FRIDTJOF NANSEN. 



This thermometer was lost by an accident simultaneously 

 with my large insulated water-bottle on August 9th, 1900. 



Twenty six Negretti and Zambra reversing thermometers be- 

 longing to the Norwegian OceanicSurvey, were examined and tested 

 by Mr. Helland-Hansen and myself, and five or six of those that 

 seemed to register most uniformly were selected for use. These 

 thermometers were of the ordinary type and divided into whole 

 degrees Centigrade, each degree of the scale being about 3 mm. 

 long. By aid of a specially constructed reading microscope, these 

 thermometers could be read without any error of parallax, which 

 is very important, as their stems are so thick that great care 

 has to be taken, if considerable errors in the reading are to be 

 avoided. The microscope, which was made by Leitz (of Wetzlar, 

 Germany), had a micrometer eye-piece, and a foot by which it 

 could be attached to the thermometer, so that its axis was always 

 perpendicular to the stem, and level with the top of the brooken- 

 off mercury. By aid of the micrometer scale the indication of 

 the thermometer could easily have been read off with an accu- 

 racy of 0*01 °G. if it had not been for the rough graduation of 

 the scales of these thermometers. 



We nevertheless obtained in this manner very accurate deter- 

 minations of the deep-sea temperatures, and these old thermo- 

 meters proved to be considerably better indicators than was at 

 first expected. 



A very important point with the Negretti and Zambra ther- 

 mometers is naturally that the temperature of the broken-off 

 mercury must be exactly known at the moment it is read off. 

 I therefore had it arranged in such a way that the thermome- 

 ters could easily be removed from the reversing apparatus when 

 they came on deck. The thermometers were then placed in a 

 water-bath of known temperature for 10 or 15 minutes before 

 they were read off; and the temperature of the water-bath was 

 introduced in the journal along with the reading. 



