502 



NATURE 



[September i8, 1890 



The weight of the engines of the Duke of Wellington and the 

 Ccesar would be approximately 400 tons and 275 tons, while 

 that of the torpedo boats is about 34 tons. 



But if these results are sufficiently remarkable, the economy 

 attained in the consumption of coal is hardly less striking. 



The consumption of coal in the early steam battle-ships was 

 from 4 to 5 lbs. per indicated horse-power per hour, and occa- 

 sionally nearly reached 8 pounds. 



At the present time in good performances the coal consump- 

 tion ranges from ij to if lbs. per indicated horse-power per 

 hour under natural draught, and from 2 to 2\ lbs. per hour with 

 forced draught. 



In war-ships the engines are designed to obtain the highest 

 possible power on the least possible weight, and this for a com - 

 paratively short time, and, further, have to work at such various 

 powers, that the question of economy must be a secondary con- 

 sideration. 



With the different conditions existing in the mercantile marine, 

 more economical results may be expected, and I believe I shall 

 not be far wrong in assuming that in special cases i^ lbs. may 

 possibly have been reached ; but I have not been able to obtain 

 exact information on this head. 



Turning now to the guns, let me refer first to those which 

 were in use thirty-five years ago, and which formed the arma- 

 ments of the ships of those days, and of the fortresses and coast 

 defences of the United Kingdom and colonies. 



The whole of these, with the exception of a few very light 

 guns, were made of cast-iron. I have already alluded to the 

 small amount of machine work (not of a very refined character) 

 expended on them. Although the heaviest gun in use was only 

 a 68-pounder, there were no less than sixty different natures of 

 iron ordnance. Of the 32-pounder alone there were as many as 

 thirteen descriptions, varying in length and weight. Of these 

 thirteen guns, again, there were four separate calibres ranging 

 from 6 "4 1 inches to 6-3 inches, and as the projectile was the 

 same for all, the difference fell on the windage. This varied, 

 assuming gun and projectile to be accurate, from about 0'I25 to 

 0*250, so that it may easily be conceived the diversity of the 

 tables of fire for this calibre of gun were very great. And 

 although from the simple nature of the guns, and the absence of 

 anything like mechanical contrivance connected with them, it 

 was quite unnecessary to give to them the care and attention that 

 are absolutely indispensable in guns of the present day, it must 

 not be supposed that they were altogether free from liability to 

 accident and other defects. 



I had occasion recently to look into the question of the guns 

 employed in the siege of Sebastopol, and found that in that 

 great siege no less than 317 iron ordnance were used by this 

 country. At the close of the siege it was found that 8 had 

 burst, loi had been condemned as unserviceable, while 59 were 

 destroyed by the enemy's fire. 



The 95 cwt. 68-pounder gun seems to have been about the 

 largest gun that could safely be made of cast-iron, and that in it 

 the limit of safety was nearly reached, was shown by the fact 

 that a serious percentage of this calibre iDurst or otherwise failed. 

 With the spherical shot the column of metal per unit of area to 

 be put in motion by the charge was small, and to this the guns 

 probably owed their safety. 



When the same charge was used, and cylinders representing 

 double, treble, or quadruple the normal weight of the shot were 

 fired, the end was rapidly reached, the guns frequently bursting 

 before cylinders four or five times the weight of the shot were 

 employed. 



But the fact that a stronger and more reliable material than 

 cast-iron was necessary, was shortly to be emphasized in a much 

 more striking manner. The great superiority of rifled to smooth- 

 bored ordnance in every respect, in power, in range, in accuracy, 

 in destructive effect of shrapnel and common shell, was in this 

 country demonstrated by Lord Armstrong and others. This 

 led to numerous attempts to utilize cast-iron for rifled ordnance. 

 The whole of these efforts resulted in failure. Although the 

 charges were feebler than with smooth-bored guns, these ex- 

 perimental guns burst one after the other with alarming rapidity, 

 generally before many rounds had been fired. The matter was 

 not made much better when the expedient was adopted of 

 strengthening these guns by hoops or rings shrunk on externally. 

 Failures with this arrangement were little less frequent, the 

 cast-iron bursting under the jackets, and the only plans in 

 which cast-iron was used with any success were those proposed 

 respectively by Sir W. Palliser and Mr. Parsons, who inserted, 



NO, 1090, VOL. 42] 



the one a coiled wrought-iron, and the other a steel tube in a 

 cast-iron gun block. 



But the country that suffered most severely from the use of 

 cast-iron was the United States. Their great civil war took 

 place just when efforts were being made in every country to 

 introduce rifled artillery. Naturally every nerve was strained 

 to manufacture these guns, and naturally the resources that came 

 most readily to hand were first employed. 



A report presented by the Joint Committee on Ordnance to 

 the United States Senate in 1869 gives the history of these 

 guns, which were nearly all either cast-iron or cast-iron re- 

 inforced with hoops in the way I have described. I have heard 

 the existence of internal strains disputed, but in this report we 

 read that ten guns burst, that is, flew to pieces, when lying on 

 chocks, without ever having had a shot fired from them, and 98 

 others cracked or became ruptured under like conditions. 



In the "Summary of Burst Guns" in the same report, it is 

 stated that 147 burst and 21 were condemned as unserviceable ; 

 29 of them being smooth-bore and 139 rifled ordnance. But 

 perhaps the most striking passage is that which relates that in 

 the action before Fort Fisher all the Parrott guns in the fleet 

 burst, and that by the bursting of five of these guns during 

 the first bombardment, 45 men were killed and wounded, 

 while only 11 men were killed or wounded by the enemy's 

 fire. 



The muzzle velocity given by the smooth-bored, cast-iron 

 guns may be taken approximately at 1600 f.s., and at the maxi- 

 mum elevation with which they were generally fired their range 

 was about 3000 yards. The 32-pounder, with a charge of one- 

 third the weight of the shot and an elevation of 10°, gave nearly 

 2800 yards, and the 68-pounder, with a charge of about one- 

 fourth, nearly 3000 yards. The same gun, with an eccentric 

 shot, and an elevation of 24°, gave a maximum range of 6500 

 yards. 



But it must not be supposed because the range tables gave 

 3000 yards as practically the extreme range of the ordnance of 

 35 years ago, that our guns possessed any high efficiency at that 

 distance. At short distances, from 300 to 500 yards, dependent 

 on the calibre, the smooth-bored guns were reasonably accurate, 

 but the errors multiplied with the distance in so rapidly increasing 

 a ratio, that long before a range of 3000 yards was attained the 

 chance of hitting an object became extremely small. 



It is desirable to give some idea of the accuracy, or, rather, 

 want of accuracy, of these guns. \ 



In 1858 I was appointed secretary to the first Committee on 

 Rifled Cannon, and the early experiments showing how extra- 

 ordinary was the accuracy of the new weapons, it became a 

 matter of importance to devise some means of comparing in this 

 respect the old and the new guns. 



The plan I proposed was one which has since been followed 

 by the artillerists of nearly all countries. It was to calculate 

 the probable error in range and the probable error in deflection, 

 and from these data the area within which it would be an even 

 chance that any given shot would strike ; or, in other words, 

 that area within which, out of a large number of rounds, half 

 that number would fall. This area was for the smooth-bored 

 gun at a range of 1000 yards, I47"2 yards long by 9"i yards 

 broad, or 1339 '5 square yards, while the similar area for the 

 rifled gun at the same range was 23'i yards long by o"8 yards 

 broad, or an area of 18 '5 square yards. But the great decrease 

 of accuracy due to an increase of range with the smooth-bore 

 guns is especially remarkable. Experiments showed that with 

 the smooth-bored gun an increase of range of only 350 yards 

 more than doubled the error in deflection, and made the area 

 selected for comparison 206 yards long by 20 "2 broad, or 4161 

 square yards, as nearly as possible trebling the area for an 

 increase in range of 35 per cent. 



But I have not done yet. These experiments were made 

 with the same lots of powder carefully mixed, and the irregu- 

 larities in velocity would be such as are due to manufacturers* 

 errors only. But the variations in the energy developed by the 

 gunpowder employed have still to be considered. In i860, 

 being then an associate member of the Ordnance Committee, I 

 carried on for the Government the first electro-baUistic experi- 

 ments made in this country. My attention was early called to 

 the great variation in energy developed by powders recently 

 made and professedly of the same make, and I pointed out that 

 in my experiments the variations between one lot of powder and 

 another amounted occasionally to 25 per cent, of the total energy 

 developed. It is unnecessary to say that on service, and when 



