1905.] IMPROVEMENT OF TOBACCO BY SEED SELECTION. 1 83 



from the tropics has been grown in the north. The change of 

 seed from one state or locality to another has produced similar 

 though less striking results, and has led to the general con- 

 clusion that the tobacco crop should be grown from seed which 

 is thoroughly adapted to soil and climatic conditions. In the 

 case of the Cuban and Sumatra crops grown in this valley in 

 1903 typical plants were selected for seed, and the seed pro- 

 tected from cross fertilization by means of a paper bag enclos- 

 ing the flowers. This seed produced uniform plants similar 

 to the mother plants, and showed that by such a method of seed 

 saving uniform crops can be secured one year after the variety 

 is introduced into a locality. If it is desirable to make a 

 change of seed the farmer should grow a few plants the first 

 season for the purpose of producing the general seed crop for 

 the following year. 



THE VALUE OF UNIFORM PLANTS. 



The value of the tobacco crop depends largely upon the 

 uniformity in type of the plants, particularly the shape, size, 

 quality, and number of leaves. The lack of uniformity in- 

 creases the cost of sorting and growing the crop, and results 

 in the production of a certain proportion of inferior tobacco. 

 It costs about the same to grow the irregular and undesirable 

 plants as typical plants of the desired type, so that it is highly 

 important to grow uniform crops. An examination of the 

 crops grown in this region showed that there was great va- 

 riety in the size of plants, number of leaves per plant, the 

 shape and size of leaves on the individual plants, the number of 

 suckers, the earliness of maturity, the quality of the leaves, 

 and other characteristics which go to make up the types of the 

 varieties of tobacco. The difference in size of plants and 

 rate of growth was most clearly seen during the fore part of 

 the growing season, at which time the plants were found to 

 vary in height from one to four feet. This variation in size 

 is not shown clearly at the time of harvest except by careful 

 examination of the tobacco stalks, by reason of the topping of 

 the plants to about the same height. The number of leaves 

 varied to a marked degree, some plants producing a large 

 number while others bear only a few desirable leaves. The 

 increase in number of leaves is not accompanied by a corre- 

 sponding increase in the height of the plant, but the plants bear- 



