Compofttton of Water. —Ekclrophort, 555 



vfien tliey produce wnter, caloric, and light. But here I muft call in the aid of tlie ima- 

 gination : accordingly I imagine that the gjzes confift of hydrogen and oxygen, which arc 

 ponderable— united to caloric, and perhaps light, which are imponderable ; that thefe pon- 

 derable particles unite with one another, and ^heir caloric and light are fet at liberty. 

 Now here 1 have not any evidence of fenfe ; for 1 cannot perceive, by the fenfes, the exig- 

 ence of the compofition of the gazes juft Hated, nor of their decompofition, and union of 

 their ponderable parts. This being the cafe, other kinds of imponderable matter may 

 unite, or efcape, bcfides caloric and li^ht ; confequently I cannot give the full and com- 

 plete demonltration in thefe inftances. But the' fame objeClions may be made to the pre- 

 tenfions to demoiiltMte fully and completely the compofition of fulphate of foda, fulphate 

 ofpot-a(h, or any double fait, perhaps, whatever; for caloric is feparated, and poinbiy 

 other imponderable matter, when thefe fubftances are compounded ; and caloric, and 

 perhaps other matter, may unite when the fulphuric acid and alkali are difut:ited. 



Hence chemiftry, in its prefent ftate, ought not to pretend to vie with mathematical phi- 

 lofophy in its demonftrations. But it does not appear improbable, that the fame certainty 

 as in mathematics may hereafter be attained in chemiftry. We are encouraged to enter- 

 tain this hope, from finding that the art of obfervation, and the invention of artifices for 

 rendering the properties of matter evident to the fenfes, have been proportionate to the 

 advances of knowledge of fa£ts ; as was predi£led by Chancellor Bacon. 



IV. 



Obfcrvalicns on the Eleclrophore, tending to explain the Means by luhieh the Torpedo and other 

 Filli communicate the EleBric Shock. 



HOSE who are converfant with eledrica! experiments know that the eIe£lrophcre 

 confifts of a flat metallic plate with an infulating handle, and another feparate plate of 

 non-conduding matter, either varnifhed glafs, or fome refinous fubftance, coated with 

 metal beneath, and placed with its uncoated fide uppermoft. When this uncoated face is 

 elcarified by fridion or otherwife, and the plate of metal placed upon it, this laft is found 

 to give a fmall fpark to the finger, which is of the fame nature as the eledricity of the 

 non-conduding furface ; but when the plate is lifted by its infulating handle, it will emit a 

 fpark to the finger which is much ftronger, and of the oppofite kind to that of the non- 

 conduding plate. This phenomenon, which was at firft confidtred as very difficult to be 

 explained, was afterwards found to be produced throughout by the fame energies, what- 

 ever they may be, which govern the eleilric cliarges. For, as I have elfewhere obferved *", 

 tlie charge confifts of two clearicitics, ufually called plus and minus, which compenfate each 

 other ; and alfo of a portion of elcftricity on the infulatcd fide, which in equal charges is 

 greater, the greater the diftance between the two ck-arificd furfaces, and in unequal mo- 

 derate charges is nearly in proportion to the charge itftlf. When the non-conduaing fur- 

 face of the elearophore is rubbed, it acquires both thefe elcaricities, namely, the charge 

 by virtue of the compenfating power of the uninfulatcd coating beneath, and alfo the por- 

 tion of fimple elcftricity requifite to maint.-iin the charge. If the metallic plate be then 



' ritilcfuliliical TrGnfailions, 17S1J, 



Z z 2 placed 



