﻿the Measurement of Light. 581 



stated that the range must be such that the entire system 

 of measurement established is satisfactory. But meantime 

 conditions can be prescribed more nearly in which the defi- 

 nition will be satisfactory even if the P.S. are not perfectly 

 matt. 



There are two such conditions to be fulfilled : (1) The 

 lines joining the two P.S. to the observing eye must make 

 equal angles with the normals to those surfaces. (2) The 

 lines joining the illuminating source (or sources) to the P.S. 

 must make equal angles with the normals to those surfaces 

 and with the lines joining the surfaces to the eye. The 

 assumption that these angles are unique implies that the 

 surfaces and sources are "points." Again, it is impossible 

 to prescribe exactly what are points, except by reference to 

 the completed system of measurement that we are about 

 to establish. But the proposition which is important for 

 our purpose is that, if the dimensions of the surfaces and 

 sources are made sufficiently small compared with the dis- 

 tance between them, the definition .applied subject to the 

 conditions just stated will prove satisfactory, even if actual 

 surfaces, not perfectly matt, are employed as P.S. We can 

 therefore give instructions whereby satisfactory measure- 

 ments can be made ; and when once such measurements 

 have been made, the range of permissible variation of the 

 conditions is determined by the agreement of other measure- 

 ments with those made under these standard conditions. 

 Difficulties such as these in describing the precise condi- 

 tions for satisfactory measurement occur in all branches of 

 physics; but they are probably more acute in photometry 

 than in most other branches. 



It will be observed that the absence of perfectly matt 

 surfaces has forced us to introduce at this stage the ideas of 

 " a source " and illumination by a source, which properly 

 belong to a later stage. We may, therefore, describe them 

 rather more closely. A surface is said to be illuminated by 

 a source when its brightness can be changed bv changing 

 the physical condition of some body {e.g., by changing the 

 current through an incandescent lamp) or by interposing 

 suitable media between that body and the surface. The 

 body, the condition of which thus affects the brightness of 

 the surface, is a source illuminating it. 



We can now state the first of the important laws of illumi- 

 nation which are not necessary laws of measurement of that 

 magnitude. If a pair of equally illuminated P.S. are re- 

 placed in position by another pair of P.S., the second pair 

 will also be equally illuminated, even if a member of the 



