FROM PROTEID SUBSTANCES. 367 



When dry, the pigments were more or less jet black in color, 

 and when dissolved in dilute alkali they yielded yellowish 

 brown, brown or black colored solutions according to the 

 degree of concentration. These solutions show no absorption 

 bands when examined before the spectroscope, but absorb a 

 certain amount of light at the violet end of the spectrum. 

 Unlike the melanin obtainable from the negro's skin,* our 

 pigments were soluble in glacial acetic acid, but were not 

 precipitable therefrom by potassium ferrocyanide. Alkaline 

 solutions of these artificial melanins were completely bleached 

 by chlorine. 



Heated on platinum foil, our products, like the pigments 

 described by Abel and Davis, gave off at first fumes of pyrrol, 



tested by a pine-sliver moistened with hydrochloric acid, 



but these soon ceased, leaving a coal-black residue very 

 difficult of combustion. 



From alkaline solutions, the pigments were precipitated by 

 cupric sulphate, silver nitrate, plumbic acetate, and baryta 

 water. By strong nitric acid the pigments were dissolved, 

 but precipitated again on addition of water. 



From these few reactions it is seen that our artificial 

 melanins or melanoidins, to use Schmiedeberg's term, differ 

 only in minor degree from the melanins or pigments obtained 

 by Abel and Davis from the skin and hair of the negro, or 

 from the melanins studied by other observers. 



Comparison of the composition of the artificial melanins with 

 that of natural melanins, etc. Our own results bearing on the 

 chemical composition of the melanin-like pigments obtainable 

 by decomposition of proteids emphasize the view that these 

 substances are many in number, and that, while having many 

 points in common, they differ widely from each other in com- 

 position, owing no doubt in part to variations in the extent or 

 intensity of the hydrolytic cleavage by which they are pro- 

 duced. This we fancy is a far more potent factor than the 

 character of the individual proteid from which they are derived. 

 Thus, in our experiments with antialbumid, the first prepara- 



* Abel and Davis, loc. cit. 



