— 367 — 
Excluding the woods and seeds, we find that over one half of 
the types are fairly distinctly Southern while not more than six, 
or about 15% of the species, lay distinct claims to even an indirect 
membership in the Northern flora (1), 
It is manifestly unwise, however, to attempat sharply to diffe- 
rentiate the Southern from the Northern floras, since in nuture there 
inevitably must have been land areas clothed with intermediate and 
partially intermingled floras, transitional between the interglacia flora 
of Australia and the Permian flora of western Europe and eastern 
North America. 
If we examine the stratigraphical occurrence or range of the 
species in the Brazilian sections we find the Gangamopteris flora 
immediately following the fbasal conglomerates or sandstones (2). 
Gangamopteris obovata is found in the roof of .black shale inter- 
bedded in the sandstone at a distance of but 6 meters above the gra- 
nite, near Suspiro, Rio Grande da Sul. Thus it appears that in Brazil, 
as in India and Sooth Africa, this genus is probably the first to 
appear with the amelioration of the climatic conditions. 
Passing to the next higher plant bed of importance, we find, 
northeast of Minas, Santa Catharina, at an horizon but 55 meters 
above the granite, Rosellinites gangamopteridis, Hysterites brasi- 
liensis, Phyllotheca Griesbachi, Phyllotheca Muelleriana, Phyllotheca 
sp., Glossopteris Brawniana, Vertebraria?, Gangamopteris obavata, 
Arberia minasica, Derbyella aurita, Noeggerathiopsis Hislopi, Cardio- 
carpon (Samaropsis) Seivasi, Cardiocarpon Moreiranum, Voltsia ? sp., 
and Hastimima White. 
This flora is to b3 regarded as a purely Southern or Gondwana 
flora, without contamination by northern types. The absence of Lepi- 
dophytes or other elements belonging lo the northern Permo-Carboni- 
ferous flora may b2 explained either on the hypothesis that the cli- 
matic or other environmental conditions had not yet become hospi- 
(1) Other Northern types listed by Professor Kurtz in collections from Argentina are 
Curdiopleris polymorpha, Adiantites antiguus, Lepidodendron selaginoidcs and Lepido- 
dendron aculcatum, The ostensible occurrence of these species mingled with forms of 
tilossopteris, Neuropteridium is so striking as to call for a detailed description. Is it 
nol possible that some of the specimens listed are from beds antedating the Gondwana 
flora? 
(2) Th will later be shown that these conglomerates may be of the same age as the 
Talehir glacial conglomerates in India, the carlier elacial deposits in Australia and the 
Dwyka e acial conglomerates in Souih Alrica, Phe appearance vf the Gangamopterts 
florain the overlying sediments is iu itself proof of the existence, in this part of the 
earth of the flora whieh immediately characterized the post-glacial deposits in the 
other revious of the older Gomlwanas. 
