ELECTRIC-SPARK PHOTOGRAPHS. 169 



spark into a band of light, but I found that while it was easy to show 

 the experiment in a small room, the amount of light was not sufficient 

 to be seen in a ^reat room like this. I must therefore be content to 

 show one or two of the photographs which were taken lately in the 

 physical laboratory at South Kensington by two of the students, Mr. 

 Edser and Mr. Stansfield, whom I now take the opportunity of thank- 

 ing. The next slide shows the drawn-out band of a particular spark 

 made between magnesium terminals by the discharge of a condenser 

 of 2i square feet of window-glass, the S[»ark being ^ inch long. Below 

 the drawn-out band I have drawn a scale of niillionths of a second. 

 If the spark had been instantaneous it would have appeared as a fine 

 vertical line. This line however has been drawn sideways to an extent 

 depending on the duration of the spark. The spark, except at the 

 ends, is extinct in rather less than oneniillionth of a second, but the 

 ends remain alight like two stars, being drawn out in consequence into 

 two lines, which have lasted, as measured by the scale, as long as six 

 or seven millionths of a second. Such a light is, therefore, seen to last 

 when tested with this very powerful instrument so long that it seems 

 absurd to call it instantaneous. It lasts too long for the purpose of 

 bullet photography. In order to get sparks of shorter duration it is 

 necessary to abolish the metal magnesium, in spite of the brilliant 

 l)hotographic effect of the two ends of the spark between knobs of this 

 material; it is well to avoid all easily volatile metals, such as brass, 

 because of the zinc that it contains, and instead to employ beads of 

 copper or of platinum. In the second place, the duration of the spark 

 proper, which in the last case was nearly a millionth of a second, can 

 be reduced by (1) reducing the size of the condenser, but one must not 

 go too far, as the light is reduced also; {2) by replacing any wire 

 through which the discharge may have taken place by broad bands of 

 cojyper as short as possible; this has the further advantage of increas- 

 ing the light; and (3) the light maybe increased without much change 

 being made in the duration by making a second gap in the discharge 

 circuit, the spark in which however must be hidden from the plate, 

 ri. II, fig. 4, shows the trail of the best spark for the purpose of 

 bullet photography that I have obtained up to the present. In this 

 case the surface of the condenser is 1 square foot, and the discharge 

 is taken througli bands of copper about 2 inches broad, and not more 

 than about 4 inches long apiece. Extra good contact is made between 

 these copper bands and the tinfoil surface by long radiating tongues 

 of copper-foil soldered to the end of the copper bands. The knobs are 

 })latinum, but this seems no better than copper. The whole of the light 

 is extinct in less than one-millionth of a second, while the first blaze, 

 which is practically the whole spark (the tail being in comparison of no 

 consequence) does not last so long as a ten-millionth of a second ; in other 

 words, itlastsosliortatinu' thatitbears the same relation to one second 

 that one second bears to four months; or again, a magazineritie bullet, 



