300 Magnetic Condition of an Alloy of Iron and Aluminium. 



Figure 2. 

 Curve I. (7th heating) . H = 4'0 c.G.s. units. — Bends corre- 

 sponding to the three maxima can be detected in this 

 curve. 

 Curve II. (12th heating). H = 4*0. — This curve does not 



show any maxima. 

 Curve III. (12th heating). H = 12'0.— This curve and the- 

 preceding one are similar to that for iron in a strong field. 

 The results of the experiments can be briefly summed up 

 as follows : — 



(i.) When the specimen had not suffered much heating after 

 casting, the curves show that the permeability attains 

 a maximum value at three distinct temperatures between 

 the temperature of the atmosphere and the temperature- 

 at which the specimen loses its magnetic properties. 

 This effect is much more marked for weak than for 

 strong fields. 



For a field of 0*5 the substance is practically non- 

 magnetic at about 300° C. 

 (ii.) For weak fields the permeability increases with each; 



heating, 

 (iii.) The heating and cooling curves for weak fields are 



never coincident, 

 (iv.) The three maxima disappear gradually as the heatings- 



and coolings are continued, 

 (v.) A new maximum is developed between 500° C. and* 



600° C. 

 (vi.) For strong fields the three maxima first observed cannot: 



be detected after several heatings, 

 (vii.) The curves exhibit temperature hysteresis*. 



These curious and interesting results would indicate if the 

 original hypothesis is correct, that there are three distinct 

 magnetic substances, corresponding to the three maxima 

 observed, present in the specimen when first made. As the 

 heatings are continued these substances become modified in. 

 such a way that the maxima disappear. This may possibly 

 be due to these substances partially or wholly breaking down. 

 The new maximum which subsequently develops seems to- 

 indicate the formation of a new substance not originally 

 present in the specimen. 



The fact that the cooling and heating curves are always 

 distinct, and that the cooling curve is always above the 

 corresponding heating curve, might be taken to indicate that 

 dissociation takes place to some extent at high temperatures. 



* There was no evidence of temperature hysteresis in the impure 

 specimens first studied. 



