A detailed procedure for the maintenance of a standard of voltage to within one or two parts in a million is described.In so far as these investigations have proceeded, neutral saturated cells have been found to be superior to acid cells as independent standards for a period of many years, though the latter are preferable for purposes of ordinary precision or shorter periods. The recommended code of procedure is briefly as follows: a number of cells should be constructed according to standard specifications with the new requirements of uniformity of container and speed of preparation, and observations made upon them every few days for a period of three months. The differences in electromotive force (at constant temperature) should be determined between each cell in the group and any one of them chosen arbitrarily as reference cell, and certain new selection and rejection rules applied. In accordance with these rules a cell should be rejected: (a) if its deviation from the mean electromotive force of the group has increased or decreased by 10 microvolts or more during the preceding two weeks; or (b) if it differs in electromotive force from the mean of the group by more than 10 +d microvolts where d is the mean deviation of the cells of the group. If d exceeds 20 microvolts the entire group should be considered untrustworthy. The selected cells should be observed for three additional months, the rejection rules again applied and if a specified proportion survive elimination, the initial reference mean of the laboratory may be established.At intervals of several months additional groups of cells, neutral and acid, should be constructed and exchanges made with laboratories possessing cells of known characteristics. Analysis of the resulting observations determines: (a) the constants in the aging* equation for the reference batch, and (b) the difference between the initial reference mean of the laboratory and the estimated value of the international reference mean.Examples of the analysis of cell observations are given, illustrating the establishment of the initial reference mean, the recapture of this value when the aging coefficients are known, and the preliminary determination of the aging equation for a given group of cells. The use of the aging equation is found to be the essential feature in the attainment of increased precision.A summary of data on standard cells is included.
A new determination of the electromotive force of Weston and Clark standard cells by an absolute electrodynamometer