108 Scientific Proceedings, Royal Dublin Society. 



of manganese steel are, before heating, undoubtedly considerably 

 more magnetic than the bar from which they are bored. This 

 cannot be wholly due to abrasion ; for, whilst filings of a 12 or 

 even 16 per cent, alloy are feebly magnetic, filings of 3G per cent, 

 are more magnetic. Heating the feebly magnetic filings of the 

 18 per cent, alloy to redness in an open platinum crucible, they 

 became black with oxide, and when cold were strongly magnetic, 

 being readily attracted by a small permanent magnet. 1 Believing 

 the attraction to be due to the formation of an oxide of manganese, 

 thus leaving the steel in its ordinary unalloyed state, I covered 

 fresh filings with a deep layer of powdered charcoal, and heated 

 them to redness in .a crucible : strong attraction still ensued on 

 cooling ; the filings became of a gray colour, showing less oxidation 

 than before. I then tried heating the filings in an atmosphere of 

 hydrogen. They were enclosed in a hard glass tube, and after 

 hydrogen had passed over for some time the tube was sealed, 

 and the filings heated to redness. Even here a slight change 

 of colour occurred, and a powerful attraction ensued when the 

 filings were cool. Numerous experiments were made to exclude 

 every trace of oxygen, and at last scarcely any discolouration of 

 the filings was obtained : nevertheless, a magnetic character was 

 conferred by the heating. These results were confirmed by the 

 experiments of Professor O'Shea, of Firth College, Sheffield, who 

 very carefully weighed the filings, before and after they had been 

 heated in hydrogen, and ultimately obtained a result where no 

 gain in weight arose on heating ; nevertheless, a higher magnetic 

 state was found on cooling. 



Experiments with a 36 per cent, alloy of manganese, 3 per 

 cent, carbon, and 60 per cent, iron, gave even more striking re- 

 sults. There was no sensible attraction with a permanent magnet 

 before heating, but a powerful attraction after heating. 



I now compared the behaviour of manganese wire and foil with 

 that of the filings. 



Suspending a fragment of manganese steel foil over and very 

 near the pole of a powerful electro magnet, I obtained an attrac- 



1 The magnetic property of iron and steel, as is well knowD, is destroyed at a bright 

 red heat, but is recovered on cooling ; but a magnetized body loses its magnetized state, 

 without recovery on cooling, when heated to redness. 



