1 1 Mr. J. Heiapath on the Indications of Thermometers. 



a repetition of the heating and cooling process immediately 

 afterwards would therefore produce no difference in the amount 

 of the variation, provided sufficient time was given in the first 

 process for the mercury to obtain from the bulb, and com- 

 municate to the glass, the necessary quantity of heat, accord- 

 ing to the distance from the bulb. For like reasons a third, 

 fourth, &c. process immediately afterwards would not alter 

 the value of the variation. Hence the reason that the 12th 

 and subsequent experiments did not exhibit any difference in 

 the apparent temperatures at the 3d, 4th, &c. observations; 

 proper allowance being made for the real variation of the 

 mercurial temperature during the intervals. 



With respect to the time required in the first heating and 

 cooling of the mercury to give the maximum effect, I have 

 not made any observations ; but, from a circumstance or two 

 that occurred in the course of my experiments, a much shorter 

 period than I took would produce a sensibly less variation. 

 Nor have I determined in how long after such an experiment 

 a like variation may be observed with the same instrument; 

 but I conceive that a half hour's careful cooling of the stem 

 would be sufficient, or less if the stem could be well cooled in 

 a less time. 



It can hardly be expected of me, with so few experiments, 

 to enter into a theoretical computation of the value of this 

 variation under all circumstances. If however the views that 

 I have taken of it be nearly correct, the variation would be 

 about half the mercurial correction for the same column of 

 mercury at the 2d and 3rd temperatures of the bidb. 



Now the correction given by the Royal Society's Committee 



range out of wat. x excess temp. . , „ ,, . , _, ,. 



is = - — — minus £ ol this value. Call- 

 ing, therefore, in round numbers the range of the last ther- 

 mometer from —4-0° to +100°, and supposing the mean tem- 

 perature of the bulb to have been 50°, we shall have for the 

 general correction 



(40+ 50) x (100- 50) /. 45) _. 3938 . 



S 10000 ^ 1Z)— 3»-«* s 



the half of which, or °'197, ought to be the amount of varia- 

 tion. Our experiments give • 17. 



Again : supposing in the first experiment there were about 

 20° or 22° of the lower degrees immersed, which I should 

 think there were though I did not attend to it carefully, the 

 correction would be 



Half 



