192 Management of Light in Illumination. 



screens ought to be preferred, and how much more one 

 of them softens the direct rays from the flame than 

 the other. 



It will likewise be possible to determine in any given 

 case, by means of experiments which are by no means 

 difficult to perform, the relative quantities of the light 

 which proceeding in straight lines from the flame pass 

 directly through the sides of the screen into the room, 

 and of that which coming from the surface of the 

 screen in all directions illuminates the surrounding 

 bodies. 



These experiments, and many others of a similar 

 nature which it is not necessary for me to describe 

 particularly, will no doubt occur to those who may 

 engage in these interesting investigations; and it is 

 highly probable that many useful improvements will 

 be derived from these researches. 



CHAPTER V. 



Of the relative Quantities of Light furnished by 

 Lamps of different Sizes, with the Combustion of 

 a given Quantity of Oil. — Of the relative Cost of 

 Light furnished by Lamps and Candles U7ider dif- 

 ferent Circumstances. — The Light furnished by a 

 good Lamp does not cost half as m-uch as that fur- 

 nished by a Tallow Candle. 



HAVING lately found, from the results of a course 

 of experiments on the light manifested in the 

 combustion of inflammable substances, of which an 



