why its effect should be so different from that produced by an other form of carbon, has not been satisfactorily explained.
The compositions we have been considering fall into one of two classes, namely, those to produce force and those to produce sparks. These two classes, with one other, namely that of colour, may be said to include all the modern recreative firework compositions. Up to the end of the eighteenth century the ingredients used in the production of the compositions of these three classes were very few in number. A considerably larger number went to supply the ingredients for a fourth class now almost extinct, these might be called the flame-producing class. The principle on which these compositions were designed was, as it were, to overload a mixture of saltpetre and sulphur with combustible material; this latter took the form of gums, resins, or fats, the object being to produce a reddish or golden coloured flame. The early writers give formulæ for variously coloured stars and fires, which must have required considerable effort on the part of the observer for identification. These belonged to the flame class.
Frézier, with more perception than most of the others, realised the shortcomings of such compositions, merely designating them greenish (verdâtre), yellowish, reddish, and russet. The only colour which he professes to produce distinct is blue, which he obtained with pure sulphur.
Progress from the earliest times of pyrotechny up to the first quarter of the nineteenth century was very gradual and very slight. The chemicals used by Bate and Babington in their actual pyrotechnic compositions were as follows:—Gunpowder and its constituents, camphor, pitch, resin, orpiment, linseed oil, both pure and boiled, oil of spike (spica lavandula), oil of petre (rock oil), an oil known either as benedict or tile, varnish (probably amber), iron scales, and aqua vitæ (spirits of wine).
