CONDUCTIVITY AND LUMINOSITY OF FI.AMKS C'oNTAININc YA1MJM .!> SAL1R 107 



No minimum is observable in the case of the sodium or lithium salts, nor in th.- 

 case of the chlorides of oa-sium. rubidium, and potassium, and (lit- bromide of potas- 

 sium. It is prolmbk-, howrvi-r. that in the case of these salts a minimum value of the 

 molecular conductivity would U- found, if higher concentrations could be investigated, 

 for this is distinctly oliservahle in the case of potassium iodide. It will also be seen 

 that with the oxysalts of potassium the minimum is only attained at higher concentra- 

 tions than in the case of the oxysalts of rubidium and ciesium. 



The intermediate character of iodide of potassium is once more evident, for the 

 minimum value of the molecular conductivity is about 6, whilst that for the bromide 

 and chloride must be considerably less, and that for the oxysalts is 11. 



In the case of the haloid salts the variation of conducting power with concentration 

 may be approximately expressed by the equation 



C = k </q, 



when c is conductivity, q the concentration, and k a constant. 

 This is shown for NaCl and KC1 in the following tables : 



The agreement which here is by no means complete cannot be recognised at all in 

 the oxysalts except at low concentrations. In this respect our results differ from those 

 of ARRHENICS, who gives the relation c = k */q as one of general applicability. The 

 d i iff cause of the difference in our results lies doubtless in the fact already mentioned, 

 that, with a solution of given concentration, much more salt was carried into the flame 

 in our cxpriiiiu'iits than in those of ARRHENIU8. 



It is of interest to determine the variation of conducting power from metal to metal 

 of the series of salts investigated. This is most satisfactorily done by making use of 

 at which the individualities of the several salts of one metal havr 

 On account of the relatively bad conducting power of the lithium and 



P2 



