Mr Stevenson on a nem Signal- Light for Baihmys. 351 



is brought into play ; and should this expectation be realized, 

 the duration of the effect of* the light, which depends on the 

 range, might be increased beyond what I have ventured to 

 state. 



Referring to the above sketch, I would propose that the 

 lens at L should be elevated 24 feet above the rails R R, or 

 about 15 feet above the level of the engineman's eyes ; and 

 that the point where the centre of the beam would intersect 

 the horizon, AC, of his vision at E, should be about 700 feet 

 from the lens. The impulse of the light would be most ad- 

 vantageously received at some point as near the lens as is 

 consistent with a full effect from a flame placed in its princi- 

 pal focus. A more remote observer would receive the rays 

 diluted by distance ; while a nearer approach of the eye to the 

 lens would render it necessary to adopt an ex-focal arrange- 

 ment, so as to cause convergence of the rays. By the latter 

 arrangement their divergence would be decreased, and the 

 space covered by the light would be lessened not only in pro- 

 portion to the decrease of divergence, but also to that of the 

 cosine of the beam's inclination to the horizon. Both these 

 circumstances would therefore combine to curtail the duration 

 of the impression on the eye. 



It may naturally be expected that I should say something 

 regarding the duration of the impulse of the light on the eye ; 

 and upon this topic I shall, in absence of actual experiment, 

 content myself with stating briefly the result of my calcula- 



