618 SCIENCE PROGRESS 



a simple or easily explained character. Probably the neatest 

 case is that afforded by the /3/3 and aft disulphonates of 

 hydroxylamine, 



HO . N (S0 2 . OH) 2 and HO . S0 2 . O . NH . S0 2 . OH 



recently described by Haga {Trans. Chem. Soc. 1906, 89, 240); 

 simpler examples such as 



(Ammonium nitrate NH, . N0 3 . 



1 or 



( Hydroxylamine nitrite NH£> . N0 2 



J Ammonium hypochlorite NJ7 t . CIO 



\ Hydroxylamine hydrochloride NH 4 . CI 



break down experimentally owing to the instability of the com- 

 pounds formulated in italics, and it is at least doubtful whether 

 compounds such as 



/Ammonium sodium sulphate (NH 4 )NaSo 4 



\ Hydroxylamine sodium sulphite (NH 4 0)NaS0 3 



have ever been crystallised out. The derivatives of cobalt, 

 chromium, and platinum have given a number of examples in 

 which compounds of identical composition can be prepared in 

 differently coloured forms, some of the simplest being a pair of 

 chromous chlorides CrCl 2 .4.H 2 (E. M. Rich, Proc. Chem. Soc. 

 1908, 24, 215), pairs of ammoniated cobaltinitrite salts such as 

 Co(N0 2 ) 2 .Cl, 4NH3, and pairs of ammoniated platinichlorides 

 and bromides such as PtCl 4 , 2NH3; but these can only be 

 explained by somewhat complex space formulae (see Stewart, 

 Stereochemistry, Part II. chaps, v.-vii), and cannot be adequately 

 dealt with in an elementary discussion of the subject. 



Isomerism of Organic Compounds. — Subsequent experience 

 has shown that organic compounds present an extraordinarily 

 fertile field for studying the phenomena now under considera- 

 tion, but the earliest work on the subject presented difficulties 

 of a very high order. Lavoisier's early organic analyses were 

 made with mercuric oxide ; those ol Gay Lussac were effected 

 by the ignition ot pellets in which the organic compound was 

 mixed with potassium chlorate and ignited, the gases produced 

 being analysed volumetrically ; those of Berzelius by a similar 

 method, using powdered chlorate in a tube closed at one end. 1 



1 Compare the estimation of nitrogen by Dumas's method as described by 

 Gattermann {Praxis des organischen Chemikers, p. 84). 



