The frictional resistance of screw propellers is always a fruitful source
of inefficiency. With a given screw, the loss due to friction may be taken
to vary approximately as the square of the speed. This is not to say that
the frictional resistance is greater in proportion to the thrust at high
than at low speeds. The blades of screws for any speed should be as smooth
and clean as possible, but for high speed screws the absolute saving of
friction may be considerable with an improvement of the surface. There is
no permanent advantage in polishing the blades. No doubt there is some
advantage for a little time, and, probably, better results may thereby be
secured on trial, but the blades soon become rough, and shell fish and weed
appear to grow as rapidly on recently polished blades as on an ordinary
surface. These screws are of gun metal. They were fitted to the ships in
the condition in which they left the foundry. It appears that within
certain limits mere shape of blade does not affect the efficiency of the
screw, but, with a given number of blades and a given disk, the possible
variations in the form or distribution of a given area are such that
different results may be realized. The shapes of the blades of these
propellers are shown in Figs. 2, 3, and 4. It will be seen the shapes are
not exactly the same for all the screws, but the differences do not call
for much remark.

[Illustration: FIG. 2., FIG. 3. & FIG. 4.]

Fig. 2 shows the blades for the A screw. C and D have the same form. Fig. 3
shows in full lines the blades of the B screw, and, though very narrow at
the tips, they, like A, are after the Griffith pattern. The blades of E and
F are of a similar shape, as shown in Fig. 4, and approach an oval form
rather than the Griffith pattern. The particulars of these propellers would
be considered incomplete without some reference to their positions with