Nov. 5, 1 8 74 J 



NATURE 



on their combustion for the musical sounds that they emit, must, 

 it appears from Count Sehaffgotch's and Prof. Tyndall's well- 

 known experiments, when placed in certain circumstances of 

 silence and indifference in an open tube, be aided by the voice at 

 a distance to commence their song. The signal-note first raises 

 certain mechanical vibrations in tlie gas-current of the narrow 

 jet, that are necessary in the outset to produce commotions 

 enough of the singing flame to make it able to continue and 

 maintain them. The sensitive sounding-flame of Mr. Geyer bears 

 a similar explanation, for not being regularly adjusted, although 

 very nearly so, to continued sounding, a rustle sufficient to flurry 

 the sensitive wire-gauze flame under the open tube creates in it so 

 many brisk explosions, that the resonance of the sounding-tube 

 is excited, and is at once exalted to a loud note by the rhythmical 

 expansions of the flame ; but witli the cessation of the external 

 sound the maintaining impulse ceases, and the wire-gauze flame 

 whose commotions must be kept up in order to maintain the note 

 immediately becomes as silent as before. 



It is remarkable that the gas-pressure used to obtain Bany's 

 sensitive flame is not sufficient to produce visible sensitiveness in 

 the taper-jet alone ; but if the gauze is raised and lowered over 

 the unlighted jet, a proper position is soon found where the cone 

 of blue flame burning on the gauze above possesses a very high 

 degree of sensibility. The use of smoke-jets instead of flames 

 in this arrangement would perhaps give more positive proofs 

 than may yet have been obtained of the cause of the impressi- 

 bility. It appears, however, scarcely probable that in the short 

 space of a few inches from the aperture the pin-hole current of un- 

 lighted gas can increase its amount of au--mixture so much by the 

 influence of external sounds, that this would account sufficiently 

 for the descent of the conical gauze-flame from the pretty stately 

 eminence of a tall and steadily-burning hill top, to little more 

 than the elevation of a stormy bed of low struggling and bustling 

 flame. The alternative supposition is that the disturbance com- 

 mences in the meshes of the gauze itself, and that it extends 

 upwards from them with such rapidly increasing agitation that a 

 perfect mixture of the gas-current with the surrounding air, and 

 its complete coml>ustion, are thus enabled to take place at very 

 short distances above the gauze. 



I have been led to offer these few reflections on some of the 

 most remarkable examples of sensitive and sounding flames 

 from a wish to distinguish in their action as well as possible 

 between the part which purely mechanical forces, and that whicli 

 the operations of heat and combustion play separalely in their 

 production. The mechanical part of the explanation appears to 

 consist in supposing the sensitive jet, when it is properly ad- 

 justed, as being in a state either bordering upon, or of actually 

 existing undulation. The hissing sound of all air-jets, if listened for 

 attentively enough, is a proof of the reality ol the disturbance ; 

 and such sounds, it has been suggested by Sir G. Airy, indicate 

 disruptions of continuity in the air round the nozzle of the jet, arising, 

 no doubt, from the rapidity with which particles of the quiescent 

 external air are there carried off by friction with the gas-current 

 of the jet. It is hardly possible that vacua so complete (when 

 they exist) should fail to supply the jet with a succession of 

 smoke-rings encircling it and probably travelling up the jet with 

 diflerent speeds according to their magnitude and the depth to 

 which they are involved in the upward current of the gas. If a 

 disposition to regular periodic action exists in the jet (and the 

 smoother its orifice, and the more steady the supply of gas to 

 the jet, the more probable this appears to be), a succession of 

 smoke-rings* of the same size, and of greater or less strength 

 according to the uniform pressure of the gas, may easily be sup- 

 posed to course each other up the flame, and being gradually 

 consumed in ascending, to leave its tall column to the top with 

 sides as smooth and even as a rod of glass. But if the gas- 

 pressure is much increased, a phenomenon like that of companion 

 cyclones observed in rotating storms, perhaps presents itself at 

 the orifice of the jet, each strong smoke-ring as it is formed being 



♦ The word " smoke-rings," as here used occasionally, is not intended to 

 imply tlie presence of smoke in the jet or flame, but to denote by a familiar 

 phrase an annular air-vortex having its rotation round a circular line or nng 

 of lower pressure than that of the surrounding air. Such annular vortices 

 are most easily seen in liquids by drawing a flat blade through them with 

 its broad side in front ; or, indeed, as was lately shown to me by Prof. 

 James Thomson, who supplied me with their explanation, in a cup of tea, 

 by drawing a spoon very gently through it. Only half of the annulus is 

 formed, encircling the edge of the blade or spoon with a curved line of low 

 pressure, round which the liquid spins as in a smoke-ring, and shows a little 

 whirlpool on the surface, one at each point of intersection of the surface 

 with the low-pressure line below it. If an oar-blade is drawn rather rapidly 

 through water, groups of two or three of these ring-vortices following each 

 other in its track can very readily be produced. 



probably followed by a weaker one (a residual offset from the 

 first) travelling after it with less velocity on the outer surface of 

 the flame. The companion rings are probably overtaken and 

 destroyed at a certain height in the flame by the next following 

 strong ring, and the succession being continuous, a puff at a 

 certain height in the flame, where the companion rings collapse, 

 throws it there into a permanent excrescence or confusion. 

 Both rings may be brolcen by the shock, and if of oval forms, as 

 they must probably be in some jets, the two projecting halves of 

 the stronger ring when struck, on springing outwards may thus 

 appear to divide the flame at a certain height above the jet into 

 two pointed tongues forking outwards from each other to a cer- 

 tain width. This form of sensitive flame was shown to be readily 

 obtainable by Prof. Barrett by means of a tapering glass quill- 

 tube jet, the edges of which on two opposite sides are slightly 

 ground or snipped away into a V-shaped notch. Besides the 

 secondary or companion ring, tertiary and higher orders of fol- 

 lowing rings may possibly be formed ; and each strong primary 

 ring may have to run the gauntlet of several weaker antagonists 

 before it at last emerges safely, or else is destroyed itself in 

 its conflicts with them. Tire flame is lowered to a bushy 

 head in the latter case ; but if the primaries outlive their 

 shocks, and if, as might sometime; happen a'so, the se;on- 

 daries alone survive, it seems possible that a sensitive 

 flame with a short continuation of steady flame overtopping 

 the region of tumult and confusion, could in this way be obtained. 

 The hypothesis seems equally applicable to gauze flame;, as 

 nothing can prevent smoke-rings after smoke-rings from rolling 

 up the contiguous sides of parallel jets nearly in contact with 

 eai;h other. Indeed, the difficulty of access of the outer air to 

 the spaces between the jets must favour the production oi vacua 

 round the orifices, and accordingly the occurrence of air-whirls. 

 This is perhaps the reason why wire-gauze flames begin to show 

 sensitive properties at gas-pressures so much lower than those 

 found necessary in the case of a single flame burning at a taper 

 jet. The whole array of jets, it may be, in a wire-gauze flame 

 behaves very nearly alike, and the flame as a body burns, 

 whether noisily or silently, in the same manner, but with greatly 

 increased susceptibility, as a single flame-jet fiom one of the 

 gauze-meshes alone would appear to do. Whatever mechanical 

 distinction may really exist between the mode of action of 

 the common taper jet .and the wire-gauze sensitive flames, it 

 appears, therefore, rather to be one of a higher degree of suscepti- 

 bility at low pressures, than of any more distantly distinct or 

 special kind. Even the mode of operation of external sounds 

 upon them is probably very similar in the two crses, for by rapid 

 vibrations of the external air, such as a hiss or shrill whistle 

 produces, the gas-jet leaving an orifice is shifted bodily to and 

 Iro over its edges, and nothing can more certainly produce partial 

 vacua, and consequently air-whirls round its circumference, 

 than sudden displacements of an air-jet laterally over the sidis 

 of its aperture, even if the tendency to develop them more or 

 less periodically did not exist already in the critical or " sensi- 

 tive " condition of the jet. Axial vibrations, also, or those im- 

 pressed by outer disturbances on the gas- current in the orifice in 

 the direction of its flow, cannot be altogetlrer without effect in 

 producing vacua and air-whirls at its mouth ; and among the 

 multitudes of them thus occurring from the impressed action of 

 external vibrations in all directions, a rhythmical selection is 

 probably made depending on tlie form of the burner and the 

 pressure of the gas. It is difficult to imagine how the partial 

 air-vacuum or aspiration constantly existing round the nozzles 

 of M.ast-apertures can bestow its energy when broken into dis- 

 continuity, rhythmical or otherwise, by a turbulent condition of 

 the jet otherwise than by producing, in the peculiar eddy 

 of its position, ring-shaped vortices encircling the blast ; 

 but it is evident that few jets and nozzles can be fasliioned 

 so smoothly in their inner and outer surfaces and edges 

 that the ring vortices will often be complete ; mere frag- 

 ments of rings are scattered from their sides, which, having 

 no stability, collapse with shocks and puffs that give the roaring 

 and blustering character to the stream. With perfectly smoothed 

 orifices there is probably every gradation according to the pressure 

 of the gas, from full continuity of the partial vacuum or rarefaction 

 round the jet, abating gradually and uniformly upwards to ulti- 

 mate disappearance by friction with the surrounding air, through 

 a condition of gentle undulations of this cone of rarefaction 

 pursuing each other up the stream with slackening strength, and 

 finally losing themselves also by friction as before, to the case of 

 turbulence where the rings of rarefaction are quite intermittent, 

 and sep.arate ring-eddies more or less distinct from each other, 



