Various prescriptions, or formulæ, have been given for the preparation of this fire. The oldest is by princess Anna Commena, in which, however, there is only resin, sulphur, and oil. Beckman observes, that the first certain mention of saltpetre will be found in the oldest account of the preparation of gunpowder, which, in his opinion, became known in the thirteenth century, about the same time that the use of the Greek fire, of which there were many kinds, began to be lost. The oldest information on this subject is to be found in the works of Albertus Magnus, and the writings of Roger Bacon. The true recipe for making the Greek fire, and the oldest for gunpowder, were found in a manuscript, preserved in the electoral library at Munich. Various copies of this manuscript were made. Bacon employed this writing, which was mentioned by Jebb, in the preface to his edition, from a copy preserved in the library of Dr. Mead. Whether the writer was Marcus Græcus, is of no moment; for Cardan observes, that the fire that can be kindled by water, or rather not extinguished by water, was prepared by Marcus Gracchus.
The former Marcus, mentions two kinds of fire-works; and the composition, which he prescribes for both, is two pounds of charcoal, one pound of sulphur, and six pounds of saltpetre, well powdered and mixed together in a stone mortar.
Friar Bacon, who lived three centuries after Græcus, was in possession of the recipe. It was concealed, however, from the people, veiled in mystery. In his treatise De Secretis Operibus Artis et Naturæ, &c. the secret of the composition is thus expressed: "sed tamen salispetræ, LURU MOPE CAN URBE et sulphuris; et sic facies tonitrum et corruscationem, si scias artificium." Luru mope can urbe, is the anagram for carbonum pulvere. Bacon supposes, that it was with a similar composition that Gideon defeated the Midianites, with only three hundred men. Besides the use of gunpowder in the 9th century, in the war between the Tunisians and the Moors, in which the former are said to have employed "certain tubes or barrels, wherewith they threw thunderbolts of fire," the Venetians employed it against the Genoese, and it was reprobated as a manifest contravention of fair warfare.
Peter Mexia, in his "Various Readings," relates, that the Moors, being besieged, in 1349, by Alphonso the eleventh, king of Castille, discharged a kind of iron mortars upon them, which made a noise like thunder. This, with the sea-combat between the Tunisians and the Moors, stated on the authority of don Pedro, bishop of Leon, places the invention much earlier than by some writers.
Polydore Virgil ascribes the invention of gunpowder to a chemist, who, having put some of his composition in a mortar, and covered it with a stone, was blown up, in consequence of its accidentally taking fire. The person here alluded to, according to Thevet, was a monk of Friburg, named Constantine Anelzen. Others, as Belleforet, with more probability, hold it to be Bartholodus Schwartz, or the black, who discovered it, as some say, about the year 1320. Du Cange, however, remarks, that there is no mention made of gunpowder in the registers of the chamber of accounts in France, as early as the year 1338. Roger Bacon knew of gunpowder, near one hundred years before Schwartz was born. (See the invention of cannon, in [military fire-works], fourth part.)
It is certain, that Albert de Bollstædt indicated the constituent parts of gunpowder, when he says, in his Mirabilis Mundi, "Ignis volans, accipe libram unam, sulphuris, libras duas, carbonas salicis, libras sex, salis petrosi, quæ tria subtilissime terantur in lapide marmorea; postea aliquid posterius ad libitum in tunica de papyro volante, vel tonitrum faciente ponatur.
"Tunica ad volandum debet esse longa, gracilis, pulvere illo optime plena, ad faciendum vero tonitrum brevis, grossa et semiplena."
Gunpowder was of a much weaker composition than that now in use, or that described by Marcus Græcus. Tartalgia, (Ques. and Inv. lib. 3, ques. 5), observes, that, of twenty-three different compositions, used at different times, the first, which was the oldest, contained equal parts of the three ingredients. When guns of modern construction came into use, gunpowder of the present strength was introduced.
The strength of powder depends upon the proportions of the ingredients, they being pure; and Mr. Napier observes, (Trans. Royal Irish Academy, ii.) that the greatest strength is produced, when the proportions are, nitre, three pounds, charcoal, nine ounces, and sulphur, three ounces. The cannon powder was in meal, and the musket powder in grain.
In the time of Tartalgia, the cannon powder was made of four parts of nitre, one of sulphur, and one of charcoal; and the musket powder of forty-eight parts of nitre, seven parts of sulphur, and eight parts of charcoal; or of eighteen parts of nitre, two parts of sulphur, and three parts of charcoal.