OF ARTS AND SCIENCES. 191 



tures above the range of this device. It was formed of a narrow brass 

 cliamber arranged to be placed conveniently between the magnet poles. 

 It was furnished with a funnel-shaped pipe extending downward, under 

 which could be set a Bunsen gas-burner, and it was surmounted by a 

 brass chimney to carry off the products of combustion. The height of 

 the flame of the burner could be regulated so as to furnish any required 

 amount of heated air, which circulated uniformly past the specimen 

 to be examined. By shielding the gas flame from air currents and 

 jacketing the outside of the bath with asbestos cloth a tolerably uniform 

 temperature was attained, A high range thermometer was inserted 

 mto the air bath, with its bulb in close proximity to the specimen. 



In many cases the change of electrical resistance with change of 

 temperature was noted by means of observations on the magnitude of 

 the difference of potential between the admission and exit ends of the 

 strip under examination. These observations served to show directly 

 that in nickel and steel at ordinary temperatures the temperature 

 coefficient of the Hall effect is greater than the temperature coefficient 

 of the electrical resistance. 



The numbers given in this paper are, in most cases, not those actu- 

 ally observed, but have been derived from the actual observations by 

 changes, usually slight, intended to make allowance for certain varia- 

 tions, in the conditions of the experiments, such as fluctuations of the 

 magnetic force or of the main current in the strip. The results 

 reached with the various substances examined will now be given. 



Copper. 



From a sheet of thin commercial copper about 0.01 cm. in thickness 

 an experimental strip was cut, which had the usual form with the 

 exception that the side arms were made very long in order that their 

 junctions with the galvanometer wires might lie outside the air-bath. 

 In the table that follows, the first column, t, gives the temperature in 

 degrees Centigrade ; the second column, a, the reading of the instru- 

 ment used to measure the magnetizing current ; the third, 6, the read- 

 ings of the tangent galvanometer used to measure the primary current 

 through the copper strip ; the fourth, H, the Hall effect deflections of 

 the mirror galvanometer. 



