﻿580 Dr. N. Campbell and Mr. B. P. Dudding on 



Photometric surfaces (P.S.) are members of pairs. A pair 

 of photometric surfaces is defined by the condition that, if 

 their positions are interchanged while everything else remains 

 unaltered, equality of brightness is undisturbed whatever 

 the nature of the illumination. P.S. which form a pair with 

 the same P.S. form a pair with each other. The condition 

 laid down ensures that the surfaces shall be of the same 

 shape, including radius of curvature, of the same colour, of 

 the same reflexion and diffusing coefficients, and of the same 

 intrinsic luminosity. If it is fulfilled, it would be possible to 

 use the pair of P.S. for light-measurement — unless, perhaps, 

 if they were absolutely black — for all or for some colours. 

 But it is convenient to choose surfaces which fulfil as nearly 

 as possible other conditions which ensure that the surfaces 

 are white, matt, and non-luminous. White is here used to 

 include grey. A white surface is one such that light reflected 

 from it is of the same colour as the direct light whatever the 

 nature of that light ; similarity of colour is judged by direct 

 perception. A pair of matt surfaces are such that, if they 

 are equally bright when viewed from one direction, they are 

 equally bright if viewed from any other. No actual surfaces 

 fulfil these conditions perfectly ; none are perfectly white or 

 perfectly matt. But some surfaces fulfil the condition veiy 

 nearly, and these are best adopted for photometry. Further, 

 it is desirable to choose surfaces which are white (in the 

 ordinary sense) rather than grey ; that is to say, if B, sub- 

 stituted for A, is less bright than A, A should be preferred 

 to B. And, lastly, it is desirable to choose surfaces free 

 from intrinsic luminosity, that is, such as can be made to 

 appear perfectly dark by placing suitable screens round 

 them. 



4. The fundamental photometric magnitude is illumina- 

 tion. A pair of perfectly matt P.S. have equal illumination 

 (or are equally illuminated) when they are equally bright. 

 But since, as we have remarked, there are no perfectly matt 

 surfaces, the law of equality, interpreted according to this 

 definition, is not strictly true ; the P.S, may be equally 

 bright if viewed from one position, but unequally bright if 

 viewed from another. If, however, a considerable change 

 in the directions from which the surfaces are viewed does 

 not change their relative brightness, the law of equality will 

 be true for observations made within this range, and the 

 definition so far will be satisfactory. The question, remains, 

 however, what is a considerable range for this purpose ; it 

 can only be answered perfectly definitely when the further 

 definitions of measurement have been added ; it can then be 



