[ iT 3 3 



like matter, figure, and capacity. It was thence inferred, that fire 

 does not fo quickly pervade thick bodies as it does thin bodies; 

 and that fire pervades water more freely than it does folid bodies, 

 and fooner- diffufes itfelf from water to air, than from any 

 other body containing water to air. 



" Thence it followed, that in reafoning on the phenomena of 

 songelation, the mafies of water, the duration of cold tempera- 

 ture in the atmofphere, and the maffes of other matter furround- 

 ing water, are to be confidered. Deep rivers and lakes do not 

 freeze fo foon as mallow rivers and lakes. Large bodies of 

 water are never frozen in any temperature of mart duration; 

 but mallow waters are often frozen in the fummer. 



M It need not be prefumed, that certain lakes, which are never 

 frozen, communicate with fubterranean fires, or hot mineral 

 ftreams ; or that they are impregnated with matter which im- 

 pedes congelation: but it is rather to be prefumed, that as fire 

 llowly pervades, enters, or quits bodies, the time neceffary for 

 its dirFuimg itfelf from deep lakes to the cold atmofphere is 

 greater than ever fuch temperature of the atmofphere continues 

 without interaiiflion below the freezing point. 



" By the like reafoning applied to maffes of earth and other 

 matter which are not fo quickly pervaded by fire as water is, we 

 can conceive why deep wells and iprings at or near their ifliung 

 from the earth are not frozen in this climate even when navigable 



O 



rivers are ice-bound. We alfo under ftand why the main pipes, 

 .buried in our ffcreets, retain the water fluid, when the pines leading: 

 from thefe to the houfes and croffing the area of each houfe, are 

 choaked with ice; and why hay-bands twifted round thefe fmall 

 .pipes prevent the freezing, &c. 



" On thefe grounds it is prefumed, that no eonfiderable con- 

 gelation ever takes place in the fea, becaufe this is the greateft 



Ct and 



