420 
ME. E. J. REED ON THE UNEQUAL DISTRIBUTION OF 
while the heavy weights of machinery, coals, powder, water, &c. are also more concen- 
trated than they would be in a long ship like the ‘ Minotaur.’ On the length of 50 feet 
between the water-borne sections R 1 R 1 and R 2 R 2 the buoyancy exceeds the weight by about 
105 tons ; and between R 3 R 3 and R 4 R 4 there is an excess of buoyancy of 410 tons on a 
length of 90 feet. There is, it will be observed, a certain amount of resemblance between 
the distribution of the weight and buoyancy in the ‘ Victoria and Albert ’ and the ‘ Belle- 
rophon,’ and in both these vessels the conditions are very different from those of the 
‘ Minotaur.’ 
A few words are needed to explain the peculiar form of the curve H H in fig. 3 of the 
‘ Bellerophon’s ’ weight of hull. In both the preceding cases this curve is very fiat, and 
throughout the length is concave to the base-line A B ; while in the ‘ Bellerophon ’ it is 
comparatively irregular, having cusps at the ordinates P P, P P. This difference is caused 
by the disposition of the armour in the ‘ Bellerophon.’ Her central battery is shut in 
by transverse armoured bulk-heads, the weights of which are included in those of the 
divisions of the ship represensed by the ordinates P P, P P ; hence it follows that there 
should be cusps in the curve II H at those ordinates. Between those ordinates the curve 
is nearly straight, as the armour on the ship’s sides is of about uniform weight for the 
length of the battery, and the weight of hull per foot of length is also nearly uniform. 
Before and abaft the central battery the armour-belt only reaches to the main deck 
(except in wake of the armoured bow-battery), and its weight per foot of length of the 
ship is considerably less than it is in wake of the central battery ; this reduction, in com- 
bination with the gradual fineing of the vessel, leads to the form of curve shown in the 
diagram. 
It will be remembered that I have already intimated that the graphical method of 
representation adopted in this paper is not minutely accurate; and the case just consi- 
dered being an extreme one, affords an excellent opportunity for illustrating this fact, 
and showing the extent of the error thus introduced. From the description of the 
method given above, it will be obvious that we have assumed spaces bounded by curves, 
which are practically continuous, to be graphic representations of the distribution of the 
weight and buoyancy. The assumption is fairly accurate for most of the divisions, 
especially in the case of the curve of buoyancy ; but it is not nearly exact for the curve 
of weight in some parts of a ship, particularly where there are any great weights con- 
centrated in a very small length. The transverse armoured bulk-heads enclosing the 
central battery of the ‘ Bellerophon ’ are each really concentrated in a space, measured 
fore and aft, of less than 2 feet ; and it is obvious that, by spreading the weight over the 
20-feet division in which such a bulk-head comes, we materially modify the distribution, 
and obtain an inaccurate curve of weight, even when we make some allowance for the 
bulk-heads, as is done in the cusps P, P in fig. 3. If we place the planes of division 18 
inches apart instead of 20 feet, and calculate carefully the weights of the 18-inch lengths, 
we shall, of course, obtain a curve of weight much more accurately representing the 
actual distribution in the ship, and shall then be able to judge of the amount of error 
