55? Experiments on the Trimfmtffiotiof Heat through Fluids. 



Soon after pouring In the hot carbonate, I obferved the opaque particles formerly men- 

 tioned, moving pretty rapidly upwards and downwards, indicating Count Rumford's 

 currents. Thefe motions did not depend upon the fides of the veflel, for particles were 

 afcending and defceuding in every pait of the liquid. They afcended with great rapidity 

 till they came within * certain diftance of the furface ; then they descended again as rapidly 

 as they had afcended, till they came within a certain diftance of the mercury ; here they • 

 became ftationary for a moment, and then afcended again with as much velocity as at firft. 

 Sometimes a few particles afcended to the very furface, and then defcended again with very 

 great rapidity. But they conftantly flopped, and began to move upwards when they came 

 to a certain diftance from the mercury ; and they all flopped at the fame diftance from the 

 mercury. At firft this diftance was about half an inch, but it became gradually greater as 

 the procefs of cooling went on, and at laft the particles did not defcend lower than the 

 middle of the liquid. If any particle happened to go beyond this boundary, and to approach 

 a little nearer the mercury, it foon became ftationary, and did not rife any more. The 

 rapidity of the motions of thefe particles gradually diminiftied, as well as the fpace through 

 which they moved : after 40' I could not obferve them moving at all ; though, when I 

 marked the fituation of a particle, and then obferved it fome time after, it became evident 

 that it had changed its place. 



Thefe motions are curious, and give us a good deal of information about the procefs of 

 cooling in the hot liquid. I pretend not to explain their rapidity ; but, waving that difE- 

 culty, it feems pretty clear that the particles near the furface giving out their heat, became 

 fpecifically heavier, and tumbled down towards the mercury. They did not reach the 

 mercury, however, becaufe at fome diftance from it they came to a part of the liquid colder 

 than themfelves. If thefe defcending.particles happened, in'confequence of a momentum 

 accidentally greater than ufual, to penetrate into this cold ftratum, they loft their excefs of 

 caloric before they could rife again, and, of courfe, coruinued ever after ftationary in the 

 cold ftratum. Thefe defcending currents neceflarily produced afcending currents ; but 

 thefe afcending particles could fcarcely reach the furface, unlefs their momentum happened 

 to be unufually great, becaufe the water being hotteft juft below the furface, they foon 

 came to particles Specifically lighter than themfelves. As the cooling advanced, the cold 

 ftratum, of courfe, increafed, and the currents moved more flowly, becaufe the water at 

 the furface, approaching to the temperature of the air^ was cooled more flowIy.^ We fee 

 then, that the defcending currents do not reach the mercury ; confequently, they cannot 

 afFe£t its heating . except indiredly; far lefs can thqy zStQ. the temperature of the 

 thermometer. 



Wifliing^ now, to try a hot liquid of a different kind, I poured lintfeed oil, at 230", over 

 the mercury, till it ftood an inch and a half above its furface. The temperature of the 

 ipercury, and of the external air^ was 41^ ° } that of the oil, as foon as it couJd be afcer- 

 taincd, after pouring it in, was 180**. The refult was follows : 



* EXPERIMENT 



