SOURCE OF ST. PETEU S RIVER. ."53 



spection of their nature shows them to be really crystal- 

 line. There is a gradual though rapid passage of the gra- 

 nite into the sienite, which proves them to be of contem- 

 poraneous formation, and which precludes the idea that 

 the rock is formed by the union of fragments of granite, 

 sienite, &c. cemented together. 



The discovery of this granitic formation here appeared 

 to us the more interesting, that its small extent might 

 easily have prevented us from observing it, had not chance 

 brought us to the river at tliat place ; for if we had been 

 travelling on the prairie, within half a mile of the edge of 

 the bank, the greater height of the bluff would have con- 

 cealed these rocky islands from our view. We feel there- 

 fore unable to decide whether they do not recur at some 

 of the other bends of the river, which we avoided ; yet from 

 the character of the stream itself, we doubt it; for we find 

 that as soon as these rocks protrude into the valley, they 

 occasion rapids and falls in the river, while otherwise its 

 course is smooth. Had we not seen the " Little rapids," which 

 we passed on the 11th, we might have been induced to con- 

 sider them as resulting from the appearance at the surface of 

 primitive rocks, but having examined with care the sand- 

 stone rocks, by which they are produced, and having as- 

 certained that no other rapids are found in the St. Peter, 

 between these and the Patterson falls, we arc induced to 

 believe that this is the only place where the granite may 

 be seen in situ. In attempting to connect this primitive 

 formation with those observed elsewhere, we find that it 

 lies in a direction about west-south-west, at a distance pro- 

 bably not exceeding eighty miles, of the "granitic and horn- 

 blende rocks," which Mr. Schoolcraft states as having seen, 

 '' occasionally rising in rugged peaks and beds," on the 



Vol. T. 45 



