E. A. Neivell Arber — The Glossopteris Flora. 347 



and representatives of a genus, of at present unknown affinities, but 

 probably belonging to the Cordaitales, NoeggeratJiiopsis. 



In India, all these four types occur in the Panchet and Damuda 

 series of the Lower Gondwanas.' In Australia, Glossopteris and 

 Phyllotheca are abundant in the Newcastle series of New South 

 Wales, and in other localities and horizons.^ Phyllotheca was first 

 described by Brongniart^ from Australian specimens. Noeggera- 

 tJiiopsis also occurs in Australia, but ScJiizoneura, although recorded,* 

 is apparently rare. 



In South Africa, three of these types are found,^ and stems 

 agreeing very closely with the Indian Schizoneura have been 

 recognized.^ It may be also pointed out that a fossil, discovered 

 by Bain in the Eoggeval (Fish Kiver), South Africa, and described 

 and figured by Sir Joseph Hooker' in 1852, has some of the 

 characteristics of a ScJiizoneura. This specimen is preserved in 

 the museum of the Geological Society of London, where I have 

 recently examined it. It shows two whorls of linear-lanceolate 

 leaves united into a cup or sheath at the base. The two whorls, 

 however, are quite separate, and on opposite sides of the rock- 

 specimen ; whereas in Hooker's figure, which is largely a restoration, 

 they are represented as if in continuity and alternate. The free 

 segments are of unequal length and breadth, and each is traversed 

 by a few parallel and distant nerves. Two of the leaves apparently 

 show some signs of splitting towards the apex. The stem characters 

 are unfortunately not shown, and it is therefore not possible to 

 definitely refer the African specimens to the genus ScJiizoneura, as 

 the evidence of the leaves alone somewhat suggests. 



In South America, all the chief members of the Glossopteris flora 

 have been recorded in recent years,^ with the possible exception of 

 ScJiizoneura. 



The floras of the various regions of Gondwana-land are therefore 

 all of the same type. Eecent researches have also shown that 

 many cases of specific identity occur among the fossil plants of 

 these widely separated areas. Glossopteris Broioniana, Brngt., is 

 found in India, Australia, and South Africa,^ Glossopteris (Ganga- 

 mopteris) cyclopteroides, Feist., and NoeggeratJiiopsis Hislopi (Bunb.), 

 in India, South Africa, and South America, and possibly also in 

 Tasmania. A close identity exists also between other members of the 

 Glossopteris flora, as is shown by the occurrence of Neiiropteridium 



1 Feistmantel : Foss. Flora Gondwana System, 1881, vol. iii : Pal. Indica. 



2 Feistmantel: Mem. Geol. Sui'v. N.S. Wales, 1890, Pal. No. 3. David: 

 Proc. Limi. Soc. JST.S. Wales, 1894, ser. ii, vol. ix, p. 249. 



3 Brongniart: Prodi'ome Hist, veget. foss., 1828, pp. 151 and 175. 

 * Etheridge : Eec. Geol. Sm-v. N.S. Wales, 1893. 



5 Zeiller: Bull. Soc. geol. France, 1896, ser. in, vol. xxiv. 



6 Seward: Q.J.G.S., 1897, vol. liii, p. 324. 



■' Hooker: Trans. Geol. Soc, ser. ii, vol. vii(pub. 1856), p. 227, pi. xxvii, fig. 1. 



8 Szajnocha : Sitz. k. Acad. Wiss. Wien, 1888, Bd. cxvii, p. 219. Kurtz : Ee-sasta 

 del Museo de la Plata, 1894, vol. vi, p. 123 ; 1899, vol. x, p. 43 ; Geol. Mag., 

 1896, p. 446. Bodenbender : Zeit. Deut. Geol. Ges., 1896, Bd. xlviii, p. 743. 



■* Zeiller: Bull. Soc. geol. France, 1896, ser. iii, vol. xxiv, p. 349. Seward: 

 Q.J.G.S., 1897, vol. liii, p. 315. 



