EXPOSURES OF THE KEWEENAWAN 327 



bed directly underlies the fourth conglomerate bed noted in the Kettle 

 River succession. A tuff underlies the fourth conglomerate bed, de- 

 scending the Chengwatana series. 



THE SNAKE RIVER LOCALITIES 



So far as seen, the rocks of the Keweenawan along the Snake river are 

 confined to the first 3 miles below Cross lake. This lake is a beautiful 

 sheet of water 5 miles long, formed through the obstruction of the belt 

 of Keweenawan rocks to stream erosion. Where the Snake empties into 

 the Saint Croix there are several fine exposures of Cambrian quartzose 

 sandstones, typical erosion cuts in modified drift, and an extensive area 

 of eolian debris lying on the modified drift and river gravels. The 

 Keweenawan lies opposite on the Wisconsin side of the Saint Croix. 



* 



The Chengwatana Series 

 general structural characters 



The lavafloivs. — The name Chengwatana has been assigned for facility 

 of description to a series of eruptive and clastic rocks of unusual extent. 

 The flood of 1898, which tore away the dam at the foot of Cross lake 

 and poured down the river gorge a vast volume of water, cleaned out in 

 an admirable manner for examination the channel of the stream for 

 several miles. 



The Chengwatana dam rests on diabase. This rock occurs as a suc- 

 cession of lava flows. Each flow is characterized by three readily dis- 

 tinguished divisions : (1) A lower massive part, finely cr3^stalline, finely 

 shattered, and only a few inches in thickness. Its rough and irregular 

 bottom layer gradually becomes more coarsely crystalline upward until 

 it mingles into (2) the main mass of the flow. This part is of medium 

 texture, dark color, and quite free from amygdules or other cavities in- 

 dicating the presence of gaseous ejecta when solidification took place. 

 There is some difl'erence in texture as different flows are compared. The 

 thicker ones are coarser and more distinct in their crystalline char- 

 acters, and the ophitic habit is far more pronounced ; they are more 

 coarsely jointed ; the rock seems fresher. (3) The upper portion, which 

 is amygdaloidal^nd in several instances tuffaceous. The upper surface 

 of each flow is rough, sometimes showing considerable erosion. On this 

 surface the next succeeding flow adjusted itself in such manner that 

 there is no difficulty in detecting the division plane between the amj^gda. 

 loid and the finely crystalline and sometimes devitrified bottom layer of 

 the next succeeding flow. The amygdaloidal holes vary in size from 



