if the band were forced on with the same tension as I obtain by putting it on hot, would deprive it of the expected accuracy.

Question 7. What is your opinion of the premature explosion of shells, and its effect upon the gun?

Answer. In rifle cannon there are many new elements and things to learn, which have not yet been solved by direct experiment. I think that the premature explosion of shells is the great cause, if not almost the only cause, of the bursting of these guns. Sand or dirt would wedge the shell in the gun, the windage being so small, which I have no doubt caused a few guns to burst at Morris island.

Question 8. What security have you that the shells are not frequently crushed in the gun by the shock of the discharge. independently of any premature explosion of the powder in the shells? Have no unloaded shells ever been broken in the guns?

Answer. I have no evidence that the shells want strength, after the experience of many thousand fires, to stand the direct shock of the discharge. No unloaded shells, so far as I recollect, have ever broken in the guns. A great many unfilled shells have been fired, and none broken in the gun. One hollow shot did break; but in these projectiles the cavity is in the rear end, and the base not quite so strong as that of the shells. Bands of shells sometimes fly off, or break, but I do not think that this endangers the gun. I think that imperfect forces may have sometimes caused accidents, but I have no knowledge that the spelter rings have ever failed; though, as first made, they might have been a source of accident by flame getting down alongside the thread of fuzehole into the shell. I have never known, after hundreds of fires, a fuze of any kind to be driven into the shell.

Question 9. What advantage do you expect to derive by an increasing twist in the rifling, over the regular or uniform twist? Is there any danger of wrenching off the muzzle by the "nip" the projectile receives at that point?

Answer. The principal advantages to be gained by the increasing twist are that the projectile takes the grooves more readily, that a higher rotation may be more easily obtained, and a stronger band used with the projectile. Bands of hard metal take the grooves fairly, which they would not do so well with a regular twist. If the groove is straight, the band enters it directly; but if curved, it has a tendency to ride over the grooves before complete expansion. No gun has ever broken at the muzzle, except by the premature explosion of a shel, within my knowledge.

Question 10. Could the weight of the projectile, and the charge of powder for the 100-pounder, be reduced, say to eighty pounds for the former and eight pounds for the latter, without mater ally diminishing the efficiency of the guns, and would not such reductions increase materially the endurance of the gun?

Answer. The short shell is a good projectile, and its use would not materially diminish the efficiency of the gun, while the endurance would certainly be increased by such reductions. The heavy shell, and ten-pound charge, give greater power, of course, and in cases where very long range and great shell power are required it may be necessary to use them, and they were designed for such purposes.

Question 11. Are Parrott shells cast with a hole in the base, for the purpose of more readily cleaning them? If so, by what means is the hole plugged, or closed?

Answer. The Parrott shells, above 60 pounders, are now cast with a hole in the base for steadying and centring the core in the flask. This hole affords facility for cleaning them, and is filled by a rivet, having a head, which effectually prevents them being driven into the shell; no instance of it having been driven in, in many hundred recovered after having been fired.

Question 12. May not premature explosions be caused by the crushing of shells in the bores of rifled guns, when the shells are porous?

Answer. I do not believe that premature explosions occur by the crushing of my shells within the bores. By the mode of casting them with the base downward, the liability to have porous metal at the base is removed. There is a solitary instance known to me of crushing a projectile within a gun (a hollow shot) in many thousands that have been fired.

Question 13. May not the imperfect adjustment of the fuze, the weakness of the spelter rings, or the imperfect manufacture of the time-fuze, be another cause of premature explosion?

Answer. I think that the original spelter ring (cast) did admit the probability of the gas entering the shell and exploding it; but those now made by being cut, and having a "shoulder," removes that probability. Imperfect fuzes, or imperfect adjustment of fuzes, would, of course, be liable to admit gases into a shell; but not more so in the Parrott than in any other shell.

I would respectfully submit to the board the following remarks as due to the importance of the subject and to my own position.

WASHINGTON, January 16, 1865.

R. P. PARROTT.

I am unable to find, on revising my correspondence, that I have failed to express, on all proper occasions, my opinion that most of the accidents to my guns have resulted from premature explosions of loaded shells taking place within the bore; although I believe that the sand blown, or otherwise accidentally carried into the guns, was an operating cause of bursting at some of the positions on Morris island.

But I do find that I have failed to give as decided an expression to my views as the importance of the subject required. The partial success of the trials made experimentally at this place to prevent premature explosions of shells, and the prospect that means entirely efficacious would soon be found, have induced me to look upon the difficulty as one which would shortly be remedied; and indeed I can say that it has been by the precautions now taken in firing loaded shells, which consist in lining the interior with a substance effectually covering the rough surface of the iron.

The great danger connected with these premature explosions is the injury which the gun may suffer from a single one, not, probably, fatal at the instant, but leaving the gun impaired, and causing its destruction under subsequent firing. Thus doubt and distrust are thrown upon the character of the guns, which burst without any cause assignable at the time of failure. I do not suppose that one or many shells could not explode in the bore of rifled cannon without serious danger to the gun, but have no doubt that it might be destroyed by a single explosion of a shell, and fear that irreparable damage would be caused by some one out of not a very large number of such accidental explosions.

I must say that the apprehensions 1 have expressed on this subject have been received with not a little doubt, and that I found most persons skeptical in regard to the causes of the explosion of shells in the bores of rifle cannon, and some as to the danger of such explosions, from whatever cause.

In referring the premature explosions of rifle shells to the friction or attrition to the powder contained in them, I do not, of course, deny that such explosions may sometimes be the result of defects of the shells or of the fuzes; but I do say that with ordinary care taken in respect to the inspection and other points, few, if any, of my shells will explode prematurely, except by the friction of the powder within; and that, if the interior surface of the shell is effectually covered, these explosions will cease.

What I state has been derived from actual trial, and is substantiated by facts officially noted in the proof and inspection of rifle guns.

It has often been asked why should the rifle shell explode prematurely more frequently than the spherical? It may be answered that if the rifle shell is charged with only the same quantity of powder as suffices to fill the spherical of the same calibre, explosions will rarely, if ever, take place. But as the rifle shells hold from three to four times as much powder as the spherical, the greater weight of powder, and that in a long column, must, by its reaction on the firing of the gun, press with much greater force, and by friction, either on the bottom or along the sides of the cavity of the shell, (if left rough as cast,) cause an explosion.

Such being the facts in respect to the explosions of rifle shells, and the means of preventing them, are they the causes of the unequal endurance and unlookedfor bursting of the guns?

I am unable to trace any connexion between the bursting of guns and the time of manufacture. The difficulties of procuring supplies of material and of labor have been unexampled within my experience; but I am not aware of any deterioration in the work, and feel confident that there has been none which can possibly account for the failure of guns, though the supposed necessity for assigning some cause, such as would be applicable to ordinary cannon, has led to the supposition that a gun which fails must have been bad from the beginning. It would be as unjust to expect of me to foresee all the difficulties which may arise in the use of a system of ordnance so new as that of rifle cannon, as it would be presumptuous on my part to pretend to do it. I cannot think, however, that the merits which mine have exhibited are merely accidental, but believe that, from the uniformity of plan and the results with the extreme sizes, no intermediate class can be wrong in principle. I know of no possible cause, other than the explosion of the shells, which would account for the bursting of my guns near the muzzle. It has been known to happen the very first round fired from the gun after a very few, after some hundreds; and in the two 10-inch guns destroyed in this way-one after twenty-seven, and the other after one thousand and four fires in actual service. In a very few instances it has happened with the 30-pounder guns after quite moderate use, and in one it did not take place up to four thousand six hundred and fifteen fires.

Again, as to those peculiar modes of bursting, by which portions of the castiron are blown off forward of the wrought-iron reinforce, or near the trunnions, after leaving all the rest of the gun still connected together, how can it be supposed that these accidents, occurring after very irregular periods of service, can be the results of any uniform and natural action of the charge? The explosion of so much additional powder as the shells contain, though undoubtedly adding materially to the strain upon the gun, might not seriously damage it; but as the base of the shell will probably be in one piece, with the ring connected, and the sides of the projectile driven laterally against the bore, it is not surprising that partial jamming of these should take place, and try the gun to a dangerous point. I have portions of the front or curved end of a shell burst in the gun, which are so marked as to show these fragments to have been violently forced into the grooves. Although I conceive that the failure of the guns at any part forward of the reinforce is certainly due to accidents occurring with the projectile, I cannot conclude that similar accidents might not cause the bursting of the gun in another part depending on the position of the shell at the time. In one instance, at Morris island, the base of the shell was actually found in the gun after the blowing off the breech.

In other cases the cast-iron may be so injured by previous explosions as, in bursting, to carry the band with it.

In conclusion, I would express my belief that I have correctly assigned the causes of the bursting of my heavy guns. I do not consider that they are less safe than ordinary cannon when subjected to the same regularity of strain.

On account of the very uncertain action of shells prematurely exploded in the bores of rifle guns, we are unable to determine the extent of injury, therefore, and cannot judge of the number of rounds which the guns can subsequently be expected to endure with safety.

At the same time I am satisfied that the means now used do effectually prevent the premature explosion of the shells, and thus remove the greatest cause of danger to the guns, it is due to the subject and to myself that I should advert to the firing of other projectiles than my own in my heavy guns. I cannot but think that many shells much inferior to those which the same makers would now supply have been used in my rifle guns, as well as many projectiles of an experimental kind.

The action of powder in the rifle takes place under circumstances very different from those existing with the spherical projectile, and, in my judgment, far too little consideration has been heretofore given to this point as connected with the durability of the guns.

R. P. PARROTT

NAVY DEPARTMENT, Washington City, June 30, 1865.

SIR: In accordance with the order of the Navy Department of the 12th instant, reconvening the board on rifle ordnance, (adjourned on the 18th January last, to await results of certain experiments suggested by said board, which have since then been conducted by Captain Joseph F. Green, under the particular instructions of the Bureau of Ordnance, hereto appended, and marked 1,) we have the honor to state that we reassembled here on the 27th instant, after having conducted several additional experiments, and inspected the condition of the guns and projectiles used in the recent trials at Cold Spring, New York, under the orders of the Bureau of Ordnance.

It appears by the course of the experiments of Captain Green that one thousand (1,000) rounds were fired from each of three 100-pounder Parrott rifle guns of nearly similar weight, density, and specific gravity of metal, and cast nearly about the same period of time, which were selected by the chief of the Bureau of Ordnance, and assumed to fairly represent the guns of that class and calibre now in service. One-half or five hundred rounds from each gun were fired with full charges of ten pounds Hazard rifle powder and the long shell of 100 to 104 pounds weight, and the remaining five hundred rounds with eight pounds of same powder and shells of 80 pounds weight, by your special order to Captain Green.

From one of these guns (No. 242) 1,000 shells "coated" were fired.

From another of these guns (No. 256) 1,000 shells not coated were fired. From another of these guns (No. 239) 1,000 shells brought to weight by sand and sawdust were fired, and with the following results in premature explosions: Coated shells, (long,) 8 exploded within the bore; coated shells, (long,) 3 exploded without the bore; coated shells, (short,) 3 exploded within the bore; coated shells, (short,) 9 exploded without the bore; not coated, (long,) 7 exploded within the bore; not coated, (long,) 1 exploded without the bore; not coated, (short,) 4 exploded within the bore; not coated, (short,) 18 exploded without the bore; and the same number (viz: 1,000) were fired from No. 239, filled with sand and sawdust, for the purpose of testing the relative endurance of this class of gun, independent of premature explosion of shells, and also to test the resistance of the shells themselves to rupture under the effect of the charge of the gun.

Four thousand and eighty (4,080) Parrott shells of 20, 30, 60, 100-pounder

and VIII-inch calibre, have been fired since the 24th June, 1863, in the proof of navy guns at Cold Spring foundry. These shells were either not loaded or had only a blowing charge-in no instance did any one of the shells break. The army during the same time fired a larger number of shells with the same results.

From this it may be fairly inferred that the Parrott shells are never crushed or broken in the gun by the shock of the discharge.

Of all these premature explosions thirty-four (34) occurred with shells prepared with the navy metal-stock time fuze, and ten (10) with the navy time fuze, with the safety plugs removed, in Parrott's metal stocks, and eleven (11) with the Parrott percussion fuze.

The annexed table gives a synopsis of the report of Captain Green: