their height trigonometrically, he remained on the summit of the pass (18,500 feet) all night, but unfortunately was so severely frost-bitten that ultimately he was invalided home.

In the year 1883 Mr. W. W. Graham started for India with the Swiss guide Joseph Imboden, on a purely mountaineering expedition; he first went to Sikkim, then attacked the group round Nanda Devi in Garhwal, and later returned to Sikkim and the mountains near Kanchenjunga.

This expedition of Graham's remains still the most successful mountaineering effort that has been made amongst the Himalaya. No less than seven times was he above 20,000 feet on the mountains, the three highest ascents being, Kabru (Sikkim), 24,015 feet, A²¹ or Mount Monal (Garhwal), 22,516 feet, and a height of 22,500-22,700 feet on Dunagiri (Garhwal). It is perhaps to be regretted that Graham did not write a book setting forth in detail all his experiences, though a short account of his travels and ascents may be found in vol. xii. of the Alpine Club Journal.

Arriving at Darjeeling early in 1883, he and Imboden made their way to Jongri just under Kanchenjunga on the south-west, and climbed a peak, Kang La, 20,300 feet. The Guicho La (pass), 16,000 feet, between Kanchenjunga and Pundim, was ascended, but as the end of March was much too early in the year for climbing, they returned to Darjeeling, and Imboden then went back to Europe. It was not till the end of June that Graham was joined by Emil Boss and Ulrich Kauffmann, who came out from Grindelwald. They started from Nynee Tal to attack Nanda Devi, travelling to Rini on the Dhauli river, just to the westward of Nanda Devi. From Rini they proceeded up the Rishiganga, which runs down from the glaciers on the west of Nanda Devi, but they were stopped in the valley by an impassable gorge that had been cut by a glacier descending from the Trisuli peaks. Obliged to retreat, they next attacked Dunagiri, 23,184 feet; after climbing over two peaks, 17,000 and 18,000 feet, they camped at 18,400 feet, and finally got to a point from which they could see the top of A²², 21,001 feet over the top of A²¹, 22,516 feet, and must therefore have been at least at a height of 22,700 feet. Unfortunately hail, wind, and snow drove Graham and Boss off the peak within 500 feet of the top—Kauffmann had given in some distance lower down—and it was only with difficulty that they were able to return to their camp, which was reached in the dark.

The weather then obliged them to return to Rini, from which place they again started for Nanda Devi. This time they went up the north bank of the Rishiganga. After illness, the desertion of their coolies, and all the sufferings produced by cold and wet weather, they reached the glacier in four days, only to find that again they were cut off from it by a perpendicular cliff of 200 feet, down which the glacier torrent poured. Their attempt to cross the stream was also fruitless; so, baffled for the second time, they were forced to return to their camping-ground under Dunagiri at Dunassau, from which place they climbed A²¹, 22,516 feet, by the western ridge, calling it Mount Monal. They then tried A²², 21,001 feet, but were stopped by difficult rocks after reaching a point about 20,000 feet. By the middle of August Graham was back again in Sikkim and got to Jongri by September 2. With Boss and Kauffmann he explored the west side of Kabru and the glacier which comes down from Kanchenjunga. But the weather was continuously bad; they started to climb Jubonu, but were turned back. Then they crossed the Guicho La to ascend Pundim, but found it impossible; more bad weather kept them idle till the end of the month. They then managed to ascend Jubonu, 21,300 feet. A few days later they went up the glacier which lies on the south-east of Kabru, camping at 18,400 feet; and starting at 4.30 A.M. they succeeded, owing to a favourable state of the mountain, in reaching the summit, 24,015 feet (or rather, the summit being cleft into three gashes, they got into one of these, about 30 feet from the true top). It was not till 10 P.M. that they returned to their camp. The last peak they ascended was one 19,000 feet on the Nepaul side of the Kang La. Thus ended this most remarkable series of ascents, carried out often under the most difficult circumstances. Graham, from his account of his travels, was evidently not a man to talk about all the discomforts and hardships of climbing at these altitudes, and this lack of information about his feelings and sensations above 20,000 feet has been urged against him as a proof that he never got to 24,000 feet at all. But any one who will take the trouble to read his account of the ascent of Kabru, cannot fail to admit that he must have climbed the peak lying on the south-west of Kanchenjunga, viz. Kabru, for there is no other high peak there which he could have ascended from his starting-point except Kanchenjunga itself; moreover, unless he had climbed Kabru, neither he nor Emil Boss could have seen Devadhunga nor the two enormous peaks to the north-west, which they distinctly state must be higher than Devadhunga. Now, if they climbed Kabru, they were at a height of 24,000 feet whether they had a barometer with them or not, for that is the height determined by the Ordnance Survey. The heights reached in all their other completed ascents are vouched for in the same way, for if a mountain has been properly measured by triangulation, its height is known with a greater degree of accuracy than can ever be obtained by taking a barometer to the summit.

The next real mountaineering expedition after that of Graham was in 1892, when Sir Martin Conway, together with Major Bruce, and M. Zurbriggen as guide, explored a large part of the Mustagh range. In all they made some sixteen ascents to heights of 16,000 feet and upwards, the highest being Pioneer peak, 22,600 feet.

Arriving at Gilgit in May, when much winter snow still lay low down on the mountains, they first explored the Bagrot nullah. Here they ascended several glaciers and surveyed the country. But huge avalanches continually falling entirely stopped any high climbing. They therefore went into the Hunza Nagyr valley as far as Nagyr. In the meantime, as the weather was bad, they investigated first the Samayar and afterwards the Shallihuru glaciers. At the head of the former a pass was climbed, the Daranshi saddle, 17,940 feet, and a peak called the Dasskaram needle, 17,660 feet. They then returned to the Nagyr valley and reached the foot of the great Hispar glacier, 10,320 feet. From here they travelled to the Hispar pass, 17,650 feet, nearly forty miles, thence down the Biafo glacier, another thirty miles. The Hispar pass is therefore the longest snow pass traversed outside the Arctic regions. About half way up the Hispar glacier Bruce left Conway and climbed over the Nushik La, but joined him again later at Askole.

From Askole the Baltoro glacier was ascended. Near its head the summit of Crystal peak, 19,400 feet, on the north side of the valley, was reached. From the summit, the Mustagh tower, a rival in height to K², 28,278 feet, was seen. To quote Conway's description: 'Away to the left, peering over a neighbouring rib like the one we were ascending, rose an astonishing tower. Its base was buried in clouds, and a cloud-banner waved on one side of it, but the bulk was clear, and the right-hand outline was a vertical cliff. We afterwards discovered that it was equally vertical on the other side. This peak rises in the immediate