at its Maximum Density. 265 



Experiment No. i. Having removed the ball, I gently 

 brushed away the warm water which in the last experi- 

 ment had been lodged in the cavity of the cork cup, and 

 which still remained there, as was evident from the indi- 

 cation of the thermometer belonging to the cup; I then 

 placed several small cakes of ice in the cylindrical vessel, 

 which ice, floating on the surface of the water in the ves- 

 sel, prevented the water from receiving heat from the 

 surrounding air, which at that time was at the temper- 

 ature of 70 F. As the cork cup had been a little heated 

 by the warm water in the foregoing experiment, time 

 was now given it to cool. 



As soon as the cup and the whole mass of the water 

 in the cylindrical vessel appeared to have acquired the 

 temperature of freezing, I carefully removed the cakes 

 of ice which floated on the surface of the water, and in- 

 troduced once more the projecting conical point belong- 

 ing to the metallic ball into the ice-cold water in the 

 vessel, placing it exactly in the same place which it had 

 occupied in the foregoing experiment ; but this ball, 

 instead of being at the temperature of 42 F., as before, 

 was now at the temperature of 60 F. 



The results of this experiment were very striking, 

 and, if I am not much mistaken, afford a direct, unex- 

 ceptionable, and demonstrative proof, not only that the 

 maximum of the density of water is in fact at a temper- 

 ature which is several degrees above the point of freezing, 

 but also that warm currents do actually set downwards 

 in ice-cold water, whenever a certain small degree of 

 heat is applied to the particles of that fluid which are at 

 its surface, as I have already announced in my Essay on 

 the Propagation of Heat in Fluids. 



The conical metallic point had been in its place no 



