250 Professor Fleming [Juno 5, 



The magnetic variety, which is mucli more brittle, is, however, in 

 this case formed by the prolonged slow heating of the non-magnetic 

 variety out of contact with air. In the non-magnetic condition the 

 material has a rather high specific resistance at 0° C, about 65,700 

 C.G.S. units per cubic centimetre ; but the magnetic variety has a 

 much lower specific resistance, viz. about 51,400 C.G.S. units at 0° C 



In all these cases it is interesting to note that the change of th 

 alloy into the magnetic variety is accompanied by a decrease in resi 

 tivity or increase in conductivity, and an increase in brittleness. 



We have tried cooling this non-magnetic variety of manganese- 

 steel in liquid air, but have not been able in that way to make any 

 change in its condition as regards magnetic suscej)tibility. 



There is a particular alloy, of copper 84 per cent., manganese 12 

 per cent., and nickel 4 per cent., called manganin, which at ordinary 

 temperature exhibits but little change of resistance with change of 

 temperature. On taking the curve of its resistance over wide ranges 

 of temperature we find that its curve is very concave downwards, and 

 the vertex of the curve lies at about 16° 0. Hence at ordinary tem- 

 peratures small changes of temperature make no change in its resis- 

 tance; but above that point its temperature coefficient is negative, 

 and below it it is positive. All alloys in which a negative tempera- 

 ture coefficient has been observed are probably instances of the 

 same mode of variation of resistance. It may be noted in passing 

 that the element manganese when present in an alloy seems to have 

 a great tendency to produce high resistivity and small temperature 

 coefficient. 



Returning then to the pure metals, we may ask, What is the mean- 

 ing of the fact that in their case the resistance lines all converge 

 so as to indicate that the electrical resistance would vanish at the 

 absolute zero of temperature ? 



We know that the passage, as we call it, of an electric current 

 through a conductor heats it, and that by Joule's law the rate of pro- 

 duction of heat in the conductor is proiDortional to the square of the 

 current' strength and to the total resistance of the conductor. 



Suppose we take two wires, say of iron and a certain copper- 

 nickel-aluminium alloy having the same resistivity at 100° C. and of 

 the same size and length. These wires will at -j- 100° C. have the 

 same resistance. A given current flowing through them will therefore 

 generate heat in them both at the same rate. 



Cool them both down, however, to the temperature of liquid air. 

 In the case of iron-wire the resistance is reduced to one-fifteenth of 

 its value at — 200° C, in the other case it is reduced by only 10 per 

 cent. Hence, at the low temperature the alloy dissipates energy for 

 the same current 13^^ times as rapidly as the pure metal. 



It is a logical deduction from all we know to conclude that if we 

 could reach the absolute zero of temperature the j)ure metal would 

 not dissipate the energy of the current at all. Imagine two iron 

 wires, then, stretched through space, say from the earth to the moon, 



