166 MOTION of Small Lighted WICKS 



this was made to reft immoveably upon the oil, the duft retired 

 in all directions, fo as to leave the fpace, adjacent to the wafer 

 quite free from every particle. But here it was obfervable, 

 that this difperfion of the duft, by the feeming repulfion of the 

 bafe of the lamp, was much more rapid at that fide which lay 

 neareft to the wick than at any other part, and leaft of all fen- 

 fible at the fide diametrically oppofite. 



The circumftances laft mentioned, feem fuificiently to account 

 both for the progrefilve motion of the lamp, and for the general 

 law of this motion, formerly defcribed. For, regarding this dif- 

 perfion of the duft, as yet, only in a general way, and as the ef- 

 fe(5l of fome repulfion between the bafe and the oil contiguous 

 to it, the fadls above mentioned plainly indicate, that, in all 

 cafes, this repulfion is ftrongeft at that part of the bafe neareft 

 the wick or flame : and as adlion and reaction arc equal and con- 

 trary, the lamp muft therefore be impelled, in the diredlion of a 

 line drawn through tlie wick, towards that part of the bafe moft 

 remote from it, and where the rea(5lion is the leaft. 



But in order to obtain a ftill more competent knowledge of 

 the phyfical caufe of thefe motions, it feemed now neceffary to- 

 inquire more particularly into this apparent repulfion between 

 the bafe of the lamp and the furrounding oil, as indicated by 

 the difperfion of the duft, in the manner above defcribed : and 

 here the following confiderations prefented themfelves. 



TfiB oil in the bafon, when of an uniform temperature, has 

 all its parts in a ftate of equilibrium and of reft. When the 

 lamp is lighted, it is evident we hare a very aClive caufe intrcK 

 duced, tending to deftroy that equilibrium. This caufe is the 

 fame, which broods over a fmall portion of the oil, and is fepa- 

 rated from it only by the intervention of a piece of paper or a 

 wafi^er. The oil, in fuch circumftances, in confequence of be- 

 ing violently heated, muft fuddenly increafe in: volume, and 

 muft now, on account of tlie dccreafe of its ipecific gravity, be 



preiTed 



