In our present Al iron vessels, it is evident, according to Lloyd's regulations, that we have only 400 inches of material at the deck to balance 690 inches at the keel, and that if suspended on rocks, in the position already discussed, the ship would inevitably be destroyed with a less weight than she is actually accustomed to carry. In this Lecture I am advocating a principle calculated to provide against such a contingency, viz. that vessels of this description should be constructed with equal sections at the deck and keel, say each about 690 square inches. They would then be equally strong, whether in the position shown in fig. 47, or whether suspended on rocks at each end, in the position shown in fig. 50, with a compressive

strain along the top, and a tensile strain along the keel. In either position there would be a surplus strength of 500 tons to spare as a margin against every contingency, or by whatever forces she might be assailed. Generally, I have contended for equal sections at the top and bottom; but cases may arise where stronger bottoms are necessary, as in screw colliers, which take the ground, but in other cases, the nearer the deck and bottom approach each other in sectional area the better.

It may be said that vessels constructed upon this prinle would be greatly increased in original cost. Το he extent, no doubt, this is true; but the material accu

mulated towards the middle should be progressively reduced towards the stem and stern. Thick plates and large masses of iron are not required at the extremities, if uniformity of strength is to be attained. It is an utter waste of material to introduce it where it is not wanted, and, moreover, where it does not add to the security and stability of the ship. In fact, I would earnestly urge upon the attention of builders, that more care should be exercised in proportioning different parts to the strain they have to bear. I do not mean that the frames and sheathing-plates should be much reduced in size or thickness, but the longitudinal stringers and side-plates may be reduced in thickness to advantage. For example, if we take the deck-stringers and longitudinal stringers, which I recommend in order to give increased security, they should be proportioned as shown in fig. 51. These numbers

approximate to the true proportions for equal strengths; and the only deviation from these ratios should be in some parts of the exterior sheathing (probably at the bows, should the vessel come in contact with floating ice), bulkheads, frames, &c. In this way a great deal of material would be saved which does not add to the strength of the vessel.

Another feature in the construction of iron vessels is the method of forming the riveted joints. I am perfectly cognisant of the custom of double riveting, which is sufficiently strong in the longitudinal, but comparatively weak in the transverse, joints. These latter ought to have long covering plates, and should be chain riveted, as shown in fig. 52. If the joints were made on this principle, it

would add twenty per cent. to the transverse strength of the ship, an important desideratum, and one which would be attained at the expense of a few more rivets and a small increase in the length of the covering plates. The secu

rity of the ship should not be jeopardised for such a consideration; and it is to be hoped that vessels will not in

future be built at a loss of one-third of the longitudinal tenacity as at present, but that the most perfect mode of uniting the plates will be adopted in every part of the

Fig. 54.

construction. If the principle of chain riveting were adopted, we might employ the constant 80, instead of 60, in the formula for calculating the strength, and the weight and sectional area of the plates might then be reduced. This is a consideration of much importance, as the frames, or ribs, might be made two feet six inches apart to admit the necessary covering plates, and secure a better system of transverse joints, as shown in figs. 52 and 53. Below water, where a water-tight joint is the first requisite, the rivets must be placed nearer together at the joint; but we may still compensate for the loss of strength caused by the closeness of the rivets, by using a covering plate thicker than the plate, and countersinking the rivets on the outside, as shown in fig. 53. It is true that builders are not allowed, by Lloyd's rules, to place the frames more than eighteen inches apart, which only permits double

riveting; but Mr. Vernon, whom I consulted on these points, and to whom I am indebted for many useful suggestions, informed me that, even with frames at this distance, each alternate tier of plates might be so riveted; th

being arranged outside and inside, as shown

wider covering plates might be introduced, as shown at a a, under the angle iron of the frames, and thus the outside plates chain riveted. This would be a great improvement, and add considerably to the strength of the ship. As respects the diameter of the rivets and their distances apart, the following table, deduced from experiment, and employed by myself and and others in extensive practice, may be relied upon.

Table exhibiting the best proportions for riveted joints.

Thickness of Diameter of Length of ri-Distance of ri-Quantity of

plates in inches.

rivets in

inches.

vets from cen- lap in single

tre to centre

in inches.

Quantity of lap in doubleriveted joints in inches.

vets from the

head in inches.

joints in
inches.

I now venture to direct attention to the plan which I propose should be adopted for securing the most effective distribution of the material which is to be added to the upper part of the ship. Iron vessels are ordinarily constructed with ribs or frames, placed from fifteen to eighteen inches apart. They are about two feet deep at the keel, and taper to the width of the angle-iron round the bilge on each side, at a, fig. 48. From that point to the top of the deck, the angle-iron is in some cases considered of sufficient strength for the reception of the sheathingplates. On the top side of the ribs a lighter description of angle-iron is riveted, and to this the flooring, whether of wood or iron, is attached. This plan of construction is not objectionable, provided two more longitudinal stringers,