250 1'KOFKSSOK W. A. TILDEN ON THK SPECIFIC HEATS OF METALS. 



Hence the absolute " atomic heat " of the two metals, cobalt and nickel, is almost 

 exactly 4. 



It appears probable, therefore, that if the experiments could t>e carried further the 

 specific heats would stand in exactly the inverse ratio of the atomic weights. 



It remains to be seen whether the value of the atomic heat for other metals 

 agrees with this.* If such turns out to be the case the original expression of the 

 law of Dui/ONG and PETIT, which is only roughly applicable at atmospheric 

 and higher temperatures, would be completely justified. 



In conclusion I desire to record my appreciation of the skilful assistance I have 

 received throughout these experiments from Mr. SIDNEY YOUNG. 



APPENDIX BY PROFESSOR JOHN PERRY, F.R.S. 



Dr. TILDEN has asked me to write the following short appendix to his paper. In 

 making the tedious calculations I was assisted by two of our students, Mr. E. R 

 VERITY and Mr. H. L. MANN. 



The law connecting the p, v, and t of unit mass of any substance in the solid 

 condition is not sufficiently well known for the general application of the laws of 

 thermodynamics. 



If we take it that there is an atomic or molecular specific heat l( of an elementary 

 substance which is constant in all states of the substance, and which is represented 



* Note added March 3. Experiments made since this paper was written, upon the metals silver, copper, 

 iron, and aluminium show that further investigation is necessary. The mean specific heat of silver, 

 for example, was found to be '0558 between 100 and 15, and -0519 between 15" and - 182'5. The 

 decrease of specific heat at the lower temperature is, therefore, much less than in the case of cobalt 

 and nickel. 



A paper by U. BEHN in the 'Annalen der Physik' (No. 2 for 1900, p. 257), issued on Feb. 16, did 

 not come into my hands till some time after the date of my communication to the Royal Society. My 

 attention was drawn by this paper to the fact that in 1898 the Author had commenced a series of 

 determinations of the specific heats of metals at low temperatures, and that about the same time some 

 experiments on the specific heats of three metals, copper, iron, and aluminium, at the temperature of 

 liquid air, had been published by C. C. TROWBRIDGE in the American periodical ' Science ' (N.S. 8, p. 6, 

 1898). 



The results of the latter cannot claim to be very important, as no information is given concerning the 

 composition of the metals, and there is great uncertainty about the temperature of the liquid in which 

 they were cooled. The numerical values given by TROWBRIDGE are considerably higher than those of 

 BEHN, as also in the one case, aluminium, in which a comparison can be made with my results ; the 

 mean specific heat between the temperature of air and that of boiling oxygen being -1833 according to 

 TROWBRIDGE, and '1676 according to my experiments. 



BEHN'S results are highly interesting, though for various reasons they must be inaccurate in some 

 cases and are probably to a slight extent inaccurate in all ; inasmuch as no special means were taken to 

 avoid access of heat in transferring the mass from the cold bath to the calorimeter, and as liquid air w;t.- 

 used instead of liquid oxygen, the temperature of the cooling liquid was somewhat uncertain. 



