_ =. 7 _—. 
a ee lls a 
44 CHELTENHAM COLLEGE NATURAL HISTORY SOCIETY, 
out fine at the end. Here they had a flame produced in this way 
about eighteen inches long. The gas being turned on just to the 
point of roaring, the flame was in a condition to roar at a very slight 
noise. The flame was only sensitive to noises of a certain kind, 
such as the noise produced by hissing, by rattling a bunch of keys, 
and generally by noises of a high pitch, Holding a watch near the 
flame its ticking was made visible to the whole room. A musical 
snuff box caused the flame to bob ina curious manner, the bobbing 
being most pronounced when the highest notes were sounded. 
An interesting observation had been made by Mr. Barrett who 
found that some flames were sensitive to notes beyond the limits of 
human audibility. It was well known that there were notes so high 
that no human ear could detect them. The limit of audibility of 
high notes was different for different individuals. This might 
readibly be shewn by means of a little instrument called a Galton’s 
whistle. It was a whistle which, by turning a screw might have its 
pitch altered and made so high as to pass beyond the limits of 
audibility. When the pitch was so high as to be quite beyond the 
limits of audibility, the flame was still found to respond to it, and 
thus to prove the fact that there was a sound, or, to speak more 
correctly, there were vibrations passing through the air though no 
human ear could detect them. In seeking for an explanation for 
this sensitiveness to sound, it was to be observed that the flame 
was brought just to the point of roaring without actually roaring. 
Thus it was in a very unstable condition and ready to roar at the 
slightest disturbance of the air. Professor Tyndall had found that 
the seat of sensitiveness was at the mouth of the burner. Lord 
Rayleigh had found that when the flame was placed in a stationary 
wave, formed by simultaneous presence of a sound wave, and its 
the air is least, but its alternating variation of pressure greatest, 
This shews that the disturbance in the flame is caused by the to 
and fro motion of the air across the mouth of the burner. This to 
and fro motion probably causes the gas to become mixed to a 
small extent with air as soon as it issues from the mouth of the 
burner and thus to become more inflammable, and consequently to 
become burnt sooner than it would when undisturbed, and there- 
fore not to rise so high before being burnt. The lowering of the 
