jjtf Epcperimtntt to /hew the eoftduffing Ponuer of EaSiet. 



EXPERIMENT I. 



^Tkermontter B rofe to 80' during the pouring in of the luater. 



Therm. B. Thern. C. Water 1 inch below the (urfice. 



«S» 67<» 174' 



90 83 163 



931 88 152 



95 



96 91 144 



96 91? ,«33 

 9<J 91I 128 



95 9»l "2 



93 89! 116 



91 88 100 

 90 87 106 

 88 85 I02 

 86 84, . 99 

 85 82 96 

 83i 8i 92 

 82 79 90 

 80 78 88 

 78 76 85 



77 75 83 



76 73 80 



74 72 7<5 



Here the thermometer B rofe In 15 minutes 16°; and in 21' the thermometer C rofe 

 14.75** : ^'^ ^^'' '^""^ '^^ thermometer A was gradually falling, and approaching nearer and 

 nearer the temperature of B, and after 95' the two were precifely of the fame temperature. 

 The maximum of B was 96°, of C 9I.75''. B firft acquired its maximum, then C : B 

 began to fall in 25', C in 28'. Now all thefe circumftances are precifely what ought to 

 happen if mercury be a conductor of caloric. The experiment, therefore, feems to prove 

 that mercury aftually condu£ls caloric : for we fee that caloric pafles downwards through 

 mercury, which could not happen if mercury were a non-conduftor. 



Though this experiment feemed decifive, I wifhed to repeat it in a different manner, in 

 order, if poflible, to determine the internal motions of the hot liquid, that I might 

 afcertain whether or not thefe motions had any influence on the rife of the thermometers. 

 Count Rumford's ingenious method readily prefented itfelf. Accordingly I poured over 

 the mercury a faturated folution of carbonate of pot-afli in water, till it flood \ inch above 

 its furface. Then boiling carbonate of pot-afli was cautioufly poured over all, till it flood 



two 



