the machine designed and built at the Her¬ 
cules Experimental Station. It is very im¬ 
portant in loading shot shells that the powder 
be uniformly compressed and confined. Our 
machine accomplishes this by causing the 
wads to lift the same dead-weight load each 
time. 
In a test of this nature it is absolutely 
essential to have the exact number of shot 
pellets as well as have the same weight of 
shot in each shell. This is accomplished by. 
the shot-counting trowel, adjustable to ac¬ 
commodate the variation that may exist from 
bag to bag of any given size shot, as well 
as changing from load to load or from one 
shot size to another. 
The ballistic range at the Hercules Experi¬ 
mental Station is admirably equipped to make 
just such an investigation as this. The shot¬ 
gun range is over 10 feet wide to accommo¬ 
date the spread of shot without possible air 
current effect or eddies caused by the blast 
of the discharge. The steel target plate 
carrying the roll of pattern paper in front 
of it is on a track and can be placed at 
various distances from the gun for measur¬ 
ing velocity and determining the pattern over 
the different ranges. For all 12-gauge work 
in this country the target is placed 40 yards 
7 he operator at the range loads and fires the gun. He removes the crusher cylinder and 
measures the breech pressure in pounds per square inch by a direct-reading dial microme¬ 
ter. The pressure and recoil readings for each shot are reported to the chronograph 
operator by phone. The pierced shells with their pasters ready to be fired 
are shown on the table 
This shows where your pellets go. Of those hitting in the 30-inch circle, 
61.5 per cent are in the inner 20-inch circle and 38.5 per cent are in 
the outer ring . Two 4%-inch circles could go through this pattern with¬ 
out being touched by a single pellet 
*7* LOST 
2 5 9 ‘Vo 
673 
ZT.Z'*!* 
from the muz¬ 
zle. Using the 
one distance as 
standard allows 
us to compare 
our work with 
existing data and 
with the work of 
other companies. 
The target paper 
is rolled up au¬ 
tomatically after 
each shot is 
fired, thus ex¬ 
posing a new 
section of paper 
for the next 
shot. This sys¬ 
tem keeps the 
shots and series 
all in order and 
prevents a mix- 
up at the time 
the patterns are 
counted, and it 
definitely con¬ 
nects each pat¬ 
tern with the 
other ballistic 
data on that 
shot. After a 
10-shot series is 
fired, the roll of 
paper is brought 
to the loading 
room, and the 
tedious job of 
counting the 
shot holes be¬ 
gins. We draw 
a 30-inch circle 
with a beam 
compass in such 
a manner that it will circumscribe the great¬ 
est number of shot holes. Sometimes the 
center has to be moved several times before 
we are satisfied. Then experienced pattern- 
counters do the actual counting by marking 
out each hole with a colored pencil. A rather 
curious error has been found to creep into 
this part of the work, in that the hand is 
apt to work faster, or in some cases slower, 
than the brain. Such a mental tally is quite 
questionable. In order to avoid this, various 
mechanical devices have been tried and dis¬ 
carded. It has been finally found necessary 
to allow each man to count up to 20 and 
then tally this on the paper and proceed to 
count another 20, etc. Although this takes 
more time we feel that the system is more 
accurate and is justifiable. 
Our proof-gun equipment is so arranged 
that the barrels can be interchanged with 
only a momentary delay. This permitted us 
to sandwich this work into the regular rou¬ 
tine testing, since a test gun equipped with 
a compensator could not well be used for 
regular proof work. The gun with its car¬ 
riage weighs 50 pounds, and is so supported 
by four wires that when it swings in its 
recoil arc on a 5-foot radius the axis of the 
barrel is parallel to its repose position. This 
arrangement allows us to measure in inches 
the free recoil of the gun after each shot. 
This measurement can readily be converted 
into units of momentum, velocity of recoil, 
energy of recoil, and used to check approxi¬ 
mately the muzzle velocity of the load. 
Since the energy of recoil and amount of its 
reduction are one of the primary objects of 
this investigation, this equipment provides 
the ideal means required in this work. In 
addition the results could be readily com¬ 
pared with our daily work. 
As we wished to measure the velocity and 
pressure on each shot as well as the pattern 
[4] 
and recoil, in the regular manner, it was nec¬ 
essary to pierce or drill a hole through the 
paper of the shell at a point 1 inch from 
the head in order that the powder gases could 
act upon the pressure piston, which is closely 
fitted in a lapped port through the top of 
the barrel. The other end of the piston bears 
upon a lead cylinder outside of the barrel, 
and compresses the lead cylinder an amount 
proportional to the maximum powder pres¬ 
sure. The drilling or piercing of the shell, 
and the paster put over the hole have a 
detrimental effect upon the patterns. This 
condition, coupled with the tight chamber 
used in all our proof guns, we believe lowers 
our patterns to the extent of 10 or 15 per 
cent. This is unfortunate as it makes our 
patterns run between 60 per cent and 70 per 
cent on ammunition that will give 75 per cent 
to 85 per cent in a normal shotgun. How¬ 
ever, we wish to deal with figures just as 
they are, and do not apply theoretical cor¬ 
rections or make separate pattern determina¬ 
tion in a regular gun. 
In order to eliminate any differences in 
guns due to their chambers, condition of 
shells, etc., we decided to use the same gun 
to compare the compensated and uncom¬ 
pensated firing. Consequently the two loads 
were tested for ten days in an unaltered 
barrel. We then journeyed to Middlefield, 
where the Lyman company had their experts 
cut off the barrel and fit the compensator 
in such a manner that the total length of the 
barrel was practically the same as before. 
After this alteration it was refired with the 
same loads. Unfortunately this arrangement 
did not permit a direct comparison to be 
made on the same day between a compen¬ 
sated and an uncompensated gun. There 
was, however, only an interval of ten work¬ 
ing days between the completion of the 
original firings and the start of the tests 
with the compensated gun. Under the cir¬ 
cumstances this is about as good as could 
own analysis, we are publishing the entire 
set of results. Each line represents a 10- 
shot test in which we show the mean values 
as well as the maximum and minimum indi¬ 
vidual shots. We wish to emphasize that in 
this work every shot has been included and 
that none were eliminated no matter how 
Ijhlvi l; 
I s V !*! 
After the series of 10 shots is completed, the paper is taken from the roll at the target, 
the 30-inch circle is drawn and the counting of the number of pellets is performed by 
experienced operators 
be expected; and since our previous expe¬ 
rience indicates that a ten-day test gives an 
average that can be relied upon to be dupli¬ 
cated, we feel that the two tests can be 
directly compared. 
In order that all figures may be available 
to those who are desirous of making their 
poor the result. Sound mathematical theory 
justifies the discarding of a shot when bal¬ 
listic results are obviously erroneous or when 
something has happened to the equipment 
causing a false record. But in this investi¬ 
gation every shot has been included. Our 
complete figures follow: 
TRAP LOAD—NO COMPENSATOR 
Velocity F. 
S. 
Pressure 
Recoil ins. 
Pattern per cent 
Loading 
Series 
Date 
40 -yd. 
gun to 
target 
Pounds 
per sq. 
inch 
50-lb. gun 
40 yds. 
room 
Ran ue 
No. 
1930 
Time 
Mean 
Max. 
Min. 
Mean 
Max. 
Min. 
Mean 
Max. 
M in. 
Mean 
Max. Min. 
Temp. 
Bare. 
Temp. 
Hum 
89 
7/5 
11:30AM 
852 
870 
837 - 
8,200 
8,900 
7,500 
10.90 
11.0 
10.7 
56.4 
61.8 47.6 
74 
29.30 
70 
64 
91 
7/7 
9 :15AM 
850 
869 
831 
7,500 
8,000 
6,900 
10.74 
10.6 
10.9 
57.0 
69.2 38.7 
74 
28.93 
74 
73 
99 
7/8 
1:20PM 
852 
872 
837 
7,800 
8,600 
7,300 
10.75 
10.9 
10.6 
65.5 
73.0 46.7 
74 
29.14 
75 
52 
5 
7/9 
10:00AM 
852 
882 
832 
7,500 
8,600 
6,800 
10.78 
11.0 
10.6 
62.2 
73.4 44.6 
73 
29.03 
71 
63 
23 
7/14 
9 :10AM 
854 
885 
825 
7,600 
8,400 
6,300 
10.82 
11.0 
10.5 
56.8 
71.4 34.8 
68 
28.80 
70 
73 
47 
7/16 
10:10AM 
861 
876 
847 
7,900 
8,400 
7,100 
10.76 
10.9 
10.6 
55.6 
63.6 47.8 
73 
29.20 
66 
64 
70 
7/21 
10 :40 AM 
845 
877 
821 
7,500 
8,000 
6,400 
10.76 
10.9 
10.6 
59.7 
71.3 33.7 
75 
28.87 
81 
64 
79 
7/22 
10 :00AM 
850 
868 
827 
7,800 
8,600 
7,200 
10.68 
10.8 
10.5 
67.6 
78.1 53.8 
70 
28.92 
80 
60 
84 
7/23 
9:50AM 
857 
873 
835 
8,000 
9,200 
7,100 
10.74 
11.0 
10.6 
57.0 
63.1 34.8 
74 
29.15 
75 
74 
86 
7/23 
2:15AM 
848 
864 
832 
7,600 
8,400 
7,100 
10.68 
10.9 
10.6 
33.2 
75.8 46.7 
78 
29.15 
81 
58 
Mean 
• ••••.• 
. 
852 
874 
832 
77.4 
10.78 
10.9 
10.62 
60.1 
70.1 42.9 
Individual extremes .... 
885 
821 
92 
63 
11.0 
10.5 
78.1 33.7 
TRAP LOAD—COMPENSATED—TUBE 
.690 
82 
8/11 
11:30AM 
855 
869 
835 
7,800 
8,500 
7,300 
9.29 
9.5 
9.0 
6.6 
71.0 61.5 
75 
29.05 
70 
38 
93 
8/12 
11:40AM 
849 
875 
829 
8,200 
9,800 
6,800 
9.37 
9.6 
9.0 
62.5 
69.5 45.3 
72 
29.11 
67 
45 
3 
8/12 
4:40PM 
852 
897 
827 
8,200 
8,700 
7,400 
9.35 
9.6 
8.9 
63.4 
72.4 49.8 
79 
29.25 
70 
37 
4 
8/13 
9:15AM 
841 
851 
814 
7,300 
7,700 
6,400 
9.41 
9.8 
9.2 
64.4 
69.6 56.8 
75 
29.45 
61 
54 
14 
8/13 
4:20PM 
847 
880 
832 
8,100 
9,000 
6,900 
9.35 
9.8 
8.8 
61.6 
71.5 37.8 
80 
29.40 
73 
39 
18 
8/15 
10:00AM 
854 
870 
835 
8,100 
8,600 
7,500 
9.36 
9.5 
9.2 
69.0 
74.1 64.5 
78 
29.15 
68 
78 
21 
8/15 
3:30PM 
850 
878 
829 
8,000 
8,900 
7,400 
9.35 
9.5 
9.2 
68.7 
73.3 60.5 
76 
29.10 
68 
84 
22 
8/16 
9:00AM 
856 
882 
835 
7,900 
8,900 
7,000 
9.42 
9.6 
9.2 
60.4 
65 9 53.8 
79 
29.05 
68 
84 
25 
8/18 
9 :30AM 
847 
873 
828 
7,900 
8,400 
7,300 
9.23 
9.3 
9.0 
60.7 
71.9 52.4 
75 
28.95 
69 
79 
34 
8/18 
2:40PM 
857 
873 
820 
8,300 
9,000 
7,600 
9.37 
9.5 
9.3 
56.0 
65.9 38.2 
80 
29.05 
75 
45 
Mean 
851 
875 
828 
79.8 
9.35 
9.57 
9.08 
63.3 
70.5 52.1 
Individual extremes . . . . 
897 
814 
98 
68 
9.8 
8.9 
74.1 37.8 
Trap Load- 
—3 drams 
E. C., lot 
No. 29. 
lM ounces No. 
7% Winch. (437) 
Shot. 
Winch. Rep. H. 
V. 2\- 
inch L. B. Shells. 
Stand. 
vel. 
over 40 
yard*. 
Maximum Herco Load —41.5 grains Herco, lot No. 11. 1V4 ounces (286) No. 6 West. Shot. 2%-inch West. Field stand, shells 
MAXIMUM LOAD HERCO POWDER—NO COMPENSATOR 
Velocity F. 
S. 
Pressure 
Recoil ins. 
Pattern per 
cent 
Loading 
Series 
No. 
Date 
1930 
Time 
40 -yd. 
Mean 
gun to 
Max. 
target 
Min. 
Pounds 
Mean 
per sq. 
Max. 
inch 
Min. 
Mean 
50 -lb. gun 
Max. 
Min. 
Mean 
40 yds. 
Max. 
Min. 
room 
Temp, Baro. 
Range 
Temp. Hum 
98 
7/8 
1:30PM 
978 
997 
957 
7,600 
8,400 
6,900 
12.59 
12.8 
12.3 
69.2 
74.3 
56.6 
74 
29 14 
75 
52 
4 
7/9 
10 :00AM 
962 
998 
934 
6,900 
8.400 
5,400 
12.31 
12.6 
12.0 
63.9 
68.0 
57.0 
73 
29.03 
71 
63 
22 
7/14 
8:50AM 
967 
982 
923 
6,200 
7,200 
4,500 
12.19 
12.5 
11.6 
64 6 
69.8 
55.6 
68 
28.80 
70 
73 
46 
7/16 
10:00AM 
963 
977 
947 
6,800 
7,300 
5,800 
12.36 
12.7 
12.0 
61.6 
74.0 
41.3 
73 
29 20 
66 
64 
69 
7/21 
10:30AM 
950 
981 
932 
6,700 
7,900 
5,000 
12.30 
12.6 
11.6 
55.6 
72.4 
24.5 
75 
28.87 
81 
64 
78 
7/22 
9:40AM 
971 
992 
950 
7,100 
8,100 
6,300 
12.36 
12.6 
12.1 
73.4 
83.6 
52.5 
70 
28.92 
80 
60 
83 
7/23 
9:30 AM 
950 
983 
893 
5,900 
7,500 
4,100 
12.08 
12.4 
11.4 
61.2 
71.4 
51.4 
74 
29.12 
75 
74 
85 
7/23 
2:00PM 
975 
997 
947 
7,500 
8,700 
6,200 
12.50 
12.8 
12.2 
58.8 
73.8 
25.5 
78 
29.15 
81 
58 
96 
7/24 
3:20 PM 
960 
979 
937 
7,100 
8,200 
6,100 
12.49 
12.9 
12.0 
64.0 
77.6 
37.8 
80 
29.08 
82 
69 
too 
7/25 
8:40AM 
965 
982 
943 
6,800 
7,300 
5,800 
12.30 
12.7 
12.0 
69.6 
75.2 
61.5 
75 
29.10 
75 
73 
Mean 
964 
988 
936 
68.6 
12.34 
12.66 
11.92 
64.2 
74.0 
46.4 
Individual extremes . 
998 
893 
87 
41 
12.9 
11.4 
83.6 
24.5 
[5] 
