FOOD-COLORING SUBSTANCES. 13 



The acid yellows (No. 8 and No. 9) although chiefly extracted by amyl alcohol from 

 the acid solution (sec. 2) always yield a large fraction in this group. MTien the colora- 

 tion of the extracted acid salt mixture is entirely due to such products it will be orange 

 red, becoming yellow on neutralization, and also will show the characteristic reactions 

 of the acid yellows with nitrous acid, etc. 



Sec. 5. The dichlorhydi-in solution is diluted with three or four times its volume 

 of carbon tetrachlorid and the color removed with a few small portions of water. 

 The combined washings should be shaken out once with carbon tetrachlorid to get 

 rid of dissolved dichlorhydrin. The aqueous solution may contain the lugher sul- 

 phonated triphenylmethane colors or perhaps sulphonated indulin. These dyes, 

 like Acid Magenta, are accompanied by large amounts of subsidiary products, and 

 their solubilities can not be established with, any definiteness. For their further 

 differentiation compare theh* properties as shown in the tables. 



Sec. 6. The amyl alcohol extract of the strongly acid salt solution, if colored, is 

 washed four or five times with fourth-normal hydrochloric acid, the washings being 

 kept separately. No. 108 and No. 692 predominate in the first washings, while the 

 acidity is still high, because of hydrochloric acid dissolved in the amyl alcohol. No. 

 106, No. 107, and No. 94 come out in large proportion when the acidity of the lower 

 layer, after the shaking, is below seven-tenths normal (usually about the tliii'd wash- 

 ing). Ob"sdously a stronger acid than fourth-normal may be used at first, but it is 

 usually better in practice to wash with this concentration and refractionate if neces- 

 sary. The dyes that may be present in the acid amyl alcohol extract show a gradual 

 transition in theii- distribution ratios relative to amyl alcohol (and other like solvents) 

 and hydrochloric acid of varying concentration. Consequently the acid normalities 

 to be chosen in working with an unknown mixture must be selected somewhat accord- 

 ing to probabilities. 



Comparison of the appearance of the different washings usually will show whether 

 more than one color is present which is extracted by fourth-normal hydrochloric acid 

 in considerable proportion. The amyl alcohol is reserved for the treatment described 

 in section 7 or 8. No. 108 may be separated from Nos. 106, 107, and 94 by fractiona- 

 tion between two-normal hydrochloric acid and amyl alcohol. Nos. 692 and 8 can be 

 separated from Nos. 106, 107, and 94 similarly with eight-normal sulphxuic acid and 

 a mixture of equal volumes of amyl alcohol and gasoline; although, since the acid is 

 somewhat difficult to remove afterwards, the procedure is better ada])ted for sepa- 

 rating the last-named dyes in pure condition than Nos. 692 and 8. For Nos. 106, 107, 

 and 94, the amyl alcohol gasoline solution is washed with a little water to take out the 

 dye. Thi.s solution is treated with one-half its volume or more of concentrated hydro- 

 chloric acid and is reextracted with amyl alcohol. This latter solution may now be 

 washed with a few portions of hydrochloiic ac-id of from four to six normal strength to 

 remove sulphuric acid. The dye is finally removed with a little water and the color 

 obtained in pure condil ion ('for the cyanid reaction, for example) l)y evaporation to dry- 

 nefiH on the steam baf li . The dyes in the sulphuric acid solution are best separated l)y 

 anilln Ccompare p. 23j; but the final removal of this solvent is tedious. No. 94 must 

 be separated from No. 100 and No. 107 by anilin and fourth-normal hydrochloric 

 acid in 5 or 6 per cent salt solution. After the fractionation the dissolved anilin in 

 the wjlutions must be carefully removed by several extractions with carbon tetra- 

 chlorid or other convenient w^lvent from tlie faintly alkaline solution. 



Commereial No. 692 and No. 8 are made l)y direct sulphonat ion of coloring matlers 

 and are rather indefinite in composition. It will often be more convenient to divide 

 the wjlution.s of the colors of tills gronj) and to destroy different dyes in the various 

 portions, liy cautious treatment with "IMankite" (Na.^S/).,) in acid solution, sub- 

 sequently shaking with air to restore the blue, No. 692 may be separated from the azo 

 colors. 



