272 ANNUAL OF SCIENTIFIC DISCOVERY. 



lie coraraentod at some length upon the dilTercnce in the 

 appoarance of the fi^lacial matL'rials in tlio valk'y of the Amazon 

 and in more Soulht'i-n })arts ol" Hiazil. In llie valloy, he observed, 

 these materials were refrnlarly stratified, while in other localities 

 they were blended. lie had tlniwn from tlu; lirst-mentioned fact 

 the conclusion, that tiie valley Avas once barred to the sea and 

 formed into a lake bj' the accumulation of a moraine; and, upon 

 investigation and incjuiry, he had found evidence to sustain this 

 theory, in the existenei; of extensive remains of the moraine. 



On the Drift and Glaciers of the Amazon Valley. — Taking the 

 regions about the provinces of Ccara and Rio for illustration, he 

 saitl the whole country within the former province is ilat ; but from 

 this plain rise hills of considenil)le height, some reaching as high 

 as three thousand feet. These mountain masses are composed of 

 metamorphic rock, and present a remarkalde degree of disintegra- 

 ion. This disintegration can be seen, in the neighborhood of llio, 

 penetrating at least tlu'ee hundred feet, even where; the rock is con- 

 tinuous from the surface downward. The solid rock is not only 

 atfected by this disintegration, but the loose materials show it, so 

 that it is dillicult to recognize their primitive condition, and to 

 trace their relation to the original material on which they rest. He 

 was satisfied tliat large masses of what we call drill rest on the 

 tropical solid rocks, as well as upon the rocks in the northern re- 

 gions ; and that these are erratic is plain from the fact that they 

 are not of the same mineral character as the rocks underneath 

 them. In this connection he stated that it was a curious fact, tiiat 

 whei-evcr rocks have btsen moulded by the power of ice they pre- 

 sent a rounded, dome-like shape. This condition of these rocks 

 and loose material is ]n-oof of the former existence of glaciers. 



But we have more direct evidence of the existence, at one time, 

 of local glaciers. In the vicinity of Mangouva he had been struck 

 ■with the character of the loose material, and upon examination 

 had found that on both sides the valle^^ on the steep slope of the 

 mountains, there were large accumulations of boulders. These 

 boulders were frnnly fixed on the slopes, Ijut none were at the bot- 

 tom of the mountains. Inquiring of the inhaljitants of this region 

 concerning these boulders, he had learned that they existed no- 

 where in the dejjressions, but were suspended along the valley on 

 the sides of the mountains. He had moreover found this to be the 

 case throughout the chain. Now, if these had been brought by 

 water they would have slid to the bottom, and could not have 

 fastened themselves upon the sides of the mountains. lie said that 

 in his mind these were pi'oofs, beyond the iiossibility of a doulit, 

 of the existence of local glaciers descending from the summit of 

 the hills to the plains, posterior to the great extension of ice over 

 the continent. 



Again alluding to these boulders, he said that he had found these 

 perch rocks, on the summit of the mountains, of an entirely differ- 

 ent character from the rocks on which they I'cst. And these must 

 have been brought by an agency none other than ice. If they 

 had been brought by flood, they Avould have been thrown over the 

 side of the hiUs. But if the boulders had been carried on the back 



