1811.] the Maximum Density of Water. 403 



the existing temperature was above, or below, that of the desi- 

 derated degree : this, therefore, affords some idea of the 

 approach to, or retrocession from, the temperature in question ; 

 but there is a better, and perhaps conclusive proof of its place in 

 the scale, which I shall now describe. 



I took a glass jar, 2 inches in diameter, and 34- in depth ; into 

 this was put distilled water to the depth of 2^V inches, and cooled 

 down to near the freezing point, but carefully prevented from 

 congealing, as the disengagement of air-bubbles from the ice, 

 when fluidity took place, would have frustrated the experiment; 

 into this was put a ball, previously well wiped with a silken * 

 cloth, and immediately, by means of a clean hook of glass, lifted, 

 but not rapidly, twice or thrice out of the water ; this cleared it 

 of any air-bubbles, which though imperceptible, might have 

 been adhering to its surface. The ball now fell to the bottom of 

 the jar, which as usual was convex, but had a small flat surface 

 on the summit, to which the ball was led, and there it rested. 

 In the water there were suspended two very accurate and sensi- 

 ble thermometers, the bulbs being at the middle of the water as 

 to depth, and just so far removed from its diametrical centre, as 

 not to be in the path of the ball when it rose. 



In these circumstances, the lower end of the ball was care- 

 fully watched with a large reading glass, and at the moment of 

 its quitting the bottom of the jar, the thermometers were exa- 

 mined, and the degree noted ; when the ball had risen about 

 one-fourth of an inch a small rod was cautiously let down, and 

 without agitating the water, gently made to touch the ball ; it of 

 course descended, but instantly rose ; this is a very delicate part 

 of the experiment, and if overdone loses its effect. It was fre- 

 quently repeated, and the ball always reascended with accele- 

 rated velocity. 



The thermometers indicating an increasing temperature, the 

 ball finally became stationary at the surface of the water ; from 

 time to time it was slightly touched as before, but in proportion 

 as the temperature rose beyond a certain point, the tendency of 

 the ball to ascend, after these strokes, obviously diminished, 

 judging by the velocity with which it did so ; its upper extre- 

 mity, when examined with the magnifier, plainly seemed to 

 press as it were more and more feebly on the surface of the 

 water, till at last, a fine thread of separation became visible ; the 

 degree by the thermometers was again marked, and as they con- 

 tinued slowly to rise, the ball gradually fell to the bottom of the 

 jar. 



From many similar experiments I have concluded that 42 is 

 extremely near the true point of the greatest density of water; 

 my most satisfactory trials never gave 3-10ths of a degree less. 



• In an experiment of such delicacy, this must be attended to, as linen never fails to 

 leave fibres on whatever is wiped with it ; these will detain air enough to render the 

 efforts of the experimenter in thiscas? abortive. 



2 i) 2 



