142 



PROCEEDINGS OF THE AMERICAN ACADEMY. 



closely adjacent regions of the same bar may have very different 

 specific magnetisms, owing perhaps to differently arranged inclusions 

 of slag. It is certain that pieces cut across the direction of the rolling 

 often show different permeabilities from those of pieces cut in the 

 same region in the direction of the length of the bar. Gumlich found 



BTnTrnp-TTraTTrr 



Tr'iv--!,!T^:rr 



;nr'r::;!:;;'r:-*"nr!::"::::;|:r::-t:rtfF 



j ! 1. I ■ i : 



OL 



Figure 10. This Figure shows the results of observations made by Roessler 

 upon an ellipsoidal piece of soft iron 50 cms. long and 1 cm. in diameter at the 

 center. 



a piece of soft "Steirisches Eisen" about 3 cms. long and about 3 mm. 

 in diameter which also showed the value 1796 for /. 



There seems to be no doul)t, therefore, that some specimens of soft 

 iron are to be found which have materially higher maximum values 

 of / than had the specimen used as a standard by Messrs. Hadfield 

 and Hopkinson. Four different observers, using solenoids for mag- 

 netizing their test pieces, and seven persons using other methods have 

 thought that they met with such pieces. This fact does not, of course, 

 make the work of Messrs. Hadfield and Hopkinson any the less valu- 

 able, but it shows, I think, since some pieces which contain consider- 

 able quantities of Fcs'C, have given values of / above 1720, that 

 material bought in the open market cannot be expected to obey the law 

 which the series of steel pieces from the Hecla Works follow. 



Still, the majority, perhaps, of pieces of iron and steel bought at 

 random will have specific magnetisms not very different from the 



