No. 640] VARIATION AND IIEnEDITY IS Lri'I.M s 447 



Additional characters when inseparably linked with an 

 old one might be due to a change in the original factor, 

 but when not linked, or so loosely so as not to maintain a 

 constant association, they would have to be considered as 

 due to a new factor. A new factor might originate by 

 the subdivision of an old one and the subsequent differ- 

 entiation of one part. In this case as well as in the case 

 of the actual loss of a gene, homologous cliromosomes of 

 the hybrid between the new and old form would present 

 the situation originally conceived in the Presence and 

 Absence hypothesis (2) of an actual gene paired with its 

 absence. In the other cases of modified factors this 

 would not be true. 



Summary. — 1. The genus Lnpinus presents an as- 

 semblage of closely related and difficultly separable 

 species. 



2. The present pajDer reports some results of a 6-year 

 field and garden study of L. apricns vaUicola, L. pipcr- 

 smithii, and L. nanus. 



3. The variations described concern tlie form and color 

 of the flower, the shape and size of the pod, and the 

 color and markings of the seeds. 



4. Dark-blue and ]Vnik-flowered races breed true. 



5. Striped-white flowered races are heterozygous for 

 a single factor, which in the homozygous condition pro- 

 duces white flowers. 



6. Light-blue flowers are duo to a single dominant fac- 

 tor, indistinguishable in the homozygous and heterozy- 

 gous condition. 



7. Dark seed coats are linked with dark-blue flower 

 color, but probably due to separate factors. 



8. The factors for light-blue and striped-white flowers 

 are both allelomorphic to that for dark-blue and not im- 

 probably constitute a system of multiple allelomorphs. 



9. Mutations are frequent, some are already known to 

 be dominant, and others appear to ])e in the nature of 

 additions of new characters and factors and so progres- 

 sive in the sense of de Vries. 



