THE ROYAL ARTILLERY INSTITUTION. 
31 
was naturally more violent tlian that of No. 2. The door was 
blown 64 yds. to the front, and portions of the magazine were 
blown 32 yds. to the right, another piece 30 yds. to the back, 
another 40 yds. to the left rear, another 52 yds. to left front. 
A piece of alloy, the melting point of which was 482° Fahr., 
was recovered unmelted and unaffected by the heat. 
No 4 furnace was extinguished at 9.15 a.m., or 22 hours after 
being lighted. 
On being opened, this magazine (No. 4) was found to be in very 
good condition ; a few of the rivets had been drawn by the expan¬ 
sion of the metal, but it was otherwise uninjured. 
The powder in the canisters was in one or two instances slightly 
damped. That in the barrel was encrusted, or superficially 
agglomerated to the depth of about -rVth to |th inch, and Pro¬ 
fessor Abel's analysis of this powder showed that this crust had 
sustained a slight loss of sulphur. Beneath the crust the powder 
was unaltered. 
I append (Appendix F) a memorandum on the condition of this 
and the other samples of recovered powder, with which Professor 
Abel has obligingly furnished me. 
One of the thermometers (Negretti's) registered 250°, the other 
registered 220°. None of the alloys had melted, nor was the 
solder on the canisters affected. 
The result of this experiment must be regarded as eminently 
satisfactory. 
Of the three magazines designed to resist six hours—a more 
than sufficient time,, in the opinion of the Superintendents of the 
Fire Brigades, to warrant a magazine being considered practically 
“ fire-proof"—one had resisted nearly 19 hours, another nearly 
17 hours, and the third, when opened at the end of six hours, had 
been found to have its contents unexploded. 
The magazine designed to resist eight or nine hours, had 
resisted 22 hours, and then showed no signs whatever of any 
approaching failure. 
Messrs. Milner have thus satisfactorily demonstrated that a 
magazine can be made which will be practically fire-proof—which 
will, indeed, resist for a longer period than it would be likely to 
be called upon to endure, a far more intense fire than in any 
ordinary conflagration could be directed against it. 
It is worth while to consider in what way the explosions of the 
magazines Nos. 2 and 3 were effected, in order to observe how far 
the safes acted as had been anticipated. The expectation was 
that the vapour from the alum would penetrate to, and thoroughly 
saturate the gunpowder, and thus render it inexplosive. In the 
case of the powder which was hermetically closed in canisters, it 
was anticipated that when the temperature rose to about 240° the 
sulphur would begin to volatilise off, and so leave the remaining 
powder explodable only at a temperature far in excess of that at 
which powder containing sulphur explodes. 
The first of these expectations was certainly not realised; the 
Result of 
experiment 
with magazine 
No. 4. 
Condition of 
recovered 
powder. 
Summary of 
results of 
experiments. 
Consideration 
of the causes of 
the explosion of 
Nos. 2 and 3 
magazines. 
