had fallen between 20 and 23 October. However, there was a brine 
covering of three-sixteenths of an inch, which appeared to cover the 
entire area of new ice. A sample of this brine, taken 50 feet south 
of the deck on 23 October, had 9 salinity value of 62.3 °/oo. Evi~ 
dently, this salinity value is the result of rapid ice formation on 
and after 12 October. The ice was able to form with comparatively low 
air temperatures, averaging about 7° F. between 12 and ll; October. Over 
a large area 300 feet south of the dock, walling was found to be very 
difficult owlng to the slippery brine covering. 
Three ice temperature and salinity profiles, as illustrated in 
figures 1), 15, and 16, were taken from ice in the immediate dock area. 
The salinity values were taken at 2-inch intervals vertically; i.e., 
O-2, 2- inches, etc. Temperature readings were taken at 2-inch ine 
tervals, starting at the surface. It will be seen that the temperature 
gradients are nearly linear, while salinity decreases irregularly with 
depth, 
Several ice thickness measuremants were made from time of forma= 
tion until the ice had attained a thickness of approximately 13 inches. 
Only one measurement was made after this time, and that was on 18 Novem- 
ber 1953. The new ice thickness at that time was 28 inches with ? inches 
of snow cover. Ice growth as a function of degree days of frost (° F.) 
and the accumulation of degree days of frost with time for the new and 
old ice are shown in figures 17 and 18. Degree days of frost are based 
on tho normal freezing point of the water at each location and may be 
expressed either in ° F. or ° GC. To illustrate the use of degree days of 
frost, a day with an average temperature of 25° F,. would accumulate h de- 
gree days of frost when the base temperature of 29° F. is used. It is the 
practice in the Hydrographic Office to use a base of 29° F. or =1.8° Ce at 
sites with salinity between 25 °/oo and 35 °/oo while a base of 32° or 31° F. 
for fresh or brackish water is used.» 
In figure 19, the ice growth as a function of degree days of frost 
is shown for two ice seasons, 1918-9 and 1953=-5h. The curves are nearly 
identical for the overlapping portion. Fhysically, this identity ex= 
presses the fact already noted that the water of North Star Bugt is well 
mixed because of the free exchange to depths. Since the composition of 
the sea water does not change greatly from year to year, the relationship 
between ice growth and the heat loss expressed in degree days ef frost 
is also the sane. There is, however, a wide variation in the accumla= 
tion of degree days of frost. Figure 20 shows the available historical 
data on degree days of frost and reveals that the extreme values ranged 
from the total of 7,950 degree days of frost on 31 May 195), to the total 
of 5,650 on 31 May 1917. This variation would be expscted to cause con= 
Siderable differences in ice thickness, assuming the other paremeters were 
unchanged. 
V. ICE GROWTH COMPUTATIONS 
In computing the ice growth a formula daveloped in the Hydrographic 
Office (Lee and Simpson, 195) is used, which tales into account the in- 
fluencing oceanographic and metecrological parameterse 
