only about one-fourth that in the Minotaur-viz. 1 ton per square inch of sectional area-when the same assumption is made as to the non-efficiency of the armour against tensile strains. The neutral axis of the equivalent girder for hogging in the Devastation is found, under these conditions, to be a little below the middle of the depth; so that the maximum compressive strain on the bottom is very nearly equal to the maximum tension on the upper deck, about 1 ton per square inch of sectional area. This again is a remarkable contrast to an ordinary iron merchant ship, where the comparative weakness of the top flange brings the neutral axis of the girder for hogging down to about two-thirds of the total depth from the top; and consequently makes the compressive strain on the bottom not much more than half the tensile strain on the top of the girder.

When sagging moments act upon an ironclad, her armour can assist the structure in resisting compressive strains, and this fact tells greatly in favour of the Devastation and similar vessels, for which the severest bending moments are those experienced when astride a wave hollow. For instance, in the Devastation herself the maximum sagging moment for this position is 40 per cent. greater than the maximum hogging moment corresponding to support on a single crest; yet the armour helps so much against sagging that the maximum tensile strain on the bottom only rises to 11⁄2 ton per square inch of sectional area. The neutral axis of the equivalent girder for sagging is, in this case, situated much higher than that of the girder for hogging, being about one-third of the total depth of the ship below the upper deck; hence the upper deck, under the maximum sagging moment, has to resist the very moderate compressive strain of about & ton per square inch. Against sagging strains the Minotaur type has the help of the armour, and (what is still more important) has to resist a maximum bending moment only equal to about one-half of that for hogging; so that astride the wave hollow there is a very much larger reserve of strength than there is for support upon a single wave crest.

Enough has been said to show that ironclad ships of the Royal Navy, notwithstanding the great weights of armour carried on their sides, and the necessity for complying with the requirements for offensive and defensive purposes, are much stronger longitudinally, in proportion to the strains brought upon them, than are merchant ships of the largest size. Recent types are also seen to gain greatly, as compared with earlier types, in their reserve of longitudinal strength. Both these results must be admitted to be highly satisfactory when the novelty and difficulty of ironclad ship construction are borne in mind. With these remarks we must take leave of this important subject.

The principles which govern the provision of transverse strength admit of being explained much more briefly than do those for longitudinal strength. In nearly all classes, the transverse frames or ribs, the deck-beams, and the planking or plating of the skin and the decks, together with the pillars under the beams, and the beam-knees, &c., connecting the decks with the sides, contribute to the transverse strength. Iron ships have the further advantage of the strength supplied by more or less numerous transverse bulkheads; and so have most composite ships, as well as many wood ships of recent types. It will be convenient, therefore, to arrange the discussion of this branch of the subject under the following heads:-(1) The strength of the transverse frames or ribs; (2) the strength of deck-beams, and their connections with the sides; (3) the strength obtained by pillars; (4) the usefulness of bulkheads in relation to transverse strength.

With each transverse frame, or rib, a portion of the skin, both inside and outside, may be considered to act in resisting changes of transverse form. For example, suppose in Fig. 103 (page 324) the ribs to be spaced 2 feet apart. If two imaginary planes of division are drawn cutting the skin

midway between the frame chosen and the frames adjacent to it on either side, this strip of skin may be regarded as forming an outer flange of a girder, the web and inner flange of which are formed by the frame. The enlarged section, placed a little below the upper deck in Fig. 103, shows the sectional form of this girder. Similarly each deck-beam may be regarded as associated with a strip of the deckplanking or plating; and, taking the beams with the frames to which they are attached, each of the combinations may be regarded as a hoop-shaped girder having in itself considerable strength to resist change of transverse form. Similarly in wood ships each rib and beam may be regarded as associated with the adjacent strips of inner and outer skins. It is unnecessary to say anything further respecting the skins, as considerable attention has already been given to their arrangements in different classes; but it is desirable to note briefly some of the chief differences in the construction of the transverse frames.

:

The ribs of wood ships are necessarily made up of several lengths (or futtocks) which are either bolted and dowelled (as shown in Fig. 102) or else connected to each other in some other way, which leaves adjacent pieces comparatively free to bend inwards or outwards in relation to one another. As a consequence no single rib can be regarded as having much strength in itself against strains tending to change its form the butts of the various futtocks are places of comparative weakness which can scarcely be avoided. The shipbuilder, therefore, has recourse to the plan shown in the inside view, Fig. 102; if any butt is taken in the diagram, it will be seen that between that butt and the next butt, similarly placed, three unbutted timbers intervene ; this is termed a "shift of butts," and the effect is to succour the ribs at the butts by the unbroken strength of adjacent ribs. This object is effected satisfactorily; but the framing must be weaker than it would be if the individual ribs could offer considerable resistance to changes of transverse form. Formerly it was the practice to fit transverse timber riders

within the ribs in order to strengthen the latter, but the practice died out when diagonal riders came into use.

The ribs of ordinary iron and composite ships are much stronger individually than those of wood ships. Fig. 103 explains their construction (see especially the enlarged sections), and it will be noted that each frame is really a Z-shaped girder, the flanged section giving it great strength to resist alterations of form. The angle-irons and plates of which the frame is made up are either obtained in one length or else welded or buttstrapped into the necessary lengths: the whole being so combined that there are no places of weakness corresponding to the butts in the ribs of a wood ship. This superiority shows itself markedly during the process of building a ship, the frame of a wood ship usually being built up piece by piece, whereas the frames and beams of iron ships are very frequently put together before being hoisted into place, and sustain no sensible change of form during that operation. Below the bilges floor-plates are fitted, gradually increasing in depth towards the keel: these floors are of great value in resisting transverse bending strains, as well as forming supports for cargo, &c.

Vessels in which the main frames lie longitudinally usually have their transverse frames spaced much more widely than in iron ships of the ordinary construction. In vessels of the mercantile marine built on the system advocated by Mr. Scott Russell, the only transverse frames-excluding the complete bulkheads-are placed from 12 to 20 feet apart, and formed by plates fitted in between the longitudinals, with stiffening angle-irons on the edges of the plates. These plate-frames are termed " partial bulkheads," resembling the outer rim of a transverse bulkhead of which all the central parts have been cut away. Their principal use is to furnish a series of sections having considerable transverse strength and situated between the complete bulkheads; also to stiffen the longitudinals, and keep them in their proper positions. The Great Eastern has no other transverse frames than such

partial bulkheads; but the existence of an inner skin adds greatly to the transverse strength, this skin forming strong inner flanges to the hoop-shaped girders, of which the outer bottom forms the outer flanges, and the plate-frames the webs. It should be added that in vessels so constructed the longitudinal frames are commonly made very numerous, in order to stiffen the bottom; but even when these frames are spaced only 3 or 4 feet apart, the spaces of bottom plating left without direct support have areas of from 40 to 60 square feet, and hence results an amount of flexibility in the bottom which may become objectionable.

To obviate this objection, and give greater support to the bottom, as well as to increase the transverse strength, the ironclad ships of the Royal Navy, built on the bracket-frame system illustrated by Fig. 104, have the transverse frames 4 feet apart. Most of these frames, within the limits of the double bottom, are formed as in the diagram, plate-brackets being fitted to connect the inner and outer angle-irons with each other and with the two skins; as well as to secure the longitudinals to the skins, and prevent any change of angle. This light and simple arrangement gives considerable transverse strength, but it is reinforced at intervals of about 20 feet by partial bulkheads similar to those used by Mr. Russell, and forming watertight partitions in the doublebottom space. Underneath the engine-room, where considerable strength is required to meet the strains due to the motions of the machinery, instead of bracket-frames, it is usual to fit plate-frames filling the spaces between the longitudinals, and to cut lightening-holes in them. Before and abaft the double bottom also, where there is no inner skin to contribute to the transverse strength, similar lightened plate-frames are fitted.

The bracket-frame system of construction was introduced by Mr. Reed when Chief Constructor of the Navy, and has been generally adopted in the construction of foreign ironclads. It differs from the system used in the Warrior and other early ironclads mainly in the adoption of the complete