PEIRCE. 



RESISTIVITY OF HARDENED CAST IRON. 



191 



For massive pieces of iron the A' mixture, which certainly is very good, 

 is said to work more uniformly than a water bath. Several specimens 

 which were chilled in iced water and iced brine developed minute 

 cracks which showed in irregularities when the rods were magnetized, 

 but these, which were tested before the construction of the special gas 

 furnace, may not have been uniformly heated. The oil bath was 

 nearly as good, so far as increasing the resistivity of the specimen, as 

 the water bath, but the hardened pieces did not seem so hard mechan- 

 ically. The melted paraffine wax, at as low a temperature as would 

 keep the wax liquid, also increased the resistivity of a specimen chilled 

 in it, provided it had not been hardened before, quite as much as the 

 water bath, but a piece thus hardened would not scratch glass. 



Most of the pieces of American cast iron which I have tested had, 

 when soft, resistivities referred to the centimeter cube, which at 0° C. 



TABLE III. 



would lie between 73 microhms and 104 microhms. These pieces when 

 hardened for the first time had resistivities which at the same tem- 

 perature lay between 80 and 126. Nine pieces of American cast iron 

 tested when soft by Barus and Strouhal had on the average a resistivity 

 at 20"" C. of about 79.1 microhms with a temperature coefficient of 

 0.00120. Four grids, typical of the softer kinds of iron which I have 

 used, gave on the average when soft at the same temperature the 

 results which appear in Table III. 



To show the effect of hardening upon the temperature coefficient of 

 the resistivity, I may instance six specimens with three different 

 coefficients when hard. (See Table IV.) 



When a number of steel bars of the same length and cut from the 

 same long rod are hardened and are then magnetized in the same 

 solenoid and aged, it frequently happens, as is well known, that the 

 ultimate magnetic moments of the bars differ somewhat widely from one 

 another ; and the same thing is true of magnets made from cast-iron 

 rods cut from the same grid. In Table V are given the magnetic 



