[ 37 ] 
n. The Specific Heats of Metals and the Relation of Specific Heat to Atvinic 
Weight.—Part II. 
By W, A. Tilden, D.Sc., F.K.S., Professor of Chemistry in the Royal College of 
Science, London. 
Received December 8,—Read December 11, 1902. 
In the Bakerian Lecture for 1900 (‘ Phil. Trans.,’ A, vol. 194, p. 233) it was shown 
that the specific heats ol very pure cobalt and nickel, when compared at temperatures 
fi-om 100° C. down to the boiling-point of licpiid oxygen, — 182°-5 C., steadily 
apjJioach each other and together tend towards a least value winch is at present 
unknown. 
It was thouglit desirable to increase the numljer of determinations at successive 
points on the tliermometric scale, and to extend the total range of the experiments so 
as to afford better data for calculation of tlie form of the curves. Tlie following is an 
account of the results obtained. 
It has not yet been possible to arrange for tlie conduct of experiments at 
temperatures lower than — 182°-5, as this could only be done with the aid of liquid 
hydiogen. The temperatures above 100° C. have been obtained bv the use of a hath 
of aniline vajiour, melted fusible metal or melted lead, and were estimated by the use 
of a platinum resistance thermometer which was carefully calibrated, and of viiich the 
fixed points 0°, 100°, and 184° (boiling-point of aniline) were verified. Tlie specific 
heats were determined m the same calorimeter and vith the same ju'ecautions as 
described in the Bakerian Lecture. 
The holder employed in the experiments at low temperatures was found equally 
useful in the experiments above 100°. Between air tenqierature and 100° the steam 
calorimeter was again employed. 
The figuies given in the following lable I. are in nearly all cases the mean values 
deduced from several experiments which were always closely concordant. 
The total amount of heat per unit mass measured in the calorimeter is equal to tlie 
pioduct of the mean specific heat and the range of temperature, beginning in eacli 
case at 15° C. Taking the values given in the table, this product, Q, may be jilotted 
as an ordinate, the liigher absolute temperature, t, being the abscissa. The result is 
Ihe nickel, cobalt, and platinum employed are the pure specimens prepared for the former series of 
experimente. Pure silver vas obtained from Messrs. John, sox and TlATTimY. For the aluminium I am 
indebted to Profe,ssor J. W. Mallet; it was described as nearlv pure. 
VOL. CC1.- --A 332. ‘ , 
Ji.o.Uo 
