30 MISC. PUBLICATION 4 3 4, U. S. DEPARTMENT OF AGRICULTURE 



Other treatments involving the use of solvents such as xylene, 

 ether, and acetone have been used experimentally. McKeever (24) 

 found that a 10- to 120-minute treatment of black locust seed with 

 these chemicals aided in dissolving the impervious waxy covering and 

 improved germination. 



Seed Testing 



Seed testing is an important phase of seed and nursery work and no 

 seed should be sown that has not been tested in some manner before 

 sowing. Germination tests are the most reliable measure of seed 

 viability, but only when made at the proper time and on seed 

 thoroughly afterripened. Consequently, cutting tests must often be 

 relied upon to indicate the percent of well-filled seed. 



SAMPLING 



No matter what kind of test of seed is to be made, it is imperative 

 that the sample used be representative. It is necessary, also, to run 

 one or preferably more separate germination tests on each lot of seed 

 of which the origin or treatment is dissimilar. If a given lot contains 

 a large quantity of seed, say several tons, it is safest to obtain a separate 

 sample from each container or bin. 



From 0.5 ounoe to 32 ounces, depending on size of seed, or from 1,000 

 to 4,000 seeds, is considered a fair sample. A seed-sampling device 

 such as is used in sampling grain will be found useful, though not 

 essential, in obtaining a uniform sample. The original sample can be 

 further reduced by mixing thoroughly, spreading out in a thin layer, 

 and resampling the seed uniformly with a spoon. The final sample on 

 which the germination test is run should contain at least 400 to 1,000 

 seeds. 



GERMINATION TESTS 



Germination-testing procedure used at the Lake States Forest 

 Experiment Station at St. Paul, Minn., since 1930, is suggestive of 

 methods that can be generally applied. On receipt of a sample for 

 testing, the number of seed per pound is determined for the standard 

 sample (including refuse) and on a clean-weight basis. A cutting 

 test is also made on part of the sample, and percents of good, rancid, 

 and empty seed are recorded. Usually several hundred seed are cut. 

 Quadruplicate samples of 250 seed each are then counted out from the 

 remainder of the sample for the germination test. 



Germination tests are conducted in wooden flats 1 foot square and 

 3 inches deep, in which one to four small samples may be sown. The 

 flats are filled almost to the top with a clean, sterile, medium sand, 

 ranging in pH from 5.0 to 5.7 8 in acidity, tamped slightly and smoothed 

 off. The seed is scattered uniformly on the surface, pressed in with a 

 small board, and covered with sand to a depth of two to three times 

 the average diameter of the seed. Water is then added in sufficient 

 amount to bring the moisture content of the sand up to 10 or 12 per- 

 cent of dry weight. 



Since damping-off may occasionally occur in a flat, despite the use 

 of sterile sand, the four subsamples of each test are always sown in 



s On what is called the hydrogen-ion scale, 7.0 is neutral, 3.5 is strongly acid, 8.2 is the alkalinity of a limy 

 soil, and 9 is very alkaline. 





