98 L)KS. J. A. HAKKEK AND P. CHAPPUIS ON A 



in the base and aides to permit free circulation of the sulphur vapour within it.* It 

 is essential for the attainment of a constant temperature that the cone should be 

 sufficiently long to completely cover the resistance-spiral and a certain length of the 

 stem immediately above it. 



During the earlier experiments we had considerable difficulty with the sulphur 

 tubes owing to their liability to crack on re-heating, after having once been used. 

 We thus found it convenient, when making several consecutive sulphur point 

 determinations, to keep the sulphur just liquid between the different sets of 

 observations, by means of a small by-pass flame. The establishment of a constant 

 temperature in the sulphur apparatus takes a considerable time ; from half-an-hour 

 to an hour was generally allowed after insertion of the thermometer, t 



The sulphur we used was obtained from Messrs. BAIRD and TATLOCK, and was 

 made by CHANCE'S process. Though we made no chemical tests of its purity, we 

 have reason to believe that the impurities present, if any, exert practically no 

 influence on the boiling-point, as a large number of determinations made at Kew 

 showed no systematic difference in the behaviour of several different samples. Addi- 

 tional evidence of the purity of the sulphur used is afforded by the remarkable 

 steadiness of the temperature of the vapour, when once the equilibrium is established. 



Three independent values for the boiling-point of sulphur were obtained under 

 different circumstances. To the first of these, obtained from the preliminary com- 

 parisons of thermometer K.9 with the gas thermometer with reservoir of " verre 

 dur," we attach less weight than to the two subsequent ones, where K.8 and K.9 

 were compared with the gas thermometer fitted with the porcelain reservoir more 

 suited for high temperatures. 



We discuss the observations of the later series, taken with K.8, as an example of 

 the method of reduction followed. 



The determinations made with this thermometer of the platinum temperature of 

 the boiling-point of sulphur were eight in number, the corresponding pressures varying 

 from 755 to 762 millims. It is obvious that, from the experiments themselves, the 

 platinum temperature corresponding to 760 millims. could be deduced by the method 

 of least squares, but a formula for the variation of the boiling-point with pressure 

 deduced from so few experiments would, however, be liable to error. We, therefore, 



* This form of protector is due to HEYCOCK and NEVILLE, and is descril>ed in their paper in ' Trans. 

 Chem. Soc.,' 1895, p. 197. 



t In the use of this apparatus there are several precautions to be observed essential for good results. 

 The liquid sulphur in the Meyer tube must extend to some few centimetres above the base plate of the 

 apparatus. The gas burner should preferably be a largo solid-flame bunsen, and the flame should be 

 screened from draughts by asbestos-card or by a number of firebricks surrounding the apparatus. The 

 cones are attached to the thermometer by fine iron wire. The asbestos becomes very hard on cooling, but, 

 if, after use, the adhering sulphur is burnt off, the cones can lie rendered sufficiently pliable to serve for 

 several determinations. 



