IRON CREEK. 159 
which tahises with steep slopes, practically uncovered by vegetation, 
descend, gradually merging with the smoother moss-covered slopes of 
the middle portion of the valley walls. 
The streams in their lower courses flow in rock-walled canyons. In 
tracing any one of the streams headward the canyon is found to 
decrease in height and to merge gradually with the older topography 
previously mentioned. The history of these features suggests that 
the former topography of gentle slopes and wide, open valleys was pro- 
duced by the long-continued erosion of rivers and weathering. Sub- 
sec jiiently uplift of the region renewed the down-cutting power of the 
streams, so that canyons were carved in the floors of the old valleys. 
This erosion allowed rapid reassortment of the old gravels and waste 
and thus effected the concentration of any gold or other heavy min- 
erals which may have been contained in the gravels. 
The canyon cutting ceased, however, before it had progressed 
beyond the lower portions of the streams. The interruption was pro- 
duced either by a movement of the land or, as is more probably the 
case, by a climatic change which decreased the amount of water trans- 
ported by the streams. Such a change may have also been respon- 
sible for the disappearance of the local glaciers which were formerly 
present in the Kigluaik Mountains. Whether the climatic change had 
anything to do with the retreat of the glaciers or not is, however, of 
slight importance in this discussion. Some change must have occurred, 
for the streams are no longer down-cutting but actually building up 
the floors of their valleys. The reason for believing that the change 
must have been one which affected the rainfall rather than the eleva- 
tion of the district with respect to sea level is based on the shape of 
the rock canyon. The canyon has a broad, swinging course which is 
so symmetrical that it could not have been produced by the straggling 
present stream, which occupies only a small portion of the floor 
between the rock walls. Many other streams in different parts of 
Seward Peninsula show this same feature. The extensive develop- 
ment of this phenomenon suggests a widespread cause, such as cli- 
matic change, rather than a local cause, such as uplift. 
EVIDENCES OF GLACIATION. 
Another feature of some theoretical interest is the presence of gran- 
ite bowlders on the divide near the low sag at the head of Mattis Creek. 
In the rapid reconnaissance it was impossible to examine the district 
with sufficient care to make a final statement as to the origin of these 
bowlders. It is known, however, that there is no granite of similar 
character south of the Kigluaik Mountains. Furthermore, the granite 
bowlders are un weathered, showing that they have not been in their 
present position a very long time geologically. Although the question 
has not been carefully studied in the field, it is suggested that possibly 
