Boorendo. Although this is considerably higher than theory or observations in other parts of the world give, yet it is not the inferior limit of perpetual snows, which in the intra-Himalayan regions depends upon a variety of causes, such as locality, the quantity of snow that falls in winter, and the reverberation of the sun's rays from a large extent of elevated land, such as is found within the Himalaya, in Chinese Tartary, and Tibet. I have measured a snow bed at 10,700 feet, which had resisted the influence of the whole summer's sun, but most of this had descended from the rugged mountains above. The limit of perpetual snows is lowest on the outer Himalaya, and here the continuous snow beds exposed to the south are about 15,000 feet. It should be observed, however, that a great deal of snow falls there; whilst in the interior, where the sun shines bright throughout the rainy season, no perpetual snow occurs at much greater altitudes. In ascending Keoobrung Pass, 18,313 feet high, in July, no snow was found on the road. In August, when I crossed Manerung Pass, 18,612 feet, there was only about a foot of snow which was new, and had fallen a few days before; and the snow beds that occurred on the road to the pass, must have been the accumulation of ages, having been precipitated from the surrounding cliffs. This, although perpetual, cannot be called the inferior limit, otherwise the snow-bed at 10,700 feet might with equal reason, for had the road to Manerung been upon the face of a range, and not in a deep glen, there would have been no snow. In October, on the ridge above Nako, we ascended to 19,411

feet, and the snow, which was all new, and no more than a few inches deep, was only met with in the last 400 or 500 feet; this was on the face of the range exposed to the west, but upon the opposite side, no snow was seen at almost 20,000 feet. Purgeool, by my measurement 22,488 feet, was quite white in August, but in October it had a very different appearance, and one face of it was entirely devoid of snow; no doubt, part of this had tumbled down, but it was evident, from the form of the mountain, that it could not have been the case with the whole. The snow, therefore, melts at this elevation, and perhaps still higher. In fact, there can scarcely be any limit accurately settled; for there are some very mild and clear winters, during which it is possible that not half a foot of snow falls in Ludak, and the sun's rays on the loftiest peaks that we have visited, are very powerful in August and September.

In latitude 31° snow falls every year at 6,000 feet, at Soobathoo in the same latitude, and 4,200 feet high, it is seen once in two or three years, but it melts almost immediately; it sometimes falls at Nahun, 3,200 feet, and it has even been observed so low as the Dehra Doon, no more than 2,300 feet from the level of the sea, and in latitude 30°20′.

REMARKS UPON BAROMETRIC HEIGHTS.

Most of the few heights mentioned in the above account of Koonawur, were observed by barometer, and calculated according to La Place's formula, and Raymond's Coefficients, from cotemporary observations taken at Soobathoo and Kotgurh, the altitudes

of both of which places were correctly ascertained, the former by the mean height of the mercury for four years.

As every thing depends upon the accuracy of the instruments employed, I shall observe, that the barometers used by my brother and myself in 1818, which were the first successfully carried through this quarter of the hills, were manufactured by a native of India, and every precaution was taken to ensure precision. The mercury was revived from Cinnabar with iron filings, in an iron retort, and boiled in the tube, which from the thinness of the glass, we found a most tedious and laborious operation, occupying from ten to twelve hours each tube, and although we succeeded in boiling fourteen, yet we broke nearly as many; these tubes were of various lengths, from twenty-three to thirty-one inches: they were compared together, and found to agree as nearly as can be expected, there not being a difference of more than of an inch, which is what the scale read off to; they were immersed in a basin of mercury, and placed perpendicular by a plummet ; and the best proof of the air being completely expelled, is that tubes of half an inch and eight inches vacuum, shewed exactly the same height, and on applying a lighted candle to the top of the shortest, the mercury seemed to rise: whereas had there been the least air, it must have sunk by the expansion, which would have been clearly perceptible in so small a space. The scales were fir rods graduated by myself to of an inch from a Trough

M

ton's standard brass scale, and they were fitted exactly to the surface of the mercury.

Two barometers were left at Soobathoo, and out of the fourteen which we took with us, only two returned in safety, and these agreed exactly with the others.

The barometer which I used in 1821 was constructed by Dollond, and was of the most improved kind, like those mentioned by Troughton in Dr. Brewster's Encyclopedia. It had a vernier divided to Too part of an inch, a glass cistern and screw to adjust the mercury to zero of the scale. I received two of these barometers, with twenty spare tubes filled and boiled by Dollond, to fit into the frames. I left one at Soobathoo, and the other which I took with me, together with six spare tubes, after an absence of five months returned in safety, and on comparison was found to be only .002 or part of an inch different from the former, and what was equally satisfactory is, that one of the barometers that we employed in 1818, which is still in existence, stood about of an inch higher than the mean of four tubes boiled by Dollond; so the heights observed in 1818, are if any thing under than above the truth. The difference may arise from the mercury which we used being of a less specific gravity than Dollond's, but from the want of a sensible balance we were unable to determine this.

At altitudes of 14,000 and 16,000 feet, we generally remarked that the mercury in the cistern of the barometer appeared as if adulterated with lead

or tin, and stuck to the fingers, but it seemed quite pure when we descended to lower places.

At lofty points of 15,000 feet and upwards, one tube was never trusted, but two or more were put up and they almost coincided. At the highest peak, where the mercury shewed 14.675 inches, three tubes gave exactly the same result.

The temperature of the mercury and air were particularly attended to, and observed with a very sensible Dollond's thermometer; the latter is by far of the most consequence; but I have even found the temperature of the mercury, from being exposed to a burning sun, 25° hotter than the circumambient air, which amounts to above sixty feet.

The temperature of the mercury was ascertained by making the bulb of a very small thermometer touch the tube of the barometer in several places, from twelve-and-a-half inches, the lowest part of the scale, to the top.

Baron De Humboldt remarks, that a considerable number of good observations made at the top of St. Bernard, prove that the whole of the barometrical calculations are too great or too little every time that the temperatures are above or below the mean temperature of the two stations, and also that the observations in the morning give the heights too little, because La Place's formula was deduced from observations made at noon.

The altitudes observed in 1818 and 1821, of which there were above a hundred at the same places, sufficiently prove this; but I shall leave others to decide how far it would be proper to compute the elevations from the mean annual tempera