158 THE ROYAL SOCIETY OF CANADA 



Specimen No. 7. 



T Correction d 



10-76 l-Ol 10-52 



21-8 1-90 11-14 



32-98 2-47 11-18 



41-59 2-90 11-94 



52-35 2-83 11-34 



61-20 2-75 11-58 



71-50 2-42 11-88 



81-04 1-90 12-37 



91-72 0-92 12-12 



The curves show how the iron temperature corrections and the 

 corresponding values of 8 vary with the comparison temperature. 

 There is a slow rise in 8 as the temperature rises. On working out 

 the probable errors which occurred in the determination of d it was 

 found that the displacement of the values of d from a straight line 

 may be entirely due to errors in the observations and apparatus. This 

 undoubtedly means the probable rectification of the points when 

 increased accuracy and improved observation are attained. 



Assuming that 5 varies linearly with the temperature, the following 

 formula will then express this variation: 



where A ^^^ ^ ^^'^ constants, and can be obtained experimentally 

 from two observations, similar to those tabulated, using two different 

 auxiliary temperatures. A complete knowledge of the thermometric 

 constants of a specimen of pure iron wire involves altogether four true 

 temperatures, two of which are the standards 0° and 100°. The com- 

 plete correction formula for an iron thermometer is — 



T-tF = (A + 6T) (I- l)L 



■^ ^11 ' 100 



The magnetic properties of some specimens of pure iron wire were ob- 

 tained and are tabulated below, namely the maximum permeability 

 H and the Steinmetz coefficient rj. 



Magnetic Properties. 



Specimen /t 35/ x 10' 



No. 1 Pure 1082 4-18 



No. 1 Annealed 2552 2-70 



No. 2 Pure (hard) 480 11-66 



No. 2 Annealed .. 1283 7-07 



No. 3 Pure 1310 3-93 



