‘i 125 
by a serew cap, connected by a rubber tube to a reservoir of compressed nit. The rubber tube was normally 
blocked by a metal blade acting as 4 pinch-valve, but this could be tripped at any moment so as to admit 
the compressed air to the gun barrel. By charging the reservoir to different pressures a wide range of 
velocities could be obtained. Four sizes of pistons were used, weighing 22h, 44, 3, and 1} ounces. ‘Lhe 
lightest of these pistons was hollow, as shown in the figure. With the reservoir charged at 75 lbs. per 
square inch above atmospheric pressure the lightest piston could be given a velocity of 150 feet per second, 
corresponding to 33 foot-pounds of energy, und the heaviest piston could be given a velocity of 50 feet per 
second, corresponding to 55 foot-pounds. For velocities from 11 to 16 feet per secoud compressed uir was 
not used, the piston being dropped from different heights inside the barrel. For velocities lower thun 
11 feet per second the piston was dropped from the end of the barrel, which was brought down toa 
suitable height above the anvil. 
When the piston was dropped from the end of the barrel its energy at the momeut of impact was 
calculated from its weight and the height of its fall, but in all other cases the velocity was measured by a 
air of electric contacts connected with a chronograph. Each contact consisted of a very light aluminium 
et just projecting into the line of flight, with u copper brush about 45-inch below it, so that contact was 
closed by the passage of the piston. ‘The two contacts were exactly 1 foot apart, the upper one being about 
1 foot below the eud of the gun barrel aud the lower one } inch above the point of impact. If I’ is the 
time recorded by the chronograph, x the distauce between the contacts, y the distance from the lower 
contact to the point of impact, and g the acceleration due to gravity, the velocity of the piston at the moment 
of impact is— 
aaa 2 
— +-9T 2 gy. 
J (e+ 5! ) + 2 gy 
The chronograph was a high-tensioy spark apparatus. The closure of each contact completed the 
cireuit of u charged condenser (4 microfarads, 200 volts) through the primary of a transformer (a small 
induction coil) causing a spark in the secondary circuit between a fixed discharge point and a drum rotating 
at a known speed. A paper strip on the druin was perforated by the sparks, and the distance between the 
perforations, mensured lengthwise on the paper strip, indicated the time-interval between the closure of the 
two contacts. The construction of the chronograph is shown in Fig. 48 and the circuits in Fig. 49. The 
method adopted for running the chronograph at correct speed was as follows: A soft iron crown wheel, 
with 10 teeth, was fixed on the chronograph shaft so that the teeth rotated past a permanent magnet wound 
with a coil connected to a telephone ; this arrangement produced an audible note corresponding to the speed 
of the drum; io a branch of the same circuit wera a battery, a “button ” microphone, aud another wound 
magnet, the microphoue and magnet being combined with a tuning-fork in the well-known way to 
constitute a retro-active system, producing a sustained pure tone; the speed of the chronograph motor was 
adjusted by a rheostat until slow steady beats were heard in the telephone, not more than one every second ; 
the frequency of the tuning-fork (by comparison with a fork determined at the National Physical 
Laboratory) was 511, so that when the above adjustment had been made it was known that the drum was 
running st 51*1 + O°1 revolutions per second. The circumference of the drum, over the paper strip, 
was 12°60 inches, so that 1 inch corresponded to 74; second. 
The primaries of the transformers had an inductance of about 3 milli-henries ; the calculated time for 
the condenser discharge current to reach its maximum was therefore ;,'55 second, lut the time necessary to 
estadlish «sparking potential was no donbt determined almost entirely by the time constant of the secondary 
cirenit, and must have been far less than this ; experiment in fact showed that the moment of sparking was 
the same, within 10-5 second, whether the capacity in the primary circuit was 2 microfarads or 8 micro- 
farads. Whatever may have been the actual lag of the chronograph action it was ouly the variability that 
mattered. A large number of experiments were made on this point, four condensers being ivchareal 
simultaneously through separate transformers hy closing a single contact, and it was found that the average 
difference between the four individual chronograph perforations and the mean of the four was only about 
10-9 second, or less than ;$5 inch. The variations were due to wandering of the sparks more than to real 
differences in the time of discharge. 
(28) Results. 
The copper cylinders were cut from rods of pure eléctrolytic copper, cold-drawn to diameter *320 inch, 
the length of each piece being from °499 inch to *501 inch. As all measurements were made to the nearest 
ten-thousandths of an inch it was necessary for the ends to be cut perfectly square. ‘The eut pieces were 
anncaled in an electric muffle furnace at 60° C., the furnace being packed with sand to avoid oxidation. 
The coppers were allowed to cool slowly in the furnace, but quenching them in water seems a better plan ; 
quenched coppers are cleauer and more couvenient to measure and their resistauce is almost exactly 
the same. 
The uniformity of these coppers was tested by subjecting 116 of them (cut from 29 ditferent rods) to a 
standard blow of 2°86 foot-pounds ; the average shortening was 27°4 x 10-8 inch, the extreme variation 
heing 26°6 x 10-3 to 27-9 x 10-3; the average difference between the mean and individual results was 
0-2 x 10-3, or less than | per cent., a very remarkable degree of consistency. 
The calibration experiments proved that the shortening of a copper measures the energy of the blow : 
a light and a heavy piston eadowed with equal amounts of energy produce the same shortening. This is 
illustrated in Fig. 50, which shows two sections of the calibration curve with the actual results given by 
pis'ons of different weights. It was observed that there was generally uo appreciable rebound of the piston, 
from which it may be concluded that the who'e of its energy was absorbed by the copper ; the exception 
was in the case of the lightest pistou at very high velocities, when there was a very perceptible rebound, 
but even in this case the energy of rebound was only about 1 per cent. of the incident energy. ‘The relation 
betweeu the shortening A and the impressed energy E is shown in the following table. 
O AS 7498 F 
