uprights. The arm should be so tightly pressed against the fender, or cushion of straw, attached to the centre of this beam, that it cannot be pulled back the least distance by hand.

If the skein of my largest catapult is fully tightened up by the winches, three strong men are unable to draw the arm back with a rope even an inch from the cross-beam, thought the windlass has to pull it down from six to seven feet when the engine is made ready for action.

When the skein is as tight as it should be, attach the slip-hook to the ring-bolt in the arm and place the stone in the sling suspended from the top of the arm.

The arm can now be drawn down by means of long spanners fitted to the windlass.. Directly the arm is as low as it should be, or as is desired, it should be instantly released by pulling the cord fastened to the lever of the slip hook.

The least delay in doing this, and the resulting continuation of the immense strain on the arm, may cause it to fracture when it would not have otherwise done so.

The plans I have given are those of my largest engine, which, ponderous as it seems - (it weighs two tons) - is, however, less than half the size of the catapult used by the ancients for throwing stones of from forty to fifty pounds in weight.

As the plans are accurately drawn to scale, the engine can easily be reproduces in a smaller size.

An interesting model can be constructed that has an arm 3 feet in length and a skein of cord about 4 inches in diameter. It can be worked by  one man and will throw a stone, the size of an orange, to a range of 300 yards.

The sling, when suspended with a stone in position, should be one third the length of the arm, as shown in fig. 2, p. 13.

If  the sling is shortened, the ball will be thrown at a high elevation. If the sling is lengthened, the ball will travel at a lower angle and with much more velocity.