172 ELECTRIC-SPARK PHOTOGRAPHS. 



poiuts,* in Older that its position every time it is fired may be the same. 

 The bnllet then traverses precisely the same course, so that wires 

 placed in the line between holes in two cards made by one shot will be 

 hit by the next. The two wires which the bullet joins as it passes by 

 are set uj) in the box seen in the middle of the figure with the lid 

 propped up so as to show the interior. The photographic plate is on 

 the left-hand side, and the spark when made is just within the rectan- 

 gular prolongation on the right-hand side. Paper tubes with paper 

 ends are placed on each side of the box to allow the bullet to enter and 

 leave, and yet not permit any daylight to fall tlirectly on the plate. 

 AH is bhick inside, and so the small amount of light which does enter 

 the box through these holes is not diffused in any harmful manner. 

 The large box at the back is a case 5 feet long, filled with bran, which 

 stops the l)ullets gently without marking them. 



The little condenser is just below the rectangular i»rolongation of 

 the i)hotogiaphic box, the large condenser is the vertical square sheet 

 seen just to the right. The electrical machine used to charge the con- 

 densers is seen on the table. It is a very beautiful lli-plate Wimshurst 

 machine, made by Mr. Wirashurst and presented to the Physical Labo- 

 rator3^ This machine not only works with certainty but is so regular 

 in its working that no electrometric ap))aratns is necessary. All that 

 has to be done is to count the number of turns of the handle which 

 are required to produce the sparks at E and E' when the gap at li is 

 not joined, and to count the. number which are sufficient to pioduce a 

 spark at E when the gap at B is suddenly closed. Then if the rifie is 

 fired after any number of turns between these, but by preference 

 nearer the larger than the smaller number, the potential will be right, 

 the spark E, inside the box, and the spark E' which is in sight outside 

 the box, will be let oft", and if a plate is exposed a photograph will be 

 taken. If by chance the E' spark is not seen then there is no occasion 

 to waste the plate; another bullet may be fired after resetting the 

 wires mid the result will be as good as if one shot had not failed. 

 When all is in order a failure of this kind is very rare. I alsoarianged 

 a tube in the side of the box with a pocket telescope fixed in it and 

 focused on the wires. If a piece of white card or paj)er is placed in 

 the line of vision and so as to l)e illuminated by a spark let off as 

 above described, but preferably much nearer the card, the bullet will 

 be seen by anyone looking through the telescope, I took this down 

 however at once, as the photograph showed more than could ever be 



* Six independent points of support are required ior a geometrical claui]). In this 

 case a V suiijiort near the muzzle supplied two, a V support near the brciccli two 

 more points. The rifle was pressed forward until a projection under the muzzle rested 

 against the front \ , thus allowing freedom of recoil, but otherwise preventing all 

 uncertainty of position except that due to rotation in the 's, which is made im- 

 possible by the sixth ])oint — that is, the lower end of the stock resting sideways 

 against a leather-covered wooden bracket lastened to the same table to which the 

 V's were attached. J 



