'' prove tiai Fluids OM Kon-conduSiprS: of HeaU jmr 



till fome part of it be cooled down to 40°. And as water is confeffedly a very bad con- 

 duftor, it is a very confidcrablc tipie before iflucb pf it is cooled down fo low. Ac- 

 cordingly we find that after 20' the boiling water at the diftance of only one inch above the 

 ice was 130°, and of Half an inch 46°. Water at 41°, then, has an advantage over boiling 

 watei, wA, ijiis, ai^^saWage. 16. tjie greater the worfe a conductor of caloric the water is. 



I fhall m?jl be fo particular in my examination of Count Rumford's experiments made in 

 order to prpse that mercury and oil are non-condu6tors of caloric ; becaufe they are not 

 fufceptible of fufHcient accuracy to decide the point. The iron cylinder would retain its 

 heat too fliort a time to melt much ice, even if thefe fluids had been excellent conduftors : 

 and I really do not fee how it was poffible to dete£t a very fmall quantity of water, even if 

 fome of the ic^ had been melted- But if the Count had made ufc of a thermometer inftead 

 of ice, it woifld have rifen faveral degrees, 4£ I have found by making the experiment, 

 even though tbe hot iron had been at a greater diftance from the bulb, than his was from 

 the nipple of ice. 



It appears, then, that Count Rumford's experiments are not decifive, or if any conclufion 

 'can be drawn {rom them, it is rather favourable to the fuppofition that fluids are condudlors 

 of caloric. It occured to me, that by varying Count Rumford's experiments a little, the 

 point in queftion might be decided with certainty. I took a tubulated glafs receiver M, 

 Fig. 2, PI. XXIII. into which there pafled the thermometer B. This thermometer was fo 

 placed, that its bulb was exaftly in the axis of' the veflel. The mouth at which it entered 

 was clofed with cement. The degrees were marked upon the tube itfelf, and were all on 

 that part of it which was without the receiver. Into this veflel I poured the fluid, wliofe 

 conducing power I wifhed to examine, till it ftood at a certain height above the thermo- 

 meter D, fuppofe at D. I then poured cautioufly over it (fuppofe to the height E) another 

 fluid of inferior fpecific gravity, and heated to a certain temperature. Now, if the ther- 

 mometer B did not rife, it would follow that the fluid was a non-condu£tor ; if it did rife, 

 that it was a conduflor. A variety of precautions were neceflary to prevent being mifled ; 

 but thefe will appear from the experiments themfelves, which I fliall now detail. 



The firfl: fluid made choice of was mercury. I filled the veflel M with mercury till it 

 ftood 0.2 inch above the bulb of the thermometer B. Over this was poured cold water, 

 till it ftood \ inch above the furfaee of the mercury. I then cautioufly poured in boiling 

 water till it ftood about two inches above the furfaee of the mercury. The thermometer A 

 was fufpended in the water, fo that its bulb approached very near the furfaee of the mer- 

 cury. There was alfo a third thermometer C, whofe bujb, which was oblong, refted' 

 againft the bottom of the veflel. The temperature of the mercury and of the external air 

 was 42°. That of the hot water, as foon as it could be obferved after pouring itj in (whichr 

 took up fome titne)> was 174°. The refult of the expeiiment was as follows : 



EXPERIMENT 



