370 Scientific Proceedings, Royal Dublin Society. 



wrought iron and of manganese steel (with 1236 per cent, of 

 manganese) deduced from Dr. Hopkinson's figures are as 1441 to 

 5-6, or 258 to 1. 



These numbers are fairly in accordance with those I have 

 obtained, viz. 330 to 1, for steel containing 1375 per cent, of 

 manganese. Considering the wide range of the figures in the table 

 given by Dr. Hopkinson, I should imagine that his ratio is some- 

 what less reliable than the one I have given. 



Other Magnetic Properties. — It was interesting to ascertain 

 whether the presence of this percentage of manganese in steel 

 deprived it of other well-known magnetic properties. As might 

 be expected, it showed no elongation under magnetisation. It did 

 not exhibit the magnetic tick or sound heard when iron, steel, 

 nickel, or cobalt is magnetized and demagnetized. A more in- 

 teresting question was whether it would exhibit the anomalous 

 expansion and after-glow which take place in iron or steel wire 

 when they cool to a certain critical temperature, after being 

 heated to whiteness. I have shown that these phenomena are 

 coincident with that temperature when the magnetic state of 

 these metals, destroyed by a high temperature, is resumed on cool- 

 ing. Careful experiments with the manganese steel wire, heated 

 to a bright whiteness, established the fact that no trace of this 

 anomalous deportment on cooling occurred with this substance. 

 Here then we have a singular and an important link between 

 the magnetic state of a body and its sudden and momentary ex- 

 pansion and reheating, when at the critical temperature. Like 

 manganese steel, the non-magnetic metals — platinum, copper, 

 German silver, silver and gold wire — do not exhibit this phe- 

 nomenon. 1 



I have good hope that the experiments here recorded will be- 

 come a starting-point for further investigation. When we remember 

 that 13 per cent, of a non-magnetic metal, mechanically mixed with 

 iron or steel, produces but a slight change in the magnetic state of 

 the latter, and then consider the profound magnetic change brought 

 about by 13 per cent, of manganese (itself a feebly magnetic metal) 



1 Nickel, I find, does not exhibit it, contrary to my expectation ; but the magnetic 

 state of nickel is lost at a temperature of 330° to 340° C, which is considerably below 

 red heat. Cobalt wire I have not yet been able to obtain. 





