Dr. E. L. Moss. Observations on Arctic Sea-water and Ice. 547 



to infer that this fragment of floe had lost little of its saltness, cer- 

 tainly none of its salt ice by drainage ; moreover, it afforded decided 

 facilities for obtaining samples of ice from every height of a consider- 

 able section, and accordingly serial chlorine estimations were made 

 from top to bottom of it. 



The lowest part of this section was probably not the bottom of the 

 floe, but samples of deep ice were obtained from an almost overturned 

 mass, exhibiting the mammillary elevations common on the under side 

 of the floes, and in one part studded with stones and grooved by 

 motion against the bottom. Our watering berg offered a section 

 where 30 inches of snow-formed ice lay over what had once been a 

 superglacial pool in the hollows of a "blue-domed" floe. It will be 

 seen, on reference to the tabulated chlorine estimations, that, while the 

 annual floe of one winter's quarters held one-sixth the chlorine of 

 sea water, the most salt parts of the polar floe held but one-fifteenth. 

 The chlorine, however, does not represent all the salts, and water from 

 such ice is quite too brackish to drink. 



If the salinity of the polar floes leaves their mode of growth doubt- 

 ful, evidence much more to the point is supplied by their structure. 

 The upper ice of upheaved segments, not only at Floeberg Beach, 

 but along the shore as far as my sledge journeys extended, displayed 

 more or less distinct horizontal stratification, like that of glacier neve. 

 The stratification was strikingly regular and parallel, decreasing in 

 width from above downwards, sometimes ending abruptly, but gene- 

 rally becoming indistinct and gradually disappearing 20 feet or 

 more from the top. Each stratum consisted of an upper white, 

 merging into a lower blue part, the difference in Colour depending on 

 the greater or less number and size of the air-cells in the ice. The 

 lower blue passed abruptly, but without break in structure, into the 

 spongy white of the layer beneath. After spring, when all exposed 

 ice- surfaces became white and granular, the stratification was to be 

 detected only by the unequal disintegration of the white and blue 

 parts, causing in well-marked floebergs a crenulated outline in the 

 profile of a section ; the blue, hard and unchanged, lying in the angle, 

 and the white, granular and swollen, in the convexity. This unequal 

 disintegration is to be attributed to the lesser transparency of the 

 white parts to solar radiation rather than to any difference in salinity 

 represented by the trifling and possibly experimental difference in 

 specific gravity, shown by estimations Nbs. 106 and 107. The in- 

 distinctness of stratification produced by summer is not permanent. 

 The old floebergs worn into "blue domes," and "pie" shapes were well 

 marked in early spring, though they had evidently borne the thaws of 

 many seasons. The stratification was most readily seen from a little 

 distance. I have marked it in a sketch of a floeberg at Cape Joseph 

 Henry, made quite a quarter of a mile off, but frequently the limits of 



