430 DR. HANS GADOW ON ALTITUDE AND 
rare, Within subarctic and temperate countries such cases are naturally 
more common, although even there the principle has been greatly exaggerated. 
In either case, relics or accidental,—restricting ourselves to the Mexican 
mountains, the species should be presumably the ‘same on the tropical tops 
and in the more distant arctic regions. But if the top flora has been derived 
from the basal, by modification, then there is every reason for suspecting 
that the plants should be specifically different. This is actually the case, 
to a great extent. Such a top flora -may assume- even an arctic facies, by 
convergence, by adaptation to similar environmental conditions. ` 
Engler's explanation of the Californian and Rocky Mountain floras applies 
also to Mexico. Without exception the highest mountains of this country 
arise from the plateau, at least with one side, whilst the other, most con- 
spicuously on Citlaltepetl, slides right into the tropics. Now, this plateau 
gives the temperate and cool flora a very good lead, so much, indeed, that the 
implied change into alpine species need be but very gradual, and the huge 
central plateau sloping from 7000 feet in the south to 3000 in the north, whilst 
maintaining continuity all along the backbone of the western Sierra Madre, 
insures an ideally perfect continuity of the same conditions from California, 
Arizona, and Sonora (the ancient Sonoraland, which stood out long before 
the rest of Mexico came into existence during the later Cretaceous epoch) to 
the truly alpine heights which now border the southern edges of the plateau. 
So far, then, it must have been easy for plants of temperate North America 
to spread to the high Mexican mountains, but the distance is great and 
physical conditions have changed, and after all the bionomie conditions 
of the Mexican highlands are not the same as those of the United States, 
even if the mean temperatures be the same: consequently most of the 
congeneric plants have become specifically distinct. But this same high 
Mexican plateau is possessed also of a great number of endemic, autoch- 
thonous plants, many of which have sent representatives into the alpine 
regions, Consequently these have received part of their flora from the 
immediate adjoining basal regions or plains. If these mountains were 
isolated, if they did arise as solitary cones out of the tropies, the investigation 
would be simple. We should have to deal only with those species, about 10 
per cent. according to Heilprin, which they have in common with some of 
the Central American peaks, e. y. Irazu, and with the distant Andes ; plants 
‘which owe their scattered, high-level distribution possibly to wind, birds, 
and similar accidental agencies, 
But the high Mexican mountains stand in a very peculiar condition, 
namely on the edge of the plateau. For instance, on the west side of 
Citlaltepetl it rises gradually from 7500 to 9000 fect before we come to the 
shoulder of the giant, while the south-eastern side glides much more rapidly 
into the tropics. These mountains have, in fact, two base levels, one of them 
