Abstract
A more complete set of observations of the relative coefficients of expansion of mercury in silica than those previously published are obtained by the use of an electrically heated oil bath. The observations comprise readings at frequent intervals up to 300°C., and are in good agreement with the earlier observations, which, when applied to the values of the absolute expansion of mercury given by Callendar and Moss, seemed to show that the cylindrical bulbs used were not isotropic. Further experiments made with a spherical bulb negatived this conclusion. Tables are included in the Paper giving representative observations and the final results. From the values of the coefficients of expansion of silica determined by Prof. Callendar upon specimens similar to those from which the bulbs were made, the coefficients of absolute expansion of mercury are calculated. These over the range of the experiments of Chappuis, who used a similar method, but with bulbs of verre dur, are in fair agreement with his results, but differ appreciably from the results of Callendar and Moss obtained by the absolute method. It seems, therefore, that there is some undiscovered fundamental difference in the two methods.