some time before using, and in starch paste similarly treated. Spontaneous ignition has also been caused by the contact of oil with finely divided carbon such as lampblack or finely divided metals, such as magnesium and aluminium, which are so largely used at the present time.
Another case of this nature is the heating up of cases after charging with gerb composition, two of the ingredients of which are sulphur and iron borings, this heating sometimes being sufficient to cause combustion. The cause of this phenomenon is the combination of the iron with sulphur to form sulphide of iron, this action being accompanied by heat. In fact, it is the same as that producing the experiment known as Lemery's volcano. As far as the knowledge of the writer extends, however, no occurrence of ignition has been definitely traced to this phenomenon, although it seems highly probable that even if ignition of the actual composition has not taken place, cases have occurred where more sensitive compositions have been fired by heat so generated where fireworks have been stored together.
The annual reports of H.M. Inspectors of Explosives published since 1876 form an interesting and instructive summary of accidents in explosive trades, an examination of which throws considerable light on our subject.
One is struck by the frequency with which explosions occur as a result of ignorance, generally on the part of amateur firework makers. In many cases, as where children are concerned, this ignorance is natural, but the want of knowledge and even reasonable care displayed by individuals whose occupation suggests at least some knowledge of the risk is indeed often extraordinary.
Such a case occurred in 1884 in Devonshire, when a local chemist who was illegally manufacturing coloured fire, instructed an assistant to grind in an iron mortar a mixture
