77 
1912-13.] Magnetic Induction in Ferric Oxide. 
iron during heating, since they have the minimum point of iron and not 
that of magnetite. 
The above curves differ from those for heat cycles in air in their extent 
of coincidence. There is here a definite magnetic effect which repeats itself 
when the test is repeated with a fresh bar. The magnetic effect is large, 
and is confined within comparatively narrow temperature limits, viz. 300° 
Double 
De Ret ho a . 
to 600° C. As these curves and their underlying changes appeared to 
follow more closely than any of the others the lines of chemical research 
on this subject, a new series of tests was instituted on the heat cycles in 
hydrogen, the details of which we hope to publish later. This new series 
has brought out with distinctness a small maximum at 425° and a higher 
one at 575°, followed by the minimum at 625° C. We consider this first 
maximum to be the same as that referred to under Curves 2 and 3 ; it also 
appears distinctly in Curve 5. 
2. Constant High Temperature in Hydrogen. (See Curves 8 and 9.) 
The two groups of curves given show, in one case, the effect of heating 
ferric oxide bars made with alcohol and dried, and in the second case the 
effect of heating bars of Bacares ore. 
With the artificial oxide the rise of the curve began 15 minutes 
after heating had commenced, and a moderately steady state was soon 
reached. No magnetic effect was obtained until the temperature rose 
above 250°. 
With the haematite bars the rise was more gradual, but it reached a 
higher maximum. This was due in part to the mass of the haematite bars 
being roughly ten times as great as that of the bars of ferric oxide. In 
