352 ME. E. SPETJCE ON INSECT-MIGEATIONS IN SOUTH AMEEICA. 



After what has been said, it is scarcely necessary to add that 

 many species of plants which grow down to the very coast in 

 Gruayana exist also in the Peruvian province of Maynas — that is, 

 at the eastern foot of the Andes, and even up to a height of a 

 few thousand feet in those mountains, — e. g. Humboldt's "Willow 

 [Salix humholdtiana, W.) and the Cannon-ball Tree {Couroupita 

 guianensis, Aubl.), called Aia-uma, or Dead Man's Head, in May- 

 nas ; "while the proportion of Orinoco plants repeated on the Ama- 

 zon is much greater than that of the plants of South Brazil. Nor 

 does this uniformity of character, and the constant recurrence of 

 certain species, preclude the possibility of the flora being won- 

 derfully rich ; for I have calculated that by moving away a degree 

 of either latitude or longitude, I found about half the species dif- 

 ferent ; while in the numerous " caatingas " I have explored I always 

 found a few species in each that I never saw again, even in other 

 " caatingas". 



The importance of inquiries of this class is obvious, even from a 

 zoological point of view ; for that an animal should flourish in any 

 region it must there find suitable food ; and there is perhaps no 

 part of the world where so large a proportion of the animals is so 

 directly vegetarian in its diet. I have reason to believe that there 

 is no carnivorous animal on the Amazon and Orinoco which does 

 not occasionally resort to vegetables, and especially to fruits, for 

 food, — not always of necessity, but often from choice. When, how- 

 ever, we come to consider and compare the distribution of the 

 various classes and subordinate groups of animals, we see that the 

 range of a fruit-eating species or tribe can rarely correspond to 

 that of one which feeds on leaves — and similarly of other pairs of 

 difi'erences or contrasts in the nature of the food,-T-that, in short, 

 the only animals which can be expected to range from sea to sea in 

 a wide continent are a few general feeders and their parasites, the 

 larger beasts of prey, and the scavengers, such as Vultures among 

 birds (and perhaps Termites among insects). 



As to the distribution of the Lepidoptera in the Amazon valley^ 

 it is plain that it can rarely correspond to the grander features of 

 the vegetation, for the simple reason that the food of caterpillars 

 is scarcely ever the foliage &c. of the loftier forest-trees, but chiefly 

 of soft-leaved undershrubs and low trees (1) which grow tinder the 

 shade of the forest and have, many of them, a restricted range, or 

 (2) which spring up where the primeval woods have been destroyed, 

 and in waste places near the habitations of men, and whose range 

 in many cases is coextensive at least with Cisandiue Tropical 



