72 
Proceedings of the Royal Society of Edinburgh. [Sess. 
place, and also what was the optimum temperature for this change. 
Further, the cooling curve indicated the final state of the bar. A cycle 
was seldom completed, as a rule, under two hours, and often required 
three to four hours. 
2. Constant High Temperature . — The bar was placed in the furnace- 
magnetometer, and its temperature was raised to a given value and kept 
constant for one, two, or more hours, during which its magnetic condition 
was noted at five-minute intervals by taking the double deflection due to 
the reversal of a known magnetic field. 
3. Heating in Gas Furnace . — In this method the bar was placed in a 
gas furnace, raised to a temperature of 850° C., and maintained at this 
temperature for ten minutes. It was then cooled, and afterwards examined 
in the magnetometer when cold. This treatment was repeated many times 
until the magnetic behaviour of the bar became steady. 
Weiss and Foex,* when preparing artificial magnetite from ferric oxide, 
enclosed the oxide in platinum, and heated it in an oxyhydrogen flame, 
and afterwards sealed the platinum in the flame to prevent re-oxidation 
on cooling. 
In our preliminary experiments it was observed that when a bar of 
haematite ore was heated to 850° in a gas furnace and cooled, there was 
always a change from the non-magnetic to the magnetic state, although no 
precautions had been taken to prevent chemical change. It was therefore 
considered that a series of heatings for short intervals might enable 
us to note the progressive changes which took place in the magnetic 
state of the bar of haematite. 
Experimental Results. 
These are given entirely in the form of curves. 
I. Ferric Oxide Bars heated in Air.f (Curves 1 to 6.) 
1. Heat Cycles. (See Curve 1.) 
In the case of five “bars” of ferric oxide heated in air to 1000° C. the 
deflections obtained were small. All the bars were initially non-magnetic 
at air temperature in a field of 43 C.G.S., and the deflection began to grow 
at a temperature of 100° C. for three of the bars, and at 200° for the other 
* Archives des Sciences , xxxi. p. 89, Feb. 1911. 
t The air here referred to means the air of the furnace in which the bars were heated. 
Usually the electric furnace was stopped at one end with wool and obstructed at the other 
by the thermo-couple. The gas furnace was open to the air at both ends. 
