Units and Useful Range in Photo-electric Colorimetry

Abstract

WE value the remarks of Drs. Edisbury, Taylor and Gridgeman¹ on a new unit in photo-electric colorimetry². Opinions do not seem at variance that measurements should be made at approximately the same optical density^{1, 2}, that—provided the Beer-Lambert law be valid—the relation of the proposed unit W⁵⁰_{1 cm.} to the usual E^1%_{1 cm.} can be expressed by : and that E (1%, 1 cm.) is certainly less cumbersome to print than E^1%_{1 cm.} Title and text of our letter², however, should have made it clear that we had photo-electric colorimeters (absorptiometers) in mind. Though it was shown by Twyman and Lothian³ that 0·4343 is the best density for this type of instrument, it was also stated that "... we may use densities between 0·2 and 0·8 without any considerable loss of accuracy". This range covers easily the proposed half-value and adjacent auxiliary readings used for interpolating W⁵⁰_{1 cm,.} An even wider range (0·2–1·2) is recommended elsewhere⁴ for an instrument other than ours. The statement¹ that E (1%, 1 cm.) is merely a standard expression that does not carry the implication that one must measure the optical density of a 1 per cent solution in a 1 cm. cell may probably be correct, but not necessarily so, and does not seem to be endorsed unanimously ; for Guild⁵ inferred regarding accuracy in spectrophotometry that "... our terminology should be kept in as close correspondence as possible with the physical magnitudes actually handled in our experiments", strongly deprecating the use of derived magnitudes. Though the proposed half-value is inversely related to quantity, this need not impair its applicability ; for a similar magnitude, the half-value layer is a standard expression in X-ray technique⁶.