NO. 17.] THE TEMPERATURE OF THE POLAR ICE. 543 



the heat-rays from the sun or the sky penetrating the ice and the melting 

 water (filling the holes), would be absorbed by the dark wood enclosing the 

 thermometers, and that thus the temperature-readings became too high. This 

 may explain the fact that temperatures above zero centigrade were observed 

 in depths near the surface of the ice during summer. 



The observations of the ice-temperatures were made by myself until the 

 middle of April, 1894. As I then thought I had formed a fairly satisfactory 

 method of making them, I asked Capt. S. SCOTT HANSEN to undertake them 

 as a part of the regular meteorological observations. During 1894, and I 

 believe until the summer of 1895, these observations were made on a flat 

 ice-floe on the port side of the Fram. This floe was formed on a wide, open 



water-lane about October 27th, 1893." 



Fridtjof Ncmsen. 



The thermometers used for the observations of the temperature of the 

 ice had been compared with the standard thermometer of the Meteorological 

 Institute before the starting of the expedition. During the drift of the Fram, 

 they were several times compared with standards and with each other, and 

 after the return of the expedition, the thermometers then remaining were 

 compared with the standards of the Meteorological Institute. The corrections 

 were found to vary somewhat with the time, and may be uncertain up to 

 some tenths of a degree C. The circumstances did not always allow of the 

 same thermometers being used in the same depth. 



The holes into which the rods with the thermometers were put down 

 required frequent cleaning, and they very often had fluid water (brine) in their 

 lower part, this being sometimes pumped out. The place of the thermometers 

 was changed several times, new holes being bored on the same ice-floe or 

 the whole system removed to a new floe. 



The thickness of the floe in which the thermometers stood, was not 

 constant. The numbers noted are the following: 



1894. January 27. 1'37 metre 



February 19. 1'87 Snow 4 cm. 



March 10. 2'01 4 



April 6. 2-22 2'8 



10. 2-28 2-8 



