174 Management of Light in Illumination. 



This is evident from the result of the experiment 

 just mentioned. 



If we examine the situation of this lighted candle 

 burning in the centre of the screen of ground glass, we 

 shall find that a considerable portion of its light escapes 

 through the open ends of this screen, and is entirely 

 lost; half of it passing upwards into the clouds, and 

 the other half passing downwards into the earth, so that 

 no part of it is usefully employed in illuminating the 

 surrounding objects. 



If now the screen, which is only 6 inches in length, 

 be removed, and another screen of ground glass, of the 

 same diameter and 12 inches in length, be put in its 

 place, the whole of the surface of this taller glass cylin- 

 der will become luminous, and the intensity of the illu- 

 mination of the surrounding objects will of course be 

 increased. A considerable portion of the light which 

 escaped through the open ends of the short cylinder 

 will be arrested by the additional length of the taller 

 cylinder, and will be usefully employed in rendering its 

 surface luminous. 



Hence we learn that the tall paper lanterns of the 

 Chinese, and those which are frequently to be met with 

 in the streets of London, in the wheelbarrows of 

 orange-women, may possibly be useful for other pur- 

 poses than merely for preventing the flame of the 

 candle from being disturbed by the wind. 



I am persuaded that they often increase the bright- 

 ness of the illumination of surrounding objects ; and 

 that they would also do so is most certain, if they were 

 properly constructed and arranged for obtaining that 

 end. They always render a service equally important, 

 or even more so ; for they defend the eye from the 



