184 Messrs. Thorpe and Riicker's Remarks on 



work; but I confine myself here to what has been actually 

 done. It is hoped that the difficulties are now for the most 

 part overcome, and that the attainment of further results is 

 not far distant. 



XXII. Remarks on Dr. Mills's Researches on Thermometry. — 

 No. II. By T. E. Thorpe, Ph.D.. F.R.S., and A. W. 

 Rucker, ALA.* 



IN a criticism on Dr. Mills's Researches on Thermometry, 

 published in the July number of the Philosophical Maga- 

 zine, we stated that we believed that he had committed an 

 error which " upsets all the inferences drawn by him from his 

 formulae " relating to the movement of the zero of a mercu- 

 rial thermometer with changes of temperature. In his reply, 

 Dr. Mills informs us that he has " so far failed to follow " the 

 section of our paper in which this remark occurs, " that he is 

 unable to discuss it„" It is evident that the question at issue 

 between us cannot thus be left unsettled. We purpose there- 

 fore briefly to notice some of the points on which Dr. Mills 

 has replied to us, and to attempt to make still clearer those 

 which he disregarded as unintelligible. 



Exposure Correction. — Dr. Mills is wrong in thinking that 

 we questioned the accuracy of his statement that the " correc- 

 tion-factors " for his thermometers were the same at certain 

 temperatures named by him. What we stated (Phil. Mag. 

 July 1881, p. 12) was, that his method of presenting his results 

 " tends to exaggerate the importance of the variation of the 

 corrections." To this we adhere — and the more strongly, 

 inasmuch as Dr. Mills's immediately succeeding remarks 

 afford an example of the confusion introduced into the discus- 

 sion of the subject by his habit of dealing, not with the quan- 

 tities which it was his object to determine (e. g. corrections, 

 positions of the ice-point), but with subsidiary quantities (cor- 

 rection-factors, total remaining ascents), by which they are 

 only indirectly measured. 



We stated that, in our opinion, Dr. Mills's experiments 

 prove that the " exposure corrections [expressed by y=z 

 (a + /3X)(T-0N] below 100° C. are in the case of three out 

 of four similar thermometers practically identical ;" and we 

 supported our view by an example calculated for 100° C. 

 Dr. Mills retorts that our inference is clearly wrong, because 

 at lower temperatures the differences of the correction-factors 

 (i. e. of a + /3N) are greater than at 100°. This we fully 



* Communicated bv the Authors, 



