LAW OF CONSTANT DENSITY RATIOS An important part of the work of Hurter and Driffield was the experimental demonstration of the statement made in the last paragraph, viz., that the ratio of the densities of any two areas of a finished negative is independent of development. A plate was exposed in three sections crosswise to a candle light placed a meter away, the first section being exposed 14 seconds, the next 21 and the third 5 seconds. The plate was then cut into three strips length wise, each strip therefore having on it three different exposures. They were then developed in the same solution for 4, 8, and 12 minutes respectively and the nine densities measured. From these the density ratios, the opacities, and the opacity ratios were calculated and all tabu lated as shown below.
The density ratios of the three exposures for all periods of development are seen to be practically the same. The densities, the opac ities and the opacity ratios all increase with the time of development. Since the exposures are as 1: 2: 4, it will be seen at a glance that the opacity ratios coming nearest to this are those resulting from the 8 minute development. It may be estimated that had strip No. 2 been de veloped about seven minutes the opacity ratios would have been about as 1 : 2 : 4, and this strip would then have been technically perfect as to its tone gradations, the several opacities being proportional to their exposures. Strip No. 1 is seen to be under-developed, while No. 3 is much over-developed.
Having exposed a plate correctly then Messrs. Hurter and Driffield conclude from the above results that it is equally important to develop it for the proper length of time. The authors
explain however how this point of "correct de velopment" must be modified in practice. On page 121 the present author shows that the point of correct opacity or correct contrast may as well be reached by means of reduction or in tensification as by development only so long as the exposure has been correctly given. As has been made clear the straight part of every characteristic curve, whatever the time of de velopment, indicates a direct proportionality between the density and the logarithm of the corresponding exposure. The slope however is seen to increase with time of development, and there is only one slope for which the various opacities are so graded that the corresponding densities are proportional to the logarithms of them; i.e.— There is only one slope, or time of development which in any particular case will result in a technically perfect negative ac cording to the definition. The trigonometric tangent of this particular slope may be called the "development factor" for the plate. The oretically this should be unity and the slope should be Practically it may and must vary from this in any degree as required to pro duce negatives whose opacity will harmonize with the printing paper or other medium used and also to obtain in the picture special effects deviating at will from the full scale of tone gradations which is usually considered desirable.