was ~1o39 C. and the warmest temperature at 50 meters was ~0.36° GC. 
The salinity curve shows the beginning of convection in the first 
10 meters. The seasonal. thermocline lies between 75 and 100 meters 
and is still fairly sharp. 
In the second station profile (fig. h) taken a week later, the 
shallow layer has cooled further, so that the warmest water now has 
a temperature of ~0.5° C. The surface convection extends to 15 meters. 
At this time, the ice was 1.6 inches thick. The seasonal thermocline 
had weakened during the week. 
In the third station profile (fig. 5) taken 6 days later on 
12 October, cooling has reduced the temperature of the upper layer so 
that it is less than that of the lower layer, thus eliminating the 
seasonal thermocline. The nearly isothermal lower layer now is the 
warmer of the two layers. Convection has produced an isohaline layer 
in the upper 20 meters, while the continuing surface cooling has brought 
the temperature to the freezing point. At this time, a 6-inch cover- 
ing of ice at site A was the result of thermohaline convection. 
Finally, 9 days later on 21 October, the station profile (fig. 6) 
shows that the surface cooling has extended below 75 meters. hLittle 
change in the salinity and in the depth of the mixed layer has occurred. 
Since the ice thickness was nearly 12 inches, it is evident that the 
loss of a moderate amount of heat through the ice produced a relatively 
large amount of ice with lithle added change in the convection of the 
water. At this point, the winter oceanographic structure is well ese 
tablished. The precipitation of salt and the process of convection dure 
ing the formation of ice are shown clearly in the four prefiles. The 
surface salinity increased from 31.20 °/co on 29 September to 31.6 °/oo 
on 6 October, to 32.00 °/oo on 12 October, and to 32.21 °/oo on 21 
October. 
IV. ICE FORMATION AND GROWTH IN 1953 
Fringe ice was first noticed on 25 September along the eastern edge 
of North Star Bugt where fresh water empties from Pitufik Valley. By 
1 October, grease ice was forming over North Star Bugt in the area north- 
east of Delong Pier; a considerable amount of slush ice formed along the 
eastern shore. By 3 October, the first new ice was formed over this 
area of the bay. Temporary patches of grease ice were forming in the 
area of sites A and B. The first sheet of young ice formed at site B 
during the morning of October and at A during the morning of 5 October. 
This new ice attained a thickness of about 2 inches by 6 October, the 
thicknesses being 1.6 and 2.2 inches at sites A and B, respectively. 
The areal distribution of ice at this tims is illustrated in figure 7. 
& polynya with some grease ice existed slightly north of site A, and a 
few smaller open-water areas were present. Some rafting had occurred. 
Ice continued to grow with little change in areal distribution, attain= 
ing a thiclmess of l.8 inches at site B by 9 October (1000 IST). The 
snow cover amounted to one-half inch. Figure 8 shows the synoptic ice 
picture at this time, The polynya north of site A had decreased in size; 
a small polynya had developed a short distance southwest of Delong Pier; 
and a small area in the immediate vicinity of the pier had becoma ice 
3 
