I02 NOTES OF A NATURALIST. 



To the botanist, the most interesting features in the 

 Andean flora are supplied by the great family of 

 CompositcB. To this belong nearly one-fourth of all 

 the plants collected by me, and nearly one-third of 

 those found in the higher Alpine region ; and, as far 

 as available materials allow me to judge, I believe 

 these to be about the true proportions for the higher 

 parts of the Andean chain. It is further remarkable 

 that of the thirteen tribes into which the 780 genera 

 and 10,000 species of this family have been divided, 

 all but the two smallest tribes — Calendidacece and 

 Arctotidece — are represented in the Andes. To the 

 European botanist, the most interesting group is that 

 of the MtLtisiacecE, which is especially characteristic of 

 the South American flora. Of 420 known species 

 belonging to this tribe, fully 350 are exclusively 

 American, the remainder being distributed through 

 Australasia, and from South Africa to Southern Asia. 

 They exhibit many unfamiliar forms very unlike what 

 we are used to find elsewhere in the world. One of 

 the first plants which I gathered was a tall, straggling 

 climber with pinnate leaves ending in a tendril. I 

 naturally thought of the vetch tribe, but I observed 

 that the leaves were without stipules, and that the 

 leaflets were not articulated to the midrib. Great, 

 however, was my surprise when, on finding a flowering 

 specimen, it revealed itself as a composite belonging 

 to the genus Mutisia. 



Next to the Composites, the grasses are of all the 

 natural orders the most largely represented in the 

 Andean flora, but with the diflerence that nearly all 

 belong to genera common to the mountain regions of 



