90 The Bulletin. 



cross poUenation it is impossible to know what to expect in the oti'spring, but when 

 seeds are produced under a bag it is safe to assume that plants from seed thus 

 produced will be like the plant upon which they grow. There is great danger of 

 disappointment in this, however. The desirable quality which we have noticed 

 may have been due to inherent qualities within itself, in which case it will have 

 the* ability to transmit these qualities. Or, what we have noticed as desirable 

 may be due almost exclusively to environment, in which event the plant will not 

 necessarily be prepotent in transmitting the qualities we recognized in it as being 

 worthy of transmission. Weather conditions may tend to overdevelop a stalk and 

 that stalk be a disappointment as a seed stalk, or a great deal of available fer- 

 tility may stimulate a growth which will cause us to select it as a parent stalk and 

 it may have no value owing to lack of prepotency to improve the subsequent crops. 



Nor is this the whole of the problem. In selection of seed corn it is essential 

 that we have as near a perfect germination as possible. We can discard by one 

 method or another the ears that look as though they would be faulty in this 

 respect, and we must actually test those ears in which the eye has failed to find 

 evidence of faulty germination. Again the analogy holds. We must eliminate 

 the tobacco seeds of poor germination, and test those which perchance we have 

 concluded were all right, that we may have the strongest and the best from Avhich 

 to produce our seedlings. Having selected the desirable stalk, it has been found 

 that in every pod there are a large percentage of seed which will not germinate 

 at all, and that there is again a large percentage which will germinate but weakly, 

 and that rarely do we find a larger amount than 40 per cent which would be 

 really desirable to use as seed. It has been found by the va^rious experiment 

 stations that the vitality of the seed is indicated by tlie relative weights of the 

 seed. The very light ones in every pod will not germinate, the medium may but 

 will not produce strong and healthy seedlings, and that only the heavy ones will 

 produce a seedling of suHicient vigor to give us a good start toward a crop. 

 Hence, the necessity appears for the separation of the heavy from the light and 

 chafTy seeds. The reason for this is readily seen, and its importance should 

 commend the practice of seed separation to every grower in the State. Tobacco 

 seed is among the smallest if not the very smallest with which the farmer must 

 deal. At best it can obtain but little food for the germ within itself, and in the 

 case of the very light there is not enough to support the germ until it can feed 

 upon the plant food in the earth into which it is cast, and it dies unavoidably. 

 In the medium there may be enough to sustain the germ, but not enough to make 

 a strong germ development, and hence a stunted plant, a disappointment during 

 the whole season. Only the very heaviest will have enough plant food against the 

 time when the germ will be sufficiently developed to draw its food from its environ- 

 ment. There are various methods by which this can be done. The method by 

 which it is done is immaterial, if the results are secured. I shall outline a few 

 methods. In the Bible we read of the winnowing process. This ought to give us 

 an idea of a simple and I believe practical method. On a very still day, when 

 there is no perceptible movement of the atmosphere, when it would appear as 

 though there were a dead calm, take your seed, a table oil cloth or a very large 

 paper, spread it on the ground, and then pour the seed out slowly upon the paper 

 or oil cloth. There will be enough air stirring to blow the light and chaffy seed 

 so that it will not fall upon the cloth or paper, only the choice heavy seeds will 

 fall upon the cloth below. If once winnowing doesn't seem sufficient, it can be 

 treated to a second winnowing. In this there must be an almost perfect calm, 

 otherwise the whole of the seed will be scattered to the winds. 



Another method, where atmospheric conditions can not be found, is to produce 

 them. An open grate with a roaring fire — with only one opening — will produce 

 a draft toward the fire. One can stand in the door and drop the seed slowly onto 

 a pai)er or oil cloth. This draft artificially made will do the same thing as the 

 wind. The desirable seed will fall directly upon the receptable, the undesirable 

 ones being drawn toward the grate. 



Another method is by casting the seed into a pan of water; the heavy ones will 

 sink and the light one's Hoat; tlie light ones can then be skimmed off. This has 

 many advocates, not as a perfect method of seed separation, but as one which 

 will be found very beneficial. It is one method suggested by the Department at 

 Washington. I have tried it and I find but little to recommend it. In fact, it is 



