CHEONOLOGY DOUGLASS AND ANTEVS 319 



large lakes were dammed between its front and higher land outside. 

 Because the central parts of the ice-covered areas were more deeply 

 depressed by the weight of the ice than the peripheral regions, many 

 lowlands and valleys that now drain southward at that time con- 

 tained large lakes. The Hudson and the Connecticut Valleys held 

 series of long and deep lakes that extended from the Narrows and 

 from Long Island Sound, respectively, to northern New York and 

 New England. The ocean did not enter these valleys because the sea 

 level of that time stood about 300 feet lower than the present and tlie 

 coast line probably some 95 miles outside Sandy Hook and 10 to 15 

 miles outside the east end of Long Island. The Great Lakes were 

 larger in late glacial times than now; the Timiskamiug-Abitibi- 

 Timmins region was covered by an enormous water body. Lake 

 Barlow-Ojibway ; more than half Manitoba was flooded by the huge 

 Lake Agassiz; and smaller ice lakes existed in all parts of the 

 glaciated area. On the other hand, the St. Lawrence and the Ottawa 

 Valleys were early inundated by a marine gulf, the Champlain Sea, 

 in which, because of the salinity, the fine mud quickly flocculated 

 and settled, forming almost massive clays. 



The sedimentation in the glacial lakes was enormous. Lakes that 

 were small in relation to their drainage areas were actually filled 

 with gravel, sand, silt, and clay. This holds for some of the lakes 

 in the New England valleys. Lakes that were ponded by the ice 

 were suddenly lowered or emptied when lower outlets were opened 

 during the withdrawal of the ice front. Other lakes were drained 

 suddenly as a result of overflow and down cutting of their drift 

 barriers. Still other lakes were emptied because of a gradual low- 

 ering of their outlets by vertical movements of the land. After the 

 disappearance of the water bodies, the lake beds have been eroded 

 and frequently deeply dissected by rivers. It is exposures of the 

 old lake sediments in river banks, in erosional bluffs on residual 

 lakes, and in clay pits and road cuts that have been used to measure 

 the clay varves (pi. 2). 



Thanks to support from the American Geographical Society of 

 New York, the Geological Survey of Canada, Harvard University, 

 the National Kesearch Council in Washington, and other institu- 

 tions, the writer has been able for several years to study the varved 

 clays in the Eastern States, southern Quebec, central Ontario, the 

 Timiskaming-Cochranc region, and northern Manitoba. The first 

 aim has been to make varve measurements at short interspaces along 

 lines running from the periphery of the ice sheet to its center in 

 Labrador, and, if possible, to connect these separate observations in 

 an unbroken record of the rate of withdrawal of the ice border, of 

 time in years, and of the variations of the total summer heat. The 



