43S igiullon from compressed Air. 



1st, That no part of the instrument is insulated ; ana 

 that insulation is a necessary condition for producing sen- 

 sible electricity with anv of the machines we know, I say 

 machines ihiit we know, because the animal electricity, 

 that manifests itself without insulation, is an exception to 

 our mechanical means, and cannot here be taken into con- 

 sideration. 



2d!y, The friction of the piston, which is a greasy bodv, 

 against a metallic substance, is not calculated to produce 

 elcctricitv. 



3dly, Experience demonstrates, that, unless during 

 storms, the atmosphere seldom exhibits anv signs of elec- 

 tricity at the height wc breathe it; and that we must 

 search foj- them willi instruments in a more elevated re- 

 gion, or when electric clouds are passing over our heads. 

 jHow then shall we estimate tlie infinitely small quantilv 

 of electric matter in a cubic inch of air, or even less, which 

 the instrument contains ? 



4thly, It is not wiihout great difficuliv that we can kin- 

 dle spunk with strong electric sparks. 1 have discharged 

 a large jar on spunk strewed with powdered resin, and it 

 has remained unkindled, though the resin caught fire, and 

 burned entirely away. 



As long as the instrument was fnade with metallic sub- 

 stances only, we were obliged to confine ourselves to the 

 exterior marks of inflamuialion alone, wiihout being able 

 to assign the true cause, <ir at least furnish proofs of it. 

 For to guess, is not sufficient in natural philosophy; we 

 must demonstrate, in order to give to facts that degree of 

 certuinty which befits science ; and this we cannot do 

 here, with(mt seeing uhat passes at the very point of in- 

 il am mat ion. 



The means are verv simple. Nothine is necessary but 

 to substitute a glass for a metal tube. Those foimd in the 

 thops being too slight, 1 applied to Mr. Laurent, the in- 

 ventor of glass flutes, requesting him to j)rocure me tubes 

 of a similar quality. This artist, as much distinguished 

 by his civililv as by his talents, furnished me with three, 

 which I fitted up. 'I'he first, eight inches long by eight 

 lines in diameter, did not kindle the spunk. The second, 

 nine inches long by si.x hues and three-quarters in diame- 

 ter, kindkd it completely. This being destroyed by acci- 

 dent, J tried the third, eight inches long by seven lines in 

 diameter, which succeeded equally well. 



When the instrument is made to act, and the spunk 

 kindles, we see a brit;ht flash, that fills llie capacity of vlite 



tube ; 



