horse power on the voyage, for a series of voyages,
would be extremely valuable.

He gave, in Table III., the consumption per indicated horse power per
hour for a number of ships. This table affords valuable data, for it
gives, in addition to the dimensions, the moulded draught of water,
the midship area, the displacement, the indicated horse power, the
speed on trial, the coefficients for the lines both from the block or
parallelopipedon, and also from the midship section prism, together
with the length and angle of entrance obtained by Kirk's rule, the
Admiralty displacement coefficient, together with the coal consumption
per day and per indicated horse power per hour.

[Illustration: Fig. 3--OREGON.]

This table, as will be seen, contains some of the most important of
the Atlantic liners, and also a number of other typical ships, which
will add a variety to its interest and a value to it. The coefficient,
which is contained in the thirteenth column of the table, viz.:

Dis 2/3 x speed cubed
-------------------------
I.H.P. x sqrt(entrance.)
---------------
10

generally comes out for ships of similar type more nearly a constant
in the true sense of the word than the corresponding Admiralty
constant. As an example, we have the curves of resistance and horse
power for the City of Rome and the Normandie, a large vessel of 6,000
tons, which the Barrow Company built for the Compagnie Generale
Transatlantique, in which the coefficient of fineness and the form of
the lines pretty closely resemble each other below water; and if we
take from the curves the corresponding speeds and horse powers, and
work out the constants by the two systems, we have at 14 knots the
Admiralty constant for the City of Rome 322.2, and for the Normandie
304.8; and taking for a modified form of constant, the City of Rome
gives 253.7 and the Normandie 251.9, which, as will be seen, are much
closer together. Similarly, at 15 knots the Admiralty constant for the
City of Rome is 310, and for the Normandie 295.2, while a modified
constant comes out for the former at 245, and for the latter 244,
again agreeing almost identically. The same at 16 knots, for the City
of Rome the Admiralty constant comes out 297.6, and for the Normandie
282.8, while a modified constant comes out for the two ships 234.4 and
233.7 respectively, again showing marked agreement. It may be
mentioned that in these two ships