Geology of the Lakes and the Valleij of the Mississippi. 209 



tial in that quarter, and entire further on where the force of the cur- 

 rent was increased by concentration in narrower channels. That 

 being conceded or established, it would follow that the course of the 

 ancient current is unerringly indicated by the course of the present 

 lakes. Now running the eye over the map from the northwest an- 

 gle of Pennsylvania to McKenzie's river, we perceive an uninterrup- 

 ted series of lakes beginning with the one of which that river is the 

 outlet, and proceeding by a direct line to the large terminating group 

 at the point from which we started. The whole tract of country is 

 studded with them as.w^ith a succession of pools left standing by a 

 retired ocean. A current setting in the direction of this line, from 

 the Pacific a little south of Bhering's Straights, w^ould sweep over the 

 primitive region at the northern extremity of the Rocky Mountains, 

 and may well have supplied the fragments in question from that point. 

 There is certainly no proof to assign them an origin elsewhere ; nor 

 can we otherwise account for their existence where we find them. 

 The angular masses occasionally, though rarely, found even within 

 the region of frost, have doubtless been borne on icebergs : but that 

 they, as well a:? the spheroids came from that quarter, there is not 

 a rational ground to dispute. It is evident from the greater degree 

 of denudation to the northeast, that the main body of the current 

 discharged itself into the gulf of St. Lawrence; but that diverging por- 

 tions of it found their way into the gulf of Mexico, is equally evi- 

 dent. Such a division of it could not but be produced by its efforts 

 to clear the obstruction presented to it by the Alleghany Mountains ; 

 to effect which it would necessarily pass round both ends of the 

 chain. The proof of it however rests not merely on the configuration 

 of the continent. A diluvium containing the same debris, is shown 

 by Dr. Drake in the volume of transactions already quoted, to over- 

 lie the country on the right banks of the Ohio and Mississippi, as 

 far down as Natchez where the pebbles are reduced to the size of 

 gravel. At Cincinnati, they rest on a horizontal shell limestone des- 

 cribed by him in his picture of that place, as 'a vast precipitate from 

 a lake or sea of salt water'. This is evidently a^continuation of the 

 lias from the lakes, whose terminal edge in that direction, makes the 

 falls of the Ohio at Louisville, and probably underlies the diluvium 

 not only at the prairies, but throughout. Dr. Drake's known accura- 

 cy gives a pecuhar value to ever^ fact asserted by him. 



Vol. XXIX,— No. 2. 27 



