SENSITIVE FLAMES. 
163 
kinds of sensitive flames. They all stand, as Professor Tyndall 
puts it, on the brink of a precipice, over which the sound 
pushes them. If the pressure be increased, they shorten and 
roar, and the sound has the same effect : tc The sonorous 
pulses, in fact, furnish the supplement of energy necessary to 
produce the roar and shorten the flame.” 
Some may object to this explanation, on the ground that it 
is inconceivable so small an amount of wave motion as in some 
cases reaches the flame could sensibly affect the pressure of the 
gas. Though this does seem a difficulty, still the amount of 
increased pressure necessary to make a gas-flame roar, when it 
is on the point of doing so, is wonderfully small. I have been 
informed by Mr. Sugg, the gas engineer, that an increase of 
pressure equal to the depression of a column of water through 
the space of only the hundredth of an inch can cause the flaring 
of a fish-tail gas-flame. Through the kindness of the same 
gentleman I have been able to determine the point at which 
a flame becomes sensitive under increased pressure. With a 
steatite burner just admitting No. 19 wire, the flame was sensi- 
tive at pressure of a inches of water, but was at its best, and 
on the point of roaring, at 3^- inches. With a smaller sized 
burner, admitting No. 21 wire, the flame began to be sensitive 
at 5 inches pressure, and was at its best at 6 inches. Further 
increase of pressure, in either case, caused the flame to shrink, 
and take precisely the appearance it has when under the 
influence of sound, as in fig. 3 on the plate.* 
It must, however, be acknowledged, that the explanation just 
given does not clear up the difficulty, for as yet its distinguished 
author has not shown how an external sound can produce an 
effect equivalent to an augmentation of pressure. Moreover, 
it does not, I apprehend, account for the fact that only notes of 
a certain pitch affect one flame, nor why different notes influence 
different flames. 
Further light seems to be thrown on the matter by a critical 
examination of the flame in its various states. To accomplish 
this, we must exalt the sensibility of our vision by adopting 
* The gas here used was cannel, hut the quality of the gas will ddiibt- 
less greatly influence the result. Issuing from the same orifice, and under 
the same pressure, cannel gas gives a far higher flame than the common coal 
gas. Indeed, this difference in the height of the flame is the principle of 
what is known as the jet photometer. This instrument consists merely of a 
jet of gas, issuing from one of these steatite burners, burning in front of a 
graduated scale. A very slight change in the illuminants contained in the 
gas causes a marked alteration in the height of the flame. If the flame 
happens to be near flaring, it is evident that shrill sounds should be avoided 
during the observations of such a photometer : in fact, the burner which at 
the least pressure gave me the most sensitive flame was the very one 
which had been discarded from photometric use on account of the irregu- 
larity of its indications. 
