1'REE CALORIC.

50

expands by heat in the same manner as solid bo-dies; the heated particles of water therefore, atthe bottom of the vessel, become specificallylighter than the rest of the liquid, and conse-quently ascend to the surface, where, partingwith some of their heat to the colder atmosphere,they are condensed, and give way to a fresh suc-cession of heated particles ascending from thebottom, which having thrown oft' their heat at thesurface, are in their turn displaced. Thus everyparticle is successively heated at the bottom, andcooled at the surface of the liquid ; but as thefire communicates heat more rapidly than the at-mosphere cools the succession of surfaces, thewhole of the liquid in time becomes heated.

CAROLINE.

This accounts most ingeniously for the propa-gation of heat upwards. But suppose you w6reto heat the upper surface of a liquid, the particlesbeing specifically lighter than those belowq couldnot descend: how therefore would the heat becommunicated downwards?

MRS. B.

Count Rumford assures us that if there was noagitation to force the heated surface downw ards,the heat would not descend. In proof’ of this hesucceeded in making the upper surface of a vessel