the risk of accident is reduced to a minimum, and there is the less need of any protective process being applied to it. It is when being handled that the risk really begins; and no policy could be more dangerous than to accustom our artillerymen and sailors to a substance that was innocuous one moment and fiercely explosive the next. The fancied security attained' would in evitably lead to neglect of precautions, and might result in those very disasters which it is the object of the invention to avert. Anything like a bead of undecomposed saltpetre, however small, should at once condemn the powder for shooting purposes.

The question of the examination and proof of gunpowder for use in Her Majesty's service is a much more complicated and serious one, and at present is in an unsettled state. Hitherto, all that was done after flashing off a little of the powder on a glass plate, when the amount of residue will indicate perfect or imperfect working, and cubing,' that is, weighing a cubic foot of the finished powder to ascertain roughly its density, was to fire a few ounces from a mortar loaded with a solid shot, the powder which threw the shot furthest being esteemed the best. But, as was pointed out many years ago, the proof mortar is entirely useless as an instrument for testing the relative projectile force of different kinds of gunpowder when employed in large charges as in cannon. Its use has been abandoned by all the continental artillerists, and is only retained in our service as a rough and ready way of comparing contract powder with samples of the same pattern of powder made at Waltham Abbey. But the results given by it with two different kinds of powder are entirely fallacious. The question of proof has assumed the highest importance since the introduction of rifled ordnance, when it soon became apparent that the cannon powder in ordinary use was altogether too strong and sudden in its action for the endurance of the guns.

The strength of exploded gunpowder depends on two things, viz. the quantity of gas evolved, and the rate at which it is evolved. No doubt these two are what mathematicians would term 'functions' of each other, for the amount of heat developed plays an important part in the action from the great expansion it produces in the gases, and the more rapid the combustion the greater will be the heat. A theoretically perfect proof, therefore, would be one which would correctly measure these two. But taking for granted that the quantity of gas produced from a certain quantity of service gunpowder is a uniform quantity, the whole question resolves itself into the best means of measuring the rapidity of combustion of charges of gunpowder, differing not only in quantity, but in the size, shape, density, and hardness of the component powder grains. The fallacy of the old mortar Vol. 125.-No. 249.

K

proof

proof lay in this, that the charges fired being only two or three ounces, the result obtained in comparing different kinds of powder were exactly the reverse of those obtained when firing large charges in cannon, a small-grain powder from its more rapid combustion giving, with the mortar, a higher range than any other. To those whose only idea in connexion with gunpowder is its intensely sudden and fiercely explosive character, by which apparently in an instant thousands of barrels of quiescent material are converted into a scorching and rending mass of flame, it may appear over refining to talk of the rate of combustion of gunpowder. But it can be demonstrated in many ways that although the explosion of a quantity of gunpowder confined in a gun may be said to be instantaneous, it is not simultaneous; for the portion next the touch-hole being first ignited, the combustion proceeds gradually through the charge. Improbable as it may appear, it is an undoubted fact, that when firing with large charges a quantity of powder furthest from the touch-hole is often blown out of the gun unignited, along with the shot, and can be seen on the ground under the muzzle of the gun. The rapidity of action of a charge of powder depends on two conditions: the rate of burning of each grain, which may be called the velocity of combustion; and the rate at which the flame is communicated from one grain to another, the velocity of ignition. As the disruptive force-the force brisante-of the powder depends on the rapidity of action, these two points have within the last few years, particularly since the failure of some of our large guns, received the most careful study from artillerymen at home and abroad; and although the whole subject is as yet to a certain extent sub judice, the following are the conclusions which are generally admitted. Assuming that the various powders in use are of the same composition, as regards the proportion and incorporation of their ingredients—any modification of which would only cause a less perfect decomposition on firing, and consequent fouling of the guns used-they may yet differ widely in the following five points, viz. size, shape, density, hardness, and amount of glaze of the constituent grains. The hardness, depending mainly on the quantity of moisture present in the meal before pressing, is generally nearly uniform, and as no test for the hardness of a body has as yet been devised, it is not allowed to complicate the question; and the question of the amount of glaze may be dismissed with the remark that a powder of which the grains are very highly polished, particularly when a little black lead has been used, appears to be slower in ignition than one of which the grains are rough and porous, the flame being able to get a better and quicker hold of the rougher surfaces. But the density of powder grains, that is the quantity

of

The

of matter actually present in each, not only may but does vary to a great extent; indeed it may be asserted that it is impossible to manufacture powder of uniform density by the present method. This, as has been explained, is caused by the impossibility of ensuring the same amount of pressure being applied to the powder-dust in the hydraulic press, not only from the difficulty of ensuring the same quantity exactly being under pressure each time, but from the impracticability of causing a uniform pressure throughout the mass. A dense powder must of course burn more slowly than a less dense one, not only from the closer texture of the grains being unfavourable to combustion, but because, obviously, there is in the case of the former a larger quantity of matter, bulk for bulk, to be consumed. questions of the size and shape of powder grains, and the effects produced by modifications of these, are extremely puzzling, and the results obtained appear at first sight to be contradictory. If a small grain of powder, as must be admitted, takes a shorter time to burn than a large one, it might be supposed that a cartridge full of small grains would be sooner converted into gas than one full of large ones. The reverse is found to be the case in practice. Again, a sphere being the smallest form in which matter can be placed, a spherical grain of powder, offering the smallest surface for combustion, will burn more slowly than the same quantity of powder arranged in the form of a flat scale. And yet practically the rate of burning of a cartridge full of spherical grains will be quicker than that of one full of flat grains packed closely together. The explanation of these apparently irreconcilable facts lies in this, that the large or spherical grains, having much larger interstices between them, allow of a far more rapid passage of the flame from grain to grain; in other words, though the velocity of combustion be lower, the velocity of ignition is increased, and the consequent result becomes the very reverse of what was intended by increasing the size and rounding the shape of the grains. The many different conditions under which charges of powder may be ignited prevent any certain rule being laid down on the subject; for what holds good with a very small charge in a gun will be reversed with large charges; and the powder which strains the gun much, when fired in short compact cartridges, may have quite a different effect when formed into long narrow ones.

As powder varies in density, so it varies in the size and shape of its grains. The system of cracking up the press-cake into grains in the granulating machine, produces fragments of all shapes and sizes; the latter are of course restricted to a certain extent by the siftings the powder undergoes, but no two powder grains are alike. It follows therefore from what has been said,

that no two charges of gunpowder made in the present way will produce exactly the same force brisante in the guns from which they are fired. The introduction of the modern guns of precision, particularly the breech-loading guns of Sir W. Armstrong, in which by using soft-coated projectiles all windage or escape of gas is prevented, and the employment of the delicate electroballistic instrument of Major Navez for measuring the velocity of shot, afford actual proof that this is the case. The problem of the day in gunpowder has therefore been to manufacture a perfect powder, in which each grain shall be the same in all respects as its fellow; the disruptive effects of which shall be, to a certain extent, under control by modifications of manufacture; and the results of which shall be uniform. The idea is not a new one; for it is distinctly shadowed forth by Congreve in one of his works, published fifty years ago. Nor on a small scale would it be difficult; for all that would be necessary, would be to compress some powder dust into small moulds, with varying pressures, till the resulting pellets which would of course be of the same shape and size, became exactly of the density required. It is only when the manufacture is attempted on a large scale that the difficulty and danger begin. However, it is satisfactory to know that these have been successfully overcome, and that the manufacture of pellet powder, as it is called, will be immediately proceeded with.

To the Americans belongs the credit of first having attempted to modify the disruptive, without interfering with the propellant force of gunpowder. They took powder meal and compressed it by hydraulic pressure into flat discs, the size of the bore of the gun. These discs or cakes, the invention of Dr. Doremus, were brought to this country and tried, but the results obtained with them were very irregular and unsatisfactory. The Russians seem next to have taken the matter up, having carried on a series of experiments with prismatic powder, which was merely powder meal compressed into small hexagonal prisms, perforated with holes; and the Belgians followed in the same direction. A similar powder, termed pellet powder, was experimented on in this country in 1865 66 and recommended at first for temporary adoption; and in 1867, as the service powder for all very large charges. The pellets differ from the Russian ones only in shape and dimensions, being small cylinders about half an inch in height, and three quarters of an inch in diameter, with a small hollow or perforation at one end. These constitute the grains of the uniform powder required for our modern rifled artillery; and the difference in manufacture between the oldfashioned, time-honoured gunpowder and this, consists merely in taking the incorporated ingredients from the gunpowder-mill

and

and pressing them by hydraulic pressure into small moulds to form the pellets; instead of pressing a large quantity of the mixture into 'press-cake,' and crushing it up afterwards into irregular grains to be afterwards dusted, glazed, and re-dusted for use. These latter operations are not required with the pellets, which pass at once from the state of meal to their finished condition.

Small quantities of pellet powder have been made from time to time in the Royal Arsenal, and the successful results it has afforded in practice have induced the Ordnance Select Committee to recommend its immediate adoption for all large charges. The manufacture will be carried on at Waltham Abbey on a large scale, and if the machinery now in course of erection be found to work rapidly and safely, a large supply of pellet powder will be forthcoming, and our artillerymen will have at their command a powder which will exercise the least possible destructive effect on their guns, and yet retain its propellant power uninjured.

ART. V.-1. Le Livre de Marco Polo, Citoyen de Venise, Conseiller Privé et Commissaire Impérial de Khoubilai Khaàn. Par M. G. Pauthier. Paris, 1865.

2. I Viaggi di Marco Polo. Firenze, 1863.

Per cura di Adolfo Bartoli.

3. I Viaggi di Marco Polo Veneziano. Da Vincenzo Lazari. Venezia, 1847.

4. Die Reisen des Venezianers Marco Polo ... mit einem Kommentar von August Bürck; nebst Zusätzen und Verbesserungen von Karl Friedrich Neumann. Leipzig, 1845.

5. The Travels of Marco Polo, greatly amended and enlarged, &c., with copious Notes. By Hugh Murray. Edinburgh, 1844. 6. Viaggi di Marco Polo, illustrati e commentati dal Conte Baldello Boni. Firenze, 1827.

7. Voyages de Marco Polo (French and Latin Texts, in Recueil de Voyages, &c., publié par la Société de Géographie), Tome 1. Paris, 1824.

8. Publications of the Hakluyt Society.

THE

prosperous activity of the Hakluyt Society indicates, we presume, that a goodly number among us still take delight in old travellers, and in the history of geography. And since this Review noticed Marsden's memorable edition of 'Marco

*

Polo,' nearly half a century ago,' so much new light has been thrown upon him, and other medieval travellers in Asia, that

*See Quarterly Review, vol. xxi. p. 177, seqq.

we