all angles of fall and will run parallel to the surface like a
locomotive torpedo. Such a projectile has yet to be invented; but I
have seen a linked shell, which has been experimented with from a
nine-inch powder gun, that partially meets this condition. It is made
of several sections united by means of rope or electric wire in
lengths of 100 to 150 feet. When fired all sections remain together
for some distance; the rear section then first begins to separate;
then the next, and so on. It is primarily intended to envelop an
enemy's vessel, and to remedy the present uncertainty of elevation in
a gun mounted in a pitching boat; but it is found that when it strikes
the water in its lengthened out condition, it will neither dive nor
ricochet, but will continue for some distance just under the surface
until all momentum is lost, when it will sink. This projectile is at
present crude, and has never been tried loaded, but it will probably
be developed into something useful in time.

I have confined my remarks in the foregoing discussion principally to
such methods of using high explosives in shells as have proved
themselves successful beyond an experimental degree, and practically
they reduce themselves to two, viz., using a sluggish explosive in
small quantities from an ordinary powder gun, and using any explosive
from a pneumatic or other mechanical gun. Naturally, the success of
the latter method will soon induce the manufacture of powders having
an abnormally low maximum pressure. There is undoubtedly a field for
the use of such powders in connection with an air space in the gun to
still further regulate the pressure; but nothing of this sort has yet
been attempted. Many methods of padding the shell have been devised
for reducing the shock in powder guns, but the variability of the
powder pressure is too great to have yet rendered any such method
successful. A method was patented by Gruson in Germany of filling a