l\)0 AriNIMUM TEMPERATURE OF WATER. 



and llill remain fluid. For my part, I adopted the maxim 

 without hefitation, until the perufal of Dr. Hope's paper, 

 given in the fupplement to your iafi volume, led me to reafon 

 in the following manner on the fubjefl. , 



When water is expofed to a freezing mixture, thofe parti- 

 cles of it which are in contaft with the fides of the veffel, are 

 loon reduced to a temperalure lower than the point of con- 

 gelation; in confequence of" this, they are probably converted 

 Tnto minute icicles, which impart a quantity of caloric at the 

 moment of their formation to the furrounding water, thereby 

 preventing its temperature from finking below 32°. Thefe 

 invifible bodies afterwards begin to afcend flowly on account 

 of the diminution in their fpecific gravity; and while they rife 

 towards the furface of the water, other particles will approach 

 tlie fides of the veffel in fuccetfion, and undergo a fimilar - 

 transformation. Thjs procefs would evidently increafe the 

 volume of the water without reducing its temperature, fup- 

 pofing it to be ice-cold at the commencement of the experi- 

 ment; for the hypothefis refts on the fuppofition that water 

 freezes as foon as it is cooled below the 32nd degree of Fah- 

 renheit's fcale. This gradual increafe of bulk will explain 

 the appearances defcribed by my friend Mr, Dalton, who 

 found that thermometers filled with water continued to rife 

 when expofed to freezing mixtures, until the enclofed wafer 

 congealed fuddenly, and fiequenlly burft his inftruments. The 

 reafon why agitation accelerated the congelation of water 

 thus circumftanced appears to be this: When the invifible 

 icicles become very numerous, the leaft motion carries thern 

 in crowds againft the fides of the vefTels ; where the fmall 

 quantity of water contained amongfl them cryfiallizes imme. 

 diately, and cements the whole into a film^adhering to the in- 

 fide of the cup. This theory, or hypothefis, call it what- 

 ever you think proper, evidently requires water to be what it 

 really is, namely, a bad condudor of heat ; and after forming 

 it, I proceeded to examine the merits of it experimentally, in 

 the following manner, 

 i^penment. ^ Experiment \fi. A fmall thermometer was fufpended at 

 was cooled bya the lower end of a wire, which could be moved in a vertical 

 furrounding direction, through a hole in a horizontal bar of wood, fixed 

 \i^camtcon. Over a (able for thepurpofe; a veffel, containing a freezing 

 gcaied at the mixture, of the tempjerature of 21®, was next placed with 



fides by ftirring, . ■. 



having never 

 f'jnlt below 3Z*. 



