MR. H. L. CALLENDAR OR THE PRACTICAL 
166 
On Pyrometry there is a valuable paper by Weinhold in ‘ Poggendorff, Annalen,’ 
1873, who gives a fairly complete list of original memoirs, nearly all of which have 
been referred to. 
On the air thermometer there is a most valuable paper by Balfour Stewart in 
the ‘Phil. Trans.,’ containing perhaps the most accurate experiments yet made on 
the dilatation between 0° and 100°. For high temperature experiments, papers by 
Deville and Troost, and E. Becquerel, are most important. 
Siemens suggested the platinum pyrometer, but I cannot find any account, worth 
mentioning, of his experiments. 
Determination of the Increase of Resistance with Temperature of the Standard 
Platinum Wire by comparison with the Air Thermometer. 
As stated above, the direct comparison is preferable to the method adopted by 
Benoit and others of heating the wire in vapour baths of substances whose boiling- 
points have to be assumed from other experiments ; a spiral of fine bright platinum 
wire is a bad radiator, and is exceedingly sensitive to slight changes in the tempera¬ 
ture of the air with which it is in contact; if, therefore, it be fixed inside the bulb of 
an air thermometer, the mean temperature of the spiral will be always very nearly 
the same as that of the air, and this is measured directly on the absolute scale. 
The Air Thermometer. 
Since the ordinary form of the instrument is in several ways inconvenient, it was 
found necessary to devise a modified form for the purposes of this investigation, a 
detailed description of which will be given. The sources of error, and the corrections 
to which it is liable, will also be incidentally considered. 
The general plan of the instrument will be easily understood on reference to the 
accompanying diagram (Plate 11, fig. 1). [An improved form is described in the 
Appendix.] 
A bulb A is connected by means of a capillary tube to a U-gauge of small bore 
(about 2 millims. diameter) containing pure sulphuric acid, which serves to confine a 
constant mass of air at nearly constant volume ; the gauge carries a bulb B on its 
other limb, which is connected by an indiarubber tube to an adjustable mercury 
manometer M with wide tubes. Attached to the limb of the U-gauge is a 
millimetre scale by which its reading can be recorded. 
The volume of the bulb A is determined by calibration with water or mercury in 
the usual way, or with air by the method of the volumenometer.* 
The coefficient of cubical expansion of the bulb enters as a small correction, 
amounting to about 1 per cent.; it is best determined at low temperatures by using 
the bulb as a mercury thermometer, and assuming the absolute expansion of mercury. 
* ! Practical Physics,’ Gt.azebrook and Shaw, p. 1G0. 
