We naturally sought an answer to the interminable query as to how much melting took place at the highest altitudes. Melting of course was always quicker on rocks. But even on the glaciers it was remarkably rapid whenever the sun shone brightly, and we were more than once surprised after a period of cloudy weather with constant snow showers to find how much the snow had consolidated. It seemed to us on more than one occasion that while snow had been falling at our camps and on the lower peaks, Everest itself must have escaped. But, generally speaking, after July 6 the mountain was remarkably white and became increasingly whiter, and only at the least two perfectly fine days, which rarely came together, made any perceptible difference. It was remarkable how little ice was ever observable on the steep Eastern face, where one would expect to see icicles hanging about the rocks. It is my own impression for what it is worth, and its value I fear is small, that though snow will melt readily enough low down, at least up to 23,000 feet during the warmer weather even on cloudy days, at greater altitudes, perhaps above 25,000 feet, it rarely melts even in bright sunshine. In September this year I doubt if it melted at all above 23,000 feet after the weather cleared. At lower elevations the direction and angle of the slope made all the difference. After one fine day the snow on a steep East slope had solidified to a remarkable degree at about 20,000 feet; on a North-facing slope at a similar elevation it had been quite unaffected; on flat surfaces 1,000 feet higher a perceptible crust had formed, but the snow remained powdery below it as on the day when it fell. After three and four fine days the snowy surface of a glacier was absolutely hard at about 20,000 feet and remained solid in the afternoon. Fifteen hundred feet higher we were breaking a hard crust and sinking in a foot or more. This condition may have been partly due to the local behaviour of clouds, which were apt to cling about a ridge overlooking the glacier and cast a shadow on this part of it. But higher, on more open ground, we met the same condition; and again the slopes facing North preserved a powdery snow which never changed before it was blown down in avalanches. Perhaps the most convincing phenomena were the powdery snow high up on the Eastern slopes under the North col and the snow on the Western slopes at a similar elevation under Lhakpa La, which was hardly more solid, while 1,000 feet lower we found excellent snow. It is difficult to resist the conclusion that altitude is a determining factor in the sun's power of melting. It is possible that a line might always be drawn on any given day above which the temperature of the air is too cold for snow to melt where it has fallen on snow, and another to meet the case where it covers rocks. From our all too limited observations in June I should judge that in the middle of summer such imaginary lines would be