242 BOTANICAL GAZETTE [OcTOBER 
the rain forests of southern Chile and of New Zealand, with perhaps 
a couple of epiphyte ferns in Japan and southern Australia, are 
indigenous in origin. The other temperate zone epiphytes of the 
Old World, of South America, and according to SCHIMPER all, 
epiphytes of temperate North America, have acquired this habit 
while in the tropical forest. ScHIMPER stated that it is the most 
xerophytic of the tropical epiphytes, those growing on the branches 
of the relatively dry roof of the forest, that have wandered out 
across the neighboring savannas and subtropical forest and onward 
sometimes 10 or 15 degrees beyond the tropics to populate with 
epiphytes the warmer and moister of the neighboring temperate 
forests. Because of the adaptation of these epiphytes to the dry 
conditions at the top of the forest, they have been able, in spite of 
the still more rigorous conditions encountered there, to colonize 
certain temperate forests. For the epiphyte that migrates from 
the tropics to the temperate zone, probably the most critical 
adverse condition encountered is not the occasional hot, dry summer, 
but the periods of low humidity during the generally wet winter 
season, when cold, dry, northwesterly winds prevail, during which 
the evaporation rate is high and water cannot be absorbed by the 
frozen roots. For example, the writer has noticed that tufts of 
Tillandsia usneoides, hung on a deciduous magnolia tree each year 
in May, thrive and grow rapidly during the summer, and even 
look fresh and green after several frosts in the autumn. They 
ultimately succumb, however, to the cold dry westerlies of winter, 
even of so moderate a winter as that of 1920-21. The precise 
measurement of the evaporating power of the air at these low 
temperatures, a factor of prime importance also to terrestrial 
plants, especially evergreen ones, must await the invention of a 
practicable frost-proof evaporimeter. Possibly the exposure of the 
epiphyte to sunlight, when the supporting tree is bare of leaves, is 
directly injurious also, although this seems hardly likely, since this 
same Tillandsta is abundant on deciduous trees only 200 miles 
south of Baltimore, where the winter sunlight would probably be 
at least as strong. The sunlight of course must work harm 
indirectly by increasing transpiration, which probably explains 
the usual restriction of polypody to the north sides of the trees. 
