FOOD-COLORING SUBSTANCES. 15 



however, that new dyes may also be formed by partial reduction in the case of polyazo 

 or nitroazo derivatives. 



The halogenated fluorescein derivatives are much more resistant to bromin in 

 acid solution than are most other colors. They tend, however, to add bromin unless 

 fully substituted. Most of the azo dyes are much more readily destroyed by bromin 

 in alkaline solution than is Naphthol yellow S . ^fixtures are made fourth-normal or 

 above with sodium carbonate and are treated with dilute bromin water very cautiously 

 until the azo dye is just destroyed or until the solution has become a clear yellow. 

 Hydrazin sulphate is now added quickly to destroy excess of bromin; the mixture 

 is finally acidified, and the yellow purified by extraction with an immiscible solvent. 

 This procedure is seldom so satisfactory as the regular extraction with ethyl acetate 

 or amyl acetate, and is not applicable in the presence of Nos. 62, 64, 65, and 188, which 

 form intensely blue substances by this treatment. 



Sec. 10. The amyl alcohol extract obtained by shaking out the original mixture 

 after adding 5 or 6 per cent salt will contain practically all of any basic dyes present. 

 Most of the acid dyes of low sulphur content are also almost completely extracted. 

 The extract is measured, diluted with an equal volume of gasoline, then washed a 

 few times with sixty- fourth-normal hydrochloric acid. The washings, if colored, are 

 treated aa directed in section 11. The extract is next shaken out with sixty-fourth- 

 normal acetic acid, these washings being treated according to section 12. Eosins and 

 (in general) coloring matters that are unsulphonated phenolic compounds are now 

 removed by a few portions of sixty-fourth-normal sodium hydroxid solution, this 

 fraction being treated according to section 13. The amyl alcohol gasoline mixture is 

 finally washed once with ver^' dilute acetic acid and, if still containing any signifi- 

 cant amount of coloring matter, is evaporated to dryness on the steam bath, the 

 residue being examined according to section 14. 



Sec. 11. The washings of sixty-fourth-normal hydrochloric acid (sec. 10) are tested 

 for basic dyes by making a small portion alkaline with sodium hydroxid, shaking 

 with ether, then treating the ether solution, which is usually colorless, with dilute 

 acetic acid.' If the latter becomes colored, indicating the presence of basic dyes, 

 the alkaline test portion may be shaken out once or twice more to determine whether 

 or not acid dyes are also present in this fraction. If these tests indicate the presence 

 of both acid and basic colors, the acid colors must be removed by making the principal 

 part of the sixty-fourth-normal hydrochloric acid extract alkahne (normal with 

 sodium hydroxid) and extracting with ether. From the combined ether portions 

 the Ijasic dyes are removed by washing — first with sixty-fourth-normal acetic acid, 

 finally with dilute hydrochloric acid. This treatment should be omitted if acid dyes 

 are absent, since most l^asic colors are unstable in alkaline solutions, Auramin, espe- 

 cially, suffering decomposition rapidly. The basic colors may be further fractionated 

 from arayl alcohol with dilute hydrochloric acid, from ether with very dilute alkali, 

 etc. The separation of basic colors from alkaline solutions by immiscible solvents is 

 rather oijjectionable, since such colors (aricording to Kehrmann, Havas, and Grand- 

 mougin ^) suffer rearrangement fnnn ortho-quinoid to para-quinoid structure. This 

 change is attended in compounds such as Crystal Violet, containing only fully allcy- 

 lated amino grouyjs, l)y eliminalion of one of the alkyl groups. The original dye 

 may not ]>o, ol)tained therefore, but, instead, the lower alkylated derivative. 



The alkaline solution, after removal of the basic dyes with ether, is made about 

 normal with hydrochloric acid and is shaken out with amyl alcohol gasoline mixture. 

 Any coloring matter extracted here probably will be a minor portion of a dye already 

 obtained by the procedure described under section 8, and its furtlier fractionation will bo 

 carrifid out as stated in that paragraph; or the solution containing it may be combined 



' O. N. Witt, Z. Antil. Chem. 20 (1887), 100. Wolngiirtnor, Z. Anal. Chom. 27 (1888), 2152. 

 » Ber. Chcm. Ges. 46 (lai.J), 2i:jl; 47 (191-1), 1881. 



