The Comparison of Thermometers. 359 



have already been used by Shaw* for measuring tempera- 

 tures. He aspirated a known volume of air saturated with 

 moisture at a temperature '/,' through tubes filled with 

 pumice moistened with strong sulphuric acid. The increase 

 in the weight of these tubes gave the data necessary for 

 calculating t. This method is, however, only applicable to 

 ordinary temperatures. 



The platinum thermometer (Fig. i.) which we used was 

 made under the direction of Mr. E. H. Griffiths, by Mr. 

 Thomas, of Cambridge. It consists of a platinum wire 

 coiled on a roll of asbestos paper. The ends of this wire 

 are soldered to thick copper leads, which communicate 

 with binding screws. The coil is protected by a thin glass 

 tube closed at the lower end. The copper leads are in- 

 sulated by narrower glass tubes. The resistance of the 

 coil at o was I0'637 ohm. 



The mercurial thermometer used was one by Hicks, and 

 was graduated in ^ degrees. The calibration of the stem 

 gave the following corrections : — 



Correction at 100° + '38 



>i 90 + *32 



» 80 + -27 



» » 70 + '22 



„ » 60 + -15 



» 50 + *i4 



„ „ 40 + 'I2 



» 



n ,» 10 + '03 



»> » 



,, 30 + 'o8 



20 + '06 



O + O'O 



The measurements of the resistance were made by 

 means of a post-office resistance box, by Elliot, and a sensi- 

 tive galvanometer. The resistances were read directly on 

 the box to '01 ohm, and the numbers in the third decimal 



•Cambridge, /%?■/. 7Va;w., 1885. 



