On Freezing-point Depressions for Electrolytes. 505 



prevents the formation of oxygen and nitrogen compounds, 

 thus blotting out soine lines : and that it permits the formation 

 of hydrogen compounds, thus accounting for the appearance of 

 new lines. 



Whatever the explanation may be, the following curious 

 relation was found to exist between the arc and spark 

 spectra, namely, all lines in the arc spectra which are affected 

 by hydrogen, whether enhanced or diminished, belong to the 

 spark spectrum also. 



To illustrate : the arc spectrum of tin in hydrogen shows 

 two strong lines at X 3352*15 and X3283'31, of which there 

 is not the faintest trace in the ordinary tin arc. But these 

 two are among the strong lines of the tin spark. See Hartley 

 and Adeney's list. In like manner, I have photographed, on 

 the same plate, the iron spark in air, the iron arc in air, and 

 the iron arc in hydrogen. In every case examined, the 

 lines affected by hydrogen are spark-lines. 



On the contrary, the lines which belong to Kayser and 

 Runge 's series are unaffected by the change from air to hydrogen. 



If these series prove equally stable in other gases, this 

 stability may form a criterion for dividing a spectrum into 

 two groups — one of which will contain all the series lines, the 

 other pf which will contain none of them. 



North- Western University, 

 Evanston, Illinois. 

 July 1900. 



XLIX. On a Diagram of Freezing-point Depressions for 

 Electrolytes. By Prof. J. G. MacG-regor, F.R.S., Dalhousie 

 College, Halifax, JS T .S* 



[Plate IV.] 



THE object of this paper is to describe a diagrammatic 

 method of taking a bird's-eye view of such knowledge 

 as we possess of the relation of the depression of the freezing- 

 point to the state of ionization in aqueous solutions of electro- 

 lytes, and to show that such diagrammatic study gives promise 

 of throwing much light upon the following questions : — (1) f 

 Has the depression-constant a common value for all electro- 

 lytes, and if so what is it? And (2) what is the state of 

 association, and what the mode of ionization, of electrolytes 

 in solution? 



* Communicated by the Author. An abstract of a paper read before 

 the Nova Scotian Institute of Science. 



t On this question see also a paper recently communicated to the Royal 

 Society of Canada, and to be published in its Transactions for 1900. 



