wear in part a gloomy aspect, the capricious and fantastic relief comes out with far more telling effect than when they are bathed in sunshine, for then a multitude of their details are obscured. The lake varies continually in breadth, being narrowest when two headlands approach each other from
the opposite shores; indeed it is everywhere so narrow that
the scenery entirely changes after but a very short stretch. In the morning before sunrise, we observed a remarkable phenomenon off the southern shore, south-east from the camp : the lake »smoking» over a not inconsiderable area. Clouds of steam, just as intensely white as the steam from a locomotive, were rising from the surface of the water and were being wafted across the lake by a gentle breeze from the south-west, until they thinned away and disappeared. But the formation of the steam ceased a short while after the sun rose. The cause is evidently a number of relatively warm springs issuing on the shore; anyway the lake water is considerably warmer than the air prior to the appearance of the sun. The narrowest reaches of the lake, beside which we travelled during the first half of the day, were covered with a thin sheet of ice; but after the lake widened out again, it was unfrozen, with the exception of a narrow fringe along the northern shore. Clearly it wanted but one or two cold nights, and this part of the Tso-ngombo would also freeze, and the Tibetans fully expected it would freeze, for once or twice they inquired of people whom they met whether the lake farther west was frozen or not.
On a small pointed promontory there were some freshly formed terraces or ramparts, arranged radially in the way shown on the accompanying illustration (fig. i 87), in that they all ran out towards the extremity of the promontory. They consisted of gravel and sand, and were hard and very regularly built. The higher, inner ones
IIIF Color HighRes
