110 Mr. H. Tomlinson on Thei*mal Conductivity [Feb. 21, 



with a galvanometer by copper wires, well covered with caoutchouc, 

 one end of each being soldered to the copper strips, so that either one or 

 both the copper strips with the soldered German- silver wires could be 

 employed as a thermo-element. In case it was found convenient to use 

 only one of these, the other was kept well covered with sawdust. 

 These elements will be called the G. S. elements. The second con- 

 sisted of two small thermo-piles, each of 12 elements, of antimony and 

 bismuth, and were each fitted into india-rubber tubes, so that they 

 could be inserted into small wooden boxes, containing water, so as to 

 fit water-tight. Sometimes these, like the G. S. elements, were used 

 separately ; at others, were made to neutralise each others effect on 

 the galvanometer. 



With these elements, a delicate Thomson's reflecting galvanometer, 

 having a resistance of 2 ohms, was employed, the scale being placed 

 about 6 feet from the mirror of the needle. 



In the circuit of the magnetising current, a tangent galvanometer 

 was placed, the needle of which was suspended by a very fine platinum 

 wire, attached to a graduated torsion circle. The two thick copper 

 wires which conveyed the current were each 27 centimetres in dia- 

 meter and 14 inches apart. 



A magnetometer was also employed to test the magnetism im- 

 parted, but, as the readings will not be given in this paper, no de- 

 scription of the instrument is necessary. 



To heat one end of the bars two large Leslie's cubes were employed, 

 having each two apertures projecting about 2 inches into the interior 

 of the cubes. The apertures in one cube were \ inch square and placed 

 in one side of the cube, about 3 inches apart from each other, and 

 half way up the side. In the other cube the apertures were circular, 

 and \ inch in diameter, and placed in the centres of opposite sides 

 of the cube. Into these cubes bars of square and circular section 

 were respectively inserted, and water filling the cubes was raised to 

 100° C. 



The first experiment was made on a soft iron bar, -J inch square in 

 section. The bar was inserted into the Leslie's cube with square 

 aperture, filled with boiling water, and into the other aperture was 

 inserted a brass bar of the same section. The length of the iron bar 

 was about 10 inches, and of the brass one, 18 inches. One of the G. S. 

 elements was fastened on to the iron bar, and the other to the brass, 

 by caoutchouc bands, the elements being insulated from the bars by 

 thin paper. The G. S. elements were made to very nearly counteract 

 each others effect by sliding one or other of them up and down the 

 bars until the heat conducted by the bars to the elements was in both 

 the same as nearly as possible. Underneath the bar was placed the 

 electro-magnet, with a piece of white cardboard lying on the upper 

 extremities of the soft iron cores, so that the cores were about 3 milli- 



