182 Mr. H. W. Bates on the Nymphaline 
free passage by this route, and that the still more singular reap- 
pearance in the south temperate zone of genera characteristic of high 
northern latitudes can be explained in the same way. We can only 
hope that future travellers in New Granada, Ecuador, Peru, and 
Bolivia will be careful to note the ranges, both horizontal and vertical, 
of all the species they collect ; for by so doing they would contribute 
greatly to the solution of this and many other kindred problems in 
geographical distribution. 
Of the 41 genera found on the banks of the Amazons, one only, 
Antigonis, is peculiar to the region. The smallness of the number 
common to the Amazons and the Old World is very remarkable, there 
being only four, namely, Melitea, Junonia, Salamis, and Apatura. No 
less than 28 out of the 41 are generally distributed in tropical and 
subtropical America. Four are common and peculiar to Guiana with — 
the Amazons (the Guiano-Amazonian province), besides three others 
which extend a little further westward into Venezuela and New Gra- 
nada. One genus, Pandora, is peculiar and common to the upper 
part of the Amazons plains and New Granada; but there is not a 
single generic group common and peculiar to the Amazons and Brazil 
proper. From these details the high degree of peculiarity of the 
Tropical-American fauna stands forth with great clearness; at the 
same time the small proportion of genera limited to one or two parts 
of the region shows how large are the diffusive powers of the insects 
composing the subfamily. If we descend from genera to species, 
however, we discover that this generality of distribution is not so 
strongly marked ; for 74 out of the 160 Amazonian species are, as 
far as at present known, confined to this part of Tropical America ; 
32 are peculiar and common to Guiana and the Amazons, whilst 7 are 
Brazilian, 9 New Granadian (or extending still further north), and 1 
Bolivian. The number of peculiar Amazonian species will no doubt 
be considerably reduced when the valleys of the Andes are better 
explored ; but still the fact of numerous species and genera common 
te the whole of Tropical America being confined to limited districts 
within the region is sufficiently well established to lead us to expect a 
large amount of peculiarity also in the Amazons plains ; and the whole 
shows that the process of distribution must be a very slow one to 
have caused so general a diffusion of genera, whilst the species, at a 
given epoch, are so commonly restricted to limited areas. 
Little can be said of general application regarding the habits and 
natural history of the Nymphalinz of the Amazons region, As may 
be seen from the foregoing remarks, the early states of the insects are 
