NORTHWESTERN LAKES OF THE UNITED STATES. 
107 
at any time. In 1912 a heavy snow fell during the first week in September. The 
high wall of the crater helps to protect the snow from the sun. The high walls 
surrounding the lake protect it from the direct rays of the sun during the early 
morning and late evening, but this is of small consequence, because the early 
morning sun and the late evening sun furnish but a small percentage of the total 
heat for the day (Kimball, 1910), 
Although only two surface temperatures were taken, it seems that the latter, 
12.1° C., must be very near the maximum surface temperature, unless it was a 
little higher before the storm, which lasted for four days immediately preceding 
September 5. During this period the waves ran so high that we were unable to 
go on the lake with a 25-foot launch. This storm is mentioned here because it 
may have lowered the surface temperature by mixing the epilimnion and increasing 
the temperatures just below the surface, and it also gives some idea of the winds 
in this crater. The first thought is that there would be little wind in such a depres- 
sion, but such does not seem to be the case, as on this occasion the lake appeared 
as completely covered with whitecaps as other lakes of similar size. If these wind- 
storms were common one would expect a thermocline, but as the maximum differ- 
ence in temperature between the surface and bottom of the lake is only 8.6° C. 
there is very little chance for a definite thermocline. The temperature drops grad- 
ually as far down as 70 to 100 m., where it reaches 4° C., the temperature of water 
at maximum density and at atmospheric pressure. This is the lowest temperature 
found in any of the deep lakes of the United States during the summer season. 
In Crater Lake the temperatures at 100 m. and below are less than 4° C. 
These temperatures were checked with all possible care, using a standardized 
Negretti-Zambra thermometer, which was then checked at the University of Wash- 
ington. For the second set of readings a Schmidt-Vossberg (Berlin) thermometer 
was used with the one above mentioned. The Schmidt-Vossberg instrument was 
calibrated to 0.1° C. and contained a small thermometer to give the temperature of 
the mercury column, so that the correction for the expansion of the latter could be 
applied. After numerous trials more concordant results were obtained by reading 
it at the temperature of the surface water. This thermometer was not standardized 
when used on Crater Lake, but the correction was afterwards determined and applied. 
Table 9. — Temperatures taken at 100 m. and below, Crater Lake, Oreg., 1913. 
Depth in meters. 
Negretti-Zambra 
thermometer. 
Schmidt- 
Vossberg 
ther- 
mometer. 
Depth in meters. 
N egretti-Zambra 
thermometer. 
Schmidt- 
Vossberg 
ther- 
mometer. 
Aug. 1. 
Sept. 5. 
Sept. 5. 
Aug. 1. 
Sept. 5. 
Sept. 5. 
100 
°F. 
35.1 
38.2 
°F. 
°C. 
3.9 
400 
°F. 
°F. 2 
°C. 
200 
500 
38.3 
3.45 
3.5 
3.5 
300 
3. 45 
3.5 
500 
300 
600 
38.3 
38.3 
A temperature below 4° C., although not found in any other fresh-water lake in 
the United States during the summer season, is readily explained by the table 
of temperatures of water at maximum density under pressure in Landolt-Born- 
