Theory' of Gunpowdei\ 493 



liquid, which at first became brownish, was again colourless. The 

 sulphurized oxide filtered off, and dissolved in fuming nitric acid, 

 gave 0"1567 grra. of sulphate of baryta free from nitrate, which 

 for 100 parts of substance corresponds to 2'239 sulphate of 

 baryta, or 1'058 sulphide of potassium, or 1*077 hydrate of 

 potash, or 09043 potash, or 0-3080 sulphur. 



3. Determination of Hyposulphite of Potash. — The whole of the 

 liquid obtained in experiments 1 and 3 was well mixed and di- 

 vided into 7 equal volumes ; one of which, corresponding to 1 

 grm. of substance, was used for this and for each of the follow- 

 ing determinations. 



The amount of hyposulphite is most readily determined by 



adding a standard solution of iodine to a volume of the liquid 



which has been mixed with solution of starch, and acetic acid 



added until a feeble acid reaction showed itself. If a. be the 



quantity of iodine in grammes in a degree of the burette, t the 



number of burette degrees added to the solution until the blue 



colour manifested itself, the c[uantitv of hyposulphite of potash is 



2K0 SO^ 

 obtained from the formula s = x- — j^ t, as is seen from the 



following equation. 



I 



KI 



/ was 40*7; a was in this, as in all following determinations, 

 0-001242 grm. Hence the fiuid investigated contained 0-001859 

 X 40-7 = 0-07566 grm. hyposulphite of potash, or 7566 per 

 cent., which corresponds to 3-747 per cent, potash, or 2-5452 

 sulphur, or 8-0434 nitre. 



4. Determination of Sulphate of Potash. — Another volume of 

 liquid, containing 1 grm. of residue, gave 0*7582 grra. sulphate 

 of baryta, which in 100 parts correspond to 56-623 sulphate of 

 potash, or 10-397 sulphur, or 30-624 potash, or 65-/21 nitre. 



5. Determination of Sulpliocyanide of Potassium. — Small quan- 

 tities of sulphocyanido of potassium may be determined with 

 great accuracy by a colorometric method. For this purpose we 

 prepared a solution which contained 00004884 grm. sulphocya- 

 nide of potassium in one degree of the burette, and we employed 

 this solution in all our determinations. As a colorometric agent, 

 we used a solution of j)erchloride of iron mixed with hydrochlo- 

 ric acid and sufficiently diluted, containing an unknown quan- 

 tity of iron. Of this solution, two measures were brought into 

 two glass cylinders of the same shape. To one of these the 

 liquid to be tested for sulphocyanide was added, and to the 

 other the standard solution of sulphocyanide of potassium was 



