158 COUNT RUMFORD. 



drew heat from the liquid in contact with it, which 

 thereby being rendered more dense, fell along the sides 

 of the tube, while, to supply its place, the lighter liquid 

 rose along the axis. The motion here described is 

 exactly that of the great geyser of Iceland. The water 

 falls along the sides of the geyser tube, and rises along 

 the axis. In this way, then, heat is propagated through 

 liquids. It is a case of bodily transport by currents, and 

 not one of true conduction from molecule to molecule. 



It immediately occurred to Rumford to hamper this 

 motion of convection. He called to mind an observation 

 he had made at Baiae, where the water of the sea being 

 cool to the touch, the sand a few inches below the water 

 was intolerabty hot. This he ascribed to the impedi- 

 ment offered by the sand to the upward diffusion of the 

 heat. The length of time required by stewed apples to 

 cool also occurred to him. He had frequently burnt 

 his mouth by a spoonful of apple taken from the centre 

 of a dish after the surface had become cool. He devised 

 thermometers with a view of bringing his notions to an 

 experimental test. With pure water he compared water 

 slightly thickened with starch, water containing eider- 

 down, and stewed apples bruised into a pulp which con- 

 sisted almost wholly of water. In all cases he found 

 the propagation of heat impeded, and cooling retarded, 

 by everything that prevented the formation of currents. 

 As he pursued his inquiries, the idea became more and 

 more fixed in his mind that convection is the only 

 means by which heat is diffused in liquids. He denied 

 them all power of true conduction, and though his ex- 

 periments did not, and could not prove this, they did 

 prove that in the propagation of heat through the 

 liquids he examined, which were water, oil, and mercury, 

 conduction played an extremely subordinate part. 



Rumford changes from time to time the tone of the 



