683 



the latter predominates ; from this solvent it is often deposited in 

 magnificent needles several inches in length. Sorbic acid fuses at 

 134 '5. It boils at a much higher temperature, and may be volati- 

 lised without decomposition. 



The alkaline sorbates are all soluble in water, the potassium, 

 sodium, and ammonium compounds are extremely soluble, and cry- 

 stallise with difficulty ; the barium and calcium salts are less soluble, 

 and may be obtained in splendid scaly crystals of the lustre of silver. 

 Their crystallisation is facilitated by the addition of a small quantity 

 of alcohol. Both salts are anhydrous, their analysis agreeing with 

 the formulae 



Sorbate of barium C 12 [H 7 Ba] O 4 



Sorbate of calcium .... C 12 [H 7 Ca]O 4 . 



The silver-salt is a white amorphous precipitate, extremely inso- 

 luble in water, readily obtained by the decomposition of the ammo- 

 nium compound by nitrate of silver. Both combustion and silver- 

 determination proved this salt to be 



Sorbate of silver C 12 [H 7 Ag]O 4 . 



The ether of sorbic acid is readily procured by treatment of the 

 alcoholic solution of sorbic acid with hydrochloric acid, or by the 

 action of chloride of sorbyl upon alcohol. It is a colourless liquid 

 of an agreeable aromatic odour resembling that of benzoic ether, 

 boiling at 191 C., and containing : 



Sorbate of ethyl C 16 H 12 O 4 =C 12 [H 7 (C 4 H 5 )]O 4 . 



The experiments which I have quoted are sufficient to fix the 

 composition of sorbic acid. I have nevertheless produced some 

 additional derivatives of the acid. 



Chloride of sorbyl is obtained by the usual processes ; by the 

 action of pentachloride of phosphorus upon the acid, or of the 

 trichloride of phosphorus upon the potassium compound. The 

 limited amount of acid at my disposal did not permit me to procure 

 this substance in a state of purity, and to establish analytically the 



formula 



C I2 H 7 2 C1 



assigned to it by theory ; but this formula is indirectly proved by 

 the deportment of the crude product, still containing chloride of 



