264 Wisconsin Academy of Sciences , Arts and Letters. 
One is on one side of a col. Three are on the outwardly pro¬ 
jecting angles of bluffs abutting diagonally on the river (Fig. 
1). One is connected with two ridges projecting from oppo¬ 
site sides, and nearly shutting in a small bluff alcove; and one 
is on the outer scarp of a foothill facing the river. 
If we compare the deposits in question with the normal talus 
we find that whereas the latter follows the general law in com¬ 
position, being a composite of material from all overlying for¬ 
mations, the former, although having an observed vertical 
range of nearly 300', is everywhere practically identical in 
composition. The only included material not derived from 
the limestone being some of the harder sandstone immediately 
sub j acent. 
Wherever I have been able to examine the base, I have found 
the deposit overlying normal talus or rock waste with well de¬ 
fined unconformity, the two deposits being perfectly distinct 
up to the line of junction. 
The hypothesis under consideration necessarily assumes a 
rather notable persistence for certain accumulations of hillside 
waste, exceeding, under certain conditions at least, that of the 
hill from which it was derived. 
It will’ be recognized that this assumption is not in harmony 
with the general law that material detached from a ledge is 
subject to more rapid disintegration than the parent ledge, be¬ 
cause in proportion to bulk a greater surface is exposed to 
weathering. In any given case it would require very clear 
evidence to overbalance this adverse presumption. I shall en¬ 
deavor in what follows, to bring out such evidence as may be 
afforded by local and specific conditions. 
It is not difficult to show for the region under consideration 
that unless it be in certain special cases, the relative endurance 
of fragmental deposits, and of the parent ledge harmonizes 
fully with the law above stated, for although the lower slopes 
of the hills up to an average height of three hundred feet are 
so thickly mantled with Loess that the true talus is rarely 
exposed, the horizon of the Madison sandstone is almost never 
so covered, and all fragmental deposits from the overlying 
