A few rude cairns were erected on the crest of the pass, covered with wands, red banners, and votive offerings of rags. I found a fine slab of slate, inscribed with the Tibetan characters, "Om Mani Padmi hom," which Meepo allowed me to take away, as the reward of my exertions. The ridge is wholly formed of angular blocks of white gneissy granite, split by frost.* [It was not a proper granite, but a highly metamorphic felspathic gneiss, with very little mica; being, I suspect, a gneiss which by metamorphic action was almost remolten into granite: the lamination was obscure, and marked by faint undulating lines of mica; it cleaves at all angles, but most generally along fissures with highly polished undulated black surfaces. The strike of the same rock near at hand was north-west, and dip north-east, at various angles.] There was no snow on the pass itself, but deep drifts and glaciers descended in hollows on the north side, to 17,000 feet. The rounded northern red shoulder of Kinchinjhow by Cholamoo lake, apparently 19,000 feet high, was quite bare, and, as I have said, I ascended Donkia to upwards of 19,000 feet before I found the rocks crusted with ice,* [Snow, transformed into ice throughout its whole mass: in short, glacial ice in all physical characters.] and the ground wholly frozen. I assume, therefore, that 19,000 feet at this spot is not below the mean level at which all the snow melts that falls on a fair exposure to the south: this probably coincides with a mean temperature of 20 degrees. Forty miles further north (in Tibet) the same line is probably at 20,000 feet; for there much less snow falls, and much more melts in proportion.* [Two secondary considerations materially affecting the melting of snow, and hence exerting a material influence on the elevation of the snow-line, appear to me never to have been sufficiently dwelt upon. Both, however, bear directly upon the great elevation of the snow-line in Tibet. From the imperfect transmission of the heating rays of the sun through films of water, which transmit perfectly the luminous rays, it follows that the direct effects of the rays, in clear sunshine, are very different at equal elevations of the moist outer and dry inner Himalaya. Secondly, naked rock and soil absorb much more heat than surfaces covered with vegetation, and this heat again radiated is infinitely more rapidly absorbed by snow (or other white surfaces) than the direct heat of the sun's rays is. Hence, at equal elevations the ground heats sooner, and the snow is more exposed to the heat thus radiated in arid Tibet, than in the wooded and grassed mountains of Sikkim.] From the elevation of about 19,300 feet, which I attained on Donkia, I saw a fine illustration of that atmospheric phenomenon called the "spectre of the Brocken," my own shadow being projected on a mass of thin mist that rose above the tremendous precipices over which I hung. My head was surrounded with