358 ARTHUR C. TROWBRIDGE 



old valley of the Wisconsin from the Lower Narrows to the north 

 end of Devils Lake gap, where its edge became stationary and 

 deposited a terminal moraine, the other lobe coming in south of 

 the South Range and advancing up the valley of the prc-Glacial 

 Wisconsin to deposit a marginal ridge across the gap east of its 

 major bend (Plate II). This left the north-south portion of the 

 gap and a part of the east-west portion confined between the two 

 edges of the ice, and in the basin so made Devils Lake was formed. 

 Connecting the ends of the two lobes, the edge of the ice reached 

 its limits of advance in an irregular line crossing the South Range 

 from the north edge of Devils Lake eastward to Sauk Point, and 

 thence southwestward to the gap east of Kirkland (Plate II). 



After its formation the lake had an interesting history during 

 the occupancy of the ice. 



Sources of supply. — When the lake was first formed, as outlined 

 above, there were at least four separate sources of water supply: 



(1) The edge of the glacier blocked either end of the lake basin. 

 There the ice melted and furnished water for the basin. Study 

 of the terminal moraine from the north end of the lake around by 

 Sauk Point and southwest to the gap east of Kirkland leads to the 

 conclusion that the water resulting from melting along this whole 

 stretch of ice front must have flowed into Devils Lake basin. 



(2) The bottom of the basin was below ground-water surface, as 

 evidenced by the fact that it had been occupied by a permanent 

 stream up to the time when this stream was blocked by the ice, 

 and ground water was a source of supply. From the inception of 

 the lake until its bottom was built up above ground-water surface 

 by fluvio-glacial deposition, if this stage was ever reached, some of 

 the lake water may have come from under ground. (3) The lake 

 must have had inlets resulting from precipitation within the 

 borders of the lake basin. For instance, Messenger's Creek with 

 its north and south forks must have flowed into the southwest 

 corner of the lake as it does today, and the stream which flows 

 west and north past the northeast corner of the lake, being blocked 

 to the north by the ice, must have contributed its waters to the 

 lake. (4) There was doubtless direct precipitation upon the 

 surface of the lake. Of these four sources of supply, the first men- 



