; 
 
 
 ■ 
 
 ili7Wiif&$iin$I: 
 
 '*.■-'..*.:.?■;:-■ ".■■;' i.-^v^v, :V:vv; ■■■■■;.:, : ..<-,-y,;\ 
 
Cornell University 
 Library 
 
 The original of this book is in 
 the Cornell University Library. 
 
 There are no known copyright restrictions in 
 the United States on the use of the text. 
 
 http://www.archive.org/details/cu31924030753440 
 
V 
 900 
 
 T5 33 
 
ARMOUR, AND ITS ATTACK BY 
 ARTILLERY. 
 
 BY 
 
 Charles Orde Browne, 
 
 CAPTAIN LATE ROYAL ARTILLERY, 
 Lecturer on Armour to the Boyal Artillery College, Woolwich. 
 
 Woolwich : 
 Pbiwtbd at thb Eotai Abtimeby Ihstiiption. 
 
 LoHDOH : 
 PUBLISHED BY MEBIB8. DuliU & Co., 37 SOHO SftViBB, LOHDOIT, W.C. 
 
 1887. 
 

 /h %y//s 
 
 , ■' r- ex ~< v r I \ \ 
 
 5 !' if i 
 
 V 
 
 
 . ^ 
 
 \ '- * - 
 
PREFACE. 
 
 This work has already appeared in the form of papers contributed to the 
 " Proceedings of the Eoyal Artillery Institution ;" it is intended to facilitate 
 instruction on the subject of armour and its destruction by artillery, and also 
 to serve as a work of reference. With these objects in view, a list of contents 
 and full index are supplied. 
 
 The work has grown out of lectures given by the author to the officers of 
 the Eoyal Naval College at Greenwich, to the Advanced Class of Artillery 
 Officers at Woolwich, and to the School of Gunnery at Shoeburyness. 
 
 The bulk of such matter as has originated with the author first appeared 
 in the form of papers and reports on experiments written by him for the 
 Engineer. The original wood blocks cut for these have been very kindly 
 placed at the author's disposal. Many of these illustrations may now be found 
 in English and foreign works, notably in Lord Brassey's " British Navy," Vol. II., 
 and " Naval Annual," in "King's War Ships and Navies of the World," and 
 Lieut. Very's "Development of Armor," Lieut. Jaques reports in the "Proceed- 
 ings of the U.S. Naval Institute," besides Continental Professional Papers. 
 
 Everything connected with armour has been determined by experiment. It 
 follows then that the greater part of this work is devoted to tracing the develop- 
 ment of armour and the means devised for its destruction through courses 
 of experiments conducted at Shoeburyness, Portsmouth, Spezia, Meppen, and 
 elsewhere. In many cases the sketches of results were made by the author 
 on the ground, in others photographs are copied. For his opportunities of 
 witnessing experiments and obtaining information, the author is indebted to the 
 Naval authorities, the Koyal Engineers, the Italian Government, Herr Krupp, 
 and particularly to officers of his old Corps, the Koyal Artillery. It will be 
 found that English trials occupy the principal place, and the author has to 
 regret that he has nothing to offer on American experiments and but little on 
 those of France. Happily, Very's admirable work is strong in these respects. 
 
 On many questions information is obtained from works which are quoted 
 verbatim, often at considerable length, the writers names being of course given. 
 Chapter III., Part II., consists of a reprint of the abbreviated report of the 
 Bucharest Turret trial made by Major O'Callaghan, E.A., and Capt. Clarke, 
 R.E. 
 
IV. 
 
 The work is divided into two parts, namely, (I.) Armour Experiments. 
 (II.) Armoured Structures, with short notes on manufacture and on results 
 produced in action. This division enables additions to be made to each part 
 without disarranging the numbering of the pages. 
 
 The author makes no attempt to predict what shape armour may take or 
 whether even it may eventually disappear. The latter contingency does not at 
 present appear likely, although a parallel might be drawn between the history 
 of body armour and that for ships. In both instances entire casing has been 
 given up, as the attack of fire-arms became too powerful for it. In both 
 instances partial armour of a superior character has been retained to protect 
 parts of the most vital importance, that is the head and chest of the soldier, 
 and the engines, magazines, and water-line of the ship. The French Curassiers 
 at Waterloo have been described 1 as appearing like men in complete armour 
 with closed visors as they bent forward and charged up the hill, just as such a 
 vessel as the Shannon is practically an armour clad when she engages head on to 
 an enemy. The parallel is indeed very complete ; for the protection afforded to 
 the ship by the sea, and by contrivances, corresponds to that afforded to the 
 soldier by the horse's head and saddle. 
 
 The author concludes by expressing the hope that must be shared with him 
 by every thoughtful christian man that the action of armour may long remain 
 chiefly the subject of scientific experiment, and but seldom be illustrated by 
 the operations of war. 
 
 1 Set Major Maoready's account — Creasy's " Fifteen Deciiive Battles.' 
 
CONTENTS. 
 
 PART I. 
 
 CHAPTEE I. Brief History of early experiments in America and England. — 
 
 French and English floating batteries in 1855. — Warrior target. 
 
 Whitworth steel shot. — Early French, English, and American armour- 
 plated vessels. — Cowper Coles' turret. — Armstrong 80-pr. — Experi- 
 mental backings. — Jones' inclined plates. — Trusty cupola. — Special 
 Committee on Iron. — Sir W. Fairbairn's investigations. — Adoption 
 of soft iron. — Palliser shot. — Laminated armour. — Bolts. — Charmer's 
 Bellerophon and Hercules targets. — Casemate shields. — Royal Sovereign 
 Turret. — Plate-upon-plate armour. — Air spaces. — pp. 1-9. 
 
 CHAPTER II. Introduction of Formulae. — Punching and Hacking. — Perfora- 
 tion. — Behaviour of various projectiles. — Formulae employed at first. — 
 Total Energy " E" and energy per inch circumference " e." — General 
 Inglis on Plate-upon-plate armour. — pp. 10-20. 
 
 CHAPTER III. Calculation and Estimation of Powers of Guns against 
 Armour. — Soft and Hard armour. — Perforation. — Colonel W. H. 
 Noble's Report. — Memorandum by Sub-Committee on Plates and 
 Projectiles in 1881. — Diagrams of General Inglis and Colonel Maitland. 
 — Formulae of Inglis, and also of Maitland, of Andrew Noble and of 
 English. — Rule of Thumb. — Practical Directions. — Table of four- 
 figure logs. — Example worked out. — Hard Armour. — Table of results. 
 — Difficulty in calculation for Hard Armour. — pp. 21-38. 
 
 CHAPTER IV. Experiments with soft Armour. — Prussian trials, 1868. — 
 Russian Hercules, 1869. — Filled shells exploded by impact when 
 stored behind armour. — Ships Deck Targets. — Water Experi- 
 ment. — 3 5 -ton Gun. — Finspong shot. — Qlatton Turret Trial. — 
 Whitworth B.L. 9-pr. at Southport. — Tegel experiment, 1872. — Wet 
 guncotton in shells, and action of charges on deck plates. — Shannon 
 Target. — 38-ton and 80-ton Gun, unchambered and chambered, fired 
 against sandwich targets, Nos. " 40 " and " 41 " — Common Shell 
 fired at Armour. — Solid plate attacked by 38-ton gun. — Air space. 
 — Krupp 12-in. gun.— pp. 39-62. 
 
VI. C0NTENT9. 
 
 CHAPTER V. Spezia Trials and steel-faced Armour .—Spezia Trials, 1876 — 
 Behaviour of Steel Plates under attack of Elswick 100-ton M.L. 
 gun. — Steel-faced (Wilson's) Armour. — Nettle Trial of Steel and Steel- 
 faced Armour. — pp. 63-72. 
 
 CHAPTER VI. Proceedings of the Sub -committee on Plates and Projectiles. 
 Competitive steel and iron shot. — Conclusions and Recommendations. — 
 Guncotton. — Wrought-iron Cap. — Investigation of Laws of Perforation 
 with New Type Guns.— pp. 73-86. 
 
 CHAPTER VII. Experiments icith Hard and Soft Armour. — Dutch Experi- 
 ments. — Spezia Trials of 1879 with Steel 27'6-in. Plates. — Krupp's 
 Meppen Trials. — High-angle attack of Inflexible Deck. — Gun muzzle 
 pivoting in Shield. — Chilled and Wrought-iron Shields. — Dillingen 
 sandwich Armour. — Nettle Steel-faced plate Trials. — 18-in. Steel-faced 
 plate at Shoeburvness. — Meppen Trial, 1882. — Direct and oblique 
 attack of Iron Armour. — Palliser Improved Projectiles. — Cammell 
 (Wilson) and Brown (Ellis) Steel-faced Plates tested on Nettle and 
 at Shoeburvness. — Major O'Callaghan on Steel under fire. — pp. 87-116. 
 
 CHAPTER VIII. Experiments with Hard and Soft Armour. Spezia competi- 
 tion, 1882. — Schneider Steel and English Steel-faced Armour. — St. 
 Petersburg (Ochta) competition. Palliser improved Projectiles. — 
 Engineer Experiments on Shields fixed on masonry at Shoeburvness. — 
 Wliitworth Southport experiment with 9-in. forged Steel projectile. — 
 Deck targets attacked at Eastney. — Copenhagen (Amager) competition 
 of Steel and Steel-faced Armour. — Hadfield Steel and Elat-headed Steel 
 Projectiles, 1884. — Spezia competition. — Steel and Steel-faced Plates 
 under attack of Elswick 100-ton B.L. gun, 1884. — Shoeburyness 
 Trials, 1885. — Admiral Class Armour. — Elswick (Ridsdale) trial, 
 1885. — Review of Shoeburyness Trials. — Powers displayed by Steel 
 and Steel-faced Armour. — pp. 117-158. 
 
 FART II. 
 
 CHAPTER I. Gruson's Chilled-Iron Armour. — Object and nature. — Pirst trial 
 at Tegel in 1868, and Perm, 1871.— Adoption in 1873.— Trials in 
 1874.— Adoption by Continental Powers.— Cupolas and Batteries.— 
 Kromhout on Cupolas.— St. Marie Battery— Buckau experiments, 
 1883-1884.— Attack by Krupp and Ternitz Steel Projectiles.— 
 Projectiles with concave ends. — pp. 1-16. 
 
CONTENTS. VII. 
 
 CHAPTER II. Trial of Griison's Armour at Spezia by 100-ton B.L. Elswick 
 gun, April, 1886. — Approval of Shield. — Behaviour of Krupp and St. 
 Chamond Projectiles. — Further trial of projectiles against Griison's 
 Shield. — General form of Turrets for defence of Spezia Harbour with 
 Griison's Armour. — Krupp's 119-ton guns and Elswick Machinery. — 
 pp. 17-34. 
 
 CHAPTER III. Bucharest Cupola Competition. (Abbreviated Report of 
 Major D. O'Callaghan and Captain Clarke). Object of Cupolas. — 
 Diary of Experiments. — German Cupola. — General Description. — 
 Results of Trials. — General Remarks. — French Turret. — General 
 Description. — Results of Trials. — General Remarks. — General Con- 
 clusions on Emplacement. — Mountings. — Mounting and Dismounting. 
 — Ventilation and Lighting. — Loading. — Elevation. — Traversing. — 
 Sighting. — Accuracy of Fire. — Rate of Fire. — Resistance. — Damage 
 to Interior. — Glacis Plate. — Visibility. — Embrasures and General 
 Conclusions. — pp. 35-72. 
 
 SUPPLEMENT TO CHAPTER III. Improved Mougin and proposed 
 disappearing French Turret. — pp. 73-78. 
 
 CHAPTER IV. Armoured Structures, Forts. — General Inglis on British Coast 
 Ports. — Casemate Shields in Masonry. — Shields in Masonry for guns 
 on Turntables. — Iron Batteries. — Iron Turret. — pp. 79-84. 
 
 CHAPTER V. Brief Summary of behaviour of various kinds of armour. 
 Wrought-iron. — Steel-faced iron. — Solid steel. — Chilled-iron. — 
 Backing and Bolts. — pp. 84-86. 
 
 CHAPTER VI. Brief notes on Manufacture. Wrought-iron armour. — Steel- 
 faced wrought-iron. — Wilson's and Ellis' systems. — Compound steel 
 plates. — Behaviour when tested. — Quality of metal for back of plate. — 
 Solid Steel Plates. — Schneider's plates tested at Gavre. — Terrible and 
 Admiral Baudin Plates. — Through Cracks and Bolts. — Spontaneous 
 fracture of Steel. — Comparison of Solid Steel and Steel-faced plates. — 
 Kinds of armour employed by each Naval Power. — Projectiles. — Cast- 
 iron. Chilled cast-iron. — Supersession of chilled-iron by steel, 1886. — 
 
 Steel Projectiles Cast and Forged.— pp. 86-103. 
 
 CHAPTER VII. Armoured Structures, Ships. — Broadside old-fashioned with 
 guns distributed. — Minotaur, &c. — Masted Sea-going Modern Types. — 
 Admiral Duperre. — Protected Ships. — Almirante Cochrane. — Mastless 
 Turret and Barbette Ships. — Italia.— Inflexible. — Builio, &c. — Admiral 
 class. — Dreadnought class and Polyphemus class. — pp. 103-114. 
 
VIII. CONTENTS. 
 
 CHAPTER VIII. Coal Protection. — Experiments conducted by Gordon's 
 committee on Ordnance on penetration of coal with and without 
 plates. — Conclusions and recommendations. — pp. 115-117. 
 
 CHAPTER IX. Remits produced hi, action — Early ships in American War. — 
 Huascar, Cochrane and Blanco. — Effects of fire on Huascar. — Lieut. 
 Madan's Report. — Naval attack of Alexandria Forts. — Ordnance and 
 armour of ships engaged, and ordnance of Forts. — Effect of Fire. — 
 Conclusions expressed by American Official Report and by Major 
 Walford and Admiral Le Hunte Ward. — pp. 117-125. 
 
 ERRATA. 
 
 .p. 37, Part I.— Footnote :— 
 
 For " Admiral Fishbourne " read " Admiral Selwyn." 
 p. 98, Part I., line 29 :— 
 
 For " shot or " read " shot on." 
 p. 139, Part I., line 1 :— 
 
 For " Palliser steel shot," read " Palliser chilled shot.' 
 
SHORT HISTORY AND DESCRIPTION OF ARMOUR 
 
 ATTACK BY AETILLEEY. 
 
 CHAPTER I. 
 
 Brief History op Early Experiments. 
 
 On the history of the development of armour, Major-General Inglis, 
 R..E., 1 remarks that " every point, down to those of the nicest detail, 
 has been based on experimental results," and that " in no subject that 
 has ever been raised has mere opinion, unsupported by practical 
 experience, proved so worthless as in this." It is not surprising, then, 
 that the early experiments in many cases led to negative results ; that 
 is, they only proved the futility of the particular design or feature 
 under trial. It is hardly likely to be generally useful to notice most 
 of these experiments, which can best be studied in the printed pro- 
 ceedings of the Committee 3 by whom they were carried out. It is 
 only necessary here to mention a few that stand out as landmarks, 
 fixing the shape afterwards taken by the armoured structures them- 
 selves and the method of attacking them. 3 
 
 In 1812, John Stevens, of New Jersey, designed a ship with a 
 battery protected by inclined armour, and by 1841 his family, working 
 at the same subject, had determined the resisting power of iron against 
 the shot of the day fairly. 4 
 
 1 See " Notes on Armoured Defences," a paper read at the E.A. Institution on April 29, 1880, 
 by Colonel Inglis, E.E. 
 
 2 "Special Committee on Iron," 1861, 1862, 1863, 1864; also "Ordnance Select Committee 
 Proceedings." 
 
 3 These are chiefly taken from "Notes on Armoured Defences,'' by Colonel Inglis. Some 
 information on foreign and naval matters is quoted from " The Development of Armour for Naval 
 Use," by Lieut. Very, U.S. Navy. The "Proceedings of the Special Committee on Iron" are also 
 quoted. 
 
 * "Development of Armour," Very, p. 385. 
 
In 1827, 1 an experiment of an unsuccessful character was made in 
 Woolwich as to the resisting powers of masonry faced with wrought- 
 iron bars, on a proposal by Major-General Ford, R.E. 
 
 In 1840, some Admiralty experiments to test the action of shot 
 against iron-plates, backed by various substances, led to the conclusion 
 that iron was a bad material for ships of war. 
 
 In 1841, General Paixhan 3 recommended the use of iron-plates in 
 the United States. 
 
 In 1853, 3 masonry strengthened with iron was tried in the United 
 States. 
 
 In 1855/ the French employed three iron-cased floating batteries 
 against Kinburn, and before the end of the war between England and 
 France with Russia the English had iron-clad batteries afloat. 
 
 From 1856 to 1859, 6 further experiments against iron plates took 
 place: the 68-pr. smooth-bore gun attacking four inches of iron thickly 
 backed with wood. In 1858, a wrought-iron shot passed through four 
 inches of iron and wood backing. In 1861, a target representing the 
 first English iron-clad frigate, the Warrior, was attacked by a Whit worth 6 
 80-pr. rifled gun, firing a steel flat-headed shot. The Warrior target, 
 which consisted of 4i inches of wrought-iron, 18 inches of teak, and 
 f-inch iron skin was indented and cracked, but not perforated by this 
 shot, 
 traduction ^ s re g ar ds the actual introduction of armour into the construction 
 ofarmour of ships of war, Lieut. Very, with great justice, observes that the 
 struction of United States, which is often credited with taking the lead, owing 
 ships of war. probably to the part played by iron-clad ships in the American War, 
 and especially to the well-known encounter between the Merrimac and 
 the Monitor, was by no means the first. He observes: 7 "The orders 
 for the construction of the Monitor, Galena, and new Iron Sides — the first 
 iron-clads built for the United States — were issued in September, 1861. 
 Prior to this time, as has been shown, England and France had each 
 constructed a squadron of floating Batteries; these squadrons were 
 quadrupled in size, and rendered doubly powerful in individual ships 
 within the next four years. In 1858, the first squadron of sea-going 
 armoured frigates — Gloire, Norwaudie, Invincible, and Couronne — was 
 commenced in France, and scarcely were their keels laid when England 
 responded to the advance with the Warrior, the Blade Prince, the Defence, 
 and Resistance. Before the United States Congress had considered 
 the question of iron-clads, England, France, Spain, Italy, Austria, 
 
 1 " Report of Special Committee on Iron," 1861-1862, p. ix., &c. 
 
 2 " Development of Armour for Naval Use," Very, p. 365. 
 
 3 " Notes on Armoured Defences," Inglis, p. 1. 
 
 * See "Armoured Defences," Very's "Development of Armour," and "Proceedings of Special 
 Committee on Iron," 1861-62. 
 
 6 See "Armoured Defences" and "Proceedings of Special Committee," 1861-62. 
 
 6 See "Proceedings of Special Committee," 1861-62. A 13-pr. shot at the present time would 
 produce nearly the effect of this 80-pr. 
 
 7 Very, " Development of Armour," p. 367. 
 
Denmark, and the Southern Confederacy either had iron-clads afloat or 
 on the stocks. Before Ericsson had submitted the design of the 
 Monitor to the Naval Commission, Captain Cowper Coles had demon- 
 strated the advantages of the turret, mounted on low freeboard iron-clad 
 hulls, in public, to the naval experts of England {see ' Proceedings of 
 the British United Service Institution/ June 29, 1660). Before the 
 United States had closed the contract with Ericsson for the Monitor, 
 the Danes had made one with Coles for the double-turreted sea-going 
 ironclad Rolf Krake, the progenitor of the Huascar, and more closely- 
 resembling her than the Nantucket resembled the Monitor." 
 
 From about 1859 the manufacture of armour-plates progressed rapidly; 
 the plates from about this date increasing steadily in thickness. In 
 1860, an 80-pr. Armstrong gurj was fired against iron embrasures with 
 plates 8 inches and 10 inches thick. 1 In 1861, various backings were 
 tried, such as timber, cork, india-rubber, layers of wire, &c. Prom 
 these, it was concluded that "while the hard materials improved the 
 resisting power of the armour, they led to its being more injured by 
 cracking, and to the giving way of fastenings." 
 
 Jones' inclined iron plates were tried about this time; when it was Jones' in- 
 concluded that a given weight of armour will protect a given vertical p^tes. 1 ™" 
 area equally well, whether in the form of a simple vertical plate, or of 
 a thinner but necessarily larger inclined plate, the penetration of the 
 shot being proportional to the sine of the angle at which it strikes the 
 plate until the glancing angle is reached. 
 
 In 1861, the cupola of the Trusty, with 4,\ inches of armour, was 
 tried at Sheerness. It was in the form of a truncated cone. It 
 resisted fire fairly well, and its machinery was not damaged. 3 
 
 In March, 1862, the Special Committee on Iron 3 made their first Special 
 report, in which they state the following conclusions, among others: on™ron. 88 
 namely, that steel and steely-iron are bad materials for armour, while 
 soft, but not necessarily costly, iron is best; that corrugations and 
 bosses, designed to break shot on impact, are undesirable ; that plates 
 should be as large as practicable ; that hard backing supported the 
 plates at the expense of the bolts, whose functions are not only to hold 
 the plates on but also to resist vibration and prevent buckling ; that 
 tongueing and grooving 4 the edges of plates tend to spread injury 
 from plate to plate, and are bad ; and that the effect of shot on plates 
 
 * " Notes on Armoured Defences," Inglis, p. 3. 
 
 2 " Notes on Armoured Defences," Inglis, p. 9. 
 
 3 The Committee consisted of the following : — President Sir J. C. D. Hay, E.N. (then Captain), 
 Major Jervois, E.E., Colonel Henderson, E.A., Doctor Percy, Sir W". Fairbairn, and W. Pole, Esq. 
 In this report the Committee divided their work into three sections : (I.) The collection and 
 classification of results of experiments already carried out. (II.) The examination of witnesses 
 possessing special knowledge. (III.) The carrying out of fresh experiments. The author thinks 
 that few are aware of what valuable data and opinions are contained in the •' Proceedings '' of this 
 Committee. 
 
 A Lieut. Very, in " Development of Armour," gives a capital illustration of complicated fitting of 
 plate edges in the Galena, which succumbed under the fire of Port Darling, p. 380. 
 
4 
 
 is not proportional to the momentum of the former, but to the " energy/' 
 
 or " stored-up work," which may be expressed by — - — , where ^stands 
 
 for the weight, and v the striking velocity of the shot, and g the force 
 of gravity. 1 
 
 Sir W. Fairbairn made some very interesting experiments on 
 punching and crushing iron, which were submitted to the Committee ; 
 the results being given in an appendix to each of their three reports. 
 From them it appeared that a flat punch clipped out a disc, which was 
 more irregular in shape, and cleft or torn across where the support 
 was less complete. A round-ended punch met with less resistance 
 than a flat one at first, but before the perforation was complete it 
 experienced about double the resistance, making a round-ended 
 impression in the plate, and eventually driving out a bent disc torn 
 across the centre. In the case of plates, which were thick in proportion 
 to the diameter of the punch, a star or cross-shaped tear was made with 
 the centre opposite to the point of the punch, but a small disc-shaped 
 fragment of plate was generally detached at the centre of the cross. 
 Hemispherical-ended pieces of iron were crushed with half the pressure 
 necessary to crush flat-ended ones : namely, 26'82 tons instead of 
 55'36 tons per square inch. 3 
 
 It may here be noticed that the soft iron having been adopted, a 
 hard-pointed projectile was sooner or later almost certain to come in, 
 in spite of the advantages at first presented by flat heads. Such a 
 projectile would meet with but little resistance as its point entered, 
 and by the time that it was so far in the plate that the resistance 
 had approached the maximum, the head of the sbot would be well 
 supported, and its condition would be almost that of a cylindrical 
 shot driving a pointed wedge in front of it. Whereas had a very hard 
 plate been adopted the shot would meet with abrupt resistance directly 
 its point touched the plate, and would experience a mechanical strain 
 somewhat analogous to the outward thrust that falls on an arch. To 
 this is probably to be attributed the ease with which a hemispherical- 
 ended piece of iron is crushed. 
 
 Paiiiser The late Sir William Palliser seized on this idea and applied it to 
 
 great advantage by the introduction of ogival-headed chilled-iron shot, 
 to which he rightly gave sharper points than were afterwards approved 
 for the Service. The advantage of a very long sharp point is apparent 
 when it is borne in mind that the plate yields by bulging and tearing 
 
 1 Fairbairn suggested the following equation for punching:— t = /_ — in w hi cn t = the 
 
 v cm' 
 
 thickness of plate punched, W and v the weight and Telocity of shot, S its semi-diameter, and 
 a constant to be determined by practice. This is the same equation as 255f = »D x t» x i 
 
 where the energy of the shot is just equal to the circumference of hole made, multiplied into 
 the square of the thickness of plate and a constant determined by practice. Nothing better than 
 this equation exists at the present time for perforation. 
 
 2 " Transactions and Beport of Special Committee on Iron," 1862, p. 5. 
 
open at the back; for, in addition to the shot bearing the shock of 
 impact better as noticed above, the actual point is applied more 
 immediately to the spot where the tear is to commence, and thus the 
 work of tearing is greatly facilitated. Palliser's projectiles were first 
 tried at Shoeburyness on November 12, 1863. 1 Up to this time it may 
 be noted that the heaviest rifled gun employed was a 10£-inch rifled 
 gun, throwing a shot weighing 3001bs., s with a muzzle velocity of 1320 
 feet. There was also a 7-inch Whitworth 130-pr., and a 13-inch 
 smooth-bore Horsfall gun. 
 
 In America, before this time, iron-clad ships had been made from new 
 designs, and also contrived by converting other vessels into armoured 
 structures covered with railway iron, and the like. The Merrimac, the 
 Atalanta, the Tennessee, the Monitor, the New Ironsides, the Weehawhen, 
 the Montauh, the Nantucket, the Nahant, the Keokuk, 5 and others, had 
 played their parts in the war. Laminated armour, of which the chief Laminated 
 recommendation had been ease of manufacture, was shown to be very 
 inferior to solid wrought-iron plate. 4 Where perforation is effected by 
 means of a pointed projectile, it may easily be seen that this is only 
 what might be expected. The shot finds a passage through by tearing 
 a ragged hole and bending back the corners, and these will bend back 
 easily in the case of laminated armour, because the layers slip on each 
 other and accommodate themselves to the bending, just as it is easier 
 to bend up the corner of a book away from the back than one next the 
 back where the edges of the leaves are fastened together. 
 
 About this time, also, much had been learned with regard to armour- Armour- 
 bolts. The early English ships, such as the Warrior, Minotaur, &c, ° s " 
 had through-bolts with conical heads holding in the armour, and 
 screwed ends nutted against the skin, the nut and conical head holding 
 armour and skin together (Fig. 1). The French had adopted wood 
 
 screws, 5 by which a bolt with a conical head held the armour to the 
 backing, into which the screw held by a projecting thread (Fig. 2). 
 The advantages of the French bolt 6 were that there was no langridge in 
 action from heads flying off, and no leakage through a bolt-hole, and 
 there was a saving of weight of nearly one-fourth on the bolt, which 
 
 1 " Report of Special Committee on Iron/' 1863, p. 169. 
 
 2 " Armoured Defences," lnglis, p. 3. 
 
 3 Vide " Development, of Armour," Very, p. 396, &c. 
 
 4 About six inches of laminated was thought to be equal to four inches solid. 
 
 5 Wood screws were tested severely with great success in America during the war, see "Develop- 
 ment ot Armour," Very, p. 402. 
 
 6 " Development of Armour," Very, p. 380. 
 
amounted to tons on the entire vessel. The importance of the second 
 advantage is evident from the fact that numbers of casualties were 
 
 caused in the American War by the bolt-ends and nuts. 1 Further, the 
 strain was better distributed throughout the length of these bolts than 
 in those of the English first pattern. The value of elastic washers to 
 bolts was soon made apparent in England; but the first important step 
 was made when, in 1862, Sir W Palliser proposed a projected thread 
 of a screw to remedy the evil of weakness caused by cutting a thread — 
 a weakness not at all measured by the mere diminution of cross section 
 by the cut, for such a cut almost instituted a commencement of fracture, 
 and certainly gave so limited and distinct a weak place that all the 
 yielding of the bolt would take place there instead of being distributed 
 along its length. This tendency to yield in one place is similarly 
 encouraged by anything which may nip the bolt at any point. On this 
 account, Major English, R.E., suggested leaving a clear space round 
 the shank of the bolt between the bearings at the ends, and further 
 proposed a spherical nut and bolt-head, and a hemispherical cup or seat 
 in the plate, so that each end of the bolt should be capable of a ball-and- 
 socket movement, which would enable it to accommodate itself to any 
 slight displacement of the plates, which are united in pairs, that is, 
 each plate is bolted only to the one next to it in plate-upon-plate 
 targets (vide Fig. 3). 
 
 liilii 
 
 jScale */E& i 
 
 1 On the Nahant, the bolt nuts, flying from the inside of the pilot-house, disabled the pilot and 
 mortally wounded the helmsman, disabling the steering-gear at the same time. In addition to 
 these, fiye men in the turret were disabled by flying bolt nuts.—" Development of Armour," p. 396. 
 
The Chalmers 1 and SelleropHon 1 targets had before this date been tried. 
 These contained the feature of angle-iron stringers in the backing, 
 intended to give support to the plates and rigidity to the structure, 
 without the evil of transmitting the shock of impact to the bolts and 
 fastenings more than necessary. 
 
 The introduction of heavier ordnance, such as the 13-3-inch gun, 
 throwing a 600-lb. shot, capable of piercing the Warrior at a range 
 exceeding two miles, caused thicker armour to come in. The Hercules, 
 consequently, had 9-inch plates at her water-line, and a target repre- 
 senting her, with massive backing of teak and iron stringers, with ribs 
 and skin, resisted the last-mentioned gun at 700 yards, and would have 
 done so at much shorter ranges. 3 
 
 In 1865, as General Inglis relates, 4 "two complete masonry casemates Masonry 
 with ports in iron shields were built at Shoeburyness. The masonry was wlth^ortem 
 14 feet thick, consisting, generally, of a face of 6 to 8 feet of stone with iron shields - 
 brick-work behind it, the side-walls and vaulting of the casemates 
 being of brick. The shield of one was a compound structure 12 feet 
 long, 8 feet high, and altogether 21 inches thick (including 7 inches of 
 wood) ; that of the other was made out of a solid rolled iron plate, 7 feet 
 high, 6 feet wide, and 13^ inches thick. After the mounting, working, 
 and firing of a 23-ton gun, and a 12-ton gun in the casemates as well as 
 on the roofs, had proved the work to be suitable in arrangement for such 
 guns, the front of the work was attacked by a battery of 7-inch, 8-inch, 
 9'22-inch, and 10-inch guns at ranges of 600 and 1000 yards, firing 
 steel and cast-iron shot, some with hemispherical and some with 
 elliptical heads. 
 
 "The general result of this trial was that after 33 hits the work 
 began to become untenable, after 54 hits its fire would have been 
 virtually silenced, and after 86 hits, of which 22 were on iron, the 
 masonry front was destroyed, but the shields still offered a fair amount 
 of protection. The aggregate of all the blows delivered came to 
 200,000 foot-tons, of which 52,000 were on iron. 
 
 " The issue of this experiment was of the utmost importance to the 
 Service, because on it were based the decisions (1) that our most 
 advanced and important sea-forts should be protected by walls con- 
 sisting wholly of iron, and (2) that for other coast batteries masonry 
 might be used, but that every gun casemate of these should have a 
 shield affording protection against fire, equal, at least, to that of its 
 own gun. A series of trials of plates of steel, and steel and iron com- 
 bined, at this time showed that at this stage of the development of 
 armour a simple rolled plate of soft iron formed the best shield." 
 
 1 Chalmers target, tried May 4, 1863, consisted of 3f-inch iron plate on lOf-inch backing ot 
 horizontal wood and iron stringers, then a l£-inch plate, and 3f-inch wood, and f-inch skin (iron). 
 — "Proceedings of Committee," 1863, p. 183. 
 
 2 The Bellerophon target, tried December 8, 1863, consisted of 6 inches iron, 10 inches wood 
 backing with angle-iron stringers, and double layer of f-inch skin. — "Proceeding of Committee," 
 1863, p. 195. 
 
 * Inglis on " Armoured Defences," p. 4. 
 4 Ibid. 
 
sandwich 
 system, 
 
 In 1866, a Royal Sovereign turret was tried at Spithead with Bellerophon 
 12^-ton guns; the turret suffered, but not in its turning machinery. 
 
 In 1864, 1 steel plates supplied by the Thames Company, Brown & 
 Co., the Parkgate Company, and Petin and G-audet were tried in 
 Russia without success. A very important feature now found its way 
 Piate-upon- into English armour, that is, the plate-upon-plate, or sandwich system. 
 Colonel Inglis relates this in the following words: 2 "In consequence 
 of the growing powers of battering ordnance, it now became evident 
 that our land-works would require walls of considerable thicknesses of 
 armour ; but there were two main reasons why very thick armour- 
 plates should not be used in them. In the first place, the manufacture 
 of a very thick plate is not so complete as that of one of moderate 
 thickness, or at least to make it as complete would involve an enormous 
 increase of cost in plant and manufacture; and next, the thicker the 
 plates the deeper the joints must be, and therefore the more points 
 of undue weakness will the armour present. It therefore became 
 important to see whether the required protection could not be gained 
 without the use of very thick plates. Against doing this was the 
 prevailing opinion, based chiefly on theoretical considerations, that a 
 single plate of given thickness would offer something like twice the 
 combined resistance of two plates each of half that thickness, or about 
 three times the resistance of three plates making up the same total 
 thickness, and so on. This view was entirely disputed by those who 
 had to deal with these questions officially, but it became our business 
 to prove its fallacy. This was done under the following circumstances: 
 In 1807, a total thickness of 7 inches of iron disposed in one solid plate 
 in two plates of 3^ inches, and in three equal thicknesses, instead of 
 giving resistances of about 100, 50, and So, gave effects more nearly 
 as 100, !">•">, and 88 respectively. Next a 10-inch plate failed to stop a 
 shot which was stopped by two 5-ineh plates, and another 10-inch 
 plate bore out this result. 
 
 "Again, in a comparison between a solid 15-inch plate and a wall 
 made up of three 5-inch plates, the result was that, although the solid 
 plate gave a somewhat better resistance to a single blow, the three- 
 plate structure stood repeated blows better than the other. Also in 
 IS 71, two targets representing portions of walls of ships 5 turrets were 
 tried at Shoeburyness. The one was protected by sino-le 14-inch 
 plates, the other by two thicknesses of armour 8 inches and 6 inches 
 respectively, with 9 inches of timber between them. In other respects 
 the targets were similar. After receiving the same amount of battering 
 the armour of both was taken off, and the effect upon the inner skin of 
 the two-plate target was unmistakeably less than that on the single- 
 plate structure. 
 
 "It may also be mentioned that, more recently still, a structure 
 
 1 "Development of Armour," Very, p. 466 ; " Notes on Armoured Defences," Inglis, p. 9. 
 
 J « Notes on Armoured Defences," Inglis, p. 5. It appears as if " plate-upon-plate " ought to 
 refer to iron plates touching each cither, and sandwich " to similar plates with wood or other 
 material between; but the words have not been so used.— See "Targets for Trial of Heaw 
 Ordnance," p. 3, Inglis. Perhaps all targets containing more than one layer of iron mav be 
 inoluded under the term plate-upon-plate, and those with the intervals between the plates filled 
 with other material further distinguished, when necessary, by the term sandwich 
 
9 
 
 composed of three thicknesses of 6| inches of iron proved rather 
 superior to a solid 16^-mch plate in stopping the 818 lb. shot of the 
 Service 38-ton gun, striking with a velocity of about 1415 f.s. 
 
 " In thus dealing with the subject, it must not be supposed that the 
 formation of iron walls made up of a number of very thin plates was 
 ever advocated by us. The trial of the boiler-plate targets, already 
 mentioned, for ever disposed of that kind of construction. 
 
 " Also, it should be mentioned that the above trials of the plate-upon- 
 plate system showed plainly that the most satisfactory results were not 
 obtained when the surfaces of the armour were in contact, but that, on 
 the contrary, some thickness of a softer and more elastic material 
 between the plates was necessary to prevent their breaking under 
 heavy blows. 
 
 " To settle the best proportions, quantity, and best nature of material 
 to be interposed between armour-plates, a series of careful experiments 
 were set on foot, and the result was that a uniform spacing of 
 about 5 inches (to be slightly modified under certain circumstances) 
 between the different plates in all structures was decided upon ; and 
 also, although an iron-concrete, made by working up together cast-iron 
 borings, asphalte, bitumen, and pitch, gave the best result, mainly on 
 account of its great weight, yet brickwork in asphalte, Portland cement, 
 concrete, and hard wood proved so satisfactory that these materials have 
 been adopted, as circumstances required, in all our armoured walls." 
 
 A word or two of explanation may be here useful. It has already 
 been explained that the shot perforates the plate by forcing its way 
 through, so that the plate is stretched over the shot's point, and forced 
 back until it tears in the shape of a cross or star; the shot's point 
 comes through in the centre, bending the corners of metal back, 
 eventually tearing them off, the shot passing clean through. In the 
 case of a wood layer between the plates, it has been established by a 
 series of experiments that the wood should be sufficiently thick to 
 prevent the plates jarring together and cracking, but not sufficiently 
 thick to give room for the shot's point to clear itself of the bent edges 
 of the first plate before it impinges on the second. Practically, 5 
 inches has been found a good thickness for the wood, and adopted 
 generally in this country for plate-upon-plate sandwich armour. It is 
 made up of two layers, one of 2^-inch planks laid horizontally, and the 
 other of 2^-inch planks laid vertically. In the course of plate-upon- 
 plate trials it was found that a very remarkable result was produced 
 on chilled-iron shot, on passing through iron plates with air-spaces Air-Bpaces. 
 between them. Projectiles which perforated the front plate were found 
 disintegrated, a small part adhering to the second plate, in a mass whose 
 consistency was rather that of metal powder pressed together than of 
 solid metal. This effect has been repeatedly produced; one or two cases 
 are noticed in experiments given hereafter (see air-space targets). 
 
 Bemarkable as this result was, it never led to any definite attempt 
 to utilize air-spaces, because there was reason to believe that by means 
 of shells the front plate might be blown entirely off, and also because 
 steel projectiles were coming into use abroad, and on these air-spaces 
 have not been found to produce the same effect. 
 
10 
 
 CHAPTER II. 
 
 Introduction op Formula for Application to Experiments. 
 
 The destruction of armour by shot is an illustration of the " rule of 
 work." The " stored-up work," or " striking energy" in the projectile 
 being converted into destructive effect on the target in the measure 
 in which the former is fairly brought to rest. There are, as might be 
 expected, many causes contributing to complicate what otherwise 
 would be a very simple question. The high velocity and great violence 
 exerted produce effects which it is very difficult to measure. The shot 
 itself often becomes shivered and heated, and work is lost by the 
 actual motion imparted to fragments. These are necessarily subject to 
 great variation, and hence it follows that there is a limit to the degree 
 of accuracy with which results may be calculated, the most favourable 
 conditions being probably such as to allow a shot to completely 
 penetrate, or, as it is termed, perforate armour with but little spare 
 force, the shot being unbroken and the pieces of plate detached 
 without being violently projected to a distance. 
 
 It is well, however, to pause here before becoming occupied with any 
 special forms of destruction of armour, each of which must be dealt 
 with according to its nature. 
 
 From the beginning there existed two distinct systems on which 
 armour might be destroyed — termed "punching " and " racking." On 
 the first system the projectiles are driven completely through the 
 armour, with the object of taking effect on the guns, men, and whatever 
 may be behind it. On the second system, the armour itself is broken 
 up and destroyed, leaving the structure it covered exposed to the effects 
 of subsequent fire. The results obtained by complete punching, or 
 perforation, are more direct and immediate. On the other hand, those 
 obtained by racking are, as regards the armour itself and the future 
 defence of the ship, more decisive, unless, in punching, shells can be 
 made to pass intact through the armour and explode in the interior. 
 The former system was originally followed in England, the latter in 
 America, as being suited to the American cast-iron heavy guns 
 discharging projectiles of great mass with low velocity. It was at 
 one time supposed that racking had become nearly obsolete. 1 The 
 early experiments with cast-iron and steel armour were very dis- 
 couraging ; and thick wrought-iron with wood backing was long 
 thought to be the only form of armour that was likely to be employed, 
 a form which is peculiarly capable of resisting racking, while it admits 
 
 1 The distortion of turrets, as attempted in the Glatton experiment at Portland, in 1872, 
 comes under the head of " racking " in the original sense. — See Glatton experiment. 
 
11 
 
 of being punched by suitable projectiles. Hence the English experi- 
 ments _ for some years consisted almost wholly of trials as to the 
 punching powers of certain individual guns and projectiles, and the 
 resisting powers of structures consisting of wrought-iron supported 
 by backing of different kinds ; the chief variation in conditions being 
 the increasing scale on which the experiments were conducted, owing 
 to the ever-increasing power of the guns and thickness of the plates. 
 Exceptions of course occm-red, and the backing and the fastenings of 
 the armour, and form and nature of the projectile were constantly 
 studied ; still, it has been mainly a question of punching wrought-iron 
 in one or another form until late in 1876, when both chilled-iron and 
 steel became the subject of more special trial on the Continent. These, 
 it shortly appeared, did not admit of being punched like wrought- 
 iron ; while, on the other hand, it was possible to shiver them in a 
 way that was impracticable with good wrought-iron. Steel and 
 chilled-iron have now both of them been adopted to such an extent 
 that racking deserves attention as fully as punching. 
 
 This racking, however, is not generally of the same character as 
 that originally advocated in America, for modern vessels or forts are 
 covered with massive steel or chilled-iron armour which is liable to be 
 broken and thus detached ; whereas the early American iron-clad ships 
 were protected not by the thick wrought-iron spoken of above, but by 
 thin iron plates superimposed in layers, forming what is termed 
 laminated armour. This, it was found, could be destroyed by bending, 
 and so tearing open the ship's side. 
 
 To calculate the stored-up work, represented by injury done in 
 racking in either form, is clearly much more difficult than in the case 
 of "perforation" or complete penetration, though, no doubt, the rule of 
 work is equally fulfilled in each case. 1 The question of partial pene- 
 
 1 The author thinks that Very must misunderstand him. In his criticism on the author's TT.S.I. 
 paper — See " Development of Armour," pp. 460-461, — Very concludes by saying : " If Captain 
 Browne's argument be followed to a logical conclusion, and it be assumed that the hardness of 
 the steel be increased, whilst its other properties remain the same, then the same blow would 
 shatter the plate much more, unless, of course, it be argued that the act of shattering absorbs 
 energy, which cannot possibly be the case." The author had stated that steel had " a remarkable 
 power of distributing into its mass the shock of impact," so as to stop the shot, but at the expense 
 of extending the area of destruction. Perhaps the meaning is best explained by an illustration. 
 At Spezia, in 1876, the 100-ton gun projectiles passed through the wrought-iron plates, fragments 
 having still some velocity left in them, in the case of one fragment, 600 feet per second. Similar 
 shot were stopped by the steel altogether, but the steel plates were completely broken up. The 
 perforation in the wrought-iron had cost a certain quantity of energy, but the shot fragments had 
 still some left in them and travelled on. The fracture of steel had absorbed all the striking energy, 
 and had stopped the shot. If the precise action of fracture were understood, the energy expended 
 in producing each part of each crack might be worked out, and the whole, with the breaking of 
 shot, the heat developed, &c, would, no doubt, equal the striking energy of the projectile. The 
 author, by the expression that the armour absorbed the shock " in the act of going to pieces," 
 meant exactly what Very expresses by the words " It is the projectile-energy which has been 
 transferred to the plate that causes it to go to pieces." The wrought-iron at point of impact gave 
 back and pulled asunder, lhe steel stood up to the work, transmitting the shock from particle to 
 particle, until it was distributed into a considerable mass of the plate, the material splitting in all 
 directions. The author had observed that steel undoubtedly differed from wrought-iron in this 
 power to distribute the shock. He had, however, doubted if a claim could be made for steel that 
 
12 
 
 tration, effected on the punching system, is also a difficult one ; it is, 
 however, of much less importance than complete perforation. A shot 
 entering wrought-iron does little injury, except at the immediate spot 
 it strikes, for the nature of soft wrought-iron causes it to yield locally 
 rather than to transmit the shock through its mass; consequently, with 
 the exception of a little tearing at the back, the injury effected in 
 wrought-iron is generally confined to the punching out of a hole ; and 
 in the case of partial penetration, this hole may often remain plugged 
 up by the projectile. Thus, it follows that in most cases wrought-iron 
 plates bear continued firing very well. They may allow projectiles to pass 
 through them, but they hold well together and suffer little loss in future 
 resisting power from perforation, and still less from partial penetration. 
 Hence, seeing that a shot which stops mid-way in its course through 
 wrought-iron plates can itself produce comparatively little injury to 
 those behind such armour 1 , and that it weakens it very little, it is the 
 least important case of effect produced that has to be considered on 
 service. 
 
 Perforation, meaning the actual passage 2 of the shot through the 
 plate, may with advantage be first considered generally, so as to 
 arrive at a formula sufficiently correct to enable the results of experi- 
 ments to be noted and compared, before discussing them in detail. 
 To this end it will also be necessary to notice generally the quality and 
 behaviour of different kinds of projectiles. 
 
 Piercing of wrought-iron plates. 
 
 Wrought-iron plates owe much of their value to the fact that they 
 do not transmit the shock to the bolts and adjacent parts of the 
 structure, but absorb the blow locally. Hence they may be penetrated 
 easily, but crack comparatively little, especially if supported by soft 
 backing. Penetration is effected in various ways, depending on the 
 form and nature of the projectile employed. 
 
 The appearance of a shot after impact accords with the supposition 
 that it is subjected to pressure exerted in lines which lie in a normal 
 direction to the surface of the head of the projectile as it forces its 
 way through the armour. A spherical cast-iron shot invariably breaks 
 
 ft given quantity of stored-up work effected actually a smaller injury on it. That is, suppose the 
 shock could be equally distributed through the wrought-iron, he doubted if it could be proved that it 
 would break it up more than steel. The undoubted property possessed by the steel was power of 
 distribution of shock from particle to particle, which was connected with its hardness. Subsequently, 
 in a Paper for the E. A. ] nstitution, April 13, 1883, the author expressed his belief that steel of some 
 kind ought, in the long run, to beat wrought-iron (see quotation in Note, Chap. XI.), having 
 greater ultimate tenacity and greater elongation ; that is, although, in 1882, he questioned if 
 actual superiority in resisting power had been proved, in 1883 he thought it ought to be found so 
 in the long run. 
 
 1 Chiefly effected in the flying off of bolt-heads. 
 
 2 Exact perforation is when the shot just passes through with no spare energy to carry it further. 
 This can hardly happen in practice, but sometimes a near approach to it is seen. If part of a 
 projectile remains lodged, and part gets through with energy sufficient to have cleared the lodged 
 portion, the energy has been the equivalent of exact perforation. 
 
13 
 
 up. The anterior part, being under pressure, as shown in Fig. 4, 
 commonly forms a wedge with the point presented towards the rear, 
 
 upon which the posterior part coming under violent tensile strain splits 
 itself, the front part being afterwards picked up as a fragment, such as 
 that exhibited in Fig. 5. Of course a great part of the work stored up 
 
 FIG. 6. 
 
 in the shot is thus wasted. A spherical wrought-iron or soft steel shot 
 is subject to the same forces, speaking generally, but yields in a different 
 way by spreading out, as shown in Fig. 6. A flat-headed projectile 
 meets with resistance directly along lines parallel to its axis, hence 
 there is no tendency to form a wedge-like anterior fragment. Sir J. 
 Whitworth's projectiles, being made of steel, hold well together, 
 flattening or setting-up slightly. Ogival-headed projectiles have little 
 tendency to form a wedge out of the anterior portion under pressure 
 of impact. At first the resistance is comparatively small, and it 
 may be seen by Fig. 7, that, by the time the head has entered 
 sufficiently far to meet with great resistance, the normal lines are in 
 such directions that the shot is nearly in the position of a cylinder 
 driving before it an ogival wedge, whose form, while available to open 
 the armour, has little splitting reaction on the shot itself. The 
 fractured head, Fig. 8, is an example of what is commonly produced 
 
14 
 
 in soft armour. When made of chilled cast-iron, the projectile has 
 comparatively little tenacity, and the posterior portion generally shivers 
 
 to pieces ; nevertheless, the density and hardness of the metal are such 
 that Palliser projectiles long held their own against all others. 
 
 -.A 
 
 FIG. 7. 
 
 These projectiles effect a passage through wrought-iron by punching 
 
 FIG. 8. 
 
 or tearing a hole. Fairbairn, at a very early stage of the investigation 
 
of the question, suggested the following equation as, in a measure, 
 representing the state of matters. This has since been employed, with 
 modifications, by most officers who have dealt with armour-plate experi- 
 ments in England. Captain Andrew Noble, Colonel W. H. Noble, R. A., 
 and other officers connected with early Experimental Committees 
 employed it. General Inglis, R.B., and Colonel Maitland, R.A., have 
 latterly adopted other formulae, and Major English, R.E., has from an 
 early date used one of his own. Fairbairn's formula is, however, the Fairbairn'i 
 one that ought first to be considered : it is as follows : — 
 
 ^-ir2M»Ki 
 
 where W = the weight of the shot, in lbs. 
 
 v = the striking velocity in feet per second. 
 
 g = the force of gravity = 32-19 feet per second. 
 
 D = the diameter or calibre of the shot, in inches. 
 
 t = the thickness of plate completely penetrated, in inches. 
 
 K = a certain constant whose value depends on the quality 
 of the plate, &c. 
 
 As it is convenient to take the weight of the shot in pounds, and to 
 give the stored-up work in foot-tons, the factor 2240 must be embodied 
 in the denominator of the fraction on the left hand side of the equation, 
 in order to bring the answer, which would otherwise be in foot-lbs., 
 to foot-tons. 
 
 It may be seen that the left hand side of this equation, sometimes 
 expressed by the letter " E," truly represents the stored-up work or 
 energy of the projectile at the moment of impact. 
 
 The right hand side is open to objection ; in fact it only claims to 
 give an approximate and partly empirical solution of the question. 
 
 It may be seen that the assumption is made that the plate yields in 
 a circle, tD, coinciding with the edge of the cross section of the 
 projectile. Some have contended that the resistance is proportional 
 to the area of the cross section, not to its circumference ; and R 3 
 therefore enters into the expression employed by them. In certain 
 cases of slight penetration into thick plates this may appear to be true ; 
 but it can be shown in cases of complete penetration, or anything nearly 
 approaching it, that the circumferential assumption is more nearly 
 correct. 
 
 Flat-headed, and even hemispherical-headed, projectiles punch holes 
 in plates by driving out the piece against which they impinge ; thus in 
 separating it from the rest of the plate they clearly tear the iron 
 through in a circle, whose circumference is expressed by irD, as seen 
 in equation above. The ogival point finds its way through the plate in a 
 line in prolongation of the shot's axis : the head tearing the plate open, 
 and bending it aside in all directions. If wrought-iron plates that have 
 
16 
 
 been partially penetrated by ogival shot be examined, it will be found 
 that the plate first yields by bending back opposite to the shot's point, 
 tearing open in the form of a star or cross. This will be particularly 
 well shown by and bye in Fig. 6, Nettle Trials, Chap. V., "Back view 
 of wrought-iron standard plate." As the shot proceeds, it bends 
 back the corners of the plate thus formed until they break off, leaving 
 a circular hole, probably less than the full diameter of the cross 
 section of the projectile, but easily enlarged so as to allow the 
 latter to pass through. ' This action requires special notice, as 
 having an important bearing on the plate-upon-plate system. On 
 this principle a plate is torn through along lines whose total length, 
 may be expressed as IB + ttD, supposing the plate opens in four cracks 
 at the back. Theoretically, then, the flat-headed shot ought to get 
 through a plate with less resistance than the ogival, if it was exactly a 
 case of clean punching ; and this has been pleaded in favour of the 
 former, when an unbacked plate is fired at. Obviously, however, the 
 rough disc of iron which is driven out in front of the flat-headed shot 
 meets with enormous resistance as it gets foul in the backing; while 
 the clean point of the ogival-headed shot, which has disposed of the 
 plate in the manner described, cleaves its way easily through backing 
 and skin ; and the case becomes stronger where the armour consists of 
 several layers of plate and backing. 
 
 Further, a sharp point has so great an advantage in commencing a 
 tear, that for direct penetration of wrought-iron flat-headed projectiles 
 have long since been abandoned, even by Sir Joseph Whitworth, who 
 has warmly advocated their use for certain other purposes. 
 
 The truth of the equation given above depends on two assumptions, 
 which are incorrect in a greater or less degree : — 
 
 1st. That the work done on the plate is proportional to the circum- 
 ference of the hole made. 
 
 2nd. That the resistance of the plate is proportional to the square of 
 its thickness. 
 
 The first of these two is a rough approach to what actually takes 
 place. The second is confessedly empirical, and is modified by almost 
 every one, according to their experience. For a long time this formula 
 was employed in the Department of the Director of Artillery, in the 
 following shape : — 
 
 IKv* 
 
 -5- = tt-D^-b x 253. 
 
 Here K= 2-53, and t is raised to the 1*6 power empirically. 
 
 The factor 2240 being, as above noticed, employed always in the 
 denominator of the fraction on the right hand side, when W is put for 
 the weight of the projectile in pounds. Of course all the constant 
 
17 
 
 parts of this expression might be included in one term ; it is convenient 
 however sometimes not to do so. 
 
 It may be required to obtain simply the total energy " B " of the 
 shot on striking, in order to ascertain the racking effect produced on 
 steel or chilled-iron, which cannot be punched. For example, if it were 
 wished to compare the relative penetrating and racking powers of the 
 English 38-ton gun and Krupp's more modern 24 centimetre (9 - 45-in.) 
 gun, it would be found that the thickness of wrought-iron plate which 
 each would penetrate would be about equal, but the relation of 
 energy of the former to that of the latter would be about Hi to 8 \, 
 which would represent their probable relative powers of smashing up 
 steel or chilled iron. 
 
 Another standard of comparison has also been employed, namely, the 
 energy per inch circumference, written as " e " 1 . The value of this 
 depends on the supposition that the plate is punched or sheared at the 
 circumference of the projectile, and therefore resists the passage of the 
 shot in proportion to the circumference of the hole that the shot makes. 
 On this supposition we may find "e," the "penetrating figure," as it is 
 called, of any shield ; that is, the work necessary to shear each inch of 
 that plate ; and it will follow that any projectile with that quantum of 
 energy per inch circumference will penetrate that shield. This is 
 sometimes applied to a structure consisting of plate and backing. 
 Thus 53 was abundantly proved to be the necessary figure for the 
 Warrior ; that is to say, it was found that any projectile from any gun 
 having 53 tons energy per inch circumference was capable of perforating 
 the Warrior target. 
 
 The formula we have presented above, then, might be used without 
 much increase of trouble, giving these three successive results, viz. : — 
 first, " E " the total energy, representing truly the actual blow, and 
 being available for racking ; second, " e " the energy per inch circum- 
 ference, or fenetraiing figure, which allows ready application to any 
 structure, single or compound, whose figure has been ascertained 
 practically : third, " t" the actual thickness of wrought-iron in a single 
 plate which can be perforated under the given conditions. 
 
 These results, obtained by the equation given above, as employed in 
 the Director of Artillery's Department, were very nearly correct for the 
 thinner kinds of armour and the projectiles used until the last few 
 years ; and, by a curious coincidence, they maintained their credit 
 better than they deserved, with any who were imperfectly informed ; 
 for, as the armour grew thicker, the so-called " plate-upon-plate " or 
 sandwich system, consisting of alternate layers of iron and wood 
 backing, came in, as described in Chapter I. ; and it happened that while 
 the increase of total thickness in iron was itself adding to the power of 
 resistance at a greater rate than allowed for in the formula as stated 
 above, the division of the iron into three or four layers gave a falling 
 off of actual power to resist perforation to an extent that very nearly 
 compensated for the extra rate of increase just noticed. 
 
 1 So expressed in " K. IS. Gunnery Manual," 1880. In Major W. H. Noble's Report, in 1866, 
 this is designated by " p." 
 
 o 
 
18 
 
 For these plate-upon-plate targets, General Inglis established the 
 following approximate rule. 1 The resistance to perforation of any 
 given thickness of wrought-iron armour, made up of single, double, or 
 three layers of iron, with about 5 inches of wood between them, is 
 proportionate to the numbers 100, 96, and 89. Thus a result obtained 
 for a single or solid plate may be corrected to apply to a double plate 
 by multiplying by 100, and dividing by 12 and 8, and sufficiently 
 correctly for three thicknesses by multiplying by 10 and dividing 
 by 9. 
 
 It was found that the modification of Fairbairn's formula with, t, 
 raised to the power 1'6, does not give correct results when the power 
 of the gun is sufficient to deal with thick plates, say equal to 1^ diameters 
 of the shot and upwards. Consequently other formulas have been 
 devised. It will be found, however, that Fairbairn's formula in its 
 original form, with t raised to the second power, gives results requiring 
 but slight correction for the perforation of thick plates, and for all 
 kinds of armour it needs probably as little empirical correction as any 
 formula known. 
 
 It may be well to illustrate what has been said, by working out one 
 example fully f a simple case is furnished by the firing of the 38-ton 
 gun, 12 - 5-inch calibre, at a solid unbacked wrought-iron plate 16"5 
 inches thick, on August 1st, 1877. 
 
 Here weight of projectile, W ' = 817 lbs. 
 
 Striking velocity v = 1410 feet. 
 
 Diameter of shot, ... B = 12-43 inches. 
 
 Fairbairn's old formula is thickness perforated, t = / — - x — x i. • 
 
 r ' >/ 2 ff vB K 
 
 Writing the factors inside the root in a convenient shape to find 
 successively the total striking energy E, that is ~ • then the energy 
 per inch circumference e, that is, — x — ; and finally the total thick- 
 ness that can be perforated t. These, as has been noticed, are all 
 useful functions, and it is easy to follow a process bringing them 
 out in succession. 
 
 1 Vide previous chapter. 
 
 2 Four-figure Logarithms are employed; they are accurate enough for the purpose Much 
 time is saved by using a card with the logs all on one side of it, so as to prevent the necessity of 
 turning over leaves. Log tables of this nature were printed by the E. A. Institution on Professor 
 Bashforth's suggestion. Separate tables are supplied for finding logs to numbers and numbers to 
 logs. This is convenient, but the single table given in the succeeding chapter is of course 
 sufficient. ° 
 
19 
 
 Here then,— Log 2 = 0-3010 
 
 Log^r* =1-5077 
 Log2240f =33503 
 
 
 Log 2y, 
 
 &c, = 
 
 5-1590 
 
 Log TT ... = 
 
 Log 12-43 ... = 
 
 0-4971 
 1-0945 
 
 
 Log ttD 
 
 
 1-5916 
 
 
 * gte taken as 32-19. 
 
 t To bring the stored-up work or energy to foot-tons. The weight of the shot is always given 
 in lbs., and unless this factor were used to bring lbs. to tons the answer would come out in ft.-lbs. 
 
 Log ^(817 lbs.) = 2-9122 
 2 Log ,(1410 fee t) ={™ 
 
 Log Wv* ... 
 
 Log 2g, &c. ... 
 Total energy E= 11263 foot-tons... 
 Log irB 
 Energy per inch e = 288-4 
 
 = 9-2106 
 = 5-1590 
 
 = 4-0516 
 = 1-5916 
 
 = 2-4600 
 
 If the square root of this were taken without any correction the 
 result would be t = 16*98 inches j but to prevent the danger of readers 
 casually using this as a type when thus imperfectly given, a slight 
 correction, K, is made on a plan explained in next chapter. 
 
 Eepeating then, — 
 
 e= 288-4 = 2-4600 
 log k = 1-9845 
 
 inches 2)2-4755 
 
 * = 17-29... 1-2378 
 
 Fairbairn's formula, as modified and used for some years in the 
 Director of Artillery's Department is, as noticed above, — 
 
 1-6/^2 1 1 
 
 ttD 2-53 ' 
 
 That is to say, it differs only from the above in the last process, when a 
 different root and constant are employed, as follows : — 
 
 e = 288-4 = 2-4600 
 log 2-5 3 = 0-40311 
 
 , 4)2-0569 
 1-6 ' 
 
 {■ 
 
 4) -51423 
 t = 19-31 (inches) ...1-2856 
 
20 
 
 The results is, in this case, much further from the truth than the 
 results given by Fairbairn's old equation, without the application of 
 any correction K. This would be supported by any of the later formulae 
 used, and it is borne out by the result on this occasiou, when the 
 projectile got through the plate, breaking to pieces and having a little 
 spare work in it, so that Inglis 1 estimated it as able to perforate a plate 
 from 17 to 17'5 inches thick. 3 
 
 It may be seen that Log 2y, &c, = 5-1590, and Log ir = 0-4971 are 
 constants available for all examples. The calculation of results of 
 experiments is further discussed in the next chapter. 
 
 1 That is to say, Inglis estimated that from 17 to 17£ inches solid plate would correspond to a 
 certain sandwich target (No. 40) which was as nearly as possible a match for this shot ; see Inglis' 
 paper on " Targets for Heavy Ordnance," E.E. Institution, 1877. 
 
 • A diagram, brought out by Colonel Maitland, gives 17-14 inches, and one by Colonel Inglis 
 17-2 inches. A rule of thumb, recommended for rough estimation, by the author, gives 17-6 inches 
 (for these see next chapter). 
 
21 
 
 CHAPTER III. 
 
 Calculation and Estimation ov the Powers of Guns against Armoor. 
 
 For the sake of distinction, Armour may be divided into two classes, 
 " soft " and " hard." Under the head of soft armour, may be included 
 all shields which admit of perforation, which, when possible, is the best 
 method of attack. Soft armour then includes all kinds consisting of 
 wrought-iron only, whether laminated, plate-upon-plate, or solid. 
 Occasionally a steel or steel-faced plate has had a hole made completely 
 through it without breaking it up, 1 but this is not generally possible. 
 Wrought-iron was universally employed until chilled iron armour was 
 adopted for certain land forts on the continent. This was first tried 
 in 1868 with success, and in 1873 it met with very marked approval. 2 
 After the Spezia trials of December, 1876, 3 steel and steel-faced armour 
 came gradually into use. Up to that time, our experiments in England 
 were almost entirely confined to the problem of the perforation of 
 wrought-iron. And even now perforation is kept in view with steel-faced 
 plates in a measure, for a shot is generally considered to be a match for a 
 steel-faced plate when it is capable of perforating some estimated equiva- 
 lent thickness of wrought-iron. Suppose, however, that such a relation 
 can be established, its application is limited to the few cases when per- 
 foration without much fracture is effected, and may greatly mislead any 
 
 i Vide trial of Palliger improved projectiles in 1882 and 1883, and Brown's plate tested at 
 Shoeburyness in 1882.— Chap. VI. 
 
 2 In Prussia against an 11-inch gun (28 cm ), and in 1874 at Tegel and Cannes. — Vide Chap. VIII. 
 
 3 Vide Chap. V. 
 
 4 
 
22 
 
 one who uses it indiscriminately, for the following reason :— Power of 
 perforation varies inversely with the diameter of the hole made. Ihus 
 a shot of less energy than another may perforate the same thickness of 
 armour if its calibre is less, because it does not require to make so large 
 a hole. Thus, as mentioned in the last chapter, the 9'45-in. Krupp gun 
 in 1879, had about the same power of perforation as the 12-5-in. Wool- 
 wich gun, namely, about 18 inches ; and would be considered a match 
 for the same plate. If, however, armour is too hard to perforate, and 
 yields by breaking up instead, it appears doubtful if the smaller 
 calibre possesses any advantage over the other. The work of " smash- 
 ing" appears to be more likely proportional to the striking energy or 
 stored-up work. Thus, in the case above quoted, the 12'5-in. shot has 
 more energy than the 9'45-in. shot, in the proportion of nearly 3 to 2, 
 and might break up armour accordingly. So, again, the 100-ton M.L. 
 gun at 2200 yards had about the same perforation as the 43-ton B.L. 
 gun at the muzzle, namely, about 24 inches of iron. 1 Its energy, how- 
 ever, is greater in the proportion of 3 to 2 ; 2 and the larger projectile 
 has this advantage with a velocity of only 1500 feet per second, while 
 that of the smaller one is over 2000 feet. It is quite conceivable then 
 that the former would hold better together, and thus deliver perhaps 
 double the blow of the smaller one before breaking up. In such a 
 case, then, it appears that the standard of perforation is a very erroneous 
 one. It is, however, the only standard which has been definitely 
 worked out, and on it our diagrams and rules are based. They must 
 therefore be considered correct only when applied to cases of soft 
 armour. Hard armour, which cannot be perforated, includes chilled 
 cast-iron, and, in most cases, steel-faced armour and steel, though the 
 latter may be made 8 so soft as to approach wrought-iron, and so be per- 
 forated in a fairly clean hole. This, however, has not been the character 
 of most of the steel plates hitherto tried. A projectile may drive its 
 point a short distance into hard steel, but eventually it acts as a conical 
 wedge splitting up the plate, and the shot breaks up before the entire 
 head enters the plate, so that the size of the calibre probably affects the 
 problem less than the shape of point, and it appears that the main 
 question is that of the striking energy, modified by the shot's power 
 to hold together, which again depends directly on tenacity of metal, 
 and possibly inversely on some function of the striking velocity. In 
 Reports on Experiments, General Inglis sometimes gives the average 
 number of foot-tons work sustained by plates per square foot, 4 but 
 
 1 As originally laid down. — vide Inglis' diagram. Of course these perforations increase as 
 powder and projectiles improve. 
 
 2 Viz.:— 31,200 against 19,800 foot-tons. In March, 1883, at Copenhagen, a Krupp 16 c » 
 (5f-in.) gun, and an Armstrong old-type 9-inch gun were fired at the same target. — vide Very. 
 Their energies per inch circumference were nearly equal, namely, 123 and 118 foot-tons, respectively, 
 but their total energies were 5,760 and 16,403 foot-tons The smashing powers would be probably 
 in the latter ratio. 
 
 3 Vide Spezia trials, 1884, hereafter. 
 
 4 Vide Eeport of Sub-Committee on Plates and Projectiles, p. 17 ; also Major O'Callaghan's 
 paper, K. A. I. Proceedings, Vol. XII., Major English's method of calculating effects took into 
 account the mass of the Plate. 
 
23 
 
 such information has never been brought into a shape to furnish data 
 to guide Officers in the attack of hard armour. In short, the detailed 
 information below will be seen to apply altogether to the perforation 
 of soft armour. At the present time it would apply to the attack of 
 the great bulk of old fashioned armour-plated ships, which are cased in 
 wrought-iron. It would not apply to the Italian vessels, Duilio, 
 Bandolo, Italia and Lepanto} nor to any French or other vessel which 
 may carry steel armour, unless the steel is very soft, as undoubtedly is 
 occasionally the case ; nor would it apply to the case of chilled-iron 
 forts. On these there is but little to be said. Some remarks and 
 calculations, however, on the results of recent trials against hard armour 
 will be found hereafter. 
 
 Perforation through Soft Armour. 
 
 An admirable official Eeport on the " Penetration of iron armour 
 plates by steel shot," was compiled in 1866 by Colonel W. H. Noble, 
 R.A. A diagram showing the ranges at which British and Foreign 
 armour-clad ships might be perforated, was brought out on a suggestion 
 made by Colonel Forde, R.A., in the " Engineer," and also in the 
 "United Service Institution Proceedings " by the author in 1872 ; and 
 in 1873, Colonel Noble prepared a diagram on a different principle, by 
 authority, which was once afterwards corrected and completed to include 
 the 80-ton gun, but all are now out of date. 
 
 A memorandum on the perforation of solid unbacked wrought-iron 
 plates by direct fire was drawn up by the Sub-Committee on Plates 
 and Projectiles in 1881, containing diagrams constructed by General 
 Inglis, R.E., and Colonel Maitland, R.A.., from which the effect of 
 projectiles, fired under various conditions, could be read off easily. 
 The text of this memorandum stated that certain conclusions ha,d 
 been arrived at by the Committee, with regard to data on which to 
 calculate perforation supplied by the very complete series of experiments 
 made with the Elswick 6-in. and 8-in. guns by this Committee. Both 
 Colonel Maitland and General Inglis based their calculations on the 
 relation borne by the diameter of shot (or calibre of gun) to the thick- 
 ness of plate perforated. The two diagrams can be best used in different 
 ways. Inglis' diagram is undoubtedly the most readily understood, 
 and most easily used, but its use is confined to the case of shot of a 
 given weight, fired from those guns which may be entered on the 
 diagram. The result is directly read off, and no explanation is here 
 required. 2 Maitland's diagram is available for problems with any 
 gun and projectile, and it should be understood that in using it an 
 actual process of calculation is performed, which may be regarded as 
 equal to any in accuracy, and when reduced to a system by the use of 
 
 1 Schneider's plates may yield by perforation accompanied by cracking, which would make 
 calculation difficult. 
 
 2 This diagram might be used on service readily if supplied. — Vide "Use" under Practical 
 Directions. 
 
24 
 
 tables it is far more rapidly and easily performed than any other, when 
 perforation is the result sought. 
 
 Maitland based his operations on the fact that the scale on which 
 plate experiments are conducted does not materially affect the 
 mechanical conditions of the question. He therefore works the problem 
 and obtains a result " K" in calibres, or more strictly in diameters of 
 the shot. Such a result can of course be brought out in inches by 
 multiplying it by the number of inches in the particular calibre or 
 diameter required. 
 
 Thus, if K be 1'5, it would imply that under those conditions a 
 6-in. projectile would perforate 9 inches, a 12-in. projectile 18 inches, 
 and so on. 
 
 Maitland took the proportion of weight and calibre (or sectional 
 
 density) of projectiles, with which the most reliable experiments were 
 
 W 
 made by the Committee, that is, -=- s = 0"37 as a standard co-efficient, 
 
 and he made for these conditions a diagram, from which was read off 
 
 the factor K, which, when multiplied by the calibre or diameter of the 
 
 shot in question, gave the perforation required. 
 
 W 
 With shot whose -r s does not equal 037, he formerly required a 
 
 second diagram to reduce the actual velocity to what would bear 
 
 W 
 
 equivalent velocity if — i were equal to - 37, as a preliminary operation. 
 
 The use of the first mentioned diagram follows. These two operations 
 are very quickly performed and give good results. 
 
 Maitland subsequently brought out a large card diagram, which by 
 means of a brass bar pivoting at one end, can be set so as to combine 
 the two operations above in one. When the striking velocity and the 
 
 the value of -^ are known, the factor K is immediately found, which 
 
 multiplied by the calibre gives the perforation in inches. In addition 
 to this, the same card, together with a wood scale, can be used to find 
 the remaining velocity at any range, when the initial velocity is 
 given, and also to find the time of flight. For practical purposes on 
 service this diagram might be reduced in size, the gain in convenience 
 would more than compensate for any loss in accuracy. Both Inglis 
 and Maitland calculate results on systems of their own, apart from the 
 use of diagrams. Between Inglis' diagram and his formula, there 
 does not appear to be any necessary relation. The diagram is based 
 simply on results obtained by practice. The fact that the curves of 
 each gun are nearly straight lines goes to prove that the terms on 
 which the abscissa} and ordmates, t and v, are based, are in direct propor- 
 tion to one another, and the same power; in other words, that v* and t* 
 enter into the question as in Fairbairn's formula. Inglis' own formula 
 differs generally from this. General Inglis lays down, in calculating 
 problems that for a given relation of t (the thickness of plate just 
 perforated) to d (the diameter of shot) a given factor multiplied by the 
 number of cubic inches of iron removed in forming a cylindrical hole 
 in the plate will give the required striking velocity. When the relation 
 
25 
 
 of t to d changes, the factor changes; but as long as this relation 
 remains the same, it is immaterial on what scale the experiment is 
 made. He gives the following table : — 
 
 Proportion of thickness of plate (t) to 
 
 diameter of shot (d) 
 
 t 
 
 d 
 
 Energy required to remove each cubic inch of iron in 
 forming a cylindrical hole through plate. — c. 
 
 •5 
 
 foot-tons. 
 2-6 
 
 ■6 
 
 3-0 
 
 •7 
 
 3-3 
 
 •8 
 
 ; 3-6 
 
 
 3-9 
 
 1-0 
 
 4-2 
 
 
 4-6 
 
 
 4-8 
 
 
 6-1 
 
 
 5-4 
 
 
 6-7 
 
 
 
 
 
 
 6-6 
 
 
 6-9 
 
 
 
 Taking W= weight, v = striking velocity, and g force of gravity, 
 Inglis gives the following formula for calculation. 
 
 =d ! X 7 X(Xei 
 
 iff 4 
 
 = d / 1 x c x 
 
 X 1g. 
 
 .. (Eq. 1) 
 
 Example. — Find the velocity at which an 8-inch shot (d = 7'97 in.) 
 of 182£ lbs. weight will perforate a 12-inch wrought-iron plate : — 
 
 Here 
 
 t 12 
 
 - = =t 1-5, whose corresponding c = 5 - 7 ; 
 
 d 7-97 
 
 .-. velocity = 7'97 / 12 x 5-7 x | X %g X 
 
 2240 
 
 182-5 
 
 = 1644 feet. 
 
 These calculations are made for projectiles with heads of 1-5 diameters 
 radius. For heads of 2 diameters radius there is a gain of from 10 to 
 5 per cent. This calculation is of course easily brought to a system 
 and worked by logarithms, some of which are constant. Correct 
 results are thus obtained. 
 
 The following objections may be used against General Inglis' system 
 of calculating results. The assumption that the relation of t to d is 
 known, confines the strict application to cases when " I " the thickness 
 
26 
 
 of armour is known, and the velocity only is required. If the velocity 
 
 is known and t is required to be found, - has to be assumed. It will 
 
 be seen subsequently that this can be done roughly, but the great 
 variation in the constant, or rather correction c, causes the error, 
 liable to accompany a guess, to tell largely. Here lies the root of the 
 principal objection to this formula; namely, that factors which should 
 be constant if the formula was theoretically a perfect one, vary from 
 2'6 to 7 - 8. In other words, the formula is in itself so incorrect that 
 it can only be made to give accurate results when corrections altering 
 the answer about 200 per cent, are employed. This objection in no 
 way applies to the diagram, which is admirable. This evil in the 
 formula will be modified in a great degree by taking Fairbairn's 
 original formula, and altering Inglis' table of factors or corrections c, 
 to correspond to it. 
 
 Theoretically, this appears to be better, for the variation required in 
 c to give correct results becomes so far reduced, that the formula 
 appears to have a considerable measure of truth in it, and practically 
 any error in assuming a relation of t to d tells comparatively little in 
 the common problem when the thickness of plate is not given, but 
 required to be found. This formula also agrees theoretically with the 
 straight lines in Inglis' diagram. 
 
 It may be shown that in cases when the relation of t to D does not 
 change, Inglis' formula is identical with that of Fairbairn. It is 
 obvious that any number of (or, if desired, all) the constants in any 
 expression may be embodied in some one symbol, and determined once 
 for all by experiment. This factor being constant, combined with the 
 variables, will satisfy all possible cases of perforation. 
 
 It may bring out the actual construction of the formulae better to 
 leave in them the constants which are identical in all, and express 
 those that are peculiar to each as kp t kj. 
 
 WiP 
 Thus Fairbairn's is -r— = ■n-JD x f x k F ; 
 
 Inglis' P'=^D»x<xc, 
 
 And when t bears a constant relation to B, we may substitute t for one 
 power of B, and express the factor required, together with c and £, 
 under the symbol kj t when the equation becomes — 
 
 -~ = "--D x P x ki. 
 
 which is identical with Fairbairn's, above, showing that ki and kp must 
 be the same when determined by practice. So then when Inglis' 
 formula requires no change in the factor c, it is identical with that of 
 Fairbairn ; but when a change is required and dealt with differently 
 according to each formula, that of Fairbairn requires a much smaller 
 variation in correction than that of Inglis. 
 
27 
 
 The following are the factors, and their logarithms, for Fairbairn's 
 formula corresponding to Inglis' table of factor c, that is, giving the 
 same results : — 
 
 The value, then, of these factors varies from 18 to 9. 
 
 For calculating results on this system, using 
 four-figure logarithms, the following are con- 
 stant logs available for all questions : — 
 log 2g X 2240*= 5-1590; 
 log it = 0-4971. 
 
 Suppose the example above be taken when 
 t = 12 ins., and d = 7-97; that is, ^ = 1-5, 
 
 nearly, log c = 7-9777, t = 12, d = 7*97, w = 
 182-5. 
 Then, 
 
 — - = ird X fi X C 
 
 when v is required, becomes by transposing — 
 
 t 
 
 factor 
 
 logarithm of 
 
 d 
 
 K. 
 
 K. 
 
 •5 
 
 1-30 
 
 0-1139 
 
 •6 
 
 1-26 
 
 0-1004 
 
 •7 
 
 1-18 
 
 0-0719 
 
 •8 
 
 1-13 
 
 0-0511 
 
 •9 
 
 1-10 
 
 0-0414 
 
 1-0 
 
 1-05 
 
 0-0212 
 
 1-1 
 
 102 
 
 0-0086 
 
 1-2 
 
 1-00 
 
 o-oooo 
 
 1-3 
 
 p-98 
 
 1-9912 
 
 1-4 
 
 0-96 
 
 1-9845 
 
 1-5 
 
 0-95 
 
 1-9777 
 
 1-6 
 
 0-94 
 
 1-9731 
 
 1-7 
 
 0-93 
 
 1-9685 
 
 1-8 
 
 0-92 
 
 1-9638 
 
 1-9 
 
 0-91 
 
 1-9590 
 
 2-0 
 
 0-91 
 
 1-9590 
 
 -J 
 
 dff- x irk X 2g 
 
 log 7-97 = 0-9015, 
 
 , (1-0792, 
 
 2 log 12 — | 1 . 0793j 
 
 log v = 0-4971, 
 log 2g = 5-1590, 
 logc = 1-9777. 
 8-6937 
 2-z613 
 
 log 182-5 = 
 
 Ans. v = 1645. 
 
 2)6-4324 
 8-2162. 
 
 This answer agrees with that obtained by General Inglis, above. 
 
 Next take a case when we do not know the thickness of the plate. 
 
 At Meppen, in March, 1 882, Krupp fired a 15 cm gun (5'87 in.), the 
 shot weighed 112-44 lbs., the velocity about 1765 feet. 
 
 Suppose it is desired to find, in succession, E, the total striking 
 energy; e, the energy per inch circumference, and t the thickness of iron the 
 shot is capable of perforating : — 
 
 We have, 
 
 '-J 
 
 %g 
 
 X — x 
 
 7TO 
 
 1 
 
 K" 
 
 Obviously it is a simple and convenient thing to use a process giving 
 E, e and t in succession. 
 
 * The factor 2240 is employed to bring the weight of the shot in lbs. to foot-tons ; g is taken 
 as 32-19. 
 
28 
 
 log W, that is log 112-4 = 2-0507 log 5-87 = 0-7686 
 
 ( 3-2467 log 7T = 4971 
 
 2 log V, that is 2 log 1765 = 1 3 . 2467 " 1-2657 
 
 8-5441 * log e = 1-9685 
 
 log (2240 X g) 5-1590 
 
 — \ = 2428 3-3851 
 
 log (* X 5-87)= 1-2657 
 
 (e) work per inch} _ / W#_ x 1\ _ m . 4 2 -1194 
 circumference C ^ 2<7 nd) —— _ 
 
 log c = 1-9685 
 inches 2-1509 
 
 [t) thickness that may be| _ , , ,g 1-0755 
 
 perforated j — 
 
 Maitland has a formula Z= '001 V- '14 when Fis greater than 700, 
 and another K= -0008 V when F is less than 700 ; 
 
 Wr? 
 
 Maitland's V= / kg-'B^M.g £ being the striking energy ^ 
 
 This may be compared with Fairbairn's formula easily by writing it 
 as follows : — 
 
 t = BK as noticed before, K being t in calibres -, 
 
 and taking D x -001 inside the root and then cancelling i) 3 in numerator 
 and denominator. 
 
 jffV 
 
 - 2 x -000,001 1 °g" 15 J 5904 - -142); 
 
 and, writing K M for the constant -000,001 log -1 5-5904, 
 
 t= /T* x I x Z3£--14D, 
 
 which is almost identical with Fairbairn's in the form for perforation, 
 putting Kp and Km for Fairbairn's and Maitland's constants ; 
 
 namely, <- J Z£ x ^ x jr,. 
 
 i c is taken on the assumption that each complete 1000 feet velocity represents a calibre 
 penetration. Thus, over 1000 feet is assumed equal to the perforation of 1 calibre, over 1600 feet 
 equal to 1-6 calibres, and so on. This assumption is explained hereafter, and is sufficiently near 
 to the truth for the purpose. Thus for 1766, if 1-7 calibres be taken on table p. 27, we haveT-9686. 
 
29 
 
 If the constant K M were modified to enable ■*■ to be combined with 
 D, Fairbairn's and Maitland's would only differ in the fact that 
 Maitland makes a constant small correction in terms of the shot's 
 diameter, and this disappears in his second equation when V is less 
 than 700. 
 
 The formulae of Inglis and Maitland give nearly the same results. 
 Pairbairn's, as corrected above, does the same, and appears simple and 
 good. It has a considerable measure of truth, but is also empirical in 
 a measure. If theoretically correct, it would not require a change of 
 the factor K, which would then be really a constant. Captain A. 
 Noble has worked out a modification of Pairbairn's formula. On the 
 other hand, Major English, R.E., has a formula, based on a very 
 searching investigation of conditions, but which cannot here be 
 discussed. 1 
 
 There are forrnulaa used abroad which, like that of Inglis', treat the 
 
 resistance of the plate as proportional to the cross section of the shot, 
 
 not to its circumference. This may be true, as noticed in preceding 
 
 chapter, in an early stage of penetration, but after the plate begins to 
 
 tear open in a cross at the back, it appears as if the remainder of the 
 
 work was proportional to the length of these lines, or rents, which are 
 
 functions of d, the shot's diameter, not of d 2 as in the case of the 
 
 area. Por a rough approximation there is a rule of thumb, which may 
 
 be noted as very easily applied, and the circumstances under which it 
 
 is liable to any considerable error, and also the probable extent of the 
 
 error maybe generally foreseen, as it is based on a formula which is true 
 
 if two assumptions are made/ assumptions which are seldom grossly 
 
 wrong, and are fairly near the truth in the most important cases, that is, 
 
 with projectiles with fairly high sectional density, fired at high velocities ; 
 
 for instance, when fired from new type guns. These assumptions are (1) 
 
 that Fairbairn's original formula, as given above, is correct. (2) that all 
 
 projectiles have the same proportion of weight to diameter (or " sectional 
 
 W 
 density"), so that -r 3 is in all cases the same. Then taking Pairbairn's 
 
 formula, viz., — 
 
 TVv* 
 
 ■n-d X 
 
 P 
 
 xXl 
 
 
 
 X 
 
 Tra 
 
 1 
 
 *1 
 
 or 
 
 writing K 2 d 3 for W, which amounts to assuming that the weight is 
 
 1 Professor Greenhill kindly went into this question with me, and concluded that it was better 
 not to use Major English's formula. 
 
 2 The author has generally had the credit given him for this rule, see Very, " Development of 
 Armour," p. 580, and Mackinlay's "Text Book on Gunnery," p. 142. The fact is, that the author on 
 examining the results obtained by Inglis' Committee, thought he discovered the law, and then 
 found the mathematical proof herewith. He afterwards found that both Inglis and Maitland had, 
 in a measure, seen the law ; General Inglis had enunciated it for three cases, roughly, in his paper 
 on "Armoured Defences," read 1880, p. 10. The author trusts that the proof herewith justifies 
 confidence, which an apparent law in one series of trials would not deserve, and he put it forward 
 boldly, and applied it generally in his paper of October, 1881. He, now, however, thinks that both 
 Inglis and Maitland must have perceived the law before he even saw the results on which it is based. 
 
 5 
 
30 
 
 proportionate to the cube of the diameter, which is a true assumption 
 in projectiles of the same sectional density ; 
 
 PsaK% ** + 
 
 3 iff X ird X-fiTj 
 
 writing Z 3 for all the constants, that is for - x *' x ^ and cancelUn ^ 
 
 d above and below — 
 
 t* = <Z» x r 2 x K s ; 
 
 and taking the square root, 
 
 t =s dv^/K$. 
 
 That is, the thickness perforated is equal to the calibre or diameter of 
 shot multiplied by the velocity in feet and by some constant yK s ; and 
 this constant turns out to be nearly TT! V?t 1 ; so that supposing this to 
 be true, the thickness of iron that may be perforated by a shot is 
 very easily found by multiplying the calibre by the number of 
 thousand feet striking velocity. Thus, an 8-inch shot to perforate 
 8 inches would require 1000 feet velocity. To perforate 12 inches it 
 would require 1500 feet velocity, each 1000 feet velocity giving 1 
 calibre perforation. Suppose then with the 26-ton B.L. gun 10'4-in. 
 calibre, the shot has a muzzle velocity of 2080 feet, it ought to perforate 
 2080xl0-t in ueg This happen to be exactly the perforation 
 
 1000 ^ 
 
 given by Inglis' diagram. For 1500 feet velocity the rule would 
 give 15-6, and the diagram 151 inches. The rule of thumb being 
 
 very near the truth when r = 2000 feet, and -^ =0'41 about. 
 
 The table herewith shows the amount of error occurring in different 
 cases from the use of the rule, supposing Inglis' diagram to be correct. 
 It will be seen that considerable error exists when the projectile is 
 
 W 
 light, as in the case of certain Armstrong guns, where ^ = 0"34 or 
 
 - 37. Projectiles for the future, on the other hand, are likely to have 
 high sectional density. An Officer, however, would surely know the 
 general character of a gun that he was firing, and make allowance for 
 a low velocity, or specially short light, or specially heavy projectile. 
 As a high limit, however, the rule may be generally safely used. That 
 is to say, it may be safely assumed that a shot will scarcely ever 
 perforate, on service, more than would be given by the rule. With 
 wrought-iron it is to be remembered that partial penetration does 
 armour no appreciable injury, for the shot generally only makes a hole 
 to a certain depth, and plugs it up, leaving the surrounding parts of 
 the shield and bolts very little injured. It is very important then not to 
 waste shot in firing at wrought-iron beyond the power of the gun. Con- 
 sequently, the following rule is suggested; it is useless to fire at 
 wrought-iron armour unless the shot has at least 1000 feet striking velocity 
 for each calibre in the thickness of the armour. 
 
 1 Agreeing with the '001 of Maitland. 
 
81 
 
 <D 
 
 
 -Q 
 
 
 >-. 
 
 
 03 
 
 
 a 
 
 
 CD 
 
 
 m 
 
 
 d 
 
 
 a 
 
 
 •rH 
 
 
 * 
 
 a 
 
 o 
 
 M 
 
 d 
 
 Co 
 
 
 ■4 
 
 
 
 d 
 
 H 
 
 3 
 
 . 
 
 be 
 
 
 <+H 
 
 y 
 
 o 
 
 •—I 
 
 tn 
 
 
 CD 
 
 
 O, 
 
 
 l>> 
 
 
 H d 
 
 H 
 
 ^.£P 
 
 £ 
 
 +3 Cfl 
 
 
 "SrS 
 
 a cd 
 
 
 d & 
 
 a 
 
 -d cd 
 
 & 
 
 E-i d 
 
 CI 
 
 O rd 
 
 o 
 
 
 
 
 w 
 
 r-H CD 
 
 M 
 
 d t3 
 
 t= 
 
 tf,3 
 
 # 
 
 o 
 
 w 
 
 CD fl 
 
 H 
 
 ^a -S 
 
 H 
 
 - 13 -» 
 
 tH 
 
 >> 2 
 
 n 
 
 nQ * 
 
 K 
 
 d cd 
 
 
 £ o 
 
 o 
 
 &D -. 
 
 H 
 
 u 
 
 
 W CD 
 
 e 
 
 © o 
 
 H 
 
 PJ 
 
 w 
 
 
 « 
 •"I 
 
 ■+3 O 
 
 
 
 
 tM r^l 
 
 fr 
 
 
 o 
 W 
 
 O CD 
 
 rti 
 
 23 
 
 M 
 
 d c3 
 
 1? 
 
 o •*= 
 
 o 
 
 of ac 
 wing 
 
 H 
 
 H 
 
 n 
 
 o 
 
 ■4 
 
 cd ■=! 
 
 H 
 
 
 
 03 CD 
 
 
 2 rd 
 
 
 a*' 
 
 
 £.2 
 
 
 fc-l 33 
 
 
 CD 
 
 
 CO 
 
 
 
 
 ^■^tto^os od t> cs t- b» co ao m h j><dj> 
 
 «s 
 
 •eDnajejjiG 
 
 hnoho 16666 1 1666 1 w i-h 6 6 6 
 
 o 
 o 
 o 
 
 i-H • 
 
 a § 
 
 
 + + + + + ++++ +++ ++I++ 
 
 ts. j 
 
 H CO 81 © H t- CO rH CO CO OS •*? l> 01 © 1> « 
 
 6 us ib ^ cb 1 hhho | 1 6a ab ab 1 i> 6 cb >b *b 
 
 ■u 
 
 
 
 
 
 CI 
 
 ■qmtn lvI j0 9p a | 
 
 NbCDOOTfUOOOOrflCOtONOOOCtiCOO 
 
 Ph 
 
 ^Nib^^COCTOlWHOOOQroQfflQOu:©© 
 
 rHr-ti-irHi-HiHi-trHi-lrHr-i.-l.-' 
 
 
 
 CO CO (N t- CO rH WU3 lO>OlQ01-*-^' l *(M'^COl>- 
 
 
 •80aa.ragtQ; 
 
 hwoooh |66 1 66666hhhh66 
 
 o 
 o 
 
 t—< . 
 
 03 . ~ 
 
 
 ++I+++ ++ ++++++++ | ++ 
 
 ■mviSmcx 
 iff' 
 
 COOCOCONO 00 US rllOU3O^HHC0NHM 
 
 CI § 
 
 lij-^MMCOl 1 J> t- luS^^^cbffJNOOOSOO 
 
 .3 "3 
 
 (M « ff) IN CI H '^HrH ' rHr-li-ti-li-li-ti-li-li-l 
 
 ■§£ 
 
 
 
 <w 
 
 
 
 a 
 
 •qramjxjoeitia 
 iff 
 
 CD©ffiOOH OO CO O © <M 00 US US O 00 -* O 
 
 Ph 
 
 cb cb cb -^ Ai 6 1 do do |us»bus^cbcbMcia>d>C3 
 
 N Cl « N « M 'i-HrH 'rHr-ti-lt-H'-li-lr-lrH 
 
 *3 
 
 
 COCOnXiHOiaHCTiOOHUJClOCOiO^QONCD 
 
 i-HtMOOOrHOOOOOOOOOi-'nHiHr-lOO 
 
 ++I +++++++ + + 1 +++I+ + 
 
 CD 
 
 
 
 
 
 
 
 
 N 
 
 a 
 
 ■ lubjSbiq 
 
 COiHtOG^O^OTI^OiOCOOir*!H^O''*C5USl>C3 
 
 
 % 
 
 «5 OD © 'O M DNOJCOiiiHO^OOOOiw^HOH 
 
 IS 
 
 "a 
 
 (MPlIliNCHlHHNHClWHHHHHHHHH 
 
 o 
 
 CD 
 
 
 
 3 
 
 
 ■qccmijx jo 
 
 CDt>C^OOt^O^COrHUSCOrH(?lC5TpeOOTeOt^05aO 
 
 3 
 
 Ph 
 
 a l n 3I % 
 
 ©oiibcb'^iictt^ibHOiit-dDOcb'ioiOH 
 
 a 
 
 
 (NCO^M'NWClHHeiHfMCl'ir-iHiyiHH r-li-H 
 
 as 
 
 3 
 
 . 
 
 
 
 V 
 
 Oionmooiflomooiyiiflooooq-iicoo 
 
 
 ■jfrjioojajL 
 
 «1 
 
 iiji>tDffitO(0'*CDO'#OOincOOO : liOO'IH>OJ 
 
 
 HHHHHHHHIMHtMC^r- 1 H (M « H CI H H r- 1 
 
 M5 
 
 
 
 © 
 
 
 
 
 t-^t^OOiCOUSOOO^OO-^cqOOOCOlMO 
 
 
 
 t>Nib6mMClCTWH66605036l03ii!iD®© 
 
 
 
 i-Hi-I.— 'rH.-|.-l.-tr-lr-ti-)>-Hi-li-l 
 
 
 Q 
 
 — 
 
 
 3 
 
 
 
 O 
 
 hjH i4 p H 
 
 
 H 
 
 s 
 
 mS : : ■ . : • : : -W : (4 : PP : : (4jW 
 
 
 ^l^hqi-iHl &JDCQ&C pp • 5P 
 
 
 EH 
 
 g m :w« : pa : = ig :^ :g i :p.«§ 
 
 
 S & g.^ & | g- | . g-S-g 
 
 
 
 
 
 
 a aaaaaaaaaaaBgggoBag 
 
 
 
 
 
 
 ■? » ■*? ■? -? "f -V "f? "? ^ " ^ T-'V "V ■? «(s°= "f-f 
 
 
 
 O HOHffiQOCOOJiOCDiOOOt-OOCONHmiO^ 
 O MBiO^Mn^NNNHHHHHH 
 
 
 
 
 
 r-i 
 
 .13 
 
 Np| 
 
 .3 
 
 5^B- 
 
 :i to 
 
 ^a 'S 
 
 — 
 
 5 
 
 ^ ^ 
 
32 
 
 Thus, with a 12-inch gun, it would be useless to fire at 12 inches of 
 armour (1 calibre) unless the shot would strike with over 1000 feet 
 velocity. 15 inches of armour (1£ calibres) would require over 1J 
 thousand, that is, over 1250 feet. To sum up shortly with one or two 
 practical directions. 
 
 Practical Directions. 
 
 To estimate quickly the power of a gun in perforating wrought-iron 
 armour with service guns and projectiles, 1 consult Inglis' diagram, or 
 Maitland's if preferred. Maitland's may be used in other cases. 
 
 Failing diagrams, take the calibre of the gun as a measure, and 
 reckon that at least 1000 feet velocity is required for each calibre in 
 the thickness of the armour to be perforated. In old type guns, with 
 light shot, considerably more velocity may be required. 
 
 For deliberate calculation, when stored-up work as well as perforation 
 is required, Fairbairn's formula with the corrections for K, given here- 
 after, is recommended. In doing this, it may be observed that the 
 formulae 2 used at Gavre and in Italy closely resemble it in form ; 
 Krupp's formula, only, assuming the resistance to be proportional to 
 area of hole perforate J. To facilitate calculation for those who may 
 only occasionally perform it, an example is worked out on Fairbairn's 
 system as here recommended, and all the necessary data for any 
 calculation given, that is, a table of four-figure logs, and a table of 
 corrections of K are supplied with it. 
 
 1 If the charge as well as the projectile is that given on Inglis' diagram, the perforation at any 
 range may be at once read off. If the charge is not that given on the diagram, the velocity at any 
 range will not agree with that given there, and it will be necessary to know the striking velocity 
 and to read off the perforation corresponding to that velocity, though found at a different range 
 from that in question. 
 
 - Tho Gavre formula is V m = 1600td /_iLin iron; in wood V m = 95td / _ L, where td = 
 
 V "k v/ ^k 
 
 thickness in decimeters ; Da = diameter in decimetres ; Wt = weight in kilogrammes ; and V m = 
 velocity in metres. 
 
 , ■">- .. 1 t c 
 
 Tho Italian formula is t = 
 
 " /^ J . ± . _> . Krupp's is HH x -L = *«_' / *• . 
 ^ 2g ttD 4-154 rr 2g A c 10 / T) c ' 
 
 where Ac — area of cross section of projectile in centimetres, D c diameter in centimetres, and 
 t thickness in centimetres. 
 
 The French formula may be written, of course, t = vk / , which differs only from Fair- 
 bairn's by k the constants, and root being different. — vide "Development of Armour" ; Very.p. 581. 
 
33 
 
 rOUK-riGURE LOGARITHMS. 
 
 No. 
 
 
 
 1 
 
 2 
 
 3 
 
 4 
 
 5 
 
 6 
 
 7 
 
 8 
 
 9 
 
 1 
 
 o 
 
 3 
 
 4 
 
 5 
 
 6 
 
 7 
 
 29 
 
 8 
 33 
 
 9 
 37 
 
 10 
 
 0000 
 
 0043 
 
 0086 
 
 0128 
 
 0170 
 
 0212 
 
 0253 
 
 0294 
 
 0334 
 
 0374 
 
 4 
 
 8 
 
 12 
 
 17 
 
 21 
 
 25 
 
 11 
 
 12 
 13 
 
 0414 
 0792 
 1139 
 
 0453 
 0828 
 1173 
 
 0492 
 0861 
 1206 
 
 0531 
 0899 
 1239 
 
 0569 
 0934 
 1271 
 
 0607 
 0969 
 1303 
 
 0645 
 1004 
 1335 
 
 0682 
 1038 
 1367 
 
 0719 
 1072 
 1399 
 
 0755 
 1106 
 1430 
 
 4 
 3 
 3 
 
 8 
 7 
 6 
 
 11 
 10 
 
 10 
 
 15 
 14 
 13 
 
 19 
 17 
 16 
 
 23 
 21 
 
 19 
 
 26 
 24 
 23 
 
 30 
 28 
 26 
 
 34 
 31 
 29 
 
 14 
 15 
 16 
 
 1461 
 1761 
 2041 
 
 1492 
 1790 
 2068 
 
 1523 
 1818 
 2095 
 
 1553 
 
 1847 
 2122 
 
 1584 
 1875 
 2148 
 
 1614 
 1903 
 2175 
 
 1644 
 1931 
 2201 
 
 1673 
 1959 
 2227 
 
 1703 
 1987 
 2253 
 
 1732 
 2014 
 2279 
 
 3 
 3 
 3 
 
 6 
 6 
 5 
 
 9 
 
 8 
 8 
 
 12 
 11 
 11 
 
 15 
 14 
 13 
 
 18 
 17 
 10 
 
 21 
 
 20 
 18 
 
 24 
 22 
 21 
 
 27 
 25 
 24 
 
 17 
 18 
 19 
 
 2304 
 2553 
 2788 
 
 2330 
 2577 
 2810 
 
 2355 
 2601 
 2833 
 
 2380 
 2625 
 2856 
 
 2405 
 
 2648 
 2878 
 
 2430 
 2672 
 2900 
 
 2455 
 2695 
 2923 
 
 2480 
 2718 
 2945 
 
 2504 
 2742 
 2967 
 
 2529 
 
 2765 
 2989 
 
 2 
 2 
 2 
 
 5 
 
 5 
 4 
 
 7 
 7 
 7 
 
 10 
 9 
 9 
 
 12 
 12 
 11 
 
 15 
 14 
 13 
 
 17 
 16 
 
 16 
 
 20 
 19 
 18 
 
 22 
 21 
 20 
 
 20 
 
 3010 
 
 3032 
 
 3054 
 
 3075 
 
 3096 
 
 3118 
 
 3139 
 
 3160 
 
 3181 
 
 3201 
 
 2 
 
 4 
 
 6 
 
 8 
 
 11 
 
 13 
 
 15 
 
 17 
 
 19 
 
 21 
 22 
 23 
 
 3222 
 3424 
 3617 
 
 3213 
 3444 
 3636 
 
 3263 
 3461 
 3655 
 
 3284 
 3483 
 3674 
 
 3304 
 3502 
 3692 
 
 3324 
 3522 
 3711 
 
 3345 
 3541 
 3729 
 
 3365 
 3560 
 3747 
 
 3385 
 3579 
 3766 
 
 3404 
 3598 
 
 3784 
 
 2 
 2 
 
 2 
 
 4 
 4 
 
 4 
 
 6 
 6 
 6 
 
 8 
 
 8 
 7 
 
 10 
 
 10 
 
 9 
 
 12 
 12 
 11 
 
 14 
 14 
 13 
 
 16 
 15 
 15 
 
 18 
 17 
 17 
 
 24 
 25 
 26 
 
 3802 
 3979 
 4150 
 
 3820 
 3997 
 4166 
 
 3838 
 4014 
 4183 
 
 3856 
 4031 
 4200 
 
 3874 
 4048 
 4216 
 
 3892 
 4065 
 4232 
 
 3909 
 4082 
 4249 
 
 3927 
 4099 
 4265 
 
 3945 
 4116 
 4281 
 
 3962 
 4133 
 4298 
 
 2 
 2 
 2 
 
 4 
 3 
 3 
 
 5 
 5 
 5 
 
 7 
 7 
 7 
 
 9 
 9 
 8 
 
 11 
 10 
 10 
 
 12 
 12 
 11 
 
 14 
 14 
 13 
 
 16 
 15 
 15 
 
 27 
 28 
 29 
 
 4314 
 4472 
 4624 
 
 4330 
 4487 
 4639 
 
 4346 
 
 4502 
 4654 
 
 4362 
 4518 
 4669 
 
 4378 
 4533 
 4683 
 
 4393 
 
 4548 
 4698 
 
 4409 
 4561 
 4713 
 
 4425 
 4579 
 
 4728 
 
 4440 
 4594 
 
 4742 
 
 4456 
 4609 
 4757 
 
 2 
 2 
 
 1 
 
 3 
 3 
 3 
 
 5 
 5 
 4 
 
 6 
 6 
 6 
 
 8 
 8 
 7 
 
 9 
 9 
 9 
 
 11 
 11 
 10 
 
 13 
 
 12 
 12 
 
 14 
 14 
 13 
 
 30 
 
 4771 
 
 4786 
 
 4800 
 
 4814 
 
 4829 
 
 4843 
 
 4857 
 
 4871 
 
 4886 
 
 4900 
 
 1 
 
 3 
 
 4 
 
 6 
 
 7 
 
 9 
 
 10 
 
 11 
 
 13 
 
 31 
 32 
 33 
 
 4914 
 5051 
 5185 
 
 4928 
 5065 
 5198 
 
 4942 
 5079 
 5211 
 
 4955 
 5092 
 5224 
 
 4969 
 5105 
 6237 
 
 4983 
 5119 
 5250 
 
 4997 
 5132 
 5263 
 
 5011 
 5145 
 5276 
 
 5024 
 5159 
 
 5289 
 
 5038 
 5172 
 5302 
 
 1 
 1 
 1 
 
 3 
 3 
 3 
 
 4 
 4 
 4 
 
 6 
 
 5 
 5 
 
 7 
 7 
 6 
 
 8 
 8 
 8 
 
 10 
 9 
 9 
 
 11 
 11 
 
 10 
 
 12 
 12 
 12 
 
 34 
 35 
 36 
 
 5315 
 5441 
 6563 
 
 5328 
 6463 
 
 5575 
 
 5340 
 5465 
 
 5587 
 
 5353 
 
 5478 
 5599 
 
 5366 
 5490 
 5611 
 
 5378 
 5502 
 5623 
 
 5391 
 5514 
 5635 
 
 5403 
 5527 
 5647 
 
 5416 
 5539 
 5658 
 
 5428 
 5551 
 5670 
 
 1 
 1 
 1 
 
 3 
 2 
 2 
 
 4 
 4 
 4 
 
 5 
 
 5 
 6 
 
 6 
 6 
 6 
 
 8 
 7 
 7 
 
 9 
 9 
 
 8 
 
 10 
 10 
 10 
 
 11 
 11 
 11 
 
 37 
 38 
 39 
 
 5682 
 5798 
 5911 
 
 5694 
 5809 
 5922 
 
 5705 
 5 521 
 5933 
 
 5717 
 5832 
 5944 
 
 5729 
 5843 
 5955 
 
 5740 
 5855 
 5966 
 
 5752 
 5866 
 5977 
 
 5763 
 
 5877 
 5988 
 
 5775 
 5888 
 5999 
 
 5788 
 5899 
 6010 
 
 1 
 
 1 
 1 
 
 2 
 
 2 
 2 
 
 3 
 3 
 3 
 
 5 
 5 
 4 
 
 6 
 
 6 
 6 
 
 7 
 
 7 
 7 
 
 8 
 
 8 
 8 
 
 9 
 9 
 9 
 
 10 
 10 
 10 
 
 40 
 
 6021 
 
 6031 
 
 6042 
 
 6053 
 
 6064 
 
 6075 
 
 6085 
 
 6096 
 
 6107 
 
 6117 
 
 1 
 
 2 
 
 3 
 
 4 
 
 5 
 
 6 
 
 7 
 
 9 
 
 10 
 
 41 
 42 
 43 
 
 6128 
 6232 
 6335 
 
 6138 
 6243 
 6345 
 
 6149 
 6253 
 6355 
 
 6160 
 6263 
 6365 
 
 6170 
 6274 
 6375 
 
 6180 
 6281 
 6385 
 
 6191 
 6294 
 6395 
 
 6201 
 6304 
 6405 
 
 6212 
 6314 
 6416 
 
 6222 
 6325 
 6425 
 
 1 
 1 
 1 
 
 2 
 2 
 2 
 
 3 
 3 
 3 
 
 4 
 4 
 4 
 
 5 
 5 
 5 
 
 6 
 6 
 6 
 
 7 
 
 7 
 7 
 
 8 
 
 8 
 8 
 
 9 
 9 
 9 
 
 44 
 45 
 46 
 
 6435 
 6532 
 6628 
 
 6444 
 6542 
 6637 
 
 6454 
 6551 
 6646 
 
 6464 
 6561 
 6656 
 
 6474 
 6571 
 6665 
 
 6484 
 6580 
 6675 
 
 6493 
 6590 
 6684 
 
 6503 
 6599 
 6693 
 
 6513 
 6609 
 6702 
 
 6522 
 6618 
 6712 
 
 1 
 1 
 
 1 
 
 2 
 2 
 2 
 
 3 
 3 
 3 
 
 4 
 
 4 
 4 
 
 6 
 5 
 5 
 
 6 
 6 
 
 6 
 
 7 
 7 
 7 
 
 8 
 
 8 
 7 
 
 9 
 9 
 
 8 
 
 47 
 48 
 49 
 
 6721 
 6812 
 6902 
 
 6730 
 6821 
 6911 
 
 6739 
 6830 
 6920 
 
 6749 
 6839 
 6928 
 
 6758 
 6848 
 6937 
 
 6767 
 6857 
 6946 
 
 6776 
 6866 
 6955 
 
 6785 
 6875 
 6964 
 
 6794 
 6884 
 6972 
 
 6803 
 6893 
 6981 
 
 1 
 1 
 1 
 
 2 
 
 2 
 2 
 
 3 
 3 
 3 
 
 4 
 
 4 
 4 
 
 5 
 
 4 
 4 
 
 5 
 6 
 5 
 
 6 
 6 
 6 
 
 7 
 
 7 
 7 
 
 8 
 8 
 8 
 
 50 
 
 6990 
 
 6998 
 
 7007 
 
 7016 
 
 7024 
 
 7033 
 
 7042 
 
 7050 
 
 7059 
 
 7067 
 
 1 
 
 2 
 
 3 
 
 3 
 
 4 
 
 5 
 
 6 
 
 7 
 
 8 
 
 51 
 52 
 53 
 64 
 
 7076 
 7160 
 7243 
 7324 
 
 7084 
 7168 
 7251 
 7332 
 
 7093 
 7177 
 7259 
 7340 
 
 7101 
 7185 
 7267 
 7348 
 
 7110 
 7193 
 7275 
 7356 
 
 7118 
 7202 
 7284 
 7364 
 
 7126 
 7210 
 7292 
 7372 
 
 7135 
 7218 
 7300 
 7380 
 
 7143 
 7226 
 7308 
 7388 
 
 7152 
 7235 
 7316 
 7396 
 
 1 
 1 
 1 
 1 
 
 2 
 
 2 
 2 
 2 
 
 3 
 
 2 
 
 2 
 2 
 
 3 
 3 
 3 
 3 
 
 4 
 4 
 4 
 4 
 
 6 
 
 6 
 5 
 6 
 
 6 
 6 
 6 
 6 
 
 7 
 7 
 6 
 6 
 
 8 
 7 
 7 
 7 
 
34 
 
 Foue-Fiqci!e Logarithms — (Continued), 
 
 No. 
 
 55 
 56 
 
 
 
 1 
 
 7412 
 7490 
 
 2 
 
 7419 
 7497 
 
 3 
 
 4 
 
 5 
 
 6 
 
 7 
 
 8 
 
 9 
 
 1 
 
 2 
 
 3 
 
 4 
 
 6 
 
 6 
 
 7 
 
 8 
 
 9 
 
 7401 
 
 7482 
 
 7127 
 7506 
 
 7435 
 7513 
 
 7443 
 7520 
 
 7451 
 7528 
 
 7459 
 7536 
 
 7466 
 7543 
 
 7474 
 7561 
 
 
 2 
 2 
 
 2 
 
 2 
 
 3 
 3 
 
 4 
 
 4 
 
 5 
 6 
 
 5 
 6 
 
 6 
 6 
 
 7 
 7 
 
 57 
 58 
 59 
 
 7559 
 7634 
 7709 
 
 7666 I 7574 
 71118 7649 
 7716 7723 
 
 7582 
 7657 
 7731 
 
 7589 
 7664 
 7738 
 
 7597 7604 
 7672 7679 
 7715 ! 7752 
 
 7612 7619 
 7686 ! 7694 
 7760 7767 
 
 7627 
 7701 
 7774 
 
 
 2 
 1 
 1 
 
 o 
 2 
 2 
 
 3 
 3 
 3 
 
 4 
 4 
 4 
 
 5 
 
 4 
 4 
 
 6 
 5 
 
 5 
 
 6 
 6 
 
 6 
 
 7 
 7 
 7 
 
 On 
 
 7782 
 
 7789 , 7796 , 7803 
 
 7810 
 
 7818 7825 
 
 7832 7839 
 
 7816 
 
 
 1 
 
 2 
 
 3 
 
 4 
 
 4 
 
 5 
 
 6 
 
 6 
 
 61 
 02 
 63 
 
 7853 
 7924 
 7993 
 
 7860 ' 7868 
 7931 I 7938 
 Soon 1 8007 
 
 7875 
 7915 
 8014 
 
 7882 
 7962 
 8021 
 
 7889 
 79-59 
 8o28 
 
 7896 
 7966 
 8035 
 
 7903 
 7973 
 8041 
 
 7910 
 
 79S0 
 8048 
 
 7917 
 
 7987 
 80.55 
 
 
 1 
 
 1 
 1 
 
 2 
 2 
 2 
 
 3 
 3 
 3 
 
 4 
 3 
 3 
 
 4 
 4 
 4 
 
 6 
 6 
 5 
 
 6 
 
 6 
 5 
 
 6 
 6 
 6 
 
 64 
 65 
 66 
 
 8062 
 8129 
 8195 
 
 8069 
 8136 
 
 82( 12 
 
 81 >75 8l 182 
 8142 8149 
 8209 8215 
 
 8089 
 8156 
 8222 
 
 8096 
 8162 
 8228 
 
 8102 
 8169 
 8235 
 
 8109 8116 
 8176 : 8182 
 8211 8218 
 
 8122 
 8189 
 
 82.54 
 
 
 1 
 1 
 1 
 
 o 
 
 2 
 
 2 
 
 3 
 3 
 3 
 
 3 
 3 
 3 
 
 4 
 4 
 4 
 
 6 
 6 
 5 
 
 5 
 5 
 5 
 
 6 
 6 
 6 
 
 67 
 68 
 69 
 
 8261 
 8325 
 
 8388 
 
 8267 
 
 8.1H1 
 8395 
 
 8271 
 8338 
 8401 
 
 8280 
 8314 
 8lo7 
 
 8287 
 8: ',51 
 8414 
 
 8293 
 8357 
 
 812(1 
 
 8299 
 8363 
 
 8426 
 
 83(Mi 8312 
 8370 8376 
 8112 8439 
 
 8319 
 8382 
 8415 
 
 
 1 
 
 1 
 1 
 
 2 
 
 o 
 2 
 
 3 
 3 
 3 
 
 3 
 3 
 3 
 
 4 
 4 
 4 
 
 5 
 4 
 4 
 
 6 
 5 
 5 
 
 6 
 6 
 
 6 
 
 70 
 
 8451 
 
 i 
 8157 8463 
 
 8170 \ 8476 
 
 8182 
 
 8488 
 
 8191 
 
 8500 
 
 8506 
 
 
 1 
 
 2 
 
 2 
 
 3 
 
 4 
 
 4 
 
 5 
 
 6 
 
 71 
 72 
 73 
 
 8513 
 
 8573 
 8033 
 
 8519 8525 ' 8,531 ' 8537 
 8579 8585 ' 8591 | 8,597 
 811 i'J 8645 8ti,31 | Si;,",7 
 
 85 13 
 8603 
 8663 
 
 8519 
 86O0 
 8669 
 
 8,555 
 8615 
 8675 
 
 8561 
 
 8621 
 8481 
 
 8.567 
 
 8627 
 86S6 
 
 
 1 
 1 
 
 1 
 
 2 
 o 
 
 2 
 2 
 
 3 
 3 
 3 
 
 4 
 4 
 4 
 
 4 
 4 
 4 
 
 5 
 5 
 
 5 
 
 5 
 5 
 5 
 
 74 
 75 
 76 
 
 8 IW 2 
 8761 
 8808 
 
 8li!l8 ' 87(11 ! 8710 8716 
 8756 8762 : 8768 8774 
 88 11 8820 8825 , 8831 
 
 8722 ' 8727 
 8779 8785 
 88 17 8812 
 
 8733 
 
 8791 
 
 8818 
 
 8739 
 8797 
 8851 
 
 8715 
 8802 
 8859 
 
 
 1 
 
 1 
 1 
 
 o 
 2 
 
 2 
 
 2 
 2 
 
 3 
 3 
 3 
 
 4 
 3 
 3 
 
 4 
 4 
 4 
 
 6 
 5 
 5 
 
 5 
 5 
 5 
 
 77 
 78 
 79 
 
 8865 1 8871 8876 888'J 
 8921 8'.'27 i 8932 8938 
 8976 8982 ! 81187 < 8993 
 
 8887 
 8943 
 89P8 
 
 88! 13 
 8919 
 9001 
 
 8899 
 8954 
 9009 
 
 8901 
 
 8960 
 9015 
 
 8910 
 8965 
 
 9020 
 
 8915 
 
 8971 
 9025 
 
 
 1 
 
 1 
 1 
 
 2 
 2 
 2 
 
 2 
 2 
 
 2 
 
 3 
 3 
 3 
 
 3 
 3 
 3 
 
 4 
 4 
 4 
 
 4 
 4 
 4 
 
 5 
 5 
 5 
 
 80 
 
 9031 
 
 9036 
 
 9042 9047 
 
 9053 
 
 9058 
 
 9063 
 
 9069 
 
 9074 
 
 9079 
 
 
 1 
 
 2 
 
 
 
 3 
 
 3 
 
 4 
 
 4 
 
 5 
 
 81 
 82 
 83 
 
 9085 
 9138 
 9191 
 
 9096 
 9143 
 9196 
 
 9096 
 9149 
 9201 
 
 9101 
 9151 
 9206 
 
 9106 
 9159 
 9212 
 
 9112 
 
 9165 
 9217 
 
 9117 , 9122 
 9! 70 1 9175 
 9222 9227 
 
 9128 
 9180 
 9232 
 
 9133 
 9186 
 9238 
 
 
 1 
 
 1 
 1 
 
 2 
 2 
 2 
 
 o 
 2 
 2 
 
 3 
 3 
 3 
 
 3 
 3 
 3 
 
 4 
 4 
 4 
 
 4 
 4 
 4 
 
 5 
 5 
 5 
 
 81 
 86 
 86 
 
 9243 
 9291 
 9345 
 
 9218 
 9299 
 
 9350 
 
 9253 
 9304 
 9355 
 
 9258 
 9309 
 9360 
 
 9263 
 9315 
 9365 
 
 9269 
 9320 
 9370 
 
 9271 
 9325 
 9375 
 
 9279 
 9330 
 9380 
 
 9284 
 9335 
 9386 
 
 9289 
 9340 
 9390 
 
 
 1 
 1 
 1 
 
 2 
 2 
 
 O 
 
 o 
 
 2 
 2 
 
 3 
 3 
 3 
 
 3 
 3 
 3 
 
 4 
 4 
 4 
 
 4 
 
 4 
 4 
 
 5 
 
 5 
 5 
 
 87 
 88 
 89 
 
 9395 
 
 9445 
 9494 
 
 9400 
 915U 
 9499 
 
 9405 
 9455 
 9604 
 
 9410 
 9160 
 9509 
 
 9415 
 9465 
 9513 
 
 9420 
 9469 
 9618 
 
 9425 
 9474 
 9523 
 
 9430 
 C479 
 9528 
 
 9435 
 9484 
 9533 
 
 9440 
 9489 
 9538 
 
 
 
 
 
 
 1 
 1 
 1 
 
 1 
 1 
 
 1 
 
 2 
 
 2 
 o 
 
 2 
 2 
 2 
 
 3 
 3 
 3 
 
 3 
 3 
 3 
 
 4 
 4 
 
 4 
 
 4 
 4 
 4 
 
 90 
 
 9542 
 
 9547 
 
 9552 
 
 9557 
 
 9562 
 
 9566 
 
 9571 
 
 9576 
 
 9581 
 
 9586 
 
 
 
 1 
 
 1 
 
 2 
 
 2 
 
 3 
 
 3 
 
 4 
 
 4 
 
 91 
 92 
 93 
 
 9590 
 9638 
 9686 
 
 9596 
 9643 
 9689 
 
 9600 
 9647 
 9694 
 
 9605 
 9652 
 9699 
 
 9609 
 9657 
 9703 
 
 9614 
 9661 
 9708 
 
 9619 
 9666 
 9713 
 
 9624 
 9671 
 9717 
 
 9628 
 9675 
 9722 
 
 9633 
 9680 
 
 9727 
 
 
 
 
 
 1 
 
 1 
 1 
 
 1 
 1 
 1 
 
 O 
 
 2 
 
 2 
 
 2 
 2 
 2 
 
 3 
 3 
 3 
 
 3 
 3 
 3 
 
 4 
 4 
 4 
 
 4 
 
 4 
 4 
 
 94 
 95 
 96 
 
 9731 
 9777 
 9823 
 
 9736 
 9782 
 9827 
 
 9741 
 9786 
 9832 
 
 9745 
 9791 
 9836 
 
 9760 
 9795 
 9841 
 
 9754 
 9800 
 9846 
 
 9759 9763 
 9805 1 9809 
 9850 , 9864 
 
 9768 
 9814 
 9859 
 
 9773 
 9818 
 9863 
 
 
 
 
 
 1 
 1 
 1 
 
 1 
 1 
 1 
 
 o 
 
 2 
 
 2 
 
 2 
 2 
 2 
 
 3 
 3 
 3 
 
 3 
 3 
 3 
 
 4 
 4 
 
 4 
 
 4 
 4 
 4 
 
 97 
 98 
 99 
 
 9868 
 9912 
 9966 
 
 9872 
 9917 
 9961 
 
 9877 
 9921 
 9965 
 
 9881 
 9926 
 9969 
 
 9886 
 9930 
 9974 
 
 9890 
 9934 
 9978 
 
 9894 ; 9899 
 9939 1 9943 
 9983 . 9987 
 
 9903 
 9948 
 9991 
 
 9908 
 9952 
 9996 
 
 
 
 
 
 1 
 1 
 
 1 
 
 1 
 1 
 1 
 
 2 
 2 
 2 
 
 2 
 
 2 
 
 2 
 
 3 
 3 
 3 
 
 3 
 3 
 
 3 
 
 4 
 4 
 
 3 
 
 4 
 
 4 
 4 
 
35 
 
 Example. — At Spezia, on November 16th, 1882, the first projectile 
 fired from the 100-ton M.L. gun struck Cammell's plate with a striking 
 velocity of 1219 feet per second; its weight was 2000 lbs., and its 
 diameter 17"64 inches. 
 
 The formula is t = I ~^~ x _ x -; 
 V iff ttB K 
 
 that is, 
 
 log t = | (log W + 2 logu — log 2$ — log ttB — log E). 
 
 Here W= 2000, v = 1219, B = 17-64. 
 The constant Hog(2ff X 2240) = 5-1590 Log 2000= 3-3010 
 
 Table of logs of 
 correction K. 
 
 V 
 
 in- feet. 2 
 
 log of 
 K. 
 
 600 
 
 01139 
 
 600 
 
 0-1004 
 
 700 
 
 0-0719 
 
 800 
 
 00511 
 
 900 
 
 0-0414 
 
 1000 
 
 0-0212 
 
 1100 
 
 0-0086 
 
 1200 
 
 0-0000 
 
 1300 
 1400 
 
 1-9912 
 1-9845 
 
 1600 
 
 1-9777 
 
 1600 
 
 1-9731 
 
 1700 
 
 1-9686 
 
 1800 
 
 1-9638 
 
 1900 
 
 1-9590 
 
 2000 
 
 1-9590 
 
 Constant log tt = 0'4971 
 
 2 log 1219 
 
 _ f 3085E 
 
 ~ (3-085J 
 
 log 17-64 = 12465. 9-4728 
 
 Here, adding these, log ttD= 1-7436. log Zg = 5-1590 
 
 E, striking euerg 
 
 
 = 20600 ft.-tons 
 
 4-3138 
 logirD=l-7436 
 
 e, energy per inch 7 ^> w 1 „_, „ ri , 
 
 X -—-=371-7 ft.-tons 
 
 circumference ) 
 
 *9 
 
 3 log IT for 1200 taken from table = 
 
 25702 
 
 00000 
 2)2-5702 
 1-2851 
 
 t, thickness of wrought-iron to be perforated 19'26 
 
 For plate-upon-plate in 2 layers t should be multiplied by f-f . 
 „ « 3 layers t « » ^ . 
 
 The rule of thumb here gives 17'6 x 1"2 = 21*12 ins., which is an 
 unusually bad result ; this gun being the one for which this rule gives 
 the largest error shown on the table on p. 31. 
 
 Hard Armour. 
 
 In compound or steel-faced armour, actual perforation has occasionally 
 been obtained ;* occasionally also a steel shot has set up and driven a 
 large disc out of a compound plate, which action may partake of the 
 character of perforation, and its diameter may bear some relation to 
 that of the projectile. Most compound and solid steel shields, however, 
 and all chilled iron shields must be destroyed by fracture, the shot's 
 point penetrating only to a certain depth, often quite an insignificant 
 one. Here, then, fracture is caused by a blow delivered, as it were, on 
 an ogival-pointed wedge which splits the shield asunder. This action 
 differs widely from perforation ; for while in both cases the stored-up 
 work or energy is the motive power, in perforation the thickness 
 
 1 g is multiplied by 2240 to bring lbs. to tons, the shot's weight being given in lbs., and the 
 energy required in foot-tons. 
 
 2 The velocity is here substituted for the relation — in this column, being approximately pro- 
 portional to it, as stated in Note 1, p. 28, and explained on pp. 29 and 30. 
 
 3 As the striking velocity generally exceeds 1200 feet, the correction for K generally consists in 
 subtracting a log with a minus 1 before it, thus slightly increasing the figures. 
 
 4 Vide Palliser improved shot trials, &c. 
 
36 
 
 perforated depends inversely on the size of the hole or diameter of 
 shot ; whereas in destruction by fracture, the point only of the shot 
 enters the plate, and its diameter can scarcely enter into the question. 
 As has been before said, when all other conditions are the same, it may be 
 supposed that fracture effected on any given plate may be simply pro- 
 portional to the stored-up work ; but this, if a fact, is of little use to 
 us, for other conditions will very seldom be the same as that of some 
 known example. It may apply to the case of different guns attacking 
 the same shield if their projectiles are of the same kind, differing only 
 in dimensions or weight, but even then, unless the shot hold together 
 on impact, which at present hardly ever occurs agaiust hard armour, 
 there are uncertain elements. A higher striking velocity may cause 
 one projectile to break up more abruptly, and deliver less of its energy 
 on the shield usefully than another with a lower striking velocity, also 
 the mechanical action of the breaking up of projectile can hardly fail to 
 be affected in some degree by its diameter, and hence even here the 
 problem may be more difficult than that of perforation. Directly 
 changes are made in the shield the problem is altogether beyond the 
 present powers of artillerists to grapple with, as may be seen easily. 
 
 The method adopted of matching a shot against a steel-faced or 
 steel plate has generally been to give the projectile sufficient energy to 
 perforate wrought-iron of the same thickness, or else 25 per cent, more 
 thickness than the compound or steel plate attacked. 
 
 The table following shows this to have been the method followed at 
 Shoeburyness, at Spezia, and at St. Petersburg in the experiments 
 referred to. In each case, except the last, 1 the power of the shot was 
 calculated evidently on the basis of perforation ; yet, as may be seen 
 by referring to the accounts of the experiments (Chap. VII.), the 
 plates were destroyed by fracture. That perforating power was the 
 measure used is made apparent by comparing columns 5 and 6 
 together, when it will be seen that a low and high test appear to be 
 framed in each case. On the low test, the shot had the energy 
 necessary to perforate a thickness of wrought-iron equal to that of the 
 compound plate attacked, and for the high test the shot had energy 
 sufficient to perforate a wrought-iron plate of 24 to 33 per cent, greater 
 thickness than the compound plate attacked. It is easy to see that this 
 is a totally wrong principle, 3 but it is not at all so easy to propose a 
 better one. 
 
 1 The last was a totally different kind of experiment, to be described hereafter in Chapter VII. 
 
 2 The author has illustrated the different actions of perforation and smashing at Woolwich, the 
 United Service Institution, and the Iron and Steel Institute by means of a dropping apparatus, or 
 pile-driver, consisting of a Irame in which a weight fell, driving punches or bullets inserted in it 
 into soft or hard substances such as admitted of perforation, or yielded bv fracture respectively. 
 
 Slabs of millboard and bard tiles were found to answer fairly well." Steel punches one inch, half- 
 
 an-inch, and a quarter-of-an-inch in diameter— were employed, with ogival heads corresponding to 
 those of service projectiles. As the velocity was that of the falling weight, the striking energy 
 -^- was equal to W S. (weight x height), and Fairbairn's equation might be written 
 
 WH = ird Pk. Nothing could be much more simple than this, and the effect of one element 
 on another could be easily seen by trial, leaving all other conditions constant. Thus in perforation 
 it could bo shown that to perforate a givcn.slab the height of fall must be nearly in direct proportion 
 to the diameter of the shot or punch. When the i-inch required 17 inches fall, the inch required 
 60. To agree exactly with the equation the three punches should require IS, 30, and 60 inches fall 
 for the same slab, and the more nearly the work was localized and brought to clean punching the 
 more nearly was this the caBe. With the hard brick or tile, on the other hand, fracture was caused 
 
37 
 
 Tab£E showing that in recent experiments shots have been matched against compound 
 plates according to their calculated powers of perforation (vide Columns 5 and 6), 
 not according to their smashing powers, for which Column 9 is suggested as a 
 measure. 
 
 1. 
 
 2. 
 
 3. 
 
 4. 
 
 5. 
 
 6. 
 
 7. 
 
 8. 
 
 9. 
 
 10. 
 
 Date and Place 
 of Trial. 
 
 Gun. 
 
 Weight 
 of Shot. 
 
 '3 
 o 
 
 .a 
 % 
 
 a 
 
 « o 
 ^ a 
 
 -*3 
 
 •ats a 
 
 ogf 
 
 §5 M 
 
 CM 
 
 6 
 
 a 
 
 3 
 "o 
 
 Is 
 
 s 
 
 o 
 
 O 
 
 a 
 o 
 H . 
 
 hs 
 
 t.,Pn 
 bo 
 & 
 fl 
 
 Remarks on 
 Effects. 
 
 July 1, 1880, 
 Shoeburyness. 
 
 1882, Shoebury- 
 ness. 
 
 November, 1882,') 
 
 November, 1882,") 
 and March, / 
 1883, Ochta, f 
 St. PeterBburgJ 
 
 Sept. 10, 1883, 
 Shoeburyness. 
 
 38-ton 
 43-ton 
 
 100-ton 
 
 11-in.B.L. 
 
 80-ton 
 
 lbs. 
 828 
 
 708-5 
 2000 ) 
 
 653-5 j 
 1700 
 
 feet. 
 1504 
 
 1870 
 
 1219 
 
 1565 
 
 1167 
 1506 
 
 1599 
 
 ins. 
 18 
 
 18 
 
 18-9 
 
 18-9 
 
 12 
 
 12 
 
 12 
 
 18-7 
 22-3 
 19-3 
 
 26-2 
 
 12-2 
 16-3 
 
 26-3 
 
 ft.-tons. 
 12,980 
 
 17,180 
 
 20,600 
 
 33,960 
 
 5,228 
 8,704 
 
 30,140 
 
 tons, 
 24 
 
 31-6 
 
 31-5 
 
 12-25 
 12-25 
 
 10-5 
 
 541 
 
 664 
 
 1046 
 
 427 
 711 
 
 2870 
 
 Face cracks and 
 
 dents only. 
 Plate completely 
 
 broken up. 
 Large portion of 
 
 corner broken 
 
 away. 
 Plate broken to 
 
 pieces. 
 /"Portion of face 
 J detached and 
 j facecrackspro- 
 (. duced. 
 Plate bent and 
 
 cracked severely. 
 
 Both, in England and in Germany a check has been attempted by 
 treating the question in the opposite way, that is to say, by considering 
 that the hard armour is fractured by delivering a shock through its 
 mass that severs it along lines of least resistance. Thus, if the striking 
 energy of the shot be divided by the weight of the armour in tons, the 
 quotient gives trie number of foot-tons energy delivered on every ton 
 of armour, vide column 9 in table. This, however, is clearly liable to 
 be as far wrong as the perforation system of calculation. There must 
 be some limit as to the distance of the mass from the point of impact 
 that could be thus dealt with. Two long narrow plates, for example, 
 might be taken, one double the leDgth the other. Both would snap 
 across, and with nearly the same blow probably, 1 yet the mass of one 
 being double that of the other, the blind application of the rule just 
 given would indicate that one would absorb twice the shock of the 
 other before fracture occurred. 
 
 The question of fracture is a difficult one. It has been said to be 
 altogether beyond mathematical calculation. It must, however, follow 
 
 by a given weight falling from a given height without regard to the size of the punch which was 
 employed as a striker. For perforation, then, the inch punch required four times the fall of the 
 ^-inch (when both were driven by the eame weight), because it made a hole of four times the 
 diameter, but for fracture, it required only the same fall, for its point only came in contact and 
 split the brick by its striking energy, without reference to the diameter of the punch. A toy of 
 this kind is only good for illustration, but the effects are simple and striking, and the holes in 
 the millboard so closely resembled perforations through iron armour, that a photograph of one 
 can hardly be detected from that of the other. Admiral Fishbourne, it appears, suggested that a 
 pile-driver might be used for illustration, many years ago, but the author had never heard of it. 
 1 Theoretically the long one might break the more easily of the two. 
 
 6 
 
38 
 
 certain laws if uniformity of quality could be secured, for even steel is 
 not really capricious. The elements certainly are troublesome ones. 
 As to dimensions, the minimum cross measurement is the line of most 
 probable fracture, but bolt-holes and other features have their influence. 
 Cracking itself is a complicated action. Clearly the first half of a crack 
 represents much more work than the completion of it, consequently an 
 increase in width of plate would not probably give a plate a propor- 
 tional increase in resisting power. It is suggested that the subject 
 might be approached on a small scale, 1 that steel bullets might be fired 
 against small slabs of steel and chilled iron. These slabs should be 
 sufficiently long to ensure a line of least resistance in one direction, 
 which would probably depend on its actual dimensions ; a disc or any 
 plate where the direction of the line of least resistance might probably 
 be determined by a flaw would be the worst kind to employ. By firing 
 at very great numbers of small plates with all the conditions fixed, 
 except the one at the moment under investigation, it is probable that 
 something might be learned of the laws of fracture under impact. In 
 the meantime, little can be said definitely about it, beyond the 
 elementary but important fact that effect is not proportional to the 
 shot's perforation, but much more nearly to its total energy, a con- 
 sideration that may actually effect the selection of the guns employed 
 against hard shields on service; 2 if, however, armour should in the 
 long run be made hard enough to resist perforation proper, that is, per- 
 foration without breaking up the armour, it follows that the destructive 
 powers of guns will depend not upon their power of perforation but 
 upon the stored-up work or energy. The value of a gun may generally, 
 therefore, be estimated on the measure used by Krupp, that is, the 
 energy per ton of gun. New type guns will not then benefit by the 
 fact that tbe reduced diameter of the projectile demands a smaller 
 hole in tbe plate; but they will benefit by the increase in total striking 
 energy. For example, the new 63-ton gun of 13"5-in. calibre of 1884 3 
 has a velocity of 2050 feet and a perforation at the muzzle of about 30 
 inches of iron. This will seldom apply, for it will seldom have to fire at 
 anything approaching 30 inches of iron ; the nearest approach perhaps 
 may be found in very soft steel. The projectile has, however, a total 
 energy at the muzzle of about 36,350 foot-tons. This will represent its 
 total smashing power against hard armour. This is about 577 foot-tons 
 per ton of gun, which represents the value of the gun as an investment 
 in artillery power. If this be compared with the energy per ton of the 
 38-ton gun, which was 360 foot-tons, the extent of the Artillery im- 
 provements in the construction of guns and burning of powder will be 
 appreciated. 
 
 1 Probably few are aware of how completely Fairbairn'a formula was worked out by him, with 
 punches, on a small scale. It seems only reasonable to apply tests on a small scale to the muoh 
 more difficult question of smashing. 
 
 2 For example, old type guns, such as the 25-ton guns, carried on board many of our vessels 
 might be found in the same attacking force as new type guns firing shot of smaller calibre and leas 
 energy but greater pcnetrat.ng power. The former guns would give the better results against 
 chilled iron shields, and the latter against wrought-iron. 
 
 1 , 1 * "Heavy Guns of 1884," by Colonel Maitland, "U.S. Proceedings," 1884. The high 
 sectional density of the projectile makes the Eule of Thumb work very badly for this eun eivinir 
 under 28 inches perforation : 2'06 x 13'5 = 27-7 inches. 
 
39 
 
 CHAPTBE IV. 
 
 Fuetheb Experiments with Soft Aemoue. 
 
 In 1868/ a Prussian trial took place between the English 9-inch and & u , 38isn 
 Krupp's 24 cm (9-27-inch) guns, against 6, 7 and 8-inch iron plates 
 with backing and skin, when the English gun maintained a decided 
 advantage. In July and August of the same year, the Krupp gun 
 and projectiles had been so far improved that they beat the English 
 ones ; the projectiles especially were thought to be good. 
 
 In August, 1869, 2 an 1 1-inch Krupp gun was fired in Russia against Russian 
 a target representing, very nearly, H.M.S. Hercules armour at the water- Hercal "- 
 line amidships. It consisted of 9 and 6-inch plates on a structure 
 consisting of 39 inches of backing, consisting of 36 inches of wood 
 and angle-iron, and 2 intermediate iron plates, each 1 inch thick, and 
 an iron skin 1 inch thick. The projectiles, which were of steel, 
 weighing 496 lbs., perforated the thickest part of this structure easily 
 at 467 yards, passing on up the range. Capt. (now Colonel) W. H. 
 Noble estimated their striking velocity at 1247 feet, and energy per 
 inch circumference at 155 foot-tons. There is therefore nothing 
 remarkable in the perforation of the target. The projectiles, however, 
 must have been excellent, and foreshadowed what Krupp' s steel pro- 
 jectiles would be. 
 
 On March 3, 1870, a structure consisting of three 5-inch plates with «no. 31" 
 iron concrete between, termed " No. 31 " target, was tested by 12-inch Tar s et - 
 guns, and others of smaller calibre. 3 
 
 On May 12, 1870, 4 an experiment was made to test the liability of km shells 
 filled shell to explode when struck by projectiles after passing through f^ot?' 1 hj 
 
 1 Vide Very's "Development of Armour, p. 441.'' For all American experiments and many- 
 foreign ones not here given, the reader cannot do hetter than consult Very's able work. 
 
 2 Vide D. of A. "Proceedings," Vol. VIII., p. 63; also Very's "Development of armour," 
 p. 437. Very gives the range as 1200 yards. 
 
 3 Vide D. of A. " Proceedings," Vol. VIII., pp. 47 and 206. 
 ' Vide D. of A. " Proceedings," Vol. VIII., p. 121. 
 
 7 
 
40 
 
 armour. The Warrior target was employed, and shells placed behind 
 it in various ways were struck by 9-inch projectiles after perforating 
 the armour. It was concluded, after this trial, that fuzed shells are no 
 more liable to explosion than plugged filled shells, also that explosion 
 may generally be limited to the shell actually struck, if live and 
 empty shells are placed alternately in the interior of the ship. It may 
 here be noticed that the impact of the penetrating projectile, whether 
 shot or shell, generally explodes any filled shell which it actually 
 strikes. 
 
 Ships Decks Targets. 
 
 peok t.nied In 1870, 1 very important experiments were commenced by the 
 
 under 3 Mor. Admiralty, to test the resisting power of ships decks. A target was 
 
 tar Are. ere cted at Shoeburyness (known as " No. 32 ") representing a ship's 
 
 deck. In order to prevent the waste of time and ammunition, arising 
 
 from the inaccuracy of vertical fire, the target, instead of being laid 
 
 horizontally, was placed in a vertical plane, and it was attacked by 
 
 horizontal fire with projectiles with reduced charges, giving striking 
 
 velocities such as they might have in falling on a deck. The target 
 
 was as follows : — Iron deck beams, 10 inches deep, of T section, and 
 
 running across the greatest breadth of the deck at 2 feet interval, 
 
 were covered with 6 inches of deck material all over. In one portion 
 
 (10 feetx 10 feet) this 6 inches was made up of 1£ inches of iron (in 
 
 two f-inch plates) and 4>\ inches of fir, and in another portion (10 feet 
 
 x 10 feet) it was made up of 1 inch of iron (in two i-inch plates) and 
 
 6 inches of fir. 
 
 A 13-inch L. S. mortar was fired at this, on April 7, 1870, from 20 
 yards range, striking at angle of 60° with the target, with a shell 
 filled with sand; 7 lbs. of L. G. powder, giving 514 feet striking 
 velocity, and 379 foot-tons energy (9 - 4 foot-tons per inch circumference), 
 two shells thus fired passed through both portions of the deck. With a 
 3£ lbs. charge and 327 feet striking velocity, and 154 foot-tons striking 
 energy, and 3 - 8 foot-tons per inch circumference, the shell passed 
 through the inch iron portion, but rebounded on the inch-and-a-half 
 plates, cracking them sufficiently to show daylight through. The 7 lbs. 
 charge, was calculated to give the striking velocity of a shell fired with 
 20 lbs. from a S. S. mortar at 45° elevation. The 3| lb. charge was 
 intended to test the actual resistance of the plate. The iron deck 
 beams were broken in the inch portion, but the first shell passed 
 between them in the l^-inch portion. 
 DeoktMget. On June 14 and 15, 1870, and May 3 and 5, 1871/ the 9-inch M.L. 
 5»™ ga " gun was fired at the deck target at 100 yards range. The target had 
 been repaired and laid on the sand ; the angle of incidence of the pro- 
 jectiles was given by raising or lowering one end of the target. The 
 sides of the vessel were represented by i-inch plates, 2 feet 6 inches 
 
 ' D. of A. " Proceedings," Vol. VIII, p. 123, and Vol. IX. p. 93. 
 a D. of A. "Proceedings," Vol. VIII., p. 209, and Vol. IX., p. 93. 
 
41 
 
 deep, worked on to projecting edge of deck plates, with angle-iron 
 3 inches x 3 inches x £ inch, and connected with each deck beam by a 
 triangular web piece. The iron work was fastened with 1-inch rivets. 
 The wood was secured with four |-inch galvanized iron bolts to each 
 plank. The target was supported on wood piles, with the deck beams 
 parallel to the line of fire. The 9-inch M. L. gun was fired with a 
 43 lbs. charge, and Palliser shell 250 lbs. weight, containing 5 lbs. of 
 powder. At 15°, and at 10°, the shell, which had a striking velocity 
 of 1328 feet, a total energy of 3060 foot-tons, or 109 per inch circum- 
 ference, passed through, tearing the decks much; at 8° the shells 
 deflected without exploding, still, however, tearing the deck consider- 
 ably. 
 
 On June 19, 20, and 21, and on July 16, 1872, 1 another deck target "N0.32A" 
 (32a) was fired at, representing upper and breastwork decks of H.M.S. " ge ' 
 Thunderer. This target measured 24 feet x 14 feet, the left half 
 representing the upper deck, the right half the breastwork deck. The 
 upper deck consisted of 4-inch oak fastened with |-inch screw bolts 
 to three thicknesses of 1-inch iron plating. The latter were secured 
 with 1^-inch rivets to iron beams 4 feet apart, 11 inches deep x £-inch 
 thickness, and two angle-irons 3 inches x 3 inches x £-inch. The 
 breastwork deck consisted of %\ inches of oak planking, fastened 
 with f-inch screw bolts to two thicknesses of 1-inch iron plating. 
 The latter were secured with 1^-inch rivets to X iron beams 4 feet 
 apart, 8 inches deep, X |-inch in thickness. 
 
 The 9-inch M. L. gun was fired with 50 lbs. charge, and Palliser 
 shell containing 4 lb. 4 oz. bursting charge, thus weighing about 
 246£ lbs. The striking energy was 3396 foot-tons total, or 121 per 
 inch circumference. At 8° incidence the shell deflected on off upper 
 deck and breastwork deck without exploding, cracking and bulging 
 deck, and tearing up wood. A flat-headed shell behaved in a nearly 
 similar way. 10-inch Palliser shells, each weighing, with bursting 
 charge of 5 lb. 12 oz., in all about 396f lbs., and having a striking 
 energy of 5055 foot-tons total, or 162 per inch circumference, deflected 
 at 8° without bursting, twice, and once burst. At 10° one shell burst; 
 of one it could not be certainly said if it burst or not. 
 
 On November 17 and 27, December 21, 1871, 3 and January 9, 1872, £££* iter 
 firing was conducted against targets, placed below water at angles of 
 depression of from 6£° to 12°, with ogival and flat-headed shot. The 
 ogival Palliser projectiles ricochetted on the water at angles less than 
 11°, but at greater angles remained under water. Plat-headed shot 
 remained under, at about 8£° depression. 
 
 On June 20, 1872, the 35-ton gun of 12 inches calibre was fired <j6-ton gun 
 against a target, known as " No. 33," at Shoeburyness, 8 which con- 33 " target. ' 
 
 1 D. of A. "Proceedings," Vol. IX., p. 249, and Vol. X., p. 170. 
 
 ■ D. of A. " Proceedings," Vol. X., p. 25. 
 
 a Eeported in " Engineer," by Author, June 28, 1872. D. of A. " Proceedings," Vol. X., p. 178. 
 No. 33 was perfoiated by 11-inch projectile, September 9, 1872. D. of A. " Proceedings," Vol. X., 
 p. 293. Sandwich targets, known as " 34" and "36," with 8 ins. iron, 9 ins. wood, 5 ins. iron, 
 and 6 ins. wood, i.e., 13 ins. iron besides skin, were perforated by 36-ton gun with e<=182. 
 
42 
 
 Finspong 
 projeotiles. 
 
 sisted of 8 inches of iron, then 6 inches backing, 5 inches iron, 6 inches 
 wood, and 1£ inches of iron skin. This structure not being a match 
 for the gun, was strengthened by an additional plate of 4 inches of 
 iron, which however was from some cause laid on with 5-inch timbers 
 round under its edges only, leaving a 5-inch space between itself and 
 the front of " No. 33 " proper. There was reason to think the target 
 itself was not up to the proper standard for such a structure, owing to 
 the metal being inferior, and the experiment would not be worth 
 notice had it not been for the air space and its possible influence on 
 the results. A Palliser shell with bursting charge of 9 lbs., weighing 
 694 lbs., was fired with a striking velocity of ]296 feet, a total energy 
 of 8090 foot-tons, and 216 foot-tons per inch circumference. It burst, 
 the head penetrating to 35 - 7 inches depth, leaving a hole in front plate 
 of from 12*1 to 12 - 5 inches diameter. A girder and some rivets were 
 broken and started in rear. The top edge of front 4-inch plate was 
 thrown back against the target proper, the supporting timbers having 
 been dislodged so that the second shot would find greatly diminished 
 air space. This was fired without bursting charge with about the same 
 striking energy as the preceding one ; it drove its head completely 
 through the target, leaving the body lodged in the 5-inch plate. The 
 base was left behind, not entering the plate at all. The perforation is 
 not much worth calculating, owing to the uncertainty as to the 
 character of the target, but it is perhaps rather more than could have 
 been expected. The most interesting feature is the fact of the pro- 
 jectile perforating the plate when fired as a shot, and failing to do so 
 as a shell. This must be attributed mainly to .the circumstance that 
 the bursting interferes with penetration in thick armour, because the 
 powder acts and blows away part of the shell before it has done its 
 work fully. Also it is possible that the larger air space may have 
 contributed to the greater shattering of the first shell. 
 
 On August 21, 1871, 1 at Shoeburyness, two Finspong 9-inch chilled 
 iron projectiles, fired at Shoeburyness, rebounded from a Millwall iron 
 shield without breaking, further than a few cracks being made at the 
 stud holes. 
 
 OtaHon 
 turret ex. 
 periment. 
 
 The Olatton Turret Experiments. 
 
 On July 5, 1872, at Portland, 3 took place the notable experiment on 
 the turret of H.M.S. Glatton. The object of this was to ascertain 
 whether the turret could be disabled either by the revolving gear being 
 put out of order, or the structure itself being so far distorted as to 
 interfere with the turning of the turret, or lastly, if the turret might 
 be riveted or wedged to the glacis by the entrance of a shot at the 
 junction of turret wall and glacis plate. The Hotspur was drawn up 
 at 200 yards distance, and fired two effective rounds from a 12-inch 
 25-ton gun, at the Glatton turret. The fact that the trial was one on 
 
 i D. of A. 
 
 Proceedings," Vol. IX., p. 174. 
 1 D. of A. " Proceedings," Vol. X., p. 175, Authors Report, 
 ! " "■ "Naval Science," January, 1878, p. 32. 
 
 also in 
 
 ' Engineer," July 12, 1872, and 
 
43 
 
 contortion or racking, and not one on penetration, must be clearly kept 
 in view, and the construction of the turret should be understood. It 
 may be seen in Fig. 1, herewith, that a single layer of armour plates, 
 
 Fig. 1. 
 
 each extending for one eighth of the circumference, surrounds a layer 
 of wood, with a skin and second or smaller skin inside. It will also be 
 seen in Fig. 2, that two rings of plates are required to extend to the 
 full height of the turret. The plates break joint as shown in Fig. 2. 
 
 HHOKEN RAILINGS 
 
 Kg. 2. 
 
 
 
 i r f ' ii 
 
 m 
 
 1 pi 
 
 liMi^ 
 
 m 
 
 iiiiiiiiiiii 
 
 :.MMlLKli 
 
 Front elevation of Turret from firing point, &c. 
 
 They are all 12 inches thick except the two plates in the upper tier, 
 containing the upper portion of the two gun ports, and in the lower tier 
 the plate between the lower portions of the two ports, which three 
 plates are each 14 inches thick. The wood backing is 17 inches 
 behind the 12 -inch plates, and 15 inches behind the 14-inch plates, 
 so as to form an uniform thickness of plate and backing of 29 inches. 
 Inside the backing is a l^-inch skin, consisting of two f-inch plates ; 
 then come vertical girders 5 inches deep, and finally on the inside 
 edges of these a mantlet skin of j-inch thick, to keep back nuts and 
 bolt heads, &c, from flying about in the turret. It is obvious that the 
 liability of the turret to contortion was most severely tested by firing 
 at the thickest plates, for against the thinner portion there was only 
 the resistance encountered by the projectile in passing through 12 
 inches, to act in bending ; for directly the shot is through the turret, 
 
44 
 
 all contorting action is past, whereas, against the thick plates, the 
 shot had to penetrate 14 inches of iron, which thickness, in fact, was 
 more than a match for it. 
 
 Calculated on the system followed for thin armour in the Director of 
 Artillery's Department, the shot might perforate 14"42 inches of iron, 
 and might therefore nearly get through the armour, but certainly not 
 entirely. The first round was intended to strike the top of the turret, 
 but actually struck a point nearly 3 feet lower down (vide Figs. 2, 3 and 
 4), producing the following effects. 1 The point of the projectile which 
 
 Kg. 3. 
 
 Vertical seotion of turret through portion struck. 
 Fig. 4. 
 
 ifctf.sw__iw.Hffi>! 
 
 1 The striking velocity of the projectile wai not taken. Its initial velocity was probably about 
 the tabular one for this gun, viz., 1300 feet, and its striking velocity about 1276. If this be 
 worked out on formula used for thin armour, Chap. II., p. 16, that is, — 
 
 1-6 IWv* 1 
 
 A / X 
 
 i _ 
 
 2-63 ' 
 
 log 2,7 - 5-1590 
 
 log » = 4971 
 
 log D (11-92) - 1-07 63 
 
 log jrD 
 
 log k (.-53) 
 
 1-5734 
 0-4081 
 
 ( 3-1059 
 " (.3-1059 
 log W (600) = 2-7782 
 
 2 logo (1276) : 
 
 WV 
 
 JFoS 
 
 log 1g = 
 
 = 6776 
 
 9 , 
 
 log irD = 
 
 _L 180-9 
 irD 
 
 logi = 
 
 8-9900 
 6-1590 
 
 3-8310 
 
 1-5734 
 
 2-2576 
 
 0-4031 
 
 1-6 
 t = 14-42 
 
 t-4) 
 
 4)1^545 
 " 46363 
 1-1691 
 
 This is an old fashioned 
 gun, and very bad for 
 the rule of thumb, p. 
 30., Chap. III. 
 
 1276 x 12 .... 
 
 -Tooo-° 15Sm8 - 
 
 On Farbairn'a formula 
 with table of *, p. 35, 
 Chap. III., the per- 
 foration is 13-62. 
 
45 
 
 was fired without any bursting charge (as a shot) penetrated to about 
 20 inches depth, driving a bolt inwards and dislodging its head. The 
 skin proper Was bent and opened to about 3 inches width, a large sheet 
 of mantlet skin was torn off, and a number of rivet heads detached. 
 Above all, the entire upper plate about the place of impact was driven 
 back bodily about 5| inches, and lifted so as to open the horizontal 
 junction of the plates about 2 inches. A kid, a rabbit, and a hen, 
 which had been put in the turret to try if they would be affected by 
 the shock, did not seem to have suffered in any way. The turret 
 revolved freely. The second round, which consisted of a projectile 
 without bursting charge, as in the previous round, was now directed 
 against junction of turret and glacis plate, and was remarkably 
 successful, it glanced along the glacis plate and entered the base 
 of the turret (vide Fig. 5), the point reaching a depth of about 15 
 
 Fig. 6. 
 
 i i 
 
 inches. The shot was easily disengaged and the turret again was 
 found to revolve and work perfectly. The vessel was nearly as fit 
 for action as before she was struck. 
 
 It is a question how far the liability to jam might have been more 
 severely tested. The vessel might have had a list given towards the 
 Hotspur, as originally intended, to represent the circumstances of 
 plunging fire. Unless the shot, however, had remained with a large 
 part projecting, and had also struck the turret well below the glacis 
 plate, it is hardly likely that the turret would have been jammed. Broken 
 pieces of shot and glacis plate, which latter is only 3| inches thick, 
 might have got into the 6-inch space round the turret, but the crew 
 would have found access to them from below, without exposing them- 
 selves. It was afterwards suggested that the revolving gear would be 
 more severely tried by a blow striking more nearly in a tangential 
 direction, so as to tend to produce sudden rotation, or to check the 
 turret suddenly when in motion. This suggestion however may be 
 
40 
 
 probably disregarded safely. The Glatton turret is nearly 81 feet in 
 diameter; a projectile with an ogival head of 1£ diameters radius, 
 would probably glance at about 48 degrees with the tangent, or 42 
 degrees with the normal line to the turret face. This would give a 
 maximum lever of about 10 feet for rotation of the turret. It is very 
 unlikely that such a blow could be made to act on the mass of the turret 
 so as to put it in motion, and certainly the gear which is at the base of 
 the turret could only be affected by the intermediate metal between it 
 and the point of impact being moved in some way. 
 
 On the whole, the direct blow at the turret top seems to be the most 
 severe test, and no 12 -inch projectile could produce a much greater 
 effect than that first fired on this occasion, for the only additioual power 
 to contort the turret, available, is that required to complete the 
 passage of the shot entirely, and this is but little. The introduction 
 of new type guns with small bores has rather favoured the turret in 
 this respect, for a shot of smaller diameter, while it would perforate 
 much more readily, would have less power to contort the structure : 
 and for a new type piece of greater diameter than 12 inches, it is 
 necessary to go above the 43-ton gun, so that the Glatton turret may 
 be safely regarded as not liable to be jammed or wedged by any 
 reasonable exposure to fire. It may here be observed, however, that 
 the condition of a turret made of harder armour is not the same. A 
 steel, or even steel-faced plate distributes the shock much more than 
 one of wrought-iron ; consequently, where the latter would be per- 
 forated, the former may be bodily driven back, and a turret covered 
 with such armour might suffer more in its structure from racking 
 than the former. 1 The total stored-up work, or striking energy is of 
 course the factor to be considered. In the above experiment it was 
 only about 6776 foot-tons, a force, however, which it may be noticed, 
 would be sufficient to lift the turret to a great height in the air, if 
 applied to such a purpose in a suitable form, 
 whitworth On October 8, 1872, Sir Joseph Whitworth fired a 9-pr. B. L. gun 
 at Southport, Lancashire, against a 3-inch Cammell wrought-iron plate 
 at an angle of 45°. The hexagonal bore of this gun was 2'72 inches 
 measuring from angle to angle, and 2 '47 inches from side to side of the 
 hexagon. The length was 8 feet. The pitch of rifling was about 1 
 turn in 18 calibres, or 50 inches. The projectile was a steel flat- 
 headed shell slightly tapered in front, made from a solid steel ingot 
 bored out. It was 5 diameters (13^ inches) long, and weighed 15 lbs. 
 14 ozs. One projectile fired at 200 yards range, nearly perforated the 
 plate, and a second one passed completely through it, penetrating some 
 depth into the earth in rear. 
 
 Sir J. Whitworth exhibits shells and plates perforated by them at 
 as oblique an angle as 30° with the face of the plate, in support of his 
 contention that flat-headed projectiles bite at a more oblique angle 
 than pointed ones. On this it is difficult to speak authoritatively. 
 
 1 At Shoeburyness entire targets of steel-faced armour have been driven back bodily many inches, 
 and the structure supporting them has suffered proportionately. 
 
 9-pr. fired 
 obliquely, 
 
47 
 
 The question has been disputed/ both on theoretical and practical 
 grounds. Occasionally pointed projectiles have bitten plates at very 
 oblique angles, but no instance of biting at 30° can be given. With 
 regard to the practical advantage of a possible superiority in the 
 flat-head in this respect, it must be understood that Sir Joseph 
 himself prefers ogival-headed shot for perforation near the direct line, 
 any possible advantage possessed by flat-headed shot would therefore 
 probably only hold for a certain number of degrees, such as could 
 hardly compensate for the admitted advantage possessed by ogival shot 
 in more direct impact. Hence ogival-headed projectiles may be pre- 
 ferred to flat-headed ones for service, even by those who admit that 
 the latter act better through water, and bite at a more oblique angle of 
 incidence against armour. 
 
 On August 15, 1872, at the Tegel in Prussia, 3 Krupp's 26™ (10i-in.) '% e *£ iaaA 
 and 28 cm (11-in.) guns were fired against 10 and 12-inch iron, backed iotI™' 
 by 18 inches of oak, and f-inch skin. The 10-inch portion was per- 
 forated with the lOJ-inch gun. The 11 -inch would penetrate the 
 12-inch portion at about 437 yards range. These penetrations are not 
 extraordinary, but the German guns were now growing longer, and 
 their power was increasing. 
 
 On February 16, 3 and March 30, 1875, some Palliser shells, filled wet gnu- 
 with wet guncotton, were fired against armour, when it was found that 
 shells so charged, with a primer of dry guncotton embedded in the wet 
 guncotton, might be fired safely from ordnance, but that on impact the 
 explosion of the primer was not communicated to the wet guncotton. 
 " The shells appeared to break up on impact outside the plate, probably 
 from the instantaneous explosion of the dry primer, the charge of wet 
 guncotton being thrown about in front of the target." 
 
 On or about July 24, 1875,* the comparative resistance of malleable Guncotton 
 iron and steel to the detonation of guncotton was tested. Two 9-oz. dlscs ' 
 discs were fired when suspended 3 inches over the centre of a Low 
 moor f-inch plate — buckling the plate 0*65 inches. A Landore steel 
 plate was more abruptly indented to 1*3 inches. 
 
 Shannon Target. 
 
 On July 7, 1875, 6 a structure (known as " No. 39") made to represent shannon 
 the unarmoured portion of the Shannon class of vessel, was fired at abffi 6 *'" "' 
 Shoeburyness. The target represented a complete section of a ship 
 with guns and dummy detachments on both sides. The iron side on 
 one portion was 1 inch thick, made up of two half -inch plates, and the 
 other portion f-inch, consisting of two f-inch plates. The total width 
 
 1 See Mallet's Papers in " Engineer," January 4, 11, 18 and 25, of 1867 ; also " Experiments on 
 Competitive Shot," vide Chap VI. Sir W. Palliser wished to compete with ogival-pointed shot 
 against Whitworth's flat-headed ones. 
 
 2 " Engineer," April 4, 1873, p. 205. Very's " Development of Armour," p. 444. 
 * D. of A. " Proceedings," Vol. XIV. p. 7. 
 
 4 D. of A. " Proceedings," Vol. XIII., pp. 206, 207. 
 
 5 D. of A. " Proceedings," Vol. XII., p. 162, and Vol. XIII., p. 141. 
 
 8 
 
48 
 
 of the ship between the sides was 56 feet. After firing with machine 
 guns and ordnance, it was concluded that both, sections of the target 
 were proof against the 065-mch Gatling, and case shot from the 64-pr. 
 gun. The inch portion was also proof against case shot from 9-inch 
 gun fired with 80 lb. charge, but the J-inch portion was sometimes 
 nearly penetrated by these at 200 yards. The greatest effect was 
 obtained with the 9-inch shrapnel. One 9-inch shrapnel killed six 
 dummy men, and wounded one at the gun close to where it entered, it 
 also dismounted the opposite gun on the far side, and blew all the 
 dummies posted at that gun to pieces. 
 
 Both parts of the sides of the section 1 fired at were now made 1-inch 
 thick, one portion consisting of two ^-inch iron plates, and the other 
 of two ^-inch Bessemer steel plates. After practice in February, 1876, 
 direct fire, and at angles of incidence of 45°, 80°, 25° and 20°, the 
 following conclusions were arrived at : — The steel resisted better than 
 the iron in two instances, deflecting the 64-pr. common shell, but in one 
 case it was more torn, and had its rivets sheared more extensively than 
 the iron plates. Common shell plugged, or fuzed with Pettman's 
 6. S. fuze, were apt to break across in few pieces without the explosion 
 of the bursting charge ; 1 3 shells, fuzed with Laboratory percussion 
 fuzes burst well into small pieces. The mantlets caught fire, but had 
 not been rendered non-inflammable. 3 It may be observed the shrapnel 
 break up well under any circumstances, and may therefore be fired 
 plugged. 8 
 
 No. 40 Plate-upon- Plate Target and 38-ton gun. 
 
 &£*£*, 0n October 11, 1876/ the 38-ton gun of 12-5-inches calibre was fired 
 at Shoeburyness, at a target identified as " No. 40," which was a typical 
 sandwich 6 section, approved by the Royal Engineers, {vide p. 49). 
 
 "No. 40 
 target 
 
 i D. of A. "Proceedings," Vol. XIV. 
 
 8 By soaking in ohloride of caloium. 
 
 » D. of A. " Proceedings," Vol. XIV., p. 86, gives the results of « few more rounds fired on 
 May 18, 1876. 
 
 * Vide D. of A. " Proceedings," Vol. XIV., p. 214; " Engineer," October 20, 1876 ; also author's 
 paper, Mechanical Engineers Institution, January, 1879 ; and Very's " Development of Armour," 
 p. 446. 
 
 Here log 2g = 5-1690 Log W (log 816) = 2-9112 
 
 logl^S Lo^(21ogl421)= {||| 
 
 log irD = 1-6912 9-2164 
 
 For ti = 1400 log K = 1-9845. log 2g = 6-1590 
 
 •9)17-42 iHio 
 
 19-36 
 
 291-8 
 
 ' 17-42 
 
 1421 x 12-6 .„„. . 
 
 iooo " 17 - 7fl ""- 
 
 Dividing by -9 for sandwich in 
 3 plates (fee Chap. II., p. 18, 
 and Chap. III., p. 36), 
 
 • 9)17-76 
 19-7 in. 
 
 'Lieut. Very employ, the word " plate-upon-plate " for iron plates in contact with each other, 
 and •andwioh for alternate iron and wood, (««, Very, p. 446). This might be good if 
 universally followed (tee note 2, p. 8). 
 
49 
 
 The projectile weighed 815 lbs., the striking velocity was 1421 feet. 
 The shell, fired without bursting charge, perforated the armour as 
 shown in Fig. herewith, breaking up, and leaving its base lodged in 
 
 MOTTLED IKON 
 
 SECTION THROUCH BOO lb. SHELL AS 
 FHACTURED 
 
 1 ' i I I 
 
 \z%kt&~-*ft---U?>li$t---e'/r--* 
 
 IRON ZBAJt 
 
 SECTION Or Nf 4-0 TABCET 
 
 the plate. It will be found that calculating the perforation by 
 Fairbairn's old formula, with the table of values for K, p. 35, and with 
 the second power of t, a perforation of solid plate of 17'42 inch is 
 obtained; and taking InghV rule of three plates being only ■& the 
 strength of one, the shot should perforate 19*36 inches. This result 
 very closely agrees with that given by the rule of thumb, namely, 
 19"7 inches, and is nearly correct, for the shell perforated 19*5 inches 
 of iron and 5 inches of wood. This case illustrates what was said in 
 Chap. II., p. 17, as to the plate-upon-plate system coming in at a time 
 when it disguised the errors caused by the use of the old D. of A. 
 form of Fairbairn's formula ; for if this example be worked out by 
 this formula, ignoring the difference between solid and plate-upon- 
 plate armour, the error arising from using the old formula and that 
 of ignoring the division of the iron neutralize one another, and give 
 as the perforation 19'5 inches, which happens to be nearly correct. 
 
50 
 
 * ir enm!nt ^ n *^ e same "^y 1 a P&Uiser projectile was fired from the 38-ton gun 
 mea " aB above, striking with, a velocity of 1436 feet, a LO-inch wrought-iron 
 plate placed 6 feet 8 inches in front of No. 33 target. The projectile, 
 of course, passed through the front plate, striking the target in rear, not 
 quite directly, and shivering to atoms against it, leaving a characteristic 
 splash or lump of pulverized metal adhering to the face of the target. 
 On March 27, 1877, 3 at Shoeburyness, the 38-ton gun fired a Palliser 
 shell weighted to 800 lbs., against a 4-inch plate, with a 10-inch plate 
 fixed with an air space of 4 feet 7\ inches between them. The shell 
 passed through the 4-inch plate and broke up, forming a splash of 
 metal against the 10-inch plate, which was very slightly bulged in 
 reai" — the splashed metal was very hot. 
 
 No. 41 Target and 80 -ton Gun. 
 
 so-'onpin On February 1, 1877, 3 the 80-ton gun was first fired against armour. 
 
 and "No. « The gun was then unchambered ; its bore was 16 inches in diameter. 
 
 target. ^ Palliser projectile, without bursting charge, weighted to 1700 lbs., 
 was fired with a charge of 370 lbs. of P 2 powder against a structure 
 known as " No. 41 " target, consisting of four thicknesses of iron each 
 8 inches, with 6 inches of teak between the layers. This is shown in Pigs. 
 1, 2, 3 and 4, pp. 51 and 52, from which the dimensions of the plates 
 can betaken. The range was 120 yards, the striking velocity was 1496 
 feet per second. The projectile penetrated the target to a depth of about 
 48*8 inches, as shown in Pigs. 3 and 4, pp. 52 and 51. The entire 
 target bent at point of impact. It will be seen that its structure was by 
 no means stiff, and the layers of metal were opened and contorted by 
 the blow. The shot itself was fractured into four longitudinally, and 
 also broken transversely. The iron of the plates was good, and the bolts, 
 which were the Palliser English pattern, holding the plates together 
 in pairs, behaved well. The target was driven back 2\ inches on the 
 right, and the wood framing in rear broken and splintered. The point 
 of the shot entered 2"6 inches into the back plate. This round is equal 
 
 » Vide D. of A. "Proceedings," Vol. XIV., p. 214, and "Engineer," October 20, 1876. 
 2 D. of A. " Proceedings," Vol. XV. p. 107. 
 
 »D. of A. " Proceedings," Vol. XV., p. 38 j " Engineer," February 9, 1877, Author's report. 
 Here W = 1700. « = 1496. B = 15-92. 
 log 2ff = 5-1590 
 
 log it = 0-4971 
 log 16-92 = 1-2020 
 
 log jtD - 
 
 For v = 1600 
 log k =T-9777 
 
 For 3 Plates— 
 
 ■ 9) 23-57 
 26-2 
 
 log 
 
 1700 
 
 = 3-2305 
 
 2 log 
 
 1496 
 
 (3-1749 
 ~ 13-1749 
 9-6803 
 6-1590 
 
 -E = 26382 
 
 
 4-4213 
 
 
 1-6991 
 
 e = 528 
 
 
 2-7222 
 
 1-9777 
 
 2) 2-7445 
 
 * = 23-67 
 
 
 1-3723 
 
51 
 
 to the perforation of 23*57 inches of iron in a single plate. No factor 
 has been worked out for four plates in sandwich, but for three the per- 
 foration is 26'3 inches, and the shot ought to perforate' four plates of 
 somewhat greater total thickness than this. The four 8-inch plates, 
 however, make up 32 inches, and the work of perforating 26 inches, 
 when constituting the entire target is much less than that of perforating 
 the front 26 inches of 32 inches of iron, because the last 6 inches are 
 stiffening and supporting the front 26 inches of metal. Hence, in 
 default of a formula for partial penetration, there is no cause of 
 surprise at the fact that the point of the shot 1 only penetrated through 
 26" 6 inches. 
 
 F 1 C . I 
 -16,0'- 
 
 I^Ml T't '\ pMi):J| 'la i ! * ^S&'!f j|f : ! -m i ' . i v : ■.. " !■ i L . ; : ■' jjjii , ) ''S^ tp ilL 
 
 ELEVATION 
 
 STARTED Is 
 
 "1E.E. BEAM CRUSHED AND 
 ™ SPLIT TO PIECES BY 
 7-J'liESSURE OF BACK PLATS 
 
 On the same day, a blind common shell was fired from the 80-ton "j^™™^. 
 gun at an unbacked 8-inch plate. The shell passed through the plate, »t armour.' 
 
 1 The point here refers, of course, to the actual tip of the shot. Perforation does not mean that 
 the point shows through, but refers to the actual passage of the entire shot. Exact perforation 
 would be when the shot just got through and was unbroken. Though this cannot exactly occur, 
 something nearly equivalent to it may do so. — Vide p. 49. 
 
52 
 
 apparently breaking up as it did so, the hole being about 20 inches in 
 diameter, and the plate broken into 5 pieces. The shell fragments, 
 which were small, ranged up to 1200 yards. 
 
 VIEW OF CRAWED AND BULCED BACK PLATE 
 OF TARCET OPPOSITE POINT OF IMPACT. 
 BEAM ENDS OF A'. A'. ARE SHEWN AS IF CUT 
 AWAY TO ENABLE CRACKED TARCET PLATE 
 
 TO BE SEEN. 
 
 BOLTS B.B.B.AFlE STARTED ABOUT Jy IN? 
 
 OWINC TO BEAMS IN FRONT OF D BEING 
 
 CRUSHED UP AND SPLIT. 
 
 common Qn March 16, 1877/ two 10-inch common shell were fired against 
 
 5-inch iron plates at Shoeburyness, one shell filled and plugged, and 
 the other filled and fuzed, in both cases the shell passed through 
 the plates exploding its bursting charge. In the first case, the shell 
 may have acted rather less quickly than the fuzed shell, as the front 
 of the plate was not blackened by the powder. Fig. 5, p. 53, shows the 
 plate struck by the fuzed shell. The torn piece is bent towards the 
 firing point, which seems to imply that the explosion acted chiefly after 
 the shell passed through. 
 
 Fig. 4, p. 54, shows a portion of the plate driven out by the plugged 
 shell. It is said 2 that discs of this kind are apt to be formed by 
 
 1 D. of A. " Proceedings," Vol. XV. p. 76. 
 3 By Mr. Forest, of the Royal Laboratory. 
 
 ' Engineer," March 28, 1877. 
 
53 
 
 common shell, and that the concentric ring gashes, and rends are due 
 to the fact that the apex of the shell yields and is first driven |in, 
 followed by successive rings of metal in the crown or head, until the 
 support of the walls is sufficient to enable the projectile to separate 
 the disc from the rest of the plate. The impression made by the square 
 key hole of the gun-metal shell plug was remarkably distinct. 1 
 
 nc.s 
 
 S INCH CASEMATE PLATE AFTER 
 BEING BROKEN BY 10 INCH COMMON 
 SHELL MASONRY LOOSENED ABOVE 
 CASE MATE. 
 
 Fig. 6, p. 54, shows the projectile fired from 80-ton gun as above 
 described when small fragments had been removed. It may be seen to 
 be fractured longitudinally in horizontal and vertical planes. A con- 
 siderable mass of metal was left in a diamond form, the angles being at 
 the longitudinal cracks. If this be compared with other fractures it 
 appears to be in conformity with some law, though it is difficult to say 
 what. 
 
 Judging by these three experiments with common shell it may be 
 concluded that a common shell will perforate iron plates of the thick- 
 ness of half their calibre. Backing would impede their further course 
 very greatly, seeing that they are broken, besides the fact that they 
 drive out a disc of plate in front of them. 
 
 1 This is an illustration of a softer metal pressing in a harder one. The Bame thing; occurs some- 
 times when gun-metal studs cut into a steel tube of a gun. Probably the gun-metal yields up to a 
 certain point, after which it offers more resistance than the metal opposed to it. 
 
54 
 
 o 
 
 ■ — 
 
 I- H 
 k. O 
 
 (J 5 
 £| 
 
 si 
 
 IC 
 
 U 
 
 5 
 
 o r- 
 
 2 O 
 
 o q; 
 
 h- 
 o 
 
 CO g 
 
 u_ "~ 
 
 O 5 
 
 o 
 
 
 
 No. 40 Target and 38-ton gun, chambered. 
 
 38-ton gun, 
 chambered; 
 
 strengthened ■ 
 
 On the same day, March 16, 1877, 1 the 38-ton gun, which had now 
 No.*otarge't an enlarged powder chamber, with a diameter of 14 inches, the bore 
 
 alrnnirthnnPtl t ^3 ^ 1 f% ~\ • TkT 1 f\ ("17 1 
 
 being 12 - 5 inches, was fired against No. 40 target at Shoeburyness. 
 The firing charge was no-w increased from 130 to 200 lbs. of pebble 
 
 1 " Engineer," March 23, 1877, also author's paper on " Construction of Armour," Institution of 
 Mechanical Engineers, 1879. D. of A. " Proceedings," Vol, XVI., p, 103. 
 
55 
 
 powder, P a , 1-5-inch cube. The gun was 70 yards from the target, 
 the striking velocity was 1525 feet, the muzzle velocity having been 
 1540 feet. A Palliser projectile without bursting charge was employed, 
 weighing about 812 lbs. It was fired against the strengthened part of 
 the target, consisting of four 6^-inch plates, with 5-inch layers of wood, 
 sandwich, between them (vide Figs. 1, 2 and 3, herewith, and p. 56). 
 The projectile penetrated deep enough to get its point probably about 
 2 inches into the last plate. The back of the target showed a slight 
 horizontal crack opposite the point of impact. The shot itself was split 
 
 -10'. 0"- 
 P LA N 
 
 longitudinally into four at the base, and was also broken transversely. 
 The whole target was bent, but the bolts and iron plates behaved well. 
 The total energy, energy per inch circumference, and thickness of plate 
 that represents the full perforating power are respectively 13090 and 
 335 - 5 foot-tons, and 18'79 inches solid plate, or 20'9 in three layers, 
 (vide note, p. 56). The point in this instance probably had penetrated 
 about 18'5 inches, which is about what might have been expected, 
 
 9 
 
511 
 
 considering that it is a case of partial penetration into iron of a total 
 thickness of 26 inches. 1 
 
 I W — 812. t>= 1525. 
 
 Log tt = 0-4971 
 
 Log 12-42 = ■ 1-0941 
 
 1-6912 
 
 For v = 1500 log K = 1-9777. 
 
 •9) 18-79 
 
 — W8 
 
 Log 812 = 
 
 1-9096 
 
 2 log 1525 = | 3;183f 
 
 E = 13090 ft. tons 
 
 e = 335-6 ft. tons 
 
 * = 18-79 ins. 
 
 1-1832 
 9-2760 
 6-1590 
 
 . 4-1170 
 1-5912 
 
 . ^-6268 
 1-9777 
 
 2) 2-6481 
 1-2741. 
 
57 
 
 Another shell was fired against the thinner portion of the plate on 
 July 20, 1877, to which was added a 3-inch plate in front, making 22J 
 inches of iron and 10 of wood. The projectile was 811 lbs. 4 oz. It 
 penetrated till the point was 28| inches, measured from the front of 
 the target proper. This would amount to about the same penetration 
 through iron as in the case of the last round. The shot was much 
 broken up, and there were marks on the second plate resembling the 
 " splashes " in air space targets. 1 
 
 No. 41 Target and 80-ton gun, chambered. 
 
 On May 4, 1877, 3 the 80-ton gun was fired a second time at No. 41 Cambered' 
 target. The gun had now been chambered to a diameter of 18 inches, and No. «' 
 the bore being 16 inches as before. The firing charge was 425 lbs. of UrgeU 
 Pa powder, 1'5-inch cube. A Palliser projectile, without bursting 
 charge, weighing 1700 lbs., was fired with a muzzle velocity of 1600 
 feet, and a striking velocity at the plate 120 yards distant from the 
 gun of 1585 feet. The striking energy, E, was therefore 29,610 foot- 
 tons. The thickness of solid iron that might be perforated was by 
 Colonel Maitland's diagram 24 - 5 inches, and by Colonel Inglis 24 - 7 
 inches. By the rule of thumb 25"4 inches. By Fairbairn's formula it 
 is 25"1 inches, and if disposed in three thicknesses 27'9 inches. The 
 point of the shot actually penetrated a distance of about 56 inches 
 from the face of the target, but owing to the giving back of the whole 
 structure {vide Pigs. 2 and 3, pp. 58 and 59), this amounted to a 
 penetration of about 3 inches into the last plate. That is about 27 inches 
 of iron in all. The hole in the front plate was slightly larger than the 
 diameter of the shot. The back plate was bulged nearly 14 inches and 
 cracked in five lines, radiating from the spot opposite to the point of 
 impact, and opened sufficiently wide to show the point of the shot. 
 The timbers were split and crushed, as shown in the Pigs. 2 and 3, 
 pp. 58 and 59. Pig. 1, p. 58, shows the front of the target, and 
 Fig. 4 the back opened by shot. 
 
 This concludes the series of rounds with the 80 and 38-ton guns 
 against their respective sandwich targets, No. 41 and 40. The following 
 round was, however, fired against a solid plate for the sake of com- 
 parison with No. 40 sandwich target, which ought to be connected 
 with them. 
 
 1 1). of A. " Proceedings 
 
 " Vol, 
 
 XVI., p. 104. 
 
 
 
 2 D. of A. " Proceedings, 
 
 Vol. XV., p. 99 ; " Engineer," May 11th, 1877. 
 
 
 
 
 
 
 Log 170H 
 
 = 3-2305 
 
 Log TtD = 
 
 1-6991 
 
 
 
 ! log 1585 
 
 ( 3-2000 
 
 ~ 13-2000 
 
 9-6305 
 
 For v = 1600 
 
 
 
 
 
 6-1590 
 
 
 
 
 E = 29610 ft. tons 
 
 
 .. 4-4715 
 
 log K = f-9731 
 
 
 
 t = 5920ft. tons 
 
 
 1-6991 
 3-7724 
 
 For 3 Plates— 
 
 
 
 ^^^^^™ 
 
 
 l-97bl 
 
 •9) 26-10 
 
 
 
 * = 26-10 ins. 
 
 
 2) 3-7993 
 1-3997 
 
 27-9 
 
 
 
 — ■ ■■ ■" 
 
 
 
58 
 
 
 VIEW OF BACK OF TARCET WITH 
 
 *<2\ l/^V 1 SH0T POmT ^ siBLE . "" OPENING 
 
 *& CRACK l'.ii"LONC CRACK ABOUT i'S'lOHC 
 
 Fl C.4- 
 
 
 e **°**.t> i0 
 
 F1C. I 
 
 FRONT ELEVATION OF Ni 4/ TARGET 
 
 IS" , 0- 
 
 ,OLD IRON PLATE LAID ON TOP OF TARCET 
 
 TT1 t \iiiiiiuiiiun \wmttiiti\ Vmiiiiiuuil \>i<miuiw\ 
 
 I J 
 
 THIS BOLT HEAD KNOCKED OUT & SEEN AS A DARK HOLE 
 ^ PROJECTILE FIRED FROM UNCHAIWBERED CUM 
 
 ! I 
 
 De 
 
 Ot FROM CHAMBERED GUN 
 
, r >9 
 
 ,S H33U13B aoeu 
 
60 
 
 Solid plate tested by 38-ton gun. 
 
 m-inch On August 1, 1877, 1 at Shoeburyness, the 38-ton 12-5-inch gun was 
 
 J^ d et plate fired at solid 16i-inch wrought-iron plate. This plate had been 
 ordered for comparison with the round tired from the gun when un- 
 chambered at the three plates forming No. 40 target, before it was 
 strengthened {see p. 49). In his paper on "Targets for the trial 
 of Heavy Ordnance," General Inglis says, " Having obtained m round 
 2039, so exact a measure of the penetrative power of the 38-ton gun 
 upon a three-plate target, it was thought that we could not do better 
 than use this result as our datum, and endeavour to obtain an equivalent 
 result with the same gun upon a single plate. Accordingly, after due 
 consideration of former trials, it was settled that the single plate for 
 comparison should be \Q\ inches thick, and an armour plate of this 
 thickness, measuring 8 f eet x 8 feet was ordered." "The finished 
 plate, as set up at Shoeburyness weighed 18| tons." For trial, the 
 plate rested against supports at its top and bottom edges to prevent 
 being driven back, and it was held from falling forwards by two bolts 
 through its upper corners." The plate was good and sound. 
 
 The projectile was a Palliser shell, 12-43 inches diameter over body, 
 weighted with sand to 817 lbs., including gas check. It was fired with 
 175 lbs. of P s powder, giving it a striking velocity of 1410 feet per 
 second, and a striking energy, E, of 11,263 foot-tons, and a power to 
 penetrate thickness of iron, t, of 17-29 inches. 2 The shell struck a 
 point 2 feet 10 inches from left edge, and 2 feet 11 inches from top, 
 and passed entirely through the plate, making a round hole 13 inches 
 in diameter, with a lip of 2 to 3 inches round the front edge. A crack 
 was formed from the left edge of the plate nearly up to the shot hole, 
 extending through the thickness of the plate at the edge. The plate 
 buckled from \ to T V of an inch, and moved forward bodily about h\ 
 inches. The projectile was broken up small, but its fragments all 
 passed entirely through the plate, including its gas check. The metal 
 was of average quality. 
 
 Comparing this result with the first round at No. 40 target, General 
 Inglis observes " It will be seen that the shell had rather more force to 
 spare with 11,263 foot-tons energy, in perforating the single 16^-inch 
 plate, than with 11,410 foot-tons in perforating the ll)£ inches in three 
 sandwich plates. He therefore considers that a 17 or 17^-inch solid 
 plate would have been the equivalent of the three 6^-inch plates 
 disposed sandwich fashion, as in No. 40 target. It will be seen 
 that according to General Inglis' factor for plate-upon-plate targets 
 (Chap. II., p. 18), that armour divided into three plates has -& the 
 resisting power of the same thickness solid. Thus, 19^ inches would 
 be equivalent to about 17-5 inches solid, as indicated above by actual 
 experiment. 
 
 1 L>. of A. " Proceedings," Vol. XVI., p. 103 ; No. of round, 2069. 
 a Worked out on p. 411, Vol. XIII. 
 
61 
 
 As to the entire series of experiments against No. 40 and 41 sand- 
 wich targets, the most striking features are the toughness and flexibility 
 of these structures, which would probably bear continued fire well, they 
 would therefore be adapted to resist the fire of any projectiles which 
 were unable to perforate them completely, and for forta where 
 additional plates could be added in front, as time goes on, to meet the 
 growing power of ordnance, this structure seems very good. The 
 Palliser English bolts stood admirably. Of course, for ships, where 
 the contortion would be serious, the case is different. 
 
 One other round was fired from the 38-ton gun, on August 1, 1877, 1 Airspace, 
 at a 4-inch plate, set up 18 inches in front of an old 10-inch plate. 
 The projectile, a Palliser shell, made a hole 13 inches in diameter in 
 the front plate, and the point entered 5£ inches into the 10-inch plate, 
 which was broken into several pieces, the base of the shell splashed 
 the face of the 10-inch plate. The effect in this case was apparently 
 considerably greater than in the round fired on March 27, previous 
 (see p. 50), the air space being reduced from 4 feet 7$ inches to 18 
 inches. 
 
 About this time a Krupp 12 -inch gun, 2 was fired at a sandwich f™|^„ h 
 target, consisting of 10 inches of iron, 8 inches of oak, 6 inches of 
 iron, 8 inches of backing, " strengthened by iron stringers and a skin 
 of two 1-inch plates." The range was 225 yards, charge 132 lbs., 
 projectile 670 lbs. Both plates were pierced, the point of the shot 
 entered the backing 7\ inches, bending the skin-plating back 3 inches. 
 
 It may appear by General Inglis' rule that the plates would represent 
 about 15^ 3 solid, but this is not the case, for the conditions are altered 
 by the thickness of the intermediate layers of wood (see Figs., Meppen 
 Trials, 1879 and 1882, Chaps. Y. and VI.) . 
 
 The projectile would get its point clear of the front plate and enter 
 the second one, attacking the plates in succession, without their 
 receiving the strength from each other that they would get with 
 5 inches of wood only. 
 
 The notable Spezia trials of steel and iron plates by the 100-ton 
 gun occurred at the end of 1876, previous to the firing of the 80-ton 
 gun, but it appears undesirable to break up the series of sandwich 
 target experiments. Moreover, the introduction of steel and other 
 hard armour so entirely changed the character of plate experiments, 
 that it seems better to commence the series of hard armour trials 
 in a separate section. Since, however, soft and hard armour were 
 occasionally compared together, it is difficult or impossible to keep 
 the two classes of experiments apart, that is, those dealing with pene- 
 tration and those dealing with shattering. In the ensuing chapter, 
 then, will be found experiments with competitive shot whose merits 
 were judged up to a certain point wholly by perforation, also some 
 experiments of Krupp's, at Meppen, both in 1879 and 1882, on per- 
 
 '• D. of A., " Proceedings," Vol. XVI., p. 103. 
 
 2 Very's " Development of Armour,'' p. 447. 
 
 3 That is, 16 * J?5 = 16-86. 
 
62 
 
 foration {see Chaps. V. and VI.). Also at Shoeburyness and Spezia 
 wrought-iron plates were used in comparison with those faced with 
 steel or wholly made of steel. 
 
 Chilled-iron armour, which constitutes a very distinct class, is kept 
 chiefly in a section of its own, but here and there comes into the 
 programme of experiments on wrought-iron and steel shields when 
 compared with them. 
 
63 
 
 CHAPTER V. 
 
 Spezia Trials, 1876, with 100-ton Armstrong M.L. Gun, and 
 First Trtals op Steel-paced Iron Plates. 
 
 The employment of steel armour at Spezia, in 1876, was attended with 
 such remarkable . results, that to these must mainly be traced the 
 stimulus given to the trial of steel in armour in this country. The 
 character of the Spezia experiments is peculiar. The object was to 
 test the power of the 100-ton M.L. gun of 17-in. calibre, made by 
 Sir W. Armstrong & Co., and delivered in Italy in 1875, and further 
 to determine the kind of armour that could best be applied to the 
 turrets of the vessels Builio and Dandolo, then in course of construction : 
 the thickness of 21-65 1 inches of plate, 29 inches of backing, strength- 
 ened by angle-iron, and 1£ inch of skin, having been previously 
 decided on. The experiments were commenced on 20th October, 
 1876, but the armour was actually fired at first on 25th October. The 
 several descriptions of targets tried are shown in the accompanying 
 series of sections. 
 
 The same tests were applied in succession to the various targets, 
 with certain exceptions, and with effects shown generally in 8 the 
 figures in successive rows. The top figures exhibit the effect produced 
 by one 10-inch projectile on every section ; with the exception of the 
 three forming the lower halves of Figs. 2, 3, and 4 A, which are shown 
 uninjured. Figs. 1, 2, 3 and 4 B represent the same sections after 
 being further subjected to the blow of one 10-inch and one 11-inch 
 projectile fired in a salvo; the same three sections (forming the lower 
 halves of Figs. 2, 3, and 4 B) again escaping any trial. Figs. 1, 2, 3, 
 and 4 C, show the effect of one projectile from the 100-ton gun on 
 every target. 
 
 The tests thus systematically applied consist, first of one round from 
 a gun of comparatively small power ; secondly of a Balvo from two guns 
 of small power, the 10-inch and 11 -inch guns being only capable of 
 piercing plates of about 13| and 14| inches thickness at the most ; and 
 
 l 65 centimetres. 
 
 • These drawings were not made on the ground by the author, who did not witness this experi- 
 ment. They are taken from the official sections of the Italian Government, on which were entered 
 the effects noted, which were afterwards submitted to an officer, who was present, to check and 
 correct them. Jt is, of course, impossible to show effects at all completely in a single section. 
 These are only intended to enable the mind to grasp and compare general effects readily. — Vide 
 " Engineer," December 29th, 1876, and January 6th, 1877. 
 
 10 
 
64 
 
 lastly of one round from a gun which was much more than a match 
 for the target, the 100-ton gun being capable of penetrating between 
 30 and 31 inches of armour. 1 The two salient features of the experi- 
 ment are, first the powers of the various sections to withstand the 
 effect of comparative light artillery ; and secondly the effect produced 
 on the same sections by the shock of artillery fire which is much more 
 than a match for them. No. 1 in each series exhibits Cammell's solid 
 22-inch wrought-iron plates in the upper half, and Marrel's in the 
 lower. No. 2 shows Schneider's solid 21 -65-inch steel plates in both 
 upper and lower tiers. Nos. 3 and 4 show sandwich targets. Of 
 these the upper portions had front plates of iron 1T81 inches (30 cm ) 
 thick, then 10 inches of wood, next 9-84 inches (25 cm ) of iron, and 
 lastly 19 inches of wood, with skin ; the iron in No. 3 section was 
 supplied by Marrel, and in No. 4 by Cammell. The lower portions of 
 Nos. 3 and 4 consisted of 14 inches of chilled cast-iron behind 8-inch 
 front plates of wrought-iron, 10 inches of wood being sandwiched 
 between the chilled and the soft iron in No. 3, while these were in 
 contact in No. 4, as shown in Figs. 
 
 The effects admit of being dealt with generally. The 10-inch 
 projectile in the first test (Figs. 1, 2, 3, and 4 A) penetrated about 10 
 inches : the depth being greatest in the softest iron, Cammell's (Fig. 
 1 A, upper portion) ; less in Marrel's iron (Fig. 1 A, lower portion), 
 but accompanied with more cracking; and still less in the steel 
 (Fig. 2 A), with considerable cracking. 
 
 When the salvo was fired, the same results relatively were produced 
 in a much more marked degree (Figs. 1, 2, 3, and 4 B). Considerable 
 penetration (about 13 inches) was obtained in Cammell's solid plate 
 (Fig. 1 B, upper portion), a large piece of plate was detached from 
 Marrell's (Fig. 1 B, lower portion), and Schneider's steel plate was so 
 much split (Fig. 2 B) that it was clearly far advanced towards 
 destruction. It thus became evident that steel was quite unsuited to 
 withstand the continuous fire of even comparatively light guns. 
 
 The projectile from the 100-ton gun, on the other hand, passed with 
 ease through the iron structures 3 (Figs. 1 C) ; while the two steel 
 
 1 The calibre of the gun then fired was 17 inches. The projectile weighed about 2000 lbs. It 
 was fired with charges of about 341 lbs., giving striking velocities of from 1476 to 1600 feet. The 
 targets faced the sea a short distance from the gun, which was mounted on a raft. 
 
 3 It broke up, however, interfering with an interesting experiment which was made to measure 
 the work absorbed in perforation, by taking the velocity of the shot after passing through the 
 entire target. The striking velocity of this round was 1478 feet, the weight of shot was 2000 lbs., 
 the energy was therefore 30,296 foot-tons. After passing through the target the velocity was 600 
 feet. This, if true of the entire shot, gives 4,993 foot-tons energy remaining in the projectile ; 
 that is, 26,303 foot-tons had been absorbed in penetrating the target. 
 
 And Log 26303 = 4-4032 
 
 log (ir x 16-92) =■ 1-7266 
 
 2^6777 
 
 log R ° 1-9912 
 
 2 ) 2-6865 
 
 t = 22-07 ins. 1-3438. 
 
 Log K is taken, supposing v to be about 1300 feet ; v might be worked out easily by subtracting 
 logs of g and W from 4-4032 ; halving and looking out the corresponding log. 
 
 The highest striking velocity against plates was 1600. The weight of shot being 2000 lbs., and 
 the diameter 16-92 inohes, the total energy JB=31,200 foot tons, the energy per inoh circumference 
 e=687 foot-tons, and perforation t=24-86 inohes. One round fired against earth had a velocity 
 of 1643 feet, and an energy .E=33,020 foot-tons. 
 
65 
 
 
 
 I A. 
 
 c 
 
 < 
 
 - 
 
 It" 
 
 SHU 
 
 M? 1. 
 
 
 '■,v, 
 
 it'.. 
 
 
 / 
 
 3 
 
 
 
 
 / 
 
 
 .■.>■■ 
 '.:."■ 
 
 WWi 
 
 r ~ 
 
 1 
 
 ^ 
 
 
 
 s 
 
 
 2 A. 
 
 1 B. 
 
 2 B. 
 
 CAIHMEU.S UARREL 
 IRON" PLATffi 
 S*SH TARGET.AFTER BEING STRUCK BV 
 I ROUNO 10" SHOT 
 
 SALVO - 10" AHO W SHOT 
 
 1 c. 
 
 51" FIRE. 
 
 SALVO ll'iio' 
 
 CUNS.ATCAMEU. Zi'pum 
 PENETRATION 13 tt'lB" 
 RESPECTIVELY- WHOLE PLATE 
 DRIVEN BACK IINCH 
 BOLT DRIVEN BACK 
 
 6V FIRE. SALVO 
 ll"&/0" DUNS AT 
 MARREL, LARGE 
 MASS 0I5L0DCED 
 RNO MUCH CRACKINC 
 
 4VFIHE. SALVO 
 11" &IO"CUNS.AT 
 SCHNEIDER STEEL 
 22* PLATE. 
 LARGE PIECE DISLODGED 
 ANO CREAT CRACKINC 
 
 SCHNEIDER 
 
 STEEL PLATES 
 
 UPPER TARGET AFTER BEINC STRUCK 
 
 1 - 10' SHOT 
 
 SALVO- to" AND It' SHOT 
 
 2 C. 
 
66 
 
 3 B. 
 
 4 B. 
 
 I3R0UND 
 salvo ii"&10"at marrel 
 (sandwichtapcet) 
 5 larce pieces dislodceo 
 cracks formed 
 
 RIVETHEAOS BROKEN & 
 
 KpFIRE 
 
 SALVO II" * 10' 
 
 CUNS. PEHETRATTOA 
 INTO REAR PLATE 
 
 RIGHT TOPCORNEh 
 OF PLATE DISLQOaC 
 
 MAR REL 
 'HON (SANDWICH) PLAies 
 
 CAM MELL 
 IRON (SANDWICH) PLAT* 
 
 3 C. 
 
 4 C. 
 
67 
 
 plates {Fig. 2 0) in the act of shivering under its blow so far 
 absorbed the work stored up, that the projectiles did not penetrate the 
 backing, the points remaining as shown in Pig. 2 0. Doubtless 
 attention will be attracted by the projectile of the 100-ton gun being 
 also shown as failing to penetrate No. 4 upper target (Fig. 4C); in 
 that instance however the shot was fired with a charge so greatly 
 reduced, that its penetration ought only to be equivalent to about ] 9 
 inches of plate, and the result accords fairly well with what would be 
 calculated. 
 
 The features to notice here are, that steel is well calculated to enable a 
 vessel to bear a single blow of the projectile from a gun which is much more 
 than a match for the same thickness in wrought-iron ; while on the other hand 
 such plating must be expected to crumble under the continuous fire of guns 
 which could not easily injure wrought-iron of the same thickness. The 
 Builio, if clad in steel 22 inches thick, might probably run past a 
 fort mounting any guns existing; but she might be gradually 
 destroyed by the continuous fire of almost any of our more recently 
 constructed ironclads. 
 
 In every instance the plates were broken across by the projectile of 
 the 100-ton gun ; consequently great fault has been found with the 
 Spezia experiments on account of the narrowness of the plates. If 
 the object was to test plates of the size which was to be used on the 
 ships, this objection scarcely holds, the design of the ship, showing a 
 belt of thick armour of the same width as the plates actually fired at, 
 namely 4 feet 8 inches. On the principle which was laid down in con- 
 sidering wrought-iron, the 10-inch wood layer in the sandwich targets 
 appears much too thick. The chilled iron was used in a manner wholly 
 unsuited to bring out its powers, and failed, as might be expected, 
 vide Figs. 3 and 4 C. The plates were easily broken up, and the results 
 do not deserve careful consideration. 
 
 The result of the Spezia experiments called the attention of the 
 authorities in England to the peculiar powers of steel to such purpose, 
 that it may be said from this date, the investigation was carried on under 
 the conviction that steel in some form must replace iron sooner or later, 
 although the breaking up en masse of steel was regarded as an evil 
 which must, if possible, be prevented. Messrs. Cammell had proposed 
 compound armour plates, consisting of a steel face united by fusion to 
 a wrought-iron foundation plate, on a patent of Mr. George Wilson's, 1 
 which offered the advantages of a hard surface, and a tough mass or 
 body. This seemed to promise well, and trials with it followed quickly. 
 
 Wilson's compound {steel-faced) armour trial. 
 
 On August 1, 1877, (the day of the trial of the 16£-inch wrought- Wilson's 
 iron plate noticed before, p. 138) at Shoeburyness, a small "compound" J£SJ££ nd 
 plate about 8 ft. 7 in. by 3 ft. 6 in. and 9 ins. thick, was fired at by a 
 7-inch gun, with Palliser shot weighing 115 lbs., with a 30 lb. charge, 
 at 70 ft. range, the striking velocity being 1456 ft. 8 The plate was 
 
 1 The total energy B here is 1690 foot-tons, e = 77'8 foot-tons and t = 9"0 inches. For manu- 
 facture vide Chapter on manufacture hereafter. 
 3 Vide " Engineer," August 17, 1877, and December 14, 1877. 
 
68 
 
 9 ins. thick, consisting of a face plate of 5 ins. of steel united to a 
 foundation or back of 4 inches of wrought-iron. The results of the 
 round are shown in Fig. 1, p. 69) . The plate exhibited one deep crack, 
 but it was by no means a through crack, extending only to a depth of 
 about 6 inches. The shot was as nearly a match for a 9-inch iron plate 
 as possible ; that is to say it was just capable of perforating 9 inches of 
 that material. It can hardly be questioned that this plate offered 
 much greater resistance than iron to a single round. The proportions 
 of steel and iron, however, were not those afterwards found best by 
 the makers, the steel generally constituting one third of the plate. 
 
 FIC.3 
 
 COMBINED IRON t> STEEL PLATE, hard steel between two layers 
 
 OF 1FONARMOURPLATE.A. WILS ONS.PATENT B'.O"* S. II x9' THIC K 
 
 fl 
 
 THIS PLATE IS INTENDED TO RESIST PENETRATION 
 AND NOT TO STAR UMDCR IMPACT 
 
 Experimental steel and steel-faced plates on board the Nettle. 
 
 On December 18, 1877, 1 on board the Nettle, took place at Portsmouth, 
 a trial of a 9-inch steel plate of Whitworth's and of 3 experimental 
 9-inch plates, supplied by Messrs. Cammell. The Whitworth plate was 
 much the smallest, it was about 6 feet square, it consisted of soft steel 
 with screw plugs of harder material {vide Fig. 1 herewith), at fixed 
 intervale. It was hoped that the steel plugs would both break up the 
 
 1 Vide "Engineer," Author'» linport, December 21, 1677. 
 
69 
 
 F1C.I 
 
 FIG. i 
 
 WBITWORTH PLOQGED'STEEL PLATE 
 
 Y1£W Of TOP OF PIKSS. 
 
 Ik 
 
 / MtAL&OOLS ARE REERESEinEJ) WIS % 
 
 tuns i 
 
 LiWAB 
 
70 
 
 projectiles and also that any craoks that might be developed in the 
 plate might run to a steel plug and there stop, on the same principle 
 that the further extension of a crack found in a boiler plate is pre- 
 vented by drilling a hole at the extremity of it. Sir J. Whitworth 
 has acted on the same principle in the construction of other steel 
 plates, it having occurred to him that the best way to limit the cracking 
 in steel would be to " manufacture the cracks" as he expressed it by 
 sub-dividing the armour into small plates. It is easy to understand 
 that a certain measure of displacement might be provided for in this 
 
 FIC.2 
 
 SOLID STEEL PLATE "Suhcarlurijed 
 
 CAMMELLS PATENT 9-&X 7-9 X 9 " THICK 
 
 THIS PLATE IS INTENDED TO OFFER A GREATER 
 RESISTANCE TO PENETRATION AND 
 ttOT TO STAR UNDER IMPACT 
 
 way, and that plates that would not bear distortion without craoking 
 might experience relief, but the question arises as to what shape the 
 work done would take, and how the structure would bear it. Messrs. 
 Cammell had the following plates, whose dimensions are shown in the 
 figures. Sub-carbonized steel, (Fig. 2) containing about 0'13 per cent, 
 of carbon, depending simply on the resisting power of the metal. 
 Fig. 3, a 6|-inch steel plate, containing 0'57 per cent, of carbon, faced 
 with f-inch of iron and built with 1^-inch of iron at the back — 9 
 inches in all. The idea here was to hold the steel from cracking in 
 front and behind. Fig. 4, p. 71, consisted of a hard steel face plate, 
 
71 
 
 containing 0'64 per cent of carbon, 5 inches thick, backed by 4 inches 
 of iron. Here the face might crack, but the plate might be held 
 together by the iron. A wrought-iron plate of excellent quality had 
 been fired at to furnish a standard of comparison. The front and back 
 of this is shown in Figs. 5 and 6, p. 72. The results of the firing are 
 
 ric. 4- 
 
 COMBINED IRON & STEEL PLATE, hard^ steel on ibon ASMOVRPLAIE 
 A. WILSONS PATENT 9.9 'x 7-'/z"xB Thick 
 
 VJ.EJN OP TOP EDGE ISP P£AT£ 
 
 THIS PLATE IS INTENDED TO RESIST PENETRATION BY OESTBO.Y1NQ OS 
 
 BREAKING OP THE PROJECTILE ON IMPACT 
 
 shown generally in the figures. They were not particularly satisfactory. 
 Pig. 4 turned out to be a bad representative of the system, and the 
 shot perforated it, hence only two rounds were fired. As to the others, 
 most Officers would prefer to have the wrought-iron plate here shown 
 on service to either of these for this particular blow, but it must be 
 borne in mind that the Spezia trials had not shown the superiority of 
 steel to iron in the case of a blow which was just a match for the 
 latter, but rather the power of steel to stop a shot which would pass 
 through the iron with a good deal of spare energy. It is probable 
 that Plates 1, 2 and 3, would have stopped a projectile, which would 
 have perforated the iron plate and passed on into the ship which it was 
 intended to protect. This experiment, however, was a crude one, and 
 it is rather given to show the devices that were tried to rival the 
 powers displayed by the steel, even at this time. The further trial 
 of compound plates in England, was conducted in connection with the 
 
 11 
 
72 
 
 investigation of the question of the best metal for projectiles by a Sab- 
 Committee under General Inglis, R.E., whose proceedings may next be 
 considered. 
 
 no. s 
 
 FRONT VIEW OF WROUGHT IRON STANDARD PLATE 
 
 FIC.6 
 BACK VIEW OF WROUCHT IRON STANDARD PLATE 
 
73 
 
 CHAPTER VI. 
 
 Proceedings of the Sub-Committee on Plates and Projectiles. 
 
 A request was made by the Admiralty on June 5th, 1877, 1 that experi- 
 ments might be instituted in order to obtain reliable information as to 
 the most suitable metal for armour-piercing projectiles, the following 
 objects being kept in view : — (1) Penetration. (2) Destructive effect 
 of the explosion of a bursting charge on armour and backing. 
 (8) Effect on the interior of armoured structures when completely 
 penetrated. (4) Liability of the shell to break up in the gun. 
 
 The improvements in steel and the experiments whioh had been 
 made with guncotton charges were the causes that led to the series of 
 experiments about to be briefly described. A Sub-Committee of which 
 General Inglis, R.E., was President, was appointed to investigate the 
 matter which was dealt with in the following manner : — 
 
 The experiments were divided into two series. (1) To test material 
 for armour-piercing projectiles. (2) To test the best combination of 
 form and material for armour-piercing projectiles. 
 
 1st aeries. (Material). It was decided to obtain projectiles from 
 various makers for the 9-inch M. L. gun, and to fire them with a 
 striking velocity of about 1500 feet, all the projectiles to be of the 
 Bervice pattern for the 9-inch 12-ton M. L. gun, that is, a projectile 
 with studs, and with an ogival head struck with 1 \ diameters radius, 
 weight 268 lbs. The projectiles were first to be tried in competition 
 against unbacked wrought-iron plates, each one 4 feet square and 12 
 inches thick, suspended on short arms fixed in them like trunnions 
 [vide Pigs. 4-5). These small plates were cut from plates 16 feet long, 
 so that one long plate would cut into four pieces, which would put 
 competing shot on as equal terms as possible. The following projectiles 
 were sent in for competition : — 
 
 Chilled Iron. (1) Service shells from store. (2) Royal Laboratory 3 CMUed iron 
 improved. (3) Gregorini iron 3 supplied from Italy, cast into pro- competitive 
 jectiles in the Royal Laboratory. (4) Shells from the Finspong Iron projeotlleB - 
 
 1 Report of Sub-Oommittee, p. 19, June 22nd, 1880. Engineer, April 12, 1878, and August 
 23rd, 1878. 
 
 2 The metal of these shells contained a small proportion of steel. 
 
 3 This metal is important, having been the material employed for the chilled shot employed in 
 most of the Spezia trials in which this metal was rather differently treated, and proved in its own 
 way to be very good. 
 
74 
 
 Company, Sweden. (5) Shells from Griison (Germany). (6) Shells 
 from Krupp, Essen. 
 
 Steel. (7) Shells from the Landore and Siemens Steel Company. 
 (8) Shells from the Terre Noire Steel Company, France. (9) Shells 
 from Messrs. Hadfield & Co. Shells with cast steel bodies and chilled iron 
 heads were sent in from (10) Messrs. Vickers & Co., and (11) Messrs. 
 Cammell & Co. Forged steel shells were submitted from (12) Messrs. 
 Firth & Co., and (13) Sir J. Whitworth & Co. 
 
 The price of the chilled shell was from £26 to £28 per ton, or 
 £2 19s. 7d. to £3 5s. per shell, except Krupp's, which was £39 per ton, 
 and the service shells which were about £10 10s. per ton. The cast 
 steel varied from £6 17s. 7d. to £8 0s. 7d. per shell. The cast steel 
 with chilled points, Cammell's, £5 14s. 8d. per shell, Vickers, £10 per 
 shell, and the forged steels from £8 to £15 each shell. 
 
 Great pains were taken to secure equality in conditions in these 
 trials. The wrought iron plates were fired at with the following 
 result : — 
 
 Chilled iron shell. (1) The service projectiles were of mottled iron 
 with close grained white chilled heads. They nearly penetrated these 
 12-inch plates, but broke up as they usually do. (2) The Laboratory 
 improved shells (containing 17 per cent, steel) appeared slightly 
 softer. They got their points through the plates or broke the 
 plates up, with the exception of one bad one ; the shells themselves, 
 however, breaking up invariably. (3) Gregorini had soft bodies, but 
 the heads were ohilled white, in fact, nearly like the service shells. 
 These shells were unfortunate ; twice a corner of plate was struck off, 
 once the plate was broken up, and in the two other rounds the 
 penetration was not good. (4) Finspong dark grey with white heads, 
 exterior turned, only three shells were fired. The points penetrated 
 the plates and the shells broke up much less than other chilled pro- 
 jectiles. One held together (vide Fig. 4, p. 75). (5) Griison chilled 
 over head and body about half way to centre. Three shells were fired, 
 one got its head well through, all broke up, (vide Figs. 1, 2, and 3, 
 p. 76). Figs. 1 and 2 do not give a fair sample, but show the 
 behaviour of a bad specimen. (6) Krupp's, heads chilled one-third of 
 depth to oentre, two shells only fired, penetration not over 11-8 inches, 
 bases of shell broke up. 
 
 Of these chilled shell the Laboratory improved were considered to 
 give " decidedly the best results, both as to depth of indent and amount 
 of damage done at the rear of the plate." As to cast steel shell — The 
 Terre Noire only held well together, (vide Figs. 5 and 6 p. 75), and 
 they set up so much that the Laboratory improved shell were preferred. 
 Of the cast steel with chilled points, Vickers was thought to be too soft, 
 but Cammell's were considered better than any of the shells above- 
 mentioned, (videYigs. 8, 9 and 10, pp. 77 and 76). Of the forged steel, 
 Firth's were considered soft, but those of Sir J. Whitworth were 
 admirable, some of them standing so well that after passing entirely 
 
75 
 
 FRONT STUDSn CUPPED OFF 
 FLDSH. REAR STUDS/UNINJURED 
 
 i.4- 
 
 FIWSPONC 
 
 PROJECTILE 
 
 CH/LLED METAL IN 
 
 PLATE 
 
 Nf a0S4 , 
 
 ^•DOTTED LINE /|[!| 
 
 * SHOWS ORIGINAL I I 
 FORM OF POINT V|i| 
 BEFORE BEING IM ]j| 
 SET UP " 
 
 ON IMPACT 
 
 ALL. BTODS 
 CLIPPED OF/ 
 FL0SH WITH 
 SHELL EOD^ 
 
 NO 
 PERCEPTIBLE) 
 SETTINC UP 
 OF BODY 
 
 F/C7 
 
 Whit worth steel projectilb 
 after passage completely 
 thhoug. 12 lm plate 
 
 FIG. 6 
 
 Terre Noire projectile 
 
 fsTEEL) N9 £039. AFTER IMPACT 
 FIG, 5 
 
 Terre Moire projectile 
 
 (STEEL) N°. 2093 
 
 "«mjjip!!ipi»^ 
 
76 
 
 FIC.I 
 IMPRESSION IN PLATE MADE BY 
 CRUSON CHILLED PROJECTILE IM? 20SS 
 
 FIC.3 
 
 ORISON 
 SHOT PO/NT 
 
 no 2 
 
 SECTION OF IMPRESSION 
 
 FIC.9 no. 10 
 
 PROFILE OF 12 IN. PLATE. AFTER VIEW OF CflMMELLS PROJECTILE 
 
 IMPACT Br CAMNIELLS PROJECTILE W 2103 HEAD WITH POINT BROKEN THRO BACK OF PLATE 
 
 VELOCITY 
 
 1520 FT 
 
 PER SECOND 
 
77 
 
 FIQ. B 
 
 Cammell's (^Wilsons) projectile 
 
 (n?2097) STEEL WITH CHILLED POINT 
 SHOWN WITH EOINT APPEARING 
 THROUGH PLATE 
 
 SHOT PO/NT~<. 
 BROKEN OFF IN \ 
 MIDST OF BEN& 
 LATER S 0f\ 
 PLATE 1 
 
 through the plate, {vide Fig. 7, p. 75), they were found uninjured, 
 re-studded, and fired a second, (vide Fig. 3, p. 78), and even a third 
 time. The Committee then selected three kinds of projectiles as the 
 best, and continued their experiments with them only, namely, the 
 Laboratory improved chilled projectile, the Cammell cast steel with 
 chilled point, and the Whitworth forged steel. 1 
 
 1 Excellent as this trial was, subsequent experience seems to show that it was imperfect in one respect 
 as it was indeed pointed out by the Committee in their report. The competition was confined by the 
 work of punching holes in toft iron. It appears to have been assumed that the projectiles that 
 performed this work best would also be the best for attacking the hard armour subsequently fired 
 at. Now, as really hard armour cannot be punched but must be destroyed by racking, this might 
 
 Eerhaps be better done by some of the soft projectiles which set up and penetrated badly but still 
 eld together, than by the chilled shell that beat them in the work of perforating soft iron, at 
 the same time breaking up. The investigations of the Committee, even to the end, were too much 
 confined to softer classes of armour. On this, vide remarks of Committee on chilled iron, p. 238. 
 
78 
 
 Against iron plates with air spaces between them. Whitworth's 
 forged steel alone held together and penetrated all plates without 
 breaking up, and that even at an angle. The plates were two 2-inch 
 plates inclined at 30° to an 8-inch plate, 4 feet 6 inches in rear of it. 
 Cammell's, when fired direct at two 2-inch plates 4 feet 6 inches in 
 front of 8-inch plate and parallel to it, broke up, the point penetrating 
 6*75 inches into the 8-inch plate. Firth's shell was fired at an angular 
 air space target like Whitworth's ; the shell made a hole through all, 
 but remained broken in the last plate. Pig. 4 shows the back of a 
 
 FIC.3 
 
 ■Plppl 
 
 awsssffl' 
 
 m 
 
 i,;., 
 
 FIC.+ 
 
79 . 
 
 wrought-iron plate after perforation by a Cammell- Wilson cast-steel 
 shot with chilled iron point, fired on July 24th, 1877 ; striking velocity 
 1440 feet. The plate, however, had suffered from being fired at 
 already. 
 
 The further experiments of this Committee consisted in firing the Fnr . u »« ex- 
 selected classes of proj ectiles, with variationsin form and make, at wrought- S£h»SwLa 
 iron plates, either single or with air spaces between them, and at steel- pr °J eetUel '- 
 faced wrought-iron plates directly and obliquely. The chief changes in 
 form and make consisted in getting rid of the studs, a rotating gas check 
 being employed on the base of the projectile, and in trying various 
 forms of head, that is, fiat-heads, ogival points struck with radii from 
 1"5 to 3 diameters, and one or two specially pointed shot. 
 
 It was found that fired directly against a 14-inch wrought-iron plate 
 a Cammell, with a specially sharp point was best, then came Whitworth, 
 Laboratory, and Cammell with two diameter heads. Owing to a failure 
 in armour plates, these were the sharpest points tried, among many 
 blunter ones, the Whitworth flat-headed, in order of merit, came last of 
 all. This implies that for direct penetration into iron a sharp point is 
 best. 1 
 
 Firing at 12-inch wrought-iron plates at angles varying from 30° to 
 37° 4' with the normal (60° and 52° 56' with the face of the plate), 
 the Laboratory improved was rather better than the Whitworth. In 
 each case the 2 diameter head beat the 1*5 diameter head. All pro- 
 jectiles broke up on oblique impact, except 2 Whitworth forged steel 
 shells out of 5 fired. At the same plate at 29° 30' to 31° 30' with the 
 normal (60° 30' and 58° 30' with the face of the plate) the Laboratory 
 2 diameter head was the best, the other 2 diameter heads were beaten 
 by the corresponding ones of 1*5 diameter. The Whitworth flat-head 
 was about midway in merit among the projectiles. 
 
 When the projectile was not a match for the plate, the projectiles bit 
 at 30° with normal (60 with plate) ; at more oblique angles (37° 4' 
 with normal) scoops only were made. When the shot was more than 
 a match for the plate it bit at a more oblique angle, namely, at 35° to 
 36° with the normal, at 40° (50° with face of plate) only a scoop was 
 made. 
 
 In direct firing 3 against steel-faced plate 4 feet square and 12 inches 
 thick, CammelPs 2 diameter head shell obtained the best penetration, 
 namely, 18 inches. Whitworth's were nearly as good. None of the 
 projectiles held together. Whitworth' s broke into the fewest pieces. 
 The only chilled shell fired broke up with a penetration of only 
 2 inches. 
 
 In oblique firing at steel-faced armour, 10 inches thick, consisting 
 of 4 inches of steel on 6 inches of iron, striking at an angle of from 
 25° to 27° with the normal, (65° to 63° with the face of the plate). 
 
 1 Sir W. Palliser, on a small scale, found that a steel bullet fired from a rather large bore small 
 arm (the Minie), with a point struck with 6 diameters radius, gave exceptionally good penetration. 
 This is not surprising. Plates are torn open generally, the tear commencing in a cross or star at 
 the back of the plate, opposite the point of the shot which is able to reach the spot, and act 
 directly on it more easily in the proportion in which the point is made sharper. 
 
 2 Heport of Sub-Committee, p. 11. 
 
 12 
 
of Sub- 
 committee. 
 
 - 80 
 
 Cammell's steel shells with chilled point, 2 diameters radius, were 
 slightly the best, but they broke up even against comparatively weak 
 armour. The Committee arrived at the following conclusions : — 
 
 concisions As to material. The Whitworth forged steel gave the best results 
 when fired direct against wrought-iron, and they alone would have 
 carried a bursting charge behind the armour. 
 
 The Oammell shells (steel with chilled point) produced most effect 
 against steel-faced armour, both indirect and oblique fire, but broke 
 up. Chilled iron gave the best results when fired obliquely against 
 iron, but invariably broke up both on direct and oblique impact. 
 "Whitworth forged steel shells striking with a velocity of from 1700 to 
 2000 feet were only slightly set up, none were fractured (against 
 wrought-iron). One Cammell forged steel shell held together after 
 striking a 16 - 5-inch wrought-iron plate. 
 
 " The Committee think it important to point out that they have had 
 no means of arriving at any conclusion as to the best form and 
 material for projectiles, with which to attack the chdled-iron armour, 
 which is being adopted by Foreign Powers particularly for Coast 
 Batteries." 
 
 On the whole, the Committee found — (1) Steel projectiles absolutely 
 necessary for the attack of steel-faced armour. (2) In mere penetration 
 in wrought-iron, steel and chilled are nearly equal, chilled being 
 slightly better obliquely. (3) If fired direct even against wrought- 
 iron, not beyond the power of the gun, with high velocity, chilled 
 shells almost invariably break up. Cast-steel may, if improved, 
 eventually hold together. Forged steel may be implicitly trusted to 
 remain unbroken. (4) The metal 1 of the Laboratory improved shells 
 is very superior to that previously employed. 
 
 With regard to the form, studs were recommended to be abolished, 
 
 and an ogival head with 2 diameters radius was preferred on the whole 
 
 for both chilled and steel shot. As to the relation of diameter of shot 
 
 to weight, the Committee recommend something heavier than that 
 
 W 
 shown by the relation -^ = , 364. 2 
 
 Compound (steel-faced) armour undoubtedly was found to possess 
 the great advantage of shattering chilled projectiles, and causing even 
 forged steel ones to set up and to deflect projectiles striking at an 
 angle well, but the Committee considered that these results were " only 
 to be relied upon when the plate is beyond the power of the gun." 
 
 In direct fire, a good compound plate between 9 and 10 inches thick, 
 was found to be about equal to a good 12-inch wrought-iron plate, for 
 a single blow ; under repeated blows, it is decidedly inferior to the 
 wrought-iron plate. In this comparison, steel projectiles are supposed 
 to be used against the compound plate, while either chilled or steel 
 may be used against the wrought-iron. In oblique fire, the 10-inch 
 compound plate was broken up by blows amounting to 554 foot-tons 
 
 i Containing 17 per oent. of steel. 
 
 W 
 3 This has more recently been brought to — - = 0-42, see Beport, p. 16. 
 
81 
 
 per square foot. The 12-inch wrought-iron plate bore 650 foot-tons 
 per square foot, and might, it was expected, bear blows amounting to 
 about 992 foot-tons in all. 1 
 
 Finally the Committee made the following recommendations : — 
 
 1. That all chilled battering projectiles should be of the kind tested "StUm 
 
 as " Laboratory improved." 
 
 2. That all battering projectiles should have heads struck with a 
 
 radius of 2 diameters. 
 
 3. That the question of a delay-action fuze to be used with gun- 
 
 cotton be further investigated. 
 
 4. That a certain proportion of forged steel shells be issued, the 
 
 proportion depending on the results of recommendation 
 No. 3. 
 
 5. That cast steel projectiles be not adopted unless greatly 
 
 improved. 
 
 6. That in armour-piercing shells, capacity for bursting charge be 
 
 considered as of less importance than strength of head and 
 walls. 
 
 7. That the proportion of weight to calibre in armour-piercing 
 
 shells be investigated with larger guns. 
 
 Three questions may be noticed separately — (1) The employment of 
 guncotton. (2) The trial of a wrought-iron cap on the point of shot. 
 (3) The investigations as to weight and velocity carried out with the 
 Armstrong 6-inch and 8-inch new type guns. 
 
 (1.) The employment of guncotton is specially important because 
 the power to carry fire through armour appears at present to depend 
 on it. As noticed above, forged steel alone give promise of holding 
 together while passing through thick armour, even of wrought-iron, 
 and forged steel shells refuse to be burst by gunpowder. Guncotton, 
 burst in the ordinary way, has disappointed expectation in Italy. It 
 is however difficult to provide for its detonation in a way that does not 
 endanger the gun. On this the Committee report that they had not 
 sufficient experience to speak positively. The delay-action fuze, tried 
 by them, enabled shells to carry fire through armour up to their power, 
 and a fuze being found which detonated the wet guncotton charge of a 
 shell striking armour with a velocity not exceeding 1554 f.s. Further 
 investigations as to a delay-action fuze were recommended. Figs. 1 
 and 2, p. 240, show the effect of a Whitworth steel shell fired with a 
 bursting charge of 4 lbs. 2 ozs. at a 12-inch wrought-iron plate. The 
 shell refuses to break up, the gas escaping at the base plug hole. The 
 shell is expanded against the plate, thus the latter is cracked, while the 
 shell, instead of passing through the plate, is held. 
 
 i Amounting to 3093, 3023 and 4614 foot-tons per ton of plate respectively for compound 10-in. 
 and wrought-iron 12-inch actual and estimated tests, assuming each to weigh 0-216 tons per cubic 
 foot. 
 
82 
 
 FIC. I 
 
 FIC. 2 
 
 iijiliiiiijiiiillwi 
 
 \ 1, 1, 1. 1, i. :, i. ,nii h in. 1 1 1 . 
 
 
 Wrought- 
 iron Gap 01 
 point of 
 projectile. 
 
 (2). The trial of a wrought-iron cap 1 on the point of the projectile 
 arose from the fact being discovered that a steel-faced armour plate 
 lost its power of breaking up chilled shot, its resistance being 
 diminished, to less than that of wrought-iron, by simply placing a 
 2^-inch wrought-iron plate over the steel face of the plate. A Palliser 
 projectile doing more work on the compound plate with this addition 
 than without it, and the effect on the shot itself being quite different, 
 General Inglis concurred with Captain English in thinking that a 
 similar result might be produced, and a chilled projectile kept from 
 
 1 Pages 43 and 125, see Report, p. ii, Report of Sub-Committee, see index, p. 176. This 
 
 experiment was suggested by Major English. 
 
83 
 
 breaking up by applying a wrougM-iron cap to the point when 
 attacking steel-faced armour. This was tried. ' The effect of the first 
 shot so fired encouraged further investigation of the subject, but further 
 trial showed that no advantage was gained by the cap. 
 
 (3). The investigations made by the Sub-Committee as to the effect I™ or** 
 of projectiles fired at armour with a sufficiently great variation in i«w\>°f per. 
 velocity to determine some law as the power of projectiles to perforate tontioa - 
 under various conditions, can only here be dealt with briefly. The 
 Armstrong B. L. 6-inch and M. L. 8-inch new type 1 guns which were 
 placed at the service of the Committee, achieved so high a velocity 
 that experiments could be made with greater variation in conditions 
 than hitherto, the striking velocities varying between 836 feet and 
 2212 feet per second. With the 6-inch gun which was first fired with 
 the following objects — (l.) 2 To determine the penetration of 6-inch 
 projectiles of uniform weight with different velocities. (2) To 
 determine the penetration with a uniform energy, whether made up of 
 high velocity and a light projectile, or of a low velocity and a heavy 
 projectile. 
 
 Experiment 1. The first object was to arrive as exactly as possible at 
 the velocity necessary to perforate unbacked and unsupported wrought- 
 iron plates varying in thickness from 4 inches upwards, with an 80 lb. 
 shell with a 1*5 diameter ogival head. A 12-inch plate is more than 
 proof against this shot at the muzzle 3 a 6-inch plate was proof against 
 the same at 2820 yards, and a 4-inch at 6000 yards. 
 
 Experiment 2. To determine penetration with uniform energy, made up 
 of weight, and velocity, in varying proportions. Shells of 60, 80, and 
 100 lbs. weight were used, and 8-inch, 12-inch, and 13-inch plates. 
 The total energies for these plates being 1080 foot-tons, 1800 foot-tons, 
 and 1900 foot-tons respectively. The 80 lb. shot gave the best 
 result, then the 100 lb. and the 60 lb., the worst result. It was not 
 found, in fact, that a uniform energy gave a uniform result, and with- 
 out basing too much on a limited series of rounds (35 in all) the 
 Committee considered that it was shown that a 60 lb. shell was too 
 light to develop the full battering power of this gun. 
 
 The trials with the 8-inch gun had the following objects in view : 4 — 
 (1.) To determine penetration with a uniform weight of shot and varying 
 velocities. (2.) To determine penetration with a uniform energy, made up 
 of weight and velocity in different proportions. (3.) To compare penetrations 
 of shells with heads struck with radii of 1'5 and 2'0 diameters, other 
 conditions being the same. 54 rounds were fired, 43 being with chilled 
 projectiles and 11 with forged steel. Some of the chilled projectiles 
 held together on impact, but the highest velocity at which they did so 
 was 1337 feet. This gun was capable of penetrating about 14-7 inches 
 
 ' Vide D. of A. " Proceedings," 1879, Vol. 17, p. 188. 
 
 5 Vide p. 84, " Beport." 
 
 • With a striking velocity of 2030 and a weight of 80-26 lbs. E (striking energy) is 2294 and e 
 (energyper inch) 121-7 and t (thickness that would be perforated) 11-67. The projectile got its 
 point through. 
 
 *Se» "Beport," p. 37. 
 
84 
 
 of iron at the muzzle, with a velocity of 2010 feet. 1 As to weight, it 
 
 appeared that with equal striking energy a shell of 132*5 lbs. was 
 
 inferior to shells of 182*2 lbs. and 232*5 lbs. The two latter produced 
 
 nearly equal effects. The 2 diameter head proved better than that of 
 
 1*5 diameters. 
 
 The experiments with the 6-inch and 8-inch guns led to the con- 
 
 W 
 elusion that with these guns -^ ought at all events not to be below 
 
 W 
 0*364. Results were not always good when ^ reached 0*47, perhaps 
 
 owing to the length of the shot causing it to set up. 
 
 With a 13-pr. gun it was found that projectiles had nearly the same 
 
 W 
 effect if their energy was the same, whether t, = 0*508 or 0*364, but 
 
 experiments were not made with lighter projectiles. 
 
 The following table shows the results obtained with the 8-inch 
 M. L. gun : — 
 
 
 Be'sumd of Experiments against Iron Plates with 8-in. Muzzle- 
 
 loading Guns. 2 
 
 
 
 
 
 
 Velocities. 
 
 
 
 
 
 Charge. 
 
 
 8hot. 
 
 Feet 
 
 Energies, foot-tons. 
 
 
 
 
 
 
 
 
 per eecond. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1° 
 
 Thickness 
 
 
 a 
 
 
 • 
 
 
 
 
 
 3 
 
 *3 
 
 
 of plates. 
 
 Remarks. 
 
 Q 
 
 
 o 
 
 
 Dt-scription. 
 
 
 ex 
 
 o 
 
 o 
 
 
 
 
 "3 
 
 2 
 
 •g 
 
 ,3 
 .2? 
 
 
 H 
 
 a 
 
 3 
 
 
 q 
 "3 
 
 
 
 
 
 *o 
 
 
 . c 
 
 
 
 
 T 
 
 
 
 "g 
 
 fe'S 
 
 
 
 K 
 
 is 
 
 a 
 
 P 
 
 
 3 
 
 CO 
 
 3 
 
 55 
 
 fc 
 
 
 
 lb. 
 
 
 lb. 
 
 
 
 
 
 
 
 inches. 
 
 
 1 
 
 24-6 
 
 B.L.G 
 
 183-5 
 
 Head 15 
 
 diameter. 
 
 Chilled shell. 
 
 1035 
 
 1023 
 
 1365-6 
 
 1324-4 
 
 63'06 
 
 6 
 
 Through easily. 
 
 2 
 
 20 
 
 
 
 n 
 
 931 
 
 922 
 
 1096-8 
 
 1076-8 
 
 43-10 
 
 6 
 
 Nearly through.. Pene- 
 tration 12'5 ins., appar- 
 ently nearly through ; 
 shot entire and uninjured 
 
 3 
 
 21 
 
 
 
 
 941 
 
 932 
 
 1120-5 
 
 1099-2 
 
 44-0 
 
 6 
 
 Nearly through. Pene- 
 tration 1'8 in., but little 
 difference from the last. 
 Shell entire and unin- 
 jured. 
 
 Through. Projectile 
 broken. 
 
 Through. 
 
 4 
 
 21 
 
 " 
 
 • 
 
 • 
 
 972 
 
 962 
 
 11932 
 
 1171-1 
 
 46-9 
 
 6 
 
 5 
 
 31 
 
 w 
 
 „ 
 
 „ 
 
 1176 
 
 1157 
 
 1760-2 
 
 1694 
 
 67-8 
 
 8 
 
 6 
 
 29 
 
 B 
 
 , 
 
 „ 
 
 1142 
 
 1123 
 
 1650-5 
 
 1596 
 
 63*9 
 
 8 
 
 Practically through. 
 
 7 
 
 46 
 
 pebble 
 
 , 
 
 , 
 
 1320 
 
 1296 
 
 2205-0 
 
 21256 
 
 88-0 
 
 10 
 
 Penetration 9'4-ins. 
 
 8 
 
 64 
 
 , 
 
 , 
 
 t 
 
 1438 
 
 1411 
 
 2617- 1 
 
 2519-5 
 
 100-9 
 
 10 
 
 Penetration 12-ins. 
 
 9 
 
 67 
 
 B 
 
 ■ 
 
 • 
 
 1497 
 
 1468 
 
 2843-0 
 
 2727-1 
 
 109-2 
 
 10 
 
 Through. Head of shell 
 with band lying in rear 
 of plate. 
 
 10 
 
 68 
 
 n 
 
 
 w 
 
 1679 
 
 1648 
 
 3567-4 
 
 34370 
 
 137-7 
 
 12 
 
 Practically through. Head 
 and first band through 
 plate, rest sticking in 
 plate ; but rear of pro- 
 jectile 2 ins. within face 
 of plate. 
 
 Penetration 14-ioe. 
 
 
 
 
 
 
 
 
 
 
 
 
 11 
 
 83 
 
 § 
 
 . 
 
 w 
 
 1894 
 
 1860 
 
 4539-6 
 
 4378 
 
 175*4 
 
 (12+2) 
 14 
 
 (12+2) 
 14 
 
 12 
 
 88 
 
 ■ 
 
 i 
 
 w 
 
 1987 
 
 1950 
 
 4096-2 
 
 4812 
 
 192-7 
 
 Through. 
 
 1 Bee " Keport," p. 43. The weights of shell of the same form will vary with the cube of their 
 
 diameter for simple mathematical reasons. Hence, — - expresses the factor on which the variation 
 
 d 8 
 
 in weight, or as it is sometimes termed, the sectional density depends. In solid shot of the 1 
 form of head any variation must be due to variation in length. In shells the size of interior 
 cavity of course affects it. Shape, and even nature of metal in some cases may affect it. 
 
 > This forms part of the remarkable series of experiments on which the systems of calculation in 
 Chap. III. were based. 
 
85 
 
 Rismmi of Experiments against Iron Plates with 8-in. Muzzle-loading Quns. 
 
 Charge. 
 
 Shot. 
 
 Velocities. 
 
 Feet, 
 per second. 
 
 Energies, foot-tons. 
 
 Description, 
 
 Thickness 
 of plates. 
 
 Remarks. 
 
 100 
 19 
 
 R.L.G 
 
 21 27-5 
 22 110 pebble 
 R.L.G 
 
 28 
 
 37 
 
 20 
 
 19 
 19 
 
 65-5 
 66 
 
 51 
 
 60 
 
 60 
 111 
 
 31 
 27-6 
 
 60 
 
 63 
 
 48 
 
 pebble 
 
 132-5 
 232-5 
 
 inches. 
 
 (12+2+2) 
 
 16 
 
 2 diameters. 
 
 1-5 diameter. 
 2 diameters. 
 1-5 diameter 
 
 2141 
 
 892 
 
 2101 
 
 884 
 
 6800-8 
 1004-7 
 
 veloci ty los t 
 
 B.L.G 
 
 pebble 
 
 1326 
 
 232-5 
 
 182-6 
 
 132-3 
 232-6 
 
 1117 
 1127 
 1342 
 
 975 
 
 1098 
 1108 
 1308 
 
 1526-6 
 1653-8 
 1654-7 1571-9 
 
 1532-5 15138 
 
 2332 2190 
 
 2 diameters. 
 1*5 diameter. 
 
 1458 
 1470 
 
 1426 
 1361 
 
 1363 
 2266 
 
 1120 
 
 1320 
 
 946 
 
 1343 
 
 1554 
 1217 
 
 6586-1 
 988-9 
 
 975-5 
 
 1525-7 
 1553-6 
 
 6069-3 
 
 922 1097 
 
 1020 
 
 987 
 
 1421 
 1441 
 
 1399 
 1335 
 
 1337 
 
 2212 
 
 1102 
 
 1286 
 
 939 
 
 1317 
 
 1613 
 1201 
 
 2113 
 
 2735 
 
 2573 
 2344 
 
 2361 
 441 
 
 1587 
 1601 
 1443 
 
 2216 
 2388 
 
 39-a 
 
 61-1 
 
 62-2 
 
 1000 
 968-6 
 
 2654-6 
 2627-0 
 
 2476-1 
 2264-7 
 
 2261-6 
 8190-1 
 
 1536-4 
 
 1519 
 
 1421-1 
 
 2194-3 
 
 2102-6 
 2324-7 
 
 43-1 
 
 40-0 
 38-7 
 
 102-0 
 
 (6 + 2) 
 
 8 
 (6+2) 
 
 (14+2+2) 
 18 
 
 10 
 
 104-9 (special) 
 10 
 
 (special) 
 10 
 
 90-3 
 247-2 
 
 61-4 
 
 60-7 
 
 66-8 
 
 87-64 
 
 84-0 
 92-9 
 
 10 
 
 10 
 
 (solid) 
 16-5 
 
 (6+2) 
 8 
 
 (6+2) 
 8 
 
 (2+6) 
 8 
 
 12 
 
 12 
 12 
 
 Penetration 12*4 ins, 
 Grazed on a plate lying 
 flat before striking. 
 
 Plate at an angle of 20 
 
 degrees. 
 Practically through. 
 
 Penetration 20-95 ins. 
 Penetration 10-8-ins. 
 
 Head 4'8-ins. in rear of 
 
 plate. Shell broken into 
 
 big bits. 
 
 Penetration 7'65-ins. 
 Penetration 7'9-ins. Shell 
 
 entire at foot of plate. 
 Penetration 9-ins. 
 Penetration 13-ins. 
 Penetration 9'5-ins., point 
 
 1'5-in. in rear of plate. 
 Through. Head hit 
 
 another plate after 
 
 passing through and 
 
 flew into dust. 
 Penetration 18-ins. Shell 
 
 paased through the 2-in. 
 
 plates breaking up tho 
 
 14-in. plate. 
 Nearly through. Point 
 
 of head through. Mould- 
 ed off to 2'6-ins. 
 Penetration 7'9 ins. Shot 
 
 entire. 
 Penetration 12'5-ins. . 
 
 6*5-ins. behind back of 
 
 plate. Shell unbroken, 
 
 but cracked; moulded off 
 
 to 1-26-ins. 
 Through. Plate broke - 
 
 shot in small pieces. 
 Shell passed olean 
 
 through ; hit target in 
 
 rear and broke into small 
 
 pieces. 
 
 Clean through. Shell 
 broke as before • gi 
 check and all through. 
 
 Penetration 13'05-ins., gas 
 check and body in four 
 pieces in front ; point of 
 shell 3'05-ins. behind 
 back; head, &c, inhole. 
 
 Penetration 16'05-ins. j 
 whole of shell in hole ; 
 point 6'05-ins. behind. 
 
 Shot apparently not good ; 
 
 Elate tell on to its face 
 •ont moulded off to 
 
 2'6-ins. 
 Penetration 10'45-hii. 
 
 rear ; moulded off to 
 
 3'4-ins. ; gape 4-inB. 
 Penetration 9'32-ins. ; 
 
 shot remained sticking 
 
 in ; moulded off to 2-ins. 
 Penetration 8-ins. ; plate 
 
 broken up and much 
 
 thrown about. 
 Penetration 9-45-inB.; 
 
 point as shown and rest 
 
 in fair-sised pieoeB. 
 Penetration 8'3-ins. ; 
 
 body in fair-lined pieces. 
 Struck near former hole, 
 
 and turned into it. 
 
86 
 
 
 RhumS of Experiments against Iron Plates with 8-in. Muzzle 
 
 •loading Guns. 
 
 
 
 
 Velocities. 
 
 
 
 
 
 Charge. 
 
 Shot. 
 
 Feet. 
 per second. 
 
 Energies, foot-tonB. 
 
 Thiekness 
 
 
 
 
 
 
 
 
 
 
 is 
 
 o 
 <- 
 
 
 ■ 
 
 
 
 
 
 ■3 
 
 3 
 
 CO s 
 
 of plate. 
 
 Remarks. 
 
 ■** 
 
 
 
 Description. 
 
 
 u> 
 
 2 
 
 s 
 
 n 
 
 
 
 <— 
 
 x 
 
 -c 
 
 o 
 
 
 A 
 
 
 
 
 
 
 a a 
 
 
 
 O 
 
 d 
 
 h0 
 
 is 
 
 
 a 
 
 GO 
 
 3 
 3 
 
 GO 
 
 © © 
 
 
 
 
 lb. 
 
 
 lb. 
 
 
 
 
 
 
 
 inohes. 
 
 
 38 
 
 46 
 
 f 
 
 V 
 
 
 1167 
 
 1152 
 
 2196 
 
 2139-4 
 
 88-6 
 
 12 
 
 Penetration 9'5-ins.; last 
 round repeated ; plate 
 broke in two ; shell 
 broke mostly in large 
 pieces ; bulge at back 
 ■83-ins. 
 
 39 
 
 28-26 
 
 H.L.Q 
 
 ■ 
 
 • 
 
 1002 
 
 994 
 
 1619 
 
 1593-0 
 
 63-8 
 
 (6 + 2) 
 
 8 
 
 Penetration 10-75-ins., 
 head in two pieces in 
 hole ; moulded off to 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 2-1 -ins. 
 
 to 
 
 29-76 
 
 • 
 
 182-6 
 
 • 
 
 1167 
 
 1137 
 
 1691 
 
 1636 
 
 66-6 
 
 (6+2) 
 8 
 
 Body of shell in large 
 pieces; small piece of 
 head through ; plate 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 moulded off to 2'7-ins. : 
 
 
 
 
 
 
 
 
 
 
 
 
 gape 6'25-ins. x 6'25-ins. 
 
 41 
 
 31-76 
 
 ■ 
 
 132-6 
 
 . 
 
 1381 
 
 1346 
 
 1752 
 
 1662 
 
 66'4 
 
 (6+2) 
 8 
 12 
 
 Penetration 9*75-ins. 
 
 12 
 
 68 
 
 pebble 
 
 182-6 
 
 Steel shell. 
 
 1682 
 
 1631 
 
 3496 
 
 3366 
 
 134-8 
 
 Through. Plate broke 
 
 
 
 
 
 1*6 diameter 
 
 
 
 
 
 
 
 into two pieces through 
 
 
 
 
 
 head. 
 
 
 
 
 
 
 
 hole. Shell unbroken. 
 Diameter at shoulder 
 before firing, 7'97-ins. j 
 after firing, 7'99-ins. 
 Diameter at rear band 
 before firing, 7 - 93-ins. ; 
 after firing, 7'932-ins. 
 
 
 
 
 
 
 
 
 
 
 
 
 13 
 
 88 
 
 • 
 
 ■ 
 
 ■ 
 
 1988 
 
 1951 
 
 5001 
 
 4817 
 
 192-9 
 
 (2+12) 
 
 14 
 
 Through. Broke up 
 plate. Shell struck a 
 plate in rear and broke 
 
 14 
 
 28'23 
 
 R.L.G 
 
 2326 
 
 Chilled. 
 
 1*6 diameter. 
 
 head. 
 
 1006 
 
 998 
 
 1631-6 
 
 1606-3 
 
 64'3 
 
 (6+2) 
 8 
 
 up. 
 Through. Shell broke 
 up ; head as shown, body 
 in fair-sized pieces ; 6-in. 
 plate hard and broke up 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 into three pieces. 
 
 46 
 
 88 
 
 pebble 
 
 182-6 
 
 Forced steel. 
 
 2 diameter 
 
 head. 
 
 1970 
 
 1934 
 
 4911-2 
 
 4732-0 
 
 189-7 
 
 14 
 
 Through. Shell broke 
 into seven pieces ; plate 
 broken through old shot- 
 hole into three pieces. 
 
 46 
 
 83 
 
 § 
 
 V 
 
 • 
 
 1874 
 
 1840 
 
 4444-2 
 
 4263-2 
 
 171-7 
 
 i 
 
 Through. Shell broke 
 into nine pieces ; gas 
 check entire ; plate 
 broken into three pieces. 
 
 47 
 
 83 
 
 • 
 
 f 
 
 do. Cammell 
 
 1*5 diameter 
 
 head. 
 
 1889 
 
 1865 
 
 4616-6 
 
 4363-3 
 
 174-6 
 
 . 
 
 Nearly through. Shell 
 broken, and plate broke 
 up into three pieceB. 
 
 48 
 
 90 
 
 * 
 
 
 do. Firth, 
 
 No. 63. 
 
 1*6 diameter 
 
 head. 
 
 1981 
 
 1928 
 
 4966'2 
 
 4702-7 
 
 188-6 
 
 e 
 
 Through. Shell through 
 plate and broke up in 
 sand, penetrating 10 ft. 
 into sand. 
 
 49 
 
 31-26 
 
 K.L.G 
 
 132-6 
 
 do. No. 49. 
 
 1*6 diameter 
 
 head. 
 
 1372 
 
 1336 
 
 1729-6 
 
 1639-4 
 
 65-7 
 
 (2+6) 
 8 
 
 Point l*27-in. through. 
 Shell entire, but cracked. 
 Penetration 9'27-ins. 
 
 60 
 
 29-76 
 
 r 
 
 182-6 
 
 do. No. 54. 
 1*5 diameter 
 
 1142 
 
 1123 
 
 1660-4 
 
 1695-6 
 
 64-0 
 
 (2+6) 
 
 8 
 
 Through easy. Shell 
 entire, striking target in 
 rear, and making im- 
 pression at head without 
 
 
 
 
 
 head. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 61 
 
 28-26 
 
 • 
 
 232-6 
 
 do. No. 67. 
 
 1*5 diameter 
 
 head. 
 
 986 
 
 979 
 
 1667-4 
 
 16447 
 
 61-9 
 
 ■ 
 
 point. 
 Through. Shell after 
 penetration hit No. 40 
 target in rear sideways 
 and broke up. 
 
 
 
 
 
 
 
 
 
 
 
 
 62 
 
 83 
 
 pebble 
 
 182-6 
 
 do. Cammell 
 2 diameter 
 
 1890 
 
 1839 
 
 4620-4 
 
 4278-6 
 
 171-6 
 
 6 
 
 Through. Shell unaltered, 
 passing through corner 
 of sand butt. 
 
 
 
 
 
 head. 
 
 
 
 
 
 
 
 63 
 
 114 
 
 ' 
 
 ■ 
 
 Chilled, 
 No. 69. 
 
 Shell 
 
 broke 
 
 up in 
 
 gun. 
 
 No re 
 
 oord. 
 
 Pressure in chamber 36' 4 
 tons per square inch. 
 
 Gape about 3-ins. Mould 
 off back of plates, 
 
 64 
 
 111 
 
 V 
 
 f 
 
 Cammell' s 
 
 2266 
 
 2211 
 
 6440-7 
 
 6190 
 
 2481 
 
 (solid.) 
 
 
 
 
 
 steel 
 
 
 
 
 
 
 1<H 
 
 
 
 
 
 hardened in 
 
 
 
 
 
 
 
 3*7-ins. ; over, 33-ine, by 
 
 
 ■ 
 
 
 
 ■oil. No. 60. 
 
 
 
 
 
 
 
 30-ins. Shell -2-ins. 
 beyond back of plate, 
 base being'6-in. from face 
 of plate. Shell un- 
 broken, but set op 
 1'8-in. of length. Pene- 
 tration 16-76-Idb. 
 
 
 
 
 
 
 
 
 
 
 
 
87 
 
 CHAPTER VII. 
 
 Experiments continued. — Hard and Soft Armour. 
 
 The series of experiments made by General Inglis' Sub-Committee 
 Has been given without interruption. The report was sent in 1880, and 
 in the meantime other experiments had been made, which may be 
 noticed. 
 
 Dutch Experiments with steel and chilled projectiles. 
 
 A report 1 printed by the Dutch Government on a trial against 
 armour of projectiles supplied by Ekman — of Pinspong — Krupp, and 
 Griison, took place at Scheveningen, on June 16th, 1879, has a special 
 interest from the fact that the projectiles were all fired with a reduced 
 velocity, calculated purposely to give bare penetration as nearly as 
 possible. In the English trials with Armstrong 8 and 9-inch guns the 
 
 FIG. 2 
 
 V-S-»^e-/oV'i° 
 
 1 Vide " original Dutch Eeport," dealt with in Engineer, Oct. 16th 1880, p. 278. See also 
 D. 6f A. " Proceedings," VoL.XVII, p. 38. 
 
 13 
 
88 
 
 same idea was carried out, beginning at about the same thickness of 
 plate, but increasing up to a very high standard. The Dutch 
 experiments, if they did not embrace an ambitious programme, were at 
 all events carefully carried oat, and are therefore instructive. The plate 
 fired at is shown in Figs. 1, 2, and 3. It consisted of a section of 
 the English Bellerophon target, namely, iron plates of 6-ins. and 8-ins. 
 above and below, (as shown in Figs. 1 and 2) on 10-in. of teak, with 
 1^-in. skin. For this, a penetrating figure of 60 foot-tons per inch 
 circumference was found to be sufficient for the upper portion, and 93 
 foot-tons per inch circumference enabled the projectile shown in Fig. 2 
 to get its point through the inner skin of the lower portion. The guns 
 employed were a cast iron and a steel gun, each of 24 centimetres 
 calibre — that is 9'45-ms. The range was 500 metres — 547 yards. 
 
 The projectiles supplied were as follows : — From Ekman, chilled iron 
 — (vide Fig. 4) of two sizes, namely, 2*6 calibres long, weighing 147" 75 
 kilogs. (325-73 lb.,) and 2-8 calibres long, weighing 152*34 kilogs. 
 (335*85 lbs.) These were chilled nearly to the centre in the head, and 
 the body was left unchilled — vide Fig. 4. From Griison, chilled iron 
 — vide Fig. 5 — of two sizes, namely, 2 - 6 calibres long, weighing 149"65 
 kilogs. (329*93 lbs.) and 2*8 calibres long, weighing 153*85 kilogs. 
 (339*18 lbs). These were chilled to a certain depth over both head and 
 
body, as shown in Fig. 5. From Krupp, steel, 2*8 calibres long, weigh- 
 ing 161 kilogs. (354-94 lbs). - - 
 
 Pica 
 
 PICS 
 
 More was expected from the steel projectiles than the chilled ones. 
 They were heavier, and were fired with a higher charge at the lower 
 portion of the target from the steel gun, which was polygrooved, the 
 projectiles being rifled by means of copper rings near shoulder and 
 base, as seen in projectile in Fig. 2, whereas the chilled projectile had 
 studs, being fired from the cast-iron gun at the upper portion of the 
 target. In all, seventeen rounds were fired, namely, four heavy and 
 one light of Ekman's, five heavy and three light of Gruson's, and four 
 of Krupp's. The table at p. 90 gives the details. 
 
 The following conclusions were arrived at as regards the calculated 
 penetration, namely, that a projectile discharged from the cast-iron gun, 
 with a velocity of 373 metres, or 1,224-ft., would pierce the 6-in. plate 
 at about 1,800 metres (1,969 yards) range; and a heavier projectile 
 from the steel gun, fired with a velocity of 472 metres (1,549-ft.) would 
 pierce an 8-inch target and backing at about 1,800 metres (1,969 
 yards). 
 
 As regards the projectiles, none of the Bkman (Finspong) chilled 
 iron projectiles were broken up on impact, except that the point of one 
 was broken off to the length of 142-mm. (5 - 6-in.) The remainder of 
 the projectiles underwent no alteration in form beyond setting up to 
 the extent of 6-mm. (0-236-in.) in length. Of the eight Griison chilled 
 iron projectiles, seven broke up into a great many fragments, while one, 
 which remained entire, showed a crack along the tapering portion 
 about the shoulder of it, but was set up in length only 2-mm. (0'079-in.) 
 The Krupp steel projectiles all remained entire. They underwent a 
 slight deformation, setting up longitudinally to the extent of from 
 
90 
 
 22-mm. to 28-mm. (0'866-in to M02-in.) while above the shoulder the 
 diameter increased to the extent of from l - 5-mm. to 2'5-mm. 
 (0'059-in to 0098-in.) ; also all the points for a length of 16-5-mm. to 
 21-5-mm. (0'64-in. to 084-in.) were deformed. The Swedish projectiles 
 of Bkman were considered very superior to those of Griison. The 
 Committee found it difficult from these experiments to say whether the 
 steel showed a superiority sufficiently decided to justify their high price 
 as compared with the chilled projectiles. 
 
 With regard to this report, it may be remarked that the behaviour of 
 the Finspong projectiles to a certain extent agreed with the results 
 obtained in English competitive trials — that is to say, the Griison's 
 projectiles were harder than the Finspong. 
 
 At Shoeburyness, in 1878, one Griison shot broke up badly, but 
 another did remarkably well {vide p. 74, Chap. VI.). The Finspong 
 metal is, of course, remarkably excellent ; but no observation is made 
 in this report as to its price. The chilling of the body, still kept up 
 by Griison, had been abandoned long ago by most manufacturers. In 
 the abstract, the unchilled metal having the advantage in tenacity 
 and the chilled metal in hardness, it would seem sound to retain a 
 little unchilled in the centre, to hold the shot together, while the entire 
 ring of chilled metal might prevent setting up. Practically it has 
 been found better to leave the entire body mottled. The crushing 
 strain there must fall much less heavily than in front, and tenacity 
 throughout the body seems the desideratum. 
 
 Details op thb 17 Hounds bibbd ok thb 16th Jr/irs, 1879, at SoHKTBHlHSBif. 
 
 
 
 
 
 Stored up 
 
 
 
 
 
 
 work in ton 
 
 
 No. of round and 
 
 Weight of 
 
 Weight of 
 
 Striking 
 
 metres per 
 
 
 maker. 
 
 charge. 
 
 projectile. 
 
 velocity. 
 
 cent, of 
 shot's cir- 
 cumference 
 
 Bemarks. 
 
 
 kilogs. 
 
 kilogs. 
 
 metres. 
 
 
 
 1. Ekman 
 
 16 
 
 147-76 
 
 266-9 
 
 7-116 
 
 Struck near top. 
 
 2. 
 
 n 
 
 162-34 
 
 266-4 
 
 7-311 
 
 Do. 
 
 3. „ 
 
 II 
 
 B 
 
 264-7 
 
 7-196 
 
 Struck junction of plates. 
 
 4. » 
 
 n 
 
 II 
 
 267-4 
 
 7-363 
 
 Fairly through. 
 
 6. Griison 
 
 18 
 
 149-66 
 
 290-3 
 
 8-627 
 
 Struck close to 4. 
 
 6. 
 
 fl 
 
 i, 
 
 290-3 
 
 8-527 
 
 Fairly through. 
 Struck top edge. 
 
 7. Ekman 
 
 n 
 
 162-34 
 
 285-9 
 
 8-416 
 
 8. Griison 
 
 // 
 
 163-85 
 
 287-3 
 
 8-684 
 
 Fairly through. 
 
 9. n 
 
 » 
 
 n 
 
 288-7 
 
 8-669 
 
 Through. 
 
 10. „ 
 
 16 
 
 IT 
 
 266-3 
 
 7-378 
 
 Struck near top. 
 
 11. „ 
 
 n 
 
 n 
 
 262-3 
 
 7-169 
 
 Just through (vide Fig. 1.) 
 
 12. ,- 
 
 " 
 
 rr 
 
 267-1 
 
 7-420 
 
 Through. 
 
 13. „ 
 
 n 
 
 149-65 
 
 271-9 
 
 7-482 
 
 Through weak place. 
 
 14. Krupp 
 
 22-3 
 
 161 
 
 3230 
 
 11-367 
 
 Through thick portion. 
 
 16. n 
 
 a 
 
 n 
 
 821-6 
 
 11-248 
 
 Barely through. 
 
 16. , 
 
 B 
 
 ii 
 
 322-6 
 
 11-326 
 
 Do. 
 
 17. n 
 
 
 
 
 ^~ 
 
 Not given. 
 
91 
 
 Spezia Trials, 1879. 
 
 In June, 1879, an important trial took place at Spezia against steel 
 armour, with the 100-ton Blswick M. L. gun of 17'72-inch calibre. 
 The general nature of the target may be seen in sectional plan, and in 
 elevation herewith. Steel plates 70 cm or 27 - 6 inches thick, were 
 
 IT V T T T T 
 Jl Jl jIl jIl Jl jIl 
 
 SECTIONAL PLAN 
 
 
 
 
 
 
 
 
 
 
 
 ' s . T SH0T 
 + + 
 
 
 
 2",° SHOT 
 + 
 
 + 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 . av, SHOT 
 
 + + + 
 
 
 
 + 
 
 + 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 FRONT CL'EVATION 
 
 enclosed in wrought-iron box frames, made of inch iron plate, as shown 
 in Figs, herewith. Each steel plate was about 9 ft. x 4 ft. 7 ins. x 2 ft. 
 3 - 6 ins. and weighed 20 - 8 tons. They were made at Terre Noire and 
 St. Chamond. An iron plate 1 inch thick covered the whole face of the 
 steel. The idea was to hold the steel in its place after fracture, so that 
 a subsequent shot might be resisted by the fragment in situ. Steel 
 projectiles were employed made by Whitworth and at Terre Noire, and 
 chilled iron from G-riison and San Vito. The weights varied from 
 about 1918 to 2099 lbs. 1 The Whitworth projectiles were of com- 
 pressed steel with points inserted like those employed in the Shoe- 
 buryness competitive trials. The Terre Noire projectiles had turned 
 bodies but rough heads. The Griisons had machined bodies and rough 
 heads. The San Vito were chiefly remarkable from their having pieces 
 of india rubber screwed to their upper sides to act as wads. Some 
 projectiles were also furnished by Armstrong & Co. from Elswick, but 
 as they were made for them by Whitworth they need hardly be 
 
 i One of the highest results in this series had striking velocity 1738 ft., 7f=2099 lbs. Con- 
 sequently the striking energy _E=4392(> foot-tons, e=792-7 foot-tons, and £=29'19 ins, 
 per ton of metal in plate 211T5 foot-tons. 
 
92 
 
 distinguished. The plates were overmatched. It is difficult to allow 
 for the iron casing on any principle. It is now recognised that a front 
 plate of iron rather facilitates perforation, and one of this thickness 
 at all events, could furnish little resisting power against racking on 
 this scale. Whether judged on the perforation or racking standard, 
 the shot outmatched the plate. The form in which destruction was 
 effected was interesting however. The first round, a San Vito chilled 
 iron shot. The shield kept it out but was much brokeu and distorted. 
 The general appearance of shot and plate are shown in Figs. 2 and 3. 
 
 Fig. 2 Fig. 3 
 
93 
 
 Bound 2 was a Whitworth steel which shattered the shield and actually 
 forced its way through, and was found entire (Fig. 4), the right hand 
 top portion of target being wrecked, the front iron casing being of 
 course perforated and thrown forward, vide Fig. 5. Round 3, 
 Armstrong steel shot, supplied to that firm by Whitworth, struck the 
 lower portion in place marked 3rd shot in Fig. showing front elevation, 
 and wrecked the whole structure, see Fig. 1. The projectile did not 
 pass completely through and was more set up than No. 2, vide Fig. 6. 
 
 Kg. 6 
 
 NOT LOOSE BUT 
 JOINT DEFORMED 
 8LCRUSHED 
 
 ■<3s 
 
 In short, the shot met with a more rigid or harder part, and more 
 was done in racking, and less in perforation than in the case of No. 2. 
 As a whole the experiment was disappointing. It showed the great 
 difficulty of holding steel shields together under blows of sufficient 
 energy to fracture them. In fact, an iron front plate assists the shot to 
 penetrate (vide p. 82) while the iron box frames, must clearly be very 
 massive to furnish efficient lateral support. 1 
 
 Krupp's Meppen Trials, 1879. 
 
 In August, 1879, Krupp carried out a notable series of experiments 
 at Meppen. The features to be here dealt with are only those con- 
 nected with armour. 2 3 Two experiments were directed against armour 
 
 1 The support afforded by box frames can only be superior to that given by adjacent steel 
 plates in the grip of the T frames, fastened round the edge of the plate, and holding over the 
 front edge, vide A in sectional plan, and C, combined with the hold of the iron covering plate. 
 Without the latter the support would be greatly decreased. 
 
 2 Vide Author's Report in " Engineer," August 15, 1879. The author was greatly indebted to 
 Herr Krupp, not only for the enjoyment of the princely hospitality displayed to all guests, in the 
 provision or board, lodgings, special trains, &e., but for what was even of more importance, namely, 
 the most ample means of examining everything, together with the supply of records, photographs, 
 and the use of telegraph wires. 
 
 3 In these trials were exhibited for the first thne publicly the greatly increased powers of new 
 type guns, and the high standard of accuracy of fire that Krupp had undoubtedly acheived, 
 partly owing to the goodness of German prismatic powder. There were also several new and 
 striking designs. Thus, Krupp's steel breech-loaders were exhibited under peculiarly favourable 
 circumstances, and an impression was produced throughout Europe. There can be little doubt 
 that the adoption of breech-loading guns, and of steel in England, was hastened by these experi- 
 ments, although in neither respect was Krupp's system followed. All this however should not 
 cause the fact to be overlooked 1 hat Colonel Inglis' Sub-Committee had for some time been trying 
 Elswick guns, embodying the features that constituted the so-called new type guns, features that 
 had been arrived at in England, by researches in which Captain Andrew Noble specially took 
 a leading_part_(ui<& also Abel and Noble's experiments). 
 
94 
 
 plates, and two concerned armoured structures. In order of time the 
 two latter came first and were as follows : — 
 
 On August 5th, 1879, was fired an 11-inch (28 cm ) rifled howitzer at 
 28° 36' elevation, at a canvas structure made to represent the deck of 
 the Inflexible, head up the range, length 328 ft. (100 metres) breadth 
 82 feet (25 metres). The actual length and breadth of the Inflexible 
 are respectively 320 and 75 feet. The range was 7327 yds. The 
 shell weighed 476-1 lbs. (216 *). The firing charge was 39-7 lb. (18 k «). 
 The vessel was struck as shown in diagram 2, herewith, that is to say, 
 
 DISTANOEW 
 
 ')■ 
 
 80 80 70 60 50 40 30 zo 
 
 6810 
 
 6600 
 
 6790 
 
 67 8 
 
 6770 
 
 B76O 
 
 67SO 
 
 67+0 
 6730 
 6720 
 
 
 
 
 
 
 
 
 
 
 i 
 
 5 
 
 
 
 
 
 
 
 
 
 5 
 
 • 
 
 
 
 
 
 
 
 
 
 ■*•! 
 
 3 
 
 \ 
 
 
 
 
 
 
 
 1 
 
 \ 
 
 
 
 
 
 
 
 
 i 
 
 N 
 
 
 
 
 
 
 
 
 1 
 
 1 
 
 fi 
 
 
 
 
 
 
 
 
 
 
 T 
 
 X 
 
 111 
 a j 
 
 
 
 
 
 
 
 
 1 
 1 
 1 
 
 z 
 
 » 
 
 
 
 
 
 
 
 \ 
 
 
 
 
 
 
 
 
 
 A 
 
 
 7 
 
 
 
 
 
 
 
 
 / 
 
 
 " 
 
 
 
 
 
 V 
 
 ' 
 
 
 
 
 
 
 
 OIAORAM.2. 
 
 5 shells pitched on the spot representing the deck of the ship, the four 
 last fired falling very near the centre of it. The importance of this 
 experiment is dwelt on hereafter. 
 
 On August 6th, 1879, 1 was fired a 6'1-inch (15-5 cm ) gun, weight, 
 64 - 9 cwt., muzzle pivoting by ball on muzzle in socket in iron shield, 
 (vide Fig. 6, p. 95). The shock of discharge was transmitted through 
 the ball and socket into the shield. However certainly this may 
 entail the gradual destruction of the structure ; with this particular 
 gun the process is so very slow that the gun ought to have done a 
 great deal of work before it becomes unserviceable. The speed and 
 accuracy of fire are remarkable. The No. 1 sits on a saddle on the chase 
 of the gun, and lays and fires from that position as fast as the piece is 
 
 1 Vide " Engineer,' - August 15th, 1879, p. 122, also Author's paper for V. S. Institution, 
 entitled " Lessons to be learned from the Meppon Trials, 1880," vide " Engineer," November 6th 
 and 19th, 1880. This design does not Beem to have been adopted, but in certain positions it 
 appears as if it might be most valuable, for example, in the salients or forts $ arret of important 
 fortresses, where it would only be exposed to the fire of medium guns, when its speed, safety, and 
 accuracy of fixe would make it specially valuable. 
 
i>5 
 
 loaded. Owing to the piece having no recoil, it is only necessary to 
 correct the laying of the piece, or leave it laid as in the previous 
 round. The rate of firing was about 1 round in 25 seconds. The 
 projectile weighed 77 lbs., the charge was 14 - 3 lbs. The plates 
 receiving the shock at the salient where they meet, the tendency is to 
 press them together, and there is little or no perceptible spring on 
 
 firing. Both muzzle of gun and disc of plate containing the socket 
 are removable. A shield can be run up, covering the salient, at any 
 
 14 
 
96 
 
 time, which renders the operation of changing muzzle ball and socket 
 of shield a safe one on service, and also an easy one, unless it should 
 happen that the whole structure is deformed by a crushing blow. 
 
 On August 7th, 1879, Krupp fired at two shields, 1 shown in Figs. 
 21 and 22, herewith, made to illustrate the behaviour, under fire, of a 
 
 " FIC.2S 
 
 chilled iron shield ot Griison's, and one of wrought-iron designed by 
 Krupp. Three rounds of chilled iron shell were fired at each, and 
 
 ■ Vide " Engineer," August 16th, 1879, p. 123. 
 
97 
 
 four of steel shell from a 6-]-lnch (15-5 cra ) gun, with the results shown 
 in Pigs. 21 and 22. Under the impact of chilled shell, the chilled iron 
 shield suffered little, but the fire, continued with steel projectiles, cracked 
 it through. Griison naturally repudiated all responsibility for the 
 behaviour of a chilled iron shield which Krupp designed and made, 
 with a view of showing its inferiority, but on the whole the shield 
 behaved well, better probably than Krupp expected. The chief lessons 
 to be learned from the experiment was, the unsuitability of chilled 
 shot for the attack of chilled iron, in other words ; the power of chilled 
 iron to resist 1 chilled shut, but its liability to fracture when steel ones 
 were employed. Finally, two rounds were fired from a 9"45-inch gun 2 
 (24 cm ) at a shield made of wrought-iron plates from Dillingen, vide 
 Figs. 23 and 24, herewith. There was a front iron plate 1 2 inches thick, 
 30 - 5 cm , then a thin layer of wood nearly 2 inches thick (5 - cm ), and 
 behind it an iron plate 8 inches (20"5 cm ) thick, for area (vide Fig. 23). 
 The projectile was steel, it weighed about 348 lbs. (158 k s) the firing 
 charge was 165'3 lbs. (75 kg ), the striking velocities were 1877 feet 
 (572 metres) and 1852 feet (564 - 5 metres). This gives, in the case of 
 
 FIC.33 
 
 
 tHH 
 
 lli!lIi!|l!!l!ijl|i!!i!|!|l»R 
 
 ...„ .....Mil lllllllllllllllllllllllllllllllllll^ ' IHIlt'ilitllllllll n ' n 
 
 MET ' ^ 
 
 F1C.S4" 
 
 1 Krupp hardly showed his usual amount of shrewdness in this trial. The 6'1-inch gun was a 
 punching gun, with high power of perforation in proportion to its energy, and therefore suited to cut 
 deep into wrought-iron, as it did, rather than shatter the chilled shield. To English Officers, hardly 
 any of whom have seen chilled iron shields under fire, the experiment was very instructive. 
 
 * Vide " Engineer/' August 16th, 1879, p. 123. 
 
98 
 
 the second round, a striking energy #=8287 foot-tons, energy per 
 inch circumference *=279-2 foot-tons, *=]7'52 ins., in a single plate, 
 and in two thicknesses t= 18-3 ins. As may be seen in the Figures 
 the projectile passed easily through the entire shield in both rounds, 
 and passed up the range over 3000 yards. These two projectiles were 
 as little deformed and injured by the work they had done as those of 
 Whitworth, under similar circumstances. To Officers who had not 
 witnessed an exhibition of the powers of new type guns, and had not 
 Been the behaviour of excellent forged steel projectiles, it almost 
 looked like a conjuring trick to see projectiles cleave their way clean 
 through more than double their diameter and passed on up the range 
 uninjured, to the eye, almost as if they had been through a wooden 
 target. Those who had had to deal with guns and projectiles of the 
 most recent kind would scarcely expect the shot to perforate the 
 target as shown by the figures above, and some explanation must be 
 sought for a projectile whose perforation works out 18 - 3 inches in 2 
 plates, passing easily through 20, and continuing its flight without 
 grazing for some hundred yards. The problem may be divided into 
 the effect that is due to the gun, supposing the projectile to be of an 
 average kind, and that is due to special peculiarity in shot or plate. 
 The perforation due to the gun's powers is a simple question of energy 
 in proportion to the bore, and, with the data established by past 
 experience, may be found by taking the velocity without firing at a 
 plate at all. The actual effect on a plate tests the qualities of projectile 
 and plate. So far as the type and powers of gun and powder go, the 
 result is 183 inches through iron arranged in 2 plates. The extra 
 effect must be traced to excellence in shot or badness in plate. The 
 latter undoubtedly may be seen to have behaved rather badly. The 
 effect, though not well rendered in the cut, is obviously that of shot or 
 badly welded plates. Some fault has been found with the bolting and 
 with the thinness of the layer of wood between the two plates. As 
 already pointed out (p. 9, Chap. I), plates may crack across from the 
 insufficiency of wood between them as these plates did, but in the actual 
 case of perforation, another cause, namely the bad welding, is more 
 likely concerned, but does not seem sufficient to go nearly far enough 
 to account for the great excess of perforating power in question. The 
 projectiles were of most excellent steel undoubtedly, and the points 
 were very sharp, being ogival struck with a radius of 2 diameters. 
 The data for English's system of calculation have been furnished with 
 points of \\ diameters radius, and this no doubt enters into the question. 
 It may not at first appear necessary to analyze an isolated experiment on 
 perforation as has been done. It is however very important to gauge 
 foreign results, especially those obtained by Krupp, in all possible ways. 
 The powder due to the gun which is registered by the velocity is a 
 matter easily dealt with. The relative powers of any gun of our own and 
 one made in Germany are patent, if the velocity is known. Any power 
 due to special excellence in the projectile is much more likely to escape 
 observation, and in the branch of development it is easy to drop 
 behind without perceiving it. (This question is dealt with again 
 hereafter). Those special qualities of a shot are of course more 
 
99 
 
 important which apply both to punching and raoking ; thus, effect due 
 to tenacity is probably more important than that due to shape of head. 
 
 Nettle Steel-faced Plate Trials. 
 
 In June, July and August, 1879, a steel-faced 9-inch plate for 
 H. M. S. Inflexible, was tried on board the Nettle, by a 9-inch 12-ton 
 gun in the usual way, but greater penetration than usual was obtained 
 by the chilled shot, which of course broke up. Some details are given 
 in the Director of Artillery's Departmental records, but it is difficult 
 to judge of the effect without sketches or photographs. 1 
 
 On April 20th, 2 1880, an experiment took place with a 9-inch 
 Cammell- Wilson steel-faced plate, in the course of proof of plates, on 
 board the Nettle, at Portsmouth, which may serve to show the progress 
 made. The service 9-inch gun fired a 257 lb. Palliser projectile with 
 a striking velocity of about 1460 feet per second, giving a total energy 
 2?=3799 foot-tons, an energy per inch e=135 - 6, and perforation /=12 
 inches, the energy portion of plate was 1039 foot-tons, supposing the 
 plate to weigh 5 tons. The plate had a steel-face of 3^ inches thick, 
 and a wrought-iron foundation plate 5£ inches thick. The area of it was 
 6 ft. x 5 ft. 8 ins., it was bent into the arc of a circle of 16 ft. 11 ins. 
 radius, measuring to the steel face. The cuts, herewith smd p. 100, 
 
 PRONT OF PLATE 
 
 i Vide D. of A. " Proceedings," 1879, Vol. XVII., ppi 151, 224. 
 : Fide "Engineer," June 11th, 1880, p. 417. 
 
100 
 
 BACK OF PLATE 
 
 show the effect produced by 3 rounds, the depth of indent of the first 
 shot which was dislodged by those subsequently fired was 5 - 8 ins. 
 The plate in this case stood admirably, having borne 3 rounds, each of 
 which would have perforated an iron plate of 12 inches thickness, and 
 still being in a condition to offer considerable resistance to further firing. 
 
 On July 21st, 1880, 1 an 18-inch steel-faced Camtnell-Wilson plate 
 was fired at by the 38-ton gun at Shoeburyness, area of plate 9^x8 
 ft., projectile, Palliser chilled, weight 828 lbs., firing charge 180 lbs. 
 P 3 powder, range 227 ft., striking velocity 1504 ft., striking energy 
 i?= 12980 foot-tons, energy per inch circumference, e = 322 - 5, perfora- 
 tion i!=18"7 ins., weight of plate about 24 tons, energy per ton of 
 plate 541 foot-tons. The effect on the plate is shown in elevation and 
 vertical section, p. 101. It will be seen that the shot broke up badly 
 against the steel face, a circle of deep indentations 1, 2, 3, 4, 5, 6, 7, 
 and 8 being made on the plate round the point of impact, two rather 
 deep horizontal cracks may be seen in elevation herewith. Obviously 
 the shot was prevented from delivering its energy well at the point of 
 
 1 Vide "Engineer," July 30th, 1880, p. 79. 
 
101 
 
 impact by being broken up abruptly by the hard surface of the shell. 
 This is an example of loss of effect from want of tenacity in chilled 
 iron shot. 
 
 Krupp's Meppen Trials, 1882. 
 
 On March 29th and 30th, 1 1882, Krupp, during the course of gun- 
 nery experiments at Meppen fired long 5'9-inch (15 cm ) gun at a target 
 consisting of two 2 6'9-inch wrought-iron plates, with 9 - 8 inches of 
 
 i "Engineer," October 6th, 1882, p. 251. 
 2 More exactly n-b™ (6'89 ins.) wood 25 om (9-f 
 p. 380. 
 
 p. sou. 
 p. oou. 
 p. oou. 
 p. sou. 
 p. sou. 
 
 ina. D. of A. " Proceedings." Vol. XX., 
 
102 
 
 wood, {vide Fig. 2, herewith), at 1 64 yards (150 m ) from the muzzle. The 
 projectile was a steel shell weighing 109-6 lbs. (49'7 k e). The charge 
 in the second round, which being the weaker, is the one to consider, 
 was 37-5 lbs. (17 k e). The striking velocity was 1750 feet (533-5 m ). 
 The striking energy i?=2328 foot-tons, e=126'3 foot-tons, the thick- 
 ness that might be perforated t=l\'7 ins., this projectile, as well as 
 the first fired, passed easily through the entire target, and went on up 
 the range, this one to a distance of 328 yards, where it was picked up 
 uninjured, {vide Figs. 2, 3, 4 and 5, pp. 102 and 103). As in the plate 
 firing in 1879, the perforation was more easily performed than can be 
 
 HORIZONTAL SECTION 
 
 rfh fig. a 
 
 FIC.3 FRONT VIEW 
 
 o 
 
 o 
 
 o 
 
 rs ji 
 
 O 
 
 -i? 
 
 $T 
 
 ^ 
 
 FI0.+ BjtCKVIEW 
 
 o 
 
108 
 
 accounted for by calculation of the power of the normal English shot 
 from this gun with the given velocity, and this must be attributed to 
 excellence in shot, badness in target, or to both combined. The 
 material of the plates was exactly that of our own, being Sheffield 
 plate supplied to the German Navy. The wood between them, 
 however, was in this case much too thick, judging from the Shoebury 
 trials (see p. 9, Chap. I.). It may seem unfair to object to the wood 
 employed in 1879 as too thin and that of 1882 as too thick, but having 
 definitely arrived at the conclusion that about 5 inches is the proper 
 thickness for a layer of wood, and having given definite mechanical 
 reasons for it, it is not captious to object to 10 inches as too thick, 
 2 inches as too thin. Nevertheless it is probable that the excellence 
 of the shot and the sharpness of the point contributed to the good 
 result. Two rounds from the same gun were also fired at an angle 
 of 55° (35° with the normal) at a wrought-iron plate 7'9 ins. (20 cm ) 
 thick, charges of 39-7 and 37'5 lbs. (18 and 17 kg ) were employed. 
 Even with the latter complete perforation was effected (vide Figs. 6 and 
 
 15 
 
104 
 
 7, pp. 103 and 104), the projectiles however were both broken up, as 
 commonly happens in oblique penetration. With the smaller charge 
 the striking velocity as in the direct firing, was 1750 ft., with which, 
 the striking energy .#=2328, energy per inch e=126"3 foot-tons, 
 thickness that might be perforated 2=11*65. This multiplied by the 
 sin 55° gives 9'5 ins., as the thickness that might be perforated at 
 this angle of incidence. The shell therefore was more than a match 
 for the plate. The angle of incidence is that which Col. liiglis' 
 
105 
 
 Committee found to be about the limit at which, a projectile would bite 
 even when more than a match with the plate. There were no further 
 plate experiments at Meppen at this time, it may, therefore be seen 
 that Krupp confined his experiments to the perforation of soft armour. 
 
 Palliser improved Projectiles. 
 On June 6th, 1882, 1 some special projectiles designed by the late 
 Sir W. Palliser to act against steel-faced plates, were tried at Shoe- 
 buryness under the direction of Captain Palliser, brother to the late 
 Sir William. This form of Palliser shot has a steel jacket or jackets 
 shrunk on to its body to hold it together at the instant of impact, 
 (see Pigs. 1, 2 and 3, herewith). It has ribs on its head running longi- 
 
 — £6 "3 >\ 
 
 tudinally, see Figs. 1, 3, 4 and A, with a view to split the armour. 
 Lastly, its jacket being left behind in penetration, the diameter of the 
 shot is reduced, and in perforation meets with a proportionally 
 decreased resistance. It may be seen that the body is slightly tapered 
 
 > See " Engineer, June 9th, 1882, pp. 411 and 421, 
 
106 
 
 to facilitate escape from the grip of the jacket. The shot in Fig. 1, 
 has its body enclosed in a second jacket, which remains on it. It may 
 be seen that in this design perforation is contemplated for in the case 
 
 FIC.4- 
 
 FIC A 
 
 of a plate which cannot be perforated by any means, the reduced 
 diameter is of no advantage. On the occasion referred to, a 13-pr. 
 gun (3-inch bore) the shot weighed about 13 lbs., the charge was 
 3£ lbs., the velocity was about 1550 feet, the striking energy E= 
 216"5 foot-tons, the energy per inch circumference e=26"7 foot-tons, 
 and the thickness of iron that might be perforated i!=4 , 9 inches, with 
 a service projectile of 2 '985 in. diameter, with Palliser' s reduced 
 diameter of 2 - 585, it would be 5'3 inches. The plate against which 
 the shot were fired was a Cammell-Wilson steel-faced plate 4 inches 
 thick. According to the conclusions of Colonel Inglis' Sub-Committee 
 this is equal to 5'05 inches wrought-iron, consequently, supposing the 
 Palliser projectile to escape from its steel jacket without loss of energy, 
 and supposing the plate to yield by perforation rather than fracture, 
 the Palliser improved shot ought just to get through. It will be seen 
 in Figs. 1 and 3, p. 107, that three projectiles fired passed completely 
 through the plate with some energy to spare, indenting an old plate 
 in rear. The shot themselves broke up. In the first round, No. 2355, 
 the projectile had a double steel jacket, and left its outer jacket behind 
 it as shown in Fig 1 , very neatly. It was impossible to say whether the 
 body of the shot remained entire before striking the old target in rear 
 or not. The second shot (round 2356) had a single jacket, and the third 
 (No. 2357) a short steel ring. A wooden gun carriage bracket was 
 placed behind the plate to prevent the shot if possible from fracturing 
 themselves, or, should they be fractured by the compound plate, catch 
 the broken pieces. It appeared that No. 2 was fractured, and No. 3 
 was broken into smaller pieces. A service chilled projectile, fired 
 afterwards at the same target, broke up without perforating the plate, 
 making a moderate bulge in the back. This trial was singularly 
 successful, therefore, and was naturally followed up by further experi- 
 ments by Capt. Palliser. The ribs appeared by the marks in the 
 perforations to have done well. The steel jacket seemed to act more 
 effectually as it increased in size. There appears, however, to be a 
 limited scope for the gain effected by decrease of diameter, as its value 
 
107 
 
 HORIZONTAL PROJECTION SHOWING OPENING AT BACK 
 
 PLATE 
 
 depends on the power of forcing the plate to yield by perforation ; this 
 cannot apparently be achieved with the hardest classes of armour, 
 namely, solid steel and chilled iron. The increase of tenacity afforded 
 by the steel jacket applies to all forms of destruction of armour. Pig. 
 4 shows the first form of projectile tried on this system by Sir W. 
 Palliser, namely, a chilled iron Minie bullet, with steel ring. 
 
 (The attack of Alexandria by the British Fleet is discussed hereafter) . 
 
108 
 
 Cammell and Brown's Plates, 1882. 
 The progress in the manufacture of compound (steel-faced) plates 
 may be illustrated by the trials made against plates by the two great 
 Sheffield firms which had taken up this work, namely Cammell's and 
 Brown's. The former 1 took place on board the Nettle, in the course 
 of proof of armour plates, the latter in experimental firing at Shoe- 
 buryness. 
 
 The Admiralty conditions of test were the following: — "A test 
 
 Fig. 1 
 
 „ _4,iKi- - ..Jrv. 5 V 
 
 L^y^.t'„.v.i- aaS ii 
 
 
 Sgrr 
 
 i Ftefc ' Engineer,' August 11th, 1882, p. 97. 
 
109 
 
 piece, not less than 8 feet x 6 feet, nor less than 9 ins. in thickness, 
 shall receive a 9-inch chilled cast-iron projectile fired from the service 
 12-ton gun with 50 lbs. of powder, at a range of ten yards without 
 cracking through, and, provided it be surrounded by a frame, no one 
 of three such projectiles shall be capable of getting through, there 
 being a fair distance between each two shots, say 2 feet, as in the trials of 
 
 
 
 
 
 Fig. 2 
 
 
 
 
 
 
 
 
 1 
 1 
 
 EHr 
 
 
 ".>■ j. .. ..... ..■...;.. 
 
 
 Hi 
 
 mm 
 
 -Hi 
 
 lliiM^^^ 
 
 'ffiyj.Jr- 
 
 '■ ^J^sS^^kSI 
 
 III 
 
 
 ■;: mi 
 
 Blllli 
 
 illy 
 
 ..." ,:;.. 
 
 n 
 
 ■ ■■'■'■ 
 
 
 
 :'i 
 
 ■H 
 
 
 
 &fi^^^^^^^&tah!^^wi*l^^^^^P- -^Kaf 
 
 Mtsr- &i&m 
 
 
 -mm 
 
 ■ v 
 
 ■ it 
 
 
 r 
 
 y 
 
 
 Stitr&^hSjk 
 
 
 ■ " 
 
 , M '■ . ■ 1 
 
 A." 
 
 
 
 
 
 Ss&ssiStSS/ 
 
 
 , ::■_..;. . :. v ... ...... v. 
 
 ■».-'.'■";', 
 
 1 .-.BB^^^^^^ :! ^"->"-jMBMfe 
 
 Jwjal * 
 
 tt.; 
 
 '■^'Ci^iH.- ''•'■ '■•'"' -~~~- ■ . - ji&'fC %Lh'' 
 
 *§&$$$%&$*& 
 
 
 
 , 
 
 .?'■"." «r'"i2S' '.>*£ 
 
 the Inflexible turret armour." _ This test is a very severe one; a 
 proieotile thus fired has a velocity of about 1420 feet, and a striking 
 
no 
 
 energy of about 3608 foot-tons, 1 being about a match for 11^ inches of 
 unbacked wrought-iron. Since Colonel InghV Committee considered 
 that a good compound plate, between 9 and 10 inches, was equivalent 
 to a good 12-inch wrought-iron plate for a single blow, but not for 
 repeated blows, the fact that this plate bore all the firing shown in 
 Figs. 1 and 2, pp. 108 and 109, without cracking at the back, shows 
 that the plate was excellent, and also that the shot must have broken 
 up badly. Steel 9-inch projectiles got their heads through 12-inch 
 unbacked plates in the competitive projectile trial at Shoeburyness, 
 some chilled shot got their points through. The resistance of this 
 9-inch compound plate was therefore remarkable. The following rounds 
 were fired at it : — three rounds with 9-inch projectiles producing only 
 slight bulges at the back; four rounds with 10-inch Palliser projectiles 
 weighing 400 lbs., with 70 lbs. firing charge, at 10 yards distance, 
 striking velocity = 1364 feet, 2 having a striking energy E=o160 foot- 
 tons, and a calculated perforation t=l'3'l inches of unbacked iron, or 
 1 0"4 of compound plate. After the whole seven rounds the plate was 
 only bulged at the back without cracks, the greatest bulge being about 
 l^e inches high, while the deepest indent was 4- 7 inches. The total 
 work borne by the plate was therefore about 31,464 foot-tons (nearly 
 the blow of the 100-ton M. L. gun shot) and the enormous strain of 
 3314 foot-tons per ton of metal, supposing the plate to weigh about 
 9\ tons. 
 
 The steel-faced plate of Messrs. Brown, above-mentioned, tested at 
 Shoeburyness, was made on Ellis' patent 3 (vide manufacture, hereafter) ; 
 it measured 10 feet x 5£ feet x 11 ins. thick, made up of about 4 
 ins. steel or 7 ins. iron. The steel contained about 075 per cent. 
 carbon. 4 The plate was backed by about 24 inches of oak, and was 
 first attacked by the 9-inch gun (12-ton), firing Palliser chilled 
 projectiles weighing 200 lbs., the powder charge being 50 lbs. of 
 pebble. Three rounds were fired with striking velocities of 1430, 
 1444, and 1432 feet, and striking energies, E, equal to 3686, 3760 
 and 3697 foot-tons respectively, energies per inch (e) of 131 - 5, 1342 
 and 131-9 respectively, and perforation (t) of 1T7, 1T8 and 11-7 
 respectively. A fourth round was fired with a studless Cammell 
 steel projectile weighing 279 lbs., a striking velocity of 1405 feet, 
 and a striking energy of 3818 foot-tons, energy per inch (e) of 136"2 
 foot-tons, and perforation tf= 11-88 inches of iron, equivalent according 
 to Colonel InghV Committee to 9-4 of compound plate. 
 
 Pig. 1 p. 1 11, shows the plate at the conclusion of this firing, it had 
 then borne a total amount of energy of 14,961 foot-tons. In October, 
 1882, the same plate was re-erected, backed by 12 inches of wood, and 
 behind that an old 12 -inch compound plate. In the meantime a star- 
 shaped crack had developed in the face near the opposite end to that 
 
 1 JT=258 lbs., JJ=3608 foot-tons, «=1287 foot-tons, <=ll-6 ins. 
 3 0=166-6 foot-tons. 
 
 » Vide Paper by Major O'Callaghan, E. A., in E. A. Institution " Proceedings," May, 1882, 
 Vol. XII, also " Engineer," September 8th, 1882, p. 182. 
 4 Both ^ the face plate and the metal by which it is attached to the iron. 
 
Ill 
 
 fired at, showing how much more extensive was the molecular action 
 developed by impact than may appear. Two rounds were then fired 
 from the 38-ton gun (12-5-inch) at a range of 93 yards, with steel 
 projectiles weighing 840 lbs. and 845 lbs., fired with charges of 160 P a 
 powder, giving striking velocities of 1425 and 1413 feet respectively, 
 
 and striking energies of about 11,820 and 11,690 foot-tons, the former 
 having an energy per inch (e) of 302*9 foot-tons, and a perforation 
 t= 19-88 ins. iron, or 15*4 compound plate. The total energy of all 
 the six blows on the plate then amounted to 38,471 foot-tons, or taking 
 the plate at 11 tons, 3497*4 foot-tons per ton of plate. 
 
 This plate may naturally be compared with the Oammell-Wilson 
 plate, whose trial on board the Nettle has just been noticed, or a Brown- 
 Ellis plate, which was tested on board the Nettle in February and July, 
 1882, might have been taken in preference to this one. It was an 
 admirable plate, and appeared to behave in the same way as its rival, 
 but as it was tested with only three 9 -inch and three 10-inch projectiles, 
 and the Cammell plate with three 9-inch and four 10-inch, the total 
 work delivered on the plates differed considerably. It may be seen 
 that the Brown plate at Shoeburyness suffered much more than the 
 Cammell on the Nettle, discs of the plate being driven out by the 12*5 
 projectiles (vide Figs. 3, 4 and 5, pp. 112 and 113), but this is accounted 
 for by the plate being more severely tried by the greater concentration 
 of the energy into individual blows. The Brown-Ellis plate had borne 
 3497 foot-tons per ton of plate, the Oammell-Wilson 3314 foot-tons. 
 As to racking or shattering strain, the difference in amount then was 
 not such as to tell so heavily as in the manner of application. An 1 1-inch 
 
 16 
 
112 
 
 plate exposed to the blow of a shot with 15"4 calculated perforation 
 is much more severely tried than a 9-inch plate exposed to one with 
 104 perforation ; and it will be seen that the Brown-Ellis plate 
 actually yielded locally. Were this however the whole difference in 
 
 FI0.3 
 
 the strain to which the plates were subjected, it would probably be felt 
 that the Cam m ell-Wilson plate had stood the best, there is however 
 the more important fact that the three last projectiles fired at the 
 Brown-Ellis plate were steel. Probably any one comparing the two 
 Nettle plates as well as this one at Shoeburyness would fail to see any 
 advantage obtained by one system of compound armour over the other. 
 In fact, the comparison of two results when the conditions are not 
 strictly similar, is more useful in calling attention to peculiar features, 
 
113 
 
 than in establishing data to measure the relative powers of the two 
 
 systems. Leaving the question of comparison to trace further features 
 in the behaviour of the plates it may be well to quote the excellent 
 
 paper of Major O'Callaghan who conducted experiments under the 
 Commandant at Shoeburyness at this time. He writes, on the Brown- 
 Ellis plate, as follows : — " A large layer of steel was found to have 
 come away from the front in the region of the indents made by the 
 9-inch projectiles, disclosing a very curious configuration of the steel 
 round these depressions. Fig. 5 shows the crater like form they 
 exhibit, and Fig. 6 is a cross section through A B, showing their depth 
 
 FIG. 6. 
 SECTION AT A. B. 
 
 and the curved form of their sides. This strange phenomenon perhaps 
 throws some light on an appearance which has given rise to much 
 speculation in former rounds fired at steel-faced armour. It has 
 
114 
 
 always been observed that a wedge-shaped layer of steel is apparently 
 separated radially round the indents, and, in early experiments, it was 
 thought that a thin part of the steel face, the absolute uniformity of 
 which cannot always be guaranteed, had been struck. The frequency 
 of the appearance, however, negatived the theory, and the denuding 
 of this plate seems to show what probably happened in other instances ; 
 the wedge-shaped layer of steel being in fact, a portion of the surface 
 or covering of the craters now removed. It is wonderful to contemplate 
 the intricate molecular movement which must have been going on 
 under the apparently undisturbed surface of the plate, and which was 
 merely indicated by the radial cracks which were developed, some at 
 the time, and some considerably after the shot had struck ; difficult, too, 
 to realise the tremendous tension to which this surface was subjected, 
 tightly stretched over the distorted metal beneath it. Yet, in spite of all 
 this the plate had still cohesive power left to withstand successfully the 
 shock of the first 38-ton projectile." Major O'Callaghan suggests 
 the following explanation of the phenomena exhibited : — " When a 
 shot strikes and is arrested by a plate, the metal its point or head 
 displaces must go somewhere. In a wrought-iron plate, it is (we 
 believe) driven forward in a cone in front of the shot, the ' bulge ' at 
 back presenting the well known features of a pronounced swelling, 
 which, if the point has nearly penetrated, is cleft by star-shaped 
 fissures. In addition to this, there is generally a high lip thrown up 
 round the entrance of the hole, (see Fig. VII) . 
 
 " In a steel or steel-faced plate the appearances are very different, 
 there is no lip, and the bulge in rear is scarcely perceptible (Fig. VIII.) 
 
 FIC.VII 
 
 or, at all events much less pronounced, henoe it is evident that the 
 metal round the point of impact is more impeded in its efforts to escape 
 when displaced. Now it is clear that the displaced metal must be 
 thrust away normally to the curvature of the head of the shot, and 
 the direction taken by the molecules is represented by the divergency 
 rays m the figure at all angles between vertical and horizontal. A 
 bursting strain or thrust will be set up somewhere between these two 
 limits, and along this line there will be a tendency on the part of the 
 metal to separate, in other words there will be a line of cleavage, 
 Fig. IX. & ' 
 
 "We should therefore be prepared to find a portion separated from 
 the rest in the form of a cone bounded by straight lines. But it is not 
 
115 
 
 so. The cone is as we have seen curved in section. How is this to be 
 accounted for ? 
 
 " When any metal is subjected to a crushing strain beyond its power 
 of resistance, it evinces a tendency to buckle or bulge outwards — this 
 is I take it, the key to the rounded aspect of the exterior of what may 
 be termed the ' craters/ The steel displaced by the advancing point 
 is crushed and yields upwards, or buckles into the cavity caused by its 
 separation, or partial separation from the rest of the plate ; thus 
 presenting the appearance of the rounded indented cumulus, before 
 described." 
 
 FIC.2. 
 
 C1RCUMFERENG 
 39.3 
 
 33.3 S» s> 
 
 General Inglis considers that a true cone is first formed (Fig. X.), 
 and that afterwards the shape of the shot's head may affect the 
 
116 
 
 modification in form taken . Fig. 2, p. 1 1 5, shows the posterior fragment 
 of the steel projectile fired at this plate. 
 
 The behaviour of compound and steel plates of the best description 
 was tested in the close of this year by more remarkable trials, both 
 because they were constructed on a much larger scale, and also because 
 they were more strictly comparative. These took place at Spezia and 
 also at St. Petersburg, and may be best dealt as a fresh subject. 
 
117 
 
 CHAPTER YIII. 
 
 Experiments continued. — Hard and Soft Armour. 
 
 The Spezia Armour-plate Experiments} 
 
 The Italian authorities having adopted for the barbette towers of the 
 Italia and Lepanto plates of a thickness of 48 cm (18"9 in.), instituted a 
 trial of such plates, which took place in 1882, by means of the 100-ton 
 muzzle -loading gun supplied by Sir W. Armstrong and Co. to the 
 Duilio, the firing charge of this gun being sufficiently reduced for this 
 purpose. The trial of the plates was partly competitive, for although 
 the armour for the Italia had been already ordered from Messrs. 
 Cammell, the order for the Lepanto plates had not been yet given, and 
 experiments were required to arrive at the best description of plate for 
 all future supplies of thick armour. It was, however, competitive in 
 a limited sense, because there had been so little experience with 
 regard to compound armour of great thickness, that neither of the 
 representatives of Messrs. Cammell's nor Brown's firms offered their 
 plates with confidence as fairly representing what they wish to manu- 
 facture. Further, Messrs. Cammell specially requested that samples 
 of their plates might be tested before fulfilling their contract for the 
 Italia. The plates to be tried may be described as follows : — The 
 dimensions of all were the same, namely, 3-3 m x 2-62 m x 48 cm (10 ft. 
 10 ins. x 8 ft. 7 ins. x 18 - 9 ins.), the weight of each plate being 
 nearly 31^ tons. Three kinds of plates were tried, one from each of the 
 firms above-mentioned, namely, Cammell's, Brown's, and Schneider's. 
 No. 1, Cammell's, consisted of a wrought-iron foundation plate, 
 with a steel face applied on Wilson's patent, being rolled down from a 
 thickness of about 30 ins. to 18'9 ins. The steel, extending to a depth 
 of about 6 ins. in the finished plate, contained about 0'65 per cent, of 
 carbon. Mr. Wilson, who represented Cammell's firm here, stated 
 before the trial began that he considered that owing to imperfect 
 means the plate was not sufficiently worked, and that to do justice to 
 the system it should have been brought down from 36 ins. original 
 thickness. 
 
 l 8m Engineer reports of author. Not. 21, and Dee. 1, 1882. — The author ii greatly indebted to 
 the Italian Government, and to Admiral Albini and Admiral Bachia, for the opportunity afforded 
 him of observing these trials. 
 
 17 
 
118 
 
 No. 2, Brown's plate, differed from the above in having a thin rolled 
 steel face plate of about 3 ins. thick attached to the wrought-iron 
 foundation plate by molten steel on Ellis' patent. The total thickness 
 of steel was the same as that of Cammell's plate, that is, about 6 ins., 
 but it was slightly harder, containing about 0"7 per cent, of carbon. 
 The remark as to insufficient rolling applies to this as well as to 
 Cammell's plate. Both of them were bolted on to backing, hereafter 
 described, by means of six bolts, each of soft steel, screwed into the 
 back of the plates to a depth of 4^ ins., in screw holes 5J ins. deep, 
 in positions shown in Fig. 10. The diameter of the bolt end was 
 
 FIG, 10 
 BACK OF CAMMELS & BROWNS PLATES 
 
 4i ins., on which was a plus thread on the Palliser system, 5 ins. in 
 diameter over the thread ; the bolt fitted the hole tight to keep out 
 water, but when clear of the plate was reduced to about 3| ins. in 
 diameter, to ensure elongation in preference to yielding in the screwed 
 part. The rear end of each bolt was secured by a washer fixed on a 
 similar screwed thread to the front end, holding against the back face 
 of the backing. 
 
 3. — Schneider's — Creusot Company — plate consisted wholly of steel. 
 It was said to contain about 0'45 per cent, of carbon. It had been 
 hammered down under a 100-ton hammer from a thickness of 7 ft. to 
 19 ins. The face was tempered by lowering it to a depth of 6 ins. into 
 oil. It was afterwards slightly aunealed, so it is said, but no authentic 
 information as to its manufacture is furnished by Messrs. Schneider 
 and Co. This plate was secured by 20 screw bolts, each 4£ ins. 
 diameter, with a thread £-in. pitch, screwed into the back of the plate 
 to a depth of 2 J ins. The position of these bolts may be seen in 
 Fig. 11. 
 
 
 
 FIC.II 
 
 
 
 BACK 
 
 or 
 
 SCHNE 
 
 DERS PL 
 
 - 
 
 O 
 
 o i a o i3-OfEa-»O*-ea-*0( i' 
 
 
 o 
 
 
 o 
 
 O 
 
 
 o — 
 
 o 
 
 
 o 
 
 
 o 
 
 
 
 o 
 
 
 o 
 
 o 
 
 O O 
 
 O o 
 
 
 
 
 
 BOLT BOLE 
 
 s *j 01V 
 
119 
 
 The structure of the target backing and supports may be seen in 
 Figs 1 and 2. Each plate was set in an iron frame made of three 
 thicknesses of strips of 6-in. armour, the width of the frame being 
 about 33 ins. and the thickness about 18 ins. These were bolted to 
 the backing as shown in Pig. 3. The supporting frames seen in 
 Pig. 1, were about 2 ft. apart from edge to edge. Each plate frame 
 had a long prop at each end, extending from the top towards the 
 front, shown in Pig. 1. On the whole it may be seen that the backing 
 was soft though well supported. Had the frames been held together 
 at the corners they would have been very powerful, but as it was they 
 were of little use. 
 
 The projectiles were of Gregorini chilled-iron. They were about 
 44| ins. long and 17"64 ins. in diameter, the head being struck with a 
 radius of If diameters, and the bottom made to take the original 
 
 FIC. 4- 
 
 ROUND N»l , 
 ROUND AT CAMMELL.S PLATE 
 
 FIG. B 
 ROUND M°S 
 FIRST ROUND AT SCHNEIDEHS PLATE 
 
 o 00000000 000 o o 
 
 00000 00000000 o o 
 to~ ol to~ol 
 
 000000000000000 
 
 FIC. 6 
 
 ROUND N£ 3 
 
 FKS.7 
 ROUND H°. 4- 
 
 FIRST 
 
 ROUND AT 
 
 BROWNS TARCET 
 
 »««4it4i)aa9<k44«a 
 
 9414»14»OIIO<>()lt> 
 
 a 
 
 BROWN. 1 y 
 
 a» a 
 
 a a 
 
 // 
 
 O o> 
 
 a a 
 
 o> o> 
 
 <k 01 
 
 a a 
 
 
 a a 
 
 a a 
 
 a a 
 
 M/a 
 
 a a 
 
 a a 
 
 
 a 
 
 a oj n» a c> a 
 
 a t ft a <k a a 
 
 SECOND 
 
 ROUND AT SCHNEIDERS PLATS 
 
 
 <»Q&o>atO»<»»»ac»c»no 
 
 
 »tt»<»o>oiaOiO>CkO»c,ao>o 
 
 
 a 
 
 SCHNEIDER. 2. / 
 
 a, a 
 
 
 a a 
 
 v> .. / 
 
 o> »| 
 
 » a 
 o> o> 
 
 
 a o 
 
 Shoes' 
 
 
 s> a 
 
 a »' 
 a a 
 
 
 9 
 
 //Is^'ifl 
 
 5 a> o 
 
 
 a a a a a a a a|f»» <» • • * I 
 
 »-iSXt'""""cRAZES ,," POINT OF N?H SHOT DISJ-ODCED 
 
 — ' FROM HOLE very little apfearange 
 
 VIEW OF S±IOTS HEAD FROM A. FIC 1 of emLL 
 
120 
 
121 
 
 Blswick gas check employed with them. The weight was 896 k s 
 (1975-3 lb.) ; with gas-check, 907 k s (2000 lbs.) Their quality is 
 best discussed in connection with their effects. Speaking generally, 
 it appeared much better than that of the competitive shot of Gregorini 
 iron cast in our own Laboratory, these projectiles resembling much 
 more nearly the Fingspong shot employed in the English competition. 
 
 On November 16, the firing was commenced. The first three 
 rounds were to be fired under strictly similar conditions, the firing 
 charge being 149 k s (328-5 lb.) of Fossano progressive powder, which 
 was calculated to give the projectile sufficient energy to perforate 
 19 ins. of wrought-iron at the distance of the target. The spots 
 aimed at may be seen in Fig. 3. They were struck almost exactly in 
 each case, the shooting being admirable, in spite of a swell causing a 
 considerable heave of the raft on which the gun was mounted, as on 
 previous occasions. The targets stood in the order shown in Fig. 3. 
 Cammell's plate was first attacked. 
 
 In round No. 1 the shot struck Cammell's plate on the spot shown in 
 Fig. 4, with a striking velocity of 371-5 m (1219 ft.), having therefore 
 20,600 foot-tons energy, or 374 - 7 foot-tons per inch circumference, 
 and a penetrative power equal to 19'28 ins. of wrought-iron. The 
 plate was completely broken through in the thick crack shown in Fig. 4, 
 while hair cracks were developed as shown in thin lines. The shot 
 itself of course broke up, but it had held well together for a chilled 
 projectile. No indentations made by fragments were to be seen round 
 the portion of the head left in the plate, which projected about 5 ins. 
 It was scored and rubbed smoother than an English shot, and felt very 
 hot to the touch, arguing tenacity and comparative softness. The 
 rear part of the shot was broken into four large, two medium, and 
 many small pieces. The plate face was flat, that is, free from bending. 
 The iron frame had yielded outwards to the extent of from 4 ins. to 
 6 ins. near the point of impact — vide Fig. 4. One long front support 
 was thrown down, and a number of bolt heads in front broken, and 
 some cut by shot, fragments, &c. The whole plate was set back 3 ins. 
 at the end struck. In rear one large plate bolt was broken, and 
 several small backing and frame bolts. The depth of the indent could 
 only be guessed by the apparent diameter of the portion of head in 
 plate. This is deceptive, especially with soft shot which set up. 
 
 Round No. 2 was fired at Schneider's plate, leaving the centre one 
 — Brown's — to the last. The charge and projectile were as before. 
 The striking velocity was 375-5 m (1232 ft.), having therefore 21,050 
 foot-tons energy, or 3798 foot-tons per inch circumference, and a 
 penetrative power equal to 19'49 ins. of wrought-iron. The plate 
 resisted admirably, showing no cracks at all. The shot had behaved 
 much as in round No. 1. The depth of indent could not be easily 
 estimated. The fragment of shot was apparently much larger than 
 that in the Cammell plate, projecting about 6£ ins., the plate being 
 slightly raised or bulged in the surrounding region. The rear portion 
 of the projectile was broken up into small pieces. The iron frame 
 was started, opening about 5 ins. near point of impact, as shown in 
 Fig. 5. At the back several small backing and bolt heads were 
 snapped off, but none of the large plate bolts. 
 
122 
 
 Round No. 3 was fired at Brown's plate, and struck near the desired 
 point, Fig. 6. The striking velocity of tbis shot was about 372*5 m 
 (1222 ft.), having therefore 20,710 foot-tons energy, or 373*8 foot- 
 tons per inch circumference, and a penetrative power equal to 19*33 ins. 
 of iron. This shot broke up more than those hitherto fired, leaving a 
 smaller portion of the head in the plate projecting about 2\ ins., the 
 indent being apparently but slight. No deep indentations were made 
 round it, though rather deeper bruises than in the other plates. The 
 plate showed a narrow long crack in the position shown in Fig. A B, 
 due apparently to a sort of wave or bend-back, made by the whole 
 plate at the end struck. Some hair cracks were also developed. The 
 plate had bodily moved back about 2 ins., and at right bottom — corner 
 struck — about 4 ins. The face appeared slightly concave in the region 
 of the point of impact. At the rear some small frame and backing 
 bolts had snapped, but no large plate bolts. 
 
 On November 17 the firing was continued. The first round — 
 No. 4 of the series — was directed at Schneider's plate, striking near 
 the bull's-eye at a spot shown in Fig. 7. The striking velocity was 
 471 m (15*55 ft.) giving a total energy of 33,500 foot-tons, or 326*9 foot- 
 tons per inch circumference, and a penetration or perforation of iron 
 equal to 24' 7 ins. This shot struck near the bull's-eye, see 2, Fig. 7, 
 p. 119, evidently penetrating deep into the plate, the following effects 
 being visible : — The portion of the projectile lodged in the target 
 measured about 1 ^ ins. across, there being a ring of closely marked 
 scores and dents, making a disc of about 2 ft. diameter. The plate 
 split vertically across in lines, shown in Fig. 7 ; the plate continued to 
 " talk " or crackle for many minutes, cracks forming and opening 
 until the main fissure running down the left branch of fork was about 
 0'9-in. wide near the bottom, 0'7-in. wide a little above the shot, and 
 the shot had opened into two parts separated about 5 ins. in the 
 widest part, rule sketch Fig. 8 ; a ring crack also ran partly round the 
 centre mass of the shot. The whole target was heated for about a 
 foot round the edge of the shot, the shot itself being intensely hot. 
 The interrupted character of the main crack near the top, as well as 
 that of the smaller ones, was very characteristic. The right hand 
 fork of this crack was about 0'2-in. wide. Hair cracks were opened in 
 the position shown in Fig. 7, p. 119. These appeared to extend to a 
 considerable depth, a very small one on the left edge visibly extending 
 completely through the plate. The head of the shot previously lodged 
 in the plate was dislodged by this blow, and lay in two pieces, shown 
 in Fig. 9. The projectile had set up considerably, the depth of indent 
 being now measurable, and proving to be only 8£ ins., which is much 
 less than the appearance of the fragment when sticking in the target 
 indicated. The shot itself was obviously of a very different character 
 from our own service Palliser projectiles, being probably softer, the chill 
 extending only to a depth of about 1 in., and the metal yielding as no 
 English service chilled shot will yield. At the same time it is possible 
 that it is a more formidable shot against hard armour, from its tenacity 
 being apparently considerably greater than that of our own. In 
 short, the disappointment felt in the behaviour of the Gregorini shot 
 
123 
 
 when fired in our English competition in 1877 was in a measure 
 explained now in seeing its performance here. Probably, however, 
 this projectile would set up, and so fail to give very good results when 
 required to perforate soft iron nearly a match for it. It will be seen 
 that two hair cracks now showed themselves extending from the point 
 of impact of the first round. The side frames were sprung wider open 
 on the left, vide Fig. 7, p. 119. The shot hole will be seen to be of 
 considerably smaller diameter than the shot fragment which was held 
 in it, the edge of the latter having turned over and flattened under 
 the mass of langridge which followed it up, the target being dented 
 and bruised in a circle of impressions beneath it. It may be remem- 
 bered that a chilled projectile in 1876 behaved in a somewhat similar 
 way here on a larger scale. The back of the target stood well; some 
 small bolts were detached and frames cracked, but no plate bolts were 
 visibly injured. 
 
 Such a tremendous shock as this must inevitably perform a great 
 deal of work on a plate. This plate may be regarded as disintegrated 
 to a great extent, but it must be pronounced to have stood admirably. 
 The pieces held well in their positions, and whatever might be the 
 effect of a third round on the plate, it could scarcely be doubted that 
 
 FIC.I2 
 
 5 T r ROUND-z'tROUND at browns plate 
 
 the shot would be kept out of a ship carrying such a plate, and the 
 question may well be asked when a single plate would ever receive 
 three such blows on service. Messrs. Schneider showed also great 
 
124 
 
 judgment in employing a large number of bolts, for it is to be observed 
 that these cracks appeared in most cases to extend through the whole 
 plate. What, then, would have become of the portion below the second 
 point of impact if this entire plate had only six bolts . 
 
 Firing was continued on Monday, November 20. Round 5 of the 
 series was fired at Brown's plate, with the same weight of charge and 
 nature of projectile used against Schneider's plate in round 4, that is, 
 a charge of 217^ (478 lb.), and a 2000 lb. projectile of Gregonm 
 chilled-iron ; initial velocity, 478 m ; striking velocity, 476-6 m (1564 ft.) ; 
 stored-up work, 33,910; work per inch circumference, 612-0 foot-tons; 
 perforation of wrought-iron, 25"17 ins. The projectile struck the bull's- 
 eye, producing the effect shown in Fig. 12, p. 123. The plate was split 
 into six main fragments — five are shown in Fig. 13 — which were all 
 
 FIG.13, 
 
 PIECES OF BROWN'S PLATE ASSEMBLED. 
 
 dislodged except No. 5, which remained supported by two bolts. 
 The shot apparently had not penetrated to any great depth, but had 
 broken the plate. Its head detached itself, vide Fig. 18 ; also the head 
 
 FIG 17 
 
 ROUND. I. 
 
 SHOT POINTS 
 BROWNS 
 
 FIC .18 
 ROUND 2 
 
 of the previous round, vide Fig. 17. The wood backing in the centre 
 was split and torn ; the side frame pieces were thrown outwards at 
 the bottom ends. The plate bolts were snapped or drawn, with the 
 exception of the two shown in Fig. 12, holding up piece 5. 
 
 Round 6 was fired with similar projectile and charge at Cammell's 
 plate: — muzzle velocity, 479 , 4 m ; striking velocity, 477-0 m (1565 ft.); 
 giving stored-up work of 33,960 foot-tons, 613 foot-tons per inch 
 circumference, and a perforation of wrought-iron of 25' 19 ins. The 
 effects are shown on Figs. 14 and 15. The shot did not penetrate the 
 the plate, but bulged it. It brought down the entire plate, however, 
 snapping or drawing all bolts. In the back, No. 6 beam from the 
 top was broken and some others started and split. 
 
125 
 
 FIC.I1- 
 
 fiW ROUND at ROUND AT CAMMELS PLATE 
 
 The seventh round was fired on November 21 at Schneider's plate, 
 against which it was decided to try the effect of a Whitworth forged 
 steel projectile. The charge was 217 k s (478 lb.), the projectile weighed 
 942-5 k * (2078 lb.). The initial velocity was 471-4 m , the striking 
 velocity 468-8 m (1538 ft.) giving a total striking energy of 34,080 
 foot-tons, or 615-1 foot-tons per inch circumference, equal to the per- 
 foration of 25-23 ins. of wrought-iron. The gun was aimed at the 
 upper right-hand portion of the plate and struck it, producing the 
 following effects : — The portion of the plate struck was broken up, 
 some fragments being driven into the backing, vide Fig. 19, and part 
 driven a little to the right, the right-hand frame being thrown aside 
 
 FIG. 15. 
 
 PIECES OF CAMMEJXS PLATE ASSEMBLED. 
 
 18 
 
126 
 
 and left hanging by its bolts nearly drawn. The top frame was thrown 
 to the front, so as to hang over the face of the target. A part of the 
 plate was brought down by the shot, which rebounded and lay in front, 
 vide Fig. 19. This shot was set up as shown in Pig. 20, the extreme 
 
 y* ROUND , 
 
 3! ROUND AT SCHNEIDER'S PLATE. 
 FIC.I9. 
 
 point being broken off. The original length of this shot was about 
 44i ins. ; it was set up to a length of 28 ins. The impression of the 
 head and point was left on a curiously shaped piece of steel shown in 
 Figs. 21 and 22, which was purple and blue with heat, as were two 
 other pieces of steel plate lying close to the shot. These details 
 should be noticed particularly, because in the contact of steels of such 
 excellent quality it is well to note every indication of the enormous 
 shock that must undoubtedly have been produced. The wood 
 backing was rent and split, as seen from the front of the target. At 
 
127 
 
 FIC 2<? 
 
 nc si 
 
 the back the horizontal beams, 3, 4, 5, 6, 7, 8, and 9 from the top were 
 seen to be broken and split; some of the upper tiers of beams were 
 lifted with the top frame. The effects are roughly shown in the small 
 sketch— Fig. 24— taken from the bank at one end of the target. One 
 
 FIC .2.4- 
 
 large plate bolt was driven far out. The entire broken portion of 
 backing beams projected about 4 ft., and long splinters much farther. 
 After the firing it was discovered that this projectile was not a 
 Whitworth, but one of Terre Noire manufacture sent over from San 
 Yito by mistake. It certainly resembled Whitworth's very closely, 
 having a point inserted, and behaving very differently from the com- 
 petitive Terre Noire shot supplied to England. 
 
 It was thought that if the target props were looked to, the effect 
 of one more shot might be tried with advantage, and for this purpose 
 a cast steel Italian projectile was selected. The same charge, 478 lb., 
 was employed. The projectile weighed 952 k s with gas check 963 - 5 k e 
 (2124 1b.). The striking velocity was 461 m (1512 ft.) This gave a 
 total striking energy of 33,670 foot-tons and 607'7 foot-tons per 
 inch circumference, equivalent to the perforation of 25'08 ins. of 
 wrought-iron. The gun was aimed at a bull's-eye made on the left 
 
128 
 
 top portion. The shot struck the portion of plate close to the edge, 
 driving it over as shown in Fig. 25, the shot burying itself in the 
 backing. The natural effect of such a shot was to render the whole 
 
 f;c ek 
 
 structure a wreck. Eventually the shot was extracted, and found to be 
 broken. It had received a more severe blow than was probably 
 supposed by most of those present. The posterior fragments appeared 
 to be of inferior quality, but the fracture and appearance of the head 
 
 eic sb 
 
 FIC B7 
 
 was very good— vide Figs. 27, 28. This is due, no doubt, to the fact 
 that this projectile had been tempered in oil as far as the head extends. 
 
129 
 
 For more ready reference is given herewith a table of charges, 
 velocities, &c, of the M. L. gun : — 
 
 a 
 
 o 
 u 
 
 o 
 6 
 
 Fired at 
 
 Charge, 
 weight of. 
 
 Projectile, 
 weight of. 
 
 '8 . 
 
 o £ 
 11 
 
 3 a 
 
 '3 
 
 M 
 
 Striking 
 ■velocity. 
 
 ii 
 
 Is 
 
 60 "? 
 
 .So 
 
 'J3 
 
 Striking energy per 
 
 ton of plate. 
 
 foot-tons. 
 
 W ork per inch 
 
 circumference, 
 
 foot-tons. 
 
 Equivalent 
 
 perforation of 
 
 wrought-iron, inches. 
 
 u 
 
 0> 
 
 1 
 
 2 
 3 
 
 4 
 
 5 
 6 
 
 7 
 8 
 
 Cammell. 
 
 Schneider. 
 Brown. 
 
 Schneider. 
 
 Brown. 
 Cammell. 
 
 Schneider. 
 Schneider. 
 
 f 149 kg. 
 1 (328-5 lb.) 
 
 If 
 II 
 
 ( 217 kg. 
 1 (478-3 lb.) 
 
 II 
 
 II 
 
 II 
 
 907 kg. I 
 (2000 1b.) 3 
 
 C 942-5 kg. I 
 I (60781b.) 3 
 
 ( 963-5 kg. I 
 I (21241b.) 3 
 
 377-5 ? 
 
 377-8 
 374-8 
 
 476 
 
 478 
 479 
 
 471-4 
 464 
 
 371-6 
 
 375-6 
 372-5 
 
 474 
 
 476-6 
 
 477 
 
 468-8 
 461 
 
 1219 
 
 1232 
 1222 
 
 1555 
 
 1564 
 1665 
 
 1538 
 1512 
 
 20,600 
 
 21,050 
 20,710 
 
 33,500 
 
 33,910 
 33,960 
 
 34,080 
 33,670 
 
 654-0 
 
 668-2 
 668-5 
 
 1064- 
 
 1076- 
 1078- 
 
 1081- 
 1069- 
 
 371-7 
 
 379-8 
 373-8 
 
 605-0 
 
 612-0 
 613-0 
 
 615-1 
 
 607-7 
 
 19-3 
 
 19-6 
 19-3 
 
 25-0 
 
 26-2 
 25-2 
 
 25-2 
 25-1 
 
 The general character of the experiments reminds a reader of those 
 conducted at Spezia in 1876. There was the same strictly compara- 
 tive trial, each maker's plate heing subjected to the same test, round 
 after round. Whatever may be said to the contrary, there is no doubt 
 that the Spezia experiments of 1876, were the immediate cause of 
 steel coming into our own armour instead of wrought-iron. To 
 consider the chief points in succession — 
 
 Plates. — Messrs. Schneider's plate was an excellent one. After 
 receiving 122,300 foot-tons energy there remained a considerable 
 portion of the plate and a considerable measure of protection on a great 
 part of the target. Three particulars in which the Schneider plate had a 
 great advantage may be noticed: — (1) The number of bolts ; (2) the 
 arrangement of bolts ; (3) the tempering of the plate. As to the 
 number of bolts, it should be noticed that a paper was originally 
 sent to the firms supplying the plates, in which a backing was specified, 
 that only allowed of three vertical rows of bolts. Messrs. Cammell 
 and Brown on this fixed their six bolts. Messrs. Schneider, however, 
 objected to the paucity of bolts, and so the backing was altered to 
 meet their requirement. Messrs. Cammell and Brown might un- 
 doubtedly have done the same. They may be considered to blame for 
 not doing so. The object, however, is not to review the makers, but 
 their plates, and it is obvious that these are made to appear at a great 
 disadvantage when the bolts are so few that fracture causes the pieces 
 to fall down in front of the target, instead of being still held up as in 
 the case of the Schneider. (2) The distribution of bolts in the 
 
130 
 
 Schneider plate is very peculiar. A diagram 1 was sent to the firms 
 concerned, on which is drawn the plate and the points to be struck, 
 with dimensions. The Schneider plate with position of bolts is shown 
 in Fig. 11, p. 118; the front of this plate is shown in Fig. 30 as 
 if it were transparent, showing the bolt holes through it, the dotted 
 circles being the spots marked to be struck. The bolt holes, it will be 
 seen, are in irregular positions, obviously adjusted to meet the case of 
 the three blows specified. This, then, is not a service condition. 
 (3) With regard to the temper,— the plate was made in a peculiar 
 manner. It was hammered — not rolled — down from a thickness of 
 7 ft. to its present thickness of 18 - 9 ins., and the front then tempered 
 in oil. Plates so treated, while they improve greatly in quality, become 
 slightly contorted, indeed, the second plate of Messrs. Schneider, which 
 was lying on the ground, showed evidence of this. It is said to have 
 measured only •ib'2" lm , or 182 ins., instead of its full 18"9 ins. in 
 thickness at the ends, and that it thus projected slightly in the face; 
 and it was suggested confidently by the English makers that it had 
 warped from the true position shown in an exaggerated way by the 
 lines A A, CC, in Fig. 31 to that shown by BB, DD, and that in 
 
 T 
 
 BROWNS PLATE AFTER I" BLOW 
 
 SCHNEIDERS PLATE AFTER a"." BLOW 
 
 FIC.30 
 
 
 
 -1+ 
 
 — -o — o — o — o — o — o— 
 
 
 
 m 
 
 O "Of 2^ 
 
 1 
 
 o ^J 
 
 -I 
 
 a 
 
 >"\ o 
 ! £ ! Mi 
 O *—'' V.„' owe£» 
 
 2 
 
 r 
 
 i 
 
 
 - 
 
 — o — o o o — -o o— - 
 
 OT 
 
 kia ' 
 
 
 1 In the author's possession at the present time. 
 
131 
 
 order to fit fair on flat backing, the portion between C C and D D had 
 been removed, leaving the plate bounded by the faces B B and C 0. 
 This, no doubt, leaves the plate practically uninjured for this particular 
 experiment, and beautifully tempered, but it is not a service condition. 
 Even in a flat face of plate it would be objectionable, and it was urged 
 that on a turret it could not be carried out ; in fact, that it would be 
 impossible to treat curved plates of any form in this way, and that as 
 a matter of fact every single plate on the Italia is a curved one. 
 Consequently Cammell must supply untempered plates, and these 
 compare with carefully tempered ones at a great disadvantage. As to 
 the behaviour of the compound plates, the fracture of the iron would 
 probably have been better had there been more rolling. One peculiarity 
 in them has been noted, namely, the existence of concentric cracks, 
 which have already attributed to a bend back of the plate — vide 
 Fig. 32 — while Schneider's plate stands up and cracks radially only. 
 It may well be questioned if a concentric crack could be produced in 
 Schneider's plate. Fig. 32 is intended to represent roughly Brown's 
 plate after the first round, and Fig. 33, Schneider's, after the first and 
 second rounds. The crack in the compound plate extended apparently 
 only through the steel face. It must be observed, however, that the 
 plate on the next blow broke along the line of fracture marked by the 
 crack of the first round — vide Fig. 13, p. 124, and Fig. 6, p. 119. 
 Similarly, the crack developed by the second round at Schneider's plate 
 running from the point of impact of the first may have been initiated 
 by the first blow. Taking the whole case of the plates into con- 
 sideration, further trial between them was desirable, when conditions 
 as to bolts and tempering should be the same, and the compound plates 
 such that no excuse or apology is to be made for them. 
 
 The next question is the projectiles. — Our Committee on Plates and 
 Projectiles declared that chilled shot are not effective against steel- 
 faced armour. Can anyone say this has proved so at Spezia ? Look 
 at Figs. 12 and 14, which show the effect of two blows each with 
 chilled projectiles, one only a match for wrought-iron the same thick- 
 ness as the plate, and one a match for this thickness of compound 
 armour or steel. We tried this Gregorini iron in England in 1878, 
 and were disappointed in the result {vide p. 74). The reason, if 
 possible, it is important to trace. We tried our metal against soft 
 wrought-iron plates. The Gregorini shot were cast in our arsenal, 
 and so were hardly representative ones. The Finspong, which a good 
 deal resemble them — being made of charcoal iron — behaved as shown in 
 Fig. 34, p. 1 32 ; that is to say, they held together but set up, and so failed 
 to penetrate as well as our own Laboratory chilled projectiles — shown 
 in Fig. 35. The fact is that for soft plates hardness is more important 
 than tenacity. Our own projeotiles were liable to break, but not in 
 such a way as to interfere with their penetration so much as did the 
 setting up of the Finspong shot. On this we declared our chilled 
 projectile to be the best. Subsequently we found them almost 
 useless against steel-faced armour. Fig. 36 shows the effect of the 
 chilled shot fired from the 38-ton gun against steel-faced armour on 
 July 21, 1880 {vide p. 132). The striking velocity was 1504 ft., the 
 
132 
 
 FIC. 38 
 
 FIC 
 
 FIC 35 
 
 weight of shot 828 lb., consequently the total stored-up work was 
 12, (ISO foot-tons, or 332"6 foot-tons per inch circumference, and the 
 power of penetration was that of about 18'7 ins. of wrought-iron. 
 The plate was but 18 ius. thick. Compare this result with Fig. 38, 
 which illustrates similarly the effect of the first blow on the Cammell 
 plate at Spezia. Surely it is apparent that the greater tenacity of the 
 Gregorini metal has held it together so as to give a more telling blow 
 <;n the target in spite of the shot setting up into a mis-shapen form. 
 The English shot, up to this time, 1 when resisted by a hard surface, 
 flew asunder because its head was not buried in the plate, and in this 
 position, for obvious mechanical reasons, an outward thrust comes on it 
 which it has not tenacity to resist; the fragments then flew into a wide 
 circle, with enough energy in them to make deep unprofitable holes, 
 while the Gregorini shot held together and makes no such marks. 
 
 Fragments of the plates were subsequently fired at, and some con- 
 clusions, arrived at in Italy, were published in the Rivista Marittima, 
 May, 1883. The principal points noticed were hardness of the Bteel 
 face of the compound plates compared with that of the Schneider's, 
 and the flexibility of the back or foundation. Penetration in the 
 Schneider's may be effected in a greater or less degree, according to 
 the power of the shot (indeed, in softer plates, such penetration may 
 be so deep as to amount to complete perforation) . On the other hand, 
 the steel face of the compound plates is hard enough to resist penetra- 
 tion proper. The yielding nature of the backing in the above- 
 mentioned experiments was thought to tell much against the compound 
 plates which bent bodily more than the steel. On a ship's side, they 
 would be more rigidly supported. 2 These conclusions at the time 
 
 1 English chilled projectiles subsequently produced some remarkable results fired against steel- 
 faced armour at Shoeburyness. 
 ' Vide Admiral Acton's opinion quoted, Engineer, June 22, 1883, p. 476. 
 
133 
 
 appeared to be sound, but need some modification when variation 
 occurs in the conditions of trial. The Spezia trials are always on a 
 clearly defined plan. When comparisons are made, the conditions are 
 strictly comparative ; conclusions on the elements under trial are 
 therefore sound generally. It will be seen in a subsequent experiment 
 at Shoeburyness how thoroughly borne out was the conclusion that 
 the behaviour of compound plates is greatly effected by the nature of 
 the backing (see pp. 143 and 144). 
 
 It is necessary, however, to be clear about the question of perforation. 
 In the last Shoeburyness compound-plate trial (given pp. 112 and 113), 
 discs of metal were driven out, and a kind of perforation established. 
 This may almost appear to be contradictory to the above, and requires 
 to be considered carefully with it. The Italian authorities appear to 
 consider that the solid steel is locally less rigid and hard than the 
 steel face of the compound plates. Penetration in degree may be 
 effected in solid steel, nevertheless the solid steel as a mass is more 
 rigid, or, as it were, has more bone in it, and is less inclined to bend ; 
 consequently rigid backing is less necessary to solid steel than to 
 compound or steel-faced armour. 
 
 Armour Plate Trials at St. Petersburg. 
 
 On November 24, 1882, at Ochta, 1 near St. Petersburg, a Schneider 
 and a Cammell compound plate, each 8 ft. long by 7 ft. wide by 12 ins. 
 thick, weighing about 12 \ tons, were tried. Both plates were backed 
 by 12 ins. of timber placed horizontally and two f-in. iron plates 
 supported by diagonal struts. An 11-inch Aboukoff breech-loading 
 gun was fired at them, chilled-iron shells weighing 553| lbs. (English) 
 being employed, made at Perm in the Ural. The full charge was 
 
 FIC.I 
 
 1 Vide Engineer, December 8, 1882, p. 423, and April 6, 1883, p. 264. 
 
 19 
 
134 
 
 132 lbs. (English) and velocity 1506 feet. The shot broke up and the 
 Schneider plate was broken into five pieces at the first round, but it 
 was not detached from the backing, being held up by 12 bolts {vide 
 Fig. 1, p. 133). Two more rounds with reduced charges, 81 lbs., 
 (English) were fired, the striking velocity in each case being 1 167 ft. ; 
 the plate then being further split up, about one quarter of it being 
 detached from the backing {vide Figs. 2 and 3, herewith), the projectile 
 
 FIC.2 
 
 FIC 3 
 
 passing on to 740 yards in rear of the target. A round corresponding 
 to the first fired at Schneider's plate, was then fired at CammeU's 
 
135 
 
 plate with 132 lbs. (English) charge, velocity 1506 feet, the shot broke 
 up, leaving the head lodged in the plate ; the plate had some slight 
 cracks in it, concentric and radial, but was intact (vide Fig. 4, herewith). 
 
 A second round was fired at this plate with a reduced charge of 81 lbs. 
 (English), when the plate became detached from the backing falling 
 on its face. This may be attributed to the fact that it had been held 
 by only 4 bolts. Face cracks were produced and a piece detached to 
 a depth of 5 inches, which was attributed to a bad weld in the iron, the 
 junction of steel and iron occurring at 4 inches depth from the face 
 not 5. The back showed a bulge J-inch high at first point of impact, 
 but no other perceptible bulges or cracks. The stored-up work in the 
 first round at each plate was 8704 foot-tons, that is, 710"6 foot-tons 
 per ton of plate, implying a power of perforation of 16'3 ins. of iron. 
 The subsequent rounds at each plate had 5228 foot-tons energy, and a 
 power of perforation of 12*21 ins. The plates then were fairly tried, 
 the Cammell standing much better than the Schneider plate. It is to 
 be regretted that Cammell's bolting in this case as well as at Spezia 
 was insufficient. 
 
 On March 8, 1883, Cammell's plate was again fired at with the 
 same charge as the second round, with much the same results, namely, 
 a few more face cracks and another piece detached from the face. A 
 fourth round produced similar effects, that is, a few more cracks and 
 another piece of the face detached and the point of the third shot 
 dislodged; Figs. 1, 2, and 3 pp. 136 and 137, show the Cammell plates 
 after the 2nd, 3rd and 4th rounds. The striking energy or stored-up 
 work in each round, except the first, at each plate being 5,228 foot- 
 tons, or 426 - 9 foot-tons per ton of plate, the Cammell plate had borne 
 
Fie 2. 
 
a total amount of 24,388 foot-tons or 1991-3 foot-tons per ton of plate, 
 and the Schneider 19,160 foot-tons, or 1564-4 foot-tons per ton of plate. 
 In this case, the Cammell-Wilson plate had defeated the Schneider, 
 reversing the results obtained at Spezia, but in justice to Schneider, it 
 must be pointed out that the scale is a much smaller one, and that 
 success on a large scale is specially important. 
 
 On April 5, 1883, 1 Palliser's improved steel jacketed shot were fired 
 from an 80-pr. gun, against a wrought-iron plate 8'73 ins. thick 
 (termed the Rupert plate, No. 75) at Sboeburyness. It had been 
 employed for some Admiralty proof firing. Five 9-inch projectiles 
 had perforated it, one with 50 lb. and two with 40 lb. charges had 
 perforated it. Two with 30 lbs. had partially got through. The 
 portion struck by the Palliser shot was sound and untouched for a 
 considerable area (Fig. 1, p. 138) ; the projectile is shown before firing 
 in Fig. 2, p. 138. The chief peculiar features are the sharp-pointed 
 ribbed head, the reduced diameter, and the steel jacket over the body. 
 The striking velocity was 1400 feet per second. A heavy projectile of 
 this calibre (6 - 3-inch) might possibly be driven through the plate. 
 This shot however only weighed 85 lbs. Its calculated perforation (I) 
 is therefore 7" 7 ins. of iron. Its energy (E) was 1155. Its energy 
 per inch circumference (e) was 58'4. The shot not only perforated the 
 plate, but passed on, striking an old target with considerable force. 
 
 A compound (steel-faced) 6-inch plate was attacked by Palliser 
 and improved projectiles, and also by special shot made in the Royal 
 
 1 Vide Engineer, August 3, 1883, p. 96. 
 
139 
 
 Laboratory. A Palliser steel shot with steel jacket perforated the 
 plate, two feet of timber, and entered three inches into a plate behind. 
 Captain Palliser thought that the shot struck obliquely from appearance 
 of hole, and breaking of shot and jacket. A second Palliser shot of 
 chilled-iron with steel jacket was fired, it broke up badly, but it was a 
 bad shot, being only chilled to a depth of £-inch. The powder charge 
 was 20 lbs. pebble, the striking velocity was only 1400 feet. 
 
 Two long pointed chilled-iron shot (being a modified form of Palliser 
 service shot) made in the Royal Laboratory, were fired from the 80-pr. 
 gun at the 6-inch steel-faced plate and perforated it, so that the 
 Palliser improved jacketed shot displayed no advantage on this 
 occasion, that made of chilled iron being bad. 
 
 Engineer Experiments on Shields fixed on Masonry. 
 
 On Tuesday, August 22, 1883, 1 an experiment of great interest, 
 especially to England, took place at Shoeburyness. It was carried out 
 by the Royal Engineers with a view to testing the amount of protection 
 afforded to granite forts by iron plates. The general nature of the 
 work (known as No. 44 target) tested is shown in the accompanying 
 sketches (Figs. 1, 2, and 3, pp. 140 and 141). 
 
 The shields fixed on the face of portions II. and III. are as follows : — 
 that on No. III., against which the first shot was to be fired, consists 
 of two plates of 8 inches thickness each, of wrought-iron, sandwiched 
 with 5 inches of wood behind each, made up of two thicknesses, that 
 is 2^-inch planks laid horizontally, nex-t behind each plate, and 2^-inch 
 planks behind them placed vertically (Fig. 3, p. 140). The dimensions 
 of each plate were as follows : length 12 feet, height 7 feet, and 
 thickness 8 inches. They were supplied by Messrs. Cammell. They 
 were held in their place by six bolts, on the Palliser-English system. 
 
 The shield on portion II. consists of 1 2 inches of Wilson's steel- 
 faced iron, in a plate 7 feet by 7 feet held up by four bolts, fixed inside 
 an iron frame, as shown in Figs. 1 and 2, pp. 140 and 141. On the 
 top of the work was laid a quantity of old broken plating, to keep the 
 masonry and concrete from rising under the force of the blow — (vide 
 Fig. 1), at portions II. and III. Structures of concrete would hardly 
 attain their full strength for many months, perhaps years to come, and 
 allowance must be made for this in this trial, which it will be seen was 
 a very severe one. 
 
 The gun employed was the 80-ton gun, M.L., which was mounted at 
 200 yards distance. One round was fired from it at portion III. — iron 
 sandwich on granite. A projectile weighing 1,700 lbs. was fired with 
 a charge of 450 lbs. of pebble powder, with a velocity of 1,588 feet. 
 This implies a total amount of stored-up-work (E) of about 29,730 foot- 
 tons or 594*4 foot-tons per inch circumference, representing a power 
 of perforating about 25 inches of iron. The shot was a service Palliser 
 chilled-iron projectile, about 3 feet Q\ inches long, fired without bursting 
 
 1 Tide Engineer, August 31, 1883, p. 166, and September 28, 1883, p. 240. 
 
140 
 
 charge, the radius of the head being about 1| inches diameter. The 
 shot struck a point 3 feet from the bottom of the plate, and 3 feet 
 8 inches from the left end looking at it. The effects were as follows : — 
 The shot cut a clean hole, passing through both plates, and breaking 
 up during penetration, turned rather to the left, the point reaching a 
 depth of nearly 1 feet, measuring from the front face of the iron. 
 The plates behaved admirably, the hole being cut almost without any 
 apparent effect in the surrounding portion of the plate. The wood was 
 driven outwards, 5 inches of the ends of the horizontal planks being 
 thrust out beyond the plate at the left end, and 3 inches on the right. 
 
141 
 
142 
 
143 
 
 The granite was pulverized all round the projectile for some distance. 
 Cracks were visible in the granite in front, as shown in Fig. 4. They 
 will be observed to radiate from the point of impact, speaking 
 generally. The stones of the course through which the shot passed 
 are, like the layers of wood, forced longitudinally, left and right, 
 projecting 3 inches at each end of the squares of masonry, Fig. 4. 
 One or two cracks also were visible in the brickwork lining of the 
 small cross passage behind the part struck. The bolts do not appear 
 to have suffered, and the general structure shows little effect beyond 
 what is here mentioned. (Fig. 7, herewith shows the projectile with 
 the pieces assembled after recovery) . 
 
 A velocity of 1,100 feet ought to be sufficient to enable this pro- 
 jectile to perforate 16 inches of iron alone. 
 
 England is the only Power that has employed wrought-iron to any 
 considerable extent in coast forts, chilled cast-iron having come in 
 generally abroad. An experiment, therefore, that shows that wrought 
 iron behaves well is specially satisfactory to us as a nation, and surely 
 this is the case here. The iron has offered a great resistance, and it 
 has suffered only locally. The latter fact is, of course, important as 
 affecting the further powers of resistance of the fort. A shield to 
 resist repeated blows of the 80-ton gunshot must of course be 
 exceptionally strong. When it yields it yields locally, and leaves still 
 a good front protection. 
 
 The second round was fired on September 11, at portion II., the 
 steel-faced plate, &c. The projectile was in this, and in every other 
 case, a Palliser shot weighing over 1,700 lbs., the striking velocity 
 being something under 1,600 feet. The effect is seen in Figs 1, 2 and 
 3, herewith), which show the plate, and in Fig. 4, p. 144 which shows 
 
 FIG. 3 
 
 FIC. 1 
 
 1 
 
 1 
 
 M 
 
 m 
 
 «•'> N 
 
 W 
 
 11 
 
 STEEL. FACE 
 CfiMMELLS(WlLSQNS) COMPOUND PLATE 
 FBONTVIEW REMOVED FROM PORTION E 
 
 FIC.7 
 
 IRON BACK 
 CAMMEL.LS [WILSONS; COMPOUND PUKTB 
 
 BACK view 
 FIC.S 
 
 6HOT FBOM WROUGHT IRON 
 
 iuii position m 
 
 SHOT FBOM CRANITE 
 PORTION X 
 
SH0W1NC PORT ON n AFTER REMOVAL OF FRONT COMPOUND PLATE 
 
 the masonry behind it. It will be seen that the plate stood wonderfully 
 well. The shot broke up, its head being fixed in the plate. — {see 
 Figs. 1, 2 and ?>, p. 113). The plate was bent and bulged, the bulge 
 and shot being pressed unusually flat against the masonry supporting 
 the plate. There were great annular rents immediately round the shot, 
 where much violent work must have been done, the radial cracks were 
 nearly all fine hair cracks — the depth of the most important may be seen 
 in Fig. 3. The bolts stood well, holding the plate up. They were sub- 
 sequently broken to enable the back of the plate and masonry to be 
 examined. Fig. 4 shows the granite with the indentation and cracking 
 made by the blow. This, it may be seen, is very slight, the deepest im- 
 pression is that made by the shot point at A. The spring of the plate 
 has opened the joints at the upper bolts B B, and if the side view 
 (Fig. 3) be examined, it will be seen that a tremendous strain must 
 have fallen on these bolts. Crack C was produced by the first round 
 fired at portion III. Hence the cracks in the masonry, as before, 
 nearly all radiate from the point of impact. 
 
 Taking this round as nearly the same as No. 1, it may be said that 
 about 30,000 foot-tons work have been delivered on this shield, and 
 that a compound 12-inch plate has, under the conditions described, 
 borne the blow of a shot capable of perforating about 25 inches of 
 iron. Taking the plate as weighing 10£ tons, the shot was 2857 foot- 
 tons per ton of plate. Of course the plate backed by granite was in a 
 very different position from one forming the chief mass of a shield. 
 This one was only the face of a mass of masonry. Nevertheless, the 
 blow is an enormous one, and the performance must seem extraordinary 
 under any circumstances. 
 
 How is this to be accounted for? The natural suggestions are 
 inferiority in shot, special excellence in plate, or special support given 
 
145 
 
 to this nature of plate by hard backing. There does not appear to be 
 any reason to call the shot bad. The plate is certainly excellent, but 
 probably the last-named cause told most — that is to say, that very hard 
 backing specially brings out the powers of steel-faced plates. This 
 supports the opinion of the Italian Committee, who considered that the 
 yielding backing at Spezia told much more against the compound 
 plate than the steel. Anyone who looks at the indication of concentric 
 hair cracks which are apt to be formed on compound plates will perhaps 
 concur in thinking that the value of hard backing to this class of 
 armour is peculiarly great. Fig. 3 may suggest what would have been 
 the effect on this plate if the backing had allowed it to bend much 
 more. Would not the line of rupture from the point of the shot to 
 the cracks about A have been completed ? The bulged back of the 
 plate and the shot have received a tremendous pressure against the 
 backing. Can it be doubted that if the backing had not been an 
 extraordinary one the plate must have snapped across ? It is not 
 necessary to detract from the qualities of the plate, which is apparently 
 beautiful. It would surely be impossible for any 1 2-inch plate to stand 
 the blow we have to consider under any ordinary conditions. Giving 
 it all credit for excellence, then, it is desirable to explain why it bore 
 much more than twice the blow that would generally smash such a 
 plate up. The following is suggested : — The plate with its hard surface 
 and hard backing resisted the shot very sharply and rigidly ; this, being 
 a chilled shot, broke under such a shock much more easily than a good 
 steel shot would do. In fact, under these particular conditions, a 
 softer shot with more tenacity might have done better. Still, an 
 
 SHOWJNQ EDRT10N XAFIER REMOVAUOEJIRANITE BUCKS ABOVE POINT HE IMPACT 
 SHOWING GENERAL CONDITION ttF CONCRETE 
 
 enormous force was at work, breaking and tearing out rings of metal 
 close round the shot, and actually crushing in the face of the granite 
 behind it, and it is to be noticed that there are no detached bruises on 
 
146 
 
 the plate face, bo that the work was delivered well at the point of 
 impact. The plate had been cracked from the front to a depth of 9^ 
 inches at the edge, as shown in Fig. 3, p. 143, but the shot was unable 
 to bend it back and tear open the remaining thickness from the 
 opposite side, and so the blow was borne. 
 
 FIC.S. 
 
 SHOWINC HOLE WHERE SNOT 
 ROUND 
 
 ENTERED WITH CONCRETEIN PROCESS OF REMOVAL 
 IT TO CET OUT SHOT "«» »>« 
 
 PORTION IV 
 
 Figs. 5 and fi show respectively the granite and concrete portions 
 fired at with projectiles possessing about the same energy. The shot 
 in the granite portion struck at A, penetrating 5 feet of granite, then 
 13 feet of concrete, which brought its point to the second layer of 
 granite, when it turned sharply to the right, and after passing 7 feet 
 along the parapet, came to rest. In the concrete the shot penetrated 
 still deeper, but did not reach the far side of the parapet. 
 
 In August, )HSo, at Southport, 1 Sir Joseph Whitworth fired a forged 
 steel shell weighing 4():J lbs. from a 20-tou 9-inch gun, 29 calibres 
 
 Fig. 1, 
 
 1 Vide Engineer, August 34, 1883, p. lfiO. 
 
147 
 
 long, at an 18-inch wrought-iron plate, strongly supported behind by 
 an iron cylinder or hoop, forming a sort of tunnel (filled with rammed 
 wet sand) (vide Pig. 4, p. 146), the plate resting against it. At the 
 opposite end of the hoop was a l|-inch plate and wood backing with 
 steel angle-plates, and a large cast-iron bed plate rested nearly 
 horizontally against it, buried deep in sand. The projectile struck 
 with a velocity estimated at 1900 feet per second, having a striking 
 energy (E) of 10,090 foot-tons, an energy per inch circumference (e) x 
 of about 357 foot-tons, and a perforation t — 19*81 inches. The pro- 
 jectile passed through the plate and tunnel, breaking up the bed plate 
 and taking a direction shown in Fig. 4, herewith. The projectile was 
 very little damaged, the point being slightly distorted. 3 This is a 
 remarkable example of perforation proper. The excellence of the 
 projectile is chiefly to be noticed. 
 
 On October 23, 1882, and on Dec. 20, 1883, deck targets consisting 
 of 3 inches of iron or else of mild steel, were fired at Bastney, laid on 
 in two or three thicknesses. The 10-inch and 9-inch old type Wool- 
 wich guns were employed with 70 lb. and 50 lb. charges, and with 
 Palliser and common shells. The decks were inclined at 10° to the 
 line of fire. Common shell could not be depended on to burst with 
 Gr. S. or R. L. fuzes, and it was questionable if the effect on bursting 
 was always greater than on breaking up, which was the invariable 
 alternative. Palliser projectiles generally broke on impact, but pro- 
 duced more effect than common shells, especially those with blunt 
 heads. Against them the deck did not afford effectual protection, 
 though the projectiles were all deflected, even when holes were broken 
 in the plate. The protection was sufficient against common shell that 
 did not burst, but not always against those that did so. The Captain of 
 the Excellent pointed out that common shell would probably break ud 
 and produce considerable effect against unarmoured sides of ships, 
 while the Palliser might pass through without either exploding or 
 breaking up. 3 
 
 A competitive trial 4 of steel and compound plates at Amager, near 
 Copenhagen, took place on March 20 and 21, 1884. The Pigs, here- 
 with show the plates. The nature of each round is entered near the 
 point of impact. 
 
 The following is a brief summary of what took place : — The plates 
 were all 6 ft. 6£ ins. long, 5 ft. high, and nearly 9 ins. thick ; they 
 were curved, so as to represent a portion of a turret 10 ft. 9 ins. inside 
 radius. The backing was oak with iron skin and bolts. Schneider's 
 plate was solid steel, held up by sixteen bolts ; Marrel's, Cammell's, 
 and Brown's were steel-faced ; Marrel's and Cammell's being made on 
 Wilson's patent, and Brown's on that of Ellis'. Marrel's had eleven 
 bolts, and Cammell's and Brown's twelve each. The position of the 
 
 1 The projectile being hexagonal in cross section, this is not actually correct. 
 - It was exhibited in Geo Street, in Sir J. Whitworth's Museum, at the Iron and Steel 
 Institute, &c. 
 
 3 Compare with pages 47 and 48. 
 
 4 gee Engineer, March 28, 1884, p. 247, and If ay 80, p. 410. 
 
148 
 
 bolts is shown by white spots. The first round at each plate was fired 
 with a Krupp's 5"9 in. steel shell, with a striking velocity of about 
 1742-2 ft. ; weight, 112-44 lb. ; and striking energy of 2364 foot-tons. 
 Estimating the plates at about 5J tons weight, this would amount to 
 about 450 foot-tons per ton of plate. The second round at each plate 
 consisted of a Krupp steel 10-in. shell, weighing 402 lb., with a 
 striking velocity of 1410 - 8 ft., and consequently a total energy of 5551 
 foot-tons, or nearly 1041 foot-tons per ton of plate. 
 
 The effect on all the plates of the first round was more considerable 
 than might have been expected. At Ochta the lightest blows were 
 427 foot-tons, and the heaviest 711 foot-tons per ton of plate. At 
 Spezia the lightest were 654, and the heaviest 1046 per ton of plate. 
 At Shoeburyness in IS SO, a blow of 541 foot-tons per ton of plate 
 produced an insignificant effect on a steel-faced 18-in. compound plate. 
 The results of this first round were decidedly greater than those of the 
 lighter Spezia or Ochta rounds. This is true of the whole of the 
 plates, and as no one could conceive for an instant that Schneider, 
 Cammell, and Brown had all deteriorated in their manufacture, it must 
 be attributed to the fact that Krupp's steel are better than the English 
 or Italian chilled iron projectiles — a very natural conclusion, but one 
 which must not on that account be overlooked. Comparing the plates 
 together, there was at this stage little to remark. The steel of 
 Schneider's plate appeared to be good, the lower crack exactly 
 resembled those made at Spezia in its character; the short broken 
 lines are very characteristic. It is considered that Cammell's plate had 
 a harder steel face than Brown's, and that the penetration in it was less 
 deep ; but Brown's exhibited less cracking, and looks peculiarly well 
 at this stage. 
 
 The second round of course tested the plate much more severely, 
 being 1041 foot-tons per ton of plate, instead of 450 only. The 
 effects are proportionally great and their shape is instructive. The 
 shot makes an absolute breach or hole through the steel-faced plates ; 
 and no doubt had the plates been harder, more of the shock would 
 have been distributed through them. They yielded, in a measure, 
 locally. On the other hand, half of Schneider's plate (Figs. 1 and 2, 
 p. 149) was bodily carried away. There are two possible reasons for 
 this — one, that, as a mass, the steel is harder than is apt to be the 
 case, and it is difficult, if not impossible, to make a hole in such 
 steel ; and, secondly, the flank target is a little less strong and less 
 well supported at its outward end. 
 
 The third round at Cammell's (Figs. 3 and 4, p. 149) broke it up, 
 being another round with the 10-in. gun. The third round at Marrel's 
 (Figs. 5, 6 and 7, p. 150) and Brown's (Figs. 8, 9 and 10) consequently 
 was fired with the 5 - 0-in. with a chilled-iron projectile. The effect 
 appears to be as great as that of the first round ; but if the fact of the 
 plates being so far broken up is taken into account, may be said to be 
 actually a weaker blow, but no comparison can really be made. 
 
 On the whole, the compound plates must be said to have held their 
 own. Brown's and Cammell's appear to have held together better than 
 Marrel's. The tests were not very well suited to exhibit the powers of 
 
149 
 
 Fig. 1. — Schneider. 
 
 Fig. 3. — Cammbll. 
 
 Pig. 4. 
 
 S"" ROUND 
 
 the plates, but may perhaps have answered the particular object of the 
 Danish Government better than something which was a more even 
 match. To exhibit the actual powers of the plates on service, the 
 backing should correspond as closely as possible to the iron or steel 
 structure of a ship, or to the wall of a fort. As armour becomes 
 harder, and as the blow is transmitted more through its mass, the 
 strength of the supporting structure is called more into play. A hole 
 in a plate generally means a hole in the backing, and the passage of 
 some langridge into the interior, but the movement of a shield bodily 
 tests the supporting frames as a structure. 
 
 Experiments at Shoeburyness in 1884 were made with Palliser 
 improved and other projectiles with specially sharp points. Some good 
 penetrative effects were obtained, the most remarkable perhaps being 
 that a cast-steel projectile of Hadfield's passed through a steel-faced 
 plate without breaking up. The plate however was too far injured 
 previously to build an opinion positively on the result, and subsequent 
 rounds did not bear it out. Experiments also pointed to the conclusion 
 
 21 
 
150 
 
 ?Pig. 5. — Marrell. 
 
 Fig. 8. — Brown. 
 
 T»T ROONP 
 
 K 1 HOUND 
 
151 
 
 that steel-faced armour transmits the shock of the supporting struoture 
 more than wrought-iron. Nothing of a decided character can however 
 be stated. 
 
 On October 16, 1884, a flat-headed steel projectile produced less 
 
 effect on a steel-faced plate than a chilled pointed projectile. 1 In 
 
 recently constructed guns the sectional density of the projectile has 
 
 been much increased. For example, in the 9 - 2-inch gun, Mark III., 
 
 W 
 
 — j = 0'488. Tbjs, with a velocity of 2000 feet means a perforation of 
 
 about 20 ins. of iron, even with the old service shaped head of 1*5 
 
 W 
 diameters radius. In the new 63-ton gun of lS'S-inch bore yr 3 = 0'51, 
 
 which implies great power of perforation as well as power to keep up 
 velocity in flight. 
 
 On Oct. 1, 1884, 2 another important competitive trial took place 
 between the steel-faced plates of Cammell, of Brown, and the solid 
 steel plates of Schneider at Muggiano, that is the Spezia experimental 
 ground. Owing to the fact that Spezia had suffered severely from a 
 visitation of cholera, very few persons were allowed to attend the trial. 
 The following is a brief account of it : — 
 
 The object in view was to learn the effect of the fire of the new 
 Armstrong 100-ton B. L. gun, 17-inch (43 cm ) calibre. When dis- 
 charging the best steel projectiles on the armour of the Italia and 
 Lepanto, that is to sav, to ascertain in what manner the plates would 
 suffer when enormously over-matched. The competitive element was 
 maintained under these conditions. The .order for the Lepanto plates 
 being said to depend mainly on the results of this trial. The 
 programme was as follows : — Cammell and Brown compound and 
 Schneider steel plates 10 ft. by 8 ft. 6 ins. by 18'9 ins. (3050 mm by 
 2600 mm by 480 mm ) were mounted on frame and backing shown in Pigs. 
 9 and 10, p. 155. One round was to be fired at the centre of each 
 plate from the 100-ton breech-loading Armstrong gun, to be followed 
 by a round at each of the four corners of each plate from a 10-inch 
 gun. 
 
 The 100-ton gun fired a projectile weighing 1841 lb. (885 k s) with 
 . a striking velocity of 1864 ft. (568 m ) per second, making a blow of 
 44,340 foot-tons, or 8337 foot-tons per inch circumference, implying 
 a calculated perforation of 30'27 ins. of iron, and taking the target's 
 weight as 29 tons, an amount of energy of 1529 foot-tons per ton of 
 plate. As the greatest amount of energy per ton of plate in the 
 
 1 It might reasonably be supposed that a flat-headed shot might at all events hold well together 
 on impact against hard armour, Fairbairn having found that a flat-headed bolt resists double the 
 crushing force required to set up a hemispherical ended one (see p. 4, Chap. I.) This 
 proiectile, however not only failed to crack the plate as much as the pointed shot, but also 
 broke up more. The author cannot explain this, and would have expected the opposite, that is, 
 he would have expected the shell to hold well together, though perhaps it might not produce a 
 great effect on the plate. As to this, however, vide trials of Griison's chilled-iron shields hereafter. 
 
 3 Discussed in Engineer of October 24, 1886, and February 27, 1885. Cuts first appeared in 
 accounts in Annates Industrielles and Mevue Maritime, and Engineering. 
 
152 
 
 previous Spezia trials was 1046 foot-tons, while the calculated per- 
 foration through iron was 25 - 2 ins., it may be seen that this test was 
 very severe indeed ; in fact, the plates were enormously out-matched 
 on any standard of estimation. The effects are shown on the Cammell, 
 Brown, and Schneider plates respectively in Figs. 1, 2, and 3, pp. 152, 
 153, 154. The steel-faced plates obviously suffered much more than 
 the steel one, but there appeared distinot reason to think that they had 
 
 CAMMELL STEEL FACED PLATE 
 AFTERIBLOW FROM 100 TON CUN STEEL PROJECTILE 
 
 CAMMELL 
 
 AFTER i S&ttKP,. FR0M '. 00 TON CUN 
 « e BOUNDS « 10. IN • 
 
153 
 
 done more towards stopping the projectile than the Schneider plate, 
 inasmuch as the shot went through the latter in a much more complete 
 condition. Fig. 8, p. 155, shows a drawing of the fragments taken out 
 of the butt subsequently. 
 
 It was argued fairly enough that a shot with less velocity might 
 have been stopped altogether by the steel-faced plate, while the steel- 
 plate would have allowed it to pass through, and further, that the 
 backing was much softer than a ship's side, and that this was in favour 
 
 fig. a 
 
 BROWN STEEL FACED PLATE 
 
 AFTER I BLOW FROM 100 TON CUN STEEL PROJECTILE 
 
 BROWN 
 
 alMW AFTER I ROUND FROM 100 TON CUN 
 U B BOUNDS u ID' IN a 
 
154 
 
 of the steel-plate, whioh, while having a less hard face, has, as a mass, 
 " more bone in it." This argument might be fairly held at this stage 
 of the trial. The plates were next subjected to the fire of the 10-in. 
 gun, with a striking velocity of about 2000 ft., with the following 
 results, on November 5, 6, and 7 : — 
 
 Brown's plate, penetration first round, 1 2'6 ins. (320°™), a great part 
 of plate detached ; second round, the plate demolished to the extent 
 shown in Fig. 5, p. 153. About three-fourths of the plate being now 
 stripped off, further firing was impossible. 
 
 FIG 3 
 
 llil 
 
 SCHNEIDER STEEL PLATE 
 AETER I BLOW FROM 100 TON GUN STEEL PROJECTILE 
 
 SCHNEIDER 
 
 AFTEB I ROUND fhdm IOO TON GQN 
 
 • 3B0UMDS . 10 W . 
 
155 
 
156 
 
 Schneider's plate then received two blows, producing cracks and 
 fracture to the extent shown in Fig. 6, p. 154, which is given for the 
 sake of comparison with the steel-faced plates. The shot is stated in 
 the French accounts to have been more broken up against the steel 
 than against the steel-faced English plates. Two more rounds were 
 fired with the 10-inch gun, the plate at the conclusion of the programme 
 being in tbe condition shown in Fig. 7, p. 155. 
 
 Cammell's plate was then attacked, broken pieces of plate being 
 thrown off to an amount estimated at nearly 5 tons — 5000 kilogs. — by 
 the first round, and the plate being brought to the condition shown in Fig. 
 4, p. 152, in the second round with the 10-inch gun — a condition which 
 made further firing useless. The Schneider steel, therefore, on this 
 occasion won a remarkable victory, and the English plates were beaten. 
 It may of course be argued that the steel-faced plates absorbed so 
 much of the energy of the first heavy blow that they were afterwards 
 in a worse position than the Schneider, which let it through more 
 easily. The natural reply is that the makers knew the tests that would 
 be applied, and should have made the plates to stand well under these 
 conditions, and few would sacrifice the power of holding together 
 shown by the steel, for the possibility that the steel-faced plate might 
 have kept out fire, while the steel admitted it in some hypothetical 
 case. 
 
 Is it not reasonable to say that, so far as we are able to form a con- 
 clusion from the above, it would be that on this scale our compound 
 armour is behind Schneider's steel when exposed to direct attack? 
 It is possible that Cammell and Brown may snatch the victory on 
 some future occasion, especially if better means are found of working 
 very thick plates, but at present, surely, it is only true wisdom to 
 recognise the liability of steel-faced iron on this scale to fail to exhibit 
 the powers of steel armour as made by Schneider. All the compound 
 armour victories have been with thinner plates. Xo one can say that 
 the compound principle, if carried out with hard and soft steel, may not 
 be the best, but when the results above given occur repeatedly with 
 iron, and with intervals of time sufficiently wide to develop manu- 
 facturing operations fairly well, it is idle to shut our eyes to them. 
 Without seeing the targets themselves, there is evidence enough in 
 the drawings to tell something to those who have seen previous 
 results. For example, the crack in the steel below the letter S precisely 
 resembles the long vertical crack in the Schneider plate fired at in 
 1882 in its character. There may be seen a succession of short, dis- 
 connected cracks, looking in the sketch like the twists in a rope, which 
 are very characteristic. 1 
 
 Early in the summer of 1885, 2 a proof trial of steel Schneider 
 plates took place at Spezia of a much less satisfactory character, 
 showing that the high quality shown in the October and November, 
 
 1 A comparison of this with author's sketch of Schneider's plate in 1882 trial, will give evidence 
 of accuracy as to features of sketch. 
 
 5 The writer has heard facta in conversation that enable him to arrive at a probable conclusion, 
 but not circumstantial or definite. 
 
157 
 
 1884, competitive plate could not always be secured. This muoh ought 
 to be said in justice to the Sheffield manufacturers of steel-faced 
 armour, but it is useless to attempt to give details from confidential 
 or doubtful sources. 
 
 Experiments at Shoeburyness during 1885, have shown that with 
 steel-faced armour, rigidity of backing is by all means to be secured, 
 even at the expense of thickness in the plate. 
 
 These experiments were made with armour and supporting structure 
 proposed for the Admiral class of ships. Steel-faced plates, 18 ins. 
 thick, under the fire of the 80-ton gun, did not stand so well as 16-in. 
 plates with the weight of the 2 ins. of plate added in the structure. 
 Less good results were of course obtained with 14 ins. of plate and 
 the same backing, and also with 18 ins., disposed in 12 ins. and ins. 
 sandwiched. 
 
 On August 18, 1885, at Ridsdale, Blswick, an Armstrong 9-2-in. B.L. 
 gun projectile, weight 392 lbs., striking velocity 2166 feet, and energy 
 12760 foot-tons, perforated an 18-inch plate of iron and entered 6£ 
 inches into a 7-inch plate in rear. The theoretical perforation is about 
 22 inches {see Oapt. Penton's paper, R. A. I. Proceedings, Vol. XIV., 
 p. 321). 
 
 In October, 1885, on the suggestion of Major English, R.E., a steel- 
 faced plate about 5 feet square and 16 inches thick, was bound round 
 by steel riband or wire laid in a large groove round the plate. The 
 plate was fixed on granite, and attacked by the 43-ton B.L. gun. It 
 was broken to pieces, and the fragments and riband all thrown about. 
 Some of the riband had been much extended, but it has been objected 
 that to act well, theoretically, the riband should be wound on to a 
 cylinder, a condition that could hardly be carried out in armour plates. 
 Steel riband is very costly, and scarcely likely to be used in this manner. 
 
 Some sort of review of the results obtained in the various trials of 
 steel-faced and solid steel plates seems here called for, though it is a 
 delicate and difficult matter. 
 
 Steel-faced armour has generally been remarkably successful when 
 well-backed, as at Shoeburyness in 1883, 1 and when its hard surface 
 was able to tell, as in the piece-meal trials after the Spezia 
 experiments of 1882. 3 On a small scale it has done remarkably well, 
 as at Ochta in 1882, 3 and often at Shoeburyness and Portsmouth.* 
 Where disappointment has occurred, it has been with thick plates and 
 indifferent support; and this statement is endorsed by experiments that 
 are not here given. The hard face is the chief advantage possessed by 
 the compound plate. This gives a great power to throw off oblique 
 blows, a work constantly called for on service, but seldom tested by 
 experiment. The compound plate has probably suffered by being 
 compared with steel under direct attack only. The hard face has not 
 always prevented discs being bodily torn out 6 when the backing 
 admitted of it ; in fact, this form of fracture is due to the hard face, 
 for soft steel or iron would have probably allowed the shot point to cut 
 
 1 Vide p. 143. 2 Vide p. 132. 3 Vide p. 133. ' Vide pp. 69, 79, 80, 99, 101, 108, 111, and 143. 
 5 Vide p. 112. 
 
 22 
 
158 
 
 through in preference. On certain occasions steel-faced armour has 
 been fairly perforated, but it resists this action much more strongly 
 than solid steel generally, owing to the face of the latter being softer 
 and allowing a hard sharp point to enter it. 
 
 Steel has shown to greatest advantage when used in great masses ; 
 it depends upon its qualities as a mass rather than on its surface, where 
 some measure of hardness has to be sacrificed to ensure the toughness 
 through the mass. To those who judge only by the results of limited 
 experiments it may probably appear that any attempt to get increased 
 hardness in the face involves uncertainty in the quality of the mass. 
 Apparently the tendency latterly has been to make steel sufficiently 
 soft to ensure toughness, and consequently to admit of perforation 
 when over-matched. So long as all fire is kept out, this may be right. 
 Most Officers would prefer to admit the quantity of dead metal that 
 would pass through such a plate as that fired at at Spezia in 1884 
 to the total demolition exhibited in the steel-faced plates on that 
 occasion, dead metal still coming through, although in smaller frag- 
 ments. Nevertheless, should projectiles improve in quality, the plates 
 may have to be made harder. In the long run it seems probable that 
 hard rather than soft armour will be preferred. Those who are not 
 themselves engaged in manufacturing plates may probably wonder that 
 compound armour, consisting of a hard steel face and tough steel 
 foundation or body, has not before this been successfully made. 1 
 
 1 The difference duo to the compound plates being rolled, and the solid steel plates hammered, 
 must not bo lost sight of. It may naturally appear that rolling is less likely to reach the interior 
 of a very thick plate than hammering with a 100-ton hammer. The very vibration of the latter, 
 though a rough process, may tend to prevent crystallization and so to preserve fibrous structure, just 
 as agitation of a matter in solution prevents the formation of large crystals. With this in view 
 the author has looked for larger crystals in the interior of a rolled compound plate after fracture 
 than near the exterior, but he cannot say that he found them hitherto. 
 
PART II. 
 
 CHAPTER I. 
 
 Gruson's Chilled-Iron Armour. 
 
 The introduction of chilled cast-iron has not been given in its 
 historical order in the plate experiments, because it constitutes a 
 distinct branch of investigation, and can be better dealt with as a 
 separate subject. In two instances only has chilled-iron been fired at 
 in the experiments dealt with, namely, once at Spezia in 1876, when 
 its application was so unsuited to its character as not to deserve much 
 notice, and once at Meppen in 1879, when a chilled-iron shield was 
 made by Krupp, which Griison repudiated, but which can hardly be 
 said to have behaved badly (vide pp. 96-97, I.). Blocks of chilled- 
 iron have been fired at at Shoeburyness, but in so imperfect a manner 
 as not to deserve notice. 
 
 Herr Griison claims to have invented chilled-iron shot, 1 the credit for 
 which he disputes with the late Sir William Palliser. He designed 
 chilled-iron armour, which has been advocated on the following 
 grounds 3 : — Wrought-iron, he considers, resists on a bad principle. 
 The material is soft and ductile, and its strength consists in its being 
 held together by tenacity rather than in any direct power of resistance. 
 Thus, a shot enters a soft material, and, as it enters, its passage is 
 opposed by the difficulty of tearing the metal open. This form of 
 resistance is objectionable on two grounds. In the first place it is 
 local in its action, that is, the metal in the immediate spot has to bear 
 the entire strain, and so perforation becomes feasible. It is true that 
 from the same cause the injury effected is local also, and an advantage 
 as to future powers of resistance may be claimed on that account for 
 wrought-iron ; but this does not affect the objection urged, namely, 
 that the power to keep out the shot must be exerted by the metal in 
 
 1 Development of Armour, Very, p. 532. 
 
 2 Author's paper in Engineer, December 26, 1879, p. 466. 
 
 23 
 
the immediate vicinity of the point of impact. The second objection 
 urged is that the plate only resists in the action of destruction, and 
 thus any shot that it has kept out has destroyed armour in the precise 
 measure in which the armour has resisted it. A shot whose power 
 falls far short of that necessary to penetrate the plate, expends all the 
 work it has stored up in it in actually penetrating the plate to a cor- 
 responding extent; for it experienced no serious resistance on striking 
 the surface, but was stopped gradually as it drove its way into the 
 plate until it had performed all its quantum of work. This, it is 
 urged, is a very unsatisfactory form of resistance, and it is claimed for 
 chilled-iron that it resists the shot at its surface, that its surface does 
 not sensibly yield at all, but transmits the shock into the mass of metal 
 behind it and around it. In this way a blow must be sufficient to break 
 the entire plate, or must fail to produce any appreciable effect. The 
 mass of metal, it is urged, absorbs the blow in the same way in which 
 an anvil stops the blows of a hammer. Consequently the plate is not 
 sensibly injured by comparatively light blows. Then, again, any 
 additional thickness in the armour strengthens chilled metal in a much 
 greater degree than wrought-iron. To illustrate this, suppose Fig. 1, 
 herewith, represents the method of yielding of a wrought-iron plate 
 
 fi c.3. 
 
 FIC.4 
 
 during the passage of a shot, a supposition which we know is 
 approximately correct. And suppose Pig. 2 to represent the trans- 
 mission of a shock through a chilled-iron plate sufficient to destroy it, 
 a supposition which is fanciful, but which may be allowable for purposes 
 of explanation. Then it may be seen that the portions of the plate 
 which come into play and take part in the resistance may be represented 
 by those shown in Figs. 3 and 4, by which it will be seen that if the 
 portion of the wrought-iron plate a be supposed to be equivalent to 
 the portion of chilled plate d, then equal additions to the thickness of 
 
3 
 
 the two plates, as shown by b and c, and e and/, increase the strength 
 of the chilled plate in much greater proportion than the wrought-iron 
 one. Of course it will be urged in reply that though the work of a 
 shot which fails to penetrate may not be visibly impressed on the 
 plate, it may have been performed in another way on the particles of 
 the metal so as to do its part towards eventual disintegration of the 
 plate, without the visible effect so apparent in wrought-iron, and that 
 the very distribution of the shock must tend to weaken all parts in the 
 vicinity, and facilitate the action of a subsequent blow, whereas in 
 wrought-iron the effect is so completely localized, that, as has been 
 said, partial penetration practically leaves the plate uninjured. To 
 pass on however to less questionable advantages offered by Griison's 
 armour. It admits of being cast in any desired shape, and with any 
 desired variations in the thickness. Owing to the difficulty of rolling 
 plates of any but the simplest form, wrought-iron turrets have nearly 
 always been cylindrical with the metal of uniform thickness, and it is 
 evident that the roof, at all events, may be much more suitably shaped 
 on Griison's system. 1 The curved form, together with the hard surface 
 of the metal, enables this armour to throw off a shot striking obliquely. 
 Then, again, chilled-iron plates may be easily cast in great masses; 
 while in wrought-iron the difficulty of manufacture rapidly increases 
 with the thickness and area. From this it follows that chilled plates 
 may often be made near their destination if skilled men are sent there, 
 chilled-iron armour is said for this and other reasons to be specially 
 cheap. Lastly, but not least by any means, when suffering under the 
 fire of Artillery capable of breaking it, there are no bolt heads, rivets, 
 &c, to fly off as langridge into the interior of a turret, and any frag- 
 ments of metal actually severed from the rest by cracks, generally 
 remain held in situ for a considerable time, because the cracks are 
 generally by no means straight, and because the arched form of the 
 structure tends to favour their retention. 
 
 Briefly then it may be said that wrought-iron and chilled-iron 
 represent soft and hard iron to the full extent. Wrought-iron resists 
 and suffers locally, hence perforation is comparatively easy. Except in 
 the immediate locality however the plate is little injured. Thus, seeing 
 that it is generally impossible to strike the same spot twice, wrought- 
 iron is singularly well suited for resisting long continued attacks from 
 the fire of such guns as have not power to perforate it completely. On 
 the other hand, it is more open to the passage of projectiles carrying 
 fire into the interior than any other kind of armour. In all this, 
 chilled-iron exhibits the very opposite qualities. It has the power to 
 distribute the shock of impact through its mass. It probably never 
 has been and never will be perforated. It yields by fracture and 
 general disintegration. Long continued fire is what tries it most, and 
 though the effect of the blows may appear to be slight, it cannot be 
 doubted that the mass into which the shock is transmitted, suffers 
 gradually, and the blows tell more and more as the firing proceeds, 
 
 i In the Schumann-Griison Cupola, tried at Bucharest in 1886-86, steel-faced plates were pressed 
 by hydraulic power into a curved shape, forming a dome. 
 
bo in almost every detail wrought-iron and chilled present the opposite 
 effects. Chilled-iron is well suited to throw off shot striking obliquely. 
 In wrought-iron they have the best opportunity of biting. A blow 
 enormously outmatching wrought-iron makes a clean hole, and effects 
 the maximum possible of mischief behind the armour and the minimum 
 on the armour itself. Such a blow will probably effect the minimum 
 of injury on material protected by chilled-iron shields, while it will 
 involve the armour itself in wholesale ruin. Lastly, wrought-iron 
 lends itself to the work of strengthening shields by additional front 
 plates, while finality is involved in the nature of a chilled-iron shield, 
 to which it would be out of the question to make additions such as are 
 contemplated in our own plate-upon-plate forts. 
 
 Griison's armour appears to have been first tried by Prussia at 
 Tegel in 1868, and by "Russia at Perm in 1871, with promising results, 1 
 and in 187o 2 such success was obtained as secured its adoption. Two 
 5171b. shot were fired from an 11 -inch (28 cm ) Krupp gun, with an 
 88 lb. charge, at a plate whose maximum thickness was 28"3 inches, at 
 10 yards distance. Both shot struck the same place, and formed on 
 the surface of the plate a rough spot a few millimetres in depth. Two 
 slight cracks were observed commencing on the inside. The shot 
 were broken to atoms. On July '27, 1874, another trial commenced. 
 Captain Grenfell says 3 of this trial, which was completed on August 21, 
 that the ten rounds fired against one plate left it far from unserviceable. 
 Firing was then continued, and although the surface of the plate lost 
 in appearance, no increase of damage, as far as the serviceability of the 
 plate was concerned, could be discovered. Three things were remarked, 
 
 (1) Throughout the whole of the practice no penetration into the 
 plates was effected ; the projectiles, were, in fact, effectually kept out. 
 
 (2) Even after the plate was split in two parts, many shot were fired 
 without producing any visible effect. (3) No injury beyond a crack 
 occurred on the interior of the plate. Fig. 5, p. 5, II., exhibits the 
 condition of the plate at the conclusion of the trial. The thickness of 
 wrought-iron " t," that might have been perforated by a 517 lb. shot 
 is about 14i inches probably, the striking energy being given as 
 7027 foot-tons, which implies a striking velocity of about 1403 feet. 
 The chilled-iron was about twice the thickness of the wrought-iron 
 that might have been perforated then. It bore 19 rounds on a very 
 confined area. At Tegel, in 1874, a turret with a shield, with a 
 maximum thickness of 21 - 65 inches (55™) withstood 277 rounds from a 
 5 - 9-inch (15™) gun, twenty from a 7-inch (17 cm ) gun, and two from an 
 11-inch (28 cm ) gun. At this period, then, Griison's metal shields 
 deservedly established their reputation in Germany. The Prussian 
 officers observed that while experiments had been going on with 
 
 1 See Development of Armour, by Very, pp. 533-534. 
 
 2 Engineer, December 20, 1879, p. 466. 
 
 3 Captain Grenfell, R.N'., read an able paper on chilled-armour at the Institution of Naval 
 Architects, in April, 1877, and advocated the trial of this armour in England. Major Kuster, of 
 the Prussian Artillery, reported the Tegel experiments in detail (vide Development of Armour, 
 Very, p. 636). 
 
wrought-iron in England through long series of years, chilled-iron 
 had been so far perfected as to be recommended in three or four years. 
 Griison's chilled iron shields in one or another form have been 
 adopted by several powers. Germany adopted cupolas and batteries 
 for coast and river defence, as, for example, those on the Lower Weser. 
 Belgium has batteries which will be noticed later on. 
 
 Fig. 5. 
 
 In Spain, turrets have been talked of for the defence of the Cadiz 
 harbour, and in Portugal, for Lisbon and the mouth of the Tagus, 
 but nothing has been actually done. Austria has also Griison's armour. 
 Italy ordered chilled-iron casemates for Alpine forts, and has recently 
 
6 
 
 adopted two chilled-iron turrets of great magnitude, each to contain 
 two 119-ton Krupp guns for the defence of Spezia Harbour. 
 France, for a time, adopted Griison's armour for inland defences. 
 This has subsequently been reversed, 1 and this class of shield in 
 France, as in most other countries, is approved only for coasts. In 
 1882, it was found that while chilled-iron projectiles were broken 
 up harmlessly against chilled-iron, this was not the case to the same 
 extent with steel projectiles, whose tenacity enabled them to deliver 
 more work at the point of impact, and produce much greater effect. 
 Hence it was concludod that chilled-iron shields might be broken up 
 under a long continued fire of steel projectiles, and on that account 
 it was better suited to coast defence, where ships could not maintain 
 a continued attack such as might be conducted in regular inland 
 breaching operations. 2 
 
 As has been seen, Griison's armour offers a shield readily made, and 
 very simple in its character, requiring no backing or supports, but 
 itself forming the structure required. The iron employed should be 
 of superior quality, free from phosphorus or sulphur, and of course of 
 a selected mixture of grey and white, so as to combine the required 
 hardness and toughness. The chilled part should not break suddenly 
 to grey, as is sometimes seen in some superior brands of Swedish iron, 
 but should shade off gradually. The cost is estimated at about £30 
 per ton. It is thought that the kind of metal required would be easily 
 procured in most countries where iron fortifications are likely to be 
 erected. Fig. 6, herewith, exhibits the general form taken by Griison's 
 
 Fig. 6. 
 
 1 This information the writer obtained incidentally when abroad, but it was verified for him by 
 the kindness of the French Naval Attached 
 » Vide pp. 96-97, I., for Krupp's experiment on a chilled-iron shield made by him. 
 
armour in a turret, or cupola as it is termed. The turret may be made 
 to revolve by hand, the men working on the levers beneath. In some 
 of the earlier examples central pivots were employed ; but it is much 
 better to dispense with them. The projectiles and cartridges are lifted 
 from beneath, the guns being in all cases breech-loaders. It may be 
 seen indeed that breech-loading guns are specially suited to this kind 
 of armour. As with breech-loading, men are not required to stand 
 upright at the muzzle of the gun, but only near the breech, the turret 
 may be brought back in a gradual curve to the height required, so as 
 to be dome-shaped, which saves metal, and adds 1 greatly to the 
 strength of the structure, and to its power to throw off shot. Each 
 turret generally contains two guns. Pig. 7, herewith, exhibits the 
 structure of a battery in course of construction. The form of the 
 
 Pig. 7. 
 
 
 mw^ 
 
 
 ■lis* 
 
 
 mm 
 
 battery is such as to give the armour the advantage of the curve of a 
 cupola as far as practicable. It will be seen that the Griison form of 
 cupola and battery alike give a great advantage in strength of roof, 
 which can be made strong enough to resist vertical fire. This is an 
 important matter, inasmuch as chilled-armour, especially in the cupola 
 form, is likely to be employed in those detached forts which play an 
 important part in detaining an enemy, but which must eventually 
 
 1 The Author thinks that anyone looking at the St. Marie Battery on the right when ascending 
 the Scheldt to Antwerp, will judge that it would be very difficult for a gun to make anything like a 
 fair hit. 
 
succumb to a regular attack, without being able, in most cases, to 
 withdraw the armament or even the men. 
 
 Lieut. -Colonel Kromhout, of the Dutch Engineers, wrote a paper 
 advocating the employment of cupolas for their "forts d'arret," as they 
 are termed. He considers that two guns in cupolas may take the 
 place of six guns in a fort, because it is rarely indeed that six guns 
 directed in converging lines would be required to fire at the same time, 
 and that two cupolas, by their power to revolve, could probably perform 
 the same work, while they have the advantage of obtaining mutual 
 cover. The saving of four pieces is not merely a matter of economy : 
 it diminishes the prize that may eventually fall into the enemy's hands, 
 not merely by the actual guns, but also the corresponding detach- 
 ments and men and stores. Approaches, bridges, coast defences and 
 the detached positions about Amsterdam, Colonel Kromhout thinks 
 cupolas are specially suited to defend. He calculates that the roof 
 might be made thick enough, not only to resist siege pieces hitherto 
 employed for vertical fire, but also a special mortar of Krupp's of 
 8"27 ins. (21 cm ) calibre, which fires a shell weighing 198 lbs., with a 
 velocity of 1000 feet per second. Provision would be made for 
 resisting such a projectile, even if its falling velocity were equal to its 
 initial, by a roof of 5 - 9 ins. (15 cm ) thick. 
 
 In Fig. 7, (p. 7, II.), the general character of the work may be seen. 
 The segments are lifted by cranes, and lowered into their places, 
 after which a mixture of zinc and lead, in all the earlier forts, 
 has been run in between the joints. This fills up the recessed space 
 seen on the ends of each portion, and gives the required solidity to 
 the entire structure, supplying a method of union which is not liable 
 to produce langridge. 
 
 Fig. 8 shows a line of cupolas erected for the Prussian Government, 
 
 Fig. 8. 
 
9 
 
 on the principal above referred to. In Germany, at the entrance of 
 the river Weser near Bremerhafen, forts were erected with cupolas, 
 one 6 and another 4, having one or two B. L. guns in each — also a 
 Gruson battery with 9 guns. 
 
 Fig. 9 shows the interior of the St. Marie Battery erected for the 
 
 Fig. 9. 
 
 defence of Antwerp. General Brialmont recommended the employ- 
 ment of chilled-iron at Malines as well as here. The armour of the 
 St. Marie Battery has a maximum thickness at the port holes of 
 27-56 ins. (70 cm ) and a minimum of 14-96 in. (38 cm ) the covering 
 varying from 13*78 to 7'87 ins. (35 to 20 cm ) thick. The ammunition 
 is stored under the casemate. The total weight of the structure is 
 800 tons. A port-hole plate weighs 35 tons, a pillar 25, and a covering 
 plate 21'5 tons. These plates were sent to Antwerp by railway in 
 special carriages. A revolving crane was used to put them on board 
 ship which brought them by Scheldt to the fort. They were then 
 carried and placed in position by means of rails and travelling and 
 steam cranes. 1 
 
 In some instances castings up to 50 tons weight have been employed. 
 On the other hand, in the Italian Passes, it was stipulated that the 
 maximum weight should be 13 tons. 
 
 An experiment was conducted on the 2 ground of Messrs. Gruson, at 
 Buckau, on October 22nd, 1883, against a chilled-iron shield. The 
 
 1 Officers have obtained leave to see the St. Marie Battery by applying to their Nation Ministers 
 at Brussels. The Author obtained permission in this way, but was prevented by an accident 
 from seeing the interior of the battery. 
 
 ' Vide original report of (Unison, Engineer, February 22, 1884, p. 141, also translation of 
 original report by Captain W. H. Bixby, V. S. Engineers. 
 
 24 
 
10 
 
 gun employed was a Krupp 30-5 cm (12 ins.) piece, 25 calibres long, 
 mounted on a Griison carriage. It fired a steel projectile of Krupp's, 
 3 - 5 calibres long, empty or blind, weighing 415 kilogs. (981 lbs.), 
 the charge being 120 kilogs. (204^ lbs). The velocity was 445 
 metres (14t>0 ft.), arid the striking energy or vis viva 4490 metre- 
 tons (14,498 foot-inns). The shield consisted of five pieces of chilled 
 iron forming half a tower — Pig. 1, p. 11, II. The rear or open part was 
 supported by means of masonry piers to which it was connected — 
 Fig. 1, p. 11, II. The entire weight of the shield was 47 - 5 tons. The 
 first round struck a spot 4 cm to the right of the centre of the shield, 
 and 90 cm from the bottom— Figs. 3 and 4, p. 11, II. The shot struck 
 the plate in a normal direction at the point marked I on Fig. 4, 
 forming a crack a. To the left of the point of impact a chipping off, 
 or bruise. Round the point of impact a chipping or bruise of a 
 maximum depth of 35 mm . The point of the shot itself was flattened 
 out into a disc with a centre core which stuck to the shield. On the 
 interior, the crack a was visible as a hair crack, running from a point 
 45 cm from the left edge to about 22 cm of the right edge, Fig. 2, p. 11, 
 II., in which, however, the crack is shown as completed by subsequent 
 firing. Other effects of a slight nature were visible. 
 
 The second round was fired with a steel projectile weighing 446 - 3 
 kilogs. (98:)'9 lb.). It struck a point on the centre vertical line of the 
 shield 100' m over the first point of impact, II., in Figs. 3 and 4, p. 11, 
 II., the axis of the shot having a striking angle of 51 degrees. 
 A long curved crack c was found running up to within 50 cm of the top 
 edge of the shield, and 17"" of the left edge, Fig 4, also another crack 
 c running to the right in a sharper curve nearly to the plate edge. A 
 vertical crack d connected points of impact I. and II. together. It 
 could not be traced beyond 1. There were also four other hair cracks. 
 A long chipping off also was made of 70 mm maximum depth, Fig. 4, 
 p. 11,11. 
 
 In the interior, the crack a was completed and opened to a width of 
 1™. The cracks c and d were not visible in the interior. The 
 horizontal plate was slightly moved back. 
 
 The third round had a projectile of 443'8 kilogs. (978 - 4 lb.) in 
 weight. This struck a point, see III., Fig. 4, p. 11, II., 86 cm from shot 
 I, and 90 cm from shot II. Striking angle, 72 degrees. Eight cracks 
 were formed radiating from the point of impact, Fig. 4, one marked e, 
 united III. and I. ; another, /, ran parallel to b for a short distance. 
 Crack c opened. The chipping-off or bruise at the point of impact 
 was about 50 mm deep. Inside cracks connecting the a and b were 
 developed in the manner shown in Fig. 2, below the point of impact 
 III. A certain amount of general disturbance of parts also was 
 effected. 
 
 The fourth round was with a shot weighing 444 - 6 kilogs. (980 - 2 lb.) 
 The striking angle was about 75 degrees. The point of impact was 
 65 cm from the bottom edge and 84 cm from I {vide Fig. 4, p. 11, II.). 
 This projectile, striking a fragment which was already detached from 
 the rest of the shield, bodily moved it back, and to a certain extent 
 dislodged and disarranged the entire structure. 
 
11 
 
12 
 
 The shield received three blows, having a total vis viva or energy of 
 13,470 metre-tons (43,495 foot-tons), or 283-8 metre-tons (916 foot- 
 tons) per ton of metal in the shield without the fragments of the 
 shield being dislodged or the protection to the interior being lost. 
 After the third round there were cracks extending entirely across the 
 plate and through the whole thickness forming large fragments, but 
 these fragments remained in situ ; the wrinkles or lines formed in the 
 surface of the cast in no way contributed to the formation of the cracks. 
 The surface of the metal was chipped off round the points of impact, 
 but in no case had any of the point of the shot entered more than 
 2 - 6 ins. (7 cm ). After the fourth shot the pieces might still have held 
 in their places had the supporting structure stood better ; but as it was, 
 the whole of the energy of the fourth shot was available for the 
 removal of the detached fragment which it struck. In all instances, 
 the steel projectiles broke up on impact. 
 
 On this experiment, Von Shiitz, in an article in the Neue Militiirische 
 Blatter, observes that " evidently a cupola is not necessarily put out of 
 action or prevented from revolving by the shattering of one of its 
 plates." He considers that it is quite unnecessary to apologize for the 
 fracture of the turret at the 4th blow ; it had done its duty fully in 
 resisting the first three blows. Nevertheless it ought to be pointed 
 out that when erected for service, a revolving cupola is mounted on a 
 strong wrought-iron frame, which, owing to its construction has con- 
 siderable elasticity, and would absorb a considerable part of the shock 
 of impact. At Buckau where no frame existed, the cracks opened 
 and the plates moved in a way that can only be explained by this 
 fact. Further the fact that a half cupola only was employed, 
 deprived the structure of many of the advantages of its arched form. 
 There was altogether, however, less spare resisting power than might 
 have been supposed from the Tegel results. It might be objected that 
 if the points of impact in this trial had been closer together a greater 
 result might have been produced, but this is not supported by 
 experience. Some Military authorities on the contrary hold that blows 
 on points one or two calibres apart, which become connected by cracks, 
 furnish the most effectual means of destroying chilled-iron. Von 
 Shiitz then remarks on the notable results obtained in perforating 
 wrought-iron by Krupp's and others, declares compound plates to be 
 no better, 1 and expresses a preference for chilled-iron shields, 
 particularly on the ground of expense. He considers that while 
 against Krupp's steel projectiles the resistance has not been so great 
 as might have been expected from the Tegel trials, it is more than 
 sufficient, and far superior to that of other systems. Moreover, the 
 Buckau experiments have yielded important indications bearing on the 
 perfecting of cupolas. The section of the first cupolas was designed 
 to make the angle of impact as small as possible, but it turned out 
 that chilled projectiles caused more havoc at oblique angles, hence the 
 manufacturers made the profile less inclined, so as to increase the angle 
 
 l " Des experiences nombreuses ont demontre que les plaques-compound ne fournissent pas de 
 meilleurs rcsultats." 
 
13 
 
 of impact. But that which was an advantage against chilled projectiles 
 i» the opposite when steel are used, and hence the profile must be now 
 modified so as to resemble that of the early cupolas. In conclusion, 
 Von Shiitz pronounces iron necessary for coast defence, especially 
 against heavy guns, and chilled-iron he considers the best description, 
 observing that whatever changes in form may be necessary, chilled- 
 iron lends itself to them without difficulty, and is reliable under all 
 circumstances. 
 
 Some experiments 1 wera made on August 28th, 1884, against a 
 shield 20 inches thick (52 cm ) and about 14-66 tons weight, (14,900 k «), 
 with a 5-9-in. (15cm) g nn fi r i n g a s t ee l shell weighing 83-8 lbs. (38 kg ), 
 striking with 1114 foot-tons energy (345 mt ), that is apparently with 
 about 1385 feet velocity. The shield bore nine of such blows well. 
 
 A trial against a side plate of a turret of Griison's chilled cast-iron, 
 constructed for two 12™ (4"7-in.) guns took place at Buckau, January 
 19th and 20th, 1885. 
 
 The object and programme of the experiment was to test the shield 
 by twenty rounds of the Prussian 15 cm (5'9-in.) gun firing hardened 
 steel shells (Ternitz) ; charge, 6'9 kilogs. (15'2 lb.) ; prismatic powder, 
 " const. 68 ; " that is to give velocity equivalent to that at 1000 metres 
 range (1094 yards). All the blows were delivered against the left half 
 of the plate. 
 
 The plate was sought to be divided by the first five blows, rounds 
 one to five, in two nearly equal parts, in order to attack the left half 
 only, and in a way free from objects. Five projectiles out of the 
 twenty to have flat heads. If the plate after fifteen blows, i.e., ten 
 per square metre of plate's vertical projection, should not be breached, 
 and its interior surface not exhibit cracks dangerous to the gun 
 detachments, the resistance should be considered sufficient. After 
 this five more rounds should complete the experiment. 
 
 The form experimented on differs materially from that of previous 
 shields, its construction being based on the results of former ex- 
 periments. The profile is debased or flattened considerably, so as to 
 avoid an angle of impact exceeding 46£ degrees, from a shell striking 
 horizontally. The plate was fixed between two other side and one 
 roof plate, so as to form nearly a half cupola. At the open side it 
 was supported by pillars of masonry by means of intermediate iron 
 coupling plates, the whole being protected from shell fire by earth and 
 wood. 
 
 Fig. 1, p. 1 4, II., gives dimensions and profile of shield. The greatest 
 width measuring round the curve — " developpee " — was 3'8 m (12 ft. 
 1'7 in.) ; that at the top edge was 2'15 m (7 ft. 0'6-in.) The weight was 
 19,918 kilogs. (19 tons 12 cwt. qrs. 7 lbs.) The Prussian 15™ (5'9-in.) 
 gun was mounted in position to deliver seven blows opposite to the 
 centre of the plate at 36 m (118"1 ft.) range, for the remaining rounds 
 at 24 degrees to the left. The projectiles employed were Ternitz hard 
 tempered steel shells filled with sand and made up to 34 - 5 kilogs. (76 - 06 
 
 1 The author ia indebted to Herr Griison for kindly forwarding to him the printed reports on 
 these 1883-84 trials, in reply to his request for information on them. 
 
!■!■ 
 
 lb. weight). The charge was 6'9 kilogs. (15 - 2 lb.) prismatic powder as 
 
 above stated. The initial velocity was 395 ra (1296 ft.) 
 
 The first seven projectiles struck at angles of impact varying from 
 46 degs. 15 mins. to 25 clegs. 56 mins. They all glanced and broke 
 up. The hits are shown in Fig. 2, herewith, and Pig. 3, p. 15, II. Hair 
 cracks were formed in the face of the shield, but nothing was visible 
 at the back of the shield. 
 
15 
 
 On January 20th 1885, 1 the gun was fired from a point 24 degs. to 
 the left of the original position, 36 m (118 - l-ft.) rangej that is, the 
 lines from axis of shield to gun in the two positions apparently formed 
 radii of a circle 24 degs. apart, so as to give lines of fire normal to the 
 horizontal section of the shield. The next thirteen rounds struck at 
 angles of impact from 46 degs. 28 mins. to 26 degs. 22 mins. All the 
 
 1 Tide article in Engineer based on reports obtained from Gruson, see Engineer of May 8, 1886. 
 
16 
 
 projectiles broke up and all glanced except No. 16 round, which was a 
 flat-headed projectile. Its head remained fixed in the plate. \ Rounds 3, 
 10, 11, 14 and 16 were flat or rather slightly concave-ended projectiles. 
 The plate in all has borne 20 blows of steel shells, each 274"6 
 metre-tons (8867 foot-tons) or 276 metre-tons (891 foot-tons) per 
 ton of entire shield, or 552 metre-tons (1 782 foot-tons) per ton of 
 half shield attacked, or 1425-6 if the shield be reckoned up to the next 
 white line, so as to include all the blows fairly, which certainly should 
 be to obtain any idea of effect, without destroying its powers of 
 resistance. The crack d, after the removal of the front shield, is 
 perceived to extend under the surface without reaching the edge of the 
 plate, so that the portion affected by it is not detached from the shield. 
 The effect of the new flattened profile is shown to be very good, all 
 the projectiles being thrown upwards. It is true that the flat-headed 
 shells have had more effect than the others — their points of impact 
 are marked on Fig. 2 with an asterisk * 1 — but they have not been able 
 to destroy the plate. The shield has greatly exceeded the resisting 
 power demanded of it, against the fifteen rounds. It is impossible to 
 say, even approximately, the number of blows necessary to break the 
 shield. The Ternitz steel shells equal the Krupp steel shells in 
 tenacity and hardness. With a fractured point of the shell it is 
 possible to scratch glass, just as with the Krupp steel. On the ground 
 of this result some advocated the application of Griison armour to 
 inland forts and lines of defence. 
 
 1 If tho depths of penetration of each projectile be divided by the sine of the angle of 
 incidence, so as to bring them to the penetration which would be obtained in the normal or direct 
 direction, that is, supposing thej can for the moment be treated like penetrations into wrought- 
 iron, the mtan of the penetrations of the pointed projectiles is 0"469 inches, and that of the four 
 uat-headed ones thus (excluding the one lodged) is 1*936 inches. The author sees all sorts of 
 objections to this method of comparison, but it appears to be the only one available. The fact 
 that the flst-headed Bhot obtained more than four times the penetration of the pointed ones, appears 
 to be a more marked superiority than can be accounted for by accident. 
 
17 
 
 CHAPTER II. 
 
 Trial ot Gruson's Armour at Spizia. 
 
 A trial of Gruson's chilled armour, on a larger scale than had hitherto 
 been attempted, was commenced at Spezia on Tuesday, April 20th, 
 1886. Before describing it, a few words of introduction may be 
 desirable. 
 
 The Italian Government, having decided to erect two cupolas, each 
 mounting two 119-ton Krupp guns, for the defence of Spezia, Herr 
 Griison was invited to construct them on his system, on the condition 
 that he should make a test shield capable of bearing three blows from 
 the projectile of the Armstrong 100-ton breech-loading gun, without 
 any portion being detached from the interior. These blows were not 
 to be within a metre of each other. A test shield was erected at 
 Spezia, and it received its first blow on Tuesday, April 20th. 
 
 This experiment was as completely opposite in character to that 
 made with the special Schumann- Griison cupola at Bucharest as can 
 well be conceived. The latter was the attack of armour by com- 
 paratively small siege guns, whose fire was continued day after day 
 from a short distance, and accurately directed. In fact, it was the 
 regular breaching of a siege battery calling for peculiar powers of 
 endurance. Soft armour is specially suited to resist a regular breaching 
 attack, and both the structures tested at Bucharest were composed of 
 soft wrought-iron, although one had a hard steel face. 1 That trial was 
 specially valuable to England, as being a test of the two kinds of armour 
 adopted by us. It was, therefore, instructive with regard to our 
 powers of defence. 
 
 The exact opposite of all this may be said of this Spezia-Griison 
 trial. Coast forts must expect to be attacked by the heaviest guns 
 existing, but it may reasonably be hoped that very few blows will be 
 actually delivered on their sides, for there is no likelihood of a ship 
 being able to remain opposite a fort many hours, much less day after 
 day. The ship's fire will consequently neither have the accuracy of 
 siege fire, which is due to a short range and a fixed position, nor will 
 it be very long sustained. The power to resist a few heavy blows is 
 therefore what is called for. This power Griison's armour of chilled- 
 iron admirably supplies. It forms the hardest shield possible. The 
 metal transmitting the shock through its mass, it is impossible to 
 injure anything covered by it, until the shield is broken and displaced. 
 A blow greatly outmatching the resisting power of the shield may 
 cause wholesale fracture, and leave a battery much more exposed to 
 future fire than would be the case with soft armour, which, under such 
 conditions, would let the projectile through it. In the case of Griison's 
 armour the blow performs the maximum work on the shield, and the 
 
 i This trial was witnessed by the author at Spezia, and reported in Engineer of April 30, 
 May 7, and May 14, 1886. 
 
 25 
 
18 
 
 minimum on the battery behind it, and in the case of soft armour a 
 shot performs the minimum of work on the shield and the maximum 
 on the battery behind it. Consequently, it may be seen, without further 
 comparison, that Griison's shield is well suited to resist a few heavy 
 blows, and is therefore well adapted to coast defence. It may generally 
 be assumed that so long as a Gruson's shield stands up in front of a 
 gun, that gun is safe against the next blow, even should that blow 
 greatly out-match the shield. And the safety of the detachment is 
 secured in a peculiar degree by the fact that there are no bolt-heads 
 to fly off, and no langridge until the shield is broken up. The 
 experiment should be peculiarly interesting to the British Government, 
 as regards offensive operations. Seeing that foreign coast armoured 
 forts are made almost without exception of chilled-iron, it follows that 
 our ships can never engage land shields of any other kind. The 
 attack of armoured coast forts is no doubt a serious operation, and is 
 only justified by strong reasons; but surely such a possibility must 
 be contemplated, or why should coast forts be built ? Certainly 
 nothing but ships can attack them. 
 
 It might appear, perhaps, that enough has been said to show the 
 importance of the trial at Spezia, but the special reason, as concerns 
 England, yet remains. English compound armour, in spite of its hard 
 skin, does not call for the same qualities in a projectile that appear to 
 be needed for the best attacks of chilled-iron. As before pointed out, 
 our chilled-iron service shot, which are of their kind excellent, have 
 exhibited such powers against steel-faced plates, that there is con- 
 siderable danger of overlooking the fact that such projectiles are 
 probably almost useless against chilled-iron. Certainly no foreign 
 chilled-iron shot has been found to be of much avail against it. The 
 shot meets with an abrupt shock directly its point touches the shield. 
 The best steel flies into pieces, but in the act of doing so, its tenacity 
 is sufficient to enable it to deliver a great shock on the shield. A 
 chilled-iron shot has little tenacity, and it is caught at a disadvantage 
 when resisted without the chance of getting its point any sensible 
 distance into the plate. Hence it breaks up without delivering a 
 blow at all proportionate to its energy. A blow that comes after 
 fracture of shot in the form of a mass of fragments, is, of course, of 
 no use against armour. Consequently, as before noticed, chilled-iron 
 armour, when tested in Prance by chilled shot only, was approved 
 for inland defences, but when afterwards it was attacked by steel 
 projectiles in 1882, its use was restricted to coast defence. Griison 
 then, as already seen, altered the profile of his shields, the better to 
 meet the attack of steel projectiles ; and the test of the shield which, 
 with ordinary steel projectiles is formidable, became much more so 
 now, because, since the time when the terms were accepted by Griison, 
 great improvements had been effected in the manufacture of steel 
 projectiles by Krupp, and consequently those now supplied by him for 
 this purpose were peculiarly formidable. 
 
 The Armstrong 100-ton B. L. gun is probably about as powerful as 
 the Krupp 1 19-ton gun. This should be borne in mind, because, while 
 many are aware that the Armstrong 110-ton gun — i.e., the Benbow 
 
19 
 
 gun — is by far the most powerful gun in the world, probably 
 few would expect that the 100-ton B. L. gun, made as long ago as 
 1882, would be as powerful as Krupp's 119-ton gun, a piece which, in 
 April, 1886, was not yet delivered in Italy, and could only be judged of 
 by data furnished to the above effect. 
 
 Herr Gruson's shield is shown in Pigs. 1, 2, and 11, p. 24, II., which 
 
 SCALE OF FEE! 
 10 IS 
 
 FIC.2 
 FRONT ELEVATION 
 
 FACE CHIPPED 
 / 
 
 CRACK AT 
 WIDEST ABOUT 
 
 Ys INCH 
 
 \ 
 \ 
 
 PLANE OF 
 FRACTURE 
 
 FIC.4 
 
20 
 
 exhibit it in section, front elevation, and plan. The chilled portion, 
 which forms the main portion and receives the blow, weighs about 
 87,950 kg., or 86"56 tons, and is of the shape shown in Fig. 7, p. 22, II. 
 In the complete turret there are to be twelve plates similar to this, and 
 three broken by the two gun ports. The interior diameter of the 
 cupola is to be about 10 metres. The shield is chilled white on its 
 exterior face. The interior and other pieces of iron are mottled. 
 The trial plate is fixed between two cheeks of iron, which are made 
 large enough to obtain a good bearing on the masonry at each side, 
 and thus to put the shield in as nearly the same condition as possible 
 as that in which it would be placed in an actual turret. The recesses 
 at the end of the plate are now made to fit, white metal not being 
 necessary to key them together. 
 
 The firing took place in St. Maria Bay, in the Gulf of Spezia, where 
 the shield was erected facing the sea. This shield being intended for 
 a land fort, the experiment was conducted by General Giovanetti, and 
 the Commission appointed for this branch of the work. The gun 
 itself was in the hands of the Navy. 
 
 The projectile was intended to strike the shield in the same way as 
 it would if the path of the shot were inclined downwards at an angle 
 of 1°, and the shield standing on a horizontal base. For this purpose, 
 as the gun fired slightly upwards, the shield was, as it were, tilted 
 very slightly forwards ; that is to say, forward to the extent of 1£ deg. 
 in comparison to its position set on a truly horizontal base, 1 deg. for 
 the supposed descending angle, and ^ deg for the difference in level 
 of shield and gun. From what has been said as to the construction of 
 the complete turret, it may be seen that the chilled plate, actually 
 receiving the blows, was the only part of the target which corresponded 
 to the turret. All other parts were substitutes. For example, the 
 two large iron side pieces served to give a bearing on the masonry 
 nearly corresponding to the support which would be afforded by the 
 contiguous portions of the turret, which should be smaller in area but 
 more rigid than masonry of the same extent. The mottled iron piece 
 at the top of the shield did duty for the circular crown of the turret, 
 while the mottled piece at the base took the place of the plate forming 
 the base of the turret. Even the small wrought-iron packing plates 
 in the joints are substitutes for the filling in of the joint by white 
 metal or by keys fitting to the front plate. Lastly, the masonry glacis 
 took the place of the glacis with chilled-iron " vorpanzer," shown in 
 the approximate cross section of the complete cupola, Fig. 12, p. 33, II., 
 which, like the plan, Fig. 12, p. 33, II., is drawn from description, not 
 copied from any drawing, and is therefore only approximate. 
 
 The first round was aimed at a point marked by — ^ 1, Fig. 2, p. 19, II. 
 The gun, as in previous Spezia trials, was fired from a raft on the sea. 
 It was therefore subject to a little movement, and the shot actually 
 struck at the point of impact shown by the bruise, with cracks radiating 
 from it, in Fig. 2, p. 19, II., and by the arrow in Fig. 1, p. 19, II. The 
 projectile broke into small fragments, which were thrown upwards. 
 The angle of incidence was 4Q degs. The point did not hold, and a 
 shallow scoop was formed with front cracks radiating as shown in Fig. 
 
21 
 
 2, p. 19, II., by the letters a, b, c, d. Crack a was perhaps £-inch 
 wide near the point of impact ; crack d was considerably narrower ; and 
 b and o were hair cracks. Inside the shield is seen a fine crack running 
 along opposite to the dotted line shown a a in Fig. 2, p. 19, II. — that 
 is, across one foot of the shield shown in Fig. 4, p. 19, II., which 
 exhibits the portion of the plate which may be separated from the 
 rest of the shield if the front cracks a and c, and the rear cracks a a 
 are one and the same. "Whether this was the case or not, the shield 
 bore the blow very well, considering the magnitude of it, and the 
 character of the projectile. The striking energy was 47,499 foot-tons ; 
 the weight of the projectile was 1000 kg. or 2204"6 lbs. (nearly a ton) ; 
 the striking velocity was 537"2 metres (1762-5 feet) ; the calculated 
 perforation through wrought-iron was 31*2 inches; the energy per ton 
 of shield was 548 - 9 tons; the charge, 375 kilogs., or 826 - 7 lbs., Cologne 
 prismatic cocoa powder; the range was about 1 33*7 metres (438 - 7 
 feet). On service such a blow as this could scarcely be given, the 
 range being specially short. On Saturday, April 24th, the firing was 
 continued. The gun on its raft was again brought into position at 
 a range from the shield of 133-7 metres, or 438-7 ft. 
 
 On April 24th, the gun was fired under conditions as exactly similar 
 to those existing on April 20th as possible. The striking velocity 
 taken was 537 - 9 metres (1764-8 ft.) the striking energy being 14,747 
 metre-tons (47,329 foot-tons). The perforation through iron would be 
 31 "2 ins. This does not apply to a hard target which cannot be 
 perforated but must be broken. The blow, estimated on the principle 
 of shock in proportion to mass of shield, is 550" 7 foot-tous per ton of 
 shield. The projectile was as before, a Krupp steel hollow projectile 
 forged and hardened. The form and dimensions are shown in Figs. 
 13 and 14, p. 23, II. Tool marks were visible from base to point. The 
 projectile had, of course, been hardened subsequently to being tooled. 
 There was a screw base plug somewhat resembling our own. The 
 pressure in the gun was not taken. The projectile of the second 
 round struck a few inches to the right, and high of the point aimed 
 at — that is, it struck rather nearer to the point of impact of the first 
 round than was intended, with an angle of incidence of 44 degs. As 
 on the last occasion, the projectile was nearly entirely broken up into 
 small fragments. One piece, however, of about 56 lbs. weight was 
 found; it had formed part of the base end. It is shown in Fig. 15, 
 p. 23, II. The quality of the steel appeared to be excellent. It was 
 pretty hard throughout, though some metal in a softer condition than 
 the rest was said to have been found about the centre. 
 
 Fig. 16, p. 23, II., shows the section of the shield. 
 
 The effect on the shield is shown in Fig. 8, p. 23, II. As will be 
 seen the shot made a more serious indentation than before, the depth 
 of it being about 4 ins., while that of round 1 was only 2 ins. Several 
 cracks were made and opened in the plate (see 2, Fig. 8). Some 
 cracks, marked A, B, C, D, were opened as wide as 1^ ins., and in 
 some places the surface of the metal was chipped off. The wide 
 cracks being low down, and cracks generally extending downwards, it 
 is quite possible that these may come out at the bottom surface of the 
 
22 
 
 shield, so that they cannot be seen at the back. One iron side-piece 
 was broken through. Fig. 8, p. 23, II., shows the position of the cracks 
 in back t, y, and 8, and a small chip off at B. These are all near the 
 bottom, but it may be seen, if their course be considered (vide 
 dotted lines in Fig. 8, shown as if the plate were transparent) that it 
 is probable that e and 8 at back corresponding with d k and g in the 
 front. If a at back corresponds with a and c in front, it follows that 
 the junction of y and a may be identified with point of impact 1, and 
 of t and 8 with 2. One iron side piece was cracked, as shown in 
 
23 
 
24 
 
 u 
 
 N99S39 A1IO0T3A M T3 
 
 u 
 
25 
 
 Pig. 8, and the supporting masonry was a very little shaken, so that a 
 little space is opened behind the bearings of the shield. The plate 
 had now received a considerable shock. It must, however, be con- 
 sidered to have stood admirably. No shield had ever yet received two 
 such blows as this. The weight of the shield is, of course, great, so 
 that the striking energy per ton is not very large. It must, however, 
 be remembered that, owing to the excellence of the steel, a much 
 larger proportion of the striking energy of projectile is impressed on 
 the plate than usual. 
 
 The cracks at the back were of course the beginning of the splitting 
 up of the shield. Up to the present time anyone might have remained 
 inside the shield in complete safety, because there is no langridge of 
 any kind, the absence of bolts being a great advantage. It has been 
 mentioned as a condition that no portion should be detached on the 
 inside. It may be urged that this was done now for a small scale 
 close to the ground at B, Fig. 9, was detached. On the other hand, 
 the actual impressions of the projectiles points were less than a metre 
 apart. Practically neither of these conditions were broken. 
 
 A third round was fired at the Griison shield on Thursday, April 
 29th, from the Blswick 100-ton breech-loading gun, the firing 
 conditions being as nearly possible the same as in the two previous 
 rounds. The third Kriipp forged steel projectile was employed, its 
 weight being made up to 1000 kg., or 2204'6 lbs. The charge was 
 again 375 kg. (826"7 lbs.) of Cologne prismatic powder; the striking 
 velocity was 536'1 metres, or 1758"9 ft. per second. The striking 
 energy being therefore 14,651 metre-tons, or 47,306 foot-tons. 
 Pressures in the bore were registered as 2010, 1973, 1985, and 2025 
 atmospheres, the mean being 1998 atmospheres, or 13'11 tons per 
 square inch. The charge in each round was made up in four cartridges 
 ribbed longitudinally with serge rolls so as to make the charge lie in 
 the bore with a space round it. The crusher gauges, tubes, and 
 obturator, were, of course, as employed at Elswick. 
 
 After the second round, as already noticed, owing it is said to the 
 impossibility of masonry and cement supporting the iron shield as well 
 as the contiguous iron plates of a cupola or iron fort, a slight space 
 had opened behind the edge of the plate on either side. In fact there 
 had been sufficient yielding to cause a little anxiety as to the conditions 
 under which the shield would receive the third shot. 
 
 This was delivered on a spot close to the centre line running down 
 the shield and about a metre from Round 1, see Pig. 17, p. 26, II., on 
 which the cracks are shown in the front of the shield by continuous 
 lines, and on the back in dotted lines seen through as if the shield 
 were transparent. The angle of incidence being more oblique than 
 before, the indent of the projectile was less, being about 1^ inch. The 
 projectile flew into smaller pieces than before apparently, but the blow 
 was sufficient to crack the shield, which is much thinner here, in lines 
 shown in Pig. 17, p. 26, II., as m, n, o,p, q, r, s and t, as well as a small 
 crack connecting q with point of impact 1. The portion of plate 
 between t and the edge was entirely separated from the rest of the 
 shield, so that it could be removed. It extended to the depth of about 
 
 26 
 
26 
 
 FIG . |- 
 
 FRDNX OF SHIELD 
 DEVELOPED 
 
27 
 
 a foot at the plate edge, but rapidly curved up to the surface at t. In 
 spite of this cracking and splitting, however, the shield appeared not 
 only to be in condition to receive another blow, but Herr Griison 
 thought its position improved, inasmuch as it had so sprung as to 
 close the opening visible at the last round, and was therefore now 
 better supported. In the inside cracks -q, k, £, 6, and X, (see Pig. 18, 
 p. 26, II.) were formed, and two very small pieces were detached near A, 
 (vide Fig. 18), where the plate had obviously felt the blow severely, the 
 fragment in the centre marked p r projecting slightly, and the cracks 
 there being deep and opened. The following table gives the details as 
 to velocity and energy of each round more exactly and fully than we 
 have hitherto stated them : — 
 
 Ballistic Power op 100-ton Breech-Loading Armstrong Gun Firbd at Spmia 
 on April 20th, 24th, and 29th, 1886. 
 
 
 Number of round. 
 
 Pirst. 
 
 Second. 
 
 Third. 
 
 Velocity at muzzle, in metres 
 
 641-2 
 
 641-9 
 
 540-1 
 
 tf n feet 
 
 1775-6 
 
 1777-9 
 
 1772-0 
 
 n 85 metres, in metres 
 
 538-6 
 
 539-3 
 
 537-5 
 
 i, 278-9 feet, in feet 
 
 1767-1 
 
 1789-4 
 
 1763-5 
 
 n striking, in metres 
 
 537-2 
 
 537-9 
 
 536-1 
 
 n n feet ... 
 
 1762-5 
 
 1764-8 
 
 1758-9 
 
 Energy at muzzle, metre-tons 
 
 14,929 
 
 14,966 
 
 14,871 
 
 ii „ foot-tons 
 
 48,207 
 
 48,326 
 
 48,019 
 
 n striking in metre-tons 
 
 14,709 
 
 14,747 
 
 14,651 
 
 i, ii foot-tons 
 
 47,499 
 
 47,629 
 
 47,306 
 
 n n per inch circumference foot-tons 
 
 892-8 
 
 896-9 
 
 888-6 
 
 Perforation in wrought-iron at striking inches 
 
 31-2 
 
 31-2 
 
 31-1 
 
 Angle of incidence of projectile with tangent to shield 
 
 40 degs. 
 
 44 degs. 
 
 34£ degs. 
 
 Energy in foot-tons at impact per ton of shield 
 
 548-9 
 
 550-1 
 
 546-4 
 
 On the whole the shield had acquitted itself admirably, for it had 
 borne the three blows without any fragment of any importance being 
 dislodged in the inside. The small pieces, from parts shown shaded, 
 which came off, appear to have been shaken and dropped down rather 
 than flown out, and a detachment of men behind this shield would 
 have remained uninjured. Bolt heads generally fly with violence, 
 because their fracture is due to a strain of the bolt, which causes them 
 to spring into the interior. This does not seem to be the case with 
 pieces that may be dislodged by cracking. At all events it must be 
 admitted that the defensive power of a shield which resists three blows 
 of the projectile of the most powerful gun in existence is remarkable. 
 
We may add that such a judgment was expressed by the Commission, 
 that Herr Griison was able to telegraph in the afternoon to push 
 forward the manufacture of the plates required for the two cupolas. 
 
 The question was raised, as it was inevitable should be the case, as 
 to the excellence of the Krupp projectiles which had thus broken on 
 the shield. One or two queer looking bits were found, and persons 
 stated that they found pieces that proved rather soft under the file. 
 There was however nothing tangible to shake belief in the projectiles 
 being excellent ones. At all events, nothing could be a fairer proposal 
 than that which was made by Krupp's agent, M. Otto Budde. A 
 competitive trial of steel 15 cm projectile had been carried on at 
 Muggiana, Spezia. In this Krupp's had proved themselves the best. 
 Shells of 15 cm had been fired point blank at very thick steel plates — 
 about 18 ins. thick. These entered to a depth of about 1^ calibres, 
 and rebounded without alteration of form, so that they could be again 
 fired. M. Budde suggested that these projectiles, whose quality had 
 been thus proved to be extraordinary, should be fired at the Griison 
 shield, against which, striking obliquely as they must do, he stated that 
 they would break up exactly in the same way as the large projectiles. 
 
 There was a wish expressed to see a French St. Chamond steel 
 projectile fired against the shield. These are excellent, but they were 
 said to be softer than Krupp's, and the question arises as to what are 
 the qualities that are most needed in short to enable them to act well 
 against chilled-iron armour. On this matter we have literally no 
 experience in England. Krupp's belief appears to be that the blow 
 delivered by the shot depends on its limit of elasticity. Not that the 
 projectile has the opportunity of recovering its form when the strain 
 falls within this limit, but rather probably because no sensible change 
 of form really takes place within it, and because directly deformation 
 commences, the resistance of the shot decreases so rapidly that little 
 more work can be got out of it. It appears probable however that 
 the ultimate tenacity, as well as the limit of elasticity, would be 
 the measure of the shot's power in this case of oblique impact. In 
 direct impact undoubtedly, on deformation commencing, all penetration 
 comes quickly to an end, but where no penetration in the ordinary 
 sense can be effected, surely as long as the projectile holds together, 
 so long it impresses its energy on the shield at the point of impact. 
 As deformation commences, its power rapidly declines, but still, after 
 commencing an injury, any following up of the blow at the exact spot 
 acts in so telling a way that it may well be supposed that between the 
 limit of elasticity and that of ultimate tenacity a sensible amount of 
 work existed, and that the latter limit, as well as the former, should 
 be considered. Indeed, slight deformation may not destroy all per- 
 forating power. In the competitive shot trials (see pp. 74-77, I.), 
 Whitworth's projectiles, after being slightly set up, perforated plates 
 which proved more than a match for most shot. 
 
 With regard to St. Chamond projectiles and other French steel ones, 
 little is known out of France, but it is stated that at Gavres great 
 results have been obtained. It is expected that steel projectiles should 
 perforate steel plates about a calibre and a quarter thick without 
 
2<> 
 
 deformation. In this country it is commonly stated that French steel 
 plates are being made soft. This appears to be true in a measure ; 
 nevertheless, the French compound plates, whose powers we know are 
 not very different from our own made on the same patent, are per- 
 forated more easily than the French steel plates. After such proof of 
 the excellence of Krupp's steel projectiles, a good deal of evidence is 
 needed to make us accept the statement that the French were as good, 
 or better ; but it may be so. One thing was unfortunately quite 
 clear, namely, that England at this time had dropped far behind in the 
 matter of steel projectiles, and it may be feared that until really 
 hard armour should be tried in this country too much confidence may 
 continue to be placed in our chilled-iron shells. These are, of their 
 
 FIG.I 
 DEVELOPMENT OF SURFACE 
 
30 
 
 sort, excellent, possibly the best in the world; but in Prance, Germany, 
 and Italy, it has been found absolutely useless to fire any chilled 
 projectiles at really hard armour. Surely it would be mad to expect 
 that our chilled shot would be of much use against a Griison shield 
 unless we prove this to be the case by actual trial. A Griison shield 
 erected at Shoeburyness would teach us much as to our neighbours' 
 armour, and as to our projectiles. A shield is surely worth trying for 
 its own sake, which, so long as it holds together, offers complete 
 security to the battery behind it, and which can be made to keep out 
 three projectiles, any one of which would have gone clean through 
 
31 
 
 almost any armour we possess either on ships or forts. It is still 
 more important to try it for the sake of the knowledge it affords us of 
 the power of our own projectiles against foreign armour. It can 
 hardly he doubted by any one that we are far behind in the matter of 
 projectiles. French steel shot go through at least a calibre and a 
 quarter thickness of steel without deformation, and Krupp projectiles 
 enter a calibre and a half into steel armour of thickness totally dis- 
 proportionate to the shot, and bound back apparently undeformed and 
 uninjured. 
 
 On July 9th, 1886, a programme was drawn up to carry out the 
 suggestions made with a view of testing the excellence of the Krupp 
 projectiles which had been used in the trial, and consequently 
 establishing the estimate which had been formed of the resisting 
 power of the shield which had been subjected to the three blows from 
 these projectiles. The structure had been patched up, the masonry 
 having yielded slightly as well as the shield. The cracks in the latter 
 were opened with steel wedges, and filled up by running in zinc. 
 
 The first firing consisted in two rounds from al5™ (3'9-in.) Armstrong 
 gun, 28 calibres long, on an Albini carriage, which had been placed on 
 the raft with a 100-ton Armstrong breech-loading gun, Lepanto type, 
 drawn up as before at a range of 133'7 metres (438-7 ft.) In both 
 rounds with the 15™ gun the projectile was a Krupp hardened steel 
 shell, made up to 36 kilogs. (79'37 lbs.). 
 
 The programme consisted of two kinds of test. (1) Krupp 15 cm 
 projectiles (5'9-in.) should be fired taken from a batch whose excellence 
 had been established in the recent competition at Muggiano. This had 
 been suggested by M. Otto Budde, Krupp's representative. If these 
 projectiles should break up in the same way as those already fired from 
 the 100-ton gun, the natural inference would be that there was no ground 
 for supposing that the large projectiles were inferior in quality, at all 
 events the fact of their breaking up affords no such ground. It must 
 be conceded that it was most reasonable that Italian officers should 
 wish for a guarantee on this head, seeing that the difficulty of making 
 good steel projectiles increases with the scale on which they are made 
 to such an extent that it is desirable that those of any weight 
 approaching 1000 kilogs., or 2200 lbs., should establish their character 
 in every possible way. 
 
 The second test was the firing of a steel projectile supplied from 
 St. Chamond for the 100-ton breech-loading gun. This would furnish 
 a comparison between the large projectiles made in France and Germany. 
 Thus, supposing it were to be concluded that Krupp's large projectiles 
 had broken up more than his small ones, or had shown an inferior 
 fracture in any part, it would be possible to see whether the French 
 manufacturers had been more successful on this scale. 
 
 This trial, then, while it had a bearing on the resisting powers of 
 Griison's shield, did not touch the question of its acceptance. This 
 had been settled at the conclusion of the trial in April. 
 
 First round with 15 om gun (No. IV. in all).— Charge 15 kilogs. (33"06 
 lbs.) progressive Fossano powder (20-24 mm ) ; striking velocity, 500 
 
32 
 
 metres (1640-45 ft.) ; striking energy, 459 metre-tons (1482-1 foot-tons) . 
 The projectile struck 13™ above the edge of the avant cuirasse (glacis 
 plate) at 86 cm to the right of the centre of the shield, at a _ striking 
 angle of 44 deg. The projectile broke up, making a small chip in the 
 shield. 
 
 Second round (No. V. in all).— Charge 18 kilogs. (39"68 lbs.), 
 progressive Fossano powder (20-24 mm ) ; striking velocity, 564 metres 
 (1850-48 ft.) ; striking energy, 584 metre-tons (1885-6 foot-tons). The 
 projectile struck 23 cm above the edge of the glacis plate and 102 cm to 
 the right of the centre of the shield ; angle of incidence, 50 degs. 
 30 mins. The projectile broke up, and produced, about the part 
 marked V. (Fig. 1, p. 29, II.), a chipping off the surface. 
 
 The fourth round from the 100-ton gun (Round VI. in all) was now 
 fired. Charge, 375 kilogs. (826-7 lbs.). YVestphalian (cocoa powder); 
 projectile, hardened steel, St. Chamond ; calibres weight, 1000 kilogs. 
 (2204-6 lbs.) ; initial velocity, 539 metres ; striking velocity, 535 
 metres (1755-3 ft.) ; striking energy, 14,603 metre-tons (4715 - 4 foot- 
 tons.) The shot was aimed high up at the point marked with a cross 
 in Fig. 1, but the movement of the raft by the swell of the sea caused 
 the projectile to strike close to the spot struck by the second round 
 fired in April (see VI. on Fig. 1.) Owing to the injury the shield had 
 suffered already, the surface was struck nearly normally — that is, at 
 between 80 and 90 degs. The projectile broke up and dislodged 
 portions of plate of different thickness up to 50™. One crack waB 
 lengthened, and some other local injury effected in front (vide Pig. 1.) 
 At the back were two new cracks /x and v and also o, also a larger scale 
 between y and 8 was detached (vide Fief. 2), the small fragments of 
 which it was composed fell vertically, and would not have injured men 
 behind the shield. A small triangular portion above a projected about 
 gem rp^g i ower portion of the plate had given back and projected 
 beyond the upper part about 3 - 5 cm (l - 4-in.) along crack y, about 6™ 
 (2-4 in.) along crack t, and 4 cm (1'6-in.) along crack o. The left 
 buttress or shoulder was slightly displaced. The general condition of 
 the shield was good. In spite of the fracture of the left bearing plate 
 or cheek, and some displacement, the shield might have borne a 
 further attack ; but there were no more available projectiles for the 
 100-ton gun. 
 
 The object of this programme may be said to have been attained, as 
 far as the chief practical bearing of it for Italy is concerned. Clearly 
 all steel hitherto known must be expected to fly in pieces against 
 chilled-iron, whether the projectile be large or small. The St. Chamond 
 projectile, however, did not strike in such a way as to admit of a com- 
 parison of its effect with that of any of Krupp's projectiles. The 
 comparatively small effect produced by the French shot, when striking 
 exactly on the most injured spot in the shield, speaks well for the 
 resisting powers of the latter even when much cracked. 
 
 To pass on to the turrets which the shield represents. As before 
 said, there are the two turrets to protect Spezia harbour, each mounting 
 two 119-ton Krupp guns, whose exact power cannot be ascertained 
 certainly, seeing that there are different estimates, and that Herr Krupp 
 
33 
 
 not long since declined to give authentic information. They have been 
 fired at Meppen with results which are not made public. So far as 
 can be learned, the power of this gun is nearly the same as that of the 
 100-ton breech-loading Armstrong, but inferior to the 105-ton and 
 110-ton breech-loading Armstrong guns. 
 
 It may be well to point out one thing about which there seems 
 occasionally to be some confusion, namely, that while the attack of 
 armour may be a very telling illustration of the power of a gun, it is 
 no test of the gun, strictly speaking. The power of the gun is fully 
 told by the energy of the projectile when leaving the muzzle. What 
 happens after that concerns the projectile and the plate, but clearly 
 has no further connection with the gun itself. In this trial, for 
 example, the gun is not concerned, at all events directly, in whether 
 the shield is broken or not. This may be a question between Krupp's 
 projectile and Griison's shield, or more really between Griison's shield 
 and what might be expected from Schneider's, Brown's, Cammell's, 
 Marrell's, or Terni armour formed into some equivalent shield. Then, 
 again, the effect of a steel projectile as compared with a chilled-iron 
 one, when attacking chilled-iron armour, naturally arises in considering 
 
 Fig. 12. 
 
34 
 
 such a trial, and this is the especial point that interests us in England, 
 and in this Elswick may be interested as manufacturers of projectiles. 
 As gun makers they may be interested in the eventual consideration of 
 what guns may be able to effect against shields, but immediately and 
 directly their success as gunmakers is limited to the energy with which 
 they can discharge a projectile of a given calibre. 
 
 Fig. 12, p. 33, II., gives a rough plan and section — drawn from verbal 
 description — of the complete turret. It will be seen that there are 
 fifteen segment or sector-shaped shields, with two centre plates forming 
 the crown. The interior diameter is 10 metres — 32"8 feet. The 
 periphery is not a circle, but is formed of fifteen arcs of circles, each 
 struck with a radius of about 15 feet or 16 feet, giving an outline 
 suggestive of that of a pomegranate. Of the fifteen plates, twelve 
 are similar to the shield under trial, each weighing about 87,950 kg. or 
 about 86"56 tons. The remaining three are lighter, being pierced 
 by the gun ports ; the lightest being that between the ports. The 
 twelve unpierced plates will thus weigh about 1039 tons. The two 
 centre or crown pieces weigh together 130,000 kg., or 128 tons. The 
 total weight of the armour is 1,400,000 kg., or 1378 tons. This 
 leaves 211 tons for the three pierced plates, two of which will be 
 something over and the other something less than 70 tons. The 
 entire running weight of shield — 1378 tons — is to be supported on 
 an iron ring. 
 
 The machinery for working the turrets is supplied by Sir W. 
 Armstrong and Co. 
 
35 
 
 CHAPTER III. 
 
 Buchakest Cupola Competition. 1 
 
 BY 
 
 Majob D. D. T. O'Callaohan, E.A., and Captain G. S. Clabke, E.E. 
 
 The general idea of the defences of Bucharest appears to be to create 
 a place of arms in which the whole military strength of the kingdom 
 can be gathered. It is recognized that the military position of 
 Roumania is distinctly dependent ; but the absence of any permanent 
 defences practically prevents the choice of an ally. The Eastern 
 frontier is so open that Bucharest could be occupied hy a Russian army 
 in a few days. It is argued, therefore, that by the creation of a place 
 of arms, time would be gained to enable any power to give practical effect 
 to an alliance. Bucharest was selected as the centre of the wealth and 
 commerce of the country, as well as the junction of its railways. The 
 position possesses no natural advantages, and the opinion has been 
 expressed that a second place of arms should at least be formed on 
 the Kronstadt Railway among the southern spurs of the Carpathians. 
 
 The decision to defend Bucharest having been taken, General 
 Brialmont was asked to furnish designs and advise generally upon the 
 scheme to be adopted. 
 
 As far as could be gathered, the following are the principal features 
 of his proposals : — 
 
 The main defence is to consist of 18 large forts, at about 4000 metres 
 interval, supplemented by intermediate batteries. 
 
 The forts are to have earth scarps and a detached wall flanked by 
 counterscarp galleries with three embrasures — two firing in front and 
 one behind the wall. The counterscarps are to have arched counter- 
 forts. The detached wall is to be protected from projectiles grazing 
 the crest of the gkcis, with a descent of 1 in 3. On the flanks, it is 
 considered that the wall is sufficiently protected if it cannot be hit at 
 an angle exceeding 45 degrees. In front, there will be a second glacis, 
 covering wire entanglements and other obstacles. On the flanks, wing 
 batteries and rifle trenches will be provided. The parapets are to be 
 constructed of sand, with a layer of vegetable mould on the top, the 
 exterior slope being about f . 
 
 An interior reduit contains three turrets or cupolas, and a battery 
 for four howitzers. The latter are to have central pivots and hydraulic 
 
 1 Abbreviated Official Eeport, printed by permission in E. A. I. Proceedings, and here reproduced 
 by permission of Authors. 
 
 28 
 
36 
 
 buffers. Between each adjacent pair there is a hollow traverse, capable 
 of containing the howitzer. At the back of this chamber there is a 
 small ammunition store. 
 
 General Brialmont, in his work " La Fortification du temps present," 
 expressed an opinion in favour of turrets for land defence, and repro- 
 duced the designs of the St. Chamond firm, including a project for a 
 moveable shielded battery running on rails laid behind a parapet. It 
 was decided, therefore, that a competitive trial should take place 
 between the rival firms of Griison and St. Chamond, from which it 
 was expected that data leading to a definite decision would be obtained. 
 The total cost of the experiment was nearly £40,000 ; but as yet no 
 final decision as to the armament of the forts seems to have been 
 taken. For the armament of the intermediate batteries on the fronts 
 most liable to siege, rifled mortars fixed in a spherical mass, pivoting 
 under a horizontal shield, and thus presenting no open port, according 
 to a design of the Griison firm, appear to have been recommended. 
 As to the ditch-flanking armament no decision has been arrived at ; 
 but some form of machine or quick-firing gun will doubtless be adopted. 
 
 Three of the new forts only had been commenced at the time of our 
 visit, situated as follows : — 
 
 1. Chitella Fort on the road and railway leading to the Austrian 
 
 frontier at Kronstadt. 
 
 2. On the road to Mogosoie, 2 kilometres from Bucharest, and 
 
 due north of tbe " chaussee." 
 
 3. At Goulaqui, on the west of the town, between Chitella and 
 
 Cotroceni. 
 
 Of these, the two first named are the most advanced, the ditch 
 excavation being nearly completed. 
 
 Diaet op Experiments. 
 
 I8tk December. — Inspection of cupolas. Fired from both cupolas 
 with " projectiles fictives." 
 
 19th December. — Fired 10 salvoes for accuracy. 
 
 2lst December. — French turret fired 25 salvoes, and 22nd December, 
 German cupola fired 25 salvoes for rapidity and accuracy. 
 
 2%rd December. — Continuation of practice for accuracy. Each cupola 
 fired 10 rounds. 
 
 2Uh December. — Practice at a target suddenly presented. Each 
 cupola fired 3 Balvoes. 
 
 26tk December. — Practice at French turret with de Bange 155""" and 
 Krupp ] 5 cm guns ; 42 rounds fired. 
 
 21th December. — Practice at French turret continued ; 9 rounds fired. 
 Practice at German turret with de Bange 155 mm and Krupp 15 cm guns ; 
 27 rounds fired. 
 
 Praotice for accuracy from French turret ; 5 rounds fired- 
 
37 
 
 28th December. — Practice continued ; 58 rounds fired. Practice for 
 accuracy from German cupola ; 6 rounds fired. 
 
 29th, 30th, and 31st December. — Practice at cupolas with 21 cm Krupp 
 rifled mortar ; 14, 56 and 44 rounds respectively. 
 
 I st January. — Ditto ; 50 rounds fired. 
 
 2nd, 3rd, and 4th January. — Dismounting guns in cupolas, and their 
 replacement by dummies, and construction of siege battery at 50 
 metres. 
 
 bth January. — Practice at embrasures of both cupolas ; 11 rounds 
 fired. 
 
 1th January. — Practice at glacis plate of French turret. 20 rounds 
 fired. 
 
 8th January. — Practice at glacis plate of German cupola ; 21 rounds 
 fired. 
 
 l\ih January. — Practice for breaching German cupola; 50 rounds 
 fired. 
 
 14^ January. — Practice for breaching both cupolas; 36 rounds fired. 
 
 15th January. — Ditto ditto French turret ; 19 rounds fired. 
 
 16t/i January. — Guns remounted in cupolas. 
 
 17tk January. — Practice for accuracy and rapidity, German cupola; 
 20 salvoes fired. 
 
 20^/i January. — Practice for accuracy and rapidity, French turret ; 21 
 salvoes fired. 
 
 22nd January. — Practice against a wrought-iron plate of German 
 cupola ; 21 rounds fired. 
 
 23rd January. — Fired 7 rounds, filled and fuzed shell, from 21 cm 
 Krupp rifled mortar ; for effect in soil. 
 
 GEEMAN CUPOLA. 
 
 Geneeal Descbiption. 
 
 The pit or emplacement consists of a single circular chamber 19 fb. 
 6 ins. in diameter, with brick walls about 5 ft. thick. The floor level 
 is 3 ft. 6 ins. below the natural surface of the ground. The pit is 
 entered by steps through a door 5 ft. wide. There are 14 recesses, 
 3 ft. high by 2 ft. broad, and 18 inches deep, evenly spaced round the 
 interior wall. Each recess has one shelf and can hold altogether 
 24 projectiles, so that a total of 168 rounds for each gun can be 
 stored in the emplacement. 
 
 The glacis plates rest on the top of the wall which ends at the level 
 of their lower surface. Outside these glacis plates, there is a mass of 
 concrete 9 ft. thick, measured along the superior slope from their inner 
 
38 
 
 edge. The exterior slope of this concrete is 35 degrees, and in front 
 is a thick parapet of sand on the side which was attacked. The roller 
 path (a) and the gunmetal training arc {b) are fixed to the inner top 
 angle of the wall. The socket of the screw pivot (c) is bedded in 
 concrete, and there is a false floor boarded over except in the centre, 
 where the fixed internal toothed wheel of the traversing gear and the 
 pivot are left bare. 
 
 There was ample space for moving about freely in the emplacement, 
 and for the operations of loading and traversing. 
 
 Description The cupola consists of two armoured portions — (1) the glacis plates 
 set in concrete and forming a complete annulus round the pit ; (2) the 
 curved shield, free to turn on a central pivot, and covering the two 
 guns and mechanism. 
 
 Gisois. 1- The glacis, or avant cuirasse, consists of eight cast-iron blocks or 
 
 segments set in concrete, and formed as shown in Fig. 8, p. 39. 
 The joints are run with zinc. The two rear blocks which were to be 
 attacked, had central ribs in addition, apparently with a view to give 
 them greater strength. The total weight of the glacis ring is 70 tons, 
 and its internal diameter 5"93 metres (19 - 45 ft.) 
 
 2. The curved shield consists of six segmental side plates and a 
 hexagonal roof plate. The front plate which is pierced for the two 
 guns, as well as the plates on either side of it, are of wrought-iron. 
 The three remaining side plates are compound. The thickness of all 
 the plates is 20 cm (8 in.) The total weight of the shield is 46 tons. 
 The curvature of the plates is such that the angle of impact of a 
 projectile striking them horizontally can never exceed 25 degrees. 
 The portion of the front plate surrounding the two gun ports is 
 strengthened by swelling out the metal round the openings at the top 
 and sides, as indicated in the section, Fig. 8, p. 39. The wrought- 
 iron plates are bent by hydraulic machinery. The compound plates 
 are made on Mr. Wilson's principle by running the molten steel on to 
 their upper surface. The plate is made up of three qualities of metal 
 as follows: — 7 cm hard steel face, 3 cm soft steel, 10 cm wrought-iron. 
 The mass is then re-heated, bent by hydraulic power, and subsequently 
 annealed. The cost of the plates is materially enhanced by the great 
 waste of metal incurred by cutting off the corners in shaping them. 
 The plates so made are united to form the shield by being bolted to an 
 inner skin composed of two layers of wrought iron-plate, each - 8 in. 
 thick. These layers are rivetted together, and through them pass five 
 bolts of steel, the female screws of which enter half through the main 
 armour plates. The bolts are rivetted up from the inside. The side 
 armour plates are further held together at each junction by one steel 
 wedge of the form shown in sketch, and abut against the central 
 hexagonal plate, but are not keyed to it. 
 
 The most striking principle in the design of this cupola is the method 
 of supporting the shield. The whole weight of the moveable portion 
 
39 
 
 of the structure is taken on a central pivot (Fig. 8) 8 ins. in diameter. 
 The pivot is screwed into the socket piece and the collar is keyed on 
 to it. This collar is provided with a worm wheel on its outer circum- 
 ference, which engages an endless screw. By means of a ratchet lever, 
 the worm wheel can be turned, thus raising or lowering the pivot 
 and with it the whole superstructure. The total vertical movement Srang&i 
 which can be given is 6 ins., and this allows the shield to be lowered mentB - 
 down upon the glacis plates or elevated into the firing position. The 
 whole structure is thus free to rock on the rounded top of the pivot 
 but the lower edges of the shield are prevented from coming in 
 contact with the glacis by legs resting on trucks running on a circular 
 racer (r) and set in journal boxes provided with strong volute springs. 
 These springs thus serve to cushion the vibratory motion set up by 
 the force of recoil, or by the impact of projectiles. The maximum 
 movement measured at the top of the shield is stated to be only about 
 2 - 5 cm or 1 in., but appeared to be considerably greater. 
 
 The whole stress of recoil is transmitted to the general structure of 
 the turret by a pair of vertical arcs (A) in rear of the breech of each 
 gun. A ring (Pig. 8) is shrunk on to the chase at a short distance 
 from the muzzle, having trunnions which rest in recesses formed in the 
 inner skin at the sides of the ports. These recesses are open in rear, Kecoii. 
 so that they allow the trunnions to pivot freely, and at the same time, 
 
 F<C 8. SCHUMANN GRUSON CUPOLA 
 
 by moving them a few inches to the rear, they clear the recesses when 
 it is required to dismount the guns. In rear of the breech of each 
 gun is a powerful capstan-headed clamping screw, by which all play 
 between the guns and the arcs is taken up, so that hammering action 
 is avoided. 
 
 Quadrant elevation is given by graduations on these arcs. The 
 preponderance of each gun measured about the front pivot, is slightly 
 overbalanced by a counterpoise weight (I) suspended by two chains, 
 each passing over two pulleys, and attached to an iron cross-bar 
 passing under the gun just in front of the breech mechanism, fixed in 
 position by a steel strap. In order to give elevation, the breech is Ele ™tin g 
 hauled down by a chain wound round a ratchet windlass. 
 
 Rotation is imparted to the cupola by a pinion gearing internally geS™ 8 
 into a horizontal toothed wheel fixed into the floor of the gun pit. 
 
40 
 
 The pinion is actuated by winch handles through simple bevel gearing. 
 This gear, by means of a clutch, can be rendered available for winding 
 in the falls of the tackle used in mounting and dismounting the guns. 
 Four bent lever arms are attached to the legs containing the journal 
 boxes, and can be used, if necessary, to assist in traversing. 
 
 Close under the circular racer on which the trucks of the shield ride, 
 is fixed a brass graduated circle, over which travels a pointer attached 
 to the shield. This circle reads to degrees and 6-minute divisions. 
 The maximum elevation is 25 degrees, and depression 5 degrees. 
 
 The guns are supposed to be parallel. In rear of the centre of the 
 top of the shield and over the medial line of the guns, is a hole 
 sufficiently large to admit of the head of the No. 1 . At the front of 
 this manhole, and inconveniently near to the eye of the No. 1, is a 
 hinged gunmetal sight, furnished with a deflection leaf, and free to slide 
 up and down a slotted bar. The movement is not for the purpose of 
 giving elevation, this being provided for as already described by the 
 vertical arc, but for convenience in laying on objects at different levels ; 
 the sighting is, in fact, merely for line. Access to the sighting hole is 
 provided for by a light iron ladder between the guns. 
 
 The loading is performed at an elevation of 14 degrees, the loading 
 position being indicated on the steel vertical arcs by a broad red band, 
 so that the guns can at once be adjusted to it. The projectile and 
 charge are inserted by means of a loading tray between the arcs and 
 through the breech clamping screw, and are separately sent home by a 
 short rammer. The guns can be loaded with equal facility at all 
 angles of training. 
 
 The guns are 15™ (6-in.) B. L. of Krupp manufacture, 25 calibres 
 in length, and weighing about 3 tons. The common shell weighs 
 69 - 3 lbs., bursting charge 2| lb. ; the steel shell 88 lb., bursting charge 
 5£ lb. ; steel armour-piercing shell 86 - 9 lb., bursting charge 1 J lb. 
 
 The maximum charge of the gun is 19 - 8 lb., cocoa powder. 
 
 The cupola above described constitutes a complete departure from 
 the earlier designs of the Griison firm. Starting with structures built 
 up entirely of chilled cast-iron blocks, wrought-iron was first introduced 
 for the roof plates. In 1882, a wrought-iron umbrella- shaped cupola, 
 designed by Major von Schumann of the Prussian Engineers, was 
 tried at Kummersdorf, the design being to some extent based on the 
 experience gained at Tegel in 1870-1, and at Maintz in 1866. The 
 Tegel turret was made of wrought-iron, contained two guns, and 
 nearly resembled the ordinary ship's turret. The desirability of 
 reducing the cost and the weight led to the adoption of the non-recoil 
 system, which allowed the protected space to be smaller in extent ; 
 while the Tegel experiment, in which projectiles striking the plates at 
 an angle of 27 degrees produced extremely little effect, pointed to the 
 utilization of wrought-iron in plates with double curvature. Thus the 
 Kummersdorf cupola was made of wrought-iron plates bent to a 
 spherical form, with an exterior radius of 1 5 ft. The maximum angle 
 at which 15 cm shells fired at 1000 metres could strike this cupola was 
 35 degrees, and the thickness of the lower edges was 7£ ins. At the 
 level at which projectiles could strike only at an angle of 25 degrees 
 
41 
 
 the thickness was reduced to 4 ins. The curved shield was formed in 
 laminse, two of 2 ins. thickness, strengthened internally by two 3f ins. 
 round the lower portion. The Kiimmersdorf cupola contained only 
 one 6-in. gun mounted in the same way as those at Bucharest. There 
 were no curved cast-iron glacis plates, but instead, a vertical ring of 
 wrought-iron 4 ins. thick. 
 
 This cupola was severely tested. A hundred rounds were fired from 
 it at elevations between 4 degrees and 12^ degrees without injury. 
 Eighteen long 6-in. shells were then fired at it, of which several hit 
 the same spot, causing no damage. Seven chilled shell were then 
 fired also without any real effect. Bight 6-7-in. (17 cm ) chilled shell 
 were tried, some of which striking at angles below 30 degrees 
 produced no effect. Others, striking low down near the rim of the 
 cupola, damaged it considerably, and proved that the protection con- 
 ferred by 7| ins. of wrought-iron was insufficient against the 17™ gun. 
 After firing 11 salvoes of chilled shell at 1000 metres range from four 
 6-in. guns, by which no satisfactory hits were obtained, the attacking 
 battery was moved up to 300 metres, and two effective salvoes were 
 then fired without injury to the structure. 
 
 The cupola was then attacked by the 8J-in. (2 1 cm ) rifled mortar at 
 65 degrees elevation, firing 176 lb. shells. Four effective hits were 
 obtained, producing merely slight indents. The turning gear of the 
 cupola was tried and found to be uninjured. A complete rotation was 
 made in 36 seconds. Finally, 50 rounds were fired from the cupola 
 gun at elevations up to 24^ degrees, the last 25 rounds for rapidity. 
 The rate of fire was a little less than one round per minute, and the 
 diagram obtained was very satisfactory. 
 
 The Commission which carried out the K'immersdorf experiments 
 made a variety of recommendations of which the following were the 
 most important : — 
 
 1. The shield should be made up of single thicknesses of plates 
 instead of laminas. 
 
 2. A flatter curvature towards the lower edge of the cupola appeared 
 desirable. 
 
 3. The lower edge of the cupola should be better protected by being 
 sunk below the glacis ring. 
 
 4. The employment of concrete under the shield to balance the 
 whole structure about the central pivot should be avoided, if possible. 
 
 These recommendations, together with others, were embodied by 
 Major von Schumann in the design of the Bucharest cupola, so that 
 the latter may fairly be regarded as representing the results of much 
 previous experience. The compound plates forming the rear half of 
 this cupola were apparently introduced in accordance with the views of 
 the Roumanian Commission. 
 
 It is to be observed that cupolas built on the principle of transmitting 
 the recoil direct to the whole structure are termed by the makers 
 " shield mountings " (Panzerlaffete), and that this principle was first 
 tried at Kiimmersdorf. 
 
42 
 
 Besults. 
 
 IBth December, 1885. — Two rounds were fired from each gun with 
 full (19"8 lb.) charges at the maximum angles — viz., 5 degrees 
 depression and 25 degrees elevation, with "projectiles fictives " 
 weighing 90£ lb. These projectiles were constructed to break up at 
 the muzzle in order to avoid the necessity for clearing the range. The 
 gear worked well throughout. 
 
 Series (a). 
 
 19(/i December, 18S5. — Trials for accuracy — Ten salvoes at a target 
 10 metres by 6 metres. Range 2,500 metres. In this series five 
 salvoes from each turret, allowed as sighting shots, were first fired. 
 
 The oscillation of the turret being considerable, and the firing being 
 by word of command, and, therefore not always synchronous, the 
 accuracy of the second gun was doubtless affected. 
 
 Series (b). 
 
 21st December, 1885. — Trials for rapidity and accuracy — After 3 
 salvoes (sighting shots), 25 salvoes were fired, the laying being 
 corrected as in Series (a) by the telephoned result of each. 
 
 Total time occupied in firing 25 salvoes or 50 rounds, 2 hrs. 27 mins. 
 20 sees. 
 
 The times lengthened towards the end of the series. This was due 
 to the traversing numbers, who were not relieved until after the 12th 
 salvo, becoming fatigued. 
 
 Series (c) . 
 
 23rd December, 1S85. — Additional trials for accuracy, apparently 
 ordered because the brown powder had now arrived. 
 
 Series (d). 
 24th December, 1885. — The practice was apparently intended for a 
 comparative trial of the relative facility for rapidly directing the fire 
 of the two turrets. This series must be regarded as unsatisfactory 
 from every point of view, and so far as testing the power of firing 
 from turrets — the object which the Committee had in view — was 
 practically worthless. 
 
 Series (e.) 
 
 21th and 28th December, 1885.— (Pig. 2, p. 43).— The object of this 
 series was to obtain 30 direct hits on two of the three compound plates 
 forming the rear portion of the cupola, at a range of 1000 metres. To 
 effect this 85 rounds had to be fired — viz., 27 on the first and 58 on 
 the second day. 
 
 The cupola was turned so that the medial line of the rear centre 
 (compound) plate faced the attacking guns. 
 
 The total number of hits obtained was 35, but of these five were 
 en ricochet. 
 
 The three guns used for the attack were — 
 
 1 de Bange, 155 mm ~) of the same model as those mounted in the 
 
 2 Krupp, 15° m J turrets. 
 
48 
 
44 
 
 The firing was sometimes by single rounds, varied by salvoes of two 
 and three guns, no apparent principle being followed. 
 
 The French gun, de Bange, was mounted on a siege travelling 
 carriage. 
 
 The German gun was mounted on a siege traversing platform of 
 the usual Krupp pattern, on A racers. 
 
 The projectiles were all steel, of Krupp and St. Chamond manu- 
 facture. 
 
 The weights were as follows, sand being substituted for the bursting 
 charge : — 
 
 Krupp— 39 kilos. (85-8 lb.) filled; bursting charge 1*7 lb. 
 
 De Bange— 40-9 kilos. (89-98 lb.) filled; bursting charge 17 lb. 
 
 The powder charge for the German gun was 9 kilos, or 19*8 lb. of 
 brown prismatic powder, for the French gun the same weight of 
 Sevran-Livry pebble. 
 
 The distribution of hits at the conclusion of the 85 rounds, and the 
 general appearance of the turret is shown in Fig. 2, p. 263. 
 
 The general effect of single shots was to make a shallow scoop, 
 Fig. 2, p. 263, with a maximum depth of indent of about | in., and it 
 would not appear that the depth varied according to the angle of 
 impact with the curved surface of the plate. 
 
 A gradual cracking of the steel face of the plate took place during 
 these rounds, but at this stage there was no actual separation from the 
 wrought-iron. 
 
 The rotation of the cupola was in no way affected by the foregoing 
 series. 
 
 No cracks were visible in the interior ; but, at the 20th hit, one of 
 armour bolts was sheared, and the nut and washer, weighing 33 lb., 
 fell on the gun floor. A cast-iron spur wheel of the traversing gear 
 was also found to be broken. One of the hand levers for turning the 
 cupola was broken off. 
 
 Series (f). 
 
 28th December, 1885. — At the conclusion of the foregoing series, six 
 rounds were fired from the guns in the turret to ascertain whether 
 the gear was in working order. The turret had to be rotated by hand, 
 as the broken wheel was not replaced till next day. 
 
 Series {g). 
 
 Zlst December, 1885 and 1st January, 1886. — This series was intended 
 to ascertain the effect of high-angle fire from the 21 cm Krupp rifled 
 mortar, with steel and cast-iron common shell, weighted up to 91 kilos. 
 (200 lb.), at a range of 2,510 metres (2,745 yards). 
 
 It will be seen that the shooting on the whole was fairly good, the 
 maximum lateral dispersion, omitting two wild rounds, being 50 metres. 
 Two shells fell on the parapet and one on the top step of the entrance • 
 unfortunately none struck the cupola itself. 
 
 Series (h). 
 hth January, 1886. — This firing was at the wrought-iron port plate 
 on which seven direct hits were obtained. ' 
 
45 
 
 The guns in the cupola had been dismounted and replaced by solid 
 cast-iron cylinders about 5 feet long, strutted against the steel arcs by 
 rounded timbers. 
 
 The general effect of these rounds will be found in Fig. 4, herewith. 
 
 ■■»■ 
 
 Inside the cupola, one armour bolt between the ports fell out, and 
 two skin bolts sheared. The threads of the armour bolt, which were 
 evidently too fine, were uninjured. 
 
46 
 
 Series (i) 
 
 8th January, 1886. — This series was intended to test the resisting 
 power of the chilled-iron glacis plate. The 15 cm Krupp gun was 
 used throughout. 
 
 The sector fired at was practically breached near the bottom. 
 The attacking gun was mounted on a piece of built-up ground, and 
 its axis was a little above the level of the glacis. The line of fire had 
 an angle of depression of 1 degree 22 minutes, giving a maximum 
 angle of impact on the curved face of about 70 degrees. 
 
 <& 
 
 * 
 
 &< 
 
47 
 
 Series (J.) 
 
 11th and 14th January, 1886. — The Committee decided to continue 
 the practice of the 27th and 28th December, 1885, at the compound 
 plates, until a breech was effected. The practice was extremely 
 accurate, 35 hits being obtained, of which 7 were ricochets. The 
 general effect may be summed up as follows : — 
 
 The steel face, which had been previously cracked in several places, 
 began to strip at the third hit, the tendency to lamination showing 
 itself between the two qualities of steel as well as between the mild 
 steel and wrought-iron. The cracking and stripping proceeded 
 gradually throughout the practice, the whole area of wrought-iron 
 exposed at the conclusion of the series being about 5 ft. 6 ins. by 2 ft. 
 
 Fig. 6, p. 46, shows the appearance of the shield and the direction 
 of the principal cracks developed. As a rule, the cracks did not 
 extend to the wrought-iron ; but, at the proper left top corner of the 
 plate, a four-sided piece, area about 12 ins. square, was completely 
 detached remaining, however, in situ, sustained by the skin. A single 
 projectile striking this portion of the shield would probably have 
 penetrated into the cupola. The total number of hits sustained by the 
 rear centre compound plate was 31, including ricochets. — (Fide Figs. 
 4 and 5, herewith). 
 
 In the interior a considerable number of skin rivets and bolts were 
 sheared, but appeared to have merely dropped down on the floor, and 
 in no case to have flown with any force. 
 
48 
 
 Series (A). 
 
 17 1 A and 20th December, 1885. This series of 30 salvoes was intended 
 to test the working of the gear after the trials to which the cupola had 
 been subjected in Series (e) and (_/). 
 
 The shooting was evidently not greatly affected by the battering the 
 cupola had received, but was indifferent. 
 
 The only damage the cupola sustained was the shearing of one nut 
 in left arc of left gun ; this was done by last salvo at 5 degrees 
 depression. 
 
 Series ({). 
 
 22nd January, 1886. — This was in continuation of Series (A), cast- 
 iron dummies taking the place of the guns as before. Five rounds 
 were fired from 15 cm Krupp gun at the wrought-iron embrasure plate 
 at a range of 1000 metres, with a view of obtaining a hit at the edge 
 of the proper left port. 
 
 The damage inflicted by the first four rounds was very slight. The 
 fifth round, had the gun been mounted, would have put it out of 
 action ; while there can be little doubt that the spar interposed between 
 the dummy chase and the arcs saved the latter from such serious 
 damage as would have prevented another gun from being mounted. 
 The uninjured gun might in this case have remained serviceable. If 
 however, either of the cupola guns were struck fair on the muzzle, the 
 injury transmitted to the whole of the gear might evidently suffice to 
 disable the cupola. As it was, the only damage done, other than to 
 dummy, was the breaking, as shown in sketch, of the small bracket 
 supporting the right auxiliary trunnion, and this could have been 
 readily repaired. 
 
 Series (m). 
 22nd January, 1886. — For this practice, the turret was turned so 
 that the centre of the left front (wrought-iron) plate was turned 
 towards the attacking gun. 21 rounds were fired, of which 15 struck 
 the plate, and two the glacis plate, in line. As far as could be 
 ascertained, the majority of these steel shells broke up on the armour 
 plate. Thi3 no doubt was due to the shell, after deflection of the point, 
 striking an independent blow on the plate with the base, and cross 
 breaking. These same shells, striking a wrought-iron plate direct 
 either penetrated, or were thrown back unaltered in shape. 
 
 General Ebmabks. 
 
 This cupola cannot be considered apart from the leading principle 
 invented by Major Schumann, in which the guns, gear and shield form 
 part of a system rigidly connected, the whole force of recoil beinff 
 absorbed in the oscillation of the compound mass. 
 
 The system possesses certain marked advantages : 
 
 1. It lends itself to the employment of a turtle-back, or umbrella- 
 shaped shield, winch is comparatively inconspicuous, difficult to hit 
 and almost proof against direct fire. ' 
 
49 
 
 2. The guns being nearly muzzle-pivoting, not only is the size of AdY»nt»g*«. 
 the ports reduced to a minimum, but the exposed portion of the guns 
 themselves becomes comparatively small. 
 
 3. By the absence of a carriage, the complicated gear entailed by 
 the necessity of providing means of controlling the recoil at all angles 
 of elevation is avoided. Thus much valuable space is saved, and there 
 is plenty of room for the operation of loading, which can be performed 
 by hand from the floor level at any position of the cupola, as well as 
 for the storage of a large number of rounds. The traversing is also 
 performed in the chamber in which the guns are mounted, so that the 
 construction of a lower chamber with speaking-tube communication is 
 rendered unnecessary. The operations of mounting and dismounting 
 the guns are remarkably simple. 
 
 4. The method of mounting allows a range of 30° of vertical arc. 
 
 The disadvantages are : — 
 
 1. The breech of the gun has to be clamped at the required elevation ^ ntagea , 
 for each round, and should the capstan-headed clamping screw be not 
 properly set up, a violent blow would be dealt to the steel arcs in rear, 
 which might distort or displace them. 
 
 2. The adjustment of the guns depends upon the maintenance in the 
 same relative position of the steel arcs and of the trunnion brackets. 
 The upper extremities of the arcs and brackets are independently 
 attached to the inner boiler plate skin, and distortion of this skin may 
 therefore destroy the adjustment. The skin being attached by armour 
 bolts to the shield plates, is liable to distortion when the latter are hit. 
 A considerable displacement of the plates took place during the trial. 
 
 Distortion of the skin may produce the following results : — 
 (a.) The parallelism of the guns may be thrown out. 
 (5.) The direction of either or both of the guns, with respect to 
 turret sights and to the horizontal arcs, may be affected ; 
 moreover, these sights being attached to the roof plates, 
 are liable to independent displacement from blows on the 
 shield. 
 (c.) Any vertical displacement of the supports of the front 
 trunnions would vitiate the elevation given on the arcs. 
 
 3. The oscillation set up by firing exercises a disturbing effect on 
 the shooting for the following reason : — 
 
 (a.) Unless both guns are fired at the same instant of time, or 
 the second when the oscillations set up by the first have 
 subsided, it is evident that the angle of elevation of the 
 last gun fired will be incorrect. Electric salvo firing is 
 therefore a necessity. This fact was apparently not 
 recognized during the experiments, and the firing through- 
 out was by word of command, the interval between the 
 rounds being, in many cases, well defined. The want of 
 accuracy shown in the practice from the cupola guns was 
 probably due in a great measure to this cause. The 
 Kiimmersdorf cupola had only one gun, and it is stated 
 that Major von Schumann greatly prefers this arrangement. 
 
50 
 
 Shield. 
 
 Compound 
 plates. 
 
 (5.) The repeated oscillation of the cupola, the blow being always 
 delivered in the same direction, tends to give a permanent 
 set to the volute springs supporting the rear guiding 
 trucks. That this occurred during the experiments was 
 evident. At the conclusion of the practice, the lower rim 
 of the shield was visibly thrown out of the horizontal 
 plane, and the whole mass of the cupola had acquired a 
 permanent cant to the rear. If then, elevation was given 
 by arc, it would be incorrect, and it would be necessary to 
 lay the guns by quadrant. 
 
 4. The error in direction, due to movement in rotation during the 
 interval between the rounds, is common to all two-gun turrets, but in 
 the case of an oscillating cupola this movement becomes more com- 
 plicated. 
 
 5. Indirect laying for direction is provided for by means of a pointer 
 attached to the inner skin, and travelling over a fixed graduated arc 
 in the inside of the pit. If this pointer is adjusted close to the arc, 
 it is liable to be bent out of shape by striking the arc on recoil ; if at 
 a safe distance, accurate reading is rendered impossible by parallax. 
 During the experiments the oscillations of the cupola caused the 
 horizontal arc to be freely hammered by the pointer. 
 
 Each shield plate is attached to the inner skin by armour bolts 
 screwed half-way into the thickness of the metal. The threads on 
 these bolts were so fine that in some cases concussion of the shield, 
 due to a blow, released the bolt without stripping the threads. In one 
 instance one of the large armour bolts was sheared close to the nut 
 and dropped into the pit. The skin to which the whole of the gear is 
 attached being, as it were, suspended from the armour plates, and 
 not rigidly attached to them, appears to have been capable of partially 
 independent movement, which serves to account for the large number 
 of bolts and rivets sheared. 
 
 The arrangement of the armour plates provided for mutual support 
 on the principle of the arch ; but, during the breaching practice, 
 individual plates moved slightly, being lifted above the general curved 
 surface of the cupola as well as opening out at the joint. One of the 
 steel dove-tail wedges uniting adjoining compound plates gave way at 
 an early period of the attack. Although this movement of the armour 
 plates did not appear to affect their resisting power, it exerted an 
 injurious effect on the whole system, as already pointed out. 
 
 The total number of hits on one compound plate was 62 direct, and 
 8 ricochets, which sufficed to crack it extensively, to strip the steel 
 face off a large area, and practically to breach it. Cracks extending 
 through the whole thickness of the face were developed at an early 
 stage, soon after which pieces of steel began to be detached. . This 
 tendency to separation seems to indicate that the weld between the 
 different metals was impaired by the molecular change in the material 
 entailed by curving the plate subsequent to its construction. It is 
 
51 
 
 doubtful whether compound plates, although they lend themselves 
 readily to high curvature, can undergo bending without loss of resisting 
 power. 
 
 The wrought-iron plate attacked received 15 direct hits which, wroueht- 
 beyond making shallow scoops, did not in the least affect its resisting p 
 power. No cracks were developed or welds opened. 
 
 The embrasure plate received 7 direct hits, and with the exception Embrasure 
 of the round which struck the edge of the port, the plate sustained no plate- 
 damage. The beneficial effect of the swell round the ports was 
 practically not tested. 
 
 FRENCH TURRET. 
 
 GrENERAL DESCRIPTION. 
 
 The pit or emplacement {see Fig. 9, p. 52) consisted of two circular Empire- 
 chambers, one above the other, and was built rather to suit the™'" ''"''"' 
 purposes of the experiment than to represent an actual structure. 
 The walls were of brick throughout, with the exception of the upper 
 portion supporting the glacis plate, which was built in cement concrete. 
 The floor of the upper or gun chamber was at the natural ground level, 
 and was supported on concrete vaulting, forming the roof of the lower 
 or winch chamber, and containing the hydraulic pivot. Round the 
 upper chamber, were arranged three recesses, 5 ft. broad by 5 ft. 
 10 ins. high ; in one of these was a lift for delivering ammunition to 
 the gun floor, worked by a small hand winch in the chamber below, 
 communication between the two chambers being maintained by means 
 of three speaking tubes. The lower chamber had the same diameter 
 as the one above, and was provided with an independent entrance by 
 a flight of steps from the ground level. This chamber contained the 
 rotating gear, a set of ventilating fans, and the hand pump and gauge 
 connected with the hydraulic pivot. The parapet was built up only on 
 the side on which it was to be attacked. 
 
 In front of the glacis plate was a ring of concrete 9 ft. 6 ins. thick, Parapet. 
 measured from the exterior surface of the turret, with a depth of 3 ft., 
 and an exterior slope of 35 degrees. The rest of the parapet was 
 built up of sand, with a total thickness of about 30 ft. 
 
 No special arrangement had been made for magazine accommodation, 
 and, as a rule, the ammunition for the day's practice was stored in the 
 lower chamber. 
 
 The turret consisted of two armoured portions — 
 
 (1.) The glacis plate set in concrete, and resting on a brick Turret. 
 
 foundation. 
 (2.) The turret itself of the ordinary cylindrical form, similar to 
 those introduced for ships during the American war, and 
 since used generally in all navies. 
 
 1. The glacis is of cast-iron, and consists of four sectors of the section Glacis plate. 
 shown in Fig. 9, p- 52, having a total weight of 72 tons. These 
 
 30 
 
52 
 
 sectors merely abut against one another, and are not bolted together 
 or attached in any way. A channel lined with asphalte, formed in the 
 concrete below the level of the lower surface of the glacis plate 
 (Pig. 9, herewith), is carried completely round the emplacement, and 
 serves to drain off any rain water. An inner curb of cast-iron, in 
 three segments and bolted to the top of the wall of the pit, forms the 
 path for 10 slightly coned steel rollers, on which the turret revolves, 
 and also for six others set on vertical axes, so as to prevent horizontal 
 movement of the turret on recoil. 
 
 2. The curved wall of the revolving turret, exterior diameter 15 ft. 
 7 ins., consists of three wrought-iron plates, 18 ins. in thickness, 
 jointed as shown in sketch {vide p. 53), lead being run into the joint. 
 
53 
 
 The front segment is pierced for the two guns, the ports being ?™°t- 
 oval; breadth, 11£ ins.; interior height, 12 - 56 ins.; exterior height, 
 17 - 34 ins. ; distance from centre to centre, 3 ft. 2 ins. The weight of 
 each plate is a little over 19 tons, making the total weight of the 
 turret, less the roof plate, about 60 tons. 
 
 J 
 
 The turret is closed at the top by a roof plate of wrought-iron, 
 7'2 ins. thick, in two semi-circular portions morticed together, and 
 weighing altogether 19| tons, thus bringing up the total weight of the 
 armour of the turret to nearly 80 tons. The roof plate is rebated 
 down on to the walls, the depth of the rebate being half the depth of 
 the latter, and secured to them by 36 - 2^-in. screw bolts with stand-up 
 hexagonal heads and washers. 
 
 The whole of the revolving turret, as above described, rests upon a 
 built-up annular box-girder, in which are set the axles of both 
 horizontal and vertical rollers. 
 
 To this substructure are attached three main brackets, forming the 
 two compartments in which the guns were mounted. These brackets, 
 stiffened with augle-irons, as shown in Fig. 9, p. 52, are bolted down 
 upon a heavy casting, which rests upon the shoulder of the central 
 pivot (A), about which the turret revolves. 
 
 This pivot is of steel, 10^ ins. in diameter, and passes down into a 
 hydraulic cylinder of cast-iron lined with steel, containing glycerine. 
 A hand pump placed in the lower chamber was connected with this 
 cylinder, and, by this means, the whole moveable portion of the turret 
 can be raised by one man, when necessary, for the purpose of so 
 adjusting the load that little or no weight rests on the rollers. 
 These rollers are intended to serve merely as guides, keeping the 
 structure steady on recoil. The total weight borne on the glycerine 
 amounts to more than 100 tons. 
 
 The turret is rotated by a pinion engaging in a large circular toothed Turning 
 arc encircling the main brackets at a level a little below that of the e8ar ' 
 roller path. The pinion is keyed on to a vertical shaft, the upper end 
 of which is held in bearings bracketed out from the interior wall of 
 the pit, while the lower end passes through the floor and gears into a 
 double purchase winch in the chamber below. The arc is built up of 
 several pieces bolted together, and to the main brackets, and is 
 furnished with lignum-vitce teeth. The above arrangement is not shown 
 in the plates, which represent a slightly different design. 
 
 In the intervals formed by the three main brackets of the turret, 
 are the gun carriages. These consist each of two parallel brackets, to 
 
54 
 
 the front of which is bolted a T-piece {c, Fig. 9, p- 52, and c, Fig. 
 10, herewith) passing into the annular box-girder which carries tb.e 
 conical rollers, and hinged at a point below the centre of the port, this 
 being the point about which the gun pivots. Projections on the 
 brackets of the carriages work in curved angle-iron guides struck with 
 radii equal to their distance from the produced axis of the front pivot 
 hinge. Eectangular pieces of steel fitted over the trunnions of the 
 guns, slide in slots cut in the upper portions of the carriage brackets, 
 allowing a recoil of 18 ins. 
 
 Elevating 
 gear. 
 
 Recoil 
 presses. 
 
 FIG 10. GUN CARRIAGE WITH RECOIL 
 BUFFER FOR M000« TURRET 
 
 On the top of the main brackets of the turret are two guide pulleys 
 on either side of each gun, carrying chains one end of which is 
 attached to the rear of the carriages, the other to a counterpoise (Q) 
 hanging in front of the central pivot. The up and down movement of 
 these counterpoise weights is steadied by hollow rods passing through 
 them, and depending from the web plates (0), connecting the under 
 surfaces of the main brackets. The weight of each of these counter- 
 poises, 7,700 lbs., was such as to leave a preponderance of 1,100 lbs. 
 to each gun. This preponderance is borne on the head of a small 
 oscillating hydraulic ram (5 Fig. 10, herewith), actuated by a hand pump 
 attached to the main bracket in a convenient position. 
 
 Between the carriage brackets and immediately under the guns are 
 small hydraulic cylinders (a Fig. 10, herewith), in which work steel 
 pistons, their heads being attached to the gun. These cylinders are in 
 hydraulic connection with larger cylinders, (6) placed immediately 
 below them. The plunger areas of the two cylinders are as 1 : 2*8. 
 To the ends of the plungers of the lower cylinders are attached cross 
 heads, carrying two spindles on which are threaded " Belleville " cup 
 springs. Both cylinders are filled with glycerine, which passes from 
 one to the other, through a one-way valve, opening automatically 
 during recoil, and permitting the fluid to pass into the lower cylinders, 
 thus forcing out their plungers and compressing the "Belleville" 
 springs. To run up, the valve is opened by hand, and the springs 
 react, forcing the glycerine back into the upper cylinders. The initial 
 
55 
 
 load on the springs is adjusted by screws, so that they have power to 
 run the guns out at the maximum angle of elevation of 20 degrees. 
 
 Neither guns nor turrets were fitted with sights, and there are no Laying. 
 Openings for either direct or indirect laying. Elevation is given by 
 quadrant. Line is given in the first place by looking at the target 
 through the gun. Round the inner surface of the cast-iron curb is fixed 
 a gun metal ring graduated in degrees and quarters of a degree. Below 
 this is another similar ring divided into millimetres, this last being used 
 chiefly for giving deflection. Contact pieces, one for each gun, ride on 
 the upper graduated arc, and can be clamped at any desired position. 
 Prom the under surface of the annular box-girder depend projecting 
 pieces, which, rotating with the turret, touch the adjustible contact 
 pieces and complete circuits through the electric tubes by which the 
 guns are fired. The Le dancae" firing battery is conveniently placed 
 between the main brackets under the guns. Correction for line, after 
 trial shots, is made by altering the position of the contact pieces on 
 the graduated rings. To effect this, the Officer in charge has to mount 
 a small step-ladder, and read the graduations by the light of a hand 
 lamp. 
 
 The turret is rotated between each salvo, and the projecting pieces 
 are hinged so as to pass over the adjustable contacts after touching 
 them. 
 
 The inventor proposes to employ a rough method of position finding 
 in which a " Planchette " or plane table with a squared map of the 
 country is used. This was not tried during the experiments. 
 
 The two guns are 155 mm de Bange pattern, weighing nearly 3 tons ; Armament, 
 total length, 13 ft. 9 ins., with a travel of shot of about 22 calibres. 
 Maximum charge, 9 kilos., or 19 - 8 lbs. Sevran-Livry powder. 
 
 Weight, 
 lbs. 
 
 Common (cast-iron or steel) shell 88 
 
 Steel armour piercing shell 89 "8 
 
 Bursting charge of cast-iron common shell ... Z\ 
 
 n a steel ;; ... 11£ 
 
 The detachment consisted of 1 officer, 3 non-commissioned officers, 
 and 18 men. 
 
 The operation of dismounting involves the removal of a great 
 amount of semi-permanent gear, is lengthy and difficult, occupying at 
 least 12 hours for each gun. 
 
 Results. 
 
 The following series {a to k) and diagrams show results of the various 
 trials to which the French turret was subjected, arranged in chrono- 
 logical order. 
 
 Before commencing the regular programme, two rounds were fired 
 from each gun, with full charges of 9 kilos. (19-8 lbs.) and "projectiles 
 fictives," weighing 51 kilos. (112 lbs.), at the maximum angle — viz., 
 5 degrees depression and 20 degrees elevation. 
 
56 
 
 Series (a.) 
 19(/i December, 1885. — Trials for accuracy. 10 salvoes at a target 
 10 metres by 6 metres ; range, 2,500 metres. In this series five 
 salvoes were fired the previous day for sighting, the results of which 
 were not published. The marking in this series was carried out as 
 described in the case of the German cupola, and was, therefore, 
 equally unsatisfactory. 
 
 Series (b.) 
 
 22nd December, 1885. — Trials for rapidity and accuracy. 11 trial 
 shots, of which the last four were salvoes, were allowed before com- 
 mencing this series. 25 salvoes were then fired, with the elevation and 
 training thus obtained. Time occupied in firing 25 salvoes or 50 
 rounds, 1 hour 27 mins. 13 sees. 
 
 Out of the 50 rounds fired there were 44 direct hits. After the 
 20th salvo the target was so much damaged that the line of fire was 
 shifted 4 metres to the left. In computing the results, these rounds 
 have been referred to the original line. There were no missfires. The 
 smoke hung in the turret, causing some inconvenience to the loading 
 numbers. 
 
 Series (c.) 
 23/v/ December, 1885. — Additional trial for accuracy. 10 rounds 
 were fired with German brown powder. 
 
 Series (d.) 
 24/// December, 1885. — Sec remarks on the corresponding series fired 
 from the German cupola (p. 42). 
 
 (Series e). 
 
 26//* and 11th December, 1885. — The object of this series was to 
 obtain 30 direct hits on the turret, at a range of 1000 metres. The 
 turret was turned backwards and forwards through an arc of 70 
 degrees during the practice, with the apparent object of obtaining a 
 distribution of the hits, such as might be expected if this system were 
 tried under Service conditions. 
 
 The same guns were used as in the similar trial of the German 
 cupola, and were fired singly, in pairs, and in salvoes of three, according 
 to no apparent system. 51 rounds were fired, giving 27 direct, and 
 three ricochet hits were obtained. 
 
 Fig. 1, p. 57, shows in development the position and appearance of 
 the hits. The principal damage was the tearing away of a large 
 triangular piece of metal at the top and at the junction of two plates, 
 extending half the thickness of the wall, and laying bare the front of 
 the roof plate. It will be seen that, owing to the movement of the 
 turret, the hits were dispersed over an area equal to f^ths of the whole 
 circumference, which, no doubt, saved the plates from the severe 
 racking they would otherwise have sustained. The maximum 
 penetration obtained was 9± ins., most of the other rounds striking at 
 angles away from the normal scooped, leaving indents of various 
 depths. At this stage, a shell striking on the exposed junction of the 
 
57 
 
58 
 
 roof plate would probably have prized it up and let daylight into the 
 turret. This state of things was practically arrived at by three rounds. 
 It should be remembered, however, that this was a specially weak 
 point, since it was the junction of the two halves of the roof plate as 
 well as that of two sectors of the wall. Only one crack was developed ; 
 this ran up to the top of the plate from the round of maximum 
 penetration. The interior of the turret and gear were unaffected. 
 
 Series {/.) 
 
 27tA December, 1885. — At the conclusion of the above series, five 
 rounds were fired from the turret to ascertain whether the gear was in 
 working order. The practice was indifferent. 
 
 Series {g.) 
 
 31«< December, 1885, and 1st January, 1886. — Mortar practice. 70 
 rounds were fired, and no hits on the cupola were obtained. The 
 shooting on the whole was fair, the maximum longitudinal dispersion 
 being 160 metres. Three shells fell on the parapet. 
 
 Series (h.) 
 
 hth January, 1886 (Fig. 3, herewith). — The guns were dismounted and 
 replaced by wooden dummies. The attacking guns were the same as 
 those used for Series (A) in the German cupola. The object was to 
 obtain hits near the ports, the position for each hit being prescribed 
 
59 
 
 by the Committee. Of the four rounds fired, the first, striking close 
 to the edge of the port, turned into it, and grazed the opposite cheek 
 with its point, rebounding uninjured. This shell destroyed the dummy, 
 and would have disabled a gun and its mounting. Nothing is to be 
 learnt from the remaining three hits. All the shells rebounded whole. 
 No damage was done to the gear. 
 
 Series (i). 
 1th January, 1886. — This series was in direct comparison with 
 Series (i), German cupola, and' the conditions were, therefore, 
 apparently intended to be the same, but the attacking gun, although 
 mounted on a built-up bank, was below the level of the glacis plate, 
 so that all the projectiles struck at an ascending angle. In 20 rounds 
 fired, nine direct hits and one ricochet were obtained. A number of 
 small cracks were developed, which did not extend through the plate. 
 This small result is attributable to the fact that, on account of the 
 rising angle of the projectiles, and the height of the point of mean 
 impact above the bottom of the plate, the shells striking at an average 
 angle with the curved surface of only about 30 degrees, glanced, and 
 did not, therefore, do their full amount of work. This explains the 
 marked difference in these results, as compared with those obtained in 
 the similar Series (i), against the German glacis plate, in which the 
 latter was practically breached. 
 
 Series (j). 
 
 IWi January, 1886. — In this series the practice of 26th and 27th Dec, 
 Series (e), was continued with a view to forming a breach. The condi- 
 tions of attack were the same, except that in this series the turret 
 remained fixed instead of being rotated backwards and forwards through 
 70 degrees of arc. Further, only one gun was used, firing one nature of 
 projectile — viz., Krupp steel armour-piercing shell. In Series (e) 
 it was decided that meridian should be regarded as the normal line, 
 the area of maximum exposure to projectiles correct for line, and, 
 therefore, striking normally, being 35 degrees on either side of this 
 line. This meridian was, however, the joint of two wall plates, the 
 upper angles of which had already been destroyed, laying bare the 
 edge of the roof plate. The Committee, presumably on this ground, 
 chose another meridian (No. 2) to the proper left, as the new medial 
 line. This would seem, at first sight, to have been a concession to this 
 turret, since the German turret, during its trial, had always presented 
 the same meridian to the attacking gun. On the other hand, the 
 French turret could have been practically breached by two or three 
 well placed shots striking on the now open joint between the roof and 
 wall plates, and it was perhaps desirable to lessen the chances of a 
 premature conclusion to the trial. 
 
 17 rounds were fired on the 14th, and 27 on the 15th January. 
 
 Of the 44 rounds there were 32 direct hits (including one which 
 grazed the roof plate, inflicting no damage), six ricochet hits, and six 
 misses. The hits were distributed over an area of 5 ft. by 3 ft. 3 ins., 
 grouping themselves mainly on the centre line, the shooting being 
 excellent. 
 
 As a final result the turret was breached as soon as a group was 
 
 31 
 
(.0 
 
 formed near the top, but the first dangerous hit was Round 6, which 
 struck the previously exposed junction of the roof plate, wedging it 
 up, shearing eight bolts, and letting in daylight. The effects are 
 shown in Figs. 2, 3, herewith, and 5 p. 61. 
 
 Fig 2. 
 Section, through, A.2. 
 
61 
 
 The maximum penetration obtained from a direct hit was about 
 8 ins. ; the projectile, in such cases, being thrown back without 
 sensible deformation. In the case of glancing shots, the damage 
 inflicted was slight, and the shell, as far as could be ascertained, 
 invariably broke up. 
 
 An examination of the turret, at the conclusion of the series 
 revealed no damage to interior or gear, the rotation being unimpaired. 
 
 ^ 
 
 £ 
 
 Series (k.) 
 
 20th January, 1886. — 20 salvoes were fired in this series to test the 
 working of the turret after the breaching trial. 
 
 The mean time per salvo in this series was 3 mins. 48 - 8 sees., as 
 compared with 3 mins. 38 sees, in Series (b). 
 
62 
 
 General Eemaeks. 
 The French turret presents the following novel features : — 
 
 1. The floating pivot. 
 
 2. The complete closing of all apertures, with the exception of 
 
 the ports. 
 
 3. The absence of any inner skin, and the supporting of the side- 
 
 plates by means of the roof and the base ring only. 
 
 4. The system of automatic electric firing, necessitated by (2). 
 
 5. The employment of Belleville springs to "run up" the guns 
 
 after firing. 
 
 1. The floating pivot unquestionably added largely to the facility of 
 traversing. The total load, of about 100 tons, could not have been 
 otherwise moved without great labour. Whether and to what extent 
 "creeping down" took place cannot be stated. The turret was 
 throughout in the hands of the artisans specially brought from the 
 St. Chamond "Works, and they were, consequently, enabled to adjust 
 the incidence of the load upon the trucks to any extent that might 
 have been required. 
 
 2. The avoidance of all apertures, by which a chance splinter might 
 enter, has been claimed as a considerable advantage for the French 
 turret. In theory it may be desirable to avoid remote contingencies ; 
 but certain risks have to be accepted in war, and, in obviating them, 
 practical disadvantages may be incurred. Notwithstanding that the 
 gun chamber of the French turret had a large door always open, the 
 accumulation of smoke was, at times, excessive. Had this door been 
 closed, or in communication merely with a passage, it is doubtful if 
 long continued firing would have been possible. Moreover, but for 
 this door, the whole of the work inside the turret must have been 
 carried on by the light of lanterns, which, considering the general 
 inconvenience of the arrangements, would have been no easy matter. 
 Finally, the impossibility of direct observations from the turret itself 
 constitutes a definite disadvantage to the system. However desirable 
 it may be, as a general rule, to regulate fire from a distant and secure 
 observing station, it appears altogether unwise to abandon the possi- 
 bility of obtaining a direct aim under the many circumstances which 
 may arise. The theoretical invulnerability of the gun detachment may 
 thus be too dearly purchased. 
 
 3. The absence of an interior skin was advantageous, inasmuch as 
 there were no rivet or bolt heads to fly when the turret was struck. 
 On the other hand, a large number of the bolts, attaching the roof to 
 the walls, were sheared by a single hit, and a shell, striking below the 
 exposed edge of the roof-plate, would, undoubtedly, have raised it, 
 thus seriously affecting the stability of the whole structure. 
 
 4. It cannot be said that the principle of automatic firing was ever 
 properly tested. There appears to be a considerable risk in thus 
 trusting to an electric contact being completed at a precise instant. 
 
63 
 
 Unless the contact pieces were always chemically clean, an uncertain 
 delay might occur which could not be corrected for. The mere wearing 
 of the rubbing surfaces would also be a source of inaccuracy, but 
 would, of course, be less a matter of chance. Theoretically, provided 
 that the pointers and arcs were correctly adjusted, the firing of the 
 two guns should have been absolutely simultaneous, which was not 
 always the case. Possibly, on account of the reasons above indicated, 
 the earlier practice, Series (a), (b), and (./), was decidedly indifferent. 
 In Series {k) better shooting was obtained, possibly because some 
 alteration of the contact pieces had been effected. It is worthy of 
 notice that, of the 21 salvoes in this last series, nine were fired by 
 friction tubes, thus involving merely an optical adjustment of a pointer 
 on the fixed horizontal arc. 
 
 5. The action of the springs in forcing the guns back into the 
 loading position appeared to be satisfactory ; but it is impossible to 
 state whether the recoil presses necessitated special care or attention 
 during the trials. 
 
 General Conclusions. 
 
 The value of these important and costly experiments was sensibly 
 modified by certain of the conditions under which they were carried 
 out. 
 
 1 . The power of the guns employed in the attack fell far short of condition 
 the standard of a modern heavy siege train. The Krupp 1 5 cm , and ° f tml ' 
 the de Bange 155 mm , practically 6-in. guns, fired a maximum charge 
 
 of 19 - 8 lbs., with projectiles varying in weight from 69"3 lbs. to 90 lbs. 
 The muzzle velocity with the armour-piercing projectiles being only 
 about 1500 ft. -sees. As a comparison, it may be stated the 6-in. B.L. 
 gun proposed for our heavy siege trains, fires an armour-piercing shell 
 weighing 100 lbs., with a charge of 34 lbs. of powder, giving a muzzle 
 velocity of 1830 ft. -sees. Thus, even discarding the probability of the 
 employment in future sieges of jointed guns of increased power, 
 existing armaments are far superior to the ordnance selected for these 
 trials. 
 
 It is to be remembered that if cupolas are ever largely employed 
 for the defence, every effort will be made to bring the heaviest siege 
 artillery against them. It is in all respects, therefore, to be regretted 
 that the most powerful weapons available were not employed for the 
 attack. 
 
 2. Guns identical with those above described were mounted in the 
 turrets. If such costly structures are to be employed at all, it is clear 
 that the best possible guns must be mounted in them, since there is an 
 obvious anomaly involved in expending £10,000 in mounting two 
 weak siege guns. Even supposing that the calibre of the guns so 
 mounted were limited to 6 ins., they would not be less powerful than 
 our Mark IV., giving a muzzle velocity of 2050 ft. -sees., with a charge 
 of 55 lb. brown powder and a 100 lb. shell. Whether it would be 
 
64 
 
 possible to mount such a gun in either of the turrets experimented 
 with cannot be said to be established. In the case of the German 
 cupola it is clear that radical structural alterations might thus be 
 entailed, since the retention of the non-recoil principle might involve 
 a prohibitory increase of weight, strength, and cost. 
 
 With the French turret similar difficulties present themselves, 
 though, perhaps, in a less degree. 
 
 3. A combination of a service and a manufacturers' trial, which 
 was practically the system adopted, is in all respects unsatisfactory. 
 The shooting was carried out under Roumanian Officers with Artillery 
 detachments ; but the artisans of the competing firms were at hand 
 to assist in any way that might be required, and throughout the 
 trials the turrets appeared to be in charge of their exhibitors, whose 
 skilled workmen, living on the spot, were able to repair damage, 
 effect adjustments, and generally make good any defects which arose 
 during the practice. Throughout the trials the respective inventors 
 were not only permitted to give directions and suggestions of all 
 kinds, but to express their mutual opinions with regard to their rival's 
 turret with perfect freedom. 
 
 4. The method of recording the results of the shooting from the 
 turrets for accuracy left much to be desired. The telephoned reports 
 were based on observations by an Officer, who, not having time to 
 measure individual hits or grazes, judged them as best he could from 
 his shelter. It was specially desirable to test the accuracy of adjust- 
 ment of the guns ; tu ascertain if their parallelism was effected ; to 
 gauge the disturbance of one gun caused by the discharge of the 
 other ; and, in the ease of the French turret, to form an estimate of 
 the practical efficiency of the system of automatic firing. For all the 
 above named purposes the recording of individual shots was essential, 
 this condition cannot be said to have been complied with. 
 
 The time of year in which the experiments were carried out, and 
 the prevalence of fog, snow and extreme cold, doubtless added much 
 to the difficulty of the conduct of the trials. 
 
 5. In the breaching trials, the system of firing salvoes of two or 
 three guns indiscriminately, rendered it impossible to identify the 
 effect produced by shells of a particular class. Moreover, speaking 
 generally, there was no opportunity for ascertaining whether particular 
 shells broke up, or the effect on the shells themselves of impact under 
 varying conditions. Again, little or no time was allowed for 
 adequately recording the effect on armour plates of successive rounds. 
 On one occasion as many as 56 such rounds were fired on a single day. 
 Thus, the recorded results cannot be regarded as sufficiently detailed, 
 while the trials fail to provide such data with regard to the behaviour 
 of steel projectiles as might have been obtained. 
 
 It is now proposed to compare the merits and performances of the 
 two turrets under separate heads. 
 
 1. Emplacement. 
 In this respect the German cupola has a marked advantage. In 
 
65 
 
 the case of the French turret, the lower chamber necessitates either a 
 considerable excavation or a high command with its attendant dis- 
 advantages of increased visibility. The absence of any sort of recoil 
 carriage in the German cupola leaves the gun chamber comparatively 
 free. Thus the whole of the operations of loading, training, &c, are 
 carried out on one floor, and therefore under the eye of one responsible 
 officer, while the recesses provide space for the storage of a consider- 
 able quantity of ammunition. Moreover, this one chamber is, by day, 
 well lighted from the back passage and the man-hole at the top, while 
 the lower story of the French turret is dependent always on artificial 
 light and ventilation. 
 
 2. Mountings. 
 
 Both systems of mounting worked well, and did not appear to be 
 affected by the trials. The German system is open to the obvious 
 objection that its applicability to a high velocity gun is still unproved, 
 while there are evident difficulties to be met. The continued recoil, 
 even of these weak guns, was sufficient to give a permanent set to the 
 springs supporting the rear trucks. The effect of this would be to 
 cause an error in elevation as given by arc, and practically to 
 necessitate recourse to the clinometer. In any case, the suspending 
 of the guns from the inner skin of the shield is open to objection. 
 Further, the vertical steel arcs, against which the breeches of the guns 
 rest, are supported above by the inner skin, and below by the built-up 
 base resting on the pivot, an arrangement capable of leading to 
 disturbance of the position of the arcs themselves relative to each 
 other, and to the axes of the trunnions. 
 
 The employment of Belleville springs, as in the French turret, to 
 run out the guns, would probably be applicable to more powerful guns. 
 
 The system of mounting adopted in the German turret, possessed, 
 however, the incidental advantage that the size of the ports was 
 reduced to a minimum, while the exposed portion of the chase of the 
 guns was less than half that entailed by the design of the French 
 turret. 
 
 3. Facility op Mounting and Dismounting Aemament. 
 
 In the design of the German cupola the desirability for providing 
 easy and simple means for mounting or replacing the guns has been 
 recognized, and the arrangements have been extremely well thought 
 out. The whole operation is quite straight forward, and can be 
 performed by ordinary gunners. There is no doubt, that, with a 
 certain amount of drill and practice, the time quoted might be 
 reduced at least one half. On the other hand, in the case of the 
 French turret, the mass of ironwork and gear which has to be removed 
 in the process of dismounting renders the assistance of skilled artificers 
 an absolute necessity. It was impossible to ascertain the exact number 
 of hours expended in removing the guns from this turret, as the work 
 was probably continued night and day, but on each occasion the 
 operation appeared to extend over two days. 
 
66 
 
 4. Ventilation and Lighting. 
 
 The ventilation of the German cupola was far superior to that of the 
 French turret. This was mainly clue to the manhole at the top of the 
 former, by which a direct draught was promoted. In the French turret, 
 after firing, the smoke accumulated in the upper portion of the turret 
 to an inconvenient extent. If the gun-chamber had not opened 
 directly into the air, this drawback would have much increased. The 
 lower chamber of the turret was provided with a fan, apparently 
 intended to assist ventilation ; but this arrangement, if it was used, 
 does not seem to have been successful. It is to be observed that 
 ventilation of the lower chamber also would have to be secured if the 
 design were applied to a work of defence. 
 
 The French turret entailed artificial lighting at all times, while the 
 man-hole in the roof-plate of the German cupola provided ample 
 illumination for day firing. 
 
 5. Loading. 
 
 The large space available in the German cupola, and the fact that 
 all the operations of loading were performed from the floor of the 
 gun-chamber, was an undoubted advantage. 
 
 In the French turret, not only had the ammunition to be sent up by 
 a lift from the gun floor to the platform above, but the whole arrange- 
 ment was extremely cramped and inconvenient. While, in the one 
 case, any gunners would in a short time have been able to carry out 
 the loading operations with the necessary rapidity ; in the other, 
 careful special training would be necessary. On the other hand, it is 
 to be observed that the German system of mounting practically limits 
 the choice of breech action to a side wedge of the Krupp type, and 
 precludes the employment of a breech-screw. 
 
 6. Elevation. 
 
 The guns in both cupola and turret being counterpoised, elevation 
 was given without difficulty. In the German cupola, the guns being 
 overbalanced, the breeches were hauled down to elevate. In the 
 French turret, there was a small preponderance, and the breeches 
 were raised by a little ram, actuated by a pump. It would be 
 interesting to know whether this ram was able to maintain the breech 
 at a constant elevation, or whether any slight variation took place. 
 No experiments were made with a view of testing this point. In the 
 German cupola, the breeches were clamped as soon as the required 
 elevation had been obtained. 
 
 7. Traversing. 
 
 The French turret appeared to traverse more easily and smoothly 
 than the German cupola, in spite of its far greater weight. The former 
 being intended always to make a complete revolution between each 
 salvo, the designers had been led to adopt the floating pivot, to which 
 the comparative ease of traversing was mainly due. The greater 
 finish of manufacture of the French turret was not without influence 
 
67 
 
 however ; while the adoption of wooden teeth in the main gearing was 
 doubtless advantageous. These features are of course equally 
 applicable to the German cupola. The latter possessed the advantage 
 that the traversing, being carried on in the gun chamber, could be 
 directed by the officers in charge by signal, without the necessity for 
 having recourse to a speaking tube. 
 
 The horizontal arc of the German cupola could be read without 
 difficulty from the floor of the gun chamber. In the French turret it 
 was necessary to mount steps and to use a hand lantern. On the 
 other hand, the arrangement of the pointer in the former case was 
 decidedly objectionable ; since, although placed at a distance so great 
 as to involve parallax, it nevertheless struck the arc violently during 
 the oscillations of the cupola on recoil. 
 
 8. Sighting. 
 
 The power of laying by sight appears to be a decided advantage in 
 favour of the German cupola. Not only can direct laying be employed 
 which, in the case of a moving object, would be of the highest 
 importance, but back laying by means of the hole in the rear armour 
 plate can be utilized. 
 
 In the French turret, unless a system of position finding is applied, 
 it is necessary in the first instance to lay the guns through the bore by 
 means of a sighting tube or cylinder. Thus, if comparative trials at a 
 moving target had been carried out, or even practice against dispersed 
 targets, the German cupola might have been expected to show a 
 marked superiority. It is to be observed, also, that this cupola lends 
 itself equally well to the employment of a position-finding system. 
 
 The complete closing in of the French cupola, and its absolute 
 dependence under all circumstances on information supplied from 
 without, is a condition which is likely to recommend itself only to the 
 pure theorist. The chance of a shrapnel ball or machine gun bullet 
 finding its way into the turret through the opening in a conning tower 
 may well be accepted, in consideration of the advantage of allowing 
 the Commanding Officer of the turret an occasional view of the distant 
 terrain, instead of rendering him completely dependent on an observer 
 who may be killed, or a system of communication which may break 
 down. Such a break down would probably result in the continuation 1 
 of a useless fire, and a waste of ammunition. 
 
 9. ACCUEACY OJ FlKE. 
 
 It is greatly to be regretted that no series were fired with the 
 object of obtaining comparative range and accuracy groups. The 
 results of each round, or salvo, were usually telephoned back to the 
 firing point, and as these results could not in many cases be other 
 than approximate, the practice does not afford a satisfactory com- 
 parative test of the shooting capacity of the cupola and turret. 
 Again, for the reasons stated on p. 49, the principle adopted in the 
 German cupola appears to demand electric salvo firing as a condition 
 of satisfactory practice, and the fact that the salvoes were all fired ftf 
 word of command, resulting in many cases in a distiUct interval 
 
68 
 
 between the rounds, doubtless told against the accuracy of the fire 
 from the cupola, while precluding the formation of a just estimate of 
 its capabilities. Generally speaking, the shooting was indifferent, 
 which must be partly attributed to the conditions of the mounting, 
 and also to the changes in direction and elevation being made on 
 information sufficiently accurate. On the other hand, the practice in 
 Series (fc) from the French turret, during which it is believed that no 
 results were telephoned, and no corrections in elevation or direction 
 were made, was decidedly better. In neither cases is there any 
 reason to suppose that the shooting was affected by the battering the 
 cupola and turret received in Series (<?) and (j) . 
 
 In the case of the French turret, the gun mounting being entirely 
 independent of the shield, there appears to be no reason to expect any 
 inaccuracy from this cause. In the German cupola, however, it is 
 evident that errors might arise from a distortion of the skin, produced 
 by movement of the supported armour plates. 
 
 The efficiency of the automatic system of firing, which forms one 
 of the special features of the turret, cannot be said to have been 
 properly tested. Judged by Series (/J), it appears to have acted fairly ; 
 but more extended trials would evidently be necessary to dispose of 
 the objections pointed out. On the other hand, the arrangements for 
 laying the German cupola for direction by arc and pointer were rough 
 and unsatisfactory. 
 
 10. Rate op Fiee. 
 
 A comparison of the two Series (b) shows that the German cupola 
 fired 25 salvoes in 2 hrs. 27 mins. 20 sees., giving a mean time per 
 salvo of 5 mins. 54 sees. ; while the French turret required only 1 hr. 
 27 mins. 13 sees., or an average of 3 mins. 38 sees. 
 
 This comparison is, however, scarcely just to the German cupola, 
 which was made to perform a whole revolution after every salvo 
 merely because the French turret had been specially designed for this 
 purpose, and could be loaded during rotation. This was claimed as 
 an advantage by the inventors, since the guns were thus rendered 
 invisible, and practically invulnerable to the attack during a portion 
 of the period of rotation. It may well be questioned whether this 
 advantage has any real existence. When facing the attacking battery 
 the guns present a small mark ; when nearly at right angles to the 
 line of the enemy's fire they offer a broadside target, which on account 
 of the protrusion of the French guns is considerable. Under service 
 conditions, the fire of the attack from the first artillery position at 
 least, would not be so extremely accurate that each round would strike 
 close to the ports, if the turret were not rotated. Hence it is doubtful 
 whether the chances of a hit on the muzzle or chase are not rather 
 enhanced by turning them round and presenting a broadside view 
 twice after each salvo. In any case, the rate of fire is much reduced 
 by the operation, while a great strain is thrown on the traversing 
 numbers, who would have to be relieved at short intervals. 
 
 In Series (£), fired after the breaching trials, the average time per 
 
69 
 
 salvo in the German cupola was 6 mins. 85 - 8 sees., while in the French 
 turret it was 3 mins. 48'38 sees. Thus the increase in the averages 
 was 41-8 sees, and 10 - 38 sees., respectively. 
 
 No opportunity was afforded to the German cupola of proving its 
 rate of fire without turning ; but in Series (6) the mean time per salvo 
 for loading and laying was 1 min. 29 sees., which must approximately 
 represent the maximum. It is evident that similar guns on siege 
 
 mountings are capable of a greater speed of fire, and the loss of time 
 is due to the inconvenience of performing the loading operations in 
 the cupola, and also to the greater weight to be traversed. 
 
 The two Series (d), apparently intended to test the performances of 
 the cupola and turret against a moving target, obviously afford no 
 data, comparative or otherwise, which can be regarded as of value. 
 
 11. Resistance of Shields. 
 
 The German cupola received in all 85 direct and 12 ricochet hits. As 
 will be seen from Pigs. 4, 5 and 6, pp. 46 and 47, by far the greater pro- 
 portion fell upon a single compound plate, and were grouped near its 
 proper left edge. The general result was the stripping off of the 
 steel face over an area of about 10 square feet. No actual breach was 
 effected, but the left hand top corner of the plate attacked was 
 detached and remained merely supported in position on the inner skin ; 
 a single shell falling on this part of the shield would have penetrated. 
 When it is remembered that this one plate received 51 hits, including 
 ricochets, the resisting power must be classed as high. This excellent 
 result is evidently due to the flat curvature of the design, and not to 
 the material. It was shown that the steel face was first cracked up, 
 and eventually detached under successive blows, proving that the weld 
 between the two metals was imperfect. This was probably due to the 
 re-heating and bending of the plates after manufacture. The wrought- 
 iron plate was unfortunately not subjected to a comparative test, as it 
 only received 15 hits, well distributed. This plate, however, exhibited 
 no signs of cracking, and individual hits merely made scoops in the 
 metal varying from - 4 in. to 112 in. in depth, the shells in most 
 cases breaking up. There seems to be no doubt that a curved wrought- 
 iron plate, even of this small thickness (8 ins.), is capable of success- 
 fully resisting a large number of hits, since it would appear extremely 
 difficult to obtain any cumulative effect. It was intended to fire a 
 series of flat-headed projectiles against the curved shield, as it was 
 considered that better results would thus be obtained. There is 
 obviously no reason to expect that this would be the case, since 
 experiments at Shoeburyness have proved conclusively the superiority 
 of ogival shot under all circumstances. 
 
 The armour of the French turret received altogether 63 direct and 
 9 ricochet hits, of which the greater number were grouped over an 
 area of 21 square feet. Projectiles striking normally, penetrated to a 
 maximum depth of about 9 ins., the steel armour-piercing shells 
 usually rebounding uninjured. When any hit took effect within about 
 18 ins. of the edge of the plate, cracks were generally developed 
 extending to that edge. The tendency to form vertical cracks was 
 
70 
 
 specially noticeable. This weakness was probably due to the initial 
 strain induced on the outer surface by bending the plate. An exami- 
 nation of the fracture showed a large crystallization towards the face 
 of the plate. In the case of oblique impact, the projectiles generally 
 broke up, forming irregular cuts. In some instances, however, a side- 
 long print of the projectile was formed in the soft metal. The turning 
 of the turret during the breaching practice in Series (e) rendered it 
 impossible to estimate the angle of impact of individual rounds. The 
 principal damage was caused by a number of shells striking near the 
 top of the wall, laying bare the junction with the roof plate, and 
 finally prizing up the latter and shearing a large number of the 
 attaching bolts. At the conclusion of the breaching practice, there- 
 fore, a projectile striking the exposed lower edge of the roof plate 
 would probably have lifted it bodily, while a portion of plate near the 
 top of the wall was so weakened that it could easily have been 
 penetrated. 
 
 It will be seen, therefore, that these results bring out the great 
 advantage of the spherical model. Although a direct comparison is 
 hardly possible, there is every reason to suppose that the 8-in. wrought- 
 iron plates of the German cupola possessed a power of resistance fully 
 equal to that of 18-in. wall of the French turret. 
 
 If a more powerful gun had been employed in the attack, the French 
 turret would have suffered in a relatively higher degree. 
 
 12. Damage to Inteeioe. 
 
 The number of rivets and small bolts detached during the attack 
 on the German cupola was considerable. Although they appeared to 
 have merely fallen without force, it is evident that this liability must 
 be guarded against in a fighting machine. The threads of the large 
 armour bolts were so fine that one of them was actually shaken out of 
 the shield, apparently uninjured. This ought not to have been 
 possible. 
 
 The employment of a rope mantlet, or of an independent inner 
 skin, would prevent the fall of rivets, &c. It would probably be 
 possible, however, to cushion the shock transmitted from the armour 
 plates to the skin, so as to obviate this great disadvantage. 
 
 The breaking of a pinion belonging to the traversing gear, appears 
 of less importance. It is clear that this pinion should never have 
 been made of cast-iron. 
 
 The absence of an interior skin in the French turret proved a great 
 advantage. There were no bolts or rivets which could, under any 
 circumstances, have fallen, and the damage was nil. 
 
 13. Glacis Plate. 
 
 The glacis plate trials, Series (i), do not lend themselves to a 
 satisfactory comparison. On account of the considerably greater 
 height of the French turret, with respect to the attacking gun, its 
 glacis ring could not be hit at so great an angle as that of the 
 cupola. The flatter curvature of the French section told further in 
 
71 
 
 its favour. Six direct hits were obtained on the German glacis plate, 
 forming a group near the bottom edge, and breaking off a large 
 fragment, which, however, was not detached. The French glacis 
 plate received nine hits, forming a dispersed group about half-way up. 
 Thus all the hits were very oblique, and, though numerous hair 
 cracks were developed, no real damage was effected. 
 
 The attack was carried out at 54 - 6 yards range, and in the case of 
 the German cupola 16 rounds were fired before the glacis plate could 
 be reached, although the spade was freely used meanwhile in clearing 
 away the sand and concrete debris. It is sufficiently evident that, 
 even at 1,000 yards, it would be perfectly useless to attack either 
 glacis with such guns as were employed in these experiments. It 
 appears doubtful, indeed, whether a cast-iron glacis ring of this form 
 could be successfully attacked by any gun or howitzer which could be 
 put into the trenches. 
 
 14. Visibility. 
 
 The German cupola possessed a marked superiority in the smaller 
 and less conspicuous target offered to the attack. The result of this 
 was that in Series (e), 85 rounds were fired at the cupola in order to 
 obtain 30 direct hits ; while in the case of the turret 27 hits were 
 obtained in 50 rounds. Taking Series (e) and (/) together the com- 
 parison is 58 hits in 137 rounds, against 59 hits in 94 rounds. In 
 other words, at a range of 1,093 yards it required about one and a 
 half times the number of rounds to obtain the same number of hits 
 on the cupola as on the turret. 
 
 The low, flat, umbrella-shaped cupola could, if properly coloured 
 and with a suitable back ground, be rendered almost invisible. 
 
 15. Embrasures. 
 
 The attack on the embrasure plates affords no comparative data. 
 Twelve rounds were fired against the German cupola, and only four 
 against the French turret. As soon as a hit wasobtained on the edge 
 of the port, the gun in each case must be considered to have been 
 destroyed. 
 
 In the case of the German cupola, however, the steel casting, sup- 
 porting the right front trunnion was also broken, but could have been 
 replaced without difficulty. To this damage the French turret was 
 not liable ; but, on the other hand, the main danger arising from a 
 hit at the port is evidently the destruction of the gun. 
 
 General Eemares. 
 
 From what has been said above, it will be evident that both the 
 structures under experiment fall far short of the ideal land service 
 cupola. The practical value of the trials was greatly reduced by the 
 weakness of the guns employed. If turrets of any form are to be 
 adopted in the main armament of land works, it is clear that much 
 heavier guns must be put into them. The great expense involved 
 
72 
 
 could not otherwise be justified. The cost of the French turret alone 
 was about £10,000, exclusive of its armament, and for this sum about 
 six overbank guns of greater power could be provided. It may fairly 
 be questioned whether the latter would not be a wiser mode of 
 expending such an amount. 
 
 Again, if cupolas of any form come into general use, the heaviest 
 possible siege guns will be brought against them. The Russians 
 already have a screw gun of 8-in. calibre, and the development of 
 this system promises to confer increased power on the attack, which 
 is limited only by the weight of the heaviest portion. In any case, 
 the Krupp and de Bange guns employed at Bucharest do not even 
 reach the standard of existing siege ordnance. For these reasons, 
 any deductions which can be made from the trials in question are 
 necessarily subject to qualification. 
 
 Summing up ; the main advantage of the German cupola was its 
 form, and the main defect the detachment of the interior bolts and 
 rivets. In the case of the French turret, the principal merits were 
 the floating pivot and the absence of an inner skin ; while the chief 
 drawbacks were the want of any means of direct laying, the bad 
 ventilation, the complication and general inconvenience of the interior, 
 and the greater cost of the emplacement. 
 
 By combining the best features of the two designs it is possible 
 that a satisfactory military machine could be arrived at. A cupola 
 with a wrought-iron umbrella-shaped shield, supported on a floating 
 pivot, and armed with guns provided with a recoil mounting and port 
 pivoting might meet all requirements. The turtle-backed shield is 
 the practical recognition of an important principle, and constitutes a 
 decided step in advance. One of the most important facts brought 
 out at Bucharest was the great resistance of a comparatively thin 
 wrought-iron plate of this flat curvature. It is significant that the 
 designer of the French turret has since proposed a structure in which 
 the German model is copied bodily as to form. 
 
 The general question of the necessity for the adoption of cupolas 
 for land defence receives no new light from the Bucharest trials. The 
 armour in both cases received a number of blows, probably representing 
 a siege of many months duration. But, on the other hand, it is 
 evident that had the guns been turned towards the attack they would 
 have been many times disabled. Against the direct frontal fire of 
 high-velocity guns, the cupola appears to afford no increase of 
 protection commensurate with its cost. Against curved fire and flank 
 fire, however, the gain of security is considerable. 
 
 If a horizontal shield, capable of providing protection against the 
 projectiles of the heaviest siege howitzers, can be provided for the 
 pits of disappearing guns, this system would seem to be preferable 
 to any land service cupola at present devised. The main point of 
 weakness, the opening through which the gun is raised and lowered 
 is perhaps more than counterbalanced by the absence of a well-defined 
 and always visible target. 
 
73 
 
 SUPPLEMENT TO CHAPTER III. 
 
 Improved Mougin and Proposed Disappearing French Turret. 
 
 Le Genie Civil, of January 15th, 1887, describes a new disappearing 
 turret to be tried shortly at Chalons. The writer (Claude Manceau) 
 observes that certain inconveniences are inseparable from the dome 
 which Major Schumann employs. It is suggested that the joints of 
 the Schumann dome offer weak places, that the bolts are liable, both 
 from shock of discharge and of impact, to fly and cause much injury 
 to the detachment, and that it is impossible to predict what dislocation 
 might be caused by the blow of even a single 21 cm. (8'27-in.) 
 projectile falling vertically on the cupola dome, seeing that at 
 Bucharest the vertical fire failed to strike either cupola. It is con- 
 sidered that Major Mougin, in the mushroom-shaped dome which he 
 has now adopted, has remedied some of the faults of the Schumann 
 construction. Dislocation is rendered less liable to occur because the 
 dome is jointed in two parallel lines, shown in Pig. 2, p. 74, which is 
 thought much better than the radial form of division, which, it is 
 urged, leaves each shield overhanging and dependent for its stability 
 on the support of the contiguous sections. The Mougin cupola will 
 then, it is remarked, resemble that of Schumann in general appear- 
 ance, but it will differ from it in its principles of construction, seeing 
 that the guns of the latter are attached to their cupola, causing 
 inaccuracy in fire as well as other evils. It is further observed that 
 at Spezia, in 1886, the Gruson turret shield threw off the Krupp 
 forged steel projectiles of proved excellent quality, breaking them up 
 in small fragments. This, it is considered, is in a great measure to be 
 attributed to the profile of the shield, against which the projectiles 
 must strike at an angle more oblique than 45 degs. ; but the extreme 
 hardness of the metal is also noticed. The effect of projectiles with 
 flat, or rather slightly concave cupped points is discussed. These 
 projectiles, advocated by Whitworth and Krupp, have cut like a 
 chipping tool into armour at oblique angles, the case in view being 
 probably that at Buckau, when Ternitz projectiles bit into the surface 
 of a Gruson cupola. It is suggested that an attack with cupped-out 
 projectiles would prepare the way for sharp pointed projectiles to 
 enter. At angles over 25 degs., it was generally found at Gavre — as 
 in England at angles over the glancing angle — that a given vertical 
 height received the same protection from a given weight of metal, 
 whether the wall was at an inclination or placed vertically. 
 
 The main disadvantage of the cylindrical form then lies in the 
 vulnerability of its sharp top edge, and in the distinctness with which 
 
 33 
 
74 
 
75 
 
 so defined a form can be seen. At Ootroceni the latter evil was 
 increased by the fact that the French turret stood on higher ground 
 than the German one — a fact, we may observe by the way, which 
 gave it a great advantage when its glacis plate — avanl cuirasse, was 
 attacked. The French turret also was painted black, and the German 
 one grey. 
 
 M. Manceau concludes that the Bucharest experiments showed that 
 no turret could resist constant artillery fire for a very long period, 
 consequently, that it is most desirable to put it under cover, if possible, 
 whenever it is not actually firing. This is what has been done in the 
 design shown in Fig. 1, p. 76. The cylindrical form has been 
 adopted • the roof is made in one piece. There is little to fear from 
 vertical fire, and the edge of the plate is fitted to the vertical wall in 
 the manner shown in Fig. 1, p. 76, which, it is thought, favours the 
 throwing off of the^projectiles and dispenses with vertical armour 
 bolts. 
 
 As to metal, it is thought that the experiments of Cotroceni did 
 not conclusively establish the relative resisting powers of steel-faced 
 and wrought-iron shields. On the whole, however, it has been 
 decided to adopt steel-faced plates in the design before us — the 
 machinery for effecting the union of the face and foundation plates 
 being of such power that a high degree of perfection in the weld 
 may be expected. 
 
 The interior diameter of the turret will be 4' 6 metres (15 ft.). The 
 vertical wall is 1*2 metres high (3 ft. 11J in.), consisting of three 
 plates of 50 centimetres (19"69 in.), which are fitted together by 
 tongues and grooves. 
 
 The roof plates resting on the wall, as above described, are 20 
 centimetres (7 - 87 in.) thick. The whole armoured structure rests on 
 a wrought-iron structure and trunk, 5 metres in height, comprising 
 two cylindrical portions of 1 and 2 metres, and a conical portion {see 
 Fig. 1, p. 76). The cylindrical parts slide, during the vertical move- 
 ment of the turret, in two guiding rings ; these keep the turret in the 
 vertical position. 
 
 The base is supported on the piston of a hydraulic press, 45 centi- 
 metres (17*72 in.) in diameter, which constitutes the pivot of the 
 turret which moves on it. The centre of gravity falls well over the 
 pivot, and rollers can be dispensed with. The piston is 1*5 metres 
 (4 ft. 11 ins.) long, the movement being 1 metre only, half a metre 
 is always left in bearing. It is pierced from end to end for the 
 passage of the liquid used in the press. 
 
 The total weight is 150 tonnes (147-63 tons). The plunger being 
 45 centimetres (17*72 in.) in diameter, a pressure of 100 atmosphere 
 (0*66 tons) will support the whole structure. 
 
 The original element in the new design, which enables so large a 
 mass to rise quickly, consists in the counterweight of an accumulator, 
 of which the cylinder is in communication with the turret pivot press. 
 The plunger of the accumulator is double the size of that of the pivot, 
 consequently 7'5 tonnes (7*38 tons) is sufficient for the counterweight. 
 
 To destroy the equilibrium, water is admitted in the annular cavity 
 
76 
 
77 
 
 round the piston. The differential effect thus produced provides for 
 the vertical movement of the turret, which can be effected as quickly 
 as the rotation, seeing that in each case it is a question only of over- 
 coming inertia. Rotation is effected by means of the horizontal wheel 
 at the base of the turret {see A, Pig. 1, p. 76), by chain gear, or, 
 better still, by hydraulic gear. 
 
 The two subterranean chambers facilitate the working of the gun 
 by keeping the chamber from becoming crowded. The ammunition 
 is brought up to the level of the breech of the gun by a lift from 
 beneath. The carriage is entirely separate from the shield. The port 
 is reduced to a minimum size by a jointed frame forming the system 
 of parallel bars shown in Fig. 3. The gun and moving parts of the 
 carriage are supported on an oscillating hydraulic press {see Fig. 3) . 
 
 The recoil is met at the trunnions (as in some of Rendel's designs), so 
 that there is no disturbing couple generated by the shock of discharge, 
 
 The muzzle of the gun projects so little that disappearance is not 
 interfered with and can be effected in all positions. When elevated 
 for a long period the turret receives additional support brought under 
 the base by hydraulic power (B Fig. 1, p. 76). It then rotates on a 
 ring of balls. 
 
 Fig. 4, p. 74, shows the new cupola adoped by Major Mougin. 
 Little description is necessary, because the interior fittings are 
 practically the same as those employed by him in his cylindrical turret 
 at Bucharest. He has adopted the German mushroom form of dome, 
 but, undoubtedly, as above" noticed, the fact that his gun is separate 
 from the shield constitutes a fundamental difference. In this, Major 
 Mougin has a great advantage ; but it is difficult to see why Schumann 
 might not employ very much the same thing — Krupp has had recoil 
 arrangements closely resembling this one. Altogether, however, this 
 modified Mougin turret is excellent. 
 
 The disappearing turret must be commended for ingenuity and 
 neatness, but there are one or two points that seem weak or doubtful. 
 The turret is in its weakest position when in action, and it is question- 
 able whether it is not a little top-heavy and delicate. _ Any contorting 
 force would act with a powerful leverage on the piston. Another 
 objection appears to be the curtailment of the length of the piece. 
 
78 
 
 Further, it must be borne in mind that the protection due to dis- 
 appearance becomes less than might at first be supposed when the 
 turret is in action. In a siege the firing is very systematic, and a 
 besieger's guns would be pointed and waiting for the appearance of 
 the turret. The latter could not move up and down sufficiently quickly 
 to make feints and escape any fire. Generally speaking, the turret 
 guns and its adversary's would be loading during the same periods, 
 and unless the rates of loading differed much, the fire of the attack 
 might be interrupted only to a very limited extent. This, however, 
 must be decided by trial. Disappearing guns in pits are admirable, 
 but in a siege there may occur spots too hot for an open pit to be 
 tenable by a detachment, when a disappearing turret, or a plate covered 
 pit with a gun coming up through an opening might be very valuable. 
 As to vertical fire on the roof of either turret, the effect would be nil. 
 Our ships with 3-in. decks are in a very different position as to pro- 
 tection. A solid 7-in. plate is about five times as strong as a 3-in. one, 
 the resistance being proportional to the squares of the thicknesses. 
 
 Incidentally, the adoption of steel-faced armour for the new turret 
 is worthy of remark. It will be found to be suggested in this work 
 that if through fracture is to be avoided, steel-faced iron armour or 
 plates compounded of hard and soft steel will, in the long run, be 
 found better than solid steel. Here is a case in point ; the plates of a 
 turret of this kind constitute the structure itself, and through fracture 
 would be ruinous. It appears that under these conditions the French 
 authorities, who surely have the greatest experience with steel, do not 
 employ even the softest steel, nor do they, as we might suppose, 
 continue to use wrought-iron — although all the turret iron armour 
 behaved very well at Bucharest ; but they go out of their way to adopt 
 steel-faced plates. This, of course, is only experimental, but still the 
 selection for trial shows their present opinion. The French are not 
 likely to be disappointed in their steel. It has performed great things. 
 No wrought-iron — not speaking of compound plates— could have 
 behaved as well as the steel plate at Spezia in 1884. But any adoption 
 of steel must probably be accompanied by the acceptance of through 
 fracture and reliance on numerous bolts to hold the shield together. 
 Sheffield makers ought to be encouraged by the preference of steel- 
 faced armour for the new turret. 
 
79 
 
 CHAPTER IV. 
 
 Armoured Structures. — Forts. 
 
 Armour has been applied by our English Engineers to coast forts only, 
 though in some cases it may be specially valuable inland. Coast forts 
 would generally be assailed by heavier guns than would be employed 
 in a siege. On the other hand, an armoured fort inland would 
 be liable to be attacked more persistently and systematically than 
 could well happen in the case of a coast fort. Thus, it naturally 
 follows that wrought-iron is admirably suited for inland works, 1 while 
 chilled-iron has been chiefly employed for coast defences. 3 
 
 Under the head of chilled-iron, the principal classes of cupolas and 
 batteries of Griison's armour have been briefly described {vide p. 1, 
 Part II.), and it has been pointed out that England differs from other 
 powers in having her coast forts protected with wrought-iron armour, 
 which, although it may be in a measure re-inforced or superseded by 
 compound armour, will for a long time to come form the main fabric of 
 our shields. A brief description of the classes of the structures used on 
 the coast of England, together with what has been already given on 
 Griison's chilled-iron shields, will be sufficient for the purpose in 
 hand. The following is taken from a paper read by General Inglis at 
 the Royal Artillery Institution, which dealt much more fully with the 
 subject : — 3 
 
 The armour takes four forms. — (1.) Casemate shields huilt in masonry. 
 (2.) Shields in masonry, with guns on turntables. (3.) Iron batteries.*' 
 (4.) Turret. 
 
 1 Vide Bucharest trials. 
 
 2 The Alpine shields, noticed in p. 5, Part II., are perhaps almost the only case of chilled-iron 
 shields adopted for inland defence. The peculiar action of wrought-iron and its value against con- 
 tinued fire, as compared with chilled-iron, is illustrated in Krupp's Meppen trials of 1879, vide 
 p. 96, Part I. 
 
 s Read at E. A. Institution, April 29, 1880.— Vide Engineer, May 6, 1881, p. 326. 
 4 That is, a battery whose entire walls are iron-armoured. 
 
 34 
 
80 
 
 Casemate Casemate shields in masonry. — Fig. 1 gives a section of Picklecombe 
 
 and Garrison Point forts as a specimen of two-tier shield fixed 
 in masonry, showing however a comparatively weak sample. At 
 Garrison Point, Sheerness, the 38-ton guns have three 10-inch plates 
 
 Fig. 1. 
 
 Shields in 
 masonry 
 with gons 
 on turn- 
 tables. 
 
 m;;m^«^m 
 
 in front of them, instead of three 5-inch. Speaking generally, the 
 shields are built into the masonry in the following way ; — A strong 
 frame of f-in. plate, and 6-in. x 6-in. angle-irons is constructed in 
 the masonry, being built in, and attached by bolts, dove-tailed pieces, 
 &c. On this is built the armour, generally on the plate-upon-plate 
 system. Between the inner plate and the shield frame there is an inch 
 interval usually filled with wood. 1 The drawing explains itself. 
 
 Fig. 2, p. 81, shows a specimen of turntable battery (Class 2), 
 such as that at Helen's Fort, Portsmouth, Malta and Gibraltar. They 
 occur where a lateral fire extending through about 120° is required. 
 Each gun has two ports, and moves to either by means of the turn- 
 table. Here again the example taken is a weak one of its class, 
 although the structure is strengthened by intermediate layers at the 
 ports, and heavy iron-cased concrete supports between the ports. 
 Three 6-in. or two 9-in. plates are used in some places. 
 
 1 Each thickness of armour is bolted to the one next behind it, and the rear plate to the shield 
 frame. 
 
Shields and 
 turntable. 
 
 Omitting Plymouth Breakwater, which is a stronger but perhaps less 
 interesting structure, (Class 3) may next be taken. Iron batteries, of 
 which the two great forts of Spithead, Horse Sands and No Man's Ir j™ tteries 
 Land, are good specimens {vide 3). Here again may be seen the 
 
 Fig. 3. 
 
 Sti^iffiti&X : &&M:K&%?: 
 
82 
 
 Turret. 
 
 plate-upon-plate armour, coupled with the principle of keeping the 
 armour separate from the floor structure, so that any contortion of the 
 armour may not affect the racers or structural parts on which the 
 working of the guns vitally depends. There are armour bar supports 
 about 10 ins. x 5 ins. x 4 ins., the greatest section of each being 
 placed normally or radially to the wall, about four inches apart, 
 extending from the lower floor to the roof. The pier cases are filled 
 with Portland cement concrete. The safe load of the upper gun floor 
 and of the roof is equal to a mass of concrete laid 10 feet thick over 
 it, or half a ton on every square foot. The middle plates of the armour 
 stand on end, being 22 ft. 6 ins. long, extending from base to roof. 
 Concrete is used between the plates except at the port frames, where 
 wood is employed. It may be seen that here again it would be 
 easy to add outside rings of armour and masonry to correspond, for 
 which work, express provision is made. The weight of iron work on 
 these forts is as follows : — skeleton per gun, 50 tons, total 2450 tons : 
 armour wall per gun, 75 tons, total 3764 tons ; total weight, 6214 tons. 
 Each sea fort at Spithead, Spitbank and Plymouth is prepared for the 
 addition hereafter of turrets on its roof, each to carry two of the 
 heaviest ordnance. The Spithead forts can carry five turrets each ; the 
 others two each. Eope mantlets soaked in chloride of calcium, to 
 render them uninflammable, are supplied to all these works. 
 
 Class 4 turret is only employed on Dover Admiralty Pier. This is a 
 tremendous structure. The foundations are laid 7 fathoms below low- 
 water mark, and the guns are 33 feet above high-water mark. The 
 entire structure is therefore about 95 feet high. The turret consists of 
 a live ring and rollers of steel running on a path of steel laid on a 
 massive cylinder of masonry. On this live ring runs the iron frame- 
 work (Fig. 4), weighing 240 tons, containing the gun chamber. 
 
 Fig. 4. 
 
 The armour consists of three 7 -inch and two intermediate 2 -inch plates, 
 making in all a weight of 460 tons ; the guns, carriages and slides 
 make up the total of the running weight to 895 tons. This throws on 
 each of the thirty-two rollers of the live ring a pressure of about 
 28 tons. The entire diameter of the turret is 37 feet, the interior 
 
83 
 
 32 feet. The interior height of the gun chamber is 8 ft. 8 ins., the 
 height of the turret armour is 9 feet. There is a massive central 
 casting holding on to the masonry, and a thick cylinder of hammered 
 Bessemer steel inside this surrounds the wrought-iron built-up 
 cylinder, which forms the centre of the turret framework. The un- 
 absorbed portion of the blows of impact may be transmitted to this 
 part which is made specially strong. The pieces are 80-ton guns, got 
 in and out through the roof, of which parts are made removable for 
 this purpose. Provision is made for the escape of smoke through the 
 turret top. The gun ports admit of 7° elevation and 2° depression. 
 The glacis round the turret is plated with 5-inch and 3-inch plates. 
 The guns are loaded from under the glacis, the muzzles being depressed 
 for the operation. 
 
 Throughout the works, above described, the probability of continual Provisionfor 
 increase being called for, has been kept ever in view. This has morease - 
 augmented the difficulties under which the Engineers have laboured. 
 This system of coast defence may naturally be compared with that of 
 Griison. In doing so it must be borne in mind that breech-loading 
 has enormously facilitated the construction of strong forts, because, as 
 may be seen in Fig. 6, p. 6, Part II., it is only necessary to give 
 the men sufficient height to stand upright some distance back from the 
 wall instead of close to the port as with muzzle-loading. Hence, the 
 forts for muzzle-loaders must of necessity be carried in a cylindrical 
 form higher than those for breech-loaders, which may be curved back 
 into a dome, thus saving wall and giving greater strength to the roof. 
 Moreover, on most systems, 1 much more space is occupied by the 
 working of a muzzle-loader, which must be run back further than a 
 breech-loader. A conical turret with steel-faced armour was tried at 
 Eastbourne, in May, 1886, but is as yet in a very imperfect condition. 
 In justice to Griison, it must be noticed that casting lends itself to the 
 shaping of a dome with each part proportioned to the work that it is 
 calculated it may have to bear in a way that no system of plates made 
 by rolling or hammering could claim to do. Probably the English 
 forts altogether cost much more than those of Griison. 
 
 In their favour must be urged that they will bear continued fire 
 better, 3 and that it is very easy to increase their strength by adding 
 plates in front of their armour, for which complete provision is made, 
 even to the drilling of holes and insertion of nuts and temporary 
 bolts, to enable the operation to be performed without removing even 
 the front plate. Griison's cupolas must, it is supposed, be broken up 
 and re-cast in order to increase their strength. With the powers of 
 Artillery ever growing this is a very important consideration. Further, 
 it is obvious that the plate-upon-plate system favours the removal of 
 plates from one part to another readily if wished, an operation that in 
 certain cases might be desirable. Under the attach of a gun completely 
 beyond the power of the fort, the wrought-iron would be perforated, and 
 the shell would pass on doing mischief, though in the case of fairly 
 
 1 Eendell's system of lowering the muzzle and loading below the glacis may save both in height 
 above the port and in running back. 
 
 2 See pp. 7, 12, and 97, Part I. 
 
84 
 
 thick steel-faced armour it would be in the form of dead metal, for 
 shells that pass through such armour before exploding are far from 
 realization at present. 1 Chilled-iron, under these circumstances would 
 probably stop every part of the projectile, possibly at the price of its 
 own wholesale destruction, but so long as the wall is fairly whole the 
 men need not fear the next round. Lastly, the vertical walls of our 
 coast forts enable tiers of guns to be mounted one over the other, so 
 as to utilize the area of the foundation to the fullest extent. 
 
 Wrought- 
 iron. 
 
 Steel-faced 
 iron. 
 
 CHAPTER V. 
 
 Brief Summary of Behaviour of Various Kinds of Plates, 
 
 Backing, &c. 
 
 1. Wrought-iron yields locally ; it is punched or perforated, a clean 
 hole being made in it. 2 The rest of the target hardly suffers appreci- 
 ably, except close to the point of impact. The entire shield, including 
 bolts, is generally capable of resisting any subsequent blow as stoutly 
 as it resisted the first one. Effect here must be obviously produced 
 by the bare power of perforation of each round taken singly. 
 Partial penetration is practically useless, however often it may be 
 repeated. 8 The shot experiences but little resistance as its point 
 enters, and hence it is well enclosed and supported round its head 
 before the full strain comes on it. Hardness and rigidity of metal in 
 the projectile here tell to the greatest extent, and tenacity to the least. 
 These conditions have favoured the use of Palliser's chilled-iron shot 
 with sharp points. The plate yields at the back opposite to the point 
 of impact by tearing in a cross or star line, letting the point of the 
 shot through the centre. 
 
 2. Steel-faced wrought-iron, Cammell's, made on Wilson's patent, or 
 Brown's on Ellis' patent. Here the steel face, which constitutes about 
 one-third of the thickness of the plate, is a harder class of steel than 
 is generally supposed, 4 and resists the point of the shot abruptly, 
 and severely tries the tenacity of its metal. Hitherto, seldom has a 
 projectile held together under these conditions ; 6 there is a mechanical 
 force acting something like the outward thrust on the sides of an arch, 
 and the shot sets up, or more commonly breaks to pieces, leaving but 
 little metal lodged in the plate. The plate yields mainly by cracking 
 in radiating lines from the point of impact. In plates badly backed 
 the plate bends slightly back, and very rigid shot occasionally get their 
 points through ; but when well backed, the entire plate must be broken 
 
 1 The excellent Krupp shells, and those of Firming and Holtzer fired at Shoeburyness in the 
 summer of 1886 promise well, but a good deal has yet to be done before they carry bursting 
 charges through steel-faced armour, more especially on service, when they would hardly ever 
 strike normally. 
 
 2 Vide p. 12, Part I. 
 8 Vide p. 4, Part I. 
 
 4 Containing perhaps about 0'7 per cent, of carbon. — Vide p. 118, Part I. 
 
 5 Firminy and Holtzer forged steel projectiles fired from the 12-inch B.L. gun, passed through 
 16-inch steel faced plates without fracture in the summer of 1886. 
 
85 
 
 in pieces and displaced before a shot gets past it in any sense. In 
 addition to the radial cracks, concentric ones are apt to be developed, 
 sometimes, perhaps, owing to the tendency of the plate to give back 
 about the point of impact. Obviously, here tenacity in the shot is 
 called for in a much greater degree than with wrought-iron armour. 1 
 
 3. Solid steel, as made by Schneider 2 and used in many foreign Soua »teei. 
 ships, though rather less hard than the steel face of compound plates, 
 
 is as a mass more rigid. It admits the point of the shot at first with 
 less resistance, but it does not yield at the point of impact even 
 when badly backed, and is, therefore, less dependent on backing than 
 compound armour. As the shot enters, it wedges and heaps up the 
 metal round it, the plate coming forward and swelling at the point of 
 impact and yielding by radiating cracks. It may be questioned if a 
 case could be found of a concentric crack being made in steel armour. 
 Nearly all cracks made extend through the metal, but by the employ- 
 ment of a large number of bolts the fragments may be held up, and 
 considerable resisting power maintained after fracture. On the other 
 hand very soft steel plates admit of perforation, and in their character 
 resemble wrought-iron more or less. 
 
 4. Chilled-iron made by Griison, s and used almost universally in Qrfison's 
 foreign coast-armoured defences, is very rigid indeed. The shot iron, 
 seldom gets its point many inches into the metal. The shield 
 transmits the shock through its mass and must be broken up bodily. 
 Chilled-iron is used in large masses, and is best suited to resist single 
 blows, especially in an oblique direction. Under the direct blows 
 
 of heavy shot of high tenacity, chilled-iron breaks up. Cracks radiating 
 from the point of impact are formed, and the whole shield breaks 
 across. An important requirement in the shot appears to be tenacity, 
 to enable it to deliver its work on the point of impact before it breaks 
 up, which it does, leaving little or no metal lodged in the shield. 4 The 
 best forged steel projectiles shiver to pieces against Gruson's shields. 
 The hardness of the metal and the angle presented by the surface to 
 the line of fire alike tend to produce this result. 
 
 Backing strengthens the plate to resist perforation better when B^^g. 
 hard, but saves bolts and fastenings from fracture better when soft. 
 Thus steel-faced iron on granite has resisted perforation wonderfully. 
 
 1 For compound steel vide p. 90, Part II. 
 
 = Vide pp. 118, 133, 147 and 161, Part I. A very remarkable series of experiments was 
 made under Sir Nathaniel Barnaby's direction at Messrs Cammell's, Cyclops Works, Sheffield, in 
 Nov., 1880, on the bending of steel plates of thicknesses of an inch and less, when it was found 
 that steel bending well both at red heat and cold would in no single instance bend without tearing, 
 at blue heat. Cammell's, Brown's, Bolton, Parkhead, landore, Siemen's, and Steel Company of 
 Scotland steel being experimented on. The author called the attention of manufacturers on 
 different occasions to this, especially with a view to steel boilers. Messrs. Hawthorne's, Newcastle, 
 found evidence of the same tendency in iron. It may be observed that it is not proved that the 
 metal yields under less force at blue heat, but only that it prefers to tear rather than bend when 
 it does yield at that temperature. 
 
 * Vide pp. 1 to 16, Part II. 
 
 4 For chipping the face by concave pointed shot see pp. 13 and 16, Part II. 
 
BoUb. 
 
 86 
 
 {vide p. 61, Part I.). For iron, concrete, brick, &c, {vide p. 9, Part I.). 
 Wood layers between plates {vide p. 9, Part I.)- Air spaces between 
 plates of armour, have the effect of breaking up chilled-iron shot {vide 
 pp. 9, Part I., and 50, Part I.). 
 
 Bolls should resist sudden shocks and bear elongation well. The 
 early ships l had through bolts with conical heads holding into the 
 armour and screwed ends nutted against the skin {vide p. 5, Part I.). 
 French bolts had wood screws with projecting threads holding into 
 the backing {vide p. 5, Part I.). Palliser bolt with projecting 
 thread was proposed in 1862, and eventually took the form of 
 the Palliser-Eaglish, bolt with shank decreased to be free from all 
 nipping to prevent local strain and with spherical screw nut fitting 
 into hemispherical bottomed socket to accommodate itself to slight 
 lateral displacement {vide p. 6, Part I.). These are used to couple 
 iron plates together in pairs on the plate-upon-plate system. For 
 steel plates Schneider introduced a screw entering a short distance 
 into the back of the plate a and leaving the surface and, indeed, the 
 mass of the plate unperforated {ride p. 118, Part I.). These bolts 
 are screwed and nutted at the reverse end. 3 In cases where the 
 backing is thin the bolt would suffer, by being too short to give scope 
 for elongation. This evil is counteracted by using an iron pipe-box on 
 the back of the shield through which the bolt is made to pass. This 
 enables it to be prolonged to the necessary length to meet the required 
 stretch, without an unreasonable per-centage of elongation. A hollow 
 iron washer filled up with india-rubber is nutted on to the bolt end. 
 This bears on the end of the pipe-box. Bolts are now generally made 
 of soft steel. They are proved by being stretched by the fall of a 
 heavy weight on a frame held up by a bolt. 
 
 CHAPTER YI. 
 
 Brief Notes on Manufacture. 
 
 The manufacturing processes employed in making armour and pro- 
 jectiles form too large a subject to be attempted here, and in all 
 probability the most important features are just those that the 
 manufacturers would not wish to publish. Nothing therefore is here 
 attempted beyond a brief sketch, indicating the general character of 
 the processes of manufacture of different kinds of armour. They are 
 taken in historical order. 
 Wrought. Armour, wr ought-iron. — General Inglis gives the following description, 
 which is quoted verbatim : — 4 
 
 "For this, as made at Sheffield, the best cold-blast pig with a 
 proportion of Cumberland hasmatite is generally used. Roughly 
 
 1 The Warrior, Minotaur, &c. 
 
 2 At Spezia in 1882, Schneider's bolts only entered 2J inches into the 19-inch plate (p. 118, 
 Part I.). 
 
 3 For behaviour of shot, see manufacture, p. 101, Part II. 
 * Paper read at K. A. Institution, April 29, 1880. 
 
87 
 
 speaking, it takes 2£ tons of pig, or 2 tons of ball furnaced iron— that 
 is, of puddle ball — to make 1 ton of finished armour plate, and in the 
 last heating of all, when the iron is in a very costly state, there is a 
 loss by the actual burning away of the iron of about 10 per cent. It 
 takes also about 6| tons of coal to make a ton of armour. 
 
 Perhaps the shortest way of giving an account of the manufacture 
 of rolled iron armour will be to say how some one plate in particular 
 has been made, and for this purpose one of the heaviest ever pro- 
 duced is selected. 
 
 This was oue of the 8-in. plates made for the target to test the 
 80-ton gun at Shoeburyness. Its finished dimensions were — length 
 16 ft., width 10 ft., and its weight 23 tons. 
 
 For its manufacture 1170 slabs 30 ins. long, 12 ins. wide, and 1-in. 
 thick were made from the puddled ball and bar. These were piled, 
 furnaced, and rolled into 65 plates about 5 ft. square, and from 1-in. 
 to l|-in. thick, called small moulds. These, again, were piled and 
 rolled into quarter-moulds, and the quarter-moulds into armour-plate 
 moulds, and the pile for the last rolling was made up of three moulds 
 about 10 ft. square, two of 7 ins. and one of 3 ins. thickness. As 
 they entered the furnace for this last heating they weighed about 
 35 tons; as they came out after about 12 hours they weighed 31£ tons. 
 
 They were rolled down in the last rolling from 17 ins. to 8 ins. ; so 
 that, in a certain sense, this plate may be said to have been made by 
 rolling a column of iron nearly 100 ft. high down to one only 8 ins. 
 high. The fibrous character of a plate depends largely upon the 
 amount of reduction in the last rolling. 
 
 The rolls are plain cast cylinders ; those used in the present case are 
 about 3 ft. in diameter, 12 ft. long, and each one of the pair weighs 
 nearly 20 tons. They are driven by very powerful steam machinery, 
 and made to reverse their running, so that the plates are sent through 
 one way and then back again, and so on, the rolls being brought 
 nearer together each time, until the plate is brought down to the 
 required thickness. 
 
 The specific gravity of a good iron armour plate is about 7 - 625. x 
 
 Armour plate iron is not made for very high tensile strength, but it 
 
 1 Contradictory statements are made on this matter. Dr. Percy {Iron ani Steel, p. 868) 
 gives the specific gravity of wrought-iron before rolling as 7-877, and after rolling as 7'86. 
 Sheffield manufacturers have told the author that the specific gravity is increased by rolling, not 
 decreased. It will be seen however that the specific gravity given by General Inglis, 7'626, ia 
 considerably less than that by Dr. Percy, in fact, it agrees with what he gives for thin sheet, 
 namely, 7'6. This suggests that the plates are very well rolled, and that Dr. Percy is correct. On 
 the other hand, Sir P. Abel, in a paper read at the Iron and Steel Institute, in October, 1886, on 
 erosion, considered that steel of 7'859 specific gravity had been worked more than one of 7-847. 
 If however the specific gravity is decreased it must be presumed to be due to the fact that the 
 crystals being elongated enormously, the intervals between them occupy a larger proportion 
 of the area of any cross section. There seems nothing unreasonable in this, though doubtless 
 Sir J. Whitworth would have considered that it supported his view that wrought-iron con- 
 sists of faggots of iron glued together with cinder. This fibrous structure ought however to offer 
 advantages. Dr. Percy states that it is twice as easy to break it along the fibre as across it. As 
 the fibre would run with the length of the plate it would give the plates the maximum strength in 
 the narrowest direction and make it very difficult to break them across. 
 
 35 
 
88 
 
 Laminated 
 
 armour. 
 
 Steel-faced 
 iron 
 
 is essential that it should draw out well, and reduce in cross section 
 some 20 or 30 per cent, before breaking. 
 
 It has been our practice to prove a very large proportion of the 
 plates made for iron fortifications by shot capable of nearly perforating 
 them. Pieces of those which could not be thus proved have been 
 broken under hydraulic presses, and otherwise treated. 
 
 Armour bolt-iron is similar to armour plate-iron (but soft steel is 
 now much used). We require reduction of area of fracture of at 
 least 40 per cent, when broken by falling weight test, and we generally 
 get a larger reduction than this. 
 
 In the early days of armour, laminated armour was used in America 
 to meet the requirements of the moment. This consisted of thin 
 sheets of iron bolted together. These sheets, sliding on each other, 
 facilitated the work of punching, especially by pointed projectiles, 
 because the corners of the torn metal bent back much more easily 
 than is the case with solid plate, and accommodate themselves 
 to the bend. The measure in which plate-upon-plate armour is 
 inferior to equally sound solid plate must be explained on the same 
 principle, a slight inferiority which is thought to be amply compensated 
 by the advantages of breaking joint, and larger surfaces with fewer 
 joints, &c. 
 
 A compound or steel-faced plate consists of a wrought-iron " foundation 
 plate " with a steel " face plate " of about half its thickness attached 
 to the front, thus giving the combination of hard surface 1 and tough 
 back or body. There are two processes on which this class of armour 
 is made namely : — Wilson's, carried on by Messrs. Cammell at the 
 Cyclops Works, and Ellis', by the firm of Sir John Brown and Co., at the 
 Atlas Works, both at Sheffield. 
 
 Wilson's system was first brought into practice, and has been taken 
 up by several firms on the continent. The foundation plate con- 
 sists of thinner plates each made up separately and finally united, 
 but not thoroughly worked up and amalgamated. It is thought 
 better to keep up the laminated structure sufficiently to cause these 
 layers to break and yield separately, and not in one continuous line. 
 It will generally be found that when a shot destroys a plate of this 
 kind, the layers tear open in steps, any breach made becoming larger 
 and larger as it reaches the back of the plate. It is held that increased 
 resistance is offered by a fracture of this kind, and much more support 
 afforded by the backing. 
 
 When the foundation plate has been completed it is at once cleaned 
 of scale, passed into a mould and secured at its edges. The mould is 
 then rotated so as to be up-ended, and steel is poured in to form the 
 face plate. After this, the plate is rolled and finished. 
 Effis'System On Ellis' system, the foundation plate is more completely worked, 
 and when ready for the steel face, it is laid horizontally, and the 
 thin steel plate which is to form the front of the shield is laid 
 on it, being held up at the proper distance to allow of the molten 
 
 Wilson's 
 System. 
 
 1 At Spezia in 1882, Brown's eteel-faoe had 0-7 per cent, carbon and Cammell's 0-65, vide pp. 117 
 and 118, Part I. 
 
89 
 
 steel entering between the two, by resting on a number of steel pins 
 standing on end. The two plates in this state are passed into a 
 furnace, and when heated are brought out, placed in a mould, and 
 steel is run in between them after they are brought to the vertical 
 position, as in the Wilson process. The steel pins are so far welded 
 to the plates in the furnace that there is no fear of the two plates 
 altering their relative position, or the pins becoming dislodged, and 
 it is probable that they are melted by the molten steel when it is run 
 in. The plate is then finished by rolling to final dimensions. When 
 a plate is required to be tapered towards one edge it is necessary to 
 narrow the space for the molten steel at this part, and to make 
 the iron proportionately thick, for working purposes. Finally, the 
 extra iron is cut away from the back. Fig. 5, herewith, 1 shows the 
 running in of the steel on a foundation plate at Brown's works. A 
 view of the same process at Cammell's would resemble it. 
 
 Pig. 5. 
 
 For Wilson's process is claimed special soundness in the body of steel, 
 and power to remove the scale, which interferes with the union of the 
 steel and iron, and also that one re-heating is saved. For Ellis' it is 
 
 i Taken from Major Maokinlay's paper to the E.A. Institution, p. 601, Vol. XII, 
 
90 
 
 claimed that the face is specially good. This is illustrated by the sharp 
 angles at which the plates may be bent without cracking, angles at 
 which wrought-iron is liable to give way. Under fire, Wilson's plates 
 are softer than Ellis'. Thus penetration is deeper and cracking less in 
 the former than in the latter. Great stress is laid on want of homo- 
 geneity by both makers. This should be understood, because this 
 quality, which is aimed at and claimed as a virtue, is pointed out as an 
 evil inseparable from the system by some writers, especially by Lieut. 
 Jacques, U. S. Navy, in an able pamphlet termed " Modern Armor 
 for National Defence." It is believed that while it is desirable to 
 unite the surfaces of the steel and iron as perfectly as possible, the 
 different nature of the substance of the iron causes cracks to stop 
 abruptly, which would otherwise have run through the mass. Figs. 1 
 and 2, p. 91, are remarkable illustrations of this action. They 
 represent Wilson's steel-taced plates tested at Amager, on Aug. 6th 
 and Sept. 10th, 1885, 5 it. square by 1T5 inches thick. The steel was 
 4^ inches, and the iron 7$ inches thick. The plates were actually cut 
 in two to furnish sections after the trial. Fig. 1 shows the effect of a 
 10-inch chilled-iron Ankarsrum projectile, weight 18 1 '5 kg. (400 lbs.), 
 with a striking velocity of J23 - 6 tn. (lo'.iOfeet) an energy of 16(30 
 metre-tons (5358 foot-tons), a perforation of 13'4 ins. of iron, an 
 energy per ton of plate of about 1040 foot- tons, supposing the plate 
 to weigh 5'15 tons. A few face cracks were formed, and the behaviour 
 of plate in stopping the shot is most satisfactory. 
 
 Fig. 2 is a more important example, because a forged steel projectile 
 was fired weighing 182 p 5 kg. (402 lbs.) with a striking velocity of 
 438 m. (1437 ft.) an energy of 1784 metre-tons (5760 foot-tons) a 
 perforation of 13 - 9 ins. of iron, and energy per ton of plate of 1118 
 foot-tons. The steel shot would deliver a much larger proportion of 
 its energy on the plate before breaking up. A plate seldom survives 
 a single blow of 1000 foot-tons per ton of plate. The behaviour of 
 this plate, which is only slightly cracked at the back, is admirable. 
 
 It naturally suggests itself that compound plates might be made 
 with very hard steel faces, and backs of soft steel instead of iron, 
 seeing that soft steel has both greater elongation and also greater 
 tenacity than iron. Experiments of this kind have been made at 
 Sheffield but without success. 
 
 Sir J. Brown and Co. once tried a steel front and back with an 
 intermediate layer of wrought-iron. The steel, however, was not soft 
 at the back, and no great stress should be laid on this trial, which was 
 merely one to test the effect of a break in homogeneity, on a steel 
 ^e™if ound plate. Two more important experiments, however, were made as 
 follows : — Messrs. Cammell made a plate, on the compound principle, 
 wholly of steel — that is to say, with hard steel in front and soft steel 
 behind. The steel face had a breaking strain of about 47 tons per 
 square inch, and an elongation of about 1*5 per cent. ; whereas the 
 back had about 23'0 tons tenacity, and 25 per cent, elongation. This 
 plate was fired at on board the Nettle, at Portsmouth, on March 14th, 
 1884, under an experimental title. Its dimensions were 8 ft. by 6 ft. 
 by 10\ ins. thick. It was attacked by a 10-inch muzzle-loading gun, 
 
93 
 
 F.o.1. 
 
 m 
 
 • 
 
 mm 
 
 IRON 
 
 P.EEJJ 
 
 ikqh. 
 
92 
 
 with a chilled-iron projectile weighing 400 lb., and a striking velocity 
 of 1364 ft. This would give a total striking energy of 5160 foot-tons, 
 or about 573 foot-tons per ton of plate, supposing the plate to weigh 
 about 9 tons. The calculated perforation would be about 13 ins. of 
 iron, or 9'8 ins. of compound armour or steel. Three blows were 
 delivered on it. The plate stopped the shot well, the projectiles only 
 entering to a depth of about 5 ins. Nevertheless, it did not behave 
 in a way that would meet with approval with English authorities, seeing 
 that it cracked all over in through cracks — vide Figs. 1, herewith, and 2, 
 p. 93. It remained held up by its bolts, but was divided into pieces, 
 and was not considered sufficiently satisfactory to encourage further 
 trials in the same direction at the time. 
 
 It may be here said that, on the first success of steel at Spezia, it 
 was proposed to try it in this country ; but our combatant naval 
 
93 
 
 advisers positively objected to it, oa the ground that it stripped off, 
 and would leave a ship's side bare. It is easy to employ a great 
 number of bolts in a target, but on a ship's side the multiplication of 
 bolts involves expense and weight in the supporting structure. Thus 
 steel-faced armour was only approved on its fulfilling certain rigid 
 conditions — namely, that no through cracks shall be made by the first 
 round fired ; it being also stipulated that no one of three projectiles 
 shall get through the plate. 
 
 fie 2. 
 
 
 
 Ml 
 
 
 'MM 
 
 Iff 
 
 mm 
 
 mm 
 
 %* 
 
 m 
 
 
 
 The experiment with compound steel is especially to be noticed, 
 because it had embodied in it the elements above suggested — namely, 
 a hard face, and a back consisting of steel of greater elongation and 
 higher tenacity than the wrought-iron usually employed, whoso 
 
94 
 
 tenacity is generally about 17 tone, aud its elongation 18 per cent. 
 A single trial is doubtless hardly likely to meet with a large measure 
 of success. Still, Messrs. Cammell are old steel makers, and their 
 first attempt, based on manufacturing investigations, ought to be very 
 good for a first trial. The exact measure of success, however, is not 
 the matter of real significance, but rather the fact that will appear 
 further on— that the behaviour of this steel plate very closely resembled 
 that of Messrs. Schneider's steel plates at Gavre, and would probably 
 have passed and been accepted by the French Government. So long 
 as such a plate meets with disapproval from our own authorities, there 
 is very little to encourage the further trial of steel. But of this, more 
 
 The second significant experiment above-mentioned, was the trial of 
 a compound plate by Sir John Brown and Co., with a wrought-iron 
 back differing from the usual sample, in the fact that the tenacity and 
 elongation of the iron used were both considerably greater. Con- 
 sequently it might be expected that the plate would have been both 
 stronger and tougher than usual. It happened, however, that it 
 cracked instead of bulging. This result, taken in conjunction with 
 that obtained with Cammell's steel, seems to show that low tenacity is 
 a necessity for the metal in the back of a plate, if it is to be kept from 
 cracking through. Probably the action of the blow is too rapid to 
 admit of the plate bending and stretching, unless the tenacity is low. 
 The great power of elongation is consequently useless under impact, 
 unless the plate elongates with but little resistance. 
 
 Steel armour has been made experimentally in this country by Sir 
 J. Whitworth (vide p. 68, Part I.), and also in thin plates supplied 
 for service, as in the turtle back scales which cover the Polyphemus 
 (vide Fig. 12, p. 114). The best heavy steel armour, however, has 
 been made abroad at Messrs. Schneider's works at Creusot. The 
 steel is not generally so hard as that employed in the face plates of 
 compound armour. 1 It is hammered down from a great thickness 
 under a 100-ton hammer. It is annealed and the face tempered in oil, 
 so it is said — but it is difficult to get authentic information — and 
 finally annealed again. It is difficult to do justice to steel armour in 
 . England. Messrs. Schneider not unnaturally decline to supply plates 
 
 steel for competition in this country unless a large order is guaranteed to 
 
 them. It is desirable however to say something on the subject. It 
 appears that solid steel plates can be made soft, to behave more or 
 less like wrought-iron, or they can be made hard, in which case they 
 resist penetration well, but break through. Hence it will be found 
 that Schneider insists on a great number of bolts, which hold up the 
 fragments and enable the plate to offer great resistance after it is 
 broken through. 
 Givre trials. Photographs may be seen of plates fired at at Gavre for the French 
 Government, which were passed as samples of lots to be accepted, which 
 would probably have been rejected by our own authorities. They are 
 plates of 18 inches thick and over. Unfortunately we have tried our 
 
 1 At Spezia, in 1882, Schneider's plate was said to contain about 0'45 per cent, of carbon. It 
 was batzunered down from a thickness of 7 feet to one of 19 inches (vide p. 118, Fart I.). 
 
 Solid steel. 
 
95 
 
 18-inch plates but seldom in this country. There has no gun larger 
 than a 10-inch muzzle-loader been fired on board the Nettle we believe ; 
 and, strange as it may appear, our supply of 18 in. plates can hardly 
 be said to have been proved at Shoeburyness or anywhere else. 18-in. 
 experimental plates have been attacked at Shoeburyness, but this is 
 not the same thing as the sample plates being tested and passed in the 
 usual way for each lot supplied to our ships. The omission of a test, 
 however, is one thing ; the approval of plates of proved inferior power 
 is another. Figs. 3, herewith, and 4, p. 96, show the front after the 
 2nd and 3rd blows, and Fig. 5, shows the back of a solid steel Schneider 
 plate, the sample of the second lot of plates which were tested for the 
 French Government at Gavre and delivered for the Terrible during 
 1881 and 1882. This sample is selected because a view of the back 
 as well as of the front is taken. The thickness of the plate was pre- 
 sumably 19 - 5 ins., the plate being for the belt of the ship. It was 
 attacked by a 32 cm (12 - 6-in.) gun; the weight of the projectile was 
 
 therefore probably 761 lb. The striking velocities of the three rounds 
 1430-8 ft., 1430-8 ft., and 1504 ft. 1 The projectile was a chilled shot, 
 as may be seen in the engraving, from the white radiating splashes of 
 metal. The gun is not a match for the plate. The perforation would 
 
 i E = 11930 foot-tons. 
 
 36 
 

 Fig. 5. 
 
 
 
 IB I' 
 11 I 
 
 
 
 WSmm: 
 
 '^M 
 
 m&mmm 
 
 
 Kf 
 
 i^i^^^ 
 
97 
 
 be only equal to 17-8 ins. of iron, or about a 13'36-in. compound or 
 steel plate. Suppose that the plate only weighed about 21 tons, the 
 shock would not be greater than that on the Wilson compound steel 
 experimental plate above-mentioned. The plate, however, probably 
 weighed 27 tons, and if so, the shock per ton was but small. Every 
 plate tried for the Terrible appears to have cracked through on the first 
 blow, and thus would have been rejected if tried at Portsmouth. An 
 18-in. compound plate at Shoeburyness received a blow nearly corres- 
 ponding to this "on July 21st, 1880, (vide p. 100, Part I.) with 
 insignificant effect from the 12 - 5-in. gun, with a chilled-iron projectile 
 striking with a velocity of 1504 ft., the weight being 828 lb., and an 
 energy per ton of 541 foot-tons. 
 
 Both Brown and Cammell's plates, when below 1 2 ins. thick, have 
 borne an amount of work out of all comparison with this. 1 An 11-in. 
 plate of Brown's, 10 ft. by 5| ft. in area, at Shoeburyness received 
 three blows from 9-in. chilled shot, one blow from a 9-inch steel 
 projectile, and two blows from 12 - 5-in. (38-ton gun) steel projectiles 
 weighing 840 lb. and 845 lb. with 1425 ft. and 1413 ft. velocity and 
 19 - 8 in. perforation, and about 11,820 foot-tons and 11,690 foot-tons. 
 In all 38,471 foot-tons, or, taking the plate at 11 tons weight, 3497'4 
 foot-tons per ton of plate. Discs of metal were nearly or quite separated 
 by the two last blows ; but the backing was sufficient to stop them, and 
 the plate held fairly together except one corner. This, of course, was a 
 result beyond all comparison with what we have given above, for here 
 an 11-in. compound plate held together and stopped blows which were 
 probably as severe as those which broke up the French steel 19'5-in. 
 Terrible plate, the respective energies are 11,820 and 11,930 foot- 
 tons, but the English plate was attacked by cast steel and the French 
 one only by chilled-iron projectiles, so that the former probably 
 delivered the more energy into the plate before breaking up. It is 
 fair to remember, however, that only a limited number of results from 
 Gavre come to us, and that these are doubtless worse than would now 
 be obtained ; although, as we have nearly the whole of the samples tested 
 for the Terrible, they represent fairly the quality of the supply at that 
 time. It is quite possible that some very excellent samples might now 
 be produced from Gavre, although hardly likely to be better than one 
 which is to be noticed presently. The Schneider plate tested at Spezia 
 in 1882 is enormously better than the Terrible samples. This plate had 
 no visible crack on the first round, and it received in four blows 3882*2 
 foot-tons per ton of plate. It was, however, by that time broken up, 
 the upper portion of the plate hanging loosely out of its place, and 
 large fragments detached altogether (vide Fig. 19, p. 126, Part I.). 
 
 The 1884 Schneider plate allowed the shot to pass clean through it, 
 so that, excellent as the plate appeared, it is impossible to argue 
 definitely on it. To do Schneider full justice, then, take the case 
 specially selected and put forward by Lieutenant Jacques, as the 
 crowning success in the shape of st6el plates fired at at Gavre, viz. 
 
 ' Vide p. 108, Part I. 
 
98 
 
 that tested for the Admiral Baudin, on March 7th, 1884. (Fig. 6). 
 Here three rounds were fired at a plate 12 ft. 9 ans. long, 8 ft. 2 ms. 
 wide, and from 18 ins. to 20 ins. thick, weighing 35-45 tons. The 32™ 
 (12-6-in.) gun attacked it with three chilled-iron projectiles, the two first 
 striking it with velocities of 1518, 1518 ft., and on places 18"5 ms. 
 thick, and the third one with 1593 ft., on a place 19'8 ins. thick. The 
 
 Bfteot of 
 bolts. 
 
 FIO 6 
 
 last gives a perforation of 19'44 ins. of iron, or 14'58 ins. of compound 
 on steel. 1 The plate looks well in the drawing, but the test is not a 
 severe one ; chilled projectiles were employed, the heaviest blow being 
 only about 378 foot-tons per ton of plate, and the whole three only 
 making up 1064 foot-tons per ton of plate. The shield, moreover, is 
 apparently broken entirely through into two pieces, and only preserves 
 its fine appearance from being well bolted up ; and this may lead on 
 to the great distinguishing characteristic of solid steel as compared 
 with compound armour when both behave fairly well. This, which 
 has been before stated, appears to be as follows : — Steel armour stands 
 up rigidly, breaking up in through cracks, but if well bolted remains 
 in its place. 
 
 How successfully bolts may be employed may be seen by an 
 inspection of the back of the steel plate given in Pig. 5, p. 96, in 
 which it may be observed that the cracks seldom run through the 
 bolts. With this experience it is no wonder that Schneider insisted 
 on having twenty bolts in the 1882 trial, when Cammell and Brown 
 competed with only six bolts each. Schneider was undoubtedly 
 
 i E = 18390 foot-ton«. 
 
99 
 
 right, Oammell and Brown should have had the same; but their 
 
 encouraging experience with smaller plates and the limited firing trial 
 
 of 18-in. plates rendered them more liable to be caught in this way 
 
 than Schneider. Now to consider another point — namely, the element 
 
 of certainty in behaviour. The variation in the quality of steel is 
 
 probably always greater than in iron ; bat there is another shape in 
 
 which the caprice of steel is shown — namely, in spontaneous fracture. 
 
 The extraordinary statement is made on authority which ought to be 
 
 good, that out of the steel plates supplied for the Terrible — about 
 
 ninety in number — eighteen actually cracked spontaneously. Of these, 
 
 fourteen broke at Creusot, either after tempering or after annealing ; 
 
 one cracked at the time of dispatch to the port, and three broke 
 
 while in the port of Brest ; but before they were placed on the vessel. 
 
 If this is true, the Terrible is clad in very capricious armour, armour 
 
 which would break and crack wholesale under fire, and possibly would 
 
 develop cracks in severe changes of temperature. If, however, it is 
 
 bolted on the ship as well as Messrs. Schneider generally bolt it, it 
 
 would protect the ship well under heavy blows in spite of its cracking. 
 
 On the other hand, the continued fire of comparatively light guns 
 
 ought to produce an effect on it disproportionate to their power. It 
 
 must be admitted, that a steel 18 - 9-in. plate used for proof of 5 - 9-in. 
 
 projectiles at Spezia recently received a great many blows without a 
 
 great loss in resisting power. It cracked through in broad lines, 
 
 but if well bolted it would resist any fire of 6-in. guns likely to fall on 
 
 it. Still, 5'9-in. guns are very insignificant pieces to attack an 18'9-in. 
 
 plate. After considering all the facts above discussed, what are we to 
 
 conclude as to the relative positions of steel and steel-faced armour ? 
 
 In spite of a natural preference for English manufacture, it will be soUd steel 
 seen that in the earlier pages devoted to this subject, written before pound™" 
 the above information had been obtained, it had been contemplated P late8 
 that steel-faced armour would be displaced by steel — steel on the oom 
 compound principle perhaps, but steel in some form, sooner or later. 
 Not so much on account of actual results, as because of the great 
 possibilities of steel generally. Just as steel had displaced iron in 
 boiler plates, guns, and projectiles, so it seemed likely to do in armour. 
 Confidence was not expressed in steel armour, however, in an 
 unqualified way, because, although there had been disappointment 
 in some larger compound plates, and better steel plates had been 
 made on a larger scale, compound plates of 12-in. or below had 
 repeatedly beat steel in competitions. Now facts have been brought 
 to light that raise the question whether steel-faced iron plates may 
 not finally prove the best, and certainly forbid admitting that steel 
 has shown itself superior up to the present time. 
 
 The facts referred to are : — (1) The cracking of our own steel 
 experimental plates, made with soft steel backing, and still more the 
 cracking of iron backing when it possessed the very qualities supposed 
 desirable — that is, high tenacity coupled with great elongation. 
 (2.) The records of the Terrible plates, and the conviction that, even in 
 the selected specimen of the Baudin sample, the plate is broken 
 entirely asunder by comparatively light blows, and those with chilled- 
 
100 
 
 iron shot, which have long since, even in England, been pronounced 
 inefficient for the attack of steel or steel-faced armour. If it be 
 denied that the plates are cracked through, why is not the back 
 exhibited ? And if the plates will bear really severe tests well, why are 
 guns of such disproportionate power used to prove them, and those 
 with chilled projectiles ? (3.) The strange account of the spontaneous 
 cracking of steel plates. (4.) The last victory of the compound plate 
 over the steel at Pola, showing that on the scale on which compound 
 plates have been continually tested and best worked, steel has never 
 been able to compete with them. These facts ought to be met with 
 actual facts and figures before we are justified in passing them over. 
 
 Happily we are not greatly concerned as a country which kind of 
 armour is eventually adopted. Cammell and Brown have the plant 
 for making steel plates to-morrow, if desired. 
 
 At Brown's a gigantic press is in course of erection, capable, it 
 is thought, of turning out sounder steel plates than could be 
 obtained by hammering. There is little prospect, however, of steel 
 displacing steel-faced plates, especially as a Brown experimental 
 steel-faced plate has recently (April, 1887) stopped a Holtzer shell 
 with remarkable success. 
 
 It would be right to try steel in this country. The question of 
 superiority is quite an open one yet. The prospects of success for 
 solid steel would, under present circumstances, much depend on the 
 power to give it a harder face than has hitherto been given to it, 
 coupled with a softer back. It must also be shown to be free from 
 the liability to spontaneous fracture or brittleness from atmospheric 
 conditions. There is something pitiful in the thermometer being con- 
 sulted, and in the circumstance of a steel plate having to be warmed 
 before firing at it, as it is stated took place at Amager in the winter 
 of 1883-84. Should the Admiralty change their minds, and conclude 
 that the fact of plates breaking through does not signify so long 
 as they are held in their place by bolts, the prospects of steel in this 
 country will be much improved; but compound steel ought in the 
 long run to have a better chance than solid steel. So long as we 
 expect to take down our plates intact and look at their backs after 
 impact ; and further, if we test them under oblique fire where a hard 
 surface tells most, and if we back them well, probably solid steel is a 
 long way from competing successfully with steel-faced plates. 
 
 At some time steel may probably become more trustworthy, it may 
 possibly, eventually, attain the property of elongating instead of crack- 
 ing under the blows of shot ; but this is almost saying what we all 
 know, that steel is daily made more and more like wrought-iron in a 
 pure form. With this a hard face does not seem attainable. At 
 present those who employ armour on their ships must take a practical 
 view of the matter and choose between the properties possessed by 
 the two kinds as they are now presented to us, or are likely to be 
 presented for some time to come. Compound armour, with a steel face 
 and soft iron bad, is likely to hold its own as the safest and most 
 reliable plate for ships under present circumstances, while for facing 
 granite or other hard material it has shown extraordinary powers. 
 
NOTE AND COBKECTION, ON P. 100. 
 
 The projectile mentioned in Line 19, p. 100, Part II, was a Firminy 
 not a Holtzer Shot. A Brown plate stopped a Holtzer experimental 
 Shot in November, 1886, but not one of the lot mentioned, p. 103, as 
 might have been inferred from the above statement. 
 
 On April 1st, 1887, a 6-in. Hadfield forged steel projectile fired 
 from a worn gun missed the mark, and buried itself in side structure 
 and granite, slightly injuring the point. The same projectile again 
 fired, perforated a 9-in. compound plate and 2-ft. of oak, and penetrated 
 32 inches into granite, when it eventually broke, but was not set up. 
 
 The ballistic powers taken from the results hitherto obtained with the 
 
 projectiles of the heaviest representative English and German 
 
 BL Guns, are as follows — 
 
 
 Weight 
 
 of Gun. 
 
 Tons. 
 
 "WEIG 
 
 Shot, 
 lbs. 
 
 HT OF 
 Charge 
 lbs. 
 
 Diameter 
 
 of Shot. 
 
 in. 
 
 W 
 D 3 
 
 V 
 ft. sec. 
 
 B 
 ft. tons. 
 
 Krupp 
 
 119 
 
 2028-25 
 
 846 -6 
 
 1 5*75 
 
 0-519 
 
 2000 
 
 50780 
 
 Elswick 
 
 110| 
 
 1800 
 
 850 
 
 16-25 
 
 0-420 
 
 2148 
 
 57580 
 
 Royal Gun Factory 
 
 67 
 
 1250 
 
 630 
 
 13-5 
 
 0-510 
 
 2025 
 
 355G0 
 
 The powers of the Krupp and Elswick Guns are thus brought more 
 nearly equal than shown on Tables XV. and XVIII. 
 
101 
 
 The following is a list of the naval Powers employing steel and com- Naval 
 pound armour at the present time :— England, compound ; Argentine ^Tying 
 Kepublic, compound; America, compound, but most likely steel will f n e d el 8 "(^i d 
 be tried ; Austria, compound ; Brazil, compound ; China, compound ; armour. 
 Denmark, compound; France, compound and steel; Germany, com- 
 pound ; Holland, compound ; Italy, Italia only compound, the other 
 vessels steel ; Bussia, compound ; Spain, steel recently ordered for one 
 ship, compound talked of for another ; Sweden, compound and steel. 
 
 The makers are for steel : Schneider, at Creusot. For compound (on 
 Ellis' patent) Brown, at Sheffield : (and on Wilson's patent), Cammell, 
 at Sheffield ; Marrel, at Bive de Gier, Loire ; La Compagnie Anonyme 
 des Forges, at Paris; La Compagnie des Hauts Fourneaux, St. 
 Chamond ; the Dillingen Works, in Germany ; and Tjora Iron-works, 
 Kolpino, St. Petersburg. 
 
 Projectiles for the attack of armour have been made of ordinary Projectiles, 
 cast-iron, wrought-iron, chilled cast-iron, and steel, and compound 
 shot of chilled cast-iron in conjunction with steel have been used 
 experimentally. 
 
 Ordinary cast-iron soon proved ineffective for thick armour. It 0rdin t ar y 
 
 ■, -. -in* i cast-iron. 
 
 may, however, be employed in common shell against thin armour. 
 Common shell have perforated unbacked iron of a thickness equal 
 to half the diameter of the shell (vide pp. 51 and 52, Part I.). 
 One common 9 -inch shell from the Shah perforated the side of the 
 Huascar (4-inch plate) and burst in the backing. Wrought-iron 
 shot hold together but spread out against the plate, it is scarcely 
 likely that they could be used with advantage even for racking 
 purposes (vide pp. 13 and 14, Part I.). Steel shells with heads 
 screwed on were proposed by Major-General Alderson in 1865 
 and acted fairly well. Chilled cast-iron were brought forward Paiuser 
 by the late Sir W. Palliser in 1863 1 and have been made in the Boyal project*™. 
 Arsenal and other factories almost without interruption since. They 
 are made from special mixtures of iron, Bidsdale, Cwmbran, pig and 
 broken projectiles have been principally used, doctored at times with 
 " ilmenite." The projectiles are cast point down in moulds of iron 
 for the heads and sand for the bodies. By means of the conducting 
 power of the metal the point and head become sufficiently chilled to 
 cause the iron to be white, while in the body the carbon separates to a 
 certain extent. 2 Chilled metal is specially hard and dense and has 
 great crushing strength, but it is brittle, the body is intended to have 
 more tenacity. The hardness and crushing strength of chilled metal 
 remain almost unrivalled to the present time. The points never 
 flatten, and the metal remains remarkably cool after impact. The best 
 
 1 Vide p. 4, Part I. Lieut. -Colonel Barlow, R.A., contributed a paper on "The manufacture 
 of chilled and other Projectiles in the Royal laboratory," in E.A.I. Proceedings, Vol. IX., No. 3, 
 1875, to which reference niay be made for details. 
 
 2 When in the Eoyal laboratory, about 1866, the author, in a very limited imperfect way, com- 
 pared the conditions of the same metal when chilled and cast in sand by taking specific gravities 
 and acting on them with nitric acid. The specific gravity appeared to be increased about 3| per 
 cent, by chilling. 
 
102 
 
 steel has only recently been made to stand as well up to its work, at 
 the point, where compression only is concerned. Chilled projectiles 
 fracture very easily if exposed to oblique blows in store, and are 
 liable to crack from molecular action. In short the metal is in a 
 constrained unnatural condition, but its excellence for punching work, 
 especially in soft armour, has caused it to be tried in points of shot 
 which are made of steel or strengthened by steel. 1 The peculiar weak- 
 ness of chilled metal was remarkably well illustrated in its pulveriza- 
 tion in air space targets. 2 Chilled shot depend on getting their heads 
 into the armour before the full shock comes on them , and are apt to 
 break up badly against the hardest kinds of plates, that is against 
 chilled-iron shields, and against solid steel. Against steel-faced plates 
 chilled-iron in this country has occasionally produced better results 
 than could have been anticipated. 3 Nevertheless, seeing that armour 
 is likely to become increasingly hard and that chilled-iron and solid 
 steel are much used abroad, the supersession of chilled-iron 4 by steel 
 supersession has been obviously inevitable; the manufacture of chilled projectiles 
 ?r f on b b led ' was discontinued in the Royal Laboratory in October, 1886, though 
 steel. since resorted to, to meet immediate demands. For sections showing 
 
 projectiles chilled in different ways vide p. 89, Part I. In judging 
 of the harshness of fracture, novices are specially warned against 
 the fracture shown by pieces chipped off near the surface, which 
 often exhibits a beautifully fine silky texture, while a deep normal 
 fracture is much coarser. Steel has been sometimes combined by 
 mixing, by melting up with iron in chilled-shot, both by Krupp 6 
 and in the Royal Laboratory, as well as in Wilson's shot, where the 
 point is cast of chilled-iron and the body of steel, 
 steel pro- Steel projectiles have been made of innumerable kinds both forged 
 
 jeoles ' and cast. Aldersou ; s, mentioned above, were some of the first tried. 
 As it is impossible at present to specify any distinct steel shot or shell 
 as service projectiles, it is better to say very little generally and refer 
 to each steel projectile under the head of particular experiments. 
 Whitworth's forged steel projectiles at first performed the most 
 remarkable feats in England and Italy. They invariably have a small 
 hollow up the axis and a point screwed into the head. They can be 
 fired through great thicknesses of wrought-iron repeatedly, almost 
 without deterioration, even when the heads are undesirably blunt. 
 They cannot be burst by powder. Cammell's have produced good 
 cast-steel projectiles which have held together under impact at very 
 high velocities; a Hadfield cast-steel projectile passed through a steel- 
 faced plate in 1884 without breaking up; 6 but more experience is 
 needed to develop cast-steel shot in this country, especially for the 
 
 1 As in Wilson's and Palliser's experimental projectiles. 
 
 2 Vide pp. 9 and 50, Part I. 
 
 a Vide p. 157, Part I. Many unpublished experiments also bear this out. 
 
 ■» For comparative effects of chilled and steel projectiles against chilled-iron shields, vide pp. 6 
 and 12, Part II. 
 6 Fragments of such projectiles at Meppen in 1879, puzzled some visitors including the author. 
 <> Vide p. 149, Part I. The plate was an injured one. 
 
103 
 
 larger natures of guns. Krupp's steel projectiles have stood up and 
 behaved generally in the same manner as Whitworth's some years ago, 
 and great progress has been made in the development of steel 
 projectiles on the Continent, while England has stood still. In a 
 competitive trial of 6-inch steel projectiles at Spezia in 1886, an 18 or 
 19-inch steel plate was attacked by direct fire. The whole of Krupp's 
 projectiles entered to a depth approaching nine inches and rebounded 
 without serious deformation. Holtzer claimed that the best result 
 was obtained by one of his projectiles. The Italian officers considered 
 Krupp had succeeded best, probably because Holtzer's shells, which 
 varied in quality and measure of success, were considered to be in an 
 experimental stage, while Krupp had achieved a fixed standard of 
 excellence. In the summer of 1886, Firminy and Holtzer's 12-inch 
 steel shells passed without fracture through 16-inch compound plates 
 at Shoeburyness. The Firminy shell struck a weak place in a plate 
 behind the target and remained unbroken. The Holtzer shell struck 
 a sound place and then broke. The best forged steel projectiles 
 appear to be made of steel, sufficiently soft to allow of finishing with 
 a cutting tool in a lathe. They have slight projecting bands, at base 
 and shoulder, to facilitate bringing the shot to exact dimensions at the 
 larger part. The front part of the projectile is then hardened and 
 tempered, and the bands of the shot ground to exact dimension. An 
 excellent lot of Holtzer projectiles were passed into the Service in 
 March, 1887. 
 
 French steel shells generally contain from J to 2 per cent of chromium. 
 Fragments of point and base of Krupp's shell fired against Griison's 
 shield at Spezia in 1886, brought to England and tested, contained 
 respectively - 891 and 0'864 per cent, of carbon but no chromium. 
 
 CHAPTER VII. 
 
 Armoured Stbuctubes. — Ships. 
 
 It would be unwise to attempt to deal with this subject, further 
 than to give a few facts and such general rules as may be useful to 
 those who only are interested in the question, so far as it concerns the 
 action of guns and armour. An armour-ciad ship is a sufficiently 
 important machine, and the number existing in the world is suffi- 
 ciently limited for each ship to have an individual character and 
 reputation, which is known and recorded in all centres of information 
 on naval matters. 1 At the Admiralty, for example, there exists a 
 quantity of information, obtained in many cases confidentially, but 
 which it is to be presumed would be utilized in time of war. It can 
 
 1 For descriptions of ships vide Brassey's " Naval Annual," " King's War Ships and Navies of the 
 "World," and " Kriegsschiff bauten Kronenfels." The author is much indebted to Sir Nathaniel 
 Barnaby, the late Director of Naval Construction, for answering all questions raised by him 
 very fully. 
 
 87 
 
104 
 
 Old- 
 
 fashioned 
 broadside 
 ships. 
 
 Modern 
 masted 
 
 sea-going 
 ships. 
 
 hardly be doubted that in case of war breaking out with any nation, 
 a description of each of the enemy's ships, with a cut showing all ner 
 characteristic features, would be issued to our vessels and coast 
 batteries. Even under these circumstances, a few notes as to the 
 classification of ships might be useful to those who have to engage 
 with ships without having made them their study. For practical 
 purposes, dealing only with their armour, armour-clad ships may be 
 divided into the following classes : — (1), Broadside, old-fashioned, with 
 guns distributed. (2), Masted sea-going modern types, including barbette, 
 central battery, and turreled masted vessels, as well as protected ships} (3), 
 Mast! ess turret. It will be found that vessels will not range themselves 
 exactly under these classes, wide as they are. 
 
 Class (1). Broadside, old-fashioned ships, with guns distributed. In 
 this class the armour is thin, running from about four and a-half to 
 six inches. It extends however in some cases nearly entirely over the 
 side of the ship. Thus, weak as such ships are, they can in hardly 
 any place be attacked by shrapnel fire or that of machine guns, or 
 even quick-firing 6-pounder guns. Their own guns are distributed 
 along the greater part of the side of each ship. The Minotaur, or the 
 Provence or (French) Heroine, fully represent this class. It is probable 
 that common shell from 1 2-inch guns might be used against them, 3 at the 
 same time it is difficult to speak with confidence, common shell having 
 been only tried against unbacked plates, and against the Huascar, 
 when a 9-inch shell got through her 4-inch plates and burst in the 
 backing. Palliser shell filled, acted with terrible effect against the 
 Huascar turrets, whose armour was about the same thickness as most 
 of this class (viz., 4^ inches of iron backed by 10 inches of teak). As 
 the armour of this class of ship is wrought-iron it must be perforated, 
 and comes under the rules for that class of attack, that is, the effect 
 depends on work per inch circumference of shot. The fire of guns 
 able only to partially penetrate the sides is useless, however long con- 
 tinued. Such guns are only available for what has been termed the 
 secondary attack (masts, &c.) Ships of this class however might be 
 perforated by almost any modern rifled gun. Even the 7-inch old 
 type Woolwich gun would perforate most of them directly (the Minotaur 
 up to over 1200 yards range). Being deficient of deck armour they 
 are specially open to attack by plunging fire or vertical fire, but as 
 their guns cannot train through a large angle these vessels would 
 engage nearly broadside on, a position favourable to the attack of 
 their sides rather than their decks. 
 
 Class (2) Masted sea-going modern types. The characteristic of the 
 armoured protection being that it is very thick in important places, 
 while many parts are left without any armour. Generally speaking, 
 the heavy guns are concentrated either in a citadel at the centre of 
 
 i Sir N. Barnaby in his Naval Review of 1886 divides ships into— (1) Armoured with armoured 
 gun positions ; (2) Protected with unarmoured gun positions ; (3) Unprotected. 
 
 2 Vide pp. 51, 52 and 54, Part I. The disc struck out by this class of projectile would be 
 impeded by backing to a great extent, and failing experiments it is difficult to speak with any 
 confidence. 
 
105 
 
 the ship or in barbette towers, or turrets, but in some instances a 
 broadside battery of medium guns exists without any armour. The 
 Admiral Duperre, for example, has 14^5^-inoh guns (vide Fig. 1) 
 without armour. Our own Shannon, Nelson (vide Pig. 2), and North- 
 ampton have their 9-inch broadside batteries unprotected. These latter 
 do not profess to take rank as fully armoured vessels, being termed "pro- 
 
 Fia. 1. 
 
 1 MA A/\ 
 
106 
 
 tected ships" which are constructed to engage "head on" with their most 
 powerful guns, in which position they are protected by armour both m 
 front of the 10-inch guns and across the bows of the vessel in a bulk- 
 head. It does not seem worth while distinguishing this class, however, 
 seeing that both in the French and Italian navies some of the most 
 powerful ships have guns similiarly exposed. It is only possible, then, to 
 direct attention to a few features that may be found, and to depend 
 on individual descriptions of each ship for further information. Speak- 
 ing generally, the French vessels have the peculiarity of mounting 
 their guns very high above the water and firing them en barbette, not 
 on the disappearing system carried out in our Temeraire, but with the 
 men exposed, above the barbette armour. There is a steel plated 
 structure overhead which protects them from the fire of machine guns 
 in the tops of vessels, but the exposure of the men to mitrailleuse and 
 shrapnel fire is a remarkable feature in the French barbette tower ships. 
 
 These ships would seldom suffer from the attack of their decks by 
 guns as they could very seldom be struck at a descending angle ex- 
 ceeding 10°, and British ships, and probably many foreign ones, have 
 deck plates to resist at 10° the same fire as the side armour resists 
 direct. At anchor their decks might be assailed in a dangerous manner 
 by mortar fire, and often great injury might be done in the structure 
 above the armoured deck. Their unarmoured parts might be attacked 
 by common and shrapnel shells. Common shell require a quicker acting 
 fuze than the Pettman G-.S. to burst violently. Shrapnel being only 
 required to break up, may be fired without any fuze, 1 as impact against 
 the thin iron side of the ship breaks them up. It would seldom be 
 possible to strike a ship below her armour belts even in the roughest 
 weather in which she is likely to fight. A ship's conning tower is 
 an obvious object for attack. There is little to be done by cutting 
 down her funnel. 2 
 
 If the primary attack of the side armour is undertaken, it is 
 important to know if it is iron, steel-faced, or steel. The former 
 suffers only by perforation ; and partial penetration, even repeated 
 many times, produces little effect in action beyond starting bolts and 
 producing langridge. Steel-faced and steel plates suffer probably in 
 proportion to the total energy of the shot striking them. Even the best 
 steel filled shell are seldom useful against thick armour ; but this is more 
 fully discussed under the head of Class 3. Fig. 3, p. 107, shows the 
 Chilian vessel the Almirante Cochrane, well known from her share in 
 disabling the Huascar [tee p. 119) ; she is armoured from bows to 
 stern at her water-line, her six heavy guns are in her central battery, 
 
 i Vide?. 48, Part I. 
 
 2 The funnel may appear to be a tempting mark for light guns, but little can be done by striking 
 it above deck, as the following consideration may show : — The boilers are placed low down in the 
 hull, and the funnels must be measured from the fire grates. Suppose the total length of funnel 
 to be 60 ft., and that it is cut down to 30 ft. The diminished power of generating steam will be to 
 the original power in the proportion of V33 to V50, that is 5-5 to 7-1. The ratio of the speed is 
 nearly proportional to the cube root of the generating power, that is, as 3 a/5-5 to 3 \/7-l, or 
 
 about 1-763 to 1-920. In other words, the vessel would only lose 1 knot in 12 by such an injury 
 
 though inconvenience might doubtless be felt from the escape of smoke on deck. 
 
107 
 
 terminating at her main-mast. She is a fairly typical ship of her 
 date (1874), designed by Sir B. Reed. 
 
 Class (3) includes all mastless 1 turret ships and mastless barbette ships Hastes 
 whether sea-going or coast defenders. Generally they arc more heavily s P ' 
 clad, both on their sides and deck, and carry heavier guns, but fewer of 
 them than any other class of vessel. The Italia and Lepanto {vide Fig. 4 
 p. 108, and Fig. 5 p. 109) which have no side plates, and depend entirely 
 
 1 That is carrying only signal masts. 
 
108 
 
109 
 
110 
 
 Figs. 5 and 6. 
 
 \ / 
 
 \ PLAN OF UPPER DECK 
 
 Pig. 
 
Ill 
 
 on barbette armour, and horizontal armour, are remarkable exceptions. 
 The Inflexible {vide Figs. 5 and 6) Duilio {rule Fig. 7, p. 110) Dandolo, 
 Ajax, Agamemnon, Colossus, and Edinburgh, are instances of citadel turret 
 ships, and the Admiral class, {vide Benbow, Fig. 8,) of barbette citadel 
 ships with unarmoured ends. The secondary attack would probably be 
 adopted against such heavily clad vessels. On the other hand, such 
 ships as the Glatton, Dreadnought, Thunderer [vide Fig. 9) and Devasta- 
 tion, are hardly open to any secondary attack. With regard to the 
 
 X) 
 
 ■6 
 
! 12 
 Fig. 9. 
 
 FIC 2 
 
 primary attack (if the side armour, it is very important to know 
 whether the ship is plated with iron, steel, or steel-faced armour. 
 If iron it is useless to attack with projectiles that have not at least 
 1000 feet striking velocity for each calibre in the thickness of the 
 armour. Thus a 12 -inch gun requires at least 1000 feet velocity for 
 each 12 inches, 1500 for 18 inches, and 2000 for 24 inches. As the 
 ship is seldom exposed to a perfectly direct blow, the chance of 
 heavy armour being perforated is small, and it would be as a rule 
 useless to fire steel filled shells, and much more chilled-iron filled 
 shells at thick armour, oblique impact almost always breaking shells 
 up. If the armour is steel-faced or steel, the chance of perforation 
 by a single blow is small in the most heavily plated ships, but 
 continued fire may break up the armour. Guns which are no match 
 tor the armour, and would be useless against it if it were wrought- 
 iron, may assist m breaking up hard armour in long continued fire. 
 Altogether the primary attack of the side armour of heavily plated 
 ships, unless made by guns of power far out-matching the armour when 
 it is steel-faced or steel, can only be expected to be successful if the 
 ship is exposed a long time to it. Wrought-iron is more liable to 
 
 iCeSn y \ b e U l be The the T^Tt '""* <* *™ ^ °»^ T™*"£ 
 it specially well. The nature of the armour might influence the choice 
 
113 
 
 of guns employed in attack. This is most obvious if the forts as well 
 as ships be taken into consideration. At the present time any foreign 
 fleet lying off a foreign armoured fort, would furnish an example of 
 the two kinds of armour, for the older ships would as certainly be 
 clad in iron, as the forts would be covered with Gruson's chilled- 
 iron armour. Wrought-iron must be perforated. Chilled-iron must 
 be fractured. The former is effected by shot in proportion to their 
 powers of perforation, the latter probably in proportion to their 
 energy. What is especially liable to mislead, is the fact that gun's 
 powers are generally estimated according to perforation. Speaking 
 generally, the new type guns would display their powers best against 
 ships, for though total energy and racking power has been increased 
 in these guns, it has not been at all in proportion to the increase in 
 power of perforation, because the small calibre assists perforation, but 
 does not affect the question of total energy and total smashing blow. 
 
 Thus the older-fashioned guns might do better against forts than 
 might be expected, while any smaller new-type guns would only dis- 
 play the full powers, with which they are generally credited, against 
 ships or other structures covered with armour sufficiently soft to admit 
 of perforation. 
 
 It appears to be generally recognised now that if ships are brought 
 to anchor, they are subject to attack of their decks by mortar fire. 1 
 
 It is hardly necessary here to discuss the ships that depend on their 
 small size and powers of attack, and in which armour is dispensed 
 with. They are sometimes protected from machine gun fire by steel 
 shutters, as in the larger cruisers built at Elswick for China. Coal is 
 utilized as far as possible, and the vital parts of the ship are kept 
 below the water line. (For coal, vide p. 115) . 
 
 The Polyphemus can hardly be called an armoured vessel in the usual p iypte: 
 sense of the term, but she is covered with curved steel deck armour. 
 The general figure of her structure in an incomplete state is shown in 
 Figs. 10 and 11, p. 114. The entire " turtle back," to five feet below the 
 water line, is covered with a Whitworth inch steel plate (45 tons 
 tenacity). From 2| feet below the water line the turtle back is 
 further protected by Whitworth steel scales {vide Fig. 12) ten inches 
 square and one inch thick, held by screws, as shown in figure (tenacity 
 65 tons) j besides this plating she is constructed of double ^-inch ship- 
 building steel of about 26 tons tenacity. The figures of the Polyphemus 
 are not correct in all details, and are only intended to show the general 
 form and character of the vessel. 
 
 1 Vide Meppen experiments, 1879, p. 94, Part I. Eifled howitzers have been supplied to 
 our coaling stations. 
 
114 
 
 2 5 
 2 I 
 
 RBiliii 
 
 
 
 Fig. 11. 
 
 VIEW FROM HUD 
 
 Fig. 12. 
 
 ARMOUR FLAT1HG 
 
 KEVS If DIA^ 
 
115 
 
 CHAPTER VIII. 
 Coal Protection. 
 
 In 1880, some experiments 1 were conducted by the Committee on 
 Ordnance, of which the late General Gordon, c.b., was President, with 
 a view to ascertaining the effect of shells on coal, in order to obtain 
 data as to the defence of unarmoured ships, such as merchant vessels, 
 and also as to the substitution of coal resistance for that of side armour 
 in battle ships. 
 
 A coal target 60 ft. thick was used in order to represent the case of 
 a ship filled with coal from side to side. 
 
 The 6-inch R.B.L. gun was employed at 100 yards range at coal 
 without any fortification plates. 
 
 3 Palliser chilled shells filled with sand, fired with 32 lb. charges, 
 penetrated from 16 to 18 feet, 2 broke up on impact. 
 
 1 common shell filled and fuzed, firing charge 34 lb. P., burst after 
 penetrating 13 feet, striking velocity about 1850 feet. 
 
 2 common shell filled and fuzed, firing charge 25 lb. P., and one 
 double shell, similarly fired, burst after penetrating about 9 feet. 
 
 Fortification plates \-iu. thick, were inserted in the coal 3, 6, 9 and 
 12 feet from the front. 
 
 1 Palliser shell (weighted) firing charge 34 lbs. P., penetrated 
 3 plates, entering 12 feet into the coal and 11 inches into the ground, 
 another, similarly fired, penetrated 4 plates and 17 feet of coal. 
 
 1 common shell filled and fuzed, fired with 25 lbs. P., burst at 6 feet 
 penetration, passed through first plate, and blew a hole through 
 second one. 
 
 1 double shell filled and fuzed, charge 25 lbs. P., burst at 8 feet, 
 having passed through first and second plates, and made a hole in third 
 plate. 
 
 1 double shell, similarly fired, burst at 3 feet, blowing hole in front 
 plate, but not reaching second plate. 
 
 With fortification plates \\-in. thick of mild steel, 12 feet from front, no 
 coal behind plates. 
 
 1 Palliser shell weighted, charge 34 lb. P., shell broke up, fragments 
 stopped by plate. 
 
 1 Palliser shell filled with powder, charge 84 lbs. ; shell apparently 
 burst, fragments did not reach back plate. 
 
 Fortification plates, \\-in. thick, mild steel, at 5 feet from front, total 
 depth of coal 15 feet. 
 
 1 Palliser shell (weighted) charge 34 lbs. ; shell broke up. Plate 
 broken, maximum penetration of fragments 10 feet. 
 
 1 Palliser shell (filled with powder) firing charge 34 lbs., plate 
 broken ; shell broke in two but powder charge not fired, total 
 penetration 15 feet. 
 
 1 Not given in D. of A. Proceedings, but in a special report printed March 29, 1881, 
 
116 
 
 The 8-inch B. L. gun was fired at 100 yards with 90 lbs. P. charge. 
 
 Without fortification plates. 1 Palliser shell weighted, penetrated 
 26£ feet, 1 common shell (filled and fuzed) burst after penetrating 
 15 feet. 
 
 With %-in. fortification plates. 1 Palliser shell, weighted, penetrated 
 4 plates and 18 feet of coal. 
 
 With plates l\-in. thick, mild steel, at 12 feet from the front. 
 1 Palliser shell (filled with powder) burst just in front of l|-in. 
 plate, blowing hole through this plate and the f-in. skin behind it. 
 
 With plates \\-i7i. thick, of mild steel, 5 feet from the front. 
 1 Palliser shell (filled) bui'st, plate broken, maximum penetration 
 of fragments 12 feet, skin plate at 15 feet driven back 20 inches. 
 
 It was concluded that a 6-iu. Palliser shell (blind) with 34 lbs. firing 
 charge, which, at 100 yards, had a striking velocity of 1850 feet, 
 penetrated about 18 feet. 
 
 Fortification f-in. plates placed vertically amongst the coal had little 
 effect, their tendency is to deflect the shot. 
 
 A 1^-in. plate of mild steel was equivalent to 3 feet in coal. 
 
 Bursting charges of common and double shell had no incendiary 
 and little or no disruptive effect. Palliser shell filled were less effective 
 than weighted with sand. 
 
 The 6-in. gun at 100 yards should perforate a solid unbacked plate 
 of wrought-iron 10£ inches thick. This may be taken as equivalent 
 to 18 feet of coal. 
 
 The penetrative power of 8-inch Palliser into coal with striking 1930 f .s. 
 was about 26^ feet. 
 
 Four f-in. plates reduced the penetration by about 8 feet. 
 
 If, as in the case of the 6-inch gun, a 1^-in. mild steel plate is 
 equivalent to 3 feet of coal, the penetration with 1-inch plate should 
 be about 23 feet. 
 
 The common shell produced no incendiary effect. 
 
 The 8-in. projectile at 100 yards should perforate 13£ inches of iron 
 this should be equivalent to 26£ feet of coal. 
 
 Conclusions and recommendations. 
 
 20 feet of coal without plates is proof against 6-in. projectiles at 
 short ranges, or 17 to 16 feet if a 1^-in. plate be inserted. 80 and 
 25 feet respectively would resist the 8-in. gun projectiles. With less 
 coal more plates must be used. Steel projectiles would probably act 
 better than Palliser projectiles, a large proportion of which broke up 
 It might be desirable to imitate the conditions of an actual ship and 
 
117 
 
 learn if the projectiles penetrate decks after deflection. Also the dis- 
 ruptive effect of large common shells and of steel shells charged with 
 guncotton should be tried. 
 
 It may be seen from the above that, roughly speaking, 2 feet of coal is 
 equivalent to 1 inch of iron} 
 
 It is to be remembered generally that vessels depending on coal 
 protection would be more vulnerable as their time at sea increases and 
 their snpply runs out. 
 
 OHAPTEE IX. 
 
 Results Produced in Action. 
 
 The first engagement that occurred between armour-clad vessels of 4perienee 
 any class was that of the Merrimac with the Monitor} The Merrimac's 
 armour was inclined at an angle of 30° with the horizontal ; it con- 
 sisted of narrow bars of railway iron in two thicknesses, making a 
 total of 3 inches, laid on 20 inches of oak. The Monitor fired 11 -inch 
 smooth-bores, with 15 lb. charges. Lieut. Very observes — "As far 
 as can be ascertained, no material damage was done, nor should it 
 have been expected." Had the 30 lb. charge been used during the 
 action .... which doubtless would have been the case if the govern- 
 ment practice had not so carefully excluded it, the muzzle-energy of 
 the projectile would have been 2730 foot-tons or more than double 
 what it was, and, as was afterwards proved by experiment, both in 
 the United States and in Europe, every fair blow planted by the 
 Monitor in the action would have smashed a hole. 
 
 On April 7, 1863, the New Ironsides, carrying 4^ in. solid wrought - 
 iron plates, backed by 21 inches of oak, took part in the general 
 action with the Charleston forts. Captain Turner in his report says — 
 " Forcing her way up the Channel, she received the fire of the enemy 
 generally obliquely, excepting when she fell off one way or the other. 
 One of these shots striking the forward facing of a port shutter, 
 carried it away instantly .... the damage done to this ship, 
 with the exception of the loss of a port shutter is not material. The 
 distance at which she received the severest fire of the enemy is about 
 1000 yards." 3 It appears that the bolts were driven out by the impact 
 of shot. The New Ironsides had a somewhat similar experience in an 
 attack on Forts Wagner and Sumter on Aug. 16, and again on 
 
 1 How rough an estimate this must be is apparent from the fact that the resistances of plates are 
 nearly proportional to the squares of their thicknesses, and that this is owing to mechanical 
 reasons which cannot apply to the penetration of coal, the resistance of which must be probably 
 nearly proportional to its thickness. 
 
 2 Development of Armour, Very, p. 387. 
 » Tbii., p. 391. 
 
actions. 
 
 118 
 
 Sept. 10, 1863. Sand bags were found valuable on both occasions to 
 prevent the flying of bolts and langridge. In a period of 6 months 
 this ship was struck 193 times, and yet was never forced to go into 
 port for repairs. 
 
 Numerous examples of the behaviour of ironclads of this date might 
 be quoted from the report of actions in the American War. Speaking 
 generally, it appears that injuries under the hottest fire hardly extended 
 to more than the subsidiary parts of the vessel's armour, such as 
 port-shutters, bolts, &c. Lieut. Very considers that evidence of 
 weakness in laminated armour was ignored or slurred over. Captain 
 (now Admiral) Simpson says — "The solid plates of hammered iron on 
 the New Ironsides, though only A\ inches in thickness, resist the impact 
 of shot much better than the 5 inches of laminated iron on the sides 
 of the monitors." "The laminated iron when disposed in a plane 
 perpendicular to the flight of the projectile, does not seem to answer 
 all demands ; but when disposed in the form of a turret, no objection 
 can be raised to it. The turrets are as near impregnable as any thing 
 can be made. The only objection to them is the "through bolts," 
 which allow the nut to fly when the head of the bolt is struck.-" The 
 most vital and dangerous part of this construction is the roof of the 
 turret, which must be apparent to every one as weak. It can never 
 be struck without causing damage. The roof of the IFeehawJcen was 
 struck at long rauge; the result was the fracturing of the thigh of 
 one man and lighter wounds to two others. 1 Captain Simpson wrote 
 a remarkable report on the evil of turret pilot house and spindle 
 system in October, 1863, strongly recommending the base ring and 
 anti-friction roller system long since adopted in England. 
 
 In February, 1868, during the war between Brazil and Paraguay 
 the Brazilian monitor Alagoas was put to a very severe test ; she carried 
 on her sides \\ inches solid wrought-iron plates backed with 15 inches 
 of teak, and a |-inch skin. Her turret carried 6-inch solid plates 
 backed by about 10 inches teak and two -|-inch skin plates. 2 "In an 
 attack on some Paraguayan batteries she was struck 200 times, her 
 side armour being pierced 12 times and her turret twice. The 
 armament of the batteries consisted of Whitworth 32-pr. rifles, 68-pr. 
 and 120-pr. smooth-bores, and the range for the greater part of the 
 time was less than 100 yards. The turret was very badly damaged ; 
 nearly all the bolts being broken, and the wood backing being 
 badly crushed in several places." 
 
 In October, 1879, took place the fight between the Peruvian vessel, 
 Huascar? and the Chilian ships Blanco, Eucalado and Almiraute Cochrane. 
 The Htiascar was a single turret vessel, carrying 4^-inch iron armour, 
 tapering down to 2 inches at bow and stern, backed by 10 inches of 
 teak, and a f-inch skin. Her turret had b\ inches iron (solid), 
 
 1 Development of Armour, Very, p. 413. 
 
 2 Development of Armour, Very, p. 437. 
 
 s The engagement between the Shah and the Kuascar does not concern the purpose of this 
 work further than to illustrate the bare fact that it is difficult to strike an adversary a full blow 
 under the condition of an actual running fight and to show the effect of a common shell. 
 
119 
 
 13 inches teak, and £-inch skin, Her conning-tower had 3-inch plates, 
 backed by 8 inches teak, with two ^-inch skin plates of iron. The 
 Huascar 1 having higher speed than any Chilian iron-clad vessel, had 
 proved herself such a scourge, that the Chilian supplies had been cut 
 off to a considerable extent. On May 21, 1879, she mercilessly riddled 
 with shot, and rammed the unarmoured Chilian ship Esmeralda, one of 
 whose boilers had been blown up, and who was almost drifting about. 
 The Captain, Captain Pratt, and part of the crew of the Esmeralda 
 desperately attempted boarding on the two occasions when the Huascar 
 rammed her, but the boarding parties were overpowered and slain, 
 and she sank with nearly all her gallant crew, with her colours still 
 flying. 
 
 To capture the Huascar the Chilians had cleaned and repaired the 
 CocJirane, and had improved her rate of speed so that she was at least 
 as fast as the Huascar. Two divisions were formed, one led by the 
 Blanco, and the other by the Cochrane, which were sister ships, carrying 
 9 inches of armour, and 124-ton 9-inch guns. The Blanco and two 
 other ships found the Huascar and another Peruvian vessel, and drove 
 them north into the path of the Cochrane, and two fast unarmoured 
 ships. Captain Grau, commanding the Huascar, tried to pass north- 
 east of the Cochrane, but was cut off and exchanged shots while the 
 Blanco was above four miles astern. It is now difficult to trace the 
 exact course followed by the ships, 2 but it appears that the Huascar so 
 far turned away from the Cochrane as to receive a most destructive fire 
 from her heavy guns firing ahead, while the Huascar 's fire was masked 
 by her poop. A 9-inch Palliser shell from the Cochrane partly destroyed 
 the conning-tower of the Huascar, killing Captain Grau and disabling 
 the steering wheel. A similar shell penetrated the turret, killing most 
 of the detachment, and a third entered the Captain's cabin, and 
 destroyed the steering tackles and some men trying to work them. 
 Now the Blanco came up and endeavoured to ram the Huascar, but 
 passed astern of her, turning the Cochrane off her course, and causing 
 her to make a complete circle to avoid collision. About the middle of 
 her circle, a large shell, believed to have come from the Blanco by 
 mistake, passed completely through the Cochrane abaft her battery, 
 which, though it did not burst, did much damage, killing 2 and 
 wounding 8 men. The Huascar was too far damaged to be able to 
 avail herself of this opportunity to escape. The Chilian ironclads fired 
 incessantly at her, their Nordenfelts cleared her deck, and " silenced 
 the G-atling on her top. Another shell penetrated her turret, and 
 bursting, splashed the left gun, killing the first officer and the men 
 who were working it." The Huascar's steering tackles were destroyed 
 a third time, and she soon after hauled down her flag. 
 
 1 This is chiefly taken from Lieutenant Madan's spirited account in U. S. Institution Proceedings, 
 1881, No. CX1L, p. 695. For Cochrane, vide fig. 3, p. 107. 
 
 2 Vide also King's War Ships and Navies of the World, p. 439, and Development of Armour, 
 by Very, p. 438. King states that the Huascar kept her stern purposely away from the enemy, 
 while Madan blames the Captain for not doing so. King states the Captain of the Kuatcar 
 endeavoured to ram the Cochrane, and also that the remains of the crew repulsed an attempt of 
 the Chilians to board and complete their capture. 
 
 39 
 
120 
 
 Both the Chilian vessels fired Palliser filled shells. Out of 76 rounds 
 from the 0-inch guns, and about 40 from lighter pieces, about 25 
 appear to have been effective hits, at least 16 of them being made with 
 the 9-inch projectiles. The Huascar fired about 40 rounds with very 
 little effect, only 3 striking the Cochrane, and inflicting no serious 
 damage. The Blanco was untouched. Lieutenant Madan, E.N., who 
 obtained his information from Chilian and Peruvian Officers, observes 
 that " the weak armour of the Huascar appears to have been worse 
 than useless, as although it deflected three or four 9-inch shells, and 
 resisted the smaller projectiles, yet it enabled the Chilians to dispense 
 with fuzes and use Palliser shell, which burst, after penetrating, with 
 terrible effect." 
 
 By the rule of thumb, it may be seen that the 9-inch shot, even with 
 
 9 X 1300 
 1300 feet velocity would be able to perforate about 10Q0 = 12-7 
 
 inches of iron unbacked. It is evident therefore that up to the 
 glancing angle these shot would perforate the armour of the Huascar, 
 and the result was what must have been expected in any regular 
 attack as far as the Hriawar's armour was concerned. It might have 
 been expected that at close ranges, the Huascar might have damaged 
 the Cochrane, but not when running from her and striking her obliquely. 
 
 Lieut. ]\ladan enumerates the following injuries to the Huascar: — 
 
 (1.) Palliser shell entered forecastle, wrecked it, and broke stem. 1 
 (2) Carried away capstan, (3) Palliser shell glanced on turret, cutting 
 a groove 1 inch deep. (4) Palliser shell entered turret compartment 
 below upper deck on port side, hur^t over magazine hatch, choked 
 turret machinery with th-hris, but it did no material damage, as it 
 revolved easily afterwards. (5) Palliser shell struck edge of upper 
 deck and burst, splashing turret with fragments. (6) Palliser shell 
 entered turret to right of right gun, and struck right carriage bracket, 
 and burst, killing, it is said, all men in turret. The gun might still have 
 been fired perhaps with one capsquare, but it ceased firing. (7) Palliser 
 shell entered turret to right of right gun near top, burst and struck 
 left gun, injuring but not disabling it, and killing or wounding all the 
 crew it is said. (8 and 9) Struck and destroyed conning-tower (3 ins. 
 iron and 10 ins. backing) No. 8 disabled fighting wheel below. 
 (10) Palliser shell glanced on funnel. (11) Palliser shell entered 
 engine room, bursting, and killing 4 men on upper platform, but did 
 not damage engines or men in charge. (12) Palliser shell struck a 
 12-pr. upper deck gun, knocking off muzzle. (13) Palliser shell burst 
 inside officer's cabin. (14) Palliser shell entered poop. (15) Palliser 
 shell burst in stern, cutting rudder chains, and killing men. (16) 
 Palliser shell burst in stern, injuring stern post and killing all men on 
 relieving tackles. There were many minor injuries. 
 
 1 These numbers aro taken from bows to stem, not in their order of time. Lieutenant Madan 
 obtained his facts from Chilian and Peruvian Officers, especially from the senior executive Officer 
 of the Cochrane. 
 
in 
 
 121 
 
 Naval Attack of Alexandria Forts. 1 
 
 The attack of the forts of Alexandria in 18S2 should be studied . 
 works specially devoted to it, such as the English aud American official 
 reports. ^ Major Watford, R.A., road a very full paper on it at the 
 United Service Institution. 3 Before discussing it briefly here, it is 
 well to make one or two remarks as to its general character. First, 
 it must be clearly understood that the case is not considered as an 
 instance of a fight between ships and really powerful sea forts. A 
 thoroughly well defended harbour would have mines laid down, 
 rendering the movement of ships dangerous. The water would, 
 perhaps, be divided up into imigiuary squares, aud mechanism might 
 be provided for automatically directing guns on each spot. Possibly 
 ships engaging at anchor might be assailed by rifled mortar or howitzer 
 fire. Torpedo boats might also be encountered. In some cases there 
 might be batteries on positions with high command. The existence of 
 such elements would make an attack by sea a very serious matter, 
 undoubtedly such an operation as would need special reasons to justify 
 it. 
 
 In the case of Alexandria these dangers were not present, but com- 
 plications and difficulties were not wanting ; for example, circumstances 
 demanded that the ships should be sparing of ammunition. The action 
 was fought under conditions rendering it very difficult to distinguish 
 the features of the forts, and to direct the fire to good purpose. 
 
 Moreover the vessels, with the exception of the Inflexible, were by 
 no means the most heavily clad or powerfully armed possessed by 
 England. The Thunderer class was wholly unrepresented. 
 
 The attack should be regarded, then, not as generally representative 
 of what might occur, but as an important naval operation of a special 
 kind whose complete success justified the judgment and skill with 
 which it was conducted. 
 
 Certain British ships, the Invincible, Jloi/arc//, and Penelope, had been 
 lying in the inner harbour of Alexandria up to July 10th, and they 
 then moved into position in the outer harbour (ride plan, PI. XIX.). 
 To this may be perhaps attributed the fact that the Egyptians laid 
 down no submarine mines, for it appears that they had an abundant 
 supply of them, although only a limited quantity of cable was found. 
 The general character of the attack may be seen from Lord Alcester's 
 orders, which were as follows : — 
 
 "There will be two attacks. (1) From the inside of the harbour, 
 in which the Invincible, Jlmuirch, and Penelope, will take part. (2) By 
 the Saltan, Superb, Temeraire, Alexandra, and Inflexible, from outside the 
 breakwater." " On the batteries opening on the off-shore 
 
 squadron in reply, every effort will be made by ships to destroy the 
 batteries on the Ras-el-Tin Peninsula, especially the light-house 
 battery bearing on the harbour. When this is accomplished, the 
 Si'ltan, Superb, aud Alexandra, will move out to the eastward, and 
 attack Fort Pharos, aud, if possible, the Silsileh battery. The 
 
 1 Extracted from Lord Brussels Naval Annual, 1SS6. 
 
 •' Vide U. S. Institution Proceedings, 18S3, Vol. XXVII., No. CXIX. 
 
122 
 
 Inflexible will move down this afternoon to the position off the Corvette 
 pass assigned to her yesterday, and be prepared to open fire on the 
 guns in the Mex lines in support of the in-shore squadron when signal 
 is made. The Temeraire, Sultan, and Alexandra will flank the works on 
 Eas-el-Tin. The gun-vessels and gunboats will remain outside, and 
 keep out of fire until a favourable opportunity offers of moving in to 
 the attack of Mex." 
 
 The strength of the ships and forts that were to be thus engaged 
 may be seen generally from the following Tables : — 
 
 Ship's Name. 
 
 Displace- 
 ment in 
 Tons. 
 
 Guns. 
 
 si 
 i ■§ 
 
 <1 - 
 
 Draught 
 of water 
 
 a . 
 
 -■? 
 
 6 £ 
 
 CO r-l 
 
 a . 
 
 s. a 
 
 u 
 
 id i-t 
 !M i-H 
 
 O0 O 
 
 p 
 o . 
 
 t-H OS 
 
 f.g 
 
 OS 00 
 
 a 
 
 o . 
 
 3d t^ 
 
 •* 
 
 CO 
 
 
 Inflexible (Turret) 
 Temeraire (Barbette)... 
 
 11,400 
 1 [8,320 ; 
 8,640 
 9,490 
 9,290 
 6,010 
 9,100 
 4,470 
 
 4 
 4 
 
 4 
 
 4 
 
 4 
 2 
 
 6 
 
 4 
 
 10 
 
 8 
 
 16 
 
 38 
 
 2 
 
 4 
 10 
 
 16 
 
 8 
 8 
 
 1 
 1 
 
 4 
 4 
 
 3 
 3 
 
 inches. 
 24 to 16 
 
 10 to 8 
 
 11 to 8 
 
 12 to 8 
 9 to 6 
 8 to 6 
 
 12 
 6 to 6 
 
 feet. 
 
 26 
 
 27i 
 26i 
 27i 
 221 
 261 
 17* 
 
 
 
 
 Total Guns 
 
 The batteries of Alexandria had the following armaments :- 
 
 EIFLED GUNS. 
 
 Forts and No. 
 on Plan. 
 
 Mounted. 
 
 Unmounted. 
 
 Total. 
 
 10-in. 
 M.L. 
 
 9-in. 
 M.L. 
 
 8-in. 
 M.L. 
 
 7-in. 
 M.L. 
 
 40-pr. 
 
 B.i. 
 
 10-in. 
 M.L. 
 
 9-in. 
 M.L. 
 
 Silsileh (1) 
 
 1 
 1 
 1 
 1 
 
 1 
 
 1 
 3 
 3 
 3 
 
 4 
 
 1 
 
 1 
 
 2 
 1 
 2 
 1 
 2 
 3 
 
 2 
 
 2 
 1 
 
 2 
 
 2 
 
 4 
 
 4 
 8 
 5 
 9 
 6 
 2 
 11 
 
 Pharos (2) 
 
 Adda (3) 
 
 Eas-el-Tin Lines ") 
 Has-el-TinFort...) 
 Oom-el-Kubeba (7) ... 
 Mex Fort (10) 
 
 Total 
 
 6 
 v.. 
 
 15 
 
 16 
 
 2 
 
 3 
 
 2 
 
 6 
 
 45 
 
 Mounted Bi 
 
 fled Guns in all ... 37 Unmounted Guns ... 8 
 
 
P1.XIX. 
 
123 
 
 SMOOTH-BORE GUNS AND MORTARS. 
 
 Port or Battery 
 and No. on Plan. 
 
 Guns. 
 
 Mortars. 
 
 i— « « 
 
 Ci <D 
 
 ■So. 
 
 
 
 
 
 
 
 
 
 15-in. 
 
 10-in. 
 
 6-6-in. 
 
 20-in. 
 
 13-in. 
 
 12-in. 
 
 11-in. 
 
 
 Fort Silsileh (1) 
 
 
 3 
 
 
 
 1 
 
 
 
 
 „ Pharos (2) 
 
 
 6 
 
 31 
 
 
 4 
 
 
 
 41 
 
 ii Adda (3) 
 
 
 14 
 
 
 
 5 
 
 
 
 19 
 
 Ras-el-Tin Lines ") 
 
 [(4).. 
 Fort Ras-el-Tin J 
 
 4 
 
 15 
 
 11 
 
 1 
 
 6 
 
 1 
 
 2 
 
 40 
 
 2 
 
 6 
 
 21 
 
 — 
 
 1 
 
 — 
 
 2 
 
 31 
 
 i, Saleh Aga (5) 
 
 — 
 
 4 
 
 8 
 
 — 
 
 — 
 
 — 
 
 — 
 
 12 
 
 Battery (6) 
 
 _.. 
 
 2 
 
 2 
 
 
 
 
 
 
 
 
 
 4 
 
 Fort Oom-le-Kubeba(7} 
 
 — 
 
 6 
 
 10 
 
 — 
 
 1 
 
 1 
 
 — 
 
 18 
 
 ii Kara aria (8) 
 
 — 
 
 2 
 
 3 
 
 — 
 
 1 
 
 — 
 
 - 
 
 6 
 
 
 4 
 
 11 
 
 9 
 
 — 
 
 — 
 
 — 
 
 — 
 
 24 
 
 „ Fort (10) 
 
 
 4 
 
 5 
 
 \ 
 
 3 
 
 2 
 
 
 14 
 
 Mounted S. B. Guns 
 
 in all .. 
 
 . 182 
 
 Mounted Mortars ... 31 
 
 
 213 
 
 It may be seen from the above Tables that while the guns of the 
 fleet were much more powerful than those of the forts, some of the 
 latter might, under reasonably favourable circumstances, perforate the 
 armour of all the ships except the Inflexible. 
 
 The guns and projectiles were supplied from Elswick, and correspond 
 to the service old type Woolwich M. L. armaments carried by the 
 British vessels. 
 
 On the morning of July 11th, 1882, at about 7 a.m., the British 
 fleet opened fire in the order detailed by Lord Alcester. The position 
 of the ships may be seen approximately in the plan. 
 
 The Invincible (flagship) and Penelope engaged at anchor in the outer 
 harbour against the earthworks at Fort Mex at 1500 yards range. 
 The Monarch acted with them under steam against Ports Mex and 
 Oom-el-Kubeba, as well as eleven gunboats. The Temeraire supported 
 the attack from the central pass into the harbour at about 4000 yards 
 range, and the Inflexible in the Corvette pass divided her fire between 
 earth and masonry forts, Oom-el-Kubeba at 4000 yards, and Ras-el- 
 Tin at 2700 yards. The Alexandra, Superb, and Sultan, engaged all 
 forts between Ras-el-Tin and Pharos, chiefly, at first under steam, at 
 from 1500 to 2000 yards, and at length anchored at about 2200 yards 
 distance. The gunboats, Helicon, Condor, Beacon, Bittern, Cygnet, Decoy, 
 assisted as opportunity offered. 
 
 At about 10.30 a.m., Ras-el-Tin and Mex Forts were to a great 
 extent silenced, and the Alexandra, Sultan and Superb attacked the 
 centre battery and other forts farther to the east, the Inflexible and 
 
124 
 
 Temeraire coming round and joining in the work. The Superb attacked 
 Adda; the Temeraire, Pharos ; and the Inflexible, both these forts. A 
 shell from the Superb blew up the magazine in Adda, and silenced it 
 about 2 p.m. The Monarch had silenced Oom-el-Kubeba about 1 p.m. 
 The Temeraire, Inflexible, and Superb, silenced Fort Pharos about 2.30 
 p.m., and the centre battery ceased to reply to the Sultan and Alexandra 
 about the same time. A battery at Hospital or Bluff Point was the 
 last one silenced at about 3 p.m. At about 1 o'clock a party in a boat 
 very gallantly landed and destroyed the guns in Port Mex. 
 
 Effect of Are The effect of the ships' fire on the various works was as follows : — 
 
 on forts. # 
 
 Port Silsileh (not much engaged), a regular fort, chiefly earthwork, 
 suffered little damage. 
 
 Port Pharos, masonry with casemates, suffered severely, casemates 
 riddled and choked. Three guns and carriages wrecked and others 
 choked. One gun on the reverse face carried away and thrown on 
 to its muzzle in the ditch 30 ft. distant. The rifled guns were silenced 
 before the smooth-bores in the casemates, although the loss of life in 
 the latter must have been out of all proportion to that in the former. 
 
 Port Adda, earth revetted with masonry. Parapets cut about but 
 not materially injured. Three guns struck and put hors de combat. 
 The fort was wrecked by the explosion of the magazine, otherwise it 
 might have continued firing. 
 
 Ras-el-Tin Lines (from Hospital Battery to Ras-el-Tin Port). 
 Masonry and earth. No guns actually disabled although struck ; 
 several put out of action by being dismounted ; or by the carriages 
 being broken, or choked with debris. 
 
 Ras-el-Tin Port (masonry and earth, but with thin parapets). All 
 rifled guns fairly put out of action, either by fire of the fleet or by 
 accident. This fort fought better and suffered more than any other. 
 The parapets were much cut about and buildings riddled. 
 
 Fort Saleh Aga an insignificant work, but with a command of 
 60 feet. One S. B. gun dismounted. 
 
 Fort Oom-el-Kubeba 80 ft. command, masonry and earth, with 
 good profile. Parapet suffered much from Inflexible 80-ton gun shells. 
 One gun destroyed. 
 
 Fort Kamaria untouched, probably did not engage in action. 
 
 Mex Sea Lines, about 25 feet highest command ; guns firing en 
 barbette over earth parapets, 15 to 18 feet thick. One or more guns 
 dismounted by fire of ships. 
 
 Mex Fort, 22 feet command ; guns en barbette ; earth parapet 
 unrevetted ; parapets hardly injured, but buildings almost swept away. 
 Three guns struck by shells and put out of action, and others by 
 machine gun or shrapnel bullets. Two 10-inch guns received great 
 protection from a wall some distance in front, which caught shells 
 aimed at the guns ; both the latter were struck, but not materially 
 injured. 
 
125 
 
 Fort Marsa-el-Kanat had no defensive power. Fort Marabout was 
 attacked by gunboats. No guns were dismounted. 
 
 The ships received the following injuries :— The Inflexible, penetrated ^ship's™" 5 
 in unarmoured parts, and a good deal knocked about aloft. Monarch 
 and Temeraire untouched. Alexandra, struck sixty times, but not 
 seriously injured. Sultan, Invincible, and Superb, armour indented, 
 and minor injuries sustained. 
 
 The Penelope hulled eight times, but no serious injuries caused. A 
 small slice cut off one gun near the muzzle. 
 
 The American Official Report, written by Lieut.-Commander Goodrich, conclusion, 
 contains among other conclusions — 
 
 (1 ) Command is important for forts. (2) 30 feet of earth stops all A ™p'ort. n 
 projectiles. (3) Embrasures should be cut deep below crest. (4) No 
 guns should be mounted en barbette. (5) Guns should be painted the 
 colour of the works. (6) The value of armour has been under-rated. 
 (7) Modern guns with flat trajectories are not the best for attacking 
 earthworks. (8) Some ships should carry howitzers. (9) Vertical 
 fire should be studied. (10) Disappearing guns firing en barbette are 
 very efficient. (11) Ships gain more in accuracy by anchoring than 
 they lose in increased exposure to injury. (12) The swell told against 
 the shootings especially of ships engaging broadside on. (13) Shot 
 not aimed at guns or magazines are thrown away. (14) Vessels do 
 not fight on even terms with forts. (15) Forts could not stop ships. 
 (16) The forts might have been repaired and action continued. 
 
 Major Walford, R.A., forms conclusions agreeing with some of the mj. w&ifora 
 above, and points out the advantage of counter-guards in advance of 
 batteries, also of traverses, parados, and mantlets, and the evil of 
 stone and brick in batteries. He recommends guns being strongly 
 pivoted, well forward. 
 
 The additional suggestion may perhaps be made, that forts might 
 mask or reserve their most powerful guns until they had drawn a large 
 quantity of the ships' ammunition. It would be difficult for a fleet to 
 detect such a practice, and the masked guns might obtain a better 
 opportunity of damaging ships, if the latter became emboldened to 
 draw in to closer range. 
 
 Special causes, as noticed before, made accurate firing from the Admiral Le 
 ships almost impossible. The conditions of sunlight and of the works ^™£, 
 made it very difficult to distinguish the guns and objects aimed at. 
 Admiral Le Hunte Warde, who commanded the Superb, states that 
 while the vessels were in motion, it was necessary, as fast as the smoke 
 cleared, to identify objects aimed at from a new point of view, and 
 then to" fire at them at an altered range. So difficult was this, 
 that to deal effectually with the forts it was judged necessary to 
 anchor. 
 
127 
 
 Explanation of Diagrams Showing Pebpoeation op Aemoub. 
 
 The accompanying diagrams PI. XVI., XVII., XVIII. and XIX., 
 appeared in Lord Brassey's Naval Annual, 1886, and with his per- 
 mission are here given. They are based on a series that owe their 
 origin to the official one prepared by Colonel Noble in 1873. 1 Foreign 
 ships were given by Captain Grenfell, R.N., in Engineering, and have 
 been shown for some years past in the Austrian Marine Almanack as 
 well as in the British Navy. To give to these diagrams a definite 
 practical character, the following course has been followed : — 
 _ On the supposition that 3000 yards is the maximum range at which 
 side armour will be systematically attacked, all projectiles perforating 
 up to 3000 yards are shown with a clear blank hole. For every 
 perforation effected only at shorter ranges the limit of range is 
 entered in or near the hole. Thus, any officer finding armour 
 corresponding to that of his enemy with a blank hole opposite his 
 gun on the diagram, may hope to perforate with a direct hit at least up 
 to 3000 yards. Should he find a number less than 3000 yards entered, 
 he must not expect to perforate beyond the range so entered. 
 
 The armour is wrought-iron. It is only approximately representative. 
 The types are chiefly French. The perforation includes backing and 
 skin. 
 
 Colonel Maitland's diagram has been used for calculating loss of 
 velocity at each range. On service each hit would have less power 
 to perforate than is indicated on the diagrams in proportion as it strikes 
 more obliquely. 
 
 The diagrams must be regarded as only approximate. They may 
 however be of practical value, for an officer may be able to see at a 
 glance that some ships armour is well within the power of his guns or 
 quite beyond them, and behave accordingly. 
 
 On this, vide Kemarks on Armoured Structures — Ships, pp. 103-113, 
 Part II. 
 
 All armour clad vessels built prior to 1876, and many subsequently, 
 are iron plated. To these the above directions apply. For example, 
 suppose 10-in. M.L. guns to attack the French vessel Redoubtable, 
 carrying 14-in. plates along her water-line, and 9^-in. plates on her 
 central battery, the above rule would indicate that it would be useless 
 to attack her water-line with less than 1 400 ft. striking velocity, and 
 her battery with less than 950 ft. Now the 10-in. 18-ton M.L. gun has 
 has a muzzle velocity of a little over 1364 ft., and even if by a special 
 
 1 The diagram prepared by the Author in 1872 was in a different form. 
 
 40 
 
128 
 
 charge it were brought up to 1500 ft., it would be right for such guns, 
 even at the shortest ranges, to attack the central battery and not the 
 water-line, for something must be allowed for backing, probable 
 obliquity of impact, and range. The Redoubtable, like some other 
 French ships, has an unarmoured battery which might be attacked by 
 any ordnance. 
 
 The question may naturally be asked why ships clad in steel and 
 compound armour are treated as if only sheathed in the same thickness 
 of wrought-iron. The reply is that the object is to furnish a low limit 
 for guns available for attack. Steel varies much in quality. Some 
 ships are known to carry very soft plates approaching wrought-iron 
 more or less in quality. Hard steel could not be perforated at all but 
 might gradually yield by fracture. In either case it seems unwise to 
 lay down an arbitrary limit, and the only definite limits that present 
 themselves are the high and low standards of attack adopted in experi- 
 ments both for steel and steel-faced armour [see p. 37). The low one 
 appearing to represent the low limit of useful attack here sought. 
 
 Thus it should be borne in mind, that while steel and steel-faced 
 armour might be attacked about the limit shown in these diagrams for 
 iron of the same thickness, they would not be perforated at all events 
 at the same limit, but would probably yield more gradually and more 
 or less by fracture. 
 
 Oblique Perforation. 
 
 It may be readily seen by drawing a figure, that if a projectile 
 perforates a plate at any angle 6 with its face, its path through the 
 plate multiplied by sin is equal to t, the thickness of the plate. Thus 
 a shot which would perforate a 10-inch plate directly {i.e. at 90°) 
 would, striking at 60° with the face, only perforate a plate 10 x sin 60° 
 =8'66 inches thick. 
 
 Practically this calculation is seldom useful because the angle of 
 incidence is seldorn known, so that the maximum effect possible 
 {i.e. direct perforation) is more important. In a siege, however 
 oblique perforation might be required. 
 
PI. XV 
 
 SS < i 
 "in * « 3 
 
 ^ te^^ tegga fa^fcLU J Euj^llluj ; 
 
 3 
 
 o 
 
 CD 
 
 X 
 
 OD 
 
 
O 
 
 -J 
 
 s 
 
 X 
 
 I/) 
 
 CD 
 
PI. XVII. 
 
 A:li ii I 
 
 ^^ ^ ^^ ^>j^ ^^^ ^t-w^ 
 
 SBg^feSESS^ pJ-U-UJJJ^ jg 
 
 pmof ^^ ^^ ftg^P^ ^WW 
 
 z 
 
 D 
 
 o 
 
 z 
 o 
 
 z 
 
 <r 
 
Ill 
 
 z 
 
 3 
 O 
 
 z 
 < 
 s 
 
 IE 
 
 in 
 O 
 
INDEX 
 
 Part I. Part II. 
 
 Adda Port _ 122-124 
 
 Admiral Duperre" 105 
 
 ii JBaudin Plate — 98 
 
 ,i Targets 157 — 
 
 Admiralty decision as to hard plates — 92,93 
 
 Experiments 40,41,42,47,51-147,167 — 
 
 Agamemnon ... ... ... ... ... ... ... — 111 
 
 Air Space Targets 9,42,60,61,78 — 
 
 Ajaoc — 111 
 
 Alcester, Lord, Orders to Pleet — 121 
 
 Alderson, General, Steel Shells — 101,102 
 
 Alexandra, in Attack of Alexandria ... — 121 to 126 
 
 Alexandria Ports Attacked — 121 to 125 
 
 n Naval Attack — American Report — 121 to 126 
 
 Alpine Chilled Iron Shield — 6 
 
 Amager Armour Competition 147 90,91 
 
 American Armour ... ... ... ... ... ... 1, 6 — 
 
 American Experiments 1,2,6 — 
 
 Angular Impact on Armour 46,79,103 128 
 
 Armour, Early History 1 — 
 
 ii Hard ... ... ... ... ... ■■. 35 — 
 
 ,, Soft and Hard 21 — 
 
 See also Chilled Iron, Compound Iron, Iron, Griison 
 Armour, Manufacture, Plate-upon-Plate, Steel 
 Armour, Targets, &c. 
 
 Armoured Structures — Forts — 79 
 
 „ „ Ships — 103 
 
 Armstrong 6 and 8-inch New Type Gun Trials 83 — 
 
 „ 100-ton B. L. Gun at Spezia 161 17 to 33 
 
 „ 100-ton M. L. Gun at Spezia 61,63,91,117 — 
 
 „ Bidsdale Experiments 157 — 
 
 Asphalte between Plates 9 — 
 
 Atalanta, &c. ... ... ■•■ •■• ••• ■•■ ° 
 
 Austrian Shields and Trials — 5 
 
 ,„, • ™ . x f 20, 38, 46 
 
 Avant Cuirasse (Glacis .flate; — £ 59,70 
 
 ( 3, 7, 9, 39, 42, 47, 50, 54, 63, 87, 91, > 
 Backing | 97] 10 2, 119, 120, 133, 139, 147, 157 > 
 
 i, General behaviour of 3 i 63 85,86 
 
 Barbette Ships - 106 to 111 
 
 Barlow, Colonel— Note — 101 
 
 Barnaby, Sir Nathaniel — 103,104 
 
130 
 
 Behaviour of Armour of various kinds ... 
 
 JBellerophon Target ... 
 
 Bessemer Steel in Dover Turret ... 
 
 Bixby, Captain — Report on Chilled Iron 
 
 Black Prince 
 
 Blanco — Action with Suascar 
 
 Bolts 
 
 n Behaviour of, in Action 
 a Temporary 
 
 u Testing of 
 
 Bosses, Condemned 
 
 Brassey, Lord — Naval Annual 
 
 Brialmont, General 
 
 Bricks between Plates 
 Broadside Ships 
 
 Brown (Ellis) Steel-faced Plates 
 
 ,, „ Plates, Manufacture of 
 
 Bucharest Turret Competition 
 „ n Conclusions 
 
 Buckau Chilled Armour Experiments 
 
 Part I. 
 
 Part II. 
 
 — 
 
 84 
 
 7 
 
 — 
 
 
 
 83 
 
 — 
 
 9 
 
 2 
 
 — 
 
 — 
 
 118 
 
 6, 6, 68, 118 
 
 80, 83, 86 
 
 7 
 
 94,98 
 
 — 
 
 83 
 
 — 
 
 86 
 
 3 
 
 — 
 
 — 
 
 103, 119 
 
 — 
 
 9, 35, 36 
 
 9 
 
 — 
 
 — 
 
 104 
 
 108, 111, 117, 147, 151 
 
 88 
 
 — 
 
 88 
 
 — 
 
 35 
 
 — 
 
 62,63 
 
 — 
 
 9,13 
 
 Cammell's Iron Plates at Spezia ... ... 
 
 Cammell (Wilson) Compound Shot 
 
 ... , c , , . (67, 68,79, 99, 100, 
 Steel-faced plates [ > ^/^ ^ 
 
 „ „ Plates, Manufacture of 
 
 Caps, Wrought-iron, on Shot points 
 Casemate Shields 
 
 Case Shot fired at Unarmoured Ships 
 Cement ... ... ... ... ... ,„ 
 
 Cha biter's Target 
 
 Chilian Actions 
 
 Chilled Iron Armour, Gruson's 
 
 First Trials of 
 
 at Meppen, 1879 
 
 at Spezia, 1876 
 
 Gruson's, Buckau 
 
 rr Spezia, 1886 
 
 n " " Projectiles ... ... ... ,.,3 
 
 Chinese Cruiser 
 
 Clarke, Captain, Bucharest Report 
 
 Coal Protection 
 
 Coast Defenders 
 
 Cochrane ... ... ... 
 
 Coles Co wper, Captain 
 Colossus 
 
 61 
 
 — 
 
 74, 79 
 
 102 
 
 106, 108, ■) 
 147-151 > 
 
 36 to 72 
 
 84, 88, 90 
 
 — 
 
 88 
 
 82 
 
 — 
 
 7, 53, 139 
 
 79,80 
 
 48 
 
 — 
 
 9 
 
 — 
 
 7 
 
 — 
 
 — 
 
 118 to 120 
 
 — 
 
 1,85 
 
 64,96 
 
 1,4 
 
 96 
 
 — 
 
 64, 67 
 
 — 
 
 — 
 
 9, 13 
 
 — 
 
 17 
 
 (4, 6,12,90, 
 42 to 161 i 93-98, 104, 
 
 — 113 
 
 — 35 
 
 — 115 to 117 
 
 — 107 
 1 106, 
 
 > 107, 118 
 
 3 — 
 
 111 
 
131 
 
 Part I. 
 3 
 
 73 
 80 
 81 
 61,62 
 40,147 
 47 
 
 Committee on Iron 
 
 » Sub- on Plates and Projectiles 
 
 " it Conclusions 
 
 " » Recommendations 
 
 Common Shell against Armour 
 
 » » ii Decks 
 
 11 » ii iron unarmoured ships 
 
 Competitive Plate Trials— 
 Amager ... 
 
 Meppen 
 
 Nettle 
 
 Ochta 
 
 Pola 
 
 Spezia 
 
 ii Projectiles 
 
 ii Turret Trial — Bucharest 
 
 Compound Armour — Brown (Ellis) 
 
 „ Cammell (Wilson) {«&«%■$&?&} 
 
 » a Powers employing it ... ... 
 
 « Steel Plates 
 
 Conclusions on Bucharest Trials 
 
 ii on Projectile Competition 
 
 Concrete 
 
 Conning-Tower — Kuasear 
 
 Constants in formulae 26 to 32 
 
 Copenhagen (Amager) Experiments 22, 147 
 
 Corrugations and Bosses 3 
 
 Cowper Coles 3 
 
 147 
 96 
 68 
 
 133 
 
 61, 63, 91, 117, 161 
 73, 87, 105, 137 
 
 108, 110, 117, 147, 151 
 
 80,81 
 9, 146, 146 i 
 
 Part II. 
 
 101, 118 
 
 100 
 
 103 
 
 36 to 72 
 
 84,88 
 
 84,88 
 
 101 
 
 90 
 
 62, 63, 
 
 37,61 
 80 to 82 
 119 
 
 Cupolas and Turrets.. 
 
 3, 42, 147- 
 
 3, 4, 8, 9, 
 
 13, 17 to 
 
 34, 35 to 
 
 72, 105 
 
 to 113 
 
 Daniolo 
 
 Danish Trials 
 
 De Bange Gun 
 
 Deck Armour Targets 
 
 ii Inflexible Target 
 
 Diagrams for Perforation 
 
 Diary of Bucharest Series of Trials 
 
 Dillingen Plates 
 
 Disappearing Turret 
 
 Disc punched out by co mm on shell 
 
 Disintegration of chilled shot 
 
 Dover Turret 
 
 Dreadnought... 
 
 Drop or Machine illustrating perforation 
 
 Duilio 
 
 ■ Target 
 
 Dutch Experiments 
 
 22, 147 — 
 
 40, 41, 147 — 
 94 — 
 
 52, 64 
 9, 42, 50, 61, 78 
 
 110, 111 
 36 
 
 36 
 
 97 — 
 73 
 
 102 
 
 82,83 
 
 111 
 
 36 — 
 
 110, 111 
 
 87 — 
 
132 
 
 " 3£" total striking energy 
 
 " e " energy per inch circumference 
 
 Eastbourne Turret, 1866 
 
 Eastney Deck Experiments 
 
 Ekman (Finspong) ... 
 
 Elswick Cruisers 
 
 n Experiments 
 
 n Guns in Prussia ... 
 
 it Guns at Spezia 
 
 i, New Type Gun Trials 
 
 ,, Turret Machinery 
 
 Ellis — Armour 
 
 Energy per Ton of Gun 
 n Striking 
 
 Engineer Experiments, 1883 
 
 English, Major 
 
 English-Palliser Bolt 
 
 Ericsson 
 
 Esmeralda sunk 
 
 Examples worked out 
 
 Part I. 
 
 Part II 
 
 15, 17 
 
 — 
 
 17 
 
 — 
 
 — 
 
 83 
 
 147 
 
 .Lit. 
 
 ... 42, 73, 74, 87, 88 
 
 — 
 
 — 
 
 113 
 
 157 
 
 — 
 
 39 
 
 — 
 
 ... 61, 63, 91, 117, 151 
 
 17 to 33 
 
 83 
 
 — 
 
 — 
 
 34 
 
 108, 111, 117, 147, 151 
 
 84,88 
 
 38 
 
 — 
 
 15, 17 
 
 
 7, 48 to 61, 139 
 
 — 
 
 ... 6,12,29,82,167 
 
 — 
 
 18, 19, 20, 25, 27, 28, 35 
 
 119 
 
 Excellent — Report on Common Shell against Decks 
 
 147 — 
 
 Fairbairn, Sir W., Formula and Investigation 
 
 4,26 
 
 — 
 
 Filled Shells exploded by impact ... 
 
 39 
 
 — 
 
 Finspong Projeotiles ... ... ... ... 42, 73 
 
 , 74, 87, 88 
 
 — 
 
 Firth Projectiles, Steel 
 
 74,78 
 
 — 
 
 Flat-headed Projectiles 41, 
 
 46, 79, 151 
 
 — 
 
 o ii fired at chilled-iron 
 
 — 
 
 16 
 
 n ii n water 
 
 41 
 
 — 
 
 Floating Batteries, 1853, 1854 
 
 2 
 
 — 
 
 Ford, Maj.-Gen 
 
 1 
 
 — 
 
 Forde, Colonel 
 
 23 
 
 — 
 
 Foreignexperiments { 1Q « JMM7, 
 
 , 91, 93, > 
 151, 156 > 
 
 4, 9, 13, 
 
 17 
 
 Forest, Mr., Note 
 
 52 
 
 — 
 
 Formulae employed by Director of Artillery's Dept. ... 
 
 16, 17 
 
 — 
 
 Formula, English's 
 
 29 
 
 — 
 
 n Fairbairn's 
 
 4, 18 
 
 — 
 
 n Gravre 
 
 32 
 
 — 
 
 ii Inglis' 
 
 24 
 
 — 
 
 ii Introduction of 
 
 10 
 
 
 
 h Italian 
 
 32 
 
 — 
 
 i, Krupp's 
 
 32 
 
 _ 
 
 ii Maitland's 
 
 28 
 
 
 
 n Noble's 
 
 29 
 
 
 
 ii Worked out 
 
 35 
 
 
 
 Forts, Alexandria 
 
 
 122 to 125 
 
 ii Armoured 
 
 
 
 5, 17, 36 
 
 ii Chilled-iron 
 
 
 
 6 
 
 ii English Coast 
 
 — 
 
 79 to 84 
 
133 
 
 Part I. Part II. 
 Fracture of Hard Armour { "^^^m^JuHW 
 
 (12, 44, 16, 47, 49, 62, 61, 70, 74, 80, •) „ ln ,. 
 Projectiles 1 83, 89, 92, 100, 103, 106, 112, 121, 124, C b ^ 1° ' A ' 01 
 (.127,128,132,139,143,153,155 ) IS, ^8, 31 
 
 French Armour Bolt 5, 6 98 
 
 » Ships _ 104 to 106 
 
 « Trial of Chilled Armour — 6 
 
 ii Trial of Plates at Gavre — 94 to 99 
 
 ;/ Turret, Bucharest 61 to 61 
 
 Funnel, Firing at ... ... ... ... ... ... \qq 
 
 a. 
 
 Garrison Point Fort — 80 
 
 Gatling, at Unarmoured Ships 
 Gavre formula 
 
 Plate Trials 
 
 German Cupola at Bucharest 
 
 Glacis Plate ... 
 
 Qlatton 
 
 a Turret fired at 
 
 Grloire Class ... 
 
 Goodrich, Lieut. -Comdr. on Attack of Alexandria ... — 125 
 
 Granite faced with Iron ... ... ... ... ... 139 — 
 
 Grau, Captain, killed — 119 
 
 Greenhill, Professor (Note) 
 
 Gregorini Iron Shot... 
 
 Grenfell, Captain 
 
 Griison Armour 
 
 » Copy of, at Meppen 
 
 ii Trial at Buokau 
 
 „ Chilled shot 
 
 „ Dutch Trial 
 
 ii Cupola at Bucharest, 1885-6 
 
 „ „ at Spezia, 1886 
 
 Gun, Armstrong, 100-ton B. 1 
 
 « » » M. L 
 
 „ 80-ton 
 
 H Krupp 
 
 „ 38-ton 
 
 i, 35-ton 
 
 ii 25-ton 
 
 n New Type 
 
 ii Smooth-bore 
 
 Guncotton, wet, in shells 
 
 , Action on thin Steel Plates 
 
 ii in Inglis' Sub-Committee Eeport 
 
 48 
 
 — 
 
 32 
 
 — 
 
 
 94 to 99 
 
 
 37 to 61 
 
 { 
 
 20, 38, 
 46, 69, 70 
 
 <1<2 
 
 111 
 
 2 
 
 
 29 
 
 — 
 
 73, 74, 119, 131 
 
 — 
 
 — 
 
 4 
 
 — 
 
 lto34 
 
 96 
 
 — 
 
 — 
 
 9,13 
 
 ...74, 87, 89, 91 
 
 — 
 
 87 
 
 — 
 
 — 
 
 37 to 51 
 
 — 
 
 17 to 34 
 
 151 
 
 — 
 
 61, 63, 91, 117 
 
 17 
 
 ... 60, 67, 139 
 
 — 
 
 22,39,61 
 
 — 
 
 48, 64 
 
 — 
 
 41 
 
 -- 
 
 42 
 
 — 
 
 ... 83,93,98 
 
 — 
 
 2,5 
 
 — 
 
 47, 81 
 
 — 
 
 47 
 
 — 
 
 81 
 
 — 
 
134 
 
 H. 
 
 Part I. Part II. 
 
 Hadfield steel Shot 74,149 — 
 
 Hard Armour ... ... ... ■■■ ■•• ••• <*6 
 
 „ Calculation of Effects 36,37,38 — 
 
 Heads of Projectiles and f 4, 12 to 16, 46, 73, 79, 81, 83, 98, 103, 1 _ 
 action in penetration...! 105,114,126,132,137,149,151 j 
 
 Flat 41,46,79,161 16 
 
 Hercules Target 7 — 
 
 a „ in Russia 39 — 
 
 Horizontal Armour Deck Plates 40,41,147 — 
 
 Horse Sands Fort — 81 
 
 Hotspur (firing at Glatton) ... 42 — 
 
 Howitzer Attack of Decks and Armour at High Angles 94 8 
 
 Huascar — 118 to 120 
 
 I. 
 
 Inclined Armour 3,46,79 — 
 
 India rubber 3 — 
 
 Inflexible — j "i^"' 
 
 n at Alexandria — 121 
 
 „ Deck Target 94 — 
 
 a Plates tested 99 — 
 
 Inglis, Major- General ...1,2,3,6,7,8,18,22,60,61,73,76,115 — 
 
 „ Diagram ... ... ... ... ... ... 23 — 
 
 „ Formula 24, 25 — 
 
 n Sub-Committee 73 to 86 — 
 
 Introduction of formula? ... ... ... ... ... 10 — 
 
 Invincible at Alexandria — 123 
 
 Iron approved for Forts ... ... ... ... ... 7 — 
 
 o Batteries — 81 
 
 „ Chilled 64,96 1,81 
 
 n Concrete 9,39 — 
 
 Ironside*, New ... ... ... ... ... ... 5 — 
 
 Iron, wrought, behaviour of — 84 
 
 „ Manufacture of ... ... ... ... ... — Q6 87 
 
 „ wrought, Trials with 1 to 9, 39 to 67, 87, 91, 96, 101 — 
 
 » Steel-faced 67 to 72, 79, 99, 106 to 168 — 
 
 Italia _ S 107,108, 
 
 I 109 
 
 Italian experiments at Spezia 63,91,117,161 17 
 
 » Ships _ 107 to 109 
 
 J. 
 
 Jones' Inclined Plates 
 
 Kamaria Fort 123 
 
 Kinbum 2 
 
 King's War Ships (Note) 103 119 
 
Part I. Part II. 
 
 — 8 
 
 — 103 
 
 74, 96 — 
 
 61, 93, 101 — 
 
 32 — 
 
 38 — 
 
 135 
 
 Kromhout, General, on chilled-iron cupolaa 
 
 Kronenfel's Kriegsschieffbauten 
 
 Krupp, chilled shot 
 
 » Experiments 
 
 » Factor for calculation perforation 
 
 " » for estimating power of guns 
 
 n Guns 039,47,61 4,10,17, 
 
 I 94 to 98 36,40 
 
 » Meppen Trials 93 101 _ 
 
 ii Rifled Mortars ... „, 
 
 '" '" '•' ot 
 
 Steel Projectiles 87,89,97,102,148,151 18,21,28 
 
 Kummersdorf Experiments _ W 4,\ 
 
 Laboratory, Competitive Shot 73 to 81, 106, 139 — 
 
 ii Manufacture of chilled shot 101 
 
 laminated Armour ... 5 _ 
 
 Land Forts, Shields 8, 9,48 to 61, 94, 139, 157 4 to 34 
 
 Landore Steel Shot 74 
 
 u » Plate and Guncotton 47 
 
 Lepanto _ iq» 
 
 Logarithms, Table of 33, 34 
 
 Lowmoor f -inch plate and guncotton 47 _ 
 
 M. 
 
 Machine Pile driving to illustrate action of shot ... 36 — 
 
 Mackinlay, Major, on Manufacture of Armour — 89 
 
 Madan, Lieut., on Huascar injuries — 119 
 
 Maitland, Colonel — Diagram 
 
 11 /, Formula 
 
 Mastless Ships 
 
 Manufaoture — Brief Notes 
 
 Marabout Fort (Alexandria) 
 
 Marsa el Kanat 
 
 Marrell's Armour 
 
 Matching Shot against Plates 
 
 Meppen Trials 
 
 Merrimac 
 
 Mex Forts (Alexandria) 
 
 Mill wall Iron Shield 
 
 Minotaur 
 
 Monarch at Alexandria 
 
 Monitors, Galena, &e 
 
 Mortar Attack of Decks 
 
 „ ,, Turrets 
 
 Mougin, Major 
 
 „ „ Turret 
 
 Muggiano (Spezia) Experiments 
 
 Muzzle Pivoting Gun in Shield 
 
 23 
 
 — 
 
 28 
 
 — 
 
 — 
 
 107 
 
 — 
 
 86 
 
 — 
 
 126 
 
 — 
 
 125 
 
 64, 147 
 
 38 
 
 36,37,86 
 
 — 
 
 ... 61, 93, 101 
 
 — 
 
 2, 5 
 
 — 
 
 
 122 to 124 
 
 42 
 
 — 
 
 5 
 
 104 
 
 — 
 
 121, 124 
 
 2, 3 
 
 — 
 
 40 
 
 104, 113 
 
 — 
 
 8,41 
 
 — 
 
 61, 62, 73 
 
 — 
 
 51 to 61 
 
 61, 63, 91, 117, 151 
 
 — 
 
 96 
 
 — 
 
 
 41 
 
Nettle Trials — steel-faced plates 
 
 » i, Steel 
 
 « r, n compound 
 
 New Ironsides 
 
 Noble, Captain Andrew 
 it Colonel 
 
 No Man's Land Fort 
 
 Normandie 
 
 136 
 N. 
 
 Part I. Part II. 
 
 68, 99, 108 90 94, 
 
 68 — 
 
 99, 108 90 
 
 2 — 
 
 16, 93 — 
 
 16, 23, 39 — 
 
 — 81 
 
 2 — 
 
 o. 
 
 Oblique perforation 
 
 ti calculations ... 
 
 a perforation, Krupp 
 
 » a Whitworth 
 
 O'Callagban, Major, on steel under impact 
 
 ,i „ Report on Bucharest Trials 
 
 Oehta (St. Petersburg) Trials 
 
 On m 1 1 Ku bcli a Fort (Alexandria) 
 
 46, 79, 103 128 
 
 — 128 
 103 — 
 
 46, 79 — 
 113 60 
 
 — 85 
 133 — 
 
 — 122 to 124 
 
 Faixham, General 
 
 Palliser, Sir W., Chilled Shell 
 i, „ n Shot 
 
 n ir n » improved 
 
 n o ii Trial of Points 
 
 ii Captain 
 
 ,, -English bolt 
 
 Parkgate Company Steel Plate 
 
 Penelope at Alexandria 
 
 Penetrating Figure " e " 
 
 Penton, Captain ... 
 
 Percy, Dr., on rolled iron 
 
 Perforation ... 
 
 n Definition of 
 
 ii Oblique 
 
 Petit and Gaudet Steel Plates 
 
 Petman Fuze... 
 
 PharoB Fort (Alexandria) 
 
 Picklecombe Fort 
 
 Pile Driver or Drop (Note) ... 
 
 2 
 
 — 
 
 41, 47 
 
 101 
 
 4, 42, 143 
 
 101 
 
 105, 137 
 
 102 
 
 72 
 
 — 
 
 103, 139 
 
 — 
 
 6 
 
 — 
 
 8 
 
 — 
 
 — 
 
 121 
 
 17 
 
 — 
 
 167 
 
 — 
 
 — 
 
 87 
 
 4, 6, 23 
 
 84,97 
 
 11,12 
 
 — 
 
 46, 79, 103 
 
 128 
 
 8 
 
 — 
 
 48, 147 
 
 106 
 
 — 
 
 122, 124 
 
 — 
 
 80 
 
 86 
 
 Plate-upon-Plate 
 
 Polyphemus ... 
 
 Portland Cement in Forts ... 
 
 n Glatton Experiments 
 
 Practical Directions 
 
 Pratt, Captain, killed 
 Projectiles, Ankarsrum 
 
 ii Cammell's 
 
 8, 9, 17, 36, 39, 41, 48 to 64, 97, 102, 139 — 
 
 9 
 
 42 
 
 74,79 
 
 113 
 
 119 
 90 
 
137 
 
 Part I. Part II. 
 
 Projectiles, Ekmau, Finspong 42,73,74,87,88 — 
 
 n Firminy jgg 
 
 " Krth 74, 78 — 
 
 » Forged Steel, Report on 80 — 
 
 n Gregorini Iron 73,74,119,131 — 
 
 " Griison 74,87,89,91 — 
 
 " Hadfield 74,149 102 
 
 a Holtzer... ... ... ... ... <t- jqq 
 
 Krupp's (filled 74; 96 _ 
 
 l ste «l 87,89.97,102,148,16118,21,28 
 
 ii Manufacture of 
 
 ii Laboratory, Palliser ... 6, 41 to 46, 49 to 61, 67 to 82 — 
 
 ii Landore, Siemen's 74 
 
 11 Palliser, Improved 105,1371 102,104, 
 
 11 a Condemned ... ... ... ... 102 
 
 11 San Vito 81 127 
 
 11 Ternitz ... ... ... ... ... ... jg 
 
 11 Terre Noire 74,91,127 
 
 « Vickers... ... ... ... ... ... 74 
 
 » Whitworth 13, 16, 74, 80, 91, 127 — 
 
 > Wilson (Cammell) 74,80 
 
 Protected Ships ^05 
 
 Prussian Trials of Griison Armour 4 9 to 16 
 
 11 n Guns, English, and Krupp 39 
 
 » 1, with Krupp Guns 39,47 4,10,13 
 
 Punching Armour, also see all trials of wrought-iron plates 4, 6, 10 97 
 
 R. 
 
 Part I. Part II. 
 
 Raoking 10, 11 — 
 
 11 GHatton Turret 10 — 
 
 Eas-el-Tin Port (Alexandria) — 121 to 124 
 
 Redoubtable — 127 
 
 Reed, Sir Edward — 107 
 
 Ridsdale Trial 167 
 
 BolfeKrake 3 _ 
 
 Royal Sovereign Turret 8 
 
 Rule of Thumb for Perforation 30,31 — 
 
 Russian Trials, Compound Armour 133 — 
 
 11 n SercvXes 39 — 
 
 S. 
 
 Saleh Aga Port (Alexandria) — 123 
 
 Sandwich Armour 8, 48 to 61, 64, 97, 101, 139 — 
 
 San Vito Projectile 91 — 
 
 Scheldt Batteries — 9 
 
 Scheveningen Trial 87 
 
 Schneider Steel Plates 64, 117, 133, 147, 161, 166 f 85 ' 
 
 Scfyuman, Major Von — 40,41,58 
 
 94 
 toiOl 
 
138 
 
 
 
 Part I. 
 
 Part II 
 
 Shannon Class 
 
 
 — 
 
 106 
 
 » Target Trial 
 
 
 47 
 
 — 
 
 Shattering hard armour 
 
 ( 11, 36, 64 to 70, 92, 97, \ 
 (ill, 121 to 137, 148, 151 j 
 
 2 to 6, 
 9,20 
 
 Shell, common, against armour 
 
 
 
 51,62 
 
 — 
 
 ii u „ unarmoured ships 
 
 
 47 
 
 — 
 
 ii n exploded by impact behind 
 
 armour 
 
 39 
 
 — 
 
 » Palliser, filled, fired at armour 
 
 
 41 
 
 104, 119 
 
 ii ii n deck armour 
 
 
 41 
 
 
 
 Also see Projectiles. 
 
 
 
 
 Shields in Masonry ... 
 
 
 7,139 
 
 — 
 
 ii Meppen Trial of 
 
 
 96 
 
 — 
 
 » _ 2, 7, 39 to 72 
 
 , 87 to 104, 139 ; 
 
 147, 151 { 
 
 4 to 34 
 79, 90, 94 
 
 Shot, Behaviour of 
 
 
 12,13 
 
 — 
 
 ii Cast-iron 
 
 
 12 
 
 
 
 ii Chilled (see Projectile, Palliser) 
 
 
 — 
 
 — 
 
 ii Mat-headed 
 
 14, 15, 16, 41, 
 
 46, 79, 151 
 
 16 
 
 ii Wrought-iron 
 
 
 13,14 
 
 
 
 Shrapnel shell fired at Shannon ... 
 
 
 48 
 
 
 
 Siemens steel shell 
 
 
 71 
 
 
 
 Silsileh Battery (Alexandria) 
 
 
 — 122 to 124 
 
 Simpson, Admiral, Laminated Armour ... 
 
 
 — 
 
 118 
 
 Sixty-eight-pounder 
 
 
 2 
 
 — 
 
 Smooth-bore experiments ... 
 
 
 2, 5, 12 
 
 — 
 
 Soft Armour ... 
 
 
 21 
 
 
 
 Southport Whitworth Trials 
 
 
 46, 147 
 
 .— 
 
 Space, air 
 
 9, 4S 
 
 i, 50, 61, 78 
 
 
 
 ii between plates 
 
 
 9 
 
 
 
 Special Committee on Iron ... 
 
 
 3 
 
 
 
 Speoific Gravity of Armour 
 
 
 — 
 
 87 
 
 ii n chilled shot 
 
 
 — 
 
 101 
 
 Spezia ohilled-iron Ports 
 
 
 
 
 6,32 
 
 ii Experiments ... 
 
 ... 61, 63, 91, 117, 151 
 
 17 
 
 Spitbank Fort 
 
 
 
 
 82 
 
 Spithead Forts 
 
 
 — 
 
 81 
 
 Steel Armour originally condemned ... ... ... 3 
 
 11 11 Powers employing it 101 
 
 T « aI s 3,8,64,117,133,147,161,156 94 to 101 
 
 „ -Facedarmour P\ 6 ?i 7 ?i 99 ' * 00 > 106 > 108 ' } 
 
 1 117,133,139,147,151 J ~ 
 
 St. Chamond Projectile 28 32 
 
 Stevens, John — Armour 1 
 
 St. Helen's Fort 
 
 St. Marie's Battery 
 
 St. Petersburg (Ochta) Trials 
 
 Studs, condemned ... 
 
 Sultan at Alexandria 
 
 Summary of behaviour of Plates 04 
 
 Superb at Alexandria 121 to 125 
 
 Swedish Shot 42, 73j 71] 87j 8g) 90 _ 
 
 
 80 
 
 
 4, 9 
 
 133 
 
 — 
 
 80 
 
 — 
 
 
 121 to 125 
 
139 
 
 Tables, Effects on Hard Armour ... 
 » Errors of Rule of Thumb... 
 
 » " K " correction 
 
 ii Logarithms ... 
 
 ii Results of New Type Guns 
 
 n Spezia Results 
 
 t 
 
 " I 
 
 Targets, No. " 31 " 
 
 ii ii "32 "Deck 
 
 ■ "32a" „ 
 
 » i, "33" 
 
 ii ,, " 34 " and " 35 " Turret 
 
 ii ii "39" Shannon 
 
 ii „ "40" 
 
 ii ii " 40 " strengthened ... 
 
 ,, „ "41" 
 
 ,, "44" 
 
 Tegel Trials 
 
 Temeraire 
 
 Ternitz Projectiles 
 
 Terre Noire Projectiles 
 
 Terrible Plates 
 
 Thames Co. Steel Plates 
 
 Thunderer 
 
 „ Deck 
 
 Trusty Turret 
 
 Turret, Eastbourne, 1886 
 
 ii Glutton 
 
 ii Armour 
 
 U. 
 
 United States Report on Alexandria Attack — 121,125 
 
 Vertical Eire 40, 94^ j^ 1 ^ 
 
 Very, Lieut 1,2,3,5,6,8,11,32,39 118,119 
 
 Vessels — Armed Structures — 103 
 
 Viokers' Steel Projectiles 74 — 
 
 VonShutz — 12 
 
 ( 20 38 
 
 Vor Panzer (glacis plate) — £ 46^69,70 
 
 W. 
 
 J? 30,80,84 — 
 
 Walford, Major, on Alexandria Attack — 121,125 
 
 Ward, Admiral Le Hunte — 125 
 
 Part I. 
 
 Part II, 
 
 37 
 
 — 
 
 31 
 
 — 
 
 35 
 
 — 
 
 33, 34 
 
 — 
 
 84, 86, 86 
 
 — 
 
 129 
 
 — 
 
 . 25, 27, 35 
 
 — 
 
 39 
 
 — 
 
 40 
 
 — 
 
 41 
 
 — 
 
 41 
 
 — 
 
 41 
 
 — 
 
 47 
 
 — 
 
 48 
 
 — 
 
 54 
 
 — 
 
 50, 57 
 
 — 
 
 139 
 
 — 
 
 39, 47 
 
 4 
 
 
 121 to 125 
 
 — 
 
 13 to 16 
 
 . 74, 91, 127 
 
 — 
 
 — 
 
 95 
 
 8 
 
 — 
 
 — 
 
 Ill, 118 
 
 41 
 
 — 
 
 3 
 
 — 
 
 — 
 
 83 
 
 42 
 
 — 
 
 .3, 41, 42, 147 
 
 36 to 72 
 
140 
 
 
 
 Part I. 
 
 Part II 
 
 Warrior, &c 
 
 
 2 
 
 — 
 
 n Target 
 
 I? Experiment with common shell. 
 
 Water, firing through 
 
 Water-line, firing below 
 
 
 17 
 
 39 
 
 41 
 
 41 
 
 — 
 
 Weser Iron Batteries 
 
 
 — 
 
 6,9 
 
 Whitworth Guns, Paraguay 
 
 n Steel shell, flat-headed 
 
 
 13,46,79 
 
 118 
 
 ■ n at air spaces 
 
 ■ n at water ... 
 
 
 78 
 
 41 
 
 — 
 
 » ii direct 
 
 
 79,146 
 
 — 
 
 n n oblique 
 ii ii ogival pointed 
 1, ii Spezia Trial 
 Wilson's Armour, Behaviour 
 
 
 46,79 — 
 
 74,80,146 — 
 
 91 — 
 
 — 84, 90, Ac. 
 
 i i, Manufacture 
 
 
 — 
 
 88 
 
 , Trials | 67 
 
 68, 
 117 
 
 79, 99, 100, 106, 108, > 
 13», 139, 147, 161 | 
 
 35 to 72, 
 90 
 
 Wrought -Iron Armour, Behaviour Trials 
 
 
 — 
 
 3,84 
 
 n cap on shot 
 
 
 88 
 
 — 
 
 • shot 
 
 
 14 
 
 — 
 
.'■.■■■•',v ' 
 
 :. .-:;:,;