108 KANSAS ACADEMY OF SCIENCE. 
of the ice sheet in Shawnee and Wabaunsee counties, so far as definitely located, 
is about 1,050 feet, or not over 1,100 feet A. T.* The planation of the surface is 
so slight that, were it not for the small amount of material left by the glacier, it 
would have been difficult indeed to have recognized the former existence of 
glaciers in that portion of the state. Consequently it seems probable that the 
ice sheet was comparatively thin at its southwestern extremity, the limit of 
which is not yet entirely known. The elevation of the divide between the Kan- 
sas and Marais des Cygnes rivers south of Topeka is 1,100 feet. This is over 400 
feet below the boulders of Battle hill, or the more elevated deposits of the Equus 
beds. How far up the Kansas river the loess is found is not known, but it prob- 
ably does not extend to the flint hills in Wabaunsee county. It seems probable, 
therefore, that the waters of the Kansas river would have flowed around the 
foot of the glacier to the east of these hills rather than rise to an elevation of 
1,550 feet, which is even higher than the divide between the Kansas and Neosho 
rivers in the flint hills of Wabaunsee county, which are now considered to be 
quite above and to the south of the terminus of the ice sheet. 
Professor Udden, in the article above referred to, suggests that the water which 
deposited these beds must have connected with another body of water in the 
valley of the Smoky Hill river to the north, and states that the river has cut its 
channel through these deposits. 
The elevation of the McPherson divide at its central point is a trifle over 1,500 
feet. To the southward, ata distance of 38 miles, the Arkansas river flows at an 
elevation of a little over 1,400 feet, or a fall of 100 feet in 38 miles. The bed of 
the Smoky Hill river, eight miles farther north, is 1,300 feet. The city well at 
McPherson, starting 1,475 feet A.T., was put down 150 feet (the present water- 
supply is taken from a depth of 140 feet) without striking the bottom of the de- 
posit. This makes the bottom of the well about 25 feet above the bed of the 
Smoky Hill river two miles south of Lindsborg, or about the same level as the 
bed four miles east of Marquette, still on the northern boundary of the Equus 
beds. The present elevation of the Arkansas river at the mouth of the Little Ar- 
kansas is 1,290 feet. The great range of elevation of the Arkansas in passing the 
beds is due to the distance it travels in passing them, while they are narrow at 
the northern end, and the Smoky Hill flows squarely across the deposit. The 
relation of the two river beds and the records of the McPherson and Halstead 
wells is shown in section 4, plate III. The section begins at the mouth of Sharps 
creek and passes along the western edge of McPherson and a trifle east of Hal- 
stead to the Arkansas. .It will be seen at a glance that the two rivers are at the 
same level at the extremities of the section and that the gravel in the McPherson 
well lies exactly in the same level, while the gravel in the Halstead well is below 
it. The fact should also be borne in mind that the Arkansas has reached its 
base level and filled its channel to some extent, though how much is not defi- 
nitely known. 
The above figures would seem to indicate that at one time the Smoky Hill river 
ran south instead of north and emptied into the Arkansas. But it is difficult to 
understand why it should have excavated so great a channel and covered so great 
a flood plain here and so narrow a valley west of Marquette. However, the en- 
croachment of the sand-hills on its southern area may offer a slight suggestion as 
to the partial choking of the southern outlet causing more rapid deposition to 
the north and thus elevating the channel and widening the flood plain. 
But there are other facts which seem to detract from this explanation of these 
* All elevations based upon U. 8. topographic sheets. 
