Dr. E. J. Mills on Thermometry. 143 



mometers. I have given a table of such agreement-points (N) 

 in the case of four thermometers ; and if the stated values of 

 N are substituted in the four equations (also given) , x has in 

 all cases the same value. This allegation I must distinctly 

 repeat ; but I understand Professors Thorpe and Eiicker to 

 say that it is unfounded. 



On page 3 of the authors' commentary, a calculation is 

 given from which it is inferred that " the exposure corrections 

 below 100° C. are, in the case of three out of four similar 

 thermometers, practically identical." The calculation, how- 

 ever, takes 100° C, and not any lower temperature ; and it 

 happens that the correction-factors are then rather close 

 together, while at lower temperatures (to which no reference 

 is made) they are much further apart. Thus the " favourable 

 suppositions " are only favourable at 100° C. to my critics, 

 with whom, at that point, I entirely agree. At lower tempe- 

 ratures their inference is clearly wrong. 



[It may be of interest for me to add that, with regard to 

 observational power, I found myself, when in great practice 

 with scale -reading, capable of estimating very fairly ^V milli- 

 metre. Verifications of this result were not unfrequently 

 made by kathetometric comparison ; but it involves a higher 

 grade of accuracy than the Leeds professors seem disposed to 

 allow.] 



My critics proceed thus : — " it is impossible for the 



correction for a thermometer with an exposed column 166 di- 

 visions long to be equal to that of another when no part of the 

 column is exposed." But my memoir contains no such state- 

 ment. They have here mistaken the "correction-factor" for 

 the "correction," which is a very different thing. A correc- 

 tion-factor may be attributed, with perfect possibility, even to 

 an unexposed thermometer. 



The authors next point out a real error, which I readily 

 admit. Among some thousands of calculations of which my 

 paper contains the results, it appears that a numerical blunder 

 has been committed. Instead of multiplying by 4, in a par- 

 ticular conversion, I have actually divided by 4. The equa- 

 tion containing this error has never once been used by myself, 

 and would instantly excite the suspicion of the most careless 

 person who employed it. The mistake is thus of no conse- 

 quence whatever ; but more than a page of the Philosophical 

 Magazine is gravely devoted to its discussion. 



2. Zero Movements. — The term " ascent " of a zero has been 

 used by me to signify its movement. Sign has been always 

 specified, or at least indicated. This is a common, and indeed 

 a necessary practice in physical writings. If it has led to 



