156 Experimental Investigations 



three thermometers began to rise, and at the end of a 

 quarter of an hour they had all reached the points from 

 which they set out at the beginning of this operation. 



To facilitate the comparison of the results of these 

 two experiments, one made with cold water at rest, the 

 other with the same water in a state of constant agita- 

 tion, I have represented them in Fig. 4. 



In the first place, we shall learn several very interest- 

 ing facts by simple inspection of this figure ; we shall 

 see, 



i st. That the progress of refrigeration or, to speak 

 more properly, the decrease of temperature was every- 

 where much more rapid when the cold water in contact 

 with the extremity, E, of the cylinder was agitated than 

 when it was at rest. 



2dly. That the extremity of the cylinder in contact 

 with this water was constantly near jo colder 'in the 

 first case than in the second. 



^dly. We shall see that the progress of refrigeration 

 was everywhere, and in both the experiments, such 

 nearly as our theory points out. 



The decrease of temperature toward the middle of the 

 cylinder was so regular that it is more than probable 

 the apparent irregularities toward the two extremities 

 were occasioned solely by the difficulty which a body of 

 water finds in communicating its mean temperature to 

 a solid with which it is in contact. 



The boiling water being in continual motion, owing 

 to its ebullition, it had a great advantage over the cold 

 water, which was at rest, in communicating its tempera- 

 ture to the extremity of the cylinder it touched; but I 

 have found, notwithstanding this, that by agitating the 

 boiling water strongly with a quill, and particularly 



