2 877-1 Southern Hemisphere. 335 
those first formed, contain some evidence of brackish water 
conditions such as might have been produced by the em- 
bracement of the salt water of bays within the area blocked 
up by the ice ; but the upper parts contain nothing but the 
remains of land animals, whilst the proof of the presence 
of water is complete in the rounded and stratified shingle 
forming the Patagonian plains. 
Mr. Wallace has shown that the cave fauna in Brazil 
differs from that of the Pampas in containing a larger pro- 
portion of existing genera, and is inclined to believe that 
this may imply that the Pampean mud is a little older than 
the cave deposits.* It may, however, be due, I think, to 
the destruction of life having been more complete in Buenos 
Ayres and Patagonia than in Brazil, and that, in pre-glacial 
times, as now, the former countries formed a part of a dis- 
tinct zoological sub-region. That the destruction of life in 
Brazil should have been less complete than farther south is 
likely— -not only because we have no proof that the great 
flood extended so far north, but because there was a large 
extent of country, neither covered with ice nor overwhelmed 
with water, where many of the species might be preserved. 
In Patagonia and Buenos Ayres the lower country was 
covered with water, the upper mostly with glaciers descending 
from the mountains, so that the retreat of the animals from 
the flood was in a great measure cut off. The peculiar spe- 
cies that have been preserved in the Chilian sub-region, 
which includes Patagonia and Buenos Ayres, are principally 
alpine forms, such as the Chinchillas, the Alpacas, and the 
Viscachas, which we may suppose found a refuge during the 
great flood in some high lands not covered with glacier ice. 
Others may have retreated northward up the western side 
of the Andes, and returned southward after the greatest 
severity of the Glacial period had passed away. The re- 
mains of an extindt llama are found on the high plains of 
Mexico, so that the genus had undoubtedly a former greater 
extension northwards. The principal extirpation of life 
appears to have been amongst the bulky species, on whom 
the changes of environment would press most heavily, and 
which were confined to low-lying districts. Thus the Mas- 
todon, whose remains are found at great heights in the 
Andes, appears to have escaped the complete destruction 
that befel its bulky companions at the time of the greatest 
extension of the ice, as I saw remains of it in Mexico that 
from their fresh appearance were probably as recent as those 
* Geographical Distribution of Animals, vol. i., p. 146. 
