INTRODUCTIO>f. 79 



ations give ground to conjecture, the constant 



temperature of 1°, 7 R., which precedes the 



freezing of the sea-water, begins, and then con- 

 tinues to uncertain depths. 



Celebrated natural philosopliers, (Franklin and 

 Humboldt,) have affirmed, building partly on ex- 

 periments, that in the vicinity of sand banks and 

 shoals, the sea is colder than where there are none 

 such, and that consequently the thermometer may 

 serve as a warning to mariners. The property of 

 the water to propagate warmth from above, quicker 

 and deeper, and the slowness with which the earth 

 is influenced by external warmth, seem to favour 

 this assertion, at least for the spring and summer 

 months ; but, in winter, the contrary is probably 

 the case ; at all events, it deserves to be accurately 

 examined by as frequent observations as possible, 

 for which we possess in the Six-thermometer so 

 convenient an instrument. 



One of the facts respecting the temperature of the 

 ocean, is the phenomenon of the formation of ice 

 in the sea, on a large scale. The first business of 

 the navigator here, is to determine, by every pos- 

 sible means, the magnitude of the floating masses of 

 ice ; respecting the extent of which, the accounts 

 of preceding navigators have given us the grandest 

 ideas. For this end it is necessary to determine 

 by experiments made on the spot the specific gra- 

 vity of this ice compared with the sea-water ; the 

 best way is, to get a largish piece of ice of a pris- 



