— 377 — 
as a result of a decrease in temperature it will retain 
either its direct or specifically modified form on the coast 
of Chile and Perû on the one side, and on the southern 
coast of Brazil on the other. This is the case with 
Bullia. In this instance we do not have to do with a 
member of an ancient common fauna of the Tertiary 
times, because species of Bullia are wanting in both 
‘hile and Patagonia as are those of Siphonaria. Laevi- 
litorina caliginosa Gld., the only Litorina of the Ma- 
gellan Province, occurs also alongside the other species 
of this genus in New-Georgia and Kerguelen Island 
while, as far as is now known, forms of Litorina are 
wholly wanting in the Patagonian Tertiary. In this way 
palaeontological and zoogeographical facts complement 
each other to establish the fact that through a recent 
-immigration of an antarctic element the ancient fauna 
of the Magellan District was greatly modified, and we 
would scarcely err if we attributed the cause of this 
migration to the ice period, concerning the extention of 
which in Patagonia Sfeënmann and Wordenskjoeld have 
recently given us extensive accounts. No wonder, then, 
that, as Philippi has said, the transition from Tertiary 
to Quaternary was a sudden and not a gradual one. 
This addition from the south has affected and modi- 
fied the Magellan and Chilian fauna to a much greater 
extent than that of the Argentinian coast, from which 
I have important new data from which to correct the 
many erroneous statements of Pfeffer. The mouth of the 
La Platta, as I have previously shown, is not a zoogeo- 
graphical barrier; the boundary between the Argentine- 
South Brazilian and the Patagonian fauna is the Rio 
Negro. 
A question intimately connected with those here 
discussed is the existence of bipolar species and genera. 
The following 5 species of Molluses from the Magellan 
Province also occur in the Arctic Regions: Saxicava 
arctica L., Lasea rubra Mont., Puncturella noochina L., 
