330 REPOET— 1886. 



With these alterations it was found that the galvanometer gave a steady 

 deflection, and the method appeared to be satisfactory. 



This residt having been obtained, the apparatus was tested for induction effects. 

 For this purpose a metallic resistance was introduced into the fourth arm of tlie 

 bridge, and its value determined when the commutator was not working ; the value 

 obtained was — 



R = 19-91 legal ohms. 



With the commutator working at varying rates of speed, the value remained 

 the same, and was not altered by changing the battery power from that of 

 2 Ijeclanche cells to a tray of 10 Daniells. This seemed satisfactorily to show 

 that there was no induction effect, though the sensitiveness of the galvanometer 

 was diminished. 



The diminution in the sensitiveness of the galvanometer when the commutator 

 was working may be seen from the following series of observations. 



The above metallic resistance being in the fourth arm of the bridge, a balance 

 was obtained when the value of the resistance of the third arm was 1991, the value 

 of the other two arms being 1000 and 10 respectively. 



(1) Commutator at rest — 



Galvanometer zero 28'3 cm. \ j n ^- v m 

 PA 1 ■ o A 00.-7 } deflection ol lo mm. 



50 ohms in -ird arm 26v ( 



Zero . . 28-3 



- 100 ohms . 25-1 f 



(2) Commutator working — 



Zero . . 28-5 



-50 ohms . 29 



Zero . . 28-5 



-100 . . 29-4 



i deflection of 32 mm. 



5 mm. 

 9 mm. 



Hence the sensitiveness of the galvanometer is decreased by about f when the com- 

 mutator is working, though the rate of driving did not appear to afl'ect the sensitive- 

 ness. The apparatus was then tested for polarisation efl'ects. For this purpose the 

 two wires leading from the commutator to the battery circuit were connected with 

 the galvanometer and an electrolytic cell with platinum plates containing copper 

 sulphate ; the commutator was started and the circuit completed, the battery con- 

 sisting of 2 Leclanche cells ; the commutator was allowed to run for two hours and 

 a half, the rate of alternation being about 120 per second ; at the end of this" time 

 the circuit was broken, and the platinum plates washed, treated with ho.t nitric 

 acid, and ammonia added after neutralisation with ammonia carbonate. There 

 was not, however, a trace of blue coloration ; it was therefore clear that no 

 copper had been deposited on either plate, and that there was no cumulative 

 polarisation effect. 



This test was again repeated, the platinum plates being replaced by platinum 

 wires, and the Leclanche cells by a tray of lO Daniells ; the commutator was 

 allowed to work for over three hours at the same rate of 120 alternations per 

 second. At the end of this time the wires were treated similarly to the plates, 

 but with the same result. This latter was a more severe test, and would tend to 

 show that the method may be expected to be free from all polarisation effects, and 

 to give satisfactory results. 



So far therefore this method seems perfectly satisfactory, and with the arrange- 

 ments as above described I have not found it in the least necessary to have a 

 second cell in the bridge, as did Macgregor. 



I find that the steadiness of the galvanometer depends — 



(1) On the commutator being perfectly clean. 



(2) On the commutator being driven at a steady speed, though permanent 

 alteration in the rate of speed does not affect it. 



Having thoroughly tested the method for polarisation and induction, I pro- 

 ceeded to examine into the question of contact resistance, a subject which has been 



