542 H. MOHN. METEOROLOGY. [NORW. POL. EXP. 



up, the temperature-reading was taken at once, and the rods were again 

 put down. The depth of the observations was measured from the surface 

 of the ice to the bulb of each thermometer. There was as a rule only very 

 little snow round the holes, and that was much trampled down. 



I received the impression that the observations on the whole gave fairly 

 trustworthy results, but one source of error is the brine contained in the ice, 

 which was apt to fill the bottom of the holes, especially the deeper ones, 

 even during the coldest season; during the summer all holes were filled with 

 briny or saline water, the salinity of which, however, decreased inversely as 

 the temperature. As this brine would come from the pores of the ice and 

 run down to the bottom of the holes, it is possible that during winter-time, 

 when the temperature of the ice increased from the surface downwards, brine 

 from the upper, coldest layers of the ice might sink down to the bottom of 

 the hole, where the temperature was much higher, and as the freezing point 

 of this very saline brine would be lower than the temperature of the bottom 

 of the hole, it might influence the latter, and the temperature-reading might 

 become too low. 1 In the warm season, when the temperature of the ice 

 near the surface was higher than that lower down, the brine would naturally 

 be less saline, and consequently lighter in the upper layers than in the 

 deeper, as the salinity of the brine varied with the temperature of the ice: 

 the freezing point of the brine would always be the same as that of the ice. 

 In the lowest layers of the ice, below the depth of the minimum temperature, 

 the temperature-reading might naturally become too low in the above manner, 

 even in the summer. Although wood conducts heat relatively slowly, it is 

 possible that the temperature-readings from the small depths near the sur- 

 face of the ice, may have been slightly influenced in this manner, especially 

 during the summer, when the upper ends of the rods projecting above the 

 surface of the ice were much warmed by the radiation of heat from the sun. 

 A still greater source of error might be due to the fact that in small depths 



1 This disturbing influence of the brine might have been avoided by a water-tight 

 iron tube, completely filling the bore-hole, and sufficiently strong to withstand the 

 pressure of the plastic or viscous ice, but unfortunately we had no material for making 

 such tubes, and moreover iron tubes might also have a disturbing influence upon the 

 temperature-readings by conducting the heat. If tubes of this kind were to be used, 

 I believe it would be better to make them of a material that conducted heat slowly, 

 e. g. glass, if it were not too easily broken. 



