ON ELECTROLYSIS IN ITS PHYSICAL AND CHEMICAL BEARINGS. 313 



The method adopted is that described in the British Association Report for 

 •Glasgow, 1876, which Professor Chrjstal, in connection with Clerk Maxwell, em- 

 ployed in the case of metal conductors, namely, that of a Wheatstone's bridge, two of 

 the arms of which are of about equal section and resistance, while the remaining 

 two, though of equal resistance, are of very diflerent sections. If Ohm's law be not 

 true, the point of balance of the bridge varies with the amount of current passing 

 through it. Balance haAing been obtained, the battery-power is altered, and if 

 balance still subsists, the deviation from Ohm's law, if there be any, is such as 

 cannot be detected with apparatus of the sensitivene.ss employed. 



The great difiiculty in experimenting in this way is that the change in the 

 current alters the temperature of the two arms of unequal section very differently, 

 and proportionately also their resistances. While the thick conductor alters little 

 in temperature, the thin one alters considerably. Thus the point of balance is 

 changed, even though Ohm's law be true. To avoid this eflect it is necessary, 

 having balanced with the large current, to immediately pass the smaller current 

 before the temperature can sensibly alter. As this requirement is practically im- 

 possible to satisfactorily fulfil, the method employed is a rapid alternation of a 

 large and small current. If Ohm's law be not true, a balance in this case can only 

 be apparent, and on reversing the direction of one of them, the two currents which 

 before neutralised each other's action now conspire to deflect the galvanometer. 

 The mode of experimenting is to find balance with a large and a small battery 

 alternately in circuit, and rapidly enough for the temperature during the smaller 

 current to be sensibly the same as while the larger one is acting, then to change 

 the direction of one of the currents and to again balance. The distance is observed 

 between the two points of balancing, but if the bridge still balances it is assumed 

 for the pui-poses of calculation that tlie distance is what the galvanometer employed 

 only just detects. From this, fi-om the two currents the section and resistance of 

 the tiiin conductor, h is calculated, where e = rc{\ - he"), as is explained in the above- 

 mentioned Report. 



The liquid, chosen for experimenting with, was a solution of copper sulphate in 

 -water. The equal arms consisted of two glass tubes bent at right angles at each 

 end, A and B, arranged syphon-like, and contained the solution. They dipped 



^-^^f^2|3: 



X 



into a long narrow trough, as shown in the diagram. The arm of large section C, 

 a tube 119 cm. long and 2-38 cm. in internal diameter, reached from the bowl D 

 into which the tube A also dipped, to the bowl E. The small arm was a hole 

 •05.5 cm. in diameter, drilled in the side of the beaker F, '063 cm. in thickness, into 

 which the tube B dipped. 



The battery poles consisted of copper plates, one in the beaker, the other in the 

 bowl D. The contact breaker was a T piece worked by an electromagnet and cell 



