platform was constructed, four feet broad, and about eighteen feet long,
which was strongly fastened with iron. The entrance was lengthways, and it
could be moved about the piece of timber, first described, as on a spindle,
and could be hoisted within six feet of the top. Round this there was a
parapet, knee high, which was defended with upright bars of iron, sharpened
at the end. Towards the top there was a ring, through which a rope was
fastened, by means of which they could raise and lower the engine at
pleasure. With this machine they attacked the enemy's vessels, sometimes on
their bow, and sometimes on their broadside. When they had grappled the
enemy with these iron spikes, if the ships happened to swing broadside to
broadside, then the Romans boarded them from all parts; but when they were
obliged to grapple them on the bow, they entered two and two, by the help
of this engine, the foremost defending the forepart, and those who followed
the flanks, keeping the boss of their bucklers level with the top of the
parapet."

From this description of the corvus, it is evident that it had two distinct
uses to serve: in the first place, to lay hold of and entangle the enemy's
ships; and, secondly, after it had accomplished this object, it served as a
means of entering the enemy's vessels, and also as a protection while the
boarding was taking place. With respect to the question, whether the
_harpagones_ or manus ferrææ; were the same with the _corvi_, it appears to
us that the former were of much older invention, as they certainly were
much more simple in their construction; and that, probably, the engineer
who invented the corvi, borrowed his idea of them from the harpagones, and
in fact incorporated the two machines in one engine. The harpagones were
undoubtedly grappling irons, but of such light construction that they could
be thrown by manual force; but they were of no other service; whereas the
corvi were worked by machinery, and served, as we have shown, not only to
grapple, but to assist and protect the boarders. We have been thus