296 Ai.F Trolle. 



Water readings Difference 



2-50-5 + 10-76 + 0-27 



2-52-5 + 10-75 — 0-01 



The difference between the air and water temperature was 11-7°. 

 The effect of the cold water was appreciable in the innermost cylinder 

 of the water-bottle after about 6 minutes. 



In addition, samples of water were drawn off from the innermost 

 cylinder and it was found on titrating, that the chlorine contents of 

 this water were unchanged before and after letting down ; thus the 

 water-bottle was water-tight. 



Of the two Ekman water-bottles ime was of an older, the other 

 of a newer model. The older bottle would not close properly and 

 was thus not used. 



The new model, on the other hand, worked well as a rule. It 

 was somewhat unreliable in the beginning, as it showed a tendency 

 not to close completely especially in currents. 



This condition improved when a heavier sinker was used. 



It was a light and practical instrument, which was used with ad- 

 vantage especially on the sledge journeys, where lightness is of great 

 importance in the matter of transport. 



In the summer of 1907 this water-bottle was tested by filling the 

 cylinder with water of a definite chlorinity, then let down to ca. ôm. 

 and swung backwards and forwards, which does not alter the chlorine 

 content of the water. 



The Richter reversible thermometers unfortunately were broken 

 on a sledge journey in the screw ice in 1906 — 07. Tliis was a great 

 loss, as it thus became impossible to make temperature determina- 

 tions with the Ekman water-bottle, nor could control temperature 

 determinations be made with the Pettersson-Nansen water-bottle. 



The thermometers of the Pettersson-Nansen water-bottle were of 

 Jena glass III, divided into tenths from — 4 to +14C. 



Control certificates from the Physikalisch-Technische Reichsanstalt 

 in Charlottenburg accompanied the instruments. 



Further, the zero correction was made onboard on ^"/j 1907 and 

 ^/2 1908 in the following manner. 



Some distilled water was laid out to freeze in an open basin. 

 The frozen distilled water was then broken into small pieces and 

 placed in a tall cylinder. The thermometer was now inserted so that 

 both the bulb and the whole of the stalk were surrounded by ice. 



The investigations gave the following corrections. 



Thermometer .... 23505 23507 25207 25208 



»»/1 1907 Corr. + 0-01 0-00 broken - 0-01 



«/a 1908 broken 0-00 - 0-01 



With the aid of a lens the thermometers could be read with 

 certainty to 0-01. 



