298 Mr. J. Orton on the Evidence of 



anfl sandstones were deposited underneath a t^it^'antic glacier, 

 whicli descended from the Andes, grinding into fine powder 

 the materials between it and the solid rock, and leaving an 

 immense moraine across the mouth of the valley. To this 

 theory we make the following objections : — 



1. The theory is short of positive proof where we need the 

 most unquestionable evidence. The confession is made that 

 " the direct traces of glaciers, as seen in other countries, are 

 wanting in Brazil." There is not a trace of fuiTOws, stride, or 

 polislied surfaces*. The answer that the rocks are so friable, 

 and disintegration in the tropics so rapid, as to render their 

 discovery hopeless, is not entirely satisfactory. The granitoid 

 rocks which border the valley, and the schists and porphyries 

 on the slope of the Andes, ought to preserve some marks of 

 the glaciationf. The pot-holes in the gneiss plains of Bahia, 

 supposed by Hartt to have been formed by glacial cascades, 

 are " exceedingly Avell preserved, and have smooth sides ;" 

 while all the ploughings and planings of the gigantic glacier 

 over the same rock have been utterly erased by disintegration ! 

 The stone structiu-es of Brazil endure remarkably well, while 

 the granite of Quebec exfoliates so rapidly in winter that oil 

 is used to protect the buildings ; yet there is no lack of stria3 

 in Canada. 



Boulders occur only along the eastern region ; none have 

 been observed in the great interior basin. This is a strange 

 inversion : if a continental glacier moved doAvn the Andes to 

 the Atlantic, we would naturally look for porphyritic boulders 

 scattered over the valley, and dwindling in number and size 

 as we near Para. We are susjiicious, also, that these so-called 

 boulders have not travelled. The only genuine erratics seen 

 by Professor Agassiz were found on the northern flank of 

 Erere ; all the others turn out to be " boulders of decomposi- 

 tion." The boulders of Tijuca, in the Bio Province, described 

 by Hartt, Mere not far-fetched ; the majority are of gneiss on 

 gneiss : still they may have been the work of local glaciers. 

 The Ererd erratics are hornblendic and without scratches ; 



* Professor Hartt likewise acknowledges, "1 have nowhere seen either 

 polislied or striated rocks." 



t The eminent explorer Dr. Spruce describes the Casiquiari lIe<non 

 as ''one great sheet of granite and gneiss. There is uovrhere any con- 

 tinuous rniigo of mountains or plateau ; and. except towards it^; bordei-s. 

 the granite has been entirely denuded of the stratified rocks that once 

 o\ I'vlay it, and is now either naked or else overspread in some places with 

 a thin covering of whit(> sand, and in others (chieily llats, hollows, ivnd 

 rifts) with a tliick deposit of the fertile 'terra roxa," or i-ed loam (de- 

 composed gneiss, mica-schist, &c.), wliicii 1 have supposed to be lacus- 

 trine, but I'rofesscu- Agassiz says is glacial drift." 



