Since the use of springs, compressed air, and steam were impossible, and the use of an explosive extremely improbable, it only remains to examine the arguments for and against water as the motive power.

The Emperor Leo VI. speaks of the “artificial fire discharged by means of siphons;”[95] the Emperor Constantine VII. speaks similarly of “the wet-fire projected by means of siphons;”[96] and if we translate siphon by tube both phrases are intelligible, but neither gives any hint as to the means by which the mixture was expelled from the tubes. But like so many other military words, siphon has (at least) two meanings, and signifies not only a tube, but a fire-engine, or water-engine, or squirt. Heron of Alexandria (cir. 130 B.C.) begins his description of a fire-engine with the words: “The siphons used for the extinction of fire are made as follows.”[97] Pliny the Younger (cir. 100 A.D.), in a letter to the Emperor Trajan about a fire which had taken place in the town of Nicomedia, observes that “there was not a sipho, nor even a bucket, at hand to quench it.”[98] Hesychius in his Greek Lexicon, about the end of the fourth century, gives under σίφων: “an apparatus for pumping water in conflagrations.”[99] If we translate siphon by water-engine, as we are perfectly justified in doing, we find that the phrases used by the two Emperors are not only intelligible, but indicate both the mode of projection and the mode of ignition of the sea-fire. The lump of quicklime-naphtha-sulphur was projected, and at the same time ignited, by applying the hose of a water-engine to the breech of the tube, which thus became an integral part of the apparatus to which it gave its name.

Two obscure passages in Byzantine works, which hitherto have never been satisfactorily explained, are made clear by this interpretation. The first occurs in the “Tactics” of Constantine VIII., where he directs “‘flexible’ (apparatus) with (artificial) fire, siphons, hand-siphons, and manjaniks” to be employed, if they are at hand, against any tower that may be advanced against the wall of a besieged town.[100] The “flexible” apparatus cannot refer to the helical springs of a later age. Neither can it mean crossbows, for the Princess Anna, who wrote a century after Constantine, expressly says: “The crossbow (tzangra) is a foreign weapon (hitherto) unknown in the Greek service.”[101] That it cannot mean longbows is quite certain from the second of the obscure passages in question, which occurs in the “Alexiad”: “In the bow of each ship he (the Admiral) put the heads of lions and other land animals, made of brass and iron, gilt, so as to be (quite) frightful to look at; and he arranged that from their mouths, which were (wide) open, should issue the fire to be delivered by the soldiers by means of (or through) the ‘flexible’ apparatus.”[102] The enemy might have exclaimed with the Jewish king: “They gaped upon me with their mouths, as a ravening and as a roaring lion.” But whatever the moral effect of these trumpery scarecrows—if ever actually used—it is certain that archers with longbows could not have shot fire-arrows through them with any success; and the meaning of Emperor and Princess remains hidden until we interpret “flexible instrument” as the leathern hose of a pump or water-engine, than which nothing can be more flexible. The import of both passages then becomes perfectly plain. Such a mode of discharging incendiaries is by no means unknown in later military history. “Dans une expérience faite au Havre, 1758, avec une pompe à huile de naphte, dont le jet était enflammé par une mèche allumée, on brûla même une chaloupe.”[103] At the siege of Charleston, 1863, not only was solidified Greek fire in tin tubes employed,[104] but coal naphtha placed in shells or pumped through hose.[105] Finally, M. Berthelot speaks of “les propositions faites, pendant le siège de Paris (1870), pour repousser les ennemis au moyen de pompes projetant des jets de pétrole enflammé. Mais cet agent ... n’a été mis réellement à l’épreuve que par la Commune, pour brûler nos palais.”[106]

When the Crusades began in 1097 Greeks were thus in possession of two species of incendiaries: the sea-fire which was distinctively and exclusively Greek, and the old mixture of the Æneas family which was known all over the East. Yet it was to the latter that the name “Greek fire” was given by the Crusaders, who, I believe, had neither experience nor knowledge of the sea-fire. The only passage I can recall among the old writers in which the two fires are discriminated and correctly named occurs in the metrical romance

“Richard Coer-de-Lion,” temp. Edward I. (1272-1307):—

“Kyng Richard, oute of hys galye,

Caste wylde-fyr into the skye

And fyr Gregeys into the see.

.....