MAXIMUM DENSITY OP WATER. ^233 



rature of 60'^, we may fee how much more rapidly the ther- 

 mometer in the cork cup acquirc^l heat when the metallic ball 

 was relativdj/ cold than when it was at the temperature ot 

 boiling water ; and it is more than probable that it was not till 

 after the conical metallic point had been confiderably cooled, 

 by a contad with the ice-cold water, that thofe ftreams of 

 moderately warmed water began to defcend from it, by which 

 the thermometer was at length heated. 



In the experiments made with the ball heated in boiling When the very 

 water, a fmall thermometer, placed juft under the furface of ^JJ^j^^"^^"^ 

 the ice-cold water, on one fide of the metallic point, began meter at the fur- 

 to rife very rapidly as foon as the hot ball was fixed in its^^^^^^* heated j 

 place; but another thermometer placed about half an inch —but a lower 

 lower, on one fide of the cork cup, remained to all appear-""^ was no.. 

 ance at perfedl reft, from the beginning of the experiment to 

 the end of it. 



The explanation of all thefe appearances is To eafy that It Explanation s 

 would be a wafte of time to fay much on that fubjeft. It may 

 however be ufeful to recapitulate the principal phenomena, 

 and {hew in what manner they tend to eftablifli the fa6ls which 

 they are brought to prove. 



Every body muft fee, at the firft glance, that in all thefe That the fl)ght!y 

 experiments the heat which caufed the thermometer to rife was defcends'-^^^'^ 

 carried down into the cork cup by defcending currents of 

 warm water; and it is evident that water which defcends 

 muft of neceffity be fpecifically heavier than that in which it 

 defcends. 



From the refults of thefe experiments we may conclude that— and thegrcat- 



the denfity of water is a maximum when that fluid is at a tem-^^ denfuy is ra- 



r II ! I /• 1 /. . , , ther below 40°. 



perature lomewhat lower than that or the Jortieth degree oH 



Fahrenheit's thermometer. 



In all the foregoing experiments, more or lefs warm water 



defcended through the ice-cold water into the cork cup; but 



the refults of the experiments which were made with the me- Hotter water is 



talllc ball heated in boiling water, ftiew evidently that when ^'Shter and af- 



the particles of water at or near the furface of a quiefcent mafs 



of ice-cold water, are by any means heated to a temperature 



feveral degrees above that of 40*^ of the fcale of Fahrenheit, 



fuch particles* fo heated, become fpecifically lighter than ice- 



,cold water, and confequently cannot defcend in that cold ^ud 



denfer liquid. 



