— 539 — 
The specimens in hand are from the Gangamopteris, bed, nor- 
theast of Minas, Santa Catharina. They occur in every instance in 
the most intimate association with the Samaropsis, both fossils 
being surrounded by great numbers of Gangamopteris obovata. 
In India the inflorescence (Arberia) is said to be associated with 
Noeggerathiopsis Hislopi. Cardiocarpon is regarded as exclusively 
a Cordaitean genus, and the actual union of these seeds with the 
dissected scale leaves would argue fora relation to Noeggerathiopsis, 
a genus so closely connected with Cordaites itself, that its generic 
separation has been questioned by some paleobotanist. The evidence 
of association, on the other hand, points strongly toward Ganga- 
mopteris. 
In any event the Gondwana type of inflorescence differs so 
widely from the Antholithus of Cordaites that, if definitely corre- 
lated with Noeggerathiopsis, it would leave no room for doubt as 
to a wide generic distinction between the leaves of the northern 
Cordaites and Gondwana Noeggerathiopsis. Incidental to the same 
correlation the very distinct specific differences between the Indian 
and Brazilian types of Arberia would involve a corresponding specific 
difference in the leaves, an alternative that, further, would be urged 
by the specific differences in the associated Cardiocarpa. 
There appears to be a varietal if not specific difference between some 
of the Indian and the Brazilian forms of Gangamopteris ; but not so 
with Noeggerathiopsis. However, it must be remembered that in the 
Northern flora the number of species of Cardiocarpon is many times 
that of the species as yet distinguished in the leaves of Cordaites. 
Pending the discovery of additional evidence the relationship of the 
type of inflorescence under consideration must remain undetermined. 
Consequently its reference to the Gymnosperms or the Pteridosperms 
will rest in doubt, the genus Gangamopetris being, in my judgment, 
Pteridospermic. Meanwhile for the systematic designation of this 
very interesting and well marked type, I propose the generic name 
Arberia, a name chosen in recognition of Professor Arber’s very 
important contributions to our knowledge of the Pteridospermic 
fruits, as well as for his valuable monographic work on the Glos- 
sopteris fiora Thetype of the genus is the Brazillian species, Arve- 
rig minasica. The Indian type as figured and described by Feist- 
mantel (1) may be specifically mamed Arberia indica. 
4) Fl. Godwana Syst., vol. III, p. 59, pl. XAVII, fig. 5. 
