BECENT EXPERIMENTS ON FOG-SIGNALS. 279 



short tube was placed at one end of the rectangle, and 

 a ' sensitive flame,' l /, at some distance beyond the 

 other end. When the reed sounded, the flame in front 

 of it was violently agitated, and roared boisterously. 

 Turning on the gas, and lighting it as it issued from 

 the slits, the air above the flames became so hetero- 

 geneous that the sensitive flame was instantly stilled, 

 rising from a height of 6 inches to a height of 18 

 inches. Here we had the acoustic opacity of the air 

 in front of the South Foreland strikingly imitated. 2 

 Turning off the gas, and removing the sensitive flame 

 to /', some distance behind the reed, it burned there 

 tranquilly, though the reed was sounding. Again 



FIG. 9. 



lighting the gas as it issued from the brass tubes, 

 the sound reflected from the heterogeneous air threw 



1 Fully described in my Lectures on Sound,' 3rd edition, p. 227. 

 c Lectures on Sound, 3rd ed., p. 268. 



10 



