206 
POPULAR SCIENCE REVIEW. 
1878, p. 373) they regard them as forming probably an order, Dolerophylleae 
distinct from the Salisburiese and Cordaitese, although specially allied to the 
latter through certain forms recently described in America by Professor 
Lesquereux. The leaves, which have generally been mistaken for leaflets 
of Neuropteroid ferns and described under the names of Cardiopteris, Cyclo- 
pteris, Nephropteris and ApMebia , differ clearly in structure from those genera. 
They are simple, sessile, oval, or rounded, eared at the base, thick, encircled 
by a cartilaginous border, and traversed by numerous dichotomous nervures, 
which radiate from the point of attachment towards the margin and often 
bifurcate several times. The nervures were enclosed between two thick 
laminae of epidermis, and are especially characterised by the abundance of 
gummous ducts, the true structure of which is not yet determined, but which 
accompany and surround the vascular bundles, often, when carbonised, form- 
ing filaments substituted for the true nervures which they conceal. This 
character recurs, although less strongly marked, in the Cordaitese, The 
insertion-scars produced on the stems by these leaves must have been rounded 
or transversely ellipsoidal, and such scars occur upon the stems of many so- 
called Calamodendrese. 
The reproductive organs ascribed to the Dolerophylleae by M. Renault are 
very singular ; and, while differing much from what are ordinarily observed 
in Phanerogamia, they indicate the existence of a series of plants in which 
fecundation was effected by means of corpuscles nearly agreeing, although of 
much larger dimensions and more complicated structure, with the pollen 
grains observed in the micropyles or pollenic chambers of several Palaeozoic 
Gymnosperms. 
Thus, according to M. de Saporta, the Dolerophylleae represent in the 
Carboniferous flora, which was formerly supposed to be so predominantly 
cryptogamic, a new phanerogamic element ; in fact, they were related to a 
whole series of prototypic phanerogams of which the Sigillarieae probably 
also formed a part. 
A Gigantic Triassic Reptile from the Cape . — Under the appropriate name 
of Titanomchus ferox Professor Owen has lately described before the Geolo- 
gical Society a gigantic reptile belonging to his order Theriodontia, derived 
from the same South African Triassic beds which have already furnished him 
with so many remarkable and interesting forms of the same class. 
He stated that among the fossils recently sent to the British Museum from 
the Cape of Good Hope by Mr. T. Bain there were two boxes containing 
specimens of a most unpromising character, there being in them no entire 
bones, but only numerous more or less water-worn fragments. Among these 
was found a portion of a maxillary showing some traces of teeth ; and sec- 
tions having been made of this bone, the remains of several teeth were dis- 
played, including a canine, the preserved portion of the socket of which was 
inches long. From the number and mode of implantation of the teeth, 
he concluded that the animal to which they belonged resembled the Therio- 
dont genera Galesaurus and Galenops. The anterior portion of the left 
ramus of the lower jaw, measuring 7\ inches in length, showed teeth pre- 
senting close analogies with those of Theriodonts, and this alliance was con- 
firmed by the study of other fragments. Some of the characters presented 
by these remains seem to suggest affinities with the carnivorous mammalia, 
