144 The American Geologist. ^^^^'^^h, 1903 
A red soil was found at the surface of the till surrounding 
the lake, but close to the immediate border this soil received 
an intensely black stain and was probably marshy. The ele- 
phant remains were found in this soil just beyond the margin 
of the silt deposit and it is reasonable to suppose that the ani- 
mal came to its deatli by approaching the lake and miring 
down into the marshy soil on its margin. The animal remains 
did not arrive at their position in this glacial deposit by the 
action of the glacier as it is beyond dispute, that the ttisks 
would not have remained unbroken or unabraded. Further, 
as already shown, the association of the remains indicates that 
the animal actually lived and died on that spot. Being within 
the glacial basin, this must have occurred after the maximtmi 
extension of the glacier. After the destruction of the lake- 
let, the soil was no longer marshy and the conditions were not 
so favorable for the incorporation within it of the remains of 
a large animal. The question of the probable age of the ele- 
phant is, therefore, reduced rather narrowly to that of the 
lakelet. 
The Coffee creek passes the obstruction at the lower end 
of the basin in a gorge cut through a massive serpentine point 
on the north of the old valley. It has an estimated length of 
500 feet, a depth of 100 feet, a width at the bottom that of the 
creek, 20 to 30 feet and at the top 200 feet. Sheared serpen- 
tine is a formation easily eroded, but massive serpentine resists 
erosion well. In this respect it is comparable with granite and 
other very hard rocks of this region. The gorge is cut through 
a particularly resistant portion of the serpentine mass and its 
steep walls are encumbered by large angular blocks of this 
rock. The gorge cutting has required a considerable time 
and this is the measure of the time since the maximum exten- 
sion of the Union creek glacier. This gorge constitutes the 
most important part of the evidence that glaciation in this 
region has been much more complex than I have heretofore 
supposed it. Unfortunately, I have no standards with which 
to compare this gorge to arrive at its absolute age and all I 
can do is to show the probability that it is much older than 
rock gorges occurring nearer the heads of the glaciated valleys. 
The silt deposit at the Nash mine is so situated with refer- 
ence to this gorge as that the lake was completely drained long 
