OF ARTS AND SCIENCES. 



193 



at very high temperatures, while the Hall effect curve, as here giro,,, 

 for French cold rolled steel begins to diminish in steepness in the 

 region of 300° C. The data for the latter curve, however, are too 

 inaccurate to warrant attaching much importance to this feature. It 

 is very likely that the true curve would continue to increase in steep- 

 ness far beyond the point corresponding to 300° C. It is evident 

 that, until very high temperatures are reached, the Hall effect in 

 French cold-rolled steel increases much more rapidly with rise of tem- 

 perature than the permeability of Wlhtworth steel does. 



50° 



100° 



150° 200° 

 Figure 1. 



250° 300° 350° 



Nickel. 



This was from a sheet of rolled nickel from Wharton, Philadelphia, 

 and was probably very pure. It was about 0.083 cm. thick. The sheet 

 fi-om which it was cut was not large enough to allow of making the 

 original side arms of the strip very long, but they were pieced out 

 with shreds from the nickel sheet, so that the junctions of nickel with 

 the copper wires lay outside the air bath. The shreds of nickel were 

 fastened to the arms of the strip by means of screw-clam ps, and these 

 clamps were insulated from the nickel by pieces of mica. At nines 

 the device of wrapping the nickel strip in .isliestos paper was tried as 

 a means of preventing sudden fluctuations of temperature. In Bpite 

 vol. xxvni. (n. s. xx.) 13 



