246 Proceedings of the Royal Society of Edinburgh. [Sess. 
scale is from 0° C. to 200° C., especially about 150° C. ; after that, further 
rises of 50° C. produce no sensitive state till a temperature of 400° C. 
is reached, after which further successive rises of 50° C. produce sensitive 
state of about 5 per cent. 
Fig. 1 (ii.), showing the effect of successive rises of 100° C., gives 
exactly the same division of the scale into susceptible and non-susceptible 
parts — the greatest effect being produced in the interval 100°--200° C., 
which induces a sensitive state of over 40 per cent. 
Fig. 2 (i.) shows curves (a), (b), and (c) — (a) showing sensitive state due 
to rise, ( b ) sensitive state due to fall, and (c) sensitive state due to rise and 
fall of temperature, (a) shows a marked maximum about the region of 
200° C., and lies above both (b) and (c) up to a temperature of 300° C., when 
it crosses (b) and finally lies below (c) also ; (b) remains pretty steadily 
between 20 per cent, and 25 per cent, for all temperatures ; and (c), which 
is below (b) for all temperatures, rises rapidly at first, then slowly till 
a temperature of 400° C. is reached, when it falls somewhat, then rises 
once more. 
Hard Steel. — Fig. 1 (iii.) shows the effect of a rise of 50° C. to be, as in 
the previous specimen, greater for the rise from 100° to 150° C. than at any 
other part of the scale, though the effect in this case is only a little more than 
half as great ; but in this specimen, on the other hand, a rise of 50° C. at any 
part of the temperature scale, except above 500° C., produces some effect. 
The effects of rises of 100° C. are somewhat similar to, but less marked than, 
those due to rises of 50° C. 
The (a), (6), and (c) curves for this specimen are somewhat similar in 
general form to those of the last, but the maximum in the (a) curve does 
not appear till aoout 400° C., and is much less marked, the maximum value 
reached being only 21 per cent, in this case, as compared with 38 per cent, 
in the other. Indeed, the sensitive state induced in this specimen by all 
varieties of thermal treatment employed is much less marked than in the 
case of medium carbon steel. 
Steel Wire . — The effect of increasing the temperature by 50° or 100° C. 
is still less marked in this case than before, but the greatest effects still 
show at 100-150° C. and 100°-200° C. The (a), (6), and (c) curves are 
again somewhat similar, but with much less definite turning-points ; and 
the (a) curve, so little is its maximum marked, never rises above the (b) 
curve, which lies throughout the range below the 15 per cent. line. 
Cast Iron. — The effect of increasing the temperature by 50° or 100° C. 
is, in the case of cast iron, greater than for steel wire, but not so great as 
for hard steel. This specimen, however, shows one point of difference from 
