1886.] Conduction and Molecular Composition, fyc. 279 



in composite electrolytes with rise of temperature. It is true that as 

 temperature rises the influence which individual molecules exert uj)on 

 each other would be lessened ; but on the other hand, the complex 

 aggregates would become more and more completely resolved into 

 their fundamental molecules, the velocity of molecular motion would 

 increase, and the tendency of the constituent atoms to remain united 

 would be lessened. From this point of view the determination of the 

 coefficient of change of conductivity with temperature in the case of 

 substances whose molecular conductivity increases considerably on 

 dilution in comparison with allied compounds which exhibit only a 

 slight variation in molecular conductivity on dilution affords an 

 interesting subject for investigation. F. Kohlrausch has already 

 pointed out (" Pogg. Ann.," 1875, 154, p. 236) that in the case of all 

 neutral salts, " der Einfluss der Temperatur auf das Leitungs- 

 vermogen mit wachsender Verdiinnung sich Anfangswerthen nahert, 

 die zwischen engen Granzenliegen," and the experiments of F. Kohl- 

 rausch and Nippoldt on solutions of sulphuric acid (ibid., 1869, 138, 

 p. 286) show that the resistance diminishes to a much greater extent 

 forequal increments of temperature in concentrated than in dilute 

 solutions. Thus : — 



Table V. 



Percentage of 

 sulphuric acid. 



Resistance 

 (Mercury = 1.) 



Percentage 

 increment of 

 conductivity 

 for 1° C. 



2 



465.100 



0-47 



8-3 



34,530 



0-653 



14-2 



18,946 



0-646 



20-2 



14,990 



0-799 



28-0 



13,133 



1-317 



35-2 



13,132 



1 '259 



41-5 



14,286 



1 -410 



46-0 



15,762 



1-674 



50-4 



17,726 



1-582 



55 2 



20,796 



1-417 



60-3 



25,574 



1-794 



As concentrated solutions would be richer in complex aggregates 

 than dilute solutions, these results are in entire accordance with my 

 hypothesis : it does not appear to me that they can be satisfactorily 

 interpreted in terms of the dissociation hypothesis. 



In cases where the influence of the one member of the composite 

 electrolyte upon the other is but slight, it may happen that the effect 

 of temperature in diminishing this influence will outweigh that due to 



u 2 



