Force in the Voltaic Arc. 355 



micrometer-screw. Notwithstanding this, the length of the arc 

 was not the same in all the experiments, in consequence of which 

 the deflections varied not inconsiderably. It was necessary, 

 therefore, to make a greater number of experiments and take 

 the mean. To give an idea of the magnitude of the variations, 

 with the mean numbers adduced below are given both the pro- 

 bable errors and the number of observations. 



Experiment 1. The battery consisted of fifty Bunsen's ele- 

 ments with carbon and nitric acid. The trough described in 

 my former investigations was placed in the circuit, and, as 

 before, was partially filled with solution of sulphate of copper. 

 The distance between the copper disks was 5 inches. By this 

 resistance the intensity of the current was materially diminished. 

 The luminous arc was formed between carbon points. When 

 the magnetometer-wires were connected in the above manner 

 with the path of the circuit, on reversing the lever there was 

 always a deflection in the galvanometer, which indicated a cur- 

 rent between the points in the opposite direction to the principal 

 current. As the mean of ten observations, this deflection 

 amounted to 18*7 divisions, with a probable error of 4:1*24. 



When the carbon points were replaced by similar ones of silver 

 and the same experiment repeated, the magnetic needle remained 

 completely at rest. When, therefore, the luminous arc was 

 formed between silver poles, there was no subsequent action 

 after the arc was extinguished. This deportment proves that 

 the cause of the current which produces the deflection is to be 

 sought in the luminous arc itself, because the deflection depends 

 upon the quality of the pole-points. This proves at the same 

 time that there was no simultaneous closing between n, n' and I, l f . 

 The mean deflection, =18*7, adduced above gives of itself no 

 idea of the magnitude of the electromotive force by which the 

 current is produced. The deflection depends not merely on the 

 delicacy of the instrument and the resistance between the poles, 

 but also on the time during which the current acts, and on the 

 electromotive force. Without information on this point, it cannot 

 be determined beforehand whether the electromotive force is 

 most nearly equal, say, to that of 0*1 of an element, or of from 

 ten to fifteen such elements. In order to determine the force 

 in question the following method was pursued. 



Experiment 2. The connexion between the principal current, 

 the magnetometer, and the commutator was altered, as repre- 

 sented in fig. 2. The mercury-cup n was joined with the wire 

 from the positive pole of the battery, and the magnetometer- 

 wire g q connected with the same wire. The conducting-wire 

 from the negative pole of the battery was directly connected with 

 one carbon point. When n and n 1 were connected by the 



2A2 



