252 Wisconsin Academy of Sciences^ Arts, and Letters. 



men is by his determination 2.752; and of the gray one, 2.829. 

 The agreement in this regard also is quite close, though perhaps 

 accidental ; for Nicollet's determination of 2.90 for the red variety 

 shows that specimens from the same quarry may vary considera- 

 bly in this particular; and so, likewise, they do in color. The 

 stone from the Minnesota quarry is not uniformly of the blood- 

 red color on which the species Gatlinite is founded, but often is 

 mottled with lighter shades of red, running into yellow; while 

 that from Devil's Lake, as thus far discovered, is all varieg'atej, 

 gray, black, yellow and red being intermingled in the same speci- 

 men, producing the veined appearance of some marbles. One 

 part of the gray specimen here submitted gives a red streak 

 ■undistioguishable from that of the red specimen, and I have seen 

 specimens from Devil's Lake in which the dark color greatly pre- 

 dominated, though sach examples are as yet rare. This diversity 

 in weight and color indicates that a quantitative analysis of speci- 

 mens from dififeren<!%ource3 would be scarcely more valuable, as 

 a means of identi6cation, than a qualitative one. I however take 

 from Silliman''s Journal^ 1839, the only analysis to which I have 

 access, that of the Minnesota pipeslone, by Dr. Jackson of Boston : 



Grains. 



"Water. , 8.4 



Silica 48.2 



Alumina 28 . 2 



Magnesia 6.0 



Perox. Iron 5.0 



Ox. Manganese 0.6 



Garb. Lime 3.(> 



Loss (pro 1 ably magnesia) 1.0 



100 

 The carbonate of lime is no*; an essential ingredient, but is 

 mixed in fine particles. 



It will be noticed that this formula agrees, in the general way,, 

 with the qualitaiive results by Prof. Daniells. 



I was led to conjecture the true character of this rock from an 

 examination, made in 1869, of the quartz'te of the falls of the 

 Big Sioux, in Dakota Territory, where the town of Sioux Falls 

 Las since been built. That outcrop is reported to be the same 



