352 Mr Stevenson on a new Signal-Light for Ticiilways. 



tions. If we suppose that an effective divergence of only 2° 

 were to be obtained (and this is just one-third of what is ob- 

 tained from Fresnel's lens with the great lamp), I find that 

 the light would spreacHtself along the horizon of the observer's 

 eye between B and C to the distance of about 1000 yards, 

 which, at the speed of 40 miles an hour, would be passed over 

 in about 50 seconds, but at the ordinary railway speed of 25 

 miles an hour, about 80 seconds, or 1^ minute, would be re- 

 quired. Such a flash of light falling upon the polished parts 

 of the engine, and upon the observer's face, would undoubt- 

 edly act as a most effective signal. If, however, it should be 

 thought advisable to increase the duration of the impression 

 by spreading it over a greater length of the line, this effect 

 could be easily produced by a slight alteration of the inclina- 

 tion of the lens, so as to cause the line of railway to cut the 

 refracted beam more obliquely ; but I by no means expect 

 that any such modification would be found necessary in prac- 

 tice. The nearness of the eye to the lens, and the brilliancy 

 of the flash, would, I am inclined to think, more than compen- 

 sate for the shortness of the impression. 



I must add a few words regarding the expense of these sig- 

 nals, which would be made up of the cost of erecting the scaf- 

 fold of carpentry, the price of the lens, and the maintenance 

 of the light. The price of the stage I shall pass over as a mat- 

 ter which may vary according to the circumstances of the si- 

 tuation and the taste of individuals ; but the cost of the great 

 annular lens does not exceed L.40 ; and if a smaller-sized lens, 

 which I think would be found quite sufficient for the purpose, 

 were employed, the expense would not be more than L.10. 

 The annual maintenance would consist of little more than the 

 supply of a gas or an oil burner. The consideration of the 

 •expense, therefore, of maintaining such a system of signals at 

 the necessary intervals on railways, is not for a moment to be 

 set against the most remote risk of the least of all the nume- 

 rous accidents, the records of which fill the public prints. 



^Edinburgh, 2Q'h January 1841. 



