196 



Photographic Rendering of Contrast. 



over a certain range of the conjugate curve B'B. Thus we 

 can reproduce correctly over a short range of contrasts in 

 the original by using both the under exposure portion of 

 the negative and of the positive printing process. This 

 portion of the curve is often used in practice, and the two 

 curves BB' and pqr shown in fig. 9 are actually those of an 

 Ilford Empress plate and a piece of Illingworth Slogas 

 paper (Vigorous) respectively. 



Fig. 9. 



>lo g E 2 



Again, for over-exposed negatives, the shape of the con- 

 jugate curve is A'B' (fig. 9). We can only get the over 

 exposure portion of the curve of a print to be of this shape. 

 Moreover, such prints would not represent any portion of 

 the original subject by white paper, and therefore such a 

 print is usually of no pictorial value. 



When this paper had been drafted and while it was being- 

 copied out, our attention was called to the Traill-Tajdor 

 Lecture, given by Mr. Renwick in 1916, in which the 

 same question of matching the negative and positive is 

 discussed. In this lecture Mr. Renwick gives what he calls 

 reciprocal curves for the positive and negative respec- 

 tively ; but he does not describe how they have been ob- 

 tained. In the absence of such a description it is not 

 possible for us to test how far his work overlaps or antici- 

 pates ours. Hut he certainly does not touch the validity 

 of Hurter and Driffield's conclusions. 



We have also found a paper by Lord Rayleigh " On the 



