" 30 " ................................ 274
" 40 " ................................ 369

Regarding the comparative effects of gunpowder and the high
explosives, I think Gen. Abbot's estimate of a varying value for
powder is more admissible than the fixed value assigned by Col.
Bucknill. Gunpowder gives a push and detonating compounds a shock; as
the quantities increase, the push reaches farther than the shock.
According to Gen. Abbot, 100 pounds of dynamite No. 1 will have a
destructive horizontal range of 16.3 feet, while the same amount of
gunpowder will only have a range of 3.3 feet. Five hundred pounds of
dynamite, however, will have a horizontal range of 35 feet, and 500
pounds of gunpowder will have 19.5 feet; the ratio has diminished from
five to two. Whether 6,500 pounds or 12,000 pounds per square inch is
necessary to crush the bottom of an armorclad will depend largely upon
how far apart the frames of the ship are spaced and what other bracing
is supplied, as well as many local circumstances. It is difficult to
judge exactly of these matters. Some four years ago the Italian
government adopted treble bottoms for their heaviest ships as a result
of experiments with seventy-five pounds of gun-cotton (the charge of
an ordinary Whitehead locomotive torpedo) against a caisson which was
a _fac-simile_ of a portion of the proposed ships. Only two of the
bottoms were broken through, and when the space between the two inner
bottoms was filled with coal, only the outer bottom was broken.
According to the formulæ of either Abbot or Bucknill, there should
have been a local pressure of at least 300,000 pounds per square inch
on the outer skin, and yet judicious interior arrangements rendered it
harmless to the target. It would not, however, be safe to conclude
that the torpedo was thus vanquished; the immediate result was simply
to create a demand for larger locomotive torpedoes for local