Abstract
The accuracy obtainable from Vierordt's method largely depends upon the establishment of numerical coefficients, which apply to the analyst's own instrument at the time of analysis. For the special case of a two component mixture, the method has been formulated in terms of extinction ratios, which can be determined from solutions of unknown concentration—so facilitating the task of obtaining up-to-date numerical coefficients. A further development of the same formulation has led to a simple theoretical criterion, which enables the analyst to avoid pairs of wavelengths and concentration ratios that are unsatisfactory for precision. This criterion has been tested in an unfavourable application of the method. A procedure for choosing the best pair of wavelengths is described in detail. The value of this new formulation and criterion of precision is discussed with special reference to a general procedure for the assay of injections which contain absorbing bacteriostatics. Any assay for a two component mixture depends upon some difference between the two components and in this respect Vierordt's method appears to be less exacting than the average quantitative separation process. For this reason, it seems to have greater scope than the latter.