THE THERMOMETER. 141 



an atmosphere of steam raised from the water under a pressure equal to that 

 of the atmosphere. This steam has the true temperature of the boiling water ; 

 and, by drawing the tube upward through the orifice in which it plays, the 

 height of the mercurial column in the thermometer may be marked with the 

 utmost accuracy, and thus the boiling point may be determined. 



The variation of the column in the thermometric tube, strictly speaking, arises 

 not from the expansion of the mercury alone, but from the difference between 

 the expansions of the mercury and glass. It is clear that, if a given change 

 of temperature dilated equally the glass of the tube and bulb, and the mercury 

 contained in it, the height of the column would not be varied ; because, in the 

 same proportion as the dimensions of the mercury would be increased, the ca- 

 pacity of the tube and bulb would also be increased. But, in fact, although the 

 tube and bulb undergo an increase of dimension from every change of tempera- 

 ture, that increase is extremely small when compared with the dilatations of 

 the mercury, and consequently, notwithstanding that more room is made for 

 the fluid by the dilatation of the glass, yet still, the room not being nearly suf- 

 ficient, the mercury rises. Nevertheless, although the variations of the mer- 

 curial column are not absolute indications of the dilatation or contraction of the 

 mercury, yet it so happens that, under all the changes of temperature to which 

 a mercurial thermometer can be submitted, the dilatation of glass is in the same 

 proportion as the dilatation of mercury, and consequently the change of volume 

 of the mercury bears a fixed proportion to the change of the capacity of the 

 tube ; and the variation in the height of the column contained in the tube bears 

 also the same proportion to the variations which it would undergo if the glass 

 suiTered no expansion or contraction. The apparent dilatation of the mercury, 

 or the difference between the dilatations of the mercury and glass, between 

 the freezing and boiling points, amounts to one sixty-third part of the volume 

 of mercury at the temperature of melting ice ; and the actual dilatation of the 

 mercury between these limits of temperature is somewhat less than this, 

 being j-ffy parts of the volume of the mercury at the temperature of melt- 

 ing ice. 



The fact that the indications of the thermometer are independent of the ab- 

 solute expansion of the glass which forms it is a matter of great importance, 

 because it shows that the accuracy of thermometers does not depend upon the 

 species of glass of which they are formed. Had it been otherwise, one of the 

 conditions necessary in the construction of a thermometer would be, that the 

 glass should be manufactured of elements precisely alike in all cases. That, 

 however, is by no means necessary. Different kinds of glass undergo different 

 degrees of expansion by change of temperature ; but they will expand propor- 

 tionally to each other, and proportionally to the expansion of mercury within 

 those limits of temperature to which mercurial thermometers are applied. 



It will be perceived, from the reasoning that has been pursued upon this sub- 

 ject, that the indications of all thermometers whatever would necessarily cor- 

 respond, even though the fluid from which they are formed were different, pro- 

 vided only that the rate of its expansion correspond with that of mercury. A 

 thermometer of spirits of wine, within that part of the scale through which the 

 dilatation of that fluid is uniform, would necessarily correspond with the mer- 

 curial thermometer. The difference would only be in the length of the scale, 

 or, ii other words, in the distances between the freezing and boiling points. 

 In the case of spirits of wine, however, the rate of dilatation approaching the 

 boiling point of water is not uniform, as has been already stated. 



It may possibly be thought that the preceding details respecting the con- 

 struction and use of thermometers may be elaborately minute, and that an in- 

 strument apparently so trifling as a glass bulb blown on the extremity of a tube, 



