VOL. LXVIII.] PHILOSOPHICAL TRANSACTIONS. 333 



ing, as well as to that of sight. Now before the great explosion was caused, 

 the fluid accumulated in the apparatus must have been diffused equally through 

 it, in consequence of its elastic principle; and, being so circumstanced, a sudden 

 application of the hand, or any other substance, which would open a door for 

 the passage of the fluid into the earth, was found to discharge the greater part 

 of that fluid; and whatever part was so discharged, the most distant particles 

 seemed to have arrived at the point where the explosion took place, at the same 

 time with those that were the nearest to it; because, immediately after the 

 explosion, there was very little of the fluid remaining in the apparatus. If then 

 the discharge of the fluid be, as it seems to be, very nearly instantaneous, the 

 particles of it must move with velocities, and consequently with forces, very 

 nearly proportional to those distances. 



From this consideration I apprehend it will appear, why the sensation on the 

 discharge from the long wire in the 46th experiment was more violent: and on 

 recollecting the 37th, 38th, and 39th experiments, and the observations I made 

 on them, I am inclined to believe that the effect depends more on the length of 

 the metallic body, than on the quantity of its matter or surface. 



It was on the idea of the velocity of the fluid being thus increased, that I 

 apprehended gunpowder might be fired without the least appearance of a spark. 

 The success of this experiment was an inducement to try Kunkell's phosphorus, 

 which was made by Dr. Higgins. The moment this inflammable substance was 

 brought very near the surface of the brass drums, it burst into a blaze; and 

 common tinder, applied in the same manner, was set on fire the instant when it 

 was brought so near as to touch the metal: but there was not the least appear- 

 ance of a spark in any of these experiments. The method taken to fire the 

 gunpowder was this: on a staff" of baked wood a stem of brass was fixed, which 

 terminated in an iron point at the top. This point was put into the end of a 

 small tube of Indian paper, made somewhat in the form of a cartridge, about 

 one inch and a quarter long, and about -^ of an inch in diameter. When this 

 cartridge was filled with common gunpowder, the wire of communication with 

 the well was then fastened to the bottom of the brass stem. Being so circum- 

 stanced, and while the charge in the great cylinder and wire was continually 

 kept up by the motion of the wheel, the top of the cartridge was brought so 

 near to the drums as frequently to touch the metal. In this situation, a small faint 

 luminous stream was observed between the top of the cartridge and the metal 

 drum. Sometimes this stream would set fire to the gunpowder at the instant 

 of the application; at others, it would require half a minute or more before it 

 took effect. But this difference in time might probably arise from some differ- 

 ence in the circumstances, for any the least moisture in the silk lines, the 

 powder, or in the paper itself, was unfavourable to the experiment. This new 



VOL. XIV. Z z 



