23 È. mvÈRS Axb t. ha<ia. 



addition that of not requiri ug voluminous solutions to be worked with. 

 For it is still more certain than that process, and gives a yield in 

 crystals of nearly 97 ])er cent, of the calculated quantity of the salt 

 when carbonate has l)een used, and nearly 95 per cent, when hydrox- 

 ide has l)een used, and with not enough nitrile with it to be visible, 

 even on recrystallising. It is tlie small solubility of the potassium 

 salt -which makes its preparation simpler than that of the sodium salt, 

 and also which, by throwing it our of solution removes most of it from 

 the influence of the sulphite and sidphurous acid, and thus lessens the 

 fbi-mation of nitrile. Our success by this method has been so much 

 greater tban that reached by either Claus or llaschig mainlv Ijecause 

 we have taken the nitrite and the hvdi'oxide in the right proportion, 

 besides "'uardino' ao'ainst local excess of sulphur dioxide, and beiniz' 

 careful to preserve a low temperature. 



Potassium nitrite, assayed for real salt, and either powdered 

 potassiiun acid-carbonate or concentrated solution of potassium hvdrox- 

 ide in equal molecular proportions are put, witli a weiglit of water 

 alxnit eight times that of the real nitrite, in a roomy flask ke[)t at 0°, 

 or even better at Ü° or so below zero. l)v means of a brine bath with 

 ice lioating in it. To preser\e this tem])erature during the jiassage of 

 the sulphur dioxide tlie flask must be kept in active motion, witli a 

 thermometer in it foi* observation. Bv a wide inlet tube terminating 

 high al)0\e the surface of tlie solution, in ordei" to guard against its 

 being choked, the sul[)hur di<jxide is i-apidlv passed in until the car- 

 bonate has ull dissohcd and etter\escence has almost ceased, or until 

 the hydroxide, if tliat has been used, is nearlv neutralised, and then 

 passed slowly until the solution has bectmie neutral to lacmoïd papei". 

 The oximidosulphonate separates during the latter part of the process 

 as ;i ci-ystalline ])Owder. T'he mother-liquoi- retains a little of it and 

 contains, 1)esides, some niti'ile along with small ([uantities of both 



