NO. 17.] THE TEMPERATURE OF THE POLAR ICE. 559 



1895 August 12. The ice-thermometers are read each evening and then 

 taken up. Next morning the holes are rebored, and the 

 thermometers lowered. 



31. The thermometers placed in new holes, and at depths 

 of 0-5, 1-0, 1-5 and 2'0 metres. The thickness of the 

 floe 3-10 m. 



October 2. a. m. The hole for 0'5 m. changed from 4 to 2 inches 



diameter, isolated by reindeer-skin. 



1896. February 15. Therm, not read, because the observer could not cross 

 the lane. 



May 20. Therm. 0*5 m. carried away by the dogs. 



23. The ice- thermometers removed to the floe close to the 



ship, and 5-inch holes bored for 1'5 m. and 2 m. 

 27. Therm. I'O m. and 1*5 m. fast in the holes. 

 30. Therm. I'O m. and 1*5 m. taken up. 

 June 15. All the instruments taken on board. 

 16. Put out again in their former places. 

 26. Thermometers down since 2 p.m. 



The above Remarks give an idea of the difficulties to which the obser- 

 vation of the temperature of the ice were subjected. 



The following Table gives the monthly means for the different years of 

 the observed temperatures for the air, the surface of the ice, and the various 

 depths. 



In order to obtain the mean temperatures for the depths of 0'4, 0'8, 1*2, 

 1'6 and 2'0 metres, I have calculated the missing values for each month and 

 year by means of the formula 



log 4 = log to f.h, 



where fo is the temperature at the depth h, t at the surface, and f a coeffi- 

 cient (the logarithmic decrement). 



The formula gives values, whose difference from the observed values 

 ranges from 0- 6 for the winter-months to 0'l for the summer months. 

 The computed values are enclosed in brackets. 



