THE ROYAL ARTILLERY INSTITUTION. 
175 
Boulenge. The chronograph remained at the proof butts from July to 
November, 1865, when it was taken down to Plumstead Marshes and placed 
in a splinter-proof, where it remained about a fortnight. Its powers to 
withstand damp and dust were w r ell tested in this manner. Eighteen shots 
were fired in all. Two of them were fired by mistake w r hen the cylinder 
was stationary. One shot carried away a screen, another cut the conducting 
wire at the second screen, &c. I am able, however, to give a good account 
of eleven out of the eighteen shots fired on November 23, November 29, 
and December 12, 1865. 
Before proceeding to give an account of the experimental trial of the 
chronograph, it is necessary to explain at some length the mode of dealing 
with the clock and screen records, because the absence of any provision to 
secure uniformity of rotation of the cylinder has appeared to many a great 
defect. The fact that the variation of velocity of a pendulum, falling from 
rest through a considerable angle, used in so many instruments, is 
enormously greater, seems to have been entirely overlooked. The reason 
for not attempting to secure uniform rotation in the present case was this: 
the thing has been often attempted, and, so far as I know, never satisfac¬ 
torily accomplished, when such small intervals of time as the one-hundredth 
or one-thousandth of a second were of importance. It is sufficient to refer 
to the chronograph constructed by M. Breguet for Colonel Konstantinoff. 
M. Gloesener 1 appears to be satisfied with the approach to uniformity of 
rotation made by that instrument, and gives four experiments in support of 
his favourable opinion. It will be sufficient to copy the second experiment, 
wdiich gives the times occupied in making five revolutions in succession:— 
2"*05 
l "-95 
l //# 90 
l "-80 
1"*60 
1"*90 
1"*90 
1"*80 
2"‘05 
K> 
6 
o 
l "-90 
i~* 
do 
o 
1"*90 
o 
<x> 
i —! 
l "-60 
2' /, 30 
2 //, 00 
l "-80 
1"*80 
l "-90 
l "-80 
l "-60 
1"*80 
2"*05 
2"-00 
l "-80 
1"*90 
l /,# 90 
o 
GO 
1 — 1 
H-* 
cd 
O 
giving a mean time of revolution of 0"*362. 
As this result w r as very far from satisfactory to me, and I had no hope 
of succeeding so well as M. Breguet, it was quite plain that the diffi¬ 
culty of making the cylinder revolve uniformly must be avoided in the 
construction of a good chronograph for experiments on gunnery. As already 
explained, in most, if not in all instruments used in Observatories for record¬ 
ing observations by the aid of electricity, a marker connected with the clock 
records the seconds. If this v r ere done, it is plain that the length of 
spiral described in successive seconds would give the rate at which the 
cylinder w r as moving. The introduction of a heavy fly-wheel would prevent 
sudden variations of velocity. The axis was placed in a vertical position, 
in order to reduce the friction as much as possible, and to avoid errors that 
might be caused by one side of the fly-wheel being heavier than the other. 
All unnecessary resistances, all disturbing causes must be removed. It may 
1 TraitS general des applications do Telectricite, &c., I. 374, 
