106 PRINCIPAL FORBES ON AN EXPERIMENTAL INQUIRY INTO 



and even of the laborious computations which have formed the basis of the pre- 

 sent paper. 



118. In the preceding Table I have completed the series for lower temperatures, 

 where the observations were less accordant, in the following way : — I have 

 assumed that the most trustworthy part of the observational curves are those 

 between the actual temperatures of 40° or 50 s and 150° or 160°, and that within 

 moderate limits, the conductivities {k) may be represented in terms of the tem- 

 perature (t), by such a formula as 



h = A + at + hi 2 



In the case of the l^-inch bar, I find for these constants 



A = -01506 a = - -0000488 b = + -000000122 



From which the conductivities corresponding to and 25° have been interpolated. 

 In the case of the 1-inch bar the constants are — 



A -01117 a=- -0000235 &=+ -000000058. 



119. I must here observe, however, that the above form of relation between 

 k and t, which has been applied by Dr Matthiessen, in his extensive and im- 

 portant researches on electric conductivity, does not satisfy the form of our con- 

 ductive curves, Plate V. fig. 4, except through a limited range. I have reason, 

 however, to think, that down to 0° of temperature it may be sufficiently exact. 

 The " percentage decrement" of the conductivity between 0° and 100° is 24-5 

 for the larger bar of iron, and 159 for the smaller one. As in the case of Dr 

 Matthiessen's electrical experiments, the " percentage decrement" diminishes 

 with the conducting power, and in almost exactly the same proportion.* The 

 numerical values in either case are, however, considerably smaller for heat than 

 those obtained by Dr Matthiessen for electricity. 



120. With this exception, however, there is an agreement in the character of 

 the metals (so far as is yet known) in conducting heat and electricity. (See Art. 2 

 of this paper.) 



§ VII. — Concluding Bernards and Suggestions. 



121. In Art. (15) of the first part of this paper, I expressed my desire to afford 

 to future experimenters every aid I possibly could to resume and extend my obser- 

 vations (confined, unfortunately, to only one metal — iron), and to furnish them 

 with such advantages as my experience afforded, as well in methods of observa- 

 tion as of reduction. 



122. It was especially with this view that I have spent what may perhaps 

 appear an undue amount of labour on the reduction of the experiments considered 



* Phil. Trans. 1863, p. 380. 



