320 



Sir W. de W. Abney. On the Variation in 



Causes of Difference of Results in the Experiments. 



We next have to consider the cause of the difference between the 

 results obtained when the intensity of the light was altered, the time 

 being fixed, and these last two sets of results. I must refer to a paper 

 which appeared in the 'Proceedings' of the Royal Society in 1893, 

 entitled " On a Failure of the Law in Photography," &c, more par- 

 ticularly to the Addendum of July 4th, when it was shown that though 

 the product of time of exposure and intensity of light remained con- 

 stant, yet when the intensity was diminished the photographic action 

 might also be less, and that when the intensity became very small, the 

 diminution was very marked. These observations were further de- 

 veloped in subsequent communications to the Eoyal Photographic 

 Society, in the same year, and it was shown that when the intensity of 

 the same light remained constant during a set of exposures, the time 

 being altered, the gradation of the plate remained the same though the 

 curves occupied very variable positions in relation to the scale of 

 abscissae. Thus if with] a light of a unit intensity exposures were 

 given to different parts of a plate for, say, 1, 2, 4, 8, &c, seconds, and 

 the light was reduced for another set of exposures on the same plate 

 to 1/100 unit, and in order to make time x intensity constant in both 

 cases the exposures were prolonged to 100, 200, 400, 800, &c, seconds, on 

 plotting the densities of the deposit in the manner described above, the 

 two curves would be strictly parallel though by no means coincident. 



In the last two sets of experiments as the relative times of exposure 

 are kept the same, though the intensity is small, the gradation of the 

 different rays would be the same, however much the intensity was 

 increased. On the other hand, where the intensity of the light is 

 small (and when we say intensity, we mean the photographic intensity), 

 the gradation would be steeper than would be the case if the 

 intensity of the light were large.. The photographic intensity of the 

 light used for the red ray is less than 1/500 of the blue: hence on 

 this account alone the " gradation factor " is larger than in the last two 

 sets of experiments. This accounts for the difference between the 

 gradation factors obtained by the two methods, from the red to the 

 blue, and also for the approximate coincidence from the blue to the 

 extreme violet when the photographic intensities of the light used are 

 nearly the same. We see, then, that the gradation factors as found 

 by the last two methods are those which really represent the difference 

 due to the alteration in wave-lengths of the monochromatic light, and 

 that the factors found by the first method are compounded between 

 this alteration and that due to diminished photographic intensity. 



As before remarked, the results of the first method of experiment- 

 ing are those which apply to camera images, for they are formed by 

 different intensities of light, and the exposure is the same for any 

 part. If, then, a plain surface were covered with a graduated scale 



