Br. P. Chappuis on Gas- Thermometry . 



Ail 



mentioned, we find values which are by no means identical. 

 The following example relates to the two series of obser- 

 vations at 82° and 3° respectively : — 



Difference of the individual estimations from the 

 means M x and M 2 = 



Grot 



ap at 82°. 



Group 



at 3°. 



Difference. 





\ 





X 



X 





2" 





2" 



2' 



A-M x 



-1-89 



A-M„ 



-172 



-0-27 



B-M x 



-1-53 



B-M., 



-1-47 



-006 



C-Mj 



-061 



C-M 



-0-52 



-009 



D-M x 



-0-35 



D-M., 



-0-39 



4-0-04 



E-M t 



+1-30 



B-M, 



+1-20 



4-040 



F-M x 



+3-10 



F-M 2 



+2-92 



4-0-18 



M x = 



=22-62 



M 2 = 



:059 





The observations made at intermediate temperatures con- 

 firm the rotation of the fringes indicated by the two series 

 just quoted; and since, in the course of the observations, the 

 direction of the temperature-change was varied frequently, 

 the fact just mentioned seems well established. 



The inequality in the dilatation of the two opposite sides of 

 the tube necessarily involves a change in its form. The 

 tube curves on expansion, which is probably the cause of the 

 divergence found between Mr. Bedford's results and mine. 



The conclusions of the study made by Dr. Harker and 

 myself of the thermometric scale of the platinum-resistance 

 thermometer and the sulphur-point do not appear to me to be 

 invalidated by the results collected by Prof. Oallendar, and 

 I think we are justified in adhering to them, whilst recog- 

 nizing at the same time the uncertainty which attaches to an 

 application of the coefficient of expansion of the reservoir 

 over an interval about four times as great as that in which it 

 was determined, and where alone it is strictly applicable. 



Consequences of a Change in the Sulphur- Point. 



The modifications resulting in the adoption of our new 

 value for the boiling-point of sulphur — namely T s = 445°'2, 

 instead of the value 444°'53, as obtained by Oallendar and. 

 Griffiths, are shown in the following table. 



