PEIRCE. — CHANGES IN INDUCTANCES OF ELECTRIC CIRCUITS. 571 



shorter iron rods of a given diameter are tested, Pf gradually decreases 

 to zero under given value of C, and then changes sign ; the inversion 

 comes with longer rods when C is weak than when it is strong. If, 

 with a given rod, C be gradually increased, the negative moment 

 finally decreases and changes sign. After this there is no inversion 

 and Pf is positive. 



It appears from Fromme's experiments that the von Waltenhofen 



Figure 32. The toroid (DIN) and the electromagnet (TP) with its poles 

 separated by a gap of about eight centimeters, were placed in parallel across 

 the terminals of a storage battery, with an oscillograph in the toroid branch. 

 QVX shows the course of the current in the toroid which approaches its final 

 value from above (See Figure 12). GCF, with its irregularities of curvature, 

 shows the current in the toroid when the battery circuit was suddenly broken. 

 AU shows the form of the temporary current induced in the toroid when an 

 iron block was suddenly dropped into the gap between the jaws of the elec- 

 tromagnet, after the currents in the circuit had become steady. 



effect is often less marked in straight, finely divided cores than in solid 

 ones, and we may inquire how greatly the division of a straight core 

 may be expected to facilitate the changes in the field (ZT) within the 

 iron, due to given changes in the exciting circuit. It is clear that if 

 the circuit of an electromagnet be suddenly broken, the decay of the 

 electromagnetic field in the core is much less rapid when the core is 

 solid and eddy currents induced in it shield the inner filaments, than 

 when it is made of wire. Indeed, if eddy currents were non-existent, 

 the field would fall instantaneously to zero, in the absence of magnetic 

 lag, when the current in the coil ceased to flow. If the exciting coil 

 remains closed and some change is suddenly made in its resistance or 



