too incomplete 

 for full description. 



Length of carpel (without wing) 4.0 mm., breadth 2.5 mm. Reuver. 



STOCKSIA? 

 PI. XI,. fig. 18. 



A small hard seed of unusual structure has been found at several localities in 

 Pliocene deposits. It was first discovered at Tegelen in 1905 (op cit. 1907, fig. 124). 

 Next several specimens were found at Raevels, in Belgium, in material collected by 

 Baron Greindl and M. Pierre Gerimont. The deposit was thought to be of the same age 

 as that found at Tegelen; but unfortunately the locality yielded only this unknown seed, 

 and a seed of Alisma. A short description of the seeds from Raevels was published by 

 us in 1908*. Now we have the same curious seed from Reuver. 



Seed small globose or slightly facetted, with a large ovate germination*valve 

 having its apex at the micropyle ; testa thick hard, with faint longitudinal granulate 

 striation, in 3 layers, an outer thin black dense, a middle thick brown spongy with 

 occasional large cavities, and a very thin inner black layer (tegmen?); hilum fitted with 

 a conical plug. The inside of the testa, in the specimens we have been able to examine, 

 shows no clear impression of the embryo. 



Length 1.2 mm. Tegelen, Raevels, Reuver. 



The large germination*valve is the most striking characteristic in these seeds. We 

 have found valves of this sort in Nyssaceae and in Sapindaceae. Nyssa in other respects 

 is quite unlike our seed ; but in Sapindaceae section Koelreuterieae we find globose seeds 

 with germinationsvalves and plugs resembling our fossil. Among the living species of 

 Stocksia, however, the smallest seed we can find at Kew is much larger than our fossil. 

 In the living species of Koekeuteria the valve is strap*shaped and shows externally as 

 a raised area cut off from the rest of the testa by a groove ; in our fossil the valve is not 

 raised and would probably be quite inconspicuous in the fresh seed. As in the recent 

 Koekeuteria, the hilum shows a conical plug, which is pushed outward by the swelling 

 of the embryo in germination, and thus forces open the hinged valve. The testa of the 

 fossil resembles that of Koekeuteria in its hardness, thickness, and division into layers ; 

 but we can find in the recent seed no counterpart of the large cavities, seen in one 



* Les Elements Botaniques de la Determination de l'age des Argiles a Btiques de Tegelen, Reuver, 

 Ryckevorsel et Raevels. Bull. Soc. Beige de Geologie, tome 21, p. 589. 



112 



