224 C. A. Perkins — Magnetic Permeability of Nickel. 



layer of asbestos, etc., by which means an insulation was pro- 

 duced which was not affected by heating to a red heat. 



The primary coil was in two layers, and as the second layer 

 was not wound immediately upon the ring, a certain error was 

 introduced, the magnetism being apparently increased by an 

 amount which was in one case -£$, and varied from this to jj-q, 

 but in rings I and II it in no case exceeded -p^. Whenever 

 it did so in ring III a correction was made. 



The magnetizing current was measured by a tangent gal- 

 vanometer constructed by Kowland. Its constants were com- 

 puted from its dimensions. The circle was 21 cm. in diameter 

 and graduated to quarter degrees. Six coils were used, each 

 containing about three times as many turns as the next pre- 

 ceding. Prom six to twenty-four cells of Bunsen chromic acid 

 battery furnished the current, which was further modified by 

 the introduction of a variable resistance. 



The galvanometer for measuring the induced currents had a 

 sensitively adjusted astatic needle whose directive force was 

 given by a magnet placed below. It had a period of about 

 nine seconds, and was deflected by one turn of the earth in- 

 ductor through nearly four degrees. It was read by a tele- 

 scope and scale one meter distant. The earth inductor was 

 introduced for the purpose of comparing the induced current 

 with one of known strength and the winding included a total 

 area of 20,716 sq. c. m. It was adjusted in a plane perpen- 

 dicular to the magnetic meridian and could be turned through 

 an arc of 180°. As in Rowland's experiments, a coil of wire 

 sliding upon a magnet was introduced into the circuit, to con- 

 trol the vibrations of the needle. 



The ring was placed in a bath of kerosene for low tempera- 

 tures, for 100° in one of paraffine and heated in a steam bath. 

 At higher temperatures it was placed in an oil bath which 

 was designed for the comparison of thermometers at high 

 temperatures. This was made of very heavy copper castings 

 and held several liters so that its temperature, when kept up 

 by a Bunsen burner, would remain constant for a long time, 

 rarely varying more than four or five degrees during the ex- 

 periment which usually took several hours. The temperature 

 was determined by a mercury thermometer. 



The course of each experiment was as follows : 



The current was allowed to run for some minutes through a 

 shunt circuit of about the same resistance as the ring, then 

 turned through the ring and the readings of the tangent gal- 

 vanometer made. 



Then the deflections of the mirror galvanometer by the in- 

 duced currents were measured, four readings being taken for 

 the temporary magnetism and four for the total. Then another 

 measurement of the primary current was made. 



