1889 - 90 .] Mr A. C. Mitchell on Conductivity of Aluminium. 301 
requiring large quantities of this substance. It seemed desirable 
that an investigation of the thermal conductivity of aluminium 
and its alloys should be made, but before obtaining the necessary 
bars a smaller bar was obtained from experiments on which a pre- 
liminary notion was got of the thermal conductivity; this being 
necessary in order to determine the most convenient dimensions for 
the long bars used in Forbes’s method of conducting these experi- 
ments. 
The present note gives the result of this preliminary experiment. 
The small bar used was furnished by the Aluminium Company, 
Birmingham, who now manufacture pure aluminium by the Deville- 
Castner process. Analysis showed that the particular sample used 
contained nearly 98 per cent, of pure metal, the chief impurity being 
iron. The method employed for finding the conductivity was sub- 
stantially that of Forbes, the only difference being that the cooler 
end of the bar was inserted in a cold bath. The bar used, being 
20 inches long and 1 inch square in section, was of such a length that 
it could be also used for the cooling experiment. Four holes were 
drilled in the bar, separated by intervals of 3 inches. The thermo- 
meters used were those employed in Professor Tait’s and my own 
previous work. The source of heat was simply a Bunsen burner, 
applied to one end of the bar. 
For such a preliminary experiment it is unnecessary to give all 
the data regarding rates of cooling, values of tangents, &c. ; and it 
will therefore suffice to state simply the result. By these experi- 
ments the thermal conductivity of aluminium was found to be 
•072 when expressed in foot-pound-minute units. This number 
represents the conductivity at (about) 100° C. The corresponding 
numbers for other metals are — 
Copper, . . . . *071 
Iron, *0116 
German silver, . . . *0078. 
Hence the thermal conductivity of aluminium (see Trans. Roy. Soc. 
Edin ., vol. xxxiii. p. 559) is slightly greater than that of the best 
conducting copper. 
Recently (see Nature , June 20, 1889), in a lecture delivered at 
the Royal Institution, London, Sir H. Roscoe stated that Faraday 
