12 H. MOHN. METEOROLOGY. [norw. pol. exp. 



The correction at 0° of No. 4993 was found to be, 



on board, April 12, 1894, 0'0° ] 



} 0-25 

 in Christiania, March, 1897, + O'S J 



The correction at 0° of No. 11001 was found to, be, 

 on board, October, 1893, + 0-25°. 



Only No. 4993 was brought back. All the zero-points may be regarded 

 as having undergone no appreciable change during the voyage. 



The Tonnelot thermometers, corrected for their errors, have been used 

 as standards in the comparisons that were made on the voyage, for finding 

 the corrections of the other thermometers for temperatures below — 20°. 

 These comparisons were only made in the air, the thermometers being 

 suspended in the thermometer-screen. 



III. 15 mercury sling- thermometers Nos. 11 to 25, made by Soderberg. 

 Their stems were divided into whole degrees C, and they had cyHndrical bulbs. 

 They were compared in Christiania in 1893, and their zero-points occasionally 

 verified during the voyage. 



IV. 10 spirit sling-thermometers, Nos. 1—10, made by Soderberg, to 

 which the same remarks apply as for III. It was observed that the fluid 

 became thick at — 40°. 



V. A number of spirit minimum (index) thermometers. Their correc- 

 tions were determined every day that they were in use, by comparison of 

 the reading of the top of the spirit-column with the thermometer for the 

 air-temperature. 



VI. 2 mercury maximum thermometers. They were compared in Chris- 

 tiania in May, 1893, and the comparisons made on the voyage showed that 

 they kept their corrections in a very satisfactory manner. 



VII. 2 sets of registering thermographs from Richard in Paris, specially 

 made for the Expedition. The clock-work turned the drum round in a week, 

 and the registered temperatures could be read to + 0"1° C. 



The thermometers used as dry and wet-bulb thermometers for observing 

 the temperature of the air and determining its humidity, were some of those 

 described above under I and II. The thermometers employed, with their 

 corrections, are as follows: 



