But there is now no longer any necessity to resort to such dangerous sources of sound as dynamite. Whistles may be made which yield tones exceeding twenty thousand vibrations per second. The wave-length corresponding to such a pitch is less than an inch. The advantage presented is that the sound is continuous, and it may be made as constant as we please by supplying the whistle from a cylinder full of compressed air, regulating the pressure by means of an appropriate gauge. The disadvantage is that the intensity is but slight, and the pitch is too high to be perceived as sound by most persons unless the ear is closely applied. An artificial indicator must hence be used, whose motion under the disturbances due to sound can be seen at a distance.

In 1857 Prof. John Le Conte discovered that an ordinary naked gas-flame, from a fish-tail or bat-wing burner, becomes an indicator of sound by vibrating in unison with an external source, provided the pressure be such that the flame is just ready to flare. This can be easily shown by blowing a shrill whistle or bowing a tuning-fork of high pitch in the immediate neighborhood of the flame, which at once becomes forked (Fig. 3) into several long, Fig. 3.—Sensitive Batwing Flames. vibrating tongues. The effect soon ceases if the pressure be gradually diminished. This result is due to the disturbance produced by sound-waves on the outflowing jet of gas at the nozzle. The high temperature of flame is therefore not necessary for the production of such co-vibration, but serves to make it more easily manifest. Nine years elapsed after Dr. Le Conte's discovery before the subject was taken up again and independently by Mr. W. F. Barrett, in London, who used small cylindrical jets, which were found to flare under similar conditions, and could be rendered far more sensitive. A "pin-hole lava-tip" may be fitted into the end of a metal tube and connected by means of India-rubber tubing to a cylinder of compressed illuminating gas. In connection with this, also, there should be a water manometer gauge for regulating the pressure of the outflowing gas. If the pin-hole is very smoothly cylindrical, the flame mounts up to the height of nearly eighteen inches (Fig. 4, x), with an apparent thickness scarcely more than that of the little finger, and burning quietly. When the pressure approaches ten inches, as indi-