50 ESTABLISHMENT OF VARIETIES IN COLEUS 



ment of green and yellow the plants ranged from pure yellow plants 

 that died within a few weeks to those that were pure green. Two 

 of the plants had much yellow in very irregular and mixed patterns. 

 Eleven had no trace of yellow for several months, when a few yellow 

 spots appeared in some of the leaves on several plants. In regard to 

 epidermal red there were numerous types ranging from solid red to no 

 red. Of those wdth red-blotched epidermis, some were uniformly 

 finely blotched, while on others the blotches were large, with a single 

 blotch sometimes covering one-fourth of a leaf. As to the shape of the 

 leaves, the series showed the same range of variation exhibited by the 

 seed progeny described above. There were but 5 that were strongly 

 laciniate. 



Besides the above plants which were grown to maturity, there are 

 at the present date (February 8, 1915) 20 seedlings of a plant that 

 had an epidermis of solid red, as shown in figure 8. The plants have 

 from 2 to 3 pairs of leaves, but it is clear that only one of the seedhngs 

 has a solid red epidermis. Red-blotched types prevail and few of the 

 seedlings show any yellow coloration. 



Summarizing, it is plain that the plants grown from seed give wide 

 variations. In respect to color patterns, there were numerous types 

 which gave very complete gradations between extremes, especially in 

 regard to epidermal red. Many of the types that had appeared as 

 bud variations appeared also in the seed progenies, such as yellow-red 

 blotched, green-yellow spotted-red blotched, green-red blotched (wide range 

 of variation in respect to size of blotches), green-yellow spotted, and green. 

 One of the new patterns could be described as yellow-green blotched- 

 solid red. Another had the red blotches of the epidermis coalescing at 

 certain points, making the red markings continuous but not solid (see 

 fig. 29), so that the underlying green showed through in blotches. 



In vegetative propagation leaf-shape remained very constant, the 

 only exception being the clones that developed laciniate leaves. In 

 the seed progeny of all plants tested, however, new types of entire 

 leaves have appeared. The laciniate leaf characters and various 

 intermediates between it and the typical entire leaf appeared in the 

 first leaves of certain plants and remained as a rather constant character 

 during the time the plants have been grown. 



The wide variations appearing in the seed progeny indicate that this 

 strain of Coleus is either of mixed parentage or that the processes 

 concerned with production of color patterns and leaf-shape are them- 

 selves subject to wide variations. The variations in respect to color 

 patterns were, however, no greater in range or extent than were those 

 that appeared in bud variations, and the fluctuations from entire 

 to laciniate leaf-shape gave extreme types of leaf-shape, with all grades 

 of intermediates, on a single plant, as is quite well shown in plant 123153 

 of plate 4. 



