Kaiifia,'^ Durinfi the Ice Ac/e. 39 



Drift-filled Channels. Of these there are two sorts, one 

 mainly filled with local material, mainly limestone ; the other 

 filled with northern erratics, consisting largely of sand and 

 gravel. The former, we may suppose, were eroded and filled 

 while the ice was quite distant — near enough to supply water, 

 but too remote, or relatively too low down, to contribute much 

 debris from the load of the glacier. The latter would be close 

 to the ice, or so much below it that drift would be washed into 

 the stream almost constantly. 



As the ice sheet advances, streams flowing toward it will be- 

 come obstructed and ponded. If such ponds fill and overflow, 

 the material first carried by them, remote from the ice, will be 

 of local origin. These channels are apt to be parallel with the 

 edge of the ice, or over a divide in some other direction. 



As the edge of the advancing ice sheet approaches the divide 

 it will greatly quicken the action of streams on the further 

 slope, especially if the melting of the ice is rapid, and it is like- 

 ly to be the more rapid at first because of heat from the newly 

 occupied earth underneath. 



As the streams radiating from the advancing ice sheet are 

 swollen the water will increase more rapidly than the debris 

 from the ice is furnished. Two things conspire to this effect : 

 first, the drift is apt to be coarser, and therefore soon dropped ; 

 second, the melting of the advancing glacier is more from the 

 upper and cleaner part of the ice than in a receding glacier. 

 As the ice recedes, though the water may increase in volume, 

 material both fine and coarse will be likely to increase also in 

 larger ratio. 



Hence, if we have reasoned correctly, radiating streams will 

 be eroded rapidly as the ice approaches ; then they will fill up 

 their channels as it becomes nearer. Then they may be cov- 

 ered with the ice and plastered over with till. Later, in the 

 recession of the ice, streams may advance headward and fol- 

 low courses approximately corresponding to which they oc- 

 cupied before the ice displaced them. 



Of the channels filled largely with limestone bowlders we 

 have found two interesting examples. One shows a stratum 

 of bowlders 15 to 20 feet thick, with its bottom about 140 feet 

 above the present Missouri. It extends for several miles along 

 the east side of the Missouri river at Weston, Mo. Its width 

 has not been determined. Mr. F. C. Greene, of the Missouri 



