3 SEPARATION OF THE CAROTENOID MIXTURES 23 



trated in vacuum, and is then subjected to chromatography on suitable ad- 

 sorbents. The aqueous methanoHc solution is diluted with water and extracted 

 with ether. After washing and drying the ethereal solution, the solvent is re- 

 moved by distillation in vacuum and the residue is dissolved in a suitable 

 solvent and also subjected to chromatography. 



3. SEPARATION OF THE CAROTENOID MIXTURES OF THE TWO PHASES 



The method almost invariably employed at the present time for the 

 separation of natural carotenoid mixtures is Tswett's adsorption analysis. It is 

 probably no exaggeration to say that the extraordinary development of 

 carotenoid research during the last 20 years is mainly due to the application 

 of Tswett's chromatographic method and it therefore appears appropriate to 

 describe it in some detail^. However, since a number of excellent monographs 

 dealing with chromatography are already available, only the practical appli- 

 cations of the method to the separation of carotenoid pigments will be dealt 

 with here. 



Although TswETT^, a Russian botanist, published the first description of 

 chromatographic adsorption analysis as early as 1906, and although he drew 

 attention to the scope and manifold applications of his method, chromatography 

 was but seldom made use of during the following 25 years^. It was not until 

 1931, when the older methods of separation had finally proved inadequate, 

 that adsorption analysis was re-introduced by Kuhn, by Karrer, and by 

 Zechmeister, although it had occasionally been used by Dhere^ during the 

 intervening period. The following short summary wiU give an indication of the 

 progress subsequently achieved. 



TABLE 5 



NUMBER OF NATURAL CAROTENOIDS ISOLATED DURING THE 



PERIOD 1922-1946 



A decisive factor in this rapid development was the fact that the adsorptive 

 capacity of the polyene pigments is strongly influenced by relatively small 

 differences in molecular structure. It is therefore possible to separate even 

 References p. 28. 



