1892.] Colour Photometry. 371 



vision, which is monochromatic. Their luminosity curve, M.'s defi- 

 ciency curve, and the normal persistency curve correspond very 

 nearly except in the part affected by the yellow spot. It therefore 

 appears as if P. and Q. had only the sensation which is looked upon as 

 the dominant sensation in the normal eye, and of which M.'s eye is 

 devoid, and that P.'s and M.'s eyes together would make up a normal 

 eye. 



The results are also given of the examination of the vision of a red- 

 blind (H. R.) and of a green-blind person (V. H.). Their "persis- 

 tency" curves, as well as that of P., nearly correspond generally 

 with each other and with that of normal vision. The "absolute 

 intensity " extinction curves of H. R. and Y. H. also do not differ 

 notably from the normal ; but in P.'s case the ordinates are larger, 

 from which it may be inferred that his sense of vision is less acute. 



Assuming blue, green, and red to be the three primary sensations, 

 P.'s and Q.'s luminosity curve or M.'s deficiency curve would represent 

 the first, and V. H.'s deficiency curve the second, but H. R.'s defi- 

 ciency curve would not quite represent the third, as he is not entirely 

 devoid of appreciation of red. 



Luminosity Curve of Spectrum of Low Intensity. The normal per- 

 sistency curve being apparently the same as the luminosity curve 

 of persons with but one colour sensation, experiments were made 

 to determine what would be the luminosity to the normal eye 

 of the different parts of a spectrum of a very low intensity. A 

 spectrum was formed of which the beam from D was equal in inten- 

 sity to 1/132'5 of an amyl acetate lamp at 1 foot. The luminosity 

 curve of this was found to correspond very nearly to that of P., and 

 to the normal persistency curve. By reducing the light in less 

 degrees, luminosity curves were produced corresponding to those of 

 persons more or less red-blind. 



Experiments were then made to ascertain whether this change in 

 the relative luminosities of the different rays would continue to vary 

 with constantly increasing intensity of the light, or whether a point 

 would be reached after which the curve, when it had the same 

 maximum ordinate, would be constant. 



A beam being taken from one point in the spectrum, it was com- 

 pared with the reference (white) beam. Rotating sectors were 

 placed in both beams, and equality of luminosity thereby produced. 

 The aperture of one set of sectors being varied, the alteration of the 

 other which was necessary to re-establish the equality of illumination 

 in each case was noted. Curves were then plotted for several rays, 

 of which the ordinates represent the apertures in the coloured beams 

 and the abscissae those in the white beam when the illumination is 

 equal. Clearly, if these curves ever became straight lines for all 

 parts of the spectrum, the luminosity curve would become constant. 



