CLASSICAL THEORY OF LIGHT 761 



ends of the ninth standard. The steps are: (1) make M coincide 

 with the end of the metre-rod, and D intersect the virtual image of M; 

 (2) draw D back to intersect with M'; (3) draw MM' back until M 

 coincides with D; (4) draw D back to intersect with M' — and so forth 

 ten times altogether, until for the last time D intersects M', and M' 

 is very near the far end of the metre-rod. The discrepancy is again 

 a fraction of a wave-length. 



The result in which all this labour culminated was: 1,553,163.5 

 wave-lengths of red cadmium light are comprised in the length of the 

 standard metre. 



Such was the process of enumerating the millions of light-waves 

 required to make up the length of that standard chosen for the measure- 

 ments of common life, and so very ill-adapted to magnitudes of the 

 scale of those in light — the distance between two scratches on the 

 bar of platiniridium alloy, conserved in the vault at Breteuil with the 

 care lavished upon a sacred relic. The achievement of Michelson 

 was the bridging of a gap, or let me say a work of translation. Nearly 

 all measurements of wave-lengths to this day are determinations of the 

 ratio of one wave-length to another, as practically all measurements of 

 objects an inch, a metre or a mile long are determinations of the ratios 

 which they bear to metre-sticks. In dealing with tangible objects we 

 use the language of the metre; in dealing with light-waves, we use 

 effectively the language of another scale of measurement. Michelson 

 was the first to make a supremely accurate translation from one to the 

 other of these languages, so making it possible to express a measure- 

 ment of either realm in the scale familiar for the other. Whether in 

 addition he may be said to have replaced a standard essentially 

 impermanent and transitory by one which in the nature of things is 

 everywhere the same and forever immutable, is a question very likely 

 never to be answered. 



