The Great Eastern is the most notable example, and her structural arrangements, due to the joint labours of the late Mr. I. K. Brunel and Mr. Scott Russell, furnish good evidence of the superiority of the longitudinal system.* Other and much smaller merchant ships have been built on very similar principles; and in all the iron-built ironclads of the Royal Navy great prominence is given to longitudinal framing. Such framing is of the greatest advantage in the lower parts of ships lying below the lower deck. The comparatively flat surfaces of the bottom plating below the bilge are best stiffened against buckling by longitudinal frames, which form strong girders well secured to the bottom plating, and contribute a very substantial addition to the lower flange of the equivalent girder for the upright position. At the bilge there is usually considerable transverse curvature in the bottom plating, a fact which gives it great stiffness in itself against buckling under compressive strains, due either to hogging moments or to the concentration of surplus buoyancy; hence immediately at the bilge longitudinal frames are not so much required for the purpose of preventing buckling. Very frequently external bilgekeels are fitted just at this part of the bottom, forming good stiffeners to the plating, besides adding their own sectional areas to the lower flange of the girder. Above the bilge, and below the lower deck, longitudinal frames are again of great use, especially in adding to the longitudinal strength when the ship occupies an inclined position, and is subject to hogging or sagging moments. When we reach the parts lying above the lower deck, other considerations enter and make the longitudinals of less importance; in fact, the decks themselves with their stringers, &c. form most

* For much interesting information concerning the construction of this ship, and her predecessors, the Great Western and Great Britain, see the life of Mr. Brunel, published by his son. It is evident from the

details therein given that, at a very early period after the introduction of iron ships, Mr. Brunel perceived the great advantages attaching to longitudinal framing.

efficient longitudinal stiffeners, and they are usually so close together as to render intermediate longitudinals unnecessary. Sometimes, where a lower deck does not extend throughout the whole length, but is broken for some reason, its stringer plate is continued in order to form a stiffener, as shown by 7, Fig. 103. It may, however, be regarded as the rule that the decks need no aid of this kind, and that the only framing required in the upper parts of ships is vertical and transverse. Such framing stiffens most efficiently the almost upright side plating, gives facilities for attaching the beams to the side, and answers other purposes. The extent to which it is adopted must of course depend upon the special conditions of each class of ship. Widely spaced vertical frames suffice in the upper parts of the Great Eastern; whereas in armoured ships these frames are very closely spaced, in order to assist in strengthening the target formed by the armoured side. Fig. 104 illustrates the last mentioned case; below the armour, the main frames are longitudinal, as shown; but behind the armour the principal frames are vertical, being spaced only 2 feet apart (see the section at cd). The longitudinal girders worked between the strakes of the wood backing are not fitted primarily with a view to increase the longitudinal strength of the structure, although they have this effect, but are intended to increase the resistance of the target formed by the side of the ship against penetration or damage by projectiles.

Looking a little more closely into the arrangements illustrated in Fig. 104, it will be evident that the lower flange of its equivalent girder must be much stronger than that of the ordinary iron ship illustrated by Fig. 103. The longitudinal frames of the ironclad are numerous and strong, as compared with the longitudinal strengthenings of the merchant ship. These frames, as already explained, are of great value in preventing buckling, and resisting the tensile strains due to sagging, even when there is only a single outer skin. But their efficiency in these respects

and the strength of the lower flange of the girder are both very greatly increased by the adoption of the inner skin plating, forming a double bottom. This cellular construction is shown by experiment to develop most efficiently the

strength of a structure formed of wrought-iron plates and bars, any one of which, taken singly, has little strength to resist bending. It is unnecessary to repeat what has already been said respecting the gain in safety due to the use of double bottoms, this being so great that, even if there were

no gain in structural strength, the shipbuilder would be fully justified in adopting the arrangement.

Cellular double bottoms require the sacrifice of some of the hold space, and this is an objection, from a commercial point of view, to their use in merchant ships, where the cargo-carrying capacity is of very considerable importance. Partial double bottoms, extending from the bilges to the keel, are, however, frequently fitted, chiefly for the purpose of holding water ballast, but the inner skins and longitudinal keelsons often fitted in these tanks add considerably to the longitudinal strength. Some of the earlier ironclads of the Royal Navy are similarly circumstanced, having only partial double bottoms; but the general practice has been for many years to fit complete double bottoms, as in Fig. 104. It has already been explained that the double bottom does not extend throughout the whole length of the ships, but usually leaves about one-sixth of the length at either end destitute of this strengthener and protection. While the safety of the ship at the extremities is provided for by the other means previously described, her strength at any crosssection outside the double bottom is much lessened by the absence of the inner skin. It is, however, to be observed that the strains to which these parts are subjected are much less severe than those borne by sections lying farther from the extremities, so that in proportion to these strains the strength is ample.

Thirdly, attention must be directed to the webs or vertical portions of the equivalent girders for different classes of ships.

In ordinary wood ships the outside planking, as well as that inside, is worked (as shown in Figs. 100 and 102) in one thickness, and made up of comparatively narrow planks, or "strakes," the butts and edge-seams of which are caulked. This planking, with the shelf-pieces under the

* See Figs. 18-25, page 30.

beams, and the diagonal strengtheners, form the web of the girder. The ultimate strength of these parts against crossbreaking strains is no doubt ample in all or nearly all cases; and what has to be regarded is rather their strength to resist the racking strains which always accompany bending.

Reverting to the case of the beam in Fig. 101, it will be seen that, although the total of the tensile forces experienced by any cross-section equals the total of the compressive forces, these two resultants act in opposite directions, and therefore tend to rack or distort the beam, this racking strain reaching its maximum at the neutral surface, and gradually decreasing to nothing at the top and bottom of the beam. So long as the beam is in one piece, or so long as the pieces forming its web are well connected together edgewise, there is no difficulty in meeting this racking strain. But if a beam were constructed of which the web consisted of strakes or narrow planks placed edge on edge, and having little connection edgewise, then obviously, as the beam bent, these planks would be made to slide upon one another by the racking strains. And if these strakes were crossed at right angles by ties, corresponding to the ribs or timbers of a wood ship, these ties would add little to the strength of the web against racking. For (to quote the well-known illustration of Sir Robert Seppings), if a field-gate be made of pieces, all lying parallel or at right angles to one another, its resistance to distortion of form will be very small. On the contrary, if the strakes forming the web are crossed by diagonal ties— corresponding to the cross-bar of the gate-there will be a great addition to the strength of the combination against racking and distortion of form.

Such are the simple principles upon which the use of diagonal "riders" or ties in wood ships is principally based. The side planking above the bilge has in itself little strength to resist racking strains; and in many cases these strains have been so severe as to show marked evidence of their