— 515 — 
line of fracture. The nervation, which is typical in the lower part, 
becomes more oblique and arched than in any other fragment in the 
collection. This relatively great obliquity is undoubtedly due in part 
to tear and strain in the upper part of the leaf, where the lamina bears 
evidence of contraction near the abraded midrib. 
The top of a leaf shownin Fig. 1, illustrates both the form of the 
apex, which is here unusually sharp pointed, and the nervation which, 
as hefore, is finely Taeniopteroid in aspect, and nearly at right angle to 
the border. 
Among other species of the genus the Brazilian plant seems to 
find its closest relations in Bunbury’s Glossopteris stricta from Nagpur, 
India, and Feistmantel’s Glossopteris taenioptcroides from the upper 
coal measures of New South Wales. It is comparable to Bunbury’s 
specimen (1) in the elongation of the leaf, the origin and attitude 
of the nervation, and the relatively great density of the nervation. 
Also the example shown by Feistmantel in Fig. 1, plate xxxviiA, of 
the second part, third volume, of the Gondwana Flora(2) is strikingly 
similar in midrib and nervation to our leaf. Feistmantel’s Glusso- 
pteris taeniopteroides (3) is Taeniopteroid, but coarsely nerved with 
meshes of equal length. The specimen from near Nagpur, illus- 
trated by Arber (4) as Glossopteris stricta has its nervation far more 
oblique than that of any specimen of our series, and is less Taenio- 
pteroid. 
In the taeniate form and the broad vascular band our species 
reminds one somewhat of the Gangamopteris Kashmirensis of 
Seward (5) and 1 am slightly inclined to regard the last named species 
as more nearly related to Glossopteris than to Gangamopteris, as shown 
by the petiole, the continuity of the central vascular band, the form of 
the leaf and the nervation. Glossopteris occidentalis, further, is readily 
distinguished from all the above named species by its broadly 
cordate sessile base, as well as, to acertain extent, by its bluntly 
acuminate apex. 
Gangamopteris (?) Kashmirensis Seward and Glossopteris ccciden- 
talis, together, perhaps, with Glossopteris decipiens, are to some extent 
——_-__—__ 
(1) Quart. Jour. Geol. Soc., vol. xvii, 1861, p. 334, pl. ix, fig. 5. 
(2) See also vol. III, pt. 1, pl. xxifg. 41. 
(3) Mem. Geol. Sur. N.S. W., Palaeont., no, 3, pl. xviii, fig, 4. 
(4) The Glossoptcris Flora, p. 76, pl. iv, fig. 1. 
(5) Pal, Indica, n. s.,vol. no. 2, 1905, pl. VIII, (see figs. 1-4). 
