of Substances upon Voltaic Action. 449 



Influence of various other circumstances. 



With the object of further improving the apparatus and 

 the method of using it, numerous experiments were made to 

 test the influence of electrolytes of different composition, of 

 amalgamating the electrodes, of including a very feeble 

 voltaic cell or a thermoelectric couple in the circuit to balance 

 the voltaic disturbance ; also of tilting the apparatus, of 

 moving the tubes to and fro during absence of the lead bars 

 and during their presence, of comparatively rough motion of 

 the tubes, of varying the length of the electrodes, of fixing 

 the electrodes in the glass tubes by means of sealing-wax and 

 of shellac, the influence of bubbles of gas or deposits of zinc 

 oxide upon the electrodes, the use of a galvanometer of 6117 

 ohms resistance, &c. 



It was found that a solution of zinc sulphate was better 

 than one of potassium chloride ; that whilst a small proportion 

 of chlorine, by forming hydrochloric acid, prevented bubbles 

 and production of zinc oxide, too large an amount caused 

 bubbles ; that a solution of ammonium nitrate produced 

 deposits of zinc oxide ; that a small proportion of hydrochloric 

 acid caused hydrogen ; that amalgamation appeared to make 

 the electrodes and, consequently, the solution more durable 

 (it probably also diminished voltaic disturbance) ; that even 

 a suitable liquid, if constantly used, required either renewal 

 or a fresh addition of chlorine after a period of about a month 

 or six weeks ; that motion of the tubes disturbed the electric 

 current ; that using the apparatus whilst it was in a tilted 

 position caused no difference ; and that the galvanometer of 

 a greater resistance gave larger deflexions. The defects of 

 " No. 2 " apparatus were that it offered too much resistance, 

 was too slow in yielding the maximum effect, contained too 

 small a stock of electrolyte, and the liquid was too quickly 

 exhausted by continuous use. 



Changes of terrestrial magnetism had very little effect ; 

 the apparatus was placed east, west, north, and south, without 

 causing any perceptible difference. The chief variations 

 w T ere due to voltaic change and temperature, and the effect of 

 the bars appeared to be a constant quantity. 



Apparatus " No. 3." 

 As in Apparatus " No. 2" the total conduction-resistance of 

 the electrolyte was very large, a third one was now made 

 offering very much less. It contained 20 tubes of the 

 annexed form (see fig. 3). Each tube was 4 inches long, 

 with its middle part 2 inches long and 1 inch diameter. 

 The zinc-wire electrodes, all cut from the same piece, were 



