194 Prof. Porter and Dr. Slade on the Fundamental 



although the projections on the exposure axes are different 

 in the two cases unless 7 is equal to unity. Hence, in order 

 that the two plates may match, the straight parts utilized in 

 each characteristic must be the same in length. Generally 

 this is possible in lantern-slides or other transparencies. In 

 the case of paper the straight line is very short ; it is there- 

 fore only the gradations corresponding to a portion of the 

 curve for the negative which will be correctly rendered. 

 The particular part reproduced correctly depends on the 

 printing exposure. Whether any part is correctly rendered 

 depends on the development, for the development factors 

 must be made reciprocals of one another (7 for the one and 



- for the other). These considerations show that if the 

 7 



subject presents a wide range of gradations and it is treated 

 so that these correspond to the straight part of the curve, 

 then development of., the negative should bfi stopped early, 

 so that the straight part utilized is as short as possible (for 

 7 increases with development). The print then should have 

 a high 7 which may be obtained partly by choice of plate or 

 paper and partly by prolonged development so as to make 

 the paper curve as lon^ as possible. For subjects with only 

 a small ransje of gradation these considerations are of less 

 moment. The three curves in fig. 6 all correspond to the 

 same range of gradation in the subject ; with increased 

 development the straight part required increases. The 

 corresponding portions of best positive curves which re- 

 quire to be straight are shown in fig. 7. The lines a, b, c 

 on fig. 6 require to be matched by the corresponding lines 

 a, b, c on fig. 7. Since the straight part on paper curves is 

 short we require the part of the curve on the negative 

 corresponding to extreme range of exposure to be as short 

 as possible, i. e. we require the negative to have a small 

 7 ; consequently the print must have a big 7. 



This conclusion is at variance with that arrived at by 

 Jones, Nutting, and Mees (Photographic Journal, liv. p. 342, 

 1914). They point out that the latitude of plates may be 

 as high as 128 in exposure units, although the contrasts in 

 most subjects is not more ihan 1 to 34 ; they conclude that 

 the negative is easily capable of satisfying what is required 

 of it. But a paper has only a latitude of 5*6 exposure units 

 (•75 in logE 2 ) and they say: "The paper then will only 

 render correctly a range from 1 to 5*6 ; hence it is obvious 

 that in order to render a negative having a range wider than 

 1 to 5*6 in transmission, it is necessary to utilize portions 



