70 ESTABLISHMENT OF VARIETIES IN COLEUS 



leaves various gradations between the extremes. The extreme yellow 

 types show some green areas of tissue. The extreme green type, which 

 appears to have no yellow, gives numerous fluctuations, and cases of 

 well-marked spontaneous development of yellow. The distribution 

 of green and yellow does not in any pattern show anatomically a 

 chimeral distribution, as both colors are much intermingled in the 

 subepidermal tissue. 



That a certain degree of chimeral relationship exists in certain 

 patterns of Coleus is evident. In the patterns with solid-red epidermis 

 the epidermal layer is rather specialized in respect to concentration 

 of red cell sap. These apparent chimeral relationships in Coleus are 

 due to intercellular development of patterns rather than to specific 

 and qualitative differences in cells as such. 



Numerous cases of variegation are induced by environmental con- 

 ditions. Cramer (1907, chap, xi) summarizes cases of the influence 

 of parasites, of soil conditions, light, and temperature in producing 

 certain types of chlorosis and variegation. These are, we may say, 

 direct reactions to external conditions, which in most cases are quite 

 apparent. At first thought this class may seem quite distinct from 

 what are considered as true hereditary types, but the difference is 

 chiefly one of degree, for there are few types of cases of variegation 

 that do not fluctuate in response to certain environmental conditions. 

 The infectious types of variegation fluctuate very much according to 

 degree of illumination and may entirely disappear from a plant if it 

 is kept in darkness for sufficient length of time. The types with green 

 and white as periclinal chimeras show, perhaps, least fluctuation in 

 regard to environmental influence. 



In this respect what I have called fluctuations in Coleus are of inter- 

 est. Fluctuations in amount of red present in blotched types is con- 

 stantly occurring. One plant, at the present writing, has some branches 

 with leaves sparsely blotched, as in figure 5, others with the red blotches 

 strongly coalescing, as in figure 29, and still others with nearly a solid 

 epidermal red. These fluctuations do not seem related to external 

 environmental conditions. In the amounts of green and yellow, the 

 various patterns possessing these two elements showed a strong ten- 

 dency to be more green in winter and more yellow in summer. The 

 degree of fluctuation was, however, not uniform for the different sub- 

 clones, for the several plants of a single generation, or even for all the 

 branches of a single plant. A few plants maintained a maximum of 

 yellow in their leaves throughout the winter. Still it is very clear that 

 many cells in the leaves are green in winter which would have been 

 yellow had the particular leaves developed during summer. 



In certain plants variegation appears periodically. These types 

 are, perhaps, in a strict sense, to be classed with the preceding, but 

 differ at least in showing marked periodicity which may well be due 



