500 



THE AMERICAN NATURALIST [Vol. XLIV 



to the keratin — the coarser the structure of the coat the 

 larger was the percentage of the decomposition product. 

 Even if this view is not correct we know that the various 

 keratins do not have the same composition, 9 and, there- 

 fore, we should look for a variation in any one decom- 

 position product. 



Method of Isolation 

 A weight of wool was boiled with a 10-per-cent. solu- 

 tion of sodium hydrate 10 in the proportion of 300 grams 

 to 1 liter for four hours. The solution was then poured 

 into water, strongly acidified with hydrochloric acid 11 

 and the precipitate allowed to settle. The supernatant 

 liquid was syphoned off and the precipitate washed by 

 decantation. The precipitate was then stirred with 

 from 5 to 10 liters of 0.2-per-cent. sodium hydrate solu- 

 tion and filtered. The filtrate was precipitated by 

 hydrochloric acid and allowed to settle, the liquid 

 syphoned off and the precipitate dissolved in one liter 

 0.2-per-cent. sodium hydrate solution and again filtered, 

 precipitated with hydrochloric acid, washed free of 

 chlorides, dried at 100 and, lastly, extracted with car- 

 bon disulphide, alcohol and ether in Soxhlet apparatus, 

 then dried at 105° and weighed. 



From the table given below it can be seen that "white 

 melanin" does not exist in either recessive or dominant 

 whites, but that there is some product formed by the de- 

 composition of the keratin, which behaves like a melanin, 

 i. e., is soluble in alkali and insoluble in acids or neutral 

 solvents; perhaps this may be shown to belong to the 

 melanin class, but it is at least common to all white 



I'hysi 



