266 HOWE <fc CROSS — GLACIAL PHENOMENA, SAN JUAN MOUNTAINS 



sisting of the early breccia of Cimarron ridge unmixed with other rocks. 

 At the time these observations were made, before the Horsefly drift had 

 been discovered, it was thought that the mounds and heaps of debris 

 might represent material which fell from the cliffs of the ridge on snow 

 fields or neve, and that it was thus transported relatively short distances 

 and deposited by the neve ice without having the form of a moraine built 

 by a well denned glacier. There is today no trace of a divide between the 

 old valleys of the Uncompahgre and Cow creek, but from the elevation 

 of the Horsefly drift, 10,000 feet, it must be assumed that the ice stood 

 at a nearly corresponding level on the east side of Cow creek, whether 

 or not the glaciers united nearer their sources than the junction of the 

 present streams. If this assumption be correct, the supposed drift be- 

 tween Cow creek and Cimarron ridge can have been deposited only after 

 the glaciers of the early stage had begun to shrink, since the drift occurs 

 quite a little below the 10,000-foot level, and the stratified gravels below 

 them, already referred to, appear to have been contemporaneously de- 

 posited — a relation which could not exist had the main valley been filled 

 with ice at the time the supposed drift was laid down. Although it 

 seems not impossible to attribute these deposits to such an origin as has 

 been suggested, yet it must be admitted that satisfactory proof is lacking. 

 On the other hand, no other explanation to account for the existing 

 conditions seems acceptable. 



EARLY TERRACE GRAVELS OF COW CREEK 



Gravel-covered terraces occur at two well marked levels below the 

 supposed old drift and above the younger valley train. The upper ter- 

 race, shown in plate 25, figure 2, on the extreme right, is merely an ex- 

 tension of the old graded surface on which the drift rests and is covered 

 by stratified and water-worn gravels separated by rather an indefinite 

 line from the drift. The terrace form and gravel cover are preserved 

 only as remnants between the streams tributary to Cow creek and in a 

 few isolated hilltops and mesas, one of which lies west of Cow creek 4 

 miles from its mouth and is shown in plate 25, figure 2, on the left. The 

 gravels of the Cow Creek drainage consist entirely of material derived 

 from the volcanic rocks and vary in size from a few inches to a foot in 

 diameter. The deposits are of variable thickness and are usually thin 

 near the old supposed morainal material. The maximum observed thick- 

 ness is 50 feet near the edges of the terraces overlooking Cow creek and 

 on the mesa a short distance from the mouth of Cow creek, which are the 

 only points where the gravels are well exposed, as they are elsewhere 

 covered by 2 or 3 feet of fine, reddish soil, possibly of eolian origin. The 



