288 NOTES OF A BOTANIST 
3000 feet, and extends thence to the very snow-Hne, going through 
more phases in external appearance than I know in any other 
genus. 
CompositcE^ 3. — So long as I herborised only in the plains, I 
could never understand how Humboldt had assigned so large a 
proportion of equinoctial vegetation to Compositse, for, from the 
mouth of the Amazon to the cataracts of the Orinoco and the foot 
of the Andes, with the exception of a few scandent Vernoniae and 
Mikaniae, and of a few herbs on inundated beaches of the rivers, 
the species of Compositse that exist are weeds, common to many 
parts of tropical America, nor did I meet with more than one 
arborescent Composita {Vernonia polycephala^ DC.) in the whole 
of that immense area. But in ascending the Andes, from 1200 
feet upwards, Compositae increase in number and variety at every 
step, and include many arborescent species. About midway of 
the wooded region, and especially in places where the trees form 
scattered groves rather than continuous woods, Compositae are 
more abundant than any other family, both as trees and woody 
twiners, and in the latter form extend nearly to the limit of 
arborescent vegetation, especially as species of the fine genus 
Mutisia ; while on the frigid paramos no frutescent plants ascend 
higher than the Chuquiraguas and Loricarias, and as alpine herbs, 
the Achyrophori, Werneriae, etc., reach the very snow-line. In 
the Red Bark woods Compositae are plentiful, and I should esti- 
mate the number of species at near 50. The trees of this order 
are chiefly Vernoniae, and they abound most in deserted clearings. 
During my stay, a plot was again brought under cultivation which 
had remained desert for twelve years, during which period it had 
become so densely and equably clad with a Vernonia, whose 
slender white stems had reached a height of 40 feet, that at a 
distance it looked like a plantation. Many of the woody twiners 
are Compositae, chiefly Senecionidae, and as herbaceous or suffruti- 
cose twiners there are several Mikaniae. The young shoots of a 
species of Mikania bear very large cordate leaves, usually white 
over the veins and purple or violet on the whole under-surface. 
. . . Among shrubby Compositae I noted some Eupatoria and 
two Baccharides, but no Barnadesia ; nor among herbs any 
Gnaphalium, although on the eastern side of the Cordillera the 
two latter genera descend nearly to 3000 feet. Tessaria legitima^ 
DC, is abundant by the Rio San Antonio. I have come on 
this tree in the roots of the Cordillera on both sides, by all the 
streams which have open gravelly or sandy beaches laid under 
water by occasional or periodical floods. 
Apocynece, 2. — One Peschiera and one Echites. This order 
rarely ascends up out of the hot region in the Andes, and in the 
temperate region I have seen only a single species. 
