Lt'iiiith of Fe}'vow(((j}ietic Suhsfances hi/ Magnetization . .'ilJ5 



the exposed })nrts were well covered with fibrous asbestos 

 and cotton in order to diminish the loss of the oenerated heat, 

 eart^ beino- taken to pi'oduce no sensible resistance to tlie 

 elongation or contraction of our specimens. 



For the measurement of the change of length in liquid air 

 the above arrangement was modified in the following way: — 

 The specimen was stretched upwards by means of a spring- 

 instead of stretching it by a suspended weight. The shaded 

 ])ortions of the protruding supj)()rter were made of brass, 

 while the unshaded })ortion was made of wood to lessen the 

 conduction of heat. Care was specially taken to stretch the 

 copper wire in the direction of the axis of the rod. 



The magnetizing coil with a Dewar's tube inside it was 

 placed coaxially with the rod. The contraction or elonga- 

 tion of the specimen was measured by a rotating cylinder in 

 contact with a vertical copper wire. The temperature of the 

 liquid air was assumed to be —186° C. 



§ 2. Metliod of Observation. 



The experiments at low temperatures were conducted in 

 the following manner: — The change of length at the tempe- 

 rature of the room was first determined and compared with 

 the corresponding result obtained by the heating arrangement. 

 The comparison showed that these two results nearly coincided 

 with each other. The hquid air was then poured into the 

 Dewar^s tube containing the specimen, and the exposed parts 

 round the magnetizing coil were carefully protected with 

 cotton. Owing to the violent evaporation of the liquid a 

 small oscillation of the image in the telescope was first ob- 

 served, but after about 15 minutes the image became steady. 

 The magnetic change of length was then measured in the 

 usual way. 



The experiments at high temperatures were undertaken in 

 the following order : — The change of length at the tempera- 

 ture of the room was first determined. Then an electric 

 current from a dynamo was passed through the heating-coil 

 for one or two hours till the temperature of the core became 

 very constant: then the current from the dynamo was replaced 

 by one from an accumulator in order to get rid of the fluctua- 

 tion of temperature due to irregularities in the dynamo-current. 

 Twenty or thirty minutes passed before the observations could 

 be taken, when the temperature had become very steady. Th(^ 

 change of temperature caused a deflexion of more than 6 cms. 

 2)er degree in the field of the observing-telescope, whereas the 

 deflexion due to maonetization was almost 5 cms. in nickel 



