﻿Vol. 64.] FOSSIL PLANTS FROM SOUTH AFRICA. 95 



example of Tli. odontopteroides represented in PI. IV, fig. 1 is 

 preserved. The Cala-River specimen (PI. Y, fig. 2) consists of 

 part of a pinna 15 centimetres long, bearing obliquely-attached 

 pinnules entire in the apical portion of the pinna, but for the most 

 part lobed. The midrib is clearly defined, but the lateral veins are 

 only faintly indicated in a few places. 



As regards the affinity of the plant, there is a close resemblance 

 in habit, which may amount to specific identity, with Th. incisa 

 figured by Saporta ^ from the Lower Lias of France, and by Count 

 H. zu Solms-Laubach '" from the Rhaetic of Chile. A dichotomously- 

 branched piece of a frond, described by Feistmantel, from New 

 South Wales, as Gleiclienia duhia^^ bears some resemblance to 

 Thinnfeldia splienopteroides . ' 



On the other hand, it is possible that these fossils should be 

 placed in the genus SpJienopteris, many species of which they 

 resemble. The discovery of more complete examples may confirm 

 this ; but the general appearance of the pinnae, with their stout 

 axes, would seem to favour the choice of Thinnfeldia. 



Thinnfeldia sp. (PL II, figs. 2 & 3.) 



The collection includes several pieces of pinnae from the Molteno 

 Beds of Maudesley, Lady Grey, which appear to be specifically 

 distinct from the other specimens referred to Thinnfeldia ; they are 

 too small to justify the institution of a new name, and it is not 

 clear whether they are more correctly named Thinnfeldia or 

 Pachyj)teris.^ 



The example shown in PL II, fig. 2 consists of a winged axis, 

 bearing entire oblique pinnules, in each of which there appear to 

 be a few diverging veins. PI. II, fig. 3 represents a fragment 

 in which the lobes are provided with dichotomously-branched 

 veins arising as a single strand from the axis of the j)inna. These 

 fragments may all belong to one species ; but the identification of 

 terminal pieces of pinnas must be attended with considerable risk 

 of error. 



Genus Dan^opsis. 

 Dan^opsis Hughesi, Feistm. (PI. YI & text-fig. 5, p. 96.) 



The very imperfect impression from the Burghersdorp Beds of 

 Lady Frere, 1500 feet below the Molteno Beds, represented in 

 PI. YI (natural size), consists of part of a rachis with portions 

 of pinnge, the longest of which measures 14 centimetres in length. 

 The axis of each pinna forms the midrib of a broadly-linear segment 

 attached to the rachis by the whole of the base. None of the 

 segments are complete ; the lamina, which reaches a breadth of 

 3*3 centimetres, is entire or slightly lobed. It is impossible to 



1 Saporta (73) pi. xlii. 



2 Solms-Laubach (99) pp 599-600 & pi. xiv, figs. 3-4. 



3 Feistmantel (90) p. HI & pi. xxvi, fig. 3. 



* Cf. Dichopteris visianica, Zigao (56) vol. i, pis. xii & xiii. 



