That, for instance, may be true for corn, for which the accurate spacing of plants is of great importance. Obviously, lower germination will mean a larger number of blank spaces between plants because of the seeds that do not germinate.

The percentage of pure live seed does help you determine the amount of seed you need. The rate of sowing is usually expressed in terms of bushels per acre or pounds per unit area. Greater importance is being attached to the spacing of plants for maximum yield. For instance, the distance between plants for maximum seed production is usually greater than the distance between plants for maximum forage production. Also, spacing of plants for maximum yield is affected by the soil, fertility, and moisture.

A pound of seed that is 50 percent of pure live seed will provide fewer plants than a pound that consists of 90 percent of pure live seed. Recommendations as to the rate of sowing are valuable guides and may be adequate for kinds of seeds that are average in purity and germination. If you want to take advantage of the increased knowledge about maximum yields, however, you must consider the spacing of plants.

You must also decide the rate of sowing and the amount needed on the basis of the quality of seed to be used. There is a good chance that you will need less seed than you assume at first. For this purpose, it would be better if recommended rates of sowing were expressed in ounces, pounds, or bushels of pure live seed per unit of area. It would then be easier to determine how much is needed.

For example, a 100-pound bag of alfalfa, with 98 percent of pure seed and 90 percent germination, contains 88.2 pounds of pure live seed. Another 100-pound bag, with 95 percent of pure seed and 80 percent germination, contains 76 pounds of pure live seed.

If you would sow 8 pounds to the acre, there is enough for about 11 acres in the first bag and 9.5 acres in the second bag. To get enough for 11 acres, 116 pounds of the quality in the second bag would be needed.

Most seed drills can be adjusted to the quantity to be sown. Fewer can be adjusted to the number of seeds per unit area or length of row. If you know how far apart plants should be in the row, you can try to get that spacing by setting the drill to drop the right number of seeds per foot, yard, or rod. You need to allow for casualties from pests, dry soil, or too wet soil, and dead seed. With 80 percent germination, you know that at best only 80 of every 100 seeds will grow. Every fifth seed will not grow. You will have to sow additional seeds to compensate. High-germination seed will have fewer skips or blank spaces, of course.

If you have the time and know the average number of seeds per ounce or pound, which is available from seed analysts, you can determine the amount of pure live seed you need. That would be an interesting project for a long winter evening.

Treating seed to prevent insect damage and to combat disease organisms on the seed and in the soil adds to the assurance of survival of the seedlings. Labeling to show that seed is treated is now required under Federal laws and many State laws. Some of the substances used are poisonous.

KINDS OF SEEDS differ in their average quality.

Perhaps the most difficult seeds to buy on the basis of detailed labeling are seeds for lawns. This is unfortunate, because the average buyer of lawn seeds is only an incidental user of seeds, and therefore does not become acquainted with the kinds, varieties, and the qualities of the several kinds. Lawn seeds generally are small-seeded grass seeds that are difficult to process and clean to a high purity.

The seed of Kentucky bluegrass, for example, is small and has hairy appendages which are hard to remove. It is in the group that we call chaffy grasses. Kentucky bluegrass seed that was 85 percent pure used to be considered to be of good quality, insofar as its purity was concerned. For seed, Kentucky bluegrass of certain improved varieties now is grown in cultivated rows, sometimes under irrigation, so that the seeds are larger, drier, and easier to clean to high purity.

Some grasses, such as fescue, Sudan-grass, and ryegrass can be cleaned to high purity. Some kinds are notorious for the weeds they often contain such as quackgrass in smooth brome and the wheatgrasses, wild onion in orchard-grass, and johnsongrass in sudangrass.

Many of the native grasses common to the Plains States are even harder to clean than Kentucky bluegrass, but special cleaning machinery has made it possible to sow them with ordinary farm drills or special grass seeders. In the past, many of these chaffy range grasses were sown in their original chaffy condition, even with such crude equipment as manure spreaders. In some of these kinds, 10 percent pure seed would be considered acceptable.

Seeds of legumes, such as the clovers and alfalfa, are smooth, heavy, and free flowing, as contrasted with the chaffy grass seeds. Legume seeds range in size from the small-seeded white-clover to the large horsebean. Removing light, chaffy material from them is relatively easy. The most serious difficulty arises with weed seeds that are of similar size and shape, such as dodder in the small-seeded legumes like alfalfa, red clover, and lespedeza. Legume seed should not contain more than one-tenth of 1 percent of weed seeds, and the pure seed should be at least 98 percent.

Germination percentages of legumes are sometimes confusing, because the hard seedcoat on some kinds excludes water. It is not unusual that legume seeds are labeled to show 10 to 20 percent of hard seeds. Some, like alfalfa, will germinate within several weeks in the soil, but the hard seeds of other kinds, such as sweetclover and crotalaria, may remain in the soil for months and years. It is possible to overcome this problem of hard seed by putting the seeds through a scarifying machine, which scratches the seed-coat to permit water to enter.

Seeds of cereals, which include the grain crops, generally are large and free flowing. Purity as high as 99.5 percent is not unusual. Seeds of other crops and sometimes weed seeds cannot be removed, however, or can be removed only with a large loss of seed and at great cost. Germination of the cereals should be 90 percent or more, unless the seeds have been frosted, have been over-treated to control insects or diseases, or have been stored for a long period or with excessive moisture.

THE RELATIONSHIP of quality and price of seeds is not well defined. We might assume that for each percentage point of difference in pure live seeds there should be a similar difference in money value. In fact, you would not want in any circumstance to buy seeds that are below a certain quality. At that point, in the range of quality for your purposes, the seeds cease to have any value. We should assume then that the reduction in money value should be proportionately more than the reduction in quality, as expressed in the terms of pure live seeds. In actual practice, lower quality seed often is sold at prices higher than it should be, if a comparison of pure live seeds is made.

An example: Lot A may sell for 10 dollars per 100-pound bag, and lot B may sell for 9 dollars per bag. Upon checking the label, we may find lot A has 99 percent pure seed and 90 percent germination, and lot B has 98 percent pure seed and 80 percent germination. From the respective 89.1 and 78.4 percentages of pure live seed in each 100-pound bag, we learn that you would pay at the rate of 11.22 dollars per 100 pounds of pure live seed for lot A and 11.48 dollars for lot B. Measured as to their ability to produce plants, lot A is cheaper and better.

Seed is never 100 percent pure with too percent germination. The price of seeds is therefore set for a quality somewhere under this ideal. A difference of 1 percent in the measurement of pure seeds and germination is really more than 1 percent. If maximum germination is 90 percent, for instance, then 80 percent germination is not one-tenth, or 10 percent, lower in quality it is one-ninth, or about 11 percent, lower in quality. The price should be more than 10 percent less.

As I said, the percentage of pure live seeds is dependable for the purpose of determining the amount of seed needed. It is not an adequate measure of quality. It is better therefore to consider the relationship between quality and money value in terms of the more definite quality factors which are percentage of germination, percentage of pure seeds, percentage of weed seeds, and rate of occurrence of noxious-weed seeds.

The specifications used by the Federal Government for buying seed allow a tolerance or margin for unavoidable variations. Beyond that, for each percentage point of germination below the requirements, a deduction of 2 percent in the purchase price is made. For each 0.5 percent of pure seed below specifications, a 2-percent deduction is made. For each 0.2 percent of weed seeds in excess of the amount allowed, a deduction of 1 percent of the purchase price is made.

It would be difficult to arrive at a mathematical deduction in purchase price that may seem reasonable for seed with noxious-weed seeds compared with seed containing none. The Federal specifications provide for rejection of any seed containing certain noxious-weed seeds at a rate of more than 45 per pound.

THE USE to be made of the crop determines the kind, variety, and quality of seeds that will best serve your needs.

One may advise buyers always to select seed of the highest quality and of the most desirable variety. That is not always possible. A choice may be forced upon the buyer as to the kind, variety, freedom from weeds, and germination. The alternative on occasion may be that of choosing between high germination of the second-choice variety and the first-choice variety that contains seeds of noxious weeds. These selections may be decided on the basis of facts available to the buyer, but in many instances it becomes a question of personal choice and opinion.

The amount of seed bought may may make it worthwhile to send a sample to a seed-testing laboratory to check the quality. One should remember that variations in results of tests must be expected, but the range of reasonable variation is pretty well established. Seed analysts can give advice.

WALTER A. DAVIDSON is Director of the Grain Division of the Agricultural Marketing Service. He is a graduate in agriculture of the North Dakota State College. He was formerly Chief of the Seed Branch, Agricultural Marketing Service, and past President of the International Seed Testing Association.