62 THE CULTURE OF TOBACCO. 



rarefull7, viz., that the ideals striven f'->r must not be radically 

 elianged with each year. If later it Ije found that the ideals c.( 

 the first years have been incoirect, then, of course, it is necessaiy- 

 to change them, but changes should not be made according to the 

 whims and fashions of each year. Characteristics established by a 

 selection of three years are at least ten times more permanently 

 fixed than those established by one year's selection. There is 

 always a tendency in all highly-developed plants to revert to type, 

 and 'this tendency is stimulated Ijy a s..il nr climate that is m 

 any waj- different from the plant's native hiime. The object of 

 selection is, therefore, not only the improyemeut of the type, l;>ut 

 its preseryation. 



Seed should not be sayed from different varieties of^ tobao.-o 

 grown in the same locality, for the pollen is likely to be carried from 

 one variety to the other,'thus producing a cross, so that the plants 

 grown from the seed will largely differ in their characteristics. Some 

 will resemble one parent and some another ; some will have the 

 i/haracteristics of both parents in different degrees, and others 

 will haye none of the yarietal characteristics of either. Where 

 seed is saved from different varieties grown on the same farm, 

 these varieties should be widely separated from each other. When 

 several planters in the same locality each grow several difi'erent 

 varieties of tobacco, they may often arrange each to save seed of 

 but one vaiiety, and to exchange each year. 



The tendency to cross fertilization, and its resulting variations, 

 is valuable where the process is directed by the hand of man, for 

 it is often necessary to produce a new variety with the character- 

 istics of two or more other varieties. When this is attempted, the 

 two plants that are to be used are carefully selected, with the end 

 sought in view, and the blossom bud covered with a small paper 

 or muslin sack. Before the blossom has matured, this sack is 

 lifted, and the pollen-bearing poition of one blossom (the anthers 

 or male organs) is removed with a small tweezer ; the sack is 

 again placed on, and when the blossom is sufficiently developed, 

 and the pistil receptive, the pollen is taken from the other plant, 

 and placed on the blos-iom of this plant bjr means of a camel-hair 

 brush, after which the sack is again adjusted, and left on the 

 plant until the .seed pod has formed. The seed from this plant is 

 sown by itself, and the plants I'esulting show hundi'eds of varied 

 characteristics and differences. These plants are gradually selected, 

 and cut out of the competition, until only the two or thi-ee plants 

 conforming nearest to the ideal are left to develop seed. The 

 process of selection must be kept up for several years before a new 

 variety may be considered as established. The results of the ciciss 

 may not pji-ove at all satisfactory, and the process may have to lie 

 repeated several times. 



The estafilishment of a seed plot is an excellent scheme. The 

 seed foi- the planting of this ci'op should lje saved from the most 

 jjerfect plant of the preceding year. Thus the fiest plant of the seed 

 plot will be the parent of the seed-bearing plants of the following 

 year and the remainder of the selected plants of the seed crop will be 

 the parents of the main crop the following yeai'. In this wa-s- a pedi- 

 greed variety of plants will be established and the plants of the seeil 

 ■nop will always lie the offspring of the same parent and the ])lants of 

 the main crop will always be the descendants of the same grandparent. 



