Wisconsin State Horticultural Society, 1912: 173-183
SEEDS AND SEED SELECTION
DR. W. W. TRACY

BUREAU OF HORTICULTURE,
U. S. DEPARTMENT OF AGRICULTURE.

The basic purpose of all cultivation of the soil is control of the character of the vegetative growth rather than a mere increase of its volume. At a cost of labor which it is hard for us to fully appreciate, our fathers cleared away the forests and broke up the prairies that the vitalizing energy of sun and soil might be directed to the production of the particular kinds of plants and in just the proportions they thought would be most useful. It is doubtful if there are more individual plants or those of better development growing within ten miles of Madison to-day than there were one hundred years ago, but there has been a gain and our orchards and fields make possible for us an incomparably broader, fuller—yes better life, than was possible for our parents on the prairies and amid the forests of the last century.

In the ease of your orchards and vineyards you direct the productive energy of sun and soil to the growth, not only of fruit but to the particular variety of fruit, which you think would be most satisfactory. With perennial plants like those of the apple and strawberry, we secure uniformity as to the character of the crop by multiplying the vegetative parts of some plant of proven desirable character, and thus secure a large culture of what is in reality but a single plant which we distinguish by a varietal name that always stands for any vegetative part of that particular seedling, and so for a definite and unchanging varietal character. To accomplish this uniformity of product of a desired type we do not hesitate to expend a considerable amount of labor to multiply by grafting, by cutting, by runners and in other ways the vegetative parts of that desirable plant, and to set them by themselves in orchard or plantation. A ten acre orchard of healthy and productive apple trees, each bearing fruit of different form and color, though all good market quality, would not be very valuable. The fruit of a strawberry plantation which was a mixture of plants of one hundred or even of twenty of the very best sorts, but of different size, shape, color and time of ripening would hardly be worth gathering.

It is true that the volume and character of the market product of such so-called different plants of the same sort may vary because of differing climatic and soil conditions. A Grimes Golden apple grown on your farm may be very different from one grown by your neighbor, or your strawberries ripen earlier and be sweeter than his, though his bed was set from yours, but such differences are most likely to be because of differing soils and cultural conditions. Were we to interchange conditions we would interchange the character of the fruit but you would not claim that such differences showed that varietal character of an apple tree or strawberry plant was of no importance.

In the case of grains and what are generally classed as farm crops which are propagated by seed, while there may be varietal differences which adapt sorts to different conditions of soil and climate or call for slightly different cultural methods in order to secure the best results, yet with possibly a few exceptions, such differences are not of such a character or so marked that a mixture of sorts in the same culture affects the cost of growing and harvesting or the value of the market product.

With garden vegetables, however, variations as to habit of growth, time of maturity, appearance and quality of market product, are much greater than in grains or even in litany fruits and have a much greater affect upon the market values, in some eases different varieties of the same species require for their best development double the space, and do not reach edible maturity nearly as quickly as others and very generally a superior lot will bring double, triple or even five times the price of inferior offerings of the same varieties. Seed of the Champion of England and American Wonder peas are so near alike that even an expert cannot be certain as to which variety many individual peas in a mixed lot of seed really are, yet the two sorts differ so much in habit of growth and time of maturity that if mixed and planted together there will be practically a total loss of one or the other. As most vegetables are annuals at least as far as crop is concerned it is not generally practical to make plantations of vegetative parts of a single plant, the varietal character of whose market product under the same cultural conditions is unchangeable. Any crop of vegetables is the product of innumerable individuals, each distinct, and the problem of uniformity of product is not as in fruits one of multiplying vegetative parts of an individual of predetermined character but of multiplying individuals which shall be as like as possible in the character of their market product. In one case a varietal name stands for a certain immutable combination of the same varietal tendencies as developed in a single plant. In the other it stands for a mutable conception of a combination of different varietal tendencies as developed in innumerable plants; each the product of an individual seed which is simply a minute plant packed for transportation.

We will leave to wiser heads the discussion of the basic cause of plant variation; why no two plants or even no two leaves on the same plant are precisely alike, and will consider the fact as an element in our problem of how to secure seed which will develop into plants of any desired varietal character. First, the varietal character of every plant with all its limitations and potentialities of development was unalterably fixed in the seed from which it grew before that seed left the parent plant, carrying within itself a balanced sum of varient tendencies inherited in different relative strength from each of its ancestors of an unknown number of generations. We can predict the varietal character of the plant any seed will develop into only in proportion to the correctness of our knowledge of what their varient tendencies are and the shrewdness of our guess as to their relative strength. We have to depend upon a mere guess, for even the most perfect knowledge of the general working of Mendelian laws does not enable us to predict with certainty the exact proportionate strength of different inheritances and the resultant character of each seed, for every seed has an individual character anti potentiality of development which may not be the same as other seeds from the same plant or even from the same fruit. I once had a very exceptionally line plant of a certain variety of tomato; I rooted cuttings of this plant until I had enough for nearly an acre which was located where there was little probability of any of the flowers being fertilized from other flowers. Although the fruit product was comparatively small it was the most uniform lot I ever saw, each fruit like every other. I saved a lot of seed, supposing it would be of exceptional value. I was disappointed, for cultures grown from this lot of seed showed much greater variation than those from ordinary stocks of seed of the variety. Why? Because my original plant was nearly perfect on account of a very fortunate combination of the proportionate influences inherited from each of many ancestors, while the seeds of that perfect plant though made up of the same inherited influences but in different proportionate strengths gave me varying results. Although by no means a perfect illustration, a suggestive one of the way the character of two plants made up of the same inherited tendencies but in different proportionate strength may vary, can be drawn from the digits—654 which in their relative positions always stand for six hundred and fifty-four—a number which cannot be expressed by the same digits in any other order or if we take away or add a single digit even if it be a one.

Variation in all plants is so often the result of conditions of soil and climate and so largely a matter of degree that it is easy for one to find instances where numbers seem to substantiate any claim of variation that may be made, but I will give an illustration from experience showing how the relative importance of the characteristics like the value of the digit does not lessen its importance in determining the value of the sum.

Some fifteen years ago there was developed a strain of a certain variety of watermelon in which an invisible but very desirable quality was correlated with one which was easily seen and shipments of this stock commanded extra prices because of quality. A block of about one-half acre was planted with this stock and about three-fourths of all of the fruits produced showed the peculiarity. Seed from the most perfect fruits of about forty of the plants in which this was best developed were saved separately and numbered in the order of the degree they showed the variations, although it was so well developed in all of these selected fruits that when the numbered melons were mixed those who had made the selection and given the numbers were unable to replace the fruits in the same order as before. The seed from each melon was saved separately and correspondingly numbered and about one-fourth of each lot of seed was planted and labeled with the number given the fruit. The general uniformity of the crop of melons was noticeable, and most of the lots showed more or less of the desired peculiarity but in only three of the lots, those numbered 3, 7 and 11, was it well developed in every fruit. All the other lots contained more or less fruits in which it was absent or imperfectly developed; and strange to say the lots from the fruits which had been marked number one and number five in order of merit in our previous year's selection contained the greatest number of fruits in which the desired quality was lacking.

The next season the balance of the seed from the fruit which had been marked number 11, was planted where there was little liability of the blossoms being crossed from other melons and a critical selection made of the fruits which not only had this peculiarity well developed, but which were as uniform as possible as to other desirable qualities, and their seed product saved for stock. The next year this seed was used to plant a 40-acre field, and when this was in full fruit, a bright, observant young man and I spent over two hours hunting in vain through it for a single fruit in which the peculiarity upon which our original selection was largely based was not well developed. It is but fair to say, however, that when the time came to gather and seed the fruit, twenty of them were rejected because they did not show this variation. Further experience with this stock shows the necessity of continued selection. Some nine years later I visited three large fields which were planted with seed which was only the fifth generation from that nearly perfect stock, and in not more than half of the fruits was the desired variation noticeable. The growers had thought that the seed from the original field was so perfect that it needed no attention and it had gone the way of all neglected things.

The characteristic upon the more or less perfect development of which the selection was made was of itself of absolutely no value, but it was very generally correlated with invisible qualities of the greatest value. Very often one has to rely upon such correlations in making his selections. This calls for fullest acquaintance with varietal differences and their correlations. Thus the shape of the leaves of a strap-leaved turnip or the shade of color in squash which might or might not be of importance of themselves may through correlation be the visible indication of invisible qualities which do most materially affect value. In the case of the watermelon referred to, the visible character was a peculiar yellow color where the melon rested on the ground. I learned of its importance from an old darkey living in the region where the variety originated, who told me that if I wanted a real sweet melon I must get a "Yellow belly" and the variety was generally known where it originated as the "Yellow belly.'' Yellow skin, however, does not always stand for quality, in some other sorts a white skin indicates the best stock. In a sort which at one time was a very popular shipping melon, the peculiar character of the blossom end depression was an indication of shipping quality, and its loss of the sort in favor as a shipper has kept pace with the disappearance of this character in the stocks offered by seedsmen. The value of varietal description is largely based on such sort of correlation of qualities,—a correlation which is rarely in accordance with any known general law and can only be learned by careful study of visible varietal differences and what they indicate. We think there is no branch of horticulture which offers such an open, and as yet, comparatively unworked field for profitable study as this. And it is an unworked field. I recently spent several days among growers of seed sweet corn in Connecticut and it was astonishing to see the want of uniformity as to the conception of what was the most desirable form for even our oldest and most widely grown varieties. I have here some sample ears of sweet corn which illustrate this.

No. 1 No. 2 Stowell's Evergreen Golden Bantam
No. 1. As originally sent out by Josiah Crosby.
No. 2. As grown in 1911.
Ideal selections made by two brothers. Two ears as grown by two seed growers living within five miles of each other, Dec. 27, 1911.

Photograph 1. 1. An ear of Crosby Corn as near as possible like one selected in 1887 by Josiah Crosby the originator, as representing his ideal of the variety. 2. An ear selected in 1911 by one of the best growers as the present ideal of the sort.

Photograph 2. Two ears of Stowell's Evergreen as selected by two brothers of the same seed growing firm without consultation as to their ideals of the sort.

Photograph 3. Two ears of Bantam as selected for stock seed by two of our best growers living within five miles of each other.

The fact that such differences in the character of the seed corn do not of themselves affect the value of the market product does not lessen their importance, for one at all familiar with vegetative forms will recognize the possibility that they are correlated with invisible characteristics which do not materially affect market value. Nor are such differences in varietal forms of stocks offered under the same name peculiar to sweet corn, they are to be found in nearly all stocks of our common garden vegetables. With such differences as to the ideals of the sort existing among the seed growers who supply our dealers with stocks of the same variety, which are liable to say the least to be mixed together before they are offered to the public—is it surprising that the planter should find that his crops are so lacking in uniformity of varietal character as to increase the cost of growing and marketing and lessen profit?

Let us turn now from this general discussion of present conditions to that of how we may grow better strains of seed than those in general use. First, .I want to refer to a condition not cry generally recognized, and still less understood, and that is the adjustment of plants to local conditions, which adjustment iii some genera at least is carried in the seed. The Office of Horticultural Investigations of the Department of Agriculture have carried on some experiments which make this seem evident. Starting with a certain strain of two varieties of Sweet Corn—there were grown for five generations in different locations, each year the seed grown in all locations was selected to a common type and the local grown seed planted in comparison with seed of the same original stock and varietal selection grown elsewhere, and in nearly every case in what was practically thirty different tests, the local grown seed gave larger and better yields. Trials made by others gave similar results with plants of certain genera. In our experience, however, plants of other genera, notably the Curcurbitaceae do not transmit such local adjustment through the seed. Enough has been proven, however, to show the general superiority of seed grown under the same environment as that of the crop desired. How then can we secure seed of the most uniform possible varietal character and which is adapted to our own conditions?

The first step in the securing of such a stock is to form a very clear and definite conception of the exact varietal form we desire. I think it is essential that this be written out and be frequently referred to in order to avoid the indefiniteness and change in the type selected, which I think is the cause of much of the variability now so common. It is a practical impossibility to write out a description which will enable the reader to recognize with certainty the exact type of plant the writer had in mind, but I also know by experience that the attempt to write such a description will always sharpen the writer's conception of the exact form he is after and be of great assistance to him in holding to that exact model from year to year, as it is essential that he should do in order to establish a strain of seed which will be of the exact character desired.

The second step is to select a score or more plants each of which come as near the exact ideal of the sort as possible, avoiding the temptation to include even superlative plants that differ in any way from the ideal of the variety. We should save the seed of each plant separately even where two or more of them are so near alike as to be indistinguishable.

The third step is to make growing tests of samples of each of these lots of seed, planting a liberal quantity of each under uniform conditions of soil and culture. As the plants develop they should be carefully studied and the lot in which each and every plant is the most like our ideal of the sort should be selected, resisting the temptation to take a lot which may contain superlative plants of the sort but also some which are inferior or which, though superior, are of a different varietal type from that aimed at.

In order to guard against the possibility of a hidden variant in our select plant and the many mishaps to which the seed grown is always liable, it is well to select one or more plants, as substitutes for our first selection. It is rarely worth while to save any seed of this trial planting, even that of our selected plants, at least not as a basis for stock seed.

The fourth step is to plant the reserved seed of the plant, which our trial has proved to be the best, and also when practical, some of the substitute plants where they will not be liable to cross-fertilization from other plants of the species. If our first selection was as good as our trial indicated and we have in, it a plant of the exact type desired, which did not carry a hidden tendency to variation inherited from some earlier generation, the battle is won, for all we need to do is to multiply the descendants of this pure plant without them being contaminated through pollen from sonic plant of a different varietal character.

Human control of conditions is imperfect. Man must still eat his bread in the sweat of his face and it is quite possible that, in spite of our failure to discover it in our trial, our foundation plant did carry a bidden variant which may crop out in some future generation so as to spoil our uniformity, in which case all we can do is to try one of the substitutes, or commence over again. Again it is generally impossible to avoid more or less crossing in succeeding generations so that sooner or later, we are likely to lose any pure strain we may have obtained, but in the case of most vegetables, we can secure several crops before this occurs and, in the meantime, we should have built up as good or a better strain to take its place.

I do not want it to be inferred from what I have said that I am in any way antagonistic to seedsmen for I am not, and believe that they are quite as well informed as to the real wants of their customers and fully as desirous of supplying just what they want and are as willing to pay for as any class of merchants, but they do and will continue to get their stocks where those which will best satisfy their customers as to quality and price can be secured at least cost, and will strive to furnish better and more uniform stocks as fast as their customers demand and are willing to pay for them. What I have tried to show is the importance of the use of seed of more uniformity as to varietal character than that which most growers are now satisfied with; that the great need and hope of betterment is not in wonderful "Burbankian creations'' of fabulous productions and of hitherto unknown superlative quality, but in truer and better stocks of the sorts we now have.

DISCUSSION.

Mr. Toole: I suppose that each one of us who has listened to this valuable paper has been impressed with the importance of some particular point that has been brought out. I am impressed with the importance of recognizing the value of the possibilities of the seeds of some particular plants. A few years ago I read before this society a paper on individual prepotency in plant breeding and each year in our pansy seed growing, I am more fully impressed with the value of that idea and if we are trying to improve one of the newer varieties we prepare them, several seeds from each plant separately that we are selecting from. if we have several that may he up to our standard, knowing that the result will he that some one plant for some reason that we do not know will have a larger proportion of what we aim for than will the progenitor of some other plant. My son has carried on experiments in corn breeding and that same result has been found, that for some reason that we do not know there are some plants that will give better results than others, that the progeny is more apt to he like the parent than are the results from some other plants, so it is of great value to you to carry out that when you. are trying to improve any variety, or to keep up the standard of any variety. Because, if you save the seed of each plant separately and experiment with them, if you have found something that has a tendency to prepotency, you can take it up. If you save for a number of the best, and then try to get your best, your best plants are simply diluted with the others.

Mr. Tracy: May I speak of another illustration? When I was living in Grand Traverse county, I spent a half day writing out an examination of just what smut nose corn should be, that was the sort we had to grow. Then, after spending a half day, I went into a seven acre field close by and spent the whole afternoon trying to find ten plants which came up to my ideal, ten plants in which the stalk and leaf and the proportion of leaf to the stalk and the height of the stalk and the character of the ear were just right. Well, I had to quit, after spending the whole afternoon in that way I did not succeed in finding the ten plants that I desired that came up to my ideal, that scored as high as I thought they ought to. I could not find ten. I only had seven. I saved each of those plants separately and planted them by themselves and continued them for seven years, while I was at that place. The seventh year I exhibited at the State Horticultural Society. I took the same field and I tied sixteen stalks one to another and took it to the County Fair, every stalk of those sixteen growing in consecutive order was as near to my ideal as the seven plants I had tried to pick out of that whole seven acres seven years before. I had secured that uniformity by simply that process, and I could give scores of such instances. I tell you, gentlemen, just as has been spoken of, there is a wonderful prepotency in certain plants to breed true and those are the ones which should be worked for and stuck to.