HALL AND CAMPBELL. — MAGNETIC EFFECTS IN SOFT IRON. C29 



meter long, were soldered to copper wires, the junctions being kept in 

 a water bath. The distance between junctions (1) and (8) could not 

 be very accurately determined. It was measured and taken as 2.78 cm. 

 The wires Ci and Cs ran down from the points of attachment, passed 

 under the lower edge of the plate, and came up across it on the other 

 face. The distance between C2 and I2 and between C^ and I4 was per- 

 haps 0.05 cm., each wire being attached to the outer end of the arm. 

 On the face shown in Figure 6, a flat sheet of asbestos about 0.15 cm. 

 thick, arranged so as not to disturb the wires, was cemented by means 

 of asphaltum or melted shellac (the latter was used with Plate 2), the 

 whole being subjected to a temperature high enough to soften the 

 cement and allowed to cool with sufficient pressure on the asbestos 

 sheet to fix it flat in place. 



The width of Plate 1 is the direction of the fibre, or grain, of the 

 original iron bar. 



Plate 2. — The measurements of Plate 2 were 



Length between coppers 5.15 cm., 



Width 2.01 " , 



Arms about 0.10 cm. wide and 0.10 cm. long, 



Thickness 



(Figure 5) 



cm. Pt. cm. Pt. cm. Pt. 



0.134 at 1 0.135 at 2 0.138 at 3 



0.138 " 4 0.142 " 5 0.145 " 6 



.0.140 " 7 0.143 •• 8 0.146 " 9 



Means 0.137 0.140 0.143 



The face of Plate 2 was smoother at the last than the face of Plate 1, 

 and we have taken as the mean thickness of the arm strip of (2) the 

 mean of the measurements as here recorded, 0.140 cm. The thickness is 

 several per cent greater on one edge of the plate than on the other, but 

 it does not appear that any serious harm can come from this inequality. 



The iron wires of Plate 1 (see Figure 6) were with Plate 2 replaced 

 by constantan (" advance ") wires, Ki, Kg, etc., about 0.020 cm. in di- 

 ameter, which had been annealed by heating to bright incandescence 

 by means of an electric current. The distance from junction (1) to 

 junction (3) was now made about 2.62 cm., in order to have them a 

 little nearer the centre of the magnetic field than the corresponding 

 junctions of Plate 1 had been. Wires Ci and C2 did not here pass 

 under the lower edge of the plate, but followed the course shown in 

 Figure 6, strips of mica fastened by shellac, melted on, being used for 

 insulation. The wires themselves were bedded in shellac in their 

 course across the plate. The plate was finally backed, on the side 



