POSTGLACIAL DEVELOPMENT OP CONNECTING EIVEES OF GREAT LAKES. 477 



more or less duny ridge for nearly the whole length. In some places the sand knolls are pro- 

 nounced, especially toward the south end, where the sand belt broadens. A few dunes have 

 apparently traveled from west of the road northwest of Roberts Landing over into the middle of 

 the swamp. Southward from sec. 10, Cottrellville Township, the sand belt widens to a mita or 

 more. 



ST. CLAIR ESKER. 



The east bank of the channel is not continuous but is nearly so and is composed in most 

 places of a narrow ridge of gravel. For 2 miles south of St. Clan the large glacial ridge men- 

 tioned forms the bank, the gravel in the ridge being fairly coarse. Southward the ridge is 

 lower, almost disappearing in some places. In places where it is low it appears to be partly com- 

 posed of clay and is generally 30 to 40 rods wide. The cemetery 1 £ miles west of Marine City is 

 on this ridge. 



The ridge in all is 11 or 12 miles long and its relations and origin are somewhat obscure. It 

 is thought to be a product of glacial dramage and is apparently an esker made in an ice tunnel 

 or canyon when the ice front was somewhere south of Roberts Landing. In one or two places 

 (for instance, in the bank of Pine River south of St. Clair and where the ridge is cut out by Belle 

 River west of East China) the gravel appears to rest on clay. The ridge is not prominent west 

 of Marine City, but grows stronger a mile or two farther south and is a strong beachlike gravel 

 ridge for about 3 miles to a point southwest of Roberts Landing. 



The difficulties in finding a clear and satisf actoiy explanation for the origin of this ridge are 

 greatly increased by its relations to three other forms of gravel deposit. In its northern part 

 just south of St. Clair it falls in line with the heavier knolls which are regarded as kaines and 

 which may possibly be related to an early position of the ice front at the time of the Port Huron 

 morainic system or possibly to a less pronounced halt of an earlier date. In its northern part 

 it is also closely associated with some of the gravelly river bars. It is here fainter and more 

 broken and is rather more sandy and clayey than it is farther south. It may have been 

 modified to some extent by the rushing of the early river over it. 



In its stronger southern part its crest happens to lie through much of its length almost 

 exactly at the level of the first St. Clair beach (correlative of the Algonquin beach), faint markings 

 of which, found at and below Courtright, on the Canadian side, show the altitude at which the 

 waves acted. This relation seems to be accidental, for all the shore markings at Courtright and 

 at other points in this part of the basin of Lake St. Clair are extremely faint, consisting merely 

 of a faintly cut notch or a sandy or gravelly belt with scarcely noticeable ridging. The ridge 

 west of Roberts Landing, on the other hand, is composed of fairly coarse gravel and stands 12 to 

 15 feet above the clay flats on either side. The ridge as seen from either east or west has the 

 appearance of a beach ridge or spit as strong as the Whittlesey and Warren ridges. Its situation, 

 however, is such that it is impossible for it to have been built primarily by wave action. Lake 

 St. Clair is so small a body of water and it was so shallow, even in the time of Early Lake Algon- 

 quin, that wave action in it must always have been relatively weak. The gravel ridge widens 

 somewhat toward its southern end in a way that is characteristic of the termination of eskers. 

 Besides, the trend of this ridge is southward along the axis of the St. Clair Valley and not at all 

 what would be expected as a product of wave action from Lake St. Clair, where the most effec- 

 tive waves would come from the south. Hence, though the crest of the ridge appears to have the 

 form of a wave-budt deposit and, indeed, has probably been fashioned somewhat by waves, the 

 ridge itself is to be regarded as an esker rather than a beach. 



From 2 or 3 miles north of Marine City southward into the St. Clair delta all the low ground 

 on both sides of the gravel ridge is composed to a depth of several feet of lake clay without stones. 



In another respect the conditions surrounding this ridge were unfavorable to production 

 by waves. Not only was there no wide deep water to provide waves sufficiently heavy to form 

 such a deposit, but the ridge stands out alone on the flat clay floor of the valley with no visible 

 source of supply for its material. No sea cliffs of stony till and nothing corresponding to a 



