278 
BULLETIN OF THE BUREAU OF FISHERIES 
from different combinations of months. The St. Germaine and Rest Lake data were 
used to determine the annual deviation from normal temperature for the periods 
May to July and May to August, both inclusive. The former period of time may be 
taken to correspond approximately to the growing season of the Trout Lake cisco, 
and the latter to the growing seasons of the Silver Lake and Muskellunge Lake 
ciscoes. The annual deviation from normal temperature for the longer Clear Lake 
growing season (May to September, inclusive) were determined from the Minocqua 
station data. In these calculations of deviations from normal temperature, records 
for a single month were found to be occasionally lacking in the data for a station. 
For these few gaps in the data the corresponding records of the nearest neighboring 
station were substituted. This substitution is justifiable for the corresponding devia- 
tions from normal temperature at the different stations were almost always close 
to each other in value. 
Table 56 shows the annual deviations from normal air temperatures as based 
on the different combinations of months and for the weather stations mentioned in the 
preceding paragraph. These data are presented graphically in figures 6 to 8 along 
with the corresponding curves of deviation from average growth. The examination 
of these curves shows that there is little evidence of a connection between annual 
deviation from average growth and annual deviation of the air temperature from the 
normal during the growing season. In the Trout Lake data (fig. 6) the years (1924, 
1928, and 1929) with better than average growth were years with subnormal tem- 
perature for the period, May to July. On the other hand, 1930 with a temperature 
slightly above normal was a poor growth year. The year 1929 was a good growth 
year in both Muskellunge Lake and Silver Lake (fig. 7), while 1930 was a year of poor 
growth. Yet the air temperature for the period, May to August, was below normal 
in 1929 and above normal in 1930. Certain other, years, however, show both sub- 
normal temperature and growth below average (for example, Silver Lake, 1924). 
There is some indication that in Clear Lake (fig. 8) there may be some slight connec- 
tion between annual variations in growth and annual variations in average air tem- 
perature over the period May to September. It may be seen that in Clear Lake all 
the calendar years with poorer than average growth (1924 to 1928, inclusive) showed 
subnormal temperatures during the growing season while 3 of the 4 years with better 
than average growth (1923, 1929, 1930, 1931) had temperatures above normal during 
the growing season. The year 1929 showed growth slightly above average, but tem- 
perature below normal. In general, however, the Clear Lake curves for annual devia- 
tion of the cisco’s growth from the average and the annual deviation of the air tem- 
perature during the growing season from normal show a rather conspicuous lack of 
parallelism. 
Table 56. — Average deviation of air temperatures from normal during the growing season of the cisco 
[Deviations are in degrees Fahrenheit; 1.8° F. = 1.0° C.] 
Stations 
Months 
1923 
1924 
1925 
1926 
1927 
1928 
1929 
1930 
1931 
St. Germain Rest Lake.. 
May, June, July 
2.0 
-5.8 
-1.8 
-1.2 
-3.5 
-0.7 
-2.4 
0.2 
-0.2 
Do 
May, June, July, 
August. 
.8 
-4.8 
-0.2 
-0.7 
-3.6 
-0.5 
-1.9 
.8 
-0.6 
Minocqua 
May, June, July, 
August, September. 
1.1 
-4.5 
-0.6 
-1.6 
-4.3 
-0.4 
-2.6 
.3 
.4 
The failure of the data discussed in the preceding paragraph to show any clear 
dependence of growth in different calendar years on the temperature of the air during 
