RECORD AND DISCUSSION OF TEMPERATURES. 



From the preceding comparisons, the following corrections have been deduced :- 



The corrections between temperatures of 20 and 40 for the two standards 

 are small, and hardly exceed 1 at the lower limit, but beyond this point, for the 

 lower temperatures, a regular increase is observed in the correction to the spirit 

 standard. The caliber of most of the spirit thermometers was tested before leaving 

 New York, and the differences exhibited in the table of comparisons for the several 

 instruments arise, perhaps, principally from the unequal contraction oftolored and 

 not chemically pure alcohol. At these extreme low temperatures, the fluid appears 

 to change its condition, and the coloring matter, with a falling temperature, remains 

 adhering to the sides of the tube. 



