1905.] IMPROVEMENT OF TOBACCO BY SEED SELECTION. 1 8/ 



been inbred for so long that they have become accustomed to 

 this kind of fertilization. 



In order to prevent this crossing in the variety the flowers 

 should be covered by a light, strong paper bag. All of the 

 sucker branches should be removed before the bag is applied, 

 together with the small, useless upper leaves. The bag must 

 be placed over the flowers before any of them open, or the 

 entire benefit of bagging will be lost. Any light bag will 

 answer the purpose, but the twelve-pound manilla bag with 

 roof-shaped bottom, sold by most grocery stores, is the best 

 kind of a bag under ordinary conditions. If the plant will 

 not support the bag a stake should be arranged to hold it up. 

 The bag should be moved up the stalk occasionally, in order to 

 accommodate the extremely rapid growth of the plant at this 

 stage of development. 



Seed saved in this way is larger, heavier, and lighter in 

 color than the seed saved in the ordinary manner. The plants 

 grown from the seed saved under the bag grows vigorously, 

 and produces the best plants. The most striking effect of this 

 method of saving seed, however, is the wonderful uniformity 

 of the plants in the field, a condition which is not found in any 

 other crop so far as is known. The number of leaves, their 

 shape and size, the number of suckers, the height of plants, 

 the size and nature of the venation of the leaves, the earliness, 

 — in fact all of the important characteristics are transmitted 

 from the parents to the progeny of the tobacco plants with al- 

 most absolute certainty. 



THE USE OF SPROUTED SEED. 



It is a common custom in the Connecticut Valley to sprout 

 a part of the seed before sowing in the seed beds. This 

 sprouting is usually done in rotted apple tree wood, or cocoa- 

 nut fibre. When the sprouts have been well started they are 

 thoroughly separated, and sowed with an equal measure of 

 dry seed. There is no doubt but that this method produces 

 the earliest plants, and the sprouted seed in the present ar- 

 rangement of seed beds means a gain of several days in earli- 

 ness for a part of the plants. However, it was observed this 

 last season that where plants from sprouted seed grew side by 

 side with plants from dry seed, other differences could often 

 be observed. In many cases the diseased and freak plants 



