built on the longitudinal cellular system, the first application, he
believed, of this system to an Atlantic liner. The plating of the
Servia is of the usual alternate outer and inner strake system, partly
double; while the third section, the Oregon, approaches more nearly to
the ordinary system of framing and plating usually adopted, but it
will be seen that she was well tied in the bottom by very heavy
intercostal and plate keelsons, as well as in the top by heavy
stringers and sheer strakes, with much of her plating doubled, and
heavy web frames inside. The author next considered the question of
stability, and went on to deal with the subject of twin screws, and
stated that the Barrow Shipbuilding Company has done more in the way
of planning and designing for the adoption of twin screws lately than
for any other mode of propulsion, and this chiefly for passenger
steamers. He did not attach much importance to the particular form of
the blade either in single or twin screws, as he believed so long as
the disk area, the surface, and pitch were properly adjusted to the
speed of the vessel, and to enable the engines to use, at the maximum
speed, just the full quantity of steam that the boilers can make, we
have got pretty nearly as far as we can get. To fix these dimensions
of the propeller accurately at the present time, and without further
knowledge of the action of the screw on the water, was, he thought,
impossible. All the rules and formulæ are empirical. The best one he
knew is given in Table IV., due to Mr. Thom, the head of the Barrow
Company's engineering drawing office, and at present acting manager,
who has used it for some years in practice. These formulæ are based
upon the assumption that the area of propeller disk should be
proportional to the indicated horse power, divided by the cube of the
speed, and the same with the projected area of the propeller and also
the surface.