8 A CONTRIBUTION TO THE STUDY OF ICE-STRUCTURES 



merely, it appeared as if the individuals in the bottom six inches had been reduced in 

 dimensions, with the production of an aggregate of very even-sized factors averaging 

 one-eighth of an inch across. Closer inspection revealed a more complicated fabric. 

 Instead of the prisms with polygonal cross-sections, which prevailed in the ice near 

 the surface, they were seen to be flattened into the form of plates set vertically, often 

 interpenetrating or twinned. The optical orientation of adjacent plates was often 

 identical. In this way composite flash-faces were met with as much as two inches square. 



Saline matter and gas-bubbles, mechanically held within the ice, had taken up 

 positions in vertical tracts between the ice-prisms. Such inclusions were quite obviously 

 elongated in a vertical direction. The saline tracts were more or less opaque and white 

 in colour. Near the bottom of the lake, where, atthe time of freezing, the mother-liquor had 

 been very saline, the salt content had materially affected the crystallisation and bulked 

 large in the solidified product. The bottom three inches, which had resulted from the 

 freezing of a liquid containing about fifteen per cent, of solid matter in solution, afforded 

 an excellent illustration. On standing at a temperature still well below 32° F. a specimen 

 of this ice slowly became freed from brine by its draining away. The interspaces once 

 occupied by brine were soon, for the most part, occupied by air, thus affording oppor- 

 tunity of examining the crystalline arrangement of the ice-plates. Upon draining for 

 a while, the specimen showed, in relief, bundles of similarly oriented thin leaves of 

 ice growing vertically downwards. In the crystallisation, as growth continued by lateral 

 expansion of the plates, much brine had become imprisoned, so producing the yellowish 

 enamel-like ice of the deeper zone. 



Coming back to the ice forming the first twelve inches below the surface; this appeared 

 on a casual glance to be quite different to that below, for a strongly marked horizontal 

 structure, outlined by layers of somewhat flattened spherical gas-bubbles, caught the 

 eye. The first of the horizontal layers of bubbles was four inches below the surface, the 

 next three inches farther ; then the layers came every half-inch, losing distinctness 

 altogether at twelve inches. The contents of the bubbles were, no doubt, chiefly air, but 

 to this it is likely that some sulphuretted hydrogen was added. The prismatic structure 

 best developed at a depth of fifteen inches persisted more or less definitely to the surface, 

 though becoming more confused in the upper layers. Stratified layers, each corresponding 

 to a particular freeze, are a regular feature of lake- and sea-ice. In this case the extreme 

 freshness of the surface layers suggests crystallisation from fresh water. Perhaps 

 abundant snow additions fell upon the surface at the time of freezing. If it were not 

 that the region is disturbed at frequent intervals by violent blizzards, one would be 

 inclined to suggest that when completely thawed the lake, in the short summer, does 

 not regain a state of even salinity. The winter freezing results in pushing the salts 

 downwards, and the extreme density of the lower layers, when thawed the following 

 summer, also tends to keep the salts to the bottom. 



3. ROUND LAKE 



For further details see Vol. I, p. 165. A tarn of very small dimensions near the north- 

 west end of Blue Lake. It averages some thirty yards across, and is only about two 

 feet deep. When trenched in early April it was found to be frozen to the bottom. The 

 water contained much chlorides and sulphates, which were so concentrated in the lower 

 portions as to give it a yellow opaline appearance. The specific gravity of the thaw- 

 water from this latter was l - 009 at 4° C, and the temperature of commencement of 

 freezing was 30-95° F. This yellow ice showed strongly the vertical tracts common 

 to saline ices. These tracts were from one to three inches long in the vertical, and 



