108 MAEIOX EXPEDITION TO DAVIS STRAIT AND BAFFIN BAY 



region is estimated to have a volume of 50,000,000 cubic feet, while 

 the average for the Grand Bank is about 0,000,000 to 8,00!),000 cubic 

 feet. The average being around Newfoundland is about 100 feet 

 high. The average berg in Disko Bay was estimated by the Marion 

 expedition to be twice as high as the average Grand Bank's berg, 

 but it is very difficult to judge accurately, because the height of the 

 land surrounding Disko Bay dwarfs them. The general form of the 

 Disko bergs appears bulkier, and there are few of the deep-valleyed 

 forms that are so common around Newfoundland. The usual melt- 

 ing signs which also appear on the underwater surface of the bergs 

 in lower latitudes are less apparent in Disko Bay. 



Structure, Color, and Density of Icebergs 



It is well known that the ice composing icebergs is formed from 

 snow crystals which have gradually sublimated into large granular 

 ice. This ])rocess of ice formation entangles and im})risons a vast 

 quantity of air in the mass producing a structure which is quite 

 different from ordinary pure ice. Barnes (1906) shows that the 

 material is relativel}' hard below temperatures of 15° to 20° F., but 

 if it be warmed, its power of resistance rapidly decreases, giving 

 away almost entirely at the melting point, where it is soft indeed. 

 An iceberg which invades the Atlantic and comes under summer tem- 

 peratures, becomes pudgy and soft on its outside. A projectile 

 plunges into the berg with a " chug *' but in cold weather it only 

 shakes down a shower of dustlike crystals. Steenstrup (1888) ob- 

 served that by reducing the tcn:perature of berg samples the air 

 bubbles inclosed in the ice were subjected to a strong compression, 

 and in that condition pricking by a needle was sufficient to burst the 

 ice with the crack of an explosion. Barnes (1927a, p. 93), however, 

 from his observations upon the size of the air bubbles in the ice. and 

 after they have been released, claims that under natural conditions 

 the air imprisoned in iceberg ice is under no greater than atm()si)heric 

 pressure. It is also interesting to note that Barnes found that the 

 air imprisoned in iceberg ice probably thousands of years ago has 

 the same composition as the atmosphere of the present day. He has 

 published the results of some recent experiments (Barnes 1928, 

 p. 345) to determine the amount of air contained in berg ice, and 

 although the work is preliminary, it was found that 7 to 15 per cent 

 with an average of 10 per cent of the co-volume was air.^^ Small 

 fragments of iceberg ice floating in sea water have been observed to 

 effervesce with a sputtering sound plainly audible a foot distant quite 

 similar to the frying of bacon. Such a reaction would ])robal)ly indi- 

 cate furthermore the release of innumerable minute air bubbles under 

 greater than atmospheric pressure. Lieutenant Connnander Ricketts 

 informs me that he has observed large bubbles of air rising to the 

 sea surface on calm days near the sides of icebergs. Undoubtedly the 

 melting of the ice under water causes many of the small bubbles to 

 collect and arrive at the sea surface as large air bubbles. The size of 

 the bubbles permeating the ice may vary considerably from a diam- 

 eter of one-fourth inch or more to a minuteness invisible to the naked 

 eye. Tlie average is less than the size of the head of a common })in. 



s^Thuras (1914, p. 68) found a contont of 1 per cent, but, he add?, the ice used for the 

 experiments was unusually clear and probably contained a minimum amount of air. 



