178 ELECTRIC-SPARK PHOTOGRAPHS. 



Tlie otlier point to wbicli I ^*oakl refer is the dark line between the 

 nose of tUe bullet and tlie wire placed to receive it. This is the feeble 

 spark duo to the discharge of the small condenser which clearly must 

 have been on the point of going off of its own accord. The feeble spark 

 precedes — or is to all intents and purposes simultaneous with, it cannot 

 follow — the main spark which makes the photograph. The feeble si)ark 

 heated the air, and the light from the main spark coming through this 

 line of heated air was dispersed, leaving a clear black shadow on the 

 plate. One spark casts a shadow of the other. Now it is evident that 

 if the spark at the nose of the bullet had followed instead of having 

 preceded the main spark by even so much as athree-huudred-millionth 

 of a second, the time that light took to travel from one to the other, 

 it would not have been able to cast a shadow. We have the means of 

 telling, therefore, which of two sparks actually took place first, or per- 

 haps the order of several, even though the difference of time is so 

 minute. Perhaps this method might be of some use in researches now 

 attracting so much interest in connection with the propagation of elec- 

 trical waves. 



On returning to the non-reflection of the air wave in the upper part 

 of the figure, we have here, 1 imagine, optical evidence of what goes on 

 in a whispering gallery. The sound is i^robably not reflected at all, 

 but runs round almost on the surface of the wall from one part to 

 another. 



We are now in a position to see how the reflection or nou-reflectiou 

 of air waves produced by a passing bullet, when they meet with some 

 solid body, may i)roduce a i)ractical result which might be of impor- 

 tance in some cases. Suppose a bullet to be i^assing near and parallel 

 to a wall. Then if the velocity of the bullet and its distance from the 

 wall are such that the head wave meets the wall at an angle at which 

 it can be reflected, especially, as in the case of Plate vii, if the 

 reflected ray can only return into the path of the bullet after it has 

 gone, then no influence whatever can be exerted upon the bullet by its 

 proximity to the wall. If however the head wave would if nndisturbed 

 meet the wall at such an angle that it could not be reflected, as for 

 instance in Plate viii, when the head wave can be reflected by 

 neither of the walls between which the bullet is passing, obviously the 

 wave will become stronger and the resistance which it ofters will, I 

 imagine, become greater, and if in this case the upper ])late be removed 

 this extra resistance will be one sided and must tend to deflect the 

 bullet. This is quite distinct from the well-known eft'ect of a bayonet 

 upon tlie path of a bullet; when a bayonet is fixed the rush of powder 

 gases between the bullet and the bayonet is quite sufficient to account 

 for the deflection which every practiced marksman all(jws for. 



I have devised a method by which a problem of some difficulty, about 

 which authorities are, I believe, by no means in accord, may be solved 

 with a fair degree of certainty. The problem is this, to find what pro- 



