778 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS the yellow-red pigments (pheomelanins)* not even the chemistry of the mother substance has been clarified. Tryptophane has been suggested as a candidate, but there is no evidence that it has any such role except in lower species (2). Studies to test whether it gives rise to pheomelanins in man have led to negative results (3). There is however in bright red human hair a characteristic entity, the "possible pink constituent" of Sorby who discovered it almost 90 years ago. His classical description (4, 5) lists most of the unique properties of this pigment: its pink color when extracted from bright red human hair with hot "20 times diluted sulfuric acid " the color change and subse- quent precipitation upon addition of alkali to the solution and its broad absorption band in the visible region of the spectrum. Sorby correctly suspected that the pink substance may be a degradation product and not the original form in which the pigment exists in the hair. In 1943 unaware of this early account, Rothman and the senior author isolated this substance again by boiling red hair with 0.1N HC1 for 1-2 hours. The pink solution turned reversibly yellow above pH 2, precipitated at the neutral point, dissolved at an alkaline pH, had an ab- sorption band in acid solution with a maximum at about 535 mu, and contained iron in the trivalent form. Very little information was ob- tained about its chemical composition. Because of its iron content, the material was named trichosiderin (6). Trichosiderin met with a less than unprejudiced acceptance by the scientific world. An attempt was made to relegate it to the status of an artifact of keratin hydrolysis (7). No mention was made of the fact that the pink colored products in keratin hydrolysates turn yellow at pH 7 and can be obtained by boiling horny products (or other proteins with aromatic amino acids) for 4-6 hours with 40-60 times stronger acids than are required for the extraction of trichosiderin. Such procedures de- stroy the red hair pigment. Nevertheless this allegation found its way into the literature (2, 8) casting doubts on the existence of trichosiderin as a valid chemical entity. The belief that trichosiderin was merely an incidental by-product of pigmentation in red hair was reinforced by the observation that after an apparently exhaustive extraction of trichosiderin, the color of red hair was not diminished noticeably. Also, the presence of iron in an epi- dermal pigment seemed incompatible with the copper enzyme-controlled synthesis of the eumelanins. * The distinction between eumelanins and phemnelanins is based on color differences, solu- bility in acids and alkalies and nature of precursors (2).

THE COLOR OF RED HAIR 779 During the past 20 years, other investigators have extracted similar pigments from red chicken feathers (9, 10). These results suggested that red iron pigments were also present. More recently, Barnicot (11), ex- tracted from human red hair a yellow solution with weak alkali at room temperature. Upon boiling in weak acids for five minutes, this solution turned red and assumed the characteristic absorption spectrum of tri- chosiderin. Barnicot's finding has provided further proof that tri- chosiderin is not an artifact of keratin hydrolysis. The progress of the past 20 years in the isolation and purification of tissue components persuaded us to resume the study of this unique sub- stance. It was hoped that the pigment of red chicken feathers would prove sufficiently similar to trichosiderin to serve as an experimental model, thus overcoming the difficulty of obtaining sufficient quantities of red hair. RECENT RESEARCH ON RED HAIR PIGMENT The first aim of the work, the isolation of various forms of iron pig- ments from chicken feathers has been achieved (12). All these forms have absorption bands at about 280 mt•. This paper describes the over- all direction and some general aspects of these studies. In all major respects the pigment of red chicken feathers behaved as its human counterpart. It had the same indicator properties in its "siderin" form, precipitated at pH 7, had the same absorption spectrum, contained iron, and could be degraded to derivatives analogous to those of trichosiderin. This finding is of considerable biologic importance. It suggests not only that the iron pigment is phylogenetically very old, but also sheds new light on the relative importance of trichosiderin. Considerably larger amounts of pigment could be extracted from feathers than from hair. Boiling for ten minutes with 0.1N HC1 yields 6-10 times more pigment than two hours of boiling of hair. This finding was not surprising, because on a macroscopic and molecular level feathers are looser structures than hair and would release their pigment more easily. However, feathers of Rhode Island Red chickens are not notice- ably redder than bright red hair. It appears that a large proportion of trichosiderin may be retained by the compact hair shafts. Conse- quently, during the prolonged extraction the pigment may be attacked by the extracting hot acid and made insoluble in the hair shaft. This hypothesis was tested by prolonging the extraction of hair and feathers with boiling 0.1N HC1 for several hours. Although the solu- tions obtained in the later stages are pale yellow or virtually colorless,