ON IT.LUMJ NATION. ^'j 



the quantity of light that thej' distribute in a room. The first 

 point seems susceptible of bein.q; pretty well determined by the 

 simple inspection; but if we would employ more precision, 

 we may muke use of the following method : — 



■ Having placed) at ecjual distances, two lamps burning with insfTO.-tlors for 

 the same degree of vivacity, in front of the photoineter, we si^^^'''<'*i""'ngtfce 

 maslc them with the two suudcs (made ot different mateniilsi, ,,.,■, ■,:,„• ,,^ 

 which are to be compared, and we place between each shade ^•■'.' ^y^ig >:iie 

 and the photometer, at the distance of about an inch iron-: I'.ie' ' ' 

 shade, a disk of thick [)asteboard, perforated in the middle bv 

 a circular hole, one inch in diameter. The diaineter of tbi^ 

 disk must be sufficiently large to mask completely the sh.ade, 

 and the center of the circular hole must be in a straight line 

 drawn from the centre of the flame to the center of the vertical 

 face of the photometer. 



We see that, in this state of things, it is almost sok-Iy tiic 

 direct rays coming from the flames of the lamps in straight 

 lines through the substance of the shades that fall upon the 

 photometer; and that, on measuring the relative intensities of 

 these rays on each side, by means of the shadows and the dis- 

 tances, we can determine not only which of the shades fulHls 

 best its principal object, that of preserving the eyes, but 

 likewise the proportion in which one of the flames is 

 more softened than the other. VVe may also determine, by ^p^j ^_f„„„g^ 

 easy experiments and calculations, the proportion v^hich exists, raHy ii'.umi- 

 in any given case, between the quantity of light that passes ''^ '"^' 

 directly from the flame in straight lines through the sides of the 

 shade, and that which is dispersed by the shade and appears to 

 issue from the shade itself. It would be too long to describe 

 in this place all these experiments, and several others that 

 might be made with the photometer, to corr.plete the researches 

 liccessary for the improvement of lamps ; but these details are. 

 the less necessary, as the experiments will not fail to present 

 tlieaiselves to those who shull have m.ade some progress in 

 these enquiries. 



I shall finish my observa ions upon shades ibr lamps by Cone msng the 

 some remarks on the size that may be given to them. And ^'"- °' =*''^*i^' 

 fii'st, it is_evident that the diameter of a shade should be .greater 

 : 5 proportion as the flame which it is intended to mask is 

 ,^ 'uter and more bright ; for if a iba,le be small, the light 



wd^iic-b 



