direction of fracture.

re-enforce would be the result, as shown in the diagram. As the heat expands the metal in the direction of the diameter also, its effect in this direction also must be considered. The expansion of length, however, is of most consequence in considering the probable

That the fracture almost always intesects the vent, has been heretofore referred to the weakness resulting from drilling away part of the metal; but on page 355, Major Wade's Reports on Metals for Guns, we find that after a gun had been put to extreme proof, and exhibited signs of fracture, a hole was drilled one inch forward of the base ring, and four inches from the line of the vent, to a depth of four inches, and of the diameter of one and a quarter inches. The gun was then fired with double charges of power, and with a bore full of balls and wads, eleven times, to bursting. Although the piece burst into more then twelve fragments, one of the fractures intersecting the vent, it did not split through the large hole, showing that the gun had strength to resist the pressure of the powder, but burst, notwithstanding the drilling away of so large a part of the metal, from the communication of heat. The true cause, probably, of the intersection of the vent by the fracture, was the communication of heat to the surface of the vent, thereby expanding the column of metal about it; for it should be recollected that the passage of a large quantity of gases through the vent would communicate more heat to its surface than would be communicated if there was no current, but the capacity of the vent only filled; in that case not much heat would be supplied to the surface, because the quantity contained within the vent would be small.

But in this example, as in all others, as is well known to ordnance inspectors, the fracture began to exhibit itself on the interior surface of the bore.

It is often noticed as a curious phenomenon when large guns burst, that notwithstanding the chase or forward part of the gun, several feet in length, may be thrown many feet end over end, the shot passes through the chase the length of the bore without being diverted from the direction of its aim. This fact corroborates the theory under consideration, as it is evident that the shot is not projected by the same force that bursts the gunthe communication of heat to the inner metal of the gun requiring a longer interval of time, and gun metals being comparatively non-conductors of heat. In Rodman, plate II., fig. 2, is shown the interior line of fracture of a 10-inch columbiad. Here a thin bit of metal, indicated by the line marked., is shown, which seems nearly to envelop the bore. Nearly one-half the re-enforce was broken off this gun in the same manner as

Fig.15

chips break off a stone door cap when a building is burning, but in this example the outside of the stone is first heated while the inside remains

colder. The outward pressure of the powder at the time of this fracture would surely have carried away so thin a piece of metal; but it remains standing to show that the pressure against the surface had been reduced before the gun broke-a remarkable evidence of the true cause of the bursting of the gun. The diagram exhibiting the place and quantity of heat shows but little heat at any of the surfaces of the gun. From this, also, we may have been hitherto deceived as to the importance of the study of its effects; and we can only appreciate it by some experiments like the following: A clean rifled musket, the barrel of which weighed about five and a quarter pounds, was fired twelve times with the ordinary charge, at intervals of five minutes between each discharge. The time during which the surface of the musket was radiating away the heat from beginning to end was, therefore, about one hour. At the end of this time its temperature was 200o. The radiation was somewhat hindered by the wood of the stock, which was a non-conductor, partially enveloping the barrel, and the burnished surface of the barrel, which was a non-radiator. The whole amount of powder was less than one ounce, and it communicated this great amount of heat to five and a half pounds of metal. There would be a material difference in the amount of heat communicated in this experiment, if the barrel were not clean inside, as the residuum of powder would be a non-conductor, and would prevent its communication to the metal of the barrel. The temperature of the gases in a large gun, say 100-pounder rifled cannon, would be much greater than in a musket; as the temperature is increased as the resistance to the expansion of the powder is increased. The work of the powder in a gun is to overcome the inertia of the shot, and to do this it presses against a certain number of inches of area. If the shot be short, the pressure is still exerted against the same area. The projectile in a 100-pounder rifle gun is about 12 inches in length, while the projectile from a common rifled musket is less than one inch in length. The resistance from the inertia would be thus about twelve times as great in the large gun as in the small one, and the expansive force or pressure, and consequently the temperature, high in proportion.

The ordinary meters, if used to measure the temperature communicated to the gun, as shown by the preceding argument, will be inefficient, as they cannot be applied at the place supposed to be the seat of the highest temperature. A meter for this purpose can be prepared in the following manner: Fig.17.

WATERS & SON.ENG

Take a gun with eight inches thickness of metal about the bore, cut off the chase at that point of its length where the metal is five inches in thick

ness, then lay out the face of the muzzle with eight radial lines, equal distances apart, and, beginning upon one, drill a hole parallel with the bore, half an inch in diameter, coming out at the breech and leaving half an inch of metal between the bore and the hole thus drilled; then drill another hole on the next radial line through lengthwise, one inch from the bore, and so on, each hole being half an inch further from the bore, until the outer one is four inches from the bore. Fit bronze rods into these holes and fasten them at the breech with screws, so that they can have no metion endwise at that end, then file off the ends of the rods flush with the muzzle, when the gun and the rods are at the same starting point of temperature, say sixty degrees, and we shall have a thermometer that will give nearly a correct indication of the quantity of heat communicated to the gun from a calculation based upon the difference of expansion of the metal of the gun and the bronze rods, and a positively correct indication of the place of highest temperature, and consequently greatest expansion, with any number of discharges. To retain a record of the place and quantity of heat at any of the successive discharges, make a number of molds, and fill them with a composition of wax and powdered charcoal with which to take an impression of the face of the muzzle. These molds should be numbered and recorded at each successive discharge. Charges should be made with heavy and light projectiles, and with no projectiles at all, and at long and short intervals between the discharges.

The deductions drawn from these experiments would give us positive knowledge on this part of the subject, and relieve it of the mystery so often referred to by ordnance inspectors; while without the knowledge thus attained, officers of our army and navy seem to be without justification, should they place large guns in our expensive iron-clad ships and fortifications, made without consideration of the important cause of failure herein presented.

When a long rifled cannon is fired at a high elevation, the gun recoils backward on a plane, represented by the deck of a ship, different from the plane of the bore. All bodies in motion resist a change of

direction, in the proportion of 1-90th of their whole momentum, or living force, for a change of direction of 1°. If one billiard ball on a table is projected against another at rest, striking it at right angles 90°, the one in motion comes to a state of rest, giving its whole momentum off, to the one before at rest. If the one at rest should be struck at an angle of 45°, the ball in motion would have its direction changed 45o and it would give onehalf its momentum to the other. Each would roll the same distance on the table. So, also, if the angle with which they came in contact was one degree, 1-90th of the momentum would be given to the ball at rest. The whole sum of the momentum of a shot projected from a rifled cannon is very great. At the muzzle of the gun, the resistance to a change of direction is sufficient to overcome the preponderance of the gun. If the bore was crooked, the shot would not be much diverted, but the gun would be

moved to conform to the direction of the shot; and many have noticed, when firing guns on ship-carriages, at high elevation, that the breech of the gun was raised, and came down again with a considerable blow on the quoin, or elevating screw. If the chase of the gun is light, the muzzle will

sometimes be broken off, instead of overcoming the inertia of the gun, or lift. ing the breech suddenly, against the resistance of the preponderance. This example is inserted as one

of the peculiarities of living force, as exhibited in gunnery, viz: resistance to changes of direction by bodies in motion, and to account for the failure of many of the Dahlgren and Parrott guns, from the breaking off of their muzzles, as has frequently happened to the Parrott rifle guns, and to the Dahlgren guns since the war began.

From the fact that solid cast guns, of the largest size now in service, have a certain strain upon them within themselves when cast, from the heat leaving the inner metal last, which is relieved by the expansion of the inner metal by the first few discharges, I hold that solid-cast Dahlgren guns, or any columbiads of large sizes, cast solid, may pass the inspection of ten service charges, and be stronger at the tenth discharge than they were at the first-that number of rounds, perhaps, being necessary to relieve them of the beforementioned strain, by communicating the proper proportion of heat to place them in the same state in which we find the hollow cast gun at the first round.

The guns in our service having great thickness of metal about the bore, should not be relied upon in rapid firing, even when exposed to the hottest rays of the sun on their very large exterior surface-the most favorable circumstances under which a gun can be fired-and should never be fired at all, if a hollow cast gun with uniform density throughout the mass, in rain or in cold weather. It may sometimes happen that a hollow cast gun, after the Rodman plan, would exhibit greater endurance than a solid cast gun, made from the same metal and at the same time. At the time of the bursting of two steel 50-pounder navy guns of my fabrication, each at the ninth round, at Staten Island, I suggested to the inspector either that the guns should be fired at longer intervals between the discharges, or that I should be permitted to give elasticity by drilling a series of small holes about the bore, having a certain position relative to each other, and a Fig 19

proper direction to permit the expansion of the inner metal without any undue strain upon the re-enforce. Captain Rodman's book, page 297, exhibits the impossibility of casting a solid projectile, cavities being formed in the center of the mass, due to the shrinkage of the inner metal

Fig 20

after the outer shell had frozen, so as to prevent any supply of metal to the center thereafter; and this is related as the cause which led to the hollow mode of casting. These cavities do not occur of necessity in the center of a casting, but of necessity in the center of the mass at equal distances from the cooling surfaces if subjected to an equal rate of cooling; and they Fig. 21

are to be found near the center of the mass in the Rodman gun as well as in any other. Their presence cannot be detected at any of the surfaces of the casting. If

they were generally distributed between the inner metal of the gun and the re-enforce, a sufficient elasticity between these parts of the gun might be had, and a similar result arrived at to that obtained by the drilling of the holes, as shown by diagram. Their presence would account for the occasional endurance of hollow cast guns, but the uncertainty of their uniformity would account for the irregularity of this endurance. The Rodman 15-inch gun, fired at Fortress Monroe, stood the indifferent test to which it was subjected, perhaps, from the occurrence of this uncertain elasticity, and from the fact that it was fired under a hot sun, with slow-burning powder, hollow shot, and windage. Even this questionable success would rarely again be obtained, as the requirements of service are unlikely to permit such favorable circumstances.

The 15-inch guns have been shown to be inefficient, therefore, for they only give a velocity of 750 feet per second; or they are unreliable (perhaps both), for they will burst as often as the accidental porosity above spoken of is not evenly distributed between the inner metal and the re-enforce. And who can say when these conditions are all fulfilled?

I here leave the subject to be considered by those who have followed me thus far, with the remark, that I have shown that no large gun yet made could be pronounced trustworthy, until it was destroyed. Yet I believe that it is possible to so construct a gun, that not a single trial shot need be fired for the purpose of demonstrating its valuable qualities. That through the light I have thrown on the subject, trustworthy guns can be constructed of any required size, and to give almost any required velocity to shot, less than the unrestrained velocity of the gases along the bore.

When I proposed to show that the heat evolved from the combustion of powder in a gun was the principal agent in bursting, I found ordnance officers opposed the theory, on the ground that but little heat was thus communicated to guns. This was said to me by Major Wade, by Captain Rodman, and indirectly by several others occupying high positions. Major Wade, however, afterwards told me he recollected having seen copper melted by the heat of powder, at a time when a few kegs of powder exploded in a powder-mill. Captain Rodman, in a part of his book before quoted, speaks of "higher temperature exhibited when larger masses of powder are burned in an 11-inch gun, than in a 7-inch," as accounting for three times the force in the larger gun than in the smaller; and Captain Dahlgren, in his report to the Secretary of the Navy for 1862, says "an 11-inch gun was so heated by firing that it was afterward eighteen hours