35

inches long; or provided the circular me-tallic canals, were one-twenty-fifth of aninch in diameter, one and one-seventh of aninch deep, and at least two inches in cir-cumference; or provided the wire gauzehad apertures of only one-one-hundred-and-twentieth of an inch. When 12 metallictubes were employed as the medium ofcommunication, one-seventh of an inch indiameter and an inch long, the explosionwas communicated by them into thebladder. Four glass tubes of the one-six-teenth of an inch in diameter and twoinches long, did not communicate the ex-plosion ; but one of this diameter and lengthproduced the effect. The explosion wasstopped by a single tube one-twenty-eighthof an inch in diameter, when it was threeinches long, but not when it was two incheslong.

The explosion was stopped by the metallicgauze of one-one-hundred-and-twentiethwhen it was placed between the explodingvessel and the bladder, though it did notpresent a surface of more than half a squareinch, and the explosive mixture in thebladder in passing through it to supply the

D 2