1912-13.] Sensitive Magnetic State induced in Various Ways. 249 
off steadily, the (6) curve differs considerably, as, instead o£ maintaining a 
fairly steady value, it rises from 10 at 100° C. to a maximum of over 20 at 
400°, then falls and rises again. The (c) curve lies entirely below the (b) 
curve, and is similar to it in form. 
Tungsten Steel . — This specimen, with which we leave the carbon series, 
displays some very decided differences in behaviour from those previously 
examined. 
First of all, the specimen shows much less sensitive state than any of 
the others, and the most susceptible part of the scale is now about 300° C. 
— from 300° to 350° C. for rises of 50° C., and from 200° to 300° C. for rises 
of 100° C. Also for low temperatures the specimen is very insensitive, rises 
of 50° having no effect in producing sensitive state below the 300° to 350° C. 
rise, while, in rises of 100° C., 0° to 100° C. has no effect ; the maximum 
sensitive state produced in any of these intervals is less than 2 per cent. 
Of the (a), (b), and (c) curves, the (b) curve shows much the more important 
effect, rising from zero for a fall of temperature from 100° C. to room 
temperature to 8 per cent, for a fall from 600° C. to room temperature ; 
while the (a) curve, beginning to rise at the same point, never rises above 
3 per cent. The (c) curve, again, only begins to rise after a rise and fall of 
more than 300° C., and it also never rises above 3 per cent. 
Coming now to compare the results given us by these five specimens, 
we note first of all the very great difference between the four belonging to 
the carbon group and the tungsten specimen. Not only is the sensitive 
state induced in the latter much less marked, and the most sensitive part 
of the temperature scale much higher up, but rises of temperature at the 
beginning of the scale have no effect at all. 
Between the different members of the carbon group very strong resem- 
blances appear. The part of the scale which is most pronounced in 
producing sensitive state is in each case very nearly the same, being only 
a very little later in the case of cast iron. It would appear, then, that the 
addition of a marked quantity of carbon tends to raise the temperature 
of maximum susceptibility, though the addition of smaller quantities has 
no appreciable effect. 
Considered generally, the specimen of the four showing the least 
sensitive state is the specimen of steel wire, while that showing most 
is the specimen of medium carbon steel. As the percentages of carbon in 
the two are very nearly the same, *755 per cent, and ’80 per cent., it is 
probable that the smaller sensitive state is connected with the greater 
percentage of manganese in the steel wire specimen. The specimen of hard 
steel shows next greatest effect, and cast iron comes third in the list. 
