■28 



Mr. J. E. Ashworth. 



[Jan. 30, 



magnet, the self-demagnetising factor being governed by the dimension 

 ratio. In general, an increment of temperature in the cyclic state 

 reduces and a fall of temperature augments the magnetic intensity, so 

 that a in the equation 



is negative, if being a greater temperature than t and l t > and l t the 

 corresponding magnetic intensities. But it was discovered that 

 magnets made of pianoforte wire in the commercial state, more than 

 50 diameters long, exhibited a positive or incremental coefficient. If 

 the magnet were much shorter the coefficient was decremental, and, 

 for an intermediate dimension ratio, zero. 



In the present paper a complete experimental investigation of the 

 temperature coefficient of a magnet is undertaken. Attention is at 

 first confined to pianoforte steel wire, and tables and curves are given 

 of the change of a, and the change also of I, as the dimension ratio is 

 increased from about 16 to 100 for the wire in three conditions, 

 namely, annealed, glass hard, and commercial drawn state. 



In the first two the coefficient is decremental and, in magnitude, 

 large for annealed and very small for glass hard ; but in the drawn state 

 a changes from decremental to incremental as the dimension ratio 

 advances. It thus appears that the drawing is responsible for the 

 peculiar behaviour of a. 



To confirm this, experiments were made on twelve samples of 

 pianoforte steel, representing every stage in the process of drawing 

 from the rolled rod through annealing and tempering to the fine- 

 drawn wire, the drawing being carried one or two steps farther than is 

 usual in commercial practice. The experiments were made on lengths 

 of 50 and 100 diameters. The curves which are traced show very 

 clearly that a changes in sign at a not very advanced stage in the 

 drawing, grows more positive, reaches a maximum, and then droops 

 towards zero again at extreme drawing. The magnetic intensity rises 

 steadily with drawing, and shows no tendency to diminish even at the 

 final stage of drawing ; it is then about 200 per cent, greater than at 

 the first drawing. 



An examination of a piece of drawn steel wire about 400 diameters 

 long then follows, in which susceptibility and intensity are determined 

 when the wire was at air temperature and when it was at 100° C. 



The effect of heating is very pronounced and, unlike the general 

 behaviour of iron and steel, the hot curve of magnetisation is always 

 -above the cold curve, as the steel is carried to its greatest intensity 

 -and brought to its residual condition by the withdrawal of all force. 

 Thus it retains more magnetism hot than cold. But on demagnetising, 

 the hot curve droops faster than the cold curve, and the curves 

 intersect when the reversed force is about 4 C.G.S. units, and at this 



