210 Plant-Breeding 



are the result of a slow evolution or direct breeding. The 

 former are indeterminate or uncertain, and the latter are 

 determinate or definite. The greater part of those in the 

 first class are plants that are multiplied or divided by 

 bud-propagation. They comprise nearly all our fruits, 

 the woody ornamental plants, and such herbaceous gen- 

 era as begonia, canna, gladiolus, lily, dahlia, carnation, 

 chrysanthemum, and the like, in fact, all those multi- 

 plied by grafting, cuttings, bulbs, or other asexual parts. 

 The original plant may be either a seedling or a bud-sport. 

 The gardener, who is always on the look-out for novelties, 

 discovers its good qualities and propagates it. 



Varieties which are habitually multiplied by buds, as 

 in those plants that have been mentioned in the last para- 

 graph, vary widely when grown from seeds, so that every 

 seedling may be markedly distinct. As soon, however, as 

 varieties are widely and exclusively propagated by seeds, 

 they develop a capability of carrying the greater part of 

 the individual differences down to the offspring. That 

 is, seedlings from bud-multiplied plants do not ''come 

 true," as a rule, whilst those from seed-propagated plants 

 do "come true." The reason of this difference will be- 

 come apparent on a moment's reflection. In the seed- 

 propagated plants, like the kitchen-garden vegetables 

 and the annual flowers, we select the seeds and thereby 

 eliminate all those variations which would have arisen had 

 the discarded seeds been sown. In other words, we are 

 constantly fixing the tendency to "come true," for this 

 feature of plants is as much a variation as is form or 

 color or any other attribute. Suppose, for example, that a 

 certain variation were to receive two opposite treatments, 



