PASSAGE OF HEAT THROUGH FLUIDS. 355 



Bat if a stratum of water, or of any other liquid, be substi and not 



tuted in place of the metallic plate, or of the board, the result through water, 

 will be very difierent. If, for instance, the cold cube being 

 placed in a large tub, restin;.'; on the middle of its bottom, the 

 hot cube be suspended over it by cords, or in any other manner 

 so that the lower surface of the hot cube be immediately above 

 the upper surface of the cold cube, at the distance of a quarter 

 of an inch, and the tub be then filled with water at the same 

 temperature as that of the cold cube, the heat will not de- 

 scend from the hot cube to the cold one through the stratum 

 of water of a quarter of an inch in thickness that separates 

 them. 



We may with propriety call silver a good conductor of heat, Silver is a good 

 and dry wood a. bad conductor; but what shall '.ve say of ^^',J^^j"^ i,^^ 

 water ? I have called it a non-conductor, for want of a more one ; and 

 suitable terra; but I always felt that that word expresses but ^^^'^.^^'^ ^""'^* 

 imperfectly the quality that was m.eant to be designated. 



In the experiment of the two cubes plunged in water, if the "Water may 

 hot cube be placed below and the cold cube above it, the heat hv'can-eats! 

 will not ouiy be communicated from the hot to the cold cube, 

 but it will pass even more rapidly than when the two cubes are 

 separated by a plate of silver. But in this case it is evi- 

 dent that the heat is transported by the ascending currents, 

 which are formed in the liquid in consequence of the heat 

 which it receives from the hot body. 



The existence of these currents, in certain cases, has been Direct conduct 

 known along time; but philosophers have not been sufficiently ^..^^^^ ex-*^'^ 

 attentive to the many curious phenomena that depend upon tremely slow, 

 them. It has not even been suspected with what extreme 

 slowness heat passes in fluids, from particle to particle, de 

 proche en proche, in cases where the effects of such communi- 

 cation become sensible. 



For some time after I had engaged in this interesting in- Liquids pro])a- 



quiry, I conceived that this kind of communication was abso- '^^^ ci^nduci 

 . . . . . oulv when 



hitely impossible in all cases ; but a more attentive exaraina- their pr.itioiPi 



tion of the phenomena has convinced me that this conclusion 



*■ to inj 



was too hasty. As early as the beginning of ISOG, in a note 



published in the third edition of my Seventh Essay, I annour.c- 



ed a conjecture that the non-conducting power of fluids might 



perhaps depend solely on the extreme mobility of their parti- 



Z z 2 cli s ; 



are not .-illowcd 



