ON ELECTROLYSIS IN ITS PUYSICAL AND CHEMICAL BEARINGS, S29 



brushes in contact with the other face of the commutator ; and, lastly, two brushes 

 in contact with the other two brass rings carry off the direct current to the gal- 

 vanometer. 



The resistance-box employed is one of Elliott's, being a legal ohm box ; the 

 galvanometer is a Thomson reflecting galvanometer, only the light mirror has 

 been replaced by a piece of mirror-glass, which is loaded ^\Tith several lead discs, 

 and to the back'is fixed a small magnet. The magnet has thus a considerable time 

 of swing, and was found to answer much better than the usual light needle. The 

 resistance of the galvanometer was about 268'8 ohms. 



Two different cells were employed to hold the electrolyte. The first was similar 

 to that of Kohlrausch— a strong glass beaker, in which the electrodes were kept in 

 fixed relative positions by two small glass plates — a thermometer was inserted be- 

 tween the two plates to register the temperature. 



Kohlrausch, to determiue the resistance-capacity of his cell, measured its re- 

 sistance when filled with pure mercury ; ' but he does not mention his having taken 

 any precautions as to the difficulty of getting a good contact between his platinum 

 plates and the mercury. Recent observers have stated that there remains a film 

 of air between the platinum and the mercury, and that to get good contact the 

 cell should be fiUed under greatly reduced pressure ; and even then I have found 

 tliat the resistance may vary. I have therefore not employed mercury to deter- 

 mine the resistance capacity of my cell, but a standard solution of copper sulphate,, 

 the absolute values of which I determined by means of my other cell. 



This consists of a glass tube of considerable cross-section which fits into two 

 glass end-pieces; the tube has been ground so as to enter about an inch into 

 the end -pieces. This apparatus was made for me by Messrs. Powell, of Whitefriara 

 Glass Works, and possesses the advantage of being entirely of glass, without any 

 corks, &c. And the electrodes were firmly pressed against the ends of the tube,, 

 and so the resistance of a given column of the solution determined. 



For electrodes, in the first instance, copper plates were platinised. For a long 

 time no satisfactory method of platinising could be found ; that which gave the 

 best results was the electrolysis of a solution of platinum chloride and nitric acid : 

 the deposit thus obtained was fairly adhesive. 



It was finally found best to obtain some platinum electrodes, and for this 

 purpose two platinum shoes were made by Messrs. Johnson and Matthey, into 

 which fitted thick copper plates ; thus a strong electrode of small resistance was 

 obtained without the expense of thick platinum plates ; these plates were slightly 

 platinised as a))ove described. 



In the earlv experiments with this method no satisfactory results could be 

 obtained. When the electrolytic cell was introduced into the fourth arm of the 

 Wheatstone bridge, sometimes the galvanometer would not give any steady 

 deflection, and it was impossible to get a balance. This was similar to Mac- 

 gregor's experience ; he states that by introducing another cell into one of the 

 other arms of the bridge he readily obtained a steady deiiection. In my own 

 case this did not in the slightest diminish the effect ; even with a metallic resist- 

 ance in the fourth arm of the bridge the same value could not be obtained as when 

 the commutator was not working. 



Under these circumstances it became necessary to thoroughly investigate the 

 action of the commutator. At first the commutator had been lubricated with a 

 small quantity of oil ; this was found to be partly the cause of the effects observed,^ 

 and it was consequently thoroughly cleaned from all oil. 



It was also found that the commutator did not run perfectly steadily ; the com- 

 mutator and its stand were therefore firmly fixed in position with wooden props, 

 and driven by a water-engine, which was supplied with water from a tank at the 

 top of the building, and thus a constant and steady speed was obtained. 



Finally, as above described, the light needle of the galvanometer was replaced 

 by one with a longer time of swing. 



' m>te hy Professor Kohlrausch, in a letter to the Editor.—' This is an error. I filled 

 the cell with zinc-sulphate solution, whose resistance had been measured in the form 

 of a cylindrical column.' 



