138 PRINCIPAL J. D. FORBES ON AN EXPERIMENTAL INQUIRY 



quite complete. The effects of my illness continuing, I contented myself during 

 the succeeding summer (1852) with deducing the best results I then could from 

 the data which I had obtained, trusting to make a more thorough examination of 

 them when the Errors of the Thermometers should be accurately known. The 

 various curves of temperature, velocity of cooling, &;c., were most carefully pro- 

 jected on a large scale, and the different stages of reduction carried out precisely 

 as I had projected in my first letter, printed above. The smoothness and regu- 

 larity of the primary curves (statical and dynamical), of Arts. 5 and G, left 

 little to desire. The values of the subtangents and various derivative quantities 

 were equalized by the aid of graphical interpolation in the manner so ably used 

 and recommended by Sir John Herschel, in several of his researches. As every 

 step of the reductions was subjected to the test of graphical representation, it is 

 not likely that any material error occurred in the calculation. 



14. In October or November 1852 (at Clifton, where I then resided), I brought 

 these provisional calculations to a close, with the results which I shall presently 

 state. I did not abandon the idea of a more rigorous reduction of the observa- 

 tions, and I was fortunate enough to induce the late Mr Welsh, of the Kew Ob- 

 servatory, to undertake the examination of the scales of the principal thermometers 

 used, and especially of that one employed for the highest temperatures, in which 

 a portion of the column being detached, for the purpose of extending the range, 

 the verification was a matter of some delicacy. Mr Welsh's tables of corrections 

 are now in my hands. The application of them will involve some labour, and 

 perhaps the entire reconstruction of the interpolating curves, and a repetition of 

 the calculations depending on them. I confess myself to blame in allowing so 

 many years to elapse without bringing these computations to a satisfactory con- 

 clusion, as well as for not extending the observations to other substances, as 

 originally proposed. The state of my health has not been favourable for such 

 tasks, which are necessarily of an irksome and tedious kind. I hope still to 

 execute, at least the desirable revision of the temperatures and reductions. But 

 upon looking again (after some years' interval) at the previous calculations and 

 their results, I feel warranted in publishing both the methods and the conclusions, 

 as worthy of considerable confidence, and as still I believe new. Indeed, I incline 

 to think that I have been perhaps too fastidious in withholding the results ob- 

 tained, until they should have received the utmost precision which I can give them. 



15. As I entertain the idea of publishing a supplement to this paper, with cor- 

 rected details, I shall not now dwell upon the intermediate steps of the inves- 

 tigation. I shall, however, insert (chiefly from my full journal-notes of 1851) 

 such a description of the apparatus and modes of observation, as may, I hope, 

 materially assist any one in pursuing the experiments for other metals from the 

 point where my labours unfortunately terminated ; as it is not very probable 

 that, under present circumstances, they will be resumed by me. 



