OF IRON, COPPER, AND GERMAN SILVER. 539 



I. Experimental Methods. 



As already stated, this work was intended to be a repetition of Prof. Tait's 

 experiments upon the same bars, and accordingly every care was taken that it 

 should be conducted under conditions as nearly as possible similar ; except, of 

 course, the important feature of the nickel films. For a full account of the 

 original experimental details, reference may be made to the papers of Forbes 

 (Trans. R. S. E., 1860-61 and 1864-65) ; and for subsequent improvements, to 

 that of Prof. Tait (Trans. R. S. E., 1878). To what is there given concerning 

 the experimental means employed little can be added. The set of thermo- 

 meters, the position and use of each, the maintenance of regularity in gas 

 pressure, and consequently in the steadiness of temperature during the last 

 few hours of the long bar experiment — all were exactly the same as in the 

 preceding work ; and any slight novelty in this part of the work was more 

 by way of addition to, than change in, the method. This was essential in 

 order that a thorough comparison might be made between the two series of 

 results. 



The four metallic bars experimented upon were, respectively, wrought 

 iron (that originally employed by Forbes), copper of superior electric conduc- 

 tivity (copper, crown), copper of inferior electric conductivity (copper, C), and 

 German silver. These, together with the short bar corresponding to each, were 

 all plated with nickel. The plating, although comparatively light, was of good 

 quality. Before beginning work on one of the bars, it was polished with fine rouge 

 and paraffin oil, and afterwards thoroughly dried ; while during the progress of 

 the experiments on the long bars these were kept free from dust, &c, or any- 

 thing likely to interfere with the normal amount of radiation. It was found 

 that, when the distances along the bars, especially between the closer holes, 

 were measured by a dividing engine, some small temperature correction would 

 require to be made for the (slight) inexactness of the position of the holes. In 

 no case was this correction greater than one-third of a degree C, and this, since 

 it was only required at the highest temperatures, was well within the limits of 

 other inevitable and undetermined errors. 



The thermometers used in this inquiry were, as already stated, those 

 used by Prof. Tait in his experiments. They were made under the direction 

 of Dr Balfour Stewart, and afterwards carefully corrected at Kew. A fair 

 idea of their quality before they had been used for any experimental work of 

 this nature may be obtained from the following table, which gives the correc- 

 tions according to the Kew certificates : — 



