at its Maximum Density. 273 



For example, in the experiments of which I have just 

 given an account, the cork cup, which received the 

 current of warm water descending from the metallic 

 point of the tin ball, was only \ of an inch below the 

 extremity of this point ; if this distance had been greater, 

 the thermometer in the cup would certainly have risen 

 to a less height : for this reason these experiments ought 

 not to be regarded as suitable for determining with great 

 exactness the temperature at which the density of water 

 is at a maximum, but rather as proving that this tem- 

 perature is really several degrees of the thermometric 

 scale above that of melting ice; and this is all that I am 

 particularly interested in showing at the present time. 



Judging from the constant temperature which. is found 

 at all seasons at the bottom of deep lakes and from the 

 results of several direct experiments, we may conclude 

 that water is at its maximum density when it is at the 

 temperature of about 41 of Fahrenheit's scale, which 

 corresponds to 4 on that of Reaumur, and to 5 of the 

 Centigrade scale. 



[This paper is translated from the Memioires de 1'Institut, etc., VII. 

 (1806), pp. 78-97. The greater part of the translation is taken from 

 Nicholson's Journal, XI. (1805), pp. 225-235.] 



VOL. II. 1 8 



