[neelin] the sensitiveness OF THE EYE TO LIGHT 227 



considered as equal to unity in all parts. The initial conditions to give 

 spectra of such intensities for each observation were the inclination of 

 the principal planes of the niçois to each other at angles of 75° 45'; 

 80° 45' and 85° 45' respectively. The results are given in Tables IV, 

 V and VI and shown graphically in Curves IV, V and VI, fig, 2, 



Tables of data. 



In Tables I to VI X indicates the wave length of light observed. 

 The column headed "Scale Pteadings" gives the observed reading on 

 the attached scale, after each rotation. With each new intensity the 

 zero of the scale was observed and taken into account to find the angle 

 between the principle sections of the niçois. These values will be found 

 in the column headed (a). Since ^ denotes the initial angle between 

 the principal sections, in the column under Cos-/3 will be found the inten- 

 sity of the spectrum considered in each case. In the columns under 

 Sin-a Table I and under Cos "/9 — Cos "a Table III will be found the 

 amount by wliich tlie initial intensity was diminished to effect the 

 minimum noticeable change of sensation. In Tables II and IV to VI, 

 under Cos- a — Cos-^ are given the amounts by which the original 

 intensity (as shown under Cos-/3) had to be increased to effect a change 

 of sensation just perceptible. Under "Sensibility" are the values of 

 the reciprocals of the amounts of minimum perceptil)le change in in- 

 tensity. In order that the various curves might be plotted in one figure 

 these reciprocals, which are the ordinates for the different curves, 

 have been multiplied by arbitrary constants as indicated in the figure 

 and then jjlotted to the scale shown. 



2. A Spectrum of Equal Luminosity in all parts. 



When measuiing the luminosity of the spectrum Allen * took obser- 

 vations upon nineteen portions of it, detei mining in each case the angle 

 which the pi'incipal plane of the polariser should make with the horizon- 

 tal, i.e., with the plane of the analyser in order that each portion as 

 viewed in the eye-piece should have equal luminosity. Therefore since 

 the apparatus used in the present investigation was also used by Allen 

 in approximately the same arrangement with the same gas jet, gas 

 pressure and slit v/idth, it was considered that, by observing these 

 same portions and setting the plane of the polariser at the corresponding 

 angle for each, equality of luminosity in the parts observed would be 

 obtained. This would it was thought be equivalent to observing a 

 spectrum of equal luminosity in all parts. The next problem considered, 

 then, was that of determining the sonsibilitv of the eve to slight 



*Phil. Mag., 1911. Vol. 21., p 604. 



