No farm practice yields more beneficial results than the careful and intelligent selection of seed for sowing. The planter who raises a special crop like tobacco, cotton, wheat, or corn usually looks carefullt to the quality of his seed, while the truch farmer is even more particular in this respect, paying a very high price for the best obtainable article. Nevertheless, it is true that in general practice, especially in the case of garden and forage plants, there is frequently very slight attention given to the real worth of the seed used for planting, and not infrequently the grower sells his marketable alfalfa or clover seed, for instance, and reserves the remainder, consisting posibly of screenings, for his own use. The folly of such a proceeding can not be too strongly condemned. Weak or otherwise inferior seed, if it comes up at all, often gives rise to sports and new varieties, and so far may be valuable for experimental use; only the very best seed, however, should be employed in the production of staple crops. Any other practice is poor economy. The grades extablished for clover and grass seed known as "prime," "choice," "extra prime," etc, take into account only its purity, that is , its degree of freedom from chaff and dirt, weed seeds, and other foreign matter. The buyer is assured in the most general terms (but not guarantied) that the seed he gets is "pure, reliable, and true to name," and selected (by the seedsman) with "reasonable care." No intimation is given, however, as to the proportion which will germinate. It is assumed that any deficiency in this respect can be readily overcome by sowing an extra amount. The still more important points as to the origin, size, and weight of the stock are seldom taken into acount.

Another serious drawback to the selection of good seed is the common practice of waiting until about time for sowing before buying. It is then too late to ascertain its origin and history or to test its vitality, even if the planter had a desire to do so. If the cultivator would secure the best possible results from his labor, the seed should be bought by sample in the fall or winter before planting. First of all, it should be examined for purity, and then a sample home germinating test should be conducted. If the sample is pure and of good germinating capacity, the purchase may be completed, after which a careful sorting should be made preparitory to planting in the spring.

    The principles governing seed selection depend largely upon the kind of seed and the objedct of the crop, whether size, quality, or earliness of the latter is most desired. It also makes some difference whether the plants are to be grown for forage or seed. Certain climates, soils, and fertilizers tend to seed production rather than to vegetative development, and a plant may be cultivated and selected for its seed-producing capacity until a strain of seed is obtained which tends to yield plants possessing similar seed fertility. If quality rather than quantity of ceop be the object, the selection of seed must follow a certain line in order to secure plants of the desired characteristics.

Seed may be selected according to its origin, color, form (considering especially whether it is plump or wrinkled), size, and weight, it being taken for granted that the selection shall be made only from sound, pure, and germinable stock. It is thought by some that the value of seed varies in certain cases according to the part of the plant or fruit from which it comes.

An experiment made in Georgia with cotton showed that the bottom bolls produced seed which gave a heavier yuield than that from the upper bolls, the yield in the former case amounting to 1,043 pounds of seed cotton per acre as compared with a yield of 750 pounds in the latter. This was undoubtedly due to the fact that the lower bolls contained larger and heavier seed, rather than because the seed came from a certain part of the plant. Many trials have been made of corn selected from the tips and butts of the ears. Sometimes one and sometimes the other kind of kernels give the best crop. It is quite likely that this variation results from the difference in size and weight of the different kernels taken frm the same ear. In the case of the parsley, carrot, parsnip, and other umbelliferous plants, it is commonly supposed that the central stalk produces the best seeds. This may be due to the fact that such seeds are frequently larger and heavier than those from the lateral shoots. the degree of maturity of seed when harvested is an important factor in determining its value. Many experiments have been made with immature seed, resulting in the conclusion that such seed produces as a general thing smaller and less vigorous plants. Professor Goff, of Wisconsin, has shown that by the use of immature tomato seed there is also a tendency to increased earliness in the maturity of the fruit. By means of continued seed selection, plants may be so developed as to show a certain flavor, capability of resisting disease, general hardiness, earliness, superior content of sugar, oil, starch, gluten, etc.

The manner of selecting seed varies somewhat with its shape, weight, and size. If heavy seed is desired, a salt solution may be used, of such density that only seeds of a desired wieght will sink to the bottom, while all the lighter seed and undesirable matter can be skimmed off and fejected. This method is open to several objections, among them being the fact that heavy seeds do not always sink, owing to bubbles of air which sorround them or to the flat surface which some species present. This however, may be obviated to a slight extent by previously boiling the water of which the solution is made. Furthermore, unless dried promptly or sown at once, such seeds may lose some of their vitality. A better way to obtain heavy seed is by making use of the centrifugal principle, applied by running the seed through some kind of apparatus which throws the heavier seed to a considerable distance, while the lighter seed and chaff drop near the machine. By the use of a current of air the same separation may be secured, in this case the lightest material being blown away. The common method of selecting large seeds is by the use of sieves, either by hand or placed in any common fanning mill.

The principal object in using the ordinary fanning mills is elimination of chaff and other foreign matter, although some of the lightest seed is plown out by the fans. It would pay the farmer when he is cleaning up seed for planting to work his seed-cleaning machines in such a manner as to blow or screen out a breat deal of the light and small seed, retaining only the largest and heaviest for planting.

Seeds are sometimes cleaned and sorted by running them through a thin metal cylinder placed in a slightly inclined position. This cylinder is provided with a series of holes of different shapes and sizes, which allow certain seeds to drop through at certain points. Machines embodying this principle are sued considerably in Europe and to a small extent in this country.

The fact needs emphasis, however, that no systyem of seed selection by mechanical means alone is adequate, although such selection, if properly practiced by the agriculturist, would invariably bring him a decided gain in the size or quality of his crop. Thorough selection must begin with the plant itself. Only thyose plants should be chosen for seed purposes which come the nearest to the type which is to be reproduced. Such plants are to be harvested and kept by themselves. After their seed is thrashed and cleaned, another and rigid selection, based upon size and weight, should be made.

In the choice of seed the place of its production should receive very careful consideration. Much of the failure to secure a desired crop of vegetables or forage plants is due to the fact that the buyer of such seed usually has no information whatever as to its origin. The siol and climate where it was produced may have been very differentfrom its own, and the seed be totally unfit for use on this account. A great deal of controversy has arisen from time to time over the alleged superiority of Northern-grown seed, and many dealers make a great point out of the statement that their seed is Northern grown. This is not a question of section alone, nor is it true that Northern-grown seed is always superior to that rased in other latitudes. As is well known, certain poants thrive better in one locality than in another. Plants adapted to Northern climates or high latitudes where the seasons are short mature more quickly than if grown under different conditions, and hence yield a strain of seed which in time tends to produce quickly maturing plants. However, such seed frequently "runs our" quickly when planted in a different climate and gives rise to very different strains form the original stock; hence, a constant renewal is necessary to maintain the type desired. In many cases by a system of careful cultivation and selection a desired strain may be secured and thoroughly acclimated, so that the introduction of outside stock becomes unnecessary. The fact that many kinds of imported seed do not produce as good crops as home-grown seed of the same variety is to some extent due to a difference of climate. The imported seed, while perhaps cheaper in the first instance, owing to the less cost of production, may be much dearer in the long run, since it is seldom so pure as American-grown seed, and frequently gives rise to noxious plants which the buyer neither desires nor pays for.

Soil, as well as climate, impresses seed with a particular character. It is not necessarily the most fertile soils which furnish the most productive seeds. If seeds are transferred to a different kind of soil from that upon which they were grown, although the climate be the same, a marked difference in crop is frequently noted. Experiments in growing oats have shown that certain varieties raised on a light soil were the most productive if sown on a similar soil, while the same varieties of seed if grown in a heavy soil showed a preference for heavy soil. These facts indicate that, in many cases at least, the farmer will get the best results from seed which he has grown himself under conditions well known to him.

The Department of Agriculture frequently receives requests from European seedsmen for seed of various kinds raised in a part of the United States the climatic and soil conditions of which correspond most nearly to those of their own localities, thus showing an appreciation of the value of a knowledge of the origin of seed. Most of our own seedsmen show a similar interest in knowing where their seed was grown. Unfortunately, however, this interest prevails at the present time among the buyers of seed only to a very limited extent. The farmer should secure from the dealer whenever possible a statement of the origin of the seed which is offered for sale. Until such requests become much more common than they are now, seedsmen will continue to offer seed accompanied by no information save its name and brief directions for planting.

Seed should be selected with reference to its ancestry as well as to the place and conditions under which it was grown, or its individual characteristics. Plants, no less than animals, inherit the qualities of their forerunners, and this applies to seed as well as to the plant taken in its entirety, especially if grown for seed alone, as in the case of the cereals and some legumes. Unfortunately the ancestry of seed can rarely be traced by the purchaser.

The main purpose of this article is to show in a brief way tht advantage of using only large and heavy seed for planting, and if possible, to establish the principle that it will pay in many cases to buy a larger quantity of seed than is to be used, in order that only a larger and heavier seed may be selected for planting.