Indications of Thermometers. 13 



tailed, we are instantly led to attribute it to the recommuni- 

 cation, from the glass of the stem to the colder descending 

 portion of mercury, of a part of that heat which had been 

 communicated from the warmer portion of the ascending co- 

 lumn. Were the bulb at every instant uniformly of the same 

 temperature, the successive portions of mercury which ascend 

 into the stem from the bulb would be successively higher in 

 temperature than those which precede or are above them ; 

 and the temperature of all that had ascended from the bulb 

 would be less almost in the precise proportion to the distance 

 from the bulb. But, from the quick conductibility of mer- 

 cury, it seems probable that there is a rapid and continued 

 transmission of temperature from the bulb to the higher parts. 

 And, in consequence of the little affinity between mercury and 

 glass, this transmission may be aided, when the stem is verti- 

 cal, even by the ascent of the warmer portions, on account of 

 the difference in the specific gravity. However this may be, it 

 is plain that the mercury being warmer than the stem, must 

 communicate a portion of its excess of temperature to the 

 glass. The loss thus sustained by the mercury will be shortly 

 made up by further transmission from the bulb. In the course 

 of a very few transmissions in this way, that is, in a very short 

 time, the interior of the glass to a small depth from the co- 

 lumn of mercury, will attain a temperature equal or nearly 

 equal to that of the adjacent mercury ; provided the tempera- 

 ture of the bulb remain nearly the same. 



Suppose now the bulb to be cooled, the mercury in the 

 stem which occupied the higher will descend to the lower and 

 wanner parts, and if this descent be very rapid it may exceed 

 in celerity the tardy communication of heat from the sur- 

 rounding glass; so that the column of mercury at the instant 

 it becomes stationary, is colder than the glass enveloping it. 

 An expansion, therefore, and a consequent rise in the mer- 

 curial column will gradually ensue by the communication of a 

 part of the excess of heat in the contiguous glass. This phe- 

 nomenon would be similar to that in the first two experiments. 



Should the bulb be brought back after being heated to its 

 primitive temperature, it is plain that the suspended column 

 of mercury, though precisely the same in quantity, yet being 

 by the heated glass raised in temperature, will settle to a 

 higher indication than before. This is the phenomenon con- 

 tained in the 14 last experiments. 



The amount of this variation of apparent temperature mani- 

 festly depends, in either case, on the difference of the second 

 and last temperatures of the bulb, and the length of the column 

 of mercury in the );>st temperature. If these remain the same, 



a repetition 



