38 



Mr. A. Campbell : Measurements 



In Table 11. are given the corresponding sensitivities ex- 

 pressed in galvanometer scale-divisions (1 mm. at 1000 mm.) 

 for 1 part in 100,000. 



Table II. 



JHethod. 



X. 



Y. Z. 



(1) 



(2) 



(3) 



(4) 



0-7 

 0-8 

 1-2 

 0-3 



0-6 1 0-4 

 0-8 0-28 

 ] -2 : 0-42 



! ! 



In conclusion I may remark that, \\ithin the limits of 

 accuracy indicated by the numbers in Table II., the results 

 given in Table I. show satisfactory agreement. 



Appendix. 



On tlie Sensitivity of T^otentiometers for Measuring 

 Resistance. 



To obtain the best results with a potentiometer it is usually 

 desirable that it should be of such resistance as to give, with 

 the available galvanometer, the maximum sensitivity. If 

 several galvanometers are available, it is important to have 

 a means of predicting which will be the most suitable in any 

 particular case. I have therefore investigated the general 

 formula for the sensitivity of a potentiometer when used for 

 measuring resistances with a galvanometer of given resist- 

 ance, extending it to the more general case by the common 

 assumption that the " current-sensitivity " of a galvano- 

 meter is proportional to the square root of its resistance. 



1 1 



In fig 



1 1 



6 let the resistances have the values sho^^n 

 1 



VIZ. 



.11.. . .^ '^ 



and — : — and - includino- the resistances of the adjacent 

 m n ^ '^ 



X m g 



batteries whose E.M.F.'s are E and <? respectively. Let 0, 

 V^ and V be the potentials at the points so marked. Let 

 t^ = current through galvanometer, so that c=^(v—Y)g. 



— is the part of the circuit under test, while 



x 



is the part 



of the potentiometer in use. 



