THE VELOCITY OF LIGHT. 225 



It depends very much on the condition of the eye and the 

 dilatation of the pupil. All experiments go to show that 

 it is very difficult to say exactly when a feeble light is 

 visible and when it is not visible. Now the method em- 

 ployed by Mr. Young avoids this completely. Instead of 

 having one collimator at the distant station, there are two, 

 one fixed about half a mile further away from the observer 

 than the other. As we increase the velocity of rotation of 

 the toothed wheel, the reflection from the more distant col- 

 limator will of course be the first eclipsed. At first two lights 

 are seen on the edge of the wheel, namely, the two lights 

 reflected from the two collimators. As you increase the 

 velocity of rotation you first see the light reflected from 

 the distant collimator eclipsed, while the other one is still 

 tolerably bright. You increase the velocity, and the 

 distant collimator comes again into view before the second 

 one has been eclipsed, and you compare the intensity of 

 those two lights, and may measure the velocity which is 

 necessary in order to render those two lights exactly equal, 

 the one having passed its eclipse and the other not yet 

 having reached its eclipse. When you have got that exact 

 velocity, and have found the two lights exactly equal, you 

 know that is the velocity required to produce an eclipse, 

 if you had one collimator fixed half-way between the two. 

 This method will probably give very much greater accuracy 

 than any other which has been employed, but no absolute 

 measurements have yet been obtained. 



There is only one step further which we can hope to go 

 in these measurements of the velocity of light. There is one 

 point which has been always necessary in the undulatory 

 theory of light, and upon which experiments are wanted 

 to render it still more certain, and that is the question of 

 dispersion. This was an objection which was raised against 

 the undulatory theory long ago. Granting these experi- 

 ments and the results to which they lead, the velocity of 

 light in water and in air ought to have a definite ratio, be- 

 cause the elasticity of the ether in water has a definite ratio 

 to the elasticity of the ether in the free air ; consequently 

 the elasticity of light in water of all colours should be the 

 same ; but we know perfectly well from experiment that 

 a beam of white light falling on the surface of a dense 

 medium is divided into its different colours, the dark rays 



VOL. II. Q 



