NEF. — BIVALENT CARBON. 189 



determination of the iron and sodium offered at first some difficulty, 

 because, on decomposing the salt with hydrochloric and nitric acids, 

 always a small amount of substance crystallizing in green leaflets is 

 formed, which is msoluble in aqua regia. The salt was finally decora- 

 posed by treating first with a small amount of dilute sulphuric acid, 

 and then evaporating to dryness and igniting. The residue was then 

 dissolved in hydrochloric acid, with addition of a small amount of 

 nitric acid, and the iron and sodium determined in the usual way. 



4.754 grams yellow salt, dried between filter paper, lost on stand- 

 ing over II2SO4 in a vacuum, after two days, 1.5806 grams HgO; after 

 four days, 1.8167 grams U^O , after six days, 1.8996 grams HgO ; on 

 further standing, no further loss of weight was noticed. 



Theory for Na^Fe (ONC)„ 2 HjO + 16 HjO. Found. 



I6H2O 39.80 39.96 



The red powder remaining is very hygroscopic, and still contains 

 two molecules of crystal water. 



1.0073 grams substance gave 0.1760 gram FcoOs- 



0.8154 gram substance gave 0.5380 gram Na2S04. 



0.6999 gram substance gave 0.1282 gram FeaOg and 0.4609 gram 



NaaSOi. 



Theory for Na^Fe (ONC)o 2 H„0 Found. 



Na 21.10 21.36 21.33 



Fe 12.84 12.12 12.73 



An attempt was made to obtain, by oxidation of this salt with bro- 

 mine and water, a fulminic acid double salt corresponding to sodic 

 ferricyanide, but without success. 



V. On Fulminic Acid and its Esters. — The Carbyloxim- 



ESTERS, C : NOR. 



The experiments described above make it clear that it is not possible 

 to obtain free fulminic acid from its salts by treatment with acids, 

 because the acids always react first on the tremendously reactive biva- 

 lent carbon atom present in these salts ; and for this reason oximes of 

 formic acid derivatives are invariably obtained, which, although they 

 can readily be converted back again into fulminic acid salts, never 

 yield, when they decompose, the free fulminic acid. In this connec- 

 tion an experiment carried out with isuretine is of interest. On heat- 

 ing alcoholic solutions of isuretine and hydroxylamine hydrochloride 

 (in molecular quantities) an interaction with formation of ammonium 



