Primary standard of light

Light is not a physical quantity, but a physiological effect, that of certain wave-lengths of radiation, and therefore can not be expressed in absolute physical units; it must be measured by comparison with an arbitrarily chosen standard of physiological effect. As a result thereof, even with the best existing primary standard of light, the amylacetate lamp, the difficulties of reproduction, and maintenance of its constancy, are such as to involve errors very far beyond those considered permissible in physical measurements. A radical increase in the accuracy of reproduction and maintenance of a primary standard of light appears possible only by relating the standard of light in such manner to physical quantities, that it can be determined by energy measurements.