elx Proceedings of the Asiatic Soc. of Bengal. [N.S., XVIT, 
of a narrower wing, are known as Cordaicarpus, but these, like 
Samaropsis, give no clue as to nature of the parent plant. The 
question of the attribution of all these seeds is therefore 
shrouded in mystery. 
From the Karhabari beds Feistmantel had figured a 
peculiar-shaped scale-like organ to which 
the name Arberia indica, and this now becomes especially 
interesting in the light of a recent discovery. Certain scale-like 
organs somewhat similar to the Indian Arberia were described 
only a couple of years ago by the Swedish geologist Lundqvist 
from rocks of a similar age in Brazil (Lundqvist 1919), and these 
show some Cordaicarpus seeds in organic connection. The point 
that I wish particularly to bring out is that these flat winged 
seeds are borne upon recurved pedicels suggesting a pendulous 
attachment favourable for dispersal by the wind. That Arbe- 
ria indica also bore reproductive organs, as hinted long ago by 
Feistmantel, and, moreover, that they were seeds of the Sama- 
ype, is at any rate a plausible sugges- 
tion. But the identity of the parent plant still remains obscure, 
for Arberia itself is only known in the detached condition. The 
stage (See Table II.) 
It will be noticed from the Table that the Karharbari flora 
contains a relatively very large number of plants whose affi- 
same neighbourhood and by Middlemiss from several other 
localities in Kashmir, is particularly interesting. Among the 
half-a dozen species discovered in a short time from Kash- 
resembles @. cyclopteroides, but the species of Psygmophyllum 
(P. Hollandi and P. Haydent) are not easily identified with any 
known Gondwana plants. It may be that these species will 
later on be discovered in more southern regions. On the other 
hand, it is possible that in this northern maritime province of 
Gondwanaland, which presumably enjoyed a more equable 
that in the interior of the continent. Any further light on this 
question would be of considerable interest. 
