42 METEOROLOGY 



condition most favourable for their growth appeared to be a sudden fall of 

 temperature during an interval of calm weather, just after a blizzard had stripped 

 the ice off the sea. 



The finest development of ice flowers, shown on Plate VII., Fig. 2, took place on 

 March 20, 1908. 



The ice under the ice flowers on March 20 was about 3 inches in thickness, and, 

 though tough, bent very much under a man's weight. The meteorological conditions 

 which led up to the development of these " ice flowers " were the following : — On 

 March 8 there was a very violent blizzard which drove out practically all the ice 

 which had formed on McMurdo Sound. On March 11 there was a light gale from 

 S.E. threatening to develop into a blizzard and keeping the sea ice-free. The wind 

 continued mostly from the S.E. The temperature was continually falling, and on 

 March 16 the sea, which was now much warmer than the air (about 20° Fahr, as 

 compared with 6 '3° Fahr.), was covered with steam or frost smoke, as though it had 

 been boiling [see Fig. 1, Plate VII.). On March 18, 19, and 20 the weather was 

 calm, and, with a mean air temperature on March 19 of — 8 '9° Fahr., the sea surface 

 froze over very rapidly, and a splendid crop of ice flowers developed {see Fig. 2, 

 Plate VII.). On that afternoon about half an inch of snow fell. The following 

 day the temperature fell further, then rose again on March 24, 25, and 26, reaching 

 a few degrees above zero Fahr. On March 27, at a temperature of 1'50° Fahr., all 

 that remained of the ice flowers were miniature domes of snow about 1-| inches high 

 and 2 inches wide. These were pierced by slits where the plates of the crystals had 

 been dissolved out. Lying around the bases of these domes and pointing outwards 

 radially, were the still frozen edges and tips of the plates, the petals, as it were, of 

 the flower. These being formed of pure ice derived from the frost smoke had 

 remained frozen, while the centre of the flower, formed of various cryo-hydrates, 

 had melted away. Thus under the centre of each dome was a tiny pool of bitter 

 saline water. Dr. Mawson has thus explained the origin of these ice flowers : 



" During the formation of the surface ice some of the sea salts are squeezed upwards 

 through capillary cracks to the surface, and there in the form of concentrated brine 

 eventually freeze as cryo-hydrates and form nuclei for additions from atmospheric 

 vapour. The net result is the production of little rosette-shaped aggregates of 

 radiating crystal blades, which were met with up to 2 inches in height." * 



We also observed ice flowers, as Murray has already recorded (op. cit. p. 341) 

 on the freshwater ice of Clear Lake and Blue Lake near Cape Royds. "They were 

 on the ice rapidly and ti'anquilly formed in the trenches sunk for the observation 

 of temperature. They were much smaller than those on the sea ice, being only half 

 an inch or less in diameter." 



Bund Structu7-e in Sea Ice. Another phenomenon due to freezing of vapour is the 

 raising of the edges of ice-floes through the freezing of vapour rising from the sea 



* " Heart of the Antaictic," vol. ii. p. 3S7. 



