formed, skipping the lard- ice stage. Dobrovol'skii related to me in January- February, 1931, 

 during the voyage of the Knipovich in the Kanin-Kolguev region, that he observed a peculiar brash 

 consisting of regular, transparent, thin plates 1 to 2 cm long and 2 to 3 mm wide. These plates, 

 which had not frozen together at all, formed a layer of brash more than a half-meter thick. On 

 top, the brash was completely dry. It packed the intake ports of the cooling system so that it was 

 necessary to stop the engine every 5 to 10 minutes. 



The precipitation of snow on the surface of the sea always accelerates ice formation. Here, 

 the surface layer always becomes fresher, cools, and moreover, introduces nuclei of crystalliza- 

 tion into the water. When snow falls on a sea surface whose temperature is below zero, the snow 

 does not melt but forms a soft, dough-like mass called sneshura ("snow slush"). 



Even when the air temperature is high, snow slush causes ice needles to form in the nearest 

 cooled water layers. When there is turbulence and wind, snow slush, like the brash, gathers in 

 bands consisting of lumps of snow saturated with sea water. With further freezing of the sea, the 

 snow slush bands differ sharply in their appearance and white color from the surrounding ice 

 formed from sea water. 



Thus, the primary forms of ice formation in the sea are ice needles and lard ice, which give 

 the sea a strange, greasy appearance. When the sea is calm, bottle ice and nilas are the next 

 stage. When the sea is not calm, pancake ice is formed, and when the sea is very agitated, brash 

 ice is found. In the majority of cases, ice formation starts from the shores, from individual ice 

 floes, and from shore ice, and extends gradually to the sea. Because of this, all stages of ice 

 formation can be traced simultaneously. 



Thus, on 15 September 1935, along the eastern edge of Franz Joseph Land, when approaching 

 and entering the windward edge of the ice at -10° , we observed from the Sadko the following 

 forms of ice in a small area: lard ice, graying, very thin ice (nilas), pancake ice, consisting of 

 discs with ridges along their edges, and the edge of old ice from which ice formation was pro- 

 ceeding. As they neared the edge of the ice, the pancake ice discs increased in size. It could be 

 seen that the large discs represented the fusion of smaller initial discs. By means of this fusion, 

 pancake ice discs can attain a diameter of two meters. 



As has already been pointed out, initial ice formations look dark steel or lead in color, which 

 is explained by the fact that they are almost completely saturated with water due to their thinness. 

 As the ice thickens, it begins to ride above the water, and at first turns gray, then white. The 

 ridges of pancake ice are the first to start turning gray. Further cooling is accompanied by thick- 

 ening and fusing of individual ice floes, and thus young ice or mol id ik is created, which is 

 light-gray and rough, and which has a surface moistened by the separated brine (figure 33). 



In the majority of cases, the upper surface of young ice is smooth and slightly rippled; the 

 lower surface, during the period of ice formation, is, on the other hand, very uneven and in some 

 cases, where there are no currents, resembles a brush of ice crystals. According to observations 

 made on the Zarya, a layer of sea water 10 cm thick or more, saturated with ice crystals 

 which gradually freeze to the ice from below and thus thicken it, lies directly under the lower sur- 

 face of young ice when it is 2 to 3 cm thick. 



It has already been pointed cut that even lard ice eliminates wind ripples completely on the 

 surface of the sea. Brash, snezhura, and pancake ice completely eliminate the secondary wind 

 waves, and very large waves gradually assume the appearance of a frozen ripple, when such ice 

 extends downward. 



99 



