August 23, 19 17] 



NATURE 



513 



:urves for these stars were compared with those de- 

 rived from visual obsenations, and it was found that 

 identical geometrical elements gave a satisfactory re- 

 presentation of both the visual and photographic 

 curves. Light-curves with the same epoch of mid- 

 eclipse, however, do not satisfy both sets of obser\-a- 

 tions, the difference amounting to as much as twelve 

 minutes in the case of S. Cancri. The differences in 

 ^'elocity of the visual and photographic rays which 

 would be required to explain the discrepancies in the 

 case of the six stars in question range from —0-9 to 

 -7-5-1 metres per sec, and are so discordant as to 

 furnish no evidence in favour of an explanation based 

 upon differences in velocity of light of different colours. 

 On the contrary, the observations prove the identity 

 of the velocity of light of the different wave-lengths 

 within a few metres per second. The discussion 

 strengthens the view that the t\-pical eclipsing binar>- 

 of large range consists of a small, bright, dense com- 

 ponent of Class A, or thereabouts, and a large, faint 

 component of much lower densit)', of Class G, or 

 redder. It is considered highly probable that the com- 

 ponent of low density represents the earlier stage of 

 evolution. In addition to the already astonishing 

 amount of information which has been derived from 

 the study of eclipsing variables, it is expected that 

 from them it will be possible to determine the rela- 

 tion of colour-index to surface brightness, and thence 

 the linear diameters of all stars of known colour- 

 index and parallax, and the angular diameters of all 

 stars of known spectral type. 



DRVM-FIRE. 



'T'HE following is an abbreviation of a letter by 

 -*■ Mr. G. F. Sleggs which appeared in the Times 

 of Tuesday last. The conclusions arrived at are the 

 result of eighteen months' experience at the Front : — 

 There is a fundamental and peculiar difference be- 

 tween .the sound emitted by a gun and that of an 

 exploding shell. When the gun is fired the sound- 

 wave produced is of a totally different nature from that 

 produced by the burst of a shell. In the former case 

 the impact of the gases leaving the muzzle, as it 

 were, "strikes" the atmosphere in the direction in 

 which the gun is pointed, but the burst from the shell 

 causes a sound-wave of uniform intensity all around, 

 as th3 gases emanating from the high explosive are 

 not confined in any direction, as is the case with the 

 cordite of the gun, the only escape being at the 

 muzzle. Every soldier who has been to the Front 

 knows that if you stand in front of a field gun or 

 naval gun whilst firing even at a considerable distance 

 (several hundred yards), the crack is painfully intense 

 to the ears, and may even cause injury, whereas it 

 is possible to stand close behind the gun with compara- 

 tive impunity. In other words, the sound-wave from a 

 gun is more concentrated along its line of fire than 

 elsewhere. No such difference is obser\'able with a 

 shell, its concussion being equally violent to the ear 

 whether it explodes in front of or behind one. 



The laws of sound say that the intensity of the 

 sound emitted from a body grows less in proportion 

 to the square of the distance of the ear from the 

 source of the sound; in other words, at double the 

 distance the sound is a quarter as great. This, of 

 course, is identical with the laws of light, and applies 

 perfectly to the shell, but not to the gun, in the same 

 way as the ordinary law of the intensity of light 

 will apply to a candle, but not to a searchlight, which 

 concentrates its light along one path instead of dis- 

 tributing it equally all around. Hence we are driven 

 to the conclusion that the wave of sound emitted by 

 a gun is closely analogous to the wave of light emitted 

 NO. 2495, VOL. 99] 



by a searchlight. The intensity of the ray from a 

 searchlight only diminishes gradually, and this analogy 

 is borne out by the peculiar fact, familiar to those 

 who have been in the trenches, that the German 

 machine-guns, or rifle shots, always seem as loud 

 whether the width of "No Man's Land" is seventy- 

 yards or 500 yards. One of the most wonderful and, 

 indeed, majestic of all sound phenomena in connection 

 with artillery is the great "roll" that follows the 

 discharge of a high-velocity gun. To hear this at 

 its best one must visit a part of the front where the 

 contour is rugged, or where the landscape is well 

 wooded, and where houses and other excrescences are 

 abundant, as at Arras. The report of the cannon is 

 followed at once and continuously by a majestic 

 echoing roll that may be compared to a mixture of 

 thunder and the music of a mighty bass orchestra. 

 This rolling sound seems to travel forward as though 

 it were following the flight of the shell, and is, in- 

 deed, mistaken bv some for the actual sound of the 

 shell. 



The real explanation, however, is that it is a series 

 of echoes from the thousands of heterogeneous ex- 

 crescences in the surface of the landscape, each of 

 which sends back its echo to the ear, the whole com- 

 bining to form a continuous trail of sound. Now the 

 fact that this continuous sound travels in the direc- 

 tion of the shell, and hence in the line of fire of the 

 gun, also fits in with the searchlight analogy; as 

 otherv\ise if the sound of the firing gun were not 

 concentrated along its line of fire this chain of echoes 

 would not appear to flow in any definite direction, 

 and thus one of the most grandiose aural phenomena 

 that the ear can receive would not exist. 



The above considerations give rise to a remarkable 

 and surprising fact, which, indeed, arises in theory 

 and is borne out in practice. This is, that at a 

 certain distance and upwards from the firing-line the 

 sound of the German guns will be greater than the 

 sound of our own, because we are in front of the 

 German guns but behind the British, and although 

 the latter are nearer to us, yet the sound of the former 

 will appear louder and sharper because of the peculiar 

 nature of the sound-wave emitted from the muzzle 

 of a gun, the noise being nearly all concentrated in 

 the direction of fire. Thus, when approaching the 

 firing-line before a big attack, the sound of the Ger- 

 man guns often appears to preponderate over our 

 own, giving one the apprehensive impression that the 

 enemv's artiller\" is in superior strength to our own,^ 

 and it is only in coming into the artillery zone that 

 the British superiority is perceived. Another point 

 illustrating this is the origin of the word "drum- 

 fire." This term {trommel-feuer) was first used by 

 the Germans to describe the effect of our massed 

 artillery on an unprecedented scale on the Somme. 

 Now to the British, who were, of course, behind the 

 direction in which their artillery was firing, this term 

 would never have occurred, for to be behind a British 

 bombardment there is but little resemblance to a 

 drummer's tattoo, the whole sound being merged into 

 a dull and heavy roar of guns ; but to the German 

 generals behind their lines every shot from the British 

 guns would stand out as a sharp staccato note, the 

 whole combining to give the impression of the rat-a- 

 tattat of a mighty drum tattoo. 



From these conclusions it will appear that the 

 further one is behind the firing-line the greater is the 

 tendency for the sound of the Gennan guns to pre- 

 ponderate over our own, although the latter mav be 

 in much greater strength, and the probability is that 

 the greater part of the noise of firing audible on our 

 coasts comes from the German artillery and not the 

 British, although the sound of shell bursts may tend 

 I to modifv matters. 



