ON ILLUMINATION, 



27 



k possible that a ray may be obliged to pass and repass in the 

 glass from one side of the shade to the other several times be- 

 fore it bfe able to escape into the room. 



If the glass were perfectly transparent, the hght would be 

 little, or perhaps not at all diminished by these repeated re- 

 flections and passages; but we know that even the finest glass 

 is verv far from being perfectly transparent. 



When crape, gauze, or other substances are employed to Shades of crape, 



make Shades for the purpose of masking the flame of a lamp, P"2e, &c will 

 ■ _ i ^ ». . .^ ' intercept light 



the loss of light will be more or less considerable m proportion in proportion 



to the greater or less degree of transparency of the solid parts '^^^^^^-'^'^P'^^^^'y- 

 of the substaiice employed. But without engaging in the very 

 delicate enquiry concerning the degree of transparence of the 

 molecules or small solid particles of the substances to be em- 

 ployed in making shades, we may determine, by simple experi- 

 ments, with ease and even precision, what are the substances to 

 be preferred for that purpose. We have onl)' to procure shades 

 of the same form and size, made of the different substances to 

 be examined, and to compare them, by pairs, by means of two 

 argand lamps, made to burn with the same degree of vivacity, 

 and of a' simple photometer, which can be constructed at a 

 very small expence. 



The photometer which I used in my experiments on the The loss may 

 comparative quantities of light produced in the combustion of >*e ascertained 

 •wax, tallow, and different kinds of oil, and of the same kind jjfg^J^^ ^' 

 of oil burned in an argand lamp and in a common lamp,* 

 would serve perfectly well for the experiments in question ; 

 but as that instrument is somewhat complicated, I shall pro- 

 pose another more simple, which I have employed since with 

 success. Its construction is as follows : — 



In the middle of the upper surface of a wooden cube of 8 or by a simpler 

 inches in diameter, composed of boards, covered with black '"^^''""'S"'^^''* 

 j)aper, there is fixed vertically a small board, 4 inches in 

 breadth, 6 inches in height, and half an inch in thickness, 

 covered on one side with white paper. In the middle of this 

 white side there is traced, with pen and ink, a slender black 

 line, from the top to the bottom, which divides the surface 

 into two equal parts. 



• See Phil. Transact, for 17 94, and my Philosopiilcal Papers, 

 vol. I. page 270. 



E 2 Before 



