Mr. W. Brown on Steel Magnets. 



273 



simple inspection of the table indicating the behaviour of each 

 specimen nnder the various treatments. 



In order to obtain data for the determination of the effective 

 lengths, the correct method would be to take a magnet the 

 length of the longest magnet required, and after tempering, 

 magnetizing, and testing it, to break it successively into a 

 number of lengths and test each part separately after remag- 

 netizing. But in doing so we would assume that every piece 

 of the same specimen would be in the same physical condition. 

 There is also the difficulty in breaking a glass-hard piece of 

 steel so as to have plane ends, and so permit accurate measure- 

 ment of the length of the magnet. 



I think, therefore, that there is less liability to error in 

 making all the magnets of the required lengths before temper- 

 ing, as was done in these experiments. 



Each specimen was cut in lengths varying from 1 to 20 

 centim., they were carefully made straight and the end- 

 planes made as accurately as possible at right angles to their 

 lengths. They were all made glass-hard by heating them to 

 a bright red heat inside an iron tube in a brisk coal fire, and 

 then dropping them end on into a vessel of water 100 centim. 

 deep, the temperature of the water being 7° C. A greater 

 number of magnets than were actually required were treated 

 in this way, and only those which were found to be glass- 

 hard throughout chosen for the experiments. 



They were then carefully cleaned, polished, measured, and 

 weighed, and, finally, magnetized to saturation in a long 

 helix giving a field of 1500 C.G.S. units intensity. 



In order to obtain the deflexions for calculating the mag- 

 netic moments, the apparatus employed consisted of a lamp 

 and scale, a modified form of the Bottomley magnetometer, 



Fi-. 1. 



40 Ct' ^ 



i p 







III 



, icr 1 ^u , 



I 1 1 



and a grooved platform for holding the magnets, as shown in 

 fig. 1. 



Phil Mag. S. 5. Vol. 27. No. 166. March 1889. T 



