1821.] an Apparatus for discharging Ordnance, ^;, 



present, therefore, it is only in my power to notice a few of th« 

 experiments in a detached form. 



The first experiments were performed by nring the new codv- 

 position, using about one grain, or rather less, at each trial, 

 through a piece of cartridge flannel tied over the hole at the 

 bottom, B, of the apparatus, when it inflamed a quantity of gun- 

 powder fixed in a tin case below the flannel. This was repeated 

 for many successive times without cleaning the apparatus, and 

 the flame never failed to pierce the flannel and fire the gun- 

 powder. 



Should this succeed as regularly, when apphed to the gun 

 itself, there could remain no doubt but that it w^ould possess all 

 the proposed advantages. There was, therefore, fixedto a six- 

 pounder, an apparatus similar to the one already described, 

 except that it wanted the long tube, A B, for which the priming 

 hole of the gun now became a substitute. It was charged with 

 cartridge, and in several of the trials with ball and cartridge ; 

 and, upon the same experiments being repeated with it, it gave 

 the same uniform results. 



The next experiments were with the view of ascertaining how 

 the results stood related to Sir Humphry Davy's theory regard- 

 ing the impervious nature of wire gauze to flame. Wire gauze, 

 of different degrees of fineness, was, in successive trials, put ia 

 the interior of the joinings, a, 6, or c, of the tube, so as to cover 

 the hole completely. When the coarser wire gauze was em- 

 ployed, the flame was found to pass through, and fire the gun- 

 powder ; but the same result never took place with wire gauze as 

 fine as that used in Sir Humphry Davy's safety lamp, unless 

 when the flame seemed to have burst a passage through the 

 gauze. But when these experiments were performed without 

 the flannel and gunpowder at the bottom, B, it was found that 

 the flame went through even three pieces of the wire gauze at 

 once. 



The next experiments, and probably the most surprising of the 

 whole, were with gunpowder placed in one of the divisions, 

 a, by or c, of the apparatus. In some of the trials, I found that 

 the flame had passed through the gunpowder at a, 6, or c, with- 

 out inflaming it, although at other times I found it did not 

 do so. 



This at first appeared to be an objection to the proposed 

 application of the apparatus. But, after repeated trials, I found 

 that the above curious result only took place when the stroke 

 with the hammer was slight ; for when a smart blow was given, 

 inflammation always took place. 



In a few of the experiments, I put gunpowder at two divisions 

 {a and Z>), and found that the flame sometimes went through 

 both, without firing either portion ; at other times one portioiik 

 was inflamed, and one left unaltered. 



In performing these experiments, I first put a small piece of 



