35 Of the Propagation of Heat 



supposed attraction, and suspecting that the equal dis- 

 tribution of the salt was owing to a very different cause, 

 the internal motions among the particles of the water, 

 occasioned by accidental changes of temperature, I 

 made the following experiment, which, I fancy, will be 

 considered as decisive. 



Experiment No. 57. 



I took a cylindrical glass jar, 4^ inches in diameter, 

 and yf inches high, and placing it in the middle of 

 another cylindrical glass jar, y% inches in diameter and 

 8 inches high, which stood in a very shallow earthen dish, 

 nearly rilled with pounded ice and water, I placed the 

 dish, with its contents, on a strong table, in an unin- 

 habited room in a retired part of the house, where the 

 temperature of the air, which was the same, with very 

 little variation, day and night, was at about 36 F. 

 Having prepared, and at hand, a quantity of the strong- 

 est brine I could make with sea-salt, which was very clear, 

 transparent, perfectly colourless, and ice-cold, and also 

 a quantity of fresh or pure water, ice-cold, lightly 

 tinged of a red colour with turnsol, and some ice-cold 

 olive oil, I first poured as much of the fresh water into 

 the small cylindrical jar as was necessary to fill it up to 

 the height of above 2 inches ; and then, by means of a 

 glass funnel, which ended in a long and narrow tube, 

 by introducing this tube into the fresh water, and resting 

 it on the bottom of the jar, I poured a quantity of the 

 brine, equal to that of the fresh water, into the jar ; and 

 in performing this operation I took so much care to do 

 it gently, and without disturbing the fresh water already 

 in the jar, that, when it was finished, the fresh water, 

 which, as it was coloured red, could easily be distin- 



