the Conductivity of Liquids. 



335 



variations in the increase are due to changes in the magnets, 

 or the wire, or both, the effect of which may therefore be 

 allowed for accurately. It is true that possible changes in 

 the distribution of the magnetism cannot be taken into ac- 

 count ; so that, if the upper magnets changed more or less than 

 the lower ones, the comparison plate would indicate too great 

 or too small a change ; but no error of any importance is 

 likely to result from such a cause. Also changes in the con- 

 ductivity of the brass due to temperature have not been con- 

 sidered, though they can be at any time. After the observed 

 conductivity of a liquid has been corrected by means of this 

 comparison plate, it is independent of changes in the magnets 

 or the wire, and so all results obtained during any length of 

 time are directly comparable. Of course, after any great 

 length of time an experiment would be made on oO-p.-c. sul- 

 phuric acid to serve merely as a check. There is one more 

 point that must be noticed before giving the results. Owing 

 to the high specific gravity of sulphuric acid, it seemed hardly 

 safe to use the full quantity of 1000 c. c. of acids stronger 

 than 60 p. c; and therefore the smaller quantity of 750 c. c. 

 was used ; and in order to compare the effect produced by 

 750 c. c. with that by 1000 c. c, 750 c. c. of 35-p.-c. acid was 

 taken, and its effect observed after correction by the com- 

 parison plate. The apparent resistance obtained was 1*66 

 times the true resistance given by 1000 c. c; and therefore, 

 of the four solutions containing respectively 70, 80, 90, and 

 95'5 p. c. anhydrous sulphuric acid (H 2 S0 4 ), only 750 c. c. 

 were taken, and the observed resistance divided by 1*66. 



The method of calculating the results is given next. The 

 speed is measured by the number of turns in one second. 

 Each of these numbers, together with the reading of the scale 



H 2 SO, 25 ; H 2 75. T. 16° C. 



Exp. 



Speed i -16-81 



Eeading 1087 



b. 

 34-60 



88-1 



c. 



+ 31-59 



165-3 



d. 



+ 14-68 



145-8 



e. 



+ 14-04 



174-3 



in the line below, is the result of at least four observations. 

 e is taken with the comparison plate attached. The first four 

 columns are then taken in pairs, and the sum or the difference 

 of the two speeds taken, according as they are of unlike or of 

 like sign, and put in the first column of the next table ; then 

 the differences in the readings are taken and put in the 

 second column. 



2B2 



