ould begin at Cape Heve (which it would consolidate) and end opposite the
entrance to the port at 1,600 meters from the jetties. Through it there
would be five passages. Afterward another dike would be constructed,
starting from the shore and running to meet the jetty designed to inclose
the Little Roadstead. On turning the angle at which it met the jetty it
would be continued as far as to Berville. Finally, a third dike, running
from Honfleur to Berville, would complete the system.

Mr. Hersent's project, which is one of the most remarkable of those that
have been proposed, has one fault, and that is that it would require
twelve years of work, and cost 158 million francs.

Mr. Thuillard-Froideville, completely renouncing masonry dikes as being
too costly and taking too long to construct, proposes to inclose the Havre
roadstead by means of floating breakwaters. As we have already seen, the
use of these between Cape Heve and the Eclat shoals had already been
proposed in 1845. As the project was abandoned, the models of these
breakwaters are rare.

In Bouniceau's "Marine Constructions" we find a curious figure, a sort of
open framework of clumsy form anchored in a singular manner, and
surmounted by rooms for watchmen, semaphores, posts for the shipwrecked,
etc. It is, indeed, the most complicated and most impracticable type that
could be imagined.

Mr. Lewis' model, which was exhibited last year at the International
Fisheries Exhibition, was, on the contrary, one of the simplest. It
consisted of a strong piece of wood of nearly triangular section, the
sharpest angle of which, being turned oceanward, was designed to cut the
waves and cause them to break over it (Fig. 2). If, by favor of divine
Providence, this breakwater, which presents absolutely plane surfaces to
the shock and pressure of the waves, is not broken to fragments in the
first tempest, it will certainly acquit itself of the _role_ for which the
inventor destined it. When we have a system of resistance to the sea,
anchored and facing a certain direction, and consequently not being able
to revolve around its axis as vessels do, care must be taken not to give
it entire surfaces.

[Illustration: FIG. 4.--FROIDEVILLE'S BREAKWATER.--MODE OF JOINING THE
PARTS.]

Mr. Froideville's breakwater consists of a framework 25 meters in length,
and 9 in height and width, and having the form of an irregular 5-sided
prism (Fig. 3). The smallest side of the prism is design