Dr. P. Chappuis on Gas- Thermometry . 439 



might become slightly deformed on heating. In that case, 

 once the scratches on it are not strictly in the plane of the 

 neutral fibres, every variation in the curvature of the tube 

 involves a change in the distance apart of the marks. 



Under these conditions it would appear to me that 

 Mr. Bedford's results, in spite of their real value, should not 

 be accepted without reserve ; and consequently I cannot 

 wholly accept the conclusions of Prof. Callendar, who, after 

 having compared the coefficients determined by me between 

 0° and 100° for " verre dur " and porcelain with those ob- 

 tained by the method just described, is of opinion that my 

 values of /3 are too large. 



Bayeux porcelain having probably a composition differing 

 slightly from that of Berlin, I thought it worth while to 

 make a determination in the Fizeau apparatus on the spe- 

 cimen of Mr. Bedford's tube, which I had obtained. The 

 specimen as prepared had a length of about 15 mm. The 

 upper and lower plane surfaces were polished, and were very 

 nearly perpendicular to the axis of the tube. The accom- 

 panying figure represents the aspect of the fringes * of 



sodium light on this specimen in the first experiment made 

 at 83°*3. At each steady temperature an estimate was made 

 of the positions of each of the points A, B, C, D, E, F marked 

 on the lens with reference to the fringes, which were numbered 

 from the centre of the system outwards. 



The irregularity of section of the tube and the defects in 

 the surface as ground and polished, have rendered it impos- 

 sible to operate on a larger number of points of reference. 



The set comprising 33 separate observations, spaced out 

 over the interval 1° to 83° gave for the linear dilatation of 

 porcelain 



l t =l (l + (2824-1* + 6-m 2 )10- 9 ), 



* In the memoirs of M. Benoit on the Fizeau apparatus published in 

 Trav. et Mem. torn. i. & vi., will be found all details regarding the mea- 

 surement of dilatations by this ingenious method.' 



