WILLIAM R. AMBERSON 525 



to connect light intensity with filni density. The dotted portions of 

 the curve represent roughly the two ends of the S, which are not 

 determined by the experimental values obtained in this particular 

 calibration exposure. It will be seen that the curve follows a straight 

 line along a great part of its course, that part, indeed which is deter- 

 mined by the experimental values. This straight line portion of the 

 curve is often spoken of as the normal exposure segment. The upper 

 shoulder of the curve forms the over exposure, the lower heel the 

 under exposure segment. I have endeavored so to control all of the 

 experiments as to hold both moving record and calibration densities 

 to the normal exposure region, although occasionally the densities 

 run out upon the upper shoulder of the S in their higher values. 



In photographic literature the contrast of a plate is called gamma 

 (7). It may be defined as the rate of growth of density with the 

 logarithmic increase of exposure. Graphically it is the slope of the 

 straight line segment of the curve of blackening. The shape of this 

 curve, and especially the slope of the straight line segment, are 

 affected by many factors of which time and temperature of devel- 

 opment, kind of developer employed, and the quality of the previously 

 incident light are the most important. In view of the changes thus 

 effected in this curve it is always highly desirable to impress calibra- 

 tion densities upon each individual record, so that such calibrations 

 may progress through all the vicissitudes of development side by 

 side with the regions of the film affected by the unknown intensities 

 of light which are to be evaluated. The two series are then strictly 

 comparable. The paper by Jones^ (1920) indicates the magnitude 

 of the changes in the value of the gamma produced by developmental 

 variations. 



The quaUty of the incident light affects the slope of the curve of 

 the blackening to some extent, although there appears still to be some 

 dispute among students of the subject as to the magnitude of this 

 effect. Shepherd and Mees (1907) observed differences in the shapes 

 of blackening curves obtained with red and violet light. Some few 

 workers have claimed that gamma does not change at all with different 



® Jones, L. A., J . Franklin Inst., 1920, clxxxix, 469. 



