New YorK AGRICULTURAL EXPERIMENT STATION. 335 
SEED SELECTION ACCORDING TO SPECIFIC 
GRAVITY.* 
Ve Aa CLARK. 
SUMMARY. 
In this report is described a variation of the method of seed 
selection by salt solutions. The variation differs from the method 
as heretofore practiced in its making separates at much shorter 
intervals, thereby permitting of determining with greater preci- 
sion the distribution of seeds with regard to specific gravity. 
With different ranges of specific gravity different cultural pro- 
perties of the seed are often found to be correlated. 
The method of a series of separates differs from the method of 
samples, which has been principally used heretofore by scientific 
investigators, in that it distinguishes between individual seeds 
of different qualities, in so far as these qualities are correlated 
with specific gravity, and separates them; and does not simply 
indicate the average specific gravity of the whole lot, as is done 
by the method of samples. 
Seeds of the same lot commonly are distributed through a con- 
siderable range of specific gravity. If the seeds are of good 
quality, the larger part of them are found within a relatively nar- 
row range near but not at the upper limit of specific gravity for 
the variety. But in the the case of oil-bearing seeds the range of 
greatest frequency of distribution is intermediate. 
Specific gravity may be utilized as a means of separating for- 
eign matter, or, occasionally, foreign seeds, as has long been 
known to practice. 
Within the limits of the variety, the lower the specific gravity, 
the greater the proportion of small seeds and vice versa. The 
separation of an unsifted lot of seeds by the method of salt solu- 
tions is in reality in part a crude separation according to size, 
so far as cultural properties are concerned. 
*Reprint of Bulletin No. 256. 
