Calorimetric Thermometers. 333 



three minutes was absolutely necessary. The extent to which 

 error from this source may accumulate may be drawn from 

 the fact that this instrument, when the mercury is rising to 

 the temperature of the bath in which it is placed, will register 

 6 millim. lower than when falling to the same temperature, 

 however long it be left in the bath, if it be not tapped (the 

 temperature of the bath in this case was such that the mer- 

 cury stood at about the middle point of the stem). Again, 

 it was found, by means of a series of experiments in which the 

 height of the column was read when the instrument was 

 inclined at different angles, that with thermometer 55082, 

 which possessed a blown bulb containing 15*6 grams of 

 mercury, a difference of 6' 6 millim. in the reading was made 

 when the instrument was placed horizontally instead of ver- 

 tically; making due allowance for the greater size of the bulb 

 and greater fineness of the bore in 62839, this latter instru- 

 ment would probably exhibit a difference of at least ten times 

 this quantity under similar treatment. This will give some 

 adequate idea of how great an effect the expansion of the 

 bulb under pressure may have on the reading. 



Now, besides the inertia of the bulb, which may be overcome 

 by a sufficient amount of tapping, there appears to be another 

 and more serious imperfection, due to what may be termed a 

 permanent alteration in size, as opposed to the other, which 

 may be called temporary, and one which no amount of tapping- 

 will overcome. It would seem that the bulb of this, and every 

 thermometer, has two different shapes, one when the column 

 is rising, the other when it is falling. The following simple 

 experiments will serve to show this. 



The thermometer being suspended in the calorimeter full 

 of water, the contents of this were heated or cooled slightly, 

 and the rate of cooling (or heating) determined during con- 

 secutive intervals. 



In the first column the time is entered, in the second the 

 reading of the instrument, and in the third the rate of cooling 

 or heating. 

 A. Experiments in which a rate of heating was determined. 



Rate. 



Time. 



Reading (arbitrary). 









-1 



Hot water added. 





1 min. 





19-96 



4 „ 





20-07 



7 „ 





20-165 



10 „ 





20-245 



Id „ 





20-38 



16 „ 





20-47 



mean =*10. 



