26 E- DIVERS AND 1'. HAöA. 



therefore to be done either at u gentle heat in the air. or in a vacuum 

 over sulphuric acid. When the solution has been somewhat con- 

 centrated, say to a weight of oOO grams, it is cooled in ice and after 

 some hours strained, still in the refrigerator, from a transparent 

 magma of the crystals which have separated of sodium sulphate. 

 Sliould the oximidosulplionate be wanted for the })reparation of 

 oxyamidosulphonate or. through this, of hyponitrite (77//.s- JonrnaL 

 3, '211). the solution at this stage is serviceable without further 

 preparation. 



The solution, deprived of much of its sulphate by cooling, soon 

 yields when restored to the vacuum-evaporator much of the sodium 

 oximidosulplionate. 90 or 100 grams of tliis having crystallised out, 

 its mother-liquor may be deprived of more sulphate by cooling and 

 then yields a further crop of oximide crystals, nearly pure, by 

 evaporation. 



Even as first obtained, disodium oximidosulplionate is nearly 

 pure, l^eing an anhydrous salt in hard and dense small thick prisms 

 deposited as a thick crust on the bottom of the crystallising pan. It 

 may be recrystaliised from its solution, made with a little hot water 

 rendered slightly alkaline by ammonia, by evaporation. 



Titration of some of the original liquor after hydrolysis of the 

 salt into hydroxylamine, shows the yield to be 90 per cent., more 

 or less, of the quantity calculated from the nitrite, but the crystals 

 obtained do not amount to much moi-e tlian 70 per cent., because 

 the magma of sulphate retains mucli of the solution, which is very 

 concentrated. 



The theorv of the formation of tlie oximidosul[)honatcs is Qon- 

 tained in the theory of the reaction between nitrites and sulphites, and 

 this we hope to discuss in a future paper. It will be enough to point 

 out here that one molecide of 'nitrite, one of hvdroxide. and two of 



