348 Warren Upham — Pleistocene Climatic Changes. 



during late Tertiary and early Quaternary time by the Scandinavian 

 peninsula, and by all the northern coasts of North America from 

 Maine and Puget Sound to the great Arctic Archipelago and Green- 

 land. The abundant long and branching fjords of these northern 

 regions, and the wide and deep channels dividing the many large 

 and small islands north of this continent, attest a very long time of 

 high elevation there. At the time of culmination of the long- 

 continued and slowly increasing uplifts at the north, they seem to 

 have extended during a short epoch far to the south, coincident with 

 the formation of ice-sheets in high latitudes. But when these lands 

 became depressed, and the ice burden of the glaciated countries was 

 removed, they, in some instances, as in Great Britain and New 

 England, returned very nearly to their original- levels, beautifully 

 illustrating the natural condition of equilibrium of the earth's crust 

 before referred to, which Dutton has named isostasy, that when not 

 subjected to special and exceptional stresses it acts as if floating on 

 a heavier plastic or molten interior. 



The wane and departure of both the North American and 

 European ice-sheets have been marked by many stages of halt 

 and oscillation, whereby the flora, including forest trees, and less 

 frequently traces of the fauna, of the temperate areas adjoining the 

 melting and mainly receding ice were covered by its drift at the 

 times of temporary re-advance of the ice-border. No better illus- 

 tration of conditions favourable for the burial of forest beds in the 

 Drift can be imagined than those of the Malaspina glacier or ice- 

 sheet, between Mount St. Elias and the ocean, explored by Eussell 

 in 1890 and 1891, and found to be covered on its attenuated border 

 with drift which supports luxuriant growing forests. Let a century 

 of exceptional snowfall cause a thickening and re-advance of that 

 ice-sheet, and sections of its drift exposed after the glacial recession 

 will show a thick forest bed of chiefly or wholly temperate species. 

 Such re-advances of the continental ice-sheets, interrupting their 

 retreat, are known by well-marked recessional moraines in both 

 North America and Europe. Near the drift boundary in the 

 Mississippi basin some of these glacial fluctuations have involved 

 long stages of time, measured by years or centuries, with important 

 though minor changes in altitude, as shown by the excellent analytic 

 studies of Chamberlin, Salisbury, McGee, and Leverett ; but farther 

 north, as in the large region of the glacial lake Agassiz, the with- 

 drawal of the ice-sheet and formation of successive moraines marking 

 slight halts and re-advances due to secular changes in temperature, 

 humidity, and snowfall, were demonstrably very rapid, the whole 

 duration of this glacial lake being probably only about 1000 years. 

 The vicissitudes of the general glacial retreat seem to me to have 

 been due thus chiefly to variations of snowfall, some long terms of 

 years having much snow and prevailingly cool temperature, therefore 

 allowing considerable glacial re-advance, while for the greater part 

 other series of years favoured rapid melting and retreat. 



Under this view we may, I think, account for all the observations 

 which have been held in America and Europe as proofs of inter- 



