COMPLEX SOLUTIONS. ARCHIBALD. 37 



zero for half an hour at a time. An error of this amount in the 

 determination of the temperature of the solution would cause 

 an error of about 0.1 per cent, in the determination of the resist- 

 ance. The temperature of the room in which the observations 

 were made was from 2 to 5 C. That one might be sure that 

 the temperature of the solution to be measured had come to be 

 that of the bath, measurements of the resistance were made at 

 short intervals, and that reading taken which was found to be 

 constant for successive interval*. The thermometer used was 

 graduated to tenths of a degree centigrade, and could easily be 

 read to twentieths. Its errors had recently been determined at 

 the Physik/alisch-Technische Reichsanstalt, Berlin. 



The factor for reducing the observed conductivities to mer- 

 cury units was found by plotting observed conductivities at 18 

 C. against concentrations, reading off from these curves the 

 conductivity values for concentrations examined by Kohlrausch, 

 and comparing them with his results. The value thus obtained 

 was found to be the same for each salt and to be practically 

 constant throughout the concentration range of my experiments. 

 As the cell was of glass the reduction factor would not be 

 appreciably different at C. from what it was found to be at 

 18 C. To make sure that no change occurred in the position of 

 the electrodes during the course of the experiments that would 

 appreciably affect the reduction factor, every second or third 

 solution was measured at 18 C. before reducing its temperature 

 to C., and the value of the conductivity obtained was compared 

 with that previously obtained at the same temperature. 



Determination of equivalent conductivity at infinite dilution 



for C. 



For this purpose a series of simple solutions of each electro- 

 lyte, of concentrations ranging from O'OL to O'OOOl gramme- 

 equivalents per litre, were prepared, and their conductivities were 

 measured both at 18 C. and at C. The conductivity of the 

 water used in their preparation was also measured at both tem- 

 eratures and subtracted in each case from the conductivity of 



