1913.] PUBLIC DOCUMENT — No. 31. 189 



lu Tables 3 aud 4 are given our results of crossing pink with 

 white-flowered varieties, the division into two tables being 

 simjDly for convenience in arrangement. A striking thing 

 shown in this table is the excess of white-flowered plants in 

 the crosses involving Golden-eyed Wax. This occurs in all 

 cases except in the F2 generation of the cross with Prolific Black 

 Wax. It is seen also in crosses with light pink-flowered sorts, 

 especially with Giant Stringless in Table 1, also in crosses with 

 Bine Pod Bntter in Table 5. The total number of hybrid plants 

 from crosses of Golden-eyed Wax with piginented flowered varia- 

 tions is 1,053, of which 407 are white-flowered, while the ex- 

 pectation of one-fourth of the total number is 262. This is a 

 proportion of 1.59:1 instead of 3:1. This departure from ex- 

 pectation is probably due to a correlation between blossom color 

 and seed-coat colors or color patterns. Another thing that calls 

 for explanation is the occurrence of light pink flowers in 

 the second generations in Table 4. These appear in a ma- 

 jority of the crosses but not in all of them. There can be little 

 reason to doiibt that nearly all, at least, of these classifications 

 of blossom color are correct, for our policy has been not to record 

 a departure from expectation in any case of doubt, but only 

 where the unexpected character is reasonably typical. As pre- 

 viously explained, these two color types are quite distinct. It 

 may be observed that White j\larrow gives peculiar results here 

 as in crosses already discussed. It is probable that the appear- 

 ance of the light pink blossoms in these crosses is connected 

 with the pigmentation of the seed-coat, and nmy be explained 

 by a study of the seed-coat color. White ]\Iarrow evidently 

 carries several factors modifying pigmentation both of seed- 

 coat and blossom. 



