380 DENSITY OF WATER. 



the afcending current ftill continues, but has produced liltie 

 effe(5t, having not added more than half a degree to the tem- 

 perature. 

 ^untilthe tem- During the next 20 minutes the top gains l°.5, the bottom 

 cxcxed? S.^^' °"'^ ^°~^' ^" ^'^'* interval we may oWerve the current has 

 turned, but not yet acquired much force. The point of great- 

 ell denfity muft therefore have exifted at the laft obfervation or 

 near it : the mean of 35^,5 and 37® is 361" for the required 

 point, as deduced from this experiment. 

 After which the ^^ ^^ minutes more the top gains 1°, and the bottom little 

 heated water af- or nothing; here we find the afcending current has become 

 cends. ^^j^.^^ ^^ ^^ manifeft its influence very fenfibly. 



In the next 10 minutes, the top gains 4**, and the bottom 



only ,25; here the afcending current has become quadruple 



what it was 2° below ; becaufe the farther the temperature is 



raifed above the ftationary point, the more powerful is the 



force of afcent arifing from the fame interval of temperature. 



Thefe fa<Ss do It would be in vain to attempt to reconcile the above ex- 



not agree with periment with the opinion that water is denfeft at 39° or 40°. 



maximum den- At the very moment when the mean temperature of the water 



fity at 39° or is 39<?, we obferve the afcending current the moft a^ive, when 



^ ' it ought to have been defcending or imperceptible. 



The eff"e£lisnot I once imagined that the experiment might be explained 



modified by the ^^ ^j^g fuppofition of 32^ being the point of greateft denfily; 



table or fupport. , ^ , / ' . r r . \ x . . . r 



that the ludden mcreale ot temperature at the bottom arofe 



from ihe heat of tha table, and that the cohefion of the par- 

 ticles of water prevented their afcent under the propulfion of 

 fo fmall a force; but having procured a large glafs jar which 

 could be fufpended, I found the fame order of differences 

 nearly :as when placed on a table, and was therefore obliged 

 to al>andon that explanation. 



Intending to fend the remainder of this inveftigation foi^ 

 a future number, I remain 



Your friend, 



JOHN DALTON. 

 Manchefier, .^jpri;7l4, 1 80.(5. 



III. Account 



