MAJOR EARTH FEATURES 
195 
is offered by the character of the Permian and Triassic conti¬ 
nental sediments in Brazil and South Africa. They are in no way, 
as has been often thought, desert deposits; but as Koken has 
shown to the contrary, indicate abundant water, river plains, 
swampy and sea areas as the environment of their deposition. 
Nothing suggests the predominance of the desert climate which 
we should expect in such a mighty continental realm as the 
majority of the paleogeographers construct for the Triassic area 
of the southern hemisphere.^^ 
A land union of anterior India with South Africa over Mad¬ 
agascar during Triassic times must be accepted on zoogeographic 
grounds, since the former has been peopled from Europe {Belo- 
don, Hyperodapedon, Thecodontosaurus, Labyrinthodonts of very 
close affinities with Metopias and Capitosaurus), as well as from 
Africa (Massospondyhis, Dicynodon, Bothriceps). The union 
with Madagascar, in view of the appearance of Megalosaurus and 
Titanosaurus, must have lasted until Cretacic time, when the con¬ 
nection with South Africa was interrupted by the Strait of Moz¬ 
ambique, which in all probability was opened during the Liassic 
epoch. Anterior India and Madagascar continued from Liassic 
into Cretacic time as a long narrow island between the Indian 
Ocean on the east and the western Ethiopian mediterranean of 
Neumayr, which was a southward extension of Tethys. 
Toward the northeast stood the continental platform of anter¬ 
ior India, with the spur of Assam extending farthest into Tethys. 
At various times during the Triassic Period this continent was 
united with the massif of Camboda — to which belonged the great¬ 
er part of the island of Borneo — and also with Australia. This 
connection, which must have lasted with interruptions at least 
into Late Jurassic times {Megalosaurus in Victoria), apparently 
lay across the Sunda Islands. 
All these Triassic land connections which have been established 
as a result of paleogeographic investigations, decrease, of course, 
the size of the Indian Ocean but only in so far as they require its 
separation from the Arabian Sea by a small peninsula, the re- 
26 Indisches Perm und permische Eiszeit: Neues Jahrb. fiir Mineral., etc. Fest- 
band 1907, p. 526. Koken in his map marks the South Atlantic continent with a 
query, and restricts the continent of Gondwana considerably in the region of the 
Indian Ocean. 
27 This exaggerated change in the distribution of continents and seas during Trias¬ 
sic times has been avoided by F. Waageni (Unsere Frde). He gives, however, no 
evidence for his reconstructions. 
