yurfh American Inlertjlacial Deposits. — Upham. 281 
an earlier and interglaeial lake with outflow southward where 
hike Agassiz afterward outtlowed to the Minnesota and Mis- 
sissippi rivers, until the outlet was deeply eroded or a very 
far retreat of the ice-border to the north allowed free drain- 
age to take its course as now to Hudson ba}-. Under these 
conditions the growth of an interglaeial forest at Barnesville 
would imply probabl}'' three to six times more glacial melting 
and recession than otherwise would sufhce to account for the 
most northern of these observed interglaeial deposits. It 
therefore seems to me more likely that during the glacial re- 
treat between the Kansas and lowan stages or epochs of the 
Glacial period the present basin of the Red river of the North, 
which was later occupied by lake Agassiz, had a considerably 
greater altitude than now% retaining a part, probably a large 
part, of its preglacial elevation, and that it was thus a land 
surface with southward descent and free drainage along the 
Minnesota river valley to the Mississippi. The recession of 
the ice-sheet before its lowan stage of renewed growth maj" 
then have reached only to the southern part of the Ked river 
valley, instead of the great farther distance to Hudson bay 
which I formerly supposed in writing of these interglaeial de- 
posits in the reports of the Minnesota Geological Survey.* 
The erosion of numerous and large interglaeial stream 
courses in the early Kansan drift sheet of southern Minnesota 
and northern Iowa, including the Minnesota river valley and 
its continuation past Brown's Valley and above the present 
bed of lake Traverse, channelled there apparently about 50 
feet below the general surface of the adjoining country to the 
level of the Herman or earliest and highest beach of lake Ag- 
gasiz,f and the deposition of thick and extensive interglaeial 
beds of sand and gravel observed at widely separated locali- 
ties along the Minnesota valley, ;J; find full explanation in this 
retreat of the ice-sheet to the vicinity of Barnesville and 
*Geologv of Mhinesota, Final Report, vol. i, 1884. pp. 402, 40G, 4(>(). 
470-48.'i, .507, 511, .552. 580, 581, 585-G, (iO!), 625; vol. ii, 1S8S. pp. 138, 180. 
187, 1!)!), 4(i(). 52!), 5.55, GG2, GG8. 
fProc. Am. Assoc, for .\(lv. of Science, vol xx.\ii. lor ISS:!. i)p. 222-227. 
Geoloo-v of Minnesota, vol. i, pp. 479-485, 507, 580: vol. ii. pp. KM. 172. 
21(). 51!)-525. 
:{:Proc. A. A. A. S., I. c. (ieol. of Minn., vol. i, pp..5S|,.")S2.(>2.">: v.. I. ii. 
pp. IRl. 171. 1:2. 
