260 
MINUTES OF PROCEEDINGS OF 
in the two former, but the probability of an explosion could have been 
equally foretold in all three instances; and though ignorance may 
perhaps be pleaded as an excuse in the case of the gas-explosion, and 
carelessness made to account for that of the powder, those pleas cannot 
be accepted as grounds for regarding the resulting disasters as legiti¬ 
mate accidents. Indeed, it is often impossible to forbear stigmatising 
as criminal the carelessness or ignorance which gives rise to explosions 
of the kind popularly accepted as accidental. 
In the present discourse it is proposed to accept the definition 
“ accidental 33 in the loose sense in which it is popularly applied to 
explosions, with the object of examining into the nature and causes of 
such explosions, and, if possible, of indicating directions in which there 
may be hope of successful efforts being made for reducing the frequency 
of their occurrence. 
The phenomena attendant upon an explosion are generally due to 
the sudden or very rapid expansion of matter, accompanied in most 
instances by its change of state from solid or liquid to gas or vapour. 
The most simple classes of explosions are those caused by the sudden 
yielding to force, exerted from within, of receptacles in which a gas is 
imprisoned in a highly compressed condition, or a liquid has been raised 
to a temperature greatly exceeding that at which its molecules have a 
tendency to fly asunder or to assume the state of vapour or gas. The 
strength or elasticity of the envelope which confines them suddenly 
yielding to pressure, the liquid passes with great rapidity into vapour, 
violently displacing by this sudden expansion the surrounding air and 
any other obstacles opposed to the expanding molecules. 
Similar explosive effects less simple in their origin are brought about 
by the sudden development of chemical activity in mixtures of gases or 
vapours, of solids and gases, or of solids only, or in chemical compounds 
of unstable character; the result in all such instances being the develop¬ 
ment of intense heat and the sudden or very rapid and great expansion 
of matter. 
Examples of the most simple class of explosions are the sudden 
failure in strength at some particular point, or generally, of the material 
composing a vessel in which a gas has either been liquefied or highly 
compressed. Accidental explosions of this character take place chiefly, 
and happily not very frequently, in the laboratory or lecture room, yet 
instances occasionally occur of disastrous explosions resulting from such 
causes in manufacturing operations, or in the practical application of 
compressed air or other gases. The most recent illustration of a serious 
accidental explosion of this kind is that which occurred in the Arsenal 
at Woolwich, in January 1874, with the air-chamber of a Whitehead 
or fish torpedo, when one man lost his life and several were seriously 
injured. In this instance, some part of the soft steel diaphragm closing 
the chamber in which the motive power of this self-propellant torpedo 
(air) was imprisoned under a pressure of about 800 lb. on the square 
inch, suddenly yielded to the efforts of the gas to return to its normal 
condition. These air-chambers were invariably tested up to a pressure 
very considerably exceeding that to which they would be subjected in 
actual service before they were allowed to form part of the torpedo; 
