Art. LX.—On the Error introduced by using a Coal-gas Flame while determining the Percentage of Sulphur in Coals, with Especial Reference to the Methods “Eschka”^* and “Nakamura.”^†.

By Dr.W. P. Evans.

[Read before the Philosophical Institute of Canterbury, 1st September, 1897.]

The total sulphur in coals is generally determined, at present, by some modification of one of the above methods. In each case an error is introduced if gas be used to heat the crucible instead of spirit, but it is only small if the gas complies with the sulphur clauses of the Metropolitan Gas Referees, and is, for technical purposes, often neglected.

As a gas flame is, if allowable, much more handy in use than a spirit flame, it seemed advisable to determine the error caused by using the fairly sulphurous gas supplied in Christchurch. For this purpose 6 grammes of pure sodium carbonate (10 gr. gave no trace of a barium-sulphate precipitate) were heated over an ordinary gas Bunsen for three hours, and then treated exactly as in a regular sulphur determination. The barium-sulphate precipitate obtained weighed 0·0057 grammes, equivalent to 0·00078 grammes of sulphur. As the average amount of coal taken for a sulphur determination is 1·5 grammes the error introduced is slightly over 0·05 per cent., and is evidently negligible in ordinary technical analyses.

As, moreover, four separate test experiments varied only by 0·004 per cent., it is evidently quite allowable, even when using a coal-gas fairly rich in sulphur, to make a blank experiment, and deduct in any subsequent actual determination the amount of sulphur found in the blank.

The above experiments were made with a platinum crucible supported by an asbestos plate in such a manner that two-thirds of the crucible projected below the plate. The