ON THE ELECTKOLYTIC METHODS OF QUANTITATIVE ANALYSIS. 303 



Although the above conditions may be considered as practically en- 

 suring a good deposit, it must not be supposed that they need always be 

 strictly adhered to. Considerable latitude is allowable in certain respects. 

 The influence which each factor exerts may next be considered. 



If ammonium sulphate is omitted or added in insufficient quantity, the 

 resulting deposit is dark and rough. The amount stated may be largely 

 exceeded, however, without influencing the character of the deposit. The 

 only effect which a considerable excess of ammonium sulphate seems to 

 possess is possibly to make it slightly more difficult to completely remove 

 the last traces of metal ; that it does so is not quite certain. 



The proportion of ammonia is more important. When nickel salts are 

 electrolysed with a platinum or other non-soluble anode, nickelic hydroxide 

 is formed on the anode. This does not take place if a sufficient quantity 

 of free ammonia is present, hence the necessity for employing the amount 

 stated. If much less is taken, a brownish-black deposit forms on the 

 anode, and causes loss of nickel if not removed. If noticed, it must be 

 dissolved by interrupting the current and adding more ammonia. If the 

 experiment has been left unattended for the usual period of four hours, 

 the formation of nickelic hydroxide involves prolonging the electrolysis for 

 a further period till all the redissolved nickel is deposited. It is therefore 

 much more satisfactory always to add sufficient ammonia at the be<^inninw. 

 The above-mentioned quantity may be exceeded by several grammes with- 

 out very marked effect, but a greatly increased quantity retards deposi- 

 tion, and may cause an unequal deposit. 



The quantity of ammonia which is necessary apparently depends chiefly 

 on the strength of current employed, not so much on the amount of nickel 

 present. During the electrolysis ammonia is neutralised at the anode, but 

 with a weak current the partially neutralised liquid is replaced by fresh 

 solution sufficiently rapidly to keep the liquid at the anode alkaline, even 

 although there may not be a great amount of free ammonia present. 



The current density may vary somewhat, but should not be ^reatly 

 increased. If that is done, the deposit suffers in quality, beino- much 

 rougher ; the rate of deposition is increased, principally in the earlier 

 stages. With a weaker current than that stated good deposits are still 

 obtained, but the operation is more prolonged. When the quantity of 

 metal to be deposited is not very great, a current density as low as 0"3 

 ampere may be employed, but it is advisable to decrease the quantity of 

 ammonia (though not of ammonium sulphate) in that case. With a 

 current density of 0-15 ampere it is not possible to get good deposits, 

 even with small quantities of metal, if the usual quantity of ammonia is 

 employed (i.e. 5 grme.) ; the metal forms irregular patches, and much 

 remains in solution even after prolonged electrolysis. By diminishino- the 

 amount of ammonia to 1 grme., or even less, these drawbacks are practically 

 removed. One grme. of free ammonia is sufficient to prevent formation of 

 nickelic hydroxide with the last-mentioned current, even when the quantity 

 of nickel is considerable, say 0-2 grme. For most of the experiments 

 in the latter part of this investigation, using 0-15 — 0'2 gi'me. of nickel 

 a current density of 0-6 — Ow ampere was adopted as generally the most 

 suitable. 



There is no special advantage in conducting nickel determinations at 

 temperatures higher than the ordinary. Deposition is then somewhat 

 more rapid, but sometimes less regular, and the quality of the deposit is 

 apt to suffer. As such determinations require more frequent attention 



