In Testing, the Sample Is All-Important

A. S. CARTER.

WE MEASURE the quality of seed by testing a representative sample of a lot. Often it is small compared to the lot itself. The test sample used to determine the content of pure seed of the 20 thousand pounds in a carload of Kentucky bluegrass weighs only 1 gram; 454 grams weigh a pound.

Yet a test of this small sample is dependable if the carload of seed was uniformly blended and if the sample is properly drawn and handled.

Seed is sampled, examined, and tested repeatedly during normal production, processing, and distribution.

A test of a sample obtained by a country buyer determines the price paid to the grower. Samples drawn before and during processing determine the processing needed to make the seed marketable. One or more additional samples are tested to obtain labeling information before seed is offered for sale under a label. If the seed is certified, seed-certification inspectors also sample and test the seed at least once. Finally, State and Federal control officials very likely sample and test the seed after it is offered for sale to determine whether the seed and its labeling meet applicable laws.

Time spent testing carelessly drawn samples may be time wasted; the test is no more accurate than the sample. Much of the variations in tests made by different seed laboratories and analysts can be attributed to samples that are not representative of the lot.

The Rules for Testing Seeds developed by the Association of Official Seed Analysts and the rules followed in the administration of the Federal Seed Act include rules for sampling. The rules are followed by official seed certification and commercial laboratories in North America and are similar to the rules of the International Seed Testing Association commonly used when sampling and testing seed that is to be shipped in international commerce.

OFFICIAL sampling rules provide for using a trier, or probe, long enough to reach all areas in a seed bag. The trier must be designed to remove an equal volume of seed from each area through which it travels. A trier with no partitions in the seed chamber must be inserted into the bags horizontally. It is hard to get samples of seeds that are not free flowing, such as certain chaffy grasses, with a trier, but one can thrust the hand into the seed and remove representative portions.

The trier that meets official requirements is a hollow tube that is pointed at one end and open at the other and has a continuous slot on one side. It is inserted into the bag horizontally with the slot down. The slot is turned up to fill, the trier is removed, and the seed is allowed to run into the sample container. The proper use of this trier will give one a representative sample.

Double-tube triers often are used, but it is doubtful whether they are superior to the single tube in sampling bagged seed. Many double-tube triers are hard to use. In those that have cross ribs, seeds may be pushed along with the trier so that too many seeds from places near the point of insertion are included in the sample.

For small seeds in bags, such as alfalfa and timothy, the trier need not be more than three-eighths inch in diameter. For larger seeds, such as cereals and soybeans, the diameter should be 7/8 inch to 1 inch.

Seed stored in bins can be sampled best by using a standard grain probe with partitions.

Bulk seed also can be sampled as it is being transferred to the bin after cleaning or bulking. That is done by removing representative portions from the moving stream of seeds falling into the bin. At intervals, a pan is moved across the entire stream with one motion. It is necessary to do it that way because seeds may segregate in the stream.

Certain instruments and methods are not recommended because they do not provide an accurate sample. A 6-or 9-inch tapered trier, frequently referred to as a thief, often is used because it is easy to use. The thief trier is inserted upward through the fabric of the bag, and seed is allowed to run into the hand or a sample container. It removes seeds only from a place immediately inside the bag and above the point of insertion. It cannot draw a representative sample.

A standard tube (or stick) trier, even one with a sleeve, inserted vertically in a bag gives a sample with a disproportionately large number of seeds from the top of the bag near the place of insertion, because the pull of gravity is greater than the slight sideward pressure exerted by seeds deep in the bag. A trier should not be inserted in a bag vertically unless the trier has partitions to lessen the downward movement of seed.

SAMPLING would be easy if seed could be blended so that it is completely uniform. Then one could take a test sample from one place in one bag or bulk lot. Complete uniformity seldom is achieved, however, and the sampler must always bear in mind that the lot may not be absolutely uniform. Many bags and many parts of bulk lots therefore must be probed to be sure that the composite sample of the lot is representative.

The Rules for Testing Seeds require the sampling of each bag in a lot consisting of five bags or fewer and (for larger lots) the sampling of five bags plus 10 percent of the bags in the lot. Regardless of size of the lot, it is not necessary to sample more than 30 bags, because the increase in precision achieved by sampling more than 30 bags is not enough to justify the extra work.

The sampler actually attempts to get a representative sample of a lot of seed. A lot of seed has been defined in the Federal Seed Act regulations and in many State laws as a definite amount of seed, identified by a lot number, every portion or bag of which is uniform, within permitted tolerances, for the factors that appear on the label.

The lot may be large or small. Usually the actual size is not limited, except that it must be reasonably uniform. The establishment of maximums for lot size has been considered but never officially adopted because one distributor may blend a carload of seed to greater uniformity than another could achieve with a much smaller amount.