306 AI^'ISrUAL EEPORT SMITHSONIAN INSTITUTION, 1912. 



formed, and the Warren beacli is therefore the beach of a raised lake. 

 The Wayne beach is generally faint and on the " thumb " of Michigan, 

 where it is gravelly, shows distinct evidence of submergence and 

 modification after it was made. Exceptmg on the ''thumb," it is 

 generally sandy and without marked characteristics. 



From the Whittlesey beach the lake level appears to have dropped 

 quite abruptly 80 or 85 feet to the Wayne beach. This drop in the 

 lake level was, of course, due to a movement of retreat on the part of 

 the ice front. The Wayne beach lies at a level in the Saginaw 

 Valley barely below the head of the channel, which had served as the 

 outlet of Lake Saginaw. At the same time it is quite certain that no 

 outlet was open toward the northwest through the Straits of Macldnac. 

 The outlet at the time of this beach seems to have been m the east 

 along the ice margin, where it rested against the hills south of Syra- 

 cuse, N. Y. 



LAKE WARREN. 



Here, as in the previous stages, the ice front readvanced, covering 

 part of the gi'ound previously vacated. This movement closed the 

 outlet which had recently been opened near Syracuse and raised the 

 lake to the level of the Warren beach. In the Saginaw Valley the 

 Warren beach passes 20 to 25 feet above the col at the head of the 

 Grand River outlet channel. The col is extremely flat and much 

 covered with dunes. 



In jSIichigan the Warren beach extends up to the vicinity of the 

 Au Sable River, north of Saginaw Bay (fig. 4), but has not been 

 identified farther. In New York it has been traced some distance 

 east of the Genesee River. Its limits in Ontario have not been 

 determined. While the Warren beach was bemg made, the Wayne 

 beach v/as being destroyed, and the Warren beach, like the Whittlesey, 

 shows in some places, but not so strongly, the characters which mdi- 

 cate rapid accumulation. The correlative of Lake Warren in the 

 Michigan basin is shown in figure 4. It is probable that a glacial 

 lake, known as Lake Duluth, also then existed m the western part 

 of the Superior basin with outflow southward through St. Croix 

 River. 



In the later stages of the lake waters, from Lake Arkona on, the 

 area covered by the lakes, included not only the basin of Lake Erie, 

 but a part of those of Lakes Huron and Ontario also. In New York, 

 Fairchild reports evidences of a readvance of the ice and raising of 

 lake level later than the one which affected Lake Warren. That 

 movement, however, appears to have been confined to the Lake 

 Ontario basin and had no effect upon the waters of the Huron-Erie 

 basin. 



The beaches of all the foregoing lakes are horizontal in the southern 

 part of the region they cover, but in the northern part they rise 



