HALL. THERMAL AND ELECTRICAL EFFECTS IN SOFT IRON. 41 



Part II : Details. 

 Preparation of the Main Bars. 



The two main soft-iron bars, a and j3 in Figure 3, were cut from the 

 same piece of metal, and each was marked at one end. The two marked 

 ends came from the same eud of the original bar, and were placed at the 

 same end of the apparatus for the main experiment. 



For an insulating layer on the narrow part of these bars, paper baked' 

 on with asphaltum varnish was at first tried ; but it did not adhere well, 

 and the varnish alone was finally used, baked on in an electric oven. 



An attempt was made to cut a groove for the fine wires of the thermo- 

 electric couples in a second layer of asphaltum, baked on at 100° C. over 

 a first layer baked on at 200° C. ; but this failed, the varnish chipping off 

 under the action of the cutting tool. 



These experiments, and various others which need not be described, 

 were made with trial bars of iron, not with the bars a and (3, as the com- 

 plete removal of the varnish from a rod in case of failure is a trouble- 

 some operation. The method finally adopted and used with bars a and (3 

 was substantially as follows : A thin layer of asphaltum varnish was 

 laid on and, before baking, was allowed to harden at the ordinary tem- 

 perature of the laboratory for three hours. This partial slow drying was 

 to prevent the varnish from running and distributing itself unevenly 

 during the baking, which lasted four or five hours at 200° C. Then a 

 second thin coating of varnish was put on and allowed to dry at the tem- 

 perature of the laboratory for two days ; then a third thin layer, which 

 also was left to dry during two days. In the unbaked double layer thus 

 provided a groove (see Figure 5) was made by means of a small, sharp- 

 edged, brass disk, held in such a way as to allow revolution with little 

 or no lateral motion and applied with moderate pressure to the rod, 

 which was carried beneath the disk in a screw-like motion by the action 

 of a gear-cutting lathe. After the grooves were thus made the bars 

 were baked about four hours at 200 J C. The varnish ran a little into 

 the grooves during this heating, but not enough to cause serious difficulties. 



The construction of the six thermo-electric couples which are placed 

 on bars a and j3 has been described in connection with Figure 5. As a 

 cement to be used temporarily during the placing of the couples in the 

 grooves prepared to receive them, a mixture of beeswax and honey in 

 equal parts was found satisfactory. After this placing, the windings of 

 wire were covered with asphaltum, which was allowed to dry over night, 



