BY THE SELECTION OF SOMATIC VARIATIONS. 47 



plant, one should look in the generations following or preceding for the 

 serial numbers. Certain plants were grown with two patterns as a 

 chimera, and these are indicated by brackets, with a record for each 

 pattern. In order to make the tabulation more compact, numbers 

 are used to represent the different patterns, and these correspond with 

 the numbers of the figures in the plate, as follows: 



1 = yellow-red blotched. 5 = green-red blotched. 



2 = green-yellow-red blotched. 6 = yellow-green-red blotched. 

 4 = green-yellow spotted-red blotched. 14 = yellow-green. 



A survey of this series of plants shoAvs that on the 91 plants 31 bud 

 variations appeared, giving a ratio of frequency of 1 to 590 against 

 1 to 860 for the entire clone 11 and 1 to 1,700 for the sister clone 111. 



A further analysis within this progeny shows that in several cases 

 similar bud variations can be traced to a common ancestry. For 

 example, 6 of the 9 cases of what has been called spontaneous develop- 

 ment of 3^ellow occurred during the summer of 1914 in plants descended 

 from plant 117142 (see table 6), which itself was constant for loss 

 of yellow during its period of growth. Another case of spontaneous 

 development of yellow was in plant 1171441. In this clone, therefore, 

 all cases of spontaneous development of yellow were in plants descended 

 from plant 11714. 



It is quite clear from such pedigrees that distinct differences in 

 tendencies in regard to the degree of variation may exist among buds 

 of branches bearing similar foliage. 



ENVIRONMENTAL INFLUENCE. 



Observations were made and pedigrees of plants examined to deter- 

 mine whether changes in ordinary environmental conditions influence 

 fluctuations and bud variations. 



To secure accurate data on the relative frequency of bud variations 

 during summer and winter is hardly feasible. In general only about 

 half of the summer plants were from cuttings made early in the spring. 

 The others were grow^n in pots in a greenhouse during the winter and 

 transplanted to the garden, where the}' grew during the summer, 

 making larger plants with many more branches than were produced 

 during the winter. Of the total number of 1 15 bud variations iuA'olving 

 color, 9 appeared during the winter, which is fairly proportional to the 

 relative number of branches that developed. 



There was a strong tendency' for plants having yellow to become 

 greener in winter and yellower in summer, and also to become greener 

 when severely pruned. At any time, however, during the winter, 

 some plants of each pattern having yellow (for example green-yellow- 

 red blotched) could be found with the pattern as pronounced as dur- 

 ing the summer. During the winter of 1913-14, two plants of each of 

 the following types, green-yellow-red blotched, green-yellow spotted-red 



