644 



H ORT I C U LT U RE 



May 19, 1906 



Seed Separation 



Per Cunt or Dh-fercimt Size© 

 IO Crammeb of Seed. 



The separation of tobacco and 

 onion seed is now quite extensively 

 u-<m1 by the growers of these crops 

 in the Connecticut valley. 



The Botanical Department con- 

 tacted with the Experiment Station 

 at Amherst has installed an equip- 

 ment for seed separation, and quite 

 a large amount of onion and tobacco 



seed lias I n separated this spring 



for growers of these crops. 



The method of separation em- 

 ployed is based upon the use of air. 

 For the separation of onion seed we 

 have employed a winnowing ma- 

 chine, which is supplied by a Ger- 

 man manufacturer. This machine 

 separates very effectually all the 

 light seed from the heavy ones. 

 and is capable of separating fifty 

 pounds in a brief period of time. 

 Tobacco seed has been separated by 

 means of a foot bellow- attached to 

 a glass tube. The lower end of the 

 tube is covered with a piece ot 

 coarse-meshed cloth, and (lie air is 

 driven up through by means of the 

 foot bellows. The chaff and light 

 seed are driven to the top of the 

 tube into another receptacle. The 

 amount of light seed and chaff in 

 tobacco and onion seeds runs all 

 the way from five to twenty per 

 cent., or more. These seeds are dis- 

 carded by the grower, and only the ^ 

 larger and better ones are used. 



The benefit derived from the practice of separating 

 seed is important enough to be more extensively recog- 

 nized. There is no reason why the practice 'of seed 



separation should not I mployed in connection with 



various lines of green-house work, such a-, for example, 

 lettuce, radish and beet crops, except perhaps when the 

 latter are 'used for greens. Many experiments have 

 shown (hat large seed produce larger plant-: they germ- 

 inate better, and will prove more desirable for use. 



The accompanying illustration give- the results of 

 separating lettuce seed. The separation in this case 

 was accomplished with sieves, provided with round 

 perforations. 



The percentage of seed ranging from 2 to 1.5 mm. 

 was 10.15 per cent.: from 1.5 to 1 mm., 66.38 per cent. 

 Prom 1 to 5 mm., 18 per cent., and below .5 mm., which 

 consisted mostly of dirt and chaff. 4.19 per cent. The 

 percentage of germination and weight of the seedlings 

 in milligrams of the various grades of lettuce was as 

 follows : 



LETTUCE. 



Size of sieves 2-1.5 mm. 1.5-1 mm. 1-.5 mm. 



No. seeds germinated. . .43 per ct. 35 per ct. 19.6 per ct. 

 Av. wt. of seedlings 43.3 mg. 41.0 mg. 31.6 mg. 



SEPARATION 

 LETTUCE SEED 



Diameter or SEEQ8 IN Millimeters 



It will be observed that the 

 highest percentage of germination 

 occurred with the largest seed, and 

 that there was quite a difference be- 

 tween the percentage of germination 

 of the seed obtained from the two 

 lower grades of sieves. The average 

 weight of the seedlings also is 

 highest in those obtained from the 

 large sieve, slightly lower in the 

 second-grade sieve, and smallest in 

 the lowest grade sieve. What passed 

 through the 1 - .."i mm. sieve was 

 merely dirt and chaff, and of no use. 

 At least 4 per cent, of the material 

 bought in this sample of seed was 

 of no value. 



For the separation of lettuce seed 

 on a practical basis, it would not be 

 necessary to employ sieves such as 

 were used in these experiments. 

 One sieve would be sufficient for all 

 purposes, and such a sieve should 

 have perforations of about 1 mm. in 

 diameter. Everything retained by 

 this sieve can lie planted : everything 

 that passes through should be 



rejected. 



In the experiment given here, 

 such a sieve would reject about 

 22.91 per cent, of seed and chaff — 

 in other words, it would reject the 

 seed and chaff in the two right- 

 hand bottles shown in the illus- 

 tration. 

 If a greater degree of separation is required, a larger 

 sieve could be employed — namely, one measuring 1.25 

 mm., or about 1-20 of an inch. These sieves can be 

 easily constructed out of colander tin, which is gener- 

 ally carried in stock by most tin-men. We have made 

 a sieve out of colander tin soldered on to the bottom of 

 an ordinary spice box, the tin having perforations of 

 about 1 mm. or 1-25 of an inch. Some of the large 

 firms who manufacture perforated metal give the size 

 of the perforations in fractions of an inch, and occasion- 

 ally millimeters. Perforated metal with 1-18 perfora- 

 tion- would he equivalent to 1.4 mm. This size sieve 

 would take quite a large number of the smaller seeds. 

 Such a sieve would probably remove 50 per cent, of the 

 lighter seed, and by this practice one would obtain a 

 very high grade of seedling-. 



The sifting of lettuce, in our estimation, is a desirable 

 thing for green-house men to practice, since it econo- 

 mize- space. It is much easier to accomplish the sep- 

 aration by sifting rather than by selecting the prickers 

 in the beds. It would be a comparatively easy matter 

 to select the proper size of perforated tin for the sepa- 

 ration of radishes, etc. The results of this separation 

 would undoubtedly eliminate manv inferior radish 





