266 Proceedings of the Royal Physical Society. 
keel or wing-like border, which it appears to me may result 
from pressure on a pericarp of softer consistency than the 
nucule, though it is possible that the seeds were provided 
with true narrow wings. Nevertheless, it is difficult to 
see how Cardiocarpus, as here defined, can be satisfactorily 
separated from Cordaicarpus. 
To sum up the matter, it would appear that of the little 
seeds under discussion, some possessed a true surrounding 
membranous wing-like structure, while in others the narrow 
wing-like border or keel with which they are usually pro- 
vided in the fossil state, probably results from the compres- 
sion of a more or less soft envelope which surrounded the 
seed. 
In these circumstances, would it not be better, either to 
suppress Brongniart’s genus Cardiocarpus as insufficiently 
defined, and to use Samaropsis, Gdpp., for the seeds possess- 
ing undoubted membranous wing-like structures, and Cordai- 
carpus, Geinitz, for those whose apparent narrow wing or 
keel is the probable result of pressure, or tentatively, to place 
all these seeds in Brongniart’s Cardiocarpus? This last course 
is that adopted here. 
CARDIOCARPUS NERVOSUS, Kidston, n. sp. 
[Plate V., Figs. 3, 4, and 5.] 
Description.—Form of complete seed, cordate, flat; nucule 
acuminate, lenticular in transverse section, broadly winged, 
membranous wing with several simple or divided vein-like 
thickenings passing from the margin of the nucule to the 
outer edge of the wing. 
Remarks.——The best preserved of these little seeds is 
shown at Figure 3; its entire width is about 1 cm., the 
nucleus measuring about 5 mm., and the wings about 
2°5 mm. each. . 
At the base of the seed the wing narrows, and a notch is 
thus formed which gives the seed a cordate outline. 
None of the specimens show the apex sufficiently well 
preserved to determine whether the wing was continuous or 
cleft at the apex. The character which at once distinguished 
