THE HEART OF TH E ANT ARCTIC 



with sacking to prevent it filling up with snow. This would 

 also reduce the circulation of air. 



On the first day, when the shaft was sunk to a depth 

 of 5 ft., there was a difference of 23° between the surface 

 (minus 21.0° F.) and the bottom (plus 2.0° F.). When 

 the bottom at 15 ft. was reached a week later the whole 

 range from top to bottom was 26° (top, minus 6.0°; 

 bottom, plus 20.0°). 



On account of the high temperatures at the bottom we 

 found these shafts very comfortable places to work in, and 

 could lie down to partake of lunch, on a luxurious couch 

 made of ice chips, in perfect comfort, when the air was 

 down between minus 30.0° and minus 40.0°. 



Series of temperatures were taken at every 2 ft. in 

 depth daily for a about a fortnight. Afterwards a series 

 was taken once a month, as we had not time to read 

 them oftener. 



In the diagram are shown the curves of temperature 

 for six months, at one reading per month, compared with 

 the curve of mean air temperature for the four weeks 

 preceding each reading of the temperatures of the lake- 

 ice. 



The scale of temperature on the left side reads up 

 and down, from plus 28.0° Fahr. to minus 20.0° Fahr. 

 The time reads from left to right, in months, July to 

 December. The lowest curve, drawn thicker, is the mean 

 air temperature. The dotted line is the surface tempera- 

 ture. The other four hues reading from below upward, 

 are the temperatures at 4 ft., 8 ft., 12 ft., and 15 ft. 

 (bottom). The thin zigzag line is the weekly mean of 

 the air temperatures. The air curve is always much 

 lower. The others maintain their relative positions pretty 

 steadily, except the surface curve, which fluctuates, and 

 becomes highest of all in December. 



These curves show some points of interest. The 



364 



