(A,n), while the
probability of their concurrence is
<^(AOxc^(A,) . . . xc^(A,„)*
* If ^1 be the probability of any event, p.^ that of any second event, p.^ that of
any third event, pm that of any m*'' event, the probability of their concurrence is
= PiXPi-^Pti • • • xp"*
This proposition is easily illustrated. If we throw in the air a penny, the
chance of throwing heads twice in succession is J x i = ^, or 3 to 1 against it. The
chance of throwing heads three times in succession is ^ x i x 4 = |, or 7 to 1 against
it, and so on.
ARTILLERY PRACTICE
which is by Equation (3)
A^ -7l2[(a;-,i,r:!+(,l;-»l,)2+ . . . («-W^j)2]
. (6)
. (7)
by virtue of Equation (5).
Now it is evident from an examination of (6) that
(A,,) = ^.e-'"""^- . . . (9)
* This will perhaps be more easily seen if we put the right-hand member of (6)
in the form
h'- 1___ _
Vl' e/'2(A,2-t-A22 . . : +Am-)
ARTILLERY PRACTICE 7
The most probable value of h is that which makes (9) a maximum.
Hence, differentiating with respect to h, and equating to zero, we have
1 - 2/<-V, =
/' = ,7^ (10)
We have above shown that that hypothesis as to x is most prob-
able for which the sum of the squares of the errors is the least pos-
sible. Now in the simple case we are discussing, the direction to
select such a result that the sum of the squares of the errors may be
a minimum, is the same as if we were told to take the arithmetical
mean. This may easily be verified. Suppose that four shots from a
9-pr. gun gave the results 950, 975, 1025, 1050, the mean of which
ranges is 1000 yards. It will be found that the sum of the squares
of the errors upon the hypothesis that 1000 yards is the true range
is less than it would be upon any other hypothesis whatever.
The law of the arithmetical mean* may also be easily deduced
from Equation (7) by the differential calculua The sum of the
squares of the errors is to be a minimum. Hence, from (5)
{x-n^- + (x -n.^- . . . +{x-ii,a)- = mmimura.
Differentiating and equating to zero,
X — ?i^ -H a- -«.,.. . +a; — n^n =
rt, + »o + • • • + Wift
or .1- = -i =
m
We may from geometrical considerations obtain a graphic view
of the law represented by Equation (3). If we take the values of A
as abscissae, and the corresponding values of 0(A) as rectangular
ordinates, we shall be enabled to trace the curve of probable error.
The general form of the curve of probabilities is shown in Fig. 1.
In this figure the abscissae, such as OA, OC, represent the errors, and
* It may be proper to advert here to a common misapplication of the principle
of the arithmetical mean, and one which sometimes leads to serious errors. Suppose
m different series of equally good observations gave m different values of the mean
range a^, a., . . . dm . Suppose further that «i is the result of p^ rounds, Oo the
result of fo rounds, am the result of jOm rounds, or as it may otherwise be expressed,
suppose tt] to have the weight p^, a^ the weight p.,, a„i the weight j)^ — the most
probable value of the range is not, as is frequently assumed,
o, + a, + . . . + a,n
.V = -^ =
m
but is ^■ = ^^^ + "^P^- •• +""'P-
P1+P2 . ■ . +P,r,
8 ARTILLERY PRACTICE
the corresponding ordinates AB, CD the comparative probabilities of
these errors.
The curve is symmetrical on each side of the axis of y, thus show-
ing that positive and negative errors are equally probable, while the
rapid decrease of the ordinate AB, as we recede from Oy, shows that
large errors are not so probable as small ones, and that when we take
a very large error, as OE, the corresponding ordinate, and, therefore,
the probability of the occurrence of that error, becomes altogether
insensible.
The curve, although close to the axis of x at E, and always
approaching nearer and nearer, only touches it at infinity, the axis of
X being an asymptote to the curve.
As said above, the curve of probabilities is always of the nature
represented in Fig. 1. The particular form it assumes is, however,
dependent upon tlie value of the constant h. An examination of
Equation (10) shows that the smaller is the mean error, that is, the
more accurate are the observations, the larger is h, while Equation (3)
shows that as li is large the more rapidly will 0(A) or the length of
the ordinate AB decrease, which is the same thing as to say that as h
increases so are large errors unlikely to occur.
On account of the ratio between the increase of exactness and
the increase of the magnitude of h, h has received the name of the
" measure of precision."
Fig. 2 is a representation of a curve in which the value of h is
large as compared with its value in Fig. 1.
If we suppose A to vary continuously, the probability of any
particular error A is infinitesimal. We have pointed out that the
probability of the occurrence of an error between A and A + dA
is (p{A)dA, or substituting for 0(A) from Equation (3)
V TT
■.f/A
(11)
ARTILLERY PRACTICE
while the probability that an error lies between OA = a and OC = c
is given by the definite integral
s]
d
e .clA
(12)
This integral represents the area ACDB, Figs. 1 and 2, and it
will also express the proportion of errors which should occur
F^^.2
between A = a and A = c, the whole area between the axis of x
and the curve, or the whole number of errors, being unity. Putting
JiA = t. Equation (12) becomes
1 .ch
dt
(13)
If we seek the probability of an error between -\-a and ~~a,
that is, if we seek the area AA'^/B'B, Figs. 1 and 2, Equation (13)
gives us
probability
1 rah
77/ .
dt
— ah
rah
J'
.dt
(14)
from the symmetry of the curve.
The value of this integral has been calculated and tabulated for
gradually increasing values of ah, and it is evident that such a
table will show by inspection the number of errors we may expect
to find between any two arbitrary limits, no regard being paid to
the sign of the errors.
A view of the distribution of these errors with regard to magni-
tude may be interesting. The number of observations is supposed
10
ARTILLERY PRACTICE
large only to show how small is the chance of the occurrence of
large errors. In 10,000 errors there will probably be —
Between
/ =
and
t =
0-5
5205 err
/ =
0-5
,,
t =
1-0
3222
/ =
1-0
„
t =
1-5
1231 ,
I =
1-5
„
t =
2-0
292
t =
2-0
.,
t =
2-5
43 ,
/ =
2-5
t =
3-0
4 „
and between t = 3'0 and t= y- there will probably not, in 10,000
observations, be a single error.
The definite integral (14) enables us also to deduce the probable
error.
By probable error we understand that error, than which there
are as many errors less as there are greater. The probability of
such an error must be i, and if then we designate the probable
error of a single round by r, Equation (1-4) becomes
1 •2_rhr
dt
(15)
and the table of the values of this integral, of which we have
spoken, shows that in this case we must have
hr = -4:76936 = p, suppose . . . (16)
But, by Equation (10),
1
hence
;• = -476936 . e, ^2 = -674489 . e.f
(17)
Before applying this last formula a correction must be made,
the reason for which we shall endeavour to explain.
We have supposed e., to have been determined from the true
errors of observation, whereas it has been determined only from
the most probable errors. Now, we have already pointed out,
•674,
* It may be useful to note, that it is an even chance that the probable error of
a single datum, r, lies oetween
, -4769 \ , ^„,. /, -4769 \
1 and -6/4.0. e., 1 - —, —
Also, if R be the mean range, and rn be the number of rounds from which it was
obtained, it is an even chance that we have not erred in our determination of R by
a quantity greater than — ^
ARTILLERY PRACTICE 11
that although we may select the most probable hypothesis, yet
the odds are strongly in favonr of our erring by a small quantity.
It will be borne in mind, that 63 was calculat-ed so as to be a
minimum, and hence the true mean error (supposing that our
hypothesis is erroneous by a small quantity) will be slightly
larger than the hypothetical mean error. Equation (17) would
then give us a probable error rather too small, and analysis shows
that the proper correction is made by substituting in Equation (8)
771 — 1 for '/;'■/, so that we obtain eo as nearly as possible from the
equation
e.,2 = ^(:^ (18)
m — i
and this value of eo must be employed in Equation (17).
^Ve are now in a position to apply these results to practice.
We shall select for illustration an actual experiment made by the
Committee on Eifled Gannon to try the relative accuracy of two
gims — a rifled 18-pr. of 12 cwt. and the service brass 9-pr.
The first of these guns gave, as regards range, the foUowing
data in yards: —
1023 1018 1005 1020 1005 1005 lOlS 1005 1026 1011
1032 1020 1025 1024 1023 1038 1032 1032 1026 1007
1002 1002 1002 1005 1018 1013 1032 1021 1005 1001
1018 1018 1025 1012 1037 1038 1032 1026 I'M- 1025
giving a mean range of 1019 yards.
Hence the errors of the preceding data, assuming the mean
range to be the true one, are, when arranged according to magni-
tude and without regard t^ sisn —
(19)
Taking the sum of the squares of these eiTors, we obtain from
Equation (18) for the mean error,
e. = x'i^ = 11-01 yards:
\ 3y •
whence from Equation (17) we have the probable error,
r = -6715 1, = 7-1 yards .... (20)
19
17
U
u
13
7
n
.-,
1
1
19
17
u
13
13
7
6
5
1
1
18
15
u
13
12
7
6
4
1
1
17
U
u
13
7
6
5
4
1
1
12 ARTILLERY PRACTICE
We pursue precisely the same course * with respect to the deflec-
tions, save that we must first reduce them all to their value at the
mean range.
In the case before us the deflections at 1019 yards were, in
inches —
32 right
2
left
G
right
30 right
30 „
20
right
4
left
42 „
24 „
38
33
„
48 „
19 „
39
„
15
right
16 „
33 „
39
15
,j
20 ,,
33 „
39
J,
15
}>
24 „
25 „
32
„
8
16 „
2 left
32
„
37
33
6 left
2 right
18
„
28
33
4 33
3 „
14
33
28
33
17 right
Hence the mean point of impact is 20 inches right, and the errors,
arranged according to magnitude and without regard to sign, are —
(21)
28
22
19
18
13
12
8
5
4
26
22
19
17
13
12
8
5
4
24
22
19
17
12
10
6
5
4
24
20
18
14
12
10
5
4
3
4 0[
3 oJ
* It is to be observed that the method here adopted is somewhat faulty. The
following would be the stricter course of procedure : —
If d be the probable angle of deflection, since is always very small, the angular
deflection of each round is given by the equations
rj^-f?j = 0, r^e-cL^O, etc,
where r,, r^, etc. are the ranges, and dj, d„, etc. the corresponding deflections.
The most probable value of d will then be that for which
{r^6 - f?i)'^ + {r^O - d.,)- + , etc. , is a minimum.
It is also obvious that the probable area is not, as is here supposed, a rectangle, but
is such as is shown in Fig. 3.
f/g.3
The error, however, induced by adopting the method followed in the text is of
a very small order, while the calculations are thereby very much simplified.
ARTILLERY PRACTICE la
Equation (8) gives us for the mean error
/8425 ^ . . . ,
Co = a/^39~ = ^^'^ inches
and .-. r = -6745 6^ = 9-8 inches . . . (22)
We have now obtained the probable error in range and in deflec-
tion. The probability of each event is |, and consequently the
probability that a shot would fall within both the probable errors is
I, that is, the chances are three to one against it.
Let us now suppose that we have found an error in range A',
within which a little more than yV^hs of the errors (-7071) will
probably fall, and an error A" in deflection, within which we may
also expect -7071 of the errors to fall.
The probability of a shot falling, so that its error in range shall
be less than A', and its error in deflection less than A", will be
= -7071 X -7071 = -5 = h
The rectangle 2A' x 2A" (Fig. 4), then, is a rectangle such
F/g 4
Ran0t
that there is an even chance of any one shot falling within it, and
this area, as before mentioned, I have named the probable rectangle.
Now, the definite integral (14) enables us to find A' and A".
For that integral has been tabulated according to an argument in
which the probable error is assumed to be unity. For the argument
A
the table gives the value of
A
-^f .c-'\dt .... (23)
that is, it shows how many errors may be expected to be less than a
given error, provided the ratio of the given error to the probable error
be known.
Conversely, the value of (23), or the number of errors which should
fall within a certain unknown limit, A being known, the table will
give the proportion which A bears to the probable error.
14
ARTILLERY PRACTICE
1-56
dt = -7071
A = 1-56/
(24)
(25)
In the case before us, we have
A
and the table shows that, in this case, we have
A
;•
Hence, from Equations (20) and (22) in the case we are discussing,
we have
A' = 11-54 yards and A" = 15*29 inches,
and the probable rectangle laid down as shown in Fig. 4, is 231
yards long by 30-6 inches broad.
The field gun which was fired for comparison with the rifled gun,
practice from which has been just discussed, gave the following
ranges : —
1038 825 1096 1078 977 1021 1014 849 1034 1001
1038 950 1033 1007 1030 910 875 953 942 1013
1053 1006 902 900 1090 1138 975 975 960 940
1008 1080 965 925 1061 932 994 979 910 912
Giving a mean range of 984-75 yards or 985 yards. Hence, the
errors arranged as before, according to magnitude, and without regard
to sign, are —
160 110 85 75 53 43 36 28 21 10^
153 105 83 73 49 43 35 25 20 9'
136 95 76 68 48 43 32 23 16 81
111 93 75 60 45 41 29 22 10 6;
and taking the sum of the squares of these errors, we have, from
Equation (18),
/190372 .^„ ,
^. = ^-39- = 69-9 yards .
whence r = -674562 = 47*2 yards
Again, the deflections in feet, at the mean range, wer(
(26)
(27)
(28)
24 right
42
right
6
12 „
3
6
22
12
„
9
24 „
12
„
9
12 „
6
„
9
33 „
9
jj
9
18 „
9
»
3
21 „
24
)>
3
9 „
6
18
18 „
6
!3
12
right
left
18 left
„
„
„
„
„
.,
„
„
„
ARTILLERY PRACTICE 15
(29)
hence the errors are-
as
20
17
11
7
7 7
5 2
1
26
19
16
11
7
7 5
4 2
1
25
17
16
10
7
7 5
2 2
1
25
17
14
10
7
7 5
2 1
1
and summing
the
squares of these
errors, '
we have
c-i
/6711
V 39
= 13-1 feet
r
= -6745 e.
= 8-8 feet .
.
• (30)
and multiplying the values of r, given in Equations (28) and (30),
by 312, we obtain for the probable rectangle in this case a space of
147'2 yards in length by 9"1 yards in breadth.
Figs. 5 and 6 (see Plate I., p. 22) show the comparative areas of
the probable rectangles of these guns at the given ranges.
For the sake of clearness, the various steps to be taken in order
to ascertain the probable rectangle by the foregoing method are here
recapitulated.
First, as regards range —
Find the mean range, and assuming it to be the true range, find
the errors of each round.
Square these errors, and calculate the mean error from the
formula
'-Vl^^ (-)
where 2)(A-) = sum of the squares of the errors and m = number of
rounds fired.
Calculate the probable error from the formula
r = -6745 €, (32)
Second, as regards deflection —
Eeduce all the deflections to their value at the mean range. Find
the mean point of impact, and thence the error in deflection of each
round. Square the errors, and the mean error will be given by
Equation (31), the probable error by Equation (32).
Lastly, multiply the probable error both in range and deflection
by 312. We shall then have the dimensions of the probable
rectangle.*
* It is probable that we do not err in our determination of the probable
rectangular area by a quantity greater than \^a~^^ + a"-^'^ where a, ^ are the sides
of the probable rectangle, and
•477 . a ^, -477 . B
m being the number of rounds fired.
16 ARTILLERY PRACTICE
The foregoing method of determining the probable error from the
sum of the squares of the errors, gives us that probable error with
greater certainty than can be attained by any other method. The
operation of squaring the errors, however, is laborious, especially
if the number of observations be large ; and the method is in truth
too great a refinement for ordinary artillery practice.
We proceed to indicate a method by which the value of r may be
obtained from a knowledge of the errors merely, and which, from its
simplicity, and from its indicating the probable error with quite
sufficient exactness, is well adapted for general application to artillery
practice. Symbolising by e^ the arithmetical mean of all the errors,
we must have in this case *
e, =1^4 (33)
'- m— i
where 2A= sum of the errors, without regard to sign; r is
determined by the equation
r = -8453 €, .... (34)
We have mentioned that when r is determined from eo, it is an
even chance that r lies between
.6745<,(l ± II??) .... (35)
In this case, it is an even chance that the value of r lies
between
.8453., (l±-^^) .... (36)
V J,n! ^ ^^
and the numerical part of the limiting values shows that we obtain r
within the narrowest limits when we determine it from e,.
Let us now apply this method for the purpose of comparison to
the cases we have already examined. In the first of these cases we
find from (19) that, with reference to range,
;(A) = 366
366
^1 - ^
gg 9-38 yards
and r = 7-9 yards . . . • (37)
See Encke, on the Method of Least Squares.
ARTILLERY PRACTICE 17
Again, with reference to deflection, from (21)
2(A) = 487
487 ,OA- u
€j = -^ = 12-4 inches
and r =-845 e^ = 10-4 inches - . (38)
and these results, it will be perceived, differ but slightly from those
obtained in (20) and (22).
In the second case, from (26) we have
2(A) = 2253
2253
ej = -^ = 57-7 yards ;
r = 48-6 yards .... (39)
Also, from (29) we have
— = 10 feet
and ?■ = 8-4 feet (40)
Eesults again differing by small quantities only from those
obtained in (28) and (30).
Hence, to obtain the probable rectangle by this method, find the
mean range, and thence the error of each round ; calculate the mean
error from the equation
, ^ ^ (41)
^ 7)1-1
where S(A) = sum of the errors without regard to sign, and m =
number of rounds fired.
Calculate the probable error from the equation
r = -8453 e^ (42)
Find the mean point of impact, and thence the error in deflection
of each round. Compute the mean and probable errors from
Equations (41) and (42).
Finally, multiply the probable error, both in range and deflection,
by 312, to give the lengths of the sides of the probable rectangle.
We shall now apply this method to solve a question which has
lately been the subject of extended practice under the direction of
Captain Haultain, viz., to find the advantage, if any, in point of
accuracy, gained by using with the service 9-pr. a charge of 3 lbs.
instead of that at present in use, viz., 2h lbs.
B
18
ARTILLERY PRACTICE
The2|
lbs. charge
gave
the following ranges wi
th 2° e
[evatic
798
844
876
893
907
921
943
963
1016
1050
798
845
880
897
908
927
944
964
1017
1050
811
850
880
897
912
930
947
967
1018
1066
818
850
881
898
913
930
947
973
1022
1082
819
850
883
899
915
931
950
974
1024
1082
821
853
884
900
916
931
950
976
1029
1089
822
857
885
901
916
932
950
983
1030
1123
825
867
885
904
916
932
952
1002
1042
1132
837
868
889
905
920
934
954
1010
1049
1139
842
869
891
905
920
935
962
1015
1050
1177
giving a mean range of 936 yards. Four of these rounds we shall
discard for the following reason: — In a considerable number of
observations, such as is here discussed, we have a right to expect
that the greatest errors in excess shall not differ very greatly from the
greatest errors in defect. In this case, however, the maximum
positive error exceeds the maximum negative by more than a
100 yards, a very improbable result; and as every officer who has
had charge of a range party knows how liable, even with the greatest
care, is the second graze to be mistaken for the first, we think we
may here safely take the liberty of expunging the four rounds which
give positive errors so much exceeding the maximum of negative
errors. This liberty should, however, be most sparingly exercised,
and never without adequate cause.
We have now as a mean range 927 yards, and the following
system of errors, arranged as before, according to magnitude and
without regard to sign : —
162
116
95
77
47
37
26
19
7
156
115
91
77
47
36
25
17
7
155
107
90
75
47
36
23
16
6
139
108
90
74
46
35
23
15
5
129
106
89
70
46
34
23
14
5
129
105
88
60
44
30
23
12
4
123
103
85
59
43
30
22
11
4
123
102
83
58
42
29
22
11
3
123
102
83
56
42
28
20
11
3
122
97
82
51
40
27
20
8
77
49
38
5240
27
55-1
20
yards
7
(43)
and
r = -8453 cj = 46-5 yards
. (44)
ARTILLERY PRACTICE
19
The deflections corresponding to the above ranges were, in feet-
11 left
20 „
• 5 „
^ 5 ,,
Si :;
2-5 right
9 le
9 ,
,
12 ,
1 ,
27 ,
26 ,
3 ,
8 ,
8 ,
1 ,
23 ,
ft 19 left
40 „
14 „
1 right
2 „
2 „
8 „
12 right
26 „
31 „
7 „
16 „
20 „
8 „
23 „
47 „
3 „
56 „
13 „
17 right 14 rig
7 „ 20 ,
63 „ 17 ,
ht 14 right
39 „
5 „
22 „
17 „
37 „
6 „
8 „
35 ,,
8 „
20 „
10 left
1 left
9 „
6 „
6 „
30 right 63 ,'
13 ,, 16 ,
.15 „ 23 ,
|19 " 27 ,
::28 „ 47 ,
^38 „ 17 ,
15 „ 15 ,
16 M 37 ,
32 „
2 left
29 „
d23 „
^15 „
£ 14 „
S17 „
4 right
6 „
7-5 „
19 left
§ 4 „
^ 7 right
^ 20 „
"57 „
. 2 left
^ 31 „
5 54 „
*- 22 ,,
and these deflections (the effects of wind being eliminated by calcu-
lating the errors for each day separately) give the system of errors —
44
42-5
41-5
39-5
36-5
35-5
29
28
28
27-5
26-5
25
25
24-5
20-5
20
20
19
18-5
18
18
18
17-5
17
17
16
15-5
15
14-5
14-5
145
14
13-5
13-5
13-5
13-5
13-5
13
13
12-5
12-5
12
12
12
11-5
11-5
11
11
11
10-5
10-5
10-5
9-5
9-5
9-5
9
9
8-5
8
7-5
6-5
6
6
5-5
5
5
4-5
4-5
4-5
4
4
4
4
4
3-5
3-5
3-5
3
3
3
3
3
3
3
2-5^
2
2
2
1-5 I
1 I
0-5
0-5 !
0-5 I
0-0 j
(45)
and from these errors,
and
1186-5
95
•8453 e.
= 12-5 feet
= 10-5 feet
. . . (46)
and the probable rectangle is 145*1 yards in length by 10-9 yards in
breadth.
rhe 3 lbs. charge from the
same
guns gave the
following
ranges
816
897
927
947
967
990
1002
1018
1033
820
900
927
950
968
991
1003
1019
1043
826
904
928
950
972
991
1003
1021
1072
839
908
933
951
975
991
1006
1022
1073
863
908
933
951
978
991
1007
1022
1076
868
910
933
952
979
992
1009
1025
1077
873
910
935
955
979
994
1010
1026
1080
877
911
936
960
980
999
1011
1026
1082
882
915
938
960
981
1001
1012
1029
1109
883
919
943
964
987
1001
1012
1030
1112
895
926
947
964
990
1002
1016
1032
1130
20
ARTILLERY PRACTICE
One round was discarded from this series for the reason men-
tioned in the discussion of the preceding case. The remaining
rounds gave a mean of 972 yards, and the following system of
errors : —
158
156
104
104
45
45
152
101
64
54
44
36
27
19
8
146
100
64
53
44
35
25
19
8
140
99
62
53
40
34
25
19
7
137
95
62
50
40
34
22
18
7
133
90
61
50
39
31
22
18
6
110
89
61
49
39
31
22
17
5
109
77
60
47
39
30
21
15
4
108
75
58
46
39
30
21
12
3
105
72
57
46
38
29
20
12
Hence
5041
'i 98
51-4 yards
and
8453 e^
= 43-5
yards
.
(47)
(48)
The errors in deflection, the wind being eliminated as in the
former example, were, in fact —
(49)
34
21
15-5
13 10
7
5
3-5
30
21
13 10
6-5
5
3
27
20
12 9
6
5
3
24
19
12 9
6
5
3
24
19
12 9
6
5
3
23
19
11-5 9
6
5
2
•5
23
19
11 8
6
4
2
•5
22
18
11 8
6
4
2
•5
22
17
11 7
6
4
2
21
16
13
11 7
6
4
1
21
16
13
11 7
5
4
1
anc
(
1008-8
1 98
10-3 feet;
whence
r = -8453 €^ =
8-7 feet
.
(50)
and hence the probable rectangle is 135-7 yards in length by 9 in
breadth.
These comparative areas are exhibited in Figs. 7 and 8 * (Plate I.,
* It will be seen by comparing Equations (4) and (40) that the probable deflec-
tion deduced from the 2i lbs. charge in Captain Haultain's practice is somewhat
larger than that obtained by the Committee on Rifled Cannon, with a similar
charge and at a similar range. This is doubtless attributable to Captain Haultain's
practice having been chiefly carried on during a wind variable and across the range.
ARTILLERY PRACTICE 21
p. 22), and it follows that the 3 lbs. charge gives results slightly
but decidedly more accurate than those of the 2| lbs. charge.*
This advantage in point of accuracy does not, however, appear to
increase in a marked manner at higher angles.
We have partially discussed the practice made with the 2h lbs.
and 3 lbs. charges at 4° of elevation.
To avoid a tedious repetition of numerical examples, we merely
give diagrams of the probable rectangles, with their dimensions, in
Figs. 9 and 10 f (Plate I., p. 22), drawing attention to the singular
decrease in accuracy caused by an increase in the range of about 400
yards, the probable deflection being in fact more than doubled, while
the range is not increased by 50 per cent. ; and it would follow that
a limit is soon reached beyond which it is mere waste of ammunition
to fire at an object even of considerable size.
There is yet another way of attaining, when the observations are
numerous, to an approximate knowledge of the probable error. We
have defined the probable error to be that error than which there are
as many errors less as there are errors greater ; hence, if the number
of observations be odd, the centre error (supposing the errors to be
arranged according to magnitude), and if the number be even, the
mean of the two centre errors, ought to give an approximation to the
probable error. The probable errors deduced in this way from (43),
(45), (47), and (49), are 44 yards, 11 feet, 40 yards, and 9 feet— results
not differing very greatly from those given in (44), (46), (48), and (50).
It now only remains to say a few words relative to the employ-
ment of the methods pointed out.
We have remarked on the rapid increase of error in the 9-pr.
field gun, but we may put the more general question, " What is the
relative accuracy of the various guns and projectiles now in the ser-
vice, and what are the limits of their effective ranges ? " A series of
experiments for the purpose would easily enable us to answer this
question, and it is clear that an accurate knowledge of the powers of
the guns would not only help to a right decision with regard to the
* It may be mentioned as a point of interest, that Captain Haultain's practice
above discussed was carried on on five different days. The probable rectangle
was calculated separately from the result of each day's practice as well as from the
combination of all the days. The differences between these probable rectangles
were very trifling, thus showing in a remarkable manner how regular in its irregu-
larities was the practice obtained from these 9-pr. guns.
t It is to be observed in comparing the relative errors of the 2^ lbs. and 3 lbs.
charges, that the errors of the 3 lbs. charge belong to a range somewhat greater
than that of the 2^ lbs. charge.
22 ARTILLERY PRACTICE
most suitable * for any particular service, but might be a valuable
guide in reducing our list of ordnance.
We should also, from a series of suitable experiments, be enabled,
as has been suggested by Captain Lyons, to determine approximately
the errors which are due to some specific causes, such as eccentricity
of shot, etc., and also determine the increase of accuracy due to a
decreased windage.
There is perhaps no branch of mathematics from which more
information of importance to practical artillerymen can be gained
than from the Theory of Probabilities. In the preceding pages, an
attempt has been made to develop one of its applications, and
although it has been impossible to enter fully into the subject within
the limits of such a paper as the present, we yet trust that the
utihty of applying its methods to the examination of artillery
practice has been sufficiently exhibited.
* To take an instance which has been the subject of considerable discussion, we
would be enabled at once to assign the advantages in point of accuracy possessed
by the 9-pr. over the 6-pr. at various ranges.
PLATE. I.
FIG. 5
^
Pro6ab/e rectao^k^ Rif/ec/ /8 P^ Mean fiance /0/8yan;k.
23-/yan*s ^
FIG 6
Prv6o6/erecian^/e, Service Bnrss 9 P^, (y?aye Z'-^zl^. Afeon /(!7r?pe985yarv(s.
I ^!
FIG 7
Pro6<76/e rechr^/e, SerWce Bmss 9P^, Ci?arye 2 ^z IM, Afeoo /^an^e 927yanak.
145-1 yare/s
FIG. 8.
PrvSoS/e iTc^n^k Sery/ce Brass 9P^, C/nryeS L6s, kf&in Pan^ 972y€/rv6
/3S-7yan7ls,
I 1 1
FIG. 9
Pro6a6/e rvchrjpk, Brass 9P'?, C/?ar^ Z'^z /.6s. Mean P(rr^ /332yarc/s.
20ej47rck
FIG. 10.
Prvdcf6/e rec^anp/e, Brass SP^, SlhCharye. Afea/^P?ar^ /409 yarak .
r34-7yarak.
[To face p. 22.
11.
EEPOET ON EXPEEIMENTS WITH NAYEZ'S
ELECTEO-BALLISTIC APPAEATUS.
{Boyal Artillery Institution Papers, 1863.)
1. In forwarding to the Ordnance Select Committee the results
of the experiments in initial velocity, which I have had the honour
of carrying on under their direction, I have to make the following
remarks : —
2. The instrument employed in these investigations was the
electro-ballistic apparatus of Major Navez,* and it may not be out
of place here to recapitulate the leading points of its construction.
The apparatus itself is merely an arrangement for measuring,
with extreme accuracy, a certain very small interval of time. Two
screens, the nearer one a short space from the muzzle of the gun, are
placed at an accurately measured distance apart, and it is the object
of the instrument to ascertain the time which the projectile takes to
pass over this measured space.
3. The apparatus consists of three parts, the pendulum, etc., the
conjunctor, and the disjunctor. The principal part is the pendulum
and graduated arc. The pendulum, before an observation, is held
suspended by an electro-magnet, the current magnetising which,
passes through the first screen. To the pendulum is attached, by
means of the pressure of a spring, an arm with a vernier. The pressure
of this spring is so regulated that the arm vibrates freely with the
pendulum, but at the same time it offers but little resistance to the
action of a powerful horse-shoe electro-magnet, which, when the
circuit magnetising it is complete, clamps the vernier arm with great
firmness.
4. The current which passes through the second screen holds, by
* Belgian Artillery.
24 REPORT ON EXPERIMENTS WITH NAVEZ'S
means of an electro-magnet, a weight suspended over a spring, a point
from which is kept just over a cup of mercury. When this weight
is permitted to fall, it presses the point into the cup of mercury, and
completes the circuit, magnetising the horse-shoe magnet, which
clamps the vernier needle. This part of the apparatus is termed the
conjunctor. The action of the instrument is very simple, and readily
understood. When the projectile cuts the wires in the first screen,
the magnet which holds the bob of the pendulum in its initial position
is demagnetised, and the pendulum commences an oscillation. When
the wires in the second screen are cut, the weight of the conjunctor
drops, completes the circuit, clamping the vernier, and the arc through
which the pendulum has moved is a datum from which may be
computed the corresponding time.
5. An important part of the apparatus (the disjunctor) remains
yet to be mentioned. It will be obvious that the arc, which we have
just supposed to be measured, corresponds to the time which the pro-
jectile takes to pass over the distance between the screens, plus the
time which the weight of the conjunctor takes to fall from its initial
position to the cup of mercury. Now, to obtain the former, the latter
of these times has to be subtracted from the reading of the instrument,
and the disjunctor enables us to do this by permitting us to break both
currents (those through the first and second screens) simultaneously.
The mode of procedure is then as follows : — The instrument being
arranged, the two currents are simultaneously broken by means of the
disjunctor, and the reading of the needle is recorded. The instrument
is again adjusted, the projectile fired (the velocity of which it is desired
to determine), and the reading of the needle again noted ; the former
arc is subtracted from the latter, and the corresponding time computed.
It will be observed that, by the use of the conjunctor, any constant
source of error (such, for example, as the error due to the time
required to clamp the vernier needle) is eliminated, as the same error
will occur both in the disjunctor and the projectile reading, and by
subtraction will disappear.
The disjunctor also enables us to ascertain the degree of regularity
with which the instrument is working, as the accidental variations of
the reading corresponding to the time are, of course, the same as the
variations which would occur in the reading corresponding to any other
time. Major Navez lays down, as a rule, that observations should
not be proceeded with when in a series of ten or twelve disjunctor
readings there is between two successive readings a difference greater
than 0°-25.
ELECTRO-BALLISTIC APPARATUS 25
6. It is of some importance to be enabled to put an exact estimate
on the degree of reliance to be placed on the results of Major Navez's
beautiful instrument ; and, to do this, let us observe that the arc from
which the required time is computed is the difference between two
arcs, in our estimation of each of which we are liable to a small error.
We have in fact the value of one arc ^ given by the equation
^ = cl>-^' (1)
where (p and (p' are each subject to probable errors (let us suppose)
r and ?-'; the probable error of $ is then Jr^ + r"^. If, after the
satisfactory working of the instrument has been ascertained and
the probable error determined, we take a single reading with the
disjunctor, and then with the projectile, r and r are equal, and the
probable error of the observation is r J 2. We have it, however, in
our power, if it be thought necessary, to reduce even this error, for
if the disjunctor reading be taken, the mean of, say five observations,
r IW
we have h' — —j^, and the probable error of $ is r^J -=, which differs
but slightly from r. An example will show how very trifling this
error generally is. With an Armstrong 12-pr. shell, whose velocity
is determined to be 1181 feet per second, the value of r is found to
be 0°"06, and the disjunctor reading being the mean of five observa-
tions, the probable error of «l> is 0°-07.
Hence the disjunctor reading being 42''"85, and the projectile
reading 107°"40, it follows that it is probable that in our determination
of 1181 '2 feet as the velocity at a point midway between the screens,
we do not make an error exceeding 1*4 feet ; that is to say, it is an even
chance that the true velocity of the single observation lies between
1179-8 feet and 11826 feet. As the round from which the above
example is selected is one of a series of ten, the probable error in
our determination of the mean velocity between the screens will be
less than one-third of that just given, or the mean velocity may be
assumed, as far as instrumental errors are concerned, to be practically
correct.
7. The experience which I have had with Major Navez's instru-
ments enables me to say, that if ordinary care be taken in their
use, and the instructions carefully followed, the instruments are so
nearly perfect as to leave little to be desired, while the ease with which
they can be manipulated and the innumerable important problems
which can be readily solved by their means, render them an invaluable,
an almost indispensable, adjunct to every school of instruction.
26 REPORT ON EXPERIMENTS WITH NAVEZ'S
8. Two instruments, Nos. 24 and 32, were used in these experi-
ments. The times of vibrations of the penduhims were carefully
determined by means of a stop-watch, and the rate of the watch was
ascertained by comparison with an astronomical clock. The observa-
tions made for this purpose are given in Appendices * Nos. I. and II.,
and from them it appears that the time of a small oscillation in
instrument No. 24 is 0-3320 seconds, while in No. 32 it is 0-3337
seconds.
9. In Appendices* III. and IV. are given corrected tables,
showing the relations between the arcs passed through and the
corresponding durations for T = 0-3320 seconds, and for T = 0-3337
seconds.
10. The experiments referred to in this report have regard
chiefly to initial velocity alone ; and for the small distance concerned,
the law of resistance adopted may be thought of small practical im-
portance, especially as before the experiments now carried on are con-
cluded, the Committee wiU doubtless be in a position to say whether
this law is better expressed by a function of the form v^-\-av^,
as proposed by General Piobert, or by one of the form v^ + (iv^,
as proposed by the Count de St Eobert and Colonel Mayevski. In
the present instance, both the law of resistance and the values of the
coefficients given by General Didion in his invaluable work have
been followed, although it may, perhaps, be inferred from a passage
in the recent edition of the TraiU de Balistiqiie, that late experi-
ments with the electro-ballistic apparatus do not give results in
quite so close an accordance with theory as might have been
expected.
11. In the first edition of General Didion's work, published in
1848, a term was introduced into the expression of the resistance
of the air dependent upon the diameter of the projectile, and this
form of the expression has been generally used upon the Continent ;
but a recalculation of the data upon which this result was founded
has led General Didion to conclude that the coefficient is independent
of the calibre, and that the resistance is represented with sufficient
accuracy by the equation
,= .037.RV.^{,+^-|g} . . . (2)
where E = radius, ■?; = velocity, o = density of the air at time of
* These Appendices, having reference only to the use of the instruments now
superseded, are omitted.
ELECTRO-BALLISTIC APPARATUS 27
observation, and Si = standard density of air ; the metre and the
kilogramme are taken as units.
In this formula the density of the air is denoted by referring
its weight to a standard of comparison, which is assumed as the
weight of a cubic metre of air at a temperature of 15° Cent.,
semi-saturated with vapour, and under a barometric pressure of
760 mm.
Now, if in Equation (2) the English foot and pound be taken as
units, the value of the numerical coefficients will be altered, and the
equation becomes
•0005 ISTttRV. All +
8, r 1426-4.
In the ordinary determinations of initial velocity it is hardly
necessary to take the variations of the density of the air into account,
and it only remains so to alter the coefficient -0005137 that the
error arising from neglecting this variation may be as small as
possible.
12. According toKegnault, the weight of a cubic foot of dry air
at a temperature of 32° Fahr., and under a barometric pressure of
30 inches, is = 566'56 grains; and according to the same author, the
coefficient of the expansion of the air for an increase in tempera-
ture of 1° Fahr. is = -002036. Hence if S be the weight in grains
of a cubic foot of dry air at any temperature t, and pressure 11,
n 566-56
30 1 + -002036 (r - 32°)
but (see Miller's Hydrostatics, p. 28) —
Weight of moist air at any teinperature and pressure _ ^ o.q>-q ^
Weight of di-y air at same temperature and pressure 11
where T = tension of the aqueous vapour. Hence the density of the
air under any circumstances will be found from the following
equation : —
1 - 0-378 i^ 566-56
^l . . . (3)
30 1 + -002036 (r- 32°)
And if we assume as the English standard of comparison the weight
of a cubic foot of air at a temperature of 60°, under a barometric
n(
28 REPORT ON EXPERIMENTS WITH NAVEZ'S
pressure of 30 inches, and if we further assume the humidity =
0'5, from (3) we find (5 = 534-3 grains, and Equation (2) becomes
P = •00052137rR-V-Yl + -- ^-i— . . (4)
^ V 1426-4; 534-3 ^ ^
and under ordinary circumstances the fraction ,„ . ^ may be taken
as equal to unity.
13. The above formula (4) applies to spherical projectiles ; in the
case of the Armstrong projectiles, the resistance of the air is repre-
sented by
„.-000M76.RV^l + j^Jj^ . . (5)
The velocity v at a point midway between the screens having
been determined by observation, the initial velocity v is deduced from
it by the equation
1 +
(l + ^).- .... (6)
where ?-= 1426-4, a; = distance, on the axis of the gun produced, of
w
the point corresponding to v, c = -^ — , w being the weight of the
projectile in lbs., g the acceleration of gravity, and n, in the case
of sj)herical projectiles, =-0005213xR^; in the case of Armstrong
projectiles, = •00034757rR2.
14. Discussion of the results. The experiments made relate
solely to the determination of the initial velocity of service pro-
jectiles fired from service guns with service charges. The detailed
results of the practice furnished in extenso give every particular
with regard to it, and the table on next page gives an abstract of
the general results.
It will be observed that the values of the "Measure of pre-
cision " for each of the series of which the result is here given, is
placed in the above table in a separate column. The value of
this constant denotes the comparative regularity of the initial
velocity.
As might perhaps be expected, from the absence of windage, the
12-pr. Armstrong has shown the greatest regularity, and I have
therefore assumed the measure of precision for this gun as unity.
An inspection of the values of the " Measure of precision " will
show how great is the amount of irregularity which exists in the
ELECTRO-BALLISTIC APPARATUS
29
F. T.
1022-2
717-2
1145-3
1169-5
1108-7
712-2
706-1
741-6
810-1
621-3
570-1
455-9
482-3
350-5
274-9
168-9
95-2
82-1
182-0
3-1
0-85
37-2
113-6
•X^IOOpA
F. S.
1292-3
940-6
1579-0
1809-9
1790-7
1308-5
1487-9
1464-4
1506-4
1690-0
1618-7
1447-5
1720-5
1690-6
1769-8
1613-7
1484-5
1163-4
1252-7
1167-6
1272-8
946-4
1190-2
•noisioajd
JO 9inSB8K
O »0 COJr^ CO r-l ^ CO ^ >0 O Tt< CO CO ^ CD CO O OO o
Cr>t^OM-*HC00 1'-0(M'*0(M'*000'-lt^CD • • -o
■jji .^ j>. eq oti >^ ip ip CO -^ o0i— ii— 11— i(Nooooo ooco loo
%%
§
1
2
10-in. gun (87 cwt. )
68-pr."(95 cwt!)
8-in. gun (65 cwt.)
32-pr. (58 cwt.)
24-pr. (50 cwt.)
18-pr. (38 cwt.)
12-pr. (18 cwt.)
9-pr. (13 cwt.) .
6-pr. (6 cwt.) .
12-pr. howitzer (61 c
24-pr. howitzer (12 c
Wall piece
Enfield rifle .
6-pr. Armstrong *
12-pr. Armstrong
30 REPORT ON EXPERIMENTS WITH NAVEZ'S
initial velocities of some of the projectiles fired from smooth-bored
guns.
To illustrate the application of these constants, we may compare
their value for the 12-pr. howitzer and the 12-pr. Armstrong, the
velocities of the projectiles fired from these guns being nearly the
same, but by the table it appears that the measure of precision in
the former case is only about one-fourth of that in the latter case,
or in other words, the mean error in initial velocity alone is
nearly four times as great. The great irregularity in the initial
velocity of the Martin shells is also very conspicuous.
15. The relation between initial velocity, weight of charge,
weight of projectile, and length of bore is given (see Didion, TraiU
de Balistique,) by the following equation : —
V = y\^ '- log- A — — — . . (7)
when V = initial velocity, /m = weight of charge, m = weight of shot,
bottom, etc., M = quantity of powder required to fill the bore, =
calibre of gun, C' = diameter of shot, y and X are constants whose
values have to be detemiined by experiment. The second term of
the right-hand member of Equation (7) represents the decrement in
initial velocity due to windage, and the value of the coefficient X
should be derived from a series of experiments expressly instituted
for the purpose. Strictly speaking, this value depends upon a great
variety of conditions, but chiefly upon the strength and physical
properties of the powder, and upon the length of the bore of the
gun. Under normal circumstances, however, the mean value of X
may, with but a very trifling error, be assumed ; and General Didion,
in his work above referred to, gives X = 2300 as the result of the
French investigations with the service gunpowder, but an analysis
of the above experiments points to a considerably higher value.
Indeed, from instances in these experiments, where the variation in
windage was sufficiently great, 3158 has been obtained as the mean
value of X, and as this number very nearly agrees with that stated
by Colonel Boxer to result from the mean of Major Mordecai's
extensive experiments on windage, I have taken as correct the value
of X, viz. 3200, given by that officer.
Assuming X as above given, y is easily computed from the data
furnished by experiments, y varies chiefly with the nature and con-
dition of the powder employed, and the annexed table given the values
ELECTRO-BALLISTIC APPARATUS
31
which have been obtained for the several guns experimented with, and
the nature of the powder used in each case.
Table 2. — Values of y for the undermentioned smooth-bored
guns, deduced from the experiments recorded in Table 1.
Nature of gun.
Nature of powder.
Value of y.
10-in. gun
L. G., W. A.
3284
68-pr. (95cwt.).
L. G.',' HaU & Sons
3491
3536
8-in. gun (65 cwt.)
3307
32-pr. (58 cwt).
L. G.,W. A.
3428
24-pr. (50 cwt.).
„
3390
18-pr. (38 cwt.).
.,
3454
12-pr. (18 cwt.).
L. G.
3561
9-pr. (13 cwt.) .
,,
3422
6-pr. (6 cwt.) .
,,
3321
12-pr. howitzer .
,,
3291
24-pr. howitzer.
"
3275
The experiments under the discussion show that the equation
J, = yy — ^
log.—
^ fX
3
3200
C2 - C'2
C2
(8)
gives the velocity due to a variation in the weight either in the charge
or projectile with great exactness, the proper value of y being used
in each series, and this equation has therefore been used to calculate
the initial velocities of the various projectiles thrown from smooth-
bored guns.
These velocities may be depended upon as correct (supposing the
same powder to be used) within very narrow limits, and the computed
velocities are in this case perhaps preferable to direct determinations,
as, unless the whole series for each gun were carried on at the same
time, and with powder of exactly the same nature and date of manu-
facture, discrepancies from variations in the strength of the powder
would be sure to arise.
[Table
32
REPORT ON EXPERIMENTS WITH NAVEZ'S
Table 3. — Showing the initial velocities of the various service projectiles fired from
the undermentioned guns. The velocities marked * are observed; the re-
mainder are calculated from the data furnished by the observed velocities.
£
1
Projectile.
i
•"^
"3 >.
'■2 u
Nature of Ordnance.
~
§
"c
■S §
'5 S
6
6
Nature.
Weight.
'^
-1
n
Lbs.
Lbs.
"
F. S.
F. T.
10-gun (87 cwt.)
10
12
Hoi. shot
88-475
-142
1270-4*
990-0
10
8
Mar. shell
117-14
-1425
930-1*
702-8
'/,
10
12
Com. shell
92-625
-15
1257-5
1015-1
10
12
Case
77-625
-18
1353^7
986-3
\\
10
12
Grape
83-375
-18
1308^1
989-3
68-pr. (95 cwt.) '.
8-12
16
Shot
66-224
-168
1579-0*
1144-9
8-12
16
Nav. shell
51-5
•17
1809-9*
1169-8
8-12
16
Com. shell
49-875
-226
1790^7*
1109-0
8-12
10
Mar. sheU
60-0
-235
1308-5*
712-3
',! \\
8-12
16
Diaph. shell
Case
60-75
-195
1627-9
1116^3
8-12
16
45-687
-265
1818-1
1047-2
1! " •
8-12
16
Grape
66-5
-3
1475-3
1003-6
8-gun (65 cwt.)
8-05
10
Hoi. shot
46-007
•21
1487-9*
706-3
8-05
10
Com. shell
49-875
•194
1464-4*
741-6
•
8-05
10
Mar. shell
51-5
•13
1506-4*
810-4
',1 \\
8-05
10
Diaph. shell
60-75
•125
1356-9
775-6
8-05
10
Case
45-687
•195
1712-7
929-3
'•
8-05
10
Grape
66-5
•23
1214-4
680-0
32-pr. (58 cwt.)
6-375
10
Shot
31-375
•196
1690-0*
621-4
6-375
8
,,
31-389
•194
1618-7*
570-3
6-375
6
^^
31-349
•195
1447-5*
455-5
11 11 '•
6-375
10
Com. shell
24-312
•198
1912-6
616-7
6-375
10
Diaph. shell
28-75
•198
1762-4
619-2
6-375
10
Case
36-094
•228
1543-8
596-5
11 11
6-375
10
Grape
36-25
•228
1510-3
596-4
24-pr. (50 cwt.)
5-823
8
Shot
23-047
•208
1720-5*
482-2
5-823
8
Com. shell
17-5
•228
1948-2
460-4
" r.
5-823
8
Diaph. shell
20-875
•228
1786-0
461-7
5-823
8
Case
25-594
-2435
1594^7
451-3
!! 11 '•
5-823
8
Grape
26-0
-253
1571-6
445-3
18-pr. (38 cwt.)
5-292
6
Shot
17-656
-205
1690-6*
349-9
5-292
6
Com. shell
13-125
-193
1971-6
353-8
5-292
6
Diaph. shell
15-875
-193
1797-3
355-6
5-292
6
Case
19-562
-218
1588-5
342-3
5-292
6
Grape
19-5
-218
1591-2
342-3
12-pr. (18 cwt.)
4-623
4
Shot
12-656
-803
1769-8*
274-9
4-623
4
Com. shell
9-0
•169
1987-4
246-5
r, "
4-623
4
Diaph. shell
10-375
•169
1854-7
247-5
M 11 '•
4-623
4
Case
16-625
•159
1469-8
249-0
9-pr. (13 cwt.)
4-2
2-5
Shot
9-359
•1
1613-7*
169-0
4-2
2-5
Diaph. shell
8-062
•12
1707-3
162-9
4-2
2-5
Case
13-0
•1315
1318-8
156-8
6-pr. (6 cwt.)
3-668
1-5
Shot
6-23
-1
1484-5*
95-2
3-668
1-5
Diapli. shell
5-125
•118
1608-8
92-0
3-668
1-5
Case
8-5
•1275
1215-8
87-1
12'pr- howr. (6icwt.)
4-58
1-25
Com. shell
9-0
•126
1144-6*
81-7
4-58
1-25
Diaph. shell
10-465
•126
1058-1
81-2
4-58
1-25
Case
8-118
•148
1185-5
79-1
24-pr.'howr. (12 cwt.)
5-72
2-5
Com. shell
17-5
•125
1222-9*
181-4
5-72
2-5
Diaph. shell
20-875
•125
1113-0
179-2
"
5-72
2-5
Case
14-014
-15
1369-9
182-4
The mean weights and windages of the various projectiles have been taken.
o
UJ
—)
o
Q.
e? ui
o cc
^5
to o
«r
3:
'^^ ^
CO
^ Si-
^§ ^
r^
:5
^ ^
(O
<^ *^
UJ
UJ
^ C
1-
1-
-^
( )
■to <,
1 U.I
cvj
.
U^
i
o _
li
o
r
S5
£|
2
5
t
s
■■^s
^S
g
^
s
c-3
t9 ~
Lb. oz.
Lbs
oz.
F.S.
F.T.
Lb
oz.
Lbs.
oz.
F.S.
F.T.
Lb. oz.
Lbs.
oz.
F.S.
F.T.
1 8
9
1380
118-8
8
22
900
123-5
1 8
35
730
129-3
1 8
10
1321
121-0
8
23
881
123-8
1 8
36
720
129-4
1 8
11
1266
122-2
8
24
864
124-2
1 8
37
710
129-3
1 8
12
1213
122-4
8
25
850
125-2
1 8
38
700
129-2
1 8
13
1174
124-2
8
26
835
125-7
1 8
39
691
129-1
1 8
14
1134
124-8
8
27
827
128-0
1 8
40
682
128-0
1 8
15
1095
124-7
8
28
810
127-4
1 8
41
673
128-7
1 8
16
1060
124-6
8
29
797
127-7
1 8
42
664
128-4
1 8
17
1027
124-3
8
30
785
128-2
1 8
43
655
127-9
1 8
18
997
124-0
8
31
773
128-4
1 8
44
646
127-3
1 8
19
970
123-9
8
32
762
128-8
1 8
45
638
127-0
1 8
20
945
123-8
8
33
751
129-0
1 8
46
630
126-6
1 8
21
923
124-0
8
34
740
129-1
1 8
47
622
126-1
Variation in initial velocity between high and low gauge ^projectiles
(Armst7'ong).
24. The experiments in Table 8 were made with a view to ascer-
tain whether there is any difference in velocity between projectiles
of the highest and lowest gauges admitted into the service.
The results are here given, and it will be seen that there exists
between the velocities no appreciable difference.
Table 8. — Abstract of the results of the experiments made to ascertain the difference
in initial velocity hetv^een high and low gauge projectiles.
Armstrong
12-pr.
No. of
rounds.
Charge.
Projectile.
Velocity
at
30 yards.
Initial
velocity
Kemarks.
Initial
energj-.
Weight.
Diam.
No. 224
5
6
11
8
Lb. oz.
1 8
1 8
1 8
1 8
Lbs. oz.
11 9
11 9
11 9
11 9
3-080
3-085
3-080
3-085
F.S.
1184-1
1177-2
1184-1
1187-8
F.S.
1193-4
1186-5
1193-4
1197-1
j Bore washed i
\ Lubricating /
j wads used \
F.T.
114-1
112-9
114-1
114-8
25. In the series of which an abstract, Table 9, follows, are given
the comparative initial velocities of the old (A) and new (Q) pattern
12-pr. shells, both with and without lubricating wads.
It will be seen that while the old pattern shell has, although
scarcely appreciable, a slightly higher initial . velocity, due to the
greater diameter at the back end, the introductiDU of the lubricating
wads adds to the velocity about 15 feet.
ELECTRO-BALLISTIC APPARATUS
37
The effect of the greater diameter at the back end will be again
referred to; but it is interesting to observe that while the initial
velocity is increased by offering, in the first instance, increased
resistance to the motion of the projectile, it is also increased by
diminishing as much as possible the resistance of the friction in its
passage through the bore. The explanation of these results is too
■obvious to require remark.
Table 9.— Abstract of experiments made to ascertain the difference in initial
velocity of old and new pattern 12-pr. shells, vnth and toithout lubricating
wads.
Armstrong
12-pr.
No. of
rounds.
Charge.
Projectile.
Velocity
at
30 yards.
Initial
velocity.
Remarks.
Initial
energy.
Weight.
Diam.
Lb. oz.
Lbs. oz.
F. S.
F. S.
F. T.
No. 224
15
14
1 8
1 8
11 9
11 9
3-072
3-085
1154-2
1157-1
1163-2
1166-1
Q. Pattern shell
lubricating wad
A. Pattern shell
lubricating wad
108-4
109-0
"
13
10
1 8
1 8
11 9
11 9
3-072
3-085
1142-0
1140-6
1150-8
1149-4
Q. Pattern shell
bore washed
A. Pattern shell
bore washed
106-1
105-9
26. Table 10 gives an abstract of the experiments made to com-
pare the initial velocities of shell of the same form and weight, fired
from rifled and smooth-bored 32-prs. of 58 ewt. The ribbed shell
was, in the first case, fired from the rifled gun. Shells of the same
form, diameter, and weight, but with the ribs removed, were then
fired from the rifled gun, and finally similar shells were fired from a
smooth-bored 32-pr.
Table 10.— Abstract of experiments to ascertain the comparative velocities of the
same shell, fired from rifled and smooth-bored 2,2-prs. of 58 cwt.
Nature of gun.
Charge.
Projectile.
Velocity
at
30 yards.
Initial
velocity.
Remarks.
Nature.
Weight.
Diameter.
32-pr., rifled .
32-pr., 58 cwt.
Lbs. oz.
5 8
5 8
5 8
PI. shell
Lbs. oz.
54
54
54
6-350
6-350
6-350
1-215-7
1122-1
1187-4
1224-5
1135-3
1201-7
^ Ribs
I of
j shell
j removed.
38
REPORT ON EXPERIMENTS WITH NAVEZ'S
The velocities in these three cases were respectively 1224-5,
1135-3, aud 1201-7 feet per second. The great difference in velocity
in thu second case is due to the escape of gas by the grooves in the
rifled gun.
27. With the same rifled 32-pr. gun, experiments were also made
to ascertain the reduction in the initial velocity due to an elongation
in the cartridge, and the results of these experiments are here
tabulated.
Table 11. — Abstract of experiments made to ascertain the initial velocities o_f
projectiles fired from a d2-pr. rifled slmnt gun, with charges made up in
cartridges of various lengths.
Nature of gun.
Xo. of
rounds.
Cartridge.
Projectile.
Velocity
at
30 yards.
Initial
velocity.
Charge.
Length.
Nature.
Weiglit.
Diameter
Lbs. oz.
Inches.
Lbs. oz.
Rifled 32-pr.
4
5 8
12
PI. shell
54
6-350
1054-6
1061-7
3
.5 8
9
„
54
6-350
1076-8
1084-2
1
5 8
8
,,
54
6-350
1102-2
1109-8
1
f. 8
7-5
,,
54
6-350
1114-5
1122-3
"
2
5 8
6
"
54
6-350
1187-9
1196-4
From the rapid increase in the initial velocity shown in this
table, due to the reduction in the length of the cartridge, the effect of
the variation in air space in the 12-pr. Armstrong, to which I have
already alluded, will be easily understood.
28. The experience of the preceding practice, together with
theoretical considerations, having pointed to a probable decrease in
velocity should the diameter of B.L. projectiles be diminished or
reduced to that of the bore, the experiments, of which an abstract is
given in Table 12, were undertaken with the object of corroborating
or discovering this view.
From the abstract of this interesting series it will be seen that
while the velocity of the projectiles under normal circumstances was
1248-2 feet per second, when their diameter was reduced to that of
the bore, with the exception of a narrow band at the back end, it
became only 1209-7 feet per second ; and when the diameter was
finally reduced throughout to that of the bore, it was reduced to
11 72 -8 feet. In the rounds fired with the reduced diameters, the
projectiles in all cases appeared to be perfectly steady in flight.
ELECTRO-BALLISTIC APPARATUS
3.)
Table 12.— Abstracts of experiments made to determine the effect on the initial
velocity of diminishing the lead on the 12-pr. Armstrong projectiles.
Armstrong
12-pr.
No. of
rounds.
Charge.
Projectile.
Velocity
at
30 yards.
Initial
velocity.
Remarks.
Weight.
Diam.
Lb. oz.
Lbs. oz.
No. 1050
4
1 8
11 9
3-074
1238-3
1248-2
Shell fired under nor-
mal circumstances.
2
1 8
11 9
3-010
1200-2
1209-7
Same shell reduced
to the diameter of
3-01, with the ex-
ception of a ring at
the base -25 inches
broad.
"
2
1 8
11 9
3-010
1163-7
1172-8
Same shell reduced
throughout.
29. The experiments with the Armstrong 12-pr. having been
chiefly carried on with the same gun, the initial velocities obtained
under similar circumstances become a measure of the variability, in
strength, of the service gunpowder, and it is somewhat surprising to
find so great a variation in powder recently made and professedly of
the same make. For illustration of this remark, I might point to
the differences in initial velocity exhibited in coloured diagram, Figs.
1 and 2, p. 32. In this case, it is true, the results were obtained from
different guns ; but under similar circumstances, these guns were
found to give nearly identical velocities. Another even stronger case,
however, may be taken from the velocities given on different occasions
by the gun numbered 1050. Thus, on the 12th March 1861, with a
service charge of powder marked (A. 4, W.A., 5/9/60, lot 288), the initial
velocity was found to be 1114-8 feet, while under precisely the same
circumstances, on the 15th March 1861, with powder marked (A. 4,
Hall and Sons, 11/7/60, lot 2), the initial velocity was 1248-2 feeb
per second. I may observe that with the Armstrong 12-pr., when
the same powder is used, the variation in initial velocity is very
slight, the extreme difference in ten rounds rarely exceeding 20 feet.
30. On actual service it is obvious that the strength of the
powder may be expected to vary considerably more than is here
indicated ; and I venture to draw the attention of the Select Com-
mittee to this point, as one seriously affecting the precision of
rifled, or indeed of any guns, and as a case in which the electro-
ballistic apparatus might be most advantageously employed.
31. My attention during these experiments was early drawn to
the ranges obtained at P. B., and at small angles of elevation, with
40
REPORT ON EXPERIMENTS WITH NAVEZ'S
the 12-pr. Armstrong. These ranges considerably exceeded those of
the smooth-bored field-service guns, although, of course, the initial
velocity in these latter is very much higher. I therefore took the
usual stejDS for ascertaining the " angle of departure," and, as much
additional trouble was not entailed, I also made arrangements for
ascertaining the ordinates at various points of the trajectory. It
will be seen by these observations that the angle of projection of a
projectile fired from a 12-pr. gun, accurately laid with its bore
horizontal, varied from 0°23'30" to 0"28'28", the mean angle of
projection being 25'33" ; while in the same gun fired with an eleva-
tion of 30', the angle of projection varied from 47'0" to 49'6", the
mean angle being 48' 18".
In coloured diagrams, Figs. 3, 4, and 5, p. 32, 1 have laid down, for
the information of the Committee, the mean results of this practice,
the observed trajectories being denoted by black, the computed by
red, and for the sake of comparison I have also shown, in blue lines,
the departure of both curves from the parabolic.
The annexed abstract. Table 13, will show how close is the
agreement between the computed and observed ordinates in the
curves delineated ; while a similar comparison for the majority of
the curves observed is made in the detailed report of the practice
furnished to the Committee.
Table IS.— Abstract of the results of the experiments made to ascertain the angle
of projection and the trajectories of the 12-pr. Armstrong projectiles when
fired P.B., and at an apparent elevation of 30'.
■J d
ft
Ordinates at
fl
II
90 feet.
150 feet.
300 feet.
450 feet.
Obs.
Com.
Obs.
Com.
Ob.s.
Com.
Obs.
Com.
30
25 44
25 55
48 35
F. S.
1188-1
1170-7
1179-6
F. S.
1197-5
1179-8
1188-9
4-832
4-834
5-431
4-832
4-832
5-441
5-083
5-085
5-097
5-118
5-444
5-472
7-421
5-459
5-443
7-44
5-2
5-202
8-359
5-255
5-207
8-329
Ordinates at- Co)
tinvnl
600 feet.
750 feet.
900 feet.
1050 feet.
1200 feet.
1355 feet.
Obs.
Com.
Obs.
Com.
Obs.
Com.
Obs.
Com.
Obs.
Com.
Obs.
Com.
4-527
4-425
8-465
4-483
4-359
8-414
3-417
2-75
8-133
3-133
2-97
8-025
-869
7-089
1-003
7-056
...
5 •4-28
3-033
0-45
0-01
ELECTRO-BALLISTIC APPARATUS
41
32. The ordinates in Table 13 were calculated on the hypothesis
that the resistance of the air was given by the equation
8 r, . V '\
resistance
.0005213.RV^^{l ^ ^
1426-4J
The accordance of the ordinates calculated on this hypothesis are,
on the whole, exceedingly close ; but it would be unwise to place too
much dependence on the results of experiments so partial, and
carried on at such low angles.
33. The following table gives an abstract of the results of
several miscellaneous experiments : —
Table 14.
Nature of
gun.
Charge.
Projectile
Velocity
at
Initial
velocity.
Remarks.
1
Nature.
Weight.
Diam.
30 yards.
Lb. oz.
Lbs. oz.
9-pr., brass!
13J cwt. /
14
1 8
Shot
9 5
9 5
4-080
4-080
1011-2
1310-9
^At 25 yards.
Experiments to
ascertain the
12-pr., ]
Si cwt., y
Armstrong]
1 2
S. shell
9
3-074
1130-0
1141-2
initial velocity
of a 9-pr. shell
with a charge
I of 1 lb. 2 oz.
Experiments to
ascertain the
r2-pr., ]
initial veloci-
6 cwt, V
1 6
^,
11 9
3-074
1103-4
1111-8
- ties of 12-pr.
Armstrong j
shells fired
from a 12-pr.
gun of 6 cwt.
12-pr., !
8^ cwt, 1
Armstrong!
Experiments to
ascertain the
1 8
"
11 9
3-084
1127-6
1136-3
difference in
initial velocity
and regularity
Armstrong (
No. 224 j
1 8
11 9
3-084
1141-8
1150-6
of two 12-prs.,
the first of
which had been
exposed to the
weather for
, several weeks.
Experiments U
) ascertain the initial velocity of the old pattern (25 I1>s. )
projectiles fired from 2Q-pr. guns.
Lbs. oz.
Lbs. oz.
Long25-pr.
Armstrong
R. G. K. r
No. 384
2 8
S. shell
25
3-830
963-8
968-8
Without lubri-
cating wads.
2 8
,,
25
3-830
1014-4
1019-8
"1
2 13
,^
25
3-830
1083-3
1089-3
1
3 2
,,
25
3-830
1136-1
1142-5
(Lubricating wads
Short 25-pr.]
( used.
Armstrong I
2 6
,,
25
3-830
874-5
878-9
No. 403 1
^
III.
ON THE EATIO BETWEEN THE FOECES TENDING TO
PEODUCE TEANSLATION AND EOTATION IN THE
BOEES OF EIFLED GUNS.
{Philosophical Magazine, Septemhcr 1863.)
The magnitude which the rifled ordnance of the present day have
attained, and the large charges which are consumed in their bores,
render it an object of great interest that we should be able to assign
the pressures on the grooves (or other driving-surfaces intended to
give rotation) due to different modes of rifling, as well as to determine
the increment in the gaseous pressure arising from the nature of
rifling adopted.
The formulae which I shall hereafter give, have, with slight
modifications, been used at Elswick for nearly three years, and are
now given, partly because no investigation of
Fi.4. 1. the question has, to my knowledge, been
^ „ published, and partly because, as several
erroneous statements on the subject have
appeared, the formulse themselves may pos-
sibly be of use to some artillerists.
The case we shall first examine will be
that in which the rotation is given by means
of grooves, the driving-surfaces of which are
such that if a section of the gun, perpen-
dicular to the axis, be made, the line drawn
from the centre of the bore to the groove is coincident with the
section of the driving-surface. A section of such a form of rifling
is shown in Fig. 1. The reader is supposed to be looking from the
muzzle towards the breech of the gun, and the direction of rotation
is shown by the arrow AB.
42
BORES OF RIFLED GUNS
43
It will be seen that the radius CD is coincident with the section
of the driving-surface DP.
In entering upon this investigation, it will be more convenient to
consider the projectile in its motion along the bore of the gun as
moving on a fixed axis, and, further, to suppose that the motion of
rotation is communicated to the projectile by a single groove. These
suppositions will not interfere with the accuracy of our results, and
will enable us very much to simplify the equations of motion.
Take (Fig. 2) as the plane of xy, the plane passing through the
commencement of the rifling at right angles to the axis of the gun.
Let the axis of x pass through the groove under consideration, and
let the axis of z be that of the gun. Let
AP be the helix, and let (see Figs. 1 and 2)
P {xyz) be the point at which the resultant
of all the pressures on the groove may be
assumed to act, the projectile being in a
given position. Let the angle AON = ^.
Let us now consider the forces which
act upon the projectile. We have, first,
the gaseous pressure acting on the base
of the shot. Let us call this force, the
resultant of which acts along the axis of
z, G. Secondly, if E be the pressure
between the projectile and the groove at
the point P, this pressure will be exerted
normally to the surface of the groove, and
if we denote by X, [x, v the angles which the
normal makes with the co-ordinate axes, the resolved parts of this
force will be
E . cos X, E . cos /x, E . cos v
Thirdly, if ^^ be the coefficient of friction between the rib of
the projectile and the driving-surface, the force /xiE will tend to
retard the motion of the projectile. This force will act along the
tangent to the helix which the point P describes ; and if a, j8, y
be the angles which the tangent makes with the co-ordinate axes,
we have as the resolved portions of this force /x^E , cos a, fi^R . cos ^,
^,E
act
cos y ; a
nd summing up these forces, we have the forces which
parallel to x
parallel to y
parallel to z
X = R j cos A - /x^ cos a ]
Y = R { cos /x - /Xj cos (3 ]
Z = G + R { cos V - /i^ cos y
I
(1)
44 TRANSLATION AND ROTATION IN THE
and the equations of motion will be
M.
d^_
G + II . { cos V - /«,j cos y }
■ (2)
• (3)
p being the radius of gyration.
We proceed to determine the value of the angles a, ^, y, X, m.
V. Let the equations to the helix described by the point P be
put under the form
r . cos <^, y = r sin (j>,
■ krcf>
(4)
]c being the tangent of the angle at which the helix is inclined to
the plane of xy. Then
dx-
and
r sin (f}d(f), dij = r cos
dx - sin (^ '
s a = -f- = —
ds
dz = krd^
cos/3
cosy
cos (^
A:
JT+lfi
(5)
To determine the values of X, //, f, we shall first seek the
equation to the driving-surface of the groove. In the case under
consideration, the surface is a well-known conoidal one, the " skew
helicoid," and is familiar to the eye as the under surface of a
spiral staircase. It is generated by a straight line which passes
through the axis of z, always remains perpendicular to it, and
meets the helix described by the point P. The equations to the
director being given in (4), if x-^, y^, %^ be the current co-ordinates
of the generator, its equations are
Hence
•Vy-^r^ = 0. "1 = ^
(6)
and the equation to the surface is
BORES OF RIFLED GUNS
or, dropping the suffixes,
y . cos -^ - X .s\n~ = Q
^ kr kr
45
. (7>
Now X, n, V being the angles which the normal to (7) makes with
the axes,
Now
cos A
cos /A
dx J
{(57) + («;) + W) }
. c /dF\ z
1 r X 2 . .V
— cos -r- + -— . sm -7-
r
ytr
£.1
(8)
but since in the case we are now considering (xyz) is a point
both in the surface given by Equation (7) and in the directing
helix, we have from (4),
and
Hence
.y
kr'
kr
+ F
cos A
COS/X
k sin (f>
Jl+k-^
k . COS (f>
1
. (9)
46 TRANSLATION AND ROTATION IN THE
Now substituting the values of the direction cosines given in
Equations (5) and (9), in (1), (2), and (3), we have as the equa-
tions of motion,
(11)
d'^(f) B.r k — [X
and hence the normal pressure on the rib of the projectile,
'• A' - /Xj * dV-
But if ft) be the angular velocity of the projectile, and h be the
pitch of the rifling, we have the following relation between the
velocities of translation and rotation.
Hence
and
A/f„2 h , h2 O^ r12r,
. . . (12)
Now, substituting in this equation the value of -^, derived from (10),
we have
rh k-[x-^ K sjl+k- J
dc}>
'' dt
27r
27r
T
'dt
d-4>
27r ^2,
" h ' dV^
J
R
_ Mp2
r
Jl+k'^
k-a,
27r
T'
dh
dF^
R _ 2Trp'^Jl+k-^
G^ ~ hr{k-iJL-^) + 27rp\[ji,^k+l)
(13)
And this equation gives the ratio between the pressures produc-
ing translation and rotation.
We now proceed to determine the increment of the gaseous
pressure due to the resistance, etc., offered by the rifling to the
forward motion of the shot. We shall imagine a smooth-bored
gun to fire a shot of the same weight as that of the rifled gun.
We shall further suppose that the two projectiles are delivered
with the same velocity; and we wish to know, the same ballistic
BORES OF RIFLED GUNS
47
effect being produced by the two guns, what is the incn
pressure which the rijEled gun has had to sustain. Now the equa-
tion of motion in the case of the smooth-bored gun is
and in the case of the rifled gun,
r^2^ ^, R
M-
df^
G'
TTTP^^^'^^^
(14)
(15)
Now, if the velocity-increments in the two cases be taken as
equal, we shall have, from Equations (14) and (15),
G' == G+ JL_(/.,^+l) . . . (16)
v 1 + /(•"
And the second term of the right-hand member of Equation (16)
represents the increment of pressure due to the rifling.
Let us now examine the pressures which subsist when a polygonal
form of rifling is adopted; and we
shall suppose the polygon to have n
sides.
The equations of motion given in
Equations (2) and (3) hold here as in
the last case, and the values of a, ^, y
given in (5) remain the same. The
driving-surface is, however, different,
being traced out by a straight line
which always remains parallel to the
plane of xy, meets the helix described
by P, and touches the cylinder whose
radius is = r cos — (see Fig. 3, where PA represents the generating
line drawn from a point P of the helix to touch the cylinder BC)
Now the equations to the helix being
x = rcos<^, i/ = rsin to (18), the co-ordinates of the point of contact (see
Fig. 3) will be
X-. = r, . cos ( - -'— )
) l{\ ■ ■ ■ ■ (1")
Now the equation to the tangent drawn through the point x-^y-^
of the circle x^ + i/ = r^ is
xx,+yy^ = r^^ .... (20)
And substituting in this equation the values of x^ and y-^^ derived
from (19), we obtain as the equations of the generator,
ar^.cosr<^ - — j+/yr^.sinf<^ - -^j = r^-, £ = krcf> . (21)
and as the equation to the driving-surface,
Now
(f ) = ^°' (& - v)' (f ) = >"" (1^ - v)
(s)^iR- + (sin^)^
sm —
n
(23)
BORES OF RIFLED GUNS
49
And putting the values of a, /3, X, y, ix, v in the equations of
motion (2) and (3), we have
d^ Rr
k cos
2 vn:^r"A^
Vl+A-2 ^ / /. ^\2r At
A- . sin —
Rr I n
Hence
M.p2
2 Vl+F
But
d^ _ irr d^z
d^
' IF
(24)
[ . (25)
and making the necessary substitutions, we obtain for the ratio
between the forces producing rotation and translation,
27rp2
(26)
/"■i
Vi+F
{2Tr p-k-rh) +
sjk^ + (sin f )'
ZTT/D-Sin
r/i/r
In precisely the same manner as in the former case, and on the
same hypotheses, we may show that if G" denote the gaseous
pressure in a bore rifled on the system we are now considering, and
D
50 TRANSLATION AND ROTATION IN THE
G denote the gaseous pressure in a similar smooth-bored gun, we
shall have
r" r ^ T? /"i^ 4- ^^^ "" — . . (27)
Hence if we have three guns of the same diameter of bore, viz.,
a smooth-bore gim ; a rifled gun, the grooves of which are similar to
those shown in Fig. 1 ; and a third, rifled polygonally ; and if we sup-
pose that the shot in each case are of the same weight, and, further,
that in each case the velocity-increments at the moment under con-
sideration are equal, then the pressures upon the base of the shot
will be as follow : — In the case of the
Smooth-bored gun, pressure = G
First rifled gun, pressure = G + —t==={ix-Jc+\)
Jl+k-^
Polygonally-rifled gun, pressure
G + R
• ^ 1
sin —
(28)
We shall now give examples of the cases we have been discussing
to exhibit numerically the above results.
Let us suppose that two 7-inch guns are rifled — the first accord-
ing to the method shown in Fig. 1, with a pitch of one turn in 294
inches, the other octagonally, with a pitch of one turn in 130 inches.
It is required to determine in each case the pressure on the driving-
surface in terms of the pressure on the base of the shot. Now, in
the first case, from (13),
Pressure on driving-surface = --— — ^^^-^ — - — - — ; — ->. . G
hr (k - /*i) + 27rp\fji^k +1)
where
7r = 3-14159, p = /-v/J= 2-475, /r=l.S-3697, A = 294, r=3-5, //i = -1666
whence we obtain
R = -0375 G (29)
In the second case, from (26),
Pressure
= ^-X G
VF + (sn, -)
where
BORES OF RIFLED GUNS 51
TT = 3-14159
= 2'350 (c = length of side of polygon)
1 - cos —
2 + cos —
1 .. n
12
yi: = 5-9117, // = 130, r = 3-5, n = 8, /*, = -1666, — = 22° 30"
n
whence R = -1706 . . . . (30)
That is, on the supposition of the same pressure on the base of the
shot, the pressure on the driving-surface is in the latter case nearly
five times as great as in the former, and is, in fact, no inconsiderable
fraction of the propelling force.
Let us now compare the gaseous pressures on the base of shot of
the same weight supposed to be fired from the guns above described,
and from a smooth-bored gun. From Equations (28) we have the
pressure upon base of shot fired from
Smooth-bored gun . . . = G
First rifled gun . . . . = 1-009 G
Polygonal gun . . . . = 1-041 G
In these calculations we have taken the coefficient of friction = \.
It is necessary, however, to observe that very little is known concern-
ing the value of this constant at pressures so high as those with which
we have here to do. It is evident that in the case of the contact of
similar metals, when the point of seizure is approached, the coefficient
of friction cannot be considered independent of pressure ; and it is
probable that when the rubbing surfaces of both projectile and groove
(or other driving-surface) are of the same hard material, the coefficient
of friction may be occasionally enormously increased.
The resistance due to this cause might under certain circumstances
be sufficient to ensure the destruction of the gun ; and this view is to
some extent corroborated by the occasional bursting of guns, the
failure of which it is difficult to attribute to any other cause ; and in
the instances referred to, the recovered fragments of the shot were
thought to exhibit decided appearances of seizure.
If in Equation (26) we substitute ^ for — , we shall have
1 = ^^^' . (31)
52
BORES OF RIFLED GUNS
Fig:. 4.
And this equation will represent the ratio of the pressures E and
G in any system of rifling, S being the angle which the radius makes
with the normal to the driving-surface. Thus in an elhptically-
bored gun (see Fig. 4) the angle OPQ represents the angle §, and we
obtain ^ by substituting in (31) the value of
this angle ; by putting S = 90°, we may derive
Equation (13) directly from (31).
We have not in this note entered into the
question of the absolute pressures existing in
the bores of ordnance of various natures, as
the subject is too extensive and of too great im-
portance to be disposed of within the limits of
a short paper.
Artillerists acquainted with the subject will
be able to form rough approximations to these
pressures from the experiments made abroad
with smooth-bored gims, with a view to the
elucidation of this important question. It is much to be regretted
that no experiments of the nature referred to have been attempted
in England under G-overnment auspices, as they are of a descrip-
tion which precludes their being satisfactorily made by private
individuals, and as the information to be derived from them would
be especially important in the case of rifled cannon, where so
many new conditions are introduced into the problem as to render
previous investigations of but little value.
We shall, however, in a future note endeavour to discuss this
subject, making use of the data at our disposal.
IV.
ON THE TENSION OF FIEED GUNPOWDER
{Trmisactions of the Royal Institution, 1871.)
Befoee entering on the investigations which will be the chief subject
of my discourse this evening, I find it necessary to give a sketch of
the means that have hitherto been adopted to determine, and the
views that have been entertained concerning, the pressure of fired
gunpowder.
The first attempt made to explain the action of gunpowder was,
I believe, that of M. de la Hire, who, in the History of the French
Academy for 1702, ascribed the force of fired gunpowder to the
behaviour of the air enclosed in and between the grains of powder.
This air he considered to be highly heated by the combustion of the
charge, and the consequent elasticity to be the moving force of the
projectile. Eobins, who followed M. de la Hire as the next writer
on the subject, and who may be considered to have laid the founda-
tion of this, as of so many other departments of artillery science,
points out how inadequate to the effect are the forces supposed to
act by M. de la Hire.
He himself instituted a carefully-planned and well-conducted
series of experiments, in which he determined the quantity of
permanent gas generated by the explosion of gunpowder ; adduced
experiments which he considered to prove that this quantity is the
same whether the powder be exploded in the air or in vacuo ; and
finally determined the increase of elasticity due to the supposed
temperature of the explosion.
The conclusions at which Eobins arrived were briefly as follow :
— 1. That the whole action of the powder on the projectile was due
to the permanent gases generated by the explosion. 2. That at
ordinary temperature and atmospheric pressure the permanent gases
occupied about 240 times the volume of the unexploded powder.
54
ON THE TENSION OF FIRED GUNPOWDER
8. That the heat of combustion increased this volume to about 1000
times that of the powder, and that hence the maximum force of
gunpowder — somewhat less with small, somewhat greater with large
charges — was about 1000 atmospheres, that is to say, about 6^ tons
on the square inch.
But although Eobins considered this pressure the maximum
exerted by fired gunpowder, it is worthy of remark that he recognised
Fig. 1.
the intensity of the local pressure which arises when the gases
generated have space sufficient to acquire a considerable velocity
before meeting with an obstacle. In a common musket he placed a
bullet 16 inches from the charge, and found that at the seat of the
shot the barrel was bulged like a bladder to twice its original dia-
meter, while two pieces were blown out of it.
But the first regular experiments which had for object the
determination of the pressure of gunpowder fired in a close vessel or
ON THE TENSION OF FIRED GUNPOWDER 55
chamber were those of Count Rumford, made in 1793, and published
in the Transactions of the Royal Society for 1797.
The apparatus used by Count Rumford is figured in this diagram
(Fig. 1), and will be readily understood. V is a small but strong
wrought-iron vessel resting on the pedestal P, and having a bore of
:|-inch diameter. The bore is closed by the hemisphere E, upon
which any requisite weight may be placed. There is a closed vent,
V, which is filled with powder, and the charge is fired by means of a
red-hot ball, B.
The modus operandi was as follows : — A given charge being placed
in the bore, a weight which was considered equivalent to the gaseous
pressure was applied on E. If the charge of powder lifted the weight
and let the gases escape, the weight was increased until it was just
sufficient to confine it, and the pressure represented by the weight
was assumed to be that of the powder.
The powder used was sporting, of very fine grain, and it is to be
remarked that its composition, there being only 67 per cent, of
saltpetre, differed notably from ordinary powder. The charges used,
moreover, were very small, the maximum being only 18 grains. In
one case, indeed, the vessel was filled : about 28 grains were necessary
to fill the chamber ; but by this experiment the vessel was destroyed.
The objects Rumford had in view were — first, to ascertain the limit
of the force exerted by the exploded powder when the gases are at
their maximum density ; secondly, to determine the relation between
the density of the gases and the tension.
The curve shown here (Fig. 2) exhibits the results of the first
and most reliable series of Rumford's experiments, and you will
observe how nearly, up to charges of 15 grains (60 per cent.), the
curve, which is expressible by the empirical equation 2/ = a;^+*^°"'*^,
passes through the observed points. Were this law assumed to be
true up to the point of maximum density,* it would give the
maximum tension at about 29,000 atmospheres, or 191 tons on the
square inch. But, great as this pressure is. Count Rumford considers
it much below the truth. In addition to the experiments graphically
represented by the diagram to which I have drawn your attention.
Count Rumford made a second series, the results of which, to use his
own words, " are still more various, extraordinary, and inexplicable."
From this diagram you will observe that the tension of the gas in
the first series of experiments was with 12 grains of powder about
2700 atmospheres ; but in this second series the pressure with the
* Considered as unity.
56
ON THE TENSION OF FIRED GUNPOWDER
same charge is repeatedly found to be above 9000 atmospheres.
Count Eumford does not attempt to explain the enormous discrep-
ancy between the two sets of experiments, unless a remark on the
heat of the weather during the second set can be so considered ; but,
relying on this second series, and on the experiment in which the
vessel was destroyed by 28 grains. Count Eumford arrives at the
conclusion that 101,021 atmospheres, or 662 tons on the square inch,
is the measure of the initial force of the elastic fluid generated by the
combustion of gunpowder. Eumford meets the objection that, if the
Fig. 2.
10 20 30 40 50 60 70 80 90 100 PARTS
tension were anything like that he names, no gun would have a
chance of standing, by assuming that the combustion of the powder
is much slower than is ordinarily supposed, and, indeed, lasts all the
time the shot is in the bore ; and he further accounts for the enormous
initial tension by ascribing it to the elasticity of the aqueous vapour or
steam contained in the powder. Supposing, from M. de Betancourt's
experiments, that the elasticity of steam is doubled by every addition
of temperature equal to 30° Fahr., his only difficulty, and one
which he leaves to his successors to explain, is why the steam
liberated by the combustion of the powder does not' exercise a much
higher pressure than the 100,000 atmospheres he has assigned to it.
ON THE TENSION OF FIRED GUNPOWDER 57
In 1843 Colonel Cavalli proposed to insert in the bore of a gun a
series of small barrels, intended to throw a wrought-iron spherical
ball. By ascertaining the velocities of these balls Colonel Cavalli
considered that he would be able to assign the corresponding pres-
sures. Colonel Cavalli's plan was actually carried out, and from his
experiments he deduced what ought to be the theoretical thickness of
the metal at various points along the bore. But a very great
improvement on Colonel Cavalli's method was introduced in 1854 by
a Prussian Artillery Committee, under the direction of General (then
Major) jSTeumann.
The plan adopted by the Prussian Committee was as follows : —
In, say, the centre, or any other point desired, of the powder
chamber, a hole was drilled, and in this hole was fitted a small gun-
barrel with a calibre of about yo of an inch and a length of, say, 8
inches. Now, suppose the gun to be loaded, and suppose further
that in the small side gun we place a cylinder whose longitudinal
section is the same as that of the projectile. On the assumption
that the pressure throughout the powder chamber is uniform, the
cylinder and the projectile will in equal times describe equal spaces,
and after the cylinder has travelled 8 inches it will be withdrawn
from the action of the gas. If, then, we ascertain the velocity of the
cylinder, we shall know that of the projectile when it has described
in the bore a space of 8 inches. Again, if we make the section of
the cylinder half that of the projectile, it will describe in the same
time double the space, and will have acquired double the velocity,
and so on; so that, for example, if the section of the cylinder be
one-eighth that of the projectile, and we ascertain the cylinder's
velocity, we know the velocity of the projectile after it has described
1 inch.
These Prussian experiments do not, however, despite the ingenuity
of their method, possess a very high interest to us, as they were
applied only to comparatively very small guns, the 6-pr. and 12-pr.
smooth-bores, and had for their chief object the comparison between
elongated and non-elongated cartridges.
Further on I shall advert to reasons which prevent this method
being altogether reliable, especially for large guns ; but I may state
that the general result seems to have been that in the 6-pr. the
maximum pressure was about 1100 atmospheres, while in the 12-pr.
it was nearly 1300 atmospheres.
I shall also further on advert to another remarkable observation
made by the Prussian Committee— namely, that in every charge
58
ON THE TENSION OF FIRED GUNPOWDER
Fig. 3.
with which they experimented two maxima of tension were distinctly
perceptible.
The distinguished Eussian artillerist, General Mayevski, who has
written an elaborate memoir on the pressure in the bores of guns,
founded on these experiments, con-
firms the results at which the
Prussian Committee have arrived,
and points out that from the ex-
periments the maximum pressure
must be attained before the bullet
is any considerable distance from
its initial position.
General Neumann's method ap-
pears to have been repeated in
Belgium about the year 1860 with
a 70-pr. rifled gun. I have not
seen a detailed report of these
trials, but the maximum pressure
with ordinary powder was stated
to be about 3000 atmospheres, or
nearly 20 tons per square inch.
In 1857-8-9 Major Eodman car-
ried on for the United States a
most interesting and extensive
series of experiments on gunpowder.
The celebrity to which Major
Rodman's ingenious instrument has
attained, the great use which has
been made of it in Europe, and the
fact that he appears to have been
the first person who experimented
on the effect of size of grain, and
proposed prismatic powder, oblige
me to describe both his instrument
and his experiments in some detail.
It is most unfortunate that experiments so well devised, and
carried out with so much care, should be rendered in many cases
almost valueless by the absence of important data, by the admission
of manifestly erroneous observations, and, finally, by results passed
over in silence which are not only frequently anomalous, but in some
cases absolutely impossible.
ON THE TENSION OF FIRED GUNPOWDER 59
The instrument which Major Rodman devised is shown in this
drawing (Fig. 3). Suppose we wish to determine the pressure in the
chamber of a gun. A hole is drilled into it, and a cylinder with a
small passage down its centre is inserted. To this cylinder is fitted
the indicating apparatus, which consists of the indenting tool g,
carrying a knife, shown in elevation and section. Against the knife
is screwed a piece of soft copper, h. You will have no difficulty in
understanding the action of this apparatus. The pressure of the gas
acting on the base of the indenting tool causes a cut in the copper,
and by mechanical means the magnitude of the force capable of pro-
ducing a similar cut can be determined. A small cup at c prevents
any gas passing the indenting tool, and the channel e provides for
the escape of gas, should any pass on account of defective arrange-
ments.
Major Eodman's first series of experiments of importance was the
determination of the pressure at different points of a 42-pr. smooth-
bored gun, two descriptions of cartridges being used — one being made
up with 10 lbs. of ordinary grained powder, the other being what he
terms an accelerating cartridge of 13 lbs., a description of which is
not given.
Major Rodman gives the mean results of this series in a tabulated
form, but I have transferred his results to this diagram (Fig. 4), and
I draw your especial attention to them. You will notice that among
the observed points I have drawn in each case a curve representing,
as nearly as may be, the observations. Remark how widely the two
curves differ. The horizontal line, the axis of abscissse, represents
the length of the bore, and by the length of the ordinates is indicated
the maximum amount of pressure existing at any particular point of
the bore.
These curves illustrate also another point. Since the ordinates
represent the pressures, and the abscissae the travel of the shot along
the bore, the areas, that is to say, the spaces between the curves and
the axis of abscissse, represent the total work done on the shot by
each of the charges experimented with. Your eye will tell you that
the area, that is the work done on the shot, is, in the case of the
grained, nearly double its amount in that of the accelerating cart-
ridge, but the actual work in each case was known to be nearly
identical.
There is here, therefore, a grave contradiction, which requires
explanation. But we have not done yet. Knowing, as we do
from these curves, the amount of the work done by each nature of
60
ON THE TENSION OF FIRED GUNPOWDER
cartridge on the shot, we are in a position to compute the velocity
with which the shot would quit the bore.
Performing this calculation, we find that the lesser area repre-
sents a muzzle velocity of about 1950 feet, while the larger one
represents a muzzle velocity of about 2620 feet— results differing
widely from the truth, and showing that the larger of the two areas
is about three times greater than it should be, while even the smaller
is at least 50 per cent, too high.
Two interesting series were fired from the same gun to determine
the pressure on the bottom of the bore when the weight of the charge
Fig. 4.
12 14
CALI BRES
was varied, that of the shot remaining constant, and when the weight
of the shot was varied, the charge remaining constant.
As far as the experiments were carried, the pressure in both
cases appeared to be nearly directly proportional — in the one
instance to the weight of the shot, in the other to the weight of the
charge.
Experiments were then made to determine the pressures in guns
of 7-inch, 9-inch, and 11-inch bore, and were so arranged that in each
gun an equal column of powder (that is, an equal weight of charge) was
behind an equal column or weight of shot. It is hardly necessary to
point out that in each gun, in the motion of the shot along the bore,
at every point, the gases would be equally expanded, and any incre-
ON THE TENSION OF FIRED GUNPOWDER
61
ment of pressure in the larger-bored guns would be attributable to
the use of the larger charge.
The mean result of these experiments is given in this diagram
(Fig. 5). As before, there are many anomalies and contradictions in
the experiments themselves. You will observe what a great increase
of pressure is credited to the larger guns, although the same column
of powder and shot exists in all cases. As before, again the work
done on the shot as indicated by these areas is enormously too
large.
The results given by these experiments are the more curious.
Fig
. 5.
40-|
A \ ^°"
X
o
z
or
(0
N
w
z.
o
-^-IlGuji
-
14
28
because, as Major Eodman himself points out, they are entirely at
variance with some subsequent experiments, in which charges of
powder of various weights, from 1700 to 11,000 grs., occupying
always one-fourth of the space in which they were fired, and the
charge escaping through the vent, gave pressures practically
identical.
The effect of the size of the grains was the next subject investi-
gated. The comparative results (care still being taken not to accept
these areas as representing the work done on the shot) are obvious
from a glance at Fig. 6 ; and Major Eodman arrives at the conclusion
that the velocities due to our present charges of small-grained powder
may be obtained with a greatly diminished strain on the gun by the
62 ON THE TENSION OF FIRED GUNPOWDER
use of powder properly adapted in size of grain to the calibre and
length of bore with which it is to be used.
With this statement I entirely agree, and can only regret that,
from the absence of information as to density and other particulars
of the various samples of powder used, these particular experi-
ments have been of no use to us in this country for comparative
purposes.
The only other experiments of Major Eodman to which I shall
draw your attention belong to a series which I am able to compare
with the experiments of Count Kumford, as to the pressure of fired
TONS
20-
Fig. 6.
DIAR OF GRAIN
5_
70 84
INCHES
gunpowder in various degrees of expansion — that is, the unfired
powder occupying a definite proportion of the space in which it is
exploded. Fig. 7 is a drawing of the apparatus Major Rodman used.
You will observe that in this apparatus the fired charge escapes
through the vent, while in Count Eumford's experiments the products
of explosion were generally more or less confined.
On the other hand, Count Eumford's charges were exceedingly
minute, while the charges we are now considering ranged from 700
to 7000 grs.
On the same diagram (Fig. 2) upon which I placed Count
Eumford's results I have placed Major Eodman's. You will per-
ceive the difference between them. But Major Eodman's experi-
ON THE TENSION OF FIRED GUNPOWDER
63
Fig. 7.
ments have not been carried far enough to possess for us much
interest.
Major Kodman, like Count Kumford, endeavoured to ascertain
the maximum force which powder was capable of exerting when fired
in its own volume. Major Eodman fired various charges in enor-
mously strong shells, through a small vent yo inch in diameter. He
considered, from some experiments with which I need not trouble
you, that in all cases the maximum pressure would be exerted before
the shell burst. His results, however, were very diverse, varying
from 32 tons per square
inch (4900 atmospheres) to
82 tons, or about 12,400
atmospheres, and, singularly
enough, the highest pressure
was given by the smallest
charge ; from the great
discrepancies, as well as
from other considerations,
I do not think we can
accept these determinations
as entitled to much weight.
Bunsen and Schisch-
koff's experiments, both
from their completeness,
and the eminent position of
the distinguished chemists
who conducted them, may
justly rank among the most
important which have been
made on our subject.
They were directed, in
the first place, to determine the exact nature, both of the permanent
gases and the solid products generated by the explosion of powder ;
secondly, to determine the heat generated by the act of explosion ;
thirdly, to determine the maximum pressure which gunpowder fired
in a close chamber would give rise to ; and, finally, to determine the
total quantity of work which a given weight of gunpowder is capable
of producing.
The apparatus adopted for obtaining the products of com-
bustion was so arranged that the powder to be analysed falls in
a very finely-divided stream into a heated bulb, in which, and in
Y )
64
ON THE TENSION OF FIRED GUNPOWDER
tubes connected with it, the resulting products are collected for
examination.
MM. Bunsen and SchischkofF, in drawing attention to their
results, and the extraordinary difference between their estimates and
those given by so eminent an authority as Piobert, point out that
many of the assumptions previously made must depend on very
faulty premises ; but their own experiments have not altogether
escaped attack, and I think we are bound to receive some of their
results with great reservation, until it can be demonstrated that the
products of combustion are the same in the bore of a gun as when
produced in the method followed in these experiments.
I shall not detain you with the results of their analysis, which
you see, however, in this table,* and shall only point out that the
permanent gases at a temperature of 0° C. and pressure of 760 mm.
occupied a volume 193 times greater than that occupied by the
powder, and represented about fVo the weight of the powder. The
remainder was solid residue, and MM. Bunsen and Schischkoff
conceive that, although a portion of these solid matters may un-
doubtedly be volatilised by the high temperature of the explosion
yet any pressure which may be exerted by such vapours is quite
insignificant. This opinion appears to be founded on the fact that
the solid residue arising from the explosion of gunpowder is not
fused when exposed to the action of a jet of inflamed hydrogen.
Piobert and other authorities, on the other hand, consider that
the pressure exerted by the volatilised residue has far more influence
on the pressure than the permanent gases.
* Transformation experienced by gunpowder in burning, after Bunsen and
Schischkoff.
'Nitre .
Sulphur
a Charcoal
. 0-7899
. 0-0984
( C 0-0769
J H 0-0041
to 0-0307^
r Residue
Gases
0^3138
0-9944
fKOSOs
0-4227
KOCOo
0-1264
KOS„0„
0-0327
KS '.'
0-0213
KC.S^
0-0030
KONO-,
0-0372
C . .
0-0073
S . .
0-0014
UnH,0,3C0
0-0286
grm.
c.c.
(N . . 0-0998
= 79-40
CO., . 0-2012
= 101-71
CO" . 0-0094
= 7-49
H . . 0-0002
= 2-34
HS . 0-0018
= 1-16
.0 . .
-00
14
= 1-00
ON THE TENSION OF FIRED GUNPOWDER 65
The temperature of the fired gunpowder was determined by
exploding a small charge of powder enclosed in a tube, which was
itself immersed in a larger tube containing water. From the
increment of temperature communicated to the water by the
explosion, it was found that one part of fired powder would raise 620
parts of water by 1° Cent., and hence it was calculated that the
temperature of gunpowder fired in a close chamber impervious to
heat is 3340° Cent., or 5980' Fahr.
Assuming, first, that the products obtained in the two methods
I have just described are identical, and, secondly, that no variations
in the products arise from the combustion of large charges, this
result would be very near the truth.
The pressure in a closed vessel is readily deducible from the
above data, and MM. Bunsen and Schischkoff compute that the
maximum tension which the gas can attain — to which it may
approximate, but can never reach — is about 4374 atmospheres, or
about 29 tons on the square inch.
I shall shortly have occasion to show that this pressure has
been undoubtedly reached in the case of heavy guns, and con-
siderably exceeded in the case of powder fired in closed vessels.
MM. Bunsen and Schischkoff also compute, from their data, the
theoretical work of a kilog. of gunpowder at 67,400 kilo-
grammetres, that is 67,400 kilogs. raised 1 metre in height.
The Committee on explosives have, however, realised in the shot
alone nearly 60,000 kilogrammetres per kilog. of powder in a
comparatively short gun ; and it may therefore be conjectured that
this estimate, like that of the maximum pressure, is considerably too
low, although undoubtedly much nearer the truth than the extrava-
gant estimates which have frequently been made.
In the year 1861-2 Sir W. Armstrong, in conjunction with myself,
made several experiments to determine the maximum pressure of
powder in the bores of what were then considered very large guns
the 110- and the 70-prs. Two methods were adopted, and although
they, like nearly every experiment connected with gunpowder, gave
results in some degree anomalous, yet the conclusion at which we
arrived — namely, that the maximum pressure with the powder
then used, in the bores of the guns I have mentioned, was about
17 tons on the square inch— is probably not very far removed from
the truth.
The first of these methods consisted of an arrangement carried
in the nose or front part of the projectile, and is shown in these
E
66
ON THE TENSION OF FIRED GUNPOWDER
drawings (Figs. 8 and 9). The apparatus itself consisted of a case
containing seven little cells, Ih. Each of these cells contained a
small pellet, a, of the same weight, and each of these pellets is
retained in the front portion of the cell 1)y means of a small wire.
Fig, S. Fig. 9.
f :
^ '
r'l
<-i- III
^ 1
r 1
f 11
k
HI
l "11
■^ taiwi!if)i]s
mmifffflj I
k
'■
h 1!
[
;M\\\\m\\\\\\\\\\s\\\m\\^^^^^^ vwww \^ \v\\\v-
Experiments were then carefully made to ascertain the exact pres-
sures that a graduated series of wires would carry. You will now
readily understand this method of deducing the maximum pressure.
If we know the maximum pressure exerted during the passage of
the shot through the bore to give motion to any known portion of the
shot's weight, we can deduce the pressure acting on the whole shot
itself. By properly adjusting the strength of the wires, we found that
certain wires would give motion
Fici- 10. to the pellets without shear-
ing ; others would not. Hence
we deduced an approximate
maximum pressure.
The other arrangement was
also carried in the front of the
projectile, and is here shown
(Fig. 10). In this case a known
weight w is supported, or rather
has motion communicated to
it, by means of a cylinder of
soft metal c. The amount of
crush on the cylinder serves as an indication of the force to which
it has been subjected.
It is not possible in anything like a reasonable time to give an
analysis of the voluminous investigations of Piobert on the question
of gunpowder.
ON THE TENSION OF FIRED GUNPOWDER G7
Generally, however, his views seem to be that he ascribes
much of the initial pressure of gunpowder to the effects of the
vaporised solid products increasing enormously the tension due
to the permanent gases.
He points out errors in some of Eumford's conclusions, but
accepts as tolerably accurate the pressures given by Eumford's first
series, which would, at maximum density, give a tension of about
29,000 atmospheres.
I have now run over hastily, but I hope intelligibly, the
principal experiments which have been made and the views which
have been entertained on the subject of the pressure of fired
gunpowder. The enormous discrepancies between the 1000 atmo-
spheres estimated by Eobins and the 100,000 atmospheres of Eumford
will not have escaped you ; and even coming to quite recent dates,
the difference of opinion between authorities like Piobert on the one
hand, and Bunsen and Schischkoff on the other, are quite startling
enough to show you the difficulties with which the subject is
enveloped. What I now have to detail to you chiefly relates to
the labours of a Committee, under the presidency of Colonel
Younghusband, recently appointed to examine into our gunpowder,
which has for some years enjoyed on the Continent the unenviable
denomination of " brutal powder."
The researches of this Committee having been devoted in the
first place to a special object — the production of a powder suitable
for the very large guns which are now required by the services —
all the experiments hitherto made have been undertaken with this
sole end in view. We have turned so far neither to the right
hand nor the left, and in consequence our knowledge relating to
many important points is very incomplete, in others altogether
defective; but, as far as my time permits, I shall lay a few of
our facts before you as concisely as I can, and where I may
venture to theorise I shall only give views which I believe to be
.shared in common with myself by the distinguished gentlemen with
whom I have the honour of being associated.
The guns we have principally used have been three in number —
a gun of 21-inch diameter, firing projectiles of 4f lbs., and charges of
9 ozs. ; an 8-inch gun, firing projectiles of 180 lbs., and charges of from
20 to 40 lbs. ; and a 10-inch gun, firing projectiles of 400 lbs., and
charges of from 60 to 70 lbs. of powder.
The means we have used to determine the pressure have been
likewise three — first, a Eodman gauge; secondly, a crusher gauge,
68
ON THE TENSION OF FIRED GUNPOWDER
Fig
designed to overcome certain faults in the Eodman gauge, which I
shall presently describe ; thirdly, a chronoscope, designed for measur-
ing very minute intervals of time.
The Eodman gauge I have already fully described. The crusher
gauge is shown in this drawing (Fig. 11), and consists of a screw plug
of steel let into the gun at any desired point, which
admits of a cylinder of copper, B, being placed in
the chamber CDEF.
The entrance to this chamber is closed by the
movable piston C, as in the case of the Eodman
gauge, and the admission of gas is prevented by the
use of a gas check.
You will have no difficulty in understanding the
manner in which results are arrived at with this in-
strument. When the gun is fired, the gas acts upon
the base of the piston and compresses the copper
cylinder. The amount of crush on the copper
serves as an index to the maximum force exerted
at that part of the bore where the plug is placed.
The chronoscope used by the Committee is
delineated in Plates VI. and VII., p, 86. It consists
of a series of thin discs, AA, each 36 inches in cir-
sEciioiM cumference, fixed at intervals on a horizontal shaft,
and driven at a high speed by the heavy descending weight B,
which is, during the experiment, continually wound up by the
handle H, and with a little practice the instrument can be made to
travel either quite uniformly or at a rate very slowly increasing
or decreasing.
The precise rate of the discs is ascertained by means of the stop-
clock * D, which can be connected or disconnected with the revolving
shaft E at pleasure.
The speed with which the circumferences of the discs travel is
in this instrument generally about 1200 inches per second. An
inch therefore represents the 1200th part of a second, and as
by means of a vernier we are able to divide the inch into 1000
parts, the instrument is capable of recording less than the one-
millionth part of a second. I may mention, by way of enabling you
to realise the extreme minuteness of this portion of time, that the
millionth part of a second is about the same fraction of a second that
a second is of a fortnight.
* An improved arrangement for registering the speed was afterwards introduced.
ON THE TENSION OF FIRED GUNPOWDER
69
I shall now endeavour to describe to you how the shot marks o)i
the instrument the record of its passage through the bore.
I need hardly remind most of you that when the primary of an
induction coil is suddenly severed, a spark under proper management
is given off from the secondary, and in the arrangement I am describ-
ing, the severance of the primary is caused by the shot in its passage
through the bore, and the record of its passage is transferred to the
discs in the following way.
The peripheries of the revolving discs are covered with strips of
white paper coated with lampblack, and are connected with one of
the secondary wires of an induction coil. The other secondary wire,
carefully insulated, is brought to one of these dischargers, Y, opposite
to the edge of a disc, and fitted so as to be just clear of it.
The mode of connecting the primary wires of the induction coils
with the bore of a gun in such a manner that the shot in passing a
Fig. 12.
^'^''^ N°^- ■• 5^. PLUGS
-1 16 17
\'i NOS.2.4.
definite point shall sever the primary current, and thereby produce
a, spark from the secondary, is shown in Fig. 12 which represents a
longitudinal section of the bore along which the shot is moving.
A hollow plug, C (see Fig. 13, p. 70), is screwed into the gun,
carrying at the end next the bore a cutter, D, which projects slightly
into the bore.
The cutter is held in this position by the primary wire, e, which
passes in at one side of the plug, then through a hole in the cutter,
and out at the other side of the plug.
When the shot passes the cutter it presses it level with the
surface of the bore, thereby severing the primary and causing the
induced spark to pass instantaneously from the discharger to the disc,
making a minute perforation in the paper-covering upon the opposite
part of the disc, and at the same time burning away the lampblack,
so that the position of the perforation is marked by a white spot.
70
ON THE TENSION OF FIRED GUNPOWDER
To prevent confusion, there is delineated in Plate VI., p. 86, only a
single induction coil and cell ; but you will understand that there is
an induction coil for each disc, and that each disc, discharger, and coil
form
an independent instrument for recording the
instant when the projectile passes
a certain point in the bore of the
gun.
It only remains to point out
that before using the instrument,,
we must be satisfied that the
various independent instruments
of which I have spoken give
corresponding results.
The best mode which occurred
to us of doing this is to get a record upon each disc of the same
event.
Thus it is obvious that if the whole of the primaries are cut
simultaneously, the sparks on all the discs should be in a straight
line, and the deviations from a straight line are the errors, either
constant or variable, and from the observations the constant errors
can, of course, be eliminated.
Two methods of securing a simultaneous rupture of the primaries
have been followed. One plan consisted in wrapping all the wires
round a small magazine of fulminate of mercury, and exploding the
fulminate. The other consisted in collecting the whole of the wires
on a small screen close to the muzzle of a rifle, and cutting them by
means of a flat-headed bullet. Both methods have given excellent
results.
Having now described the instruments, I turn to the guns. The
arrangements in all the guns were similar in character, but I have
given to you here (Fig. 12) a drawing of the 10-inch M. L. gun as repre-
senting the most perfect arrangement used in the early experiments.
We have, in the first place, the power of firing the cartridges in different
positions. Eodman's gauges, or the crusher gauges, are always placed
in the holes marked ABC, and in such other holes as we may desire,
while 8 holes every round are reserved for use with the chronoscope.
Suppose, for example, we wished to experiment with a charge of
70 lbs. of powder, our usual course would be : the chronoscope plugs
would be placed alternately in the holes 4 to 11, and in 11 to 18,
while the crusher gauges would be alternately in the holes ABC, 1,
14, 17, and in the holes ABC, 1, 4, 10.
ON THE TENSION OF FIRED GUNPOWDER 71
The pressures derived from either the Rodman or the crusher
gauge are read off from tables at once, but the determination of the
pressure from the time curve given by the chronoscope is a very
different matter.
I am aware that there are many authorities who consider it almost
impossible to obtain from a time curve such as is given by the chrono-
scope reliable indications of the pressure, and I cannot wonder that
many should so think.
We who have been investigating this subject, are quite alive to
the fact that a cause of error far graver than any chronoscopic error
lies in the difficulty, I might almost say impossibility, of assuring
ourselves that the projectile in successive experiments should
describe precisely the same space in passing between any two suc-
cessive plugs ; but, fortunately, errors of this description can gener-
ally be removed by known methods of interpolation and correction.
Again, if we relied for the determination of our maximum pres-
sure on the observation of two velocities only at very short intervals,
as trifling errors in the determination of the velocity would give rise
to considerable variations in pressure, our results would be open to
considerable doubt, but the fact is that, with the assumptions we are
at liberty to make, I have found that it is not posssible materially to
alter our pressure without setting our records altogether at nought.
The time curve — that is, the curve whose ordinate at any distance
up the bore represents the time the shot has taken to arrive from
zero at that spot — being drawn through the observed points, what
may we assume respecting the curve representing the velocity?
According to theory, we may assume that it commences by being
convex to the axis of abscissae, then becomes concave — that the radius
of curvature becomes greater and greater as x increases, and, were
the bore long enough, would be finally asymptotic to a line parallel
to the axis of x.
Again, as regards the curve representing pressure. We know
that the pressure will run up with extreme rapidity until it attains
a maximum, and that after attaining a maximum the ordinates will
rapidly decrease, the curve after passing the maximum being always
convex to the axis of abscissse.
These considerations, joined to the observations themselves, are
amply sufficient to give us the information required. At the com-
mencement of motion the plugs are very close to one another (about
2 inches apart), and the distances are gradually increased as they
approach the muzzle ; but close as they are at the seat of the shot
72 ON THE TENSION OF FIRED GUNPOWDER
they could advantageously be closer still — say, half the distance —
and they would have been so had we not been afraid to add more to
the many holes we have bored in a gun destined to be so severely
tested.
In working out the results for the first 6 inches of motion, the
times, velocities, and pressures are interpolated for every sVth of a
foot ; after that distance up to 3 feet, for every yV^h of a foot ; and
for the remainder of the bore, for every 6 inches.
Our experiments with the 2-inch gun do not call for much remark,
save that in this calibre the differences between samples of the same
class of powder of different manufacture were very strikingly shown,
the maximum pressure of one sample of powder of professedly the
same make being in some cases nearly double that of other samples.
But when we commenced our experiments with the 8-inch gun we
were at once brought in contact with some very singular anomalies.
Our first experiments with this gun were made with the Eodman
gauge and the chronoscope only, and our attention was directed
chiefly to two points — the different action of various kinds of powder,
and the effect on the same kinds of powder of lighting the cartridge
in a different position. On firing 20-lb. charges of the service powder
— technically known as E. L. G. — with the vent in the position in
which it is generally used in service, that is, at a distance of yV^hs
the length of the battering charge from the bottom of the bore, not
only did we find the Eodman gauges placed as I have described differ
very materially in their results one from the other, but they all indi-
cated a pressure very much higher than that shown by the chrono-
scope, the maximum chronoscope pressure being 17 tons per square
inch, while the maximum pressure of the Eodman gauges varied from
28 tons to 33 tons on the square inch.
We then fired a series with the same charge and powder, using
instead of the service vent a vent lighting the cartridge in the rear
and here the results were still more anomalous. The chronoscope
showed a maximum pressure differing but very slightly from the
result when the service vent was used, while the Eodman gauge at
the point C indicated a pressure of 50 tons.
These discrepancies threw some doubt on the accuracy of the
indications of the Eodman gauge, and we were led to ascribe this
inaccuracy to two causes — first, to the position* of the gauge on the
It must be remembered that this defect, due to position, has no existence in
many of the experiments with the Rodman gauge made on the Continent, because in
the Continental experiments breach-loading guns have been generally used and the
ON THE TENSION OF FIRED GUNPOWDER 73
outside of the gun ; secondly, to what appeared to us to Ije a slight
defect in the design of the gauge.
You will easily see our grounds for suspecting the effect which
the position of the gauge might have if I recall to your recollection
the experiment of Mr Eobins, to which 1 alluded early this evening
— namely, the enormous local pressure he found in a musket-barrel
when he placed the bullet a considerable distance in front of the
charge. In the Eodman gauge the indenting piston may be taken to
represent Eobin's bullet, and you will observe the distance the gas
has to travel before it reaches the indenting tool.
The slight defect I have mentioned in the design of the Eodman
gauge I may thus explain. Suppose the indenting tool, instead of
pressing against the copper as shown, was removed from it by any
given space, the gun then fired, and the gas allowed to act, it is
obvious that the indication given by the copper could not be relied
on, because, in addition to the pressure, the indenting tool would
express on the copper the vis viva due to the velocity it had acquired
when moving freely. In the Eodman knife the resistance to the
motion of the indenting tool commences at zero and rapidly in-
creases; but it is possible to conceive that the velocity* imparted
to the tool when the resistance is but small may to some extent
affect the amount of the indicated pressure.
The crusher gauge which I have described, and which admits of
being applied either close to the interior of the bore or at the exterior
of the gun, was thenceforth generally substituted for the Eodman
gauge ; and I may mention, as a proof of the correctness of our views,
that in quick-burning powders this gauge, when applied at the out-
side of the 8-inch gun, gave pressures about double of those it indicated
when applied to the inside.
The powders with which we have experimented maybe divided
into four classes — 1. The old quick-burning, violent powders, such
as E. L. G. and L. G. ; 2. Pellet Powder ; 3. Pebble Powder ; and 4.
Prismatic Powder. (See Pig. 14, p. 74.)
Here is a sample of the service E. L. G., and I will only remark
gauge has been applied to the wedge which closes the breach, and in this position
would give satisfactory results ; on the other hand, the pressure would only be
obtained at one point, and such a determination, our experiments show, is not to
be relied on.
* I was informed by General Gadolin, in Paris, that the results of the experi-
ments made with the Rodman gauge in Russia were found to be uniform and satis-
factory, only when prior to the experiments an indent was made in the copper a
little less than that expected to be produced in the gun. This fact may be explained
by the considerations referred to in the text.
74
ON THE TENSION OF FIRED GUNPOWDER
that our old rule of proof for powder, that of the eprouvette mortar,
seems, with our present lights, to be specially designed to produce in
powder those qualities whose absence we most desire. Here are
samples of pellet and pebble powders. You will notice that the
former are regular cylinders formed in moulds, while the latter are
tolerably regular lumps of powder cake, about the size of large
pebbles ; and, lastly, here is a sample of the prismatic powder which
has attained so considerable a reputation on the Continent.
Fig. 14.
RUSSIAN
PRISMATIC POWDER
PELLET
FULL SIZE
PEBBLE
Any one of the three last classes is very much superior to the
first. There is, in fact, no great difference, except as regards process
of manufacture, between the pellet and pebble. Both give, when
properly made, good results, although there seems to be a greater
probability of attaining uniform results with the pellet than the
pebble ; but the prismatic differs considerably from these in being a
less dense powder, and possessing the property of lighting with
extreme slowness, as you will see by comparing its velocity or time
curves with those of either pebble or K. L. G. 1 might characterise,
perhaps, 11. L. G. as a quick-lighting and extremely quick-burning
ON THE TENSION OF FIRED GUNPOWDER
75
powder ; pellet and pebble as quick-lighting, slow-burning powders ;
and prismatic as slow-lighting and quick-burning powder. It is
probable that the prismatic powder owes it extreme slowness of
lighting to the deposition of a heavy coating of saltpetre, due to the
moisture present in the process of manufacture.
Although we find that almost inappreciable differences in the
manufacture cause occasionally great differences of acbion when the
powder is submitted to the test of firing, we are able to point to
several causes which are of the greatest importance in modifying the
behaviour of the powder in the gun. These points are — 1, Specific
gravity; 2. Length of time during which the component charcoal
Fig. 1
5.
X
/'
.
X^
/'
.•^'
r'
##
t
^|C
X"
^
►^"^
nl*
^
^'
9^
9
J
^
/
^
/
./
.• /J'
<
/
' ,'•
r
a
'006 z
o
o
u
«/>
•004
-•002
S FEET.
has been burned ; 3. Degree of moisture employed in manufacture ;
4. Hardness ; 5. Size of grain.
I have arranged on diagrams curves intended to illustrate the
differences between three of the classes of powder I have been
describing, and in each case I have selected an example which I
believe to be as nearly as possible a type of the class. For the
purpose of comparison, they are all taken from experiments with the
10-inch gun.
On this diagram, Fig. 15,* are delineated the time curves, that is,
* In this and the following figure the black dots denote the observed points.
In each figure, however, to prevent confusion, the dots are omitted in the case of
one curve.
76
ON THE TENSION OF FIRED GUNPOWDER
the indications given by the chronoscope itself ; the
represent the lengths of bore ; the ordinates, the total time the shot
takes to reach those lengths from the commencement of motion.
Tliis curve represents R. L. G-., this pebble, and this prismatic. Note
p5!
ii^*^
1400
1200
_ 1000
800
_ 600
i- - 400
1 2 3 4-5
Fig. 16.
8 KEET
800-
!
_,_--
__,-;cr?=*'*'
z-
-6'0"0^
-^
'''" ^-^''
©
<-)
^^-' — ' — "Z^
'"' ^^^
'''
U
J^
Ui^-^
^ -' "',
^"
w
/^ ^'^
1^'
"^
u
Q.
400-
y
y^
U
200-
,*
//
_^ ■''
"
'^mStZZ^
1
Fig. 17.
how much less is the time taken by the shot in the earlier parts of its
motion in the case of R L. G. and pebble than in that of prismatic.
It may be interesting to mention that the total time taken by a
projectile, when fired with a battering charge, to reach the muzzle of
a 10-inch gun is about the nr^th part of a second.
ON THE TENSION OF FIRED GUNPOWDER
77
The velocities at each point of the bore, deduced from these time
curves, are here exhibited. Figs. 16 and 17.* Observe how, in the
pebble and prismatic powders, the velocity commences by being con-
siderably lower than the K. L. G. velocity ; how they gradually reach
it, pass it, and the projectile finally quits the gun, possessing a very
considerably higher velocity.
The curves towards the muzzle pass very nearly through the
observed velocities. Near the origin of motion the curves pass
above the observed points, as they necessarily would do.
These curves, again. Fig. 18, represent the pressures correspond-
Fig.
18.
R.
L.G
'RISMATIC
PEBBLE.
ing to those velocities, and their area is the measure of the work
done by the respective powders on the shot.
You will note that with both the prismatic and pebble powders,-
although the maximum pressure is considerably less than with the
E. L. G., this area is considerably more than the E. L. G. area.
Hence follows the important fact— not only by the use of pebble
powder, for example, is the gun much less strained than by the use
of E. L. G., but we actually obtain from our gun, with the charges
* As, owing to the small scale of Fig. 16, giving the velocities throughout the
bore, the differences in velocity near the commencement of motion are not readily
perceptible, the same curves for the first 6 feet of motion have been laid down to
a larger scale in Fig. 17.
78 ON THE TENSION OF FIRED GUNPOWDER
we are enabled to use, nearly 20 per cent, more effect, the work done
by the former powder being about 5700 foot tons, while by the latter
it is only 4900 foot tons.
The pressures indicated by these curves are obtained from the
chronoscope indications, and I now propose to examine what are the
corresponding indications with the crusher gauge. They are as
follow: — With the pebble, pellet, and prismatic powders, under
ordinary circumstances, that is to say, with ordinary or battering
charges of the service and with service vents, the pressiire indicated
by the crushers placed in the powder chamber in the positions
marked A, B, C, do not differ materially from one another, and any
of them, or the mean of the whole of them, agree tolerably closely
with the maximum pressure indicated by the chronoscope. But
when we come to E. L. G-. or L. Gr. powders, a striking difference
manifests itself; not only do the pressures in E. L. G. differ very
materially from the indications given by the chronoscope, but they
differ widely from one another. It is hardly necessary for me to
point out to you that on the ordinary theory of the distribution of
gas in the powder chamber in the first moments of motion, the
density and consequent tension of gas should be least next the shot
and should gradually, but not very greatly, increase towards the
bottom of the bore. This, however, was not at all so. Thus, for
example, with one specimen of E. L. G., while the chronoscope pres-
sure was found to be 28 '3 tons, the pressure indicated by the crushers
at C was 280 tons, at B was 31-3 tons, and at A was 47-9 tons.
From other circumstances we were well aware that when similar
■charges and powder were fired with a rear centre vent, the destructive
action on the gun was much reduced, but unfortunately with the
destructive action was reduced also the useful effect. On our
making the experiment, however, we found the chronoscope maxi-
mum pressure 19 tons instead of 28 tons, while the crusher
pressure indicated at B was 26 tons instead of 31 tons, and
at C 39 tons instead of 28 tons. What then was the cause of
these striking differences? I may point out that there is no
manner of doubt as to the reality of the facts indicated by the
•crushers ; not only do they appear, round after round, with unfail-
ing regularity, but we have tested the correctness of the results in
every way our ingenuity could suggest. We are therefore met in
the case of the destructive powders with difficulties which do
not exist in the case of slow-burning powders, and as we are com-
pelled to admit that some of those pressures are entirely local, or
ON THE TENSION OF FIRED GUNPOWDER 79
confined to certain portions of the gun, we give the following
explanation.
I need hardly again recall to your memory the early experiment
of Eobins, and the high local pressure he obtained by placing the
musket-bullet at some distance from the charge. The explanation
of this phenomenon doubtless is that the inflamed gas, vapours, or
other products of explosion arising from the combustion of the
powder attained a very high velocity before encountering the
resistance of the bullet, and the reconversion of the vis viva into
pressure accounts for the intense local pressure that Eobins
observed. The local pressure we have observed can be similarly
explained. The vis viva of the products of combustion of the
first portion of the charge ignited is in like manner converted
into pressure at the seat of the shot, and as we know that the
rapidity of combustion of powder is enormously accelerated by
the tension under which it is exploded, it is possible that this
pressure may be increased by a violent disengagement of gas
from the unconsumed powder at the seat of the shot.
The crusher pressure indicated with the rear vent is, as we
might expect from the increased run, considerably higher than
when the service vent is vised.
The time during which this abnormal pressure is kept up must
be exceedingly minute, even when compared with the infini-
tesimal times we are considering, for we find the chronoscope
pressure, which may be regarded in the case of these " poudres
brutales" as representing the mean of pressures of a violent
oscillatory character, hardly altered at all, even although the
local pressures — as, for instance, when the rear vent is used — are
increased 50 per cent.
Other indications also, which I shall shortly notice, lead to the
same conclusion ; but it is worthy of remark that, when violent
local pressures are set up, waves of pressure, so to speak, appear
to sweep from one end of the inflamed gases to the other, and to
continue more or less during the whole time the shot is in the,
bore.
We are led to this conclusion from the following : —
With pebble and other powders, where no wave action is set
up, the pressures indicated by the crushers throughout the bore
agree satisfactorily with those indicated by the chronoscope, and
the area of a curve drawn through the observations represents
with tolerable accuracy the work done on the shot, but when
80
ON THE TENSION OF FIRED GUNPOWDER
wave action is set up this no longer holds. The velocity of the
shot may be the same, or even less, and of course the area of which
I have spoken should correspond. On the contrary, however, it is
always greater — frequently enormously so — representing 60 to 70'
per cent, more work than is really done on the shot.
I have drawn on this pressure curve. Fig. 19, belonging to
R. L. Gr., an imaginary line showing the way in which we may
suppose these violent oscillations to exist; you will observe that
oscillations of this character would not only explain the anomalies
obtained with the crusher, but would explain also the double
maxima invariably observed by General Neumann's Committee.
\
Fig. 19.
J
A
'
W
"^
--
-
-^
—
—
3
0-
2
~
^-
0-
6
ftH
I will only add that the chronoscope and crusher in these inves-
tigations appear to me to be complementary one to the other.
The chronoscope hardly recognises the existence of the local
pressures; on the other hand, the crusher frequently gives no
clue whatever to the mean pressure existing in the chamber.
The above remarks as to local pressures apply to quick-firing
powders. With service vents and service charges this wave
action scarcely seems to exist in the other powders I have
discussed; but if the charge be greatly increased in length,
more especially if the cartridge be lighted from the rear, it
again appears. It must be remembered that, objectionable, for
many reasons, as this action is, it is in no way so serious a&
if the local pressure extended simultaneously throughout the
ON THE TENSION OF FIRED GUNPOWDER
81
chamber. In fact, certain considerations, with which I need not
trouble you, led me to the conclusion that it was possible that
under certain circumstances the maxima of the local pressure
might be confined, not only to a certain portion in the longi-
tudinal section of the bore, but even to a certain small arc in
the transverse section through that portion.
I therefore caused the records of proof of certain 10-inch guns
which have been proved at Elswick in a manner calculated to
produce in a high degree local pressures, to be examined, and
found that out of 26 guns 16 had, after proof, no expansion at
Fig. 20.
all, 2 were expanded in a very narrow rim all round at the seat
of the shot, and the remainder, 8 in number, had small enlarge-
ments technically called dents, but the whole of these dents were
confined to the seat of the shot, and to that portion of the section
nearly opposite the vent which I have indicated in this diagram.
Fig. 20.
Again, it is almost certain that the high local pressure indicated
at the bottom of the bore in the 10-inch guns is confined to the par-
ticular point where the crusher is placed, and is due to the contrac-
tion of the bore towards the end.
To one difficulty I must allude.
In the quick-burning powders, at all events, it seems to be certain
F
82 ON THE TENSION OF FIRED GUNPOWDER
that all, or at least all but a very trivial quantity, of the powder is
converted into gas by the combustion of the powder before the pro-
jectile has been materially moved from its initial position. A glance
at one of these pressure diagrams must convince you of this fact ; but
this being the case, how are we to account for the great loss of work
which results when, under ordinary circumstances, a charge is ignited
from the rear vent ? This loss is very variable, but in one instance
in our own experiments the work realised in the shot was reduced
from 78 foot tons to 58 foot tons per lb. of powder.
The cause of this great loss of work, in an instance where it is
difficult to believe that any quantity of powder can have escaped
ignition, may, perhaps, be sought either on the hypothesis that under
this peculiar mode of ignition the products of combustion differ
materially from those arising under ordinary circumstances, or, as
heat plays so important a part in the pressure of fired gunpowder, it
may possibly be surmised that with the rear vent a much greater
waste of heat has resulted than in the case of the service vent. I
believe it is generally assumed that the loss of work arising from the
heat communicated to the gun is altogether insignificant. This is,
however, not so.
Careful experiments were made on this head some years ago in
Italy with rifles, the rifles being fired under three conditions — viz.,
with the bullet as usual, the bullet very considerably removed from
the charge, and with no bullet at all.
The results were that in all cases the heat communicated to the
barrel represented considerably more than one-third of the total work
developed, according to Bunsen and Schischkoff, by the combustion
of the powder, being greatest when the ball was placed at some
distance from the charge, least when the rifle was loaded in the
ordinary manner.
The loss of heat would be very different in the case of the large
charges with which we are dealing, but it is still much too large to
be neglected, and it is certain that where the wave action, to which I
have so often adverted, is set up, there is always a considerable loss
of useful effect.
We are not, however, disposed to theorise too closely on the
anomalies to which I have referred, as I believe I may say we have
reasonable hopes of being able to solve some of our difficulties.
Collaterally with the researches of the Committee on the action
of gunpowder in guns, I have made at Elswick a series of experi-
ments on the tension of the gases in closed vessels.
ON THE TENSION OF FIRED GUNPOWDER
83
On the same diagram (Fig. 2, p. 56) in which I have placed
Rumford's and Eodman's experiments I have plotted down our Elswick
experiments, a portion of which were undertaken at the suggestion
of General Lefroy.
Eumford only succeeded in determining the tension of the powder-
gases when the powder occupied less than 70 per cent, of the space
in which it was fired. His charges also were insignificant, and his
results, possibly from faults arising from his mode of operation, are
extravagantly high. Rodman's results, owing to the defect I have
pointed out in his instrument, are also high, but he did not determine
the tension where the powder occupied a larger proportion of the
space than 50 per cent.
At Elswick, however, we have been so fortunate as not only to
determine the tension of the gases at various densities, but we have
exploded charges filling entirely the chambers of close vessels, and
have altogether retained, and, by means of a special arrangement,
discharged at pleasure the gaseous products of combustion.
The results of our experiments, all with Government R. L. G., are
shown in the diagram, and it only remains for me to describe the
apparatus with which we obtained our results. It is here shown
(Fig. 21):-
FiG. 21.
The inflamed products are confined in the chamber by means of
this gas check. The pressure is determined by means of a crusher
arrangement fitted at A. The charge is exploded by means of one
of Mr Abel's fuzes. The curren passes through this insulated cone,
B, which, the moment the charge is fired, destroys the insulating
material and effectually closes the passage. The details of one or
two of these experiments will be interesting to you. When we first
made the arrangement for confining the powder absolutely, I thought
84 ON THE TENSION OF FIRED GUNPOWDER
that the best method of stopping the escape of the gas was to make
a steel vent, closing it with a gun-metal plug faced with tin. This
arrangement was apparently successful. When I had just got up to
the cylinder, and was stooping down to feel its heat, the charge
suddenly made its escape with considerable violence. When the
cylinder was opened for examination it was found that the escape of
the gas was due to the heat of the explosion having melted the tin
between the conical plug, and through the melted tin the gas readily
Another most remarkable occurrence was noted in the examination
of this cylinder. On taking out the crusher apparatus, to my sur-
prise I found that a portion of the solid steel projecting into the
charge had been melted, and apparently run; also the head of a
hardened steel screw had evidently fused. I hold in my hand these
evidences of fusion, and call your attention to the exceedingly short
time, 32 seconds, in which these effects were produced. By way of
comparison, I put, for 37 seconds, into one of the hottest of Siemen's
regenerative furnaces, at a temperature probably of about 3300° Fahr.,
a similar piece of steel. It was raised only to a heat of about 180" Fahr.
I must warn you, however, that the temperature of this fusion
may have been seriously affected by chemical changes through which
the fused metal may have passed ; but an examination which I hope
to have shortly made will settle this point.
With one other experiment I must trouble you. In the experi-
ment I have just related I determined the tension of three-quarters
of a pound of E. L. G. powder, completely filling the chamber in
which it was fired, and having no escape whatever, to be about 32
tons on the square inch. For the purpose of my lecture this evening,
I determined to make a similar experiment with F. G-. and pellet. I
have done so, and the results were completely successful. The gas
was entirely confined. In the first case, when I got up to the cylinder
it was making a singular crepitating noise, due probably to the
sudden application of great internal heat. The temperature of the
exterior of the cylinder rose rapidly to 111° Fahr., and then remained
nearly stationary for some time. I then let the gases escape, which
they did with a sharp, hissing noise, rising to a scream when any
obstacle was placed on the orifice. With the escaping gases there
was not the slightest appearance of smoke, vapour, or colour of any
kind. The pressure indicated by the F. G. was 37 tons on the square
inch, or about 5600 atmospneres.
Here, in those sealed bottles, are the solid residues of combustion
ON THE TENSION OF FIRED GUNPOWDER
85
from the R G. and also from pebble. In each cylinder had been
platinum wire and foil of different degrees of thickness. These have
disappeared, and I am unable to say in what state they now are,
until the residues have been examined.
I look upon the success of these experiments as being of great
importance.
Not only, with the assistance of my friend and colleague (on the
Committee), Mr Abel, so well known for his researches in explosive
substances, shall we be able to determine the various products of
combustion when the powder is fired at its maximum pressure, but
r— 30
Fig. 22.
R.L.G.I0"CUN.
.C.8 CUN.
,10". GUN.
we shall be able to determine whether any, and if so what, change in
the products is due to combustion under varying pressure ; we shall
also be able to determine the heat of combustion, and solve other
important questions.
To a remarkable coincidence and singular confirmation of the
Committee's results I must draw your attention.
Upon my obtaining this curve, giving the relation between the
tension and density of the powder-gases in a close chamber, I was
anxious to see how these results would conform with similar ones
obtained from our observations of the tension in the bores of guns.
Accordingly I laid down these curves anew, Fig. 22, representing
86 ON THE TENSION OF FIRED GUNPOWDER
pebble-powder fired in 10-inch and 8-inch guns, and K. L. G. fired in
10-inch, 8-inch, and 2-inch guns, the ordinates as before representing
the tension of the powder, but the abscissae representing the density of
the gas. You will perceive, under this view, how closely the 10-inch
and 8-inch pebble and K. L. G. approximate. But when I came
to put on the same diagram, as indicated by the crosses, the
tensions I had obtained from powder fired in a close vessel, they
were nearly absolutely identical with the results obtained in the
10-inch gun from pebble-powder.
The coincidence, you will agree, is too remarkable to be
accidental.
The practical conclusions to be deduced from the investigations
forming the subject of this lecture may be arranged as follows : —
1st. — The maximum pressure of fired ordinary gunpowder
density being unity unrelieved by expansion, is not much above 40
tons to the square inch.
2nd. — In large guns, owing to the violent oscillations produced
by the ignition of a large mass of powder, the pressure of the gas
is liable to be locally exalted, even above its normal tension, in a
perfectly closed vessel, and this intensification of pressure endangers
the endurance of the gun, while detracting from the useful effect.
3rd. — Where large charges are used, quick-burning powder for
the same energy greatly increases the strain upon the gun.
4th. — ^The position of the vent or firing point exercises an
important influence upon the intensity of wave action, and in
further enlarging the dimensions of heavy guns we must look to
improved powder, and improved methods of firing the charge, so
as to avoid as much as possible throwing the ignited gases into
violent oscillation.
5th. — In all cases it is desirable to have the charges as short as
possible, and the cartridge so lighted as to reduce the run of the gas
to the shortest limit.
But I must conclude, and, while regretting the imperfect and
incomplete information which I have been able this evening to give
you, I trust you will remember that our investigations are still
proceeding, and that, should the subject be of interest to you, and
our work seem of sufficient importance, I or some other member
of our Committee may yet be able to Jay before you the results of
our further researches.
PLATE VI.
PLATE VII.
ON THE PEESSUEE EEQUIEED TO GIVE EOTATION
TO EIFLED PEOJECTILES.
{Philosophical Magazine, 1873.)
1. In a paper published in the Philosophical Magazine for 1863
(vol. XX vi.), and subsequently in the B,evue de Technologie Militaire
I gave some investigations on the ratio between the forces tending to
produce translation and rotation in the bores of rifled guns.
2. My object in these investigations was to show, 1st, that in the
rifled guns with which experiments were then being made the force
required to give rotation was generally only a small fraction of that
required to give translation ; 2ndly, that in all cases (and this was a
point about which much discussion had taken place) the increment
of gaseous pressure (that is, the increase of bursting force) due to
rifling was quite insignificant.
3. In the paper referred to, although the formulae were sufficiently
general to embrace the various systems of rifling then under con-
sideration in England, they did not include the case of an increasing
twist, which has since been adopted for the 8-inch and all larger guns
of the British service ; neither was our knowledge of the pressure of
fired gunpowder sufficient to enable me to place absolute values on
either of the forces I was considering.
4. Since the date at which I wrote, an extensive series of experi-
ments has been made in this country ; and the results of these
experiments enable me to give with very considerable accuracy both
the pressure acting on the base of the projectile and the velocity at
any point of the bore. I am therefore now able not only to assign
absolute values where in my former paper I only gave ratios, but
also to show the amount by which the studs of the projectiles of
heavy guns have been relieved by the introduction of the accelerat-
ing twist known as the parabolic system of rifling.
88 ON THE PRESSURE REQUIRED TO GIVE
5. Very little consideration will satisfy any one conversant with
the subject, that in the ordinary uniform spiral or twist the pressure
on the studs or other driving-surface is a maximum when the pressure
on the base of the shot is a maximum, and becomes greatly reduced
during the passage of the shot from its seat to the muzzle of the gun.
In my former paper I showed, in fact, that in a uniform twist the
pressure on the studs was a constant fraction of the pressure on the
base of the shot, the value of the fraction of course depending on the
angle of the rifling ; and as it is evident that the tension of the
powder-gases at the muzzle is very small when compared with the
tension of the same gases at the seat of the shot, it follows that in
such a system of rifling the studs may have scarcely any work to do
at the muzzle, while they may be severely strained at the commence-
ment of motion.
6. If, then, the defect of the ordinary or uniform system of rifling
be that the studs are severely strained at the flrst instants of motion
and are insignificantly strained at the instant of quitting the gun,
it is obvious that it is possible to remove this inequality and at the
same time allow the projectile to leave the bore with the same angular
velocity by reducing the twist at the seat of the shot and gradually
increasing it until it gains the desired angle at the muzzle. In fact,
if we know the law according to which the pressure of the powder
varies throughout the bore, it is theoretically possible to devise a
system of rifling which shall give a uniform pressure on the studs
throughout the bore.
7. These reasons doubtless led the late Ordnance Select Com-
mittee, to whom the application of the increasing twist to the service
guns is due, to propose its introduction ; and they selected as the
simplest form of an increasing spiral the curve which, when developed
on a plane surface, should have the increments of the angle of rifling
uniform. This curve is, as is well known, a parabola ; and as con-
siderable advantages have been claimed for the parabolic system of
rifling, I propose in this paper to examine and evaluate them.
I may add that I should not have given the results I now give,
before the full experiments made by the Committee of Explosives, as
well as some investigations undertaken by Mr Abel and myself are
published, were it not that several groundless assertions concerning
the Woolwich rifling have recently appeared, and have led to much
discussion and very unnecessary uneasiness.
8. The argument commonly advanced against an accelerating
twist is based upon the fact of the shot moving slowest at first, it
ROTATION TO RIFLED PROJECTILES
Fig. 1.
being supposed that while moving slowest the shot will require less
force to make it rotate ; but there is a fallacy in this argument,
which lies in confounding velocity with rate of acceleration. The
shot undoubtedly moves slowest at first, but it acquires velocity most
rapidly at first, and it is the gain of velocity that determines the
strain upon the stud.
9. The first question, then, which I propose is, to determine the
pressure on the studs of a projectile fired from a gun rifled on a
parabolic or uniformly increasing twist ; and in this investigation I
shall adopt the notation used in my former paper.
10. Take, then, as the plane of xy a plane at right angles to the
axis of the gun. If the angle of rifling commence at zero, increasing
to, say one turn in n calibres, let the plane
of xy pass through the commencement of
the rifling ; but if the rifling do not com-
mence at zero, it will be found more con-
venient to make the plane of xy pass through
the point where the twist would be zero
were the grooves sufficiently prolonged. Let
the axis of x pass through one of the
grooves ; and, for the sake of simplicity, we
shall suppose the rifling to be given by one
groove only. Let the axis of z be coincident
with that of the gun; let AP (see Fig. 1)
be the groove or curve described by the
point P, and let P {x, y, z) be the point at
which the resultant of all the pressures
tending to produce rotation may be assumed to act at a given
instant. Let the angle AON = 9!).
11. Now the projectile in its passage through the bore is acted on
by the following forces : —
1st. The gaseous pressure G, the resultant of which acts along
the axis of z.
2nd. The pressure tending to produce rotation. Calling this
pressure E, and observing that it will be exerted normally to the
surface of the groove, we have for the resolved parts of this pressure
along the co-ordinate axes, E cos X, E cos n, and E cos i/ ; X, /x, and v
being the angles which the normal makes with the co-ordinate axes.
3rd. The friction between the stud or rib of the projectile and the
driving-surface of the groove. This force tends to retard the motion
of the projectile ; its direction will be along the tangent to the curve
90 ON THE PRESSURE REQUIRED TO GIVE
which the point P describes. If wi be the coefficient of friction, and
if a, (3, y be the angles which the tangent makes with the co-ordinate
axes, the resolved portions of this force are /x^R . cos a, /UjE . cos /3,
yu^E . cos y.
12. Summing up these forces, the forces which act
parallel to x are X = R . { cos A - /a^ cos a } '\
>, j/ „ Y = R.jcos/x-/^^ cos/3} I • • (1)
„ 2 „ Z = G + R . { cos V - /Xj cos y } j
and the equations of motion are
M. J= G + R{cosv-/x, cosy} . . . (2)
M.^-l^ = Jl=ll (3)
p being the radius of gyration. Equations (1), (2), and (3) are
identical with those I formerly gave.
13. Now, in the case of a uniformly increasing twist, the equations
to the curve which when developed on a plane surface is a parabola
may be put under the form
.r = r cos ^ ; y = r sin ^ ; z^ = kr(fi . . . (4)
Hence
dx= -r sin <^ . c?<^ ; dy = r cos <^ . d; ds = 2" n/402 + k-^.dcf>
and we have, to determine the angles which the tangent to the curve
described by P makes with the co-ordinate axes, the equations
dx - 2s . sin
^ dy 2z . cos (h
cos/? = -j^ = ■ ^ .
ds Jiz^ + k^
dz k
cos y = — - = — .
ds Jiz^ + k'^
(5)
14. In the Woolwich guns the driving-surface of the groove may
be taken, without sensible error, as the simpler form of surface where
the normal to the driving-surface is perpendicular to the radius, the
surface itself being generated by that radius of the bore which,
passing perpendicularly through the axis of z, meets the curve
described by the point P ; but in the first instance I shall examine
the more general case, where the normal makes any assigned angle
with the radius.
ROTATION TO RIFLED PROJECTILES 91
Assume then that on the plane of xy the normal makes an angle
S with the radius of the gun. The driving-surface of the groove is
then swept out by a straight line which, always remaining parallel to
the plane of xy, intersects the curve described by P, and touches the
right cylinder whose axis is coincident with that of z, and whose
radius =r . cos S.
Now, the equations to the director being given by (4), and that
to the cylinder, which the generator always touches, being
x^ +y-i = (r COS 8)- . . . • (6)
it is easily shown that the co-ordinates x^, y^ of the point of contact
of the tangent to the cylinder drawn from P parallel to the plane xy,
.r-j = r . cos 8 . cos ((/> - 8)^
^^ = r . cos 8 . sin (^ - 8) /
and that the equation to the driving-surface is
(7)
S\ + y.sinj^ - SJ = r.cos8 • (8)
A-r J ' " Kkr
15. The angles which the normal to this surface make with the
co-ordinate axes are given by
cos A
with similar expressions for cos fx and cos v. But
ms for cos /j. and cos v. .
Therefore the angles which the normal to the driving-surface makes
with the axes are given by
. sin 8
/S2
k . cos y—
cos A
COS fl
J 4:z%sin 8y + k'^
k . sin (J^ - 8)
V4s2(sin8)2 + F
2s . sin 8
JizXsin8y + ¥
(9)
92 ON THE PRESSURE REQUIRED TO GIVE
16. Substituting in (2) and (3) the values given for a, /3, y. A, ijl, v
in (5) and (9), the equations of motion become
M ^'' n Rf 2.^^in8 f.,k
• (10)
. (11)
''• dV^ ^ I V402(sin Kf + F ^ 74^2 +
rf2<^ R . r (- yt . sin 8 2/XjS
' df^ p^ lV422(sinS)2 + F 74^2 +
and from (11),
M.^2 rf2^
. (12)
C A-, sin 8 2//^s )^ rf/2
17. To determine -^
From (4) ^r<^ = 22
A,. '^'^ 0. dz
^''-i^--r-im^\dt)]
d^ 2 i d'^z , )
./^2 -krV-di^^'^l
. 03)
and substituting this value of -^ in (12)
2Mp2 ( d
^^2|_^^sin8^ _ , ^^1^^
i J4z%sin 8)2 + F Jiz^ + F
for brevity,
V-m*'"}
A
or, substituting the value of j.^ derived from (10)
1 M M VV4s2(^sinS)2 + F ^ V4^2 + ^.2; /
and from this expression may be deduced
R 2p2iGc+M.2}
V4.a2(sin8)2 + F ^ s/4^2:fl2
18. Equation (14) gives the pressure acting between the studs or
rib of the projectile and the driving-surface of the groove at any
ROTATION TO RIFLED PROJECTILES 93
point of the bore, and for any inclination of the driving-surface ; but,
as before stated, in the Woolwich guns the normal to the driving-
surface (that is, the line of action of E) may, without material error,
be considered as perpendicular to the radius.
If in (14) S be put = 90°, the equation is simplified ; and the
resulting expression gives the total pressure on the studs for the
Woolwich guns.
Putting then ^ = 90°, (14) becomes
h'Xk - 2/1,^2) + 2p^c(2:: + fx^k)
19. Compare now (14) and (15), the equations giving the pressure
on the studs for parabolic rifling, with the equations subsisting where
a uniform twist is used.
For a uniform twist we have, as I formerly showed,
R '^ G. . (16)
li^{2ivp^k - rh) (2 V + rhk) sin 8
Jl+¥ "*" V^' + (shi8>^
where h is the pitch of the rifling, k the tangent of the angle which
the groove makes with the plane of xy, the other constants bearing
the meaning I have already assigned to them in this investigation.
20. In the Woolwich guns, where ^ = 90°, (16) becomes
R= 2iTp''J\+lc^ _G . . . (17)
hr{k - /Xj) + 2Trp\fj.^k + 1)
21. I proceed to apply these formulae, and propose to examine
what are the pressures actually required to give rotation to a 400-lb.
projectile, fired from a 10-inch gun with battering charges, under
the following conditions :— 1st. If the gun be rifled with an increasing
twist as at present. 2nd. If it be rifled with a uniform pitch, the
projectile in both cases being supposed to have the same angular
velocity on quitting the gun. As the calculations for the uniform
pitch are the simpler, I shall take this case first.
22. I have before remarked that with a uniform twist the pres-
sure on the studs of the projectile is a constant fraction of that on
the base of the shot, and represents, so to speak, on a reduced scale,
the pressure existing at any point in the bore of the gun. Calling
the fraction in equation (17) C, we have
R = C.G (18)
94
where
ON THE PRESSURE REQUIRED TO GIVE
27r/D2Vl+l2
Ar(^-/Xj) + 277/3^1^+1)
•04426
• (19)
the values of the constants in (19) being in the case of the 10-inch
gun as follow : —
P = -312ft., k
Hence
12-732, A = 33-333ft., /•=-417ft., /Xj = -167
R = -04426.0
(20)
23. But the values of G are known with very considerable exact-
ness from the investigations of the Explosive Committee under the
presidency of Colonel Younghusband. The following Table gives the
value of G (that is, the total pressure in tons acting on the base of
the projectile) for a charge of 70 lbs. of pebble-powder at various
points of the bore, and the corresponding values of E. It will be
remarked how high the pressure on the studs is when that on the
base of the shot is a maximum, and how rapidly the strain decreases
as the shot approaches the muzzle.
Table showing the pressure, on the studs in a 10-inch British-service f/un
rifled with a uniform tvrist, calculated from (17).
Travel of shot,
in feet.
Total pressure G
on base of shot,
Value of C.
Value of R, or
total pressure on studs.
in tons.
in tons.
0-000
•04426
0-333
1547
,,
68-5
0-945
1077
47-7
1-8:34
781
34-6
2-723
621
27-5
3-612
510
22-6
4 500
424
18-7
5-389
-356
15-8
6-278
305
13-5
7-167
268
11-8
8-055
240
10-6
8-944
220
9-7
9-833
205
9-1
24 The results in the Table show the pressures required to give
rotation, if the 10-inch gun be rifled on a uniform twist. I turn
now to the rifling as it actually exists, and which is defined to
be a parabolic twist, commencing with one turn in 100 calibres
and terminating at the distance of 9 '833 feet with a twist of one
ROTATION TO RIFLED PROJECTILES
95
turn in 40 calibres; and first to determine the equation to the
parabola.
Let the origin be at the point where the twist vanishes when the
curve AB is sufficiently prolonged — that is, at the vertex of the
parabola. Let Oz and Oy' be the axes of co-ordinates ; let A' = z-^,
Fig. 2.
OB' = 23; let tan Oi be the tangent of the angle which the curve
makes with O2: at A, and tan 0^ be the corresponding tangent at B.
Then, from the definition of the parabolic twist,
dz
constant = c, suppose
and
But, from (21),
tan 6^ = cz.-,, and tan 6^ = cz-^
tan 0^ - tan (
0047925
(21)
(22)
Comparing (22) with the form of this equation given in (4),
2
z^ = kr(f), we have y =r(f) and k = — = 417-3
Hence the equation to the development of the parabolic rifling
s2 = 417-3r0 (23)
and z^ the distance of the origin from the commencement of the
rifling
tan
6-555 feet.
25. As in the last case, I place in the form of a Table the results
96 ON THE PRESSURE REQUIRED TO GIVE
given by (15) for different values of z. The values of the constants
are,
r= -417 feet, ^ = 417-3, p = -.312 feet, /x^ = -167, M = -00555
Table shoviing the pressure on the studs in a 10-inch British-service gun rifled with a
parabolic twist, commencing at one turn in 100 calibres and terminating at one
turn in 40 calibres, calculated from (If)).
Value of z, the
distance from
the origin, in
Corresponding
travel of the shot
in the bore, in
feet.
Corresponding
velocity of shot,
in feet.
Total pressure
on base ot shot,
in tons.
Value of R,
or total
pressure on
studs, in tons.
6-5.55
0-000
6-888
0-333
411
1547
31-2
7-500
0-945
675
1077
28-7
8.389
1-834
873
781
29-0
9-278
2-723
992
621
30-2
10-167
3-612
1078
510
31-4
11-055
4-500
1146
424
32-3
11-944
5-389
1200
356
33-0
12-833
6-278
1245
305
33-8
13-7-22
7-167
1-282
268
34-5
14-610
8-055
1311
240
35-2
15-499
8-944
1333
220
35-8
16-388
9-833
1349
205
36-3
26. From an examination of the values of E given in this Table,
it will be seen that the total pressure on the driving-surface reaches
about 31 tons shortly after the commencement of motion, and the
projectile quits the bore with a pressure of about 36 tons. With the
view of making the variations which the pressures undergo more
readily comparable, I have laid down in the coloured Plate facing
page 98 the curves derived from Equations (15) and (17) for the
battering charge of pebble-powder.
From these diagrams the pressures on the driving-surface at any
point of the bore, both for the uniform and parabolic twists, can be
seen by simple inspection. The axis of abscissae gives the travel of
the shot, and the ordinates give the corresponding total pressure on
the studs.
The curves show that with the uniform spiral the pressure on the
studs reaches nearly 70 tons after a travel of '3 feet, rapidly falling
to about 9 tons at the muzzle, while with the parabolic rifling the
pressure at '3 feet of travel, corresponding to the point of maximum
pressure, is only 31 tons. The pressure then falls slightly, and
amounts to 28 tons at about 1 foot travel; thence it gradually
increases to a maximum of 36 tons at the muzzle.
By way of comparison, I have added in the Plate a curve showing
ROTATION TO RIFLED PROJECTILES 97
the pressures required to give rotation to a 400-lb. projectile fired
from the 10-inch gun with uniform twist when E. L. G-. instead of
pebble-powder is used.
The curve in this case is of the same nature as that derived from
the pebble-powder ; but the variation is greater, the maximum pres-
sure being much higher, and the muzzle-pressure, owing to the
smaller charge, somewhat less.
27. To one more point it is worth while to call attention.
If the gun were a smooth-bore gun, the equation of motion would
be
M.g = G' .... (24)
and comparing this equation with (10), we have, on the supposition*
that the velocity increments in both cases are equal,
G = G'
or, in the case of the Woolwich gun, where o = 90°,
G = G' + R.f4^±fli| . . . (26)
and the interpretation of these equations is that the gaseous pressure
in the rifled guns (rifled with the parabolic twist) is greater than
that in the smooth-bored gun by the second term of the right-hand
member of the equation.
28. The corresponding equations for a uniform twist are
or, if ^=90';
G = G' + R|4t±l| .... (28)
29. I shall now put these results in actual figures, and shall
again take for illustration the 10-inch gun, supposed (as before) to
be rifled, 1st, on the uniform, 2nd, on the parabolic or service twist.
With the uniform twist, G- (see Table) = 1547 tons; and using
Equation (28) and the values of the constants given in 22,
G' = G--24.5R
= -989G (29)
* Were the velocity increments not supposed equal, the reduction of pressure
due to the suppression of rifling would be less than that given in the text.
G
98 ROTATION TO RIFLED PROJECTILES
Hence the decrement of pressure due to the sujDpression of rifling is
only about 1 per cent. ; that is, the total pressure on the base of the
shot is reduced from 1547 tons to 1530 tons, or the bursting pres-
sure is reduced from 19-7 tons per square inch to 19*5 tons per
square inch.
At the muzzle of the gun in the same manner we find that the
total pressure is reduced from 205 tons to 202 8 tons, and the pres-
sure per inch in a corresponding proportion.
30. Similarly, from Equation (26) and the values of the constants
given in 25, the values of Gr' at the point of maximum pressure and
at the muzzle of the gun are obtained; and I find that with the
parabolic twist the pressure on the base of the shot would be reduced
from 1547 tons to 1541 tons, or the bursting pressure would be
reduced from 197 tons to 19-62 tons per square inch.
At the muzzle the corresponding reductions are from 205 tons
total pressure, to 196 tons, or from 2-61 tons to 2-49 tons per square
inch.
31. For the sake of clearness, I recapitulate the results at which
I have arrived. They are as follows : —
1st. That the pressures actually exerted at all points of the bore
to give rotation to the 10-inch British-service projectile, compared
with the pressures which would be exerted were the gun rifled on a
uniform twist, are very approximately exhibited in the diagrams on
opposite page.
2nd. That in the 10-inch gun (and other guns similarly rifled)
the pressure on the studs due to rifling is but a small fraction (about
21 per cent.) of the pressure required to give translation to the shot.
3rd. That the substitution of the parabolic for the uniform rifling
has reduced by about one-half the maximum pressure on the studs.
4th. That the increment of the gaseous pressure, or the pressure
tending to burst the gun, due to rifling is exceedingly small,* both
in the case of the uniform and parabolic rifling. This result is
entirely confirmed by the experiments of the Explosive Committee,
who have found no sensible difference of pressure in the 10-inch gun
fired in the rifled and unrifled states.
5th. That, small as the increment in gaseous pressure due to
rifling is, it is still less in the parabolic than in the uniform system
of rifling.
' Although the increase of strain due to rifling is inconsiderable, yet the decrease
of the strength of the structure of a gun inseparable from rifling may be, and in
many systems is, considerable ; but the discussion of this question is outside of the
.scope of my paper.
r
"
^
I
>c>
^
^^
—
1
•^
.t
\
i
^
%
^
■<^
1 1
^-
T)
I|
—
"^
"^
1
t?
>%
1
1
a
;i
1
^
1 1
-i
ft
"~~~
'
^
~^
■1
;'/
§
C9
o
Gases
fK.,CO,
•1264
grm.
0-6806-
K.S .
KCNS
•0327
•4227
•0213
•0030
KNO,
•0372
(NHJXOg .
O . . . .
•0286
•0014
Ic . . . .
•0073
rsH, .
grm.
•0018
= 1-16
0-3138
. .
CO .
•0014
•0094
= 1-00
= 7-49
0-9944
CO., .
H : .
•2012
•0002
= 101-71
= 2-34
In . .
•0998
= 79-40
193^10
In Table 3 a comparative statement is given of the foregoing
results with those of other recent experimenters, and with those
furnished by our investigations. (See p. 130.)
Bunsen and Schischkoff determined the number of units of heat
generated by combustion, by exploding a small charge of powder in a
tube immersed in water. They found that the combustion of a
gramme of powder gave rise to 620 gramme-units of heat ; and hence
they calculated that the temperature of explosion, in a^ close chamber
impervious to heat, was 3340° Cent. (5980° Fahr.).
From the above data the pressure in a close vessel is deducible ;
and they computed that the maximum pressure which the gas can
attain, which it may approximate to but can never reach, is about
4374 atmospheres, or 29 tons on the square inch.
Bunsen and Schischkoff further computed the total theoretical
work which a kilog. of gunpowder is capable of producing on a
projectile at 67,400 kilogrammetres.
In the course of our paper we shall have frequent occasion to refer
to these very important researches.
In 1858, D. J. Linck* repeated, with Wurtemburg war-powder,
Bunsen and Schischkoff analysis of the products of combustion,
which were obtained by the same method. The composition of the
powder used is given in Table 2, p. 128.
Linck's results, which we have placed in the same Table as those
of Bunsen and Schischkoff, differed in several points from the results
of the latter chemists, but chiefly in the much smaller quantity of
* Annalen der Chemie, vol. cix. p. 53.
110 RESEARCHES ON EXPLOSIVES
potassium sulphate found. Linck considered that 1 grm. of the
powder used generated 218"3 c.c. of gas.
In 1863, M. von Karolyi * examined the products of combustion of
Austrian musket- and ordnance-powder.
M. von Karolyi's method of obtaining the products of combustion
consisted in suspending in a spherical shell a small case containing a
charge of the powder to be experimented with. Before firing the
charge, the air contained in the shell was exhausted ; the powder was
fired by electricity.
The arrangement will readily be understood from the sketch
shown in Fig. 3, Plate XI. (p. 230).
After combustion, the gases were obtained for examination by
means of the stop-cock, while the solid residue remaining in the shell
was removed with water and filtered.
The composition of the powders used is given in Table 2 (p. 128),
and the results of analysis in Table 3, p. 130. Von Karolyi computed
that the gases resulting from 1 grm. of small-arm powder generated
2266 c.c, and from 1 grm. of ordnance-powder 200'9.
The Astronomer Eoyal, Sir G-. B. Airy, in a paper f published in
1863, " On the Numerical Expression of the Destructive Energy in
the Explosions of Steam-boilers, and on its comparison with the
Destructive Energy of Gunpowder," considers that " the destructive
energy of 1 cubic foot of water (6223 lbs. = 28-23 kilogs.) at the
temperature which produces the pressure of 60 lbs. to the square inch
is equal to that of 1 lb. of gunpowder, and that the destructive
energy of 1 cubic foot of water at the temperature which produces the
pressure of 60 lbs. to the square inch, surrounded by hot iron, is pre-
cisely equal to the destructive energy of 2 lbs. of gunpowder as fired
in a cannon."
Airy takes the energy of a kilog. of powder as fired from a gun at
56,656 kilog. metres = 82894 foot-tons per lb. of powder ; so that the
total energy of gunpowder would be somewhat less than double the
above value. He states, however, that this estimate does not pretend
to be very accurate.
In 1869 were published, in the Zeitschrift fur Cheviie,% the results
of some experiments made by Colonel Fedorow to determine whether
the products varied materially with the mode of combustion.
Fedorow experimented (1) by firing a pistol with a blank charge
* Pogf?endorff's Annalen, April 1863. PhUosophkal Mdgazive, ser. 4, vol. xxvi.
p. 266.
t PhilosophkaL Mmjaziw, ser. 4, vol. xxvi. p. o29. % Und. vol. v. p. 1 2.
RESEARCHES ON EXPLOSIVES 111
into a glass tube 4 feet long, (2) and by firing a shotted 9-pr. bronze
gun with 3 lbs. of powder ; the residues were in each case dissolved in
water and analysed.
The composition of the powder employed by Fedorow is given in
Table 2, and his analytical results are shown in Table 3.
From the experiments with the gun, Fedorow calculated that the
gaseous products were 826 c.c. N, 1621 c.c. CO.2, and 14 c.c. SO.^ and
O, He considers that several successive reactions take place during
combustion, that potassium sulphate and carbonic anhydride are first
formed, while the excess of carbon reduces the sulphate to carbonate,
hyposulphite, and carbonic anhydride.
In 1871, Captain Noble,* one of the present writers, in detailing
to the Eojal Institution his earlier researches on the tension of fired
gunpowder, stated that the conclusion at which he had arrived from
the results of his experiments, where the products of combustion were
entirely or partially confined, was, that the maximum pressure of
fired gunpowder, of the usual gravimetric density, when unrelieved
by expansion, did not greatly exceed 6100 atmospheres (40 tons to
the square inch). Upon the same occasion a curve was exhibited,
showing the relation between the tension and the density of the
exploded products. These results have been confirmed by our
present more extensive and exact investigations.
Captain Noble also stated that, by means of a special apparatus
which was fully described at the time, he had not only determined
the tension of the gases at various densities, but had exploded con-
siderable charges filling entirely the chambers of close vessels, and
had altogether retained and at pleasure discharged the gaseous and
other products of combustion.f
Berthelot J published, in 1872, a collection of theoretical papers
upon the force of powder and other explosive substances.
Berthelot does not attempt to evaluate the force of fired gun-
powder, but evidently accepts as tolerably correct § the tensions
assigned by Eumford and Piobert, and accounts for the discrepancy
* Proceedings of Royal Institution, vol. vi. p. 282. Revue Scientifique, No. 4S,
p. 1125.
t In the present paper, in Section K, the results of some of Capt. Noble's
earlier experiments are given. They accord, as will be seen, exceedingly well with
the series we have discussed at length ; but a few experiments made with a fine-
grained powder are excluded, both because the powder, being sporting, was not
comparable with the fine-grain used in the present researches, and because the
differences in their composition are unknown, the sporting-powder not having been
analysed.
X Sur la Force de la Poudre. Paris, 1872. •^j Loc. cit. p. 80.
112 RESEARCHES ON EXPLOSIVES
between their conclusions and those of the modern chemists by
assuming that the laws of Mariotte and Gay-Lussac lose all physical
significance for pressures so enormous as those developed in the
combustion of gunpowder.
Berthelot is disposed * to think that dissociation plays a consider-
able role during the expansion of the products in the bore of a gun.
He supposes that the phenomena of dissociation do not exercise their
influence only during the period of maximum effect, but that, during
the expansion of the gases, a cooling effect is produced, by which a
more complete combination is effected and more heat disengaged.
Taking Bunsen and Schischkoff's experiments as a basis, Berthe-
lot expresses the decomposition experienced by gunpowder by the
equation f
I6KNO3 + 6S + 13C = 5K,SO^ + 2K,C03 + K.S + 16N + 1 ICO,
which he considers represents their resvilts with sufficient exactness.
In 1873, M. de TromenecJ communicated to the Academy of
Sciences a short memoir on the means of comparing the absolute
force of varieties of powder. His method was based upon the
principle that, when a body is exploded without producing mechanical
effect, the " force disponible " is converted into heat, and that it is
only necessary to explode a given weight in a close vessel and
determine the heat produced.
The apparatus used by De Tromenec was closed in much the
same manner as was that employed by Captain Noble in his earlier
experiments already alluded to. The three kinds of powder experi-
mented with gave results varying between 729 and 891 calories
generated by the combustion of 1 kilog. of powder.
In the same § number of the Comptes Rendus in which De
Tromenec's memoir is given, appears a note by MM. Eoux and
Sarrau, in which, and in a subsequent note,j| are determined, with
small charges, some of the points to which our own investigations
have been specially directed.
MM. Eoux and Sarrau have given, for five species of powder, the
number of calories and volume of gas generated by a given weight of
powder, and have from these data calculated the temperature of com-
bustion and tension of the gas.
With one of the powders, representing closely the composition of
those chiefly experimented with by us, the number of calories and
* Sur la Force de la Poudre, p. 83. f Loc. cit. p. 91.
I Comptes Rendus de VAcadimie des Sciences, torn. Ixxvii. p. 126.
§ Comptes Rendus, torn. Ixxvii. p. 138. || Ibid. p. 478.
RESEARCHES ON EXPLOSIVES 113
volume of the gas agree nearly exactly with the numbers found by
ourselves. There is, however, a considerable difference in our deter-
minations (both theoretical and experimental) of the tension of the
gas and also of the temperature of explosion, the temperature being
estimated by Koux and Sarrau at about 4200" Cent., and the tension
at about 4700 atmospheres.
"We shall return, however, to these points when discussing our
own experiments.
(&) OBJECTS OF EXPERIMENTS.
The chief objects which we had in view in making these investi-
gations were : —
First. To ascertain the products of combustion of gunpowder fired
under circumstances similar to those which exist when it is exploded
in guns or mines.
Second. To ascertain the tension of the products of combustion at
the moment of explosion, and to determine the law according to
which the tension varies with the gravimetric density of the powder.
Tliird. To ascertain whether any, and, if so, what well-defined
variation in the nature or proportions of the products accompanies a
change in the density or size of grains of the powder.
Fourth. To determine whether any, and, if so, what influence is
exerted on the nature of the metamorphosis by the pressure under
which the gunpowder is fired.
Fifth. To determine the volume of permanent gases liberated by
the explosion.
Sixth. To compare the explosion of gunpowder fired in a close
vessel with that of similar gunpowder when fired in the bore of a
gun.
Seventh. To determine the heat generated by the combustion of
gunpowder, and thence to deduce the temperature at the instant of
explosion.
Eighth. To determine the work which gunpowder is capable of
performing on a shot in the bore of a gun, and thence to ascertain
the total theoretical work, if the bore be supposed of indefinite length.
(c) METHODS OF EXPERIMENT.
1. Explosion-apparatus.
We propose, in the first place, to describe the principal apparatus
used in these investigations, and shall commence with that portion
H
114 RESEARCHES ON EXPLOSIVES
which is of primary importance, viz., the vessel in which the
explosions were produced. Two sizes of vessels were used, the
larger being capable of holding about 2^ lbs. (1 kilog.) of powder,
the other being about half that capacity.
Both vessels were of the same general construction, and similar
to that described in Captain Noble's Lecture at the Eoyal Institu-
tion already referred to. A drawing of the apparatus is given in
Plate X., Figs. 2 and 3 (p. 230).
A (see Figs. 2 and 3) is a mild steel vessel of great strength,
carefully tempered in oil, in the chamber of which (B) the charge
to be exploded is placed.
The main orifice of the chamber is closed by a screwed plug
(C), called the firing-plug, which is fitted and ground into its
place with great exactness.
In the firing-phig itself is a conical hole, which is stopped by
the plug D, also ground into its place with great accuracy. As
the firing-plug is generally placed on the top of the cylinder, and
as, before firing, the conical plug would drop into the chamber if
not held, it is retained in position by means of the set-screw S,
between which and the cylinder a small washer (W) of ebonite
is placed. After firing, the cone is, of course, firmly held, and
the only effect of internal pressure is more completely to seal the
aperture. At E is the arrangement for letting the gases escape;
the small hole F communicates with the chamber where the
powder is fired, and perfect tightness is secured by means of the
mitred surface G. When it is wished to let the gases escape,
the screw E is slightly withdrawn, and the gas passes into the
passage H.
At K is placed the "crusher-apparatus" for determining the
tension at the moment of explosion.
When it is desired to explode a charge, the crusher-apparatus,
after due preparation, is first carefuUy screwed into its place, and
the hole F closed. The cone in the firing-plug is covered with
the finest tissue-paper, to act as an insulator.
The two wires LL, one in the insulated cone, the other in the
cylinder, are connected by a very fine platinum wire passing
through a small glass tube filled with mealed powder. Upon
completing connection with a Daniell's battery, the charge is fired.
The only audible indication of the explosion is a slight click ;
but frequently, upon approaching the nose to the apparatus, a
faint smell of sulphuretted hydrogen is perceptiljle.
RESEARCHES ON EXPLOSIVES 115
The difficulties we have met with in using this apparatus are
more serious than might at first sight appear.
In the first place, the dangerous nature of these experiments
rendered the greatest caution necessary, while, as regards the
retention of the products, the application of contrivances of well-
known efficacy for closing the joints, such as papi&r-mdclU wads
between discs of metal (a method which has been successfully
employed with guns), are inadmissible, because the destruction of
the closing or cementing material used, by the heat, woiild
obviously affect the composition of the gas. Every operation con-
nected with the preparation of the apparatus for an experiment
has to be conducted with the most scrupulous care. Should any
of the screws not be perfectly home, so that no appreciable amount
of gas can escape, the gases, instantly upon their generation, will
either cut a way out for themselves, escaping with the violence of
an explosion, or will blow out the part improperly secured, in
either case destroying the apparatus.
The effect produced upon the apparatus, when the gas has
escaped by cutting a passage for itself, is very curious. If, for
example, one of the plugs has not been sufficiently screwed home,
so that the products of combustion escape between the male and
female threads, the whole of these threads at the point of escape
present the appearance of being washed away, the metal having
been evidently in a state of fusion, and carried over the surface
of the plug by the rush of the highly-heated products.
Again, the difficulty of opening the vessel after explosion, when
large charges have been used, is very great. This will be readily
understood when the temperature and pressure of explosion are
considered. The exploding-chamber being filled with products
intensely heated and under an enormous pressure, there is an
expansion of the interior surface of the cylinder. Hence small
portions of the fluid products become forced in between the
threads of the screws. These solidify into a substance of intense
hardness, which cements together the metal surfaces, and, on
cooling, the contraction of the cylinder puts such a pressure on
the screw, that, in attempting to open it, seizure is very difficult
to avoid. In one or two cases it was found impossible to open
the cylinder until melted iron had been run round it, so as to
cause it to expand.
This difficulty has been in a great measure avoided, in the
more recent experiments, by making the screws conical, so that
116 RESEARCHES ON EXPLOSIVES
when once started clearance is rapidly given, and they are removed
with comparative ease.
2. Measurement of Pressure.
The apparatus used for the measurement of the tension of the
gas was precisely similar to that which has been used by the
Committee on Explosives, and consists of a screw-plug of steel
(Plate X., Figs. 4 and 5, p. 230), which admits of a cylinder of copper
or other material (A) being placed in the small chamber (B). The
entrance to the chamber is closed by the movable piston (C), and
the admission of the gas is prevented by the use of the gas-check
(D). When the powder is fired, the gas acts upon the base of the
piston and compresses the cylinder. The amount of compression
of the cylinder serves as an index to the force exerted, the rela-
tion between the amount of crush and the pressure necessary to
produce it being previously carefully determined.
3. Measurement of the Volume of the Permanent Gases.
The apparatus used for the measurement of the permanent gases
is shown in Plate XI., Figs. 1 and 2 (p. 230). A is a vessel the annular
space (B) of which is filled with water ; on the surface of this a thin
film of oil is floated, to prevent any slight absorption of the gas
which might otherwise take place.
Immediately after the explosion of a charge, the gas from which
it is desired to measure, the cylinder (0) containing the products i&
placed on the table (D), and the gasometer (E) is placed over the
cylinder ; the height of the water on the glass scale (F) being then
registered, the escape-screw (G) of the cylinder is turned, by means
of a turn-cock passing through the stuffing-box (M).
When the gas has all escaped, the height indicated on the glass
scale being again registered, the cubic contents are known, and the
thermometer (H) and height of barometer being noted, the necessary
data are available for reducing the volume of the gas to a tempera-
ture of 0° Cent, and a barometric pressure of 760 mm.
4. Measurement of Heat.
To determine the heat generated by explosion, a charge of powder
was weighed and placed in one of the smaller 'cylinders described,
which was kept for some hours in a room of very uniform tempera-
RESEARCHES ON EXPLOSIVES 117
ture. When the apparatus was throughout of the same temperature,
the thermometer was read, the cylinder closed, and the charge
•exploded.
Immediately after explosion the cylinder was placed in a calori-
meter containing a given weight of water at a measured temperature,
the vessel being carefully protected from radiation, and its calorific
value in water having been previously determined.
The uniform transmission of heat through the entire volume of
water was maintained by agitation of the liquid, and the thermometer
was read every five minutes until the maximum was reached. The
observations were then continued for an equal time to determine the
loss of heat in the calorimeter due to radiation, etc. ; the amount so
determined was added to the maximum temperature.
In this method there is a possible source of error ; the walls of
the cylinder being of very considerable thickness, it is obvious that,
although the outer surface of the cylinder must be of the same
temperature as the water, it by no means follows that this is true of
the internal surface ; consequently the loss of heat due to radiation,
etc., may be in some degree compensated by a flow of heat from the
interior.
We had reason, from some experiments we made, to believe that
the error due to this cause was very small ; and our views were con-
firmed by finding no appreciable rise of temperature on placing some
water from the calorimeter into the chamber of the cylinder immedi-
ately after an experiment.
5. Collection of Gaseous Products.
To collect the gases for analysis, a small pipe was screwed into the
escape-passage (H) of the cylinder (Plate X., Figs. 2 and 3, p. 230),
and an indiarubber tube, terminating in a glass nozzle, was led to a
mercurial trough. Before the gas was taken, a sufficient quantity
was allowed to escape to clear the tubes of air ; the gas was then
collected in tubes over mercury, and confined in the usual manner by
sealing them with the blowpipe.
The gas was generally collected in from five to fifteen minutes
from the time of explosion. Owing to the dangerous nature of the
experiments, and the precautions necessary to be adopted in explod-
ing such considerable charges of powder, it was not generally possible
to collect the gases more rapidly ; but a comparison of the analysis
of different tubes taken from the same experiment has shown that,
118 RESEARCHES ON EXPLOSIVES
at all events within moderate limits, no change takes place in the
composition of the gas by its continued contact with the solid pro-
ducts.
6. Collection of Solid Products.
The collection of the solid products presented much more difficulty
than that of the gaseous products. On opening the cylinder, the
whole of the solid products were found collected at the bottom, there
being generally an exceedingly thin (in fact, with large charges,
quite an inappreciable) deposit on the sides. Upon the firing-plug
there was usually a button of deposit, which differed considerably
both in appearance and in chemical composition from the rest. In
the button a crystalline structure was quite apparent, some of the
crystals being large and transparent. The surface of the deposit was
generally perfectly smooth, and of a very dark grey, almost black,
colour. This colour, however, was only superficial, and through the
black could be perceived what was probably the real colour of the
surface, a dark olive-green. The surface of the deposit, and the sides
of the cylinders, had a somewhat greasy appearance, and were indeed
greasy to the touch. On the smooth surface were frequently observed
very minute particles, in , appearance like soot, but of the greasy
texture to which we have alluded.
The removal of the deposit was generally attended with great
difficulty, as it formed an exceedingly hard and compact mass, which
always had to be cut out with steel chisels. Lumps would frequently
break off, but a considerable portion flew off before the chisel in fine
dust. In various experiments, on examining the fracture as ex-
hibited by the lumps, the variation in physical appearance was very
striking, there being marked differences in colour, and also, frequently,
a marked absence of homogeneity, patches of different colours being
interspersed with the more uniform shade of the fracture. There
was no appearance of general crystalline structure in the deposit;
but, on examination with a microscope and sometimes with the naked
eye, shining crystals of metallic lustre (sulphide of iron) were observed.
On the whole, the general appearance of the deposit was attended
with such considerable variations, that, for minute details, we must
refer to the account of the experiments themselves. The deposit
always smelt powerfully of sulphuretted hydrogen, and frequently
strongly of ammonia. It was always exceedingly deliquescent, and
after a short exposure to the air became black on the surface, gradu-
ally passing over into an inky-looking pasty mass. As in physical
RESEARCHES ON EXPLOSIVES 119
appearance, so in behaviour of the solid, when removed from the
cylinder, there were considerable differences between the experi-
ments. The deposit was transferred to thoroughly dried and warm
bottles, and sealed up as rapidly as possible. In most cases, during
the very short time that elapsed while the transference was being
made, no apparent change took place ; but in some a great tendency
to development of heat was apparent ; and in one instance, in which
a portion of the deposit (exhibiting this tendency in a high degree)
was kept exposed to the action of the air, the rise of temperature
was so great that the paper on which it was placed became charred,
and the deposit itself changed colour with great rapidity, becoming a
bright orange-yellow on the surface.
This tendency to heating always disappeared when the deposit
was confined in a bottle and fresh access of air excluded.
The portion of the residue which could not be removed from the
cylinder in a dry state was dissolved out with water, the solution
being reserved for examination in well-closed bottles.
{d) ANALYSIS OF THE PEODUCTS OF EXPLOSION.
1. Gaseous Prodiiets.
The method pursued for the analysis of tlie gaseous products of
explosion presented only one important point of difference from that
pursued by Bunsen and Schischkoff. The volume of gas at command
being more considerable than was the case in the investigations of
those chemists, it was found more convenient to have recourse to
methods for determining the sulphuretted hydrogen differing from
that which they adopted — namely, its estimation by oxidation of the
sulphur in the ball of potassium hydrate employed for absorbing the
carbonic anhydride and sulphuretted hydrogen together. In some
instances the volume of this gas was ascertained by absorption with
manganese balls, but generally the following indirect method was
pursued. The combined volume of carbonic anhydride and sulphu-
retted hydrogen was determined in one portion of the gas by means
of potassium hydrate ; another portion of gas was then treated with
a small quantity of cupric sulphate, and the volume of carbonic
anhydride determined in the gas thus freed from sulphuretted
hydrogen.
The following numerical data relating to the analysis of the gases
obtained by the explosion of 190"5 grms. of E. L. G. gunpowder (of
120
RESEARCHES ON EXPLOSIVES
Waltham-Abbey manufacture) in five times its own space, are given
in illustration of the detailed result obtained : —
1. Original volume of gas .
2. After absorption of CO.,
andSH, . . '.
3. After absorption of oxygen
I.
Volume.
Tempe-
rature.
Pressure.
Volume corrected
for temperature
and pressure.
144-4
13-3
0-7243
99-80
78-2
13-3
0-6727
50-16
76-9
14-4
0-6795
49-64
Volume oforiginal gas after
absorption of sulphu-
retted hydrogen .
After absorption of CO., .
After absorption of oxygen
144-2
82-2
80-6
14-2
16-3
18-8
0-7293
0-6672
0-6735
99-97
51-76
50-79
III.
7. Portion of 3 transferred to
eudiometer .
8.*After addition of air
After addition of oxygen .
After explosion with oxy-
hydrogen gas
After absorption of CO^ .
Portion of 11 transferred
to clean eudiometer
After addition of hydrogen
After explosion (dry)
174-8
15-4
0-1983
32-81
248-4
15-5
0-2712
63-75
319-5
15-6
0-3427
103-58
310-8
15-8
0-3302
97-02
291-6
18-3
0-3271
89-39
301-5
18-6
0-3141
88-66
550-8
18-9
0-5642
290-85
416-0
18-8
0-4295
167-16
By calculation from the above data, the composition of this gas,
in volumes per cent., was found to be as follows : —
Carbonic anhydride
46-17
Sulphuretted hydrogen
3-91
Oxygen ....
0-52
Carbonic oxide
11-46
Marsh-gas ....
003
Hydrogen ....
2-72
Nitrogen ....
35-18
* Air was added to dilute the gas in this and one or two subsequent explosion
experiments ; but this precaution was found to be unnecessary, and was therefore
not continued.
RESEARCHES ON EXPLOSIVES 121
The gas in each experiment was generally collected in three or
four large tubes. The contents in one tube sufficed, in most in-
stances, for the complete analysis; but the results obtained were
always controlled by determinations of several, if not of the whole, of
the constituents in the contents of another tube. Only in one in-
stance were the contents of different tubes, collected from one and
the same experiment, found to differ materially in composition ; in
this particular instance the proportion of sulphuretted hydrogen in
the different tubes was discordant. The mean of the results furnished
by the contents of the three tubes was taken to represent the com-
position of the gas.
2. Solid Residue. — Preparation of the Hesichie for Analysis.
The residue, as collected for analysis, consisted of one or more
large masses, besides a quantity in a more or less fine state of
division which had been detached from the sides of the vessel. The
appearance presented by the large pieces themselves indicated that
they were by no means homogeneous, and they evidently differed in
some respects from the smaller particles just referred to ; moreover,
the foreign matters (metal and glass) could not be expected to be
uniformly distributed throughout the mass, and a chemical examina-
tion of the latter clearly indicated that certain constituents existed in
different proportions in the upper and lower parts of the residue.
For these reasons, in order to ensure the attainment of results
correctly representing the composition of the residue, it appeared
indispensable to operate upon the entire quantity at one time, with
the view of determiuing the total amount of matter insoluble in
water, and of preparing a solution of uniform composition in which
the several components of the residue could be estimated. As the
investigation proceeded, much inconvenience and delay were experi-
enced from the necessity of working with very large quantities (from
400 to 100 grms.), which rendered the filtration s and washings pro-
tracted operations, and necessitated dealing with very large volumes
of liquid. It was therefore attempted to expedite the examination of
the residues by so preparing them that only portions might be
operated upon at one time in conducting the individual deter-
minations of the constituents. The impossibility of pulverising
and mixing the residue by any ordinary mode of proceeding, on
account of the rapidity with which oxygen and water were absorbed
from the air, was demonstrated by two or three attempts. An
122 RESEARCHES ON EXPLOSIVES
arrangement was therefore devised for performing the operation in an
atmosphere of pure nitrogen. The gas employed was prepared in the
following manner : —
A gasometer filled with air was submitted to a gentle pressure
causing the air to flow very slowly through a delivery-pipe to a
porcelain tulie filled with copper turnings and raised to a red heat.
To remove any traces of oxygen, the nitrogen passed from the tube
through two Woulfe's bottles containing pyrogallic acid dissolved in a
solution of potassium hydrate ; and, finally, to remove moisture, it
passed through two U-tubes filled with pumicestone moistened with
sulphuric acid. The nitrogen thus obtained was collected in india-
rubber bags ; the residue was placed in a closed mill, connected by
an indiarubber tube with the gas-bag, which was subjected to a
considerable pressure to establish a plenum in the mill. The substance
was then ground, and allowed to fall into bottles, which were at once
sealed. By this treatment a sufficient degree of uniformity in
different samples of any particular residue was generally attained ;
in some cases, however, the state of division of the substance was
not sufficiently fine to secure such intimacy of mixture as would
preclude the occurrence of discrepancies in the analytical results
furnished by different samples. It was therefore found necessary to
return occasionally to the employment of the entire residue obtained
in one experiment for determining its composition.
3. Analysis of the Solid Residue.
Qualitative analysis indicated that the proportions of the following
substances had to be determined in the solid residue.
a. Portion insoluble in loater. — This consisted of steel (unavoidably
detached from the interior of the vessel during removal of the
residue) and of small quantities of other metals, besides glass, which
were used in the construction of the electric igniting arrangement.
The weight of these substances was deducted from the residue, as
foreign to the research.
In addition to these substances, the residue insoluble in water
contained generally traces of charcoal, besides sulphur, which was
combined with iron and portions of the other metals, and the amount
of which is included in the statement of results as free sulphur,
together with the proportion which was found, in combination with
potassium, in excess of the amount required to form the mono-
sulphido.
RESEARCHES ON EXPLOSIVES 123
h. Portion soluble in water. — In this, the chief portion of the
residue, there existed the potassium sulphide, sulphocyanate,
hj-posulphite, sulphate, carbonate, and nitrate, besides ammonium
carbonate, and, in very exceptional cases, potassium hydrate. The
estimation of the proportions in which these several constituents
existed in the residue was conducted as follows:—
c. Water contained in the residue. — It is obvious that the highly
hygroscopic nature of the powder-residue rendered it impossible to
transfer the product of an explosion from the iron cylinder to suitable
receptacles for its preservation out of contact with the atmosphere
without some absorption of moisture, however expeditiously the
operation was performed. Moreover, any water produced during the
explosion, or pre-existing in the powder, would necessarily be retained
by the solid residue after explosion, as the gas remained in contact
with a large surface of this powerful desiccating agent for some time
before it could be collected. In some instances the water was
expelled from the residue by exposing it for some time to a slow
current of hydrogen at 300^ Cent., the gas and volatile matters being
passed into solution of lead acetate, for the purpose of retaining
sulphur, and the weight of the dried residue determined. The
amount of residue, however, was generally too considerable for this
operation to be satisfactorily performed ; there was therefore no
alternative in such cases but to assume that the difference between
the total weight of the residue and the combined weights of its
several solid constituents, ascertained in almost every instance by
duplicate and check determinations, represented the amount of water
present in the substance.*
d. Separation of the portion insoluUe in water, and determination
of sulphur in it.— The separation was accomplished by thoroughly
washing the entire residue, or about 7 grms. of the ground residue,
with well-boiled water until no discoloration was produced in the
washings by lead acetate. Boiled water was employed to avoid
oxidation of any of the constituents. After drying and washing the
residue, it was introduced, with its filter, into a small flask ; a little
potassium bichromate was added before addition of nitric acid, to
guard against violent reaction and the possibility of minute quantities
of sulphur escaping as sulphuretted hydrogen. The oxidation was
completed by the addition of potassium chlorate; the liquid, after
* If discrepancies existed between the results of determination of the several
constituents and the check-determinations, the water was estimated, as described,
in a portion of the residue.
124 RESEARCHES ON EXPLOSIVES
sufficient dilution, was filtered and evaporated, the residue redis-
solved in water, with addition of chlorhydric acid, and the sulphuric
acid determined in the solution by the usual method.
The proportion of charcoal contained in the insoluble residue
was, in most instances, so small that no importance could be
attached to any attempt to determine the quantity. In a few
cases its amount was determined by combustion.
e. Fotassium monosulphide. — The method pursued differed but
very slightly from that adopted by Bunsen and Schischkoff. The
aqueous solution, separated from the insoluble portion, was digested
with pure ignited cupric oxide in a well-closed flask, with occa-
sional agitation, until it became colourless. The oxide containing
sulphide was then filtered off, thoroughly washed, and the sulphur
was determined in it by oxidation according to the method just
described (d).
f. Fotassium sulphate. — The filtrate obtained after the treat-
ment with cupric oxide just described (or a measured quantity of
it, if the entire residue was operated upon at one time) was
mixed with chlorhydric acid and boiled to expel the sulphurous
acid resulting from the decomposition of hyposulphite; the liquid
was then separated by filtration from liberated sulphur, and the
sulphuric acid determined as barium sulphate.
g. Potassium hyposulphite. — The solution obtained by treatment,
as above described, of about 4 grms. of the residue (or a sufficient
volume prepared from the entire residue) was acidulated with
acetic acid; 3 or 4 c.c. of starch solution were added, and the
hyposulphurous acid determined by means of a standard iodine
solution.
h. Potassium sulphocyanate. — A solution of the residue, after
separation of the insoluble portion and the soluble sulphide, was
carefully acidified with a measured quantity of dilute chlorhydric
acid, so as to avoid separation of sulphur. The oxidation of the
hyposulphite was then effected by the gradual addition of a very
dilute solution of ferric chloride until the liquid exhibited a
permanent pink tint. A measured quantity of the ferric solution
was afterwards gradually added until the greatest attainable depth
of colour was produced. To determine what was the amount of
sulphocyanate thus arrived at, a volume of water corresponding
to that of the original solution tested was mixed with equal
volumes of the dilute chlorhydric acid and ferric chloride to those
used in the previous experiments. A solution of potassium sulpho-
RESEARCHES ON EXPLOSIVES 125
cyanate of known strength was then gradually added until a depth
of colour corresponding to that of the actual assay was produced.
i. Potassium carbonate,. — After the usual treatment of a solution
of the residue with cupric oxide, pure manganous sulphate or chloride
was added to the liquid in excess ; the resulting precipitate might
generally be washed by decantation in the first instance; after
complete washing it was transferred to a small flask suitably
fitted for the liberation of carbonic anhydride from it, by addition
of sulphuric acid, and for the transmission of the gas through
small weighed absorption-tubes containing respectively sulphuric
acid, calcium chloride, and solution of potassium hydrate. The
increase in weight of the latter corresponded to the proportion of
carbonic anhydride in the solid residue.
y. Potassium sulphide, potassium carbonate, and potassium hydrate.
— Pure manganous chloride or sulphate was added in excess to the
aqueous solution of the residue, and the amount of manganese, in
the thoroughly washed precipitate, determined as red oxide. If
the amount obtained exceeded those which would be furnished by
the potassium sulphide and carbonate (deduced from the previous
determinations), the excess was taken to correspond to potassium
hydrate existing in the residue. If it was less, the sulphur exist-
ing as monosulphide of potassium was calculated from the weight
of the manganous oxide, and the difference between it and the
sulphur found in the cupric oxide (in determination e) was taken
to represent excess of sulphur, or fi^ee sulphur, and was added to
the result of determination d, the necessary correction being made
in the number furnished by determination e.
h. Total amount of potassium. — The solution of the residue, after
treatment with cupric oxide, was evaporated with excess of sulphuric
acid, and the residue repeatedly treated with ammonium carbonate
and ignited, until the weight of potassium sulphate was constant.
Or water and sulphuric acid were added to about 4 grms. of the
residue, and after boiling to expel sulphurous acid, two or three
drops of nitric acid were added to peroxidise the little iron in
solution and excess of ammonia to precipitate the latter. The
precipitate and insoluble matters (glass, etc.) were then filtered off,
and the solution evaporated, the weight of potassium sulphate being
ascertained by treatment of the residue as already described. In
this way the amount of potassium arrived at indirectly, by the
determinations of the several substances with which it existed in
combination, was controlled by direct estimation.
12G rp:searches on explosives
I. Ammonium sesquicarhonate. — The solution of about 12 grms.
of the residue was diluted to 1 litre; the liquid was then
carefully distilled until about 250 c.c. remained in the retort,
the distillate being allowed to pass into dilute chlorhydric acid.
As some minute quantities of potassium salt might have passed
over, the distillate was returned to a retort, mixed with excess of
sodium carbonate and again distilled, the product passing into
dilute chlorhydric acid. This second distillate was evaporated,
and the ammonium determined as platinum salt with the usual
precautions, the weight of the latter being controlled by ignition
and determination of the weight of the platinum.
m. Potassium nitrate. — The portion of solution remaining in
the retort, after the first distillation above described, was acidified
with sulphuric acid ; a piece of thin sheet zinc was then placed in
the liquid and allowed to remain for a week, a small quantity of
sulphuric acid being occasionally added. After the lapse of that
time the zinc was removed, and the ammonia produced from any
nitrate existing in the liquid was determined exactly as at I.
(e) COMPOSITION OF THE GUNPOWDERS EMPLOYED.
The method pursued in determining the proportions of proximate
constituents in the samples of gunpowder present but very few
points of difference from those ordinarily adopted, and need
therefore not be detailed.
It may be mentioned, however, with reference to the deter-
mination of the proportion of saltpetre, that a very appreciable
amount of the most finely-divided particles of the charcoal generally
passes through the filter during the final washings, however care-
fully the operation be conducted.
These last washings, which contain only a very small proportion
of the saltpetre, were therefore evaporated separately, and the
residue was carefully heated until the small quantity of charcoal
was completely oxidised. The resulting carbonate was then con-
verted into nitrate by careful treatment with dilute nitric acid,
and the product added to the remainder of the saltpetre previously
extracted.
The composition of the charcoal contained in the powders was
determined by combustion, after as complete a separation of the
other constituents as possible. There was, of course, no difficulty
in completely extracting the saltpetre; but the sulphur cannot be
RESEARCHES ON EXPLOSIVES 127
entirely removed from the charcoal by digestion and repeated
washings with pure carbon disulphide. The amount remaining
was therefore always determined by oxidation of the charcoal, and
estimation of sulphuric acid produced; the necessary correction
thus arrived at was made in the amount of charcoal used for
analysis. The latter was dried by exposing it for some time (in
the platinum boat in which it was to be burned) to a temperature
of about 170° in a current of pure dry hydrogen; it was allowed
nearly to cool in this gas, and dry air was then passed over for
some time, the boat being afterwards rapidly transferred to a
well-stoppered tube for weighing. The dried charcoal was burned
in a very slow current of pure dry oxygen, the resulting products
being allowed to pass over the red-hot cupric oxide, and finally
over a layer of about 8 inches of lead chromate, heated to incipient
redness. The efficiency of this layer in retaining all sulphurous
acid was fully established by preliminary test experiments.
The following tabular statement (Table 2, p. 128) gives the per-
centage composition of the five samples * of gunpowder employed
in these investigations as deduced from the analytical results.
In every instance at least two determinations were made of each
constituent, the means of closely concordant results being given in
the table.
This table also includes the results of analysis by Bunsen and
Schischkoff, Karolyi, Linck, and Federow, of the gunpowders
employed in their experiments.
It will be seen that the several English service-powders of
Waltham-Abbey manufacture did not differ from each other very
importantly in composition ; the most noteworthy points of difference
are the somewhat low proportion of saltpetre in the F. G. powder
and the slightly higher proportion of carbon in the pebble-powder.
The charcoals contained in these powders presented some decided
differences in composition, as is shown by the following comparative
statement : —
Pebble.
R. L. G.
E. F. G.
F.G.
Carbon
85-26
80-32
75-72
77-88
Hydrogen .
2-98
3-08
3-70
3-37
Oxygen
10-16
14-75
18-84
17-60
Ash .
1-60
1-85
1-74
1-15
* The authors are indebted to Colonel C. W. Younghusband, R.A., F.R.S., the
Superintendent of the Waltham-Abbey Gunpowder Works, for having selected and
furnished to them the samples of English gunpowder employed in their investiga-
tions.
128
RESEARCHES ON EXPLOSIVES
1c
O t- C^ (M '*< »C
t^
OC' OS 00 o
II
>C O O (>] ^ o
•c
rH OO GO ,;H
•^- Oi Tj< Ah
ftO
t» ft
lO 00 QO co^
11
lO 00 tH rji
•S-S
CD (JO <» 5p
1^,^ cpoo
g-2
00 O A( o
O O 00 o
^'
^^
«
i
■^
1
g"
O ^ • (M Ol 00
u
CO «» l~
^
• ^
>o CO : o >.* 'f
fcf-S
CD -^ o-j :
i
}l
K) o : o 4n ,1h
Si'?'
4tt eq c-i :
"^
4J
a
00 TH o ,^
£
3|o
j^ ;S
s
S
— o c^ o
2
k|
ea b
s
o,
.£P
.2
S
■§
e
•S'-
■«
w
(D
|1
-f< tJ< • 00 CTi O
S
^
U-; 0?. i>-
1
lo o : Ci ^ o
l^ I—I
5
o
Bi|
c-^ ?^ :
t^ 00 -3< :
„
*
^
2§a
S^
w <>] CO ^
^.2 ft
OO (M OS CO
^
1
o
■pip ^ Oi
o o c^ o
1
M i3 a
W ^ l^ CM
•S 1
ft
a
s
O
ll
1
^
"S
1
ll
a
ft
.= £•
lO »C ■ 1- (M r-l
1
A
00 o o> :
t4
Oi r^ . C tV» ©tc, of the vessel was
successively filled with powder, which was then fired, and the
resulting pressures determined.
Later on it was found that, as with each description of powder the
gravimetric density varied, it was more convenient to refer the pressure
not, as at first, to a density arrived at by taking the weight of powder
which completely filled a given space as unity, but to the specific
gravity of water as unity. The densities given hereafter must therefore
be taken to represent the mean density of the powder inclusive of the
interstitial spaces between the grains, or, what is the same thing, the
mean density of the products of explosion referred to water as unity.
The gravimetric density of the modern pebble jDowder closely approxi-
mates to 1 ;* that of the old class of cannon-powders, such as L. G.,
E. L. G., etc., varied generally between f "870 and -920 ; that of R G.
and sporting-powders was still lower.
The results of the whole of our experiments, as far as they relate
to tension, arranged according to the three descriptions of the powder
used and to the density of the products of explosion, are given in
Table 5. The experiments numbered with an asterisk are taken
from the earlier series made by Captain Noble. They accord very
well with the present experiments ; but the powder used in the first
series not having been analysed, we are not prepared to say that it
was of exactly the same constitution as the corresponding kind of
powder used in the present experiments, although the difference could
of course be but very trifling, it being gunpowder of Waltham-Abbey
manufactui^e, which, as shown by the analyses given in Table 2,
varies very little in composition.
* This statement applies only to the powder taken in considerable bulk. In our
explosion-vessels, the gravimetric density, when they were completely filled, did
not exceed, with pebble-powder, -92 or -93. The statement, therefore, that the
powder was fired in so many per cent, of space does not actually refer to the space
occupied in the chamber, but to a chamber of a size that would hold powder of our
standard density.
t Boxer, Gen., R.A., Treatise on Artillery, 1859, p. 21. Mordecai, Major,
U.S.A., Report on Gunpoicder, Washington, 1845, p. 187.
[Table
160
RESEARCHES ON EXPLOSIVES
Table 5. — Giving the pressures actually observed, in ions j)er square inch, with
F. G., R. L. G; and Pebble jwwders for various densities of the products
of explosion.
Mean
Nature of Powder.
Mean
Nature of Powder.
F. G.
R. L. G.
Pebble.
F. G.
R. L. G.
Pebble.
density of
density of
products
products
of explosion.
Pressure
Pressure
Pressure
of explosion.
Pressure
Pressure
Pressure
in tons
in tons
in tons
in tons
in tons
in tons
per
per
per
per
per
per
sq. inch.
sq. inch.
sq. inch.
sq. inch.
sq. inch.
sq. inch.
•0940
1-6
•5000
10-48
10-48
•1064
1-66
1-39
•5000
10-20
10-70
•1064
1 ^35
1^26
•5000
11-10
•1064
0-96
r28
*-5300
*ll-80
•1973
...
2-67
•5322
11-48
12-20
•2000
2 •70
•6000
14-14
14-36
13^78
■2114
2-93
•6000
13^50
•2129
3-70
•6000
...
14-80
•2129
3-58
*-6100
*15-6
•2129
3-00
*-6200
*16-8
•2963
6-40
•7000
18-2
19-54
18-60
•3000
5 •40
•7000
tl8-9
J17-00
•3171
4-90
*-7500
*2i-90
•3193
6-75
•8000
23-20
24-40
28^60
•3193
6-32
•8000
27-10
23-20
24^20
*^3800
*8-5
*7^7
-9000
•9000
27-20
35-6
33-40
31-60
•3860
7-68
•9000
31-40
•3947
8-1
^^•9000
*33-l
•4258
9-34
8-40
*30-7
•4258
9^10
*31-9
•4615
8-68
•9150
34-5
•4893
10-14
•9300
36-2
•4934
11-50
•9300
*34-0
§35-0
R. F. G. powder.
X Spanish spherical pellet.
^ Pellet.
We have laid down on Plate XII., p. 230, the whole of these experi-
ments. The pressures given by the pebble and the E. L. G. are nearly
identical ; we have therefore considered them so, and have drawn but
one curve to represent their mean residts. The curve representing
the pressures given by the F. G-., although nearly identical with the
pebble and E. L. G. at the lower densities, does not coincide at the
higher densities. A separate curve has therefore been drawn for
this powder. The lower tension is perhaps accounted for by the
difference between the quantity of permanent gas yielded by it and
l)y the other two powders.
The corrected values of the tension, in terms of the density of
the different powders, as indicated by the curves, Plate XII., p. 230,
are given in the following table : —
RESEARCHES ON EXPLOSIVES
161
Table 6. — Showing the pressure corresponding to a given density of the products
of explosion of F. G., B. L. G., and Pehhle poiaders, as deduced from actual
observation, in a close vessel. The pressures are given in tons per square
inch, atmospheres, and kilogrammes per square centimetre.
Corresponding pressures for Pebble
Corresponding pressures for
and E. L. G. powders.
F. G. powder.
Mean density of
products of
explosion.
In tons
In
In kilos.
In tons
In
In kilos.
per
sq. inch.
atmospheres.
per sq.
centimetre.
per
sq. inch.
atmospheres.
per sq.
centimetre.
•05
0^70
107
110-2
0-70
107
110-2
•10
1-47
224
231-5
1-47
224
231-5
•15
2-33
355
367-0
2-33
355
367-0
•20
3^26
496
513-4
3-26
497
513-4
•25
4^26
649
670-9
4-26
650
670-9
•30
5-33
812
839-4
5-33
812
839-4
•35
6-49
988
1028-1
6-49
988
1022-1
•40
7^75
1180
1220-5
7-74
1179
1219-0
•45
9^14
1392
1439-5
9-10
1387
1433-2
•50
10-69
1628
1683-6
10-59
1614
1667-8
•55
12-43
1893
1957-6
12-22
1863
1924-5
•60
14-39
2191
2266-3
14-02
2136
2208-0
•65
16-60
2528
2614-3
16-04
2445
2526-1
•70
19-09
2907
3006-5
18-31
2790
2883-6
•75
21-89
3333
3447-5
20-86
3179
3285-2
•80
25-03
3812
3942-0
23-71
3613
3734-1
•85
28-54
4346
4495-0
26-88
4096
4233-3
•90
32-46
4943
5112-1
30-39
4632
4786-1
•95
36-83
5608
5800-4
34-26
5190
5335-6
1-00
41-70
6350
6567-3
38-52
5870
6066-5
In considering the pressures indicated, the question naturally
arises as to how their value would be affected if the charges were
greatly increased ; or, to put the question in another form, it may be
inquired whether the tensions indicated by our experiments are
materially affected by the cooling influence of the vessel in which the
explosion is conducted.
We think there are very strong grounds for assuming that the
pressure is not materially affected by the above circumstances, except
in cases where the density of the products of explosion is low, and
the quantity of powder therefore very small as compared with the
space in which it is fired.
Thus it will be observed that the pressures obtained in Experi-
ment 2 and in Experiments 65, 66, and 68 compare very well (the
density being about the same), although the quantity of powder
fired in the first case is only half of that fired in the last three
experiments.
162 RESEARCHES ON EXPLOSIVES
Again, if there were any considerable decrement of pressure due
to loss of heat, we should expect to find that the tension indicated
would be higher when means are taken to ensure rapidity of com-
bustion. Such, however, is not the case; for if reference be made
to Experiments 70 and 71, in which the charges were fired by
means of mercuric fulminate, it will be observed that the tension
realised in these experiments was not materially higher than when
the powder was fired in the ordinary way.
We may cite also, in support of our view, some interesting obser-
vations made during some earlier experiments, in which charges of
10,500 grains (680-4 grms.) E. L. G-. and pellet powder were fired in
chambers entirely closed with the exception of a vent '2 inch (5-08
mm.) in diameter.
With the former powder the pressure realised under these
circumstances was 36-2 tons per square inch (5513 atmospheres),
with the latter 17*3 tons (2634 atmospheres). This large difference
was due to the slower combustion of the pellet-powder, upon the
ignition of which, therefore, a large part of the products of com-
bustion escaped by the vent before the whole of the powder was
fired. When, however, the same powders were fired in vessels
absolutely closed, the pressure indicated by the pellet-powder
was more than doubled (being 35 tons per square inch, or 5330
atmospheres), while the pressure indicated by the E. L. G. was
practically the same (being 34 tons per square inch, or 5178
atmospheres).
From the experiments made by the Committee on Explosives, we
are able to name approximately the absolute time that would be con-
sumed in burning a charge of E. L. Gr. and of pebble, assuming that
the powder be confined. With E. L. G-. the time would be approxi-
mately -00128 second, with pebble approximately -0052 second. Of
course these figures must vary greatly with different powders, as
they depend not only on the nature, size of grain, and density of the
powder, but also on the mode of ignition. They are interesting,
however, as indicating the minuteness of the times involved, and the
relatively much larger time required for the decomposition of the
pebble-powder. It follows, from the accordance of the pressures in
the experiments just referred to, when powders dijEfering so con-
siderably in rapidity of combustion are fired in close vessels, that
there is no very appreciable difference in pressure due to the
longer time taken by the pebble-powder to consume under these
conditions.
RESEARCHES ON EXPLOSIVES 163
But the strongest, and at the same time an altogether inde-
pendent, corroboration of our view is derived from the experiments
upon the pressures exerted in the bores of guns by the action of the
charge.
Not only do these pressures, as obtained by observation, agree
with most remarkable accuracy with the theoretical pressures
deduced from the experiments in a close vessel, but, when in large
guns the tensions due to different charges are compared (not with
reference to the position of the shot in the bore, but with reference
to the mean density of the products of explosion), a most striking
accordance is found to exist. We may therefore conclude that,
where powders such as those we have experimented with are
employed, there is but a trifling correction to be made in the observed
pressure when the powder entirely fills the space in which it is fired,
or, indeed, whenever it occupies a considerable percentage of that
space. But though the pressure may not be seriously affected when
the generated gases are of a high density, it is more than probable
that some very appreciable correction should be made in the results
we have observed when experimenting with gases of low density. In
this latter case the cooling influence of the vessel would be greatly
increased, not only from the higher ratio which the cooling surface
bears to the charge, but also from the slowness of combustion due
to the comparatively feeble pressure ; and we think the effect of slow
combustion is clearly traceable in the low tensions observed with
pebble-powder (See curve, Plate XII., p. 230) at densities of "1, "2,
and "3, as compared with those given at corresponding densities by
F. G. powder, the combustion of which would be much more rapid.
But we shall return to this point when we compare our results with
those demanded by theory.
Upon the same plate (Plate IX., p. 230), on which we have given
curves representing the experiments of Eumford and Eodman, there
is also laid down a curve representing our own experiments. The
very high results obtained by Eumford are probably in great measure
attributable to his method of experiment. The charges being placed
at one end of his little vessel, while the weight to be lifted, so to
speak, closed the muzzle, the products of combustion acquired a high
vis viva before striking the weight, and thus indicated a much higher
pressure than that due to the tension of the gas, just as in Eobins's
well-known experiment a musket-barrel may be easily bulged or
burst by a bullet placed at some distance from the charge. That
Eumford's and even Piobert's corrected estimate of the tension of
164 RESEARCHES ON EXPLOSIVES
fired gunpowder was very excessive, is of course indisputably proved
by our experiments, as the vessels in which they were made were
quite incapable of resisting pressures at all approaching those
assigned by these eminent authorities.
Eodman's results are also too high, from a defect in the applica-
tion of his system of measurement, which has elsewhere* been
pointed out ; and his experiments on the ratio of tension to density
were not carried sufficiently far to admit of comparison in the more
important portion of the curve.
(I) DETEEMINATION OF HEAT GENEKATED BY THE COMBUSTION OF
GUNPOWDER.
The amount (that is the number of units) of heat liberated by
the combustion of gunpowder is determined from Experiments Nos,
46, 47, 48, 49, and 63. (See pp. 221, etc.).
The powder used was the E. L. G. and F. Gr. ; but as it was found
that there was no material difference in the heat liberated, we have
drawn no special distinction between the experiments made with the
two brands.
In each of the Experiments ISTos. 46, 48, and 63, 3800 grains
(246*286 grms.) were exploded ; and when the necessary reductions
were made to convert the alterations in temperature which were
observed into their equivalents in water, it was found that in Experi-
ment 48 the explosion of 246*286 grms. F. Gr. was sufficient to raise
173,077-4 grms. of water through V Cent. In Experiment 48, the
explosion of the same quantity of E. L. G. was equivalent to raising
172,569 grms. of water through 1° Cent., and in Experiment 63 to rais-
ing 171,500 grms. through 1" Cent. ; or, expressing these results in a
different form, it appears that the combustion of a grm. of powder
gave rise to quantities of heat represented by raising a grm. of
water through 702°-80 Cent., 700°-69 Cent., and 696°-50 Cent,
respectively.
In Experiments 47 and 49, the charge used was 393*978 grms. ;
and it was found that in Experiment 47 the heat generated by the
explosion of the F. G. was sufficient to raise 277,994*1 grms. of water
through 1" Cent. ; and in Experiment 49 the explosion of the same
quantity of E. L. G. generated heat represented by the raising
of 278,185*5 grms. through 1° Cent., — or, in the two experiments,
* Noble, loc. cit. p. 25 ; Revue Scientifique, No. 48, p. 1138.
RESEARCHES ON EXPLOSIVES 165
1 grm. of powder gave rise respectively to 705'61 and 706'09
grm. -units.
The mean of the whole of these experiments gives 702-34 grm.-
units of heat generated by the explosion of a grm. of powder, and we
shall probably have a very close approximation to the truth in
assuming it at 705 grm.-units.
From this datum the temperature of explosion may be deduced, if
we know the mean specific heat of the products of combustion. We
have only to divide 705 by the specific heat, and the result is the
required temperature.
The specific heat of all the gaseous products of combustion are
known ; and they have also been determined for the principal solid
products at low temperatures, when they are (of course) in the solid
form.
Bunsen and Schischkoff, from their experiments, deduced the
temperature of explosion on the assumption that the specific heats of
the solid products remain invariable over the great range of tempera-
ture through which they pass.
With every deference to those distinguished chemists we think
their assumption is quite untenable. Without, we believe, any
known exception, the specific heat is largely increased in passing
from the solid to the liquid state. In the case of water, the specific
heat is doubled ; the specific heats of bromine, phosphorus, sulphur,
and lead are increased from 25 to 40 per cent, and those of the
nitrates of potassium and sodium nearly 50 per cent., while it is more
than probable that, even with liquids, the specific heat increases very
considerably with the temperature.
We shall, however, deduce from our experiments the temperature
of explosion on Bunsen and Schischkoff' s hypothesis, both with the
view of enabling our results to be compared with theirs, and for the
purpose of fixing a high limit to which it is certain the temperature
of explosion cannot reach. We shall afterwards endeavour to
estimate more accurately the true temperature.
The data necessary for computing the specific heat of a grm. of
exploded powder are given in the subjoined table.
[Table 7.
166
RESEARCHES ON EXPLOSIVES
Table 7. — Shoimng the specific heats and
proportions o
/ the product
s generated
hy the combustion
of gunpowder.
1.
Products of combustion.
2.
Proportion
in a
gramme.
3.
Specific
heat.
4.
Authority.
5.
Products
of columns
2 and 3.
Solid -5684.
Potassium carbonate
•3382
•206
Kopp
•06967
„ hyposulphite
0355
•197
Pape
•00700
„ sulphate .
0882
•196
Kopp
•01729
sulphide .
0630
•108
Bunsen
•00680
„ sulphocyanatc
0009
nitrate .
0006
•239
Kopp
•00014
Ammonia carbonate .
0006
Sulphur
0414
•171
Bunsen
•00708
Carbon
•0000
•242
Regnault
At constant
Gaseous -4316.
volume.
Sulphuretted hydrogen
•0113
•184
Clausius
•00208
Oxygen . '.
■0000
•155
Carbonic oxide .
•0447
•174
•00778
Carbonic anhydride .
•2628
•172
•04520
Marsh-gas .
•0005
•468
•00024
Hydrogen .
•0010
2-411
•00241
Nitrogen ....
•1113
•173
•01925
•18494
Adding up the numbers in column 5, we obtain "18494 as
the mean specific heat of the products of explosion of a grm. of
powder at ordinary temperatures; and since, as we have said, the
temperature of explosion is obtained by dividing the grm. -units
of heat by the specific heat, we have the temperature of explosion
= :^^ = 3812° Cent.; and we may accept this figure as indicating a
temperature which is certainly not attained by the explosion of gun-
powder. We defer until further on the consideration of the actual
temperature.
(m) DETERMINATION OF VOLUME OF SOLID PKODUCTS AT OKDINARY
TEMPERATURES.
The space occupied by the solid products of combustion at
temperatures but little removed from 0°, is deduced from experiments
Nos. 46, 48, 49, 57, 58, 60, 61, and 62. From these experiments it
appears that
246"29 grms. R. L. G. gave rise to 76*46 q.c. solid residue.
246-29 „ F. G. „ 67*26 „
393^98 „ R. L. G. „ 123^12 „
RESEARCHES
ON EXPLOSIVES
386-21 grms.
386-21 „
386-21 „
386-21 ,,
386-21 „
F. G. gave
R. L. G.
P.
R. L. G.
E.G.
rise to 115*34 c.c, solid residue.
110-81 „
111-78 „
105-30 „
108-54 „
167
Or, stating the results per grm. of powder, it appears that in the
several experiments the solid products arising from the combustion
of a grm. of powder occupied respectively -3105, -2731, '3125, -2987,
-2869, -2894, -2726, and "2810 c.c.
The mean of these figures is -2906 ; and we may thence conclude
that at 0° Cent, the soHd residue of 1 grm. of burned powder occupies a
volume closely approximating to '29 c.c. ; therefore, since the solid
products represent 57 per cent, of the original weight of the powder,
it follows that at 0° Cent, the specific gravity of the residue is about 1-4.
(n) PEESSUEE IN CLOSE VESSELS, DEDUCED FEOM THEOEETICAL
CONSIDEEATIONS.
From the investigations we have described, it appears that in a
close vessel, at the moment of explosion, or at all events shortly
afterwards, the results of the decomposition of a given charge (say
1 grm.) of powder such as we have experimented with, are as
follows : —
1. About 43 per cent, by weight of permanent gases, occupying
at 0° Cent, and under a pressure of 760 mm., a volume of about
280 c.c.
2. About 57 per cent, by weight of liquid product, occupying,
when in the solid form and at 0° Cent., a volume of about -3 c.c.
Now, if we assume that the conditions known to exist shortly
after explosion obtain also at the moment of explosion, we are able,
with the aid of our experiments, to compute the pressure, tempera-
ture of explosion, and volume occupied by the permanent gases. We
propose to make these calculations, and then, by comparison with the
results obtained under the varied conditions adopted in our experi-
ments, to form an estimate of the correctness of our assumption.
And, first, to establish a relation between the tension and the mean
density of the products of explosion at the moment of ignition, —
Let ABCD (Plate XL, Fig. 4, p. 230) represent the interior of the
vessel, of volume v, in which the experiments were made. Let CDEF
represent the volume of a given charge of powder placed in the
vessel. Let <5 be the ratio which the volume CDEF bears to ABCD,
168 RESEARCHES ON EXPLOSIVES
and let CDHG {vaS suppose) be the volume occupied by the liquid
products at the moment and temperature of explosion.
It is obviously, for our present purpose, a matter of indifference
whether we suppose the liquid products collected, as in the figure, at
the bottom of the vessel or mixed with the permanent gases in a
finely divided state.
Our conditions on explosion, then, are: — We have the space
CJ)B.Gc = vaS occupied by the fluid residue, and the space ABHG =
v{l — aS) by the permanent gases.
Hence, since the tension of the permanent gases will vary
directly as their density, we have, if p represent the pressure and D
the density,
P = RD3 (1)
where E is a constant.
Now suppose the charge exploded in the chamber to be increased.
In this case, not only is the density of the permanent gases increased
on account of a larger quantity being generated, but the density is
still further added to, from the gases being confined in a smaller
space ; the liquid residue CDHG- being increased in a like propor-
tion with the charge (D, in fact, varying as ^_ ), we have
p = ^-r\ (2)
1 — ao
or if Pq, Sq, be corresponding known values of p and S,
^ 8^ 1 - aS •
Poi^ -^ ^ .... (3)
In taking the tension of the permanent gases to vary directly as
their density, we have of course assumed that the temperature,
whatever be the value of S, is the same.
In our experiments, the charges exploded have varied in quantity
from that necessary to fill entirely the chamber to a small fraction of
that quantity; but whatever the charge, it is obvious that if the
vessel be considered impervious to heat (and we have already pointed
out that only with the lower charges is there a material error due to
this hypothesis), the temperature at the moment of explosion would
be the same ; for, as in the case of Joule's celebrated experiment,
any heat converted into work by the expansion of the gases would
again be restored to the form of heat by the impact of the particles
against the sides of the vessel.
Eeturning to (3), the value of the constant a in this equation has
RESEARCHES ON EXPLOSIVES
169
yet to be found. If, from Table 6, we take out a second pair of
corresponding values ^i, Sj^, a is determined, and will be found = -65,
very nearly. Taking a = "65, and from Table 6, or the curve, Plate
XII. (p. 230), taking Sq=-Q, Po = 14:-4: tons, Equation (3) becomes
14-63
l-a8
(4)
Substituting in this equation successively values of S'05, 1, "15, etc.,
we obtain computed values of p, which we compare with those
derived directly from observation in Table 8.
Table S.— Showing the comparison, in atmospheres and tons per square inch,
between the pressures actually observed in a close vessel, and those calculated
(l-«5o) 5
5n ' 1 - a5
from the formula p = Po-
1.
■2.
3.
4.
5.
6.
7.
Value of p
Value of n
Value of p
Value of p
Density
Value of })
deduced from
direct
deduced
deduced
Value of p
deduced
deduced
of
products
from
Equation (3)
from
Equation (8)
deduced from
direct
from
Equation (3)
from
Equation (8)
of com-
bustion.
observation.
when
a =-65.
when
a =-60.
observation.
when
a =-65.
when
a =-6.
Tons
per sq. inch.
Tons
per sq. inch.
Tons
persq. inch.
Atmospheres.
Atmospheres.
Atmospheres.
•05
0-70
•758
•855
107
115
130
10
1-47
1-565
1-765
224
238
269
15
2-33
2-432
2-734
355
370
416
20
3-26
3-363
3-771
496
512
574
25
4-26
4-367
4-879
649
665
743
30
5-33
5-452
6-071
812
830
924
35
6-49
6-628
7-350
988
1009
1119
40
7-75
7-908
8-732
1180
1204
1330
45
9-14
9-305
10-228
1392
1417
1557
50
10-69
10-837
11-851
1628
1650
1805
55
12-43
12 -.524
13-620
1893
1907
2074
60
14-39
14-390
15-554
2191
2191
2369
65
16-60
16-466
17-679
2528
2507
2692
70
19-09
18-791
20-024
2907
2861
3049
75
21-89
21-410
22-625
3333
3260
3445
80
25-03
24-383
25-525
3812
3713
3887
85
28-54
27^789
28-780
4346
4232
4383
90
32-46
31-728
32-460
4943
4831
4943
95
36-83
36-336
36-654
5608
5538
5582
1-00
41-70
41-698
41-477
6.350
6350
6316
Now if the figures given in columns 2 and 5, being those derived
from the observations themselves corrected by differencing, be com-
pared with the values given in columns 3 and 6, computed on the
value a = "65 (that is, on the assumption that at the temperature of
170 RESEARCHES ON EXPLOSIVES
explosion the liquid residue of 1 grm. of powder occupies "65 c.c), it
will be found that the two columns are practically identical, thus
affording a confirmation of the strongest nature of the correctness of
our assumption. The closeness of agreement will be best seen by
examining the graphical representations in Plate XIII. (p. 230). We
have already, however, had more than once occasion to remark that
there is reason to suppose that the observed pressures are slightly in
defect, at all events at low densities. Other considerations have led
us to the conclusion that a value of a not far removed from "6 would
more nearly represent the truth, were all disturbing influences
removed. We have therefore added to the above table the pressures
computed on this hypothesis ; and Plate XIII. (p. 230), shows at a
glance the comparison between the three curves.
(o) DETERMINATION OF THE TEMPERATURE OF EXPLOSION OF
GUNPOWDER.
We are now in a position to compute the temperature of
explosion.
Since p, v, and t are, in the case of permanent gases, connected by
the equation of elasticity and dilatability,
jw = Rt (5)
(where E is a constant and t is reckoned from absolute zero), t will be
known if p, v, and E be known.
Now if we assume a = '6, it follows that in the combustion of
1 grm. of powder (gravimetric density = 1) the gaseous products will,
if the powder entirely fill the chamber in which it is placed, occupy
a space of '4 c.c. But we know that, at 0° Cent, and under a baro-
metric pressure of 760 mm., the gaseous products of 1 grm. occupy a
space of about 280 c.c. Hence at 0° Cent., if the gaseous products
are compressed into a space of '4 c.c, we have a pressure of 700
atmospheres ; and since absolute zero = — 274° Cent., we have, in the
equation ^Q-y^ = Ei^o, the values p^ = 700, Vq = % ^^ = 274;
.-. R=™°^*= 1-0218
Hence (5) becomes
pv = 1-0218/ .... (6)
Now under the above conditions, but at the temperature of explosion,
we have, from Table 8,^ = 6400 atmospheres, and, as bef ore, -y = '4.
RESEARCHES ON EXPLOSIVES 171
Therefore
'-'-^-'''^ . . ■ w
and this is the temperature of explosion reckoned from absolute
zero. Subtracting 274° from this temperature to reduce the scale to
Centigrade, we have temperature of explosion = 2231° Cent.
If we assume a = '65, the temperature of explosion deduced in
the same way would be 1950° Cent. ; but this temperature, as we
shall shortly show, would be somewhat too low.
We have now three points to consider : —
1. Is this temperature a probable one? and can any direct
experimental facts be adduced to corroborate this theoretical
deduction ?
2. What is the mean specific heat of the solid or Hquid products
which the above temperature implies ? and
3. Can any corroboration be given to the high rate of expansion
of the soHd residue implied by assuming the value of a as = "6 ?
With regard to the direct estimation of the temperature of
explosion, we have made several experiments with the view of
obtaining this result, by ascertaining the effects of the heat developed
on platinum. For example, in Experiment 78 we introduced into
the charge of R F. Gr. a coil of very fine platinum wire and also a
piece of thin sheet platinum. After the explosion the sheet platinum
was found much bent, but unmelted ; but on examination with a
microscope there were evident signs of a commencement of fusion on
the surface, and a portion of the fine platinum wire was found welded
on to the sheet. The coil of wire was not to be found, but portions
of it were observed welded to the sides of the cylinder.
Now we know that platinum is readily volatilised when exposed
to the hydrogen-blowpipe at a temperature of about 3200° Cent., and
therefore, if the temperature of explosion had approached this point,
we should have expected the very fine wire to be volatiHsed; re-
membering the low specific heat of platinum, we should furthermore
have been warranted in expecting more decided signs of fusion in the
sheet metal.
Again, in Experiments 84, 85, and 68, pieces of platinum wire,
•03 inch (0-75 mm.) in diameter and 4 inches (100 mm.) long, were
placed in the cylinder with considerable charges of E. L. G. and
F. Gr. In none of these experiments did the platinum melt, although,
as in the case of the sheet platinum, there were signs of fusion on
172 RESEARCHES ON EXPLOSIVES
the surfaces of the wires. In Experiment 79, however, in which
platinum wire was placed with a corresponding charge of the Spanish
powder, the wire was fused, with the exception of a small portion.
With this powder, indeed, which is of a very different composition
from the Enghsh powders, and decidedly more rapidly explosive in
its nature, it is quite possible that a somewhat higher heat may have
been attained. But, as in one case the platinum wire was nearly
fused, and in others it only showed signs of fusion, the conclusion we
draw from the whole of these experiments on the fusion of the
platinum is, that the temperature of explosion is higher than the
melting-point of that metal, but not greatly so. Now, according to
Deville, the melting-point of platinum is nearly 2000° Cent. ; and
hence we have a strong corroboration of the approximate accuracy of
the theoretical temperature of explosion at which we have arrived,
viz., 2231° Cent.
(p) MEAN SPECIFIC HEAT OF LIQUID PRODUCTS.
We have already given the specific heat of the liquid products
when in the solid form. If we assume the temperature above
specified to be correct, a mean specific heat of the liquid product of
•4090, or a total mean specific heat of the entire products of '3094,
would result, being an increment of about 67 per cent. ; and this,
judging from the analogy of the case we have cited, does not appear
an improbable conclusion.
{q) PROBABLE EXPANSION OF NON-GASEOUS PRODUCTS BETWEEN ZERO
AND TEMPERATURE OF EXPLOSION.
So far as we are aware, there were, prior to our experiments, no
data existing as to the behaviour of the non-gaseous products of com-
bustion at the high temperature involved, except perhaps the experi-
ment made by Bunsen and Schischkoff, who exposed on platinum foil
the solid residue in an oxyhydrogen jet, and concluded, from there
being no ebullition, that at the temperature of 3300" Cent, the tension of
the resulting vapour did not reach one atmosphere. Taking the circum-
stances into account, we may indeed doubt if the residue itself actually
reached the temperature we have named ; but the experiment would
at all events prove that, at the temperature which we find to be that
developed by explosion, the solid or liquid products are not in the
state of vapour, or at least that the small portion volatilised had but
an insignificant tension. To test, however, the behaviour of the
RESEARCHES ON EXPLOSIVES 173
residue for ourselves, we placed in one of Siemens gas-furnaces, the
temperature of which was estimated at about 1700° Cent., several
crucibles containing powder-residue. The behaviour of the residue
was in all cases the same ; at first there was a little spirting (prob-
ably due to escape of water), which, however, soon diminished, and
in time the contents of the crucibles became perfectly quiet, but up
to the end of the experiment only a very slight volatilisation could
be observed. In the case of three of the crucibles, two of which
contained powder-residue, the other a mixture of potassium carbonate
and liver of sulphur, when removed from the furnance after being
exposed to the full heat for about a quarter of an hour, the volumes
of the contents in the highly heated state were observed without
difficulty. The contraction in cooling was evidently very great,
especially at first. The contents set at a temperature of between
700° and 800° Cent., and when cool the expansion was measured by
calibration with mercury. The first crucible gave an expansion of
77"8 per cent, between 0° Cent, and 1700° Cent. ; the second (potass
carb. and liver of sulphur) an expansion of 93-3 per cent. The third
(powder-residue) gave a considerably higher rate of expansion, above
100 per cent. ; but we have not included the result, as, owing to the
presence of a piece of platinum put in to test the temperature of the
furnace, we were unable to make a very accurate measurement.
Of course the expansions, under the conditions we have just
named, cannot be strictly compared with those which would have taken
place in a close vessel under the high tension we know to exist ; but
they tend to confirm the results arrived at by a perfectly inde-
pendent method. The experiments also show that, at a temperature
approaching that developed by explosion, and under atmospheric
pressure, the liquid products are still in that condition; and our
experiments so far confirm those of Bunsen and Schischkoff to which
we have alluded.
(r) OBSERVED PEESSURES IN THE BORES OF GUNS.
The data which we shall use for the discussion of the phenomena
attending the combustion of gunpowder in ordnance are nearly
entirely derived from the experiments carried on by the Committee
on Explosives, under the presidency of Colonel Younghusband, F.E.S.
Two methods, of an entirely distinct nature, were employed by
the Committee for the elucidation of the questions they had to
consider.
174 RESEARCHES ON EXPLOSIVES
One method consisted in determining the tension of the gas at
various points in the bore, by direct measurement. The other mode
consisted in measuring the time at whicli the projectile passed certain
fixed points in the bore, thence deducing the velocities from the seat
of the shot to the muzzle, and finally obtaining, by calculation, the
gaseous pressure necessary to generate the observed velocities.
The apparatus used for determining the tension by direct
measurement was the crusher-gauge, which we have already
described; that for ascertaining the velocity was a chronoscope,
specially designed for measuring very minute intervals of time. As
the construction of this instrument has been fully explained else-
where, we shall only here give a very general description of it.
.Its most recent form is shown in plan and elevation in Plate
XIV., Figs. 1 and 2 (p. 230). The mechanical part consists of a series
of thin discs. A, A, etc., 36 inches in circumference, keyed on to a shaft,
S, and made to rotate at a very high and uniform velocity through
the train of wheels F, by means of a very heavy descending weight
at B, arranged, to avoid an inconvenient length of chain, upon a
plan originally proposed by Huyghens. This weight is continually
wound up by means of the fly-wheel and handle at T. The stop-
clock D, which can be connected or disconnected with the shaft E at
pleasure, gives the precise speed of the circumference of the discs,
which is usually arranged at about 1250 inches a second.
The recording arrangement is as follows : — Each disc is furnished
with an induction-coil, G, the primary wire from which is conveyed
to any point, K, in the gun where we may wish to record the instant
at which the shot passes. There is at each such point a special con-
trivance, by which the shot in passing severs the primary wire,
thereby causing a discharge from the secondary, which is connected
with the discharger, Y. The spark records itself on the disc by
means of paper specially prepared to receive it. The instrument is
capable of recording the millionth part of a second, and, when in good
working order, the probable error of a single observation should not
exceed 4 or 5 one-millionths of a second.
The guns were arranged for the experiments as shown in Fig. 3
in the same plate. Holes were drilled in the powder-chamber in
the positions marked A, B, C, and in the bore in the positions
marked 1 to 18.
In A, B, and C, crusher-gauges were always placed ; the holes,
numbered 1 to 18, were fitted with crusher-gauges or the chrono-
scope-plugs at option.
RESEARCHES ON EXPLOSIVES 175
It would be beside our object in this paper to enter into a
discussion of the special experiments undertaken by the Committee
on Explosives. The chief object of their investigations was to
determine the nature of powder most suitable for use with heavy
guns — that is to say, the powder which will allow of the highest effect
being realised without unduly straining the structure within which
the explosion is confined. A number of experiments were therefore
made with powders of abnormal types, interesting and instructive
only to artillerists; and these experiments will doubtless be fully
reported on at a later date, by the proper authorities.
In our present paper we shall confine our attention chiefly, if not
entirely, to the results obtained with the well-defined and weU-known
powders which have been admitted into the service for use with
rifled guns, and which are known under the names of " Eifled Large
Grain" and "Pebble." These powders are, moreover, the same as
were used by us in our experiments in closed vessels, and therefore
allow of a strict comparison with the tensions so obtained. But
before giving the details, we cannot pass without notice certain
diS"erences in the results obtained by means of the two modes of
experimenting to which we have alluded.
With pebble and other powders, where a slow and tolerably
regular combustion takes place, the maximum tension of the gas,
obtained both by direct measurement and by the chronoscope, agrees
remarkably closely. There is generally a very slight difference
indeed between the indicated pressures; but the case is greatly
different where the powder is of a highly explosive or quickly burn-
ing description. In such a case, not only are the pressures indicated
by the crusher-gauge generally much above those indicated by the
chronoscope, but they differ widely in various parts of the powder-
chamber, in the same experiment, and even in different parts of the
same section of the bore. They are also locally affected by the form
of the powder-chamber, and frequently indicate pressures consider-
ably above the normal tensions that would be attained were the
powder confined in a close vessel.
It is not difficult to explain these anomalies. When the powder
is ignited comparatively slowly and tolerably uniformly, the pressure
in the powder-chamber is also uniform, and approximates to that due
to the density of the products of combustion.
The crusher-gauges, then, give similar results throughout the
powder-chamber, and they accord closely with the results deduced
from the chronoscope observations. But when a rapidly lighting or
176 RESEARCHES ON EXPLOSIVES
" brisante " powder is used, the products of combustion of the portion
first ignited are projected with a very high velocity through the
interstices of the charge, or between the charge and the bore ; and
on meeting with any resistance their vis viva is reconverted into
pressure, producing the anomalous local pressures to which we have
drawn attention.
We have pretty clear proof that, when this intense local action is
set up, the gases are in a state of violent disturbance, and that waves
of pressure pass backwards and forwards from one end of the charge
to the other, the action occasionally lasting the whole time that the
shot is in the bore. In fact, with the rapidly burning, and in a less
degree even with the slower burning powders, motion is communi-
cated to the projectile not by a steady, gradually decreasing pressure
like the expansive action of steam in a cylinder, but by a series of
impulses more or less violent.
The time during which these intense local pressures act is of
course very minute ; but still the existence of the pressures is
registered by the crusher-gauges. The chronoscopic records, on the
other hand, which are, so to speak, an integration of the infinitesimal
impulses communicated to the shot, afford little or no indication of
the intensity of the local pressures, but give reliable information as
to the mean gaseous pressure on the base of the shot.
The two modes of observation are, as we have elsewhare pointed
out, complementary one to the other. The chronoscope gives no clue
to the existence of the local pressures which the crusher-gauge shows
to exist ; while, on the other hand, where wave or oscillatory action
exists, the results of the crusher-gauge cannot be at all relied on as
indicating the mean pressure in the powder-chamber.
An interesting illustration of this distinction was afforded by
two consecutive rounds fired from a 10-inch gun, in one of which
wave-action was set up, in the other not. In both cases the pro-
jectile quitted the gun with the same velocity, and the mean pres-
sure throughout the bore should of course have been the same. The
chronoscopic records were, as they ought to be, nearly identical for
the two rounds; but the pressures indicated by the crusher-gauge,
were in the one round, at the points A, B, C, 1, 4 (Fig. 3, Plate XIV.,
p. 230), respectively 63-4, 41-6, ST'0, 41-9, and 25-8 tons on the square
inch ; in the other, at the same points, respectively 28"0, 29"8, 30*0
29'8, and 19"8 tons on the square inch.
Where no wave-action exists, the chronoscopic pressures are
generally somewhat higher than those of the crusher-gauge. The
RESEARCHES ON EXPLOSIVES 177
difference is not generally greater than about 5 to 7 per cent.,
although, in the case of some exceptionally heavy shot, this variation
was considerably exceeded. Among the causes tending to produce
this difference may be cited : — 1. Friction in the parts of the crusher-
gauge. 2. Slight diminution of pressure due to windage.* 3. Vis
viva of particles of the charge and products of combustion, a portion
of which would be communicated to the shot, but would not take
effect on the crusher-gauge. On the whole, however, the accordance
of results derived from methods so essentially different was quite as
close as could reasonably be expected, and entirely satisfactory.
We now pass to the consideration of the tensions actually found
to exist in the bores of guns. Two series of experiments were made
by the Committee on Explosives with the 10-inch 18-ton gun. The
one series was with charges of 70 lbs. (31'75 kilos.) of pebble-powder.
The weights of the shot were made to vary, the first rounds being
fired with projectiles of 300 lbs. (136*05 kilos.), and the weights being
successively increased to 350 lbs., 400 lbs., 450 lbs., 500 lbs., 600 lbs.,
800 lbs., 1000 lbs., and concluding with projectiles of the weight of
1200 lbs. (544-20 kilos.).
In the other series, charges of 60 lbs. (27-21 kilos.) E. L. G. were
used. The projectiles were of increasing weights, as above ; but the
experiments were not carried so far, the heaviest projectile in this
series being of 600 lbs. (272 kilos.) weight.
As we shall have occasion more than once to refer to these ex-
periments, and as the powder used was carefully selected to represent
as nearly as possible the normal service-powder of each description,
it appears to us convenient, in order to illustrate the methods
followed in determining the powder-pressures, to take an example
from each series.
This plan will further enable us to compare the difference of
behaviour of pebble and E. L. G. powder in the bore of a gun.
Commencing, then, with the charge of 70 lbs. (31*75 kilos.) pebble-
powder and the projectile of 300 lbs. (136-05 kilos.), the results given
by the chronoscope, to which we shall turn our attention in the first
instance, are given in Table 9.
In this table, column 1 gives the distances of the various plugs
* In the experiments with the 38-ton gun, an opportunity occurred of determining
the differences in pressure due to the escape of the gases by the windage, and it
was found that a reduction of windage of '07 inch (175 mm.), i.e., the difference
between '01 inch and "08 inch windage, reduced the maximum pressure indicated
by the crusher-gauge by about 1 ton per square inch. Of course the mean pressure
on the base of the projectile was not reduced in anything Uke the same proportion.
M
178
RESEARCHES ON EXPLOSIVES
from the seat of the shot in feet (see Fig. 3, Plate XIV., p. 230) (the
distance from the seat of the shot to the bottom of the bore being
2 feet = '6 10 metre). Column 2 gives the same distances in metres.
Column 3 gives the observed time of passing each plug. Column 4
gives the corrected time from the commencement of motion, the time
from the commencement of motion to first plug being interpolated.
Column 5 gives the differences of time — that is, the time taken by
the projectile to traverse the distance between the plugs. Column 6
gives the mean velocity of the projectile over the space between the
plugs, in feet ; and column 7 gives the same velocities in metres.
Table 9. — Giving data obtained with chronoscope for calculating velocity and
pressure in the bore of a 10-inch \8-ton gun. Charge, 70 lbs. (31-75 kilos.)
pebble-powder. Weight of projectile^ 300 lbs. (136^05 kilos.). Muzzle-
velocity, 1^27 feet (465^4 metres).
1. 2.
Distance from seat
of shot.
3.
Time observed
at plugs.
4.
Total time
from
seat of shot.
Time taken
by shot
to traverse
distance
between
plugs.
0. 7.
Mean velocity over
spaces between plugs.
Feet.
0^00
0^06
0^26
0^46
0-66
0-86
1-06
1-46
1-86
2-26
2-66
3-46
4-26
5-06
6^66
8^26
Metres.
0^000
0-018
0-079
0-140
0-201
0-262
0-323
0-445
0-567
0-689
0-811
1-055
1-298
1^542
2^030
2-518
Seconds.
•000000
•001096
•001611
•001967
•002272
•002548
•003036
•003469
•003869
•004244
•004947
-005605
-006234
-007426
-008554
f
Seconds.
000000
002683
003779
004294
004650
004955
005231
005719
006152
006552
006927
007630
008288
008917
010109
011237
Seconds.
•002683
•001096
•000515
•000356
•000305
•000276
•000488
•000433
•000400
•000375
-000703
-000658
-000629
-001192
-001128
Feet
per second.
22
183
388
562
656
725
820
924
1000
1065
1138
1215
1273
1342
1418
Metres
per second.
6-7
55-8
118-3
171-3
199^9
221^0
249-9
281-6
304-8
324-6
346-9
370-3
388-0
409-0
432-2
From these data are deduced, by correction and interpolation,
the times given in Table 10, pp. 180 and 181. From the differences
of the times are calculated the velocities, and from the velocities
the pressures necessary to produce them are obtained.
We have not space within the limits of our paper to enter upon a
discussion of the methods of calculation and correction necessary to
arrive at the results tabulated ; they are attended with very great
labour, and a full consideration of the question would necessitate a
RESEARCHES ON EXPLOSIVES 179
separate paper. As we shall hereafter show, it is not difficult, if we
were to suppose the powder entirely converted into gas on the instant
of explosion, to lay down the law according to which the pressure
would vary in the bore of the gun ; but the case under consideration
is a much more complicated one. The charge of powder is not
instantly exploded, but is generally ignited at a single point ; the
pressure (commencing at zero) goes on increasing at an extremely
rapid rate until the maximum increment is reached. It still goes on
increasing, but at a rate becoming gradually slower, until the
maximum tension is reached, when the increase of density of the
gas, aided by the combustion of the powder, is just counterbalanced
by the decrease of density due to the motion of the projectile. After
the maximum of tension is reached, the pressure decreases, at first
rapidly, subsequently slower and slower.
If these variations in pressure be represented by a curve, it would
commence at the origin convex to the axis of x, would then become
concave, then again convex, and would finally be asymptotic to the
axis of X.
In the same way, the curve representing the velocity would
commence by being convex to the axis of abscissae ; it would then
become concave, and, were the bore long enough, would be finally
asymptotic to a line parallel to the axis of x.
The results of Table 10 are graphically represented in black lines
in Plate XV. (p. 230), the space described by the shot being taken as
the equicrescent or independent variable, and the two curves giving
respectively the velocity and pressure at any point of the bore.
From the table (or curves) it will be seen that the maximum
pressure attained by the powder is 18 tons per square inch (2745
atmospheres), and that this pressure is reached when the projectile
has moved -5 feet (-153 metre) and at '00437 second from the com-
mencement of motion.
The results given in the table have, as we have said, been arrived
at liy special methods of correction and interpolation ; and their
general correctness can be tested by examining whether a material
alteration of pressure or velocity at any point can be made without
seriously disturbing the times actually observed. It will be found
that they cannot. But another question here presents itself for
consideration. We have, in the curves on Plate XV. (p. 230), taken s
as the independent variable ; but if t were taken as the independent
variable, and the relation between s and t were capable of being
expressed by the explicit function s =/ {€), the velocity corresponding
180
RESEARCHES ON EXPLOSIVES
•s^
"^
£
Atmospheres.
262
585
928
1259
1503
1705
1857
1998
2119
2220
2311
2381
2442
2498
2543
2583
2623
2652
2680
2702
2720
2733
2734
2741
2745
2741
2726
2711
2575
2381
Ph
Tons
per square inch.
1-723
3-843
6-096
8-270
9-873
11-198
12-192
13-120
13-915
14-578
15-174
15-638
16-036
16-407
16-698
16-963
17-228
17-414
17-599
17-745
17-864
17-944
17-957
17-997
18-023
17-997
17-904
17-800
16-910
15-637
>
go cooj05t-^t^<»rH(rqo>n(»<»i^oow07HOop.po-?OioOT»T'T'°0':-'?
Feet
per seco
11
35-
65-
94-
123
150
176
200
223
246
267
288
308
327
345
363
.380
397
413
429
444
459
473
488
501
515
528
540
576
629
a
>
1
5
(MO00OCD(rC0-a<-#'*tD
S §§g§§§§§8§8§§§§§§88§gS§§§§§S§§
1
£ OCDtMOOTHOl-^OOOJlOi-lt^COOivOi-HCOTjIOCDC^OO^OOC^lOOiOi— li-H
■g OO.-H.-l(MMC0TM^vnCDt0Wt-C00i0JO.-l.-l(M(MC0':H^l0l0C0WO
iS OOpOOpOpppppppppprHi^r-l>^i^i^r^r^rHi^i^>;-lC-1
-t5 O CM tH to 00 O (M ^ CD OO O (M Ttt «D 00 O (M ^ CD 00 O pp
c=^ o o o o o o o o o o o o o o o o o oo o o o o o o o o o o o
RESEARCHES ON EXPLOSIVES 181
. t^OiOICOoOMt^OS«OeD'X>Jr^i-(OOOOOTt O 00
(M ^ 5D
CO CO 30 CO t^ t^ t^ l^ t^ O ^ O O O -* Tfi ^ CO Otl CO CO CO
000-*OiacOOCOCOi— lOi— lO.— li-"5C0>-HC0l0r-lC0^O'X>-*^'-c0005-*t^ONOSOJvr5t^(MiOu^t^OiiOCOC>5CO.-ICOO
,-it^cqi>cot^5^co.:HiciKc^«prH;^>jr5io-^(MOi^'*iO!mi-H
-pcpl>.OC0i;-10iCT)iq,pU.;-i^-jhQp
a5Tj<'*oo5tNa5>nocoTj-coooc.i-ii^icOSi-ICOOt-Oii-((M-^ir5t^COO.— l(MCOmO>COC005C-t^t^00GCI00C0020»050505OOOOOO'— li— l.-(>-lr-li— i(M(MCt^OTt050iOcoQoco^co-*oocDO0it^C0l01>-00l^0i
.... , ,, . __...___. _l>.'X5COOiCOJ>-OiOOOOi^COOOCOO>COOt^O<:D0305iOOO
^OC0'X>C0OC^C't)'^^:JliOiO^Cur^i^'^*^(^0CNC^i-H05G0C0^CiOr-(C^i— lOCsr^vOCOO
0005I— liMCO^tOt-^OOOJOi— l(MCO'*iOOt^00050i— l>-l!M'-OC5!M^a5C^iOCOOCO;005(M
-*-#voiOiO>i-5\Oir5iOiO?0«DeD«DC05DCDC0500t^t^t^t-I:^t^cOOOCOOJCT)OJOOOOi— I
OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOrHi— 'i-li-lrH
ooooooooooooooooooooooooooooooooooooo
Cq(NCOCOCOTt.J:^0000050500i-li-l.-((M(M'#COCOO(N'*'»I^01t-(OOlOt~
COC003tNlOOOrH-^lr^000500StN»OOOT-l»r500T-l-^r~OCOvni>-OJIM-*<»000(Mmt— OSi-H
c<^o^|^^cococo■^■*'*oo>no50Xl«o^~^^^^^^oooooooiOJOl;^(^I■Jt^>p^^^-osOI^
0^^4-iAhi--(4-Ii-Ii— li--li--IC^C^C^C^Cai'^t:^i-ti-ir-i<^oz<^oooot^'^ait-~
OTOCDt---l^tDOt-(Mi-CCOOC>DOJt^r-,00«5-H
O CO O rH ,^ O Oi CO CO -^ O .p Oi ,^ rH Oi .p W t;- •*
t-ooo(^^cO'*lTJ^^n■■o^^ooooooo505Gooo^^~to»ri
III
0Jt-HC0l0t^i:D^I>-O05t^Ttr;CO(Nt->ncOOt^COt^COCO
coc^cooscooor^cocoot^coinaicoooc^cs^
(MrHOOOOCO^-^OOCOt-OiOiCOOOCDCO^lOO
CO«DOiCN--«COOOOi,-l(MCOTtMcocococo{M
«0Cn as U-; ,-1 CO CO 05 CO
COr-HCq,— (0-#i-Hi-<(MlOCOi— liOCOOSTXOi-^COCO t
0500cp(».;^t~-yO'^-^t--^'>pC^tp05O957*<^-
22S^S2S^^^^§S^SSS||S j
FU
i-lC<10.-lOCD.-H--rMi0 05CO.-ia5.-H005COCncOCOIr^CO
p-(
coa>o.-io5cqt^m^oo05co>ooovoc2-*t^coio
COC005000iOu-2 02io.-ICOCi-i j
CO(M(MO-*OJt^COCOCO(NCOC-
rHr-i.-i(MC^CSl(McococO'*-*T}r:u-5cocDcot-i>.w
Ol-^OJinOOTt^u-lCMt^oOt^WOOt-^lMt^OiOCO'^OO
C<305COCOIr^vni>-J>->OOi-ii— (COCOOllNCOlMmOJCOCO
oic^irtc^.-H(Mioo:)i-ioinoioco(M(MOcoocoi:nc<)coMi-,-(^co^-t^
O(MC0T(i'*Tj<^lO>OiC^^-*'^C0C0(NC^T-HOOS00
^-:*H^-*-*-*'#T*iT:HrJ<^-H-*'*^-:H-*-:H^-#COCO
+
os(Mr-Hioco^co^r-^t^mvn>ct-cococoicco^Oi
i-ii-ivot^cocoooot>-C3coooi-(Ot^»ncoi— fcoco^ir:
C0C0rHt-t-01C0OC5(MCiOC0C0OOu-. CO>rtTt0 >r5 >0 >-. >0 CO CO CO
C^OCOOO
T-HCOOlO^COCOCO^i— ICOOT-l^-^tNOSiOOCOirtlO
O^C005l-^0i>Oi0 0U:^00C0C0C0C0^00t^OC0C0O
lr^C0r^0i(Mr^C0r-lOO(Mioaii0C0r-lr-lC0
(MOCOi^C^t^Cc;
0500.-l(M(NCOCO'*tlC0OO-<*lt^C0i-lt^C0J:^OO
c6ol-5oicoeqt^(Ncornt^cow>nTHCT>t-(»co^0505
rHl^ l--^t^'cfW050COOOt^-i^-:»<(N^Oicocooco>r3 0oioiooococOi-iooa5(?oos
t:^O:Coco»C(McDco(M-*coi^co^i-~-coi-icoo-^eoo5t-i
05(N05.-H01:-COOO.-lCOCO(MOCO>-O(MC0^.0C0t-00C0
OSOi-l'— llM(N(M(MCOCOCOCOCO^-*'*<-*-*'*-^'ti-*
(NCOCOCOCOCOCOCOCOCOCOCOCOCOCOOOCOCOCOCOCOCO
a50e?^iOcor-co050rH(Ncp-^ip-op05prHoq
■"
C0
RESEARCHES ON EXPLOSIVES
185
The pressures given by the crusher-gauges (which can be com-
pared with those given in either of the Tables 10 or 11) at the points
A, B, C, 1, 4, are respectively 17-2, 15-6, 15-6, 12-8, and 111 tons
per square inch; or in atmospheres, 2169, 2376, 2376, 1949, and
1690.
We now pass to the consideration of the results furnished by
E. L. a powder. Taking, as in the case of pebble-powder, the
particular set of experiments where shot of 300 lbs. (136-05 kilos.)
were used, the data furnished by the chronoscope are given in
Table 12.
Table 12.— Giving data obtained with chronoscope for calculating the velocity and
pressure in the hore of a 10-inch 18-ton gun. Charge, 60 lbs. (27-2 kilos.)
B. L. G. Weight of projectile, 300 lbs. (136-05 kilos.).
Distance from base
Time observed
otal time
from
It of shot.
Time taken by
shot to traverse
Mean velocity over spaces
of shot.
at plugs. gg
distance
between plugs.
betweei
plugs.
Feet.
Metres.
Seconds.
Seconds.
Seconds.
Feet per
second.
Metres per
second.
0-00
0-000
...
000000
•nnn7fi7
78-2
23 •S
0-06
0-018
-000000
000767
000596
336
102-4
0-26
0-079
-000596
001363
000411
488
148^7
0-46
0-140
-001007
001774
000316
633
192-9
0-66
0-201
•001323
002090
000278
719
219-1
0-86
0-262
•001601
002368
000255
781
238-0
1-06
0-323
•001856
002623
000469
855
260-6
1-46
0-445
•002325
003092
r\f\r\A9.r\
935
285-0
1-86
0-567
•002755
003522
From these data, in the same manner as in the case of pebble-
powder, are calculated the velocities and pressures exhibited in
Table 13, p. 186.
The velocity and pressure obtained with the R. L. G. powder are
graphically represented by the dotted curves in Plate XV. (p. 230) ;
and by comparing these with the similar curves furnished by pebble-
powder, the advantages obtained by the use of the slow-burning
pebble-powder are clearly seen.
Thus it will be observed that the muzzle-velocity obtained with
the pebble-powder is 1530- feet (466-3 metres), while the maximum
pressure in the bore is 18 tons per square inch (2745 atmospheres).
The velocity given by the R. L. G-. powder is, on the other hand, only
1480 feet (451-1 metres), while the maximum pressure is 22-07 tons
per square inch (3360 atmospheres).
186
RESEARCHES ON EXPLOSIVES
Table IZ.-— Giving the total time from commencement of motion, velocity, and
tension of products of explosion in bore of 10-inch 18-ton gun, deduced from
Table 12.
Feet. Metres.
0-00
0-02
0-04
0-06
0-08
0-10
0-12
0-14
0-16
0-18
0-20
0-22
0-24
0-26
0-28
0-30
0-32
0-34
0-36
0-38
0-40
0-42
0-44
0-46
0-48
0-50
0-52
0-54
0-56
0-66
0-76
0-86
0-96
1-06
1-16
1-26
1-36
1-46
1-56
1-66
1-76
1-86
•000
•006
•012
•018
•024
•030
•037
•043
•049
•055
•061
•067
•073
•079
•085
•091
•098
•104
•110
•116
•122
•128
•134
•140
•146
•152
•158
•165
•171
•201
•232
•262
•293
•323
•353
•384
•415
•445
•475
•506
•536
•567
Seconds.
•0000000
•0005164
•0006615
•0007674
•0008548
•0009310
•0009994
•0010621
•0011204
•0011750
•0012267
•0012760
•0013231
•0013685
•0014123
•0014547
•0014959
•0015360
•0015751
•0016132
•0016505
•0016870
•0017229
•0017580
•0017925
•0018264
•0018598
•0018927
•0019250
•0020802
•0022262
•0023649
•0024976
•0026253
•0027487
•0028685
•0029851
•0032102
•0033193
•0034264
•0035318
Over Intervals.
•0005164
•0001451
•0001059
•0000874
•0000762
•0000684
•0000627
•0000583
•0000546
•0000517
•0000493
•0000471
•0000454
•0000438
•0000424
•0000412
•0000401
•0000391
•0000381
•0000373
•0000365
•0000359
•0000351
•0000345
•0000339
•0000334
•0000329
•0000323
•0001552
•0001460
•0001387
•0001327
•0001277
•0001234
•0001198
•0001166
•0001138
•0001113
•0001091
•0001071
•0001054
Velocity.
137^8
188^8
228-7
262-5
292-2
318-9
343-3
365-7
386-6
405-9
423-9
440-6
456-4
471^2
485-3
498-7
511-4
523-7
535-4
546-8
557-7
568-4
578-7
588-8
598-5
608-1
617^5
644-3
684-9
721^0
753^5
783^1
810-1
834-8
857-6
878^7
898^2
916^3
933-2
949-1
Metres pe:
second.
11^80
42^00
57^55
69^71
80-01
89-06
97-20
104-64
111-46
117-83
123-72
129-20
134-29
139-11
143-62
147-92
152-00
155-87
159-62
163-19
166^66
169^99
173-25
176^39
179^46
182^42
185 ^35
188^21
196-38
208-76
219-76
229-66
2.38-69
246-92
254-44
261-39
267-83
273-77
279-29
284^44
289-28
Tons per
sq. inch.
7-950
21-204
22-065
22-039
21-999
21-840
21-628
21-403
21-138
20-767
20-276
19-746
19-216
18-713
18^249
17^851
17^440
17^096
16-778
16-499
16-261
16-036
15-863
15-691
15-558
15^439
15^320
15^201
14^700
14^286
13^451
12-722
12-060
11-384
10-774
10-204
9-701
9-210
8-720
8-296
7^885
Atmo-
spheres.
1211
3229
3360
3356
3350
3326
3293
3259
3219
3162
3088
3007
2926
2850
2779
2718
2656
2603
2555
2512
2476
2442
2416
2389
2369
2351
2333
2315
2238
2175
2048
1937
1836
1734
1641
1554
1477
1402
1328
1263
1201
If, as in the case of pebble-powder, we express for the first
instants of motion the relation between s and t by an equation of the
form of that given in (9), we obtain
g ^ .57g37^3-42802--02336« + -000245;2 _ _ (14)*
* In this equation and Table 14, the unit of time is, for convenience, the one
ten-thousandth part of a second.
RESEARCHES ON EXPLOSIVES 187
and the values of s, v, T corresponding to those of t are given in the
scheme shown in Table 14, p. 188.
The results of Table 14, in comparison with those of the other
mode of calculation (Table 13), are graphically compared in Plate
XVII. (p. 230). It will be observed that, as in the case of pebble-
powder, the two methods give values closely accordant ; and if Plate
XVII. (p. 230) be compared with Plate XVI. (p. 230), the differences
in velocity and pressure at the commencement of motion between the
two natures of powder are very strikingly shown. Thus it will be
observed that with pebble-powder the maximum pressure, 2745
atmospheres, is reached when the projectile has moved "5 foot (152
metre), and at about '00437 second after the commencement of motion.
With E. L. G. powder the maximum pressure, 3365 atmospheres, is
reached when the projectile has moved only '05 foot ("015 metre), and
at about "00070 second from the commencement of motion. The first
foot of motion is, with the one powder, traversed in about "0025
second, with the other in about "0051 second.
The pressure given by the crusher-gauges in the experiments with
R. L. G. under discussion (and these pressures should be compared
both with those given in Table 13 and with the crusher-gauge pres-
sures furnished at the same points by pebble-powder) were, at A, B,
C, 1, and 4, respectively 44-2, 30*3, 22-5, 13-5, 12 tons per square inch ;
or, in atmospheres, 6731, 4614, 3426, 2056, and 1827.
In deducing the pressure from the velocity, we of course assumed
that the gaseous products of combustion acted on the projectile in
the manner in which gases are generally assumed to act.
With the slower-burning powders this hypothesis appears to be
not far from the truth; but with the more explosive powders the
crusher-gauges show that the powder acts on the shot, as we have
already observed, by a succession of impulses ; and in this case the
curve of pressures derived from the chronoscopic observations must
be taken to represent the mean pressures acting on the projectile
throughout the bore.
With the various powders experimented on by the Committee on
Explosives, there have of course been very great variations in the
pressures indicated.
The highest mean pressure indicated by the chronoscope was 30'6
tons, 4660 atmospheres ; and this pressure was attained with a charge
of 60 lbs. E. L. G., and a projectile weighing 400 lbs. In the same
series, the highest local or wave-pressure exhibited by the crusher -
gauges was 57"8 tons, 8802 atmospheres ; but this excessive pressure
188
RESEARCHES ON EXPLOSIVES
» -w
■» s
^ •
.SCi5
1^
,^(i^
-^
947
2057
2849
3230
3311
3342
3356
3356
3318
3251
3131
3003
2866
2727
2593
2362
2199
2127
(M,-li-l.-lTjlio^'J<010CO— ICOOO(M^'a'Cii-ltOOCOCDOi— li-l(N
I— l\rimOiOi— lrHrHOOOOt^lOCO(NO>t^lX>
t^C0i--;c3iOi— l!M!MOOl-~>J-l>-l^r-{COt:^tO(N
p;
03ir5a5>nu:i>r2^xtii-ioo(M-#t^cciTf.T-lrHlOCr30i— li— It^CQ^'^COOO'fflt^
05C5(M.-ioo5aicocor-iooo(Masc^TjHcD
?99"?P9'?99?9999?999
t^C0'#t^C00500CX)^OOCsC0a5oot-.i-.j.-.---!t^.o^050«Dvnoo«r-jco^
lOi— looco^L-mtocooiosoOi— ii— ii-Hoo
r-^,-l(^^c-IOOCDr-H«D
0 (M O OO l-^ 1.- Ttl CO 1— 1 CD CD O CO (M o ^ in
o ^ o o ^ c<] >n ^ I- -*< «-. oj r- a> rH lo o .o .-1
(N>ncDt^C0O5C^lOO5C0J:^(Mt^0ii— IVOOSIO
rH>-lr-l(N(MCOCO'HCOt-000
s
ooot^r-(vn(MO.-io>t^coojco«oo5eoT-iinvOi-i
i^s§Srtsss§Sf^gS5g2^;s§f^S
C0C0O5>Or»— IO0CCD^C0t-lOt^'#0Q05t^
•«•
in »n
RESEARCHES ON EXPLOSIVES 189
was exhibited only in the crusher marked A in Plate XIV., Fig. 3
(p. 230), and was probably confined to that particular point. The
pressures exhibited by the same powder in the same round, at the
points B and C in the powder-chamber, were respectively 37 tons,
5634 atmospheres ; and 29-6 tons, 4507 atmospheres.
But although, in the various guns and with the various charges
and special powders experimented with, the pressures at different
points of the bore exhibit, as might be expected, marked differences,
these differences almost altogether disappeared when powders of
normal types and uniform make were experimented with, and when
the pressure was referred, not to fixed positions in the bore of the
gun, but to the density of the products of combustion.
We have already referred to the experiments made with cylinders
gradually increasing in weight in the 10-inch gun. A similar series
was made in the 11-inch gun with charges of powder of 85 lbs. (38"56
kilos.); and as the series in both guns were made with great care and
under as nearly as possible the same conditions, we selected, in the
first instance, the experiments with pebble-powder in these guns, to
test the accordance or otherwise of the tensions, under the varied
conditions of experiment, when taken simply as functions of the
density.
The results of these calculations are graphically represented in
curves 1 and 2, Plate XYIII. (p. 230) ; and it will be observed that
with these different calibres and charges the tensions developed are
as nearly as possible identical.
Curves 3 and 4 on the same plate exhibit the results of similar
calculations for 60 lbs. Pt. L. G. fired in the 10-inch gun, and 30 lbs.
R. L. G. fired in the 8-inch gun. In this case also, although there
are differences between the curves representing the pebble and R. L. G.
powders, to which we shall allude further on, the accordance between
the same description of powder fired from the different guns is
almost perfect.
(s) EFFECT OF INCREMENTS IN THE WEIGHT OF THE SHOT ON THE COM-
BUSTION AND TENSION OF POWDER IN THE BORE OF A GUN.
In our preliminary sketch of the labours of previous investigators,
we alluded to the views held by Ptobins and Paimford upon the
rapidity of combustion within the bore. The latter, relying chiefly
upon the fact that powder, especially when in very large grains, was
frequently blown unburned from the muzzle, concluded that the com-
190 RESEARCHES ON EXPLOSIVES
bustion was very slow. Kobins, on the other hand, considered that,
with the powder he employed, combustion was practically completed
before the shot was materially displaced ; and it is not easy to see why
the unanswerable (if correct) and easily verified fact of which he makes
use has received so little attention from artillerists.
Eobins, it will be remembered, argues that if, as some assert, a
considerable time is consumed in the combustion of the charge, a
much greater effect would be realised from the powder where heavier
projectiles were used, but that such is not the case.
The Committee on Explosives have completely verified the correct-
ness of Eobins' views.
In the 10-inch gun, with a charge of 60 lbs. (27-2 kilos.) E. L. G.
powder, the work realised from the powder is only increased by about
5 per cent, when the weight of shot is doubled.
In the slower-burning pebble-powder, with a charge of 70 lbs.
(31"75 kilos.), with a similar increase in the shot, the greater effect
realised was about 8^ per cent. ; but when the weight was again
doubled (that is, increased to four times the original weight), the
additional effect was barely 1 per cent.
Piobert's views, moreover, that the pressure exercises but a trifling
influence upon the rate of combustion, appears to us entirely unten-
able. With a particular sample of service pebble-powder, we found
the time required for burning a single pebble in the open air to be
about 2 seconds. The same sample was entirely consumed in the
bore of a 10 -inch gun, and must therefore have been burned in less
than -009 second.
(t) EFFECT OF MOISTUKE UPON THE COMBUSTION AND TENSION
OF POWDER.
It is perhaps unnecessary to say that we do not share the views
of those who consider that the presence of water in powder may
increase the tension of the products of explosion. We have made no
experiments upon this head in closed vessels ; but the following table
exhibits the effect of moisture in gunpowder upon the velocity of
the projectile and the tension of the gas when the powder is fired in
a gun, the proportions of moisture varying from 0'7 to 1"55 per cent.
The powder from which these results were obtained, was pebble, care-
fully prepared by Colonel Younghusband, and was the same in all
respects, except as regards the quantity of moisture.
RESEARCHES ON EXPLOSIVES
191
Table 15. — Showing the effect of moisture in the powder upon the
velocity of the projectile and pressure of the gas.
Percentage of
Moisture.
Velocity.
Maximum Pressures.
Feet.
Metres.
Tons per
square inch.
Atmospheres.
0-70
1545 i
470-92
22-02
3353
0-75
1541
774-50
21-70
3304
0-80
1537
468-47
21-38
3256
0-85
1533-5
467-41
21-07
3208
0-90
1530
466-34
20-77
.3163
0-95
1526-5
465-30
20-47
3117
1-00
1523-5
464-40
20-18
3073
1-05
1520-5
463-44
19-90
3030
1-10
1517-5
462-53
19-63
2989
1-15
1514-5
461-61
19-37
2949
1-20
1512
460-85
19-12
2911
1-25
1509-5
460-10
18-87
2873
1-30
1507
459-33
18-63
2837
1-35
1504-5
458-60
18-40
2802
1-40
1502
457-80
18-18
2768
1-45
1499-5
1 457-04
17-97
2736
1-50
1497-5
456-43
17-76
2704
1-55
1495-5
455-82
17-55
2672
From this table it will be seen that, by the addition of consider-
ably less than 1 per cent, of moisture, the muzzle-velocity is reduced
by about 60 feet, and the maximum pressure by about 20 per cent.,
pointing obviously to a much more rapid combustion in the case of
the drier powder.
(u) LOSS OF HEAT BY COMMUNICATION TO THE ENVELOPE IN WHICH
THE CHAEGE IS EXPLODED.
We have now given a hasty sketch of the means that have been
adopted to determine the pressures actually existing in the bores of
guns, and of the general results we have arrived at ; and before pro-
ceeding to the theoretical consideration of the relation which should
then exist between the tension and the density of the gases, we must
direct attention to an important point — and that is, " what loss of
heat do the gases suffer ? or, in other words, what proportion of energy
in the powder is wasted by communication to the envelope in which
the powder is fired, that is, to the barrel of the gun ? "
Every one is aware that if a common rifled musket be very
rapidly fired, as may easily now be done by the use of breech-loading
arms, the barrel becomes so hot that it cannot be touched with the
192 RESEARCHES ON EXPLOSIVES
naked hand with impunity, and, even with a field-gun, the increment
of heat due to a few rounds is very considerable.
So far as we know, the Count de Saint-Eobert* made the first
atten^t to determine the amount of heat actually communicated to
a small arm.
De Saint-Eobert made three series of experiments with service
rifled muskets, firing the ordinary charge of 4-5 grms. In the first
series, the muskets were loaded in the usual manner ; in the second
series, the ball was placed near the muzzle ; in the third, the muskets
were loaded with powder alone. The results at which De Saint-
Eobert arrived, and which are not difficult to explain, were, that the
greatest quantity of heat was communicated to the musket when the
ball was placed near the muzzle, that the quantity communicated
when no projectile at all was used, stood next in order, and that
least heat was communicated when the musket was loaded in the
usual manner.
He further found that the quantity of heat communicated in this
last case, with the powder and arm used, was about 250 grm.-units
per grm. of powder fired.
We found ourselves unable, however, to adopt Count de Saint-
Eobert's important results for the guns and charges we have been
considering, because conclusions derived from small arms could
hardly be applied to large ordnance without modificatiun.
We therefore instituted the experiments described under Nos. 72
and 73. The gun used was a 12-pr. B.L., and in the first Experiment
(No. 72) nine rounds were fired with If lb. (794 grms.) and a projectile
weighing nearly 12 lbs. (5330 grms.).
Prior to the rounds being fired, arrangements were made for
placing the gun, whenever the series should be concluded, in a vessel
containing a given weight of water ; and before the experiment was
commenced the gun and water were brought to the same temperature,
and that temperature carefully determined.
After the firing, the gun was placed in the water, and the rise of
temperature due to the nine rounds determined. This rise was found
to be equivalent to 236,834 grms. of water raised through 2°-305 Cent.,
or the heat communicated to the gun by the combustion of 1 grm. of
the charge was equal to 764 grm.-units.
Of course an addition has to be made to this number, on account
both of some loss of heat in the determination and of the unavoidable
loss of heat between the rounds.
* TraiU de Thermodnamique (Turin, 1S65), p. 120.
RESEARCHES ON EXPLOSIVES 193
The second Experiment (No. 73) was made with five rounds of
1| lb. (680-4 grms.) of the same powder with the same weight of
projectile. The heat communicated to the gun by the five rounds
was, when expressed in water, sufficient to raise 112,867 grms. through
2°-694 Cent., or 1 grm. of the charge, in burning, communicated to
the gun 894 grm.-units of heat.
Considering the difficulty, in an experiment of this nature, of
avoiding a considerable loss from radiation, conduction, and other
causes, we do not think we shall be far wrong in assuming that in the
case of the 12-pr. B.L. gun, fired under the conditions named, the
heat communicated to the gun is about 100 grm.-units for each
gramme of powder exploded.
To arrive at the amount of heat communicated to the gun when
still larger guns are employed, there are two principal points to be
considered — 1st, the ratio which the amount of the surface bears to
weight of the charge exploded ; and 2nd, the time during which the
cooling effect of the bore operates upon the products of explosion.
The first of these data is of course exactly known, and from our
experiments the second is also known with very considerable exact-
ness. Computing, therefore, from the data given by the 12-pr., the
loss of heat suffered by the gases in the 10-inch gun, we find that
loss to be represented by about 25 grm. -units ; and hence we find
that the quantity of work in the form of heat communicated to the
gun varies approximately from 250 grm.-units per grm. of powder
in the case of a rifled musket, to 25 grm.-units in the case of a
10-ineh gun.
Similar considerations lead us to the conclusion that in a close
vessel such as we employed for explosion, lohen filled with powder,
the loss of pressure due to the communication of heat to the envelope
would not amount to 1 per cent, of the total pressure developed.
(-y) PKESSURE IN THE BORES OF GUNS, DERIVED FROM THEORETICAL
CONSIDERATIONS.
We now pass to the theoretical consideration of the question.
Suppose the powder to be fired, as is the case in the chamber of a gun,
and suppose, further, that the products of combustion are allowed to
expand, what will be the relation between the tension of the gases and
the volume they occupy throughout the bore ?
For the sake of simplicity, we shall, in the first instance, assume
that the gravimetric density of the powder is unity, that the powder
N
194 RESEARCHES ON EXPLOSIVES
fills completely the space in which it is placed, that the whole charge
is exploded before the projectile is sensibly moved from its initial
position, and that the expansion takes place in a vessel impervious to
heat.
In our preliminary sketch we alluded to the results of Button's
investigations as to the relations existing between the density and
tension of the gases and the velocity of the projectile at any point of
the bore. Hutton, however, assumed that the tension of the inflamed
gases was directly proportional to their density, and inversely as the
space occupied by them. In other words, he supposed that the
expansion of the gases, while doing work both on the projectile and on
the products themselves, was effected without loss of heat.
Kecent research, which has demonstrated that no work can be
effected by the expansion of gases without a corresponding expen-
diture of heat, has enabled modern artillerists to correct Hutton's
assumption ; and the question of the pressure exercised and work per-
formed by gunpowder in the bore of a gun has been examined
both by Bunsen and Schischkoff, and by the Count de Saint-Eobert.*
De Saint-Eobert, like Hutton, supposed that the whole of the
products of the explosion were, on ignition, in a gaseous state, and
that hence the relation between the pressure and the volume of the
products followed from the well-known law connecting the tension
and volume of permanent gases.
Bunsen and Schischkoff, on the other hand, who, like ourselves,
have arrived at the conclusion that at the moment of explosion a
large part of the products is not in the gaseous state, have deduced
the total work wlrich gunpowder is capable of performing, on the
assumption that the work on the projectile is effected by the expan-
sion of the permanent gases alone, without addition or subtraction of
heat, and that, in fact, the non-Q;aseous products play no part in the
expansion.
Sufficient data were not at the command of either of the authori-
ties we have named, to enable them adequately to test their theories ;
and we propose in the first place, with the data at our disposal, to
compare their hypothesis with actual facts, by computing the tensions
for different volumes and comj^aring the calculated results, both with
the tensions in a close vessel and with those derived from actual
experiments in the bores of guns.
Assuming, in the first place, with De Saint-Eobert, that the whole
of the products are in the gaseous form, —
*Trait6 de Thermodynainique, p. 164.
RESEARCHES ON EXPLOSIVES 195
Let p be the value of the elastic pressure of the permanent gases
by the combustion of the powder corresponding to any
volume V, and let p^^, v^ be the known initial values of p and v. Let
also C be the specific heat of these gases at constant pressure,
and C„ be the specific heat at constant volume. Then, from the
well-known relation existing between p and v, where a permanent
gas is permitted to expand in a vessel impervious to heat, we have
P^\
■^y^ . . . . (15)
and this equation, upon De Saint-Eobert's hypothesis, expresses the
relation between the tension of the gases and the volume occupied by
them in the bore of a gun.
Taking p^ from Table 8, at 41 "477 tons per square inch, and
assuming at unity the space v^ occupied by the charge when at a
C
gravimetric density of 1, taking, further, the value of -^ = 1-41 as
computed by De Saint-Eobert, Equation (15) becomes
;. = 41-477(1) .... (16)
If we now take Bunsen and Schischkoff's view, that a portion
only of the products is in the form of permanent gases, and that they
expand without addition or subtraction of heat, we are able, from
Equation (15), to deduce the law connecting the tension and the
pressure. For if we call v and v'^ the volume at any instant and
the initial volume of the permanent gases, we have from (15)
pA'^V" .... (17)
but if a be the ratio wliich the volume of the non-gaseous products
at the moment of explosion bears to that of the unexploded powder,
we have
^''o = Vl-«)» V = v-av^ . . . (18)
and Equation (17) becomes
"^i^y- ■ ■ ■ (-)
196
RESEARCHES ON EXPLOSIVES
and this is the relation between p and v on Bunsen and Schischkoff s
hypothesis.
Taking, as before, ^o = 41 -477, -^0 = 1, and remembering that we
have found the value of a to be '6, we have
•4 \c„
'' - "-"'C^cr"
(20)
The value of the exponent -^ can be deduced from the data
given in Table 16.
Table 16. — Showing the percentage iveights, specific heats at constant volume, and
the specific heats at constant pressure of the permanent gases produced hy
the explosion of powder.
Nature of gas.
Percentage weight
of gas.
Specific beat at
constant pressure.
Specific heat at
constant volume.
Sulphuretted hydrogen
Carbonic oxide
Carbonic anhydride
Marsh-gas ....
Hydrogen ....
Nitrogen ....
•0262
•1036
•6089
•0012
•0023
•2579
•2432
•2450
•2169
•5929
3-4090
•2438
•1840
•1736
•1720
•4680
2^4110
•1727
From the data in this table the value of C^ is found to be =
•23528, of a = 1782, and that of the fraction -rf = 1-3203; and
Equation (20) becomes
P
41-477
C^eT
(21)
The results of (16) and (21) are given in Table 17; and in the
same table are given the values of p, both as deduced from actual
experiment in the bore of the 10-inch and 11 -inch guns (see Plate
XVIII., p. 230), and also as deduced from our experiments in a close
vessel. The results of the experiments upon the tension of different
densities in a close vessel represent of course the elastic force which
would exist were the gas allowed to expand in a vessel impervious to
heat, without production of work.
RESEARCHES ON EXPLOSIVES
197
Table 17. — Shovnng in terms of the density (1) the tension actually found to exist
in the bores of guns ; (2) the tension which would exist were the gases suffered
to expand without production of work; (3) the tension calculated upon
De Saint-Robert's hypothesis ; (4) the tension calculated on Bunsen and
Schischkoff's hypothesis.
Tension observed
Tension observed
Tension calculated
Tension calculated
in bore
where the gases
upon
Count
upon Bunsen
of 18-ton gun
expand without
De St.-Robert's
and Schischkoff's
Mean density
(pebble-powder).
production of work.
hypothesis.
hypothesis.
of products of
combustion.
Tons
Atmo-
Tons
Atmo-
Tons
Atmo-
Tons
Atmo-
per
sq. inch.
spheres.
per
sq. inch.
spheres.
per
sq. inch.
spheres.
per
sq. inch.
spheres.
1-00
41-48
6320
41-48
6320
41-48
6320
•90
20-35
3101
32-46
4946
35-75
5448
30-00
4572
•SO
17-01
2590
25^52
3889
30-14
4593
21-85
3330
•70
14-03
2133
20-02
3051
25-08
3822
15-85
2416
•60
11-33
1722
15-55
2370
20-18
3076
11-62
1771
•50
8-87
1352
11-85
1806
15-61
2378
7-93
1209
•40
6-65
1019
8-73
1330
11-40
1736
5-30
808
•30
4-67
722
6-07
925
7-60
1157
3^28
500
•20
2-93
459
3-77
574
4-29
653
1-75
267
•10
1-77
270
1-61
246
•64
98
The graphical representation of this table is given in Plate XIX.
(p. 230) ; and by examination either of the table or of the curves, it is
obvious that neither Formula (16) nor (21) gives results which can be
taken as at all representing the truth. The values of the elastic force,
calculated on the assumption that the whole of the products of com-
bustion are in the gaseous state, and that the effect on the projectile
is produced by such expansion, are largely in excess of the pressures
observed in the gun, and very greatly in excess even of the pressures
observed when the gases were expanded without production of work
On the other hand, the pressures calculated on the assumption that
the work is caused by the expansion of the permanent gases alone,
without addition or subtraction of heat, are considerably in defect of
those actually observed, and this too, although no allowance is made
for the absorption of heat by the gun.
At an early stage in our researches, when we found, contrary
to our expectation, that the elastic pressures deduced from experiments
in close vessels did not differ greatly (where the powder might be
considered entirely consumed, or nearly so) from those deduced from
experiments in the bores of guns themselves, we came to the con-
clusion that this departure from our expectation was probably due
to the heat stored up in the liquid residue. In fact, instead of the
expansion of the permanent gases taking place without addition of
198 RESEARCHES ON EXPLOSIVES
heat, the residue, in the finely divided state in which it must he on
the ignition of the charge, may be considered a source of heat of the
most perfect character, and available for compensating the cooling
effect due to the expansion of the gases on production of work.
The question, then, that we now have to consider is — What will
be the conditions of expansion of the permanent gases when dilating
in the bore of a gun and drawing heat, during their expansion, from
the non-gaseous portions in a very finely divided state ?
To solve this question we must have recourse to certain well-
known principles of thermodynamics.
Let dB. be the quantity of heat added to, or drawn from, the non-
gaseous portion of the charge by the permanent gases, while the
latter pass from the volume v and temperature t to the volume
v' -\- dv' and temperature t -f dt, we then have *
dH = t.dcf, . . . . (22)
^ being Eankine's thermodynamic function.
But if X be the specific heat of the non-gaseous portion of the
charge, and if ^ be the ratio between the weights of the gaseous and
non-gaseous portions of the charge, and if we assume further, as we
can do without material error, that X is constant, we shall have
dH = - /3\dt .... (23)
and by integration
d^=-(3X± . . . . (24)
But the value oi (p — (pQ for permanent gases is well known, being
readily deduced from the general expression for the thermodynamic
function.
This expression being f
c/> = Clog^M-i^.||.rf.' . . . (26)
(J being Joule's equivalent), and ~ being readily obtained from the
equation of elasticity and dilatability of perfect gases,
pv' = R/
* Rankine, Steam Engine, p. 310. De Saint-Robert, loc. cit., p.
t Rankine, loc. cit., p. 311. De Saint-Robert, loc. cit., p. 72.
RESEARCHES ON EXPLOSIVES
we deduce from (26), by integration,
^-4>,
.--(i)"
'■(0
since* J =
=c,-a.
=-.a)'
'■(3)"'-'
Hence,
equating (25) and (27),
Therefore
1 =
-0
and
or, since v
'o = ^o(l-
-a), v
= v-aVQ,
p (-Vo(l-a)|c^+^X
199
(27)
(28)
(29)
(30)
and Equation (30) gives the true relation connecting p and v when
the gaseous products expand in the bore of a gun with production
of work.
The values of the constants in this equation we have already
determined ; they are as follow : —
C„=-1782, C^ = -2353, p,==il-An, A = -4090, ^ = 1-3148.
The results of Equation (30) are given in Table 18, p. 200 ; and, as
before, for comparison we give similar values of p both as derived
from experiments with heavy ordnance and on the supposition of
expansion without performance of work.
The results of Table 18 are graphically represented in Plate
XX. (p. 230) ; and on the axis of abscissse are figured, for convenience,
both the density of the products and the volume they occupy.
* Rankine, loc. cit., p. 318. Clausius, loc. cit., p. 39. De Saint-Robert, loc. ciU,
p. 93.
200
RESEARCHES ON EXPLOSIVES
Table 18. — Giving, in terms of the density, the tensions actually found to exist in
the bores of guns toith pebble and It. L. G. powders ; giving, further, (1) the
tensions calculated from Equation (30) ; (2) the tension which toould exist
loere the gases suffered to expand without production of toork.
Tension observed
Tension observed in
Tension calculated
Tension observed
when the gases
expand without
production of work.
Mean density
in bores of guns.
Pebble-powder.
bores of guns.
R. L. G. powder.
from
Formula (30).
of products of
combustion.
Tons
Atmo-
Tons
Atmo-
Tons
Atmo-
Tons
Atmo-
per
sq. inch.
spheres.
per
sq. mch.
spheres.
per
sq. inch.
spheres.
per
sq. inch.
spheres.
1-00
41-48
6316
41-4,8
6316
•95
36-30
5528
36 ^65
5581
•90
20-35
3099
27-33
4162
31-84
4848
32^46
4943
•85
18-63
2837
24-63
3751
27-95
4256
28^78
4383
•80
17-01
2590
22-01
3352
24-56
3740
25^53
3888
•75
15^48
2357
19-50
2969
21-56
3283
22^63
3446
•70
14-03
21.36
17-16
2613
18-89
2877
20-02
3049
•65
12-65
1926
15-05
2292
16-51
2514
17-68
2692
•60
11-33
1725
13-21
2011
14-38
2190
15-55
2368
•55
10^07
1533
11-61
1768
12-46
1897
13^62
2074
•50
8-87
1351
10-18
1550
10-72
1632
11-85
1804
•45
7^73
1177
8-87
1351
9-15
1393
10-23
1558
•40
6^65
1013
7-65
1165
7-71
1174
8-73
1329
•35
5^63
857
6-49
988
6-40
975
7-35
1119
•30
4^67
711
5-39
821
5-21
793
6-07
924
•25
3^77
574
4-34
661
4-11
626
4-88
743
•20
2-93
446
3-33
507
3-11
474
3-77
574
•15
2-15
327
2-35
358
2-20
335
2-73
416
•10
1-37
209
1-76
268
The curve marked A represents the tensions deduced (with a
slight correction for loss of heat) from actual observation in a close
vessel, and may, as we have already said, be taken to represent the
pressures that would exist were the products of combustion allowed to
expand in a vessel impervious to heat and without production of work.
The curve marked B, derived from Equation (30), denotes the
tensions that would exist in the bore of a gun, if we suppose the
powder, of a gravimetric density = 1 and filling entirely the chamber,
to be completely consumed before the projectile is moved from its
place, and to expand in a gun impervious to heat. By comparison
with the curve A will be seen the difference in tension arising from
the loss of heat due to the work expended. The great importance of
the heat contained in the non-gaseous portion of the charge is
rendered apparent by comparison of curve B with curve 4, Plate
XIX. (p. 230), or Table 17, where, on Bunsen and Schischkoff 's hypo-
thesis, the permanent gases are supposed to expand without deriving
any heat from the non-gaseous portion of the charge.
RESEARCHES ON EXPLOSIVES 201
The area comprised between curve B and the axis of abscissae
represents the maximvim work that it is possible to obtain from
powder.
Curve C represents the mean results obtained with R. L. G. powder
from the 8-inch and 10-inch guns, and curve D represents the mean
results obtained with pebble-powder from the 10-inch and 11-inch
guns.
It is interesting to study the differences exhibited by these curves
B, C, and D. The curve C, representing the pressures obtained with
E. L. G., denotes tensions not far removed from the theoretic curve,
while the densities are still very high ; before the volume is much
increased, the two curves slide into one another and become almost
coincident.
The curve D, on the other hand, is at first very considerably
below both the E. L. G. and the theoretic curve. It is still consider-
ably lower even when the E. L. G. curve is practically coincident
with the theoretic curve, and it retains a measurable though sHght
inferiority of pressure even up to the muzzle of the gun.
These differences are without doubt due to the fact that with the
E. L. G. powder, at least under ordinary circumstances, the whole or
a large proportion of the charge is consumed before the projectile is
greatly removed from the seat of the shot. With the slower-burning
pebble-powder, on the other hand, a considerable quantity of powder
remains unconsumed until the projectile approaches the muzzle ; and
the curve indicates in a very striking way the gradual consumption
of the powder, and the portion of the bore in which the slow-burning
powder may be considered practically burned.
It might perhaps be expected that the difference between the
theoretic curve B and the observed curves C and D near the muzzle
would be greater than is shown, since the curve B has been obtained
on the supposition that the expansion has taken place in a vessel
impervious to heat.
We have pointed out, however, that although in muskets and
small arms the loss of heat arising from communication to the bore
is very considerable, it is comparatively unimportant in very large
guns. In our calculations also we have taken X, the specific heat of
the non-gaseous portion of the charge, at its mean value. It should,
however, be taken at a higher value, since the specific heat must
increase rapidly with the temperature ; and this difference no doubt
more than compensates for the loss of heat to which we have referred
as not being taken into account.
202 RESEARCHES ON EXPLOSIVES
Our hypothesis as to a portion of the charge remaining uncon-
sumed until the projectile approaches the muzzle, is confirmed by
the well-known fact that in short guns, or where powder of high
density or very large size is employed, considerable quantities some-
times escape combustion altogether.
The appearance of pellet or pebble powder which has been
ignited and afterwards extinguished in passing through the atmo-
sphere, is well known to artillerists.
The general appearance (and in this appearance there is wonderful
uniformity) is represented in Plate XI., Fig. 5 (p. 230), and gives the
idea of the combustion having proceeded from centres of ignition.
If we imagine a grain, or rather (taking into account the size of
the grains of the present day) a pebble, of powder arriving uncon-
sumed at a point a little in advance of that of maximum pressure, it
is not difficult to conceive that such pebble will traverse the rest of
the bore without being entirely consumed, when the great influence
of diminished pressure, combined with the shortness of time due to
the increasing velocity of the projectile, is considered.
Thus, by reference to Table 10, it will be found that the time
taken by the projectile to describe the first foot ("305 metre) of
motion is about "005 second, while the time taken to describe the
remaining length of the bore, 7'25 feet (2-21 metres), is only about
"Oil second.
The mean powder-pressure over the first foot, again, is about 15
tons per inch (2300 atmospheres), and over the remainder of the
bore is only 5-25 tons (800 atmospheres).
(w) TEMPEKATUKE OF PRODUCTS OF COMBUSTION IN BOKES
OF GUNS.
The temperature in the bore of the gun during the expansion of
the products is given by Equation (28), or, restoring the values of v'
and v'q,
"-{ v-ar
Y^"-^^ . . . (31)
The temperatures calculated from this formula are given in
Table 19. It is hardly necessary to point out that the values given
in this table are only strictly accurate when the charge is ignited
before the projectile is sensibly moved ; but in practice the correction
due to this cause will not be great.
RESEARCHES ON EXPLOSIVES
203
Taui.e 19.— Giving the temperature in degrees Centigrade, and in terms of the
density, of the products tvhen expanded, with production of work, in the bore
of a gun supposed impervious to heat.
Number of
Temperature.
Mean density
Number of
Temperature.
of products of
volumes of
Degrees
of products of
volumes of
Degrees
combustion.
expansion.
Centigrade.
combustion.
expansion.
Centigrade.
1-00
1-0000
22°31
•50
2-0000
2019
95
1-0526
2209
•45
2^2222
1996
90
1-1111
2188
•40
2-5000
1971
85
1-1765
2167
•35
2-8571
1943
80
1-2500
2146
•30
3-3333
1914
75
1-3333
2126
•25
4-0000
1881
70
1-4286
2105
•20
5-0000
1843
65
1-5385
2084
•15
6-6667
1796
60
1-6667
2063
•10
10-0000
1734
•55
1-8182
2041
•05
20-0000
1637
{x) WOEK EFFECTED BY GUNPOWDER.
The theoretic work which a charge of gunpowder is capable of
effecting during the expansion to any volume v is, as we have said,
represented by the area between the curve B, Plate XX. (p. 230), the
ordinates corresponding to v and v^^, and the axis of abscissae. In
mathematical language, it is expressed by the definite integral
dv
(32)
Replacing in this equation the value of p derived from Equation (30)
we have for the work done by the powder in expanding from v^ to v,
3(l-a) -.C^+^X^^, .... (33)
^0 J
W
7^0
Po^o(l-«)(C. + W h
c -c
V v-av^J
%{l-a) \%+^X\
(34)
The values of all the constants in this equation have already been
given ; but for our present purpose it is convenient to determine the
work which 1 grm. of powder is capable of performing for different
degrees of expansion. Assuming, then, that a gramme of powder is of
the gravimetric density of unity (that is, that it occupies a volume
of 1 CO.), we have Vq = 1; and expressing the initial pressures 41*5
tons (6320 atmospheres) in grammes per square centimetre, we have
^0=6,532,450 grms. per square centimetre.
We have calculated W from (34) from various values of v up to
and inclusive oi v = 20. The results are embodied in the following
204
RESEARCHES ON EXPLOSIVES
table, and are expressed both in kilogrammetres per kilogramme
and foot-tons per lb. of powder.
Table 20. — Giving the total work that gunpowder is capable of per-
forming in the bore of a gun, in kilogrammetres per kilogramme
and foot-tons per lb, of powder burned, in terms of the density of
the products of explosion.
Total work that the gunpowder
is capable of realising.
Density of products
of
Number of volumes
Per kilogramme
burned in
Per lb. burned
combustion.
of expansion.
kilogrammetres.
foot-tons.
•95
1-0526
3210-8
4-70
90
1-1111
6339-6
9-29
85
1-1768
9412-8
13-79
80
1-2500
12443-8
18-23
75
1-3333
15460-8
22-65
70
1-4286
18488-1
27-08
65
1-5385
21544-9
31-56
60
1-6667
24650-8
36-11
55
1-8182
27841-9
40-78
50
2-0000
31153-7
45-62
45
2-2222
34614-0
50-70
40
2-5000
38290-0
56-08
35
2-8571
42234-7
61-86
30
3-3333
46565-9
68-21
25
4-0000
51414-8
75-31
20
5-0000
57031-7
83-53
17
5-8824
60952-1
89-35
16
6-2500
62368-1
91-45
15
6-6667
63884-4
93-64
14
7-1429
65470-1
95-94
13
7-6923
67138-4
98-39
12
8-3333
68940-1
101-00
11
9-0909
70855-4
103-82
10
10-0000
72903-7
106-87
9
11-1111
75214-5
110-18
8
12-5000
77679-9
113-81
7
14-2857
80462-1
117-85
6
16-6667
83582-1
122-42
•5
20-0000
87244-4
127-79
The results embodied in this table are of very considerable
importance. They enable us to say by simple inspection what is
the maximum work that can be obtained from powder such as is
employed by the British Government in any given length of gun.
To make use of the table, we have only to find the volume occupied
by the charge (gravimetric density =1) and the number of times this
volume is contained in the bore of the gun. The maximum work*
* It is hardly necessary to point out that the velocity of the projectile at any
point of the bore is directly deducible from Equation (34).> For the velocity being
connected with the work by the equation
velocity =./-^.W
\ w
RESEARCHES ON EXPLOSIVES 205
per kilogramme or pound which the powder is capable of performing
during the given expansion, is then taken out from the table ; and
this work being multiplied by the number of kilogrammes or pounds
in the charge, gives the total maximum work. Thus, for example, in
an 18-ton 10-inch gun, a charge of 70 lbs. (31'75 kilos.) pebble-
powder is fired, and we wish to know what is the maximum work
that the charge is capable of performing. "We readily find that the
length of the gun is such that 'y=5'867 vols.; and from the table
we find that 89-4 foot-tons or 61,000 kilogrammetres is the maximum
work per lb. or per kilog. ; multiplying by the number of pounds
or kilos., we find that 6258 foot-tons or 1,936,750 kilogrammetres
is the maximum work which the whole charge is capable of
performing.
As a matter of course, this maximum effect is only approximated
to, not attained ; and for actual use it would be necessary to multiply
the work so calculated by a factor dependent upon the nature of the
powder, the mode of firing it, the weight of the shot, etc. ; but in
service-powders fired under the same circumstances, the factor will
not vary much. In the experimental powders used by the Com-
mittee on Explosives, there were, it is true, very considerable
differences, the work realised in the same gun varying from 56
foot-tons to 86 foot-tons per lb. of powder; but with service-
powders fired under like conditions this great difference does not
exist.
We have prepared at once, in illustration of the principles we
have just laid down, as a test of the general correctness of our views
and as likely to prove of considerable utility, a table in which we
have calculated, from the data given, first, the total work realised
per lb. of powder burned for every gun, charge, and description of
powder in the English service ; second, the maximum theoretic work
per lb. of powder it would be possible to realise with each gun and
charge ; and third, the factor of effect with each gun and charge —
that is, the percentage of the maximum effect actually realised.
w being the weight of the shot, we have only to take out, from Equation (34) or
Table 20, the value of W for any given expansion, multiply it by the " factor of
effect " (see p. 206) for the particular gun, charge, etc. , and use in the above equation
the value of W so found.
As an illustration, if it be required to determine the velocity at the muzzle of
the 10-inch gun under the circumstances discussed at p. 205, the total work, as
shown in the text, which the charge is capable of effecting, is 6258 foot-tons ;
multiplying this by the factor for the gunpowder and weight of shot, we have
W = 4880 foot-tons ; substituting this value of W in the above equation, we obtain
i5 = 1532 feet, or nearly identical with the observed velocity.
206
RESEARCHES ON EXPLOSIVES
^•1
1^ «
1%
I
1
rHt^C0C00C.-IOOl.^l^r-lrHcrit^C0-*C0t-HrH,-lrH(M-*'*0000C0
CO(Mt^OO'H^OO'i<'*(M(M(M05CO(NOaJ05I>-t~000000?0«Ot-
O5cooot^i^oocoooooaoooooooi>.i>-oooot^t^t~i>.t~i^j>.t^t^t^
•nimmx'Bni
pa^Binoiuo
ii
0(NasojOi<»c»oo(jq(Noo<»T)<-^t;->pr;Tt<^oocoiniocoeoco(NoooicDoooo.io
O'*IM^r-(«0(N0>lO»O0000C0.-l051>.CDOOC-5)<-^05i-lT-(00t^t^Tj<-*t^t^O-^05t^X.^-^-alOOOiOOO'^CDCO-*'^^'*COOOCOCOiOu-;'N
coooiaicDt^i:^i>~iocoi-i.-ioooococcc^t^t^co^iOT-i,-i>-ii-(t^
t^to.iovoasoo>jr^ioi»':D05'i=50t^t^o
£
>
Ii
^COC0C0rHrH(Mi-IC0C0Cq(Mt0CO>-IO5rH'*Tt(t0C0rH'*^eOC0r-l
1
1-3
1
1
i
ooooooooooooooooooooooooooo
1
odd dd dd ddd ddd
Oh' 0^ e^' 0^ a; J J h4 Oh' iih' J J p.; a; 0^ h4 J Ph* a; J J 1-3 clh' eu J ,j J
}
1
6
>p ipo cq >p oi CO
i 03 (?q M 4t< CO -*
g|
i
%
C^ (N iM --1 O 05 CO
i
so
3
1
CO
en «3 «3 tn to C
c c c c c o 52
BBS B B ■ *l §
SSS ^ S 2 ^
RESEARCHES ON EXPLOSIVES 207
00■^•*cO'XlOlplpOrHOO^^-l^3'pr^a3co■Jt^■JHOc
QnooaiOii-HOOOOOJOST-li-li-lrHi-ll-HOjasOaiasOOOi-IOOl-IOSprHl-HC^rHi-HOi-Hi-HOOi-Hr-H
oot-hi— ic^OiOsfMCNC^-^Ocoooasaji
- oo oo oo CO < "' "" " " "" " " •- — '
■ 00<:DOC005ai rH.— I
»Oin?DtOl»5DCD^iCC^OOCQi»C305rHC50005KrtO^
I§-^SoiCDlO»OMCOOCOCRt^c»t^.-ia5t^.--10COOr-iOr-10a3050JrtiIOC^CO-^COOio0^2^
lO^-*CO(rOC^(Mi-H.-l.-l>-l l^OOCDCOrHi
<»cDooooocO(^5^^-50'aDT(<(MO»OCO(Ni— (^
,_lrHi-H.-IOi— li-li-lT-(OOOOOOi^i^rH>7l>;H7^'7l
_i-jmcncrilo\r^OOOO(MOilOOt^':OlOmOOOC— lvoCO(M^>0-*COCOiOOOOOOt^tO
SSSSSScqSmM§SS-*MLnSo>nc.«Dcoc30cgco>ovQi-rH2S^2g^^;2;
ioio»o>o>ovo»nioioo^-*^-*oooioi
0000000000000000000000000(NOO^ ^ (N ,-1 OS Oi t^ t^t^ CD-*!N!NrH05!D
208 RESEARCHES ON EXPLOSIVES
If the factors of effect be examined, it will be observed how, in
spite of the use of slow-burning and therefore uneconomical powders
in the large guns, the percentages realised gradually increase as we
pass from the smallest to the largest gun in our table — the highest
factor being 93 per cent, in the case of the 38-ton gun, the lowest
being 50"5 per cent, in the case of the little Abyssinian gun.
This difference in effect is of course in some measure due to the
communication of heat to the bore of the gun, to which we have so
frequently referred.
{y) DETERMINATION OF TOTAL THEORETIC WORK OF POWDER
WHEN INDEFINITELY EXPANDED.
To determine the total work which powder is capable of perform-
ing if allowed to expand indefinitely, the integral in Equation (33)
must be taken between oo and v^. If so taken, we have
Total work =2»(i^?ifff±^). . . (35)
= 332,128
gramme-metres per gramme of powder (486 foot-tons per lb. of powder).
Bunsen and Schischkoff's estimate of the work which powder is
capable of performing on a projectile, if indefinitely expanded, we
have already given ; but their estimate (being only the fifth part of
that at which we have arrived) is altogether erroneous, as these
eminent chemists appear to have overlooked the important part which
the non-gaseous portion of the charge plays in expansion.
It is interesting to compare the above work of gunpowder with
the total theoretic work of 1 grm. of coal, which is about 3,400,000
grm. -units. The work stored up in 1 grm. of coal is therefore more
than ten times as great as that stored up in 1 grm. of powder.
The powder, it is true, contains all the oxygen necessary for its
own combustion, while the coal draws nearly 3 grms. of oxygen from
the air. Even allowing, however, for this, there is a considerable
inferiority in the work done by gunpowder, which is doubtless in
part due to the fact that the coal finds its oxygen already in the
form of gas, while a considerable amount of work is expended by the
gunpowder in placing its oxygen in a similar condition.
In an economic point of view also the oxygen stored up in the
gunpowder is of no importance, as that consumed by coal costs noth-
ing, while the oxygen in the powder is in a most expensive form.
The fact is perhaps worth noting as demonstrating the impractica-
RESEARCHES ON EXPLOSIVES 209
bility of making economic engines deriving their motive power from
the force of gunpowder.
(z) SUMMAKY OF EESULTS.
It only now remains to summarise the principal results at which
we have arrived in the course of our researches ; (a) when gunpowder
is fired in a space entirely confined ; (b) when it is suffered to expand
in the bore of a gun.
(a) The results when powder is fired in a close space are as follow,
and for convenience are computed upon 1 grm. of powder occupying
a volume of 1 c.c. : —
1. On explosion, the products of combustion consist of about 57
per cent, by weight of matter, which ultimately assumes the solid
form, and 43 per cent, by weight of permanent gases.
2. At the moment of explosion, the fluid products of combustion,
doubtless in a very finely divided state, occupy a volume of about
•6 c.c.
3. At the same instant the permanent gases occupy a volume of
4 c.c, so that both the fluid and gaseous matter are of approximately
the same specific gravity.
4. The permanent gases generated by the explosion of a gramme
of powder are such that, at 0" Cent, and 760 mm. barometric pressure,
they occupy about 280 c.c, and therefore about 280 times the volume
of the original powder.
5. The chemical constituents of the solid products are exhibited
in Tables 3 and 6.
6. The composition of the permanent gases is shown in the same
tables.
7. The tension of the products of combustion, when the powder
fills entirely the space in which it is fired, is about 6400 atmospheres,
or about 42 tons per square inch.
8. The tension varies with the mean density of the products of
combustion according to the law given in Equation (3).
9. About 705 grm.-units of heat are developed by the decomposi-
tion of 1 grm. of powder such as we have used in our experiments.
10. The temperature of explosion is about 2200° Cent, (about
4000° Fahr.).
(b) When powder is fired in the bore of a gun, the results at
which we have arrived are as follows: —
1. The products of explosion, at all events as far as regards the
210 RESEARCHES ON EXPLOSIVES
proportions of the solid and gaseous products, are the same as in the
case of powder fired in a close vessel.
2. The work on the projectile is effected by the elastic force due
to the permanent gases.
3. The reduction of temperature due to the expansion of the
permanent gases is in a great measure compensated by the heat
stored up in the liquid residue.
4. The law connecting the tension of the products of explosion
with the volume they occupy is stated in Equation (30).
5. The work that gunpowder is capable of performing in expand-
ing in a vessel impervious to heat is given by Equation (34), and the
temperature during expansion is given in Equation (31).
6. The total theoretic work of gunpowder when indefinitely ex-
panded is about 332,000 grm.-metres per gramme of powder, or 486
foot-tons per lb. of powder.
With regard to one or two other points to which we specially
directed our attention in these investigations, we consider that our
results warrant us in stating that : —
1. Very small-grain powder, such as F. Gr. and E. F. G-., furnish
decidedly smaller proportions of gaseous products than a large-grain
powder (E. L. Gr.), while the latter again furnishes somewhat smaller
proportions than a still larger powder (pebble), though the difference
between the gaseous products of these two powders is comparatively
inconsiderable.
2. The variations in the composition of the products of explosion
furnished in close chambers by one and the same powder under
different conditions as regards pressure, and by two powders of
similar composition under the same conditions as regards pressure,
are so considerable that no value whatever can be attached to any
attempt to give a general chemical expression to the metamorphosis
of a gunpowder of normal composition.
3. The proportions in which the several constituents of solid
powder-residue are formed, are quite as much affected by slight acci-
dental variations in the conditions which attend the explosion of
one and the same powder in different experiments as by decided
differences in the composition as well as in the size of grain of
different powders.
4. In all but very exceptional results the solid residue fur-
nished by the explosion of gunpowder contains, as important
constituents, potassium carbonate, sulphate, hyposulphite, and sul-
phide, the proportion of carbonate being very much higher.
RESEARCHES ON EXPLOSIVES 211
and thai of sulphate very much lower, than stated by recent
investigators.
Abstract of Experiments.
In this abstract the following abbreviations are used : —
S to represent the mean density of the products of explosion ; A
the area of the piston of the crusher-gauge ; a the sectional area of
the crushing-cylinder.
Experiment 1, April 20, 1871.— The cylinder (Fig. 2, Plate X.,
p. 230) having been prepared for the experiments, was cahbrated and
found to contain 14,000 grs. (907-20 grms.). A charge of 1400 grs.
(90*72 grms.) E. L. G. powder was then placed in the cylinder and fired.
The gaseous products of combustion were collected in tubes and
sealed.
On opening the cylinder the solid products of combustion were
found adhering to the sides pretty uniformly, but thicker at the
bottom ; they had to be scraped off for collection.
Crush, copper
Pressure per
cylinder.
square inch.
•009
1-6 ton.
•0940 •1667 -0417
Experiment 2, April 4, 1871.— Fired 3500 grs. (226-80 grms.)
K. L. G. powder as above, in a similar cylinder, the powder exactly
filling the space in which it was confined.
The gas was retained in the cylinder for about a second, and
then, owing to a want of accurate fit in the coUecting-screw, made
its escape with a considerable explosion, completely, so to speak,
washing away every trace both of the male and female screw along
the channel it cut out for itself.
On opening the cylinder but little solid residue was found, and
that uniformly distributed over the surface, and about '07 inch thick.
Its colour was of a very bright vermiHon red, rapidly changing
to black on the surface, and was similar in all respects to the deposit
so often seen in the powder-chambers of heavy guns.
Eesidue collected and sealed up in a test-tube.
Crush, copper Pressure per
0. ^- '^'- cylinder. square inch.
•91.5 ^1667 -083.3 -293 34^5 tons.
Experiment 3, April 29, 1871.— Cylinder No. 6 cahbrated and
found to contain 14,702 grs. (952-68 grms.). 2940 grs. R. L. G.
(190-54 grms.) were fired and the gases collected within fifteen
minutes after firing.
212 RESEARCHES ON EXPLOSIVES
On opening the cylinder the solid products were found to be
collected at the bottom, only a very thin light -coloured deposit being
on the sides.
The appearance of the deposit was very different from any yet
obtained, being grey on the smooth surface and very bright yellow
in fracture. It was exceedingly hard and very deliquescent.
The interior surface of the cylinder appeared quite bright when
the deposit was removed.
A portion of the deposit, whitish on the surface, dark grey next
the cylinder, was collected and sealed in separate test-tubes.
A tin cylinder was substituted for copper, to measure the crush
in this experiment.
Crush, tin Pressure per
"• • "■ cylinder. square inch.
•1973 -1667 -0833 -165 2-67 tons.
Experiment 4, May 10, 1871.— 4411 grs. (285-5 grms.) of E. L. G.
powder were fired in cylinder No. 7. Gases were collected, com-
mencing seven minutes after explosion.
On opening the cylinder the solid products were found in a mass
at the bottom ; and the sides of the cylinder were also as noted in
the last experiment.
The residue, however, was of intense hardness, and the difficulty
of removing it was very great. Hardly any could be got off in
lumps, but it flew off like sand before the chisel.
Copper firing-wire fused off and dropped in the form of a button.
Crush, copper Pressure per
0. ^' '■'' cylinder. square inch.
•2963 -1677 ^0833 ^033 6^4 tons.
Experiment 5, June 22, 1871. — Cylinder No. 6 calibrated and
found to contain 15,859 grs. P. powder. Eired 1586 grs. (102-77
grms.) P. ; but, owing to the low pressure, the cylinder did not
become closed up very tightly, and most of the gas slowly
Solid products at the bottom, and easily removed. Colour light
grey on surface, dark grey next steel, shading into light grey near
the surface.
Crush, tin Pressure per
"• ^- "■ cylinder. square inch.
•1064 •1667 •0833 ^042 h39 ton.
Experiment 6, June 28, 1871.— -Fired 1586 grs. (102-77 grms.)
pebble in same cylinder (No. 6) as that used in the last experiment.
Crush, tin
Pressure
cylinder.
in tons.
•032
1-26
RESEARCHES ON EXPLOSIVES 213
Nearly all the gas escaped from the same cause (defect of pressure).
Products of combustion not collected.
5. A. a.
•1064 •1667 -0833
Experiment 7, June 28, 1871.— Fired 3150 grs. (20412 grms.)
pebble-powder in cylinder No. 6. Gas collected immediately. Solid
products at bottom as usual, and tolerably easily detached. Colour
whitish grey on the smooth surface, almost black next steel.
Fracture yellowish green with splotches of grey.
s \ a Crusl), tin Pressure
"• • ■ cylinder. in tons.
•21U -1667 •0833 -188 2-93
Experiment 8, June 29, 1871.— Fired 1586 grs. (102-77 grms.)
pebble-powder in cylinder No. 6. There was a slight escape of
gas at first, but the plug soon tightened. Gas collected and sealed
immediately.
On opening the cylinder, the deposit was found principally at the
bottom. It adhered very firmly, and was removed with great difficulty.
The colour of the smooth surface was light grey and green, buff
in one or two places. Fracture yellowish green.
The portions of the residue that could not be removed with a
chisel were dissolved out.
The firing copper wires '07 in diameter were melted and had
formed a button, having, however, rather long stumps.
^- A. a. Crush, tin. Pressure in tons.
•1064 ^1667 -0833 -033 1-28
Experiment 9, June 29, 1871.— Fired 4725 grs. (30618 grms.)
pebble in cylinder No. 4.
On firing there was a slight escape of gas past the crusher-gauge.
The gases were collected within five minutes of the explosion ;
and after the tubes were sealed a rough measurement was made of
the remaining quantity of
gas, which amounted to ^
59,000 c.c. t
The residue was very V-
easily detached from the ^•^^;.S^,=^=as£ ^ ,„- ^^J:--,
cylinder. It was darker
grey on the surface than in the last experiment. The fracture was
a deep olive-green with a stratum of light grey in the middle, thus
(see figure).
214 RESEARCHES ON EXPLOSIVES
The deposit was all on the bottom, excepting a very thin coating
on the sides. Firing-wires fused level with the plug.
Crush, copper Pressure
8. A- "• cylinder. in tons.
-.3171 -1667 -0833 -018 4-90
Experiment 10, July 5, 1871.— Fired 6344 grs. (411-09 grms.)
P. powder in cylinder No. 6. Most of the gas escaped before enough
could be collected.
Eesidue was found, when the cylinder was opened at the bottom,
not in the usual hard compact mass, but much looser in texture. On
the surface there were three large spongy projections, presenting an
appearance as if the surface had been broken by the escape of
occluded gas, thus (see figure).
Colour of surface grey in parts, also light yellow shading into
dark yellow. Colour of fracture grey, shading off into dirty yellow
and occasionally into gamboge. Powerful odour of sulphuretted
hydrogen.
Crush, copper Pressure
5. ^- "• cylinder. in tons.
•4258 -1667 -0833 -054. 8-4
Experiment 11, July 5, 1871.— Fired 5881 grs. (381-09 grms.)
E. L. G. in cylinder No. 4. Some little escape of gas past crusher-
plug. Eesidue very hard and adhering strongly to the side ; a portion
obtained in solid lumps. Colour grey on surface, black next steel.
Fracture olive-green.
A good deal of the deposit was chiselled off in the form of fine
dust, and this, when it had lain for a minute or two, heated very
much, say to about 80° or 90° Cent., agglomerating into loose lumps
and changing from a light greenish-grey colour to a bright yellow.
A portion of this last deposit was collected in a separate bottle.
When the crusher-gauge was taken out, the plug at the end was
found to be broken right through transversely.
The fracture was perfectly clean and bright; it was therefore
RESEARCHES ON EXPLOSIVES 215
concluded that it
must
have
broken after the
great heat had
sub-
sided.
5.
•3947
A.
•1667
•0833
Crush, copper Pressure
cylinder. in tons.
•051 8^10
Crush, copper
Pressure
cylinder.
in tons.
•091
11-5
Experiment 12, July 8, 1871.— Fired 6344 grs. (411-09 grms.)
P. powder in cylinder No. 6. A good deal of leakage past the crusher-
plug. Gas collected. Eesidue very hard, but it split off tolerably
easily. The colour was grey throughout ; fracture much the colour
and appearance of slate. The difference in physical appearance
between this residue and that in the last experiment was very great,
the colour of the fine dust being grey, while in the last experiment
it was a light yellow.
- , Crush, copper Pressure
"• ■^- "• cylinder. in tons.
•4258 -1667 -0833 -063 9-1
Experiment 13, July 12, 1871.— Fired 7351 grs. (476-34 grms.)
E. L. G. in cylinder No. 6. The products cut away the screw of the
pressure-gauge and escaped.
8. A.
•4934 -1667 ^0833
Experiment 14, July 12, 1871.— Fired 7930 grs. (513-86 grms.)
P. in cylinder No. 4. Gas and residue collected as usual. Cylinder
tight.
5. A.
•5322 •1667 ^0833
Experiment 15, July 22, 1872.— Fired, in cylinder No. 6, 1586
grs. (102-77 grms.) of F. G. Cylinder perfectly tight. Gas and
residue collected.
. ^ Crush, copper Pressure
°- "' cylinder. in tons.
•1064 •1667 ^0467 -003 1-66
Experiment 16, July 22, 1872.— Experiment 15 repeated with
tin cylinder.
. Crush, Pressure
0- ^- "• tiu. in tons.
•1064 -1667 -0467 •US 1^25
Experiment 17, July 24, 1872.— Fired, in cylinder No. 6, 3172
grs. (205-55 grms.) F. G. Collected gas and residue. Residue very
Crush, copper
Pressure
cylinder.
in tons.
•100
12-2
216 RESEARCHES ON EXPLOSIVES
hard, but not so dark in colour as that in Experiment No. 16.
Surface dark grey, but of a lighter colour when fractured. A very
thin coating on the sides of the cylinder.
Small bright yellow crystals pretty uniformly distributed through
the residue.
■ush, copper
Pressure
cylinder.
in tons.
•0475
3-70
•2129 -1667 ^0417
Second experiment.
•2129 -1667 ^0417 •0435 3-58
Experiment 18.— Fired 4758 grs. (308-32 grms.) F. G. in cylinder
No. 6. Cylinder perfectly tight. Collected gas and residue.
On opening the cylinder the residue was found all collected at
the bottom ; and it had evidently run down the sides in a very fluid
state, the deposit on the side being very thin. Colour on surface
dark grey. Fracture more uniform than usual, there being no patches
of yellow and but few of a lighter colour.
^ . Crush, copper Pressure
"• ■ ■ cylinder. in tons.
•3193 -1667 -0467 ^132 rv75
Experiment 19, August 26, 1872.— Fired, in cylinder No. 6, 6344
grs. (411 "09 grms.) F. G. Cylinder perfectly tight. Colour and
fracture dark grey, nearly black ; but in places both surface and
fracture light grey. No appearance of yellow anywhere in this
deposit. All the residues, so far, of F. G. differ very considerably in
appearance both from pebble and E. L. G.
The deposit on the sides was exceedingly thin, not more than
•01 inch in thickness.
. . Crush, copper Pressure
"• ■ ■ cylinder. in tons.
•4258 -1667 -0417 ^222 9-98
(This pressure rejected.)
Experiment 20, August 28, 1872.— Fired, in cyhnder No. 6, 7930
grs. (513^86 grms.) F. G. Cylinder was absolutely tight. Gas
collected in the usual manner. On opening the cylinder and remov-
ing the firing-plug, observed that the little button of residue adher-
ing to the firing-plug, when cut into, had a large well-defined
crystalline structure, the crystals being transparent although the
surface of the button was dark grey. Sealed a portion in a tube for
examination.
Crush, copper
Pressure
cylinder.
in tons.
•145
15-8
RESEARCHES ON EXPLOSIVES 217
Eesidue in mass at bottom of cylinder as usual ; next to nothing
on sides. Colour and fracture much the same as in the last experi-
ment, but the centre much lighter grey.
5. A. a.
•5322 •1667 •0834
(This pressure rejected.)
Experiments 21 to 24. — These experiments discarded.
N.B. — From Experiment 16 inclusive, the crusher-gauge was put
loose in the charge of powder to be fired ; but it having been found
that the crusher-gauge was heated to such an extent as to soften the
copper cylinder and thereby affect the observations, these experiments
were repeated, as far as regards the pressure determinations, in
Experiments 25 to 32.
Experiment 25, October 1, 1872.— Fired 2974 grs. (192-72 grms.)
F. G. in cylinder No. 7.
. Crush, copper Pressure
"• ' ■ cylinder. in tons.
•3860 -0834 -0417 •OSl 7^68
Experiment 26, October 17, 1872.— Fired 1586 grs. (10277 grms.)
F. G. in cylinder No. 6.
S. A.
•1064 -0834 -0417
Experiment 27, October 18, 1872.— Fired 3172 grs. (205-55 grms.)
F. G. in cylinder No. 6.
5. A. a.
•2129 -0834 -0417
Experiment 28, October 18, 1872.— Fired 4758 grs. (308-32 grms.)
F. G. in cylinder No. 6.
Crush, tin
Pressure
cylinder.
in tons.
•016
0^96
Crush, copper
Pressure
cylinder.
in tons.
•008
3-0
Crush, copper
Pressure
cylinder.
in tons.
•032
6^32
•3193 -0834 -0417
Experiment 29, October 19, 1872.— Fired 6344 grs. (411-09 grms.)
F. G. in cylinder No. 6.
5. A. a.
•4258 ^0834 ^041 7
Crush, copper
Pressure
cylinder.
in tons.
•074
9^34
218 RESEARCHES ON EXPLOSIVES
Experiment 30, October 21, 1872.— Fired 7930 grs. (513-86 grms.)
F. G. in cylinder No. 6.
Crush, copper
cylin
y
/;
.^
'^,'
■■
y
^
<:
'■'
-^
f»^
-1
'
/
^^
y-
^
i>
s
ia3
HdS
ow.
V
H,
;
an;
S3*
d
'i
I
J
1
1
1
1
1
§
3
1
I
J
1
J
J
1
1
J
§
§
1
=j
PLATE XIII.
T—
T
r
n
s
R
j
il
1
1
n
3a
s
a:
1
T
OJ.
; s
5
3
n
an
;s3
r
r
r
n
1
li
J
II
z
o
o
^
1
■
; ~
a
,
'
^
I
K)
1
1
i
o
«>
■t
<^
Cy
/;
-^
'■f
s
•^
i'l
1
'^
=
/
/;
i
z
V
^
o
4
^
■■'
V
T^
fj
7
<
uJ
Ss
/,
/
5
1
,
//
JU
1
1
/
^
Z;-'^
^
e^
c
§
J
J
M
§
p
1
§
J
J
J
1
1
J
§
«)
J
1'
§
PLATE XIV.
PLATE XV.
\t
i 1
ri
i i 1 g S i ^ '-nf
'fJO 3-izzntM
rNOoas asd ±33i — axiooi3a
,1 1
■
1 111
' ;
^ b
q:
\\
^ '
E
•
^S
\
,
■g 1
\\
^ ^
Y
1 "^
s
\
a.
\
1 ^
\
',
f^ s
\
^ f^
',
~ 1 /
\ \
e j^
\
\ \
^ 1 ^
\\
^lll /
"* ^ „ «^ /
\
1 II
i
1
\
^ \ -^ %
\
5. X ^ V
y
/
o
u
f ^ » »- /
a
t
Ss i /
\
H 1 /
/
^•■
III /
^
\
/
s|
\
,/
i
UIUUUI
§
1
\s3y3Hd90W±V NI
1
"1
§
i
1
s
1
^ «
ss
3
ob
t
ii
ti
i
ji
■^
«. ,V
^
en
^
i
i
i
i
ei ^ "
^
s las
li s
^
aar
4s
lad
'i
V
/
/'
•■'
i
1
i
^
/
X
X
s
t
1
-
-^
-^
—
-
■^
\
S
,
s
y
ir:
=i
■JT
-
—
;
b
;;:■
~~
■^
rj
-
—
'Z
=.
=
ii
^
:^
:s
PLATE XVI.
PLATE XVII.
I'LATE XVIII
?
I
?"
^
T
b ^
T
H3
.,
yt
s
J3c
p
7
NO
L
Nl
3J
ns
S3i
r
]
1
-T=
m
fjj
< ^
''1
?
«^
^ ^
I;/
1
^5
'.U
i'i
■S
1
.^
1
1
^
s; i
i MD
1
3
m
•7
1
^
^
/
1
S
i
1
1
;
/
1
1
o
^
^
/
>■
p
•y
f
1
t
^
M
/
■/
'4
"N
.
,
/
7
A
^§
/
'/
a
1
i
^
1
1
/
/
/
/
2
<
Ul
1
^
//
/
/
/
<>
CV Of
5 ■*
.^
/
."
w ?
. ^
/.'
V/
b
G
<^.
s^
^
<^
^
^'
^
^
^
;>^
,.;= ^
"
^
>'
jj
- ,
"
_^
, _^
— ■■
_,
^
-■;;;; --
'■
_..
'_^
-^
33 i
3H
JSC
Wl
/
Nl
3M(
s«
3h
rt
.
^
PI
P,
99'
Q
^ n
P
p
^
« ^
!?;
o
6
S
^
ci
J^
s
^^
s
r^:
*
9
P
f^
y
H
}=;
!.
OC.
■^
. H «
_j
M
N
K
i
-J
-
-^
1 —
PLATE XIX.
"■"
i=i
"^
n^
~T
r
^
n
-r
"^
^~^
T
"^
r t
^^
F
^
"^
'
ijpil
ONI
■M
s
I3d
s
101
Mr
"
lUd
,
!
1
1
i
j
i
j
'\
;l
1
1
i
i
?>
•e
\
1
^
g
j^
?
^
^
g
^'
1
1
fe
'1'
IsS
i^
■^
§
1
«>
1"
•5s
1
1
8
i
\l
1 "'
1
1
s?
\
i
is
^
1
1
^
!
f
1
1
1
\
?
rl
1
^
1
1
1
i
/
,' 1
/
/
2
s
«
10
•)
r>
!
^
.§
-^
^^
'
M.
fi
"rJ
*J
/
(
/
f!^
f>
fj;
Si
/
/
*
-
c
«•
*
/
/
/
i
./
' /
/
**•
/
/
'
'/
/
Ik.
^'
'/
''■
,'
^
/
^ ■
.''
^
09
=
=:
=
zz:
:=
=
=
=
=
=
—
Z^
^r:
zIZ
1^
^
^
E::
—
r^
=
rr
=^
OT-
=
is
^
m
^
2=
=
=Sg
=
=
=
==
=
^
^
—
-^
sSZ
M 1 1 i 1 1 i 1 1 1 1 1 1 1 1 1 1 1 i i i i i 1 i 1 i 1 i
saasHdsowxv ni aanssiad
PLATE XX.
=n=
r
=i=
r
H'
Nl
ub
5
biac
T
NO
aa
rr
1S«
3«
i
r
r
j
i
1
i
i;
z
I
1
^i
9
1-
ti
!
if
3
s
^
ij(
f^
•^
o
■f
■1
*
i
!
V.
'fe
^
^
\
I
U)
1
<6
■a
\
\
i
13
o
1^
is
^
1
1!
li.
o
^
1
k
;io
^
ie
'•'i
1
j
■~§
^
1
(0
1
^
^
//
u
Q
s
V
,
',■
1
1
1
1
/
t
ll
UJ
5
V
/
:J
1
"3
^
^
II,
1
1
/
/'
1
tsi
Ki
K
t-«
\' '
''/
(.1.
<
CO
o
Q
/■
7
/
/
V
/
I>fr.
/
/
s*.
.^'
--;•■
/
^,
.-^
..-
^-
,
00
^
•a^BnoqiBoinbSGS
O t^ «D
v" r' o
CD
CO
CT>
Oi
cN-^ :
(N lO •
r-ICO p
'^l
^
?^
T'
1
amissT!^o
00
1
uinissB:}opc
o
(M oo T-H
(M CO .-(
CO -^ CO
CO
CO
'-'
CO
lO
1-1 OO
CO Tj<
05 O
CO
CO
CO
CO
a
•sbS-iisjbk
co=> Oi
O -* CO
o
__,
o
CD
CO
a
•nsSoi:)!^
r-ll^ CO
ppCN
CO CD CO
lO
°p
o
8
^
CO
CO
1-1
005t-
»o
a
l^Oop
-^pp
»o
00
"?
lO
t-
§
oiuoqj'BO
CO CD O
cot- 00
O CDt^
■*
00
■^
t-
CO
^
r-l
•apijpjfqtiB
0 05
CO ^ CD
a>
o
^
CO
»o
copcq
pw^.
CO
cp
?'
oiuoqjuo
OO i-l ■*
iO
lO
lO
CO
■ "S -M
• 4J •
"S-*^
.
.
.
S-S ^
S-c g
s^ ae
s'jonpojd
sl^l
s^
J
;
J
^
JO
X^isnep WBSM
^K^
o
o
o
o
o
" ,^ '
" , '
<
<
d
6 ■ be •
^ ..s s
g-
:§
HJ
o ^
Oi-
^0=^
l>CO .
Ph
Pi
fo ■'^
I-l
05
236 RESEARCHES ON EXPLOSIVES
Table 2.* — Composition hy weight of the products of combustion of 1 (jramme
-S
Nature of
i
Proportions by weight of gaseous products.
.2
p,
powder.
.2i
.2 .
j3
u
1
a
•
Bi
1
if
II
5°
z
1
i
1
>>
o
■7 3
II
8
Pebble, W. A. .
10
•2553
•0514
•1140
•0133
•0007
•308
7
20
•2494
•0570
•1108
•0182
•0009
::: i
•318
9
30
•2595
•0545
•1113
•0124
•0007
...
•328
12
40
•2624
•0471
•1085
•0069
•0006
•0006
•311
14
50
•2743
•0469
•1128
•0083
•0012
•0005
•306
37
60
•2654
•0470
■1087
•0093
•0011
•0005
•321
38
70
•2604
•0415
•1065
•0128
•0007
■0005
•285
43
80
•2690
•0396
•1084
•0080
•0006
•0004
•332
77
90
Means
•2684
•0359
•1080
•0085
•0013
•0005
•364
•2627
•0468
•1099
•0109
•0006
•0006
...
•319
Highest
•2743
•0570
•1140
•0182
•0013
•0009
•364
1
R. L. G.,W. A.
Lowest
10
•2494
•0359
•1065
•0069
•0004
•285
•2569
•0300
•1188
•0164
•0006
•0005
•297
3
20
•2477
•0389
•1189
•0148
•0001
•0006
•0022
•309
4
30
•2582
•0386
•1096
•0126
•0007
...
•298
11
40
•2595
•0356
•1125
•0077
•0005
•0006
•278
70
50
•2494
•0558
•1011
•0065
•0016
•0007
•348
39
60
•2648
•0467
•1065
•0066
•0007
•0005
•359
96
60
•2457
■0490
•1090
•0176
•0007
•0005
•370
41
70
•2576
•0441
•1053
•0114
•0011
•0004
•343
44
80
•2672
•0401
•1060
•0062
•0014
•0004
...
•377
68
90
Means
•2720
•0352
•1074
•0076
•0015
•0003
•371
•2580
•0414
•1095
•0106
•0008
•0005
•0002
•335
Highest
•2720
•0558
•1189
•0176
•0016
•0007
•0022
•377
16
F. G.,W. A. .
Lowest
10
•2457
•0300
•1011
•0062
•0003
•278
•2423
•0561
•1122
•0095
•0004
•0010
•0006
•277
17
20
•2475
•0410
•1074
•0153
•0010
•34C
18
30
•2586
•0370
•1050
•0088
•0008
•0010
•264
19
40
•2689
•0334
•1055
•0079
•0008
•257
75
50
•2611
•0338
•1080
•0087
•0005
•0007
•32C
40
60
•2651
•0312
•1080
•0089
•0003
•0006
•23£
42
70
•2678
•0253
•1100
•0080
•0009
•0005
•0006
•246
47
80
•2598
•0265
•1105
•0101
•0008
•0005
•0007
•25]
69
90
Means
•2698
•0247
•1088
•0116
•0003
•0005
•28J
•2596
•0343
•1084
•0099
•0003
•0007
•0003
•27(
Highest
•2698
•0561
•1122
•0153
•0009
•0010
•0010
•34(
78
R. F. G.,W. A.
Lowest
70
•2423
•0247
•0285
•1050
•0079
•0005
•23<
•2652
•1110
•0063
•0002
•0006
•34^
79
Spanish spherical
70
•2424
•0133
•1091
•0096
•0003
•0007
•21
196
Curtis and Har-
vey's No. 6
30
•2576
•0245
•1124
•0082
•0046
•0008
•33
194
Mining powder .
30
•2254
•1508
•0849.
•0385
•0070
•0017
•19
See Tabic 13.
RESEARCHES ON EXPLOSIVES 23T
fired (funpowder of the undermentioned natures, and of various gravimetric densities.
Proportions
by weight of the solid residue.
°i
=5
Si
1
1 ft
§1
"h bo
11
If
11
1
p ■
It
11
ij
"^1
s .
■3 5
II
"o =t
o °
11
< ^
1
6
•0835
•1152
•0412
•0005
•0027
•0009
•0034
•0095
•4347
•5558
•0095
•0760
•0206
•1001
•0003
•0005
■0381
•0095
•4363
•5542
•0095
•0696
•0239
•0911
•0012
•0002
•0007
•0372
•0095
•4384
•5521
■0095
•0745
•0794
•0548
•0014
•0005
•0077
•0004
•0342
•0095
•4261
•5644
•0095
•0652
•0335
•1045
•0013
•0011
•0003
•0337
•0095
•4440
•5565
•0095
•0752
•0556
•0641
•0019
•0017
•0003
•0384
•0095
•4320
•5585
•0095
•0726
•1827
•0127
•0022
•0014
•0003
•0110
•0095
•4224
•5681
•0095
•0584
•1167
•0220
•0026
•0018
•0005
•0303
•0095
•4260
•5645
•0095
•0518
•0754
•0218
•0032
•0025
•0007
•0480
•0095
•4226
•5679
•0095
•0696
•0781
•0569
•0016
•0013
•0009
•0005
•0306
•0095
•4314
•5601
•0095
•0835
•1827
•1045
■0032
•0027
■0077
•0009
■0480
■0095
•4440
•5681
•0095
•0518
•0206
•0127
•0003
■0002
•0003
■0034
■0095
•4224
•5521
•0095
•1160
•1154
•0228
•0032
■0003
■0041
•0072
•0106
•4232
•5662
■0106
•1364
•0326
•0541
•0003
•0007
•0004
•0323
•0001
■0106
•4232
•5662
■0106
•1380
•0732
•0334
•0003
•0002
•0002
•0260
•0106
•4197
•5697
■0106
•1310
•1379
•0116
■0009
•0007
•0002
•0118
•0106
•4164
•5730
•0106
•0266
•1455
•0204
•0017
•0029
■0006
■0284
•0106
•4151
•5743
•0106
•0619
•0365
•0559
•0015
■0006
•0475
•0106
•4258
•5636
■0106
•0608
•0392
•0434
•0016
•0026
•0491
•0106
•4225
•5669
•0106
•0600
•1059
•0219
•0028
•0024
■0001
•0329
•0106
•4199
•5695
•0106
•0501
•0175
•0514
•0014
•0010
•0006
•0684
•0106
•4213
•5681
•0106
•0482
•0486
•0409
•0021
•0011
•0009
•0521
•0106
•4240
•5654
•0106
•0829
•0752
•0356
•0013
•0015
•0004
•0353
•0007
•0106
•4211
•5683
•0106
•1380
•1455
•0559
•0028
•0032
•0009
•0684
•0072
•0106
•4-258
•5743
•0106
•0266
•0175
•0116
•0041
•0106
•4151
•5636
•0106
•1005
•1338
•0193
•0011
•0308
•0004
•0148
•4221
•5631
•0148
•1388
•0304
■0294
•0001
•0005
•0009
•0328
•0148
•4122
•5730
•0148
•1302
•1583
•0161
•0004
•0006
•0001
•0038
•0148
■4112
•5740
•0148
•1250
•1640
•0193
■0004
•0005
■0002
•0067
•0148
•4115
•5737
•0148
•1186
•0768
•0248
•0004
•0005
•0005
•0301
■0148
•4128
•5724
•0148
•1269
•18-22
•0009
•0010
•0183
•0001
•0026
■0148
•4141
•5711
•0148
•1202
•1836
•0013
•0011
•0170
•0002
•0026
•4131
•5721
•0148
•1218
•1995
•0014
•0015
•0002
•0005
•0148
•4089
•5763
•0148
•1046
•1474
•0152
•0014
•0015
•0002
•0112
•4157
•5695
•0148
•1207
•1418
•0138
•0007
•0009
•0074
•0003
•0100
•0148
•4135
•5717
•0148
•1388
•1995
•0294
•0014
•0015
•0308
•0009
•0328
•0148
•4221
•5763
•0148
•1005
•0304
•0005
•0001
•4089
■5631
•0148
•1268
•0472
•0245
•0002
•0004
•0005
•0386
•0080
•4118
•5802
•0080
•2943
•0470
•0196
•0003
•0058
•0002
•0348
•0065
•3754
•6181
•0065
•1243
•0228
•0582
•0017
•0005
•0332
•0117
•4081
•5802
•0117
•0028
•0277
•1578
•0138
•0004
•0084
•0614
•0095
•0161
•5083
•4756
•0161
1238 RESEARCHES ON EXPLOSIVES
regards as satisfactorily explaining (" and definitely reducing to five
simple reactions ") the formation of carbon dioxide, carbon monoxide,
potassium sulphate, sulphide and carbonate, from a powder of what
we call normal composition.*
After giving further equations which apply to the extreme results
{in regard to the chief products only) assumed to be obtainable from
the introduction, on the one hand, of excess of saltpetre, on the other
of excess of charcoal, into the composition of powder, M. Berthelot
passes to what he terms the accessory products and, excluding from
these potassium hyposulphite, which he deals with separately, he
first gives two equations to account for the production of sulphocyanide;
then two more to explain the existence of ammonium sesquicarbonate
(which he believes to be formed by the action of water-vapour on
potassium cyanide). The existence of sulphuretted or free hydrogen
are explained by two more equations, and marsh-gas is assumed to
result from " the pyrogenous decomposition of the charcoal in the
powder." Lastly, an equation is given to account for the possible
formation of traces of hyposulphite, which Berthelot however regards
* The five simple reactions in question are thus explained : —
1.
NOfiK + S + C,
= K.S. + 3C0o + N
2.
NOeK + S + C3
= K.C.O3 + CO. + CO0 + N + S
3.
NOfiK + S + Cs
= KCOj + UC.O.o + N + S + iC
4.
NOfiK + S + C,
= KSO4 + 2CO. + N + C
5.
NOeK + S + Cs
= KSO4 + CO.0 + N + C2
When sulphate is formed in such small quantities that it may be neglected, the
simultaneous reactions supposed to occur are 1, 2, and 3, by quantities of the
powder proportionate to the numbers J, A, and J ; but when the sulphate amounts
to 12 or 14 per cent, the simultaneous reactions supposed to occur are Nos. 1, 3, 4,
and 5, with quantities of powder corresponding to the numbers i, about h, g, and yV
As there is only one single instance out of twenty-nine analyses of powder-residues in
which the sulphate was found to amount to as little as 4-6 per cent, of the sohd
products (the next lowest proportion being nearly double that amount), it can
scarcely be assumed that M. Berthelot's first arrangement of reactions can represent
any but a most exceptional result. Again, the acceptance of his arrangement of
four equations in the proportions he indicates as accounting for the formation of
the chief products when the sulphate amounts to 12 or 14 per cent, of the total con-
stituents, involves the assumption that a somewhat considerable proportion of
charcoal should remain unoxidised ; in fact, nearly 2-5 per cent, of carbon should be
found in the residue. The detection and determination of such a constituent of
powder-residue does not involve any difficulty, yet there were only three instances
out of eighteen residues (in which the sulphate was considerable in amount) where
the charcoal was present in estimable quantities ; in two of these it was below 1 per
cent, in the other it was only 0-01 per cent In a few other residues only traces of
charcoal were discovered ; the larger number contained none.
These points are referred to in illustration of how imperfectly M. Berthelot's not
very simple arrangement of theoretical reactions correspond to the results actually
obtained, even so far only as the chief products are concerned.
RESEARCHES ON EXPLOSIVES 239
entirely as a product formed during the collection and analytical
treatment of the solid residue, but which we n^ertheless beUeve we
shall conclusively prove * to be formed in very notable quantities
before the solid residue can have undergone alteration from external
causes.
It will be seen from the foregoing outline of M. Berthelot's
theoretical explanation of the chemical changes involved in the meta-
morphosis of gunpowder, that the simplest form of expression which
he can give to the formation of the products of explosion consists in
the incorporation of nine or ten distinct reactions occurring simultane-
ously, but in very variable proportions, which have to be supplemented
by three or four other chemical equations, by which the formation,
during the process of cooling, of certain products believed to be
secondary, is explained. Now, although such speculations as the
above are unquestionably interesting, and, it may be added, of a
nature which must occur to those who desire to give some kind of
definite explanation, for purposes of elementary instruction, of the
chemical changes involved in the explosion of powder, we fail to see
that beyond this they do more than afford the strongest confirmation
of the correctness of our conclusion, that " no value whatever can be
attached to any attempt to give a general chemical expression to the
metamorphosis of a gunpowder of normal composition."
With regard to the potassium hyposulphite which is included in
our statement of the composition of the soHd products of explosion,
we have to submit the following considerations.
In the analytical results furnished by the soHd residues, as
detailed in our first memoir, the hyposulphite ranged in amount from
3 to 35 per cent. ; and on comparing the results of different analyses
it is observed that in most instances the proportion of monosulphide
was small when the hyposulphite was large in amount, and in a few
instances — all of them F. G-. powder-residues — in which the proportion
of the latter was very high, there was no sulphide at all.
Being fully alive to the possibility of the existence of potassium
polysulphide in the solid residue giving rise to the production of
some hyposulphite through the agency of atmospheric oxygen, great
precautions were taken, especially in the latter experiments, in
collecting and preserving the residue and in submitting it to treat-
ment for analysis, to guard against this possible source of error.
In the first place, it should be mentioned that the residue con-
sisted in nearly all cases of fused, very hard masses, collected at the
* See note at end of this memoir (p. 309),
240 RESEARCHES ON EXPLOSIVES
bottom of the explosion-vessel, the sides of which were, moreover,
generally covered with very thin films. The action of atmospheric
oxygen upon the fused solid could only be superficial, but would vary
in extent with the amount of surface of the residue exposed to the
air during removal from the explosion apparatus or subsequent
exposure. The latter was avoided as much as possible, as the
residues were transferred at once, as they were detached from the
surfaces of the explosion-vessel, into small bottles, in which they
were carefully sealed up. It was only in one or two instances that,
before opening the bottles, an odour of sulphuretted hydrogen,
distinctly perceptible at the sealed surfaces of the mouths, indicated
a slight imperfection in the sealing of the bottles.
The difficulties in the way of reducing to a minimum the ex-
posure to air of the residues during their detachment from the
explosion-vessel were, however, very much greater. We pointed out
in our first memoir that in almost all cases the residues were in the
form of exceedingly hard and compact masses, which had to be cut
out with steel chisels, and that although portions of the mass were
detached in the form of lumps, a considerable amount of it flew off
before the chisel in fine dust. The utmost care was taken to avoid
exposure of the detached residues to the air, but it was of course
impossible to avoid their being more or less attached by atmospheric
oxygen during the period of their collection. There is no doubt,
moreover, that the residues, which differed greatly from each other
in structure and in their tendency to absorb moisture and to become
heated upon exposure to air, were susceptible in very variable degree
to atmospheric oxidation. We, therefore, are quite prepared to
admit that, of the large amount of hyposulphite found in a number
of the analyses, a proportion, and in some instances possibly a large
one, may have been produced by the agency of atmospheric oxygen
during the removal of the residue from the apparatus; and the
results of some special experiments, which we shall presently quote,
appear to favour the conclusion that in those instances where no
sulphide was discovered, its absence may have been ascribable to
atmospheric oxidation. We regret having neglected to make any
reference to this probable source of error in describing the results
of our analyses, our belief being at the time that any important
alteration of the residue by atmospheric action was sufficiently
guarded against ; at the same time, it is right we should point out
that, in several instances in which the circumstances attending the
manipulation of the solid residue and its consequent mechanical
RESEARCHES ON EXPLOSIVES 241
condition were apparently most favourable to its accidental oxidation,
the proportion of hyposulphite formed was comparatively moderate
in amount.
On the other hand, we cannot concur in M. Berthelot's view that
the existence of hyposulphite among our analytical results is also
ascribable in part to accidental oxidation of potassium sulphide
during the analytical manipulations. These were carried out with
great uniformity so far as certain preliminary operations were con-
cerned, which consisted, firstly, in dissolving the residue in water
which had been carefully boiled to expel air, and secondly, in
filtering the solution in closed vessels — both of these being
rapidly completed operations. The receiving vessel contained pure
ignited cupric oxide, with which, as soon as the filtering opera-
tion was completed, the solution was agitated until it became
colourless.
The fact that in some of the analyses, all of which, we repeat,
were uniformly conducted in regard to the above points, from 3 to
10 per cent, only of hyposulphite were found, while the proportion
of monosulphide in these analyses ranged from 7 to 19 per cent,
(being above 9 per cent, in eight instances), appears to afford
substantial proof that accidental atmospheric oxidation during
the collection and analysis of the residues is not sufficient to
account for all but the very small quantities of hyposulphite which
M. Berthelot considers could only have pre-existed in the residue
examined by us. That chemist appears, moreover, to have overlooked
the following facts given by us in our first memoir : —
1. Separate examinations (conducted precisely alike) of the upper
and lower portions of some of the residues showed that considerably
larger proportions of hyposulphite existed in the upper portions.
In one case quoted by us in our first memoir, the upper portion con-
tained 1714 per cent, of hyposulphite, while the lower portion only
contained 4-34 per cent. At the same time there was only a
difference of 1-27 per cent, in the proportions of monosulphide
existing in the two portions of the residue (6-03 in the upper part,
and 7"3 in the lower), while there was a very great difference in the
amount of free sulphur (4-88 in the upper part, and 10-09 in the
lower).
2. One of the small buttons of the fired solid products, of which
there was generally one found attached to the firing plug in the
cylinder, was examined for sulphide and hyposulphite (it having
Q
242 RESEARCHES ON EXPLOSIVES
been detached without fracture, and at once sealed up in a small
tube). It contained the latter, but none of the former, while the
mass of the residue of this particular experiment contained a some-
what considerable proportion of sulphide.
3. The production of high proportions of hyposulphite was but
little affected by any variations in the circumstances attending the
several explosions (i.e., whether the spaces in which the powder was
exploded were great or small), excepting that the amount was high
in all three cases when the powder was exploded in the largest
space. On the other hand, a great reduction in the size of grain of
the gunpowder used appeared to have a great influence upon the
production of hyposulphite, as when passing from a very large-grain
powder (pebble or E. L. G.) to a fine grain-powder (F. G.).
Thus the production of hyposulphite exceeded 20 per cent.
3 experiments out of 9 with pebble-powder (Nos. 8, 38, 43). (pp. 211,
3 „ „ 10 „ R. L. G. „ (Nos. 1, 11, 70). [etc.)
„ a J, r KNos. 16, 18, 19, 40, 42,
7 „ ,, 9 „ 1^. (j. „ '^ 47,69).
It was below 10 per cent, in —
4 experiments out of 9 with pebble-powder (Nos. 7, 9, 11, 37).
5 „ „ 10 „ R. L. G. „ (Nos. 3, 39, 44, 68, 96).
1 „ „ 9 „ F. G. „ (No. 17).
There were no circumstances connected with the carrying out of
the explosions, or with the collection and analysis of the residues,
to which the above great differences between the results furnished
by fine-grain powder and by the two large grain powders could be
ascribed.
While, however, certain of the great variations in the proportions
of hyposulphite and sulphide, which cannot be accounted for by
variations of structure of the residue or of manipulations favourable
to oxidation by atmospheric agency, appear to us to demonstrate
that the hyposulphite is formed in the solid residue before the
explosion-vessel is opened, and indeed in such amount that it must
be regarded as an important product (whether it be a primary or a
secondary one), we have been anxious to obtain, if possible, some
more decisive evidence as to the probable proportions of hyposulphite
actually existing in the residues furnished by the explosion of
gunpowder in closed vessels. We therefore varied the method of
RESEARCHES ON EXPLOSIVES 243
collecting and preparing the residues for analysis, in the experiments
of which the following is an account : —
1. 5960 grains (386-2 grms.) of the K. L. G-. and pebble powders
used in these researches were fired in the large cylinder under a density
of 0-40.
Immediately on opening the cylinder in each case, the solid
products were as nearly as possible divided into two equal portions,
consisting of the top and the bottom. Each of these portions was
again divided roughly into two equal parts, one of which, in large
lumps, was, as rapidly as possible (being but for a few seconds exposed
to the air), sealed in dry bottles freed, or nearly so, from oxygen, the
other moieties being finely ground and freely exposed to the air for
48 hours.
The only point of difference calling for remark in the appearance
of the two residues was the difference in colour, the residue from the
pebble being decidedly the lighter in colour, both on the surface and
in fracture ; but there were material differences in the behaviour of
the ground portions of the two powder-residues.
"With both powders, the bottom ground portion heated very
decidedly more than the top ; but while, in the E. L. G-., this tendency
was exhibited in a remarkably low degree, with the pebble the
tendency to heat was, we think, abnormally high. In the latter case,
the ground deposit from the top began to heat immediately on being
placed upon paper. The deposit on the apex of the cone and in the
interior, where the heat was highest, changed rapidly in colour to a
light yellow, tinged with green.
The ground bottom part of the residue darkened considerably
during the development of heat, and an orange-coloured deposit was
condensed on the surface.
When the heat was highest, a considerable quantity of vapour
was given off. Its smell was very peculiar ; SHg was distinctly per-
ceptible, but was by no means the dominant odour.
The maximum temperature appeared to be reached at about twenty
minutes after exposure. A thermometer placed in the centre indicated
a temperature of over 600° Fahr. (315° Cent.), and the paper on which
the residue was placed was burnt through. After half an hour's
exposure the deposit cooled very rapidly.
It should be observed that the physical characteristics of the
ground deposit were altered very materially by the heating.
When the residue is taken out of the exploding cylinder, it is
difficult to pound in the mortar, being somewhat unctuous; but
244 RESEARCHES ON EXPLOSIVES
after the development of heat it becomes crisp, and is readily
powdered.
2. In the examination that we have instituted of the products of
explosion of a sample of sporting powder (Curtis and Harvey's No.
6), and of mining powder, the following course of proceeding was
adopted for the removal of the solid residue from the explosion-vessel,
and its preparation for analysis :— Distilled water which had been
freed from air by long-continued boiling, was siphoned into the
explosion-vessel when the latter had cooled, so that air was never
allowed to come into contact with the solid residue. When the
cylinder was thus quite filled with water, it was closed, and set aside
for sufficient time to allow the residue to dissolve completely. The
solution was then decanted into bottles freed from oxygen, which were
quite filled with the liquid, and carefully sealed up until required for
analysis, in carrying out which the course already described was
pursued.
The products obtained by the first of these modifications of the
ordinary course of procedure were submitted to partial examination,
the chief object being to see to what extent the proportions of hypo-
sulphite and sulphide varied in the upper and lower portions of the
residue, and the extent to which they were affected by the great
difference in the mode of treatment sustained by the different portions
of one and the same residue. The proportion of hyposulphite was
determined in every instance, and the products were also examined
in all cases for sulphide. In the first experiment the exact propor-
tion of this latter constituent was ascertained only in one of the three
portions of the residue in which it existed ; it will be seen that one
of the ground portions contained none. The sulphate was determined
in all instances, and, in the second experiment, the proportions of
carbonate existing in the upper and lower portions of the (unground)
residue were ascertained. The analytical results obtained are given
in the following table : —
RESEARCHES ON EXPLOSIVES
245
—
9. p
CO to
10 »o
tn «*-<
Se . Si •
i; 3 : -M 3 :
r« §1
ll II
S Ph c3 Ph
w^
6
Ti
4J
"m
c
s
s
>-< Ci
1
a
:
ITS
00 • t^ •
-%
ii
o
'■a
4J
5
§
%
^
^
c
4)
o
p-a
c
s
s
3 aj
«D 4;
4) (M «
8
' 1
be
I 1 § i
-s
3
S
il
CO c^
,-1 -*
^ CO CD ^
•s »
00 >p
ip 00
-1< 05
sg
!>•
CO CD
u-5 Ah i (M
II
(M
CO
CO CO
""
So;
'Ji (N
10 1^
00 (M IM
•i"S
oo c»
Cq CO >
1
i
i
1
III
•04193
•00244
•01898
•00178
•00072
•00011
•04476
•05805
•00932
•00213
•00006
•00139
•00007
•00598
00
CO^OSOS -OO l>.«Dt^OSOu:5 -OiO •
"^i-HOo :oo .-iCT>->*i-ioo :oco :
CN0,;H0 00 -IOO 'Oi-iO
000000 opoooo 000
CO
4ii
.-1 ,-1 i-i i-il;^ 00 .-H (M 'i* ir.coro '
OSCOtOOSr^rH - J^(NOO.-(MiO -OOCNOS
l*^"
(M i-lW to 10 CO CO CD VO 5D .-H 0010
§
<»
«D0Ot-CO«D0O to 00 i-( OS t^ in CD
tj< CO 00 ^ • CO in CO 00 • >-( • CO ■
eo(Mr-iooo : ^(M(Min :o : oco :
-<
6
iisiiii liiiiiiii '-
CD
05
4:i
^.- to
^ 00 CO t^ CO ':t< ic r^ oi in CO rM
C0-*<00000 0(M^-*<00t^00 •
CDOT^^OOO CO(M rl^ .-: OrH :
C-05pQO^-00
6554-
5839-
5208
4647
4144
3691
3283
2909
2569
2257
1970
1705
1460
1231
1019
821
635
461
298
144
55
CD.-ICDC005CD^C^O00f-«DlOC0C0C^rHOO
4^■*l^0O0(^^(N(^^(^t^lt-l,-lAl.-^1^1-^T-^r^.-l1-l
1*
OT-lt-as05(N-*05CDr-IC00300C005<3it-CO^vO
OMr-l^,-H,-HOOyM.;HO0J0i
M<»4jHOj>.4j<,-iosi)-^(N.-iC5oo»o-*co.-i
Tj00(M>0-*p5pOSt-IJ^J>-niC'X).-IOaiC-OiOcot-^.-ic^^wo(Nepooc
r-70
127°
l°-77
|28°
l°-83
1^29°
l°-90j
"33 o
RESEARCHES ON EXPLOSIVES 299
Experiments for the determination of the specific heats of the vessels
used for determining the heat generated hy explosion.
In all cases the vessel was boiled, and then kept for five minutes
suspended in the escaping steam; it was then transferred to the
calorimeter containing 30,000 grs. of distilled water.
Experiment 141.— Weight of vessel, 21,311-6 grs. (1381-0 grms.).
P'ahr.
Temperature of air . . . . . . . 58°-8
„ steam 2ir-65
„ calorimeter before immersion of
vessel 58°-88
„ calorimeter after thei'mometer be-
came stationary . . . . 70° '08
Loss of heat in vessel . 2ir-65-70°-08 + 0°-2 =14r-77
Gain of heat in water . . . ir-2 +0°-44= ir-64
Hence specific heat of vessel = 1156.
Experiment 142. — The same vessel.
Temperature of air ....... 59°-0
„ steam 2ir-65
„ calorimeter before experiment . 62°-21
„ „ after „ . . 73°-20
Loss of heat in vessel . 2ir-65-73°-20 + 0°-2 =138°-65
Gain of heat in water . . . 10°-99 + 0°-44 = ir-43
Hence specific heat of vessel = -1158.
Experiment 143. — The same vessel.
Temperature of air ....... 60° '20
steam 212°-20
„ calorimeter before experiment . 61°'ll
after „ . . 72°-20
Loss of heat in vessel . 212°-20 - 72°-20 + 0°-20 = 140°-20
Gain of heat in water . . . ll°-09 + 0°-44 = ir-53
Hence specific heat of vessel = 'IISS.
Experiment 144. — The same vessel. ^^^^
Temperature of air 60° -20
steam 212°-10
„ calorimeter before experiment . . 65°' 10
after „ . . 75°-95
Loss of heat in vessel . 212°-20 - 75°-95 + 0°-2 =136°-45
Gain of heat in water . . . 10°-85 + 0°-44 = ir-29
Hence specific heat of vessel = -1163.
Hence mean specific heat of vessel from four experiments = -1158
300 RESEARCHES ON EXPLOSIVES
Experiment 167.— Weight of vessel, 52,931-6 grs. (3430 grms.).
Falir.
Temperature of air . . . . . . . 60°-0
„ steam 2ir-14
,, calorimeter before experiment . 55°-75
after „ . . 84"-52
Loss of heat in vessel . 2ir-14 - 84°-52 + 0°-2 = 126°-82
Gain of heat in water (25,000 grs.) 28°-77 + T '86 = SO^-GS
Hence specific heat of vessel = -1140.
Experiment 168. — The same vessel. p^j^^.
Temperature of air . . . . . . • 62° '0
„ steam 2ir-14
„ calorimeter before experiment . 55° -48
after ,, . . 84°-10
Loss of heat in vessel . 2ir-14 - 84°-10 + 0°-2 =127°-24
Gain of heat in water . . . 28°-60 + 1°-85 = 30°'40
Hence specific heat of vessel = •1132.
Experiment 169. — The same vessel. p^j^^.
Temperature of air 62°-0
„ steam 211°-14
,, calorimeter before experiment . 55-°55
after „ . . 84°-30
Loss of heat in vessel . 2ir-14 - 84°-30 + 0°-2 =127°-04
Gain of heat in water . . . 28°-75 + r-86 = 30°-61
Hence specific heat of vessel = 'IISS.
Hence mean specific heat of vessel from three experiments = •1137.
Deter7ni7iatio7i of heat evolved hy the various poivders.
A. — Small explosion-vessel.
Grs.
Weight of water, 30,000 grs. . equivalent in water 30,000^0
explosion-vessel, 21,311 •G grs. „ „ 2,465^8
„ powder products . . . „ „ 28 ^5
Equivalent in watei*, of contents of calorimeter . . . 32,494^3
When 200 grs. of powder used, the equivalent in water of
the contents of the calorimeter is . . . . . 32,503"8
Experiment 146. — Exploded 150 grs. Curtis and Harvey's No. 6.
Falir.
Temperature of calorimeter before explosion, . . 61°"50
after „ . . 67°^78
Hence difference 6°-28 + 0°^24= 6°^52
Hence heat evolved = 784"0 grm. -units Cent.
RESEARCHES ON EXPLOSIVES 301
Experiment 147. — Exploded 150 grs. Spanish.
Fahr.
Temperature of calorimeter before explosion . . 65°"10
after „ . . 7r-20
Hence difference 6°-10 + 0°-24= 6°-34
Hence heat evolved = 762-5 grm. -units Cent.
Experiment 148. — Exploded 150 grs. E. L. Gr.
Fahr.
Temperature of calorimeter before explosion . . .56°'28
after „ . . 62°-07
Hence difference 5°-79 + 0°-24= 6°-03
Hence heat evolved = 72.5-1 grm. -units Cent.
Experiment 149. — Exploded 150 grs. pebble.
Fahr.
Temperature of calorimeter before explosion , . 60° -42
after „ . . 66°-10
Hence difference 5°-68 + 0°-24= 5^-92
Hence heat evolved = 711-9 grm. -units Cent.
Experiment 150. — Exploded 150 grs. F. Gr.
Fahr.
Temperature of calorimeter before explosion . . 65°-16
after „ . . 70°-80
Hence difference .... 5°-64 + 0°-23= 5°-87
Hence heat evolved = 706-45 grm. -units Cent.
Experiment 153. — With 200 grs. pebble.
Failure ; the plug being spoiled by the explosion.
Experiment 154. — Exploded 150 grs. F. Gr.
Fahr.
Temperature of calorimeter before explosion . . 49°-55
„ „ after „ . . 55° -45
Hence difference 59°-0 + 0°-24= 6°-14
Hence heat evolved = 738-9 grm. -units Cent.
Experiment 155. — Exploded 150 grs. E. L. Gr.
Fahr.
Temperature of calorimeter before explosion . . 86° -GO
after „ . . 9r-73
Hence difference 5°-73 -|-0°-24 = 5°-9r
Hence heat evolved = 718-4 grm. -units Cent.
302 RESEARCHES ON EXPLOSIVES
Experiment 156. — Exploded 150 grs. pebble.
Temperature of calorimeter before explosion . . 56°'07
after „ . . 6r-95
Hence difference 5-°87 + 0°-23= 6°-10
Hence heat evolved = 734*1 grm. -units Cent.
Experiment 157.— Exploded 150 grs. Spanish.
Temperature of calorimeter before explosion . . 56°-90
after „ . . 63°-06
Hence difference 6°-16 + 0°-25- 6°-41
Hence heat evolved = 771-4 grm. -units Cent.
Experiment 158. — Exploded 150 grs. Curtis and Harvey's No. 6.
Falir.
Temperature of calorimeter before explosion . . 57°-92
after „ . . 63° -87
Hence difference 5°-95 + 0'-24= 6°-19
Hence heat evolved = 744-9 grm.-units Cent.
Experiment 159.— Exploded 150 grs. mining.
Temperature of calorimeter before explosion . . 58° -50
after „ . . 62°-54
Hence difference 4°-04 -hO°-16 = 4°-02
Hence heat evolved = 505-5 grm.-units Cent.
Experiment 160.— Exploded 150 grs. K. L. G. ^^^^
Temperature of calorimeter before explosion . . 62° -42
after „ . . 68°-06
Hence difference 5°-64 + 0°-24= 5°-88
Hence heat evolved = 707-5 grm.-units Cent.
Experiment 161.— Exploded 150 grs. pebble.
Temperature of calorimeter before explosion . . 49° -05
after „ . . 54° '59
Hence difference 5°-54 + 0°-23= 5°-77
Hence heat evolved = 694-4 grm.-units Cent.
Experiment 162.— Exploded 150 grs. mining. ^^^^^
Temperature of calorimeter before explosion . . 56° -80
after „ . • 60°-82
Hence difference 4°-02 + 0=-20= 4-22
Hence heat evolved = 507-9 grm.-units Cent.
RESEARCHES ON EXPLOSIVES 303
Experiment 163. — Exploded 150 grs. Curtis and Harvey's No. 6.
Fahr.
Temperature of calorimeter before explosion . . 52°'80
after „ . . 58° -65
Hence difference 5°-85 + 0°-24= 6°-09
Hence heat evolved = 732-9 grm.-units Cent.
Experiment 164. —Exploded 150 grs. F. G.
Fahr.
Temperature of calorimeter before explosion . . 57° -42
after „ . . 63°-26
Hence difference 5°-84 + 0°-24= 6°-08
Hence heat evolved = 731*7 grm.-units Cent.
Experiment 165. — Exploded 150 grs. Spanish.
Fahr.
Temperature of calorimeter before explosion . . 55"-70
after „ . . 6r-72
Hence difference 6°-02 + 0°-24= 6''-26
Hence heat evolved = 753-4 grm.-units Cent.
Experiment 166. — Exploded 150 grs. pebble.
Fahr.
Temperature of calorimeter before explosion . . 61°-12
after „ . . 66°-80
Hence difference 5°-68 + 0°-22= 5°-90
Hence heat evolved = 710-0 grm.-units Cent.
Experiment 151.— Exploded 200 grs. mining.
Fahr.
Temperature of calorimeter before explosion . . 60°-38
„ „ after „ . . 65°-87
Hence difference 5°-49 + 0°-22= 5°-7l
Hence heat evolved = 512-7 grm.-units Cent.
Experiment 152. — Exploded 200 grs. Curtis and Harvey's No. 6.
Fahr.
Temperature of calorimeter before explosion . . 64° -95
after „ . . 7 3° -00
Hence difference 8°-05 + 0°-32= 8°-37
Hence heat evolved = 755*7 grm.-units Cent.
B. — Large explosion-vessel.
Grs.
Weight of water, 25,000 grs. . . equivalent in water 25,000-0
„ explosion-vessel, 52,931-6 grs. . „ „ 6,018-3
„ powder products, 400 grs. . . „ „ 76-0
Equivalent in water, of contents of calorimeter . . . 31,094-3
304 RESEARCHES ON EXPLOSIVES
Experiment 171. — Exploded 400 grs. pebble.
Fahr.
Temperature of caloi'imeter before explosion . . 54° '38
after „ . . 69°-43
Hence difference .... 15°-05 + r-24 = 16°-29
Hence heat evolved = 703-41 grm. -units Cent.
Experiment 172.— Exploded 400 grs. E. L. G.
Fahr.
Temperature of calorimeter before explosion . . 57° "08
after ,. . . 72°-44
Hence difference .... 15°-36 + r'26 = 16°-62
Hence heat evolved = 717"7 grm. -units Cent.
Experiment 173.— Exploded 400 grs. E. L. G.
Fahr.
Temperature of calorimeter before explosion . . 57°'22
after „ . . 72°-74
Hence difference .... 15°-52 4- 1°-26 = 16°-78
Hence heat evolved = 724-7 grm. -units Cent.
Experiment 174. — Exploded 400 grs. pebble.
Fahr.
Temperature of calorimeter before explosion . . 57°'13
after „ . . 7 2° '42
Hence difference .... 15°-29 + r-26 = 16°-55
Hence heat evolved = 714-7 grm. -units Cent.
Experiment 175. — Exploded 400 grs. pebble.
Fahr.
Temperature of calorimeter before explosion . . 56°-40
after „ . . 72°-00
Hence difference 15°-6 -h r-27 = 16°-87
Hence heat evolved = 728-5 grm.-units Cent.
Experiment 176.— Exploded 400 grs. E. L. G.
Fahr.
Temperature of calorimeter before explosion . . 67°-12
after „ . . 82°-38
Hence difference .... 15°-26 4- r-26 = 16°-52
Hence heat evolved = 713'4 grm.-units Cent.
Experiment 177. — Exploded 400 grs. mining.
Fahr.
Temperature of calorimeter ])efore explosion . . 53°-27
after „ ' . . 64°-13
Hence diflerence .... 10°-86 + 0°-71 = ll°-57
Hence heat evolved = 499-65 grm.-units Cent.
RESEARCHES ON EXPLOSIVES 305
Experiment 178.— Exploded 400 grs. mining.
Fahr.
Temperature of calorimetei* before explosion . . 57°-25
after „ . . 68°-27
Hence difference .... ir-02 + 0°-72 = ir-74
Hence heat evolved = 507*0 grm.-units Cent.
Experiment 179. — Exploded 400 grs. mining.
Fahr.
Temperature of calorimeter before explosion . . 64°-73
after „ . . 76°-03
Hence difference .... ir-30 + 0°-74 = 12°-04
Hence heat evolved = 520-0 grm.-units Cent.
Experiment 181. — Exploded 400 grs. Spanish.
Fahr.
Temperature of calorimeter before explosion . . 5r-62
after „ . . 67"-76
Hence difference .... 16°-14 + 1' -32 = 17°-46
Hence heat evolved = 7540 grm.-units Cent.
Experiment 182. — Exploded 400 grs. Spanish.
Fahr.
Temperature of calorimeter before explosion . . 66°-90
after „ . . 83°-21
Hence difference .... 16°-31 + T -33 = 17'^-64
Hence heat evolved = 761-8 grm.-units Cent.
Experiment 183. — Exploded 400 grs. special mining.
Falir.
Temperature of calorimeter before explosion . . 52°-72
after „ . . 63° -58
Hence difference .... 10°-86 -hO°-71 = ir-57
Hence heat evolved = 499-65 grm.-units Cent.
Experiment 184. — Exploded 400 grs. Spanish.
Fahr.
Temperature of calorimeter before explosion . . 58°"52
after „ . . 75°-03
Hence difference .... 16°-51 + r-35 = 17°-86
Hence heat evolved = 771-3 grm.-units Cent.
Experiment 185. — Exploded 400 grs. Curtis and Harvey's No. 6.
Fahr.
Temperature of calorimeter before explosion . . 53°-19
after „ . . 69° -26
Hence difference .... 16°-07 + 1°-31 - 17°-38
Hence heat evolved = 750-6 grm.-units Cent.
U
306 RESEARCHES ON EXPLOSIVES
Experiment 186. — Exploded 400 grs. Curtis and Harvey's No. 6.
Fahr.
Temperature of calorimeter before explosion . . 55°*80
„ „ after „ . . 72°-19
Hence difference .... 16°-39 + r-33 = 17°-72
Hence heat evolved = 765*3 grm. -units Cent.
Experiment 187. — Exploded 400 grs. Curtis and Harvey's No. 6.
Fahr.
Temperature of calorimeter before explosion . . 53°-48
after „ . . 69°-58
Hence difference .... 16°-10 + r-32 = 17°-42
Hence heat evolved = 752'3 grm. -units Cent.
Experiment 189.— Exploded 400 grs. F. G.
Fahr.
Temperature of calorimeter before explosion . . 55°*61
after „ . . 7r-26
Hence difference .... 15''-65 + r-28 = 16°-93
Hence heat evolved — 731*1 grm.-units Cent.
Experiment 191.— Exploded 400 grs. F. G.
Fahr.
Temperature of calorimeter before explosion . . 53°'04
after „ . . 68° -49
Hence difference .... 15°-45 + r-27 = 16°-72
Hence heat evolved = 722-1 grm.-units Cent.
Experiment 192.— Exploded 400 grs. F. G.
Fahr.
Temperature of calorimeter before explosion . . 53°-95
after „ . . 69°-59
Hence difference .... 15°-64 + r-28 = 16°'92
Hence heat evolved = 730*7 grm.-units Cent.
Experiment 193. — Fired 5960 grs. = 386-2 grnis. mining powder
in cylinder No. 6, containing 14,900 grs.
Temperature of gas = 55°-5 Fahr. = 13°-1 Cent. Bar. 30*025 = 762*35 mm.
Amount of gas = 27"*94 - 8'''*35 = 19"*59.
= 444-8 X 19*59 -I- 18 cub. inches.
= 8731*63 cub, inches.
= 143,076-49 c.c. at 13°*1 Cent., and 762*35 mm.
= 143,518*9 c.c. at 13°*1 Cent., and 760 mm.
= 136,944*2 c.c. at 0" Cent., and 760 mm.
= 354*6 times original volume.
RESEARCHES ON EXPLOSIVES 307
Experiment 194. — Fired 4650 grs. (301*3 grms.) mining powder
in cylinder containing 15,500 grs.
In letting the gas escape, foimd for the first time that the gas
lighted, giving rise to an intensely suffocating smell of sulphurous
acid, showing, as was indeed otherwise apparent, that sulphydric acid
was present in large quantities. Sealed up gases for examination.
Did not take out the deposit as usual, but after the gases had
escaped, filled the cylinder, by displacement, with distilled water,
entirely freed from air by long boiling.
On the water touching the deposit, it decrepitated with consider-
able sharpness. When the cylinder was full it was entirely sealed,
and reopened after an interval of about forty-eight hours. The solution
was then decanted into bottles, freed from oxygen, and sealed for
examination.
5. A. a. Crush. Pressure.
•30 -0833 -0417 -015 5-04 tons per square inch.
Experiment 195. — Fired 5960 grs. = 386"2 grms. Curtis and
Harvey's No. 6, in cylinder No. 6.
Temperature of gas = 60°-8 Fahr. = 16°-0 Cent. Bar. 30"'430 = 772-9.
Amount of gas = 444*8 x 13-10 -t- 18 cub. inches.
= 5,844-88 cub. inches.
= 95,774-2 c.c. at 772-9 and 16°-0 Cent.
= 97,399-8 c.c. at 16°-0 Cent, at 760 mm.
= 92,004-6 c.c. at 0° Cent., and 760 mm.
= 238-23 vols.
Experiment 196. — Fired 4650 grs. = 301'3 grms. Curtis and
Harvey's No. 6, in cylinder containing 15,500 grs. Took all the
precautions described in Experiment 194 ; observed gas issuing from
vessel would not light.
5. A. a. Crush. Pressure.
•30 -0833 -0417 -015 5-04 tons per square inch.
Experiment 197. — Fired 10,000 grs. of mining powder in cylinder
containing 15,500 grs. of water. A good deal of gas escaped past
crusher-plug.
5.
•700
•0833 -0417
Crush. Pressure.
A -220 = 20-8 tons per square inch.
B -221 = 20-8
C -226 = 21^2
Deposit approximately = 2025 grs. water.
308 RESEARCHES ON EXPLOSIVES
Experiment 198. — Fired 10,000 grs. Curtis and Harvey's No. 6.
The greater proportion of the gases escaped, the gas getting
between the steel barrel and the coil, by the screw of the crusher-
plug, causing the coil to crush, and indenting the steel and the coil in
a very remarkable manner.
5. A. a. Crush. Pressure.
•70 -0833 -0417 A -214 = 19-95 tons per square inch.
B -197 = 19-11
C -197 = 19-11
Experiment 199. — ^A series of experiments in guns of various
calibres with mining powder.
Experiment 200. — Fired cylinder hooped with B. E. iron No. 3,
and with a capacity of 11,000 grs. water, with a charge of 8750 grs.
pebble, and 2250 grs. F. G. — 11,000 grs. in all. Copper in crusher-
plug crushed beforehand to 35 tons.
Head of crusher-plug broke off by the explosion, and gas escaped,
taking, as nearly as could be guessed, from one to two seconds to escape.
Outside diameter of cylinder before firing . lO'^-SeS
after „ . 10"-393
Pressure developed over 39 tons, but not reliable, owing to the
escape of the gases.
Ex^Deriment 201. — Fired cylinder hooped with Siemens' mild
steel No. 2, and with a capacity of 11,200 grs. water, with a charge of
11,200 grs. powder, consisting of 8750 grs. pebble and 2450 grs. F. G.
Gas escaped with great rapidity past the firing cone, which was
of course destroyed ; great difficulty found in extracting the crusher-
piston, which had been jammed by the compression of the chamber
in which it was placed. Its record was therefore valueless.
Experiment 202. — Fired same cylinder with a charge of 13,640
grs. powder, of which 8375 grs. were pebble, the rest F. G. Crusher-
plug blew out before charge fully fired.
Experiment 225. — Fired 9000 grs. pebble and 3000 grs. F. G. in
cylinder No. 2, containing 12,680 grs. water, less 670 grs. occupied
by ijiternal crusher-gauge. One internal crusher used ; gas escaped
slowly.
0. A. a. Crush. Pressure.
1-0 -0417 -0833 B crusher -193 = 42-52 tons per sq. inch.
C „ -193 = 42-72
RESEARCHES ON EXPLOSIVES 309
Experiment 230. — Fired in N"o. 2 cylinder (Siemens'), containing
12,680 grs., 11,360 grs. mining powder, same as that tested in the
2-5-inch B.L. gun on 4th September 1878. The gas escaped
through the insulated cone, almost at once. It did not escape with
any violence.
Pressures indicated were as follow : —
3_ A. (t. Crush. Pressure.
1 •0417 •OSSS
A •leS = 36"8 tons per square inch.
(A, doubtful ; piston being jammed.)
B -200 = 43^9 tons per square inch.
= 43^9
C ^200
Experiment 233.— Eired 9000 grs. pebble, and 4000 grs. F. G-.,
total 13,000 grs. (842-4 grms.), in No. 2 Siemens' cylinder, cubic
contents, 12,680 grs. — 2000 grs. for two internal crusher-gauges;
total contents, 10,680 grs. The pressure forced out the closing-plugs
by shearing the threads.
5.
1-21
A.
•0417
•0833
A crusher
B
c „
Crush. Pressure.
•256 = 55-6 tons per sq. inch.
•256 = 55-6
•260 = 56^8
Note.— {Added 9th March, 1880.)
Since this memoir was submitted to the Society, we have been
led, in consequence of a communication made to us by Dr Debus, to
modify considerably our views with regard to the formation of
hyposulphite.
The experiments rendered necessary by Dr Debus's discovery are
fully described and discussed in a note submitted to the Eoyal
Society,* but as the facts there given have led us to the conclusion
"that although it would seem that in certain cases and under
certain exceptional circumstances potassium hyposulphite does exist
as a secondary, it exists in no case as a primary product, and should
not, therefore, be reckoned among the normal constituents of powder-
residues," we have recalculated the whole of our analytical results,
and we append two tables, Nos. 12 and 13, giving for each experi-
ment the products of decomposition calculated on the hypothesis
that prior to removal from the explosion-vessel the whole of the
hyposulphite found was in the form of mono- or polysulphides.
* Proc. Roy. Soc, vol. xxx., p. 198.
310
RESEARCHES ON EXPLOSIVES
Table 12. — Showing the mean analytical results ohtained from an examination of
powders ; showing also the same particulars
2
Percentage composition by
g
S
Mean density
of products
'
K
Nature of powder.
of
combustion.
O X
.2
.g
o
Yl
II
1
o
ItJ
•^
H
1°
'A
ft c4
1
8
Pebble, W. A
•10
46-66
14-76
32-75
3-13
7
20
44-78
16-09
31-31
4-23
9
30
47^03
15-51
31-71
2-90
12
40
49^52
13-95
32-16
1-70
14
50
49-82
13-36
32-19
1-96
37
60
49-48
13-75
31-83
2-24
38
70
49-93
12-51
32-08
3-18
43
80
51-54
11-88
32-61
1-96
77
•90
Means
51^75
10-87
32-72
2^13
48-95
13-63
32-15
2-60
Highest
51-75
16-09
32-75
4-23
1
R. L. G.,W. A
Lowest
•10
44-78
10-87
31-31
1-70
49-00
8-98
35-60
4-06
3
•20
46-56
11-47
35-13
3-58
4
•30
49-35
11 •eo
32^96
3-11
11
•40
50-25
10-84
34^23
1-93
70
•50
47-21
17-04
30^29
1-61
39
•60
46-29
14-52
32^40
4-29
96
•60
50-22
13-93
31^74
1-62
41
•70
49-75
13-38
31-94
2-85
44
■80
51-62
12-16
32-16
1-56
68
•90
Means
52-65
10-73
32-65
1-90
49-29
12-47
32-91
2-65
Highest
52-65
17-04
35-60
4-29
16
F. G.,W. A
Lowest
•10
46-29
8-98
30-29
1-56
44-76
16-25
32-57
2-26
17
•20
47-41
12-35
32-35
3-76
18
•30
50-45
11-33
32-22
2-21
19
•40
51-79
10-31
32-54
2-00
75
•50
51-04
10-38
33-15
2-20
40
•60
5200
9-60
33-28
2-26
42
•70
53-02
7-91
34-26
2-03
47
•80
51-80
8-32
34-64
2-61
69
•90
Means
53-34
7-71
33-81
2-95
50-63
10-47
33-21
2-48
Highest
53-34
16-25
34-64
3-76
78
R. F. G., W. A. ...
Lowest
•70
44-76
7-71
32-22
2-00
52-40
8-86
34-51
1-60
79
Spanish spherical
•70
53-34
4-62
37-80
2-74
196
Curtis and Harvey, No. 6 .
•30
50-22
7-52
34-46
2-08
194
Mining powder .... -30
32-15
33-75
19-03 7-10
RESEARCHES ON EXPLOSIVES
311
the solid and gaseous products of decomposition of Fehble, B. L. G., and F. O.
toith respect to four other poxoders.
volume of the gas.
Percentage composition by weight of the .solid residue.
i
1
W
O
u
II
4
If
li
S .
|.|
(2 "
S
ll
r
11
1
1
0-32
0-58
0-55
0-35
0-34
0-68
2-70
3-59
2-84
2-35
2-08
2-15
1-95
1-67
1-85
58-56
58-01
60-09
57-25
57-04
59-00
54-64
62-35
66-43
15-84
13-85
12-74
13-69
12-12
13-82
13-91
10-94
9-45
20-50
20-41
19-24
18-52
23-02
17-68
22-72
16-84
11-92
0-09 1 0-51
0-06 ! ...
0-21 i 0-03
0-25 j 0-08
0-23 1 0-20
0-36 1 0-32
0-41 , 0-26
0-06 ! 0-33
0-59 0-44
0-17
0-09
0-17
0-07
0-08
0-06
0-06
0-08
0-12
4-33
7-58
7-52
8-74
7-31
8-76
8-00
9-40
11-05
1-40
0-31
0-68
2-35
3-59
1-67
...
59-26
66-43
54-64
12-93
15-84
9-45
18-98
23-02
11-92
0-25 0-24
0-59 0-51
0-06 0-03
0-10
0-17
0-06
8-08
11-05
4-33
0-16
1-40
0-29
0-07
0-28
0-84
0-36
0-35
0-55
0-77
0-80
2-07
2-62
2-98
2-47
3-01
2-14
2-14
1-53
1-72
1-27
0-57
55-41
55-47
54-16
51-82
64-77
66-43
64-88
63-25
67-00
67-16
21-58
24-44
25-03
24-35
4-96
10-90
11-16
11-04
8-88
8-71
16-68
13-08
13-76
17-00
19-47
11-85
13-91
15-34
10-92
12-50
0-05
0-05
0-17
0-30
0-28
0-26
0-51
0-25
0-38
0-59
0-12
0-03
0-13
0-53
0-46
0-44
0-18
0-20
0-06
0-06
0-04
0-04
0-11
0-09
0-11
0-08
0-11
0-15
4-93
6-76
6-93
6-49
9-86
9-99
9-68
9-34
12-66
10-90
0-75
0-02
0-43
0-84
0-07
2-19
3-01
1-27
0-06
0-57
61-03
67-16
51-82
15-10
25-03
4-96
14-45
19-47
10-92
0-22
0-51
0-05
0-27
0-59
0-03
0-08
0-15
0-04
8-74
12-66
4-93
0-08
0-75
0-18
0-27
0-18
0-50
0-41
0-16
3-83
4-13
3-51
3-36
2-96
2-68
2-13
2-04
2-04
0-15
0-28
o'-15
0-18
52-43
60-20
47-17
48-37
57-97
45-55
48-39
47-80
54-17
19-00
24-55
23-24
23-46
21-45
24-15
23-61
23-15
19-64
18-30
8-30
19-23
21-50
12-55
20-12
20-90
21-98
18-88
0-02
0-07
0-08
0-07
0-17
0-26
0-26
0-27
0-21
0-08
0-10
0-10
0-09
0-18
0-21
0-28
0-28
5-74
3-49
0-07
0-15
0-01
0-04
0-08
0-01
0-03
0-04
0-03
4-25
6-70
10-18
6-45
7-79
6-33
6-60
6-49
6-73
0-19
0-50
2-96
4-13
2-04
0-08
0-28
51-34
60-20
45-55
22-47
24-55
19-00
17-97
21-98
8-30
0-13
0-27
0-17
0-28
0-08
1-02
5-74
0-05
0-15
0-01
6-83
10-18
4-25
0-12
2 •46
2-73
2-51
1-29
3-26
5-24
0-21
60-17
35-66
59-10
41-36
22-35
48-55
21-65
0-59
9-14
7-72
12-42
37-10
0-04
0-04
2 '-95
0-06
0-95
0-29
0-09
0-05
0-04
0-09
1-78
8-19
7-04
6-45
14-11
2-02
312
RESEARCHES ON EXPLOSIVES
Table 13.
-Composition by weight of the products of combustion of 1
ijravimetric
3
Proportions by weight of gaseous products
1
s
!
1
Nature of powder.
^1
It
II
Nitrogen.
Sulphhydric
]'
1
1
5
8
Pebble, W. A. .
•10
•2634
•0530
•1176
•0137
•0007
7
•20
•2505
•0572
•1114
•0183
•0009
9
•30
•2609
•0548
•1120
•0124
•0007
12
•40
•2683
•0481
•1109
•0071
•00()7
•0006
14
•50
•2768
•0472
•1137
•0084
•0012
•0005
37
•60
•2695
•0477
•1103
•0094
•0011
•0005
38
•70
•2748
•0438
•1124
■0135
•0007
•0005
43
•80
•2785
•0409
•1121
•0082
•0007
•0004
77
•90
Means
•2743
■0367
•1103
•0087
•0014
•0005
•2685
•0477
•1123
•0111
•0006
•0006
Highest
•2785
•0572
•1176
•0183
•0014
•0009
1
R. L. G., W. A. .
Lowest
•2505
•0367
•1103
•0071
...
•0004
10
•2653
•0309
•1226
•0170
•0006
•0005
3
20
•2497
•0391
•1198
•0148
•0001
•0006
•0022
4
30
•2633
•0394
•1119
•0128
•0007
11
40
•2702
•0371
•1172
•0080
■0006
•0006
70
50
•2601
•0581
•1053
•0068
■0017
•0007
39
60
•2480
•0495
•1101
•0177
•0007
•0005
96
60
•2672
•0471
•1074
•0067
•0007
•0005
41
70
•2655
•0454
•1085
•0118
•0011
•0004
44
80
•2651
•0397
•1051
•0062
•0014
•0040
68
•90
Means
•2760
•0358
•1089
•0077
•0015
•0003
•2630
•0422
•1117
■0109
•0008
•0009
•0002
Highest
•2760
•0581
•1226
•0177
•0017
•0040
•0022
16
F. G.,W. A.
Lowest
•10
•2480
•0309
•1051
•0062
•0001
•0003
•2512
•0580
•1163
•0098
•0004
•0010
•0006
17
•20
•2490
•0413
•1081
•0153
•0010
18
•30
•2621
•0374
•1065
•0089
•0008
•0010
19
•40
•2765
•0350
•1105
•0082
•0008
75
•50
•2665
•0344
•1102
•0089
•0005
■0007
40
•60
•2782
•0327
•1133
•0093
•0003
■0007
42
•70
•2804
•0266
•1152
•0083
•0010
•0005
•0006
47
•80
•2752
•0281
•1171
■0107
•0008
•0005
■0007
69
•90
Means
•2812
•0259
•1134
■0120
•0003
•0005
•2689
•0355
•1123
■0101
•0004
■0007
■0003
Highest
•2892
•0580
•1171
■0153
■0010
■0010
•0010
78
R. F. G., W. A. .
Lowest
•70
•2490
•0259
•1065
•0082
■0005
•2686
•0289
•1126
•0064
•0002
•0006
79
Spanish spherical .
•70
•2457
•0136
•1108'
•0097
•0003
•0007
196
Curtis&Harvey.No.e
•30
•2593
•0247
•1132
•0083
•0046
•0008
194
Mining powder
•30
•2279
•1522
•0858
■0389
•0070
•0017
RESEARCHES ON EXPLOSIVES
313
qramme of fired gunpowder of the undermentioned natures, and of various
densities.
Proportions by weight of the solid residue.
oi
P
It
||
1
. 1
S 6
a^-
s|
4
B .
a
§3
i"!
-:i
II
ll
.3§
CO >>
II
.S5
II
1°
II
1
1
a-3
si
2g
^ M
a
^1
"o o
a 3
m
o
Is.
gs
•3174
•0858
•1111
•0005
•0027
•0009
•0234
•4484
•5418
•0095
•3203
•0765
•1127
•0003
•0005
•0419
•4383
•5522
•0095
•3303
•0700
•1058
•0012
•0002
•0009
•0413
•4408
•5497
•0095
•3176
•0760
•1028
•0014
•0004
•0004
•0485
•0077
•4357
•5548
•0095
•3096
■0658
•1249
•0012
•0011
•0004
•0397
•4478
•5427
•0095
•3257
•0763
•0976
•0020
•0018
•0003
•0484
•4385
•5520
•0095
•2977
•0758
•1238
•0022
•0014
•0003
•0436
•4457
•5448
•0095
•3428
•0601
•0926
•0003
•0018
•0004
•0517
...
•4408
•5497
•0095
•3711
•0528
•0666
•0033
•0025
•0007
•0617
•4318
•5587
•0095
•3258
•0710
•1042
•0014
•0013
•0005
•0445
•0008
•4409
•5496
•0095
•3711
•0858
•1249
•0033
•0027
•0009
•0617
•4484
•5587
•0095
•2977
•0528
•0666
•0003
•0003
•0234
•4318
•5418
•0095
•3059
•1191
•0921
•0033
•0003
•0272
•0041
•4369
•5520
•0111
•3121
•1375
•0736
•0003
•0007
•0003
•0380
•0001
•4263
•5626
•0111
•3037
•1403
•0772
•0003
•0002
•0002
•0389
•4281
•5608
•0111
•2877
•1352
•0944
•0010
•0007
•0002
•0360
•4337
•5552
•0111
•3601
•0276
•1083
•0017
•0030
•0006
•0548
•4327
•5561
•0112
•3739
•0614
•0667
•0016
•0026
•0005
•0562
•4265
•5629
•0111
•3629
•0624
•0778
•0015
•0006
•0541
...
•4296
•5593
•0111
•3519
•0614
•0853
•0028
•0024
•0004
•0520
•4327
•5562
•0111
•3802
•0504
•0620
•0014
•0010
•0006
•0718
•4215
•5674
•0111
•3764
•0488
•0700
•0021
•0011
•0008
•0611
•4302
•5603
•0111
•3415
•0844
•0807
•0013
•0015
... 1 -0004
•0490
•0004
•4298
•5591
•0111
•3802
•1403
•1083
•0021
•0033
•0008
•0718
•0041
•4369
•5674
•0112
•2877
•0276
•0620
•0002
■0272
•4215
•5520
•0111
•2872
•1042
•1003
•0011
•0315
•0004
•0233
•4372
•5480
•0148
•3434
•1401
•0473
•0001
•0005
•0009
•0382
•4147
•5705
•0148
•2683
•1321
•1093
•0003
•0005
•0001
•0579
•4167
•5685
•0148
•2680
•1300
•1192
•0004
•0006
•0002
•0358
•4310
•5542
•0148
•3269
•1210
•0708
•0004
•0005
•0005
•0439
•4212
•5640
•0148
•2508
•1330
•1108
•0009
•0010
•0192
•0001
•0349
•4345
•5507
•0148
•2674
•1305
•1155
•0014
•0012
•0002
•0364
•4326
•5526
•0148
•2640
•1278
•1214
•0014
•0015
•0002
•0358
•4331
•5521
•0148
•2989
•1084
•1042
•0015
•0016
•0002
•0371
•4333
•5519
•0148
•2861
•1252
•0999
•0007
•0009
•0056
•0003
•0381
•4282
•5569
•0148
•3434
•1401
•1214
•0015
•0016
•0315
•0009
•0579
•4372
•5705
•0148
•2508
•1042
•0473
•0005
•0001
•0233
•4147
•5480
•0148
•3458
•1285
•0525
•0002
•0003
•0003
•0471
•4173
•5747
•0080
•2186
•2975
•0473
•0002
•0058
•0002
•0431
•3808
•6127
•0065
•3413
•1250
•0717
•0017
•0005
•0372
•4109
•5774
•0117
•1945
•0028
•1745
•0139
•0004
•0084
•0664
•0095
•5135
•4704
•0161
314 RESEARCHES ON EXPLOSIVES
Note on the Existence of Potassium Hyposulphite in the
Solid Eesidue of Fiked Gunpowdek.
In our second memoir on fired gunpowder we have discussed in
detail that part of M. Berthelot's friendly criticism of our first
memoir, which relates to the potassium hyposulphite found by us, in
variable proportions, in our analyses of the sohd products obtained
by the explosion of gunpowder in the manner described. While
pointing out that we had taken every precaution in our power to
guard against the production of hyposulphite by atmospheric action
upon the potassium sulphide during the removal of the hard masses
of solid products from the explosion-vessel, and had effectually
excluded air from them, when once they were removed until they
were submitted to analysis, we admitted the impossibility of
guarding against the accidental formation of some hyposulphite
during the process of removal, especially in some instances in which
the structure of the residue had certainly been favourable to atmo-
spheric action, and in which a more or less considerable development
of heat had afforded indications of the occurrence of oxidation.
We contended, however, that the method of analysis, and the
precautions adopted by us in carrying it out, precluded the possi-
bility of accidental formation of hyposulphite at this stage of our
investigations. With respect to the precautions, we could, and still
do, speak with perfect confidence ; and we certainly have beheved
ourselves fully justified in being equally confident with respect to
the process adopted by us for the determination of the proportions
of sulphide and hyposulphite, inasmuch as we accepted and used in
its integrity the method pubHshed in 1857 by Bunsen and Schischkoff
in their classical memoir on the products of explosion of gunpowder,
and adopted since that time by several other investigators who have
made the explosion of gunpowder the subject of study, and whose
results are referred to in our first memoir.
Imposing implicit confidence in the trustworthiness of this
method of analysis, emanating as it did from one of the highest
authorities in experimental research, we considered ourselves fully
justified in maintaining that the very considerable variations in the
amount of hyposulphite found in different analyses, carried out as
nearly as possible under like conditions, and the high proportions o f
sulphide obtained in several of those analyses, afforded substantial
proof that accidental oxidation during the collection and analysis of
the residues was not sufficient to account for all but the very small
RESEARCHES ON EXPLOSIVES 315
quantities of hyposulphite, which, in M. Berthelot's view, could have
pre-existed in the powder-residues. Other facts, estabhshed by the
exhaustive series of experiments detailed in our first memoir, were
referred to by us in our second memoir, in support of the above con-
clusion (from which we have still no reason whatever to depart).
At the same time we described a series of supplementary experiments
which had been instituted by us, with a view to obtain, if possible,
further decisive evidence as to the probable proportions of hypo-
sulphite and sulphide actually existing in the residues furnished by
the explosion of gunpowder in closed vessels.
In the first place, the residues obtained by the explosion of
charges of E. L. G. and pebble powders were submitted to special
treatment. Portions of each, consisting exclusively of large masses,
were very speedily detached and removed from the explosion-vessels,
and sealed up in bottles freed from oxygen, having been exposed to the
air only for a few seconds. Other portions of the same residues were
very finely ground, and exposed to the air for 48 hours. As was stated in
our recent memoir, the portions of the residues treated in the last-named
manner contained very large proportions of hyposulphite (although
in one of them there still remained about 3 per cent, of sulphide),
while those portions which had been for only a brief period exposed
to air (and which presented but small surfaces) were found to contain
from 5 to 8-5 per cent, of hyposulphite. As, throughout our entire
series of previous experiments, no accidental circumstances had
occurred which even distantly approached the special conditions
favourable to the oxidation of the sulphide presented in these
particular experiments, we considered ourselves fully justified in
concluding that the non-discovery of any sulpliides in the analyses
of residues furnished by the fine-grain powder in three out of the
whole series of experiments, was not due to accident in the manipula-
tions ; and that in those instances, in our several series of experi-
ments, in which large quantities of hyposulphite were found, the
greater proportion of that substance must have existed before the
removal of the residues from the explosion-vessel.
Not suffering the question to rest there, however, we proceeded,
in the second place, to adopt new precautions, in two special experi-
ments, for guarding against the possible formation of hyposulphite in
the removal of the residues from the explosion-vessel, and their pre-
paration for analysis.
Distilled water, carefully freed from air by long-continued
boiling, was syphoned into the vessel when it had cooled after the
816 RESEARCHES ON EXPLOSIVES
explosion, and thus no air was ever allowed to come into contact
with the solid products. "When the vessel was quite filled with
water it was closed, and, after having been left at rest for a sufficient
time to allow the residue to dissolve completely, the solution was
rapidly transferred to bottles which had been freed from oxygen.
These, when completely filled with the liquid, were hermetically
sealed until the contents were submitted to analysis in accordance
with the usual method, when they furnished respectively 4 and 6
per cent, of hyposulphite. These results corresponded closely to
others obtained by the analysis of seven residues obtained in experi-
ments with P., E. L. Gr., and L. G-. powders, in which there were no
pecuHarities assignable as a possible reason why the proportions of
hyposulphite should be so much lower in these cases than in other
experiments carried out with the same powders under as nearly
as possible the same conditions.
By the results obtained under the various conditions pointed out
in the foregoing, we are forced to the conclusion that the discovery
of a small or a larger proportion of hyposulphite by the analysis of
the powder-residue, obtained as described, is consequent upon some
slight variation (apparently not within the operator's control)
attending the explosion itself; but that hyposulphite does exist,
though generally not to anything like the extent we were at first led
to believe, as a normal and not unimportant product of the explosion
of powder in a closed space.
Some time after the submission of our second memoir to the
Eoyal Society, we received a communication from Professor Debus,
which has led us to institute a further series of experiments bearing
upon this question of the existence of hyposulphite, and the results
we have arrived at have led us so greatly to modify our views on
this point, that it is our duty to communicate them without loss of
time to the Eoyal Society.
As introductory to these, it is necessary to repeat the account,
given in our first memoir, of the method pursued by us for determin-
ing the proportions of potassium monosulphide and hyposulphite in
a powder-residue.
The solution of the residue, prepared by the several methods
already described, was separated by filtration, as rapidly as possible,
from the insoluble portion, the liquid being collected in a flask, in
which it was at once brought into contact with pure ignited copper
oxide. The solution and oxide were agitated together, from time to
time, in the closed flask, the two beings allowed to remain together
RESEARCHES ON EXPLOSIVES 317
until the liquid' was perfectly colourless. In a few instances the
oxide was added in small quantities at a time, in others the sufficient
excess was added at once, with no difference in the result obtained.
The only points in which this method differed from that described by
Bunsen and Schischkoff in their memoir, was in the employment of
a flask well closed with an indiarubber bung for the stoppered
cylinder which was employed by them ; and in occasionally curtailing
somewhat the prescribed period (two days) for which the liquid and
the copper oxide were allowed to remain together, the operation
being considered complete when the solution had become colourless.
Bunsen and Schischkoff prescribed that the liquid when separated by
filtration from the mixed copper oxide and sulphide obtained in the
foregoing treatment, is to be divided into seven equal volumes, in one
of which the amount of hyposulphite may be most sunply estimated
by acidifying it with acetic acid, and then titrating with a standard
iodine solution. This course was adopted by us, and it will therefore
be seen that we departed in no essential point whatever from the
method of Bunsen and Schischkoff, which we had considered ourselves
fully warranted in adopting, without questioning its trustworthiness.
We were informed, however, last July by Dr Debus, that in sub-
mitting potassium polysulphides to treatment with copper oxide, he
had found much hyposulphite to be produced, even when air was
perfectly excluded, it having been in the first instance ascertained
that the several polysulphides experimented with did not contain any
trace of hyposulphite. We proceeded at once to confirm the correct-
ness of his observations by submitting potassium polysulphides to
treatment with copper oxide, proceeding exactly according to the
method prescribed by Bunsen and Schischkoff for the treatment of
powder-residues. In one experiment we obtained as much as 871
per cent, of potassium hyposulphite (calculated upon 100 parts of
potassium monosulphide). Even in an experiment with pure potas-
sium monosulphide, we obtained 11"6 per cent, of hyposulphite upon
its treatment for the usual period with copper oxide.
We next proceeded to convince ourselves that by substituting
zinc chloride solution for copper oxide, the sulphur existing in solu-
tions of potassium mono- and polysulphides might be abstracted,
according to the usual method of operation, without producing more
than the very small quantities of hyposulphite ascribable to the access
of a little air to the sulphides before or during the method of treatment.
Having confirmed the validity of Dr Debus's objection to Bunsen
and Schischkoff's method, and established the trustworthiness of a
318
RESEARCHES ON EXPLOSIVES
modification of that method (zinc chloride being substituted for
copper oxide), we proceeded to submit to precisely similar treatment
with these two reagents portions of solutions obtained by dissolving,
with total exclusion of air (in the manner described in our last
memoir and the present note), the residue furnished by special
experiments with P., E. L. G., and F. G. powders, exploded under the
usual conditions obtaining in our researches, and in quantities rang-
ing from 4200 to 35,000 grs. (272-2 grms. and 2268 grms.). The
following is a tabulated statement of the results obtained by the two
modes of treatment, and of the differences between the proportions
of hyposulphite obtained by treatment of portions of one and the same
residue with the two different reagents under conditions as nearly
alike as possible : —
Table 1.
Amount of hyposulphite furnished
Quantity used.
by 100 parts of powder with
No. of
experiment.
Description
of powder.
Density
of
employment of—
cliarge.
Grains.
Grams.
Zinc
chloride.
Copper
oxide.
Difference.
245
P.
3,396
220-05
0-3
-12
1-93
1-81
241
P.
5,660
366-76
0-5
-07
2-46
2-39
246
R. L.G.
4,200
272-16
0-4
•05
1-43
1-38
244
R. L. G.
5,250
340-19
0-5
•06
1-58
1-52
243
F.G.
4,523
293-41
0-4
-07
1-56
1-49
242
F.G.
6,300
408-23
0-6
-27
2-26
1-99
247
P.
35,000
(5 lb.)
2,267-97
0-23
-78
2-82
2-04
For purposes of comparison, we subjoin a statement of the lowest
proportions of hyposulphite furnished by 100 parts of the three
powders used in our general series, and also the proportions, similarly
expressed, which were obtained in the experiments with sporting and
mining powder, the residues of which were dissolved with the same
special precautions adopted in the case of the experiments given in
Table 1.
Table 2.
No. of
experiment.
Nature of
powder.
Amount of hyposulphite in
100 parts of gunpowder used.
Remarks.
7
44
17
196
194
Pebble
R. L. G.
F.G.
(Curtis and
\ Harvey
/ No. 6
f Mining
\ powder
2-06
1-75
3-04
V 2-28
1 2-77
] Lowest proportions fur-
y nished by the respective
J powders.
Special precautions taken
in collecting the residue.
RESEARCHES ON EXPLOSIVES 319
In reference to the foregoing numerical statements, we have to
offer the following observations : —
1. Substituting zinc chloride for copper oxide as the precipitant
of the sulphur which existed in the form of sulphide in solutions of
powder-residues to which air had not had access at all until the time
of its treatment with the zinc chloride, the amount of hyposulphite
existing in solution after such treatment was found to range from
0'05 to 0*78 in 100 parts of gunpowder, while the treatment of por-
tions of the same solutions with copper oxide, in the precise manner
adopted in our series of experiments, yielded proportions ranging
from 1-43 to 2'82 per 100 of powder used. Comparing the results
furnished by the two modes of treatment, it will be seen that in the
case of the parallel experiment (Experiment 246), which exhibited
the least considerable difference in the amount of hyposulphite
found, that existing after the copper oxide treatment was about
twenty-eight times greater, while in the case of the highest differ-
ence (Experiment 241) it was about thirty-four times greater than
that found after the treatment with the zinc chloride.
2. It would appear from these results that, in four or five out of
seven experiments, no hyposulphite, or at any rate only minute
quantities, existed in the residues previous to their solution, and
although it would seem to have existed in very appreciable amount
in two out of seven residues, the highest proportion found after the
zinc chloride treatment was less than one-half the lowest proportion
found in our complete series of analyses in which the copper oxide
treatment was adopted.
3. A comparison of the results among each other leads, therefore,
to the conclusion that potassium hyposulphite cannot be regarded as
a normal constituent of powder-residue (obtained in experiments such
as those carried out by us), and that M. Berthelot is correct in
regarding this salt as an accidental product, which, if existing
occasionally in appreciable amount in the solid matter previous to its
removal from the explosion-vessel, is formed under exceptional con-
ditions, and then only in comparatively small proportions.
While submitting this as the conclusion to be drawn from our
most recent experiments, we are of opinion that the following points
deserve consideration in connection with the question whether hypo-
sulphite may not, after all, occasionally exist, as the result of a
secondary reaction, in comparatively large proportion in the explosion-
vessel before the residue is removed.
It will be observed that although the copper oxide treatment.
320 RESEARCHES ON EXPLOSIVES
when applied to the sulphide in the pure condition {i.e., undiluted
with the other potassium compounds found in powder-residue), gave
rise to the production of very large proportions of hyposulphite, when
polysulphides were used, the highest proportion of that substance
found, after the treatment of the particular residues used in the
experiments given in Table 1, only amounts to 2'82 per cent, upon
the gunpowder (pebble-powder) employed, which corresponds to
about 14'5 per cent, of the average proportion of monosulphide exist-
ing in the residue furnished by that powder. In observing this, it
must be borne in mind that the sulphide existing in powder-residue
is always present, in part, and sometimes to a considerable extent,
in the form of polysulphide, also, that the experiments with the sul-
phides were conducted precisely according to the method pursued in
the treatment of the powder-residues. It would appear, therefore, as
though the mixture of the sulphide with a very large proportion of
other salts in solution rendered it less prone to oxidation by the
copper oxide than when the undiluted sulphide is submitted to its
action.
In comparing with the results furnished by the zinc chloride those
obtained by the copper oxide treatment, in the special experiments
given above, it is observed that, omitting one exceptional result
(Experiment 241), for which we do not attempt to account, the
highest proportions of hyposulphite are furnished by those residues
which also gave the highest with the zinc chloride, the differences
])etween the results furnished by the two treatments being likewise
the highest in these three cases ; so also the lowest proportions fur-
nished by the copper oxide treatment correspond to the lowest
obtained with the zinc oxide, and the differences between the results
furnished by the two methods are in the same manner the lowest in
these. It would almost appear, therefore, as though the existence of
a very appreciable proportion of hyposulphite in the solution of the
residue had some effect in promoting the production of hyposulphite
when the residue is submitted to treatment with copper oxide.
In a recalculation of the results of our analyses of the powder-
residues upon the assumption that the whole of the hyposulphite
obtained existed originally as monosulphide, it is found that, in
several instances in which the proportion of hyposulphite was high,
the analytical results are much less in accordance with each other
than when it is assumed that the hyposulphite found, or at any
rate a very large proportion of it, existed as such in the residue
before removal from the explosion-vessel. Thus, taking the F. G.
RESEARCHES ON EXPLOSIVES 321
series, in which the mean quantity of hyposulphite actually found
is about double of that obtained either from the pebble or R L. G.
powders, selecting from this series the three experiments which gave
the highest proportions of hyposulphite, and calculating in the
manner described in our first memoir the total quantities both of
solid and gaseous products ; first, from the basis of the analysis of the
solid products ; secondly, from the basis of the analysis of the gaseous
products ; and, further, on the assumption that the hyposulphite found
existed as hyposulphite either as a primary or secondary product
prior to removal from the explosion-vessel, we have as follows : —
Experiment No. 40, F. G. powder. — Density, "6 ; hyposulphite
found, 18'24 per cent.
Calculated ,
solid products.
Calculated
gaseous products.
Grms.
Grms.
170-268
125-220*
172-509*
122-979
From analysis of solid products
From analysis of gaseous products
Experiment No. 42, F. G. powder. — Density, '7; hyposulphite
found, 18-36 per cent.
Calculated
solid products.
Grms.
200-191
Calculated
gaseous products.
Grms.
144-547*
200-220*
144-520
From analysis of solid products
From analysis of gaseous products
Experiment No. 47, F. G. powder. — Density, '8; hyposulphite
found, 19-95 per cent.
Calculated Calculated
solid products. gaseous products.
Grms. Grms.
From analysis of solid products . 231-652 162335*
From analysis of gaseous products . 229*392* 164-595
* Water included.
Calculating now in the same manner the quantities of solid and
gaseous products on the assumption that the hyposulphite found
was, prior to removal from the explosion-vessel, in the form of mono-
or polysulphide, we have from the same experiments : —
Experiment No. 40.
Calculated Calculated
solid products, gaseous products.
Grms. Grms.
From analysis of solid products . 157-273 133-842
From analysis of gaseous products . 168-136 122-979
X
322 RESEARCHES ON EXPLOSIVES
Experiment No. 42.
From analysis of solid products
From analysis of gaseous products
Calculated
solid products.
Grms.
185-914
Calculated
gaseous products.
Grms.
155-722
197-118
144-520
Experiment No. 47.
Calculated
solid products.
Grms.
211-462
Calculated
gaseous products.
Grms.
176-694
223-561
164-595
From analysis of solid products
Fi-om analysis of gaseous products
Lastly, we still think that the following facts, given in our second
memoir, must not be overlooked in considering the question of
possible occasional existence of considerable proportions of hypo-
sulphite, viz. : — That " the production of high proportions of hypo-
sulphite was but little affected by any variations in the circumstances
attending the several explosives (i.e., whether the space in which the
powder was exploded were great or small), excepting that the
amount was high in all three cases when the powder was exploded in
the largest space ; on the other hand, a great reduction in the size of
grain of the gunpowder used appeared to have a great influence upon
the production of hyposulphite, as when passing from a very large-
grain powder (P. or E. L. Gr.) to a fine-grain powder (F. G.). Thus,
the production of hyposulphite exceeded 20 per cent, (on the solid
residue) in only three out of nine experiments with P. powder, in
three out of ten with E. L. G-., and in seven out of nine with P. G. ;
while it was below 10 per cent, in four out of nine experiments with
P. powder, in live out of ten with E. L. G., and in only one out of
nine with P. G. powder." The experiments made with these several
powders followed in no particular order, and no circumstance existed
in connection with them to which these great differences in the
results obtained could be ascribed.
We append a recalculation of the mean results of our three series
of analysis, adding the values of the hyposulphite found, as mono-
sulphide, to the amount of sulphide actually found, and we hope to
be allowed to add to our second memoir a similar recalculation of the
whole of our analyses.
This recalculation develops (as we pointed out in our second
memoir must necessarily be the case) a more or less considerable
RESEARCHES ON EXPLOSIVES 323
deficiency of oxygen in the total products of explosion; there is,
however, in every instance, also a deficiency of hydrogen, and it may,
therefore, be reasonably concluded that the deficiencies in the total
quantities of the oxygen and the hydrogen in the powder used, which
are unaccounted for in the products found, on the assumption that
variable proportions of the total hyposulphite found actually existed
in the residues as sulphide, are properly accounted for by assuming
that the missing quantities of these elements actually existed among
the products as water, the amount of which it was obviously impos-
sible to determine.
In conclusion, we have to state that we considered it right, in
consequence of the error discovered in the method adopted for the
examination for hyposulphite, to repeat the experiments described m
our first memoir as having been made by us, with the view of ascer-
taining whether hyposulphite could exist at temperatures approach-
ing those to which the solid products of explosion were actually
subjected in the explosion-vessels in our experiments.
To test this point, we submitted, for between ten minutes and a
quarter of an hour, to the highest heat (about 1700'' Cent.) of a
Siemens' regenerative furnace, two platinum crucibles, one filled with
powder-residue, the other with potassium hyposulphite. At the
conclusion of the exposure, and wliile the crucibles were still red hot,
they were plunged into water, deprived of air by long-continued
boiHng, and at once sealed. The powder-residue was found still to
contain 1-27 per cent, of hyposulphite, while the crucible with the
pure salt consisted of a mixture of sulphate and sulj)hide, but with
an amount of 2'1 per cent, of hyposulphite.
It is probable that, if the exposure had been still longer con-
tinued, the hyposulphite would have altogether disappeared, and the
experiment can only be taken as proving that the hyposulphite,
especially if mixed with other salts, is neither quickly nor readily
decomposed, even at very high temperatures.
[Table S.
324
RESEARCHES ON EXPLOSIVES
&5 ~
1-1 »
S 5
i ^
•J81B.\i
•0095
•0111
•0148
•s^onpojd puos
JO !}q9raAi j£q uoi^odoj,!
•5496
•5593
•5569
•s^onpoad sno9SBS
JO !Hi3t8AV Xq uoi^iodoj,!
•4409
•4298
•4282
§
o
s
%
•IBOOJBUO
•0008
•0004
•Jtiqding
•0445
•0495
•0381
■8|Buoqjv'0inbs9S
ranmouitnv
■0005
•0004
•0003
•apixo
uimssB^oj
CD
9
tnntssB^od
•0013
•0015
•0009
•apmB.tooqdins
uinissTJ^oa
•0014
•0013
•0007
•spiqdinsouom
mnissB^oj
•1042
•0787
•0999
•a^Bqdius
raniBSB^od
•0710
•0844
•1252
•a^BuoqaBO
ramss'B:^0(i
■3258
•3415
•2861
!
t
1
pLi
■ubSAxq
•0002
•0003
•uaSojpXH
•0006
•0009
•0007
•SBS-qs.iBM
•0006
•0008
•0004
•piOB oijp^qqdins
•0111
•0109
•0101
•ueSomx
•1122
•1117
•1123
•opixo oitioqjBO
•0480
•0422
•0355
•opiJiiXquB
omoqaBO
•26S5
•2630
•2689
•jepAiod JO 8irnB>j
Pebble
R.L.G.
F.G. .
VII.
HEAT-ACTION OF EXPLOSIVES.
{Lecture, delivered at the Institution of Civil Eiigineers, 1884.)
Examples of explosive substances will readily occur to all of you.
The saKent peculiarities of some of the best known may roughly be
defined to be the instantaneous, or at least the extremely rapid,
conversion of a solid or fluid into a gaseous mass occupying a volume
many times greater than that of the original body, the phenomenon
being generally accompanied by a considerable development of
measurable heat, which heat plays a most important part not only
in the pressure attained, if the reaction take place in a confined
space, but in respect to the energy which the explosive is capable of
generating.
Fulminates of silver and mercury, picrate of potassa, guncotton,
nitro-glycerine, and gunpowder, may be cited as explosives of this
class.
But you must not suppose that substances such as I have just
named are the only true explosives. In these solid and liquid
explosives, which consist generally of a substance capable of being
burnt, and a substance capable of supporting combustion, in, for
example, guncotton or gunpowder, the carbon is associated with the
oxygen in an extremely condensed form. But the oxidisable and
oxidising substances may themselves, prior to the reaction, be in the
gaseous form; as, for instance, in the case of mixtures of air or
oxygen with carbonic oxide, of marsh-gas with oxygen, or of the
mixture of hydrogen and oxygen forming water, which, if regard be
had to the weight of the combining substances, forms an explosive
possessing a far higher energy than is possessed by any other known
substance.
But these bodies do not complete the list, and, under certain
circumstances, many substances ordinarily considered harmless must
be included under the head of explosives.
326 HEAT-ACTION OF EXPLOSIVES
Finely-divided substances capable of oxidation, or certain
vapours, form, when suspended in, or diluted with, atmospheric air,
mixtures which have been unfortunately the cause of many serious
explosions.
Minute particles of coal floating in the atmosphere of coal-mines
have either originated explosions, or in a very high degree intensi-
fied the effects of an explosion of marsh-gas. Flour-dust and
sulphur-dust suspended in the air have produced like disastrous
results. Lines of demarcation are generally difficult of definition,
and the line between explosive and non-explosive substances forms
no exception to the rule ; but, from the instances I have given, you
will note that an explosive may be either solid, liquid, or gaseous,
or any combination of these three states of matter.
In the course of my lecture, I propose, in the first instance, to
give you a short account of the substances of which some explosives
are composed, illustrating my meaning by giving you the composi-
tion of one or two which may be considered as types, and which are
well known to you.
I shall, in the second place, show the changes which occur when
our explosives are fired ; and shall endeavour to give you some idea
of the substances formed, of the heat developed, of the temperature
at which the reaction takes place, and of the pressure realised, if the
products of our explosive be absolutely confined in a strong enough
vessel, as well as of the experiments which have been made, and the
apparatus which has been used either directly to ascertain or to
verify the facts required by our theory.
I shall in certain cases suppose our explosives to be placed in
the bore of a gun, and shall endeavour to trace their behaviour in
the bore, their action on the projectile, and on the gun itself. I
shall, at the same time, describe to you the means and apparatus
that have been employed to ascertain the pressure acting on the
projectile and on the walls of the gun, and to follow the motion of
the projectile itself in its passage through the bore.
Let us take, suppose at the temperature 0° Cent., and at the
pressure 760 mm. of mercury, two equal volumes of the gases
hydrogen and chlorine, which when combined produce hydrocliloric
acid. I have the gases in this tube, and let us apply a light ; you
will observe that the mixture explodes violently, with considerable
evolution of heat. Now this is perhaps as simple a case of an
explosive as we can have.
If we suppose the gases to be exploded in an indefinitely long
HEAT-ACTION OF EXPLOSIVES 327
cylinder, closed at one end, and with an accurately fitting piston
working in it, and if we suppose the gases (fired, you will remember,
at 0° Cent, and atmospheric pressure) to be again reduced to the
temperature and pressure from which we started, the piston will
descend to its original position, and the gases will occupy the same
space as before they were exploded.
If we now suppose that we had, in a calorimeter, measured the
quantity of heat produced by the explosion, that quantity of heat,
about 23,000 grm. -units per gramme of hydrogen, or about 600 grm.-
units per gramme of the mixture, expresses, without addition or deduc-
tion, the total amount of work stored up in the unexploded mixture,
and from that datum, knowing the specific heat, we are able to
deduce not only the temperature at which the explosion takes place,
but the maximimi pressure produced at the moment of explosion,
and the work which the gases, in expanding under the influence of
the heat evolved, are capable of performing.
If, instead of a single volume each of hydrogen and chlorine, we
take two volumes of hydrogen and one of oxygen (which when com-
bined produce water), or by weight two parts of hydrogen and
sixteen of oxygen, and explode them as I now do, you will observe
that there is a still more violent explosion, and I may add that there
is a still greater development of heat.
If, as before, we supposed the explosion carried on in an indefi-
nitely long cylinder, the piston, on the gases being brought back to
the temperature and pressure existing before the charge was fired,
would no longer stand at its original height, but at two-thirds of
that height, the three volumes would be condensed into two, and the
heat determined by our calorimeter, about 29,000 grm.-units per
gramme of hydrogen, about 3300 grm.-units per gramme of the gaseous
water produced by the explosion is increased above what may be
considered the true heat of the explosion by the condensation
which the aqueous vapour has suffered in passing from three to two
volumes.
From the heat determined, however, we are able as before to
deduce the temperature of explosion, the pressure exerted on the
walls of a close vessel at the instant of maximum temperature, and
the energy stored up in the exploded gases.
I have mentioned that the potential energy stored up in this
mixture of hydrogen and oxygen is, if taken with reference to its
weight, higher than that of any other known mixture, and it may
fairly be asked why should such an explosive, whose components
328 HEAT-ACTION OF EXPLOSIVES
are so readily obtainable, not be more largely employed as a pro-
pelling or disruptive agent ?
There are several objections ; but you will readily appreciate one
when I point out that if we assume a kilog. of gunpowder forming a
portion of a charge for a gun, to occupy a litre or a decimetre cubed,
a kilog. of hydrogen, with the oxygen necessary for its combustion,
would at zero and at atmospheric pressure occupy a volume sixteen
thousand times as great.
Let us now pass to guncotton, known also as pyroxyhn or trinitro-
cellulose. This substance, as you probably know, is prepared by
submitting ordinary, but carefully purified, cotton to the action of a
mixture of concentrated nitric and sulphuric acids at ordinary
temperatures, where a proportion of the hydrogen in the cellulose is
replaced by an equivalent amount of nitric peroxide.
Nitro-glycerine is in like manner formed by the action of a
mixture of nitric and sulphuric acids on glycerine ; but we shall for
the present confine our attention to guncotton.
The formula representing guncotton is C6H73(NO.,)05, and gun-
cotton itself may be employed in several forms in the flocculent or
natural state ; or it may be made up into strands, yarns, or ropes ; or
it may be granulated or made into pellets ; or it may be highly com-
pressed into slabs or discs, in which last form it is almost invariably
used for industrial or military purposes, and for which we are so
largely indebted to the labours and researches of my friend and
colleague. Sir Frederick Abel.
Samples of all these forms are on the table before you.
When guncotton is fired, practically the whole of its constituents,
which before ignition were in the solid, assume the gaseous form,
and this change is accompanied by a very great development of heat.
I now fire a train of different forms of guncotton, and you will note,
in the first place, the small quantity of smoke formed, and this may
be taken as an indication of the small amount of solid matter in the
products of combustion. You will observe, also, that instead of the
explosions which took place when our gaseous mixtures were fired,
guncotton appears rather to burn violently than explode. This,
however, is due to the ease with which the nascent products escape
into the atmosphere, so that no very high pressure is set up.
Were we, by a small charge of fulminate of mercury or other
means, to produce a high initial pressure, the harmless ignition that
you have seen would be converted into an explosion of the most
violent and destructive character.
HEAT-ACTION OF EXPLOSIVES
329
You will finally note that this transformation differs materially
from those which we have hitherto considered. In both of these
the elements were, prior to the ignition, in the gaseous state, and
the energy liberated by the explosion was expressed directly in the
form of heat. In the present instance, a very large but unknown
quantity of heat has disappeared in performing the work of placing
the products of explosion in the gaseous state.
Let me try to show you how large an amount of heat may be
absorbed in the conversion of solid matter into the gaseous state.
You are aware that if a gramme of carbon be burned to carbonic
anhydride there are about 8000 grm. -units of heat evolved, whereas
if a gramme of carbon be burned to carbonic oxide, there are only
evolved about 2400 grm.-units. Now a priori we may certainly
suppose that the assumption by the carbon of the two atoms of the
oxygen should result in equal developments of heat, but you will
note, from what I have stated, that in the combination with the
second atom of oxygen about two and a third times more heat is
developed. Whence, then, comes the difference, and where has the
heat disappeared which our calorimeter declines to measure ? The
missing heat may be assumed to have disappeared in performing the
work of placing the solid carbon in the gaseous state.
In the case which we have been considering, the oxygen which
supports the combustion of the carbon is already in the gaseous
state ; but with guncotton all the gases are, prior to combustion, in
the solid state. Their approximate w^eights are exhibited in the
following table : —
Composition.
Products of Explosion.
Carbon
24-89
Carbonic anhydride .
,, oxide .
0-424
Hydrogen
2-69
0-280
Nitrogen .
13-04
Hydrogen
0-011
Oxygen .
.56-66
Nitrogen .
0-145
Ash . . . .
0-36
Marsh-gas
0-003
Moisture .
2-36
Water
0-116
Formula— CfiH^SCNOoPg.
Original moisture
0-021
Carbonic oxide and anhydride, nitrogen, hydrogen, aqueous
vapour, and a little marsh-gas, are the products of explosion, and
their quantities are such that a kilog. of guncotton, such as that
with which Sir F. Abel and I have each made so many experiments,
will produce, when the gases are reduced to atmospheric pressure and
to a temperature of 0° Cent., about 730 litres. In this volume the
water produced by the explosion is not included, being at that
temperature and pressure in the liquid form. In estimating either
330 HEAT-ACTION OF EXPLOSIVES
the pressure exerted on the walls of the close vessel, or the potential
energy of the gimcotton, we have to add to the work done, that is,
to the heat absorbed by the great expansion from the solid state
into the number of volumes I have indicated, the potential energy
due to the heat at which the reaction takes place.
As might be expected from the definite nature of the chemical
constitution of guncotton, the constituents into which it is decom-
posed by explosion do not very greatly vary ; the chief point to be
observed being that the higher the tension at which the explosion
occurs, the higher is the quantity of carbonic anhydride formed, that
is, the more perfect is the combustion.
Gunpowder, the last and most important example I shall select, is
also by far the most difficult to experiment with, as well as the most
complicated and varied in the decomposition which it undergoes.
To begin with, it is not, like guncotton, nitro-glycerine, and other
similar explosives, a definite chemical combination, but is merely
an intimate mixture, in proportions which may be varied to a con-
siderable extent, of those well-known substances, saltpetre or nitre,
charcoal, and sulphur ; and in this country the proportions usually
employed are 75 parts of saltpetre, 10 of sulphur, and 15 of charcoal.
They do not during manufacture undergo any chemical change, and it
is perhaps owing to this circumstance that gunpowder has for so many
generations held its place as the first and principal, indeed almost the
only, explosive employed for the purposes of artillery and firearms.
One great advantage for the artillerist which gunpowder possesses
in being a mixture, not a definite chemical combination, is that
when it is fired it does not explode in the strict sense of the word.
It cannot, for example, be detonated as can guncotton or nitro-
glycerine, but it deflagrates or burns with great rapidity, that
rapidity varying largely with the pressure under which the explosion
is taking place. As an instance of the difference in the rate of
combustion due to pressure, we have found that the time necessary
for the combustion of a pebble of powder in free air is about two
seconds. The same pebble in the bore of a gun is consumed in
about the aw part of a second ; but a more striking illustration of
the effect of pressure in increasing or retarding combustion is shown
by an experiment devised by Sir F. Abel, and which by his kindness
I am able to repeat. It consists in endeavouring to burn powder
in vacuo, and you will see for yourselves the result of the experi-
ment. The powder refused to explode.
But although the composition of gunpowder is in this country
HEAT-ACTION OF EXPLOSIVES
331
approximately what I have said, the requirements or experiments
of the artillerist have for certain purposes modified in a high degree
both the constituents and the physical characteristics of gunpowder.
In the following table are exhibited the composition of the
numerous powders with which Sir F. Abel and I have experimented ;
and the samples which I have upon the table, many of which will
be new to some of you, illustrate the irregular forms into which we
mould the mixture, which by a misnomer we still call gunpowder.
Here you see the forms with which all are familiar, and which
are called fine grain and rifled fine grain. Here, a little larger, you
see rifled large grain, which at the introduction of rifled guns was the
powder then used. Here these small lumps are called pebble-powder,
and this powder is that generally used in this country with rifled
guns of medium size. Here is a still larger size of service pebble.
Table 2.-
-Showing the composit
Ion of various gunpowders.
u
25
o
I.
<
<
6
;
^
^
^;
S
^
£
1
M
1
1
i
6
hi
6
.1
1
i
•6192
Saltpetre .
•8130
•7783
•6374
•7724
•7883
•7476
•7456
•7391
•7559
•7468
Sulphur .
■0018
•0028
•1469
•0615
•0204
•1007
•1009
•1002
•1242
•1037
•1506
Charcoal .
•1671
•1972
•2018
•1543
•1780
•1422
•1429
•1459
•1134
•1378
•2141
Water .
•0181
•0217
•0139
•0118
•0133
•0095
•0106
•0148
•0065
•0117
•0161
This form, prismatic, differing from the others both from its
regular shape, and from the hole or holes traversing the prisms, is
perhaps the most convenient form in which powder can be made up
in large charges, while these blocks exhibit still larger masses,
representing powders which have been used with success in very
large guns. The object of the holes in the prismatic and other
powders, is to obtain more uniform production of pressure to ensure
the more complete combustion of powder by increasing the burning
surface, as the prism is consumed, and consequently diminishes in size.
I draw your particular attention to these samples, because I
shall have, before I conclude, something to say about them. You
will observe that they are in the prismatic form, and that they differ
from the other prisms, with which you can compare them, in being
brown in colour instead of black.
Let us now apply a light to trains of different natures, and to
some other samples of powder — experiments which I daresay at one
332 HEAT-ACTION OF EXPLOSIVES
time or another you have made for yourselves — and observe the result.
You will note, in the first place, that an appreciable time is taken by
the flame to pass from one end to the other ; but you will also note
an essential difference between this combustion and that I showed
you a short time ago with guncotton, viz., that there is a large quan-
tity of what is commonly called smoke slowly diffusing itself in the air.
Now this so-called smoke is really only finely-divided solid
matter existing as a fluid, or volatilised only to a very slight extent
at the moment and temperature of explosion, and if, adopting means
which I shall presently describe to you, we had exploded in a close
vessel the powder which we have just burned in the air, and allowed
the vessel to stand for a few minutes, the products would be divided
into two classes — one, a dense solid, generally very hard, and always
a disagreeably smelling substance ; the other, colourless gases, the
odour of which is, I must confess, not much more fragrant than that
of the solid matter to which I have referred.
These large bottles on the table contain a portion of the so-called
smoke of a charge of 15 lbs. of powder, collected in the manner I
have described, in a closed vessel. You will see it is a very solid
substance indeed ; but as these products are sometimes very protean
in their characteristics, I have upon the table one or two other
specimens of these residues differing considerably in appearance.
I have also in this steel vessel the products of combustion of
2 lbs. of powder. I shall not now let the gases escape ; but after
the lecture shall be glad to do so for the benefit of those who have
no objection to a disagreeable smell.
If the gases produced by the combustion be analysed, they will
he found to consist of carbonic anhydride, carbonic oxide, and
nitrogen, as principal constituents, with smaller quantities of
sulphhydric acid, marsh-gas, and hydrogen, with — this point depending
much on the constitution of the charcoal — always small quantities,
and occasionally considerable, of aqueous vapour.
The solid substances are found to consist of, as principal ingredi-
ents, variable quantities of potassium carbonate, sulphate, and
sulphides, with smaller quantities of sulphocyanate, and ammonium
sesquicarbonate.
The annexed table shows by weight the products of combustion
in the different powders examined by Sir F. Abel and myself, and I
call your special attention to the considerable variations in the
decomposition of powders which are intended practically to have the
same chemical constitution.
HEAT-ACTION OF EXPLOSIVES
333
4
•enpisaj pnog
•5812
•5773
•4692
•5591
•5825
•5496
•5591
•5569
•6127
•5774
•4201
•s^onpoad snoes^o
•4187
•4216
•5307
•4409
•4175
•4409
•4298
•4282
•3808
•4109
•5135
3
s
1
1
s
■s
•()n8;S!X8-8jd JQ^BAi
r-l t- Oi CO CO O CD CO lO I- r-l
Sj^Sn^oSi^loI^S
•jnitdins
•0627
•0352
•0445
•0490
•0381
•0431
•0372
•0664
■8^BnoqjBombs9S ranmouimv
OOOi-lCD vO-HCO(71in-l'
OC100 -oooooo
oooo oooooo
•apixo mnissB-joj
::.::::§:::
•a^Bj^m mnissB^od;
SSgg^S
■■■■■■■■ :§§8§§§
•a^BUBXooqdins ranissB^oj
■0187
•0005
•0014
•0013
•0007
•0003
•0139
•apiqdinsonoui runissBC^o,!
•0034
•0006
•1646
•0382
•1042
•0807
•0999
•0473
•0717
•1745
•e^iUdtnsodXq ratnissB^oj
•0048
•0008
•0073
•0029
•s;Biid[ns uinissB^oa
•0017
•0006
•0042
•0230
•1332
•0710
•0844
•1252
•2975
•1250
•0028
•9:)Buoqi'BO umissB^oj
•5474
•5296
•2036
•4579
•4360
•3258
•3415
•2861
•_'186
•3413
•1945
1
ea
O
S
■s
•j8;BAi
•0225
•0026
•0832
•uaS/CxQ
ooo
•neSoip^H
7, . ,-, rH CD CD Oi t^ CO CO l^
§§§g§§§§§ig
•SBS-qsJBH
lllllll 11
•piov oijp^uqdins
•0191
•0091
•01 11
•0109
•0101
•0096
•0083
•0389
•u8.goj'}!jv[
•1158
•1200
•0875
•1188
•1049
•1123
•1117
•lliiS
•1108
•1132
•0858
•apixo oiaoqjBQ
Oi .* in 03 CD t^ w in CD I- .71
(71 1— 1 i-l (7) oo i:^ (7) in CO ^ (M
•apupjCiinB oraoqjBO
•2253
•1915
•2723
•2467
•2198
•2685
•2630
•2689
•2457
•2593
•2279
Powder " A "
Cocoa-powder
Pebble, W.A., means.
R.L.G., W.A., means
F.G., W.A., means .
Spanish
C. & H. No. 6 .
Mining
334
HEAT-ACTION OF EXPLOSIVES
Considerations such as are suggested by this table led Sir F. Abel
and myself to make a statement which has been somewhat misunder-
stood, and which has been the subject of a good deal of controversy,
viz., that, except for instructional purposes, but little accurate value
can be attached to any attempt to give a general chemical expression
to the metamorphosis of a gunpowder of normal composition.
Now by this statement, to which, after many years of research,
we most emphatically adhere, we did not mean to say that, given
precisely the same conditions, the same products would not follow ;
but we did mean to say that the circumstances under which gun-
powder, nominally of the same composition, may be exploded, are so
varied — the nascent products may find themselves under such varied
conditions both as to pressure, temperature, and the substances with
which they find themselves in contact — this last point depending much
on the physical characteristics of the powder — that it is not wonder-
ful if considerable variations in the products ensue.
I need only refer in illustration of my remarks to the very
interesting decomposition experienced by cocoa-powder. Observe the
very small quantity of carbonic oxide, and the large quantity of water
formed, while the solid constituents are reduced in number to two.
Let me now call your attention to another point. The table
giving the decomposition of gunpowders shows also the ratio between
the weights of the solid and gaseous products ; but it is necessary
that we should know the volume of the gases at ordinary tempera-
tures and pressure. A kilog., then, of these powders, at a tempera-
ture of 0" Cent, and a barometric pressure of 760 mm., would give
rise to the following quantities of gases : — The numbers in the table
expressing litres per kilog., or c.c. per gramme of jDowder exploded.
Table 4. — Shoxoing the, volumes of permanent gases evolved by the combustion
of 1 (jramme of the undermentioned fowders.
^
a
<
-
IS
^
y^
1
1
•2
1
1
1
i
6
2
d
.1
1
d
c
Vols, of gases .
254
315
347
282
198
278
274
263
234
241
360
That is to say : — Assuming that a kilog. of each of these powders
occupied a decimetre cubed, the figures in the table represent for each
description of powder the number of similar volumes occupied by the
liberated gases when at the temperature and pressure I have named.
HEAT-ACTION OF EXPLOSIVES
335
I have, in the case of each explosive that I have described, given
to you the number of heat-units produced by the explosion. Follow-
ing the same course with these powders, the number of grm.-units
of heat evolved by the combustion of a gramme of each of the
powders with which we have experimented is given in Table 5.
Table 5. — Showing the units of heat evolved hy the combustion of 1 gramme
of the undermentioned powders.
'
/]
\\
? 1
■ft
/
k
( /
'
P
^
?' K
H
t
^
^ ^
i<
^
/
i|
'^
/
^ ^
Vl
^
'
/
A
^ ^..
§
'^
/
/
\
//
\
/
\
^
\
\
/
/
X
•^
/
^i>'
>P<
X
':^
k!^
^v
(
■
^.
^-T
r.rz
~-
.
_1
^^^
>^^
[To face p.
PLATE XXII.
1
».■
^
i
f
.■^
^
?
i
'•^
'':
/
1
•! "^
i .C
^
1
^
{^ ^
^
lO
/
^
Os
r,^
5
§
"*
IS
-0
0"
/
>.
^
11
/ /
5
5;
'i
^
^
^
^
/
1
t
""l
^
r
/
: .C
1
^
ft
A
■ s"
^
^
1 ~s
■^
<5
/
7
i
; is
^
N
/
/
^0
^ '
1 ^
f:
'
-f-
/
i
^
J r-H ^ O-J 0 CD 00 lO l-^ i-t
S^CO.-HCOaiO-45 ;2; ?2! 5" S S!
§§§§ =§§8§§§
•apixo rantssB^oa
\Ci . . .
:::::::§:::
•8:}w:jm ranissBioj
CO >c 05 oo t^ ^
■ ■ ■ ■ '999999
"'
•a^BtrejCoondins
ranissE^oj
s
.^
■apiqdinsonora
' mnissB^Od:
r# .a^ CD (M (M i^ 03 CO 1^ m
COO'tflOO .^OOit^i— 1^
1
•acnqdinsodi^il
OOTOCOOi
llll :::::: :
1
•8;Bi[dins umissB^oa
t-MO(MO^(MlOOOO
oooc-iSoricoe^osiMO
OOOOi— lOOi— iC^li-HO
?^
•8^Bnoq.reo mnissi^'jo.i
'*(MO>OC0(M'*00i-4.^05
»OlO(MM<^COCOC0(^^
CI .^H (>i (>i 91 c^i o^ 91 C-1 (N c^
r r s :: S^-<: . . • .
INTERNAL BALLISTICS 407
appreciable time is taken by the flame to pass from one end to the
other ; but you will also note that there is a large quantity of what
is called smoke slowly diffusing itself in the air.
Now this so-called smoke is really only finely-divided solid matter
existing as a fluid, or volatilised only to a very slight extent at the
moment of explosion ; and if the powder you have just seen fired had
been exploded in larger quantity in a close vessel such as I have
described, nearly 60 per cent, of the weight of the powder would have
lieen converted into the so-called smoke, and when the products had
cooled would have been found at the bottom of the cylinder in the
shape of a dense, hard, evil-smelling substance, generally very
difficult of removal, with a smooth, dark surface, and an olive green
fracture.
In the bottle which I hold in my hand, I exhibit to you a portion
of the so-called smoke of a charge of 15 lbs. of powder fired in a close
vessel in the manner I have mentioned.
I need hardly call your attention to the magnitude of the charge
which has thus been entirely confined. At the date of the Crimean
war the highest charge of the 56 cwt. 32-pr., the principal heavy gun
of the service, was only 10 lbs., but I have fired and succeeded in
absolutely retaining in one of these vessels a charge of no less than
23 lbs.
The principal constituents of the solid residue are potassium
carbonate, potassium sulphate, and potassium sulphide, with small
quantities of the other substances you see in Table 2. There are
considerable fluctuations in the proportions of the principal con-
stituents, two charges fired as nearly as possible under the same
circumstances frequently differing more in the products of decom-
position than do others in which both the nature of powder used and
the gravimetric density at which they are fired have been changed.
For these fluctuations it is difficult to assign any cause, unless it be
that in a combination and decomposition of such violence, the nascent
products find themselves in contact sometimes with the products of
explosion, sometimes with powder not yet consumed.
But it may interest you to know the appearances presented by
the solid products when cool, and after the opening of the cylinder in
which the powder was exploded. The whole of the solid products
were usually found collected at the bottom of the cylinder, there
being but an exceedingly thin deposit on the sides. The surface of
the deposit was generally quite 'smooth, and of a very dark grey,
almost black, colour; this colour, however, was only superficial, as
408 INTERNAL BALLISTICS
through the black could be perceived what was probably the real
colour of the surface, viz., a dark olive green. The surface of the
deposit and the sides of the cylinder had a somewhat greasy appear-
ance, and were, indeed, greasy to the touch. When the charge was
large and the confined gases at a high pressure were allowed to escape
rapidly, the surfaces, especially in the vicinity of the point of escape,
were covered with a deposit of solid carbonic acid, this deposit arising
from the cooling effect due to the rapid expansion.
In cases where the gas had escaped before the deposit was cold,
the surface was rough, and the deposit somewhat spongy, as if
occluded gas had escaped while the deposit was still in a semi-fluid
state. In various experiments, on examining the fracture as exhibited
by the lumps, the variation in physical appearance was very striking,
there being differences in colour and texture, and also frequently a
marked absence of homogeneity, patches of different colour being
interspersed.
There was no appearance of general crystalline structure in the
deposit, but shining crystals of sulphide of iron were frequently
observed. The deposit had always a powderful odour of sulphuretted
hydrogen, and frequently smelt strongly of ammonia. It was always
extremely deliquescent, and small portions, after a short exposure to
the air, became black, gradually passing into the inky-looking, pasty
substance familiar to you all as resulting from the residue left in the
bores of guns after practice.
As in physical appearance, so in behaviour, when removed from
the cylinder, the solid products presented great differences. In most
cases, during the short period that elapsed while the deposit was
being transferred to thoroughly dry and warm bottles, no apparent
change took place, but in some a great tendency to development of
heat arising from the absorption of oxygen from the air was apparent.
In one case where a deposit exhibited this tendency to heat in a high
degree, a portion was ground, placed in the form of a cone on paper,
and observed. The action proceeded very rapidly, the deposit on the
apex and in the interior, where there was greatest heat, changing
rapidly in colour to a light sulphury yellow, with a tinge of
green.
During the development of heat, the residue gave off a good deal
of vapour, and an orange-coloured deposit, probably resulting from
this vapour, formed on the surface. The staell was very peculiar,
sulphuretted hydrogen being distinctly perceptible, but being by no
means the dominant odour.
INTERNAL BALLISTICS 409
The maximum temperature occurred at about twenty minutes
after the commencement of the exposure — a thermometer placed in
the centre showing a temperature of over 600° Fahr. The tempera-
ture was no doubt somewhat higher, but the thermometer had to be
removed, to avoid fracture. The paper on which this deposit was
placed was entirely burned through.
From an examination of the cylinder when opened after an
explosion, it was easy to see that the solid products had been in a
fluid state; but to ascertain the state of the contents at different
periods, the following experiments were made. The cylinder, being
about two-thirds full of powder, was fired, and 30 seconds after
explosion was tilted so as to make an angle of 45°. Two minutes
later it was restored to its first position. On opening, the deposit
was found to be lying at the angle of 45°, and the edges of the deposit
were perfectly sharp and well-defined.
Again, the cylinder being about three-fourths filled with powder,
was fired, and allowed to rest for 1 minute after explosion. It was
then placed sharply at an angle of 45°, and 45 seconds later was
returned to its first position. Upon opening, it was found that when
the cylinder was tilted over, the deposit had just commenced to
congeal, for upon the surface there had been a thin crust, which the
more fluid deposit underneath had broken through. The deposit was
at an angle of 45°, but the crust through which the fluid had run
was left standing like a sheet of ice.*
Another experiment with the vessel completely full of powder
showed that at a minute and a minute and a quarter after explosion
the non-gaseous products were still perfectly fluid, and that it was
nearly 2 minutes before their mobility was destroyed ; and my con-
clusion from the whole of the experiments is, that immediately after
explosion the non-gaseous products are collected at the bottom of the
vessel in a fluid state, and that some time elapses before the products
assume the solid form.
The existence of this fluid residue in the bore of a gun is some-
times clearly shown by the occurrence of large splotches of residue,
frequently close to the muzzle, and which indicate that considerable
masses of the residue, travelling at a high velocity, had been arrested
* Note.— In consequence of this action, in later experiments the deposit was
not removed by chisels, but distilled water, freed from air by long-continued
boiling, was siphoned into the explosion-vessel, so that air was never allowed to
come into contact with the sohd residue— when the cylinder was thus quite filled
with water, it was closed, and allowed to stand until the residue was completely
dissolved.
410 INTERNAL BALLISTICS
by striking the sides of the bore. In the chambers of guns, again,
considerable masses are frequently found, the residue having evidently,
while in a fluid state, run down the sides and collected at the bottom
of the chamber. In the 100-ton gun chamber masses of about three-
quarters of an inch in thickness have been found. One of these
specimens is before you.
Turn now to the gaseous products. These do not exhibit the
variations shown by the solid products; on the contrary, if the
powder be of similar composition, as, for instance, in the case of the
Waltham- Abbey powders, the gases are remarkably uniform in com-
position. In weight they amount to about 43 per cent, of the
unexploded powder, and consist chiefly of carbonic anhydride, nitrogen,
carbonic oxide, and sulphuretted hydrogen, with small quantities of
marsh-gas and hydrogen. The proportion of carbonic acid was found
slightly, but decidedly, to increase as the gravimetric density of the
charge was increased; this, of course, corresponding with increased
pressure in the explosion-vessel, and pointing, under this condition,
to a more perfect oxidation of the carbon.
The quantity of permanent gases generated by explosion difiers
very considerably with the nature of the powder, and even with the
size of grain. Thus the quantity of gas generated by a gramme of
dry pebble-powder was found to be 278-3 c.c. ; by a gramme of E. L. G.,
274-2 ; and by a gramme of F. G., 263-1. All the above volumes are
reduced to the standard barometric pressure of 760 mm., and the
temperature of 0'^ Cent.
I ought perhaps to explain, that the statement that a gramme of
powder generates so many cubic centimetres is equivalent to the
assertion that, at the temperature and pressure stated, the gases
occupy the same number of times the volume that the powder
occupied in the unexploded state, the gravimetric density of the
unexploded powder being supposed to be unity.
You will observe that there is an appreciable difference in the
volume of the permanent gases generated by Waltham- Abbey pebble-
powder and F. G., two powders which are intended to be of precisely
the same composition, and which in reality differ but slightly. But
if I take some other powders I have experimented with, you will find
that the differences in the volumes of the gases produced are very
striking. Thus 1 grm. of Curtis & Harvey's well-known No. 6
powder generated 241 c.c, 1 grm. of English mining 360-3 c.c, while
1 grm. of Spanish pellet generated only 234-2 c.c.
Table 3 shows the volumes of permanent gases evolved by the
INTERNAL BALLISTICS
411
combustion of 1 grm. of the powders whose composition was exhibited
in Table 1.
Table 3. — Shovnncf the volumes of permanent gases evolved hy the comhustion of
1 gramme of the undermentioned powders.
5
1
(2
1
Q
i
8
<
6
hi
<
i
■s
''A
S
Volumes of gases .
254
315
347
282
198
278
274
263
234
241
360
Observe, now — for I shall shortly have occasion to draw your
attention to the point — the arrangement of these six last powders on
the list. If I place them in ascending order of magnitude with
respect to the volumes of gas they respectively generate, first we have
the Spanish pellet with 234 volumes, next comes the Curtis & Harvey
with 241 volumes, then F. G. with 263, and so on, while mining-
powder with 360 volumes closes the list.
You will remember I have explained to you what I mean by the
expression " quantity of heat." All the powders in Table 1 have, by
carefully conducted calorimetric experiments, had the number of
units of heat they were capable of evolving carefully determined, and
the results of these determinations are given in Table 4.
Table 4.
—Showing the units of heat evolved hy
the combustion c
f
1 gramme of the undermentioned powders.
<
m
d
«■
<
pi
<
6
f5
fe
.^
S
a
oT
"
t?
■5
W
ti:
'i
•%
5
&
1
3
^
o
■^
■=
^
^
s
^
S
rt
fa
m
"
s
Units of heat .
800
715
525
745
837
721
726
738
ni
764
517
As in the case of the quantity of gas, so with the heat evolved
there is a great variation, but with this peculiarity, that the powders
producing the largest quantity of gas evolve the least quantity of
heat. Take, for example, the six last powders on the table to which
I recently drew your attention. You will remember that when I
gave you the volumes of gas generated by the different powders, I
arranged the powders in an ascending order of magnitude. Now
that I give you the quantities of heat evolved by the same powders.
412
INTERNAL BALLISTICS
I arrange them in descending order, and I want you to observe, as
you will see from Table 5, that not only do the same powders head
and close the list, but that the order of arrangement taken from the
two sets of data is absolutely identical.
Table 5.
Nature of powder.
Units of heat
per gramme exploded.
Cubic centimetres
of gas
per gramme exploded.
Spanish pellet .
Curtis & Harvey's No. 6 .
W.A.F.G.
W.A.R.L.G. .
W.A. pebble .
Mining ....
767-3
764-4
738-3
725-7
721-4
516-8
234-2
241-0
263-1
274-2
278-3
360-3
Observe also, that although the mining-powder generates about
50 per cent, more gas than does the Spanish spherical, on the other
hand about 50 per cent, more heat is generated by the Spanish than
by the mining-powder. As a matter of fact, the products of the
quantities of heat multiplied by the volumes of gas generated (which
may be taken approximately as a measure of the potential energy
stored up) do not differ very greatly from a constant quantity, and
as a further matter of fact, the pressures developed by the powders at
various gravimetric densities are very much the same, and this
circumstance is remarkable when the variety in the composition of
the powders and the decomposition which they experience is taken
into account.
But another question here arises, and that is, are we able in any
way to account for the great difference in the quantity of heat
measured when two powders, such as mining and Spanish, are
exploded ? I believe we are, and I shall endeavour to make my
meaning intelligible.
You are all aware that when, for example, ice at 0" Cent, is
converted into water at 0" Cent., or when water at 100" Cent, is
converted into steam at 100" Cent., a large quantity of heat has to
be commvmicated to the ice or water as the case may be, — and as this
heat produces no effect on the thermometer, it has received the name
of latent heat.
But the modern theory of heat — I need not detain you with an
explanation of this theory — has shown that the heat which was
supposed to have become latent has really disappeared in performing
work of one sort or another — in doincj work ag-ainst molecular forces,
INTERNAL BALLISTICS 413
or in communicating motion to the molecules of water. In placing
a gramme of water at 100° Cent, in the form of steam at 100" Cent.,
no less than 537 units of heat are absorbed.
Again, you all probably know that when a gramme of carbon
unites with a single equivalent of oxygen, the gas carbonic oxide is
formed ; and that when a second equivalent of oxygen is taken up,
carbonic anhydride, or in the old nomenclature carbonic acid, is
formed. But the quantities of heat generated when carbon burns
to carbonic oxide, and when carbon or carbonic oxide burn to
carbonic acid, are well known, and it appears that while the union
of one gramme of carbon with an equivalent quantity of oxygen
burning to carbonic oxide gives rise only to about 2445 units of heat,
the assumption of the second equivalent quantity of oxygen gives
rise to 5615 units, or to 8060 units in all.
Now I think we may regard it as certain that the great difference
indicated in the heat shown by the figures I have given you is due
to the fact, that when carbon burns to carbonic oxide a very large
proportion of heat escapes measurement, because its potential energy
has been expended in placing the solid carbon in a gaseous form, just
as in the case of water and steam which I cited just now, a large
amount of heat is absorbed in placing the steam in a gaseous form ;
but when carbonic oxide, which is already a gas, burns to carbonic
acid no such expenditure of heat is necessary, and we are able to
measure, and if need be to utilise, the whole quantity, or at all events
the greater proportion of the heat generated.
You will now, I think, have no difficulty in understanding the
interpretation I put upon the relation between the quantities of heat
generated and the quantities of permanent gas evolved by the
various powders with which I have experimented.
The case is by no means a simple one, as a glance at Tables 2
and 3 will show you how numerous are the substances which play
a part in the metamorphosis, and in addition the powders themselves
differ very considerably ; but I venture to lay down the broad rule,
that the heat measured by the calorimeter in the case of the mining-
powder is greatly less than that measured in the case of the Spanish
powder, because a much larger quantity of heat has been expended
in placing the solid constituents of the powder in a gaseous state.
Coming now to specific heat ; since we know the specific heats at
ordinary temperatures of the solid products of explosion, and since
we know also the specific heats of the permanent gases, we should
be in a position to determine the actual temperature of explosion if
414 INTERNAL BALLISTICS
we could assume that the specific heats of the soHd products
remained invariable over the great range of temperature through
which they pass.
Our distinguished predecessors in researches on gunpowder
(Bunsen and Schischkoff) made this assumption, and from it calcu-
lated the temperature of explosion to be 3400° Cent., about 6150° Fahr.
Sir F. Abel and I are, however, agreed in considering this hypothesis
(piite inadmissible.
I know of no exception to the general experience that the specific
heat is largely increased in passing from the soHd to the liquid state.
The specific heat of water, for example, in so passing is doubled, and
in addition it is more than probable that even with liquids the
specific heat increases very considerably with the temperature.
For these reasons, I consider it certain that the temperature of
explosion calculated in the manner followed by Bunsen and Schisch-
koff would give a temperature much higher than that really attained.
But the determination of the real temperature is a matter of extreme
difficulty and doubt. I employed two methods to settle the point,
and these two methods gave approximately the same temperatures.
The first method I can only briefly describe, as its basis rests on
theoretical considerations. It may be thus described. If we know
the space in which a given quantity of permanent gases are confined,
if we know also the pressure they exert upon the walls of the chamber
in which they are confined, we have the necessary data for determin-
ing at what temperature the gases must be. Now the course of our
researches (and to these I shall presently refer) led us to a pretty
definite conclusion as to the space that the permanent gases occupied
at the moment of explosion when confined in a close vessel, and a
calculation from the data I have mentioned gave a temperature of
nearly 2200° Cent.
To check this theoretical temperature I made numerous experi-
ments with sheet and wire platinum of various degrees of thickness,
also with similar wires of iridio-platinum. Now if the platinum,
when the vessel is opened, were found completely melted, this would
be a proof that the temperature of explosion is considerably higher
than the melting-point of that metal. In the experiments in which
I have placed platinum wire or sheet in the explosion-vessel, although
in nearly all cases the surfaces of the sheet or wire showed signs of
fusion, there was only one instance in which the platinum was
completely melted, and this was in the case of the explosion of a
charge of Spanish powder, which, you will remember, in the heat
INTERNAL BALLISTICS
415
experiments developed a larger qviantity of heat than any of the
other powders.
The conclusion I draw from the whole of these experiments with
platinum and iridio-platinum is, that since in nearly every case the
surface of the metal was melted, or showed signs of fusion, and since
in one case (and that the case where we know the greatest heat was
developed) the platinum did fully and entirely melt, we may conclude
that with these powders the temperature is above the melting-point
of platinum, but not very greatly above it.
Now the melting-point of platinum is about 2000^ Cent., and of
the iridio-platinum still higher. Hence, I should infer from these
direct experiments that the temperature of explosion of powders like
the W.-A. powders is between 2100° Cent, and 2200° Cent., thus
confirming the theoretical determination to which I referred.
The apparatus used for the determination of the tension or
pressure existing in the closed cylinder at the moment of explosion is
shown in diagram Fig. III., and its action is easily understood. The
Fio. III.
CRUSHER GAUGE
gauge consists of a small chamber in which is placed a cylinder of
copper of fixed dimensions and well-determined hardness. This
copper cylinder is acted on by a piston of steel, the piston itself being
acted on by the tension of the gases. The pressure corresponding to
a given compression of the copper cylinder being known and registered
in tables, it is only necessary after each explosion to ascertain the
altered length of the small pillars. In the earlier experiments a
416 INTERNAL BALLISTICS
single crusher-gauge only was employed ; but in nearly all the more
recent experiments two or three gauges were used so as to check the
accuracy of each determination.
To ensure accurate results with these gauges, certain precautions
under varying circumstances have to be carefully attended to, or
pressures may be obtained which are very wide of the truth ; but as
this subject is of great importance and has caused no inconsiderable
amount of discussion, I shall endeavour to explain the conditions
which under the indications given by the crusher-gauge may be
safely relied on.
It will be easily understood that if a pressure of, say, 20 tons per
square inch is suddenly applied to the piston, and if this pressure be
resisted by a copper pillar which initially is only capable of support-
ing, without motion, a pressure of 4 tons per square inch, a certain
amount of energy will be communicated to the piston, and the copper
pillar when taken for measurement will have registered not only the
gaseous tension, but, in addition, a pressure corresponding to the
energy impressed upon the piston during its motion.
To get rid of this disturbing cause, it has been found necessary,
when high pressures are being measured, to employ cylinders which
are capable of supporting, without motion, pressures very near to
those which it is desired to measure.
Thus, if it be desired to measure expected pressures of 15 tons
per square inch, cylinders which would support 14 tons would be
selected. If it were desired to measure 20 tons pressures, cylinders
of between 18 and 19 tons per square inch would be taken, and so on.
But there is another cause which may seriously affect the indica-
tions given by the crusher-gauge. If we could suppose that an explo-
sive was homogeneous, that it filled the chamber of the gim or
explosion-vessel in which it was confined completely, that it could be
instantaneously and simultaneously exploded right through its mass,
and that when so converted into gas or other products of explosion
there was no motion of any of the particles ; in that case, a properly
adjusted crusher-gauge would give an accurate measurement of the
pressure ; but the actual state of the case is very different. The explo-
sive is generally hghted at a single point. The products of explo-
sion are projected at a very high velocity, occasionally with large
charges, through considerable spaces, and impress their energy upon
any bodies with which they may come in contact. If that body
happen to be the piston of a crusher-gauge, to the mean gaseous
pressure existing in the chamljer will or may be added the pressure
INTERNAL BALLISTICS 417
due to the action I have just explained. I may illustrate my mean-
ing by asking you to imagine the effect of a charge of small shot fired
into the crusher-gauge, — the products of explosion projected with a
high velocity act in a precisely similar manner.
But there is yet another point to consider. In the ignition of
very large charges, especially when the explosive is transformed with
great rapidity, it is a -priori in every way probable that in different
sections of the chamber very different pressures may exist, and experi-
ments have shown that this is the case. In such instances the crusher-
gauges may give approximately the pressures that actually existed
during an infinitesimal portion of time, but such pressures must not be
taken as correctly indicating the pressure due to the density and
temperature of explosion. It was to escape action of this sort that the
very large grain slow-burning powders of the present day were elabo-
rated, and with such powders the pressure in the powder-chamber is
also tolerably uniform ; but with the old " brisante " powders a portion
of the charge was exceedingly rapidly decomposed, the products of
explosion were projected with a high velocity to the other end of the
chamber, and on striking the shot or other resisting body this vis viva
was reconverted into pressure, producing intense local pressures.
When this intense local action was set up, it commonly happened
that waves of pressure swept backwards and forwards from one end
of the chamber to the other, and crusher -gauges placed at different
points would register the maximum pressure of these waves as they
You will fully understand — I must not detain you by going into
great detail — that these high local pressures act only upon a small
section of the chamber at the same instant, and therefore are not
very serious as far as the radial strength of the gun is concerned.
They are however exceedingly serious in breech-loading guns, as the
breech screw or other breech arrangement has to sustain the full
efiect of all such wave action as I have just been describing. In the
" brisante " powders of many years ago it was frequently a matter of
doubt whether or not this wave action would be set up; but to
illustrate my remarks I give you an instance which I have before
cited.
In experiments with a 10-inch gun in which a rapidly lighting
powder was used, two consecutive rounds were fired, in one of which
wave action was set up, in the other not. The two rounds gave
practically the same velocity, so that the mean pressure in the bore
must have been the same ; but five crusher-gauges, three of which
2 D
418 INTERNAL BALLISTICS
were in the powder-chamber, one in the shot-chamber, and one a few
inches in front of the shot-chamber, gave the following results : —
With wave action, 63-4, 41-6, 37-0, 41-9, and 25-8 tons per square
inch.
With no wave action, 28-0, 29-8, 30-0, 29-8, and 19-8 tons per
square inch.
Chronoscopic observations of the velocity were simultaneously
taken in these two rounds, and were, as they ought to be, nearly
identical.
When experimenting with high explosives, I have found it
necessary to use a special form of gauge; with guncotton, for
example, which detonates with great readiness under certain con-
ditions, the gauges were so formed as only to allow the gases to act
on the piston after passing through an extremely small hole in a
shield or cover protecting the piston.
To illustrate the difference between a gauge protected as I have
described, and a gauge such as that shown in Fig. 3, I need only refer
to an experiment in which I employed four crusher-gauges, three of
which had shielded pistons, and indicated pressures of respectively
32-4, 32-0, and 33-6 tons per square inch. The unshielded gauge,
which was placed at the end of the chamber, and was free to receive
the full energy of the wave action, indicated 47 tons per square mch,
or over 7000 atmospheres.
For similar reasons, although I do not deny that crusher-gauges
placed in the chase of a gun may give valuable indications, I still
consider that unless confirmed by independent means the accuracy of
their results is not to be relied on. Where gases and other products
of combustion are in extremely rapid motion, there is always a prob-
ability of a portion of these products being forced into the gauge at a
high velocity, when too high a pressure would be indicated, and there
is a possibility that occasionally a pressure somewhat in defect might
also be registered.
The experiments made to determine the tension at various gravi-
metric densities have been very numerous, and on the whole exceed-
ingly accordant.
The highest density at which I have been perfectly successful in
retaining absolutely the products of combustion and obtaining a
perfectly satisfactory determination has been unity. Even to obtain
these I have at this density had several failures, all arising from the
gas at one point or other succeeding in cutting its way out. It is
worth while mentioning that although with such explosives as gun-
INTERNAL BALLISTICS
419
cotton, cordite, ballistite, etc., the explosion-vessels have been subjected
to much higher pressures than with gunpowder, the difficulty in
retaining the products is not nearly so great.
The reason probably is, that under the first violent action of the
explosion, portions of the non-gaseous products are immediately forced
between the surfaces intended to be closed by the pressure. Perfect
closure is thus rendered impossible, and the destruction of the surfaces
is an immediate consequence.
The tensions obtained in the experiments with service-powders
gave for a density of unity a pressure of about 6500 atmospheres, or
43 tons per square inch, the tensions at lower densities representing
the pressures, and the axis of abscissae the densities.
But, having determined the tensions by direct measurement, it
became important to ascertain how far these tensions were in accord
with those deduced from theoretical considerations.
It is not possible for me here to explain to you the details of the
calculations by which the formula connecting the density and the
tension is arrived at; but I have placed the theoretical curve and
that arrived at from actual experiment in juxtaposition in Fig. IV.
(see p. 420), and you will note that the two are practically identical.
The fi2;ures from which the curve is drawn are given in Table 6.
Table 6. — Pressures in closed vessels observed and calculated.
Density of
products
of combustion.
Volumes
of expansion.
Pressures
observed in
explosion-vessels.
Pressures
calculated.
Tons per sq. inch.
Tons per sq. inch.
90
1-11
32-46
32-460
80
1-25
25-03
25-525
70
1-43
19-09
20-024
60
1-66
14-39
15-554
50
2-00
10-69
11-851
40
2-50
7-75
8-732
30
3-33
5-33
6-071
20
5-00
3-26
3-771
10
10-00
1-47
1-765
•05
20-00
0-70
-855
We are now in a position to give answers to the questions relating
to gunpowder which I propounded to you a short time ago, and for
the sake of clearness I shall give these answers categorically.
I say, then, —
1. That the substances produced by the explosion of the different
natures of gunpowder of which I have to-day spoken are shown in
Table 2, and that they occur in the proportions there stated.
420
INTERNAL BALLISTICS
2. That, with service-powders, about 57 per cent, by weight of
the products of explosion are non-gaseous.
3. That, with the same powders, about 43 per cent, of the products
of explosion are in the form of permanent gases, and that these gases,
at a temperature of 0° Cent, and at a barometric pressure of 760 mm.,
occupy about 280 times the volume of the unexploded powder.
4. That, at the moment of explosion, the non-gaseous products
are in a liquid state.
5. That a gramme of dry ordinary service-powder, by its explosion,
generates about 720 grm. -units of heat.
Fio. IV.
PRESSURES IN CLOSED VESSELS OBSERVED AND CALCULATED
K
a rs
u °
a 16
V) 14
z
O ,2
\
\
\
V,
\
\
\
u e
N1'\
^
^^^
:i:^
ft
~
"^"^
^^^^
^^^
-■^^■^
_
-0 2 3 4 5 6 7 8 3 10 II 12 13 14. 15 16 17 IB 19 20
VOLUMES OF EXPANSION.
6. That at the moment of explosion the temperature of the
products is about 2200° Cent, or nearly 4000° Fahr.
7. That the mean specific heat of the products of explosion at the
temperature of explosion is about "31.
8. That the relation between the tension of the products of
explosion and their mean density is as exhibited in the curve in
Fig. IV. or Table 6.
9. That the changes produced in the products by a variation in
the gravimetric density of the charge are slight — the one of most
importance being the increase in the quantity of carbonic acid, and
the corresponding decrease in carbonic oxide.
10. That the effect of changes in the chemical composition of
gunpowder on its metamorphosis are very considerable, the propor-
INTERNAL BALLISTICS
421
tions of the products, the quantity of heat, the amount of permanent
gases, being all materially altered, but that these variations do not
alter as much as might be expected the tension of the products in
relation to the gravimetric density of the charge.
11. That the form and size of the grain affects to some extent the
quantity of permanent gases formed, as well as the proportions in
which the various products occur.
Guncotton, known also as pyroxylin or trinitro-ceUulose, is
prepared by submitting cotton to the action of a mixture of con-
centrated nitric and sulphuric acids, a portion of the hydrogen in the
cellulose being replaced by an equivalent quantity of nitric peroxide.
The formula representing guncotton is given in Table 7, and in
this table I have also given both the ultimate composition of gun-
cotton and the metamorphosis which it undergoes on explosion. It
is employed in several forms. For the most useful, the compressed,
and for many other improvements in guncotton, we are indebted to
the labours of Sir F. Abel. Several of the forms of guncotton are
before me. Here is granulated guncotton, here is guncotton in yarns,
strands, and ropes. Here it is in pellets, here in discs, here in slabs ;
and in these last two forms it is generally used for military and
industrial purposes.
Table 7. — Showing the composition and metamorphosis of pellet guncotton.
Composition. •
Products of Explosion
Carbon .
24-89
Carbonic anhydride .
oxide .
0-424
Hydrogen .
2-69
0-280
Nitrogen .
13-04
Hydrogen .
0-011
Oxygen .
56-66
Nitrogen .
0-145
Ash .
0-36
Marsh-gas
0-003
Moisture .
2-36
Water
0-116
Formula-C6H73(NOo)05.
Original moisture
0-021
I may explain that these last two specimens, which represent
considerable quantities of guncotton, are wet, and perfectly safe unless
treated in the manner I shall presently describe.
Guncotton differs from gunpowder in this, that when fired, prac-
tically the whole of its constituents assume the gaseous state, and
the transformation is accompanied by a much higher temperature.
Something is to be learned by observing the ignition. I fire here
a piece of yarn ; observe the time that the flame takes to traverse
the train. Here, again, is a strand : you will note that the ignition is
much more rapid, while if I fire this piece of rope you will observe
that the rapidity of combustion is so great as to amount almost to an
explosion. The slowness of combustion of the yarn and strand gun-
422 INTERNAL BALLISTICS
cotton is due to the ease with which the nascent products escape, so
that no very high pressure is set up.
The rapidity of combustion of the rope is due to the higher
pressure arising from the greater compression and the much larger
quantity of gas liberated in a given section.
Were I, by using a few grains of fulminate of mercury, to produce
a high initial pressure, the harmless ignition you have seen would be
converted into an explosion of the most violent and destructive
character. This disc I hold in my hand would blow a hole in a
tolerably thick iron plate, and I need not say would make an end of
myself and any who had the bad fortune to be very near me.
One great advantage that guncotton possesses lies in the fact that
we are able to keep it and use it in the wet state, and in that state to
produce quite as effective an explosion as if it were dry. It is only
necessary that a few ounces of dry guncotton be in close juxtaposition
to the wet, and that the dry guncotton be detonated, as I have
described, with a few grains of fulminate of mercury.
I may mention as a curious fact that Sir F. Abel has shown, by
means of a chronoscope I shall presently describe, that whereas the
detonation of dry guncotton travels at the rate of about 18,000 feet
a second, or about 200 miles a minute, the detonation of wet gun-
cotton is at the rate of about 21,000 feet a second, or 240 miles a
minute. When the fact is known, it is not difficult to understand
the cause of this increased rapidity.
The effect of pressure in increasing the rapidity of combustion of
explosives may be very well illustrated by comparing the rapidity of
combustion of pebble-powder under different circumstances. A
pebble such as I hold in my hand is generally, in the bore of a gun,
and under a pressure of from 15 to 20 tons per square inch, entirely
consumed in less than the 200th part of a second. In free air the
time taken for such a pebble to burn is about two seconds, and ioi
vacuo it will not burn at all.
A beautiful experiment for showing this phenomenon has been
devised by Sir F. Abel, but for want of the necessary apparatus I am
unable to show it here. In an exhausted receiver a platinum wire,
which can be heated by an electric battery, is arranged, touching
either gunpowder or guncotton. On raising the wire to a red heat
the gunpowder or guncotton in contact with the wire burns, but the
cooling effect of the immediate expansion of the gases is so great that
the combustion is confined to the explosive in actual touch with the
heated wire
INTERNAL BALLISTICS 423
The effect of pressure in increasing the rapidity of combustion
will enable you to understand the action of fulminate of mercury on
guncotton, nitro-glycerine, picric acid, and other high explosives.
I ought to explain that, destructive as are the effects of fired gun-
powder, I should not myself include gunpowder in my list of true
explosives. It is not Hke guncotton, nitro-glycerine, and other
similar explosives, a definite chemical combination in a state of
unstable equihbrium ; but it is merely an intimate mixture in
proportions which, as you see from Table 1, may be varied to a very
considerable extent, of those well-known substances, nitre, sulphur,
and charcoal. These constituents do not, during the manufacture of
the powder, undergo any chemical change, and being a mere mixture,
gunpowder cannot be detonated; but it deflagrates or burns with
great rapidity — that rapidity, as I have pointed out, varying largely
with the pressure under which the explosion is taking place. Gun-
cotton, on the other hand, when, by means of fulminate of mercury
an extremely high local pressure has been set up, transmits that
pressure to the adjacent guncotton with extreme rapidity. A charge
of pebble-powder in a gun would be consumed in about the 200th
part of a second, but a charge of 500 lbs. of these slabs would, if
effectively detonated, be converted into gas in somewhere about the
20,000th part of a second.
Eeverting again to Table 7, you will observe that carbonic
anhydride, carbonic oxide, nitrogen, and water are the principal
products of the decomposition of guncotton. The composition of
these gases does not vary much with the pressure, but, as in gun-
powder, with the higher pressures a larger proportion of carbonic
anhydride is formed. The permanent gases, when reduced to 0° Cent,
and 760 mm. barometric pressure, measure between two and three
times the number of volimies given off by gunpowder, 1 grm. of
guncotton generating about 730 c.c. of permanent gas, while the
temperature of explosion is at least double that of gunpowder.
From these data it is obvious that the tension of fired guncotton
is very high, and, provisionally, I have placed it at about 120 tons
per square inch, or nearly 20,000 atmospheres ; but all efforts actually
to measure with any degree of accuracy these enormous pressures
have so far proved futile. The highest pressure I myself have
reached was, with a density of "55, about 70 tons per square inch ;
but all the crusher-gauges used having been more or less destroyed,
this measurement must be accepted with a good deal of reserve.
I have experimented with so many varieties of amide powder
424 INTERNAL BALLISTICS
which have differed considerably in their composition, that I would
find difficulty in giving you in a few words the somewhat varied
results obtained from them.
I shall, therefore, take one only as a sample, and I select this
because I shall have occasion, shortly, to refer to some results
obtained with it. The powder in question consisted of a mixture
of 40 per cent, of potassium nitrate, 38 per cent, of ammonium
nitrate, and 22 per cent, of carbon.
Its explosion generates a considerably larger quantity of
permanent gases than ordinary powder, and the quantity of heat
developed is also greater. The permanent gases consist of 30 per
cent, of carbonic acid, 13 per cent, of carbonic oxide, 27 per cent, of
hydrogen, and 30 per cent, of nitrogen. The powder cannot truly
be called smokeless, but the smoke formed is much less dense and
more rapidly dispersed than that of ordinary powder. Its potential
energy, as I shall shortly show you, is also much higher, and it
further, as far as my experiments on that subject have gone, appears
to possess the invaluable property of eroding steel to a much less
degree than any other powder with which I have experimented.
The main objection to its extended use is the tendency to deliques-
cence, arising from the use of ammonium nitrate in its composition,
and necessitating the powder being kept in air-tight cases. As,
however, at aU events on board ship, all powders are supposed to be
so kept, I do not know that this undoubtedly serious objection would
be an insurmountable difficulty if the other advantages of the powder
should be fully established.
I come now to the last class of explosive with which I shall
trouble you. It is a new explosive, of which you have probably
heard, known by the name of "Cordite," and for which we are
indebted to Sir F. Abel and Professor Dewar. I have on the table
several samples of this explosive. This, which you see looks like a
thick thread, is for use in rifles. This size, a little thicker, is used
in field guns. These two sizes, like thick cords, a resemblance to
which they owe their name, are for 4-7-inch and 6 -inch guns.
As with guncotton, I burn one or two lengths. Like guncotton,
you will observe there is no smoke ; but, unlike guncotton, you will
note that there is not the striking difference in the velocity of
combustion that you observed with that explosive. I drew your
attention to the power we possessed of detonating guncotton by
means of fulminate of mercury. This property you will readily
understand makes guncotton a most valuable explosive for torpedoes,
INTERNAL BALLISTICS 425
or other cases where a maximum of explosive effect is desired ; but
it imfits it for safe use in large charges in a gun, because under
certain abnormal circumstances a detonation might be set up, when
the failure of the gun would be the almost infallible consequence.
It would be too much at present to say that under no condition is
it possible to detonate cordite ; but I am able to say that at all events
it is in a very high degree less susceptible to detonation than gun-
cotton, and that so far, even by the use of fulminate with the charges
with which I have experimented, I have not succeeded in detonating
it. The explosive has other advantages ; it is, as you see, made in
a form specially suitable for making into cartridges. It is not
injured by being wetted. I dip it into water, and on removing the
superficial water with my handkerchief to allow it to light, you see
it burns much as before.
The samples of cordite I have shown you consist approximately
of 58 per cent, of guncotton, trinitro- and dinitro-cellulose, 37 per
cent, of nitro -glycerine, and a small percentage of a hydrocarbon.
From explosive experiments I have made, it has been found that
the products of combustion, which are all gaseous, consist of
approximately 27 per cent, of carbonic acid, 34 per cent, of carbonic
oxide, 27 per cent, of hydrogen, and 12 per cent, of nitrogen, and to
these permanent gases has to be added a considerable quantity of
aqueous vapour.
The volume of the permanent gases, at 0° Cent, and 760 mm.
barometric pressure, is as nearly as possible 700 c.c. per gramme of
cordite, while the quantity of heat developed is 1260 grm.-units.
These figures show that the potential energy of this explosive must
be very high, as will be demonstrated to you when I come to treat
of its action in a gun.
On the diagram Fig. V. I exhibit to you the pressures that I have
measured with this explosive up to gravimetric densities of '55. For
purposes of comparison, I have placed on the same diagram the
corresponding curve for gunpowder, and you will note how with the
cordite the pressures are much higher in relation to density. With
this explosive I have made experiments at higher densities than are
shown on the curve, and have in fact measured pressures up to
90 tons per square inch ; but certain anomalies and difficulties in the
interpretation of the results prevent my relying on their exactness
until confirmatory experiments have been made.
I believe I mentioned to you that ballistite, an explosive of some-
thing of the same nature as cordite, is now being introduced on the
426
INTERNAL BALLISTICS
Continent for military purposes. In appearance it is very different,
as you will see from the samples on the table ; but it possesses the
same property of smokeless combustion. The gases generated are
the same as those of cordite in somewhat different proportions, but
their quantity is less, being about 615 c.c. per gramme, while the
quantity of heat is higher, being 1356 grm. -units. The potential
energy is somewhat less than that of cordite, and I have placed on
the same diagram as the cordite the pressures with ballistite for
different densities.
I return to gunpowder, which we shall now consider under very
different conditions, and shall study the behaviour of a charge when
placed in the chamber of a gun, and allowed to act upon a projectile.
Fio. V.
pressures observed in closed vessels with
\arious explosives.
65
fiD
/
'
fir
/
/.
/
>
i =
/
'X
au
^-i^
^
n?
f^'^
TT
^
t'
^^
_^
^
^-^
..-^
^r
^
rW^^-
J--^
p9>t
ti^
IS
^
'
^
^
■—^
-
s
""
^
'
1=T
•
lO -i
5 •
30 •;
55 •<
-i
5 .
50 -S
>5 •€
-C
5 -7
O •
i> •
0. -8
DENSITY OF PRODUCTS OF EXPLOSION
You will remember, and forgive me for repetition, that the charge
when fully exploded, and at the moment of explosion, consists of
about 57 per cent, of liquid products in an extremely 'fine state of
division, and about 43 per cent, of permanent gases ; also that when
the gravimetric density of the charge is unity the tension when
unrelieved by expansion is 6770 atmospheres, or 43 tons per square
inch.
Let us suppose in the first instance that the gravimetric density
of the charge is unity, and that the charge is entirely consumed
before the projectile is removed from its seat. Now under these
circumstances, if we represent the relation between the tension and
volume occupied by that charge in the bore by a curve, representing
the tensions by the ordinates, the volumes or the distances travelled
INTERNAL BALLISTICS 427
by the shot corresponding to these volumes by the abscissae, you will
at once see that the curve will be of the form shown in the diagram
Fig. IV. This curve, in fact, is the same as that indicated in the
diagram Fig. V., and represents the relation between the tensions
and densities of the products of combustion, where the gases expand
without cooling or production of work ; but the densities, instead of
being as in the former curve taken as the independent or equicrescent
variable, are here dependent on the volume, or numbers of expansions
occupied by the charge in the bore, these expansions in this instance
being taken as the independent variable.
The maximum tension under the circumstances supposed would,
as I have said, be 43 tons per square inch, but the tensions at other
points would not, for reasons I shall explain later on, be as great as
are shown on this curve.
But many of you are aware that in reality, and especially with
the slow-burning powders now introduced for large guns, we cannot
consider the charge to be instantaneously exploded, and although the
determination of a theoretic curve of pressures to include the first
moments of combustion would be a work of the very greatest
difficulty, it is yet tolerably easy to see what the general form of this
curve of pressures must be.
The charge of powder is generally ignited at a single point. No
doubt, especially with pebble, P2, or prismatic powders, the flame is
very rapidly communicated to all the grains, pebbles, or prisms. In
each individual pebble, supposing it to be ignited on the whole of its
surface, the burning surface is of course a maximum at its first
ignition ; but the quantity of gases generated will depend so much
on the pressure at any particular moment, that it does not necessarily
follow that the greatest quantity of gas is given off at the moment
when there is the largest burning surface.
Be this as it may, however, it is obvious that before the charge is
fuUy consumed there will be a great decrement in the quantity of
gases given off, unless, as in prismatic and some other powders,
arrangements are made by means of holes to keep as large as possible
the ignited surfaces, or unless the interior of the large pebbles is
composed of more explosive and easily broken up material.
The pressure, then, when the charge is ignited commences, with
muzzle-loading guns, at zero; with breech loading-guns, with a
pressure of 2 to 4 tons per square inch, or with whatever pressure
may be necessary to force the driving-ring of the projectile into the
grooves and into the bore of the gun. The pressure increases at an
428
INTERNAL BALLISTICS
extremely rapid rate until the maximum increment is reached. It
still goes on increasing, but at a rate gradually becoming slower, until
the maximum tension is reached, and this tension is attained when the
increase of density of the gases due to the combustion of the powder
is just balanced by the decrease of density due to the motion of the
projectile. After the maximum tension is reached the pressure
decreases, at first very rapidly, subsequently slower and slower.
Now, if these variations in pressure be represented by a curve, it
is easy to see that the curve will commence at the origin by being
convex to the axis of the abscissae. It will then become concave,
then convex, and will finally be asymptotic to the axis of x. It will
be in general form similar to the gun-pressure curve shown on the
diagram Fig. VI.
Fig. VI.
DIAGRAM SHOWING THE FORMS OF PRESSURE AND VELOCITY CURVES IN
A GUN, AND OF CURVE OF PRESSURES IN A CLOSED VESSEL.
\
""""
"
"^
\
IML/M
19- OVON
■RESSUf
s
)9
^
.
/
\ \
-
5!r
/
\
^
..oOi
i — —'
/
\
\
%
&^
\
\
^^
\ \
^^
\ /
\^
< IZ
y\
M>.
1200*
3 ,,
/
. ^
/
\
1000]
fA
1
/
\
'"/^ """^e
\ ,
«00
FS^
^^^f.
^UNrr.
0. 7
*^.
■~-~ia
6
600
-^■^
^/.,
■=^
5^£
"S
/
^^r
t____^
/
i^OO
^iC^c
1- 7
/
~~-— -_
/
200
— -^__
C
) 1
FE
ET 1
-RAVE
i A
L
F SH
or
1
N. B
J 3
OF C
1
UN,
3 1
1 1.
2 13
13
i
In like manner, the curve representing the velocity in the bore of
a gun would commence by being convex to the axis of abscissae. It
will then become concave, and were the bore long enough, would be
finally asymptotic to a line parallel to the axis of x. The velocity
curve is shown in the same diagram, Fig. VI.
Such, then, would be the general form of these curves. Let me
now describe to you the means which have been taken to obtain the
data necessary for the construction of such curves by actual
observation.
INTERNAL BALLISTICS 429
It is obvious that if we desired to know the pressure exercised by
the gases on the projectile at various points of the bore, the velocities
of the projectiles at the same points, and the times taken by the
projectile to reach these points from the commencement of motion,
there are two courses open to us. We may either, first by suitably
prepared gauges (if it be possible to construct such gauges) determine
the pressure at various points of the bore, from the pressures deduce
the velocities, and thence the times, or we may follow the inverse
method. We may measure the times at which the projectile passes
certain known points in the bore. From these times we may calcu-
late the corresponding velocities, and finally calculate the pressures
necessary to produce these velocities at all points along the bore.
Now, both these methods have been followed, and I shall
endeavour to describe the instruments and methods in as few words
as possible.
To Eodman, as far as I know, is due the merit of having first
used a pressure-gauge to determine the pressure in the bore of a gun.
Eodman's own experiments with his gauge are, however, as I have
elsewhere shown, unreliable, his pressures being generally much
higher than any possible actual pressures, since the pressures given
by him as existing at various points along the bore would, if assumed
to act on the projectile, give energies in some cases nearly three
times as great as .in reality.
In this country the crusher-gauge I have already described is
almost universally employed. The results it gives are, with proper
precautions, reliable when the gauge is placed in the chamber ; but
it cannot, as I have endeavoured to explain, be depended upon with
any accuracy when placed in positions where the products of explo-
sion are moving at a high velocity.
Two methods have been employed for obtaining the pressure by
the inverse method. One method consisted in shortening the gun by
successive calibres, and at each length determining the velocity
imparted to the projectile by the same charge.
This method, however, is a very rude one, and is open to several
very serious objections, and a preferable method is to measure the
time at which the projectile passes certain fixed points in the bore.
To effect this object a chronograph was employed, with certain
peculiarities of construction designed to measure very minute inter-
vals of time.
This instrument is shown in the diagrams Figs. VII.. VIII.,
and IX.
430
INTERNAL BALLISTICS
Fia. VITI,
END ELEVATION.
It consists of a series of thin discs, each 36 inches in circum-
ference, made to rotate at a very high and uniform velocity through
the train of wheels F by means of a heavy descending weight B,
arranged, to avoid an inconvenient length of chain, upon a plan
originally proposed by Huyghens, the weight being, during the
experiment, continually wound up by the handle C, and thus the
instrument can be made to travel either quite uniformly or at a rate
very slowly increasing or decreasing. The speed with which the
circumference of the discs
travels is usually about 1200
inches per second ; an inch,
therefore, represents the
1200th part of a second;
and as by means of a vernier
we are able to divide the inch
into 1000 parts, the instru-
ment is capable of recording
less than the one millionth
part of a second. You will
appreciate the extreme minute-
ness of this portion of time, if
I point out that the millionth
of a second is about the same
fraction of a second that a
second is of a fortnight. The
precise rate of the discs is
ascertained by means of the in-
termediate shaft, which in the
earlier arrangement worked a
stop clock, but in the more
recent, by means of a relay,
registers the revolutions on a subsidiary chronoscope (each revolution
of the intermediate shaft corresponding to 200 revolutions of the
discs), upon which subsidiary chronoscope a chronometer, also by
means of a relay, registers seconds.
The recording arrangement is as follows. The peripheries of the
discs we cover with specially prepared paper, and each disc is
provided with an induction coil. You are aware that when the
primary of an induction coil is suddenly, severed, a spark, under
proper management, is given off from the secondary, and in the
arrangement I am describing the severance of the primary is caused
INTERNAL BALLISTICS
431
by the shot in its passage through the bore, and as each successive
wire is cut the induction coils record on their own discs the instant
at which the shot cuts the wire, that is, passes the particular point
with which the primary wire is connected.
To prevent confusion, there is delineated in the diagram only a
single induction coil and cell ; but you will understand that there is
an induction coil for each disc, and that each disc, discharger, and
cell form, so to speak, independent instruments for recording the
Fia. IX.
SIDE ELEVATION.
instant when the projectile passes a certain point in the bore of the
gun. The diagram will give you an idea of the manner in which the
primary wires are conveyed to the interior of the gun.
By experiments with which I need not now trouble you, I have
found that when the instrument is in good working order the
probable instrumental error of a single observation does not exceed
from two to three miUionths of a second.
Let me now give you some of the actual results obtained by the
means I have just described.
I have given you a short account of two or three of the more
432
INTERNAL BALLISTICS
modern explosives — I mean amide powder, cordite, and ballistite —
and have, I fear, somewhat wearied you with the details into which
I have gone in respect to the old powder.
Now on this diagram. Fig. X., are exhibited the results obtained
with three of these powders, all fired from the same gun, under the
same conditions, with approximately the same maximum pressures ;
but, as you see, with very different results as regards velocity and
energy.
You will note that the axis of abscissae. Ox, denotes the length of
Vw. X.
CURVES FROM
CHRONOSCOPE
co?.93:^-^--
..ajfP-
.2036..
B.-"""'..„A.^-''''-:
....IP.??-
...- .Bn--- PEBBLE
il7
,--' ..--".-— b;
U
Z 16
A o^^->5-'''' '''''''' e ^^^
'^
^
IS
111
E 11
5,2
t\ S^^' .^<:^^
^^^
-A^
S.^
wC ^
^>^0^^
■
II
V^.-*
K
<-^ ^
^ 10
Av% ^-
^■^^
'/ \2»^ \> ^
^-^
Z
il
/ Wv^
f
3
O
:!• x\/*^
t- 7
'/'•' >yy
V
^
2 6
! /yy^
\
^\^^
" 5
'' /y
\
^\^C,
u
u
1 A
^.
■\.
_I
0:4
^3
r
^\
^
^
^^«a
^
N
2 2
m
"~--C!^!;;^
^^^i:::::::;:;--^,^
S
I'
J\\
PL
UCS
1
N
CUN
1
£ 1
3 ~
/2 3 -. .5 , e ^ e g >
A
^0 1 2 J 4 5- 6 / a 3 10 «/ /2 ■
a «4
FEI
:t
TR
AVEL
OF
■ Sl-
OT
the l3ore of the gun traversed by the shot under the action of the
explosion, and that the ordinates to the crosses on the curves A^, A,,
Ag denote the times occupied by the shot in reaching the various
points shown. The total time taken by the shot in traversing the
whole bore is "01 second in the case of pebble-powder, -0095 seconds
with amide powder, and '009 seconds with cordite.
The velocities at all points of the bore are indicated by the curves
Bp B,, Bg, and you see that the muzzle velocity of ordinary powder
is 1839 feet per second, of amide powder 2036 feet per second, and
cordite 2150 feet per second.
INTERNAL BALLISTICS 433
In like manner, the pressures at all points of the bore are shown
by the curves C^, C2, C3. Observe that the maximum pressure with
pebble-powder is 16 "4 tons per square inch, with amide powder 16
tons per square inch, and with cordite 14*4 tons per square inch.
The area between the curves C, the axis of abscissae Ox, and the
ordinate at the muzzle represent for each curve the energy developed
by that explosive in the bore. It is easy by mere inspection to see
that the total energy of the amide powder and of the cordite is much
higher than that developed by the pebble ; but if we make the required
calculations and put the result in figures, we shall find that the
energy developed by the pebble-powder was 1055 foot-tons, by the
amide 1293 foot-tons, and by the cordite 1435 foot-tons, or, with less
maximum pressure, nearly 40 per cent, higher than in the case of
pebble-powder, and an examination of the pressure curves will show
in what portions of the bore this great additional energy is reaHsed.
It is important to observe that these greatly higher energies
developed by the modern powders are obtained with considerably
reduced charges, the weight of the charge of amide powder being
reduced to 85 per cent., and of cordite to 47 per cent, of that used
when the service pebble-powder is employed.
These results have been obtained by actual experiment, and it
remains to ascertain what accordance there is between them and what
we have a right to expect from theoretical considerations.
Taking, again, gunpowder as our illustration ; Hutton appears to
have been the first person who attempted anything like a theoretical
explanation of the action of gunpowder on a projectile, but he not
unnaturally fell into the error of assuming that the whole of the
products were in a gaseous state, and, further, that their tension was
directly proportional to the density, and inversely as the space they
occupied. In other words, he supposed that the gases in expanding
and performing work accomplished that work without expenditure
of heat.
De Saint Eobert, who was the first to apply to the question the
modern theory of thermo-dynamics, corrected Hutton's error, but, like
Hutton, he assumed that the whole of the products were in a gaseous
state, and, as gases, doing work on the projectile.
Bunsen and Schischkoff, in their well-known researches on gun-
powder, pointed out that although it was probable that there might
be a slight volatilisation of the solid products, yet it was in the
highest degree improbable that such volatilisation would ever reach
a single atmosphere of pressure, and that any effect on the projectile
2 E
434 INTERNAL BALLISTICS
would be perfectly insignificant. They therefore, in their calculations,
disregarded the solid residue altogether, and calculated the total work
which gunpowder is capable of perforndng on the assumption that
such work is done by the expansion of the gases alone without
addition or subtraction of heat, and that, in fact, the non-gaseous
products played no part in the expansion.
The effect of these erroneous assumptions was that the tensions
calculated on De Saint Eobert's hypothesis were considerably higher
(for given densities) than those which were observed in a close vessel
where the gases expand without production of work, while the
tensions calculated on Bunsen and Schischkoff's hypothesis were
greatly in defect, not only when the tensions were taken from those
observed in a close vessel, but also in defect of the pressures actually
observed in the bores of guns.
At an earlier stage in the researches carried on by Sir F. Abel
and myself, I came to the conclusion, when I found that the pressures
deduced from experiments in close vessels did not differ so much as
I anticipated from those taken in the bores of guns, that this
departure from expectation was probably due to the heat stored up
in the liquid residue ; and it must be noted that this liquid residue
being in an exceedingly finely divided state, and thoroughly mixed
up with gases, constitutes a source of heat of the most perfect
character, immediately available for compensating the cooling effect
due to the expansion of the gases when employed in the production
of work.
On correcting the assumptions I have referred to, and calculating
the tensions that would exist in the bore of a gun, it was found that
the anomalies to which I have drawn your attention were entirely
removed, and that theory and observation were in accord, the
pressures obtained with Waltham-Abbey powder being, even while
the densities were still very high, not greatly removed from the
theoretic curve, while when the powder may be considered entirely
consumed the two curves slide into one another, and for all practical
purposes become coincident.
In my address to the mechanical section of the British Association
I drew attention to the extraordinary stagnation that had existed in
guns and artillery during a period of more than two centuries — a
stagnation that was the more remarkable because the mind of this
country during the long period of the Napoleonic and earlier wars
must have been to a large extent fixed on everything connected with
our Naval and Military Services.
INTERNAL BALLISTICS
435
It is not too much to say that the changes and improvements in
artillery made during the ten years that followed the Crimean "War far
exceeded in importance all the improvements made during the previous
200 years.
What the future may produce it is difficult to say ; but to show
you that during the last ten or fifteen years great progress has been
made both in guns and the explosives which are used in them, I draw
your attention to the results obtained from a 7-inch 7-ton rifled gun
of fifteen years ago and those obtained from a 6-inch 6|-ton quick-
firing gun with a charge of cordite. To show you the difference in
the appearance of the guns, I have placed side by side half sections of
Fio. XI.
COMPARISON BETWEEN A 7-INCH OLD GUN AND A 6-INCH NEW GUN
2680 TEET
FER-'St'cONil
^^^.^^y'-'--'
r
^X ,,-"""
IStl FEET
PER SECOND
/ \-
' ^^^»^
/ '''
V ^^-^^
W
\^<.,^^^
■~-~~-«as .
—- -___
}i
^"~^~.^,.,^_^
1
,
the two guns (see Fig. XI.). Note the difference in the length of the
guns. Taking first the velocity curves, note the enormous difference
in velocity, the old gun having a velocity of 1560 feet per second,
while no less than 2680 feet per second were realised with the 6-inch.
But the energy of the projectile shows in the most striking way the
difference between the guns. The maximum pressure being about the
same, the energy of the 7-inch projectile is only 1943 foot-tons, while
that of the 6-inch gun is 5000 foot-tons, or not far off three times as
great.
To emphasise what I have said as to the magnitude of the advances
in artillery that have been made since 1856, it is enough to point out
that smce that date the charges of gunpowder fired in guns have
436 INTERNAL BALLISTICS
increased from 16 to 1000 lbs., the weights of the projectiles from 68
to 2000 lbs., the velocities from 1600 feet per second to 2700 feet per
second,* and the energies developed in the projectiles from 1100 foot-
tons to 62,000 foot-tons.
In referring to the diagram to which I just now called your atten-
tion, I pointed out the great difference in the length of the two guns
compared, but I am bound to admit that as a thermo-dynamic machine
the old guns were more economical than the new ones ; the reason
being, that as the charges are proportionally much larger in the new
ones, the tension of the gas at the muzzle is also larger, and the
products are discharged with a larger proportion of their total energy
unrealised.
It may interest you to know what this total energy amounts to.
Knowing the permanent gases formed, knowing also the specific heats
and the tensions at the moment of explosion, the ordinary laws of
thermo-dynamics enable us to calculate the total energy which will be
developed.
In the case of gunpowder the calculation is somewhat complicated
by the large proportion of non-gaseous products, but, as I have else-
where shown, with certain modifications the ordinary laws are appli-
cable, and the total energy obtainable if the charge be indefinitely
expanded is about 34,000 kilogrammetres per kilogramme of powder
or, in English measure, nearly 500 foot-tons per pound of powder.
Cordite would give, approximately, under the same conditions, a
total energy four times as great, or, say, 2000 foot-tons per pound of
cordite.
When we consider the destructive effects realisable by even a
small charge of gunpowder, it is somewhat surprising to reflect that
this potential energy of gunpowder is only about one- tenth of that of
one pound of coal, and is not even equal to the energy stored up in
the carbon which forms one of its own constituents.
At the same time, it must not be forgotten that the gunpowder
has stored up in it the oxygen necessary for the oxidation of its
carbon and other oxidisable substances, while one pound of carbon,
in burning to carbonic acid, has to draw from the air nearly 3 lbs. of
oxygen.
You may, possibly, desire to know what proportion of the total
theoretic work of gunpowder is realised in modern artillery.
* Were it necessary, with our new explosives still higher velocities and energies
might be obtained. The highest possible veloc-ity, however, interesting as it may
be in a scientific, is not always desirable in a practical point of view.
INTERNAL BALLISTICS 437
A gun being, as I have said, an extremely simple form of the
thermo-dynamic engine, the coefficient of effect is high. The actual
energy realised varies considerably, dependent on circumstances, but
may be taken as something between 50 and 90 foot-tons per pound of
powder, or, say, from about one-tenth to one-fifth of the total theo-
retic effect. The average coefficient of effect, comparing the energy
expressed in the projectile with that due to the expansion of the gases,
may, I think, be taken as somewhere near 80 per cent. It rarely falls
below 70 per cent., and, occasionally, with large guns and charges, is
considerably above 90 per cent.
But I must conclude, and conclude as I began, by emphasising
the indebtedness of my own department, as well as of nearly all depart-
ments of knowledge, to the great man whose anniversary my lecture
to-night is intended to commemorate.
It must ever be a subject of pride to this country that the two
inventions — I mean the steam engine and the locomotive — which in
my judgment have done far more than any other which can be named
to advance civilisation and the welfare of the human race, are due to
her sons.
These inventions have for some generations brought to this country
great wealth, and employment for thousands upon thousands. But
other nations are now running us close, and unless the patient industry,
laborious search after truth, and energy in overcoming difficulties,
which were the distinguishing characteristics of James Watt and
George Stephenson, be preserved in some degree among all classes in
the generation which shall carry on our work, the days of England's
manufacturing pre-eminence are numbered.
In the following table I have given the values of certain constants
which are of common occurrence in questions connected with " Internal
Ballistics." Tliis table requires no explanation, but to it I have added
some other tables which I have calculated, and which in my own work
I have found exceedingly useful.
These tables are as follow : — First, a table giving the work in foot-
tons that 1 lb. of service gunpowder is capable of performing, in
expanding from a volume whose gravimetric density is unity, to any
given number of volumes, up to forty. As an example of the use of
this table, suppose that in an 8-inch gun, a charge of 100 lbs. of
powder, with a gravimetric density of unity, is fired, and suppose
further, that the number of expansions that this charge suffers, when
the base of the projectile reaches the muzzle, is 4*29; what is the
438 INTERNAL BALLISTICS
maximum energy that the powder is capable of generating ? From
the table it will be seen that the work corresponding to an expansion
of 4-29 volumes is 85'068 foot-tons per lb. of powder, and as the charge
is supposed to be 100 lbs., the maximum energy, under the conditions
stated, which that charge would be capable of generating, would be
8506'8 foot-tons. The maximum effect is of course never realised, and
for proved powders a factor of effect is generally approximately known.
If we suppose this factor to be 0-84; then 8506-8 x 0-84= 7145-7 foot-
tons represents the energy which will be realised. Should the density
not be unity, a correction has to be made. Thus if the gravimetric
density of the charge were 0-87, which density corresponds to 1-15
volumes of expansion, from the value 85'068 foot-tons per lb. of powder
given above, would have to be subtracted* 12-625 foot-tons, the
energy due to the expansion of 1*15 volumes, the maximum energy
realisable would be (85-068-12-625) 100 = 7244-3 foot-tons, while
assuming the same factor of effect, the energy which would be actually
realised would be 6085 foot-tons.
The second table gives the energy in foot-tons stored up in 1 lb. in
weight, moving at any velocity, up to 3000 feet per second. For
example, if we desired to know the energy stored up in a 100-lb. shot
moving with a velocity of 2182 feet per second, from the table we see
that that energy is 3301-4 foot-tons; or if we wished to know the
velocity with which a projectile 200 lbs. weight, possessing 7145-7
foot-tons energy, was moving, —„-^ = 35-7285, and from the table
>dJ' o' 200
the velocity required is 2270 feet per second.
Tables 3 and 4 differ from Tables 2 and 1 only in the metre and
kilogramme being employed to replace the foot and pound as the units
of length and weight.
Table 5 is for converting cubic inches per pound of powder into
densities and volumes, and vice versd.
* See Phil. Trans, of the Roy. Soc, part i., 1880.
INTERNAL BALLISTICS
439
ft 3
I i
« o.'
s
1 1
o a
tS ^ d
.5 £
Hi
II
S 2
a s
_ 3 a
■a o o a
, C3 « _• =3 CS
a g I .S c
s'-a
.a £
o a a a ^
-M c3 ei c4 J-!
aa
"2 ^
6 d
C» CD O 0-^'*ClCa
>p^
.|^«iii.|;
s s a
S :b M 6
£1. a
II
o o o o a
O O Pi Pi H Ph O
a d
O O
• S
. a
■s a
-ti -ti o o d g
S M-g^
§ a
a S
<4-. O
o a
§3
:ti ^ ^ .^^ »_
a^
K .- .ii a
a a -a -5 ^ a
O O Pi Oh 3Q Pi
440
INTERNAL BALLISTICS
Table shoioing the work in foot-tons that 1 lb. of gunpowder is capable of performing
in expanding from volume = \, to any given number of volumes up to 40.
i
•00
•01
■02
•03
•04
•05
•06
•07
•08
•09
I
Work in Foot-tons.
Work in Foot-tons.
1
1
2
0-000
8-852
16-063
•978
9-637
16-716
1-934
10-406
17-359
2-868
11-160
17-992
3-780
11-899
18-614
4-672
12-625
19-226
5-545
13-338
19-828
6-399
14-038
20-420
7-234
14-725
21 -003
8-051
15-400
21-577
3
4
5
•22-142
27-380
31-986
22-699
27-867
32-417
23-248
28-348
32-843
23-789
28-823
33-264
24-323
29-291
33-681
24-850
29-753
34-093
25-370
30-211
34-500
25-882
30-663
34-903
26-388
31-109
35-301
26-887
31-550
35-695
6
7
8
36-086
39-778
43-133
36-473
40-128
43-452
36-855
40-474
43-769
37-233
40-817
44-083
37-608
41-156
44-394
37-979
41-493
44-703
38-346
41-827
45-009
38-709
42-158
45-313
39-069
42-486
45-614
39-425
42-811
45-913
2
9
1
46-209
49-050
51-673
46-503
49-321
51-927
46-795
49-590
52-179
47-085
49-857
52-429
47-372
50-121
52-677
47-657
50-383
52-923
47-940
50-643
53-167
48-221
50-902
53-410
48-500
51-160
53-652
48-776
51-417
53-893
3
4
54-132
56-439
58-605
54-370
56-662
58-814
54-606
56-883
59-022
54-840
57-103
59-229
55-073
57-322
59-435
55-304
57-539
59-639
55-534
57-755
59-842
55-762
57-970
60-044
55-989
58-183
60-245
56-215
58-395
60-444
5
6
7
60-642
62-563
64-385
60-839
62-750
64-562
61-035
62-936
64-738
61-230
63-121
64-913
61-424
63-304
65-088
61-616
63-486
65-262
61-807
63-667
65-435
61-997
63-848
65-607
62-186
64-028
65-778
62-375
64-207
65-949
3
8
9
66-119
67-771
69-347
66-288
67-932
69-501
66-456
68-092
69-654
66-623
68-251
69-806
66-789
68-410
69-958
66-955
68-568
70-109
67-120
68-725
70-259
67-284
68-882
70-409
67-447
68-938
70-558
67-609
69-093
70-706
1
2
3
70-854
72-301
73-690
71-001
72-442
73-826
71-148
72-583
73-962
71-294
72-723
74-097
71-440
72-863
74-231
71-585
73-002
74-365
71-730
73-141
74-498
71-874
73-279
74-631
72-017
73-417
74-764
72-159
73-554
74-896
4
5
6
75-027
76-315
77-553
75-158
76-441
77-674
75-289
76-566
77-795
75-419
76-691
77-916
75-548
76-816
78-036
75-677
76-940
78-156
75-805
77-064
78-275
75-933
77-187
78-394
76-061
77-310
78-513
76-188
77-432
78-631
7
8
9
78-749
79-905
81-024
78-866
80-019
81-134
78-983
80-132
81-244
79-100
80-245
81-353
79-216
80-357
81-462
79-332
80-469
81-570
79-447
80-581
81-678
79-562
80-692
81-786
79-677
80-803
81-893
79-791
80-914
82-000
4
82-107
83-157
84-176
82-213
83-260
84-276
82-319
83-363
84-376
82-425
83-466
84-476
82-530
83-568
84-575
82-635
83-670
84-674
82-740
83-772
84-773
82-845
83-873
84-872
82-949
83-974
84-970
83-053
84-075
85-068
3
4
5
85-166
86-128
87-064
85-263
86-223
87-156
85-360
86-317
87-248
85-457
86-411
87-340
85-554
86-505
87-432
85-650
86-599
87-523
85-746
86-692
87-614
85-842
86-785
87-705
85-938
86-878
87-795
86-033
86-971
87-885
6
7
8
87-975
88-861
89-724
88-065
88-948
89-809
88-154
89-035
89-894
88-243
89-122
89-979
88-332
89-209
90-063
88-421
89-295
90-147
88-509
89-381
90-231
88-597
89-467
90-315
88-685
89-553
90-399
88-773
89-639
90-482
5
9
1
90-565
91-385
92-186
90-648
91-466
92-265
90-731
91-547
92-344
90-814
91-628
92-423
90-896
91-708
92-501
90-978
91-788
92-579
91-060
91-868
92-657
91-142
91-948
92-735
91-223
92-028
92-813
91-304
92-107
92-891
2
3
4
92-968
93-732
94-479
93-045
93-807
94-553
93-122
93-882
94-627
93-199
93-957
94-701
93-276
94-032
94-774
93-352
94-107
94-847
93-428
94-182
94-920
93-504
94-257
94-993
93-580
94-331
95-066
93-656
94-405
95-138
•00
•01
•02
•03
•04
•05
•p6
•07
•08
•09
INTERNAL BALLISTICS
441
Table shoxmng the work hi foot-tons that 1 lb. of f/unpowder, etc. — continued.
Work in Foot-tons.
Work in Foot-tons.
95-210
95-925
96-625
97-310
97-981
99-282
99-915
100-536
101-145
101-744
102-333
102-912
103-480
104-038
104-586
105-125
105-655
106-176
106-688
107-192
107-688
108-177
108-659
109-133
109-600
110-060
110-514
110-962
111-404
111-840
112-270
112-695
113-114
113-528
113-937
114-341
114-739
115-133
115-521
115-905
116-284
95-282
95-996
96-694
97-378
98-047
98-703
99-346
99-978
100-598
101-205
101-803
102-391
102-969
103-536
104-093
104-640
105-178
105-708
106-228
106-739
1U7-242
107-737
108-226
108-707
109-180
109-646
110-106
110-559
111-007
111-448
111-883
112-313
112-737
113-156
113-569
113-978
114-381
114-779
115-172
115-559
115-943
116-322
116-659 116-696 116-733
117-029 117-066 117-103
117-395 117-432 117
95-354
96-066
96-763
97-446
98-113
98-768
99-410
100-041
100-659
101-265
101-862
102-449
103-026
103-592
104-148
104-694
105-231
105-760
106-280
106-790
107-292
95-426
96-136
96-832
97-513
98-179
98-833
99-474
100-103
100-721
101-325
101-921
102-507
103-083
103-648
104-203
104-748
105-284
105-812
106-331
106-841
107-342
109-227
109-692
110-152
110-604
111-051
111-492
111-926
112-356
112-779
113-197
113-610
114-018
114-421
114-818
115-211
115-598
115-981
116-359
95-498
96-206
96-901
97-580
98-245
99-537
100-165
100-782
101-385
101-980
102-565
103-140
103-704
104-258
104-802
105-3.37
105-864
106-382
106-892
107-392
95-570
96-276
96-970
97-647
98-311
98-962
99-600
100-227
100-843
101-445
102-039
102-623
103-197
103-760
104-313
104-856
105-390
105-916
95-641
96-346
97-038
97-714
98-377
99-026
99-663
100-289
100-904
101-505
102-098
102-681
103-254
103-816
104-368
104-910
105-443
105-968
106-433 106-484
106-942 106-992
107-442 107-492
107-786 107-835 107-884 107-933
108-274 108-323 108-371 108-419
108-755 108-803 108-851 108-898
109-274
109-738
110-198
110-649
111-096
111-536
111-969
112-399
112-821
113-239
113-651
114-059
114-461
114-858
115-250
115-636
116-019
116-397
116-770
117-140
117-505
109-321
109-784
110-244
110-694
111-140
111-580
112-012
112-442
112-863
113-280
113-692
114-099
114-501
114-897
115-289
115-675
116-057
116-434
116-807
117-176
117-541
109-368
109-830
110-289
110-739
111-184
111-624
112-055
112-485
112-905
113-322
113-733
114-140
114-541
114-937
115-328
115-713
116-095
116-472
116-844
117-213
117-577
107-982
108-467
108-945
109-415
109-876
110-334
110-784
111-228
111-668
112-098
112-527
112-947
113-363
113-774
114-180
114-581
114-976
115-367
115-752
116-133
116-509
116-881
117-249
117-613
95-712
96-416
97-106
97-781
98-443
99-090
99-726
100-351
100-965
101-565
102-157
102-739
103-311
103-872
104-423
104-964
105-496
106-020
106-535
107-042
107-541
108-031
108-515
108-992
109-462
109-922
110-379
110-829
111-272
111-711
112-141
112-569
112-989
113-405
113-815
114-221
114-621
115-016
115-405
115-790
116-171
116-547
116-918
117-286
117-649
95-783
96-486
97-174
97-848
98-508
99-154
100-413
101-025
101-625
102-216
102-797
103-368
103-928
104-478
105-018
105-549
106-072
106-586
107-092
107-590
108-080
108-563
109-039
109-508
109-968
110-424
110-873
111-316
111-754
112-184
112-611
113-031
113-446
113-856
114-261
114-660
115-055
115-444
115-829
116-209
116-584
116-955
117-332
117-685
113-487
113-897
114-301
114-700
115-094
115-482
115-867
116-247
116-622
116-992
117-359
117-721
442
NTERNAL BALLISTICS
Table showinf/ the loork in foot-tons that 1 lb. of c/unpowder, etc. — continued.
Work in Foot-tons.
Work in Foot-ton.s
10-0
11
VI
13
14
If.
16
17
18
19
20
21
22
23
24
25
26
117-757
121-165
124-239
127-036
129-602
131-970
134-168
136-218
138-138
139-944
141-647
143-258
144-788
146-242
147-629
148-960
150-232
151-452
152-622
153-743
154-819
155-857
156-856
157-824
158-771
159-678
160-556
161-411
162-241
163-046
118-114
121-486
124-531
127-302
129-846
132-195
134-382
136-413
138-326
140-117
141-813
143-415
144-937
146-383
147-765
149-090
150-356
151-571
152-736
153-852
154-925
155-959
156-954
157-919
158-863
159-767
160-643
161-495
162-323
163-125
118-468
121-804
124-820
127-566
130-089
132-419
134-594
136-608
138-513
140-289
141-977
143-571
145-085
146-524
147-900
149-219
150-480
151-690
152-850
153-961
155-030
156-060
157-052
158-014
158-955
159-856
160-729
161-579
162-404
163-204
118-818
122-119
125-107
127-828
130-330
132-642
134-804
136-802
138-698
140-461
142-140
143-726
145-233
146-664
148-035
149-348
150-603
151-808
152-963
154-069
155-135
156-161
157-150
158-108
159-046
159-944
160-815
161-663
162-485
163-283
119-164
122-431
125-391
128-088
130-570
132-863
135-012
136-995
138-881
140-632
142-302
143-880
145-380
146-803
148-169
149-476
150-726
151-926
153-076
154-177
155-239
156-262
157-247
158-202
159-137
160-032
160-901
161-746
162-566
163-361
119-506
122-739
125-671
128-346
130-809
133-083
135-218
137-187
139-063
140-803
142-463;
144-033
145-526
146-942
148-302
149-603
150-848
152-043
153-188
154-285
155-343
156-362
157-344
158-296
159-228
160-120
160-987
161-829
162-647
163-439
119-845
123-045
125-949
128-602
131-040
133-302
135-422
137-379
139-243
140-973
142-623
144-186
145-671
147-080
148-435
149-730
150-970
152-160
153-300
154-393
155-447
156-461
157-441
158-390
159-319
160-208
161-072
161-912
162-727
163-517
120-180
123-347
126-224
128-856
131-281
133-520
135-624
137-570
139-421
141-143
142-782
144-338
145-815
147-218
148-567
149-856
151-091
152-276
153-411
154-500
155-550
156-560
157-537
158-483
159-409
160-295
161-157
161-995
162-807
163-595
120-512
123-646
126-507
129-107
131-513
133-737
135-824
137-760
139-597
141-312
142-941
144-489
145-958
147-355
148-699
149-982
151-212
152-392
153-522
154-607
155-653
156-659
157-633
158-586
159-499
160-382
161-242
162-077
162-887
163-673
120-840
123-944
126-777
129-356
131-743
133-953
136-022
137-949
139-771
141-480
143-099
144-639
146-100
147-492
148-830
150-107
151-332
152-507
153-633
154-713
155-755
156-758
157-729
158-679
159-589
160-469
161-327
162-159
162-967
163-751
INTERNAL BALLISTICS
443
Tahle giving in foot-tons the energy stored up in 1 Ih. in vmght, moving at any velocity
between 10 and 3000 feet per second.
1
>
1
'
3
4
5
6
7
8
g
10
20
•0007
•0028
•0008
•0031
-0010
-0034
•0012
•0037
-0014
-0040
•0016
•0043
•0018
•0047
•0020
•0050
-0022
-0054
•0025
•0058
30
40
50
•0062
•0111
•0173
•0067
•0117
•0180
•0071
•0122
•0187
•0076
•0128
•0194
•0080
•0134
•0201
•0085
•0140
•0209
-0090
-0147
-0217
•0095
•0153
•0225
-0100
•0160
•0233
•0105
•0166
•0241
60
70
80
•0250
•0339
•0444
•0258
•0349
•0455
•0267
•0359
•0466
•0275
•0369
•0477
•0284
•0380
•0489
•0293
•0390
•0501
-0302
-0400
-0513
•0311
•0411
•0525
•0321
•0422
•0537
•0330
•0433
•0549
90
100
110
•0561
•0693
•0839
•0574
•0707
•0854
•0587
•0721
•0870
-0600
-0735
-0885
•0613
•0749
•0901
•0626
•0764
•0917
-0639
-0779
-0933
•0652
•0794
•0949
•0665
•0809
•0965
•0679
•0824
•0982
120
130
140
•0999
•1172
•1359
•1015
•1190
•1379
•1032
•1208
•1398
•1049
•1226
•1418
•1066
•1245
•1438
•1083
•1264
•1458
-1101
-1283
-1478
•1118
•1301
•1498
-1136
-1321
•1519
•1154
•1340
•1539
150
160
170
•1560
•1775
•2004
•1581
•1797
•2028
•1602
•1820
-2051
•1623
•1842
-2075
•1644
•1865
•2099
•1666
•1888
•2124
-1687
•1911
-2148
•1709
•1934
•2172
•1731
•1957
•2197
•1753
•1980
•2222
180
190
200
•2247
•2503
•2774
•2272
•2530
•2801
•2297
•2556
•2829
-2322
-2583
-2857
•2348
•2610
•2886
•2373
•2637
•2914
-2399
-2664
-2942
•2425
•2691
•2971
•2451
•2718
•3000
•2477
•2746
•3029
210
220
230
•3058
•3356
•3668
•3087
•3387
•3700
•3116
•3417
•3732
-3146
-3448
-3764
•3176
•3479
•3797
•3205
•3510
•3829
•3235
•3542
•3862
•3265
•3573
•3895
•3295
•3604
•3928
•3326
•3636
•3961
240
250
260
•3994
•4334
•4687
•4027
•4369
•4724
•4061
•4403
•4760
-4095
-4438
-4796
•4128
•4474
•4833
•4162
•4509
•4869
•4196
•4544
•4906
•4230
-4580
-4943
•4265
•4616
•4980
•4299
•4651
•5018
270
280
290
•5055
•5436
•5832
•5092
•5475
•5872
•5130
•5514
•5912
-5168
-5553
-5953
•5206
•5593
•5994
•5244
•5632
•6034
•5282
•5672
•6075
-5320
-5712
-6116
•5359
•5751
•6158
•5397
•5791
•6199
300
310
320
•6241
•6664
•7100
•6282
•6707
•7145
•6324
•6750
•7190
-6366
•6793
-7234
•6408
•6837
•7279
•6450
•6880
•7324
•6493
•6924
•7369
-6535
-6968
•7415
•6578
•7012
•7460
•6621
•7056
•7506
330
340
350
•7551
•8016
•8494
•7597
•8063
•8542
•7643
•8110
•8591
-7689
-8158
-8640
•7735
•8205
•8689
•7782
•8253
•8738
•7828
•8301
•8788
•7875
-8349
-8837
•7922
•8397
•8887
•7969
•8446
•8936
360
370
380
•8986
•9493
1^0013
•9036
•9544
1^0066
•9086
•9596
l^OllS
-9136
-9647
1-0171
•9187
•9698
1^0225
•9238
•9751
1^0278
•9289
•9803
1^0331
-9339
-9855
1-0385
•9390
•9908
1^0439
•9441
•9960
1^0493
390
400
410
1^0547
1^1095
1^1656
1^0601
P1150
1-1713
1-0655
1-1-206
1-1,770
1-0710
1-1262
1-1827
r0764
1^1318
1^1884
1^0819
1^1374
1^1942
1^0873
1^1430
1^2000
1-0929
1-1486
1^2058
1^0984
1^1543
1^2115
1-1039
1-1599
1-2174
420
430
440
1^2232
1-2821
1-3424
1^2290
1^2881
1-3485
1-2348
1-2941
1^3547
1-2407
1-3005
1-3608
1^2466
1^3061
1^3670
1^2525
1^3121
1^3731
1^2584
1^3181
1^3793
1^2643
1^3242
1-3855
1^2702
1^3303
1^3917
P2762
1 •3363
1^3979
i
1
2
3
4
5
6
7
8
9
444
INTERNAL BALLISTICS
Table of enen/ies — continued.
1
!3
1
1
2
3
4
5
6
7
8
9
440
450
460
1-3424
1 -4042
1-4673
1-3485
1-4104
1-4736
1-3547
1-4167
1-4800
1-3608
1-4229
1-4865
1-3670
1-4292
1-4929
1 -3731
1-4355
1-4993
1-3793
1-4418
1-5058
1-3855
1-4482
1-5122
1-3917
1-4545
1-5187
1-3979
1-4609
1-5252
470
480
490
1-5317
1-5976
1-6649
1-5383
1-6043
1-6717
1-5448
1-6110
1-6785
1-5514
1-6176
1-6853
1-5579
1-6243
1-6922
1-5645
1-6311
1-6990
1-5711
1-6378
1-7059
1-5777
1 -6445
1-7128
1-5843
1-6513
1-7197
1-5910
1-6581
1-7266
500
510
520
1-7335
1-8036
1-8750
1-7405
1-8106
1-8822
1-7474
1-8177
1-8894
1-7544
1-8248
1-8967
1-7614
1-8320
1-9039
1-7684
1-8391
1-9112
1-7754
1-8462
1-9185
1-7824
1-8534
1-9258
1-7894
1-8606
1-9331
1-7964
1-8678
1-9404
530
540
550
1-9478
2-0220
2-0976
1-9551
2-0295
2-1052
1-9625
2-0370
2-1128
1-9699
2-0445
2-1205
1-9773
2-0520
2-1282
1-9847
2-0596
2-1359
1-9921
2-0672
2-1436
1-9996
2-0747
2-1513
2-0070
2-0823
2-1590
2-0145
2-0899
2-1668
560
570
580
2-1745
2-2529
2-3326
2-1823
2-2608
2-3407
2-1901
2-2687
2-3487
2-1979
2-2767
2-3568
2-2057
2-2846
2-3649
2-2135
2-2926
2-3730
2-2214
2-3006
2-3811
2-2292
2 -3086
2-3893
2-2371
2-3166
2-3974
2-2450
2-3246
2-4056
590
600
610
2-4138
2-4963
2-5802
2-4219
2-5046
2-5886
2-4302
2-5129
2-5971
2-4384
2-5213
2-6056
2-4466
2-5297
2-6141
2-4548
2-5380
2-6226
2-4631
2-5464
2-6312
2-4714
2-5549
2-6397
2-4797
2-5633
2-6483
2-4880
2-5717
2-6569
620
630
640
2-6655
2-7521
2-8402
2-6741
2-7610
2-8491
2-6827
2-7696
2-8580
2-6913
2-7784
2-8669
2-7000
2-7872
2-8758
2-7086
2-7960
2-8847
2-7173
2-8048
2-8937
2-7260
2-8136
2-9027
2-7347
2-8225
2-9117
2-7434
2-8313
2-9206
650
660
670
2-9297
3-0205
3-1127
2-9387
3-0296
3-1220
2-9477
3-0388
3-1313
2-9568
3-0480
3-1406
2-9658
3-0572
3-1500
2-9749
3-0664
3-1593
2-9840
3-0757
3-1687
2-9931
3-0849
31781
3-0022
3-0942
3-1875
3-0113
3-1034
3-1969
680
690
700
3-2063
3-3013
3-3977
3-2157
3-3109
3-4074
3-2258
3-3205
3-4171
3-2347
3-3301
3-4269
3-2442
3-3397
3-4366
3-2537
3-3493
3-4464
3-2632
3-3590
3-4562
3-2727
3-3686
3-4660
3-2822
3-3783
3-4758
3-2918
3-3880
3-4856
710
720
730
3-4955
3-5946
3-6952
3-5053
3-6046
3-7053
3-5152
3-6146
3-7155
3-5251
3-6247
3-7256
3-5350
3-6347
3-7358
3-5449
3-6447
3-7460
3-5548
3-6548
3-7562
3-5647
3-6649
3-7664
3-5747
3-6750
3-7766
3-5847
3-6851
3-7869
740
750
760
3-7971
3-9004
4-0051
3-8074
3-9108
4-0157
3-8177
3-9213
4-0262
3-8280
3-9317
4-0368
3-8383
3-9421
4-0474
3-8486
3-9526
4-0580
3-8589
3-9631
4-0686
3-8693
3-9736
4-0792
3-8797
3-9841
4-0898
3-8900
3-9946
4-1005
770
780
790
4-1112
4-2187
4-3276
4-1219
4-2295
4-3385
4-1326
4-2404
4-3495
4-1433
4-2512
4-3605
4-1540
4-2621
4-3715
4-1648
4-2730
4-3825
4-1755
4-2839
4-3934
4-1863
4-2948
4-4044
4-1971
4-3057
4-4155
4-2079
4-3166
4-4266
800
810
820
4-4378
4-5495
4-6625
4-4489
4-5607
4-6739
4-4600
4-5720
4-6853
4-4712
4-5832
4-6967
4-4823
4-5945
4-7081
4-4935
4-6058
4-7195
4-5046
4-6171
4-7310
4-5158
4-6284
4-7424
4-5270
4-6398
4-7539
4 -.-.382
4-6511
4-7654
830
840
850
4-7769
4-8927
5-0099
4-7884
4-9044
5-0217
4-7999
4-9160
5-0335
4-8115
4-9277
5-0453
4-8230
4-9394
5-0571
4-8346
4-9511
5-0690
4-8462
4-9628
5-0809
4-8578
4-9745
5-0927
4-8694
4-9863
5-1046
4-8811
4-9981
5-1165
860
870
880
5-1285
5-2i83
5-3696
5-1404
5-2604
5-3818
5-1523
5-2725
5-3940
5-1643
5-2846
5-4062
5-1763
5-2967
5-4185
5-1883
5-3088
5-4308
5-2003
5-3209
5-4431
5-2123
5-3330
5-4554
5-2243
5-3452
5-4677
5-2364
5-3574
5-4800
1
2
3
4
5
6
7
8
9
INTERNAL BALLISTICS
Table of energies — continued.
445
1
1
2
3
4
5
6
7
8
9
880
890
900
5-3696
5-4925
5-6166
5-3818
5-5048
5-6291
5-3940
5-5172
5-6416
5-4062
5-5296
5-6541
5-4185
5-5420
5-6666
5-4308
5-5544
5-6792
5-4431
5-5668
5-6918
5 -4554
5-5792
5-7043
5-4677
5-5917
5-7169
5-4800
5-6041
5-7295
910
920
930
5-7421
5-8690
5-9973
5-7547
5-8818
6-0102
5-7674
5-8946
6-0231
5-7800
5-9074
6-0360
5-7927
a-9202
6-0490
5-8054
5-9330
6-0620
5-8181
5-9458
6-0749
5-8308
5-9587
6-0879
5-8435
5-9715
6-1009
5-8563
5-9844
6-1139
940
950
960
6-1270
6-2580
6-3905
6-1400
6-2712
6-4038
6-1531
6-2844
6-4171
6-1661
6-2976
6-4305
6-1792
6-3108
6-4438
6-1923
6-3241
6-4572
6-2054
6-3373
6-4700
6-2186
6-3506
6-4840
6-2317
6-3639
6-4974
6-2448
6-3772
6-5108
970
980
990
6-5243
6-6595
6-7961
6-5377
6-6731
6-8098
6-5512
6-6867
6-8236
6-5647
6-7003
6-8374
6-5782
6-7140
6-8512
6-5917
6-7276
6-8649
6-6052
6-7413
6-8787
6-6187
6-7550
6-8926
6-6324
6-7687
6-9064
6-6459
6-7824
6-9202
1000
1010
1020
6-9341
7-0735
7-2142
6-9480
7-0875
7-2284
6-9619
7-1015
7-2426
6-9758
7-1156
7-2567
6-9897
7-1296
7-2709
7-0036
7-1437
7-2851
7-0176
7-1578
7-2994
7-0315
7-1719
7-3136
7-0455
7-1860
7-3278
7-0595
7-2001
7-3421
1030
1040
1050
7-3564
7-4999
7-64 48
7-3706
7-5143
7-6594
7-3850
7-5288
7-6740
7-3993
7-5433
7-6886
7-4136
7-5577
7-7032
7-4279
7-5722
7-7178
7-4423
7-5867
7-7325
7-4567
7-6012
7-7471
7-4711
7-6157
7-7617
7-4855
7-6302
7-7764
1060
1070
1080
7-7911
7-9388
8-0879
7-8059
7-9537
8-1029
7-8206
7-9686
8-1179
7-8353
7-9834
8-1329
7-8501
7-9983
8-1480
7-8648
8-0132
8-1630
7-8796
8-0281
8-1780
7-8944
8-0431
8-1931
7-9092
8-0580
8-2082
7-9240
8-0730
8-2233
1090
1100
1110
8-2384
8-3903
8-5435
8-2535
8-4055
8-5590
8-2687
8-4208
8-5743
8-2838
8-4361
8-5897
8-2990
8-4514
8-6052
8-3142
8-4667
8-6206
8-3294
8-4820
8-6361
8-3446
8-4974
8-6516
8-3598
8-5127
8-6671
8-3750
8-5281
8-6826
1120
1130
1140
8-6981
8-8541
9-0116
8-7137
8-8698
9-0274
8-7292
8-8855
9-0432
8-7448
8-9012
9-0590
8-7604
8-9169
9-0749
8-7760
8-9327
9-0908
8-7916
8-9484
9-1067
8-8072
8-9642
9-1226
8-8228
8-9800
9-1385
8-8385
S-9958
9-1544
1150
1160
1170
9-1703
9-3305
9-4921
9-1863
9-3466
9-5083
9-2023
9-3627
9-5246
9-2183
9-3788
9-5408
9-2343
9-3950
9-5571
9-2503
9-4111
9-5734
9-2663
9-4273
9-5897
9-2823
9-4435
9-6060
9-2984
9-4597
9-6223
9-3144
9-4759
9-6387
1180
1190
1200
9-6550
9-8194
9-9851
9-6714
9-8359
10-0018
9-6878
9-8524
10-0184
9-7042
9-8690
10-0351
9-7206
9-8855
10-0518
9-7370
9-9021
10-0685
9-7535
9-9186
10-0852
9-7699
9-9352
10-1019
9-7864
9-9518
10-1187
9-8029
9-9685
10-1354
1210
1220
1230
10-1522
10-3207
10-4906
10-1690
10-3376
10-5077
10-1858
10-3546
10-5247
10-2026
10-3715
10-5418
10-2194
10-3885
10-5589
10-2363
10-4055
10-5761
10-2531
10-4225
10-5932
10-2700
10-4395
10-6103
10-2869
10-4565
10-6275
10-3038
10-4735
10-6447
1240
1250
1260
10-6619
10-8345
11-0086
10-6791
10-8519
11-0261
10-6963
10-8692
11-0436
10-7135
10-8866
11-0611
10-7308
10-9040
11-0786
10-7483
10-9214
11-0961
10-7653
10-9388
11-1137
10-7826
10-9562
11-1312
10-7999
10-9737
11-1488
10-8172
10-9911
11-1664
1270
1280
1290
11-1840
11-3608
11-5390
11-2016
11-3786
11-5569
11-2193
11-3964
11-5748
11-2369
11-4141
11-5928
11-2546
11-4319
11-6107
11-2722
11-4498
11-6287
11-2899
11-4676
11-6466
11-3076
11-4854
11-6646
11-3254
11-5033
11-6826
11-3431
11-5211
11-7006
1300
1310
1320
11-7186
11-8996
12-0820
11-7367
11-9178
12-1003
11-7547
11-9360
12-1186
11-7728
11-9542
12-1370
11-7909
11-9724
12-1553
11-8089
11-9906
12-1737
11-8271
12-0089
12-1921
11-8452
12-0271
12-2105
11-8633
12-0454
12-2289
11-8814
12-0637
12-2473
1
2
3
4
5
6
7
8
9
446
INTERNAL BALLISTICS
Ta1)le of energies — continued.
1320
1330
1340
1350
1360
1370
1380
1390
1400
1410
1420
1430
1440
1450
1460
1470
1480
1490
1500
1510
15:20
1530
1540
1550
1560
1570
1580
1590
1600
1610
1620
1630
1640
1650
1660
1670
1680
1690
1700
1710
1720
1730
1740
1750
1760
12-0820
12-2657
12-4509
12-6374
12-8253
13-0146
13-2053
13-3974
13-5908
13-7857
13-9819
14-1795
14-3785
14-5789
14-7807
14-9839
15-1885
15-3944
15-6017
15-8104
16-0205
16-2320
16-4449
16-6592
16-8748
17-0919
17-3103
17-5301
17-7513
17-9739
18-1979
18-4232
18-6500
18-8781
19-1076
19-3385
19-5708
19-8045
20-0395
20-2760
20-5138
20-7531
20-9937
21-2357
21-4791
12-1003
12-2842
12-4695
12-6561
12-8442
13-0336
13-2244
13-4167
13-6103
13-8052
14-0016
14-1994
14-3985
14-5991
14-8010
15-0043
15-2090
15-4151
15-6225
15-8314
16-0416
16-2533
16-4663
16-6807
16-8965
17-1136
17-3322
17-5522
17-7735
17-9962
18-2203
18-4458
18-6727
18-9010
19-1306
19-3617
19-5941
19-8279
20-0631
20-2997
20-5377
20-7771
21-0178
21-2600
21-5035
12-1186
12-3026
12-4881
12-6749
12-8631
13-0526
13-2436
13-4360
13-6297
13-8248
14-0213
14-2192
14-4185
14-6192
14-8213
15-0247
15-2295
15-4358
15-6434
15-8524
16-0627
16-2745
16-4877
16-7022
16-9181
17-1354
17-3541
17-5742
17-7957
18-0186
18-2428
18-4684
18-6955
18-9239
19-1537
19-3849
19-6174
19-8514
20-0867
LA) -3235
20-5614
20-8011
21-0420
21-2842
21-5279
12-1370
12-3211
12-5067
12-6936
12-8820
13-0717
13-2628
13-4553
13-6491
13-8444
14-0411
14-2391
14-4385
14-6393
14-8415
15-0451
15-2501
15-4564
15-6642
15-8733
16-0838
16-2958
16-5090
16-7237
16-9398
17-1572
17-3761
17-5963
17-8179
18-0409
18-2653
18-4911
18-7182
18-9468
19-1767
19-4081
19-6408
19-8749
20-1103
20-3472
20-5855
20-8251
21-0661
21-3086
21-5524
12-1553
12-3396
12-5253
12-7124
12-9009
13-0907
13-2820
13-4746
13-6686
13-8640
14-0608
14-2590
14-4585
14-6595
14-8618
15-0656
15-2707
15*4772
15-6850
15-8943
16-1050
16-3170
16-5305
16-7453
16-9615
17-1791
17-3980
17-6184
17-8402
18-0633
18-2879
18-5137
18-7410
18-9697
19-1998
19-4313
19-6641
19-8983
20-1340
20-3710
20-6094
20-8491
21-0903
21-3329
21-5768
12-1737
12-3581
12-5440
12-7312
12-9198
13-1098
13-3012
13-4939
13-6881
13-8836
14-0806
14-2789
14-4786
14-6797
14-8821
15-0860
15-2913
15-4979
15-7059
15-9153
16-1261
16-3383
16-5519
16-7668
16-9832
17-2009
17-4200
17-6405
17-8624
18-0857
18-3104
18-5364
18-7638
18-9927
19-2229
19-4545
19-6875
19-9218
20-1576
20-3947
20-6333
20-8732
21-1145
21-3572
21-6013
12-1921
12-3766
12-5626
12-7500
12-9387
13-1289
13-3204
13-5133
13-7076
13-9033
14-1003
14-2988
14-4986
14-6998
14-9025
15-1065
15-3119
15-5186
15-7268
15-9363
16-1473
16-3596
16-5733
16-7884
17-0049
17-2228
17-4420
17-6627
17-8847
18-1081
18-3329
18-5591
18-7867
19-0156
19-2460
19-4777
19-7108
19-9454
20-1813
20-4185
20-6572
20-8973
21-1387
21-3815
21-6-258
12-2105
12-3952
12-5813
12-7688
12-9577
13-1479
12-2289
12-4137
12-6000
12-7876 12-8065
12-9766 12-9956
13-1671 13-1861:
13-3396 13-3588
13-5327 13-5520
13-7271 13-7466
13-9229
14-1201
14-3187
14-5187
14-7200
14-9228
15-1269
15-3325
15-5394
15-7477
15-9574
16-1684
16-3809
16-5948
16-8100
17-0266
17-2446
17-4640
17-6848
17-9070
18-1305
18-3555
18-5818
18-8095
19-0386
19-2691
19-5010
19-7342
19-9689
20-2049
20-4423
20-6812
20-9214
21-1629
21-4059
21-6503
13-9426
14-1399
14-3386
14-5388
14-7403
14-9432
15-1474
15-3531
15-5601
15-7686
15-9784
16-1896
16-4022
16-6162
16-8316
17-0483
17-2665
17-4860
17-7069
17-9293
18-1529
18-3780
18-6045
18-8323
19-0616
19-2922
19-5242
19-7576
19-9924
20-2286
20-4662
20-7051
20-9454
21-1872
21-4303
21-6748
INTERNAL BALLISTICS
Tahle of eiierc/ies — continued.
447
1760
1770
1780
1790
ISOO
ISIO
1820
1830
1840
1850
1860
1870
1880
1890
1900
1910
1920
1930
1940
1950
I960
21-4791
21-7238
21-9700
22-2175
22-4665
22-7168
22-9685
23-2216
23-4761
23-73-20
23-9892
24-2478
24-5079
24-7693
25-0321
25-2963
25-5618
25-8288
26-0972
26-3669
26-6380
21-5035
21 -7484
21-9947
•22-2424
22-4915
22-7419
■22-9938
23-2470
23-5016
23-7576
24-0150
24-2738
24-5340
24-7955
25-0585
25-3228
25-5885
25-8556
26-1241
26-3940
26-6652
1970 26-9105 26-9379
1980 27-1844 27-2119
1990 27-4597 27-4873
2000
2010
2020
2030
2040
2050
2060
2070
2080
2090
2100
2110
2120
2130
2140
2150
2160
2170
2180
2190
2200
27-7364
28-0145
28-2939
28-5747
28-8570
29-1406
29-4255
29-7119
29-9997
30-2888
30-5794
30-8713
31-1646
31-4593
31-7554
32-0529
32-3517
32-6520
32-9536
33-2566
33-5610
21-5279
21-7730
22-0194
22-2672
2-5164
22-7670
23-0190
23-2724
23-5272
23-7833
24-0408
24-2997
24-5601
24-8217
25-0848
•25-3493
25-6151
25-8824
26-1510
26-4210
26-6924
26-9652
27 '2394
27-5150
21 -5524
21-7975
22-0441
22-29-21
22-5414
22-7922
23-0443
23-2978
23-5527
23-8090
24-0667
24-3257
•24-5862
24-8480
25-1112
25-3758
25-6418
25-9092
26-1780
26-4481
26-7196
26-9926
27-2669
27-5426
21-5768
21-8221
22-0689
22-3170
'22-5664
22-8173
23-0696
23-3232
23-5783
23-8347
24-0925
24-3517
24-6123
24-8743
25-1376
25-4024
25-6685
25 -9360
26-2049
26-4752
26-7469
27-0199
27-2944
27-5702
27-7641
28-0423
28-3219
28-6029
28-8852
29-1690
29-4541
29-7406
30-0285
30-3178
30-6085
27-7919 27-8197 27-8475
28-0702 28-0981 28-1261
28-3500 28-3780 28-4061
28-6311
28-9136
29-1974
29-4827
29-7694
30-0574
30-3468
30-6377
30-9299
31-1940 31-2234
31-4889 31-5184
31-7851 31-8148
32-0827
32-3817
32-6821
32-9839
33-2870
33-5916
32-1125
32-4117
32-7122
33-0141
33-3174
33-6221
28-6593
28-9419
29-2259
29-5113
29-7981
30-0863
31-2529
31-5480
31-8445
32-1424
32-4417
32-7423
33-0444
33-3478
33-6526
28-6875
28-9702
29-2544
29-5399
29-8269
30-1152
30-4049
30-6960
30-9885
31-2823
31-5776
31-8742
32-1723
32-4717
32-7725
33-0747
33-3782
33-6832
21-6013
21-8467
22-0936
22-3418
22-5914
22-8425
23-0949
23-3487
23-6038
23-8604
24-1184
24-3777
24-6384
24-9005
25-1640
25-4289
25-6952
25-9628
26-2319
26-5023
26-7741
27-0473
27-3219
27-5979
27-8753
28-1540
28-4341
28-7157
28-9986
29-2829
29-5686
29-8556
30-1441
30-4339
30-7251
31-0178
31-3118
31-6072
31-9040
82-2021
32-5017
32-8026
33-1050
33-4087
33-7138
21-6258
21-8714
22-1184
•22-3667
22-6165
22-8677
23-1202
23-3741
23-6294
23-8861
24-1442
24-4037
24-6646
24-9268
25-1904
25-4555
25-7219
25-9897
26-2589
26-5294
26-8014
27-0747
27-3494
27-6256
27-9031
28-1820
28-4622
28-7439
29-0269
29-3114
29-5972
29-8844
30-1730
30-4630
30-7544
31-0471
31-3413
31-6368
31-9337
32-2320
32-5317
32-8328
33-1353
33-4391
33-7444
21-6503
21-8960
22-1431
22-3917
22-6416
22-8929
23-1455
23-3996
3-6550
23-9119
24-1701
24-4297
24-6907
24-9531
25-2169
25-4820
25-7486
26-0165
26-2858
26-5566
26-8287
27-1021
27-3770
27-6533
27-9309
28-2099
28-4903
28-7721
29-0553
29-3399
29-6259
29-9132
30-2020
31-0765
31-3708
31-6664
31-9635
32-2619
32-5617
32-8630
33-1656
33-4696
33-7750
21-6748
21-9-207
22-1679
22-4166
22-6666
22-9181
23-1709
23-4251
23-6806
23-9377
24-1960
24-4558
24-7169
24-9794
25-2433
25-5086
25-7753
26-0434
26-3129
26-5837
26-8559
27-1296
27-4046
27-6810
27-9587
28-2379
28-5185
28-8004
29-0837
29-3684
29-6545
29-9420
30-2309
30-5212
30-8128
31-1058
31-4003
31-6961
31-9933
32-2919
32-5918
32-8932
33-1959 33-2263
33-5001 33-5305
33-8056 33-8362
448
INTERNAL BALLISTICS
Tahle of energies — continued.
2200
2210
2220
2230
2240
2250
2260
2270
2280
2290
2300
2310
2320
2330
2340
2350
2360
2370
2380
2390
2400
2410
2420
2430
2440
2450
2460
2470
2480
2490
2500
2510
2520
2530
2540
2550
2560
2570
2580
2590
2600
2610
2620
2630
2640
33-5610
33-8668
34-1740
33-5916
33-8975
34-2048
34-4826 34-5135
34-7925 34-8236
35-1039 35-1351
35-4166
35-7307
36-0462
36-3631
36-6814
37-0010
37-3221
37-6445
37-9684
38-2936
38-6202
38-9481
39-2775
39-6083
39-9404
40-2739
40-6089
40-9452
41-2829
41-6219
41-9624
42-3043
42-6475
42-9921
43-3381
43-6855
44-0343
44-3845
44-7360
45-0890
45-4433
45-7990
46-1561
46-5146
46-8745
47-2358
47-5984
47-9625
48-3279
35-4480
35-7622
36-0779
36-3949
36-7133
37-0331
37-3543
37-6769
38-0008
38-3261
38-6529
38-9810
39-3105
39-6414
39-9737
40-3074
40-6424
40-9789
41-3167
41-6559
41-9965
42-3385
42-6819
43-0267
43-3728
43-7203
44-0693
44-4196
44-7713
45-1244
45-4788
45-8347
46-1919
46-5505
46-9106
47-2720
47-6348
47-9990
48-3645
33-6221
33-9282
34-2356
34-5445
34-8547
35-1663
35-4793
35-7937
36-1095
36-4267
36-7452
37-0651
37-3865
37-7092
38-3588
38-6856
39-0139
39-3436
39-6746
40-0070
40-3408
40-6760
41-0126
41-3506
41-6899
42-0307
42-3728
42-7161
43-0612
43-4075
43-7552
44-1042
44-4547
44-8065
45-1597
45-5144
45-8703
46-2277
46-5865
46-9467
47-3082
47-6711
48-0355
48-4011
33-6526
33-9588
34-2664
34-5754
34-8858
35-1976
35-5107
35-8252
36-1411
33-6832
33-9895
34-2973
34-6064
34-9169
35-2288
35-5421
35-8568
36-1728
36-4585 36-4903
36-7771 36-8091
37-0972 37-1293
37-4187
37-7415
38-0658
38-3914
38-7184
39-0468
39-3766
39-7078
40-0403
40-3743
40-7096
41-0463
41-3844
41-7239
42-0648
42-4071
42-7507
43-0958
43-4422
43-7900
44-1392
44-4898
44-8418
45-1951
45-5499
45-9060
46-2635
46-6225
46-9828
47-3444
47-7075
48-0720
48-4378
37-4509
37-7739
38-0983
38-4240
38-7512
39-0797
39-4097
39-7410
40-0737
40-4077
40-7432
41-0801
41-4183
41-7580
42-0990
42-4414
42-7852
43-1304
43-4769
43-8249
44-1742
44-5249
44-8771
45-2306
45-5854
45-9417
46-2994
46-6584
47-0189
47-3807
47-7439
48-1085
48-4745
33-7138
34-0203
34-3281
34-6374
34-9480
35-2601
35-5735
35-8883
36-2045
36-5221
36-8410
37-1614
33-7444
34-0510
34-3590
34-6684
34-9792
35-2914
35-6049
35-9199
36-2362
36-5539
36-8730
37-1935
33-7750 33-8056 33-8362
34-0817 34-1125 34-1432
34-3899 34-4208 34-4517
34-6994
35-0103
35-3226
35-6363
35-9514
36-2679
36-5858
36-9050
37-2256
37-4831 37-5154 37-5477
37-8063 37-8387 37-8711
38-1308 38-1633 38-1959
38-4567
38-7840
39-1127
39-4427
39-7742
40-1070
40-4412
40-7768
41-1138
41-4522
41-7920
42-1332
42-4757
42-8196
43-1649
43-5117
43-8597
44-2092
44-5600
44-9123
45-2656
45-6210
45-9774
46-3352
46-6944
47-0550
47-4169
38-4894
38-8168
39-1456
39-4758
39-8074
40-1404
40-4747
40-8105
41-1476
41-4861
41-8260
42-1673
42-5100
42-8541
43-1996
43-5464
43-8946
44-2442
44-5952
44-9476
45-3014
45-6566
46-0131
46-3711
38-5220
38-8496
39-1786
39-5089
39-8406
40-1737
40-5082
40-8441
41-1814
41-5201
41-8601
42-2015
42-5444
42-8886
43-2342
43-5812
43-9295
44-2793
44-6304
44-9830
45-3369
45-6922
46-0489
46-4069
34-7304
35-0415
35-3540
35-6678
35-9830
36-6176
36-9370
37-2578
37-5799
37-9035
38-2284
38-5547
38-8824
39-2115
39*5420
39-8739
40-2071
34-7615
35-0727
35-3853
35-6993
36-0146
36-3314
36-6495
36-9690
37-2899
37-6122
37-9359
38-2610
38-5874
38-9153
39-2445
39-5751
39-9071
40-2405
46-7304 46-7664
47-0911 47-1273
47-4532 47-4895
47-7803 47-8167
48-1450 48-1816
48-5111 48-5478
47-8531
48-2181
48-5845
40-5418 40-5753
40-8778 40-9115
41-2152 41-2490
41-5540
41-8942
42-2358
42-5787
42-9231
43-2688
43-6159
43-9644
44-3143
44-6656
45-0183
45-3723
45-7278
46-0846
46-4428
46-8024
47-1634
47-5258
47-8896
48-2547
48-6212
41-5880
41-9283
42-2700
42-6131
42-9576
43-3035
43-6507
43-9993
44-3494
44-7008
45-0536
45-4078
45-7634
46-1204
46-4787
46-8384
47-1996
47-5621
47-9260
48-2913
48-6580
INTERNAL BALLISTICS
Table of enerc/ies— continued.
449
1
>
1
2
3
4
5
6
7
8
9
2640
2650
2660
48-3279
48-6947
49-0629
48-3645
48-7315
49-0998
48-4011
48-7682
49-1367
48-4378
48-8050
49-1736
48-4745
48-8418
49-2106
48-5111
48-8786
49-2475
48-5478
48-9155
49-2845
48-5845
48-9523
49-3215
48-6212
48-9892
49-3585
48-6580
49-0260
49-3955
2670
2680
2690
49-4325
! 49-8035
50-1758
49-4695
49-8407
50-2132
49-5066
49-8778
50-2505
49-5437
49-9150
50-2878
49-5807
49-9523
50-3252
49-6178
49-9895
50-3625
49-6549
50-0267
50-3999
49-6920
50-0640
50-4373
49-7292
50-1013
50-4747
49-7663
50-1385
50-5122
2700
2710
2720
50-5496
50-9247
51-3012
50-5870
50-9623
51-3390
50-6245
50-9999
51-3767
50-6620
51-0375
51-4145
50-6995
51-0752
51-4522
50-7370
51-1128
51-4900
50-7745
51-1505
51-5278
50-8120
51-1881
51-5656
50-8496
51-2258
51-6035
50-8871
51-2635
51-6413
2730
2740
2750
51-6792
52-0584
52-4391
51-7170
52-0965
52-4773
51-7549
52-1345
52-5154
51-7928
52-1725
52-5536
51-8307
52-2106
52-5918
51-8686
52-2486
52-6300
51-9066
52-2867
52-6682
51-9445
52-3248
52-7064
51-9825
52-3629
52-7447
52-0205
52-4010
52-7829
2760
2770
2780
52-8212
53-2047
53-5895
52-8595
53-2431
53-6281
52-8978
53-2815
53-6666
52-9361
53-3200
53-7052
52-9744
53-3584
53-7438
53-0128
53-3969
53-7824
53-0511
53-4354
53-8211
53-0895
53-4739
53-8597
53-1279
53-5124
53-8984
53-1662
53-5510
53-9370
2790
2800
2810
53-9757
54-3633
54-7523
54-0144
54-4022
54-7913
54-0531
54-4410
54-8303
54-0919
54-4799
54-8693
54-1306
54-5188
54-9083
54-1694
54-5577
54-9474
54-2081
54-5966
54-9864
54-2469
54-6355
55-0255
54-2857
54-6744
55-0645
54-3245
54-7134
55-1036
2820
2830
2840
55-1427
55-5345
55-9277
55-1819
55-5738
55-9670
55-2210
55-6130
56-0065
55-2601
55-6523
56-0459
55-2993
55-6916
56-0853
55-3385
55-7309
56-1248
55-3776
55-7702
56-1642
55-4168
55-8096
56-2037
55-4560
55-8489
56-2432
55-4953
55-8883
56-2827
2850
2860
2870
56-3222
56-7182
57-1155
56-3618
56-7578
57-1553
56-4013
56-7975
57-1951
56-4409
56-8372
57-2350
56-4804
56-8769
57-2748
56-5200
56-9167
57-3147
56-5596
56-9564
57-3545
56-5992
56-9961
57-3944
56-6389
57-0359
57-4343
56-6785
57-0757
57-4743
2880
2890
2900
57-5142
57-9143
58-3158
57-5541
57-9544
58-3560
57-5941
57-9945
58-3962
57-6341
58-0346
58-4365
57-6741
58-0747
58-4768
57-7141
58-1148
58-5170
57-7541
58-1550
58-5573
57-7941
58-1952
58-5976
57-8342
58-2354
58-6380
57-8742
58-2756
58-6783
2910
2920
2930
58-7187
59-1229
59-5286
58-7590
59-1624
59-5692
58-7994
59-2029
59-6099
58-8398
59-2435
59-6505
58-8802
59-2840
59-6912
58-9206
59-3246
59-7319
58-9610
59-3651
59-7726
59-0015
59-4067
59-8133
59-0419
59-4473
59-8541
59-0824
59-4879
59-8948
2940
2950
2960
59-9356
60-3440
60-7538
59-9764
60-3849
60-7949
60-0172
60-4259
60-8359
60-0580
60-4668
60-8770
60-0988
60-5078
60-9181
60-1396
60-5487
60-9592
60-1805
60-5897
61-0004
60-2213
60-6307
61-0415
60-2622
60-6717
61-0827
60-3031
60-7128
61-1238
2970
2980
2990
61-1650
61-5776
61-9915
61-2062
61-6189
62-0330
61-2474
61-6603
62-0745
61-2886
61-7016
62-1160
61-3299
61-7430
62-1575
61-3711
61-7844
62-1991
61-4124
61-8258
62-2406
61-4537
61-8672
62-2822
61-4950
61-9086
62-3237
61-5363
61-9501
62-3653
3000
62-4069
62-4486
62-4902
62-5318
62-5734
62-6151
62-6568
62-6985
62-7402
62-7818
1
2
3
4
5
6
7
8
9
2 F
450
INTERNAL BALLISTICS
Table givimf in dinamodes the PMerffy stored up in 1 kilogramme in weight, moving at any velocity
between 1 and 1000 m,etres per second.
'5
1
2
3
4
5
6
7
8
9
10
20
•00000
•00510
•02038
•00005
•00617
-02247
•00020
•00734
•02466
•00045
•00861
■02696
-00082
-00999
-02935
-00127
•01147
-03185
-00183
-01305
•03445
•00249
•01473
-03715
■00326
-01651
-03995
-00413
-01840
•04285
30
40
50
•04586
•08154
•12740
•04897
-08566
•13255
•05218
•08989
•13780
•05549
•09422
-14315
•05891
•09866
•14860
•06243
•10319
•15415
•06604
•10783
•15981
-06976
-11257
-16557
-07359
-11741
-17143
•07756
•12235
•17739
60
70
80
•18346
•24970
•32614
•18962
•25689
•33435
•19589
•26418
•34265
•20226
•27157
•35106
•20873
•27906
-35957
•21531
•28665
•36819
•22198
•29434
•37690
-22876
-30214
•38572
-23564
•31004
-39463
•24262
•31804
•40365
90
100
110
•41278
•50960
•61662
•42200
•51984
•62788
•43132
•53019
•63924
•44075
•54063
•65071
-45028
•55118
•66228
•45991
•56183
•67395
•46965
•57258
■68572
•47948
•58344
•69789
■48942
•59440
-70957
•49946
•60546
•72164
120
130
140
•73382
•86122
•99881
■74610
•87452
]^01314
•75849
•88793
1^02756
•77097
•90143
1^04208
•78356
•91504
1-05671
•796-25
•92875
1-07143
•80904
•94256
1-08626
•82193
•95647
ri0129
-83493
-97048
1-11623
•84802
•98460
1-13136
150
160
170
1^14660
1 •30458
1^47274
M6194
P32093
1^49012
1^17738
1-33740
1-50760
1-19292
1-35396
1-52518
1-20857
1-37062
1-54286
1-22431
r38739
1^56065
r24016
1^40425
r57854
r25611
1-42122
1-59652
1-27216
1-43829
1-61461
1-28832
1-45547
1-63281
180
190
200
1^65110
1 •83965
2^03840
1^66950
1^85907
2^05883
1^68880
1^87859
2^07937
1-70660
1-89821
2-10001
1-72530
1-91793
2-12075
1-74410
1-93775
2-14159
1-76301
1-95768
2-16254
1-78201
1-97771
2-18358
1-80113
1-99783
2-20473
1-82034
2-01807
2-22598
210
220
230
2^24734
2 •46646
2^69578
2^26879
2^48894
2^71928
2^29035
2^51151
2^74287
2-31200
2-53419
2-76657
2-33376
2-55697
2-79036
2-35563
•2-57985
2-81427
2-37759
2-60283
2-83827
2-39965
2-62592
2-86237
2-42182
2-64910
2-88658
2-44409
2-67239
2-91089
240
250
260
2 •93530
3-1 8500
3-44490
2^95980
3^21053
3^47145
2^98442
3^23616
3^48810
3-00913
3-26190
3-51485
3-03395
3-28773
3-54170
3-05887
3-31367
3-57867
3-08389
3-33971
3-60572
3-10902
3-36586
3-63288
3-13424
3-39210
3-66015
3-15957
3-41845
3-68752
270
280
290
3-71498
3^99526
4^28574
3-74255
4-02385
4-31534
3-77022
4^05254
4^34505
3-79800
4-08133
4-37486
3-82587
4-11023
4-40478
3-85385
4-13923
4-43480
3-88192
4-16832
4-46491
3-91010
4-19752
4-49513
3-93839
4-22683
4-52545
3-96678
4-25623
4-55587
300
310
320
4^58640
4-89726
5-21830
4^61703
4^92890
5^25097
4^64775
4-99065
5-28374
4-67859
4-99250
5-31660
4-70952
5-02445
5-34958
4-74055
5-05650
5-38265
4-77169
5-08866
5-41582
4-80293
5-1-2092
5-44910
4-83426
5-15328
5-48248
4-86571
5-18574
5-51596
330
340
350
5-54954
5-89098
6-24260
5 •583-22
5^92568
6^27832
5-61701
5-96048
6^31415
5-65090
5-99539
6-35007
5-68489
6-03040
6-38610
5-71899
606551
6-42223
5-75318
6-10073
6-45847
5-78748
6-13604
6-49480
5-82187
6-17146
6-53124
5-85637
6-20698
6-56777
360
370
380
6-60442
6-97642
7-35862
6-64116
7^01418
7^39740
6-67800
7-05205
7-43629
6-71495
7-09001
7-47526
6-75200
7-12808
7-51435
6-78915
7-16625
7-55354
6-82640
7-20452
7-59283
6-86375
7-24289
7-63223
6-90120
7-28137
7-67172
6-93876
7-31994
7-71132
390
400
410
7-75101
8-15360
8^56639
7^79081
8^19442
8^60822
7-83073
8^23534
8^65016
' 7-87072
8-27636
8-69221
7-91083
8-31749
8-73435
7-95103
8-35871
8-77660
7-99134
8-40004
8-81894
8-03175
8-44147
8-86139
8-07226
8-48300
8-90394
8-11288
8-52463
8-94660
420
430
440
8^98935
9^42251
9^86585
9 •03-221
9-46639
9-91075
9^07517
9^51037
9^95575
9-11823
9-55445
1000084
9-16139
9-59863
10-03605
9-20466
9-64292
10-08135
9-24803
9-68730
10-13676
9-29149
9-73178
10-18227
9-33506
9-77637
10-22787
9-37873
9-82106
10-27359
1
1
2
3
4
5
6
7
8
9
INTERNAL BALLISTICS
451
Table giving in dinamodes the energy stored up in 1 kilogramme in weight, etc. — continued.
440
450
470
480
490
500
510
5:^0
530
540
550
560
570
580
590
600
610
640
650
660
670
680
710
720
730
740
750
760
770
780
790
800
810
820
840
850
9-86585
10-31940
10-78313
9-91075
10-36531
10-83007
11-25706 11-30502
11-74119 11-79016
12-23550 12-28550
12-74000
13 25470
13-77964
14-31472
14-86000
15-41546
12-79101
13-30673
13-83269
14-36879
14-91508
15-47157
15-98112 il6-03824
16-55697 16-61511
17-14301 17-20218
17-73924
18-34567
18-96220
19-58911
20-22611
20-87330
21-53071
22-19828
22-87605
9-95575110-00084
10-41133 10-45745
10-87710 10-92424
11-35307111-40123
11-83924 111-88842
12-33558:12-38578
12-84212 12-89334
13-35891 113-41114
13-88584113-93909
14-42296 14-47723
14-970?7 15-02556
15-52778 15-58409
16-09547 16-15280
16-67336 jl6-73171
17-26144117-32081
10-03605 [10-08135 10-13676 11018227
10-50367 I1O-54999 10-59642 |l0-64294
10-97148:11-01883 11-06627 11-11381
11-44949111-49785
11-93769 [11-98708
12-43607 !l2-48647
12-94465
13-46348
13-99245
14-53160
15-08096
15-64040
12-99607
13-51592
14-04590
14-58608
15-13645
15-69701
17-79942
18-40687
19-02451
19-65235
20-29036
20-93858
21-59710
22-26560
22-95439
17-85971
18-46818
19-08683
17-92010
18-52958
19-14926
16-21023 16-26777
16-79016 16-84872
17-38028 17-43985
17-98059 18-04118
18-59109 J18-65271
21179119-27442
19-71569 19-77914
20-35473 20-41919
21-00396 21-06944
26-41779
27-15672
21-66351
22-33300
23-02283
23-70283
24-40303
•25-11341
25-83399
26-56476
27-30572
21-73001
32-40053
23-09137
23-56402 23-63338
24-26217 124-33255
24-97052 25-04192
25-76148
26-49123
27-23117
27-90583 27-98130 28-05688
128-66514 i28-74163 28-81822
129-43464 ;29-51215 29-58976
30-21433 '30-29286 30-37149
31-00422 ;31-08376 31-16342
31-80429 131-88486 31-96553
23-77240 23-84206
24-47361 24-54429
25-18501 125-25671
25-90661
26-63840
27-38038
28-13255
28-89492
29-66748
32-61456132-69615
33-43502133-51763
34-26567 j34-34929
35-10652 135-19116
35-95755 36-04322
36-81878 36-90546
37-69020
38-57181
39-46362
37-77790
38-66053
39-55336
32-77784
33-60034
34-43303
35-27591
36-12898
36-99225
37-86571
38-74936
39-64320
19-84269
20-48376
21-13503
21-79641
22-46815
23-15001
25-97933
26-71214
27-45514
28-20833
28-97172
29-74529
30-45023 30-52906
31-24317 31-32302
32-04630 32-12718
32-85963
33-68315
34-51685
35-36076
36-21485
37-07914
37-95361
38-83828
39-73315
32-94152
33-76606
34-60079
35-44571
36-30082
37-16613
38-14162
38-92731
39-82319
19-90634
20-54843
21-20072
22-53590
23-21876
23-91182
24-61507
25-32852
26-05215
26-78598
27-53000
28-28421
29-04862
29-82321
11-54631
12-03656
12-53697
13-04759
13-55845
14-09946
14-64066
15-19205
15-75363
16-32540
16-90737
17-49953
18-10188
18-71442
19-33715
19-97009
20-61320
21-26651
11-59488
12-08614
12-58758
13-09922
13-61109
14-15312
14-69534
15-24775
15-81035
16-38314
16-96613
17-55930
18-16267
18-77623
19-39999
20-03394
20-67807
21-33240
10-22787
10-68957
11-16146
11-64355
12-13583
12-63828
10-27359
10-73630
11-20921
11-69232
12-18562
12-68909
13-15094 13-20277
13-66383 !l3-7l668
14-20689 14-26075
14-75012 14-80500
30355 15-35945
15-86717 115-92409
16-44098 '16-49892
17-02499:17-08395
17-61918 17-67916
18-22357118-28457
18-83815 18-90017
19-46292(19-5-2596
20-09790 I2O-I6I95
20-74304 20-80812
21-39821 21-46441
21-93003 21-99694
22-60372I22-67165
23-28761 23-35656
23-98169
24-68596
25-40042
26-12508
36-85992
37-60496
28-36019
29-12562
29-90123
30-60800130-68704
31-40298 31-48304
32-20815 32-28923
33-02352
33-84907
34-68482
35-53076
36-38689
37-25322
38-12973
39-01644
39-91334
33-10561
33-93219
34-76896
35-61591
36-47307
37-34041
38-21794
39-10567
40-00359
22-06395
22-73969
23-42561
24-05166
24-75695
25-47243
26-19810
26-93397
27-68003
28-43628
29-20272
29-97935
30-76618
31-56320
3-2-37041
33-18781
34-01541
34-85319
35-70117
36-55934
37-42770
38-30626
39-19501
40-09394
22-13107
22-80782
23-49476
24-12173 24-19190
24-82804 24-89923
25-54454 25-61675
26-27123
7-00812
7-75519
28-51246
29-27992
30-05758
30-84542
31-64346
32-45169
33-27011
34-09873
34-93753
26-34446
27-08237
27-83046
28-58875
29-35723
30-13590
30-92477
31-72382
33-35251
34-18215
35-02197
35-78653 35-87199
64572 36-73220
51510 37-60260
38-39467138-48319
39-28444,39-37398
40-18440 ko-27495
452 INTERNAL BALLISTICS
Table giving in dinamodes the energy stored np in 1 kilogramme in weight, etc. — continued.
1
1
2
3
4
5
6
7
8
9
880
890
900
910
920
930
940
950
960
970
980
990
1000
39-46362
40-36561
41-27780
42-20018
43-13275
44-07552
45-02848
45-99162
46-96497
47-94850
48-94222
49-94614
50-96025
39-55336
40-45637
41-36958
42-29298
43-22657
44-17035
45-12433
46-08850
47-06286
48-04741
49-04216
50-04709
51-06222
39-64320
40-54723
41-46146
42-38588
43-32049
44-26530
45-22029
46-18548
47-16086
48-14643
49-14219
50-14815
51-16429
39-73315
40-63820
41-55345
42-47888
43-41451
44-36034
45-31635
46-28256
47-25896
48-24555
49-24233
50-24930
51-26647
39 82319
40-72927
41-64553
42-57199
43-50864
44-45548
45-41251
46-37974
47-35716
48-34477
49-34257
50-35056
51-36875
39-91334
40-82043
41-73772
42-66520
43-60286
44-55072
45-50878
46-47702
47-45546
48-44409
49-441:91
50-45192
51-47113
40-00359
40-91170
41-83001
42-75850
43-69719
44-64607
45-60514
46-57441
47-55386
48-54351
49-54335
50-55338
51-57361
40-09394
41-00308
41-92240
42-85191
43-79162
44-74152
45-70161
46-67189
47-65237
48-64304
49-64389
50-65495
51-67619
40-18440
41-09455
42-01489
42-94542
43-88615
44-83707
45-79818
46-76948
47-75098
48-74266
49-74454
50-75661
51-77888
40-27495
41-18612
42-10749
43-03904
43-98078
44-93272
45-89485
46-86717
47-84969
48-84239
49-84529
50-85838
51-88166
9
1
2
3
4
5
6
7
8
INTERNAL BALLISTICS
453
Table showing the work in dinainodes that 1 kilogramme of gunpo^nder is capable of performing in
expanding from volume = 1, to any given number of volumes up to 40.
1
■3
>
■00
•01
•02
•03
•01
•05
•06
•07
•08
■09
1
1
2
•000
6-038
10-957
-667
6-574
11-402
1-319
7-098
11-841
1-956
7-612
12-273
2-578
8-117
12-697
3-187
8-612
13-114
3-782
9-098
13-525
4-365
9-576
13-929
4-934
10-044
14-327
5-492
10-505
14-718
3
4
5
15-104
18-676
21-818
15-483
19-009
22-112
15-858
19-337
22-403
16-227
19-661
22-690
16-591
19-980
22-974
16-951
20-030
23-256
17-305
20-608
23-533
17-655
20-916
23-808
18-000
21-220
24-080
18-340
21-521
24-348
6
7
8
24-615
27-133
29-422
24-879
27-372
29-639
25-140
27-608
29-856
25-397
27-842
30-070
25-653
28-073
30-282
25-906
28-303
30-493
26-157
28-531
30-702
26-404
28-757
30-909
26-630
28-981
31-114
26-893
29-202
31-318
2
9
1
31-520
38-458
35-247
31-764
33-643
35-420
31-920
33-826
35-592
32-118
34-008
35-763
32-313
34-189
35-932
32-508
34-367
36-100
32-701
34-545
36-267
32-892
34-721
36-432
33-083
34-897
36-597
33-271
35-073
36-761
2
3
4
36-924
38-498
39-998
37-086
38-650
40-118
37-248
38-801
40-260
37-407
38-951
40-401
37-566
39-100
40-542
37-724
39-248
40-681
37-881
39-396
40-819
38-036
39-542
40-957
38-191
39-688
41-094
38-345
39-832
41-230
5
6
7
41-365
42-675
43-918
41-499
42-803
44-039
41-633
42-930
44-159
41-766
43-056
44-278
41-899
43-181
44-398
42-029
43-305
44-516
42-160
43-428
44-634
42-289
43-552
44-752
42-418
43-675
44-868
42-547
43-797
44-985
3
8
9
45-101
46-228
47-303
45-216
46-338
47-408
45-331
46-447
47-512
45-445
46-555
47-616
45-558
46-664
47-720
45-571
46-772
47-823
45-784
46-879
47-925
45-896
46-986
48-027
46-007
47-024
48-129
46-117
47-130
48-230
1
2
3
48-351
49-318
50-265
48-431
49-414
50-358
48-531
49-510
50-451
48-631
49-606
50-543
48-730
49-701
50-634
48-830
49-796
50-726
4R-92S
49-891
50-817
49-027
49-985
50-907
49-124
50-079
50-998
49-221
50-173
51-088
4
5
6
51-177
52-056
52-900
51-267
52-142
52-983
51-356
52-227
53-066
51-445
52-313
53-148
51-533
52-398
53-230
51-621
52-482
53-312
51-708
52-567
53-393
51-795
52-651
53-474
51-883
52-735
53-555
51-969
52-818
53-636
7
8
9
53-716
54-505
55-268
53-796
54-583
55-343
53-876
54-660
55-418
53-956
54-737
55-492
54-035
54-813
55-567
54-114
54-889
55-640
54-192
54-966
55-714
54-271
55-041
55-788
54-349
55-117
55-861
54-427
55-193
55-934
4
1
2
56-007
56-723
57-418
56-079
56-793
57-486
56-151
56-864
57-554
56-224
56-934
57-623
56-295
57-003
57-690
56-367
57-073
57-758
56-439
57-143
57-825
56-510
57-211
57-892
56-581
57-280
57-960
56-652
57-349
58-027
3
4
5
58-093
58-750
59-388
58-160
58-814
59-451
58-226
58-878
59-514
58-292
58-942
59-576
58-358
59-007
59-659
58-423
59-071
59-701
58-489
59-134
59-763
58-554
59-198
59-825
58-620
59-261
59-887
58-685
59-324
59-948
6
7
8
60-001
60-614
61-203
60-071
60-673
61-261
60-132
60-732
61-318
60-192
60-792
61-376
60-253
60-851
61-434
60-314
60-910
61-491
60-374
60-969
61-548
60-434
61-027
61-606
60-494
61-086
61-663
60-554
61-144
61-720
5
9
1
61-776
62-335
62-882
61 -833
62-391
62-936
61-889
62-446
62-990
61-946
62-501
63-043
62-002
62-556
63-097
62-058
62-610
63-150
62-114
62-665
63-203
62-170
62-720
63-256
62-225
62-774
63-310
62-280
62-828
63-363
2
3
4
63-415
63-936
64-446
63-468
63-988
64-496
63-520
64-039
64-547
63-573
64-090
64-597
63-625
64-141
64-647
63-677
64-192
64-697
63-729
64-243
64-747
63-781
64-294
64-797
63-833
64-345
64-846
63-885
64-395
64-896
•00
•01
•02
•03
•04
•05
•06
•07
•08
•09
454
INTERNAL BALLISTICS
Table showing the work in dinamodes that 1 kilogratnme
of gunpowder^ e.uivioo«oo = M ii^/f-i^fifTS, eB
-
. _| 1
^"
---v--.-^
"~~~-
■"-
—•-...
!
\
1
^1
1
i
— •
...JX
..'U
^1^ .^..-.
__,
'7'
---r
- "'<"-
*
^,
-^
0' ,.
«
1
„.-
.......
g'..
0.
«
0< kCH
•
/
,
•/
^^
/
y
/.
/
y^
/
.'
V
y
Cx-
/
J..
/
......
y
/
^
n
W
r
■n
m
/
/
/
6
P''
/
/
m
/ /\
p
/
H o.-
1
/
1
y
,/1
i
H
>
]
-\
/
S/ir,
0.
/
' /
■'/
/
< s>-
j
/
/
m
O
n
DO
>
m
1
/
!
/
/
1
r
1
1
1
/
• f
/
— -si -
i i /
: 1
/
'
/
i/ i /
1
CD
o
\i
/
X) 00-
i
•' +U-i
-{■-•■
/
— .
1
o
; ^ / *
/
2
i»
CD
■>
">
-n
' / / 5
/
s
>
J
D !
/ 1 / / * i /
p
- -
PD 1
O
C
///I/
p r vi
B G
z
1 (
• !\
1 11
/ ,
/'
1/
H
o-
//
i
n
z jz
^ 2
! i*
/
•4
"^
-
f7f^
*
3
I'/
i
^ I
fsJ
•1//
/
to
i
Tl
Si'
}
1
■
1
i
i
1 1 1 ! i
1
r\i O) 4> (!•
TONS PER SQUARE INCH.
O = IV o< -^ 0> <}^
DEVELOPED BY SOME NEW EXPLOSIVES 4G7
those obtained from crusher-gauges. To facilitate this comparison
I have added to the pressure curves I have described a curve showing
the pressures developed by cordite when fired in a close vessel, and
I have further added the results of five rounds of cordite fired for
the Explosives Committee in the crusher-gauge gun, the pressure of
each individual round at each point of observation being indicated.
The sample of cordite used in these experiments was not of the
same make as that employed in my own. The pressures given on the
axis of y denote those taken in the powder chamber, and are compar-
able with the crusher-gauge pressures I have given as derived from
my own experiments. It will be observed that the mean chamber
pressure indicated is two- or three-tenths of a ton higher than that
I obtained ; but it will be further observed that if I attempt to draw
a pressure curve through the mean of the crusher-gauge observation,
such curve would indicate pressures far higher than are necessary
and sufficient to develop the work impressed on the projectile.
Again, the pressures indicated after the projectile has moved
1 foot are about 2 tons per square inch higher than those observed
in the powder chamber, and it will be further noticed that not only
are these observations, at all events at certain points, considerably
too high, but they exhibit in the forward part of the bore variations
quite unknown when the pressures are taken in the powder chamber.
Thus, in the particular experiments I am discussing, the mean
pressure in the powder chamber being about 13 '5 tons, the extreme
^•ariation in the five rounds amounts only to about IJ tons per square
inch, while the crusher-gauge placed in the chase at a point about
8i feet from the seat of the shot, gave in the same number of rounds
an extreme variation of 3 tons per square inch, the mean pressures
being only about 4 tons ; and it will further be noted that, while some
of the rounds indicated pressures below those deduced by the method
I have described, other rounds at the same point indicated pressures
even exceeding those which would have existed under the same gravi-
metric densities in a close vessel. It may also be noted that from the
crusher-gauge experiments, round 5 should have given the lowest
muzzle energy of the series ; as a matter of fact it gave the highest.
My conclusion, therefore, is that, although crusher-gauges placed
in the chase may, and doubtless do, give valuable comparative results,
they cannot be relied on for absolute determinations, unless confirmed
by observations altogether independent in their nature.
XII.
EESEAECHES ON EXPLOSIVES. PEELIMINAEY NOTE.
{Proceedings of the Koyal Society, 1894.)
The researches on which I, in conjunction with Sir F. Abel, have
been engaged for very many years, have had their scope so altered
and extended by the rapid advances which have been made in the
science of explosives, that we have been unable to lay before the
Society the results of the many hundreds of experiments under varied
conditions which I have carried out. We are desirous also of clear-
ing up some difficulties which have presented themselves with certain
modern explosives when dealing with high densities and pressures ;
but the necessary investigations have occupied so much time, that I
am induced to lay a few of our results before the Society, trusting,
however, that before long we may be able to submit a more complete
memoir.
A portion of our researches includes investigations into the trans-
formation and ballistic properties of powders varying greatly in
composition, but of which potassium nitrate is the chief constituent.
In this preliminary note I propose to refer to powders of this
description chiefly for purposes of comparison, and shall devote my
attention principally to guncotton and to those modern explosives
of which guncotton forms a principal ingredient.
In determining the transformation experienced during explosion,
the same arrangements for firing the explosive and collecting the
gases were followed as are described in our earlier researches,* and
the gases themselves were, after being sealed, analysed either under
the personal superintendence of Sir F. Abel, or of Professor Dewar,
and to Professor Dewar's advice and assistance I am indebted, I can
hardly say to what extent.
The heat developed by explosion, and the quantity of permanent
gases generated were also determined as described in our researches,
but the amount of water formed plays so important a part in the
* Phil. Trans., vol. clxv., p. 61.
RESEARCHES ON EXPLOSIVP^S 469
transfonnation that special means were adopted in order to obtain
this product with exactness.
The arrangement employed was as follows : —
After explosion the gases formed were allowed to escape through
two U -tubes filled with pumice-stone and concentrated sulphuric
acid ; when the gases had all escaped the explosion cylinder was
o})ened, and the water deposited at the bottom of the cylinder was
collected in a sponge, placed in a closed glass vessel, and weighed.
The cylinder was then nearly closed and heated, and a measured
quantity of air was, l)y means of an aspirator, drawn slowly through
the (J -tubes till the cylinder was perfectly dry. This was easily
ascertained by observing when moisture was no longer deposited on
a cooled glass tube through which the air passed.
The U -tubes were then carefully weighed, the amount of moisture
absorbed determined, and added to the quantity of water directly
collected. The aqueous vapour in the air employed for drying was,
for each experiment, determined and deducted from the gross
amount.
Numerous experiments were made to ascertain the relation of the
tension of the various explosives employed, to the gravimetric density
of the charge when fired in a close vessel ; but I do not propose here
to pursue this part of our inquiry, both because the subject is too
large to be treated of in a preliminary note, and because approximate
values have already been published * for several of the explosives
with which we have experimented.
With certain explosives, the possibility or probability of detona-
tion was very carefully investigated. In some cases the explosive
was merely placed in the explosion-vessel in close proximity to a
charge of mercuric fulminate by which it was fired, but I found that
the most satisfactory method of experiment was to place the charge
to be experimented with in a small shell packed as tightly as possible,
the shell then being placed in a large explosion-vessel and fired by
means of mercuric fulminate. The tension in the small shell at the
moment of fracture and the tension in the large explosion-vessel
were in each experiment carefully measured.
It may be desirable here to explain that I do not consider the
presence of a high pressure with any explosive as necessarily denot-
ing detonation. With both cordite and guncotton I have developed
enormous pressures, close upon 100 tons per square inch (about
15,000 atmospheres), but the former explosive I have not succeeded
Noble, Internal Ballistics, 1892, p. 33; Roy. Soc. Proc. vol. Hi., p. 128.
470 RESEARCHES ON EXPLOSIVES
in detonating, while guncotton can be detonated with the utmost
ease. It is obvious that if we suppose a small charge fired in a
vessel impervious to heat, the rapidity or slowness of combustion will
make no difference in the developed pressure, and that pressure will
be the highest of which the explosive is capable, regard being of
course had to the density of the charge. I say a small charge,
because, if a large charge were in question, and explosion took place
with extreme rapidity, the nascent gases may give rise to such whirl-
winds of pressure, if I may use the term, that any means we may
have of registering the tension will show pressures very much higher
than would be registered were the gases, at the same temperature, in
a state of quiescence. I have had innumerable proofs of this action,
but it is evident that in a very small charge the nascent gases will
have much less energy than in the case of a large charge occupying
a considerable space.
The great increase in the magnitude of the charges fired from
modern guns has rendered the question of erosion one of great
importance. Few, who have not had actual experience, have any
idea how rapidly with very large charges the surface of the bore
is removed. Great attention has therefore been paid to this point,
both in regard to the erosive power of different explosives and in
regard to the capacity of different materials (chiefly different natures
of steel) to resist the erosive action.
The method I adopted for this purpose consisted in allowing
large charges to escape through a small vent. The amount of the
metal removed by the passage of the products of explosion, which
amount was determined by calibration, was taken as a measure of
the erosive power of the explosive.
Experiments have also been made to determine the rate at which
the products of explosion part with their heat to the surrounding
envelope, the products of explosion being altogether confined. I
shall only briefly allude to these experiments, as, although highly
interesting, they have not been carried far enough to entitle me to
speak with confidence as to final conclusions.
Turning now to ballistic results. The energies which the new
explosives are capable of developing, and the high pressures at
which the resulting gases are discharged from the muzzle of the
gun, render length of bore of increased importance. With the
object of ascertaining with more precision the advantages to be
gained by length, the firm to which I belong has experimented with
a 6-inch gun of 100 calibres in length. In the particular experi-
I
RESEARCHES ON EXPLOSIVES 471
merits to which I refer, the velocity and energy generated has not
only been measured at the muzzle, but the velocity and the pressure
producing this velocity have been obtained for every point of the
bore, consequently the loss of velocity and energy due to any
particular shortening of the bore can be at once deduced.
These results have been obtained by measuring the velocities
every round at sixteen points in the bore and at the muzzle. These
data enable a velocity curve to be laid down, while from this curve
the corresponding pressure curve can be calculated. The maximum
chamber pressure obtained by these means is corroborated by simul-
taneous observations taken with crusher-gauges, and the internal
ballistics of various explosives have thus been completely determined.
Commencing with guncotton, with which a very large number
of analyses were made, with the view of determining whether there
was any material difference in the decomposition dependent upon
the pressure under which it was exploded, two descriptions were
employed : one in the form of hank or strand, and the other in the
form of compressed pellets. Both natures were approximately of
the same composition, of Waltham-Abbey manufacture, containing in
a dried sample about 4-4 per cent, of soluble cotton and 95 ^G per cent,
of insoluble. As used, it contained about 2-25 per cent, of moisture.
The following were the results of the analyses of the permanent
gases. They are placed in five series, viz. : —
first. — Analyses showing the decomposition of the strand or
hank guncotton. Second. — Analyses showing the decomposition
of pellet guncotton.
In both these series the analyses are arranged in the order of
the ascending pressures under which the decomposition took place.
Third and fourth. — Examples of the decomposition of strand
and pellet guncotton when exploded by means of mercuric
fulminate. And, fifth, a series showing the decomposition
experienced by pellet guncotton saturated with from 25 to 30 per
cent, of water, and detonated by means of a primer of dry guncotton
and mercuric fulminate.
I leave these resvilts for discussion in the memoir which Sir F.
Abel and I hope before long to submit, and will only remark that,
in Tables 1 and 2, the same peculiarity we have before remarked
upon in reference to gunpowder, is again exhibited ; I mean the
marked manner in which the carbonic anhydride increases with
the pressure. It will be noted that in Table 1 the volumes of
carbonic anhydride and carbonic oxide are nearly exactly reversed ;
472
RESEARCHES ON EXPLOSIVES
P
i .2
i|
il
1'
'I
1
36-18
27-57
16-76
16-15
3-34
9
34-77
28-66
17-48
16-05
3-04
p
34-70
28-60
16-56
16-83
3-31
p
O
33-01
30-32
18-25
16-60
1-82
CO
33-63
31-20
17-99
16-23
0-95
CO
3
32-70
31-36
19-23
16-25
0-46
9
(M
32-23
30-65
20-38
16-43
0-31
p
31-00
32-76
18-80
16-90
0-54
9
30-95
32-27
19-10
17-20
0-48
ip
29-62
35-03
17-13
18-18
0-04
^2
26-49
36-66
19-68
16-85
0-32
-3
1 ^ ^ ^ ^
I
9
00 'H O CO
00 lo OS
f
S
CO u-5 O 'J'
CM CO (M i-(
9
^s^gg
(M
25S?:2<=
9
s^sgg
t^
s^^ss^
p
:SS^§S
^
CO lo 00 CO I-l
Kl CO !-• r-l
1
9
25-75
38-00
19-71
15-26
1-28
p
1-1 CO t^ lO 'S*
-ti ^) CO CO CO
1
t^ t^ OS lO o
C
^1^SS-*
lO
CO lO O 00 T*"
poopoo<>5
'"'
^^;^s-
V
o
ssss^
S^i^S^
^
1 . . : :
i
s
s
RESEARCHES ON EXPLOSIVES
473
Table 3. — Results of the analyses of strand gunrotton token fired in a
close vessel hy detonation.
Pressure * jier sq. incli.
1 ton.
3 tons.
CO., (vols.) .
19-21
29-08
CO" „ .
41-25
32-88
H „ .
23-07
20-14
N „ .
16-21
17-50
CH, „ .
0-26
0-75
The pressures given are those due to the gravimetric density of the charge.
Table 4. — Similar results for pellet guncotton.
Pressure per sq. inch.
3 tons.
10 tons
COo(vols.) .
25-76
26-50
CO „ .
39-34
37-48
H „ .
18-71
20-97
N „ .
16-19
15-05
CH4 „ .
Nil
Nil
Table 5. — Results of analyses of saturated pellet guncotton fired
close vessel hy detonation.
', per square inch.
Under 10 tons.
10-5 tons.
16 tons.
16-5 tons
32-14
33-25
32-93
35-60
27-04
25-90
27-25
23-43
26-80
26-53
25-76
24*22
13-83
14-32
14-06
15-25
0-19
Nil
Nil
1-50
CO., (vols.)
CO" „ .
H „ .
N „ .
CH, „ .
again, considering that the composition of the pellet and strand
guncotton is practically the same, the distinct difference between
the proportions of these products in the two series is sufficiently
remarkable. It not improbably is connected with the rapidity of
combustion of the two samples. Another striking peculiarity is
the manner in which the COo is increased (as exhibited in Table 5)
when saturated pellet cotton is detonated.
Such are the average analyses of the permanent gases generated
by the decomposition of guncotton under the various conditions I
have described, and it will be evident from these analyses that the
volumes of the permanent gases may be expected to ditfer to some
very appreciable extent, depending both upon the density under
which it is exploded, and also upon the mode of explosion. I have
found it most convenient to explode the charges, the permanent
gases from which were to be measured, under a pressure of about
10 tons per square inch (1524 atmospheres), and, under these
circumstances, the average of several very accordant determinations
474 RESEARCHES ON EXPLOSIVES
gave, at 0' Cent, and 760 mm. of mercury, 689 c.c. per gramme of
strand guncotton and 725 c.c. per gramme of pellet guncotton.
At the temperature of explosion the whole of the water formed
is in the gaseous state. It is therefore necessary, in order to obtain
the total gaseous volume, to add to the above volumes of permanent
gases the equivalent volume of aqueous vapour at the temperature
and pressure stated. Now the quantity of water formed by the
explosion of 129"6 grms. of guncotton was found to be 16'985
grms. ; hence 1 grm. of guncotton generated 01311 grm. of water,
equivalent to 162"6 c.c. of aqueous vapour, and the total volume
of gaseous matter at the temperature and pressure stated is for
strand guncotton 852"2 c.c. per gramme, for pellet 887'6 c.c.
The heat measured reached, with strand guncotton, 1068 grm.-
units (water fluid), or 988 grm.-units (water gaseous), while with pellet
guncotton these figures were 1037 or 957 grm.-units respectively.
Pellet guncotton made at Stowmarket generated 738 c.c. of
permanent gas and 994 units of heat per gramme, while dinitro-
cellulose containing 12-8 per cent, of nitrogen generated 748 c.c.
of gas and 977 units of heat, the water in both cases being fluid.
Guncotton, both pellet and strand, I have detonated by means
of mercuric fulminate with ease and certainty. The effect of
employing this means of ignition in a close vessel is very striking,
and the indications of intense heat are much more apparent than
when the charge is fired in the ordinary way. This effect is probably
partly due to an actual higher temperature, caused by the greater
rapidity of combustion. I allude elsewhere to the extreme rapidity
with which the gases part with their heat, but this higher heat is,
I think, clearly indicated by the surfaces of the internal crusher-
gauges becoming covered with innumerable small cracks, and by
thin laminae occasionally flaking off exposed surfaces ; but perhaps
the most striking proof of the violence of this detonation is shown
by its action on a cast-iron shell fired as I have described ; where
no detonation takes place the shell is broken into fragments of
various sizes, such as are familiar to all acquainted with the burst-
ing of shell; but when detonation, with guncotton, for example,
takes place, the whole shell is reduced to very minute fragments,
and, what is more remarkable, two-thirds of the total weight are
generally in the form of small peas and of the finest dust.
The ease with which guncotton can be detonated renders it
unsuitable for use as a propulsive agent, unless this property be in
some way neutralised. I have, therefore, made but few experiments
RESEARCHES ON EXPLOSIVES 475
ill this direction, and shall not further allude to them in this note,
as more suitable explosives — explosives also of which guncotton is
a principal component — have been elaborated ; and these not only
possess to the full the high ballistic properties of guncotton, but are
more or less free from the tendency to detonate, which, however
useful it may be in other directions, is a fatal objection to the
employment of guncotton for propelling purposes.
Turning now to cordite ; cordite consists, as is well known, of
nitro-glycerine and guncotton as its main ingredients. As now
made, it contains 37 per cent, of guncotton (trinitro-cellulose with
a small proportion of soluble guncotton), 58 per cent, of nitro-
glycerine, and 5 per cent, of a hydrocarbon known as vaselin. On
account of the importance of this explosive, I have made numerous
experiments, both with large and small charges, to determine the
relation of the tension to the density of the charge. Up to densities
of 0-55 the relation may be considered to be very approximately
determined : above that density, although many determinations have
been made, these determinations have shown such wide variations
that they cannot, until certain discrepancies are explained, be
assumed as at all accurate.
The average results of some of the analyses of the permanent
gases are given below : —
The first four analyses were made from experiments with the
earlier samples of cordite when tannin formed an ingredient of
cordite. They are not, therefore, strictly comparable with the later
analyses. There appears also to be a difference in the transforma-
tion, slight but decided, which the same cordite experiences, depen-
dent upon the diameter of the cord ; and this difference is shown at
once in the analyses, in the volume of permanent gases, in the heat
developed, and, I think, in the amount of aqueous vapour formed.
The following are some of the analyses: —
Table 6.
I'res.sure per square inch.
^
0-048 Cordite.
0-220
Cordite.
2-5 tons.
6 tons. 10 tons.
14 tons.
10 tons.
12 tons.
11 tons.
14 tons.
COo
29-9
30-4 32-0
31-6
27-0
28-4
23-9
26-3
CO"
28-3
30-7 32-9
32-1
34-2
33-8
37-2
35-8
H
19 -S
20-0 18-0
21-6
26-9
24-4
28-4
26-1
N
22 '5
18-9 17-1
14-8
12-0
13-4
10-4
11-8
CH,
traces.
In the whole of these analyses the water formed by the explosion
smelt stronglv of ammonia.
476 RESEARCHES ON EXPLOSIVES
The quantity of permanent gases measured, under the same condi-
tions as in the case of guncotton, was found to be : —
For tlie earlier cordite, 655 vols.
For the present service cordite, 0'255 inch in diameter, 692
vols., and for that 0-048 inch in diameter, 698 vols. In the two
latter samples the aqueous vapour was determined, and was found
to amount to 20-257 grms. for the 0-255-inch cordite, and to 20-126
grms. for the 0-048-inch cordite ; or, stating the result per gramme,
these figures are respectively equivalent to 0-1563 grm., or 194 c.c.
aqueous vapour, and to 0-1553 grm., or 192-5 c.c. per grm. of
cordite.
Hence the total gaseous products generated by the explosion of
cordite amount per grm. to 886 c.c. for the 0-255-inch cordite,
and to 890-5 c.c. for the 0-048-inch cordite, the volumes being,
of course, taken at 0° Cent, and 760 mm, atmospheric pressure.
The heat generated was found to be: — For the earlier cordite,
1214 grm. -units water fluid ; for the service 0-2 5 5 -inch cordite, 1284
grm. -units water fluid or 1189 units water gaseous ; for the service
0-048-inch cordite, 1272 units water fluid or 1178 units water
gaseous.
From my very numerous experiments on erosion, I have arrived
at the conclusion that the principal factors determining its amount
are : (1) the actual temperature of the products of combustion, (2)
the motion of these products. But little erosive effect is produced,
even by the most erosive powders, in close vessels, or in those
portions of the chambers of gims where the motion of the gas is
feeble or nil ; but tJhe case is widely different where there is rapid
motion of the gases at high densities. It is not difficult absolutely
to retain without leakage the products of explosions at very high
pressures, but if there be any appreciable escape before the gases
are cooled, they instantly cut a way for themselves with astonishing
rapidity, totally destroying the surfaces over or through which they
pass. Among all the explosives with which I have experimented,
I have found that where the heat developed is low, the erosive effect
is also low.
With ordinary powders, the most erosive with which I am
acquainted is that which, on account of other properties, is used for
the battering charges of heavy guns: I refer to brown prismatic
powder. The erosive effect of cordite, if considered in relation to
the energy generated by the two explosives, is very slightly greater
than that of brown prismatic ; but very much higher effects can, if
RESEARCHES ON EXPLOSIVES 477
it be so desired, be obtained with cordite, and, if the highest energy
be demanded, the erosion will be proportionally greater. There is
however, one curious and satisfactory peculiarity connected with
erosion by cordite. Erosion produced by ordinary gunpowder has
the most singular effect on the metal of the gun, eating out larg&
holes, and forming long, rough grooves, resembling a ploughed field
in miniature, and these grooves have, moreover, the unpleasant
habit of being very apt to develop into cracks ; but with cordite, so
far as my experience goes, the erosion is of a very different character.
The eddy holes and long grooves are absent, and the erosion appears-
to consist in a simple washing away of the surface of the steel
barrel.
Cordite does not detonate; at least, although I have made far
more experiments on detonation with this explosive than with any
other, I have never succeeded in detonating it. With an explosive
like cordite, capable of developing enormous pressures, it is, of
course, easy, if the cordite be finely comminuted, to develop very
high tensions, but, as I have already explained, a high pressure does
not necessarily imply detonation.
The rapidity with which cordite gases lose their temperature, and
consequently their pressure, by communication of their heat to their
surrounding envelope, is very striking. Exploding a charge of about
If lb. of cordite in a close vessel at a tension of a little over 6 tons
on the square inch, or say 1000 atmospheres, I have found that the
pressure of 6 tons per square inch was again reached in 0"07 second
after explosion, of 5 tons in 0171 second, of 4 tons in 0-731 second^
of 3 tons in 1-764 second, of 2 tons in 3-523 seconds, and of 1 ton
in 7-08 seconds. The loss of pressure after 1 ton per square inch
was reached, was, of course, slow, but the figures I have given were
closely approximated to in two subsequent experiments. With
ordinary gunpowder the reduction of pressure was very much
slower, as was to be expected, on account of the charge being much
larger ; on account, also, of the temperature of explosion being much
lower.
These experiments are now bemg continued with larger charges
and higher pressure.
It only remains to give particulars as to ballistics, that is,,
as to the velocities and energies realisable by cordite in the
bore of a gun; but these will be most conveniently given with
similar details regarding other explosives with which I have
experimented.
478 RESEARCHES ON EXPLOSIVES
The ballistite I have used has, like the cordite, been changed in
composition since the commencement of my experiments. The
sample I used for my earlier experiments was nearly exactly
composed of 50 per cent of dinitro-cellulose (collodion cotton) and
50 per cent, of nitro-glycerine. The cubes were coated with
graphite, and the nitro-cellulose was wholly soluble in ether
alcohol.
The second sample was nominally composed of 60 per cent, of
nitro-cellulose and 40 per cent, of nitro-glycerine. The proximate
analysis gave —
Nitro-glycerine . . . .41-62
Nitro-cellulose .... 59'05
as before the whole of the nitro-cellulose was soluble in ether
alcohol.
The earlier sample gave the following permanent gases under
i^ressures of 6 and 12 tons per square inch respectively: —
CO
H
N
CH^
37-3
38-49
27-8
28-35
19-1
19-83
15-8
13-32
traces.
One gramme of this ballistite gives rise to 610 c.c. of permanent
gases, and to 01588 grm. of aqueous vapour, corresponding to 197
c.c. at O'Cent. and 760 mm.
Hence the total volume of gas is 807 c.c, and the heat generated
by the explosion is 1365 grm.-units (water fluid), 1269 grm.-units
(water gaseous).
Although I have not made nearly so many experiments on detona-
tion with ballistite as with cordite, those I have made with the
earlier samples (50 per cent, guncotton and 50 per cent, nitro-
glycerine), neither detonated, nor did they show any tendency to
detonate ; but the case is different with respect to a sample of ballistite
consisting of 60 per cent, guncotton and 40 per cent, nitro-glycerine.
This sample, 0'2-inch cubes, detonated with great violence on two
occasions; but I am unable, without further experience, to say
whether this result was due to the change in the composition of the
ballistite or to defective manufacture.
The erosive action of ballistite is, as might perhaps be anticipated
RESEARCHES ON EXPLOSIVES 479
from the higher heat developed, greater than with cordite, but the
remarks made with respect to the action of cordite apply also to
ballistite.
The French B. K powder consists of nitro-cellulose partially
gelatinised and mixed with tannin, with barium, and potassium
nitrates.
When exploded under a pressure of 6 tons per square inch, the
permanent gases were found to consist of
CO., . . . . .28-1 vols.
CO" 32-4 „
H . . .... 21-9 „
N . . . . . 16-8 „
CH^ . . . . . 0-8 „
These permanent gases occupied at the usual temperature and
pressure a volume of 616 c.c. ; the aqueous vapour formed occupied
in addition 206 c.c, so that the total gaseous volume was 822 c.c.
The heat generated was 1003 grm.-units (water fluid) or 902 grm.-
units (water gaseous) ; the ballistics obtained with this powder are
given along with those furnished by other explosives.
For purposes of comparison, I have introduced among the ballistic
results those obtained with amide prismatic powder, and with 11. L. G.
Particulars as to both these powders have already been given * and
need not here be repeated.
In a preliminary note, like the present, the most convenient mode
of comparing the velocities and energies developed by the new
explosives is by the aid of diagrams.
Accordingly, in Fig. 1 (coloured diagram, p. 480), I show the
velocities of seven different explosives from the commencement of
motion to the muzzle of the gun ; the position of the points at wdiich
the velocity is determined are shown, and on the lowest and highest
curves the observed velocities are marked where it is possible to do
so without confusing the diagram. Lines are drawn to indicate the
velocities that are obtained with the lengths of 40, 50, 75, and 100
calibres.
Fig. 2 (coloured diagram, p. 480) shows the pressures by which
the velocities of Fig. 1 were obtained. The areas of these cm^ves
represent the energies realised, and the lines intersecting the curves
indicate the pressures at which the gases are discharged from the
muzzle for lengths of 40, 50, 75, and 100 calibres respectively. The
* Eoy. Soc. Froc, vol. lii., p. 125 ; FIiH. Trans., part i., 1880, p. 278.
480 RESEARCHES ON EXPLOSIVES
chamber pressures indicated by crusher-gauges are also shown in
Fig. 2, and it will be observed that the two modes of determining
the maximum pressure are in general in close accordance.
It will further be observed that with the slow-burning powders
the chronoscopic maximum pressures are somewhat, though not
greatly higher, than are those indicated by the crusher-gauges. This
observation is not new.* It was noted in the long series of
experiments with black powders carried on by the Committee of
Explosives.
The result is widely different where an explosive powder or a
quickly-burning powder, such as E. L. G., giving rise to wave-pressure,
is employed; the crusher-gauge in such cases f gives considerably
and frequently very greatly higher pressures, and this peculiarity is
illustrated in the curve from K. L. G. in Fig. 2.
It is, perhaps, hardly necessary to point out that the results given
in Fig. 1 have to be considered in relation to the facts disclosed in
Fig. 2. Thus it will be noted that the velocities and energies realised
by 22 lbs. of 0-35-inch cordite and 20 lbs. of 0-3-inch cordite are
practically the same; but reference to Fig. 2 shows that, with the
0-3-inch cordite, this velocity and energy has been obtained at the
cost of nearly 30 per cent, higher maximum pressure.
A similar remark may be made in regard to the French B. N.
powder if compared with the ballistite. Its velocity and energy are
obtained at a high cost of maximum pressure, and it is interesting to
note how the velocity curve of B. N., which for the first 4 feet of
motion shows a velocity higher than that of any other explosive,
successively crosses other curves, and gives at the muzzle a velocity
of 500 feet per second under that of cordite.
The velocities and energies at the principal points indicated in
Figs. 1 and 2 are summarised in the annexed table, which shows for
each nature of explosive the advantage in velocity and energy to be
gained by correspondingly lengthening the gun.
Fig. 3 (coloured diagram) is an interesting illustration of a point
to which I have elsewhere adverted. Cordite and ballistite leave no
deposit in the bore. Eound 1 with E. L. G. was fired with a clean
bore. The difference in velocity between round 1 with a clean bore
and rounds 2 and 3 with powder deposit in the chase, is very clearly
marked, and it will be noted that in this instance the effect of the foul
bore is only distinctly shown when the length exceeds 40 calibres.
'■ Noble and Abel, Pliil. Trans., vol. clxv., p. 110.
t Compare Noble and Abel, loc. cit., p. lOP.
Velocity in f* per Sec
11. . . ■ li. . ,.r , ■ .1. .. .1.
OC2
T>OoO
■DOO
OOOOOOCn
r S p 5 O O J
- w rv ixjfufuoi
O (J> too 1»
33 o»
*-5i
■ I I I I te T
Velocity in f* per Sec.
Pressure m tons per sguare Inch.
o ■ ■ ■ ■ X
Tons per sq inch.
Pressure in
tons per square inc
h
g
° ' ' L...UJ' • ' '
, ,6 , , , ,
a.
f^^^^^^^T
'a
''•-
ffi .y''^^
^^^,
2 /^'
^^N
■c-
1 #
Ot
V o
o>-
°5 // 0/
-
.'^ V <<>
\\
^ / cT
\
5-
;I ^
^
\
I.
INJ-
~ 'f
\'
-
;!
•D
-1
V
•f>~
^:.'L
TO
1
5-
^;/ 40Ca//bres
ss
\
_;
\
:!
T -N
;
5 0-5
l\
;/
3 o 2. :c
'l\
°'
;(
-1 O "
0«^ S r-
\\
( 50 Calibres
?c..=^0
'A
fs-
"t
o 5 -- '^
TT
ro
I
i
g-S?5
''\
■»•"
ii
8^-01
"<3 a Wi 3'
\
ro
i;
"
NU
IVJ.
1;
"^5*° c^
G»
•';
a> T T c
III
lis"
It
ii
'1
i
•^ tft o
j;
Co
1 1
»•
Ii 75 Calibres
Co C)
<^-s
1 1
ft :r-
! 1
s-
i;
1
j;
ss-
s-
1
-
t-
ii
ii
^
1
— 1 — 1 \ — 1 1 1
... ,„ ... -.
ii
4-
Velocity in feet per second
RESEARCHES ON EXPLOSIVES
481
From 40 calibres onwards, the loss of velocity due to a bore
encrusted with deposit is very distinctly shown.
Table 7. — Showing the velocities and energies realised in Q-inch gun loith the
undemoted explosives.
Length of bore,
40 calibres.
Length of bore,
50 calibres.
Length of bore,
75 calibres.
Length of bore,
100 calibres.
weight of charge.
1
>
2
1
!>
1
>
i
1
>
i
s
Cordite, 0-4-in. dia.,
27-5 lbs. .
2794
5413
2940
5994
3166
6950
3284
7478
Cordite, 0-35-in. dia.,
22 lbs.
2444
4142
2583
4626
2798
5429
2915
5892
Cordite, 0-3-in. dia.,
20 lbs.
2495
4316
2632
4804
2821
5518
2914
5888
Ballistite, 0-3-in. cubes,
20 lbs.
2416
4047
2537
4463
2713
5104
2806
5460
French B.N., 25 lbs. .
2422
4068
2530
4438
2700
5055
2786
5382
Amide Prismatic, 32 lbs.
2225
3433
2331
3768
2486
4285
2566
4566
R. L. G..„ 23 lbs. .
1533
1630
1592
1757
1668
1929
1705
2016
2 H
XIII.
ON METHODS THAT HAVE BEEN ADOPTED FOE
MEASUEING PEESSUEES IN THE BOEES OF GUNS
{Paper read 'before the British Association, Oxford, 1894.)
The importance of ascertaining, with some approach to accuracy,
the pressures which are developed at various points along the
bores of guns by gunpowder or other propelling agent is so great
that a variety of means have been proposed for their determination,
and I purpose, in this paper, to give a very brief account of some
of these means, pointing out at the same time certain difficulties
which have been experienced in their employment, and the errors
to which these methods have been in many cases subject.
The earliest attempt, by direct experiment, to ascertain pressures
developed by fired gunpowder, was that made by Count Eumford
in his endeavour to determine the pressures due to different
densities of charge. He assumed, the principles of thermo-dynamics
being then unknown, that charges fired in a small closed gun-barrel
would give pressures identical with those given by charges doing
work both on the projectile and on the products of combustion
themselves ; but even this error was a small one compared with
that which led him to adopt, as correct, his extravagant estimate
of the pressures developed.
For a density of unity — or, in other words, for a charge
approximately filling a chamber in which it was fired — he estimated
the pressure at over 101,000 atmospheres, or at 662 tons per
square inch.
He adopted this pressure notwithstanding the great discrepancy
which he found to exist between the two series of experiments
which he made, and he meets the objection that, were the pressure
anything approaching that which he gives, no gun that ever was
made would have a chance of standing, by assuming that the
MEASURING PRESSURES IN THE BORES OF GUNS 483
combustion of powder is exceedingly slow, and lasts the whole time
occupied by the projectile in passing through the bore.
It is sufficiently curious that a man so eminent for his scientific
attainments as was Eumford should have fallen into so great an
error, both because any attempt at calculation would have shown
him his mistake, and because Eobins, sixty years earlier, had
conclusively proved that with the small-grain powders then used —
and it must be remembered that Eumford's powder was sporting
of very fine grain — the whole of the powder was fired before the
bullet was very greatly removed from its seat. Eobins's argument
— and it is incontrovertible — was, that were it otherwise a much
greater energy would be realised from the powder when the weight
of the projectile was doubled, trebled, quadrupled, etc. ; but his
experiments showed that under these circumstances the work done
by the powder was nearly the same.
Eor other objects, on a much larger scale, and with appliances
far superior to those which the great man I have named had at
his disposal, I have had occasion to repeat Eobins's experiment,
and the results are interesting. With a charge of 10 lbs. of the
powder known as E. L. G. 2 and a shot weighing 30 lbs., a velocity
of 2126 feet per second, representing an energy of 97l"6 foot-tons,
was attained. The same charge being used, but the weight of the
projectile being doubled, the velocity was reduced to 1641 feet per
second, while the energy was increased to 1125 foot-tons. With
a shot weighing 120 lbs. the velocity was 1209 feet per second, and
the energy 1196 foot-tons. With a shot of 150 lbs. the velocity
was 1080 feet per second, and the energy 1191-5 foot-tons; while
with a shot of 360 lbs. the velocity was reduced to 691 feet per
second, representing a muzzle energy of 1191*9 foot-tons. These
energies were obtained with maximum chamber pressures respectively
of 13'5 tons, of 17"25 tons, of 19 tons, of 20 tons, and of 22 tons
per square inch. It will be noted that the maximum energy
obtained was realised with the shot of 120 lbs. weight, the energy
given by a shot of 360 lbs. — i.e., three times that weight, or twelve
times the weight of the original shot — being nearly exactly the
same.
Very different, however, were the results when one of the modern
powders, introduced with the special object of insuring slow
combustion, was compared with the E. L. G. 2 experiments which
I have just quoted.
With brown prismatic or cocoa powder, an exactly similar series
484 ON METHODS THAT HAVE BEEN ADOPTED FOR
was fired. The 30-lb. shot gave a velocity of 1515 feet per second,
and an energy of 493-4 foot-tons; the 60-lb. shot gave 1291 feet
per second, or an energy of 693-4 foot-tons; the 120-lb. shot,
1040 feet per second, or 877-5 foot-tons; the 150-lb. shot, 948 feet
per second, and 920-7 foot-tons ; while with the heaviest shot, the
360-lb., the velocity attained was 654 feet per second, equivalent
to an energy of 1064*7 foot-tons. The maximum chamber pressures
in this series varied from 4*8 tons per square inch with the lightest
projectile, to 9*6, with the heaviest ; and with this powder it will
be observed that the energy developed increased steadily and
considerably with each increment in the weight of the shot, while
the low chamber pressure shows that, even with the heaviest shot,
the projectile must have moved a considerable distance from its
seat before the charge can be considered to have been entirely
consumed.
I have mentioned the discrepancy between Eumford's two series
of experiments. This discrepancy was very great, the one series
giving, for a density of unity, a tension of about 190 tons per square
inch, or 29,000 atmospheres, the other series giving a tension of
over 101,000 atmospheres. It is remarkable that Eumford makes
no attempt to explain this discrepancy, but, as he deliberately
adopts the higher tension, it is not improbable that he was led to
this conclusion by an erroneous estimate of the elastic force of the
aqueous vapour contained in the powder or formed by its explosion.
He considered, relying on M. de Betancourt's experiments, that
the elasticity of steam is doubled by every addition of temperature
equal to 30° Fahr., and his only difficulty appears to have been —
he expressly leaves to posterity the solution of the problem —
why the tension of fired gunpowder is not much higher than
even the enormous pressure which his experiments appeared to
indicate.
It will be remembered that Ptumford's apparatus consisted of
a small but strong wrought-iron barrel, terminated at one end by
a small closed vent, so arranged that the charge could be fired by
the application of a red-hot ball. At the other end it was closed
by a hemisphere upon which any required weight could be placed.
His method was as follows: — A given charge being placed in the
bore, a weight judged to be equivalent to the expected gaseous
pressure was applied. If the weight were lifted, it was increased
until it was just sufficient to confine the gases, and the pressure
was then assumed to be that represented by the weight.
MEASURING PRESSURES IN THE BORES OF GUNS 485
It seems probable that Euiiiford's erroneous determinations
were mainly due to two causes : —
1st. To the weight closing the barrel being lifted, not by the
mere gaseous pressure, but by the products of explosion (produced, it
will be remembered, from a very " brisante " powder, and consider-
ably heated by the red-hot ball), being projected at a high velocity
against it. In such a case, the energy acquired in traversing the
barrel would add notably to the pressure due to the density of the
charge ; and it is again remarkable that the augmentation of pressure
due to this cause was clearly indicated by an experiment designed
for the purpose by Eobins.
2nd. To the gases acting on a much larger area than was allowed
for in his calculations ; and this view appears to be confirmed by the
r4sum4 he gives of his experiments.
No attempt was made for very many years either to corroborate
or amend Count Eumford's determinations ; but, in 1845, General
Cavalli endeavoured indirectly to arrive at the pressure developed by
different kinds of powder in a gun of 16 cm. calibre. His method
consisted in drilling holes in the gun at right angles to the axis, at
different distances from the base of the bore, in which holes were
screwed small barrels of wrought iron, so arranged as to throw a
bullet which would be acted on by the charge of the gun while
giving motion to the projectile. By ascertaining the velocities of
these bullets he considered that the theoretical thickness of the
metal at various points along the bore could be deduced. His
experiments led him to some singular results.
He believed that with some very brisante Belgian powder with
which he experimented a chamber pressure of 24,022 atmospheres
(157"6 tons per sqviare inch) had actually been reached, while with
an ordinary powder and a realised energy of nearly the same amount
the maximum chamber pressure was only 3734 atmospheres (24"5
tons per square inch). With the brisante powder this erroneous
conclusion was doubtless due to two principal causes, viz. : —
1st. To the seat of the small bullet being at a considerable
distance from the charge. Under these circumstances, as later on
I shall have occasion to describe experiments to prove, a far higher
pressure induces motion in the bullet than is due to the tension of
the gases in a state of rest.
2nd. To the brisante nature of the powder. With such powders,
especially in large charges, it has been proved that great variations
of pressure exist in the powder chamber itself, in some cases the
486 ON METHODS THAT HAVE BEEN ADOPTED FOR
pressure indicated at one point of the chamber being more than
double that at others.
It has further been proved that with brisante powders waves of
pressure of great violence sweep from one end of the chamber to the
other, and if Cavalli's small bullet were acted on by one of these
waves an exceedingly high pressure would, without doubt, be
indicated.
3rd. A third cause of error, but much slighter, is due to the
muzzle pressure, when the small bullet quits its barrel, being both
abnormally high and also abnormally sustained ; hence there will be
a considerable increment of velocity after the bullet quits the gun.
It is but fair to add that the results obtained by Cavalli with the
powders which he terms "inoffensive" are, if some correction be
made for the third cause of error alluded to above, not far removed
from the truth.
A Prussian Artillery Committee, under the presidency of General
Neumann, made, in 1854, a great improvement on the plan proposed
and employed by Cavalli.
Their mode of procedure consisted in drilling a hole in the
powder chamber of the gun to be experimented with, in which hole
was placed a small barrel of about 6 inches in length. Now when
the gun was loaded, if in the small barrel were placed a cylinder of
a length equal to that of the projectile, it is clear that, on the
assumption that the pressure in the powder chamber is uniform, the
cylinder and the projectile will describe equal spaces in equal times ;
hence, if we determine the velocity of the cylinder when it quits
the small barrel, we know the velocity of the projectile when it has
moved 6 inches from its seat. By altering the length of the column
of the cylinder placed in the small barrel, and ascertaining the
resultant velocity, the velocity of the projectile at any desired point
of the bore can be determined.
General Neumann's Committee carried out their experiments
only in very small guns and with the grained powder used in those
days. Their results were probably not far from the truth, although
subject to one of the defects to which I alluded in reviewing General
Cavalli's experiments. Indeed, these results were examined and
entirely confirmed by the distinguished Eussian artillerist General
Mayevski, in a very elaborate memoir ; but the experiments of the
Prussian Committee were chiefly remarkable for being, so far as T
know, the first to recognise the variations of pressure which may
exist in the powder chamber itself, variations which may, under
MEASURING PRESSURES IN THE BORES OF GUNS 487
certain circumstances, attain great magnitude, and to which I have
akeady drawn attention.
The results of the Prussian experiments showed, with every
charge fired, two distinct maxima of tension. Other relative maxima
no doubt existed, but the mode of experiment was not sufficiently
delicate to render them perceptible.
Before passing to the more modern methods adopted for deter-
mining the tensions in guns, I must advert to one which has been
repeatedly resorted to during the last one hundred and fifty years.
I mean the method of firing the same weight of charge and projectile
from guns of the same calibre but of different lengths, or, as has
sometimes been done, by successively reducing the length of the same
gun by cutting off a determinate number of cahbres from the muzzle.
It is obvious that if, under the circumstances supposed, we know
the muzzle velocities of a projectile from a gun of, say, 25 caUbres in
length and from a gun of 30 calibres in length, we are able from the
increased energy obtained to deduce the mean pressure acting upon
the projectile over the additional 5 calibres.
The earhest experiments with different lengths of guns appear to
have been made in England as far back as 1736. These experiments,
however, have but little value, as the velocities were not directly
determined, and could only be deduced from the observed ranges.
The same objection applies to the long series of experiments carried
on in Hanover in 1785, and those cited by Piobert in 1801 ; but the
interesting observation that the ranges obtained from guns of 12,
15, 19, and 23 calibres in length were relative maxima cannot be
relied on in any way as showing abnormal variations in the muzzle
pressure accompanying variations in length.
In Hutton's experiments, made with guns varying in length from
15 to 40 calibres, the muzzle velocities were obtained by means of
the ballistic pendulum; and, between these limits of length, the
mean powder-pressure he realised can with sufficient certainty be
deduced.
This remark appHes also to the numerous similar experiments
where the muzzle velocities have been obtained by the more accurate
chronoscopes that have been for many years in common use ; but this
mode of determining the pressure has many inconveniences, and
ceases to be reliable when the bore is of a very reduced length and
the pressures approach their maximum value.
To the important and extensive series of experiments carried on
by Major Eodman for the United States Government in 1857 to
488 ON METHODS THAT HAVE BEEN ADOPTED FOR
1859, the main object of the expenments being to ascertain the
effect which the size of grain of the powder used has upon the
pressure, we are indebted for that officer's most ingenious pressure
gauge ; and the crusher gauge, which is now so extensively used, can
only be considered a modification of Major Eodman's instrument
designed to remove certain difficulties attending the use of the
original instrument.
Major Eodman's gauge is well known, but its construction is
Fio. 1.— Rodman's Pressure Apparatus.
shown in the accompanying drawing (Fig. 1). Major Eodman
applied his gauge in the following manner: —
Desiring to ascertain the pressure at various points along the
bore of a gun, he bored at these points channels to the interior
surface of the bore, and in these' channels cylinders with small holes
drilled down the centre were inserted ; to this cylinder is fitted the
indicating apparatus, carried by Major Eodman on the outside of the
gun, and consisting of an indenting tool G with its knife (shown in
elevation and section). Against the knife is screwed a piece of copper
H. The pressure of the gas acting on the piston I forces the knife
into the copper ; by mechanical means a similar cut can be produced,
and hence the magnitude of the cut gives the measure of the pressure
which has produced it. A small cup at c prevents any gas passing
the indenting tool.
The great improvements that Major Eodman made in gunpowder
are well known. To him we are indebted both for the earliest
experiments on the effect of the size of grain on the maximum
MEASURING PRESSURES IN THE BORES OF GUNS 489
pressure and for the powder adopted by all nations for large guns, I
mean prismatic powder ; but it is a question whether he was not in
some degree led to these great improvements by an erroneous estimate
of the pressures produced, this erroneous estimate being mainly due
to the necessity of placing the Eodnian gauge at the exterior of the
gun; and the effect of this objectionable position would be greatly
exaggerated if the powder experimented with were of a " brisante "
nature.
It is curious that so distinguished an artillerist as Major Eodman
should never have taken the trouble to calculate what energies the
pressures which his instrument gave would have generated in a
projectile ; had he done so he would have found that many of the
results indicated by his instrument were not only improbable but
were absolutely impossible.
As an illustration of Major Eodman's method I take an interest-
ing series of experiments made in smooth-bored guns of 7-inch, 9-inch,
and 11-inch calibres, and so arranged that in each gun an equal
column or weight per square inch of powder was behind an equal
column or weight per square inch of projectile. Under these con-
ditions, in each gun, during the passage of the shot along the bore,
the gases would be equally expanded, and the energy per unit of
column developed at every point in the three guns should be the
same, except for slight differences on account of increased temperature
and pressure in the larger guns, due to the smaller cooling surface in
proportion to the weight of charge.
Major Eodman measured his pressures at the base of the bore
and at every 14 inches along it, and his results are given in the
annexed table, which is a most instructive one : —
Dia-
Weight
Weight
1
meter
of
rt;
Pressure at different Distances from Bottom of Bore |
of
Charge
Shot
°
in Tons per sq. m., at |
Bore.
in oz.
in lbs.
>
In
Sq. in.
Sq. in.
F.S.
Bottom.
14 in.
28 in.
42 in.
56 in.
70 in.
84 in.
7
2-13
1-973
904
16-26
7-08
3-74
3-01
3-06
3-59
3-00
9
2-13
1-99.5
888
29-96
9-42
7-92
6-65
13-16
9-36
10-19
11
2-1.3
1-997
927
38-73
13-04
12-41
10-01
12-68
15-11
11-18
Examining this table, it will be observed, in the first place, that
the muzzle velocities of the equal column projectiles are nearly the
same; that of the 11-inch gun being, as it should be, somewhat the
higher ; hence the energies per square inch must be nearly the same,
490 ON METHODS THAT HAVE BEEN ADOPTED FOR
and the mean pressures per square inch, inducing these energies,
must likewise be the same.
But, for example, comparing the 7-inch and the 11 -inch guns, it
will be noted that in the latter gun the pressures are always twice
and sometimes more than four times as great as in the 7-inch gun,
the mean pressure being nearly three times as great.
The energy should be in the same proportion; hence, if the
pressure observations had been correct, the observed velocity
should have been 1570 feet per second, instead of 927 feet per
second.
It will be noted also that the forward pressures not only differ
greatly in the several calibres, but, for instance, in the 9-inch gun
the pressure at 56 inches from the bottom of the bore is double the
indicated pressure measured at 42 inches. Eodman accepts the
pressures up to and including 42 inches as correct, but ascribes the
irregular pressures in the chase to the vibrations of the metal due to
the discharge.
Some experiments made by the earlier Explosive Committee fully
explain the cause of the differences between the pressures exhibited
by the 7-inch and 11 -inch guns.
In the first of the experiments of this Committee, they used simul-
taneously Eodman's gauge and the chronoscope to which I shall
presently advert. In the former case, of course, the pressure was
determined directly. In the latter it was deduced from the motion
communicated to the projectile. The results were quite irreconcil-
able, as a few examples will show.
In an 8-inch gun, with a charge of 32 lbs. of Kussian prismatic
powder and a projectile of 180 lbs. weight, fired from a vent a little
in advance of the centre of the charge, and called the forward vent,
the chronoscope gave a maximum pressure of 204 tons, while the
Eodman gauge gave maximum pressures in the powder chamber
varying from 267 to 337 tons per square inch. In the same gun,
under similar conditions, a similar charge of pellet powder gave, with
the chronoscope, a maximum pressure of 19 '2 tons per square inch,
while the chamber pressures given by the Eodman gauge varied from
41-6 tons to 49'2 tons per square inch.
But perhaps more striking discrepancies were exhibited by two
series of experiments with E. L. G. of Waltham-Abbey make, fired
from the same gun, and developing in the projectile approximately
the same energies. In the first of these series, with a charge of 20 lbs,
fired from a forward vent, the maximum chronoscope pressure was
MEASURING PRESSURES IN THE BORES OF GUNS 491
13-3 tons, while the Eodman gauge gave pressures varying from 24-6
to 38'9 tons per square inch.
In the second series, all conditions being the same, except that
the charge was fired from the extreme rear, the maximum chrono-
scope pressm-e was 14-3 tons, while the Eodman pressure varied from
31-6 tons per square inch to over 50 tons per square inch, that
pressure being the highest which the instrument was capable of
registering, every observation in this series with the gauge placed at
the seat of the shot being over fifty tons.
Shortly afterwards the Eodman gauges were destroyed, two of
them being blown from the gun.
These discrepancies led the Committee to investigate with certain
powders the variation in pressure indicated when a guage was placed
at the surface of the bore and at the exterior of the gun, as with the
Eodman gauge.
For this purpose they used the crusher-gauge, which admits of
being placed in both positions.
With pebble-powder the gauge placed at the interior of the bore
gave 14-5 tons ; placed under precisely the same conditions at the
exterior it gave 27 tons per square inch. With E. L. G-. the similar
figures were respectively 20 and 57 tons, and with L. G. respectively
19-5 and 45-5 tons per square inch.
The error I have just discussed was due to the position of the
gauge; but Eodman's pressures and the pressures of the Explosive
Committee were exaggerated from another cause. It will be readily
understood that if a pressure of, say, 20 tons per square inch be
suddenly applied to a gauge, and if the resistance to the motion of
the knife be initially trifling, a certain amount of energy will be
communicated to the piston and knife ; and the copper when measured
will indicate not only the gaseous pressure, but in addition a pressure
corresponding to the energy impressed upon the piston during its
motion.
This cause of error can, however, be eliminated by producing
beforehand by mechanical means a cut indicating a pressure a little
less than that to be expected.
Eodman admits that his chase pressures are erroneous; their
exaggeration is no doubt greatly due to the causes I have just pointed
out ; but in my opinion, based upon long experience, no gauge of this
description placed in the chase, where the products of explosion are
moving with a very high velocity, can be depended upon to give
reliable results.
492 ON METHODS THAT HAVE BEEN ADOPTED FOR
If we disregard the energy of the moving products and suppose
the gauge to be acted on by pure gaseous pressure, with a projectile
moving at the rate of 2500 feet per second (and such velocities are
now quite within the range of practical ballistics), the projectile
would pass the entrance to the Eodman gauge in something like the
TO oVoo-l'h part of a second. It is difficult to imagine that the full
indentation could be given to the copper in this small fraction of
time, and, if it were not so given, the gauge would indicate the
pressure at a point considerably in advance of the gauge.
On the other hand, if, as would generally be the case, the products
of explosion moving at a high velocity acted on the piston, the energy
of these products would be reconverted into pressure, and the gauge
would in this case give too high a result.
Major Eodman appears to have considered it impossible that any
gauge could rightly indicate a pressure higher than that indicated by
another nearer to the seat of the shot. This, however, is not so;
nothing is more certain than that, with the powders known as
" Poudres brutales," and, possibly, in a less degree with all explosives,
motion is communicated to the shot by a series of waves or impulses :
and it is easy to see that, if the position of a gauge coincided with
the " hollow " of a wave, while that of a more forward gauge coincided
with the " crest," the latter might easily show the higher pressure.
Later on I shall revert to this point.
The crusher-gauge is a modification of the Rodman gauge, designed
to overcome some of the defects of that instrument, and it is now
Fig. 2. — Crusher-Gauge,
almost universally used for the direct measurement of pressure : it is
shown in the diagram exhibited (Fig. 2), and its action is easily
understood. The powder gases act upon the base of the piston, com-
pressing the copper cylinder; the amount of crush on the cylinder
serves as an index to the maximum tension acting on the piston. It
MEASURING PRESSURES IN THE BORES OF GUNS 493
is usual, where possible, to employ in each experiment two or three
gauges so as to check the accuracy of the determination. Properly
used, very great confidence may be placed in their results; but, as
may be gathered from my remarks on the Eodman gauge, this and all
similar gauges will cease to give reliable information as to the energy
that can be impressed on a projectile, or as to the mean pressure on
the surface of the bore, if there be any probability of the products of
explosion being projected into them at a high velocity. In such a
case the pressure indicated would not be the true gaseous pressure,
such as, for instance, would exist were the products of ignition
retained in a vessel impervious to heat until the waves of pressure
generated by the explosion had subsided. But I defer an examina-
tion of the results given by the crusher-gauge until I compare these
results with those given by the indirect method of deducing the
pressure from the motion of the projectile within the bore.
The method I have adopted for this purpose consists in register-
ing the times at which a projectile passes certain fixed points in the
bore of a gun. The chronoscope (Figs. 3 and 4, p. 498), which I have
designed for this purpose has been so often described that I shall
only here briefly allude to it. It consists of a series of thin discs
made to rotate at a very high and uniform velocity through a train of
geared wheels. , The speed with which the circumference of the discs
travels is between 1200 and 1300 inches per second, and, since by
means of a vernier we are able to divide the inch into thousandths,
the instrument is capable of recording the millionth part of a
second.
The precise rate of the discs' rotation is ascertained from one of
the intermediate shafts, which, by means of a relay, registers the
revolution on a subsidiary chronoscope, on which, also by a relay, a
chronometer registers seconds. The subsidiary chronoscope can be
read to about the -g-oVotb part of a second.
The registration of the passage of the shot across any of the fixed
points in the bore is effected by the severance of the primary of an
induction coil causing a spark from the secondary, which writes its
record on prepared paper gummed to the periphery of the disc. The
time is thus registered every round at sixteen points of the bore.
In the earlier experiments with this instrument the primary was
cut by means of the arrangement shown in Fig. 5, and this was
entirely satisfactory when velocities of from 1400 to 1600 feet per
second were in question. But with the very high velocities now
employed, with velocities, for example, between 2500 and 3500 feet
494 0\ METHODS THAT HAVE BEEN ADOPTED FOR
per second, the knife, instead of being knocked down, frequently cuts
a long groove in the cast-iron projectile, on some occasions reaching
the driving band of the shot before being forced into its place.
FlO. 6.— Original Apparatus for Cutting Wire by Moving Shot.
On account of this defect I have in all recent experiments adopted
the arrangement shown in Fig. 6, which gives extremely satisfactory
results, if care be taken that the plug is sufficiently secured to
Fig. 6.— Improved Apparatus for Cutting Wire by Moving Shot.
prevent its being forced out of its place by the rush of compn
air displaced by the passage of a projectile.
I have ascertained by experiments which I need not here describe
that the mean instrumental error of this chronoscope, due chiefly to
the deflection of the spark, amounts only to about three one-millionths
of a second.
MEASURING PRESSURES IN THE BORES OF GUNS 495
I must not conceal the fact that the determination of the pressure
by this method is attended with very great labour. As an illustra-
tion I have prepared a diagram (Fig. 7, p. 498) of a recent set of
experiments. Usually the pressures are deduced from the mean of
three consecutive rounds fired under the same circumstances.
In this case, owing to the bore being clean, a much higher velocity
was obtained from the first round, and the velocities and pressures
were therefore calculated both for the mean and independently for
each of the three rounds.
The first curves represented in the diagram are the time curves.
So far as the eye can see, the time curves in all cases pass through
the observed points. From the time curves the velocity curves are
deduced, and I have given for each velocity curve the observed
velocities, so that the accordance of the computed curve with the
observed velocities will be seen. The velocity curve being fixed, the
pressure curve of necessity follows, and the diagram shows both the
accordance of the two rounds fired under the same circumstances
and the slight discordance in the forward part of the curve of the
round with the bore clean is very distinctly shown.
Comparing now the methods of determining the pressures which
have been chiefly used in this country — I mean the chronoscope
and the crusher-gauge — if the object sought be merely to determine
the maximum pressure developed with the powders now generally
in use, no instrument can be simpler than the crusher-gauge, and,
when properly used, its indications may be taken as very approxi-
mately correct, but it cannot be relied on to give accurate results
when, placed in positions where the products of explosion are
moving with a high velocity.
The maximum pressures under the conditions I have supposed
are very approximately confirmed by the chronoscope, as may be
seen by comparing the pressures shown on the diagram giving the
results as to pressure obtained with certain new explosives, to
which I shall presently advert. As a general rule, it may be said
that, where the powders are slow in lighting and no wave action
exists, the chronoscope pressures are generally somewhat higher
than those of the crusher-gauge; but the case is very different
where the powder is of a highly explosive or quick-burning
description. With such powders, not only are the crusher-gauge
pressures greatly above those of the chronoscope, but the widest
difference frequently exists between the pressures indicated in
different parts of the chamber in the same experiment. The
496 ON METHODS THAT HAVE BEEN ADOPTED FOR
pressures, moreover, are often greatly above those which would
exist were the charge absolutely confined in a close vessel.
A very striking instance may be cited from the early experi-
ments of the Explosives Committee with a M. L. 10-inch gun (Fig. 8).
The first round was fired with a charge of 87^ lbs. Belgian Pebble,
the charge being lighted in two places. The maximum pressure
with the chronoscope was 25-2 tons. With the crusher-gauge the
pressure in the chamber varied from 22*2 to 24'8 tons per square
inch, while the energy developed by the powder on the shot was
6240 foot-tons. With the second round, all conditions being the
same except that the charge was fired at a single point, the
chronoscope pressure was as nearly as possible the same ; but the
chamber pressure was, at the rear, 79'1 tons; in the middle, 52*0
tons ; at the seat of the shot, 39'5 and 48'0 tons per square inch.
Fig. 8.— Position oi Pressure Plugs in 10 inch Gun.
C 2 4
A similar large excess of pressure was shown at points 1 foot and
2 feet in advance of the seat of the shot, and the crusher-gauges
did not show their normal pressures until points 5 or 6 feet from
the seat of the shot had been reached.
Yet with the violent difference in pressure shown between the
crusher-gauges in this round and in the previous round (which I
have just cited), the difference of energy developed in the shot
was exceedingly trifling, being only 6249 foot-tons, as against
6240.
I believe I have expressed pretty clearly my views that crusher-
gauges placed in the chase are for absolute determination not of
much value, and their main use, if used at all, is to give comparative
results. But the same remark does not apply to crusher-gauges
placed in the chamber.
Gases moving at a high velocity in the chase are, so to speak,
performing their proper function ; but the same is not true of those
MEASURING PRESSURES IN THE BORES OF GUNS 497
violent waves of pressure in the chamber which appear to
accompany the explosion of all brisante powders, and which occur
either when the projectile has hardly moved at all or when it is
moving with a comparatively slow velocity.
It is our object, and in this we have had great success, to avoid
these waves as much as possible; and in attaining this end our
indebtedness to the crusher-gauge is very great, as this instrument
has made plain to us not only the extreme violence but the
variability of these oscillations.
I have heard it urged that these waves of pressure are, after
all, not of high importance, because their maxima act at the same
time only upon a very small section of the bore, and the continuity
of the metal is amply sufficient to resist the stress.
This is no doubt true, but it is not true of the base of the bore,
which in modern guns is almost invariably a movable piece, and
which under certain circumstances might have to sustain the full
force of the violent pressures, a sample of which I have cited.
To ascertain the mean pressure throughout the bore, it seems
to me that there is no method so satisfactory, despite its attendant
labour, as that of making the projectile write its own story. In
that case we cannot fall into the error of making the pressures
three or four times as great as are necessary to generate the energy
the projectile has actually acquired, while occasional errors, due
to causes I have not time to explain, are easily detected and
eliminated.
To give an idea of how great is the range of velocity over which
these experiments have been carried, I exhibit here diagrams (Figs. 9
and 10, p. 498) showing the velocities and pressures obtained with
several of the new explosives which in recent years have attracted so
much attention. Observe also how closely, with the exception of
the one somewhat brisante powder, the results given by the
chronoscope accord with those given by the crusher-gauge. Where
these differ, as I have elsewhere pointed out, the two modes of
research so widely different are complementary to each other.
The chronoscope takes little or no note of the violent oscilla-
tions of pressure acting during exceedingly minute intervals of time.
On the other hand, if with the explosives I allude to we trusted
to the indications of the crusher-gauge, we should arrive at a most
erroneous idea of the energy communicated to the projectile.
In concluding, if I may venture to quote the excuse of a much
more eminent man than myself, I have only to express my regret
2 I
498 MEASURING PRESSURES IN THE BORES OF GUNS
that I have not had time to condense the remarks with which I
fear I have fatigued you, while at the same time I am aware that
there are many important points in connection with my subject
which I have left altogether untouched, and others upon which I
have touched that require further elucidation.
f=^=0
IMjCDO:
K}C3a
^2
MXII30
ONONIS M3d
O
C
r 4
uJ\\
2
1
ocV.
sajqiieQ ofr ^'
I
V- t
4p
i
i
1 :
" !^
\\
/'
2
\\
\
///
2„
> <*
CM
.=...,=.^
^.^^.^-^^^^^-^^'"^^
■,^, _^
o
*n -t tf\
^ i i J> .i
,1 «,
rf) * <«>
ci^r~o
HONI 3MVnC)S y3d
SNOl Nl
3bnss3ad
O CT>0 OO t~ '
'honi ayyncss y3d snoi
XIV.
THE EISE AND PKOGEESS OF EIFLED NAVAL
ARTILLEEY
{Paper read at the Institution of Naval Architects,
Neiocastle-upon-Tyne, 1899.)
At the last meeting which the Institution of Naval Architects held
in this city, an invaluable paper by Lord Armstrong and Mr
Vavasseur was read and discussed, and it appears to me that, using
this paper as a landmark, it may be convenient, and not uninteresting,
to give a brief sketch of the subject of my paper during the fifty
years of my connection with artillery, and to note the striking
progress which artillery science, in common with other applied
sciences, has made during the last years of the century now drawing
to a close.
In the paper to which I have referred, Lord Armstrong and Mr
Vavasseur draw attention to the primitive carriages on which the guns
of the first half of the present century were mounted ; but the
guns themselves were nearly as primitive, differing in little, except in
size and power, from those with which the fleet which met the
Great Armada were armed.
It is both interesting and instructive to compare the guns which
in 1850 formed the principal armament of the most powerful frigates
or line-of -battle ships, with the guns which now form the principal
secondary armament of first-class cruisers and battleships.
In the year I have mentioned, and it will be remembered that
within a short period the long peace which succeeded the Napoleonic
wars was broken, the principal guns with which our ships were
armed were 32-prs. They were, we must admit, of very rude con-
struction, mere blocks of cast iron, the sole machining spent upon
them being the formation of the bore and the drilling of the vent.
The velocity of the shot was about 1600 feet per second, and the
energy developed in it by the charge was about 570 foot-tons.
500 RISE AND PROGRESS OF RIFLED NAVAL ARTILLERY
The carriage upon which this rude gun was mounted was even
more rude. It was made, as described by Lord Armstrong and Mr
Vavasseur, entirely of wood ; generally, in later years, of teak or
mahogany. It was carried on wooden trucks, or sometimes the rear
trucks were replaced by a chock. The recoil was controlled by the
friction of abnormally large wooden axles, and sometimes by wedges
acting on the trucks, and was finally brought up by the breeching by
which the gun was attached to the vessel's side. The elevation was
fixed by quoins resting on a quoin bed, and handspikes were used
either for training or for elevating. For the running out, at the date I
have mentioned, blocks and tackle were generally employed.
To work, with any degree of smartness, such rude weapons, a very
strong gun's crew was necessary, and, indeed, the gun and its
carriage were absolutely surrounded by its crew. For the sake of the
younger members present, who may probably not have seen the
weapons I have been describing, I show in Plate I. * (p. 520), a
32-pr. gun of H.M.S. Excellent, with its crew at practice.
In the year 1858 the first great step in artillery progress was
made. In that year the Committee on Eifled Cannon recommended
the introduction of the rifled Armstrong guns into the service, and
the experiments which were made with these and other rifled guns
opened the eyes of all who gave attention to the subject to the great
advantages possessed by the new artillery.
With regard to range, accuracy, and penetrative power, the
superiority of rifled guns was so conspicuous that nearly all artillerists
were at once convinced that the days of smooth-bored guns were
numbered. The advantage in range at high angles of elevation at
first excited some surprise, as the velocity of the projectile was, from
causes to which I shall later advert, very much lower than in the
case of the smooth-bored guns.
Comparing, for example, the velocities and energies of the 32-pr.
smooth bores, to which I have adverted, and the 40-pr. E. B. L. guns,
which, on the introduction of rifled guns replaced them, the com-
parative muzzle-velocities were respectively 1600 and 1200 feet per
second, and the comparative muzzle-energies respectively 570 and 400
foot-tons. It is hardly necessary to point out that at comparatively
short distances the superiority of the rifled gun, both in regard to
range and penetration, manifested itself.
But in these early days of rifled artillery, the point which
attracted most attention was the great increase of accuracy. The writer,
* From Sir Howard Douglas's Naval Otmnerif.
RISE AND PROGRESS OF RIFLED NAVAL ARTILLERY 501
in using the method of " Least Squares " to determine the relative
accuracy of rifled and smooth-bored guns of approximately the same
weight, showed that, at a range of 1000 yards, half the shot fired
from a rifled gun fell in a rectangle of about 23 yards long by 1 yard
wide, while in the case of the smooth bore the similar rectangle was
about 145 yards long by 10 yards broad.
The objection to the use of iron and steel as a material for
carriages was shown by careful experiment to be founded on prejudice,
and the introduction of iron carriages has been so fully described in
the paper by Lord Armstrong and Mr Vavasseur that I need not
here further refer to it.
I have alluded to the reduction in velocity of projectiles fired
from rifled guns when these weapons were first introduced, and this
reduction arose from two causes. First, because the flatness of the
trajectory and increased penetrative power of rifled projectiles at
long ranges were at first supposed to be sufficient ; and, secondly,
because the numerous failures of rifled guns, with the materials and
modes of construction then in vogue, rendered artillerists cautious as
to the stresses to which rifled guns, especially those with deep
grooves, were subjected.
But the battle between guns and armour rendered it a matter of
first-rate importance to increase the potential energy of our rifled
guns, and the first steps in this direction were made by the Explo-
sives Committee, who, by their experiments, showed that, with
improved forms of ]iowder, the velocities, which had generally run
between 1200 and 1300 feet per second, could, in the same guns,
be raised to 1600 feet per second, the increase in velocity being at the
same time accompanied by a very considerable reduction of maximum
pressure.
But perhaps the most important step was made by my firm, who,
acting upon certain experiments carried out at Elswick, and which
formed the subject of memoirs by myself and Sir F. Abel, made, in
1877, 6-inch and 8-inch guns, with which, while the maximum
pressures remained the same, the velocities of the rifled projectiles
were at a single bound raised from 1600 to 2100 feet per second,
thereby increasing the energies by nearly 75 per cent., and immedi-
ately another reconstruction of guns and their mountings became a
necessity.
At the same time, from the increase in the length of guns, which
the slow-burning powders and high energies then introduced de-
manded, a return to breech -loading from the somewhat retrograde
502 RISE AND PROGRESS OF RIFLED NAVAL ARTILLERY
change to muzzle-loading, which had some years before been adopted,
was also necessitated.
Nearly all these breech-loading guns were arranged for velocities
somewhere about 2000 feet per second, the mountings also were
greatly improved ; but it is unnecessary for me here further to allude
to these improvements, as they have been fully described in the
paper by Lord Armstrong and Mr Vavasseur, to which I have so
often referred.
At about the date of that paper, however, experiments were being
made in three directions, the whole of which experiments were
destined to have a most important bearing on the progress of naval
artillery. The first of these had reference to the question as to
whether gunpowder, which had since the days of Eoger Bacon, that is
for nearly seven centuries, no serious competitor as a propelling
agent for artillery purposes, was to retain its pre-eminence.
The second was due to my own initiative. Seeing the great
advantages that had attended the introduction of the small rifled
guns designed by Hotchkiss and Nordenfeldt, my firm had con-
structed 4-7-inch and 6-inch quick-firing guns, and submitted them
to the Admiralty. The success of these guns, both in our own and in
foreign navies, was rapid and complete ; and it is not too much to
say that, for cruisers and the secondary armaments of battleships,
their adoption amounted to another reconstruction of artillery.
The third series of experiments were on the question of the
introduction of high explosives as bursting charges for shell— a
question of great importance, both in an artillery point of view, and
as affecting naval construction.
With reference to the first of these questions, it is unnecesary to
tell you that the long pre-eminence of gunpowder has come to an
end. In this country, for artillery purposes, it has been replaced by
the cordite of Sir F. Abel and Professor Dewar, and this explosive
has also been used extensively abroad. Many other nations employ
ballistite or kindred explosives, giving results generally similar, but
having a somewhat less potential energy. Having spent many years
in experimenting on gunpowder, I cannot quit that interesting
subject without regret; but, as I have also experimented largely
with cordite and other kindred agents, I am obliged to confess
that the new explosives have many and great advantages. The
absence of smoke, and an increase of energy, with the same maximum
chamber pressure, of about 50 per cent., are advantages much
too great to be overlooked. There is one point, however, to which I
RISE AND PROGRESS OF RIFLED NAVAL ARTILLERY 503
ought to allude, and which is, I believe, at present exercising the
minds of the authorities to a considerable extent. I mean the rapid
destruction of the bores due to the erosion by cordite. It must be
borne in mind, however, that if taken in relation to the energy
developed, the erosion of cordite differs but little from that of brown
prismatic powder, which is also very erosive, and gives rise to erosion
of a much more objectionable character. Erosion is, in my opinion,
caused by three factors — the heat of combustion, the pressure,*
and the motion of the products of combustion — not to any chemical
action. This view is borne out, not only by my numerous experi-
ments on this subject, but by the state of the surface of close
vessels in which large charges have been fired, and by the examina-
tion of the chambers of guns from which a large number of charges
have been fired.
In the forward part of the chamber, where the gases are in rapid
motion, the erosion is decided ; but in the rear of the chamber, where
the temperature and pressure are highest and longest continued, but
where there is little or no motion, there is no trace of erosion. Let,
however, but a slight leakage past the pad occur, and the effects of
erosion are immediate and decided.
The object, then, at which we have to aim is to diminish the
temperature of explosion, and I am not without hopes that this
greatly-to-be-desired end may before long be achieved.
The velocities obtainable with cordite are very high. There
would be no difficulty, should it be desired, in approximating with
ordinary projectiles to 3000 feet per second ; but, for many reasons,
I consider very high velocities objectionable, and, if a given energy
be required, would prefer to see it represented by a lower velocity.
I may here mention that, with a 100-calibre 6 -inch gun, and with a
projectile of the dimensions of the ordinary 6-inch projectile, but of
aluminium, I have obtained a muzzle velocity of close upon 5000
feet per second.
Turning now to the quick-firing guns, I think it will be most
convenient to consider the guns themselves in connection with their
mountings ; because these last, when rapidity of fire is in question,
are quite as important as the arrangements of the guns themselves.
Early in the year 1887, the gun and mounting shown in Plates
IV. and V. (p. 520) were submitted for trial on board H.M.S. Handy.
The gun was the first mounted on the Elswick cradle, having the
* Subsequent experiments showed that pressure exercised but little influence on
the erosion.
504 RISE AND PROGRESS OF RIFLED NAVAL ARTILLERY
recoil-press and spring box beneath the cradle, the piston rods and the
attachments for compressing the springs during recoil being fixed to
a horn on the gun. The weight of the gun and mounting was taken
on balls under the pivot, and the mounting and shield were carefully
balanced. The whole weight of the gun and mounting was about
4 tons 12 cwt., and it could be trained quite easily by the shoulder,
no gear being used. The sights were placed on the cradle, and did
not recoil with the gun. No. 1 could with ease train with the
shoulder-piece, work the elevating gear, lay, and fire by means of an
electric pistol. During the operation he was quite clear of the breech,
and could keep the gun pointed continuously on the object.
With this gun and mounting a very great advance in rapidity
of fire was obtained. The breech mechanism was of the three-motion
type, and was very quick and handy; but the great speed was
obtained by the careful design of both gun and mounting, in such a
manner that the movements of one did not interfere with those of
the other. At the trial above mentioned ten rounds were fired in
47| seconds, and later as many as fifteen rounds per minute were
obtained. An interesting incident connected with this gun and
mounting may be mentioned. The gunboat Mastiff was ordered to
fire ten rounds as rapidly as possible from her service 5 -inch B. L.
gun. The time taken for the ten rounds was 6 minutes 16 seconds,
so that the quick-firing gun fired its ten rounds before the then
service gun fired its second shot.
About the same time a great improvement was made in the mode
of mounting of the smaller 3-pr. and 6-pr. quick-firing guns. Up to
that date they were movmted on crinoline, or so-called elastic, stands ;
but, with this pattern, the strains on the decks and holding-down
bolts were very severe. The mounting shown in Plates VI. and VII.
(p. 520), in which the gun recoils in the line of fire, was submitted
for trial at Portsmouth, and proved itself so successful that it was
at once adopted in our own and many other navies.
In 1890 an important improvement in quick-firing mountings was
introduced, viz., the pedestal mounting shown in Plates VIII. and
IX, (p. 520). The cradle is of the same type as that of the 4-7-inch
quick-firing gun above mentioned. The carriage is of forged steel in
the form of a " Y," having a long shank which fits into the pedestal
and forms the pivot. The whole weight is taken on the end of the
pivot, and the mounting can be trained with ease by a few pounds
applied at the shoulder-piece. The pedestal is very solid, is of forged
steel, and affords excellent protection to the pivot ; the base is also
RISE AND PROGRESS OF RIFLED NAVAL ARTILLERY 505
small, and, there being no rollers or roller-paths, the deck may be
considerably distorted without interfering with the working of the
piece.
The shield is of a very substantial character, 3 inches thick,, and
perfectly balanced ; it is attached to the carriage by means of flexible
stays, so arranged that, if the shield be struck, the stays yield, and a
very reduced shock is transmitted to the carriage.
This mounting was the first to be fitted with the bar and drum
sight, also shown on Plates VIII. and IX. (p. 520).
In 1891 an experimental mounting of this type was made for a
47-inch gun. It was fitted with a 3-inch shield with sloping roof,
carried by yielding stays, and with this mounting a firing trial was
carried on to compare its resistance to injury with that of a centre
pivot roller-path mounting, in which a shield 3 inches thick formed
an integral portion of the mounting, which had in addition an outer
shield 1 1 -inch thick. The latter mounting was disabled after two
rounds, one each from a 3-pr. and a 6-pr. This trial showed con-
clusively that steel castings, although giving excellent tests, could
not withstand a severe blow from a projectile. The pedestal mount-
ing received no less than twelve rounds before it was disabled, four
from a 3-pr., six from a 6-pr., and two from a 4-7-inch gun ; and it
would not then have been disabled, had the pedestal been made, as
they are now, of forged steel. In the experimental mounting the
pedestal was of plate and angle ; the last projectile fired penetrated
the pedestal and jammed the pivot. Even then the damage was not
serious, and could have been rectified in a few hours, but with this
exception, the mounting in all other respects was as good as ever.
This type of mounting for guns up to 6-inch calibre is now
almost universal in our own and many other services.
In Plates X. and XI. (p. 520) are shown a 6-inch mounting of the
latest type as arranged for a casemate between decks. It differs
chiefly from those previously described in having training gear fitted
on both sides, and in having a special arrangement made for removing
the balls which form the bearing at the base of the pivot from the side
instead of from below. A small jack is provided to take the weight
of the gun and mounting, and in a few minutes the balls can be
removed, examined, and replaced. The arrangements permit the gun
to be elevated through the whole angle of 22° in 11 seconds, and to be
trained through the whole angle of 120" in 16 seconds by one man.
An important improvement in the cradles for 6 -inch and larger
mountings was the result of a trade dispute. In 1894 a strike of
506 RISE AND PROGRESS OF RIFLED NAVAL ARTILLERY
moulders took j)lace. The cradles up to that date had been made of
cast steel, and, as at Elswick, considerably more than 100 mountings
were stopped for want of cradles, it was determined to substitute
forged for cast steel. A new design was consequently made, and a
much more satisfactory cradle, lighter and more reliable, was pro-
duced. At Elswick cast-steel cradles are not now made, unless
specially ordered.
I am now in a condition to make the comparison referred to in
the opening sentences of this paper. In Plate III. (p. 520), I have
placed side by side diagrams of the 6"3-inch 32-pr. of 1850 and of the
6-inch 100-pr. of the present day, while Plates I. and II. (p. 520) show
the crews necessary to work the guns. You will observe the diagrams
give the pressures, velocities, and energies of the two guns. The velo-
city and energy given by the 32-pr. are, respectively, about 1600 feet
per second and 570 foot-tons. The corresponding figures for the 6 -inch
Q. r. are 2570 feet per second and 4580 foot-tons. But the rapidity
of fire and accuracy of the modern gun are even more remarkable.
Most of you are doubtless aware of the conditions under which
target practice is carried on in the navy. Each gun's crew has
3 minutes to fire as many rounds as they can with accuracy, the
variable range commencing at about 2200 yards, diminishing to about
1600 yards, and again increasing to 2200 yards. In H.M.S. Blahe
the best gun's crew fired eighteen rounds, hitting the target fifteen
times, while the total number of rounds fired by her ten guns was
one hundred and forty-eight, the target being hit one hundred and
ten times. H.M.S. Royal Arthur did nearly as well, the best gun
having fired eighteen rounds, striking the target fourteen times.
In Plates XII. and XIII. (p. 520) are shown two systems of dis-
mounting gear for 6 -inch guns. The bogie system is used for the upper
deck, or for casemates where it is not necessary to run the gun back for
stowing. It is only used for purposes of examination, and is found to
be very convenient. One pair of bogies is usually supplied per ship.
The between deck dismounting gear is shown on the same plate.
It consists of a lever L mounted on rollers on an overhead rail, which
can be run backwards or forwards by means of an endless chain on a
sprocket wheel, worked by means of worm gear and hand chain as
illustrated. The lever L is readily attached to the cradle at aliout
the centre of gravity, and the screw J to the breech end. Then, by
means of the capstan head, the lever L takes the weight of the gun,
and gun and cradle are run back together, rested on chocks, and
secured as shown. This system of dismounting gear has been
RISE AND PROGRESS OF RIFLED NAVAL ARTILLERY 507
rendered necessary from the great projection of the muzzle when
guns are mounted on the broadside, due to the length of the guns.
The time occupied from commencement to " gun secured " is about
4 minutes, and from casting loose to gun in firing position, aliout
3 minutes.
The above arrangement is that now in the service, but a new
design has recently been made, fitted up, and experimentally tried at
Elswick. From the results of these experiments, it seems probable
that the above times will be reduced to something between a half
and a third of those I have mentioned.
Plates XIV. and XV. (p. 520) show an 8-inch C. P. mounting for
swift cruisers. The man at the sights can look over the top of the
shield, thus commanding a good field of view, his head being protected
by a hood. Electric and auxiliary hand training gear is provided,
either of which can be applied at once, should the other be disabled.
The elevating gear is worked entirely by hand, the trunnions being
mounted on Mr Brankston's anti-friction arrangement, with knife
edges supported on springs to relieve the shock when the gun is fired.
So easily does this gear work, that one man can elevate or depress
the gun at the rate of 2° per second. With the hand training gear
one man can train the mounting through 60° in 25 seconds, and with
the electric gear through 180' in 30 seconds. The shield is 4| inches
thick, and is supported on elastic stays in the usual manner. The
powder-supply is brought up the centre, ancl is delivered at the side
under cover of the shield. The axial hoist for this purpose is shown
in Plate XVI. (p. 520), and is so arranged that, when one charge
is going up, the empty case is going down, thus effecting a great
saving of time and labour, as the weight of the two cases balance
each other, and there is thus only the actual weight of the charge
to lift. Four rounds have been fired in a minute from this gun.
In 1889 Mr Vavasseur and the writer submitted to the Admiralty
the design of a mounting so arranged that the gun could be fired at
all elevations up to 35° or 40", the firm having been requested by a
foreign government to consider whether or not such an arrangement
was feasible.
The naval authorities were much pleased with the design : but,
as the arrangement was altogether novel, it was not unreasonably
stipulated that, before it could be introduced into the service, its
success must be proved by an experimental mounting being made,
and by passing a satisfactory firing trial.
My firm agreed to the stipulation, and a high angle mounting for
508 RISE AND PROGRESS OF RIFLED NAVAL ARTILLERY
a 9'2-inch gun was made at Elswick at the firm's expense, and fitted up
in the Handy. It passed a most satisfactory firing trial in April 1890.
The total weight of the gun and mounting was 54 tons, and it
could he trained quite easily by hand power. The gun had a range
of elevation from 5° depression to 40"^ elevation, and an arc of 45°
could be traversed in 30 seconds by one man. At the trial, roimds
were fired at angles varying from 5° depression to 39° elevation, and
the results were most satisfactory. The range of three of the rounds
at 39° was estimated to be about 10 miles, but the shot could not be
seen to strike the water.
In this mounting the slide was horizontal, and the carriage was of
the Vavasseur type, the recoil-press and carriage being in one piece of
forged steel ; the gun consequently, did not recoil in the line of fire,
but horizontally, and was returned to the firing position by means
of springs, the force of the springs being regulated by means of a
controlling ram in the recoil-press. Illustrations of this type of
mounting are shown in Plates XYII. and XVIII. (p. 520). A con-
siderable number of vessels, chiefly in foreign navies, are fitted with
this form of mounting.
Plates XIX. and XX. (p. 520) show the type of armoured gun-
house arranged for the very powerful Chilian cruiser O'Higgins. This
mounting affords excellent protection to the gun's crew, having
8-inch armour in front, and 5-inch on the sides and rear ; the trunk
for the supply of cartridges being also protected by 5-inch armour.
The gun and mounting can be trained either by electric gear or by
hand power. A store of projectiles is carried in the gunhouse for
ready supply. The cordite charges come up the central trunk by
means of a hydraulic motor; arrangements are also provided for
bringing shell up this trunk to replace the ready supply.
Plates XXI. and XXII. (p. 520) show the type of twin-armoured
gunhouses supplied to several Japanese cruisers. The following
points may be mentioned. The training gear can be worked by
hydraulic, electric, or hand power. Sighting gear for both guns is
supplied to both sighting stations, and the mountings can be trained,
and both guns elevated or depressed, from either station. A good
supply of projectiles is carried in the gunhouse (30 per gun), and an
electric bollard is provided to enable this supply to be replaced. The
cordite is supplied through a central trunk, protected by an armoured
barbette.
The armour of the gunhouse is attached to the turntable by means
of elastic stays.
RISE AND PROGRESS OF RIFLED NAVAL ARTILLERY 509
Before passing to the arrangements connected with the guns,
which form the principal armament of battleships, I may mention a
discussion which illustrates in a striking manner how widely sepa-
rated are the ideas held by those who now rule the Queen's Navy and
by those who held a similar position forty years ago. About thai
time I was secretary to a long-forgotten committee, called the
" Committee on Plates and Guns," and among the subjects discussed
was the design of a rifled gun of 7 tons weight. The naval
authorities, however, vere very strong in insisting that no gun
weighing more than 6 tons could be safely carried on board ship,
and I believe that the weight selected for that extraordinary weapon
called the " Somerset Gun " was due to a compromise betw^een the
weights I have mentioned.
Turning now to guns of larger calibre, I propose to draw attention
to some of the designs of the last twelve years. At the beginning of
this period was designed the Be Umherto, with two barbettes, each
having a pair of 13|-inch 68-ton guns. The mounting of these guns
is principally noteworthy, because of two features which have again
come to the front in more recent ships, viz. : what is known as all-
round loading of the guns, which, in the Re Umherto, were protected
by a circular barbette, and the provision of what is known as a
working chamber below the turntable, into which depend, from the
rear of the gun, hoists which are charged from this working chamber,
the charges being first brought up by a central hoist terminating at
the floor of the working chamber. As will be seen from Plates
XXIIL and XXIV. (p. 520), the guns had trunnions, partly in order
that they might be also available for land service. They were
exactly balanced on these trunnions, in order to reduce the work of
elevating and depressing the guns (which in this design is entirely
done by hand) to a minimum ; that is, so far as the work of lifting-
weight is concerned.
To reduce the work of overcoming the friction of the trunnions,
a special device is placed under the trunnion of the gun between the
plates of each cheek of the carriage. It consists of the arcs S,
supported on the spring T, Plate XXIY. (p. 520). The springs are
made powerful enough to lift, say, 98 per cent, of the weight of the gun,
so that, although the gun is not thereby lifted off the usual trunnion
bearing in the carriage, the majority of the weight is transferred to
the rolling surface of the arc U, and to its point at W, where the
knife-edge friction is insignificant. The recoil-presses are made on
the Vavasseur principle. The piston rods pass out of their cylinders
510 RISE AND PROGRESS OF RIFLED NAVAL ARTILLERY
at opposite ends, and are attached to the gun sHdes, so that one may-
be used as a means of running the gun out, and the other be used for
running it in. The hydraulic pressure for this purpose is passed to
and from the cylinders by means of a passage drilled up the centre
of each piston rod, so that the connection of the hydraulic system is
only made to that end of each cylinder which never receives a high
recoil pressure.
The ammunition hoists behind the gun are carried upon the
centre girder of the turntable to which their guides are attached.
The powder-tube is inclined at the loading angle, and is partially
blocked up at its bottom end, so that the powder when passed into it
may not slide too far through. The powder-charge, delivered by the
central hoist, is passed over by hand in separate parts to this tube.
The shot trough is also fixed at the loading angle, and is pivoted, so
that it may be slung round to receive the shot from the central hoist.
The cylinders for working these ammunition hoists are telescopic,
the smallest ram having insufficient power to lift the powder and
shot, so that it is not till after the shot is rammed into the gun that
the hoist has power to lift the cage to the height required for
ramming the powder home.
The breech mechanism is hydraulic, and is carried in the turn-
table within the protection of the armour, so that, although the guns,
as will be noticed, are almost entirely exposed, there is no very
vulnerable part about them.
The central ammunition hoist passes up from the shell-room and
magazine passages to the battery through an armoured tube. The
cage is almost cylindrical, and is provided with a turntable top.
Before sending the hoist down to receive the charge, it is necessary
to turn this table top into one particular position, and this position
will present the shell and powder receptacles in the correct direction
for charging the hoist down belov/.
If, when the hoist comes up with the charge, it is found that the
gun is so trained that the rear hoist is not in line with the centre
hoist, the turntable top can be revolved to the proper position, and if
the gun turntable is in motion the hoists can be locked together
while the shot is passed through from one to the other.
The trial of this mounting took place on 26th April 1893, and
later on those of the Slcilia and Sardinia, which were of the same
design.
Six ships of the Royal Sovereign class had their guns mounted in
oval barbettes with one fixed loading station. These guns also were
RISE AND PROGRESS OF RIFLED NAVAL ARTILLERY 511
almost entirely devoid of armoured protection. Then came the desire
to have a second or alternative station for loading, together with
breech mechanism carried on the gun, and armoured shields protecting
as much of the guns as possible, which found expression in the design
shown on Plates XXV. and XXVI. (p. 520).
This design, like some of the earlier ones, had a single fixed load-
ing station, but had an important alteration to the hydraulic loading
rammers. These were made with a trough, attached to and moving
with the second, or larger, ram.
The ammunition cage, instead of coming up between the rammer
and the gun, was placed alongside the rammer, so that the shot and
powder charges could readily be rolled out into the trough carried l)y
the rammer. Moreover, as this trough advances towards the gun it
acts as a locking bolt to secure the gun turntable in position, and then
to secure the gun-slide in position, and finally it bridges over the
breech-screw threads in the gun.
The smaller ram of the rammer next advances, pushing the
charge from the trough into the gun. This arrangement removed all
fear of damage to the rammer by movement of the ammunition cage,
training of the turret, or depressing the gun.
For the alternative method of loading, the gun was placed in line
with another hydraulic rammer fitted at the rear end of the gun shield
where space was allowed for a small chamber, giving room to work
projectiles from a small bin to a trough in line with the rammer.
This loading gear can, of course, be used in any position of train-
ing of the guns.
A concurrent advantage of the shield is that the sighting hoods are
placed well above the guns, thereby giving a better all-round view.
At the same time, however, it should be noted that the men at the
guns are not so well protected, and there is a possibility that they
may suffer from small projectiles entering at the gun ports.
The ships fitted according to this design are seven, of the Majestic
class, also the Japanese ships Fugi and Yashima. The last five of the
Majestic class might have been fitted with all-round main loading
positions, had it not been, I believe, that the frames for the oval bar-
bettes were well advanced, and it was feared that any alteration of
design might cause delay.
The Canopus mounting, designed at Openshaw, was therefore the
first for the English Admiralty to be carried out with all-round
loading.
I regret that time has not allowed me to have lithograph drawings
512 RISE AND PROGRESS OF RIFLED NAVAL ARTILLERY
of this prepared. The chief feature is the provision of a roomy work-
ing chamber below the gun turntable, into which the powder and
projectiles are brought by suitable central hoists. These hoists are
fixed in relation to the ship. In order to transfer the shot to the
gun hoists, hydraulic cranes are fitted.
Also in the working chamber there are shell bins holding twenty-
four rounds per gun, which are also commanded by the hydraulic
cranes, and these shell bins could all be exhausted first, and be
replenished by the central hoists at leisure.
There is no doubt some disadvantage in having two sets of hoists,
the central ones and those behind the gun, as it involves an additional
set of operations to transfer the charges from one to the other ; but
against this must be set the fact that, having a large store of pro-
jectiles immediately under the gun turntable, an ordinary action
might be fought before this store could be used up, so that the central
hoists might never be required for use during the action. We have
indeed brought forward the idea that, by still further increasing the
storage of shell below the turntable, the provision of central shot
hoists and shell-rooms at the bottom of the ship would be unnecessary.
I am assured by my colleague, Mr Watts, the Chief Constructor
at Elswick, that, although it might not be possible to carry out this
idea in existing ships, there would be no difficulty in designing a
ship to meet this requirement, and I would point out that there
would be a considerable total saving of weight if the central hoists
and the shell-room gear could thereby be dispensed with. This is,
perhaps, particularly observable in the next 12 -inch design to which
I now draw your attention, and which is shown on Plates XXVIL,
XXVIII., and XXIX. (p. 520).
This mounting is fitted with a pair of main hoists, each carrying
a projectile and powder-charge from the bottom of the ship to the
rear of the guns. It also has, as an alternative, a pair of shot
hoists reaching from the bottom of the ship to the rear of the
gun shield, and arranged to deliver or pick up shot from the working
chamber.
These latter hoists work either by hydraulic power, or alter-
natively, by hand. Also a pair of powder hoists reaching from the
bottom of the ship to a platform placed between the two guns ; these
work by hand only. In addition, it is provided that either shot or
powder can be hoisted from below to the working chamber by hand
tackle as a last resource.
Very great precautions are, therefore, as }'0U see, taken to make
RISE AND PROGRESS OF RIFLED NAVAL ARTILLERY 513
sure that the projectiles stowed below can be brought up to the gun.
Nevertheless, three rounds per gun are stowed in the gunhouse, and
eight rounds per gun in the working chamber.
The weight of gear for transporting shell in the shell-room and
the weight of tlie shell hoists and their gear is about equal to 54 tons
per ship, or equal to half the weight of the projectiles stowed in the
shell-rooms.
In order to charge the main hoists a revolving platform is pro-
vided in the shell-room, having on each side trays for carrying a
couple of projectiles. This revolving platform is first locked in one
particular position to the ship, and shell are placed in the trays by
overhead tackle in the shell-room. The platform is then unlocked
and moved to whatever position is necessary to bring the shot trays
opposite the hoist doors. It is then locked to the hoist trunk until
the shot are required to be passed into the hoist cages.
To manage this heavy platform in a seaway, it has been thought
necessary to revolve it by hydraulic engines. These platforms and
the gear for working them have a total weight per ship of 9 tons.
The alternative method of charging the hoist cages, which I myself
prefer, shown on Plates XXX. and XXXI. (p. 520) and Figs. 1 and 2,
Plate XXXVa. (p. 520), and which is being fitted to four Japanese
ships, is by a pair of overhead circular rails, the outer one of which
is fixed to the ship and the inner one to the trunk of the hoist.
A small four-wheeled chariot runs upon these rails. The point of
suspension of the supporting tackle carried by this chariot can be
shifted so as to throw the weight entirely on the one rail or the other.
While picking up shot on the ship the load is on the wheels which
run on the fixed rails. The suspension point is then shifted to throw
the load on to the moving rail, so that, while the shot is being placed
in the hoists, any movement of the hoists carries the projectile with
it. I have already referred to the danger, which I consider exists, of
small shot entering the gun ports. In these four ships this is met
by providing on the top of the gun a port protector, indicated at A,
Plate XXXIII. (p. 520). Alternative electrical training gear, con-
trolled and worked by the same hand-wheel as is ordinarily used for
the hydraulic training, is also being fitted for these Japanese ships.
The design for the FormidabU and three sister ships, Plate XXXII.
(p. 520), also has a working chamber below the gun turntable, and a
pair of hoists in the rear of the gun. The central hoists are contained
in a cylindrical casing, 6 feet 6 inches diameter, extending from the
under side of the working chamber to within 2 feet of the ship's bottom.
2 K
514 RISE AND PROGRESS OF RIFLED NAVAL ARTILLERY
This casing revolves with the mounting, and contains a pair of shot
hoists and a pair of powder hoists.
The bottom of the casing is fitted with rails, on which a pair of
bogies carrying shot trays, can run. These are arranged to be locked
to the ship while being charged, and to the hoist casing while dis-
charging into the shot cage.
The shot on arriving at the working chamber are automatically
rolled out into an inclined trough leading to the gun hoists.
A new departure in this design is the loading of the guns at an
elevation of only 4^°. I believe there is an impression that time can
be saved if the guns can be loaded at any angle without coming to a
fixed position. If, however, the gun has to be washed out after each
round, it would have to be placed at about 4° or 5° of elevation, to
allow the water to run out of the chamber. This, and the provision
of something to catch the water, seems to make it desirable to place
the gun on a stop at this position. On the comparatively rare
occasions when more elevation is required, the stop can be easily
removed.
In the design of mounting shown on Plate XXXIII. (p. 520) for
the Japanese ship Mihasa, the outer casing of the hoists is built water-
tight at the middle, lower, and platform decks, each of which will
therefore be strengthened and bound together. The interior and
bottom portions of the hoist are practically the same, and revolve
within the fixed casing. This design of hoist is also adopted for the
Italian ships Regina Margherita and B. Brin.
On Plate XXXII. (p. 520) is shown a chain rammer. About the
use of these chain rammers there is some difference of opinion among
authorities. In a fixed loading arrangement it is quite possible that
a chain rammer might be advantageously used to save room by
reducing the lengths of the oval barbette ; but, as now applied to all-
round loading arrangements, its u.se appears to me to be doubtful.
In the hydraulic rammer we have a machine placed in a line with the
work to be done, and making a stroke in a straight line, so that
nothing more direct-acting could be devised for the purpose ; and I
confess, under the existing circumstances, I fail to see the advantage
of a mechanism which first converts rectilinear motion into circular
motion, and then converts the circular back again to rectihnear.
Moreover, the hydraulic rammer shown on Plate XXXIII. (p. 520),
in a similar position to that occupied by the chain rammer, is made
up of far fewer pieces, and weighs only one-fourth as much as
the chain rammer of similar power.
RISE AND PROGRESS OF RIFLED NAVAL ARTILLERY 515
I have gone rather fully into the central hoist question, because,
when the all-round loading of guns is to be arranged for, the difficulty
at once presents itself of how to get the projectiles and charges
transferred from the ship to the gun turntable, not only in every
position the latter might take up in relation to the ship, but also
while the gun turntable is moving or is liable to be moved at any
moment.
You will notice that in the Be Umherto this difficulty is met by
using a turntable top to the central hoist, and sliding the projectiles
radially outwards into the gun hoists. In the Canopus, by the
employment of overhead travelling cranes placed above the central
hoists ; in the Albion, by surrounding the bottom of the ammunition
trunk by a revolving platform running on rails on the ship's bottom,
and capable of being locked either to the ship or the hoist trunk ; in
the ShiUshima, by using a double overhead rail, half of which moves
with the hoist and the ..other half a fixture to the ship; and in the
Formidable, by having two shot carriages running on rails carried at
the bottom of the trunk of the hoist.
There are objections to each of these systems, and perhaps they
all make too much of what is, after all, a very simple matter. If the
hoist cage carrying the projectile can be made to vary its position
according to the training of the gun during its ascent from the shell-
room all difficulty will be overcome. I wish, therefore, to draw your
attention to the design in Plate XXXIV. (p. 520), due to my friend
Mr Murray, which accomplishes, I think, exceedingly satisfactorily
this end.
In this design there is a small fixed central trunk, 2 feet 9 inches
outside diameter, which forms a strong pillar guide for a pair of
ammunition cages. The back of each ammunition cage is curved to
fit partially round this pillar. There is also an outer trunk of about
6 feet 6 inches diameter built to the ship. This outer casing is
smooth inside, and the ammunition cages are prevented from falling
away from the central pillar because their outer edges are in contact
with the outer casing. It will thus be seen, as the cages move up
and down, they could be slewed round the pillar or travel up it in a
spiral line.
In order to make the cages follow the desired path, the central
pillar is clasped at convenient intervals by several rings simk in the
thickness of its plate, so as not to prevent a free passage. Each of
these rings carries an arm on either side, and a pair of ropes kept
taut by springs are stretched from top to bottom and pass through
516 RISE AND PROGRESS OF RIFLED NAVAL ARTILLERY
an eye at the end of each arm. The stretched ropes are secured at
the bottom to the ship, and at the top to the under side of the gun
turntable. As the turntable is trained to right or left, the ropes take
up a spiral position, and, by means of the arms upon the rings round
the pillar, guides for the cage, which are also carried on the arms, are
likewise compelled to take a spiral form.
With this arrangement a most satisfactory method of charging
the cages in the shell-room can be employed. This wiU be seen on
Plate XXXV. (p. 520).
Troughs are provided in line with the position to which the cages
always descend. Overhead hydraulic and hand-worked runners
command these troughs and shell bins.
Hydraulic rammers are placed in line with the troughs for pushing
the shot which has been set in the troughs into the cage. On the
other side of the trunk the magazine handing-rooms are arranged ; so
that, while the shot are being placed in the cage on one side, the
powder can be placed from the other side.
With this arrangement any shot from the bins can be picked up
and put into either cage, and the whole arrangement is simpler and
more complete than has yet been fitted on any ship in this respect.
In the working chamber the cages always arrive by the side of
the trough in line with the gun hoists, into which the shot is
automatically rolled. A hand or hydraulic rammer can be used to
slide the shot down into the gun hoists, and the powder is transferred
by hand in quarter-charges.
This is, of course, assuming that the current opinion is in
favour of a transference of the ammunition in the working
chamber.
In Plates XXXVI. and XXXVII. (p. 520), is shown a design in
which the central hoist is not stopped at the working chamber, but
is carried on to the rear of the guns. There are, I believe, those who
fear that this arrangement would give too direct a path for fire to the
magazine in case of any accident at the gun. The difference between
this and any other system is so very slight that, with proper precau-
tionary measures, I do not think there need be any fear. It seems to
me that, if it is decided that the shell ought to be at the bottom
of the ship, the most perfect arrangement is that in which any shot
can be conveyed from any shot bin to either gun in whatever
position the gun may be, entirely by mechanical means, and without
having to handle it.
In concluding this part of my subject, I venture to draw
RISE AND PROGRESS OF RIFLED NAVAL ARTILLERY 517
attention to one point. In an earlier part of my paper I have
alluded to the rough and ready appliances with which the navy
of the past achieved such great things, and I myself have heard
distinguished naval officers urge that mechanical contrivances
which could not at sea be repaired by the crew were out of place
on board men-of-war.
All this is now changed. A battleship carries well on to a
hundred machines of the most varied, and some of the most
compHcated character. I have elsewhere expressed my admiration
of the ability and zeal with which naval officers of the present day
have mastered, and the skill with which they use their varied
machinery, but I think there is some tendency to push automatic
arrangements too far. The blue-jacket will lose much, if he is
■degraded into a mere machine, and, in regard to the heavy mountings
I have been describing, our aim should be to obtain efficiency with
as great simphcity and as few complications as possible.
The number of explosives which have been used or proposed
as bursting charges for shell, is very large, but in this short sketch
I shall confine my attention to three — gunpowder, guncotton, and
melinite, including under this latter head the form known to our
service as lyddite. Mr Vavasseur and the writer were placed in
a position to communicate to the authorities of this country full
details concerning this last explosive, and the whole of the first
experiments with it were made either by, or under the superin-
tendence of, my firm. Guncotton and lyddite are not only
capable of detonation, but also possess a potential energy very much
higher than that of gunpowder.
Fired against unarmoured structures, shell charged with gun-
powder do not generally explode until they are some short distance
within the side of the vessel, but with guncotton and lyddite two
alternatives have to be considered. The shell may either be fired
with a fuse and detonator so arranged that the shell will burst
immediately on impact, or it may be so arranged as to give rise to
a slight delay, or hang fire.
In the first alternative the shell will burst instantaneously on
impact, a result impossible to obtain with gunpowder ; and in such
cases a hole of very large dimensions, and impossible to plug, will
be made in the side of the ship, while the innumerable small frag-
ments to which the shell is reduced sweep the deck in the wake of
-the shell.
In the second alternative the shell will probably burst inside.
518 RISE AND PROGRESS OF RIFLED NAVAL ARTILLERY
making only a small hole in the side of the vessel ; but the full effect
of the explosion, and the destruction to the crew from the fragments
of the shell would undoubtedly be serious, and the cone of dispersion
of the fragments much larger, from the explosion taking place inside
the vessel.
Shell charged with gunpowder fired against unarmoured structures
possess, however, one great advantage. The shell will probably
burst from 2 to 4 feet inside the vessel, and, although the dispersion
of the fragments is not nearly so great as with high explosives, the
large fragments into which the shell parts, are capable of doing
much more serious damage to any portion of the ship's structure-
with which they may come in contact.
If fired at armoured structures, the results will greatly depend
upon the thickness and resistance of the plates, and on the size
and energy of the attacking projectile.
Generally, it may be stated that armour is a most effective
protection against high explosives, the shell in the large majority
of circumstances bursting comparatively harmlessly against the
armour. Even if unfused, but with detonator, and possessing
sufficient energy to penetrate the plate, the shell will burst in passing
through, but the dispersion of the fragments is not very great.
If fired without fuse or detonator, wet guncotton will not
explode, but mehnite or lyddite probably will, the result to a great
extent depending on the thickness of the armour.
From the numerous experiments we have made, either ourselves
in this country or elsewhere, I draw the following conclusions : —
(1) To attack unarmoured structures, I have no doubt that shell
charged with high explosives are a most formidable weapon. The
large quantity of explosive that can be carried, and the power of
immediately detonating the shell, permit the vessel to be attacked,,
either by making large holes at or near the water-line, or if the shell
should burst inboard, the effect of the explosion and the destruction
to everything in the wake of the shell would be very serious.
(2) But with high explosives the shells are reduced to very small
fragments, and even very thin steel plates resist penetration. Hence
the importance of traverses; and, supposing a first-class cruiser to
engage two smaller cruisers firing high explosives, one on each broad-
side, a longitudinal traverse of very moderate thickness would be a
protection, the importance of which could hardly be overrated.
(3) Having regard to the size of the holes made by high explosives,
in unarmoured structures, I regard it of great importance that, where-
RISE AND PROGRESS OF RIFLED NAVAL ARTILLERY 519
€ver possible, the water-line should be protected from stem to stern
with a belt of armour, and that side armour should be provided where
guns are carried on the main deck. On the upper deck effective
shields, and as thick as can be conveniently carried, should Ije
attached to the mountings.
(4) Where an attack is made against thin armour, shell charged
with gunpowder are more effective than high explosive shell, as,
dependent on circumstances, the former can be got to pass through
thin armour and burst inside. I doubt if shell charged with any
explosives can be got to pass through thick armour without bursting.
(5) There is one serious objection to certain high explosives, as
bursting charges, which is not shared by wet guncotton, and that is,
the liability to detonate if struck by another projectile, or even by a
large fragment. Wet guncotton is quite safe in this respect, and
yet, if fired, for example, by a fulminate, it detonates even more
rapidly than in the dry state. This property has led certain govern-
ments to adopt it as the high explosives for use on board ship.
In concluding this paper I desire to defend our Elswick practice,
which I have sometimes heard attacked, of mounting as many guns
on the broadside as can be conveniently carried. Personally, I share
strongly the opinion which a distinguished admiral once expressed
to me: that, supposing a fight between two cruisers equally ably
commanded, the victory would remain with the ship that got in first
her second broadside, and the victory would be more assured if her
broadside were the more powerful. It must also be remembered that
with our modern weapons, allowance must be made for a gun, or two,
being disabled without altogether crippling the broadside. For these
reasons I prefer to carry as many guns as possible, even if the
number of rounds carried per gun be reduced.
I feel that I ought perhaps to apologise for the length of this
paper ; I may, however, make the excuse which I have before heard,
that I have been so much pressed with other work, I had not time to
make it short. I must, however, express my obligations to my
friends and fellow-workers, Mr Murray and Mr Brankston, who, with
this paper, as in many other ways, have given me most valuable
assistance.
PLATE I.
PLATE II.
PLATE I!
COMPARISON BETWEEN A 32 Pr OLD GUN AND A 6 INCH NEW GUN
—
_ —
— -
— -(•
_,_.
-''
'
,--•
''
^'
'
-
sleV
5>!-'J
,-
^-
¥ i
/
fe«on,
/-"^
S
/
f\'
.-"
-"
<
\
/
:¥
i '
<
/
\
\
e
//
\
,' /
N
1
\
1
^
^>i
'
i 1
^
1 1
^
"
^
~~~~
■
n
\
~
_^
!
V
-r TTMVl
*0T
i-^
*- ? 3
^^^f^^^^- 1
PLATE IV.
o
o
o
LU
oc
H
O
>
ui
H
Z
UJ
O
z
o
o
6
ft.
(0
W
z
o
z
s
rs
PLATE V.
o
o
o
Ul
DC
I-
o
>
Q.
UJ
oc
H
z
UJ
o
o
ul
&
0.
to
PLATES VI. AND VII.
47M/M3 PR QUICK FIRING GUN ON ELSWICK PEDESTAL RECOIL MOUNTING.
ELEVATION
40CalibrRGun
47M/M 3 PR QUICK FIRING GUN ON ELSWICK PEDESTAL RECOIL MOUNTING
PLAN
40 Calibre Gun
1^3
PLATE VIII.
,0!
z
p
z
o
-I
o
u
z
o
X
o
z
/
ro
. /
J '
>.
\
' ci '
s
/v /
/ "v/ ■
3
•■ o
O H
u. in
o 9
O -1
Z OQ
; 5i
»- u
llJ -1
a: u.
I- <£
o z
UJ o
<
PLATE XXXVII.
"•I-
O 52
z o
H =
z ui
ii
I"
< CD
Z U
UJ
2 i
UI =
o ^
< X
OF THE A
rJiVERSlTY I
OF /
XV.
SOME MODEEN EXPLOSIVES
{Pcij)e,r read at the Royal Listitution, 1900.)
Neaely thirty years ago, in the Eoyal Institution, I had the honour
of describing the great advances which had then recently been made
both in our knowledge of the phenomena which attend the decom-
position of gunpowder, and in its practical application to the
purposes of artillery.
I described the uncertainty which up to that date had existed as
to the tension developed by its explosion, the estimates varying
enormously from the 101,000 atmospheres (about 662 tons on the
square inch) of Count Eumford to the 1000 atmospheres (6-6 tons
per square inch) of Eobins, or, taking more modern estimates, from
the 24,000 atmospheres (158 tons per square inch) of Piobert and
Cavalli to the 4300 atmospheres (about 29 tons per square inch) of
Bunsen and Schischkoff.
These uncertainties were, I think I may say, set at rest by
certain experiments carried out both in guns and close vessels at
Elswick, by the labours of the Explosive Committee appointed by
the War Office, and by researches conducted by Sir F. Abel and
myself. These researches were conducted on a large scale, with the
view of reproducing as nearly as possible in experiment the conditions
that exist in the bore of the gun. You may judge of the magnitude
of the experiments, when I tell you that I have fired and completely
retained in one of my cylinders a charge of no less than 28 lbs. of
ordinary powder.
The result of the discussion of the whole series of experiments
led to the following conclusions : —
1. That the tension of the products of combustion at the moment
of explosion when the powder practically filled the space
in which it is fired — that is, when the density is about
unity — is a little over 40 tons on the square inch, or about
6400 atmospheres.
521 2 L
522 SOME MODERN EXPLOSIVES
2. Although changes in the chemical composition of powder, and
even changes in the mode of ignition, cause a very con-
siderable change in the metamorphosis experienced in
explosion, as evidenced by the proportions of the pro-
ducts, the quantity of heat generated, and the quantity of
permanent gases produced, being materially altered, it is
somewhat remarkable that the tension of the products in
relation to the gravimetric density is not nearly so much
affected as might be expected from the considerable altera-
tion in the above factors.
3. The work that gunpowder is capable of performing in ex-
panding in the bore of a gun was determined both by
actual measurement and by calculation, and the results
were found to accord very closely.
4. The total potential energy of exploded gunpowder supposed to
be fired at the density of unity was found to be about
3.32,000 grm.-units per grm., or 486 foot-tons per pound
of powder.
I must confess that when I gave the lecture I have referred to,
seeing the many centuries during which gunpowder had held its
own as practically the sole propelling agent for artillery purposes,
seeing also that gunpowder differs in certain important points from
the explosives to which I shall presently call your attention, I had
serious doubts as to whether it would be possible so far to modify
these latter as to permit of their being used in large charges and
under the varied conditions required in the Naval and Military
Services.
Gunpowder is not, like guncotton, cordite, nitro-glycerine, lyddite,
and other similar explosives, a definite chemical combination in a
state of unstable equilibrium, but is merely an intimate mixture of
nitre, sulphur, and charcoal, in proportions which can be varied to
a very considerable extent without striking differences in results.
These constituents do not, during the manufacture of the powder,
suffer any chemical change, and being a mixture it cannot be said
under any condition truly to detonate. It deflagrates or burns with
great rapidity, varying very largely with the pressure and other
circumstances under which the explosion is taking place — a train
like that to which I set fire taking, as you see, an appreciable time to
burn, while in the bore of the gun a similar length of charge would
be consumed in less that the hundredth part of a second.
You will further have observed the heavy cloud of smoke which
PLATE I.
PLATE 11.
SOME MODERN EXPLOSIVES 523
has attended the deflagration you have seen. Nearly six-tenths of
the weight of the powder after explosion remains as a finely-divided
solid, giving rise to the so-called smoke familiar to many of you,
and of which a good illustration is shown in this instantaneous
photograph, Plate I. By way of comparison, I hurn similar lengths
of guncotton in the form (1) of cotton, (2) of strand, (3) of rope, and
you will observe the different rates at which these varied forms of
the same material are consumed, the rate depending in this case
upon the greater aggregation and higher density, consequently higher
pressure of the successive samples.
Although the names of cordite and ballistite are prol3ably
familiar to all of you, the appearance may not be so familiar, and I
have here on the table samples of the somewhat protean forms
which these explosives, or explosives of the same nature, are made to
assum.e.
Here, for instance, are forms of cordite, the explosive of the
service, for which we are indebted to the labours of Sir F. Abel and
Prof. Dewar. This, which is in the form of fine threads, is used in
small arms, and here are successive sizes, adapted to successive
larger calibres, until we reach this size which is that employed for
the charge of the 12-inch 50-ton guns.
A couple of the smaller cords I burn, both for purposes of com-
parison and to draw your attention to the entire absence of smoke.
The smoke of the gvmpowder you see still floating near the
ceiling, but little or no trace of smoke can be seen from such
explosives as guncotton, cordite, or ballistite, their products of
combustion being entii'ely gaseous. See photograph, Plate II.
You will have observed that in the combustion which you have
just seen there is no smoke, but I must explain, and I shall shortly
show you, that this combustion is not quite the same as that which
takes place, for instance, in the chamber of a gun. Here the
carbonic oxide and hydrogen, which are products of explosion, burn
in the air, giving rise, with the aid of a little free carbon, to the
bright flame you see, and somewhat increasing the rate of com-
bustion. In a gun, however, owing chiefly to pressure, the cordite is
consumed in a very small portion of a second.
In brder krTltvistrate the effect of pressure upon the rate of
combustion, I venture to show you a very beautiful experiment
devised by Sir F. Abel. It has been shown in this room before, but
it will bear repetition.
In this globe there is a length of cordite. I pass a current
524 SOME MODERN EXPLOSIVES
through the platinum wire on which it is resting, and you see the
cordite burns. I now exhaust the air and repeat the experiment.
The wire is red-hot, but the cordite will not burn. That the
failure to burn is not due to the absence of oxygen, is shown by
plunging lighted cordite into a jar of carbonic acid, where, although
a match is instantly put out, the cordite continues to burn — but
observe the difference. There is no longer any bright flame, although
the cordite is being consumed at about the same rate as when burned
in air ; and when a sufficient quantity of the COo is displaced, I can
make the inflammable gases ignite and burn at the mouth of the jar.
Another illustration is- also instructive. I have here a stick of
cordite wrapped round with filter paper ; I dip it in water and light
the end ; you may note that at first you see the bright flame. But,
as the combustion retreats under the wet filter paper, there appears a
space between the flame and the cordite, the flame finally dis-
appears, hot gases with sparks of carbon alone showing.
One other pretty experiment I show. I have here a stick of
cordite, which I light. When fairly lighted, I plunge it in this
beaker of water. The experiment does not always succeed at the
first attempt, but you now see the cordite burning under the water
much as it did in the jar of carbonic acid. The red fumes you
observe are due to the formation of nitric peroxide caused by the
decomposition of the water by the heat.
I have on the table samples of certain other smokeless explosives
of the same class. Here is a ballistite used in Italy. Here is some
Norwegian ballistite. Here again is ballistite in the tubular form,
and in these bottles it is seen in the form of cubes. Here is some
gelatinised guncotton in the tubular form, and here are some interest-
ing specimens with which I have experimented, and which up to
a certain pressure gave good results, but which exhibited some
tendency to violence when that pressure was exceeded. Here also
are some samples of the Trench B. K powder, consisting of nitro-
cellulose partially gelatinised and mixed with tannin, and with
barium and potassium nitrates. Lastly, I show you here a sample
of picric acid, a substance which has been used for many years as a
colouring material, but which will be of interest to you, because it
is used as the explosive of lyddite shell, concerning which I shall
presently have more to say; it differs from all the other explosives
in being, in the crystalline form, exceedingly difficult to light. I
fuse, however, in this porcelain crucible a small quantity. I pour
a little on a slab, and on dropping a fragment into a red-hot test-tube
SOME MODERN EXPLOSIVES
525
you see with how much violence the fragment explodes. I also burn
a small quantity, and you will observe that, unlike guncotton, cordite,
and ballistite, it is not free from smoke, the smoke in this case being
simply carbonaceous matter. You will observe also how much more
slowly it burns.
The composition of these various explosives (although in the case f>f
both cordite and ballistite I have experimented with samples differing
widely in the proportion of their ingredients), may be thus stated : —
The guncotton I employed was of Waltham-Abbey manufacture,
and when dried consisted of 4-4 per cent, of soluble cotton and 95 '6
per cent, of insoluble — as used, it contained 2-25 per cent, of moisture.
The service cordite consists of 37 per cent, trinitro-cellulose with
a small proportion of soluble guncotton, 58 per cent, of nitro-
glycerine, and 5 per cent, of the hydro-carbon vaseline.
The ballistite I principally used was composed of 50 per cent,
dinitro-cellulose (collodion cotton) and 50 per cent, of nitro-glycerine.
The whole of the cellulose was soluble in ether alcohol, and the
ballistite was coated with graphite.
The French B. K powder consisted of nitro-cellulose partly
gelatinised, and mixed with tannin, with barium and potassium
nitrates. The transformation experienced by some of these
explosives is given in Table 1, while the pressures in relation to the
gravimetric densities of some of the more important are shown in
Fig. 1.
Table 1.
Constituents.
Cordite.
Ballistite.
B.N.
Lyddite.
CO., ....
CO" .
H . . . .
N . . . .
HoO ....
CH, . . . .
Vols.
20-5
23-3
16-5
14-6
23-6
1-5
Vols.
29-1
21-4
15-0
10-1
24-4
trace
Vols.
21-1
24-2
16-4
12-6
25-0
0-6
Vols.
12-8
49-7
13-8
19-6
3-8
0-3
Quantity of gas in
c.c. per gramme
Units of heat
890-5
1272
807
1365
822
1003
960-4
856-3
The decomposition experienced by these high explosives on being
fired is of much greater simplicity than that experienced by the old
powders, and is moreover not subject to the considerable fluctuations
in the ultimate products exhibited by them.
The products of explosion of guncotton. cordite, ballistite, etc..
526
SOME MODERN EXPLOSIVES
are at the temperature of explosion entirely gaseous, consistino' of
carbonic anhydride, carbonic oxide, hydrogen, nitrogen, and aqueous
vapour, with generally a small quantity of marsh-gas.
The water collected, after the explosion-vessel was opened,
always smelt, occasionally very strongly, of ammonia, and an appreci-
able amount was determined in the water.
In examining the gaseous products of the explosion of various
samples of gunpowder, it was noted that as the pressure under which
the explosion took place increased, the quantity of carbonic anhydride
also increased, while that of carbonic oxide decreased. The same
Fi^ 1.
PRESSURES OBSERVED IN CLOSED VESSELS WITH
VARIOUS EXPLOSIVES.
•15 -20 25 -SO -35 -40 -AS SO 55 60 -65
DENSITY OF PRODUCTS OF EXPLOSION
75 80 85
peculiarity is exhibited by all the explosives with which I have
experimented. I show in Table 2, p. 527, the result of a very com-
plete series of a sample of guncotton fired under varying pressures,
and it will be noted that the volumes of carbonic oxide and carbonic
anhydride are, between the highest and lowest pressures, nearly
exactly reversed.
There are slight changes as regards the other products, but they
do not compare in importance with that to which I have referred.
But before drawing your attention to other points of interest, it
is desirable to give you an idea of the advances in ballistics which
have been made, both by improvements in the manufacture of tlie
old powders and by the introduction of the new.
VELOCITY IN FEET PER SECOND,
VELOCITY IN FEET PER SECOND.
o o o
SOME MODERN EXPLOSIVES
527
On Fig. II. are placed the results as regards velocity of nine
explosives, commencing with the E. L. Gg powder, which was in use
in the latter part of the fifties, and terminating with the cordite of
the present day.
Table 2.
Under pressure of E
xplosion, tons per square inch.
Constituents.
2 tons.
8 tons.
12 tons.
18 tons.
20 tons.
45 tons.
50 tons.
Vols.
CO, .
21-44
25-06
26-27
27-21
26-75
28-13
29-27
CO . . .
29-66
26-31
25-08
25-24
24-53
23-19
22-31
H . . . .
1.5-92
15-33
16-03
14-56
14-77
14-14
13-56
N . . . .
13-63
13-80
13-22
13-13
13-43
12-99
13-07
H2O . . .
19-09
19-09
19-09
19-09
19-09
19-09
19-09
CH. . . .
-26
-41
-31
-77
1-47
2-46
2-70
The experiments I am now referring to were made in a gun of
100 calibres in length, and were so arranged that in a single round
the velocities could be measured at 16-points of the bore. The
chronoscope with which these velocities were taken has been already
described, and I will now only say that it is capable of registering
time to the millionth of a second with a probable error of between
two and three millionths. One curious fact connected with the mode
of registration I may mention. In the early experiments with the
old powders, where the velocities did not exceed 1500 or 1600 feet
per second, the arrangement for causing the projectile to record the
time of its passing any particular point was effected by the shot
knocking down a small steel knife or trigger which projected slightly
into the bore, but when the much higher velocities, with which I
subsequently experimented, were employed, this plan was found to
be unsatisfactory, the steel trigger, instead of being immediately
knocked down by the shot, frequently preferred, instead, to cut a
groove in the shot, sometimes nearly its whole length, before it
acted. Hence another arrangement for cutting the primary wires
had to be adopted.
The diagram I am now showing you is, however, both interesting
and instructive. The intention, among other points, was to ascer-
tain, for various calibres in length in a 6 -inch gun, the velocities
and energies that could be obtained, the maximum pressures, whether
mean or wave, not exceeding about 20 tons on the square inch. The
horizontal line or axis of abscissse represents the travel of the shot
528 SOME MODERN EXPLOSIVES
in feet, the ordinates or perpendiculars from this line to the curve
represents the velocity at that point.
The lowest curve on the diagram gives, under the conditions I
have mentioned, the velocities attainable with the powder which
was used when rifled guns were first introduced into the service, and
you will note that with this powder the velocity attained with 100
calibres was only 1705 feet per second, while with 40 calibres it
was 1533 feet per second. Next on the diagram comes pebble-powder,
with a velocity of 2190 feet per second; next comes brown prismatic,
with a velocity of 2529 feet per second.
The next powder is one of considerable interest, and one which
might have risen to importance had it not been superseded by
explosives of a very different nature. It is called amide powder,
and in it ammonium nitrate is substituted for a large portion (about
half) of the potassium nitrate, and there is also an absence of sulphur.
You will observe the velocity in the 100-calibre gun is very good,
2566 feet per second. The pressure also was low, and free from wave
action. It is naturally not smokeless, but the smoke is much less
dense, and disperses much more rapidly than does the smoke of
ordinary powder. Its great advantage, however, was, that it eroded
steel very much less than any other powder with which I experi-
mented, while its great disadvantage was due to the deliquescent
properties of ammonium nitrate, necessitating the keeping of the
cartridges in air-tight cases.
Next on the diagram comes B. N. or Blanche Nouvelle powder,
an explosive which, while free from wave action, is remarkable, as
you will note if you follow the curve, in developing a much higher
velocity than the other powders in the first few feet of motion, and
less in the later stages of expansion.
Thus, if you compare this curve with the highest curve on the
diagram, that of the four-tenths cordite, you will note that the B. N.
curve for the first eight feet of motion is the higher, and that at
about eight feet the curves cross, the B. N. giving a final velocity of
2786 feet per second, or 500 feet below the cordite curve.
Then follows ballistite, which, with much lower initial pressure,
gives a velocity of 2806 feet per second, or somewhat higher than that
of B. N. Then follow three different sizes of cordite, the highest of
which gives a muzzle velocity of 3284 feet per second, or a velocity
nearly double that of the early E. L. Go.
In the somewhat formidable-looking table (Table 3) I have
placed on the wall, are exhibited the velocities and energies
SOME MODERN EXPLOSIVES
529
realised in a 6-inch gun with the various explosives I have named,
and the table, in addition, shows the velocities and energies in guns
of the same calibre but of 40, 50, and 75 calibres in length, as well
as in that of 100 calibres.
Table 3.—Q-inch gun, 100 calibres long. Velocities and energies realised with
high explosives. Weight of projectile, 100 Ihs.
Nature and weight of
explosive.
Length of bore,
40 calibres.
Length of bore,
50 calibres.
Length of bore,
75 calibres.
Length of bore,
100 calibres.
>>
1
>
1
iS
1
S
1
>
>>
Cordite, '4 in. (27*5 lbs.)
Cordite, 0-35 in. (22 lbs.)
Cordite, 0-3 in. (20 lbs.)
Ballistite, 0-3 in. cubs.
(20 lbs.) .
French B. N. (25 lbs.).
Amide prism (32 lbs.) .
Brown prism (50 lbs.) .
Pebble-powder (36 lbs.)
R.L. G.2(23lbs.).
F. S.
2794
2444
2495
2416
2422
2225
2145
1885
1533
F. T.
5413
4142
4316
4047
4068
3433
3190
2464
1630
F. S.
2940
2583
2632
2537
2530
2331
2257
1980
1592
F. T.
5994
4626
4804
4463
4438
3768
3532
2718
1757
F. S.
3166
2798
2821
2713
2700
2486
2435
2110
1668
F. T.
6950
5429
5518
5104
5055
4285
4111
3087
1929
F. S.
3284
2915
2914
2806
2786
2566
2529
2190
1705
F. T.
7478
5892
5888
5460
5382
4566
4485
3326
2016
If you compare the results shown in the highest and lowest lines
of this table, that is, the results given by the highest and lowest
curves on the diagram, you will see that the velocity of the former
is nearly twice as great as that of the latter, while its energy and
capacity for penetration is nearly four times as groat.
I need hardly remind most of you that in artillery matters it is
the energy developed, not the velocity alone, that is of vital import-
ance. I venture to insist upon this point, because so many of those
who desire to instruct the authorities, write as if velocity were the
only point to be considered. In a given gun with a given charge,
if the weight of the shot, within reasonable limits, be made to vary,
the ballistic advantage is greatly on the side of the heavier shot, and
for three principal reasons : —
1. More energy is obtained from the explosive.
2. Owing to the lower velocity, the resistance of the air is greatly
reduced.
3. The heavier shot has greater capacity for overcoming the
reduced resistance.
You will observe that on this velocity diagram, upon which I
have kept you so long a time, is shown, not only the travel of the
shot in feet, but the position of the plugs which gave the velocities.
2 M
530 SOME MODERN EXPLOSIVES
Further, on the higher and lower curves, the observed velocities are
shown where it is possible to do so. Near the origin of motion the
points are so close that it is not possible to insert them without
confusing the diagram.
At the risk of fatiguing you, I show, in Fig. III., curves showing
the pressure existing in the bore at all points, these pressures being
deduced from the curves of velocity.
You will note the point to which I drew your attention, with
regard to the powder called B. N. You will remember that in the
early stages of motion it gave velocity to the shot, much more
rapidly than did the other powders. You see the effect in the
pressure curves, the maximum being considerably higher than any
of the other pressures, while the pressure towards the muzzle is, on
the other hand, considerably below the average.
I fear you may think I have kept you unnecessarily long with
these somewhat dry details, but I have had reasons for so doing.
In the first place I desire to demonstrate to you the enormous
advances which have been made in artillery by the introduction of
the new explosives, and which we in a great measure owe to the
distinguished chemists and physicists who have occupied themselves
with these important questions.
Secondly, I desire to show you that the explosive which has
been adopted by this country, and which we chiefly owe to the
labours of Sir F. Abel and Prof. Dewar, is in ballistic effect inferior
to none of its competitors. I might go further, and say that it is
decidedly superior.
Lastly, at a time when the efficiency of all our arms, and
especially our artillery, is a question which has been deeply agitat-
ing the country, I may do some good by pointing out that the
authorities are well aware that any practicable velocity or energy
they may desire for their guns is at their disposal.
They have such guns, I mean guns with high velocity and high
energy — whether they have enough of them, and whether they are
always in the right place, is another matter, for which perhaps the
military authorities are not altogether responsible. But velocity and
energy is not the only thing that is required under all circumstances
in war, and I ask you to believe that if the War Office authorities
have, for their field guns, fixed on a velocity very much below what
is possible, they have had sound and sufficient reasons for so doing.
My firm and I, individually, have had much to do with the
introduction of the larger high-velocity and quick-firing guns into
PRESSURE IN TONS PER SQ INCH
PRESSURE IN TONS PER SQ INCH
SOME MODERN EXPLOSIVES 531
our own and other services ; but as an old artillery officer, in no way
responsible for our field guns, I may perhaps be allowed to say that,
whether as regards maUriel or personnel, our field artillery is
inferior to none anywhere, and I venture to add that in the present
war it appears to have been handled in a way worthy of the reputa-
tion of the corps.
I fear the causes of some of our military failures at the commence-
ment of the war must be looked for in other directions, and the present
unfortunate war will turn out to be a blessing in disguise, if it
should awaken the empire to the necessity of correcting serious
defects in our organisation, possibly the natural result of our Con-
stitution, and in that case the invaluable lives that have been lost
will not have been sacrificed in vain.
I now pass to points which have to be considered when weighing
the comparative merits of explosives for their intended ends.
You will easily understand that between explosives which are
intended to be used for propelling purposes, and those which are
intended to be used, say for bursting shell, a wide difference may
exist.
In the former case, facility of detonation would be an insuper-
able objection ; in the latter, the more perfect the detonation the
better, certain special cases, to which I have not time to refer,
excepted.
There exists, I think, considerable diversity of opinion as to what
does, and what does not, constitute true detonation. I find many
persons speak of a detonation, when I should merely consider that
a very high pressure had been reached. This guncotton slab on the
table affords me, I think, a fair opportunity of explaining my
meaning. Were I to set fire to it, except for the large volume of
flame and the great amount of heat generated, we in this room would
not suffer ; we should probably experience more inconvenience did I
fire a similar slab of gunpowder, as detached burning portions would
probably be projected to some distance.
But if I fired this same slab with two or three grammes of
fulminate of mercury, a detonation of extreme violence would follow.
The detonation would be capable of blowing a hole in a tolerably
thick iron plate, and would probably put an end to a considerable
proportion of the managers in the front row.
I mentioned to you some time ago the time in which a charge
would be consumed in the chamber of a gun — if a charge of 500 lbs.
of these slabs were effectively detonated, this charge would be con-
532 SOME MODERN EXPLOSIVES
verted into gas in less than the twenty-thousandth part of a
second.
No such result would follow were I to try a similar experiment
with a slab of compressed gunpowder of the same dimensions. I
do not say the experience would be pleasant, but there would be
nothing of the instantaneous violent action which marks the decom-
position of the guncotton.
To give you an idea of the extraordinary violence which accom-
panies detonation, I have fired, for the purpose of this lecture,
with fulminate of mercury, a charge of lyddite in a cast-iron shell,
and those who are sufficiently near, can see for themselves the
result. By far the greater part of the cast-iron shell, weighing
about 10 lbs., is reduced to dust, some of which is so fine that I
assumed it to be deposited carbon until I had tested it with a
magnet. I may add that the indentation of the steel vessel by pieces
of the iron which were not reduced to powder, would appear to
indicate velocities of not less than 1200 feet per second, and this
velocity must have been communicated to the fragments in a space of
less than 2 inches.
For the sake of comparison, I place beside it a cast-iron shell
burst by gunpowder. You will observe the extraordinary difference.
I also have on the table two small steel shells exploded, one by a
perfectly detonated the other by a partially detonated charge. I
may remark that in the accounts of the correspondents from the
seat of war, frequent mention is made of the green smoke of lyddite.
This appearance is probably due to imperfect detonation — to a
mixture, in fact, of the yellow picric with the black smoke ; I do not
say, however, that imperfect detonation is necessarily an evil.
To another experiment I draw your attention.
For certain purposes, I caused to be detonated in the chamber of
a 12-pr. a steel shell charged with lyddite. The detonation was
not perfect, but the base of the shell was projected with great
violence against the breech screw. You may judge of how great
that violence was, when I tell you that the base of the shell took a
complete impression of the recess for the primer, developing great
heat in so doing ; but what was still more remarkable, the central
portion of the base also sheared, passing into the central hole through
which the striker passes. This piece of shell is upon the table, and
open to your inspection.
One other instance, to illustrate the difference between com-
bustion and detonation, I trouble you with. Desiring to ascertain the
SOME MODERN EXPLOSIVES
533
difference, if any, in the products of explosion between combustion
and detonation, I fired a charge of lyddite in such a manner that
detonation did not follow. The lyddite merely deflagrated. But a
similar charge, differently fired, shortly afterwards detonated with
such extreme violence as to destroy the vessel in which it was
exploded. The manner in which the vessel failed I now show you
(Fig. IV.), and I have on the table the internal crusher-gauge which
was used, and which was also totally destroyed.
The condition of this gauge is very remarkable, and the action on
the copper cylinder employed to measure the pressure was one to
Fig. IV.
EXPLOSION VESSEL
Plug Containing
Crusher Gauq^.
which I have no parallel in the many thousand experiments I have
made with these gauges. The gauge itself is fractured in the most
extraordinary way, even in some places to which the gas had no
access, and the copper cylinder, which when compressed usually
assumes a barrel-like form (that is with the central diameter larger
than that at the ends, as shown in the diagram. Fig. V.), in this experi-
ment, and in this only, was bulged close to the piston, as you see.
It would appear as if the blow was so suddenly given that the
laminae of the metal next the piston endeavoured to escape in the
direction of least resistance, that being easier than to overcome the
inertia of the laminae below.
The erosive effect of the new explosives is another point of
534 SOME MODERN EXPLOSIVES
first-rate importance in an artillery point of view. The cordite of
the service is not, if the effect be estimated in relation to the energy
impressed on the projectiles, more erosive than, for example, brown
prismatic, which was itself a very erosive powder ; but as we are
able to obtain, as you have seen, very much higher energies with
cordite than with brown prismatic, the erosion of the former is, for a
given number of rounds, materially higher.
There is, however, one striking difference: by the kindness of
Colonel Bainbridge, the Chief Superintendent of Ordnance Factories,
I am enabled to show you a section of the barrel of a large gun
eroded by 137 rounds of gunpowder. Beside it is a barrel of a
4-7-inch quick-firing gun eroded by 1087 rounds of gunpowder, and
another eroded by 1292 rounds of cordite. You will observe the
difference. In the former case the erosion much resembles a
COPPER CYLINDERS
ploughed field. In the latter, the appearance is more as if the
surface were washed away by the flow of the highly-heated gases.
But take it in what way you please, the heavy erosion of the
guns of the service, if fired with the maximum charges, is a very
serious matter, as with the large guns, accuracy and in a smaller
degree energy, are rapidly lost after a comparatively small number
of rounds have been fired.
Cordite was first produced for use in small arms only, where,
owing to the small charges employed, the question of erosion is
not of the same importance as with large guns ; but its employment,
from the great results obtained with it, was rapidly extended to
artillery, and the attention of my friends. Sir F. Abel and Prof.
Dewar, has for some time been devoted in conjunction with myself to
investigating whether it is not possible materially to reduce this
most objectionable erosion.
With this object I made the following series of experiments.
Energy in Foot Tons
Heat in Units
Gas in C,C
Erosion in Inches.
PressureinTons.
SOME MODERN EXPLOSIVES 535
1 had cordite of the same dimensions prepared with varying
proportions of nitro-glycerine and guncotton. The nitro-glycerine
being successively in the proportions of 60, 50, 40, 30, 20, and 10
per cent., and with each of these cordites I determined the following
points : —
1. The quantity of permanent gases generated.
2. The amount of aqueous vapour formed.
3. The heat generated by the explosion.
4. The erosive effect of the gases.
5. The ballistic energy developed in a gun, and the corresponding
maximum pressure.
6. The capacity of the cordite to resist detonation when fired
with a strong charge of fulminate of mercury.
The results of these experiments were both interesting and
instructive.
To avoid wearying you with a crowd of figures, I have placed on
Fig. VI. the results of the first five series of experiments.
On the axis of abscissae are placed the percentages of nitro-
glycerine, while the ordinates show the quantities of the gases
generated, the amount of heat developed, the erosive effect of this
explosive, the ballistic energy exhibited in a gun, and the maximum
gaseous pressure.
You will note that with the smallest proportion of nitro-glycerine
the volume of permanent gases is a maximum, and that the volume
steadily decreases with the increase of nitro-glycerine. On the other
hand, the heat generated as steadily increases with the nitro-
glycerine, and if we take the product of the quantity of heat and the
quantity of gas as an approximate measure of the potential energy
of the explosive, the higher proportion of nitro-glycerine has an
undoubted advantage ; but in this case, as in the case of every other
explosive with which I have experimented, the potential energies
differ less than might be expected from the changes in transforma-
tion, as the effect of a large quantity of gas is to a great extent
compensated by a great reduction in the quantity of heat generated.
This effect is, of course, easily explained, and was very strikingly
exhibited in the much more complicated transformation experienced
by gunpowders of different compositions, a long series of which were
very fully investigated by Sir F. Abel and myself.
Looking at this diagram, you will have observed that the energy
developed in the gun is very much smaller with the smaller pro-
portions of nitro-glycerine, but if you will look at the corresponding
536 SOME MODERN EXPLOSIVES
maximum -pressure curve you will note that the pressures have
decreased nearly in like proportion. Hence it is probable that the
lower effect is mainly due to a slower combustion of the cordite, and
it follows that this effect may be, to a great extent, remedied by
increasing the rate of combustion by reducing the diameter of the
cordite to correspond with the reduction in the quantity of nitro-
glycerine.
To test this point, I caused to be manufactured a second series of
cordites of the same composition, but with the diameters successively
reduced by '08, as you see with the samples I hold, and this diagram
(Fig. VII.) shows at a glance the result. The energies you see are
roughly practically the same, but if you look at the pressure curve,
you will observe that I have obtained a curve in which, on the whole,
the pressures vary in the contrary direction, that is to say, in this
case the pressures increase as the nitro-glycerine diminishes.
Taking the two series into account, they show that by a proper
arrangement of amount of charge and diameter of cord, it would be
possible to obtain the same ballistics and approximately the same
pressure from any of the samples I have exhibited to you.
But I have to draw your attention to another point. From the
curve showing the quantities of heat, you will note that in passing
from 10 per cent, nitro-glycerine to 60 per cent., the heat generated
has increased by about 60 per cent. But if you examine the curve
indicating the corresponding amount of erosion, you will see that
while the quantity of heat is only greater by about 60 per cent., the
erosion is greater by nearly 500 per cent.
These experiments entirely confirm the conclusion at which I
have previously arrived, viz., that heat is the principal factor in
determining the amount of erosion.
In experimenting with a number of alloys of steel, the greatest
resistance was shown by an alloy of steel with a small proportion of
tungsten, but the difference between the whole of these alloys
amounted only to about 16 per cent.
The whole of these cordites were, as I have mentioned, subjected
to detonation tests. None of them, so far as my experiments went,
exhibited any special tendency in this direction.
I will now endeavour to describe to you a most interesting and
important series of experiments, which I regret to say is still a long
way from completion.
The objects of these experiments were — (1) to ascertain the time
required for the combustion of charges of cordite in which the
ENERGY IN FOOT TONS
— N> to
Pressure in Tons.
PLATE VII.
SOME MODERN EXPLOSIVES 537
cordite was of different thicknesses, varying from 0"05-inch to 0"6 of
an inch ; (2) the rapidity with which the explosives part with their
heat to the vessel in which the charge is confined; and (3) to
ascertain, if possible, by direct measurement, the temperature of
explosion, and to determine the relation between the pressure and
temperature at pressures approximating to those which exist in the
bore of a gun, and which are, of course, greatly above any which
have yet been determined.
As regards the first two objects I have named, I have had no
serious difficulties to contend with ; but as regards the third, I have
so far had no satisfactory results, having been unable to use Sir W.
Eoberts-Austen's beautiful instrument, owing to the temperature at
the moment of explosion being greatly too high — high enough indeed
to melt and volatilise the wires of the thermo-junction.
I am, however, endeavouring to make an arrangement by which
I hope to be able to determine these points when the temperature
is so far reduced that the wires will no longer be fused.
The apparatus I have used for these experiments (see Plate VII.) is
placed on the table. The cylinder in which the explosions were made
is too heavy to transport here, but this photograph will sufficiently
explain the arrangement. The charge I used is a little more than
a kilogramme, and it is fired in this cylinder in the usual manner.
The tension of the gas acting on the piston compresses the spring
and indicates the pressure on the scale here shown. But to obtain
a permanent record, the apparatus I have mentioned is employed.
There is, you see, a drum made to rotate by means of a small
motor. Its rate of rotation is given by a chronometer acting on a
relay, and marking seconds on the drum, while the magnitude of the
pressure is registered by this pencil actuated by the pressure-gauge
I have just described.
To obtain with sufficient accuracy the maximum pressure, and
also the time taken to gasify the explosive, two observations, that is
two explosions, are necessary.
If the piston be left free to move the instant of the commence-
ment of pressure, the outside limit of the time of complete explosion
will be indicated ; but on account of the inertia of the moving parts
the pressure indicated will be in excess of the true pressure, and the
excess will be, more or less, inversely as the time occupied by the
explosion.
If we desire to know the true pressure, it is necessary to compress
the gauge beforehand to a point closely approximating to the expected
538 SOME MODERN EXPLOSIVES
pressure, so that the inertia of the moving parts may be as small as
possible — the arrangement by which this is effected is not shown in
the diagram, but the gauge is retained at the desired pressure by a
wedge-shaped stop, held in its place by the pressure of the spring,
and to the stop a heavy weight is attached — when the pressure is
relieved by the explosion, the weight falls, and leaves the spring free
to act.
I have made a large number of experiments with this instrument,
both with a variety of explosives and with explosives fired under
different conditions. Time will not permit- me to do more than to
show you on the screen three pairs of experiments to illustrate the
effect of exploding cordite of different dimensions, but of precisely
the same composition.
I shall commence with rifle cordite. In this diagram (Fig. VIII.),
the axis of abscissae has the time in seconds marked upon it, while
the ordinates denote the pressures ; and I draw your attention to the
great difference, in the initial stage, between the red and the blue
curves. You will notice that the red curves show a maximum
pressure some 4-| tons higher than that shown by the blue curve ; but
this pressure is not real, it is due to the inertia of the moving parts.
The red and blue curves in a very small fraction of a second come
together, and remain practically together for the rest of their course.
The whole of the charge is consumed in something less than fifteen
thousandths of a second.
In the case of the blue curve, the maximum pressure indicated is
obtained in the way I have described, and is approximately correct —
about 9 tons per square inch. The rapidity with which this
considerable charge parts with its heat by communication to the
explosion-vessel is very striking. In 4 seconds after the explosion the
pressure is reduced to about one-half, and in 12 seconds to about
one-quarter.
I now show you (Fig. IX.) similar curves for cordite 0"35-inch in
diameter or about fifty times the rifle cordite section. Here you see
that the time taken to consume the charge is longer. The effect of
inertia is still very marked although much reduced. The true
maximum pressure is a little over 8-5 tons, but after the first third
of a second the two curves run so close together that they are indis-
tinguishable.
Again you see the pressure is reduced by one-half in 4 seconds, and
in a little more than 12 seconds again halved.
The last pair of curves I shall show you (Fig. X.) was obtained
co-
rn
O
O
z
o
CO
I
i
PRESSURE IN TON
V3 oj -^ oi cn ^
S PE
RSQ
INCH
5 =
io
w
1
1
^
;:^
!
\
1
/
^
!"
y
/
■f>
/I
/
//
//
o
c
?
1 »
o m
o
o
o
r-
2
O
1
a.
1
ro
■t>
1
('
CD
(S
1
1
PRESSURE IN TONS PER SQ INCH
_ -
—
1.
^
c
o:
^
c
(0
o
.
M -
CO -
/
^
1
■t> -
—
f^ —
/
/
" 1
C
X
<
!
1
H
9 S 1
? X
3- 5
1 o
O -n
2 ~
o
2<^
D
CO
ll
^ 8
o
n
z
•^
'
Ol
'
1
5
1
1
1
(O
o
i
i
1
1
s
M
i
1
1
i
I
t2l
aro>
SOME MODERN EXPLOSIVES 539
with cordite 0'6-inch in diameter, or nearly 150 times the section of
the rifle cordite. With this cordite the combustion has been so slow
that the effect of inertia almost disappears; it is reduced to about
half a ton per square inch. The maximum being nearly the same
as in the last set of experiments. The time of combustion indicated
I have called slow, but it is about '06 of a second, and the whole of
the experiments show a most remarkable regularity in their rate of
cooling, the pressures at the same distance of time from the explosion
being in all cases approximately the same, as indeed they ought to
be. The density being the same and the explosive the same, the only
difference being the time in which the decomposition is completed.
It appears to me that, knowing from the experiments I have
described, the volume of gas liberated, its composition, its density, its
pressure, the quantity of heat disengaged by the explosion; and
knowing all these points with very considerable accuracy, we should
be able, from the study of the curves to which I have drawn your
attention, and which can be obtained from different densities of gas,
to throw considerable light upon the kinetic theory of real, not ideal
gases, at temperatures and pressures far removed from those which
have been the subject of such careful and accurate research by many
distinguished physicists.
The question, as I have said, involves some very considerable
difficulties, nevertheless I am not without hope that the experiments
I have been describing may, in some small degree, add to our
knowledge of the kinetic theory of gas.
That wonderful theory faintly shadowed forth almost from the
commencement of philosophic thought, was first distinctly put
forward by Daniel Bernoulli early in the last century. In the latter
half of the century now drawing to a close, the labours of Joule,
Clausius, Clerk Maxwell, Lord Kelvin, and others, have placed the
theory in a position analogous and equal to that held by the
undulatory theory of light.
The kinetic theory has, nowever, for us artillerists a special
charm, because it indicates that the velocity communicated to a
projectile in the bore of a gun is due to the bombardment of that
projectile by myriads of small projectiles moving at enormous speeds,
and parting with the energy they possess, by impact, to the
projectile.
There are few minds which are not more or less affected by the
infinitely great and the infinitely little.
It was said that the telescope, which revealed to us infinite space,
540 SOME MODERN EXPLOSIVES
was balanced by the microscope, which showed us the infinitely small ;
but the labours of the men to whom I have referred, have introduced
us to magnitudes and weights infinitesimally smaller than anything
that the microscope can show us, and to numbers which are infinite
to our finite comprehension.
Let me draw your attention to this diagram (Fig. II.) * showing
the velocity impressed upon the projectile, and let me endeavour to
describe the nature of the forces which acted upon it to give it its
motion. I hold in my hand a cubic centimetre, a cube so small that
I daresay it is hardly visible to those at a distance. Well, if this
cube were filled with the gases produced by the explosion at 0° C.
and atmospheric pressure, there would be something over seven
trillions, that is, seven followed by eighteen cyphers, of molecules.
Large as these numbers are, they occupy but a very small fraction
of the contents of the cubic centimetre, but yet their number is so great
that they would, if placed in line touching one another, go round many
times the circumference of the earth, a pretty fair illustration of
Euclid's definition of a line.
These molecules, however, are not at rest, but are moving, even
at the low temperature I have named, with great velocity, the
molecules of the different gases moving with different velocities
dependent upon their molecular weight. Thus, the hydrogen mole-
cules which have the highest velocity move with about 5500 feet
per second mean velocity, while the slowest, the carbonic anhydride
molecules, have only 1150 feet per second mean velocity, or about
the speed of sound.
But in the particular gun under discussion, when the charge was
exploded there were no less than 20,500 c.c. of gas, 'and each centi-
metre at the density of explosion contained 580 times the quantity of
gas, that is, 580 times the number of molecules that I mentioned.
Hence the total number of molecules in the exploded charge is 8?;
quadrillions, or, let us say, approximately for the total number,
eight followed by twenty-four cyphers.
It is difficult for the mind to appreciate what this immense number
means, but it may convey a good idea if I tell you that if a man were
to count continuously at the rate of three per second, it would take
him 265 billions of years to perform the task of counting them.
So much for the numbers ; now let me tell you of the velocities
with which, at the moment of explosion, the molecules were moving.
Taking first the high-velocity gas, the hydrogen, the molecules of the
* See p. 527.
i
SOME MODERN EXPLOSIVES 541
gas would strike the projectile with a mean velocity of about 12,500
feet per second. You will observe, I say, mean velocity, and you
must note that the molecules move with very variable velocities.
Clerk Maxwell was the first to calculate the probable distribution of
the velocities. A little more than one-half will have the mean
velocity or less, and about 98 per cent, will have 25,000 feet per
second or less. A very few, about one in 100 millions, might reach
the velocity of 50,000 feet per second.
The mean energy of the molecules of different gases at the same
temperature being equal, it is easy from the data I have given to
calculate the mean velocity of the molecules of the slowest moving
gas, carbonic anhydride, which would be about 2600 feet per second.
I have detained you, I fear, rather long over these figures, but I
have done so because I think they throw some light upon the
extraordinary violence that some explosives exhibit when detonated.
Take, for instance, the lyddite shell exploded by detonation I showed
you earlier in the evening. I calculate that that charge was con-
verted into gas in less than the one 60,000th part of a second, and it
is not difficult to conceive the effect that these gases of very high
density suddenly generated, the molecules of which are moving with
the velocities I have indicated, would have upon the fragments of the
shell.
The difference between the explosion of gunpowder fired in a close
vessel and that of guncotton or lyddite when detonated, is very
striking. The former explosion is noiseless, or nearly so. The latter,
even when placed in a bag, gives rise to an exceedingly sharp
metallic ring, as if the vessel were struck a sharp blow with a steel
hammer.
But I must conclude. I began my lecture by recalling some of
the investigations I described in this place a great many years ago. I
fear I must conclude in much the same way as I then did, by
thanking you for the attention with which you have listened to a
somewhat dry subject, and by regretting that the heavy calls made
on my time during the last few months have prevented my making
the lecture more worthy of my subject and of my audience.
INDEX
Abel, Sir Frederick, 83, 85, 88, 99, 108,
146, 328, 329, 331, 332, 334, 337, 34-_',
345, 348, 373, 386, 402, 414, 421, 424,
434, 462, 463, 468, 471, 501, 502, 521,
522, 530, 534, 535 ; his article in Philoso-
phical Transactions, 480.
AcacUmie des Sciences, 112, 232, 234.
Accelerating twist, 87-98.
Airy, Sir G. B. , Astronomer Royal, his
paper in Philosophical Magazine, 110.
Albion, H.M.S., 515.
Aloncle, Colonel, 267.
Amide gunpowder, 372, 373, 386, 389,
390, 397, 400, 423, 424, 432, 433, 464,
481, 526, 528, 529.
Ammonia, picrate of, 397.
Ammonium carbonate, 136, 149, 166.
Ammonium sesquicarbonate, 126.
Ammunition hoists, 510.
Analytical results from examination of
solid and gaseous products (Researches
on Explosives), 130.
Annalen, Poggendorff's, 108, 110.
Annalen der Chemie, 109.
Archiv fiir die OJiziere der Koniglich
Preussischen Artillerie- una Ingenieur-
Corps, 106.
Arithmetical mean, law of, 7.
Armaments of battleships, past and pre-
sent, 366-383.
Armour of ships, 518, 519.
Armour-piercing guns, 373.
Armstrong, Lord, 65, 363, 367, 370, 499,
502.
Armstrong projectiles, 25, 28-30, 33-36,
39-41.
Artillery, rise and progress of rifled
Naval, 499-515.
Artillery practice, application of theory
of probabilities to, 1-22.
Ash, 127-129.
Austrian cannon- and small-arra-powder,
128-130, 134.
Baker, Sir B., 357.
Ballistics, internal, 397-461.
Ballistite, 386, 397, 425, 464, 478, 481,
523-529.
Battleships, past and present, 362 ;
armament of, 366-383 ; their guns in
1850, 499, 500.
B. Brin, gimpowder Italian, 514.
Belgian " brisante " gunpowder, 106,
176, 417, 465, 485.
Benhoio, H.M.S., 378.
Bernoulli, Daniel, 102, 539.
Bernoulli, John, 102.
Berthelot, M., Sur la Force de la Poudre,
111, 112, 157, 158; his article in
Comptes Bendus de VAcadAmie des
Scieiices, 232-234, 238, 239, 241, 255,
314, 315, 319.
Betancourt, M. de, 56, 484.
Blake, H.M.S., 506.
Blanche Nouvelle (B. N.) French gun-
powder, 481, 524, 525, 528, 529.
Bloxam, C. L., Chemistry, Inorganic
and Organic, 100.
Boxer, R.A., General, 30, 146; Treatise
on Artillery, 159.
Boyle, 337.
Brarawell, Sir F., 355.
Brankston, Mr, 519; his anti-friction
gear, 507.
Brisante gunpowder, Belgian, 106, 176,
417, 465, 485.
British Association, 482 ; Mechanical
Science Section of the, 355.
British-service 10-inch gun, 94, 96, 98.
Bunsen and Schischkoff, 63-65, 67, 82,
108, 109, 112, 119. 127, 134-137, 140,
141, 144, 145, 165, 166, 172, 173, 194-
196, 200, 208, 234, 314, 317, 34?, 349,
414, 433, 434, 521 ; their sporting gun-
powder, 128-130.
Ccesar, H.M.S., 359, 365.
Calorimeter, 256, 297-306.
Canopus, H.M.S., 511, 515,
Carbon, 127-129, 134, 138, 139, 147, 166,
247-249, 329, 421.
Carbonate, potassium, 111,125, 135, 139,
141, 143, 147, 148, 166, 244,250,251,
332, 407.
Carbon dioxide, 238.
Carbon monoxide, 238.
Carbonic anhydride, 104, 111, 119, 120,
134, 138, 139, 142, 143, 148, 166, 249,
329, 332, 413, 421, 423, 471, 526,
541.
Carbonic oxide, 104, 120, 134, 138, 139,
143, 147, 148, 166, 249, 329, 332, 413,
421,423,471,526.
Cast iron, used for guns, 367.
Cavalli, General (article in Bevue de
Technologic Militaire, Memoire sui- les
Eclatements des Canons, etc.), 57, 105,
106, 485, 486, 521.
544
INDEX
Charcoal in gunpowders, 127, 128, 134,
137, 238, 247, 331, 405.
Chevreul, M., article in Dictionnaire des
Sciences Naturelles, 104.
Chilworth Co., 372.
Chlorhydric acid, 126.
Chloride, potassium, 128, 248; zinc, 319,
320.
Chlorine, 326, 327, 337, 398-400.
Chronograph, 430, 431.
Chronoscope, 68, 70-72, 78-80, 174-176,
347, 432, 493-497; data for calculating
velocity and pressure in the bore of a
gun obtained with, 178-186.
Clausius, 166, 199, 539.
Closed vessels, pressures in, 167, 419,
420, 423-429.
Cocoa (brown prismatic) gunpowder,
331, 333-335, 405, 406, 411, 483, 529.
Colossus, H.M.S., 378,
Combustion in bores of guns, tempera-
ture of products of, 202.
" Comite des Poudres et Salpetres," 103.
Committee on Explosives, 116, 146, 162,
173, 175, 177, 187, 190,205,372,373;
on Plates and Guns, 509 ; on Rifled
Cannon, 368, 500.
Oomptes Rendus de VAcadimie des
Sciences, 112, 232, 234, 263.
Conjunctor, of Navez's electro-ballistic
apparatus, 24.
Cordite, and experiments with, 386, 390,
396, 397, 424, 425, 432, 433, 436, 462,
481, 503, 523-529, 534, 536-539; an-
alyses of the permanent gases gene-
rated by, 475, 476; non-detonating,
477.
Cowper-Coles turrets, 374.
Crimean War, 359.
Crusher-apparatus, 114.
Crusher-gauge, 67, 68, 70, 71, 78-80, 84,
174-177, 337, 339, 340, 346, 403, 415-
418, 462-464, 467, 480, 492, 493, 495-
497, 533.
Cupric oxide, 125, 127, 241.
Cupric sulphate, 119.
Curtis and Harvey's No. 6 gunpowder,
2.36, 244, 248-250, 257-262, 294, 302,
303, 305-308, 310-313, 318, 331, 333-
335, 405, 406, 410-412.
Curve, time, 71.
Cylinders, recoil-, 377-383.
Debus, Professor, 309, 316, 317.
Decomposition of gunpowder, 148, 149.
Deflection and range of guns, 11-21.
dela Hire, M., 53, 101.
Deville, M., 172.
Dewar, Professor, 386, 424, 462, 468, 502.
Dictionnaire des Sciences Naturelles, 104.
Didion, General, Traits de Balistique,
26, 30.
Disjunctor, of Navez's electro-ballistic
apparatus, 24.
Douglas, Sir Howard, Naval Gunnery,
358, 500.
Driving-rings, 385 ; result of experiments
with, 387, 392-394.
Driving-surface, 42-44, 47, 50-52, 89-93.
Duke of Wellington, H.M.S., 359, 360,
Eiffel Tower, 356.
Electro-ballistic apparatus, experiments
with Navez's, 23-41 ; Noble's experi-
ments (1860), 369.
Elswick, experiments at, 42, 81-83, 501,
506-508, 521.
Encke, 3Iemoir on the Method of Least
Squares, 4, 6, 16.
Encyclopcedia Britannica, 104.
Enfield rifle, 29.
English-service gunpowder (Waltham-
Abbey), 127.
Eprouvette mortar, 74, 369.
E. R. gunpowder, 29.
Erosion, cause of, 503 ; from new ex-
plosives, 534-536.
Excellent, H.M.S., 380, 500.
Expansion in closed vessels, volumes of,
419, 420.
Explosion, permanent gases generated
by, 53; the phenomenon of, 105;
results deduced by calculation from
analytical data, 151 ; condition of pro-
ducts at the instant of or shortly after,
156 ; of gunpowder, determination of
the temperature of, 170; the products
of, 348; temperature of, 414.
Explosion-apparatus, 113-115.
Explosion-vessel, 337, 402, 403, 533.
Explosive Substances, Committee on, 146.
Explosives Committee, 88, 94, 98, 463,
501, 521.
Explosives, researches on, 99-324, 468-
481 ; heat-action of, 325-354 ; pressures
observed in closed vessels with various,
426-429, 526 ; pressure developed by
some new, 462-467 ; some modern,
521-541.
Fedekow, Colonel, his article in Ziet-
schrift der Chemie, 110; his Russian
powder, 127-130, 137.
Fine-grain (F. G.) gunpowder, 84, 128-
130, 139-143, 145, 147, 150, 151, 155,
156, 159-161, 164, 166, 167, l7l, 207,
210, 215-219, 221, 222, 224, 226, 227,
230, 235, 239, 242, 246, 249-252, 257-
261, 277, 289-293, 301, 303, 306, 308-
313, 315, 318, 321, 322, 331, 333-335,
404-406, 410-412.
Flour dust, 398.
Formidable, H.M.S., 513, 515.
Fort Fisher, 368.
Forth Bridge, 356, 357.
Fossano powder, 265.
Fowler, Sir John, 357.
French B. N. (Blanche Nouvelle) gun-
powder, 481, 524, 525, 528, 529.
Friction in the bores of rifled guns, 385-
INDEX
545
Fuji, Japanese battleship, 511,
Fulminate of mercury, 422, 423.
Fulminates, 397.
Gadolin, General, 73.
Gas, marsh, 104, 120, 139, 166, 238, 250,
329, 332, 393, 421.
Gas pressure (Explosives), measurement
of, 116.
Gaseous products of explosion. 111,
117, 119, 130-134, 139, 140, 149, 151,
153, 155, 166, 254, 321, 322, 406, 408 ;
possibility of dissociation among,
157.
Gases, permanent, generated by ex-
plosion, 53, 104, 410, 411 ; measure-
ment of volume of, 116, 150; from
explosion of guncotton, 472 ; of cordite,
475 ; of ballistite, 478.
Gauge, Rodman's, 58, 67, 70-73 ; crusher,
67, 68, 70, 71, 78-80, 84, 174-177, 337,
339, 340, 346, 403, 415-418, 462-464,
467, 480, 492, 493, 495-497, 533; pres-
sure. 429.
Gay-Lussac, M., 103, 104, 112, 337.
Government Committee on Gunpowder,
146.
Graham, article in Encyclopwdia Brit-
annica, 104.
Gravimetric density, 102, 104, 111, 159,
256, 401, 418.
Greenock Philosophical Society, 397.
Gun-carriages, mountings, turrets, etc.,
374-384.
Guncotton, 328, 397, 469, 471, 517-519,
523 ; composition and metamorphosis
of pellet, 329, 421-423; experiments
with, 339 ; temperature of explosion of
gunpowder and, 340-345 ; results in
volumes of the analyses of permanent
gases generated by explosion of strand,
472 ; results of analyses of strand and
pellet, fired in a close vessel by detona-
tion, 473, 474 ; difference between ex-
plosion of gunpowder and, 541.
Gun-house, armoured, 508.
Gunpowder, tension of fired, 53-86 ;
observed in a close vessel, 158 ; de-
composition of. 100-105, 112, 148, 149 ;
its constituents, 104, 521 ; Govern-
ment Committee on, 146 ; specific
heats and proportions of the products
generated by the combustion of, 166 ;
determination of the temperature of
explosion of, 170; determination of
heat generated by combustion of, 164,
255 ; effect of moisture upon the
combustion and tension of, 120 ; work
effected by, 203 ; when indefinitely
expanded, determination of total
theoretic work of, 208 ; note on the
existence of potassium hyposulphite
in the solid residue of fired, 314 ;
temperature of explosion of guncotton
and, 340-344 ; its advantages, 343 ;
pressure of, 343-345 ; total energy
stored up in, 352 ; curves showing
pressure and work developed by
expansion of, 349 ; its destructive
effects, 436 ; shells charged with, 517;
difference between explosion of
guncotton and, 541.
Gunpowders, employed in researches on
explosives, composition of various,
126, 331 ; results of analysis of, 128 ;
decomposition of various, 333 ;
permanent gases and units of heat
evolved by combustion of various, 334,
335; "A," "B," " C," and " D,"
354, 355, 405, 406, 411.
Guns, ratio between the forces tending
to produce translation and rotation in
the bores of rifled, 42-52 ; smooth-
bored, rifled, and polygonal, 50, 51 ;
observed pressures in the bores of,
173 ; effect of increments in the
weight of the shot on the combustion
and tension of powder in the bores of,
189; pressure in the bores of, derived
from theoretical considerations, 193 ;
temperature of products of combustion
in the bores of, 202 ; comparison
between early rifled and modern
rifled, 350 ; armour-piercing, 373 ;
comparison between 7-inch old and 6-
inch new, 435 ; methods for measuring
pressure in the bores of, 482-498 ;
mounting of, 503-509 ; cradles of, 505,
506 ; of larger calibre, 509-511.
Hall & Sons, 29, 31.
Handy, H.M.S., 503,508.
Haultain, Captain, 17, 20, 21.
Heat-action of explosives, 325-354.
Heat (explosives), measurement of,
116 ; generated by the combustion of
gunpowder, determination of, 164,
255 ; of liquid products, mean specific,
172 ; its loss by communication to the
envelope in which the charge is
exploded, 191 ; " quantity of," 399 ;
units of, 411, 412; specific, 413, 414.
Heats, and proportions of the products
generated by the combustion of
gunpowder, 166.
Hedon, Commandant, 267.
Helmholtz, 354.
History of Explosive Agents, 234.
History of the French Academy, 53.
Hogue, H.M.S., 359.
Hoist, ammunition, 510.
Hotchkiss, 502.
Humphreys & Tennant, 359, 364.
Hutton, Dr, Mathematical Tracts, 101,
102, 194, 348, 433, 487.
Huyghens, 430.
Hydrate, potassium, 119, 125.
Hydraulic rammers and cranes, 376.
Hydrochloric acid, 326.
Hydrogen, 120, 127-129, 137, 149, 166,
247-250, 326-329, 332, 337, 398-400, 421.
Hyposulphite, potassium, 111, 124, 135-
546
INDEX
137, 141-145, 148, 166, 232, 238-242,
244-247, 249-251, 253-255, 314-324.
Increments in the weight of the shot,
their effect on the combustion and
tension of powder in the bore of a
gun, 189.
Initial velocity, experiments in, 23-41.
Institution of Civil Engineers, 325.
Institution of Naval Architects, 499.
Internal ballistics, 397-461.
Iridio-platinum, 414, 415.
Joint Committee on Ordnance to the
U.S. Senate, 367.
Joule, Professor, 168, 198, 539.
KAROL^ia, M. von, his article in
PoggendoriTs Annalen, 110; his experi-
ments with Austrian small-arm
powder, 127-130, 134, 135, 137, 138,
148, 250.
Kelvin, Lord, 539.
Kopp, 166.
Large-grain (L. G.) gunpowder,
29, 31, 207, 316.
Leclanche battery, 403.
Lefroy, General, 83.
Linck, D. J. , his article in Annalen der
Chemie, 109 ; experiments with
Wurtemburg powder, 127-130, 134-136,
144, 145, 148, 250.
Liquid products, their mean specific
heat, 172.
Lyddite, 517, 518, 524, 525, 532, 533, 541.
Lyons, Captain, 22.
Majestic, H.M.S., 511.
Maralunga, Spezia, 383.
Marsh-gas, 104, 120, 139, 166, 238, 250,
329, 332, 398, 421.
Mastiff, H.M. gunboat, 504.
Mathematical Tracts (1812), 101.
Maudslay Sons & Field, 359.
Maxwell, Clerk, 539, 541.
Mayevski, General, 26, 58, 486; his
article in Revue de TechnologieMilitaire,
107.
Measure of precision, 8, 28.
Mechanical Science Section of the
British Association, 355.
Medusa, H.M.S., 364.
Melinite, 517, 528.
Mercury, fulminate of, 422, 423.
Mikasa, Japanese battleship, 514.
Mill, J. S. , System of Logic, 4.
Miller, Hydrostatics, 27."
Mining gunpowder, 236,248-250, 257-266,
277, 294, 302-387, 310-313, 318, 331,
333-335, 405, 406, 410-412.
Moisture, its effect on the combustion
and tension of powders, 190; its
effect in the powder upon the velocity
of the projectUe and pressure of the
gas, 191.
Moncrieff, Colonel, 382.
Monosulphide, potassium, 124, 241, 245,
250, '^52, 254.
Moorsom's concussion fuse, 360.
Mordecai, Major (U.S.A.), Report on
Gumpoioder, 159.
Morgan, Professor de, works on Proha-
hilities, 4.
Morin, General, article in Comptes
Rendus, 232, 233, 255.
Mounting of guns, 503-509.
Murray, Mr, 525, 529.
Naval and Military Services, mechanical
science in relation to the, 355-384.
Naval Artillery, rise and progress of
rifled, 499-519.
Navez, Major, experiments with his
electro-ballistic apparatus, 23-41.
Neumann, General, 57, 58, 80, 486.
Nile, H. M.S., 375.
Nitrate, potassium, 126, 136, 139, 144,
148, 149, 166.
Nitrogen, 104, 120-122, 143, 166, 329,
332, 421, 423.
Nitro-cellulose, 478.
Nitro-glycerine, 328, 397, 478, 535.
Nitrous oxide, 104.
Noble, Captain, 153, 159. 164; his
article on " Tension of Fired Gun-
powder " in Proceedings of Royal
Institution, 108, 111; Internal Ballistics,
469 ; in Philosophical Transactions,
480.
Non-gaseous products, their probable
expansion between zero and tempera-
ture of explosion, 172.
Nordenfeldt, 502.
O'lligqins, Chilian cruiser, 508.
Oil hardening for gun barrels, 370, 371.
Ordnance Select Committee, 23, 33, 88,
369.
Orlando, H.M.S., 380.
Owen, R.A., Lieut. -Colonel, Principles
aiid Practice of Modern Artillery, 100.
Oxide, potassium, 142.
Oxygen, 120, 127-129, 136, 144, 145, 149,
166, 247-249, 254, 255, 327-329, 337,
421.
Palliser, Sir W. , 367.
Pape, 166.
Parabolic rifling, 87-98, 387-396.
Parrott guns (U.S.A.), 368.
Parsons, 367.
Pebble gunpowder, 73-79, 85, 94, 128-
133, 138, 140, 142, 143, 145, 147, 149-
152, 160, 161, 167, 175, 178, 184-187,
189, 190, 200, 201, 206, 210, 212-215,
218-220, 223, 225, 228, 235, 242, 243,
245-247, 249, 251-253, 257-261, 279-283,
301-304, 309-313, 316, 318, 322, 331,
333-335, 385, 386, 391, 400, 405, 406,
411, 412, 427, 432, 433, 462, 463, 528,
529.
INDEX
547
Pellet gunpowder, 73-75, 78, 84, 160, 162.
Pendulum of Navez's electro-ballistic
apparatus, 23.
Penn & Sons, J., 359.
Permanent gases, generated by explo-
sion, 53, 104, 410, 411 ; measurement
of volume of, 116, 150; generated by
explosion of guncotton, 472 ; of
cordite, 475; of ballistite, 478.
Philosophical Magazine, 42, 87, 110, 388.
Philosophical Transactions of the Royal
Society, 102. 248, 254, 255, 264-266,
438, 466, 468, 479, 480.
Picrates of ammonia and potassa, 397.
Picric acid, 524.
Piemonte, Italian cruiser, 364.
Piobert, General, 26, 64, 66, 111, 163,
190, 487, 521 ; Traiti d' Artillerie
Theorique et Experimentale, 100, 103,
104 ; Traits d'Artillerie, PropriHis et
Efets de la Poudre, 105.
Platinum, 171, 414, 415.
PoggendoriTs Annaleu, 108, 110,
Polysulphide, potassium, 141, 144, 239.
Potassa, picrate of, 397,
Potassium, carbonate. 111, 125, 135,
139, 141, 143, 147, 148, 166, 244, 250.
251, 332, 407; chloride, 128, 248;
hydrate, 119, 125 ; hyposulphite. 111,
124, 135-137, 141-145, 148, 166, 232,
238-242, 244-247, 249-251, 253-255, 314-
324; monosulphide, 124, 241, 245,
250,252,254; nitrate, 126, 136, 139,
144, 148, 149, 166 ; oxide, 142 ; poly-
sulphide, 141, 144, 239 ; sulphate. 111,
124, 128, 135-137, 141, 143-145, 148,
166, 238, 244, 245, 247, 248, 250-253,
332, 407 ; sulphide, 125, 135-137, 141-
145, 148, 166, 241, 242, 244, 246, 249,
250, 252-254, 3o2, 407 ; sulphocyanate,
124, 149, 166, 332.
Precision, measure of, 8, 28.
Pressure-gauge, 429.
Pressure in close vessels, deduced from
theoretical considerations, 167 ; in the
bores of guns and measurement of,
173, 482-498.
Pressures in closed vessels with various
explosives, 420, 426, 429, 526,
Prismatic gunpowder, 73-78.
Probabilities, theory of, its application
to artillery practice, 1-22,
Proceedings of Royal Institution, 111 ;
of the ^ Royal Society, 309, 385, 462,
468, 469.
Products, gaseous. 111, 117, 119, 130-
134, 139, 140, 149, 151, 153, 155, 166,
254, 321, 322, 406, 408; possibility of
dissociation among, 157; solid, 118,
130-135, 138, 144, 146, 147, 149, 151,
153-155, 166, 254, 321, 322, 407, 409.
Products, their condition at the instant
of or shortly after explosion, 156 ;
generated by the combustion of gun-
powder, their specific heats and
proportions, 166 ; mean specific heat
of Hquid, 172 ; of combustion in bores
of guns, their temperature, 202.
Projectiles, rifled, pressure required to
give rotation to, 87-98.
Prussian Artillery Committee, 57, 58,
106, 486, 487.
Pyroxylin (guncotton), 328.
" Quantity of heat," 399,
Range and deflection of guns, 11-21,
Rankine, Steam Engine, 198, 199.
Ravenhill Miller & Co., 359.
Recoil-cylinder, 377-383.
Reqina Marqherita, Italian battleship,
Regnault, 27, 166. [514.
Rendel, George, 378.
Rennie Brothers, 359.
Researches on explosives, 99-324, 468,
481 ; list of contents. Part I., 99, 100;
Part II., 231; summary of results,
209, 211 ; abstract of experiments,
211-230, 279-309, 321, 322.
Re Umberto, 378, 509, 515.
Revue de Technologie MiUtaire^ 87, 105-107,
Revue Scientifique, 164.
Reynolds, Professor Osborne, 390.
Rifled Cannon, Special Committee on,
1, 11,20.
Rifled guns, translation and rotation in
the bores of, 42-52 ; energy absorbed
by friction in the bores of, 385-396.
Rifled Naval Artillery, rise and progress
of, 499-519.
Rifle fine-grain (R. F, G.) gunpowder,
128-133, 171, 210, 229, 249, 310-313,
400, 404, 405.
Rifle large-grain (R. L. G.) gunpowder,
72-74, 76-78, 80, 86, 128-133, 138-143,
145, 147, 150, 151, 156, 159-162, 164,
166, 167, 171, 175, 177, 185, 187-190,
200, 201, 210-212, 214, 215, 219-221,
223-227, 230, 235, 242, 243, 24.5-247,
249-253, 257-26:., 274-277, 284-288,
295, 301, 302, 304, 310-313, 315, 316,
318, 322, 331, 333-335, 400, 405, 406,
410-412, 481, 483, 490, 491, 527-529.
Rifling, of Woolwich guns, 88, 90, 93,
97 ; uniform and parabolic, 387-396 ;
polygonal, 51.
Roberts- Austen, Sir W., 537.
Robins, New Principles of Gunnery, 53,
54, 67, 79, 100-102, 163, 189, 190, 483,
485, 521.
Rodman, Major, Experiments on Metal
for Cannon and qualities of Cannon
Poioder, 58, 59, 61-63, 83, 107, 108,
163, 164, 347, 429, 487 ; his pressure
apparatus, 488-493.
Rodney, H.M.S., 375,
Rotation, in the bores of rifled guns, 42-
52 ; to rifled projectiles, pressure
required to give, 87-98 ; of modern
breech-loading projectiles, 385.
Roux and Sarrau, MM., article in
Comptes Rendus, 112, 233.
548
INDEX
Royal Arthur, H.M.S., 506.
Royal Artillery Institution, 1, 23, 108.
Royal Institution, 53, 55, 108, 111, 158,
521.
Royal Society, 99, 309, 316, 462, 479.
Royal Sovereign, H.M.S., 510.
Rumford, Count, 55, 56, 62, 63, 67, 83,
100, 102, 103, 105, 111, 163, 189, 482-
485, 521.
Saint Robert, Count de {Trait6 de
Thermodynainique), 26, 194, 195, 198,
199, 348, 433, 434.
Saltpetre, 128, 134, 138, 139, 147, 238,
247, 248, 331, 405.
Sardinia, battleship, 510.
Science, mechanical, in relation to the
naval and military services, 355-384.
Schischkoff, Professor, 63-65, 67, 82, 108,
109 ; see also Bunsen and Schischkoff.
Sebastopol, guns employed at siege of,
366.
Sheffield, and gun steel-making, 370.
Shell-fire, importance of, 360, 517, 518.
Shells, high explosive, 361.
Shikishima, Japanese battleship, 515.
Sicilia, battleship, 510.
Siemens furnace, 143, 173.
Sinope, battle of, 360.
Solid products, 118, 130-135, 138, 144,
146, 147, 149, 151, 153-155, 166, 254,
321, 322, 407, 409.
SoHd residue (explosion), analysis of,
121-126, 130-133, 406.
Somerset gun, 509.
Spanish gunpowder (spherical pebble
and pellet), 128-133, 135, 139, 151,
160, 229, 236, 249, 257-266, 276, 294,
301-305, 310-313, 331, 333-335, 405,
406, 410-412.
Special Committee on Rifled Cannon, 1,
11,20.
Sporting gunpowder, 55, 128-130, 236,
244, 248-250, 257-262, 294, 302, 303,
305-308, 310-313, 318, 331, 333-335,
405, 406, 410-412.
Steel, for gunmaking, 370, 371.
Stephenson, George, 437.
Sulphate, potassium. 111, 124, 128, 135-
137, 141, 143-145, 148, 166, 238, 244,
245, 247, 248, 250-253, 332, 407.
Sulphide, potassium, 125, 135-137, 141-
145, 148, 166, 241, 242, 244, 246, 249,
250, 252-254, 332, 407.
Sulphocyanate, potassium, 124, 149,
166, 332.
Sulphocyanide, 238.
Sulphur, 128, 134, 140, 145, 147, 149,
166, 245, 248, 250, 251, 331, 405 ; dust,
398 ; free, 122, 124, 141, 142, 144, 250.
Sulphuretted hydrogen, 104, 118-121,
123, 134, 166, 238, 250, 408.
Sulphuric acid, 332.
Temperaturk of explosion, 414 ; its
determination, 170.
Tension of fired gunpowder, observed
in a close vessel, 158.
Terrible, H.m.S., 360.
Theory of Probabilities applied to
artillery practice, 1-22.
Theseus, H.M.S., 360.
Time curve, 71.
Torpedo boats, 365, 381.
Trafalgar, battle of, 359.
Trafalgar, H.M.S. , 362, 363, 378.
Transactions of the Royal Institution, 53,
55 ; of the Royal Society, 99.
Translation in the bores of rifled guns,
42-52.
Trinitrocellulose (guncotton), 328.
Tromenec, M. de, 233, 263 ; article in
Comptes Rendus, 112.
Turntables, 376-378, 510, 512.
Turrets, revolving, 374 et seq.
Twist, accelerating, 87-98; no, 387, 392,
394, 396 ; uniform, 94, 97, 98 ; uni-
form and parabolic, 392, 393.
Uniform rifling, 387-391, 393-396.
Uniform twist, 94, 97, 98, 392, 393.
United States, use of cast-iron guns in,
367.
Units of heat, 411,412.
Variations of fire in artillery practice,
1,2.
Vavasseur, Mr, 378-380, 383, 385, 499-
502, 507-509, 517.
Velocity, experiments in initial, 23-41.
Vessels, pressure in close, 167, 419, 420.
Victoria, H.M.S., 362, 363.
Victoria, Queen, 357.
Victory, H.M.S., 357, 359, 360, 363.
Walli'iece, 29.
Waltham-Abbey gunpowder works, 29,
31, 119, 120, 127, 159, 245, 248, 249,
257-266, 277, 310-313, 331, 333-335,
405, 410-412, 415, 434, 471, 490.
Watt, James, 397, 437.
Watts, Mr, Chief Constructor at
Elswick, 512.
Whitworth & Co., Sir J., 370.
Woolwich, rifling of guns, 88, 90, 93,
97 ; experiments at, 462, 463.
Woulfe's bottles, 122.
WiJrtemburg war-powder, 109; cannon-
powder, 128-130, 134, 235.
Yarrow, 365.
Yashima, Japanese battleship, 511.
Younghusband, R.A., Colonel, 67, 94,
127, 128, 173, 190.
Zeitschrift der Chemie, 110.
Zinc chloride, 319, 320.
'HINTED BV OLIVER AND BOVD, EDINBURGH.
YD 00551
I 7" ^9:-