and it is not difficult to see that the impulse is of a kind to encourage the motion, and to produce sound.

Fig. 23.-Trevelyan's instrument. Cause of vibratory movements.

Sondhauss's Experiment.

When a bulb about of an inch in diameter is blown at the end of a narrow tube 5 or 6 inches in length, a sound is sometimes heard proceeding from the heated glass. It was proved by Sondhauss that vibration of the glass is no essential part of the phenomenon. An explanation of the production of sound has been given by Lord Rayleigh, which will be fully detailed in Chapter VIII.

Singing Flames. Although the series of small explosions by which the combustion of gas is marked have contributed some brilliant experiments to physicists, they can hardly as yet be said to be a practical source of musical sound. The flame of hydrogen has long been known as a means of originating a note; and of late a second form of the experiment has utilized the ordinary flame of coal-gas as a very delicate consonator and test for sounds produced extraneously in its neighbourhood. The former are termed singing, the latter sensitive, flames. The sensitive flames may best be considered elsewhere; but a short account of the singing-flame is required to complete the series of sound-producers.

It is easy, whenever a jet of hydrogen is inflamed, or even when coal-gas issues from a burner with some force, to hear an unmusical hissing or roaring accompanying the process. Even in this case, the noise often puts on a more or less definite form of vibration, and an impure note of coarse quality makes itself manifest. But if the jet be surrounded by a resonating tube, this has the power of reducing the irregular vibrations to a greater uniformity, and of selecting those which synchronise with its own vibration-period. note in this case is often very pure and powerful. general type of the process may be studied in the ordinary Bunsen burner. In this very convenient laboratory appliance, ordinary coal-gas is allowed to issue by a small orifice into a larger tube, perforated at its lower extremity with several

The

The

holes admitting more or less atmosphéric air, the admission of which is regulated by a slide. An explosive mixture of gas and air is thus made, which is prevented from communicating with the source of gas by the cooling effect of the surrounding tube, just as occurs in the wire-gauze envelope of a Davy safety-lamp. If only a small amount of

Fig. 24.-Philosophical lamp or chemical harmonicon.

air is admitted the mixture burns with a semi-luminous flame and silently. But as the proportion of air is allowed to increase by opening the slider, the flame loses its luminosity and at length begins to roar. The combustion gradually becomes discontinuous, and is indeed composed of a series of short successive explosions. At length the mixture becomes too explosive even for this, it lights throughout its whole

length with emission of a single report, and carries the combustion down the tube to the small gas-jet itself. If the Bunsen burner, freely supplied with air, be introduced into a tall vertical tube, the roaring is toned to consonance and gives a pure note instead of an irregular roar. It is not always easy with quietly burning coal-gas in a tube to obtain any sound at all. But if the flame be reduced in size, and moved up and down the consonating tube, it may be observed to become tremulous at certain spots, and if left there, suddenly, by increase of the pulsations, bursts into sound. This action may be determined by the method first named as a means of exciting musical oscillation in a pipe, namely, by striking a slight blow with the palm of the hand on the upper orifice of the tube. The wave thus sent downwards is instantly taken up by the flame, and the note starts out, sometimes with such vehemence as to extinguish it altogether. it can also be set going by the voice. Many of the older works on chemistry and physics give the simple experiment with a hydrogen flame and a glass tube. Dr. Higgins names it in 1777; but Wheatstone was the first who attempted successfully to produce a definite scale of notes by this method. His instrument is now preserved in the Museum of King's College, and was recently shown at the Loan Exhibition at South Kensington. A series of glass tubes, with metal sliders for the purpose of tuning, are arranged in a row like the pipes of an organ. Within each of these is a fine conical tube pierced above with a capillary orifice. The lower end slides air-tight in a second tube connected with a supply of gas. In front is a short keyboard like that of a piano. Each key, on being depressed, lifts the small gas-jet from its position at the bottom of the tube to about the junction of the lower and two upper thirds, which, being a sensitive point, immediately originates the fundamental note of the tube.

The Pyrophone.-M. Kastner has endeavoured to utilize this principle in a musical instrument, but on a slightly different system. He says, "If two flames of a certain size be introduced into a tube made of glass, and if they be so disposed that they reach the third part of the tube's height, measured from the bottom, the flames will vibrate in unison. The phenomenon continues as long as the flames remain apart, but as soon as they are united, the sound ceases. By means of finger-keys the flames are united and separated so that a inelody can be played. There is some uncertainty about the instrument, depending on the pressure of gas; so

that although announced for performance in Paris, it has not hitherto been used.

A curious accidental source of sound appears to have been several times discovered. It possesses no practical importance, but affords an apt illustration of the theory of harmonics. If a piece of the ordinary vulcanite tubing, such as is used for conveying gas, and which is prevented from collapsing by a spiral wire coiled round its internal surface, be cut to a length of about 18 inches, and gently blown into, a soft musical note of feeble but reedy quality is produced. On pressing the force of wind, it rises successively to higher notes, which will be found on examination to follow roughly the order of the common chord to the foundation tone. It is obvious that the wire coiled inside the tube produces a series of equidistant obstacles, competent to throw the air into regular vibration; the rapidity of which vibration, and the consequent pitch of the note produced, vary with the speed at which the air is blown into the calibre of the tube.

The only remaining source of sound is the human voice; but this is of so much importance that it will be considered separately in a later chapter.

CHAPTER II.

MODES OF PROPAGATION OF SOUND. VELOCITY. WAVE-MOTION. REFLECTION. REFRACTION.

The Propagation of Sound appears to take place to some extent through all bodies, but in very different amounts and with varying degrees of velocity. This factor has been found to vary directly as the square root of the bodies' elasticity, and inversely as the root of its density. The formula

therefore serves for all forms of matter. Solids, however, being liable to many kinds of strain, and fluids, whether liquids or gases, to only one, we may have different values of E, and different velocities of transmission for the same solid. In a perfectly free solid this value of E is identical with Young's modulus. The great majority of solids however transmit sound more rapidly in one direction than in others. In solids, moreover, the thermal correction, to be spoken of presently, is very small, as it is also in fluids, whereas in air it is large.

By the Earth.-There is distinct evidence of its transmission through the solid mass of the earth itself for long distances. Humboldt says, "At Caracas, in the plains of Calabozo, and on the borders of Rio Apure, one of the affluents of the Oronoko; that is to say, over an extent of 130,000 square kilometres, one hears a frightful report, without experiencing any shock, at the moment when a torrent of lava flows from the volcano St. Vincent, situated in the Antilles, at a distance of 1,200 kilometres. At the time of the great eruption of Cotopaxi in 1744, the subterranean reports were heard at Honda, on the borders of Magdalena; yet the distance between these two points is 810 kilometres ;