Indications of Thermometers. 1 1 



the bulb having at each repetition, after being raised to 105°, 

 been cooled down to 47° or lower, and quickly and carefully 

 wiped with a silk handkerchief before retaking the tempera- 

 ture of the mercury. 



In consequence of the weather being fine, its temperature 

 manifestly increasing, and the mercury at first so much colder 

 (one degree) than the atmosphere, I concluded that the suc- 

 cessive augmentations of temperature observed in the mercury 

 at the several repetitions of the experiment, were due rather 

 to the gradual rise of this body's temperature than to any 

 continued increase in the indications of the thermometer. As 

 a confirmation of this, I observed that the temperature of the 

 air and mercury taken at 9 h 40', or one hour after, was 49 0, 8; 

 which, compared with the temperature of the mercury at the 

 first and last of the preceding observations, shows nearly a 

 uniform increase of 2 0, 2 per hour after allowing for the excess 

 of the apparent above the real indication in the observation at 

 9 h 10'. 



Deducting therefore from all the observations except the 

 first a quantity proportional to the time, at the rate of -^-^ of 

 a degree per minute, the observations will stand, corrected for 

 the increase of the mercury's temperature, as below : 



This experiment seems to show that the difference of indi~ 

 cations in the thermometer is produced at the first heating of 

 the bulb, and not increased by any following similar opera- 

 tions. 



Exp. 14. During the preceding experiments I was near 

 the mercury at the times only of my taking its temperature ; at 

 other times I was many feet from it ; and not unfrequently 

 several things stood between me and it, so as to cut off all 

 rectilinear communication. But having after the last experi- 

 ment observed that the heat of the body, when standing within 

 about two feet from the unprotected glass, raised the tempera- 

 ture of the mercury in one instance as much as 3-10ths of a 

 degree in 5', I was desirous of repeating the experiment un- 

 der circumstances in which this proximity could have no 

 effect, or at least no sensible effect. For this purpose I ob- 

 tained the assistance of my friend Mr. Mervyn Crawford, to 

 whose precision the accuracy of the following experiments is 

 much indebted. 



The mercury having been for several days in our experi- 

 B 2 menting 



