73 
1912-13.] Magnetic Induction in Ferric Oxide. 
two. On the average, there was a maximum at 500°, a minimum at 650°, 
another maximum at 750°, and a minimum at 1000°, but these are by no 
Double 
Dcf-let-Ko t?. 
o 
Average Curve for 5 bars of Ferric Oxide Seated iq Air, 
Heat 
i n«i 
CooiinQ rr> © 0 
Tem perature 
C. 
means distinct, as the whole effect is very small. The average curve for 
these five bars is given. The bars were made up as follows : 
(1) Yellow oxide mixed with alcohol and dried. 
(2) Red oxide mixed with alcohol and dried. 
(3) Bar of compressed red oxide. 
(4) Red oxide powder in glass tube. 
(5) Brown oxide powder in copper tube. 
As a rule, the bars were non-magnetic when cooled down after heating ; 
but the majority of them showed a slight maximum when cooling 
through 400°. 
Bars of Haematite Ore. (See Curves 2 and 3.) 
When heated in air, bars of haematite ore containing about 60 per cent, 
of iron also exhibited an increase of magnetic quality. 
Curves 2 contain the results for heating fresh bars of red Seriphos, 
Lucinena, and brown Seriphos in the furnace-magnetometer. The behaviour 
of these bars seems irregular, but it may be noted that all three curves 
agree in passing first through a low maximum and afterwards through a 
higher one, and it would appear that the same succession of changes had 
gone on in all three bars, but at different rates. No importance is placed 
on the absolute value of the deflections at present, because the three bars 
were of different weights and were heated for different lengths of time. 
When the hsematite bars were cooled in the furnace-magnetometer, they 
followed a course similar to that given for Luc. 1 in Curves 3 ; but no 
accurate data are at hand to illustrate this behaviour, the cooling curves 
having been affected by furnace trouble and by coil effect. 
