350 MR. R. SPRUCE OK INSECT-MIGRATIONS IN SOUTH AMERICA. 



trees such as sprinkle the sayannas, or are gathered into groves, 

 along both the northern and southern borders of the great Ama- 

 zonian forest-belt, wherein they now barely exist on the bits of 

 " campos " that at wide intervals break the monotony of the wood- 

 land — although they probably at some antecedent period ranged 

 contiauously from north to sotith. 



In other cases, closely allied species occupy distinct areas. One 

 of the finest fruits of Equatorial America, the Cocura {Pourouma 

 of Aublet), is borne in large grape-like bunches on trees of the 

 Breadfruit tribe, having large palmatifid hoary leaves, quite like 

 those of their near allies the Cecropias. Now the Cocura of the mid- 

 region of the Eio Negro, of the Japura, an^ of the Upper Amazon or 

 Solimoes, is owe species {Pourouma cecropi<sfolia, Mart.), while that 

 of the mouth of the Eio Negro and adjacent parts of the Amazon 

 is a very distinct and smaller-fruited species (P. return. Spruce), 

 and that of the Uaupes is a third species (P. apiculata, Spruce), 

 all three being so plainly diverse that the Indians distinguish 

 them by adjective names, although that diversity or divergence, 

 as in a vast many parallel instances, is but a measure of the time 

 that has elapsed since their derivation from a single stirp. 



But the most general cause of resemblance lies in this fact, 

 that there are many orders and families of plants whereof many 

 of the species are confined to limited areas, and yet, throughout 

 the Amazon valley, each order, or family, will be everywhere re- 

 presented by about the same number of individuals and species, 

 having to each other n^arly the same correlation, as regards aspect 

 and sensible properties — provided always that the conditions of 

 growth (as above defined) be the same ; so that a plant which 

 serves as food for any particular animal or tribe of animals in a 

 given locality, is pretty certain to have its congener (or at least 

 its coordinate) in any other locality of the same region. 



The Eiparial Plants of the Amazon (such, namely, as grow be- 

 tween ebb- and flood-mark, or within the limits to which the an- 

 nual inundations extend) range in many instances from the very 

 mouth of the river up to the roots of the Andes ; and I do not 

 yet know of a single tree which is not found both on the northern 

 or Gruayana shore and on the southern or Brazilian*. The most 

 notable example of this extensive range is the Pao Mulatto, or 

 Mulatto Tree {EnTcylista, Benth.), a tall elegant tree allied to the 

 Chinchonas, and conspicuous from its deciduous brown bark, 



* Hence I suspect that those insects of tlie south side of the Amazon which 

 have been identified with Gruavana species, belong chiefly to the Eiparjnl forests- 



