- THE PRINCIPLES AND METHODS OF ARMOUR PROTECTION IN MODERN WARSHIPS - W.H. White - BRASSEY 1904

THE PRINCIPLES AND METHODS OF ARMOUR PROTECTION IN MODERN WARSHIPS - W.H. White - BRASSEY 1904

, -: : THE PRINCIPLES AND METHODS OF ARMOUR PROTECTION IN MODERN WARSHIPS - W.H. White - BRASSEY 1904 : jpg, rar : 8 rapidshare: R A P I D - L I N K : tsushima PS: NMD, , .

- 22

, -: The Principles and Methods of Armour Protection in Modern Warships. Discussions on the defensive qualities of modern war-ships frequently disclose lack of knowledge of the history of the applications of armour, and a failure to appreciate the principles underlying its use. It may be useful therefore briefly to review the subject in both aspects. At first, armour was used simply for the purpose of protecting ships of war against the destructive effects of percussive shells. When General Paishans introduced horizontal shell fire (about 1824). he also indicated the necessity for armour protection on the sides of ships, and proposed the use of iron plates which would keep out shells. After investigation by a special Committee, the conclusion was reached by French authorities that the load of plating required would prove too great to be carried by sea-going ships. This conclusion was generally accepted, and acted upon in all Navies until the destruction of the Turkish Fleet by Russian shell-fire at Sinope (in 1854) recalled attention to the warnings of Paixhans. Napoleon III., aided by that great naval architect Dupuy de Lome, broke away from this tradition ; and, by the construction of the first armoured floating batteries (built during the Crimean War), and subsequent sea-going ironclads, of which La Gloire was the pioneer, he introduced a new era in warship construction. At the outset the arrangement of armour was extremely simple, as all ships were armed on the broadside principle. Ironclads were frigate-built, carrying nearly all their armament on the main deck. Their sides were cased with armour from the upper deck down to 5 or 6 feet below the waterline. For a time the system of so-called " complete protection " alone prevailedthat is to say, throughout the limits of depth just mentioned the sides were clothed over with armour from stem to stern ; the numerous guns were mounted at broadside ports cut through the armour between the main and upper decks. The Warrior and other of our first sea-going ironclads departed from this principle of complete protection. A considerable length at each end was left without armour, and the main deck battery was completed by sthwart-ship armoured bulkheads. Undoubtedly the principal reason for this adoption of " unarmoured ends " was the feeling that the seagoing qualities of the ships would be improved greatly, if the ends were relieved from heavy loads of armour required with complete protection. The serious dangers resulting from exposure of the steering gear, and the liability to easy perforation of the unarmoured ends in the region of the waterline, by even the lightest projectiles as well as by shells, soon led to the abandonment of the Warrior system of armour, and the adoption of complete protection in the vessels of the Minotaur class and in the "converted" ironclads built in 1861-63. As the weights of guns were increased, in order to obtain greater perforation aud shell power, their numbers were diminished ; shorter batteries consequently sufficed. Hence resulted the " belt and battery" system of armour, which, the French used as early as 1859 in the Magenta and Solferino; these ships had double-storeyed central batteries amidships, associated with a belt of armour rising only to the height of.the main deck before and abaft the battery. The Achilles and Northumberland of 1861 also had belts and batteries, ■aud that system was adopted by Sir Edward Eeed in his designs for broadside ships from 1862 to 1870. The batteries had a single storey in most cases, although in some instances the central batteries were double storeyed and were supplemented by detached batteries placed at the bow and stern to protect " chase " guns. In none of these ships was prominence given to horizontal (or deck) armour; the protection consisted simply of vertical side armour associated with bulkheads enclosing the batteries. The decks afc the top of the armour belts, before and abaft the batteries, were of ordinary construction. In American " monitors " armed on the turret principle, introduced by Ericsson about 1862, the freeboard was exceedingly small. The vertical side armour, therefore, became a narrow belt, and the upper decks were very strongly plated.; The few but heavy guns were mounted in revolving turrets. A very small target was exposed to gun-fire, but this involved slight " reserve of "buoyancy," accompanied by great risk of foundering. The " cupola " turret-ships, introduced almost simultaneously by Captain Cowper Coles, had a greater freeboard than the American " monitors "; although this freeboard was very moderate when compared with, that of contemporary broadside ships. In all these " monitors" and turret-ships the hulls were completely protected throughout the length to the upper deck. The armoured cupolas or tur

, -: In the " breastwork monitor " system, introduced by Sir Edward Reed about 186G, after the Miantonomoh had visited this country, the low freeboard (to the upper deck) of the American " monitor *' type was supplemented by a central armoured " breastwork " standing upon the strongly-plated upper deck. This breastwork was narrower than the upper deck and extended only over a portion of the length. The turret-bases were protected by the breastwork, and an iron protective deck was fitted on the top of the breastwork. When the Dreadnought was re-designed by Sir Nathaniel Baraaby in 1873 the breastwork system was modified. The vertical armour was brought out to the full width of the ship, so forming a. central battery within which the turret bases were placed. In the latest broadside ships also, such as the Alexandra, which were built in 1873-74, protective iron decks placed at the top of the belt were associated with vertical side armour. Speaking broadly, therefore, it may be said that as the result of the first 14 or 15 years' experience with armoured ships there was a growing appreciation of the necessity for and value of horizontal or deck armour, while there was practically universal agreement that what may be termed the " hull protection" by vertical armour, should extend from the height of the main deck down to five or six feet below water. The disposition of the remainder of the armour was controlled by the disposition of the armament, and the protection demanded for guns and crews. In broadside ships this vertical armour was arranged in the form of batteries; in turret ships it wa& concentrated in turrets and in protection to the bases of the turrets, and to the mechanisms needed for revolving turrets, transporting ammunition, or working the heavy guns. Armoured " conning towers " had also grown in favour, having; first become prominent in connection with turret ships. This brief statement must suffice to summarise the condition of affairs which prevailed at the time when Lord Dufferm's Committee on Designs (appointed in 1871 after the sad loss of the Captain) "took stock" of the Royal Navy and made suggestions in regard to the principles which should guide future construction. As a. consequence of the recommendations of this Committee the u central-citadel" system was introduced in 1873. It was first embodied in the Inflexible and subsequently adopted in many other vessels. The thick armour was concentrated on a citadel of moderate longitudinal extent in relation to the total length of the ship. Before and abaft the citadel the defence consisted of a strong under-water armoured deck, above which the " unarmoured ends " were minutely sub-divided and reinforced by cork-packing, designed to assist the maintenance of buoyancy and stability should the ends be seriously injured by shot and shell (or " riddled ") in action. French designers, when the reconstruction of their fleet was -undertaken after the conclusion of the war with Germany, decided ; to build battleships in which the hull-protection was restricted to a shallow water-line belt, rising only about 2 to 3 feefc above water, but carried to the usual depth below water. This belt was extended from stem to stern, and was associated with a protective deck situated at the height of the top of the belt. The vessels were built with considerable heights of freeboard. Their heavy guns were mounted iu isolated armoured barbettes high above water, with shallow rings of vertical armour protecting the turn-tables. The floors of the barbettes were strongly plated, and armoured ammunition tubes of small diameter were fitted from the belt-deck up to the floors of the barbettes. The sides of the ships above the armour belt and the support to the barbettes were formed by thin structural iron or steel plates and stiffeners. There was a secondary armament of unprotected light guns on the main deck. Trench designers, in fact, boldly surrendered to easy destruction by shell fire, the sides and structure of the ships above the shallow armour belts; they did not protect the secondary armaments, restricting protection to the heavy guns. a time they made no endeavour either by minute watertight subdivision or by cork packing to limit the spaces to which water could find entry if the thin sides were perforated near to the waterline. Their intention was to produce vessels carrying heavy guns high above water, so that they could be efficiently fought in a seaway, when the lower-placed guns of British turret ships with moderate freeboard would be either fought with difficulty or not be capable of fighting. To secure this superior offensive power they were prepared to sacrifice defensive power, as compared with contemporary British turret ships.

, -: Judged by smooth water tests, such as are ordinarily applied in discussions of the efficiency of armour protection, French ships of the period of 1872 to 1878 were sadly deficient in protection, and this feature was emphasised in consequence of the fact that most of these ships when completed exceeded their designed draught. The very small height above water originally contemplated fortthe armour belts was not obtained. On the other hand, when trials came to be made between the Imperieuse of the Eoyal Navy (which bore a close resemblance to the French type) and a turret ship of moderate freeboard, armoured on the central systemwith turrets placed en echelon to give greater horizontal commandit was shown that, in comparatively moderate weather, when the Imperieuse could fight ner high-placed guns with perfect efficiency the ship of moderate freeboard with lier guns near the water was practically unable to make an effective reply. This experience had much, to do with the introduction into the Eoyal Navy of vessels of the Admiral class, and the practical abandonment of the turret type which (with certain modifications in the arrangement of armour) had persisted in the Royal Navy from the design of the Thunderer and Devastation in 1868-69. In the Admiral classas I pointed out many years ago when controversies arose as to the merits of the designthe governing conditions were as follows:First, definite limits of cost had to be conformed to; secondly, provision had to be made for certain. qualities of speed, armament and protection, which it was considered would make the Admirals more than a match for contemporary vessels in the French and other foreign Navies, The hull armour consisted of a narrow waterline belt, and was associated with a strong steel deck at the top of the belt. This belt, however, was not carried to the bow or stern, extending over about 45 per cent, of the length, and having athwart-ship armoured bulkheads fitted at its ends. Before and abaft these bulkheads the protection consisted simply of a strong steel under-water deck, like that of the central citadel ships, and the space above the deck was minutely subdivided and assigned to the stowage of coal, stores, water, chain cables, etc., so as to diminish the unoccupied spaces to which the water could find access if the sides were "riddled " in action. The restriction of the belt within these limits has been strongly condemned, and alleged to be markedly inferior to the continuous belt of French ships. In these criticisms, however, there was no fair appreciation of many important features in which the English ships were undoubtedly superior. For example, the Admiral class had cellular sides formed by longitudinal bulkheads extending from the belt to the main decks, and placed about ten feet away from the side plating. Cellular compartments were thus formed, and were minutely subdivided by numerous transverse bulkheads. It was also arranged to pack these spaces with coal or patent fuel as an additional protection. The principal watertight transverse bulkheads were extended to the main deck, instead of being stopped below the belt deck as in the French ships. The armoured barbettes for the heavy guns and the armoured ammunition tubes, also gave much better protection, than that afforded by the corresponding features of French ships. Experience has proved that the disposition of armament in the British ships is more efficient; and although the French maintained their own system for many years, and critics -gud us to adopt it, in all recent French constructions our system has been followed. All other Navies have taken a similar course. Another feature, in which the Admiral class marked a new departure in British practice, was the association with heavy guns (placed high above water) of a powerful "secondary" armament of 6-in. guns and numerous quick-firers of small calibre. This secondary armament Itad only shield protection, while the heavy guns and their mechanisms were protected by thick armour.

, -: It is interesting to note that when the reconstruction of the German Fleet was begun, about 1895, the designs of the battleships were largely influenced by considerations very similar to those above described for the Admiral class. In view of the sizes of the docks available, moderate dimensions were essential for the ships. Relatively high speed and powerful armaments were decided upon, and the defensive qualities were necessarily diminished as compared with those of the larger battleships then building for other Navies. The hull armour of these German ships took the form of a narrow water-line belt rising only two or three feet above water with a protective deck at its top; an arrangement similar to that of the French ships designed twenty-five years before, except that at the stern there was an "imarmoured end." The guns were placed at a great height above water in isolated and shallow barbettes or casemates, with small armoured ammunition tubes reaching down to the belt deck, and with unarmoured supports to the gun emplacements. In many published descriptions of these CJermaa vessels their considerable offensive powers have been dwelt upon, while no notice occurs of the great sacrifices made in defence in order to increase-offence. This is only one example, out of a multitude that might be-mentioned, where criticisms have been limited to certain selected features, and a comprehensive or balanced appreciation, of all the qualities of a design has been lacking.

, -: When the first great programme for the reconstruction of the Kcyal Navy had to be considered in 1888, new conditions had come into play. The whole situation was then reviewed afresh; as to the disposition and character of armour and of armaments, speed, coal endurance and sea-going capability. The discussions which followed the reading of the papers which I contributed to the " Transactions of the Institution of Naval Architects," in 1889-92, contain the opinions of many leading authorities on war-ship design, and can still be consulted with advantage. Here it must suffice to say that - after full discussion by the Board of Admiralty, who called into Council a Special Committee of distinguished Naval Officers - it was decided to adopt the features of armour and armament embodied in the design of the Eoyal Sovereign class. Many changes wore made from the Admiral class, including the following features:-First, high-freeboaid throughout the length, instead of low extremities like those of the Admiral class, which had been found seiiously to compromise the power of maintaining speed in. rough water. Second, the association of a strong armour belt in the region of the water-line, and the usual protective deck at its top, with thin side armour reaching to the height of the main, deck, about 9 or 10 feet, above water. Third, the construction of barbettes for the heavy guns, which were fully armoured right down to the belt deck, instead of repeating the shallow barbettes and ammunition tabes of the Admiral class. Fourth, the adoption of the "casemate" system, giving protection to some of the 6-in. quick-firing guns. These changes in the distribution of the armour, and in the nature of the armaments, grew out of the then recent introduction of high explosives and quick-firing guns of larger natures. They involved the assignment of much greater weights to the armour, but they added enormously to the defensive power. The offensive power was also greatly increased, in the form of a powerful secondary armament. My personal representations to the Board had much to do with the change in the design for the barbettes and with, the introduction of casemates. As regards the barbettes, it was obvious thatplaced as they were in previous ships, high above the true armoured hulls, with only armoured ammunition tubes connecting them with, the belt deckthere were great risks of serious damage and derangement in consequence of the introduction of high-explosives and of larger quick-firing guns. Consequently, it became essential to carry the armoured walls of the barbettes down to the belt-deck, notwithstanding the very large additions of weight and cost which were involved. Further, it appeared in the highest degree desirable that there should be no undue restriction of the space required for the effective working of the heavy guns as well as the rapid and safe transport of their ammunition. The barbette system was primarily governed by these considerations, and a heavy price had to be paid for the increased safety and efficiency which was secured. There were at that time, and have since been, many proposals fur diminishing the dimensions of barbettes, and so lessening the weight and cost of armour. doubt, it might have been possible to work . the guns. in. smaller barbettes, or to have restricted the dimensions, thickness and weights of the lower portions of the armour. This has been done in many foreign ships; but, in my opinion, the policy which the Admiralty has consistently maintained in battleships of the Eoyal Navyof making liberal provision of space and weight for the armoured enclosures and shields of heavy gunshas been wise and well-considered, although the size and cost of ships have been necessarily increased.

, -: The work of the naval architect is, of course, simplified if any arrangement can be devised by which the weight of material neede.1 for protection of the armament can be diminished. It is a fact that iu many designs which have been proposed or adopted in other Navies for mounting and working heavy guns, considerable economies of weight have been effected as compared with the practice of the Eoyal Navy. My profound conviction is, however, that in resisting the introduction of similar arrangements into H.Jl.'s ships, the responsible administrators at the Admiralty have taken the best possible course; and that the additional expenditure which has thus been incurred and the larger dimensions of the ships which have been consequent, have resulted in greater fighting efficiency and better protection. I trust that the time may never come when designs of the ships of the Eoyal Navy will be largely influenced by the remarks of amateur critics, imperfectly informed as to what is involved, and judging the merits of designs simply or chiefly by what naval architects have been able to accomplish on reputed or tabulated displacement tonnages. Eepeatedly and publicly I have expressed the opinion that something much larger is involved in these debates than the credit or discredit of a particular naval architect. My belief is that when working under the same conditions of speed, coal-endurance, offence and defence, the skill and inventive power of naval architects in this and other countries, applied to a particular problem at a given date, will almost always result in the production of vessels of practically identical displacement. "When extraordinary results are said to be achieved on relatively small displacements, and certain features have been unusually developed in a design, it is well to inquire closely - What has been sacrificed as compared with other designs?

, -: One of the subjects most debated in connection with the distribution of armour is that relating to its longitudinal extension in the region of the water-line. It has been explained that, except in the Warrior, and a few other early ironclads of the Eoyal Navy, the water-line belt of armour extended throughout the length, up to the time when the "central citadel" system was introduced. The Committee on Designs of 1871 were of opinion that alternative methods of protection towards the extremities were preferable; provided they were associated with strongly armoured citadels occupying a considerable portion of the length amidships, these citadels rising to about ten feet above water. The controversy m regard . to the supposed dangers of " unarmoured ends " arose in connection with the design of the Inflexible, the first citadel ship-, and was referred by the Admiralty to a Special Committee, which reported in favour of the central-citadel system. This deliverance, however, was not accepted by opponents of unarmoured ends; and prophecies of disaster in action to such vessels were numerous and gloomy. Less has been heard on this point since, at the battle of the Yalu, two Chinese ironcladsbuilt in Germany on the central-citadel systeminstead of being destroyed or made to "turn turtle," remained effective to the close of the day, and actually followed up the Japanese squadron on its withdrawal. Considering that the Japanese ships were armed with quick-firing guns, and were far superior in marksmanship to the Chinese, this object-lesson has great importance, particularly when it is known that the conditions of stability, and the extent of armour protection, in these vessels were much less favourable than the corresponding features in British ships. In previous discussions of this subject I have expressed the opinion which I still hold, namely, that the longitudinal extension of the armour is a matter of much less importance than its vertical extension, when considered in regard to the capability of battleships to maintain buoyancy and stability when attacked by modern gun fire. This critical featureextreme narrowness of belt armourwas common both to French and British ships of the period uuder consideration. If armour protection is to be treated as a guarantee for the maintenance of buoyancy and stability in action, then it is a matter of certainty, demonstrable by calculation, that the vertical extension of the armourin other words, the height to which it rises above wateris a matter of much greater importance than is the longitudinal extension of a narrow belt. It is unfortunately true of a very great number of these narrow-belted ships (both British and foreign) that, owing to additions made during construction or to other causes, the intended height of armour belt above water has not been attained in completed ships. As a result many of these vessels, when fully laden, have the top of the armour-belt less than a foot above water; and in some instances the belts are practically awash or under water. Under these circumstances it, need hardly be said that the question of unarmoured ends becomes-insignificant. Although many thousands of pounds have been spent on the hull-armour of such vessels, yet it has practically no value in maintaining buoyancy and stability. But this is not all that may be. said on this matter. When the vessels are under way in smooth water the tops of the belts at certain parts of the length are continuously under water, because of the rise of the waves on the sides. In a seaway, of course, the case is much worse, owing to the greater rise of the sea-waves and the rolling motions of the ships. Indeed, this type of ironclad really resembles a "monitor" in its armoured freeboard; and the towering superstructurebuilt above the belt to provide accommodation, and to carry high above water the armoured or other enclosures containing the gunswhile of the greatest value from many points of view, does not contribute to the protection.

, -: The great height at which these heavy weights of armour and armament are caraed also raises the centre of gravity of this type so much, thatapart from the contribution of the unarmoured structure to stabilitythe vessels can have no practical range of stability. This is equally true whether the armour belt is carried to the bow or is stopped short and associated with unarmoured ends. There is a periodical recurrence of this onslaught on unarmoured ends, and a complete ignoring of the fact that successive Boards of Admiralty and Committees of Advice, after investigating all the conditions, have affirmed the principle that it is best to concentrate defence towards the middle of the length, and to leave the ends either unarmoured or lightly armoured. In 1889 when the designs of battleships to be built under the Naval Defence Act were referred to a special committee, this matter of protection towards the extremities was most carefully considered; after the controversies above mentioned in regard to central citadels and the Admiral class had occupied the public mind. My personal recommendation then was that the bow should be lightly armoured before the citadel in vessels of the Eoyal Sovereign class, chiefly with the intention to give greater power of maintaining speed should the bows of the vessels be struck when chasing an enemy. But the conclusion reached by the Committee, and affirmed by the Board after full consideration, was, that both ends of the ships should be left unarmoured and that an under-water steel protective deck with sub-division was the best arrangement, giving ample security against damage in action. It lias been explained that, so far as the maintenance of stability is concerned, the presence or absence of armour at the ends is of comparatively small importance. For the maintenance of " trim " when the ends are riddled it is, of course, conceivable that injuries forward might interfere both with manoeuvring power and with the maintenance of speed. An unarmoured stern, however, could only produce changes of M trim" of a moderate character, not interfering with either speed or manoeuvring power. In the later vessels of the Formidable class and in the King Edward these considerations Lave led to the extension of the armour to the bow and to the abolition of the forward bulkhead of the citadel; the thickness of the side armour being gradually diminished towards the bow, the side armour carried to the height of the main deck, and strong protective deck plating being fitted at the top of the side armour. In some recent armoured cruisers a similar arrangement has been made and the bow has been lightly armoured for a certain length up to the height of tho upper deck, in order to prevent riddling by light quick-firing guns-and consequent check to the speed when chasing an enemy. Having carefully studied the adverse criticisms of the British system of disposing armour towards the extremities of Avar ships, I am distinctly of opinion that the condemnation rests upon no solid ground, and leaves out of account many important considerations which have influenced Admiralty practice and added to the efficiency of His Majesty's ships. In regard to the " casemate" system introduced into the Royal Navy in 1888-89, and since largely imitated in foreign Navies, there has been much misapprehension, and it may be useful to make a. detailed statement.

, -: When the use of high explosives, and the introduction of larger natures of quick-firing guns, made it obvious that protection must be found for the guns and crews of the secondary armament, two views-were taken of the best method of procedure. One of these found an early and excellent illustration in the armoured cruiser Dupuy de Lome; for which the design was prepared by my friend the late M. dc Eussy, who has recently died full of years and honours. Id the Dupuy de Lome he reverted to the system of "complete protection," clothing over the sides with steel armour, about 4 inches in thickness, extending from the upper deck down to a few feet below water. Upon this armoured hull he placed a number of armoured turrets in which the guns were installed. This was a perfectly logical system; but like all arrangements it has drawbacks as well as advantages. For example, while this thin armour prevented explosion, within tha ship, of high explosive shells having large "bursters," its presence actually determined the explosion of such shells at their points of impact and so tended to produce wholesale destruction of th& armoured sides. Further, it was experimentally demonstrated that,, with the small thickness of armour use:!, even of the best quality then obtainable, there was no need to use high explosive shells; with which there were unavoidable risks to the users. Chilled iron (Palliser) or suitable steel projectiles of small cost sufficed to perforate and drive into the vessel an amount of mitraille which would inflict greater damage, over a larger area, than high explosive shells would cause. Furthermore, the great weight of side armour required for this system of protection and disposition of armament was useful only for the protection of the hull : separate protection, in the form of turrets, had to be devised for the guns. M. de Bussy fully recognised the possibility of serious damage being done, if the guns were placed in a battery formed by the side armour; and he preferred to accept the necessarily large weight of side armour rather than, to incur the risks incidental to a battery, with concentrated guns and unavoidable points of weakness produced by cutting " ports " in the armour. What was being done in Trance was, of course, perfectly well known to us, and it was not without the most careful consideration and discussionincluding the collection of opinions from the highest British authorities in and outside the Navythat my proposal to adopt casemates was approved. The principal reasons which led to this approval were as follows: "We had only a moderate number of 6-ineh guns to mount. It was desired to distribute these as widely as possible. It was preferred not to adopt the turret system, because that involved mechanical appliances for the revolution of the turrets, . and did not give the same facility for the application of manual power, which was obtained when the protecting armour was fixed and the guns were mounted on centre pivots. In addition it was. seen that the service of the ammunition could be more readily and certainly dealt with in the larger casemates than in the restricted sp.ice of a turret. The isolation of each gun and its crew in a separate enclosure was considered to be a great advantage ; and the habitability of the crews' quarters, on the main deck amidships, was much increased, as the absence of side armour made easy the provision of natural light and air for the spaces in which a large proportion of the crew lived. This latter consideration, of course, was subordinate to the others mentioned. The character of the protection to be afforded in the casematesincluding the thickness of the plating on the backs, as well as the frontwas decided after careful trials in the Resistance, when 6-in. guns were used with high explosive shells, as well as with common shells and powder-bursters. There has been much criticism of the relative weakness of the backs of the casemates; but this has, to a large extent, arisen from ignorance of the results obtained in the Resistance, to which results it would obviously be undesirable for me to refer in detail. Looking back on the decision reached fifteen years ago, and having regard to all subsequent experience, as well as to the adverse opinions that have been expressed, it does not appear that there is any reason to regret the adoption and extensive use of the casemate system in British ships.

, -: One other point deserves notice, namely, the fact that the upper deck 6-in. quick-firing guns of the Koyal Sovereign class were not fitted with casemate protection. Here again there was no oversight in preparing the design. At that time great importance was attached to keeping the upper deck as free and unencumbered as possible; and it was for that reason that casemates were not fitted. In the Kenown, which I designed in 1892, four casemates were used on the upper deck, at the angles of the screen bulkheads, where they caused little obstruction. This has since been made the rule in later vessels. As is well known the number of the upper deck casemates has been increased in the latest battleships and cruisers, with obvious advantages to fighting efficiency at sea ; and the old objection to obstruction on the upper deck has disappeared as the result of experience. In their recent refits effect has been given to the idea of fitting casemates on the upper decks in the Royal Sovereign class, the ships having been so designed that they could carry them safely. For many years past, in some Navies, there has been a reversion to central batteries, instead of casemates or turrets, for the protection of secondary armaments. In 1878 I prepared a design for a modified Inflexible, .having turrets placed as in the Dreadnought and with a central battery containing a number of lighter guns. The Nile and Trafalgar (of 1885) had this disposition of armament and armour. But there were serious objections to so concentrated a ■ battery, and it was not adopted in the Iioyal Sovereign class. French designers introduced a central-battery into the Brennus, launched in 1891. The Italians, in battleships and armoured ■cruisers built from 1890 to 1896, also adopted central batteries for ■C-in. quick-firing guns; and similar batteries were introduced (about 1899) into designs for the Maine class of the United States Navy. A more recent example of this battery system is found in the Japanese battleship- Mikasa, built by Messrs. Tickers, Maxim & Co,, where the central battery is sub-divided by thick steel traverses running longitudinally in addition to traverses of the uffial character placed transversely. Of course some advantages are obtained from central batteries, particularly as regards economy in weight of deck armour; since it is unnecessary to thickly plate the battery deck when there is substantial side armour protecting it; but there are also disadvantages. These considerations were not overlooked in Admiralty practice when casemates were adopted: but the governing consideration, so long as the number of guns to be mounted permitted wide-distribution, was that by using separate casemates there was au avoidance of serious risks run when a number of guns are mounted in a single-armoured battery, with necessarily weak positions at the gun ports, and with large numbers of men and considerable amounts of explosives crowded into one enclosure. The use of traverses, of course, tends to reduce this risk: but does not by any means remove it. Suggestions have been made that, being the originator of the casemate system, I was naturally averse to change. This is absolutely incorrect, so also is the statement that a change in Admiralty practice followed my resignation. Tlie point is of small importance ; but, as a matter of fact, it may be stated that in my last battleship designs, which eventually took the form embodied in the King Edward class, I proposed the adoption of a central main-deck battery for the 6-in. guns. This recommendation was made because the adoption, of 9*2 in. guns, as supplements to the 12~in. guns on the upper deck, made l:>rge demands on space, and greatly restricted the positions in which the 6-in. guns could be mounted. These guns were necessarily placed dose together, and there was therefore no longer reason for locating them in separate casemates. The balance of advantage, in short, lay in the adoption of a central battery, from the point of view of the protection of the secondary armaments. It had the further advantage of facilitating the protection for the 9-2-in. guns and of enabling a considerable reduction to be made in the weight of protective plating on the main deck.

, -: From published accounts of the Duke of Edinburgh class of armoured cruiserthe first design prepared by my friend and successor, Mr. Philip Wattsit appears that the Admiralty has adhered to its fixed policy of giving to first-class cruisers a secondary armament identical with that adopted for first-class battleships of the same date. The arrangements for the 9 2-in. gun mountings and shields and the C-in. gun battery in that design are understood to be identical with those introduced into the King Edward class. It is unfortunate that writers on these subjects so often treat matters of policy in construction, governed by decisions of the Board of Admiralty, as if they were dependent on the personal opinions of gentlemen who may, for the time, be the responsible naval architects at the bead of the Construction Department, but so much has been said on this subject that it is desirable to place the facts on record. Unanimity of opinion has not been reached, and probably never will be, as to the best system of protection and disposition for the secondary armament. In our latest designs, and in those of the United States and Germany, central batteries have been adopted, while Italy (which led the way in central batteries) as well as France and Eussia are adopting isolated turrets, and Austria has passed from central batteries to casemates. The explanation of this diversity of opinion is to be found in the fact that armaments differ in character and disposition : while each system of protection has advantages and disadvantages. No general or permanent solution of the problem is possible under the conditions, and it is absurd to claim absolute superiority for any system.

, -: Resuming the historical sketch of the disposition of hull-armour in. war ships, reference must next be made to the Majestic class, laid down in 1894. In these ships the hull-armour was arranged differently from that in the Eoyal Sovereign class. These differences are shown in detail on cross-sections appearing in plates, Nos. 4 and 5. The protective deck towards the sides in the Majestic is turned down to the lower edge of the side-armour, instead of being continued horizontally across to meet the top of the armour belt as in the Eoyal Sovereign class. The thick armour belt is abolished > and axmour of uniform thickness is fitted throughout the length of the citadel up to the height of the main deck. This armour is also wrapped round the bases of the barbettes. This system of protection was proposed by me as an alternative to the Eoyal Sovereign class. = (in 1888), but the Board of Admiralty then preferred to adopt another arrangement. When the Eenown was designed in 1892 I again submitted the modified system, and it was accepted by the Board. The weight of armour in both cases is practically the same: the cost of armour of a given quality is less when the belt is abolished ; the protection to buoyancy and stability, as well as to the vitals, is no doubt much improved. All experience shows that the chance of hitting becomes greater as the height above water increases. It is therefore unreasonable, to mass armour in the form of a thick belt, extending over a very limited vertical space in the immediate region of the water-line. Comparing the Royal Sovereign and the Majestic, on the assumption of still water, it will be seen that, for a height of about three feet above the load-line, and to about one foot below, the Eoyal Sovereign has 18-in. armour, as against 9-in. in the Majestic. The established practice at the time of the Eoyal Sovereign design permitted the thickness of this armour to be rapidly diminished (in the vertical sense) towards the lower edge of the belt from about one foot below the load-line. From three feet above the load-line up to nine and a half feet (i.e., to the height of the main deck) the protection in the Eoyal Sovereign consists of 4-in. steel, with suitable backing and supports; whereas, in the Majestic, the corresponding thickness is nine 1, "\Vith the same quality of armour the resistance to perforation is about five-fold greater in-the Majestic than in the Eoyal Sovereign for this important zone. Against shell-fire the difference is still more in favour of the later arrangement. In addition, it must be noted that by turning down, the protective deck towards the side, as is done in the Majestic, the defence given by the vertical side-armour to the vitals is greatly reinforced. It is also possible, at the discretion of the commanding officer, to retain coal in the angular spaces above the protective deck behind the side armour, and so, with a moderate weight of coal, to increase defence considerably. The ammunition supply to the 6-in. guns has very much greater protection given to it in the Majestic, and the important longitudinal communications on the lower deck within the coal bunkers are very strongly defended, thus supplementing the still more valuable longitudinal ammunition-passages which have been provided below the protective deck. This system of protection has now found almost universal favour in all battleships and armoured cruisers, and the changes that have been introduced during the subsequent twelve years, since its earliest application in the Eenown, chiefly have relation to the protection given to the extremities or to the arrangement of deck-armour.

, -: Although it is a common practice to restrict the use of the term "armour" to thick vertical plating fitted on the sides, batteries, barbettes, casemates or conning-towers of ships, it is necessary to recognise the fact that from a very early period of the ironclad reconstruction great importance has attached to horizontal armour and to protective decks. Beginning with vessels of low freeboard the use of such decks has become general. Italian designers had the courage of their opinions and a quarter of a century ago in the design of the Italia and Lepanto, Signor Erin entirely abandoned side armour, and trusted for the protection of " vitals," buoyancy and stability simply to an under-water armoured deck, with minute subdivision of the water-line region. The great class of "protected" cruisers subsequently constructed also depend entirely for the protection of their vitals and the maintenance of buoyancy and stability upon similar arrangement. In most British ships of that class the central portions of the protective deck rise above water instead of being wholly below it in the Italia class. Strong condemnation lias been expressed of the policy maintained by the Admiralty up to 1895, in continuing the construction of protected cruisers after many foreign Navies had introduced various types of so-called "armoured'* cruisers, with comparatively thin armour plating protecting more or less considerable areas of the sides, or with narrow water-line belts of greater thickness. These adverse criticisms usually ignore important facts. Before coming to the conclusion to build protected cruisers the Admiralty and its technical advisers fully considered what was being done abroad, and, rightly or wrongly, reached the conclusion that the balance of advantage for the cruiser classes was secured by the adoption of protective decksassociated with coal protection, and minute sub-division, and, in the larger classes, with casemate-protection added for the G-in. gunsrather than in the use of thin and soft steel armour plating on the sides, or narrow belts. It might be thought from some utterances on this subject that a magic virtue attached to the presence of vertical armour on the sides of a ship in any form, even if narrowly limited in extent. In tabulations of the fleets of the world it has often been the fashion to include in the " armoured" section, vessels having only small patches of armour in the region of the water-line, which armour was more often under water than above, and rarely constituted an important contribution to protection. As a matter of fact, there is, and can be, no practical difference in regard to protection, and the maintenance of buoyancy and stability, between a narrow-belted ship (in which side armour is associated with a horizontal protective deck) and a vessel of the protected class, in which the deck rises as high as the belt, and its lower edge is as deeply immersed as the belt, while coal protection is arranged at the sides. But there is a sensible saving in weight, with equal protection, and a considerable saving in cost of protective material, when decks are used and belt armour is omitted.

, -: Prior to 1895 there were few vessels which could fairly claim to be described as " armoured " cruisers, in the sense that their sides were effectively protected over considerable areas by strong vertical plating. The Dupuy de Lome of the French Navy was one of the first and most notable examples ; she was followed by smaller French vessels (such as the Bruix) where the protected area was very greatly reduced, and the thickness of the plating somewhat diminished. In all these cases, as has been already stated, the thicknesses and defensive powers of the side-armour were distinctly inadequate. Palliser or other cheap forms of projectiles were most effective against these vessels. High explosives were not required, and true "armour-piercing'* projectiles were unnecessary. Those who have seen the results of artillery experiments against lightly armoured and unarmoured structures, attacked by high-explosive shells- and other varieties of projectiles, will know that, so long as the armour used was not more than 4 in. thickalthough it was of the best quality obtainable up to 1896there was no sufficient reason for preferring such a system of defence to that embodied in the designs of British protected cruisers. My own recommendation during this period was always in favour of the limitation of side-armour to cases where the thickness and quality of armour and the extent of the protected area enabled an adequate defence to be provided against the attack of 6-in. quick-firing guns. In my judgment, the protected cruisers of the British Navy, designed prior to 1896., are in every way qualified to meet, on more than equil terras, the contemporary cruisers built for any foreign Navy and still on service. Their offensive and defensive, qualities were determined upon with the intention of fulfilling this condition ; and the results obtained in the completed ships justify the opinion that the intentions of the design have been fully realised. The relative fighting efficiencies of the great protected cruisers Powerful and Terrible, as compared with the Eurik and Hossia, which they were designed to meet, have been discussed elsewhere. In regard to the vessels of the Diadem class, which have been severely criticised as embodying inadequate offensive and defensive qualities in proportion to their dimensions, it may again bo asserted that, for the services they were designed to fulfil, and when compared with foreign cruisers built or projected at the time of their design (1S94), the Diadem and her sister ships were efficient protectors of commerce and communications. Owing to various circumstances, which had no relation to the design, the last completed Diadems occupied an extraordinarily long period in construction; and some of them were only recently brought into active service. But it is obviously unjust and unreasonable to compare the Diadem class with vessels designed many years after them, in which all subsequent improvements of materials, machinery, and armaments could be introduced. Yet some writers have maintained that there has been deliberate acceptance on the part of the Admiralty of an inferior type, and have left unnoticed the fact that these later foreign vessels have been met and overmatched by later British cruisers.

, -: Italian designers deserve and should receive credit for initiating the type of modern armoured cruiser. Financial limitations had much to do with the decision to build this class, which were much less costly than battleships, but superior in speed, although inferior in the power of the principal armament, and carrying only moderate thicknesses of armour. The Italians extended this moderate protection over considerable areas of the sides, and associated it with central batteries. Their aim was obviousto produce a class of cruisers that could take part also in fleet-actions if required. It has always been a pleasure on my part to acknowledge these and other valuable suggestions, influencing my professional practice, which have arisen from the study of Italian methods. But until the improvements of Harvey and Krupp in the manufacture of armour were brought into practical form, it was not possible, in cruisers of permissible cost and dimensions, to associate high speed and large coal-endurance with such thicknesses of armour over sufficiently large areas as were required to secure a proper standard of protection. Immediately after this condition was fulfilled, by the production of Krupp armour about six inches in thickness capable of resisting the attack of 6-in. quick-firing guns under fighting conditions, the design of the Cressy class was put in hand at the Admiralty, and followed in due course by the Drake class. The County class, as has been previously stated, were introduced subsequently as a reply to the smaller swift armoured cruisers (of 23 knots maximum speed) then building for Foreign Navies, and to the protected cruisers or " Corsairs " of equal speed, which had been constructed or proposed for the avowed purpose of destroying British commerce. The armament of the County class, as well as the protection, was regulated by the standard of power existing in or projected for these foreign vessels, and was decided upon by the Board of Admiralty. It needs no further comment from me. On the side of protection there can be no doubt that, taking into account the protected area and the defensive power of the armour employed, the County class are much superior to the smaller French armoured cruisers, such as the Desaix, and closely approach or equal the Mont calm class, and the Russian Bayan. The County class are immensely superior to recent protected Russian cruisers, such as the Variag, or the French and American protected Corsair cruisers, both in regard to powers of offence and defence, and these are the classes they were specially designed to meet, although their larger dimensions fitted them to deal also with the other classes mentioned. Having given this explanation of the governing conditions of the design for the County class, I may be permitted to add that my personal opinion has always been adverse to the construction of armoured cruisers with less than six inches of side armour, and this opinion was well known to the Board of Admiralty when it was decided to construct the County class. In my designs for later vessels of the class, 7*5-in. guns were introduced; and since my retirement, the design has been further modified in order to provide 6-in. armour, dimensions and cost having been increased while speed has been diminished. The fighting efficiency of these later vessels has thus been increased, partly by sacrificing speed; but it still remains true that the original design provided vessels of ample power for the particular service contemplated, and every intention of the design has been more than fulfilled. The screw propellers tried in the earliest vessels were unsuitable. At small cost, other propeller-blades of greater area have been made and fitted, with the result that the designed speed of 23 knots has been considerably exceeded.

, -: It is noteworthy that, at present, the construction of protected cruisers is limited to vessels of small dimensions, and that there is a free use of armour on the sides and decks of battleships and first-class cruisers. Under these circumstances it is interesting to remember that about 18 years ago, in Parliamentary debates, high authorities expressed the opinion that the days of armoured ships were near an end, and that swift protected ships would be used instead. These utterances occurred almost immediately before the introduction of high explosives, and the larger natures of quick-firing guns, as well as the remarkable series of improvements in armour and projectiles, fundamentally altered the conditions of warship design. He is a bold man who ventures to predict what may happen in this department of construction. " One step " is here " enough " for most persons who have to undertake responsibility. Irresponsible critics manifest no doubt or hesitation in statements of policy or suggestions for programmes of construction. Those who have to take the responsibility for decision in regard to naval affairs, both in this country and abroad, are too well-informed to be influenced by hasty or incomplete reviews of the situation of affairs, and the lines on which advance must be made. The subject of the relative importance of deck protection has been touched upon, but it seems desirable to call attention to some considerations which are hardly appreciated even by writers who have given great attention to this matter, and whose opinions command respect. Amongst these may be included the author of the able articles on " Armour and Ordnance " which have appeared iu. recent issues of the Naval Annual. It is too much the fashion to treat the subject of armour protection from the point of view of smooth-water fighting. Much is made of the fact that, in the Mediterranean, battleships rarely roll heavily; but a small experience of Mediterranean weather suffices to prove that circumstances may occur in that sea producing considerable rolling in the largest battleships. Mail steamers of the largest size often have to reduce speed in the Mediterranean because of heavy weather. Nor must it be overlooked that the British Fleet can never be designed for exclusive service iu the Mediterranean, but must be equally available in the Atlantic, or iu other seas where smooth water is the exception. These are mere truisms, but they need to be repeated, as they are often overlooked. Let it be assumed, however, that a very moderate roll takes place, say only 10 on each eide of the vertical. Further, let it be assumed that, under modern conditions the flight of projectiles is practically horizontal, within the ranges that need to be considered. Let us take a modern battleship of, say, 75 ft. beam and 400 ft. in length. The total deck area in such a ship is over 22,000 square feet. A roll of 10 involves the exposure to attack, by the projectiles of an enemy, of a deck-target which has an obliquity of 10 to the horizontal flight of the projectile, and which on the full width of the ship has a projected height of about 13 ft. If the vessel in question was a British first-class ironclad the vertical height of the side-armour above water -when the ship is. upright and at rest in still water is about 9^ ft. Consequently, under the assumed conditions, the deck target resulting from the moderate roll of 10 exposes to attack a vertical (projected) height of target exceeding by 3^ ft. (nearly 40 per cent.) that of the side armour when the ship is upright.

, -: We have exact information of what such a deck target is capable of bearing, when attacked by ordinary projectiles or by high-explosive shells; and it is known that very serious damage will be inflicted unless adequate strength is given to the deck plating. Large fragments of the deck plating and supports must be blown down by the impact of large projectiles with powerful bursting-charges. It is therefore necessary to do two things. First to make the exposed protective deck at the top of the vertical armour of sufficient strength. Secondly, to reinforce this deck by another protective deck, curved down at the edges to meet the lower edge of armour, and made capable of stopping debris and fragments from the upper protective deck. This idea of a debris deck is a very old one; it originated with the French; but in many cases the debris deck has been either placed too close to the upper protective deck, or else so rigidly connected thereto, that injury to the upper protective deck would almost inevitably have involved serious damage to the lower or debris deck. These were the considerations which led me, from the time of the design of the Canopus. class in 1896 onwards, to recommend to the Admiralty a new-arrangement of protective decks both for battleships and for armoured cruisers. Persons familiar with the designs of battleships are aware of the very considerable weights which have to be assigned to steel plating on protective decks in relation to the weights assigned to hull armour. It will be seen that, if the vertical wall of armour on the side of a battleship is, say, 15 ft. deep and her extreme breadth, is 75 ft., the same weight will be required for a steel deck 1 in. thick as would provide side-armour 5 in. thick. In a modern1 British battleship of the first-class, considerably more than 25 percent, of the total weight of protective material is worked into deck.-armour; whereas in many foreign designs the weight assigned to deck-armour is proportionately much less, and the vertical armour is strengthened by accepting much weaker protective decks. This, in my judgment, is a serious defect in design, and our practice is worthy of continuance, although it involves more weight and larger ships. Tubulated comparisons ignore this fundamental difference.

, -: Moreover, for reasons stated above, it is in the highest degree important that the stronger of the two protective decks should b& situated at the upper edge of the side armour. High explosive common shells with large bursters must obviously burst on impact on the upper deck. The lower protective deck is situated so far below the main deck that there is little risk of serious injury to the lower deck from fragments of plating or projectiles. In our latest armoured vessels the vertical side armour is carried forward to the bow, and the transverse armoured bulkheads at the fore ends of the citadels are abolished. The upper protective deck then becomes of even greater importance to the defence of the bases of the barbettes, and to that of the vitals placed below it in the holdsengines, boilers, magazines, etc. Further, in discussing this matter, it is important to remember that when a vessel is cither chasing or being chased the great longitudinal extent of the deck'target, and the slow change-in relative position of the ships engaged, tend to greatly increase the risk of injury to decks by gun fire. On many previous occasions I have drawn attention to the fact that defence in modern warships is always at a disadvantage as compared with offence. Defence by armour forms an essential part of a design, and is practically fixed and unalterable; whereas the attack can be varied and possesses great flexibility. Given a certain quality and thickness of armour, and assuming an attack by guns whose penetrating power is superior to^ the defensive power of much of the armour, then clearly it is not impossible, and usually it' is not difficult, to so vary the character of projectiles and the weights of bursting charges, so as to produce increased destructive effects. Difficulties have occurred no doubt in the designs of projectiles and fuses with high explosive charges; great improvements have been made also in recent years with armour manufacture; but in the end it is certain that the attack will gain upon the defence, because of its greater possible modification. Artillerists anticipate that large explosive charges will be carried1 through armour, and made to take effect within ships. This outlook furnishes sufficient reason for fitting substantial plating on lower protective decks. But obviously there can be no comparison between the destructive effect possible when projectiles have only passed. -through the light steel sides of a ship, and-the corresponding damages produced by projectiles which have been modified so as to carry smaller charges through armour of considerable thickness. Recent designs of ships, with their more extended armoured areas and stronger protective decks, have well matched the farthest advances ia artillery and explosives. To reduce deck armour is not a change that can be justified. Horizontal armour is now ordinarily fitted on the decks forming crowns to batteries, and is necessary to complete the protection of guns, crews and ammunition. Being placed high above water, it is usually much thinner than the protective plating on main decks; but it is unquestionable that in a ship rolling through very moderate angles there must be great risks of penetration of these crowns to batteries and of serious damage to the secondary armament. One result of the adoption of the battery system is that these crowns have large areas and offer large targets, while injury at any point is likely to produce widespread damage in the enclosed space. The corresponding risks with isolated casemates or turrets are obviously much smaller, aud this feature is not unimportant, although it is of a secondary character in the defence. We are now witnessing the development of another chapter in the never-ending struggle between guns and armour, and in a broad sense are seeing a repetition of history.

, -: .The principal armaments of battleships are now almost universally formed by 12-in. guns. Their secondary armaments are being reinforced by 7^-in., 8-in., 9-2-in. and, in the latest Japanese battleships, 10-in. guns. Improvements in gun mountings and. in mechanical appliances for the operation of the breech, as well as changes in the arrangements for loading heavy guns, have enabled much more rapid rates of fire to be obtained. Protection of the secondary armaments has been strengthened by the use of modern armour of superior quality and greater thickness. These changes have led to the relative degradation of 6-in. quick-firing guns, or to their entire supersession. As one who has witnessed many experiments with, guns and armour, and has carefully studied the results, I would express the opinion that notwithstanding all that has been achieved in the direction of rapid loading for heavier guns, 6-in. guns still retain great value as items in the secondary armaments of battleships and large armoured cruisers. More especially does this appear true when it is remembered that the tendency to_adopt batteries for secondary armaments and to give large horizontal training to guns, inevitably carries with it the necessity for gun ports which necessarily form points of weakness in the defence. The presence of 7-in. or thicker armour between these points of weakness in no sense increases the strength at the gun ports, although it is undoubtedly important as an element in defence. Against batteries, therefore, the 6-in. gun still remains a valuable weapon, and against turrets it may be effective in producing "jamming" or obstruction. Nor should it be forgotten that the 6-in. gun is, in a certain sense, the largest possible " true quick-firer,'* having projectiles which can be "man-handled," and being capable of rapid training and evolution by manual power alone. The 7^-in., 9-2-in., and 10-in. gun all require the provision of mechanical power for working and loading them rapidly. With the utmost care and skill in its device and maintenance, the entire avoidance of accidental injury or breakdown to such gear is impossible. When it occurs manual power can only be applied in a slow and unsatisfactory manner, so that fighting efficiency is impaired. Not unfrequently it. happens too that the defence of guns above G in. in calibre is not reasonably proportioned to their ballistic characteristics. To mount 9-2-in. guns with only 6-in. protection, or to adopt 10-in. guns with the same defence, is not a reasonable arrangement. By general agreement hitherto the thickness of armour-protection given to guns has borne a fair relation to their penetrative power, and sufficient reasons for abandoning this principle have not been shown. In short, the practice of using thin armour for protection to guns of high power and great weight grows out of the desire to keep down the weight of armour, and sacrifices defence to an increased power of offence. Another matter of importance is the desirability of mounting tliese more powerful and much longer guns at considerable heights above-water, in order that their more efficient use in a seaway may be secured. I do not propose to dwell on or illustrate this point, but it must greatly influence designs in the .immediate future, and affect the distribution of armour.

, -: The small depth below water to/w-hifh armour plating is carried in all classes of warships is a-feature which has been so long accepted, that its full bearing upon the; question, bf armour protection is apt to be left unnoted. Taking again_the_caj3e of _a-ty;pieal first-class battleship of 75 feet beam, if :the armour belt extent's six feet below water an angle of inclination of less than 10" in still water would emerge the imarmoured portion of ..the bottom. If the'coals were consumed and other stores or ammunition expended, of course, a less angle of inclination would produce the same result in still water. When a vessel is moving at speed in still water, the wave-profile along her sides, at certain portions of the length, considerably lowers the water leve], and so diminishes the depth of the armour below the surface; ven exposing the unarmoured bottom at certain points. In a seaway a very moderate rise and fall of the wave-surface exposes many portions of the unarmoured bottom; and rolling or pitching of moderate amount may produce greater exposure. Here again there is no difference of opinion as to the facts, but a very frequent forgetfulness of what the facts involve. It is generally assumed that projectiles which do not impinge upon the hull of the ship because their range is a little short can have little or no chance of penetrating below water and reaching unarmoured portions of the hull. Long ago it was proposed to use " flat-headed" projectiles for this purpose, but the objections to that form of head were obvious, and no practical effect has been given to the proposal. As remarked above there is nothing more difficult than to obtain the exact range required in order that a vessel may be struck at, or very close to, the waterline, and this is especially true in the long-distance firing which now finds favour; but it is interesting to note that in the action at Port Arthur the Japanese are said to have made most excellent practice against the liussian ships and to have seriously damaged some of them at or below the water-line. The increasing use of range-finders of a trustworthy character may have had something to do with this result: and it is known that the Japanese were well equipped in this respect; but other means may have been taken in this special case for accurately determining the range against ships which were at anchor and in fixed relation to well-known positions on shore. However this may be, enough has been said to show that it would be folly to regard the unarmoured portions of the bottoms of ships as safe from all attack of gun-fire. At the same time that risk is hardly worth mentioning when compared with the dangers resulting from under-water attack by torpedoes, of which the speed, effective range and maintenance of course have been so remarkably developed in recent years.

, -: There have been many proposals to extend the use of armour plating to the submerged portions of ships, more especially as a protection against torpedoes. Experiments have shown that, if fitted on- the outer bottom, armour is ineffective against the - explosion, of large charges of explosives in contact with, or very near to, the ships. As a result it has been suggested to fit armoured inner bottoms (or wing bulkheads) at a few feet inside the outer skins, which were to be formed as usual of thin steel plating. Sir Edward lieed has patented arrangements of this nature, and Russian designers are reported to have constructed armoured " wing " bulkheads in their latest' battleships. The Borodino class are said to have vertical and longitudinal armour bulkheads 4-in. thick in wake of engines and boilers, extending from the armoured deck down to the turn of the bilge-fabeut twenty-three feet "below water) find placed about' five to six feet from the outer skin, and 4-in. armour on the outer'bottom. The latter - item seems doubtful. In ■ the Cesarevitch it is stated that1 there are armoured wing bulkheads, but ordinary outer skins. This ship is one of those attacked by the Japanese torpedo flotilla at Port Arthur and very seriously damaged. As yet our information is incomplete, but st> far as is known the armour protection to the 'under-water portions of the Cesarevitch have been altogether ineffective, and&he has been put out of action as completely as her companion ship, the Ketvisan, which was constructed in the ordinary1 manner;''Personally, I have been always opposed to this application of under-water armour, chiefly because it appeared probable that,-even with external armour fitted at some distance from the outer bottom^ the effect of the explosion, of a modern torpedo would produce- 3uch serious shocks and jars upon (he armoured inner bottom or bulkhead and its fastenings, that leakage into the interior would result even if the armour itself was not perforated or driven in. So far as can be judged from accounts hitherto available, this form of injury really proved fatal to the Cesarevitch, for it would seem that she was kept afloat for a considerable time by her pumps, and this suggests that the leakage which occurred was through openings having a small aggregate area such as would result from "jarring" of the rivets and fastenings, rather than from any large holes in the water-tight portions of the structure, through which water would pour in large quantities and overpower the pumps. Another serious danger from Lhe use of external armour, supposing that this material did its work and maintained its integrity, ■obviously arises from the " water-logging" of spaces outside the armour protection, but within the ship; which water-logging would inevitably result in serious heeling or possibly in absolute instability. -Calculations made for typical vessels protected in the manner suggested, have shown conclusively that this danger from water-logging is real and great. On the whole, therefore, my opinion remains unchanged, and, in my judgment, the use of under-water armour as a protection against torpedoes is not to be recommended. Torpedo attacks have no doubt been greatly developed in recent years, by improvements in the speed and increase in the charges of locomotive torpedoes, and it may be anticipated that the school of writers who take exception to the construction of large and costly battleships or armoured cruisers will be disposed to point to recent events at Port Arthur as additional proofs of the soundness of their contention that it is unwise to "put too many eggs into one basket." Before the incidents at Port Arthur, the same opinions were expressed, in consequence of the results of the experiments on the Belleisle, which was sunk by a torpedo explosion. To quote words-used by me in my recent address at the Civil Engineers': " History is thus repeating itself, as it is apt to do. It appears to be forgotten that many years ago ships were similarly sunk by torpedoes and the same arguments used. Of course, it is necessary, in view of improvements in torpedoes, carefully to consider how the defence may be strengthened, and this is doubtless being attended to. My own conviction is, however, that these improvements in torpedoes are matched by the advances made in rapidity, range and accuracy in gun-fire, and in projectiles and ' bursters.' So that relatively the gun and the torpedo stand much in the same relation as before. In these and in 11 questions of war-ship design it is necessary to take a broad andl comprehensive view, not to narrow the discussion to a single feature of offence or defence." W. H. White.

, -: word - http://rjw.iatp.org.ua/doc/white-armor.rar -