182 GLACIOLOGY 



appearance of the brine pools formed through the thawing of the centres of the 

 ice flowers surrounded by their fringe of ice plates is shown on the accompanying 

 sketch. 



In regard to the history of the cui-dling over of McMurdo Sound with ice 

 and its final congelation, it may be stated that the surface of the ice first became 

 crusted over during the early days of March. This ice attained a thickness of 

 from 9 to 10 inches when it was broken up and removed by a blizzard. It 

 was on this occasion that the ice flowers just described were formed. The next sheet 

 attained a thickness of 18 inches, when it was broken up by a northerly swell, and 

 removed by a southerly gale at the beginning of April, with the exception of 

 a small patch at the extreme north of Backdoor Bay. On April 2 a fresh sheet 

 of ice began to form, again accompanied by ice flowers. The fragments of ice that 

 had been driven out at the beginning of April now became recemented, as shown 

 in the accompanj-ing photogi-aph, Plate LIV. Fig. 2. 



A few days later McMurdo Sound was permanently frozen over for the winter. 



Scale of Inches 





Fig. 58 



Nevertheless, fiom time to time, during blizzards, large areas of ice broke out from 

 near Cape Royds and towards the centre of McMurdo Sound, leaving a long wide 

 tongue of dark open water extending several miles south of Cape Royds towards the 

 " pinnacled ice " in the direction of Mount Discovery. At the beginning of July the 

 thickness of the ice, measured at a crack extending from Flagstafi^ Point to the 

 stranded berg, and about 400 yards distant from the point, was 4 feet. The thick- 

 ness of the sea ice at the dredging ground in Backdoor Bay at the end of July was 

 between 4 feet 6 inches and 5 feet. 



On August 23 the ice at a spot very near the one just mentioned, 400 yards 

 southerly from Flagstafi" Point, was trenched for the purpo.se of putting down a fish 

 trap, and proved to be 6 feet thick. At the end of September the bay ice in the 

 extreme northern corner of Backdoor Bay proved to be 7 feet thick, the maximum 

 measured by us, but there seems no doubt that this thickness would steadily increase 

 until probably about the end of November. It may be mentioned that we observed 

 that the delicate plate-like crystals, which develop ultimately into pancake if the 

 surface of the sea at the time of theii- formation is rippled, or into smooth clear ice 

 surmounted by ice flowers if the weather at the time is calm, either form at the 

 surface or possibly at a slight depth below the surface, though there is some doubt as 



