222 WARREN UPHAM, ESQ., ON CAUSES OP THE ICE AGE. 



42 Somewliat analogous Avith the foregoing is the second of 

 these objections, namely, the ivlly proved low altitude of 

 the glaciated lands when the ice-sheets attained their 

 maximum extent and during the diversified and fluctuating 

 history of their recession. It must be recognized, however, 

 that we have in the complex series of drift deposits left for 

 our examination only a representation of the later and 

 closing phase of the Ice age, while tlie land was low or 

 near its present level. In North America the comparatively 

 much longer early phase of high altitude leading to the ac- 

 cumulation and slow extension of the ice-sheets is not clearly 

 represented by the drift and numerous moraines of the 

 glacial retreat or of the extreme limit of glaciation, but by 

 the earlier fluvial Lafayette formation, in which, according 

 to Hilgard, coarse gravel from the Archean areas near the 

 head of the Mississippi was carried down by that stream 

 quite to the shores of the Gulf of Mexico. 



43 The wane and departure of both the North American and 

 European ice-sheets have been marked by many stages of 

 halt and oscillation, Avhereby the flora, inclurling iuveat 

 trees, and less frequently traces of the fauna, of the tem- 

 perate areas adjoinmg the melting and mainly receding ice 

 were covered by its drift at the times of temporary re- 

 advance of the ice-border. No better illustration of condi- 

 tions 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 

 Russell in bSyO 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 both in 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 Chaniberlin, Salisbury, and Leverett; but farther 

 north, as in the large region of the glacial Lal^e Agassiz, the 

 withdrawal of the ice-sheet and foi-mation of successive 

 moraines marking shght halts and re-advances due to secular 



