Dr. Mills's Researches on Thermometry. 11 



the stem of which was graduated in millimetres, was exhausted 

 by means of a SprengePs air-pump and filled with mercury. 

 It was then placed in melting ice ; and when the position of 

 the mercury in the stem had become constant, it was noted. 

 The bulb and its contents were then immersed in boiling water, 

 and the reading in melting ice was again determined. The 

 zero was found to be considerably depressed; but if the bulb 

 was left at the temperature of the air, this depression steadily 

 diminished. The experiments of heating the bulb once and 

 observing the zero were repeated after an interval of four, 

 and again after another interval of thirteen days, with the result 

 that after the second and third heatings the zero fell to the 

 same position as it had taken up after the first. The method 

 of experiment was then changed. The bulb was raised to 

 about 100° C; and the water with which it was surrounded, 

 having been allowed to boil for a few minutes, was run off, and 

 the bulb was left to cool for half an hour or thereabouts. This 

 operation was repeated during three consecutive days, in all 

 twenty times ; and the zero-point was determined as before. 

 The mean depression of the zero immediately after the three 

 previous boilings had been 12*38 millim.; it was now found 

 to be only 9*35 millim. Twelve days afterwards the alternate 

 heating and cooling was again commenced, and in the course 

 of five days was repeated fifty times. The depression of the 

 zero immediately after the last boiling was 7*71 millim. It 

 had therefore been still further reduced. 



The experiments are summed up in the following table : — 







Table II. 







I. 



II. 



in. 



IV. 



V. 



VI. 



Days after 

 heating. 



i. 



ii. 



in. 



IV. 



v. 







1240 

 9-80 

 8-50 



12-35 



7-08 

 6-68 



6-20 



12-39 

 9-66 

 8-49 

 8-30 

 7-98 



9-35 

 7-61 



5-04 

 4-23 



7-71 



2-7 



1 



2 



3 



4 



5 



6 



8 



9 



13 









There were in all, as has just been stated, 5 series of ex- 

 periments. Column I. gives the number of days which had 

 elapsed since the last heating to which the bulb had been 

 subjected in any particular series, at the time when the ob- 



