140 Mr. Ritchie on the permeability of transparent 
tin-plate. This may easily be done by applying electric 
cement to the circumference of the aperture, and then laying 
the film of glass on it when properly melted. Let a delicate 
air thermometer be placed opposite the disc of glass on one 
side of the plate, and a heated iron ball opposite to the bulb 
on the other. Let a current of cold air be made to play 
constantly against the disc of glass, which will keep it uni- 
formly below the temperature of the ambient air. Things 
being thus arranged, the following facts will be observed. 
ist. When the temperature of the ball is low, ho sensible 
effect is produced on the thermometer. 
2nd. When the temperature of the ball is high, though 
still invisible in the dark, the effect on the instrument is very 
considerable, even if the ball should be placed at a greater 
distance than formerly. 
Here we have two sources of heat, which, on account of 
the change of distance, would produce equal effects in the 
naked bulb of the thermometer ; but, by the intervention of 
a cold screen, the effect of the former is almost annihilated, 
whereas the effect of the other is still very considerable. This 
difference cannot possibly result from the difference of tem- 
perature in the screen, which is kept as near as possible at 
the same temperature by the influence of the current of cold 
air. We are therefore unavoidably led to the following con- 
clusion ; viz. That the progress of the heat was, in the first 
experiment, arrested by the screen ; whereas in the other, a 
portion of it freely radiated through the screen, and found its 
way directly to the bulb of the thermometer. 
