erywhere several hundred feet above the level of the lakes.
This expedition incurred some loss of time, and when night came we were
still on the ridge.

From our camping-ground we saw fifteen black tents on the hillside. To
the east, on the lake shore, there was a large Gomba, or Lamasery, with
a temple and a number of mud houses. I estimated the distance between
ourselves and the Gomba at only eight miles, a cheering fact, because I
hoped to get there fresh provisions that would enable us to proceed more
rapidly on our journey. We were now quite out of reach of the Gyanema
soldiers, as well as of such troublesome officials as the Barca Tarjum
and the Jong Pen of Taklakot. If we could only obtain a sufficient
quantity of food during the night, and proceed across country early the
next day, there would be little danger of being overtaken by our
pursuers. The Shokas were again shaking with fright at the idea of
entering a Tibetan settlement. I told them firmly that we must reach
Tucker Gomba and village that night.

We had below us the two great lakes. Before I left this magnificent
panorama I could not help taking a last long look at the marvellous
scene. The Devil's Lake, with its broken, precipitous shores, its rocky
islands and outstretching peninsulas, was far more enchanting to me than
the sacred lake by its side, in which, according to tradition, dwelled
Mahadeva and all the other good gods. Although the water was equally
blue and limpid, although each lake had for a background the same
magnificent Gangri chain, Mansarowar, the creation of Brahma, was not
nearly so weirdly fascinating as its neighbor. Mansarowar had no ravines
rising precipitously from its waters. It was almost a perfect oval
without indentations. There was a stony, slanting plain some two miles
wide between the water's edge and the hills surrounding it, except along
the ridge separating it from the Rakastal, where its coast was slightly
more rugged and precipitous.

Directly south of the lake was a chain of high peaks covered with snow,
from which several streams descended. From where we stood we could see
evident signs, as in the case of the Rakastal, that the level of the
lake must at one time have been at least thirty feet higher than it was
when I visited it. The slanting bed of small, rounded, smooth stones,
which extended from one and a half to two miles beyond the water-line,
was evidence enough that the level of the water must have been