n
their excavation, can be found in their bottoms. Others, which are
shallower, are not so easily explained. Perhaps where they are found
softer spots existed in the sandstone, places that yielded more readily
to atmospheric degradation, the loose sands being carried away by the
winds.

Just before sundown I attempt to climb a rounded eminence, from which I
hope to obtain a good outlook on the surrounding country. It is formed
of smooth mounds, piled one above another. Up these I climb, winding
here and there to find a practicable way, until near the summit they
become too steep for me to proceed. I search about a few minutes for an
easier way, when I am surprised at finding a stairway, evidently cut in
the rock by hands. At one place, where there is a vertical wall of 10 or
12 feet, I find an old, rickety ladder. It may be that this was a
watchtower of that ancient people whose homes we have found in ruins. On
many of the tributaries of the Colorado, I have heretofore examined
their deserted dwellings. Those that show evidences of being built
during the latter part of their occupation of the country are usually
placed on the most inaccessible cliffs. Sometimes the mouths of caves
have been walled across, and there are many other evidences to show
their anxiety to secure defensible positions. Probably the nomadic
tribes were sweeping down upon them and they resorted to these cliffs
and canyons for safety. It is not unreasonable to suppose that this
orange mound was used as a watchtower. Here I stand, where these now
lost people stood centuries ago, and look over this strange country,
gazing off to great mountains in the northwest which are slowly
disappearing under cover of the night; and then I return to camp. It is
no easy task to find my way down the wall in the darkness, and I clamber
about until it is nearly midnight when camp is reached.

_July 30.--_We make good progress to-day, as the water, though smooth,
is swift. Sometimes the canyon walls are vertical to the top; sometimes
they are vertical below and have a mound-covered slope above; in other
places the slope, with its mounds, comes down to the water's edge.

Still proceeding on our way, we find that the orange sandstone is cut in
two by a group of firm, calcareous strata, and the lower bed is
underlaid by soft, gypsiferous shales. Sometimes the upper homogeneous
bed is a smooth, vertical wall, but usually it is carved with mounds,
with gently meandering