If conditions of its use are strictly specified, especially with regard to the spectral composition of the illumination and the nature of the medium in which it is suspended (which should be water rather than beer), the E.B.C. formazin unit possesses a definite and reproducible absolute turbidity. Because of the complexity of the light-scattering phenomenon, it is in principle impossible for conventional nephelometers to measure true absolute turbidity, and they must therefore be calibrated empirically; formazin is a very convenient material to use for this. Nevertheless, nephelometers of varying designs will still distort this unit, each in its own way; hence, different nephelometers calibrated against formazin will, when used to estimate beer turbidity, yield results which are liable to be inconsistent not only with the true turbidity but also with each other. The extreme extent of the discrepancies which may thus arise is probably large. Hence, while formazin may fairly be used to establish an arbitrary scale for individual nephelometers, and to check their calibrations periodically, it must not be thought that a universal quasiabsolute scale has thereby been achieved. This is not a shortcoming of the formazin standard itself, but of practical nephelometry: any haze standard will be subject to similar limitations.