70 Mr. W. Gannon. 



ence. In ten experiments, with the sulphate solution containing no 

 free acid, nine gave a greater deposit in vacua than in air, one giving- 

 a negative difference. In this experiment (February 18) the deposit 

 was not of the usual colour, and it is possible it was oxidised. Great 

 care was necessary in washing the vacuum deposit, as often the 

 copper was deposited in a loose form at and near the edges ; in the 

 experiments of March 17 and 18 a little of the vacuum deposit was 

 lost in drying between the folds of blotting paper. It was not to 

 be expected that the experiments in this table would show very con- 

 cordant results; as, even presuming that the deposit in vacuo is 

 constant, the inconstancy of deposits in air from neutral solutions 

 would cause an inconstant difference. 



In order to get a constant deposit in air, I next added acid to the 

 air solutions and compared the deposits from these with the deposits 

 from a neutral solution in vacuo. Table II includes experiments 

 conducted in this way : in the experiments of February 24 and 29 

 some of the vacuum deposit adhered to the blotting paper and was 

 lost. I cannot account for the negative result of February 25. The 

 remaining experiments show fairly concordant differences. It will be 

 observed from Tables I and II that, with the exception of those 

 experiments in which some of the deposit was mechanically lost, the 

 percentage difference is greater with acid solutions in air than with 

 neutral solutions, which agrees with Gray's observations that the 

 deposit (in air) in neutral solutions is generally higher than in acid 

 solutions. 



I have brought together the experiments included in Table III, 

 although the results are very inconsistent and puzzling. It will be 

 seen that, if the neutral solutions have been used in previous experi- 

 ments and acid added to them for a fresh experiment, the difference 

 between the weights of the two deposits varies not only in amount 

 but also in sign. An explanation of this may be that electrolysing a 

 neutral solution changes its chemical constitution, and that adding 

 acid afterwards does not remove all basic salts. Accepting this possi- 

 ble explanation, I abandoned for the present using neutral sulphate 

 in either voltameter. In all the remaining experiments described in 

 this paper, both solutions were freshly prepared, not previously used 

 in any experiment, and contained 1 per cent, free sulphuric acid. 

 The results I have obtained in this way are more concordant and 

 very interesting. A deposit obtained from a neutral solution in vacuo 

 shows the same looseness in deposition as in air ; but the addition of 

 a little free acid causes the copper to be deposited in a compact form, 

 which is not liable to be lost in the washing or drying. 



