OF LAKE SUPERIOR. 



189 



with a much smaller proportion of protoxide. Both taken together, range from 

 five to twenty per cent. In all this, these rocks resemble the red sandstones of the 

 older States, and differ essentially from the more common varieties of sandstone, 

 appertaining, on the Mississippi and its tributaries, to F. 1, the equivalent of the 

 Potsdam Sandstone of New York. This Lower Silurian rock, taking its general 

 mass, is of a white or buff colour, and is a siliceous (quartzose), not an argillaceous or 

 ferruginous sandstone. With the exception of subordinate beds, of red colour, 

 usually in the immediate vicinity of igneous ranges, it commonly contains ninety- 

 two per cent, and upwards of silica; while of alumina and oxide of iron, taken 

 together, it has seldom more than three per cent. 



In determining, however, the synchronism of two or more formations, it is, as 

 every geologist knows, wholly unsafe to rely upon lithological appearance and 

 chemical composition, except as corroborating indications. Equivalent formations 

 in distant countries, or even at comparatively short distances, vary greatly in this 

 respect. Indeed, beds of strictly cotemporaneous origin (as indicated by the identity 

 of their fossil remains), may, at one locality, be true limestone, at another pure 

 sandstone. 



The test of superposition, on the contrary, is satisfactory and conclusive, provided 

 it be unequivocally obtained. How far Dr. Norwood, who was specially charged with 

 this investigation, succeeded in obtaining it, will be seen from his Report. I myself 

 have had opportunity personally to witness what, even if it be not admitted to be 

 unequivocal proof of the true position of these sandstones with reference to another 

 well-defined formation, must certainly be taken as very strong presumptive evidence, 

 not only in favour of the opinion that the sandstones of Lake Superior are of Lower 

 Silurian date, but that they underlie the palceozoic base of the Mississippi Valley. 



The phenomena to which I refer, came under my notice chiefly on the upper 

 waters of the St. Croix ; and are especially worthy of attention in this connexion. 



It will be remembered, that in Chapter I., when speaking of the palaeontology of 

 F. 1, I noticed the occurrence, on the St. Croix, just below its Falls, of Lingula and 

 Orbicula beds (Tab. I., B, & F. 1, h) ; and stated, that these strata, constituting 

 the palaBozoic base of the Mississippi Valley, were found in horizontal layers, 

 in juxtaposition with the trap range. Here we have a point of departure, deter- 

 mined by unquestionable paloeontological evidence ; all admitting these beds to be 

 of Lower Silurian date. 



Now, as the general dip, throughout this entire region of country, is southeasterly, 

 it follows that, in ascending the St. Croix, above the Falls, so long as the formation 

 remains continuous, we gradually reach lower members of F. 1. Accordingly, we 

 find, from the Falls up as high as the mouth of Snake, the white and buff quartzose 

 sandstone beds of F. 1, a, preserving all their characteristic peculiarities, in place, 

 at brief intervals on both sides of the river. From this point, still ascending the 

 river, occurs a marked change, challenging the attention of the geologist. Instead 

 of the light or yellow quartzose sandstone, which he has left behind him, he is met, 

 commencing on the St. Croix about five miles above the mouth of Snake, by red 

 sandstone, argillaceous, ferruginous, similar in tint and composition to the red sand- 

 stones of Lake Superior, and associated, like them, with coarse red conglomerates, and 



