144 BOTANICAL GAZETTE | AUGUST 
ent plant world to allow time enough for the isolation of such 
groups as Chorizanthopsis, Malvastrum § Phyllanthophora, or 
even for the Chilian development of Borrag.-Eritrichiex. 
But how does it occur that the high andean flora is chiefly 
boreal? And how have the arctic-alpine plants reached the 
southern high Andes from the Rocky mountains? Further, 
how have the sharply defined and isolated species of Larrea, 
Frankenia, Prosopis, etc., come to be in both regions ? 
Gray and Hooker supposed that in the glacial time there was 
a driving of boreal and warm temperate elements southward, as 
a result of which some plants were placed favorably for migra- 
tion farther south. Engler suggests that the southward migra- 
tion of animals caused by the glacial encroachment was very 
notable in aiding the distribution of plants southward over the 
isthmus. 
It has been suggested that geological conditions have allowed 
a more general extension along the west American coast of an 
arid plateau similar to that of middle and northern Chili. 
Referring to a chart of ocean depths along the Pacific coast it is 
clear that by an uplift of 3000 feet a series of abrupt step-offs 
or shelves would be exposed extending to the Californian coast, 
and making the isthmus region a broad belt of land. This, 
besides offering a highway for xerophytic elements, would also 
bring about that union of the Pacific islands off the Californian 
coast with the continent which Sereno Watson” supposed must 
have prevailed in order to give the similarity of island flora to 
that of California. He suggested that the relations to the adja- 
cent continent indicate a former flora which spread over a wide 
region now submerged, from which ancient flora the elements 
common to California and Chili were derived. 
Such a condition of emergence along the west coast would 
be very favorable as an explanation for many of the phenomena 
of distribution, but I am not convinced that it is either necessary 
or possible to assume this. One must bear in mind that an ele- 
vation of the coast of South America by 3000 feet would in all 
* On the flora of Guadaloupe islands. Proc. Amer. Acad. 11: 112. 1876. 
