82 OS THE POWER OF FLUIDS TO CONDUCT HEAT. 



N. B. This defcent of half a degree was gradual, but did 

 not commence till long after the beginning of the experiment. 

 After this the piece of ice was inclined to one fide, by which 

 nearly one half of it was immerfed in the cooling liquid, and 

 the inclofed bulb of the thermometer was now not more than 

 an inch from the cold mixture. 



Therm, in the 

 ice. 

 28° 

 23 

 Ice along with the 

 therm, flipped down 

 into the cold liquid. 

 The ice now weighed 12§ ounces : the reft had been liqui- 

 fied by the operation of the faline liquor. 



This experiment, I think, decidedly proves that ice is a 

 worfe conductor of heat than water : indeed this is not won- 

 derful ; for it is faid, that ice at a low temperature becomes 

 an electric. 

 Mixture is much It is certainly a remarkable circumftance, but not at all in- 

 fo?? ualifin 81 confiftent with t he known laws of heat, that in a mixture of 

 temperature, hot and cold liquids, the uniform temperature fliould be fo 

 becaufehotand f oon induced by agitation and fo flowly by reft : but when we 

 are brought to- confider, that in the former cafe, hot and cold particles are 

 gether. brought together, and that in the latter there is a feties of 



particles one upon another, gradually riling in temperature, 

 but differing by infenfible degrees, we fhall not wonder at 

 The gradual dif- t j ie fa& St When any one particle of water, or any other bo- 

 perature renders dy, has one above it, warmer by an infenfible degree, and 

 water a bad another below it, colder by an infenfible degree, its power to 

 tranfmit heat muft be very fmall t. Thefe confiderations 

 gave rife to the two following experiments. 



Experiment 10. 

 Exp. io. An A mercurial thermometer was taken, its bulb -,% inch in 

 rial thermometer diameter, and hanging clear of the fcale : it was heated by 

 expofed to air, the flame of a candle to 600°, and then laid upon a table with 

 gave out mo ft 



!fe hi^fem- * Frcm the be ? innin g of the experiment. 



peratures. t Will not this argument apply to folids univerfally ? contrary 



tofaft, N. 



the 



