Volume Number: 1  2  3  4  5  6  7  8  9  10  11  12 




Very good illustration of directive plant breeding is furnished by the case of the Empson peas. This was a case in which I received an order for the development of a new variety of pea that would fulfill certain definite specifications, somewhat as a manufacturer of cloth or of electric dynamos or of machinery of any sort might receive an order for a new product to meet a special condition. It is gratifying to record that I was able to meet the specifications, and "deliver the goods," as a manufacturer might say, about as accurately and satisfactorily as if the product had been one to be turned out by factory machinery instead of by selective breeding of a living plant. The specifications were these: A variety that shall mature all its pods at the same time; bearing individual peas of reduced but uniform size, sweet, and of superior flavor. Here, it will be observed, there are several quite distinct characteristics to be borne in mind. Perhaps the most important, or at least the ones most difficult to attain and fix, were the uniform time of ripening and uniform size of the peas themselves. How these difficulties were met will be detailed presently. First, however, let me tell just how it came about that the order for peas having just these specifications was received.


The order was given by a large canning factory, located originally in Colorado, but now having branch factories in other regions, with capacity to handle in the aggregate forty-six thousand cans of peas per hour. The head of this company, Mr. J. H. Empson, is a man who has made a study of his public, and who aims to give the public what it wants. He discovered that there was a demand for canned peas of very small size. This had come about, probably, through the example set by the French, who can the peas when they are half grown, at which stage they appear to be sweeter than when more fully ripened. The American public developed a liking for these small peas, and a willingness to pay more for them than for the larger ones, but no American canner could duplicate them in size and quality. The American canners are themselves convinced that peas of medium size are really better; but they were desirous that the public should have what it wanted. So it came about that I received a letter from the management of the canning company asking me to undertake the work of developing a pea that would meet the specifications as to size, and yet would mature in such quantities and with such uniformity that there would not be great loss in handling, as there would be if the pods matured at different times. The reason that this specification is imperative is that peas for canning, according to modern methods, are not gathered by hand. Indeed they are not touched with the hand at any stage of their existence, even in planting. The crop must be ready all at once, because the vines themselves are harvested. A machine is drawn along the rows cutting off the roots about an inch underground, and raking four rows together in a windrow. Cutting below the ground keeps the peas fresh and also ensures getting the entire crop. A wagon immediately follows, gathering up the pod-laden vines like a load of hay, and hauling them to the factory, where they are fed by machinery into a sheller, which consists of two big cylinders with vulcanized rubber cups on their surfaces, so arranged that the air pressure splits the pods open without crushing them. The peas roll down an inclined plane with perforations of different sizes, and are thus automatically sorted into five grades, just as oranges of different sizes are sorted in California. The peas all fall into clean running water and are immediately canned without being touched. It may be interesting to add that a factory of this type has a record of putting canned peas on the shelves of the grocer within two hours of the time when they were growing on the vine in the field. Peas in cans under these circumstances may be fresher than those purchased in the pod usually are. These details as to canning obviously have no direct bearing on the methods of the plant developer. But they explain the specifications that were given along with the order for the new variety. In attempting to meet the specifications, I followed the methods of rigid and systematic selection. There was no occasion for cross-fertilization, as the peas were of superior quality, and showed enough variation as to all of the desired characteristics to offer material for selection. My plan was to pick out in successive generations the vine that came nearest to meeting specifications as to number of pods, uniformity of ripening, and small size as well as uniform size of the peas themselves. It was necessary, as in some other experiments of a similar kind, to watch the individual plants, selecting the very best individual plants, and harvesting them by themselves, counting the pods and counting the peas, and making careful record of results. Fortunately it is possible with the pea to raise two crops in a season. Thus I was enabled to progress very much more rapidly than otherwise could have hoped to do. We could do two years work in one. So we were able to deal with six generations of peas in three years. And yet by that time the undesirable qualities had been so systematically excluded and desirable ones so persistently sought for that the educated pea vines fulfilled the specifications exactly. I find in my files a letter bearing date of February 29, 1908, that may be quoted here as summarizing the results of the experiments: "By express today," I wrote, "I send you all the peas raised from the one best of all my selections, This one is the one which produced the most peas to the pod, the most pods to the vines, had the most uniformly filled pods, and in all respects was the most productive and best; on the whole, the best pea, taking quality, quantity, and everything into consideration, which I have ever seen. They are fifteen percent smaller on the average. One other thing which I have added to them is that they are sweeter than the pea which you first sent me. They all came from one single vine which was the best in all respects and the seed has been reselected through six generations."


Of course, the selected pea, as thus produced, existed only in small quantities. But it had been fixed as to type and could be depended on to breed absolutely true. It was necessary, however, for the company to multiply the seed for a number of years before there was enough of it in existence to use for the purposes of the canner. By growing the crops in California, however, where from two to four crops could be raised each year, and by using the entire product for the seed in successive years, the progeny of the single vine from which I developed the new variety had been multiplied by 1912 so that material enough was at last in hand to plant hundreds of acres and supply the cannery with the small, sweet, uniform-sized and uniform-ripening pea that was desired. I have cited this case in detail, not because it is of exceptional importance in comparison with other of my plant developing experiments, but simply because it illustrates the possibility of developing quite rapidly a particular plant to meet a specific commercial need. But to understand fully the conditions met even in this single experiment, it is necessary to add that I did not confine attention to the production of the single variety just described, even in the line of experiments that were undertaken specifically for the purpose of producing that variety. On the contrary, while scrutinizing the vines for small peas of uniform size, I kept vigilant watch also for other vines that varied in the opposite direction.


By carrying forward several series of selections at the same time, a number of varieties were simultaneously developed that differed widely both from one another and from the original stock. I found, for example, in the observation of the early generations grown from the seed, that some plants would produce four or even five times as much as others. This habit of productiveness was carried to the next generation with a good deal of certainty. So it proved possible, by careful selection, in three years, to develop new forms of peas which produced regularly four and five times as much as the average production of the parent form. Of course, this quality of productivity was combined with the various other qualities and was manifested in the perfected pea that was delivered along with the letter just quoted. But there were other qualities which obviously could not enter into the combination, because of variation in exactly the opposite direction from the one in which we were developing the little canning pea. Thus, for example, one variety instead of having small peas had exceptionally large ones. Another variety produced lozenge-shaped peas. These seemed to be unusually sweet, and as they were also among the most productive, I made two strains of this selection alone. One of these is a very large lozenge-shaped pea, circular, and indented on the flattened sides. Both are practically fixed, coming true to type from seed. In point of fact, by having different ideals and bearing them in mind all along, I developed four strains of new varieties that the canners were glad to purchase, in addition to the one that they had specifically ordered. And all this was done, as just noted, by selection, without the aid of hybridizing experiments. It should be explained that the pea is normally self-fertilized, so that there is the closest inbreeding, and it is therefore relatively easy to fix a type. Moreover the pea is a very pliable plant, producing new varieties with little care and labor as compared with many other plants. Although I have devoted much less time to it than to many other plants, I have developed numerous varieties that are specially modified for color, for productiveness, for size, for quality, or for resistance to mildew and other affections. And other experiments are under way that will probably lead to still further developments.


Although much may thus be accomplished with the pea by mere selection, it should be remembered that this plant offers exceptional opportunities also for development by hybridization. In particular it should be recalled that the extraordinary experiments through which the Austrian monk, Mendel, made the discoveries that have created such commotion in the biological world, were made with the common garden pea. Reference to these experiments has been made more than once, but it will be worth while to examine them a little more in detail in the present connection. The discovery that Mendel first made, to which we have already referred, was that certain qualities of the pea are grouped into very conspicuous pairs. His investigation led him to believe that there are at least seven differentiating characters that could be relied upon to reproduce themselves with certainty in the offspring of the pea. These characters, which he came to speak of as "unit" characters, are the following: (1) The form of the ripe seed, which may be roundish, either with or without shallow wrinkles, or angular and deeply wrinkled. (2) The color of the reserve material in the cotyledons or little leaves that first appear when the seedling comes out of the ground; the colors being pale yellow, bright yellow, orange, or green. (3) The color of the seed coats; white, as is usual in peas with white flowers, or gray, gray-brown, leather-brown, with or without violet spots, etc. (1) The form of the ripe pods, whether inflated or constricted or wrinkled. (5) The color of the unripe pods, whether light or dark green or vividly yellow, these colors being correlated with colors of stalk, leaf, vines, and blossoms. (6) The position of the flowers, whether axillary or terminal. (7) The length of the stem of the plant itself, whether tall or dwarfish. It is obvious that in each case the different qualities named are antagonistic or mutually exclusive. The seed cannot be at the same time round and angular; it cannot be at the same time smooth and wrinkled; cotyledons cannot be at once yellow and green; the pods cannot be at once inflated and constricted. And as each race of peas, when inbred, holds true to its type, there was opportunity to observe the effects of crossing the different races in relation to these different fixed characters. The results Mendel obtained have already been outlined, and more than once referred to in this and in previous volumes. It will be recalled that, as regards the various pairs of antagonistic characters, he found that one or the other proved prepotent or dominant in the first generation; but that in the second generation (when the first generation hybrids were inbred) the submerged or recessive character would reappear in one case in four on the average. Thus he found that in the pea tallness of stalk is dominant to shortness of stalk; that yellowness of seed is dominant to greenness of seed, etc. This was demonstrated by the fact when a tall pea was crossed with a short one all the offspring were tall, but one-fourth of the offspring of the second generation were short. Similarly when a pea with yellow pods was crossed with one having green pods, all the plants of the first generation had yellow pods; but one-fourth of their offspring of the next generation had green pods.


A second very important feature discovered by Mendel was that the different antagonistic pairs of qualities are transmitted quite independently of one another. For example, the relations of tall and short peas, blended in heredity, are quite independent of the question of yellowness versus greenness of pod. So observation may be made as to two or more qualities in the course of the same experiment. Thus if a tall variety of pea that bears green pods is crossed with a short variety bearing yellow pods, all the offspring will be tall peas with yellow pods-therefore unlike either parent. But the offspring of the next generation will show a recurrence of each of the recessive factors in one case in four, so that one-fourth of them will be short and one-fourth will have green pods. But it appeared, so far as Mendel could determine, to be a mere matter of chance-like the throwing of dice-as to the exact number of cases in which shortness of stalk would be combined with the bearing of yellow pods.


If we assume-as Mendel finally came to do-that each of the different qualities about which we are speaking is represented in the germ plasm by a definite mechanical factor which must be paired with another factor, either like or unlike itself, in order to stimulate the development of the character it represents, then at least a provisional explanation of the observed facts might be found in supposing that a dominant factor when mated with a recessive one hides or obscures the recessive one in that particular combination; but does not eliminate it. And when the factors are again mixed to produce a new generation, they are still equal in number, and if we think of the factors as tangible things-let us say like black or white checker men-it will appear that if equal numbers of each are mixed together and taken from a bag in pairs at random or blindfold, it will come about, according to the mere theory of chances, that one time in four two of the white checkers will be paired. This accounts in a crude and mechanical but on the whole a rather satisfactory way for the appearance of the recessive character-say shortness of vine or greenness of pod-in one individual out of four of the second-generation progeny. And when we apply the same reasoning to the case where two pairs of factors are under consideration-tallness versus shortness, and yellowness versus greenness in the present case-it appears that each pair of factors will follow precisely the same law, so that one in four of the second generation descendants will be short and one in four will be green; but that the same law of chances, applied to this more complex case, gives us only one case in sixteen in which two factors for shortness are combined with two factors for greenness in the same group. In other words, one pea in sixteen descended in the second generation from the tall pea with green pods and the short pea with yellow pods will have a short vine and at the same time will bear green pods. This will be a new variety. It has no new quality, but it has the old qualities in a new combination. Extending the experiment one stage further, Mendel found that the second-generation peas that show the recurrence of the recessive factor will breed true to that factor. And this, again, is quite what might be expected on the theory just outlined. For the pea that contains two factors for shortness will obviously have no propensity to grow tall, and the pea that contains two factors for greenness of pod will obviously have no capacity for the production of pods other than green. So our short pea vine with its green pods, although it represents a new variety, which, for the sake of argument we assume never to have existed before; and although it appeared suddenly as what might be considered a mutation, yet is fixed from the outset, and will breed true, and constitute an established variety. All this we have referred to in earlier chapters, and we have seen many illustrations of this so-called Mendelian inheritance in the case of a good many of our plant developments-the white blackberry, for example, the stoneless plum, and the thornless blackberry among others. But it seemed worth while to make specific reference to Mendel's work with the peas, in the present connection, in particular because this work doubtless represents the most important thing that has been done with the pea at any recent stage of its development.


It was perhaps fortunate that the Austrian monk chose the pea for his investigation rather than the bean, for, notwithstanding the fairly close relationship between these two, there is a rather marked difference between them as to their practical response to the efforts of the plant developer. Perhaps because the pea has been cultivated under varied conditions, and selected for a wide variety of qualities, this plant shows a marked tendency to vary, suggesting in this regard the evening primrose and the godetia, and the new varieties are often practically fixed from the outset. With beans it is less easy to trace and classify the opposing "unit" characters, and in practice it is often necessary to select rigidly and continuously for five or six successive generations in order to fix a new variety. An illustration of the complexities that may result when beans of different kinds are crossed was given me at the outset of my work as a plant developer.


Almost my first experiment in hybridizing was made by crossing the horticultural pole-bean or wren's egg with another variety of pole-bean known as the red cranberry bean. The hybridization was effected with some difficulty, inasmuch as only one blossom in perhaps fifty responded to cross-pollenization and a part of the offspring seemed to lack vitality, as I succeeded in bringing but one plant to maturity. But this was in some respects the most astonishing bean plant that I have ever seen. It bore long black pods and the beans within them were as black as ink. Yet one of the parent beans had produced a crimson pod with a red seed, and the other a crimson and white striped pod, with red and white striped seed. Here, it will be seen, there was no such sharp differentiation of the color-factors for pod or seed into opposing pairs, with dominance in one and recessiveness in the other, as was shown by the peas in Mendel's experiments. On the contrary, the union of red beans with red and white striped ones produced something totally unlike either-namely, a jet black bean. But in the succeeding generation the offspring of the black bean showed a breaking up into new groups of characters suggestive of Mendelian heredity. Some of them were black, some red, some speckled, and some white. There were corresponding variations also as to size and shape of the beans, some being large and some small, some round and some flat. And there was marked diversity in time of ripening. As to the vines themselves, the original hybrid showed the enhanced vitality that commonly characterizes the offspring of rather widely separated parents. The original first-generation vine (which bore the black beans) grew enormous, outstripping either parent by eight or ten feet, and rivaling the growth of a hop vine. The vines of the second generation were as diversified as the seed. Some of them were long and vigorous, while others were extraordinarily dwarfed, some being so stocky as to grow pods that almost immediately touched the ground and were obliged to bend back like hairpins to find room for growth. There were corresponding variations in size, shape, and color of the leaves. All this suggests that the beans originally hybridized were themselves of very mixed ancestry, and that a large number of hereditary traits that had been blended in them were permitted to make themselves manifest through the recombination and segregation of hereditary factors. The reader cannot fail to note a similarity here between the results obtained and those that were obtained when the Persian walnut and the California black walnut were hybridized. There, as in the case of the beans, the immediate offspring were of gigantic growth, but their progeny in turn showed both giants and dwarfs. The interest of both cases (and of a number of other allied ones that will be recalled) in illustrating the Mendelian principle of the segregation of recessive factors for size, leading to the production of a race of dwarfs, will be obvious. Another hybridizing experiment with the beans, also undertaken in the early day of my investigations, brought together two varieties that are even more distantly related.


In this experiment I hybridized the horticultural pole-bean, or "wren's egg," with the lima bean. It proved exceedingly difficult to make this cross, but after many fruitless efforts I at last succeeded in securing a single pod containing four sound beans by using the pollen of the lima on the pistil of the horticultural bean. When these beans were planted, in the summer of 1872, a very strange result was observed-the beans themselves had in all respects the form, size, and appearance of the horticultural bean, but when their sprouts broke ground it was at once observed that the upper part of their cotyledons (varying from one-quarter to three-quarters of their length in different specimens) were indubitably those of the lima bean; while the lower part of each cotyledon was precisely that of the horticultural pole-bean. These parts were connected with serrated edges, which at last separated, allowing the lima bean part to drop away. Such separation, however, did not occur until the vines had made a foot or more of growth. The cotyledons on each side were divided uniformly in every case. Thus the influence of the pollenizing parent was very markedly shown in the young vines from the moment of their appearance. But after the cotyledons had fallen, all evidence of the paternal parentage of the plants disappeared. The vines did, indeed, show unusual vigor throughout the season, this, of course, suggesting their hybridity. But as to appearance and characteristics in general, with this exception, they were essentially horticultural pole-beans like their maternal parent. The experiment was carried on for several succeeding generations, but the progeny showed no reversion to the traits of the lima bean. The characteristics of the pole-bean had seemingly been prepotent or dominant to an overwhelming degree. This, then, would appear to be another case in which a new race was formed in a single generation by the mingling of two widely divergent racial strains. These hybrids of the lima and the pole-bean may be compared, in that regard, to the Plumcot and the Primus berry, to name only two of the various allied instances that have come to our attention. This is what I call a seed-graft-hybrid. This and one other instance elsewhere described are the only two similar ones that ever came under my observation, and they never, so far as I know, have been duplicated before or since. But the fact that the lima bean, the conspicuous traits of which were submerged and subordinated in the mature hybrid, should have made its influence strongly felt in the seedling at the beginning of its growth is peculiarly interesting. One recalls the similar case of the raspberry plant hybridized with pollen from the strawberry. In that case, the young hybrids at first bore close resemblance to the strawberry plant, yet subsequently shot up into the air and took on the aspects of the raspberry vine. In both cases, then, the influence of the seed plant was at first submerged but ultimately preponderant. We cannot be sure, however, that this was more than a coincidence. To determine that point it would be necessary to make a reciprocal cross. It has been pointed out that as a rule it appears to make no difference in the ultimate character of the hybrid as to which of its parents is the staminate and which the pistillate one. Yet the cases of animal heredity in which there is a marked tendency to cross-inheritance, from father to daughter and from mother to son, suggests that there may be analogous cases in plant life. In any event, the analogy between the hybrid beans and the hybrid strawberry-raspberry, each following first the staminate and then the pistillate parent, is not without interest.


After an interval of many years, during which I did not experiment further with the bean, I have somewhat recently found time to turn attention again to this very interesting plant, and have developed a large number of new varieties of unusual qualities. The recent experiments have had to do with the bush bean, and I have paid attention to a large number of attributes, including form of the plant, color of bean, and the quality and flavor. The new experiments have involved the crossing of many varieties and have brought to light many interesting developments, although none perhaps as striking as those just outlined. I have found that it is feasible to segregate and recombine the traits of different varieties of beans in almost any desired combination. Thus, for example, it is perfectly feasible to put the pod of one bean on the vine of another, quite as Mendel did with his peas. Observation will show what qualities or characteristics are prepotent or dominant even without directive effort on the part of the plant experimenter. It will be observed that in the second, third, and fourth generation plants will appear that show the pods and beans of one of the original parents combined with the leaves and vine of the other, in all possible combinations. As I have operated with about forty varieties of beans in the course of these experiments, it will readily be surmised that the number of new combinations that have been presented is almost infinite. Among the hybrid stock can be found beans of almost every color and combination of colors, black, brown, blue, slate, yellow, green, and white; mottled, striped, and otherwise variously marked and shaded. Moreover, if beans of one color are selected and planted, as a rule all the other colors appear in the progeny. One finds the offspring bearing beans that are speckled, spotted, striped, and shaded in every conceivable way. Yet beans that show this diversity of color may be quite uniform as to size of the beans and time of ripening, as well as in regard to the size and general appearance of the plants on which they grow. In other words, a certain number of characters may have become fixed while other characters are still variable. And here the obvious explanation is supplied, at least provisionally, by the supposition that the plants in question are unmixed as to their Mendelian factors for size and character of vine, but retain mixed factors for color of seed. No one as yet, however, has worked out in detail the combinations of hereditary factors for the bean as Mendel worked it out in the case of the pea. Such an investigation would constitute one of the most interesting experiments in plant breeding that any one who has time for it could undertake. It is true that the hybridizing of the plant of this genus is rather difficult, inasmuch as the flowers must be opened and the stamens removed with a pair of small forceps to avoid self-fertilization. But, on the other hand, once cross-fertilization has been effected there are obvious advantages in later generations in working with a plant that is normally self-fertilized, the pollen of which is inaccessible to insects. All in all, I think the bean offers as many inducements for improvement as any other plant under cultivation.

-Although much has been and may be accomplished with peas and beans by mere selection, these plants offer exceptional opportunities also for improvement through hybridization.

This text is from: Luther Burbank: his methods and discoveries and their practical application. Volume 7 Chapter 3