PEIRCE. — RESISTIVITY OF HARDENED CAST IRON. 203 



difficult to make upon cast iron and often disagree with the resistivity 

 test. A repetition of the annealing and hardening process increases 

 somewhat the size of a specimen and increases the resistivity for both 

 the annealed and the chilled states, but in the hardened state the iron 

 is never so hard mechanically as at the first hardening, and the bar 

 loses in great measure its magnetic retentiveness, as do most kinds of 

 tool steel which have been through the same experience. Many kinds 

 of chilling liquids serve to make cast iron glass-hard, but for massive 

 pieces cold water seems not to give such uniform results as the acid 

 bath used by some professional hardeners. The temperature coefficient 

 {a) of the resistivity of every one of my specimens was decreased by 

 the hardening, though this does not seem to have been the case for the 

 special cast iron used by Barus and Strouhal, which had a larger coef- 

 ficient (120) than any I used. The coefficient a is not always smallest 

 in that one of a number of specimens of cast iron which has the 

 largest resistivity. 



Castings from different sources often show when glass-hard a very 

 close agreement in magnetic permeability, though their resistivities and 

 the temperature coefficients of the resistivities may differ widely. The 

 temperature coefficient of the magnetic moment of a cast-iron bar mag- 

 net is usually not more than one third as large as that of a similar 

 magnet made of tool steel. 



A uniformly hardened cast-iron or steel rod may have been irregu- 

 larly magnetized, but if it be thoroughly demagnetized and then care- 

 fully remagnetized in a solenoid, its magnetism will become regular. 

 Only irregular hardening seems to lead to persistently irregular mag- 

 netization in the case of a bar magnet, though the use nowadays of 

 electromagnetic crane lifters sometimes magnetizes iron and steel rods 

 in a manner which is difficult to deal with in the laboratory. Even an 

 irregularly hardened slender rod may usually be demagnetized well 

 enough for all practical, purposes in a solenoid which carries currents 

 alternating in direction and gradually decreasing in intensity, but 

 large thick pieces are very tenacious of charges once given to them. 

 The shield of a certain Rubens Panzer galvanometer in use in the 

 Jefferson Laboratory was twice heated white hot and was kept hot for 

 some time in a vain attempt to get rid of a slight magnetization. 

 The resistivity of different portions of a casting gives trustworthy 

 information about the uniformity of the hardening. Occasionally, as 

 in a case cited above, an irregularly hardened piece of tool steel may 

 be magnetized nearly normally, but usually irregular hardening leads 

 to an irregular distribution of the magnetism which shows itself in an 

 abnormal iron-filing diagram. An unusual filing diagram does not 



