174 THE ROYAL SOCIETY OF CANADA 



ducibility of these cells by independent workers, as well as the con- 

 stancy of given types. 



Further details concerning the construction and early records of 

 our cells may be found in the paper by Bronson and Shaw loc. cit. 



t 

 §3. Suggested Rules With Reference to the Construction 



AND Use of Modified Standard Cells for 



Precision Work. 



If an accuracy of at least two or three parts in a hundred thousand 

 is to be expected from a small batch of standard cells, it will be found 

 useful to follow the regulations suggested below : 



1. In Construction (a) prepare the mercurous sulphate paste 

 either chemically or electrolytically with great care, dry it and cover 

 with a special solution to prevent hydrolysis; and make it slightly 

 basic. If possible, however, obtain at least this one ingredient from a 

 "standards" laboratory,^ as slightly impure or acidic mercurous 

 sulphate leads to subsequent variations in e.m.f. which are both 

 progressive and variable in their development, (b) One distillation of 

 mercury, bought as c.p., will be found to be sufficient. It should be 

 noted that if the mercury is slightly impure, the resultant variation 

 in e.m.f. is usually of a constant character, and after comparison the 

 cell is as useful as others, (c) The cadmium sulphate should be the 

 best quality commercial product c.p., rendered slightly basic with 

 cadmium oxide. The exceptionally tedious recrystallizations specified 

 for its purification do not appear to be necessary, (d) Cadmium is 

 prepared in sticks of exceptional purity by many firms, and the com- 

 mercial c.p. product has been found quite satisfactory. 



2. In choosing standards for reference a batch of at least five cells 

 should be chosen and of these there should be representatives of at 

 least three separate constructions, each from difïerent samples of 

 ingredients. 



3. Any cell the e.m.f. of which difïers from the mean of the re- 

 mainder by more than 100 microvolts should be rejected and replaced 

 by a satisfactory substitute. It has been found that a cell which 

 changes its e.m.f. some time after its construction to an amount 

 exceeding 100 microvolts, usually becomes considerably worse. Unless 

 they have had that value from the date of their construction, appar- 



^ The distribution of ready-made Hg2S04, purified carefully by one of the best 

 recognized methods, dried with alcohol and sent out in hermetically sealed tubes to 

 various laboratories, is a very simple method of avoiding the chief source of trouble 

 in standard cell construction. 



