PRECIPITATION 41 



feet. The plateau here is also jiractically free from rock dust. It may therefore be 

 questioned whether even near the summer solstice the temperature there rises to 

 thaw point. There can be no doubt that these large ice crystals, where they are 

 developed on the higher plateaux, are formed by a process of vaporisation and 

 recrystallisation at a tempei'ature considerably below freezing-point. Near sea level 

 granulation below the snow surface is to bo attributed to partial thaw in many 

 instances as well as to vaporisation. As granulation is of course much more active in 

 summer than in winter, the drift bi'ought up b}' a blizzai'd in summer consists largelj'' 

 of finely granulated snow. This fact was specially experienced and noted by the 

 Western party. They found that in summer the drift snow on the sea ice of McMurdo 

 Sound was composed in its upper portion of loosely coherent granular snow, the 

 individual grains being much larger than those accompanying the winter drifts. It 

 was from such snow that the largest quantity of the low-flying drift was derived 

 which accompanied the southerly gale of December 7, 8, 9 of 1908. It was a 

 distinctive feature of the summer blizzards, experienced during the Western journey, 

 as opposed to the winter ones experienced at Cape Royds, that the dritt accompanying 

 them was always low-flying and heavy, and if it did reach as high as tlie face — a 

 most unusual thing — it felt like fine gravel. This is no doubt due to the thawing 

 action of the sun on the freshly fallen snow lying on tlie sea ice and glacier 

 surfaces. 



On January 2 the Western party observed a curious type of hoar-frost precipitation 

 from water- vapour or ice-vapour — which is described as follows in the diary of one 

 of us (R. E. Priestley) : 



" On January 2 a new type of precipitation was observed, and one so unusual that 

 we have noted it as follows : 



" Yesterday and the day before, this part of the Sound was traversed by a layer of 

 moisture-laden air moving slowly towards the north-west. To-day the snow and ice 

 are covered by little snow-trees about an inch to an inch andahalf high, and consisting 

 of a central axis with six branches, three short ones to leeward separated from each 

 other by a very acute angle, and these in turn separated by an obtuse angle from 

 three long branches, similarly disposed, to windward. These branches were directed 

 upwards at an angle of 30°, and as both they and the central axis increased in size 

 as the height above the ground increased, the resulting structure looked decidedly 

 top-heavy. Sometimes the trees were compound, having several other single main 

 axes with thoir six branches of exactly similar appearance, except that the branches 

 on those axes nearly at right angles to the stem were of equal length on either 

 side. 



" These peculiar ' trees ' seem to be due to an accretionary growth of ice from the 

 moisture in the air, but the most interesting point about tlieni was undoubtedly their 

 sino'ularlv svnmietrical form. ' 



Ice Flowers. Somewhat analogous in their oriijin are the ice flowers. The 



