in Fhtids. 359 



us to form a just opinion respecting a matter of fact 

 which has been the subject of a good deal of dispute. 



Though many accounts have been published of ice 

 found at the bottom of rivers, yet doubts have been en- 

 tertained of the possibility of its being fanned in that 

 situation. From the result of the foregoing experiment 

 it appears to me that we may safely conclude, that, if 

 after a very long and a very severe frost, by which the 

 surface of the ground has not only been frozen to a con- 

 siderable depth, but also cooled several degrees below 

 the freezing-point, a river should overflow its banks, and 

 cover the surface of ground previously so cooled^ ice would 

 be formed at the bottom of the water ; but all the ex- 

 periments that have been made on the congelation of 

 water show the absolute impossibility of ice being ever 

 formed, in any country, at the bottom of a river which 

 constantly fills its banks, or which never leaves its bed 

 exposed, dry, to the cold air of the atmosphere. 



By reflecting on the various consequences that ought 

 to follow from the peculiar manner in which Heat ap- 

 pears to be propagated in Fluids, we are led to conclude, 

 that it is almost impossible that any Fluid exposed to 

 the action of light should ever be throughout of the 

 same temperature, though its mass be ever so small ; 

 and that the difference in the Heat of its different par- 

 ticles must occasion perpetual motions among them. 



Suppose any open vessel, — as a common glass tum- 

 bler, for instance, — containing a piece of money, a small 

 pebble, or any other small solid opaque body, to be 

 filled with water, and exposed in a window, or elsewhere, 

 to the action of the sun's rays. As a ray of light can- 

 not fail to generate Heat when and where it is stopped 

 or absorbed, the rays, which, entering the water, and 



