324 PROP. A. C. SEWARD AND MP.. J. WALTON ON [vol. IxxiX, 



The late Prof. E. Zeiller, who spoke with exceptional authority 

 on the taxonomy of fossil plants, included his Tongking specimens 

 in G. indica, and it is noteworthy that his definition 1 of the species 

 fits the characters exhibited by such a leaf as that reproduced in 

 our fig. 15. Glossopteris decipiens Feistmantel and the leaf 

 from Vereeniging, originally described as Gangamopteris cyclo- 

 pteroides, are from Lower Gondwana rocks, and are associated 

 with the oldest members of the Permo-Carboniferous flora ; 

 Gangamopteris (?) angarica Zalessky is from beds believed to be 

 of Permian age, and the Tongking specimens included by Zeiller 

 in Glossopteris indica are from the highest Glossopteris- bearing 

 strata, probably llhsetic in age. In view of these facts, it is 

 obvious that the type of leaf that we have called G. indica, 

 cf. G. decipiens, cannot be regarded as a decisive criterion of 

 geological age. It is tempting to interpret the variation from the 

 more typical G. -indica type in the direction of Gangamopteris as 

 evidence of greater antiquity, on the ground that leaves with a 

 more complete midrib were probably later developments than forms 

 without a clearly marked distinction between midrib and lateral 

 veins. On the other hand, if the beds from which the aberrant 

 specimens were obtained were homotaxial with the plant-beds of 

 Tongking, one would expect to find in association with them other 

 members of an Upper Triassic or Rhsetic flora. The balance of 

 evidence is, perhaps, in favour of assigning the type of leaf from 

 North Arm (PI. XXI, fig. 15) to a position below that of the beds 

 containing the more typical examples of G. indica. 



In the centre of an incomplete leaf reproduced in fig. 14 

 (PL XXI), identical in venation with that shown in fig. 15, there is 

 a shallow linear depression 3 cm. long, separated by a constriction 

 from an approximately circular and deeper depression higher on 

 the lamina. This feature may be due to the pressure of a young 

 and partly expanded leaf against the base of the larger frond, the 

 deeper circular depression being the impress of the infolded apex of 

 the immature leaf. 



Some specimens collected by Dr. Baker from G-eorge Island and 

 the Bay of Harbours are clearly identical with the single impres- 

 sion figured by Dr. Halle, from a locality south of Dos Lomas, as 

 Gangamopteris cyclopteroides var. major Feistmantel 3 ; but, in a 

 few of the recently discovered examples, both the basal and the 

 median portions of the leaf are preserved. At the base of the 

 leaves the lamina is narrow, and there is no separation into lateral 

 veins and midrib ; but higher in the leaf a midrib is clearly 

 shown, and the lateral veins form anastomoses of the Glossopteris- 

 Broioniana type : that is, the meshes are less uniform in size, and 

 their upper and lower boundaries are not so straight as in G. indica. 

 We are of opinion that if more of the specimen figured by 



1 Zeiller (03) p. 85. 



2 Halle (11) pi. viii, figs. 8 & 9 



