' 



MAXIMUM DENSITY OF WATER, 235 



When not only the whole of the conical point, but a part The conical 

 <iiro of the (hort cylinder were immerCed in the ice-cold water, fofVhe^immerfei 

 the warmed water appeared to be thrown into eddies in its de- mccal. 

 fcent, which difperfing about prevented its falling regularly in 

 one continued ftreara into the cork cup. 



To conclude, I would juft obferve, Ihat although the fore- It »s not eafyta 



t /•CM i* make conclufive 



going experiments appear to me to be perlectly unexception- experiments at 

 able, and that their refults afford demonflrative proof of the elevated temper- 

 fa6ls which they were contrived to eftablifli; yet, when at-^^"*^"' 

 tempts are made by experiments fimilar to Ihefe, to determine 

 whelher heat can be made to pafs dozvniuards in water which 

 is at a higher temperature than that at which its denfity is a 

 maximum, difficulties occur which appear to me to be quite 

 infurmountable. 



The fluidity of water Is fo perfe6l, or the mobility of its For the fides 

 particles fo great, that the liquid at the furface, which is firft greatly modify 

 heated and rarefied, immediately fpreads far and wide, and the effefts by 

 meeting with the fides of the containing veHTel, heats them, ^eaf "^c? ^"^ 

 and this heat, fo acquired, making its way downwards (as 

 well as upwards) in the folid fubftance of which the velTel is 

 conftruded, raifes the temperature of the lower ftrata of the 

 fluid in contad with it, which moving towards the axis of the 

 vefifel, communicates heat to a thermometer placed there, be- 

 low the furface of the water. 



That thefe various operations do in fadl take place, nobody Fa£ls which 

 can doubt ; and it appears to me to be the more probable that biil^ ' hat^con " 

 all the heat which a thermometer placed below the furface of fiderable heats 

 warm water acquires when a great degree of heat is applied ^'? ^ j tranf- 

 to the particles of that fluid which are at the furface, is in fafl wards through 

 received from the fides of the containing veflel, not only be- ^"'^" ^^r^t 

 caufe the thermometer acquires heat fo very flowly, but alfo, veffeU 

 and more efpecially, becaijfe this heat is acquired much more 

 flowly when the containing velTel is wide than when it is nar- 

 row; and alfo when it is made of a fubftance which is a good 

 condu6tor of heat tiian when it is conftrudted of a fubftance 

 which is a bad conductor of heat, as I have found by experi- 

 ment. But as this particular enquiry is foreign to my prefent 

 purpofe, I fliall not enlarge on it in this place, 



Defcription 



k 



