22 



Scientific Proceedings, Royal Dublin Society. 



the salt solution. A current of air, previously passed througli a glass wool filter 

 and through distilled water to saturate it with aqueous vapour, was drawn 

 througli the solution under examination. The inlet tube jeaehed to the bottom 

 of the solution ; and the apparatus was immersed in a thermoslat. The current 

 of air was continued for a sufficient time to ensure equilibrium being reached 

 throughout the tube at the observed temperature. The dissolved nitrogen and 

 the sodium chloride in the solution were then estimated. 



The experimental results are shown by the accompanying curve (fig. 1). 



\ 



Saturation value of Nitrogen in c.cs. per litre. 

 P'lQ. 1. — Saturation values of Kitrogen in solutioiis of varying NaCl concentration at lo''C. 



Method of expressing Experimental liesults for the purpose of Comparison. 



The columns of water employed in these experiments were so deep — about ten 

 feet --and the exposure of them to the air had to be continued over such extended 

 periods of time — two to eight weeks — that it was decided not 1o attempt to control 

 the tempei'ature obtaining during the conduct of the experiments beyond 

 preserving the room in which they were carried out, and which was exceptionally 

 well circumstanced for such purpose, at as closely uniform temperature as possible 

 during a set of experiments. 



To have provided the means for controlling, at will, the temperature at which 

 these experiments could be started and continued would have added very 

 considerably to the practical difficulties of an already sufficiently laborious 

 problem ; and it was not thought necessary to attempt to do so, since the authors 

 had good reasons, which will be explained later on, for assuming that when the 

 gas concentrations of a de-aerated column of distilled water, or of a solution of 



