probably would have solid timbers and fillings, forming a compact mass, very difficult of penetration, the iron ship having no similar backing to the thin plating. It is therefore easy to see why wooden ships are, as a rule, capable of standing more of the wear and tear incidental to grounding than ordinary iron ships with a single bottom. To attempt to increase the thickness of the bottom plating in order to meet this comparative disadvantage would be wasteful and unwise; the preferable course is to fit an inner skin within the frames, forming a double bottom. Then, if the outer plating is broken through, and the inner still remains intact, no water enters the hold, and no serious damage ensues, as explained at length in the first chapter.

Such a cellular construction of the double bottom has a further advantage well deserving consideration. Thin iron or steel plating, stretching over the spaces between transverse frames, not unfrequently shows signs of bending or "buckling" between these supports when subjected to the upward or sideways pressure of the water; and this effect may be aggravated by the strains due to hogging. By means of longitudinal frames or keelsons running along upon the plating, and attached to it, buckling may be prevented; but when, in addition, an inner bottom is worked, buckling becomes almost impossible. The experiments made before the construction of the tubular railway bridge across the Menai Straits was begun first demonstrated the great advantages obtained by the cellular system applied to wrought-iron structures, especially in those parts subjected to compressive strains. Since then the knowledge of this fact has been made generally useful, both in ship and in bridge construction; but even yet not so fully as it might be, for the use of a double bottom involves some loss of internal cargo-carrying space in a merchant ship, which owners are loth to sacrifice.

When a ship sags, the upper deck and topsides are subject to compressive strains; to meet these, as well as hogging strains, more efficiently, a cellular construction of the deck has in

some few cases been adopted. The Great Eastern is a case in point, to which reference will be made hereafter. Longitudinal supports are not commonly fitted to decks; the wood planks usually assisting to prevent buckling in the iron or steel plating, if any is fitted.

The local strains on the decks of ships constitute another important group. Very heavy weights are placed upon certain parts of the decks, resting only upon a certain number of the deck-beams; and no little care is needed in connecting the beams with the sides of the ship, arranging the pillars beneath them, or taking other means to distribute the load. If the loads to be carried were known, and the kind of pillaring determined, it would be a comparatively easy matter to fix the dimensions of the beams required to support the loads. In practice, however, these conditions are not commonly fulfilled, and the breadth of the ship amidships, or some other dimension, is had recourse to in proportioning the sizes of the beams. Special cases occur, especially in war-ships, where the loads to be carried are excessively great, and their positions can be fixed; as, for example, the turrets of a vessel like the Devastation, or the guns in the battery of a broadside ship. Beams of exceptional strength, or beams spaced more closely than at other places, are often employed in such cases; but even then it is not sufficient to regard the beams as girders supporting certain loads, with the assistance of the pillars. Both beams and pillars, besides meeting these local strains, have to assist in the maintenance of the transverse form of the ship, as will be shown in the next chapter. Sometimes it happens, especially in wake of the machinery or boilers, that it is difficult to fit pillars under some of the beams; but these beams are easily supported by longitudinal girders extending a sufficient distance fore and aft to have their ends upheld by very strong pillars.

Another class of local strains, of special importance in a war-ship, are those brought upon the bows by collision with another vessel. The importance of ram attacks is now so

generally recognised that the great majority of the ironclad ships of all navies have been constructed with bows specially designed for delivering an effective blow upon an enemy without receiving serious damage themselves. Spurbows, protruding forward under water in such a fashion as to be able to strike the comparatively weak bottom below the armour of the ironclad attacked, are those which find most favour. Whatever may be the form of bow adopted, it must be made exceptionally strong if it is to successfully withstand the shocks and strains produced by ramming. These strains may be arranged in three divisions: (1) direct strains, tending to drive the stem and bow bodily backwards into the ship; (2) twisting strains, tending to wrench the bow off when the blow is struck obliquely, or the vessel attacked has motion across the bow of the ram-ship; (3) strains tending to perforate the skin of the ram-bow, resulting from the jagged parts of the hull of the vessel which has been struck pressing upon the ram, while the two vessels are locked together, and while the wrenching just mentioned takes place. Similar strains act upon the bow of any ship which comes into collision with another; and unfortunately there are too numerous instances of the truth of this statement in the records of accidental collisions between vessels of the mercantile marine, or other ships not built for ramming. In fact, it is to these ordinary vessels, and not to ships specially designed for ramming, that one must look for the fullest evidences of the character of the strains incidental to collision. The bows of many ships have actually been crushed in; or the skin has been penetrated; or wrenching strains—as in the ill-fated Amazon, of the Royal Navy-have been so serious in proportion to the strength of the bow as to twist the latter and cause the ship to founder. On the other hand, we have ample evidence that the special arrangements of ram-bows provide satisfactorily against strains which are fatal to weaker bows.

At Lissa, the Austrian ram Ferdinand Max, a wood ship

with a strengthened ram-bow, struck and sank the Re d'Italia, besides making other less successful attacks on other Italian ships; yet her bow sustained no serious damage, although it suffered more than an iron-built ram would have done under similar circumstances. The improvised Confederate ram Merrimac sank the Federal wooden frigate Cumberland at Hampton Roads, but wrenched her own spur badly in consequence of its faulty construction, and is said to have been consequently far less efficient in her subsequent fight with the Monitor. The disastrous collision between the Vanguard and the Iron Duke furnished one of the severest tests yet put upon the strength of the ram-bow in one of the modern types of iron-hulled ironclads. To understand the severity of the test, it is necessary to note a few facts given in evidence before the courtmartial. At the time of the collision the Iron Duke is said to have been going 7 knots, her course being six points off that of the Vanguard; the direct force of the blow delivered was at least 12,000 foot-tons. Fig. 26, page 35, illustrates the damage done to the Vanguard, the armour being driven in bodily and the outer bottom pierced by a huge hole some 20 or 30 square feet in area. Such a blow, of course, reacted on the bow of the Iron Duke,

tending to drive it back

into the ship; and meanwhile the Vanguard had a speed athwart the bow of the Iron Duke of no less than 6 knots, the motion producing a tendency to twist and wrench the bow, as well as to perforate the skin. The simple and comparatively light arrangements of the ram-bow answered admirably when thus severely tested, subsequent examination proving it to be so little damaged that the Iron Duke could, in action, have ventured safely on a repetition of the blow, and yet have remained efficient. Special interest attaching to this matter, Figs. 93-95 have been drawn to illustrate the principal features in the framing of an ironbuilt armoured ram-bow; and only a few explanatory remarks will be required.

The stem is a solid iron forging, weighing several tons.

Against direct strains tending to force it backward, it is supported by the longitudinal frames or breasthooks (1,1,

in Fig. 93), as well as by the armour-plating, backing, and skin-plating, all of which abut against the stem. The breast

FIG 95.

hooks are very valuable supports, being very strong yet light; their construction is shown in Fig. 95; and the foremost ends of the decks are converted into breasthooks