~ 574 — 
the presence in the Brazilian coal measures of coniferous types re- 
lated to those of the Permo-Trias elsewhere. Three of the best pre- 
served of these fragments areshown in figs. 11, 12 and 13, on Pl. 
VII. The leaves, as shown in fig. 13, are lanceolate, broadest a little 
above the decurrent base, then tapering upward to a narrow acute 
apex. They are concavo-convex, rourided dorsally but not carinate, 
and often somewhat incurved or even somewhel uncinate. The mi- 
drib is clear and persistent. In one fragment, too obscure for iltus- 
tration, some of the leaves are much more open and slightly recurved, 
suggesting Volfsia heterophylla. 
The fragment with broader and shorter leaves shown in Pl. VIII, 
fig. 11, lies on the same stone with that last mentioned and the 
original of fig. 12. Since its uppermost leaves are elongating, and 
since it occurs associated with the twigs with longer leaves, it is per- 
missible to regard it as presumably belonging to the same type, in 
which the leaves are perhaps dimorphic. 
Among other conifers described from the regions of the Gondwana 
flora, our fragments with the longer leaves are comparable to the 
most slender Karharbari specimens, described and illustrated by 
Feistmantel (1) as Voltzia heterophylla Brongn., or the material pu- 
blised by the same author (2) as Palissya indica from the upper 
Gondwana series. 
Notwithstanding the close relationship of our fragments to those of 
the Indian Volézia, or with the same especies as described by Schimper 
and Mougeot (3) from the Bunter of the Vosges, I hesitate, in the ab- 
sence of the larger forms of the species, to refer it to this type. 
Comparison may also be made with the Palissya Braunit Endl. 
as described by various authors, and with Goeppert’s Walchia linea- 
rifolia from the Permian (4) of Saxony and Bohemia. 
The fact that the Brazilian twigs do not appear to branch disti- 
chously stands against a reference to Walchia, though the aspect of the 
fossil is very close to that of some of Géppert’s specimens. 
The characters offered by the fragments in hand make it probable 
that further discoveries will show them to belong to either the Arau- 
(1) Fl. Gondwana Sys‘. vol. III. p. 30, pl, XXIV, fig. 4. pt. %, p. 122 pl.XLVIL\, 
figs. 20, 22, 24, 
(2) Op. cit., vol. I. pt. 2, p. 27, pl. XV. fig. 15. 
(3) Monogr. pl. foss. grés bigarré d. Vosges, p. 25, pl. vill, pl. ix, pl. x. 
(4) Foss. Fl. d. Perm, Form. pag. 242, pl. I figs. 7, 41. 
