60 ZOOLOGY 



flower may in some measure parallel. There is, how- 

 ever, this important difference : the dahlia can be 

 propagated vegetatively, by tubers, and hence it is 

 possible to preserve and increase the various heterozy- 

 gous forms. The sunflower, propagated by seed, will 

 not remain constant unless homozygous ; and then, 

 under ordinary circumstances, it must be hand-polli- 

 nated. Consequently, it will be practically impossible 

 to maintain a large number of pure strains representing 

 the variations of the sunflower. Some of the best semi- 

 doubles appear to be necessarily heterozygous, and con- 

 sequently incapable of producing seed that will regularly 

 come true. 



Variation in 12. Finally, it must be added that even when the color 

 expression an j mar k m g factors are those desired, there may be 

 great variations in "expression." These may in some 

 cases be caused by the presence of what are called 

 "dilution factors," having the property of causing the 

 color to be relatively faint, or diluted. Such factors 

 have been demonstrated even in animals. They may 

 be due also to purely environmental causes, having 

 nothing to do with heredity. For example, it appears 

 that the red tends to fade out in very hot weather, 

 and it has been claimed that its appearance is greatly 

 affected by the soil. 



Plant breed- 13. Plant breeding, in the light of our present knowl- 

 occupation ec ^g e ? ' ls a fascinating subject, and may be carried on 

 very well in a small garden. It is especially suitable 

 for amateurs, who do not expect to earn their living in 

 this manner. They can experiment as they will, with- 

 out being obliged to consider financial, results. The 

 best way is to select some species or genus of plants, 

 and study it intensively, becoming acquainted with all 

 its known variations and special peculiarities of struc- 



