Septbmbee 24, 1920] 



SCIENCE 



285 



the waning glacier. Evidence is found in the 

 excessive lime content, and in the occurrence 

 of large ice-rafted bovplders imbedded in the 

 fine clays. Also, the clays often rest directly 

 on smooth glaciated rock, or on the glacial 

 drift. 



The Hudson-Champlain valley is the proper 

 line for the study, and not southern Canada, 

 for in the stretch from l^Tew York City 

 to the foot of Lake Champlain, some 300 

 miles, we have the continuous record of the 

 removal of the Quebec (Labradorian) gla- 

 cier from the United States. The St. Law- 

 rence valley clays are subsequent in time. Of 

 the long period of ice advance we have no 

 record, nor of its long standstill at the Long 

 Island-Staten Island terminal moraine. The 

 existing record is only that of the diminishing 

 ice sheet and the recession of its border. 



The laminated clays were derived largely 

 from the glacial outwash, being the rockflour 

 from the glacial mill. Partly they came from 

 the land wash, by streams eroding the freshly 

 uncovered glacial drift. At the time the clays 

 were deposited the glaciated territory was 

 much lower than to-day, having been depressed 

 by the long-continued weight of the deep ice 

 cap. With the lifting of the burden the land 

 slowly rose. With the land rise the tidal cur- 

 rents and river flow in the shallowing waters 

 swept away most of the clay deposits, or buried 

 them along the sides of the valley under sand 

 and gravel beds. The clays remaining and 

 open to examination are only scattered and 

 minor fragments of the original more or less 

 continuous deposits. These remnants are 

 found along the sides of the valleys; the bulk 

 of the beds which occupied the prisms of the 

 valleys has gone out to the sea. 



In the Hudson we have po considerable beds 

 south of Tappan Sea and Haverstraw Bay, al- 

 though some clays in New Jersey may repre- 

 sent the early time. In the narrow stretch 

 above Peekskill erosion has removed most of 

 the clay, but we find good remnants in the 

 wider valley of the Newburg district. North- 

 ward the next massive clays are at Kingston, 

 with only small remnants beyond. In the Al- 

 bany district and northward the sealevel waters 

 were wide but shallow, and silt and sand plains 



laid down. In the deeper Champlain Valley 

 massive clays were laid, not only in the broad 

 basin but also in the valleys of tributary 

 streams. 



A similar history pertains to the Connecticut 

 Valley and to every deep valley in New Eng- 

 land and the maritime provinces having a deep 

 estuary. 



I The lamination of the clays was produced 

 by interrupted deposition, or variation in the 

 rate of deposit. In the Lower Hudson and in 

 the St. Lawrence the oceanic tides were a 

 periodic factor. In the upper beds of any dis- 

 trict the stream inwash of severe storms was 

 an irregular factor. The periodic element on 

 which emphasis is placed is the seasonal varia- 

 tion of the glacial outwash, due to differences 

 of summer and winter melting. In non-tidal 

 areas this may doubtless be discriminated from 

 the superposed day and night variation; and 

 from tidal variation in the open estuary dis- 

 tricts. The irregular storm factor may not be 

 a serious complication. 



As a time record the clays must be regarded 

 as overlapping, south to north, or as consti- 

 tuting a theoretical vertical column, the south- 

 ern beds at the bottom of the column and the 

 northern beds at the top. It becomes neces- 

 sary, therefore, to correlate the laminae; that 

 is, to determine which particular series of 

 laminEe in a southward section is identical in 

 time with some lower laminse in a northward 

 section, so as to eliminate duplication. With 

 the fragmentary character of the Hudson- 

 Champlain beds it would seem impossible to 

 identify horizons. And correlation of far sep- 

 arated localities with the indefinite moraine 

 lines will be extremely difficult in either the 

 Hudson-Champlain or the St. Lawrence 

 valley. 



DeGeer's estimate of Postglacial time as 

 20,000 years is certainly an underestimate for 

 America. It should be understood that the ice 

 sheet did not diminish steadily, or the front 

 back away continuously. The ice margin had 

 many oscillations, readvanees and reretreats, 

 each probably covering many thousands of 

 years. In the Mississippi Valley between Cin- 

 cinnati and Mackinac F. B. Taylor has 

 mapped fifteen frontal moraines, each of which 



