joined by Mr. Bourget, with his fine collections from Tabatinga. He, like both the other parties, has been hindered, by want of alcohol, from making as large collections as he might otherwise have done; but they are, nevertheless, very valuable, exceedingly well put up, and embracing a great variety of species, from the Marañon as well as from the Hyavary. Thus we have a rich harvest from all the principal tributaries of the Upper Amazons, within the borders of Brazil, above the Rio Negro, except the Purus, which must be left unexplored for want of time and a sufficient working force. On leaving Teffé I should say something of the nature of the soil in connection with Mr. Agassiz’s previous observations on this subject. Although he has been almost constantly occupied with his collections, he has, nevertheless, found time to examine the geological formations of the neighborhood. The more he considers the Amazons and its tributaries, the more does he feel convinced that the whole mass of the reddish, homogeneous clay, which he has called drift, is the glacial deposit brought down from the Andes and worked over by the melting of the ice which transported it. According to his view, the whole valley was originally filled with this deposit, and the Amazons itself, as well as the rivers connected with it, are so many channels worn through the mass, having cut their way just as the igarapé now wears its way through the more modern deposits of mud and sand. It may seem strange that any one should compare the formation of these insignificant forest-streams with that of the vast river which pours itself across a whole continent; but it is, after all, only a reversal of the microscopic process of investigation. We magnify the microscopically small in order to see it, and we must diminish that which transcends our apprehension by its great size, in order to understand it. The naturalist who wishes to compare an elephant with a Coni (Hyrax),[76] turns the diminishing end of his glass upon the former, and, reducing its clumsy proportions, he finds that the difference is one of size rather than structure. The essential features are the same. So the little igarapé, as it wears its channel through the forest to-day, explains the early history of the great river and feebly reiterates the past. ----- Footnote 68: It is a curious fact, that though a large number of cows were owned in Teffé, and were constantly seen feeding about the houses, milk was among the unattainable luxuries. Indeed, milk is little used in Brazil, so far as our observation goes. It is thought unhealthy for children, and people will rather give coffee or tea to a two-year-old baby than pure milk. The cows are never milked regularly, but the quantity needed for the moment is drawn at any time. Footnote 69: TEFFÉ, September 22, 1865.