20 Bulletin 392 



Type 8 Light purple (Plate I) 



Description of the type. — The type called light purple is not homogeneous, 

 but is so variable as to be difficult of classification. Typically it is between 

 dark purple and magenta. It may appear as a dark purple streaked 

 or washed with magenta, or, when fresh, it may be clear and pure and 

 little darker than a typical dark blue. It seems to be in a very unstable 

 condition, and quickly changes to magenta on a warm day or as the flower 

 begins to wither. Very often this color in a flower seems to indicate 

 a hybrid condition of the plant that bears it. 



Theoretical consideration of data. — Thirty-seven Fi pedigree plants 

 and 4 others from commercial seeds were classified as light purple, and 

 when selfed gave an aggregate'' progeny of 70 dark purples, 37 light 

 purples, 36 dark blues, 17 magentas, 9 mauves, i pink, 6 light blues, 

 14 whites, 9 cream whites, and 3 tinged whites. 



When the data of light purples selfed are examined in the aggregate 

 it is seen that this is characteristically a heterozygous type. It breaks 

 up into all the other types, including dark purple, which is higher in 

 the color series. It does not breed true. An analysis of the type is 

 impossible as a whole, but must be worked out in each individual case 

 by analysis of progeny and by synthesis from other plants of known 

 constitution. 



Type 9 Dark purple (Plate I) 



Description of the type. — The type considered as dark purple is a very 

 rich pansy, or bluish purple, color. The dark purple type may be supposed 

 to contain the intensifying gene I in addition to all the other genes 

 of the other types. 



Theoretical consideration of data. — The data for the dark purple plants 

 that were selfed may be considered in two classes — those that bred true 

 and produced only dark purples, and those that broke up and threw 

 other types. 



In the first group there were 9 plants, which gave an aggregate progeny 

 of 54 plants. These were all dark purples; no other types appeared. 

 It may be assumed, then, that these plants were duplex for all genes that 

 concern the dark purple type (CCRRBBXXII). It is possible, however, 

 that they may have been simplex for either X or I and still bred true. 



Other dark purples split up in various ways when selfed and threw 

 all the types lower in the scale. Such behavior is to be expected when 

 the purples are heterozygous as to various genes. Three of these plants 

 threw whites. This, too, is to be expected if R is simplex. 



* The data are here considered in the aggregate because the progeny in any single case was too meager 

 ♦,o be treated separately. 



136 



