62 ESTABLISHMENT OF VARIETIES IN COLEUS 



rings to quite a definite band or zone of red cells extending through the 

 stem. Also the stems of plants, especially those with blotched patterns, 

 may have large irregular streaks of red-colored tissue and we may say 

 are internally blotched. In general, plants with a decided nodal 

 zone of red were those with epidermis free of coloration, and on the 

 other hand, plants with solid epidermal red or heavily blotched epider- 

 mal red had as a rule slight localization of red in zones. There was also 

 the marked localization of red in upper center of leaves (see fig. 10), with 

 almost complete absence of red below, although in many leaves of this 

 type small blotches of red were evident on the under side (see fig. 10a). 



All these conditions indicate that the possibility of producing and 

 possessing red cell sap is a specific property of all cells and that the 

 distribution giving localization at nodes, in streaked areas in stems, and 

 in subepidermal tissues, and in the epidermis either as a uniform red or 

 a blotched red are dependent on two fundamental conditions : (a) total 

 amount of pigmentation, and (6) the appearance of it in certain centers 

 of concentration. The facts as described for Coleus seem to indicate 

 the such conditions are determined largely by intercellular relations. 



Chemical studies, in general, show that differences in quality, 

 quantity, and distribution of pigments in flowers and leaves are cor- 

 related with changes in quality, amount, or distribution of any one 

 of such substances as chromogens (glucosides, phenols, tannic acid, etc.), 

 oxydases, enzymes, oxygen, etc. 



That marked changes in color quite comparable to those I have de- 

 scribed for Coleus may result from slight chemical changes is well shown 

 by the chemical studies of various members of the genus Monarda. 

 The results of these studies are fully summarized by Wakeman (1911). 

 The isolation and determination of yellow and red pigments and a study 

 of their chemical relatives have given rise to the quinhydrone hypoth- 

 esis of plant pigmentation. It is considered that the plant oxidizes 

 thymol or carvacrol to a series of oxidation products of yellow, orange, 

 and red colors, but all closely related to each other. Of these, hydro- 

 thymoquinone, thymoquinone, and dihydroxythymoquinone have been 

 definitely isolated and determined. These in turn have the capacity of 

 adding phenols yielding highly colored phenoquinones and quinhydrones. 



Wakeman (1911, p. Ill) states: 



"Taking into consideration only those compounds that have been isolated 

 or whose presence has been indicated in the monardas thus far, the number of 

 possible pigments becomes truly bewildering." 



Furthermore, some of the pigments are phenolic in character and 

 can combine with metallic constituents of the plants, giving rise to 

 different shades of the original pigment. 



The highly colored red and purple pigments of the stems and leaves, 

 and the yellow and purplish pigments of the flowers in Monarda fistulosa 

 are thus quite definitely identified as mixtures of quinhydrones which 



