20 A CONTRIBUTION TO THE STUDY OF ICE-STRUCTURES 



began to grow up from the surface neve. In a few hours it was just as if we were 

 walking on a velvet-pile carpet. The wind sprang up later on and mowed down the 

 crystal plates. They reminded me at the time of the flakes of flake anthracene. These 

 delicate plates were formed from vapour rising out of the comparatively warm porous 

 neve below, and becoming chilled by the air temperature. 



The " ice -flowers " of the newly frozen sea are composed of plates similar to 

 those of the Low-temperature Laboratory. Each " flower " is composed of an inter- 

 penetrating bunch of these composite plates, and may be so well developed as to reach 

 a diameter of two inches {see Plate V, Fig. 2). In the case of a typical crop which 

 formed over Back-Door Bay in the autumn, the thickness of the plates was found to 

 be a two-hundred-and-fiftieth of an inch. Favourable conditions for the growth of ice- 

 flowers are a dead calm and a sudden fall of the temperature to zero or lower. 

 Growth takes place by additions to the plates from atmospheric moisture replenished 

 from the comparatively warm thin ice layer over the sea-water. It was found that, 

 normally, the nuclei for the building up of these rosettes is brine — drops of brine 

 exuded by the hardening sea-ice below. If, therefore, a sample of ice-flowers be 

 collected and thawed, brine and not fresh water is the result. As indicating the 

 degree of salinity may be mentioned determinations of the beginning of freezing 

 of the brine obtained by thawing two samples. One, marked " large old ice-flowers," 

 commenced to freeze at 27-7° F. ; the other, distinguished as " small young ice- 

 flowers," gave 20° F. Obviously, as they grow in size, the original brine is continually 

 diluted. A more irregular growth of the kind is that which throughout the winter 

 is seen to form around the margins of cracks in the sea -ice and over newly frozen 

 leads. In such situations the " flowers " are more fluffy and fern-like, and the brine 

 nuclei may be dispensed with altogether, the deposition taking place more or less at 

 random on any exposed objects. Like growths are seen on the surface of sheets of 

 fresh water — in all cases freezing takes place at low temperatures. For instance, on the 

 freezing surface of water exposed by trenching operations in the winter on the lakelets 

 around Cape Royds. 



Akin to these plates formed from depositions from rarefied vapour are the leaf-like 

 plates which are typical of ice forming as separations from concentrated brines. Crystal- 

 lisations of the kind have been mentioned as forming the lowest and most briny zones of 

 Green Lake and of Shallow Lake. Separations of the kind were also well illustrated 

 as developing from sea -water under certain conditions ; for example, adhering to ropes 

 left freely suspended in sea-water. An instance of the kind is figured on Plate XXVI 

 of Vol. I. A sample of such plates was allowed to drain at a temperature of about zero F., 

 then thawed, and the freezing-point of the liquid determined. Freezing was found to 

 commence at 31-7° F., illustrating the freshness of the crystals. 



(c) Crystallisations from Rarefied Vapour. 



A glance into Wild's Store supplied valuable information in this connection. 

 Adhering to cans of preserved food and on the ceihng were beautiful crystals of clear 

 ice with more or less perfect shape. The ideal form was that of a short hexagonal prism 

 limited by basal planes. In situations where the air current was greater, skeleton forms 

 were the result, and some of these were found so large as to represent portions of hexagons 

 three-quarters of an inch across. 



A good example of the same type of formation was met with in the spring on digging 

 up a seal liver which had been buried in the snow during the winter. The lower side 

 of the liver was covered with perfectly formed short prisms of clear ice (see Plate VII, 

 Fig. 2). 



This form of crystallisation is universal on the walls of crevasses, and in such situations 



