PEIRCE. — PERMEABILITY OF A MASS OF FINE IRON PARTICLES. 89 



four shunts. When the switch T was closed to the right, it was possi- 

 ble, by manipulating the rheostat arm and the commutator K, to put 

 Q through any desired hysteresis cycle in the usual manner. The 

 current came from a battery of storage cells, any number of which 

 could be used at pleasure. A current of 1 ampere in the solenoid 

 gave rise to a field of 54.8 gausses in the space within it. The field 

 about M'b needle had to be artificially strengthened to suit the circum- 

 stances, and a piece of soft Bessemer steel rod of almost exactly the 

 same dimensions as the column of filings was used to determine the 

 sensitiveness of the apparatus at any time. By means of a coil of 20 

 turns of extremely fine insulated copper wire wound directly on this 

 piece at its centre and connected with a carefully standardized ballistic 

 galvanometer, the induction flux through the centre of the rod could 

 be found and the corresponding deflection of the magnetometer needle 

 determined. 



The work was undertaken in order to compare the magnetic prop- 

 erties of masses of iron particles, as they came from the milling-machine, 

 with those of masses of particles from the same lot which had under- 

 gone the treatment used in hardening iron castings for permanent 

 magnets. These " filings " were prepared by Mr. G. W. Thompson, 

 the mechanician of the Jefferson Laboratory, who has had much ex- 

 perience in the process. A completely closed iron crucible with thin 

 walls, containing a mass of the particles to be treated, was heated 

 white hot under a power blast in a gas furnace, and then suddenly 

 chilled in an acid bath. After this experience, during which the 

 particles had been carefully excluded from the air, they had a some- 

 what altered color and lustre, but under a microscope of low power 

 showed very little difference from the untreated particles ; at best all 

 such particles cut by machine tools from iron castings are most irreg- 

 ular in form, and are so much seamed by deep furrows and pits as to 

 look like clinkers from a furnace. All the particles were kept quite 

 free from oil or dirt, and the surfaces of the " hardened " ones were 

 only very slightly tarnished, but it was not possible to pack quite so 

 large a mass of the material into a given space after the treatment as 

 before. This might have arisen from changes of shape, but it is a 

 suspicious fact that the induction flux through the centre of a column 

 (of given dimensions) of the filings under a given excitation was almost 

 exactly the same whether the filings had been hardened or not. Two 

 uniform glass tubes from the same piece and practically of the same di- 

 mensions were filled respectively with 103 grams of untreated particles 

 and 96 grams of the others, and each was put several times through a 

 hysteresis cycle, using about 250 gausses as the intensity of the maxi- 



