2gg MAXIMUM DENSITY OF WATER. 



into all the re- outfide of the brim of the brafs cup, betweeh it and the 



roaming (pace, vertical fide of the cylindrical veflel, which vertical pieces 



of ice reached upwards to within about y^ of an inch of the 



top or brim of the cylindrical veflel ; and when this was 



done, ice-cold water was poured into this veflel till the furface 



of that cold fluid flood juft one inch above the level of the 



brim of the cork cup. — Both cups were of courfe fubmerged 



and filled with ice-cold water, and furrounded on all fides by 



folid cakes of ice. 



The cups re- After things had remained in this fituation more than an 



merged for one bour, during which time the cold water in the cylindrical 



hour. velfel, and lliat in the cups, was frequently agitated with 



the foft end of a flrong feather, and the cups and the water 



in every part of the vefl^el appearing to be exa6lly at the 



temperature of freezing : I proceeded to make the following 



decifive experiment. 



J 



Experiment. No. 1 . 

 Experiment!. A foIid ball (F) of tin having, been provided, two inches 

 A cone of metal jj^ diameter, with a cylindrical projedlion on the lower fide of v I 

 plunged in the '*» one inch in diameter, and half an inch long, ending in a 

 ice-cold water, conical point, which proje6led (downwards) half an inch 

 iork cup : farther ; this ball (to which was fixed a flrong iron wire fix 



inches in length, which ferved as a handle to it,) having been 

 made to acquire the temperature of 4-2° F. by keeping it 

 immerfed near half an hour in a large quantity of water at 

 that temperature, was placed as expeditioufly as poflible over 

 the middle of the cork cup, and in fuch a fituation that tliQ 

 whole of the defcending conical point (half an inch in length) 

 was immerfed in the ice-cold water in the cylindrical velfel, 

 the extremity of that point being juft half an inch perpen-, 

 dicular above the upper fide of the bulb of the thermometer 

 which lay in the cork cup. 

 ^t was fore feca, I knew that the particles of ice-cold water which werQ 

 that if the water ^i^yg brought into conta6l with the conical point, could not, 



m con tadt with . r r n i r t /I 1-1 



the cone became iail to acquire lome Iraall degree of heat from that relatively^ 

 denier by the warm metal; and I concluded, that if the particles of water 

 dcfcend and lo warmed fliould in fad become heavier than they were be- 

 raife the thcr- fore, in confequence of this fmall increafe of temperature, 

 inometer. ^^^^ ^^^^ necelfarily defccnd in the furrounding lighter ice^ 



cold liquid, and as the healed metallic point was placecj 



directly 



