Account of a felf-tnov'ing Ldmfi. 1.7 1 



Vphilft the floating niotcf, or particles of i<i{^^ keep nearly their relative places ; siid the 

 fame thing happens, though the point of the iron touches the water fo as to make it fimmer, 



I do not well know how to account for thif, unlefs it may be a confeqiience of the known 

 much k'fs expanfibility of water by heat, compared to that of the inflamnnabJe fluids, and 

 which may be fo inconfiderable as not to deftroy the equilibrium, fo f.ir as to produce aa 

 ciflux from the lighter and expanded fluid immediately under the jjeatcd body. Pofiibly too 

 the parts of the water, as foon as heated, may tranfmit the furplus temperature to the con- 

 tiguous colder water, much more rapidly than the inffamifiable flufd?' do in like circum- 

 ftances, and thereby refift the high temperature neceflary to that degree of expanfion 

 which would difturb the equilibrium and produce an efflux ; not to mention that the max- 

 imum of this temperature can never, at any rate, exceed 212 degrees, the boiling point of 

 water. 



That the equilibrium, however, atnongfl the parts of water is difturbed by the local ap- 

 plication of heat, though in a much fmaller degree than what obtains among the inflam- 

 mable fluids, appears from an experiment I was led to make with a fmall thin cup fwim- 

 ming on water, and fo contrived as to carry and feed with oil a wick placed a little way 

 down from the lip in the infide, fo as to be on a level with the water. The confequence of 

 this conftruftion was, that the cup moved upon the water very flowly.but always with the 

 flame evidently fl:crnmoft. The fame cup, when taken from the water and put into a bafon 

 of ftrong rum, failed a great deal faflrer, and according to the fame ufual law. 



I am much afraid that, by this time, I have wearied you by fuch a detail of minute facis 

 and circumftances, and by thofe frequent repetitions which every new fubjed more or lefs 

 requires.— And I ever remain, dear Sir, your moft obedient faithful fervant, 



PAT. WILSON, 



P. S. Should you be inclined to repeat any of the experiments, the following diredions 

 and mifcellaneous obfervations may be attended to : The thread I made ufe of for the wicks 

 was of that foftkind commonly employed in the flowering of muflin. After making the punc- 

 ture in the bafe, you put through a bit of the thread, which clip fliort off below, and with a 

 pin force in the burr gently round the thread, to give the bafe a proper hold of it, Thea 

 clip away the fuperfluous thread above, leaving the wick about a quarter of an inch long ; 

 and fo the lamp is comp'eted. Set it then upon the oil by taking hold of the wick, that the 

 paper bafe may not be bent or diftorted by handling it ; and, after the wick is touched with 

 a drop of oil, it is ready for being lighted. For this purpofe a bit of packthread, which has 

 been fteeped in oil, is a cleanly and convenient match, and fheds no impurities on the oil, 

 as a candle or wax taper would do. 



When you want the lamp to circulate, the oil muft be very pure, and brought into full 

 contadt with the fides of the glafs. The oil and the bafon, or falver, fbould all be allowed 

 to come to the fame temperature, between 55 and 60 degrees of Fahrenheit. For, if any 

 part of the brim be much hotter than the refl, the lamp, on arriving there, will leave the 

 fide, by the current iffuing from the heated part forcing it away. 



Sometimes the lamp, when failing, veers a little into a different diredVion, by the bafe 

 altering or warping by the fcorching heat of the flame, which determines the ftream to 

 flow out nioft copioufly at a different part of the bafe. 



Z2 In 



