NITKILOSULPHATES. 5 



more so than the Perchlorate (Fremy). It forms slender needles 

 of pearly lustre which in crystallising iill its mother-liquor. The 

 work of later investigators has added nothing to the account of 

 it given by Fremy, except that the crystals arc rhombic (Easchig 

 and Fock). ::: His analyses gave results which agree well with 

 calculation for the true composition of the salt, though not with 

 his own formula for it. By preference he prepared it by passing 

 sulphur dioxide into a solution, not too concentrated, of the nitrite 

 and the hydroxide. With the neglect of details characteristic of 

 Ins celebrated memoir, he fails however to mention the essential 

 potassium hydroxide. Probably on account of this, later workers, 

 have only made use of his other process, which consists in mixing 

 solutions of the nitrite and pyrosulphite. 



Sodium salt, N(S0 8 Na) 3 , 50H 2 . 



A strong solution of sodium pyrosulphite poured upon solid 

 sodium nitrite furnished Easchig (Annalen, 1887, 241, 180 and 

 229) with a useful solution of the nitrilosulphate, although it 

 contained much hydroximidosulphate, sulphite, and unchanged 

 nitrite ; sodium hydroxide he represented also to be present, but 

 that could not have been the case (this Journal, 1900, 13, -J84). 

 Sodium nitrilosulphate being very soluble in water, he did not 

 isolate it in the solid state. We have found that instead of 

 three mois, pyrosulphite, which was the quantity used by Easchig 

 to two mois, nitrite, at least four mois, must be taken if most 

 of the hydroximidosulphate is to be sulphonated.v By preparing 



* Misled by the faulty translation in the Annalen (1845, 56, 342) of Fremy's paper, 

 Clans ami Koch supposed that he had said that red fumes were evolved when the salt is 

 heated. Fremy's statement is that thesalt does not evolve red fumes, which is correct. 



t Claus and Koch found it best to take still mure pyrosulphite when working with the 

 potassium salts (Annalen, 1869, 152, 336), and they were right, for in the case of that in- 

 soluble nitrilosulphate, great excess of sulphite did not matter; here it does. 



