OF WESTEHK SOrTlI AMERICA. 147 



of tropical America, are clearly of Andean origin, some species 

 of which haA'C been adapted to the climate of the lower region. 

 To tliis categorj^ belong several genera of Vacciniaccce, of which 

 an examj)le is found in the new species of Anihoj^teriis above 

 enumerated. 



The conclusion to which I am led from a study, doubtless very 

 incomplete, of the flora of tropical America, is that the vast 

 region including the warm and moist parts of South and Central 

 America sliould be regarded as a single Botanical Province, 

 in Avhich the same generic types are represented by species of 

 which a large proportion are endemic, and confined to compara- 

 tively small areas. Along with these we find in yarious parts of 

 the same region a few forms so distinct as to be ranked as 

 separate genera, mostly represented by one, or very few, species, 

 and nearly allied to types of wide distribution. 



If along with these elements of tlie flora we find representatives 

 of very ditlerent types of vegetation in many parts of the same 

 great region, closer examination sliows that, with rare exceptions, 

 these are more or less modified forms of groups whose origin is 

 either in the Andes or in the mountains of Gniaua and Bivizil. 



If this view be correct, it undoubtedly entails the inconvenience 

 that it is impossible for the phyto-geographer to malvc the 

 natural divisions of the flora agree with dellned geographical 

 boundaries. But this is an inconvenieiice which has become 

 aj)parent elsewhere in the world as well as in Smith America. 



In a broad sense it may be said tliat the most natural divisions 

 of the vegetation of the earth are wide areas of low country, over 

 which, with more or less of modification, a limited number of 

 types have extended, with islands of high land, which are tlie 

 homes of the special types that form the charucleristic features 

 of the floras of different regions, l^rom these mountain-centres 

 we ordinarily find some representatives of these special types 

 that extend through the low country to a greater or lesser 

 distance from their original home, but as to who^-e mountain- 

 origin there is little room for doubt. I am well aware that real 

 or apparent exceptions to this view may readily he cited. Such, 

 for instance, are the VocTiysiacecs of South America, which are now 

 confined to the warm tropical zone. It is a fair matter for 

 discussion whether it he more probable that they have originated 

 in that zone, or whether the existing species are the modified 

 descendants of ancestors that inhabited the ancieiit, and now 



LlNIf. JOURN. BOTA^'^; VOIi. XXII. ^i 



