of Edinburgh, Session 1880-81. 127 
wenigstens innerhalb der bei den Beobachtungen vorkommenden 
Temperaturgranzen, nicbt bedeutend sind, ...... so miissen 
dieselben den Werth des Warmecoefficienten nur unbedentend 
afficiren konnen.” In my paper on Thermal and Electric Conduc- 
tivity (Trans. K.S.E., 1878), I said that “the change of specific heat 
with temperature would increase the values of h at higher tempera- 
tures, and thus reduce the change in conductivity in iron, and 
increase the small changes indicated for the other substances.” But 
I had not at hand the means of applying these corrections. Becent 
discussions as to the comparative merits of different experimental 
methods have led me to investigate the amount of this effect, by the 
aid of the best data I could procure. A comparison of these seems 
to leave no doubt that the specific heat of iron increases by some- 
where about T -jjo of its amount for each degree of rise of tempera- 
ture ; at least from 0° to 300° C., between which limits the investi- 
gations of conductivity have hitherto been carried on. 
Besides this result, which I have gathered from various scientific 
journals, I may adduce from my Laboratory Book for 1868 the 
following determinations : which were made with great care by the 
late Mr J. P. Mchol, by means of the method of mixtures. The 
nature of the process employed is such that the results must all err 
in defect, and the more so the higher the temperature. The iron 
was heated sometimes in oil, sometimes in paraffin. 
Specific Heat of Iron. 
15° to 100° C., 
15° to 150° C., 
15° to 200° C., 
15° to 250° C., 
15° to 300° C., 
QT154 \ 
0-1127 I 
0-1158 j 
0-1168 ) 
0-1193] 
0-1189 ' 
0-1186 ) 
0-1208 
0T214 
0T218 
0-1234 
0T240 
0T274 
0T276 
Mean. 
0T152 
0-1189 
0-1213 
0-1237 
0-1275 
VOL. XI.' 
