208 CAROTINOIDS AND RELATED PIGMENTS 



fresh portions of pure ether. Even in these cases extraction of the 

 pigment could be facilitated by diluting the serum with two volumes 

 of water and adding an equal volume of methyl alcohol before shaking 

 with ether. 



In the writer's experience with cattle and horse serum the evidence 

 for a carotin-albumin combination of some kind rests upon a number 

 of easily demonstrated facts, among which are the following. The 

 fat solvents will extract little if any pigment from serum even after 

 great dilution with water. When the globulins and albumins in the 

 serum are fractionally precipitated by increasing concentrations of 

 ammonium sulfate, the carotin follows the albumin fractions. In 

 fact it is possible to roughly isolate an albumin which carries the 

 carotin in firm combination, which, like the serum itself, will not 

 give up its pigment to the fat solvents until first treated with alco- 

 hol, indeed unless alcohol is present. The lead, silver and mercury 

 salts of the protein act in the same manner. After coagulation with 

 alcohol and drying it was found, in one test at least, that alcohol had 

 to be added before petroleum other would extract the pigment from 

 the protein. This albumin, moreover, seems to have a more or less 

 definite heat-coagulation point of 86 C., when in half-saturated am- 

 monium sulfate solution. This property can therefore be used for 

 the isolation of the pigment-carrying protein. 



The isolation of the carotin-albumin complex can be carried out 

 as follows: The serum is first freed from globulins by adding an 

 equal volume of saturated ammonium sulfate solution. These are 

 filtered off on a Biichncr funnel, using suction, and thoroughly washed 

 with half saturated ammonium sulfate solution. The combined filtrate 

 and washings are then carefully heated to a temperature of 79 C., 

 at which temperature the bulk of the albumins are coagulated. 

 Some carotin is lost in this coagulum, but with serum rich in carotin 

 the filtrate from these proteins will have a golden yellow color. The 

 carotin-albumin fraction is secured from this filtrate either by salting 

 it out by any of the albumin precipitants (complete saturation with 

 ammonium sulfate is best) or by heating to 86 C. In either case 

 the precipitate will have a deep yellow color and the amount ob- 

 tained will be very small in comparison with the proteins which 

 have been precipitated as globulins and albumins in the preliminary 

 operations. The protein can be redissolved after salting out, and the 

 aqueous solution thus obtained exhibits all the properties of blood 

 serum so far as its relations to fat solvents and the extraction of 



