﻿910 Mr. R. A. Mallet on the Failure of 



From the foregoing are obtained the following mean 

 values of p : — 



Paget " Half-Tone " plates 0865 + O005 



Wratten " Instantaneous " plates . 0*846 ±0*005 

 Imperial " Eclipse " plates 0*846 ± 0*005 



The probable error, which is certainly not due to errors in 

 measurement of the lengths of the lines, was calculated in 

 the usual manner. 



It is noteworthy that the mean values for the fairly slow 

 Wratten plates and for the very fast Imperial plates agree 

 exactly. Becker and Werner (loc. cit.) have stated that the 

 value of p tends to increase with sensitivity, but this does 

 not appear to be confirmed by the present investigation. 



The question then arose as to whether p was really a 

 constant for any one type of plate for any variations of 

 intensity and duration of exposure within the normal range 

 (great over-exposure, i. e. solarization, being left out of the 

 question), or whether it was dependent on the intensity or on 

 the time of exposure, or on both. Evidently, if p is constant 

 for all conditions of time and intensity, the gradation of two 

 lines on the plate of about equal length, but made under 

 differing conditions, will be the same. 



To ascertain this, two plates were prepared, having on 

 them lines of approximately the same length, but in which the 

 intensity and the duration of exposure were so altered that 

 IT^ = const. Six lines were made on the two plates. The 

 intensity of the incident light was as 1 in the case of three 

 lines to 6*25 in the other three, and the time of exposure 

 adjusted so that IT P = 8 in two cases and =12 in the other 

 four. 



The plates used were Paget " Half-Tone." They were fully 

 developed but not " cut," and the relative blackening at 

 different points of the lines was compared by means of a 

 photometer, involving the use of a photoelectric cell and a 

 string galvanometer which had been designed by Mr. G. M. 

 B. Dobson, to whom I am greatly indebted for the loan of 

 this instrument and for his assistance and advice. Owing to 

 the limited range of the galvanometer it was possible to 

 measure the densities of the lines only over a range of 10 

 to 14 mm. 



The distances from the ends of the lines were plotted 

 against the actual galvanometer readings, which have been 

 used as an arbitrary scale of blackness. It will be seen from 

 the figure that the gradations of the lines are identical 



