WEIGHT AND SUPPORT IN SHIPS. 
421 
in the curve W W obtained previously. This I have done in the present case for a length 
of about 40 feet in the neighbourhood of one of the armoured bulk-heads, and it has 
been found that the curve of weights between the stations e and /‘in fig. 3 (Plate XVI.) 
assumes the form W' W' W'. The concentrated weights (such as the armoured bulk-head, 
the armoured pilot-tower, the engines, and the mainmast) are in the calculations made 
for W' W' W' put very nearly into the spaces they really occupy, instead of being spread 
over 20-feet spaces as in the curve W W ; and a further approximation to accuracy might 
be made by taking the planes of division still closer together. In practice, however, this 
is quite unnecessary ; and it will be seen from the diagram in fig. 3 that even in this 
exceptional case no very considerable error results from taking the curve W W as a fail- 
representation of the distribution of the weight. This follows from the fact that the 
curve W W averages, so to speak, the inequalities of W' W' \Y', and, on the length of 40 
feet considered, includes a nearly equal area ; that is, represents as nearly as possible the 
same weight. On this account, when considering the strength of a ship as a whole, we 
may avail ourselves safely of the method previously described, taking the planes of divi- 
sion, say, 20 feet apart in large ships; but excessively concentrated weights necessarily 
cause severe local bending- and shearing-strains, which must be specially provided against. 
It will suffice here to say that shipbuilders depend chiefly upon longitudinal bearers or 
keelsons to distribute such concentrated loads over a considerable length, and prevent 
their exercising a prejudicial effect on the structure. 
My fourth example of the distribution of weight and buoyancy in a fully laden ship 
floating in still water is taken from the class of broadside ironclads of which the 
4 Audacious ’ and 4 Invincible ’ were the first examples. The calculations have been con- 
ducted in the same manner as those for the 4 Victoria and Albert,’ and the results are 
represented by the curves shown on fig. 4. It is necessary to remark that in these 
vessels there is an armoured upper-deck battery amidships above the central main-deck 
battery, and that there is neither a bow nor a stern battery protected by armour, as the 
fore-and-aft fire, so essential in ironclads, is obtained with the central battery guns on 
the upper deck. By this means nearly all the principal weights are brought towards 
the middle of the length, and the extremities are much less heavily burdened than in 
most other ironclads, not excluding even ships like the 4 Bellerophon.’ In short, the 
present case is one of a ship with weights very unusually concentrated at the centre. 
The curve II H of weight of hull in this case bears a certain resemblance to that of the 
4 Bellerophon ’ in fig. 3, only the cusps P, P are more pronounced on account of the 
armoured bulk-heads of the upper-deck battery. This fact also accounts in part for the 
difference between the curves of total weight W W for the two ships. It is only neces- 
sary to add that the length of the 4 Audacious ’ is 280 feet, and that the same scales 
have been used for lengths and weights in fig. 4 as in the preceding diagrams. The 
distribution of weight and buoyancy for this ship may be summarized as follows : — 
On the 35 feet of length between the bow and the foremost balanced section (R 1 in fig, 4) 
mdccclxxi. 3 N 
