THE DATE OF THE GLACIAL PERIOD. 559 



preglacial channel till it was driven out by the encroaching 

 moraine. It was the easier removed from its old channel by 

 reason of its reduction in volume. When it began its course 

 in its new channel, it flowed over a broad plain of gravel and 

 sand, the then latest accumulations of glacial torrents. This 

 plain of gravel and sand extended throughout the adjoining 

 space now occupied by such drift deposits. The same kind of 

 deposits filled the whole Minnesota Valley, from side to side, 

 and rose as high as the plains back of Fort Snelling. The 

 river, being comparatively small, had but little effect on these 

 deposits. If it excavated any channel, the torrents from the 

 ever-present glacier-ice filled them at once — indeed, it exca- 

 vated, it refilled, as it was glacier-born.. It was on the retire- 

 ment of ice, bringing a greater drainage area into contribution 

 to swell the main streams at this latitude, that these rivers 

 began to deposit the fine loam -sand which covers the coarse 

 gravel and sand of these terraces. It was still later, when 

 the rivers were shrunk, by the partial or complete with- 

 drawal of the glaciers from their remote sources, that they 

 began to excavate through the loam and the gravel and 

 sand and finally entered on the slow erosion of rock-gorges. 

 Thus it appears that the date from which the recession of 

 the falls must be reckoned was after the outlet of Lake 

 Agassiz had been opened toward the north, one of the last 

 acts of the Ice age. . . . 



Finally, if all the supposed irregularities be allowed their 

 full force, and all the elements of doubt be admitted, their 

 combined effect would not, at the most, more than slightly 

 modify the result. And even if it should double the first result, 

 or should reduce it to one half, the chief value of the calcula- 

 tion is not impaired. That consists in showing the lateness of 

 the last Glacial epoch compared with the enormous time that 

 has sometimes been supposed to have elapsed since its de- 

 parture. 



If the occurrence of our winter in aphelion, caused by the 

 precession of the equinoxes and the revolution of the line of 

 the apsides, about eleven thousand three hundred years ago, 

 was the cause of our last Glacial period, it follows that it re- 

 quired about thirty-five hundred years for the withdrawal 



