70 On the TjXpanJioii of Water during Congelation. 



at the extremities, except by a fniall pipe in each iflliingfrom 

 the upper fiirfacc. One of thefe 1 filled with pure but ordi- 

 nary rain-water; the other with water which I had carefully 

 diftilled. The apertures of both were clofed alike by well 

 fitted plutrs of wood, which I had previoufly boiled in oil, 

 I expofed them to the open air, on the furface of the fnow, 

 in Quebec, the night of the 5th of January 1799, when the 

 thermometer ftood at 28 degrees below zero of Fahrenheit's 

 fcale. About twenty mimiles afterwards, I found both the 

 cylinders inclining to the horizon at an angle of about thir- 

 teen degrees; being fo far rent from their bafes, which yet 

 relied on the fnow. The intermediate fpace in each was 

 filled w ilh rude mafles of ice, which, on being examined 

 with the microfcope, (in a roonj,) appeared to be compofed 

 of cryitals, chieily in the form of parallclopipeds and trun- 

 cated pyramids. The whole (if the appearances, in both 

 cafes, were cxaol:ly liniilar. I afterwards repeated thefe ex- 

 periments, without arty variety in the refult. 



From hence we perceive evidently that Dr. Black's folu- 

 tiou is inadequate to account for thefe remarkable pheno- 

 mena. At the fame time 1 think it impoffible not to admit the 

 original principles u]ion which his ingenious rationale reds. 

 It is inipoflil)le ihat water can be converted into ice without 

 the evolution of one-tenth of the whole quantity of its fpe- 

 cific fire, which is equal to 146 degrees of Fahrenheit's ther- 

 mometer. Now, we know that water can be converted into 

 vapour, under particular circumftances, by much inferior tem- 

 perature. — Qiui-rd. Wuit not part of the water have, been con- 

 verted into vapour in the vciy procefs of congelation ; and 

 may not this caufe be adequate to the produtSlion of the molt 

 violent expanlive force of troll? 



. QuihiC) Janntiry 24, 1 799. 



IX. The 



