xanthophyll is 0.07 to 0.12 percent and carotin 0.03 to 0.08 percent, or about I molecule of carotin to 1.5 to 2 molecules of xanthophyll.

Xanthophyll has the formula, C40H56O2, and thus may be considered an oxid of carotin. Nothing is known of the function of the oxygen atoms; they are considered ether-like, since xanthophyll does not give a reaction for =COH, =CO or -COOH. It appears to give a very easily dissociable addition product when an ether solution is treated with methyl alcoholic potash. It shows a tendency to crystallize with alcohol of crystallization and is best obtained solvent-free by precipitation from chloroform with petroleum ether. The typical crystal forms are long tables and prisms. which are pleochroic and often show a steel blue luster. In transmitted light they are yellow, and are red only where several crystals cross. This distinguishes them from carotin, for the colors of the solutions are very similar. It melts at 172°. Xanthophyll is relatively stable towards oxygen in dilute sol., but the pulverized substance takes up 36.5 percent of its weight of oxygen, giving a compound, which, precipitated from methyl alcohol by ether, has the formula, C40H56018. Like carotin, it gives a di-iodid, tufts of thin, dark violet, prisms with metallic luster. It is easily decomposed. The bromid, C40H40 Br22, is also similar to that of carotin. It gives the same color reactions with conc. sulfuric acid and alcoholic hydrochloric acid sol.

Lutein.31 As mentioned above, a compound isomeric with xanthophyll has been found in lutein, the coloring matter of eggyolk. This was first isolated in a pure state by Willstätter and Escher, 32 who obtained 4 gm. of very crude pigment from 6000 eggs (110 k.). The yolks were coagulated with alcohol (7 1. to 6 k. of eggs) and the coagulum extracted with acetone (5.4 k. were shaken with 3 1. of acetone and filtered; 2.8 k. of the residue were shaken with 2 1. of acetone for one hour and then washed on a filter with 21. of acetone). Phosphatids were removed by shaking the acetone with petroleum ether, washing with water, and mixing the petroleumether sirup with two vol. of acetone; the acetone was then removed by washing with water, the petroleum ether concentrated to about 2 1., filtered from cholesterol, diluted to about 6 1. and cooled.

The lutein that separated was purified by repeated crystallization from methyl alcohol (1 gm. required 1000 cc. for solution) or from carbon disulfid. It formed dark, brownish-yellow, compact prisms with blue surface-luster, melting at 195-6°. It differed from xanthophyll only in its higher melting point, and was called xanthophyll b by Willstätter.

Fucoxanthin. Fucoxanthin is the carotinoid characteristic of the Phaeophyceae, or brown algae. It differs from the other yellow pigment, in its high oxygen content, having the formula, C40H5406. Many investigators have had more or less pure solutions of this pigment, but Willstätter and Page34 were the first to obtain a crystalline product.

Fucoxanthin was isolated from the mother liquor of chlorophyll a (extracted with 85 percent acetone). Four liters of the extract were treated with I 1. of a mixture of petroleum ether (30-50°, 3 vol.) and ether (I vol.) and then with 1.5 1. of water. The ether mixture was then carefully washed free from acetone, concentrated to about 11⁄2 1. and shaken with 1 1. of methyl alcohol (saturated with petroleum ether) four times, then twice with 1⁄2 1. of alcohol. The xanthophyll was removed by shaking with an equal vol. of a mixture of 5 vol. of petroleum ether and I vol. of ether. The fucoxanthin was then transferred to a large vol. of ether, and the ether concentrated to a thick sirup. Fucoxanthin crystallized out upon the addition of low-boiling petroleum ether. The yield from 20 k. of fresh algae was about 2 gm. of a product 85 percent pure.

The use of all reagents containing mineral matter must be avoided, if ash-free preparations are desired. It is also essential that all extracts and solutions be kept from the light and from moisture as much as possible. If the algae are dried previous to the extraction, the yield is very much smaller; and if this dry material is kept for some time before being used, little if any fucoxanthin can be isolated.

The crude product may be recrystallized from methyl alcohol, forming bluish, glistening, brownish-red, long, monoclinic prisms, containing 3 molecules of alcohol. From methyl alcohol or acetone, in the absence of air, it forms dark red six-sided tables containing 2 molecules of water. These forms are interchangeable. It is obtained free from solvent by precipitation from absolute ether with low-boiling petroleum ether, forming compact needles, melting at 159.5-160.5°, depending upon the rate of heating. The ether solution is orange-yellow; the alcoholic sol. more red, with a brownishyellow tinge; while the carbon di-sulfid sol. is quite red. The pure pigment is stable in an atmosphere of oxygen, but is oxidized in benzene or dilute alcoholic sol., giving a product of approximately the composition, C40H54016

Fucoxanthin does not show acid properties; it is not extracted from ether by aqueous potassium hydroxid sol. and is not changed by 50 percent hydroxid sol., solid barium hydroxid or metallic sodium. It reacts with alcoholic potash, forming an addition product. This is decomposed by water but gives, instead of fucoxanthin, a product with increased basic properties and a different absorption spectrum. 35 The ether sol. gives a deep blue salt with dilute hydrochloric acid sol. This behavior may indicate the existence of a pyrone ring in the fucoxanthin. The reaction with alcoholic potash appears to consist in the decomposition of a part of the pyrone nucleus; the hydroxyl group thus formed would account for the increase in basic properties. 36 A characteristic of fucoxanthin is its marked basic properties. The other carotinoids give a deep blue color only with conc. sulfuric acid. Fucoxanthin reacts as a weak base with dilute mineral acids. Thirty percent hydrochloric acid sol. decolorizes the ether sol., itself becoming violet blue in color. With a solution of the acid in dry ether the hydrochlorid, C40H54O6.4HCl, is obtained as blue flakes with a copper luster. When shaken with ether and sodium bicarbonate, a compound is formed with one atom of chlorin, which gives a greenish yellow solution. The iodid, C40H540614, forms violet-black, short pointed prisms with a copper luster.

One k. of fresh algae (Fucus) contains 0.169 gm. of fucoxanthin, 0.089 gm. of carotin, 0.087 gm. of xanthophyll and 0.503 gm. of chlorophyll a.

PLANT PIGMENTS

Their color and interrelationships

B. HOROWITZ

Introduction. In attempts to explain the action of ammonia on thymol,1 Prof. Gies and the author were led to review the work of Liebermann on the influence of ammonia upon orcinol.2 Liebermann's suggestion that ammonia combines with oxygen of the air to form nitrous acid, and that the latter is the effective agent in the production of pigment, strengthened our view, as already held, that many plant pigments are synthesized in a similar way. Miss Wakeman's study of the pigments in the Monardas, whereby she came to the conclusion that these are probably oxidation products of thymol, and its isomer, carvacrol, and Wurster's suggestive paper on the rôle of hydrogen peroxid in color formation, afforded further evidence in support of this idea. During the past year the thymol problem has been studied side by side with an investigation into the chemistry of some plant pigments (the botanical side of which is engaging the attention of Dr. A. B. Stout, of the N. Y. Botanical Garden). As an introduction to a description of these studies, we present herewith a brief review of the theories on color and chemical constitution, as well as an outline of the possible chemical interrelationships of some of the more important plant pigments.

Color and chemical constitution. Perhaps one of the most fascinating chapters in the development of organic chemistry has been the attempt to correlate the chemical constitution of substances with their physical properties. With the rise of synthetic chemistry, and especially as a result of the pioneer work of Graebe, Liebermann and Baeyer in the production of synthetic dyes, color

1 Gies: BIOCHEM. BULL., 1912, ii, p. 171. Horowitz and Gies: Ibid., 1913, ii, P. 293. Horowitz: Dissertation, Columbia Univ., 1913, pp. 68.

2 Liebermann: Ber. d. d. chem. Gesell., 1874, vii, p. 247.

3 Wakeman: Bulletin of the Univ. of Wisconsin, No. 448; Science series,

1911, iv, p. 25.

4

Wurster: Ber. d. d. chem. Gesell., 1887, xx, p. 2934.

and chemical constitution began to attract attention. Witt's chromogen-chromophore theory, as well as the quinone theory of Armstrong, held absolute sway for many years; and though their usefulness is far from exhausted, the tendency at the present time seems to be to rely less on the influence of the radical in the alteration of color, and more on the relationship of color to the absorption spectra produced.

As is well known, colored substances exhibit the phenomenon of selective absorption; that is, whenever a body vibrates so as to emit waves of certain definite periods, any waves of these periods falling upon the body will be absorbed. This gives rise to the absorption spectra that have so often been of use in the identification of complex colored compounds. Introduction of radicals into a compound, transforming it from a colorless to a colored substance, with consequent exhibition of absorption in the visible part of the spectrum, may be explained by assuming that the oscillation-frequency has been altered; for example, benzene, though colorless shows absorption bands in the ultra-violet portion. Introduction of the azo group, - N=N-, gives red azo-benzene, with absorption bands in the visible part of the spectrum. What apparently occurs in this case is a change of the short wave-length with its high oscillation-frequency (as found in the ultra-violet region) into a longer wave-length and a consequent slower oscillation-frequency.

Application of the inductive method in attempts to draw general conclusions has been but partially successful. Even early in the course of these studies it was recognized that unsaturation in a compound is essential to the development of color. Attempts were also made to trace relationships between the molecular weights of compounds and the probable colors produced. This culminated in Nietzski's rule: "The simplest colored substances are in the greenish yellow and yellow, and with increasing molecular weight the color passes to orange, red, violet, blue and green." Like most of the theories in this field, this is at best highly imperfect.

Undoubtedly the most fruitful theory which has so far been advanced connecting color with chemical constitution is that due to Witt. He considered color to be due to the presence of a "chromophore" group in the molecule. The resulting "chromogen"

5

Witt: Ber. d. d. chem. Gesell., 1876, ix, p. 522; 1888, xxi, p. 325.