OF ARTS AND SCIENCES : JANUARY 29, 1873. 505 



apparatus of the dimensions described, a two-gallon bottle is large 

 enough to hold all the excess which escapes before the water is satu- 

 rated. To insure perfect saturation, the water in each of the bottles 

 B and C should be frequently shaken up with the gas, especially 

 towards the end of the process. 



We constantly use an apparatus mounted as above, with a water- 

 head of about thirty feet. It would undoubtedly stand a much greater 

 pressure, but a solution saturated under a pressure of two atmospheres 

 is as strong as is desirable. For example, one hundred cubic centime- 

 tres of such a solution is more than sufficient to precipitate a gramme 

 of antimony. For saturating four litres of water in an apparatus of 

 the dimensions described above, the charge should be about 200 

 grammes of ferrous sulphide, about two litres of water, and 288 

 grammes, or 160 cubic centimetres, of sulphuric acid. As this 

 amount of acid water, when at a low temperature, is insufficient to dis- 

 solve all the ferrous sulphate formed, we place the generator in front 

 of a hot-air register. In dismounting the apparatus, we close first the 

 inlet tube of the bottle B, and then remove the generator and wash 

 bottle ; but care must be taken to relieve the pressure on the generator 

 very slowly. Otherwise the escaping gas will cause the acid solution 

 of ferrous sulphate, left in the bottle, to boil over. 



By the use of a solution of hydric sulphide in place of the gas, the 

 consumption of ferrous sulphide in our large laboratory has been 

 reduced twenty-fold ; and when it is remembered that by the previous 

 waste the air of the room was constantly poisoned, and the waste- 

 pipes clogged with the undissolved sulphide of iron, carelessly washed 

 into the sinks, the advantage will be appreciated. The gain in those 

 processes of quantitative analysis where hydric sulphide is required is 

 hardly less important. The bubbling of a gas through a liquid inevi- 

 tably entails loss, which can be wholly avoided by using the solution ; 

 and, by regulating the pressure on the tap, the reagent can be delivered 

 in the proportions required, and at the exact point where it is wanted. 

 Complete precipitation, moreover, is effected in a few minutes ; and, if 

 the liquid is constantly stirred as the reagent slowly flows in, the 

 precipitate will settle in a condition admirably adapted for filtering. 

 Lastly, the separation of sulphur, which is often so excessive when the 

 gas is employed, is diminished, if not prevented, by using the reagent 

 in solution. 



VOL. VIII. 64 



