TlIK BUI-LETIN. 91 



the method of last resort. The best method, however, is by means of a blower. 

 The Government Department at Wasliin^ton has devised an apparatus which 

 seeina to nie to ho almost ideally adapted for tiie purposes herein mentioned. It 

 consists of a glass tube about one incli in diameter, inside measurement. This is 

 set upon a stand with a receptacle for the glass tube. On one end of the glass 

 tube is a very line screen to prevent the seed from failing through. Into this end 

 by means of a foot bellows is introduced a current of air. 'Jhe seed is then 

 poured into the glass tube and the air is turned on. The light cliad'y seed is 

 blown out of the top of the tube and the heavj' seed remain i)i the tube. This is 

 an almost perfect separation, and while the machins co-^ts something, it will last 

 a whole community for years, and hence we think it will justify any locality to 

 buy one of these separators. In almost every tobacco district there is rejiorted a 

 gradual deterioration both in yield and textuie from year to year. In son:e cases 

 this deterioration is fancied, not real, but in a large number of instances there is 

 a decided deterioration, and this is due almost entirely to a lack of systematic 

 and careful seed selection, and when it is constantly shown that such deteriorati<m 

 is a source of great loss to producers, greater care should be taken in the selection 

 of individual seed stalks. I have seen the seed stalks selected one after another 

 in a row, 4 to 10, and these up close to the shed to be hand v. This will never do. 

 The chances are extremely remote that you will ever find two stalks together 

 representative of the best types. They are likely to be found one here, one there, 

 and hence they must be selected Avherever they may be found. A farmer should 

 by all means select his own seed, not de])ending npon a seedsman or the other 

 fellow for it. It is a wise thing to secure a little foreign seed every year, try it 

 out, acclimatize it, and if perchance you receive anything promising, further 

 improve by selection. By this method tliere may be a marked improvement iu the 

 type, both along lines of increased yield and quality. 



Ihere is an idea that it is a wise practice to save enough seed one ycur for 

 several subsequent years, the notion prevailing that there is no deterioration in 

 germination due to the aging of the seed. Such a notion became prevalent 

 among a class of farmers, and indeed among such class of farmers it has several 

 things to recommeml it — those who do not pay any attention to the beliavio\ir 

 of plants in the field, who do not consider anything but size or yield, unmindful 

 of disease resisting qualities or quality or texture when they make their seed 

 stalk selections; but among those who are cognizant of the many problems enter- 

 ing into seed selection it is never advisable to select seed for more than one year. 

 Save that this is always a wise plan, to select enough for one, two or three years 

 ahead against the possibility of hail, frost, or an abnormally poor tobacco season, 

 and hence poor seed. One year without this precaution might undo the work of 

 years of careful and painstaking selection. 



In conclusion, we would urge a very careful study of the class of tobacco the 

 local market demands. Having found this type, then give a most careful and pains- 

 taking attention to the improvement of this type, to the end that there maj' be 

 an increased yield of better quality. This will result in economical production. 

 Part of the land now devoted to tobacco can be utilized for other crops, and part 

 of the time now devoted to tobacco can be utilized in the solution of the many 

 and varied problems continually and insistently before the farmer. 



