102 LIGHT SCIENCE FOR LEISURE HOURS. 



although the considerations on which the matter de- 

 pends are exceedingly simple, yet the case is by no 

 means the first in which exceedingly simple considera- 

 tions have been lost sight of by students of science. 

 Suppose we have a brightly-white globe encased sym- 

 metrically within a globe of some imperfectly trans- 

 parent substance as green glass. Now, if the white 

 globe is an inch in diameter and the green glass globe 

 a yard in diameter, the brightness of the white globe 

 will be more or less impaired according to the trans- 

 parency of the glass ; but it will not be much more 

 impaired at the edge of the inner globe's disc than 

 near the middle. For clearly, when we look at the 

 middle, we look through a foot and a half of glass 

 (wanting only half an inch), and when we look at the 

 edge of the inner globe's disc, we also look through a 

 foot and a half of glass (wanting only a small fraction 

 of an inch). Neither the half inch in the one case, 

 nor the small fraction of an inch in the other, can 

 make any appreciable difference, so that the enclosing 

 globe of glass cuts off as much light when we look at 

 the centre of the inner globe's disc as when we look at 

 the edge. But now suppose that the enclosing globe 

 forms a mere shell around the inner one. Suppose, for 

 instance, that the inner globe is a yard in diameter, 

 and the shell of glass only half an inch thick. Then 

 in this case, as in the former, the brightness of the 

 inner globe will be more or less impaired according to 

 the transparency of the glass ; but it will no longer be 

 affected equally whether we look at the middle or at 



