Drushol and Elston — Sulphide Sulphur. 



157 





Table I. 







Sulphide sulphur 

 taken 



Sulphide sulphur 

 found 







per cent 



per cent 



Error 



1. 



0-0025 



0-0020 



0-0005 



2. 



0'0015 



0-0012 



0-0003 



3. 



0-0000 



0-0006 







4. 



0-0004 



0-0003 



o-oooi 



5. 



0-0004 



000G4 







6. 



0-004 



0-005 



o-ooi 



1 . 



0000 



0000 



- - - - 



8. 



0-0002 



0-00015 



0-00005 



9. 



0-008 



0-009 



o-ooi 



10. , 



0-002 



0-0025 



0-0005 



11. 



0-001 



0-0015 



00005 



12. 



o-oooi 



o-oooi 



_ _ 



per cent. These errors may be reduced by repeating the 

 determination and taking- the mean of several values found. 

 In this way in the practical applications of the method the 

 errors may be kept within reasonable limits. 



Practical ajyplications. — (1) In gas analysis. The method 

 was used for the estimation of the hydrogen sulphide in the 

 laboratory atmosphere and in illuminating gas. On a day 

 when no hydrogen sulphide generator was being used in the 

 laboratory the air contained barely a perceptible trace of 

 sulphur; on another day when a class two floors below was 

 using a hydrogen sulphide generator the amount of hydrogen 

 sulphide in the air of the upper laboratory was found to be 

 J part in 5,000,000. Twenty-five liters of air were slowly 

 drawn through a Geissler bulb of the most modern type con- 

 taining dilute potassium hydroxide solution. This solution 

 was then washed into a measuring flash and made up to the 

 mark with nitrite free distilled water and aliquot portions of 

 this solution were used for determining the sulphide sulphur 

 as previously described. The same procedure was used for 

 estimating the hydrogen sulphide in illuminating gas. The 

 colorimetric method gave 1 part of sulphur in 1,000,000 of gas. 

 The same result was obtained by oxidizing the larger portion 

 of the solution from the Geissler absorption apparatus with 

 bromine water and weighing the sulphur as barium sulphate. 

 Duplicate determinations on the same illuminating gas gave 

 precisely the same result. 



(2) In coke analysis. The simplest method of estimating 

 sulphur in coke given by Fresenius is to boil 5 grams to 10 

 grams of powdered coke in dilute hydrochloric acid, and to 

 absorb the hydrogen sulphide evolved in dilute potassium 

 hydroxide solution. The sulphur is then oxidized to the sul- 



