328 SECTIONAL TRANSACTIONS.— C. 



of the Jurassic. The isolation of the fragments was complete by the end of the 

 Cretaceous, and evolution proceeded independently in each portion. Particular 

 biological affinities noted by zoologists and botanists may be explained by migration 

 along temporary connections formed bj' folding in advance of the drifting crustal 

 blocks. 



This new concept of Gondwanaland revolutionises all extant ideas of palaeobiology 

 and seems to provide a golden key to the otherwise puzzling distribution of life in the 

 southern hemisphere. 



Dr. A. B. Walkom. — The remarkable similarity in the development of the 

 Glossopieris and Thlnnfeldia floras, in AustraUa, South America and South Africa, 

 proves land-connection between these continents in late Palaeozoic and early Mesozoic 

 times. The conception of ' Gondwanaland ' as an extensive east-west continent on 

 the southern side of an extended ' Tethys ' sea is difficult of proof or disproof. W. D. 

 Matthews has suggested the idea of the dispersal of faunas and floras along radial land- 

 masses from an arctic continent. If there had been an Antarctic land-mass connected 

 from time to time with South America, South Africa, Australia and New Zealand, 

 faunas or floras originating in the south would be able to migrate into those land- 

 masses with which there was connection at the time. Little or no mixture would be 

 possible in the polar regions, but would be progressively more complete away from 

 the polar continents. This suggestion is supported by admixture of northern plants 

 with the Glossopteris flora in certain of the southern land-masses. 



The Ohssopteris flora originated in the south in Upper Carboniferous times and 

 spread northwards, occurring in post-Permian rocks. The Thinnfeldia flora also 

 originated in the south in Lower Triassic times, and spread northwards, becoming 

 extinct at the end of the Triassic in the south, but persisting until the Jurassic in 

 Germanjf where the species are, however, considerably different. It appears possible 

 to account for most problems of distribution by migration from two polar continents 

 along longitudinal continental masses, remnants of which are seen in the existing 

 continents, and which were in temporary communication with the polar continents, 

 but which were not always continuous from north to south. 



Dr. R. Broom, F.R.S. — The resemblances of the faunas and floras of South 

 America and South Africa are infinitely greater than any resemblance they have to 

 the Permian faunas and floras of the north. It is unnecessary to discuss whether the 

 parts of Gondwanaland were originally in the positions they now occupy, or whether 

 they have drifted apart from a central mass as on the Wegener hypothesis ; the fact 

 remains that there tcere southern lands which had very similar floras and faunas. The 

 cases of transference of northern types to the south and vice versa are quite excep- 

 tional, and for long periods the northern and southern lands were completely separated. 

 If we knew nothing of Gondwanaland we should have to invent it as a theory to 

 explain the sudden appearance of new life-forms in the north. The Lnciu-sion of 

 higher types of hfe from the south appears to have given rise to the Mesozoic epoch 

 in the north, and almost certainly it was another incursion from the south that ended 

 the Mesozoic and ushered in the Tertiary. In America, in the basal Eocene, a large 

 number of small mammalian types appear suddenly. They have no ancestors in 

 the northern Cretaceous. Later, another wave of immigration resulted in the intro- 

 duction of a large number of new and higher mammals in the Eocene : aU these had 

 originated in a great southern continent. Gondwanaland has been far more important 

 in the world's historj' than any of the northern lands, and most of the great advances 

 have originated in the south. 



Prof. A. C. Seward, F.R.S. — The geographical distribution of Glossopteris cannot 

 be explained without serious interference with the present relations in space of the 

 several parts of Gondwanaland, a once continuous continental sheet that has become 

 disrupted into isolated units. The earliest strata containing the Glossopteris flora are 

 almost certainly Upper Carboniferous : the genus flourished "wdthin a few hundred 

 miles of the pole, in the Falkland Islands, in South America, South Africa and 

 Madagascar, in India and Afghanistan. The genus Linguifolium, from the lower 

 Mesozoic rocks of New Zealand, is possibly generically identical with Glossopteris. 

 By the Rhaetic Glossopteris had migrated to Tonkin, Southern Sweden and Eastern 

 Greenland. The association of Gangamopiteris and other Gondwanaland genera with 

 northern Permian plants in the Kusnezk flora of Siberia resulted from the wandering 

 of southern types across the Equator. Prof. Halle's memoir on the Shansi flora and 

 the discovery of outhers of the northern Permo-Carboniferous flora in Sumatra and 

 the Malay Peninsula show the wide distribution of the late Palaeozoic forests, and 



i 



