[ 97 ] 



" At the fame time I placed another gallon of the fame fea- 

 vvater in a glafs body. The column of water in this veffel was 

 about thirteen inches high, about fix inches diameter at the bafe. 

 and about three inches at the mouth of the veffel. I placed this 

 body with the fea-water clofe by the veffel marked A ; fo that 

 both were equally diffant from the adjoining houfes; and after 

 marking the glafs body B, I covered the veffels A and B with 

 glafs bafons in fuch a manner, that the air might communicate 

 with the furface of the water, but rain or fnow might be ex- 

 cluded. 



" A Thermometer was placed between thefe veffels. 



" From the 2d to the 7th of January, the mercury in the 

 Thermometer ftood, at various times, as low as thirty-one of 

 Fahrenheit; and Thames water in mallow wooden veflels, placed 

 on the ground, near the wall above-mentioned, was often frozen 

 to the thicknefs of a crown piece. But an earthen oil-jar con- 

 taining twenty gallons of Thames water, and a like jar contain- 

 ing twenty gallons of diftilled water, and each covered with a 

 pewter difii, preferved the water contained in them from free- 

 zing during this interval. 



" About the 7th of January, the mercury in the courfe of 

 twenty-four hours did not rife above thirty-one, but fometimes 

 funk to thirty. Ice was formed in the veflel marked A ; but none 

 in the veffel marked B. Ice was at the fame time formed in 

 the great jars containing Thames water and diftilled water; and 

 to a thicknefs much greater in the Thames water than in the 

 water diftilled. The ice obtained from the veffel A was all 

 formed on the furiace of the water ; and confifted of thin 

 laminae adhering to each other weakly, and intercepting in their 

 interfaces a fmall portion of water, which was feline to the tafte- 

 This ice beaten gently with a glafs peftle to divide the lamina-, 



O then 



