a loss, would
follow from giving them much greater beam than had been proposed, and
this was amply verified in the actual ships.

So long ago as the last decade of last century, an extended series of
experiments with variously shaped bodies, ships as well as other
shapes, were conducted by Colonel Beaufoy, in Greenland dock, London,
under the auspices of a society instituted to improve naval
architecture at that time. Robert Fulton, of America, David Napier, of
Glasgow, and other pioneers of the steamship, are related to have
carried out systematic model experiments, although of a rude kind in
modern eyes, before entering on some of their ventures. About 1840 Mr.
John Scott Russell carried on, on behalf of the British Association,
of which he was at that time one of its most distinguished members, an
elaborate series of investigations into the form of least resistance
in vessels. For this purpose he leased the Virginia House and grounds,
a former residence of Rodger Stewart, a famous Greenock shipowner of
the early part of the century, the house being used as offices, while
in the grounds an experimental tank was erected. In it tests were made
of the speed and resistance of the various forms which Mr. Russell's
ingenuity evolved--notably those based on the well-known stream line
theory--as possible types of the steam fleets of the future. All the
data derived from experiment was tabulated, or shown graphically in
the form of diagrams, which, doubtless, proved of great interest to
the _savants_ of the British Association of that day. Mr. Russell
returned to London in 1844, and the investigations were discontinued.

It will thus be seen that model experiments had been made by
investigators long before the time of the late Dr. William Froude, of
Torquay. It was not, however, until this gentleman took the subject of
resistance of vessels in hand that designers were enabled to render
the results from model trials accurately applicable to vessels of full
size. This was principally due to his enunciation and verification by
experiment of what is now known as the "law of comparison," or the law
by which one is enabled to refer accurately the resistance of a model
to one of larger size, or to that of a full sized vessel. In effect,
the law is this--for vessels of the same proportional dimensions, or,
as designers say, of the same lines, there are speeds appropriate to
these vessels, which vary as the square roots of the ratio o