274 



NATURE 



\yan. 23, 1890 



zero, we may in this state try any experiment we please. 

 Reversing the current on the ring, we shall find that it is 

 in all cases non-magnetic. Suppose next that we allow 

 the ring to cool without any current in the primary, when 

 cold we find that the ring is magnetized ; in fact, it has a 

 distinct recollection of what had been done to it before it 

 was heated to the temperature at which it ceased to be 

 magnetic. When steel is tried in the same way with 

 varying temperatures, a similar sequence of phenomena 



Fjg. 9. 



Fig. 10. 



Fig. II. 



is observed ; but for small forces the permeability rises to 

 a lower maximum, and its rise is less rapid. The critical 

 temperature at which magnetism disappears changes 

 rapidly with the composition of the steel. For very soft 

 charcoal iron wire the critical temperature is as high as 

 880" C. ; for hard Whitworth steel it is 690" C. 



The properties of an alloy of manganese and iron are 

 curious. More curious are those of an alloy of nickel 

 and iron. The alloy of nickel and iron containing 25 per 



cent, of nickel is non-magnetic as it is sure to come from 

 the manufacturer ; that is to say, a substance compounded 

 of two magnetic bodies is non-magnetic. Cool it, how- 

 ever, a little below freezing, and its properties change : 

 it becomes very decidedly magnetic. This is perhaps 

 not so very remarkable : the nickel steel has a low critical 

 temperature — lower than we have observed in any other 

 magnetizable body. But if now the cooled material be 

 allowed to return to the ordinary temperature it is mag- 

 netic ; if it be heated it is still mag- 

 netic, and remains magnetic till a tem- 

 perature of 580° C. is attained, when 

 it very rapidly becomes non-magnetic, 

 exactly as other magnetic bodies do 

 when they pass their critical tempera- 

 ture. Now cool the alloy : it is non- 

 magnetic, and remains non- magnetic 

 till the temperature has fallen to 

 below freezing. The history of the 

 material is shown in Fig. il, from 

 which it will be seen that from — 20° C. 

 to 580^ C. this alloy may exist in either 

 of two states, both quite stable — a 

 magnetic and a non-magnetic — and 

 that the state is determined by whether 

 the alloy has been last cooled to —id" 

 C. or heated to 580° C. 



Sudden changes occur in other pro- 

 perties of iron at this very critical tem- 

 perature at which its magnetism dis- 

 appears. For example, take its elec- 

 trical resistance. On the curve. Fig. 

 12, is shown the electrical resistance 

 of iron at various temperatures, and 

 also, in blue, the electrical resistance 

 of copper or other pure metal. Ob- 

 serve the difference. If the iron is 

 heated, its resistance increases with 

 an accelerating velocity, until, when 

 near the critical temperature, the rate 

 of increase is five times as much as 

 the copper ; at the critical tempera- 

 ture the rate suddenly changes, and it 

 assumes a value which, as far as expe- 

 riments have gone, cannot be said to 

 differ very materially from a pure 

 metal. The resistance of mangan- 

 ese steel shows no such change ; 

 its temperature coefficient con- 

 stantly has the value of o'ooi2, 

 which it has at the ordinary tem- 

 perature of the air. The electrical 

 resistance of nickel varies with 

 temperature in an exactly similar 

 manner. Again, Prof. Tait has 

 shown that the thermo-electric 

 properties of iron are very ano- 

 malous — that there is a sudden 

 change at or about the tempera- 

 ture at which the metal becomes 

 non-magnetic, and that before this 

 temperature is reached the varia- 

 tions of thermo-electric property 

 are quite different from a non- 

 magnetic metal. 

 Prof. Tomlinson has investigated how many other pro- 

 perties of iron depend upon the temperature. But the 

 most significant phenomenon is that indicated by the 

 property of recalescence. Prof. Barrett, of Dublin, ob- 

 served that if a wire of hard steel is heated to a very 

 bright redness, and is then allowed to cool, the wire will 

 cool down till it hardly emits any hght at all, and that 

 then it suddenly glows out quite bright again, and after- 

 wards finally cools. This phenomenon is observed with 



