ignition from compressed Air. 42? 



thrown from it into the eye. This painful accident once 

 happened to me. 



They who imagine that electricity kindles the spunk, 

 consider these sparks as an incontrovertible proof of their 

 bpinion. J think they are tnistaken in this case; yet I 

 must not conceal a fact conimnnicated to me by Mr. Veau- 

 Delannay, which seems to confirm this opinion, of which 

 he is a partisan. Out of twelve times, when he operated 

 with the instrument without any spunk in it, he saw sparks 

 emitted three times. There are strong reasons, however, 

 for suspecting that electricity is not the cause of the in- 

 flammation here. These I shall give in the sceond part of 

 this paj>er, ci)ncludins; the present with an important ob- 

 servation on the construction of pistons. 



If we could find an elastic substance sitfFciently com- 

 pact to be turned in a lathe, we should have perfect pistons, 

 that would spring and adapt themselves to the inequalities 

 of the tube, without suffering a bubble or air to escape. 

 I have made some with caouichouc, softened before the 

 fire, in order to give it a decree of elaslicity more obedient 

 to the inequalities of the tube. But on attempting to turn 

 it in a lathe, it bent under the tool. Even the edge of a 

 razor wcn'.ld not take hold of it ; so that the piston re- 

 mained uneven and almost ragged, and yielded like soft 

 wax under the fingers. In this imperfect state it so far 

 prevents the air from escaping, that a column of three 

 inches is sutncient to kindle the spunk ; but after a fevp 

 strokes of the piston the heat dilates it to such a degree, 

 that it cannot be moved without considerable force. If a 

 droj) of oil be put on it, it moves easily ; but this soon 

 spoils the instrument; for the oil dissolves the caoutchouc, 

 and forms a varnish, which, as the piston grows hot, makes 

 i)t adhere still more strongly to the sides of the tube. 



Might not these inconveniences be avoided, by arming 

 the piston rod with caoutchouc, and covering this with 

 leather? If this process succeeded, it might be applied with 

 advantage tu all sorts of pumps, 



H. To attain, if possible, a knowledge of the principle 

 of inflammation in the pneumatic tinder-boX, four thing)* 

 are to he considered — the materials of the lube, the maltef 

 Contained in the tube, the maierials of the piston, and the 

 friction. Among the materials of the piston 1 include the 

 grease, with which it is coated, to make it ttiove vaox4 

 easdy, and render it fitter to intercept the passage of the air. 



In examining the cjuestion whether the ipunk be kindled 

 by electricity, I consider, 



J St, That 



