io8 Management of Light in Illumination. 



prised me not a little; but, after meditating more at- 

 tentively on the subject, I perceived that there was 

 nothing in this result that could not easily be ex- 

 plained. 



Although ground glass appears to us to be opaque, 

 it cannot be so in fact. In the operation of grinding 

 it, its surface, which was smooth and even, is so 

 ploughed and broken up as to present an assemblage 

 of asperities which are invisible to the naked eye on 

 account of their extreme smallness, but which have all 

 their sides smooth and shining, as may be seen by 

 examining them with a microscope. 



Now it is quite evident that a ray of light which 

 arrives at the smooth surface of one of those little 

 asperities must enter the glass with the same facility 

 (at the same angle of incidence) as it would penetrate 

 the surface of the largest sheet of polished glass ; and 

 it is likewise evident that the ray, having passed 

 through the surface, must continue its course in the 

 glass, and pass out of it on the other side, in the same 

 manner in the one case as in the other. 



If a collection of parallel rays of light, forming a 

 small cylindrical bundle, fall perpendicularly on the 

 polished surface of a large sheet of glass, they will pass 

 through the glass in straight lines, and will continue 

 their courses without suffering any change in their 

 direction ; but, if these rays fall on a sheet of ground 

 glass, they will be dispersed, and having passed 

 through it they will diverge in all directions. 



The final direction of each individual ray will de- 

 pend on the refractions it will have experienced in 

 passing into the glass and in passing out of it; and 

 these refractions will depend on the positions of the 



