The Interval Between the Glacial Epochs. 83 
glacial in the sense that they were not contemporaneous with the glacial 
incursion at its earliest maximum. They may have been contemporaneous 
with the very earliest stages of glaciation before the ice reached the Mis¬ 
sissippi Valley and was able to mingle its deposits with those of valley. 
Now these gravels occupy a wide area stretching across the basin of the 
lower Mississippi from some distance back in Tennessee, Kentucky and 
Mississippi to the high lands upon the Arkansas side, appearing in Crow- 
, ley’s ridge that bisects the present bottoms of the Mississippi. The gravel 
stratum undoubtedly was originally horizontal, but it now undulates more 
or less conformably with the surface. The explanation of this, it seems 
to me, is found in the gradual creep of the soft material of the hills as they 
were slowly carved out by erosion. The brows of the hills in some cases 
have obviously crept down on the slopes, for on the summit we find the 
gravels compact and firm and the constituent pebbles lying with their 
maximum diameters in a horizontal position while the stratum has level 
upper and lower bounding planes. On the slopes of the hills however the 
gravel beds are more or less broken up, and the pebbles have been dis¬ 
turbed and displaced and tumbled into various attitudes, such as we 
might naturally expect under the hypotheses of a creeping movement on 
the slope. It seems impossible to suppose that this stratum of gravel was 
originally deposited in the undulatory form in which it is now found. It 
might be supposed that the silt which overlies this gravel bed was depos¬ 
ited as a mantle over an undulatory surface, but gravel does not lend it¬ 
self to such a method of distribution. 
This overlying mantle, which now claims attention, consists of fine silt 
and embraces the loess deposits of the lower Mississippi. It spreads out 
broadly over the gravel stratum and extends somewhat beyond it, es¬ 
pecially on the east. This stratum is in places differentiated into two 
parts and separated by a soil-like horizon. This differentiation is not com¬ 
mon to the entire valley. This silt mantle may be traced almost in un¬ 
broken continuity northward to the border of the glacial drift, whence it 
spreads itself over the drift, reaching up on it some hundreds of miles to 
the northward. In this northern stretch the silt mantle is correlated with 
n second episode of the earlier glacial epoch. It graduates down into a 
stratum of boulder clay that overlies a bed of vegetable material, which in 
turn overlies another till. Both of these tills I have been accustomed to 
correlate with the earlier glacial epoch. I do not wish, however, to raise 
differences of opinion on that point here. It is unimportant to the main 
conclusions which we desire to reach. 
Besides this continuity there is a further reason for regarding these silt 
deposits as contemporaneous with the ice invasions. They are made up in 
part of glacial particles; that is, particles derived from the mechanical 
abrasion of a glacier. These particles consist of decomposable silicates, 
dolomites and limestones, and were rasped from rocks of these varieties 
lying further north. Such decomposable particles do not abound in resid¬ 
uary clays but are abundant constituents of glacial clays. 
