280 The American Geologist. May, 1895 
and upper till, the latter being usuall}' from five to twenty 
feet thick. ^'• 
Continuing northward into Minnesota, frequent or occa- 
sional observations of this forest and peat bed are afforded b}^ 
wells, as described by Prof. N. H. Winchell, in Fillmore, 
Mower, and Freeborn counties on the southeastern border of 
this state, adjoining Iowa, a thickness of peat varying from 
one to eight feet being found enclosed between deposits of 
till at depths from 20 to 70 feet below the surface. f 
Farther west and north in Minnesota, where I have ex- 
plored and mapped the drift and its moraines, well sections 
encountering remnants of these interglacial deposits are very 
rare, but sometimes several are found within a few miles of 
each other, testifying to a partial preservation of the old land 
surface over considerable tracts, to Lyon, Renville, and Mc- 
Leod counties, 60 to 90 miles north of the Iowa line. Still 
rarer occurrences of probably the same forest bed are also 
shown b}^ my records of wells as far northwestward as Fergus 
Falls and Barnesville. The second of these towns, situated 
218 miles north from the northwest corner of Iowa, is the most 
northern locality w^here I have evidence of an interglacial bed 
apparently belonging to the time of glacial recession preced- 
ing the lowan stage or epoch of ice accumulation and exten- 
sion. 
It seems very significant that Barnesville is three miles 
west of the highest or Herman beach of the glacial lake Ag- 
assiz and about 75 feet below it (plate X). Another localify 
of an interglacial bed, containing many small gasteropod 
shells beneath 26 feet of till, is in Mitchell township, Wilkin 
county, Minnesota, also lying within the area of lake Agassiz 
and nearly 100 feet below its highest beach. If the surround- 
ing countr}^, when these interglacial beds were formed, had 
the same relative altitudes and slopes of its drainage as at the 
time of final retreat of the ice-sheet and existence of lake Ag- 
assiz, the land in Barnesville and Mitchell must have been in- 
capable of forest growth or swamp deposits, being covered by 
*"Pleistocene Historv of Nc)rtlieast(M-n Town," P^levo'iilh AnnuHl Re- 
port, U. S. Geol. Survey, for lS89-'00, Part I, pp. ISO-.")';'?, with plates ii- 
Lxi, and 120 figures in the text: the forest bed being described, willi 
numerous sections, in pages 480-49(). 
fGeology of Minnesota, Final Report, vol. i, 1SS4, pp. 313, 363, 3U0. 
