THERMOMETER AND PYROMETER. 



1.3 



M. Flaugergues aif ributes this change 

 to the effect of long continued atmo- 

 spheric pressure on the bulbs of ther- 

 mometers, in which there is no air to 

 counteract it. De la Rive and Marcet 

 give the same explanation, and remark 

 how this circumstance must affect the 

 result of all experiments on the cold 

 produced in vacuo. 



Arago is not inclined to attribute this 

 elevation of the zero to atmospheric 

 pressure on the bulb ; since he found it 

 equally affecting thermometers with 

 very thick and very thin bulbs. He in- 

 clines to ascribe it to the disengagement 

 of air, which either adhered to the glass 

 or the mercury, and its accumulation in 

 the upper pail of the bulb, so as to 

 affect the column in the stem. 



The most complete observations on 

 this point are those of Bellani,* who 

 acknowledges two sources of variation 

 in the zero of thermometers. That ele- 

 vation of the zero, first noticed by Flau- 

 gergues, according to him, goes on 

 gradually increasing for a limited pe- 

 riod, but ceases after a year or two. He 

 ascribes it to the extreme slowness with 

 which glass once softened has the equi- 

 librium among its particles restored. 

 He found, that some months after gra- 

 duation, a thermometer did not sink 

 quite to the freezing point when im- 

 mersed in melting ice; if laid by for 

 some months, and again tried, its zero 

 will be still higher ; but after some time 

 this irregularity ceases. He found that 

 this effect was not diminished by leav- 

 ing the thermometer open at the top, 

 and it was sensible even in spirit ther- 

 mometers. 



The other irregularity noticed by 

 Bellani is detected in the following 

 manner. Let a thermometer, having 

 such a range that ^ of a degree is 

 appreciable, after lying by for some 

 months, be plunged into melting ice, and 

 its height accurately noted, then into 

 boiling water, and again into ice, it will 

 now stand lower by about T 1 5 of a de- 

 gree than at its first immersion in the 

 liquefying ice. This effect he ascribes 

 to the extreme slowness with which the 

 expanded glass can regain its former 

 state of contraction, compared to the 

 mercury. 



These deductions appear to be per- 

 fectly just ; and we are further indebted 

 to Bellani for an ingenious method of 

 showing that the air, if not wholly, is 



* Bellani, Giornale di Fisica, torn, v. 



chiefly retained in thermometers and 

 barometers by the glass, not by the 

 mercury. He introduced a portion of 

 unboiled mercury into a bulb, contain- 

 ing mercury which had ceased to give 

 out any air, and found that this intro- 

 duction did not renew the agitations 

 which the first application of heat to 

 the bulb had occasioned. 



The difficulty of freeing thermometers 

 from air is admitted by Arago, while he 

 recommends boiling the mercury in the 

 bulb as the best method of effecting the 

 expulsion of the air; and he quotes 

 some unpublished experiments of Du- 

 long, to show the tedious manipulations 

 which are necessary for this purpose. 



We would recommend the boiling to 

 be performed in the manner stated, un- 

 til the agitation of the fluid caused by 

 the air ceases ; and after the tube is 

 closed, the observations of Bellani 

 would incline us to recommend, for 

 delicate instruments, that the attempt 

 to fix the freezing point should be de- 

 ferred, until the glass might be supposed 

 to have contracted to its state of equi- 

 librium ; after which, there would pro- 

 bably be little change in the dimensions 

 of the bulb. 



6. We come now to the last and most 

 delicate step of the process, the adapta- 

 tion of the scale to the instrument. 



In the manufacture of thermometers 

 this is conveniently done by plunging 

 the new instrument, along with a stand- 

 ard thermometer, into two liquids at 

 different temperatures : but the gradua- 

 tion of this standard instrument is a 

 work of such nicety and importance, 

 that a committee of seven members of 

 the Royal Society was formed to in- 

 vestigate the subject, and their elabo- 

 rate report is given in vol. Ixvii. part ii., 

 where all the requisite circumstances 

 are distinctly noticed, and the best ma- 

 nipulations minutely described. 



Two fixed points are sought ; and the 

 freezing and boiling points of water are 

 most convenient for that purpose. To 

 find the first, nothing more is necessary 

 than to place the thermometer to be gra- 

 duated, after it is filled, in melting snow 

 or ice, in such quantity around the ball 

 and tube, as to bring it to the desired 

 temperature. When the mercury has 

 become stationary in the tube, a mark 

 is to be made on the tube with a file, 

 just opposite to the top of the mercurial 

 column ; and that mark fixes the freez- 

 ing point of the scale of the instrument. 

 The determination of the boiling point 



