ABSOLUTE THERMAL CONDUCTIVITY OF NICKEL. 



378 



depend on that of the determination of specific heat. While I have no confidence in 

 the values found for the specific heat of the nickel, I give the values of the conductivity 

 found by using them. I hope to be able at a future time to supplant these figures by 

 others which can be relied on. 



The following table gives the values of the ratio of the conductivity to the specific 

 heat after applying the end correction to the rates of cooling given in a previous table, 

 and the values of the conductivity using the values of the specific heat in the adjacent 

 column. No corrections have been applied for changes of the dimensions of the nickel 

 with temperature as these are really negligible. 



Temperature. 



Ratio of Conductivity to Q 



Specific Heat. BpeCI 



fic Heat. Cone 



i 



uctivity. 



/ 



40 



1-19 



098 



118 





50 



1-19 



102 



121 





60 



1-19 



105 



125 





70 



1-20 



108 



130 





80 



1-20 



111 



133 





90 



1-18 



114 



135 





100 



1-16 



118 



137 





110 



1-14 



121 



138 





120 



1-12 



124 



139 





130 



1-09 



•127 



139 





140 



107 



130 



140 





150 



1-06 



134 



142 





160 



1-04 









170 



1-01 









180 



•99 









190 



•97 









200 



•96 









§ 8. Experiment. 



\L AND OTHER ERRORS IN FORBES 



/~\ 1 1 v-VT-* rr* • 



j's Method. — The sou 



rces of 



error 



Statical Experiment. 



(1) Thermometric errors. \ 



(2) Errors in reduction of results, for example in differentiating the temperature 

 :urve. 



(3) Want of uniformity or regularity in the substance or surface of the Forbes bar. 



Cooling Experiment. 



(4) Radiation from ends of bar. 



(5) Lag of thermometer behind bar due to gradient of temperature necessary to 

 ause flow of heat from thermometer to bar. 



(6) Thermometric errors. 



(7) Errors of observation in taking cooling readings. 



(8) Errors in reduction of rate of cooling from these readings. 



