•::■■.'■,- ,■■■: - % |L p. PtU ^tbrarg SBI23 BI6 Date Due SB133 B16 ■Yl PHILOSOPHY OF CROSSING PLANTS. select the best from among the crosses, in order to maintain a high degree of usefulness and to make any advancement ; and it sometimes happens that the selection is much more important to the culti- vator than the crossing. I do not wish to discour- age the crossing of plants, but I do desire to dispel the charm which too often hangs about it. Further discussion of this subject naturally falls under two heads : the improvement of existing types or varieties by means of crossing, and the summary production of new varieties. I have already stated that the former office is the more important one, and the proposition is easy of proof. It is the chief use which nature makes of crossing, — to strengthen the type. Think, for instance, of the great rarity of hybrids or pronounced crosses in nature. No doubt all the authentic cases on record could be entered in one or two volumes, but a list of all the individual plants of the world could not be compressed into ten thousand volumes. There are a few genera, in which the species are not well denned or in which some character of inflorescence favors promiscuous crossing, in which hybrids are conspicuous ; but even here the num- ber of individual hybrids is very small in compari- son to the whole number of individuals. That is, the hybrids are rare, while the parents may be common. This is well illustrated even in the willows and oaks, in which, perhaps, hybrids are RARITY OF HYBRIDS. 53 better known than in any other American plants. The great genus carex or sedge, which occurs in great numbers and many species in almost every locality in New England, and in which the species are particularly adapted to intercrossing by the character of their inflorescence, furnishes but few undoubted hybrids. Among one hundred and eighty-five species and prominent varieties inhabit- ing the northeastern states, there are only eleven hybrids recorded, and all of them are rare or local, some of them having been collected but once. Species of carex of remarkable similarity may grow side by side for years, even intertangled in the same clump, and yet produce no hybrid. These instances prove that nature avoids hybridization, — a conclusion at which we have already arrived from philosophical considerations. And Ave have reason to infer the same conclusion from the fact that flowers of different species are so constructed as not to invite intercrossing. But, on the other hand, the fact that all higher plants habitually propagate by means of seeds, which is far the most expensive to the plant of all methods of propa- gation, while at the same time most flowers are so constructed as to prevent self-fertilization, proves that some corresponding good must come from crossing within the limits of the species or variety ; and there are purely philosophical reasons, as we have seen, which warrant a similar conclusion. 54 PHILOSOPHY OF CROSSING PLANTS. But experiment has given us more direct proof of our propositions, and we shall now turn our atten- tion to the garden. Darwin was the first to show that crossing within the limits of the species or variety results in a con- stant revitalizing of the offspring, and that this is the particular ultimate function of cross-fertiliza- tion. Kolreuter, Sprengel, Knight, and others had observed many, if, indeed, not all, the facts obtained by Darwin ; but they had not generalized upon them broadly, and did not conceive their relation to the complex life of the vegetable world. Dar- win's results are, concisely, these ; self-fertilization tends to weaken the offspring ; crossing between different plants of the same variety gives stronger and more productive offspring than arises from self-fertilization ; crossing between stocks of the same variety grown in different places or under dif- ferent conditions gives better offspring than cross- ing between different plants grown in the same place or under similar conditions ; and his re- searches have also shown that, as a rule, flowers are so constructed as to favor cross-fertilization. In short, he found, as he expressed it, that " nature abhors perpetual self-fertilization." Some of his particular results, although often quoted, will be useful in fixing these facts in our minds. Plants from crossed seeds of morning-glory exceeded in height those from self-fertilized seeds as 100 exceeds INCREASED VIGOR OF CROSS-BREEDS. 55 76, in the first generation. Some flowers from these plants were self-pollinated and some were crossed, and in this second generation the crossed plants were to the uncrossed as 100 is to 79 ; the opera- tion was again repeated, and in the third generation the figures stand 100 to 68 ; fourth generation, the plants having been grown in midwinter, when none of them did well, 100 to 86 ; fifth generation, 100 to 75 ; sixth generation, 100 to 72 ; seventh genera- tion, 100 to 81 ; eighth generation, 100 to 85 ; ninth generation, 100 to 79 ; tenth generation, 100 to 54. The average total gain in height of the crossed over the uncrossed was as 100 to 77, or about 30 per cent. There was a corresponding gain in fertility, or the number of seeds and seed- pods produced. Yet, striking as the results are, they were produced by simply crossing between plants grown near together, and under what would ordinarily be called uniform conditions. In order to determine the influence of crossing with fresh stock, plants of the same variety were obtained from another garden, and these were crossed with the ninth generation mentioned above. The off- spring of this cross exceeded those of the other crossed plants as 100 exceeds 78, in height; as 100 exceeds 57, in the number of seed-pods ; and as 100 exceeds 51, in the weight of the seed-pods. In other words, crosses between fresh stock of the same variety were nearly 30 per cent more vigorous than 56 PHILOSOPHY OF CROSSING PLANTS. crosses between plants grown side by side for some time and over 44 per cent more vigorous than plants from self-fertilized seeds. On the other hand, experiments showed that crosses between different flowers upon the same plant gave actu- ally poorer results than offspring of self-fertilized flowers. It is evident, from all these figures, that nature desires crosses between plants, and, if pos- sible, between plants grown under somewhat dif- ferent conditions. All the results are exceedingly interesting and important; and there is every reason to believe that, as a rule, similar results can be obtained with all plants. Darwin extended his investigations to many plants, only a few of which need be discussed here. Cabbage gave pronounced results. Crossed plants were to self-fertilized plants in weight as 100 is to 37. A cross was now made between these crossed plants and a plant of the same variety from another garden, and the difference in weight of the resulting offspring was the difference be- tween 100 and 22, showing a gain of over 350 per cent, due to a cross with fresh stock. Crossed lettuce plants exceeded uncrossed in height as 100 exceeds 82. Buckwheat gave an increase in weight of seeds as 100 to 82, and in height of plants as 100 to 69. Beets gave an increase in height represented by 100 and 87. Maize, when full grown, from crossed and uncrossed seeds, INCREASED VIGOR OF CROSS-BREEDS. 57 gave the differences in height between 100 and 91. Canary-grass gave similar results. I have obtained results as well marked as these upon a large and what might be called a commer- cial scale. I raised the plants during the first generation of seeds from known parentage, the flowers from which they came having been care- fully pollinated by hand. In some instances the second generations were grown from hand-crossed seeds, but in other cases the second generations were grown from seeds simply selected from the first-year patches. As the experiments have been made in the field and upon a somewhat extensive scale, it was not possible to accurately measure the plants and the fruits from individuals in all cases ; but the results have been so marked as to admit of no doubt as to their character. In 1889 several hand-crosses were made among egg-plants. Three fruits matured, and the seeds from them were grown in 1890. Some two hundred plants were grown, and they were characterized through- out the season by great sturdiness and vigor of growth. They grew more erect and taller than other plants near by grown from commercial seeds. They were the finest plants which I had ever seen. It was impossible to determine productiveness, from the fact that our seasons are too short for egg-plants, and only the earliest flowers, in the large varieties, perfect their fruit, and the plant 58 PHILOSOPHY OF CROSSING PLANTS. blooms continuously through the season. In order to determine how much a plant will bear, it must be grown until it ceases to bloom. When frost came, I could see little difference in productive- ness between these crossed plants and commercial plants. A dozen fruits were selected from various parts of this patch, and in 1891 about twenty-five hundred plants were grown from them. Again the plants were remarkably robust and healthy, with fine foliage, and they grew erect and tall, — an indication of vigor. They were also very pro- ductive ; but, as the cross had been made between unlike varieties, and the offsprings were therefore unlike either parent, I could not make an accurate comparison. But they compared well with com- mercial egg-plants, and I am satisfied that they would have shown themselves to be more produc- tive than common stock could they have grown a month or six weeks longer. Professor Munson, of the Maine Agricultural College, grew some of this crossed stock in 1891, and he told me that it was better than any commercial stock in his gardens. In extended experiments in the crossing of pumpkins, squashes, and gourds, carried on dur- ing several years, increase in productiveness due to crossing has been marked in many instances. Marked increase in productiveness has been ob- tained from tomato crosses, even when no other results of crossing could be seen. BENEFITS FROM CHANGE OF SEED. 59 b. Change of Seed and Crossing. Bearing in mind these good influences of cross- ing, let us recall another series of facts following the simple change of seed. Almost every farmer and gardener at the present day feel that an occasional change of seed results in better crops, and there are definite records to show that such is often the case. In fact, I am convinced that much of the rapid improvement in fruits and vege- tables in recent years is due to the practice of buying plants and seeds so largely of dealers, by means of which the stock is often changed. Even a slight change, as between farms or neighboring villages, sometimes produces marked results, such as more vigorous plants and often more fruitful ones. We must not suppose, however, that be- cause a small change gives a good result, a violent or very pronounced change gives a better one. There are many facts on record to show that great changes often profoundly influence plants, and when such influence results in lessened vigor or lessened productiveness we call it an injurious one. Now, this injurious influence may result even when all the conditions in the new place are favorable to the health and development of the plant ; it is an influence which is wholly indepen- dent, so far as we can see, of any condition which interferes injuriously with the simple processes of 60 PHILOSOPHY OF CROSSING PLANTS. growth. Seeds of a native physalis or husk-tomato were sent to me from Paraguay in 1889 by Dr. Thomas Moroug, then travelling in that country. I grew it both in the house and out of doors, and for two generations was unable to make it set fruit, even though the flowers were hand-polli- nated; yet the plants were healthy and grew vigorously. The third generation grown out of doors set fruit freely. This is an instance of the fact that very great changes of conditions may injuriously affect the plant, and an equally good illustration of the power to overcome these condi- tions. Now, there is great similarity between the effects of slight and violent changes of conditions and small and violent degrees of crossing, as both Darwin and Wallace have pointed out, and it is pertinent to this discussion to endeavor to dis- cover why this similarity exists. It is well proved that crossing is good for the resulting offspring, because the differences be- tween the parents carry over new combinations of characters or at least new powers into the crosses. It is a process of revitalization, and the more different the stocks in desirable characters within the limits of the variety, the greater is the revitalization ; and frequently the good is of a more positive kind, resulting in pronounced characters which may serve as the basis for new varieties. In the cross, therefore, a new combina- BENEFITS FROM CHANGE OF SEED. 61 tion of characters or a new power fit it to live better than its parents in the conditions under which they lived. In the case of change of stock Ave find just the reverse, which, however, amounts to the same thing, — that the same characters or powers fit the plant to live better in conditions new to it than plants Avhich have long lived in those conditions. In either case, the good comes from the fitting together of new characters or powers and new environments. Plants which live during many generations in one place become accustomed to the place, thoroughly fitted into its conditions, and are in what Mr. Spencer calls a state of equilibrium. When either plant or conditions change, new ad- justments must take place ; and the plant may find an opportunity to take advantage, to expand in some direction in which it has before been held back ; for plants always possess greater power than they are able to express. " These rhythmical actions or functions [of the organism]," writes Spencer, "and the various compound rhythms resulting from their combinations, are in such adjustment as to balance the actions to which the organism is subject. There is a constant or periodic genesis of forces which, in their kinds, amounts, and directions, suffice to antagonize the forces which the organism has constantly or peri- odically to bear. If, then, there exists this state 62 PHILOSOPHY OF CROSSING PLANTS. of moving equilibrium among a definite set of internal actions, exposed to a definite set of ex- ternal actions, what must result if any of the external actions are changed ? Of course there is no longer an equilibrium. Some force which the organism habitually generates is too great or too small to balance some incident force ; and there arises a residuary force exerted by the environ- ment on the organism, or by the organism on the environment. This residuary force, this unbal- anced force, of necessity expends itself in produc- ing some change of state in the organism." The good results, therefore, are processes of adaptation, and when adaptation is perfectly com- plete the plant may have gained no permanent advantage over its former condition, and new crossing or another change may be necessaiy ; yet there is often a permanent gain, as when a plant becomes visibly modified by change to another cli- mate. Now, this adaptive change may express itself in two ways : either by some direct influence upon the stature, vigor, or other general character, or indirectly upon the reproductive powers, by which some new influence is carried to the off- spring. If the direct influences become heredi- tary, as observation seems to show may sometimes occur, the two directions of modification may amount, ultimately, to the same thing. For the purposes of this discussion it is enough CHANGE OF STOCK AND CROSSING. 63 to know that crossing within the variety and change of stock within ordinary bounds are bene- ficial, that the results in the two cases seem to flow from essentially the same causes, and that crossing and change of stock combined give much better results than either one alone ; and this benefit is expressed more in increased yield and vigor than in novel and striking variations. These processes are much more important than any mere groping after new varieties, as I have already said; not only because they are surer, but because they are universal and necessary means of maintaining and improving both wild and cultivated plants. Even after one succeeds in securing and fixing a new variety, he must employ these means to a greater or less extent to maintain fertility and vigor, and to keep the variety true to its type. In the case of some garden crops, in which many seeds are produced in each fruit and in which the operation of pollination is easy, actual hand- crossing from new stock now and then may be found to be profitable. But in most cases the operation can be left to nature, if the new stock is planted among the old. Upon this point Dar- win expressed himself as follows : " It is a common practice with horticulturists to obtain seeds from another place having a very different soil, so as to avoid raising plants for a long succession of gen- erations under the same conditions ; but with all (J4 PHILOSOPHY OF CROSSING PLANTS. the species which freely intercross by the aid of insects or the wind, it would be an incomparably better plan to obtain seeds of the required variety, which had been raised for some generations under as different conditions as possible, and sow them in alternate rows with seeds matured in the old garden. The two stocks would then intercross, with a thorough blending of their whole organi- zations, and with no loss of purity to the variety, and this would yield far more favorable results than a mere change of seed." c. The Outright Production of Neiv Varieties. But you are waiting for a discussion of the second of the great features of crossing, — the sum- mary production of new varieties,, This is the sub- ject which is almost universally associated with crossing in the popular mind, and even among horticulturists themselves. It is the commonest notion that the desirable characters of given parents can be definitely combined in a pronounced cross or hybrid. There are two or three philo- sophical reasons which somewhat oppose this doc- trine, and which we will do well to consider at the outset. In the first place, nature is opposed to hybrids, for species have been bred away from each other in the ability to cross. If, therefore, there is no advantage for nature to hybridize, we may PRODUCTION OF NOVELTIES. 65 suppose that there would be little advantage for man to do so ; and there would be no advantage for man did he not place the plant under conditions different from nature, or desire a different set of characters. We have seen that nature's chief barriers to hybridization are total refusal of species to unite, and entire or comparative seedlessness of offspring. We can overcome the refusal to cross in many cases by bringing the plant under cultivation ; for the character of the species be- comes so changed by the wholly new conditions that its former antipathies may be overpowered. Yet it is doubtful if such a plant will ever acquire a complete willingness to cross. In like manner we can overcome in a measure the comparative seedlessness of hybrids, but it is very doubtful if Ave can ever make such hybrids completely fruit- ful. It would appear, therefore, upon theoretical grounds, that in plants in which seeds are the parts sought, no good can be expected, as a rule, from hybridization ; and this seems to be affirmed by facts. 1 It is evident that species which have been differentiated or bred away from each other in a given locality will have more opposed qualities or powers than similar species which have arisen quite independently in places remote from each 1 See definition of hybrids, crosses, and other terms in the Glossary. bb PHILOSOPHY OF CROSSING PLANTS. other. In the one case the species have likely struggled with each other until each one has at- tained to a degree of divergence which allows it to persist ; while in the other case there has been no struggle between the species, but similar con- ditions have brought about similar results. These similar species which appear independently of each other in different places are called representa- tive species. Islands remote from each other but similarly situated with reference to climate very often contain representative species ; and the same may be said of other regions much like each other, as eastern North America and Japan. Now, it follows that, if representative species are less opposed than others, they are more likely to hybridize with good results ; and this fact is re- markably well illustrated in the KiefTer and allied pears, which are hybrids between representative species of Europe and Japan ; and I am inclined to think that the same may be found to be true of the common or European apple and the wild crab of the Mississippi valley. Various crabs of the Soulard type, which I once thought to con- stitute a distinct species, appear upon further study to be hybrids. We will also recall that the hybrid grapes which have so far proved most valuable are those obtained by Rogers between the American Vitis Labrusca and the European wine grape ; and that the attempts of Haskell and others to hybrid- INSTABILITY OF HYBBIDS. 67 ize associated species of native grapes have given, at best, only indifferent results. To these good results from hybrids of fruit trees and vines I shall revert presently. Another theoretical point, which is borne out by practice, is the conclusion that, because of the great differences and lack of affinity between par- ents, pronounced hybrid offsprings are unstable. This is one of the greatest difficulties in the way of the summary production of new varieties by means of hybridization. It would appear, also, that, because of the unlikeness of parents, hybrid offspring must be exceedingly variable ; but, as a matter of fact, in many instances the parents are so pronouncedly different that the hybrids represent a distinct type by themselves, or else they approach very nearly to the characters of one of the parents. There are, to be sure, many instances of exceedingly variable hybrid offspring, but they are usually the offspring of variable par- ents. In other words, variability in offspring appears to follow rather as a result of variability in parents than as a result of mere unlikeness of characters. But the instability of hybrid off- spring when propagated by seed is notorious. Wallace writes that " the effect of occasional crosses often results in a great amount of varia- tion, but it also leads to instability of character, and is therefore very little employed in the 68 PHILOSOPHY OF CROSSING PLANTS. production of fixed and well-marked races." I may remark again that, because of the unequal and unknown powers of the parents, we can never predict what characters will appear in the hybrids. This fact was well expressed by Lindley a half century ago, in the phrase, " Hybridizing is a game of chance played between man and plants." V. Characteristics of Crosses. Bearing these fundamental propositions in mind, let us pursue the subject somewhat in detail. We shall find that the characters of hybrids, as compared with the characters of simple crosses between stocks of the same variety, are ambiguous, negative, and often prejudicial. The fullest dis- cussion of hybrids has been made by Focke (see Lecture IV.), and he lays down the five following propositions concerning the character of hybrid offspring : — 1. " All individuals which have come from the crossing of two pure species or races, when pro- duced and grown under like conditions, are usually exactly like each other, or at least scarcely more different from each other than plants of the same species are." This proposition, although perhaps true in the main, appears to be too broadly and positively stated. 2. " The characters of hybrids may be different HYBRIDS AND CROSS-BREEDS. 69 from the characters of the parents. The hybrids differ most in size and vigor and in their sexual powers." 3. " Hybrids are distinguished from their par- ents by their powers of vegetation or growth. Hybrids between very different species are often weak, especially when young, so that it is difficult to raise them. On the other hand, cross-breeds are, as a rule, uncommonly vigorous ; they are dis- tinguished mostly by size, rapidity of growth, early flowering, productiveness, longer life, stronger reproductive power, unusual size of some special organs, and similar characteristics." 4. " Hybrids produce a less amount of pollen and fewer seeds than their parents, and they often produce none. In cross-breeds this weaken- ing of the reproductive powers does* not occur. The flowers of sterile or nearly sterile hybrids usually remain fresh a long time." 5. " Malformations and odd forms are apt to appear in hybrids, especially in the flowers." Some of the relations between hybridization and crossing within narrow limits are stated as follows by Darwin : " It is an extraordinary fact that with many species flowers fertilized with their own pollen are either absolutely or in some degree sterile ; if fertilized with pollen from another flower on the same plant, they are some- times, though rarely, a little more fertile ; if 70 PHILOSOPHY OF CROSSING PLANTS. fertilized with pollen from another individual or variety of the same species, they are fully fertile ; but if with pollen from a distinct species, they are sterile in all possible degrees, until utter sterility is reached. We thus have a long series with absolute sterility at the two ends ; at one end due to the sexual elements not having been sufficiently differentiated, and at the other end to their having been differentiated in too great a degree, or in some peculiar manner." The difficulties in the way of successful results through hybridization are, therefore, these : the difficulty of effecting the cross ; infertility, in- stability, variability, and often weakness and monstrosity of the hybrids ; and the absolute im- possibility of predicting results. The advantage to be derived from a successful hybridization is the securing of a new variety which shall combine in some measure the most desirable features of both parents ; and this advantage is often of so great moment that it is worth while to make re- peated efforts and to overlook numerous failures. From these theoretical considerations it is apparent that hybridization is essentially an empirical sub- ject, and the results are such as fall under the common denomination of chance. And, as it does not rest upon any legitimate function in nature, we can understand that it will always be difficult to codify laws upon it. HYBRIDS AND BUD-PROPAGATION. 71 Among the various characters of hybrid off- spring, I presume that the most prejudicial one is their instability, their tendency still to vary into new forms or to return to one or the other parent in succeeding generations. It is difficult to fix any particular form which we may secure in the first generation of hybrids. At the outset, we notice that this discouraging feature is manifested chiefly through the fact of seed-reproduction, and we thereby come upon what is perhaps the most important practical consideration in hybridization, — the fact that the great majority of the best hybrids in cultivation are increased by bud-propa- gation, as cuttings, layers, suckers, buds, or grafts. In fact, I recall very few instances in this country of good undoubted hybrids which are propagated with practical certainty by means of seeds. You will recall that the genera in which hybrids are most common are those in which bud-propagation is the rule ; as begonia, pelargonium, orchids, gladiolus, rhododendron, roses, cannas, and the fruits. This simply means that it is difficult to fix hybrids so that they will come " true to seed," and makes apparent the fact that if we desire hybrids we must expect to propagate them by means of buds. This is a point which appears to have been over- looked by those who contend that hybridization must necessarily swamp all results of natural se- 72 PHILOSOPHY OF CROSSING PLANTS. lection ; for, as comparatively few plants propagate habitually by means of buds, whatever hybrids might have appeared would have been speedily lost, and all the more, also, because, by the terms of their reasoning, the hybrids would cross with other and dissimilar forms, and therefore lose their identity as intermediates. Or, starting with the assumption that hybrids are intermediates, and would therefore obliterate specific types, we must conclude that they should have some marked de- gree of stability if they are to swamp or obliterate the characters of species ; but, as all hybrids tend to break up when propagated by seeds, it must follow that bud-propagation would become more and more common, and this is associated in nature with decreased seed-production. Now, seed-pro- duction is the legitimate function of flowers ; and we must concede that, as seed-production de- creased, floriferousness must have decreased ; and that, therefore, pronounced intercrossing would have obliterated the very organs upon which it depends, or have destroyed itself ! But I may be met by the objection that there is no inherent reason why hybrids should not become stable through seed-production by in-breed- ing, and I might be cited to the opinion of Darwin and others that in-breeding tends to fix any va- riety, whether it originates by crossing or other means. And it is a fact that in-breeding tends to IN-BREEDING OF CROSSES. 73 fix varieties within certain limits, but those limits are often overpassed in the case of very pro- nounced crosses, whether cross-breeds or true hybrids. And if it is true, as all observation and experiments show, that sexual or reproductive powers of crosses are weakened as the cross be- comes more violent, we should expect less and less possibility of successful in-breeding ; for in-breed- ing without disastrous results is possible only with comparatively strong reproductive powers. As a matter of fact, it is found in practice that it is exceedingly difficult to fix pronounced hybrids by means of in-breeding. It sometimes happens, too, that the hybrid individual which Ave wish to per- petuate may be infertile with itself, as I have often found in the case of squashes. It is often advised that we cross the hybrid individual which we wish to fix with another like individual, or with one of its parents. These results are often successful, but oftener they are not. In the first place, it often happens that the hybrid individuals may be so diverse that no two of them are alike ; this has been my experience in many crosses. And, again, crossing with a parent may draw the hybrid back again to the parental form. So long ago as last century Kolreuter proved this fact upon nicotiana and dianthus. A hybrid between Nicotiana rus- tica and N. paniculata was crossed with N. pani- culata until it was indistinguishable from it ; and 74 PHILOSOPHY OF CROSSING PLANTS. it was then crossed with N. rustica until it became indistinguishable from that parent. Yet there is no other way of fixing a hybrid to be propagated by seeds than by in-breeding, and by constant at- tention to selection. Fortunately, it occasionally happens that a hybrid is stable, and therefore needs no fixing. In this connection I may cite some of my own experience in crossing egg-plants and squashes ; for, although the products were not true hybrids in the strict interpretation of the word, many of them were hybrids to all intents and purposes, because made between very unlike varieties, and they will serve to illustrate the difficulties of which I speak. Offspring of egg-plant crosses were grown in 1890, and upon some of the most promising plants some flowers were self-pollinated. But these self-pollinated seeds gave just as variable offspring in 1891, as those selected almost at random from the patch; and, what was worse, none of them reproduced the parent, or " came true to seed," and all further motive for in-breed- ing was gone. My labor, therefore, amounted to nothing more than my own edification. My experience in crossing pumpkins and squashes has now extended through many years ; and, although I have obtained about one thousand types not named or described, I have not yet succeeded in fixing one. The difficulty here is an aggravated EXPERIMENTS IN IN-BREEDING. 75 one, however. The species are so exceedingly variable that all the hybrid individuals may be unlike, so that there can be no crossing between identical stocks ; and, if in-breeding is attempted, it may be found that the flowers will not in-breed. And the refusal to in-breed is all the more strange because the sexes are separated in different flowers upon the same plant. In other words, in my expe- rience, it is very difficult to get good seeds from squashes which are fertilized by a flower upon the same vine. The squashes may grow normally to full maturity, but be entirely hollow, or contain only empty seeds. In some instances the seeds may appear to be good, but may refuse to grow under the best conditions. Finally, a small number of flowers may give good seeds. I have many times observed this refusal of squashes (Cucurbita Pepo) to in-breed. It was first brought to my attention through efforts to fix certain types into varieties. The figures of one season's tests will sufficiently indicate the character of the problem. In 1890, one hundred and eighty-five squash flowers were carefully pollinated with staminate flowers taken from the same vine which bore the pistillate flowers. Only twenty-two of these produced fruit, and of these only seven, or less than one-third, bore good seeds, and in some of these the seeds were few. Now, these twenty-two fruits represented as many different varieties, so that the inability to set 76 PHILOSOPHY OF CROSSING PLANTS. fruit with pollen from the same vine is not a peculiarity of a particular variety. The records of the seeds of the seven fruits in 1891 are as follows : — Fruit No. 1. — Four vines were obtained, with four different types, two of them being white, one yellow, and one black. Fruit No. 2. — Twenty-three vines. Fifteen types very unlike, twelve being white and three yellow. Fruit No. 3. — Two vines. One type of fruit Avhich was almost like one of the original parents. Fruit No. 4. — Thirty-two vines. Six types, differing chiefly in size and shape. Fruit No. 5. — Twent}^ vines. Nineteen types, of which ten were white, eight orange, one striped, and all very unlike. Fruit No. 6. — Thirteen vines. Eleven types, — eight yellow, two black, one white. Fruit No. 7. — One vine. These offspring were just as variable as those from flowers not in-bred, and no more likely, apparently, to reproduce the parent. These tests leave me without any method of fixing a pro- nounced cross of squashes, and lead me to think that the legitimate process of origination of new kinds here, as, indeed, if not in general, is a more gradual process of selection, coupled, perhaps, with minor crossing. ATTEMPT TO FIX A CROSS. 77 I will relate a definite attempt towards the fixa- tion of a squash which I had obtained from cross- ing. The history of it runs back to 1887, when a cross was effected between a summer yellow crook- neck and a white bush scallop squash. In 1889 there appeared a squash of great excellence, com- bining the merits of summer and winter squashes with very attractive form, size, and color, and a good habit of plant. I showed the fruit to one of the most expert seedsmen of the country, and he pronounced it one of the most promising types which he had ever seen ; and, as he informed me that he had fixed squashes by breeding in and in, I was all the more anxious to carry out my own convictions in the same direction. It is needless to say that I was very happy over what I regarded as a great triumph. Of course I must have a large number of plants of my new variety, that I might select the best, both for in-breeding and for cross- ing similar types. So I selected the very finest squash, having placed it where I could admire it for some days, and saved every seed of it. These seeds were planted upon the most conspicuous knoll in my garden in 1890. It was soon evident that something was wrong. I seemed to have everything except my squash. One plant, how- ever, bore fruits almost like the parent, and upon this I began my attempts towards in-breeding. But flower after flower failed, and I soon saw that 78 PHILOSOPHY OP CROSSING PLANTS. the plant was infertile with itself. Careful search revealed two or three other plants very like this one, and I then proceeded to make crosses upon it. I was equally confident that this method would succeed. When I harvested my squashes in the fall and took account of stock, I found that the seeds of my one squash had given just as many different types as there were plants, and I actually counted one hundred and ten kinds distinct enough to be named and recognized. Still confident, in 1891 I planted the seeds of my few crosses, and as the summer days grew long and the crickets chirped in the meadows, I watched the expanding squash blossoms and wondered what they would bring forth. But they brought only disappoint- ment. Not one seed produced a squash like the parent. My squash had taken an unscientific leave of absence, and I do not know its where- abouts. And when the frost came and killed every ambitious blossom, my hope went out and has not yet returned! Let us now recall how many undoubted hybrids there are, named and known, among our fruits and vegetables. In grapes there are the most. There are Rogers' hybrids, like the Agawam, Lindley, Wilder, Salem, and Barry ; and there is some reason for supposing that the Delaware, Catawba, and other varieties are of hybrid origin. And many hybrids have come to notice lately LIST OF COMMON HYBRIDS. 79 through the work of Munson and others. But it must be remembered that grapes are naturally ex- ceedingly variable, and the specific limits are not well known, and that hybridization among them lacks much of that deflniteness which ordinarily attaches to the subject. In pears there is the Kieffer class. In apples, peaches, plums, cher- ries, gooseberries, and currants, there are no im- portant commercial hybrids. In blackberries there is the blackberry-dewberry class, represented by the Wilson Early and others. Some of the rasp- berries, like the Philadelphia and Shaffer, are hy- brids between the red and black species. Hybrids have been produced between the raspberry and blackberry by two or three persons, but they pos- sess no promise of economic results. Among all the list of garden vegetables (plants which are propagated by seeds) I do not know of a single important hybrid ; and the same is true of wheat, — unless the Carman wheat-rye varieties become prominent, — oats, the grasses, and other farm crops. But among ornamental plants there are many ; and it is a significant fact that the most numerous, most marked, and most successful hy- brids occur in the plants most carefully cultivated and protected, those, in other words, which are farthest removed from all untoward circumstances and an independent position. This is nowhere so well illustrated as in the case of cultivated orchids, 80 PHILOSOPHY OF CROSSING PLANTS. in which hybridization has played no end of freaks, and in which, also, every individual plant is nursed and coddled. 1 With such plants the struggle for existence is reduced to its lowest terms ; for it must be borne in mind that, even in the garden, plants must fight severely for a chance to live, and even then only the very best can persist, or are even allowed to try. I am sure that this list of hybrids is much more meagre than most catalogues and trade-lists would have us believe, but I am sure that it is approxi- mately near the truth. It is, of course, equivalent to saying that most of the so-called hybrid fruits and vegetables are myths. There is everywhere a misconception of what a hybrid is, and how it comes to exist ; and yet, perhaps because of this indefinite knowledge, there is a wide-spread feel- ing that a hybrid is necessarily good, while the presumption is directly the opposite. The identity of a hybrid in the popular mind rests entirely upon some superficial character, and proceeds upon the assumption that it is necessarily intermediate be- tween the parents. Hence we find one of our popular authors asserting that, because the kohl- rabi bears its thickened portion midway of its stem, it is evidently a hybrid between the cabbage and turnip, which bear respectively the thickened parts 1 Consult E. Bohnhof, " Dictioimaire des Orchid^es Hy- brides," Paris, 1895. INFLUENCE OF PARENTS. 81 at the opposite extremities of the stem ! And then there are those who confound the word hybrid with high-bred, and who build attractive castles upon the unconscious error. And thus is confu- sion confounded ! But, before leaving this subject of hybridization, I must speak of the old yet common notion that there is some peculiar influence exerted by each sex in the parentage of hybrids ; for I shall thereby not only call your attention to what I believe to be an error, but shall also find the opportunity to still further illustrate the entanglements of hybri- dization. It was held by certain early observers, of whom the great Linnaeus was one, that the female parent determines the constitution of the hybrid, while the male parent gives the external attributes, as form, size, and color. The accumulated experi- ence of nearly a century and a half appears to con- tradict this proposition, and Focke, who has recently gone over the whole ground, positively declares that it is untrue. There are instances, to be sure, in which this old idea is affirmed, but there are others in which it is contradicted. The truth ap- pears to be this, — that the parent of greater strength or virility makes the stronger impression upon the hybrids, whether it is the staminate or pistillate parent; and it appears to be equally true that it is usually impossible to determine be- forehand which parent is the stronger. It is cer- 82 PHILOSOPHY OF CROSSING PLANTS. tain that strength does not lie in size, neither in the high development of any character. It appears to be more particularly associated with what we call fixity or stability of character, or the tendency towards invariability. This has been well illustrated in my own experi- ments with squashes, gourds, and pumpkins. The common little pear-shaped gourd will impress itself more strongly upon crosses than any of the edible squashes and pumpkins with which it will effect a cross, whether it is used as male or female parent. Even the imposing and ubiquitous great field pumpkin, which every New Englander associates with pies, is overpowered by the little gourd. Seeds from a large and sleek pumpkin which had been fertilized by gourd pollen, produced gourds and small hard-shelled globular fruits which were entirely inedible. A more interesting experiment was made between the handsome green-striped Ber- gen fall squash and the little pear gourd. Several flowers of the gourd were pollinated by the Bergen in 1889. The fruits raised from these seeds in 1890 were remarkably gourd-like. Some of these crosses were pollinated again in 1890 by the Ber- gen, and the seeds were sown in 1891. Here, then, were crosses into which the gourd had gone once and the Bergen twice, and both the parents are to all appearances equally fixed, the difference in strength, if any, attaching rather to the Bergen. POLLINATION IS UNCERTAIN. 83 Now, the crop of 1891 still carried pronounced characters of the gourd. Even in the fruits which most resembled the Bergen, the shells were almost flinty hard, and the flesh, even when thick and tender, was bitter. Some of the fruits looked so much like the Bergen that I was led to think that the gourd had largely disappeared. The very hard but thin paper-like shell which the gourd had laid over the thick yellow flesh of the Bergen, I thought might serve a useful purpose, and make the squash a better keeper. And I found that it was a great protection, for the squash could stand any amount of rough handling, and was even not injured by ten degrees of frost. All this was an acquisition, and, as the squash was handsome and exceedingly productive, nothing more seemed to be desired. But it still remained to have a squash for dinner. The cook complained of the hard shell, but, once inside, the flesh was thick and attractive, and it cooked nicely. But the flavor ! Dregs of quinine, gall, and boneset! The gourd was still there ! VI. Uncertainties of Pollination. We have now seen that uncertainty follows hybridization, and, in closing, I will say that uncertainty also attaches to the mere act of pollination. Between some species which are 84 PHILOSOPHY OF CBOSSING PLANTS. closely allied and which have large and strong flowers, four-fifths of the attempts towards cross- pollination may be successful ; but such a large proportion of successes is not common, and it may be infrequent even in pollinations between plants of the same species or variety. Some of the failure is due in many cases to unskilful opera- tion, but even the most expert operators fail as often as they succeed in promiscuous pollinating. There is good reason to believe, as Darwin has shown, that the failure may be due to some selec- tive power of individual plants, by which they re- fuse pollen which is, in many instances, acceptable to other plants even of the same variety or stock. The lesson to be drawn from these facts is that operations should be as many as possible, and that discouragement should not come from failure. In order to illustrate the varying fortunes of the pollinator, I will transcribe some notes from my field-book. Two hundred and thirty-four pollinations of gourds, pumpkins, and squashes, mostly between varieties of one species (Cucurbita Pepo), and in- cluding some individual pollinations, gave one hundred and seventeen failures and one hundred and seventeen successes. These crosses were made in varying weather, from July 28 to August 30. In some periods nearly all the operations would succeed, and at other times most of them would RECORDS OF POLLINATIONS. 85 fail. I have always regarded these experiments as among my most successful ones, and yet but half of the pollinations " took." But one must not understand that I actually secured seeds from even all these one hundred and seventeen fruits, for some of them turned out to be seedless, and some were destroyed by insects before they were ripe, or they were lost by accidental means. A few more than half of the successful pollinations — if by suc- cess we mean the formation and growth of fruit — really secured us seeds, or about one-fourth of the whole number of efforts. Twenty pollinations were made between potato flowers, and they all failed ; also, seven pollinations of red peppers, four of husk-tomato, two of Nico- tian a affinis upon petunia and two of the reciprocal cross, twelve of radish, one of Mirabilis Jalapa upon M. longiflora and two of the reciprocal cross, three Convolvulus major upon C. minor and one of the reciprocal, one muskmelon by squash, two musk- melons by watermelon, and one muskmelon by cucumber. This is but one record. Let me give another : — Cucumber, ninety-five efforts: fifty-two suc- cesses, forty-three failures. Tomato, forty-three efforts: nineteen successes, twenty-four failures. Egg-plant, seven efforts : one success, six failures. Pepper, fifteen efforts : one success, fourteen fail- ures. Husk-tomato, forty-five efforts: forty-five 86 PHILOSOPHY OF CROSSING PLANTS. failures. Pepino, twelve efforts: twelve failures. Petunia by Nicotiana affinis, eleven efforts : eleven failures. Nicotiana affinis by petunia, six efforts : six failures. General Grant tobacco by Nicotiana affinis, eleven efforts : eight successes, three fail- ures. Nicotiana affinis by General Grant tobacco, fifteen efforts : fifteen failures. General Grant tobacco by General Grant tobacco, one effort : one success. Nicotiana affinis by Nicotiana affinis, three efforts : two successes, one failure. Tuberous be- gonia, five efforts : five successes. Total, three hundred and twelve efforts : eighty- nine successes, two hundred and twenty-three failures. Conclusion. And now, the sum of it all is this : encourage in every way crosses within the limits of the variety and in connection with change of stock, expecting increase in vigor and productiveness ; hybridize if you wish to experiment, but do it carefully, systematically, thoroughly, and do not expect too much. Extend Darwin's famous proposition to read: Nature abhors both perpetual self-fertiliza- tion and hybridization. LECTURE III. HOW DOMESTIC VARIETIES ORIGINATE. " The key is man's power of accumulative selection : nature gives successive variations ; man adds them up in certain directions useful to him." This, in Darwin's phrase, is the essence of the cultivator's skill in ameliorating the vege- table kingdom. So far as man is concerned, the origin of the initial variation is largely chance, but this start or variation once given, he has the power, in most cases, to perpetuate it and to modify its characters. There are, then, two very different factors or problems in the origination of garden varieties, — the production of the first de- parture or variation, and the subsequent breeding of it. Persons who give little thought to the sub- ject, look upon variation as the end of their endeav- ors, thinking that a form comes into being with all its characters well marked and fixed. In reality, however, variation is but the beginning ; selection is the end. I. Indeterminate Varieties. There are two general classes of garden varie- ties as respects the method of their origin, — 87 88 HOW DOMESTIC VARIETIES ORIGINATE. those which come into existence somewhat sud- denly and which require little else of the hus- bandman than the multiplication of them, and those which are the result of a slow evolution or direct breeding. The former are indeterminate or uncertain, and the latter are determinate or definite. The greater part of those in the first class are plants which are multiplied or divided by bud-propagation. They comprise nearly all our fruits, the woody ornamental plants, and such herbaceous genera as begonia, canna, gladiolus, lily, dahlia, carnation, chrysanthemum, and the like, — in fact, all those multiplied by grafting, cuttings, bulbs, or other asexual parts. The original plant may be either a seedling or a bud- sport. The gardener, who is always on the look- out for novelties, discovers its good qualities and propagates it. Varieties which are habitually multiplied by buds, as in those plants which I have mentioned in the last paragraph, vary widely when grown from seeds, so that every seedling may be markedly distinct. As soon, however, as varie- ties are widely and exclusively propagated by seeds, they develop a capacity of carrying the greater part of the individual differences down to the offspring. That is, seedlings from bud- multiplied plants do not " come true," as a rule, whilst those from seed - propagated plants do INDETERMINATE VARIETIES. 89 " come true." The reason of this difference will become apparent upon a moment's reflection. In the seed-propagatecl plants, like the kitchen-gar- den vegetables and the annual flowers, we select the seeds and thereby eliminate all those varia- tions which would have arisen had the discarded seeds been sown. In other words, we are con- stantly fixing the tendency to "come true," for this feature of plants is as much a variation as form or color or any other attribute is. Suppose, for instance, that a certain variation were to re- ceive two opposite treatments, the seeds from one- half of the progeny being carefully selected year by year, and all those from untypical plants dis- carded, whilst in the other half all the seeds from all the plants, whether good or bad, are saved and sown. In the one case, it will be seen, we are fixing the tendency to " come true," for this is all that constitutes a horticultural variety, — a brood which is very much like all its parents. In the other case, we are constantly eliminating the tendency to " come true " by allowing every modifying agency full sway. So the very act of taking seeds only from plants which have " come true," tends to still more strongly fix the heredi- tary force within narrow limits. Working against this restrictive force, however, are all the agencies of environment, so that, fortunately, now and then a seed gives a " rogue," or a plant widely 90 HOW DOMESTIC VARIETIES ORIGINATE. unlike its parent, and this may be the start for a new variety. With bud-multiplied varieties, however, the case is very different. Here every seed may be sown, as in the illustrative case above, because the seedlings are not wanted for themselves, but simply as stocks upon which to bud or graft the desired varieties. So there is no seed selection in the ordinary propagation of apples, pears, peaches, and the other orchard fruits. The seeds are taken indiscriminately from pomace or the refuse of can- ning and evaporating factories. But every annual garden vegetable is always grown from seeds more or less carefully saved from plants which possess some desired attribute. There is no reason why the tree fruits should not reproduce themselves from seeds just as closely as the annual herbs do, if they were to be as carefully propagated by selected seeds through a long course of generations. There is excellent proof of this in the well-marked races or families of Russian apples. In that country, grafting has been little employed, and conse- quently it has been necessary to select seeds only from acceptable trees in order that the off- spring might be more acceptable. So the Russian apples have come to run in groups or families, each family bearing the mark of some strong ancestor. Most of the seedlings of the Duchess of Olden- burg are recognizable because of their likeness to PLANT-BREEDING. 91 the parent. We may thus trace an incipient ten- dency in our own fruits towards racial characters. The Fameuse type of apples, for example, tends to perpetuate itself ; and a similar tendency is very well marked in the Damson and Green Gage plums, the Orange quince, Concord grape, and Hill's Chili and Crawford peaches. But inas- much as bud-multiplication is so essential in nursery practice, we can hardly hope for the time when our trees and shrubs, or even our per- ennial herbs, will u come true" with much cer- tainty. In them, therefore, we get new varieties by simply sowing the seeds ; but in seed-propa- gated varieties we must depend either upon chance variations or else we must resort to definite plant- breeding. II. Plant-breeding. The breeding of domestic animals is attended, for the most part, with such definite and often precise results that there has come to be a gen- eral desire to extend the same principles to plants. It is not unusual to hear well-informed people say that it is possible to breed plants with as much certainty and exactness as it is to breed animals. The fact is, however, that such exact- ness will never be possible, because plants are very unlike animals in organization, and because, 92 HOW DOMESTIC VARIETIES ORIGINATE. also, the objects sought in the two cases are fun- damentally unlike. Plants, as we have seen, are made up of a colony of potential individuals, and to breed between two plants by crossing means that we must choose the sex-parents from amongst as many individuals as there are flowers or branches on the two plants, whilst in animals we choose two definite personal parents. And these personal parents are either male or female, and the union is essential to the production of offspring, whilst in plants each parent — that is, each flower — is generally both male and female, and the union of two is not essential to the production of offspring, for the plant is capable of multiplying itself by buds. The element of chance, therefore, is one hundred, or more, to one in crossing plants as compared with crossing animals. Then, again, the plant-parents are modified profoundly by every environmental condition of soil and temperature and sunshine, or other external condition, since they possess no bodily temperature, no choice of conditions, and no volition to enable them to overcome the circumstances in which they are placed. Animals, on the contrary, have all these elements of personality, and the breeder is also able to control the conditions of their lives to a nicety. In view of all these facts, it is not strange that animals can be bred by crossing with more confidence than plants can. But there is another ANIMALS AND PLANTS. V6 and even more important difference between the breeding of animals and the breeding of plants. In animals, our sole object is to secure simply one animal or one brood of offspring. In plants, our object is, in general, to secure a race or genera- tions of offspring, which may be disseminated freely over the earth. In the bovine race, for example, our object in breeding is to produce one cow with given characters ; in turnips, our object is to produce a new variety, the seed of which will reproduce the variety, whether sown in Penn- sylvania or Ceylon. It is apparent, therefore, that any comparisons drawn between the breeding of animals and plants are fundamentally fallacious. Is there, then, any such thing as plant-breeding, any possibility that the operator can proceed Avith some confidence that he may obtain the ideal which he has in mind? Yes, to a certain extent. It is apparent that the very first effort on the part of the plant-breeder must be to secure indi- vidual differences ; for so long as the plants which he handles are very closely alike, so long there will be little hope of obtaining new varieties. He must, therefore, cause his plants to vary. In plants which are comparatively unvariable, it is frequently impossible to produce variations in the desired direction at once, but it is more important to " break " the type, — that is, to make it depart markedly from its normal behavior in any or 94 HOW DOMESTIC VARIETIES ORIGINATE. many directions (page 19). If the type once begins to vary, to break up into different forms, the operator may be sure that it will soon become plastic enough to allow of modification in the manner which he desires. But whilst it is im- portant or even necessary to break a well-marked type into many forms, it would no doubt be un- wise to encourage this tendency after it once appears, lest the plant acquire a too strong habit of scattering. This initial variation is induced by changing the conditions in which the plant has habitually grown, as a change of seed, change of soil, tillage, varying the food supply, crossing- and the like. As a matter of fact, however, nearly all plants which have been long cultivated are already suf- ficiently variable to afford a starting-point for breeding. The operator should have a vivid mental picture of the variety which he designs to obtain ; then he should select that plant in his plantation which is the nearest his ideal, and sow the seeds of it. From the seedlings he should again select the individuals which most nearly approach his type, and so on, generation after generation, until the desired object is attained. It is important, if he is to make rapid progress, that he keep the same ideal in mind year by year, otherwise there will be vacillation and the prog- ress of one year may be undone by a counter ANTAGONISTIC FEATURES. 95 movement the following year. In this way, it will be found that almost any character of a plant may be either intensified or lessened. This is man's nearest approach to the Creator in his dominion over the physical forms of life, and it is great and potent in proportion as it sets for itself correct ideals in the beginning and adheres to them until the end. When beginning this selection or breeding for an ideal, it is important that impossible or contra- dictory results should be avoided. Some of the cautions and suggestions which need to be con- sidered are these : — 1. Avoid striving after features which are antag- onistic or foreign to the species or genus with which you are working. Every group of plants has be- come endowed with certain characters or lines of development, and the cultivator will secure quicker and surer results if he works along the same lines, rather than to attempt to thwart them. Nature gives the hint : let men follow it out, rather than to endeavor to create new types of characters. Let us take some of the solanaceous plants as examples. There are certain types of the genus solanum which have a natural habit of tuber-bearing, as the potato. Such species should be bred for tubers and not for fruits. There are other solanums, however, like the egg-plants and the pepinoes, which naturally vary or develop in 96 HOW DOMESTIC VARIETIES ORIGINATE. the direction of fruit-bearing, and these should be bred for fruits and not for tubers ; and the same should be true in the related genera of tomatoes, red peppers, and physalis. Those ambitious per- sons who are always looking for a tuber-bearing tomato, therefore, might better concentrate their energies on the potato, for the tomato is not devel- oping in that direction ; and even if the tomato could be made to produce tubers, it would thereby lessen its fruit production, for plants cannot main- tain two diverse and profitable crops at the same time. It is more reasonable, and certainly more practicable, to grow potatoes on potato plants and tomatoes on tomato plants. 2. The quickest and most marked results are to be expected in those groups or species ivhich are nor- mally the most variable. There are a greater num- ber of variations or starting-points in such species ; but it also follows that the forms are less stable the more the species is variable. Yet the varia- tions, being very plastic, yield themselves readily to the wishes of the operator. Carriere puts the thought in this form : " The stability of forms, in any group of plants, is, in general, in inverse ratio to the number of the species which it contains, and also to the degree of its domestication." The most variable types are the most dominant ones over the earth ; that is, they occur in greater numbers and Tinder more diverse conditions than VARIABLE TYPES. 97 the comparatively invariable t}^pes do. The corn- posits, or sunflower-like plants, comprise a ninth or tenth of the total species of flowering plants, and the larger part of the subordinate types or genera contain many forms or species. Aster, goldenrod, the hawkweeds, thistles, and other groups, are representative of the cosmopolitan or variable types of composites. Whenever, for any reason, any type begins to decline in variability, it also begins to perish ; it is then tending towards extinction. Monotypic genera — those which con- tain but a single species — are generally of local or disconnected distribution, and are, for the most part, vanishing remnants of a once dominant or important t} r pe. As a rule, most of our widely variable and staple cultivated species are mem- bers of large, or at least polytypic genera. Such, for example, are the apples and pears, peaches and plums, oranges and lemons, roses, bananas, chrysanthemums, pinks, cucurbits, beans, potato, grapes, barley, rice, cotton. A marked exception to this statement is maize, which is immensely variable and is generally held to have come from a single species ; but the genesis of maize is un- known, and it is possible, though scarcely proba- ble, that more than one species is concerned in it. Wheat is also a partial exception, although the original specific type is not understood ; and the latest monographers admit three or four other spe- 98 HOW DOMESTIC VARIETIES ORIGINATE. cies to the genus, aside from wheat. There are other exceptions, but they are mostly unimportant, and, in the main, it may be said that the domi- nant domestic types of plants represent markedly polytypic genera. 3. Breed for one tiling at a time. The person who strives at the same time for increase or modi- fication in prolificacy and flavor will be likely to fail in both. He should work for one object alone, simply giving sufficient attention to sub- sidiary objects to keep them up to normal stand- ard. This is really equivalent to saying that there can be no such thing as the perfect all- around variety which so many people covet. Va- rieties must be adapted to specific uses, — one for shipping, one for canning, one for dessert, one for keeping qualities, and the like. The more good varieties there are of any species, the more Avidely and successfully that species can be cultivated. 4. Bo not desire contradictory attributes in any variety. A variety, for example, which bears the maximum number of fruits or flowers cannot be expected to greatly increase the size of those organs without loss in numbers. This is well shown in the tomato. The original tomato pro- duced from six to ten fruits in a cluster, but as the fruits increased in size the numbers in each cluster fell to two or three. That is, increase in size proceeded somewhat at the expense of n inner- SELECT FOR THE ENTIRE PLANT. \)\) ical productivity ; yet the total weight of fruit per plant has greatly increased. The same is true of apples and pears ; for whilst these trees bear flowers in clusters, they generally bear their fruits singly. Originally, every flower normally set fruit. The reason why blackberries, currants, and grapes do not increase more markedly in size, is probably because the size of cluster has been given greater attention than the size of berry. Plants which now bear a full crop of tubers cannot be expected to increase greatly in fruit- bearing, as I have already explained under Rule 1. This fact is illustrated in the potato, in which, as tuber production has increased, seed production has decreased, so that potato growers now complain that potatoes do not produce bolls as freely as they did years ago. 5. When selecting seeds, remember that the char- acter of the whole plant is more imjjortant than the character of any one branch or part of the plant ; and the more uniform the 'plant in all its parts, the greater is the likelihood that it will transmit its characters. If one is striving for larger flowers, for instance, he will secure better results if he choose seeds from plants which bear large flowers throughout, than he will if he choose them from some one large-flowering branch on a plant which bears indifferent flowers on the remaining branches, even though this given branch produce much larger 100 HOW DOMESTIC VARIETIES ORIGINATE. flowers than those borne on the large -flowered plant. Small potatoes from productive hills give a better product than large potatoes from unpro- ductive hills. The practice of selecting large ears from a bin of corn, or large melons from the grocer's wagon, is much less efficient in produc- ing large products the following season than the practice of going into the fields and selecting the most uniformly large-fruited parents would be. A very poor plant may occasionally produce one or two very superior fruits, but the seeds are more likely to perpetuate the characters of the plant than of the fruits. The following experiences detailed by Henri L. de Vilmorin illustrate my proposition admirably: " I tried an experiment with seeds of Chrysanthe- mum carinatum gathered on double, single, and semi-double heads, all growing on one plant, and found no difference whatever in the propor- tion of single and double-flowered plants. In striped verbenas, an unequal distribution of the color is often noticed ; some heads are pure white, some of a self color, and most are marked with colored stripes on white ground. I had seeds taken severally from all and tested alongside one another. The result was the same. All the seeds from one plant, whatever the color of the flower that bore them, gave the same proportion of plain and variegated flowers." UNIFORMNESS IN THE PARTS. 101 The second part of my proposition is equally as important as the first, — the fact that a plant which is uniform in all its branches or parts is more likely to transmit its general features than one which varies within itself. It is well known that bean plants often produce beans with various styles of markings on the same plant or even in the same pod, yet these variations rarely ever perpetuate themselves. The same remark may be applied to variations in peas. These illustra- tions only add emphasis to the fact that intending plant-breeders should give greater heed than they usually do to the entire plant, rather than confine their attention to the particular part or organ which they desire to improve. At first thought, it may look as if these facts are directly opposed to the proposition which I emphasized in my first lecture, that every branch of a plant is a potential autonomy, but it is really a confirmation of it. The variation itself shows that the branch is measurably independent, but it is not until the conditions or causes of the vari- ation are powerful enough to affect the entire plant that they are sufficiently impressed upon the organization of the plant to make their effects hereditary. There is an apparent exception to the law that the character of the entire plant is more impor- tant than any one organ or part of it, in the case 102 HOW DOMESTIC VARIETIES ORIGINATE. of the seeds themselves. That is, better results usually follow the sowing of large and heavy seeds than of small or unselected seeds from the same plant. This, however, does not affect the main proposition, for the seed is in a measure independent of the plant-body, and is not so directly influenced by environment as the other organs are. And, again, the seed receives a part of its elements from a second or male parent. The good results which follow the use of large seeds are, chiefly, greater uniformity of crop, increased vigor, often a gain in earliness and sometimes in bulk, and generally a greater ca- pacity for the production of seeds. These results are probably associated less with any innate he- reditable tendencies than with the mere vegeta- tive strength and uniformness of the large seeds. The large seeds usually germinate more quickly than the small ones, provided both are equally mature, and they push the plantlet on more vigorously. This initial gain, coming at the most critical time in the life of the new indi- vidual, is no doubt responsible for very much of the result which follows. The uniformity of crop is the most important advantage which comes of the use of large seeds, and this is obviously the result of the elimination of all seeds of varying degrees of maturity, of incomplete growth and formation, and of low vitality. PROGENY OF IMMATURE SEEDS 103 Another important consideration touching the selection of seeds is the fact that very immature seeds give a feeble but precocious progeny. This has long been observed by gardeners, but Sturte- vant, Arthur, and Goff have recently made a critical examination of the subject. " It is not the slightly unripe seeds that give a noticeable increase in earliness," according to Arthur, " but very unripe seeds, gathered from fruit [tomatoes] scarcely of full size and still very green. Such seeds do not weigh more than two-thirds as much as those fully ripe. They germinate readily, but the plantlets lack constitutional vigor and are more easily affected by retarding or harmful influences. If they can be brought through the early period of growth and become well estab- lished, and the foliage or fruit is not attacked by rots or blights, the grower will usually be re- warded by an earlier and more abundant crop of slightly smaller and less firm fruit. These char- acters will be more strongly emphasized in sub- sequent years by continuous seed propagation." Goff remarks that the increase in earliness in tomatoes, following the use of markedly immature seeds, "is accompanied by a marked decrease in the vigor of the plant, and in the size, firmness, and keeping quality of the fruit." These results are probably closely associated with the chemical constitution and content of the immature seeds. 104 HOAV DOMESTIC VARIETIES ORIGINATE. The organic compounds have probably not yet reached a state of stability, and they therefore respond quickly to external stimuli when placed in conditions suitable to germination ; and there is little food for the nourishment of the plantlet. The consequent weakness of the plantlet results in a loss of vegetative vigor, which is earliness (see Rule 11). Still another feature connected with the choice of seeds is the fact that in some plants, as in some Ipomoeas, for example, the color of the seed is more or less intimately associated with the color of the flower which produced them and also with the color of the flowers which they will produce. 6. Plants which have any desired characteristics in common may differ widely in their ability to transmit these characters. It is generally impos- sible for the cultivator to determine, from the appearance of any given number of similar plants, which of them Avill give progeny the most unvari- able and the most like its parent ; but it may be said that those individuals which grow in the most usual or normal environments are most likely to perpetuate themselves. A very unusual condi- tion, as of soil, moisture, or exposure, is not easily imitated when providing for the succeeding gen- eration, and a return to normal conditions of envi- ronment may be expected to be followed by a more or less complete return to normal attributes on the SELECT SEVERAL STARTING-POINTS. 105 part of the plant. If the same variation, there- fore, were to occur in plants growing under widely different conditions, the operator who wishes to preserve the new form should take particular care to select his seeds from those individuals which seem to have been least influenced by the imme- diate conditions in which they have grown. Again, if the same variation appears both in uncrossed and crossed plants, the best results should be expected in selecting seeds from the former. We have already seen, in the second lecture, how it is that crosses are unstable, and how the instability is apt to be the greater the more violent the cross. " Cross-breeding greatly increases the chance of wide variation," writes Henri L. de Vilmorin, " but it makes the task of fixation more difficult." It is very important, therefore, when selecting seeds from plants which seem to give promise of a new variety, to sow the seeds of each plant separately, and then make the subsequent selec- tions from the most stable generation; and it is equally important that the operator should not trust to a single plant as a starting-point, when- ever he has several promising plants from which to choose. 7. The less marked the departure from the genius of the normal type, the greater, in general, is the likelihood that it will be perpetuated. That 106 HOW DOMESTIC VARIETIES ORIGINATE. is, widely aberrant forms are generally unstable. This is admirably illustrated in crosses. The seed-progeny of crosses between closely related varieties, or between different plants of the same variety, is more uniform and generally more easy of improvement by selection than the progeny of hybrids. In uncrossed plants, the general ten- dency is to resemble their parents, and the greater the number of like ancestors, the greater is the ten- dency to "come true." There is thought to be a tendency, though necessarily a weak one, to return to some particular ancestor, or to "date back." This is known as atavism. The so-called atavistic forms are likely to be unstable, to break up into nu- merous forms, or to return more or less completely to the type of the main line of the ancestry. The following statements touching some of the rela- tions of atavism to the amelioration of plants, are the results of an excellent study of heredity in lupines by Louis Leveque de Vilmorin : — "1. The tendency to resemble its parents is generally the strongest tendency in any plant; "2. But it is notably impaired as it comes into conflict with the tendency to resemble the general line of its ancestry. " 3. This latter tendency, or atavism, is con- stant, though not strong, and scarcely becomes impaired by the intervention of a series of gen- erations in which no reversion has taken place. CROSSING NOT AN END. 107 "4. The tendency to resemble a near pro- genitor (only two or three generations removed), on the other hand, is very soon obliterated if the given progenitor is different from the bnlk of its ancestors." *K 8. The crossing of plants should be looked upon as a means or starting-point, not as an end. We cross two flowers and sow the seeds. The result- ing seedlings may be unlike either parent. > Here, then, is variation. The operator should select that plant which most nearly satisfies his ideal, and then, by selection from its progeny and the progeny of succeeding generations, gradually ob- tain the plant which he desires. It is only in plants which are propagated by asexual parts — as grafts, cuttings, layers, bulbs, and the like — that hybrids or crosses are commonly immediately val- uable; for in these plants we really cut up and multiply the one individual plant which pleases us in the first batch of seedlings, rather than to take the offspring or seedlings of it. Thus, if any particular plant in a lot of seedlings of crosses of cannas, or plums, or hops, or strawberries, or potatoes, is valuable, we multiply that one in- dividual. There is no occasion for fixing the variety. But any satisfactory plant in a lot of seed- lings of crosses of pumpkins, or wheat, or beans, must be made the parent of a new variety by sow- ing the seeds of it and then by selecting for seed- 108 HOW DOMESTIC VARIETIES ORIGINATE. parents, year by year, those plants which are best. " The unsettled forms arising from crosses," Focke writes, " are the plastic material ont of which gardeners form their varieties." But even in the fruits, and other bud-propa- gated plants, crossing may often be used to as good advantage for the purpose of originating variation as it can in peas or buckwheat. It only requires a longer time to fix and select variations because the plants mature so slowly. Ordinarily, if the operator does not find satisfactory plants amongst the seedlings of any cross of fruit trees, he roots up the whole batch as profitless. But if he were to allow the best plants to stand and were to sow seeds from them, the second gen- eration might produce something more to his liking. But it is generally quicker to make another cross and to try the experiment over again, than to wait for unpromising seedlings to bear. This repeated repetition of the experiment, however, — continual crossing and sowing and uprooting, — is gambling. Throwing dice to see what will turn up is a comparable proceeding. The sowing of uncrossed seed is little better. Peter M. Gideon sowed over a bushel of apple seed, and one seed produced the Wealthy apple. 1 1 The facts in the origination of the Wealthy apple, as re- lated to me by Mr. Gideon, are these : he first planted a bushel of apple seeds, and then each year, for nine years, he planted GUIDES TO CROSSING. 109 D. B. Wier raised a million seedlings of soft maple, and one plant of the lot had finely divided leaves, and is now Wier's Cutleaved maple. Teas' Weeping mulberry, which is now so deservedly popular, was, as Mr. Teas tells me, " merely an accidental seedling." So this explains why the production of new varieties of fruits is always chance, whilst a skilled man can sit in his study in the winter time and picture to himself a new bean or muskmelon, and then go out in the next three or four summers and produce it. 9. If it is desired to employ crossing as a direct means of producing new varieties, each parent to the proposed cross should be selected in agreement with the rules already specified, and also because it pos- sesses in an emphatic degree one or more of the qualities which it is desired to combine; and the more uniformly and persistently the parent pre- sents a given character, the greater is the chance that it will transmit that character. It has already been said that crossing for the instant production of new varieties is most certain to give valuable enough seed to give a thousand trees. At the end of ten years, all the seedlings had perished (this was in Minnesota) except one hardy seedling crab. Then a small lot of seeds of apples and crab apples was obtained in Maine, and from these the Wealthy came. There were only about fifty seeds in the batch of crab seed which gave the Wealthy ; but before this variety was obtained, much over a bushel of seed had been sown. 110 HOW DOMESTIC VARIETIES ORIGINATE. results in those species which are propagated by buds, because the initial individual differences are not dissipated by seed-reproduction. This is especially true of hybridization, or crossing between distinct species ; for in such violent cross- ing as this the offspring is particularly likely to be unstable when propagated by seeds. The re- sults of hybridization appear to be most certain in those plants which are grown under glass, and in which, therefore, the selection of the seed- parents is most carefully made, and where the conditions of existence are most uniform. The most remarkable results in hybridization which have yet been attained are with the choicer glass- house plants, such as orchids, begonias, anthu- riums, and the like. (Lecture II.) The more violent the cross, the less is the likeli- hood that desirable offspring will follow. Species which refuse to give satisfactory results when hybridized directly or between the pure stocks, may give good varieties when the " blood" has become somewhat attenuated through previous crossings. The best results in hybridizing our native grape with the European grape, for ex- ample, have come from the use of one parent which is already a hybrid. Two notable examples are the Brighton and Diamond grapes, raised by Jacob Moore. The Brighton is a cross of Con- cord (pure native) by Diana-Hamburg (hybrid of IMPORTANT HYBRIDS OF FRUITS. Ill impure native and European). Diamond is a cross of Concord by Iona, the latter parent un- doubtedly of impure origin, containing a trace of the European vine. T. V. Munson's Brilliant is a secondary hybrid, its parents, Lindley and Dela- ware, both containing hybrid blood. Others of his varieties have similar histories. Even when the cross is much attenuated — or three or four or even more times removed from a pure hybrid origin by means of subsequent crossings — it may still produce marked effects in a cross without introducing such contradictory characters as to jeopardize the value of the offspring. Amongst American fruit plants there are com- paratively few valuable hybrids. The most con- spicuous . ones are in the grapes, particularly the various Rogers varieties, such as Agawam, Lind- ley, Wilder, Barry, and others, which are hybrids of the European grape and a native species. Other hybrids are the Kieffer and allied pears (between the common pear and the Oriental pear), the Transcendent and a few other crabs (between the common apple and the Siberian crab), the Soulard and kindred crabs (between the common apple and the native Western crab), a few blackberries of the Wilson Early type (between the blackberry and the dewberry), the purple-cane raspberries (between the native red and black raspberries, and possibly sometimes 112 HOW DOMESTIC VARIETIES ORIGINATE. combined with the European raspberry), the Utah Hybrid cherry (between the Western sand cherry and the sand plum), and probably a few of the native plums. There is undoubtedly a fertile field for further work in hybridizing our fruits, particularly those of native origin, and also many of the ornamental plants ; the danger is that persons are apt to expect too much from hybridization, and too little from the betterment of all the other conditions which so profoundly modify plants. Violent hybridizations generally give unsatisfactory and unreliable results ; but subsequent crossings, when the "blood" of the original species to the contract is considerably attenuated, may be expected to correct or over- come the first incompatibility, as explained above. 10. Establish the ideal of the desired variety firmly in the mind before any attempt is made at plant-breeding. If one is to make any progress in securing new varieties, he must first be an expert judge of the capabilities and merits of the plants with which he is dealing, otherwise he may attempt the impossible or he may obtain a variety which has no merit. It is important, too, that the person bear in mind the fact that a variety which is simply as good as any other in cultivation is not worth introducing. It should be better in some particular than any other in existence. The operator must know the PRODUCE AN INITIAL VARIATION. 113 points of his plant, as an expert stock-breeder knows the points of an animal, and he must possess the rare judgment to determine which characters are most likely to reappear in the offspring. Inasmuch as a person can be an ex- pert in only a few plants, it follows that he can- not expect satisfactory results in breeding any species which may chance to come before him. Persistent and uniform effort, continued over a series of years, is generally demanded for the production of really valuable varieties. Thus it often happens that one man excels all competitors in breeding a particular class of plants. The hor- ticulturist will recall, for instance, Lemoine in the breeding of gladiolus, Eckford in peas, Crozy in c annas, Bruant in pelargoniums, and others. There are now and then varieties which arise from no effort, but because of that very fact they reflect no credit upon the so-called originator, who is really only the lucky finder. So far as the originator is concerned, such varieties are merely chance. If, however, the operator — him- self an expert judge of the plant with which he deals — chooses his seeds with care and discrimi- nation, and then proposes, if need be, to follow up his work generation by generation by means of selection, the work becomes plant-breeding of the highest type. First of all, therefore, the operator must know 114 HOW DOMESTIC VARIETIES ORIGINATE. what he can likely get, and what will likely be worth getting. Most persons, however, begin at the other end of the problem, — they get what they can, and then let the public judge if the effort has been worth the while. 11. Having obtained a specific and correct ideal, the operator must next seek to make his plant vary in the desired direction. This may be done by crossing, or by modifying the conditions under which the plant grows, as indicated in Lectures I. and II. If there are any two plants which possess indications of the desired attributes, cross them: amongst the seedlings there may be some which may serve as starting-points for further effort. A change in the circumstances or environment of the plant may start the desired attribute. If the plant must be dwarfer, plant it on poorer or drier soil, transfer it towards the poles, plant it late in the season, or transplant it repeatedly (see pages 25 and 143). Dwarf peas become climb- ing peas on rich, moist soils. If the plant must have large fruits, allow it more food and room, and give attention to pruning and thinning. Cer- tain geographical regions develop certain charac- ters in plants, as we have seen (page 24); if, therefore, the desired feature does not appear spontaneously or as a result of any other treat- ment, transfer the plant for a time to that region PRODUCE AN INITIAL VARIATION. 115 which is characterized by such attributes, if there is any such. The importance of growing the plant under conditions or environments in which the desired type of characters is most frequently found, is admirably emphasized in the evolution of varieties which are adapted to forcing under glass. Within a century, — and in many instances within a decade or two, — species which were practically unknown to glass-houses have produced varieties which are perfectly adapted to them. This has been accomplished by growing the most tractable existing varieties under glass, and then carefully and persistently selecting those which most com- pletely adapt themselves to their environment and to the ideals of the operator. One of the most remarkable examples of this kind is afforded by the carnation. In Europe it is chiefly a border or out-door plant, but within a generation it has> produced hosts of excellent forcing varieties in America, where it is grown almost exclusively as a glass-house flower. So the carnation types of Europe and America are widely unlike, and the unlikeness becomes more emphatic year by year because of the rapid aberrant evolution of the American forms. Sowing the seeds of hardy annual plants in the fall often generates a tendency to produce thick- ened roots. The plant, finding itself unable to 116 HOW DOMESTIC VARIETIES ORIGINATE. perfect seeds, stores its reserve in the root, and it therefore tends to become biennial. In this manner, with the aid of selection and the varia- tion of the soil, Carriere was able to produce good radishes from the wild slender-rooted charlock (Raphanns Raphanistrum) . Lessened vigor, so long as the plant continues to be healthy, nearly always results in a compara- tive increase of fruits or reproductive organs. It is an old horticultural maxim that checking growth induces fruitfulness. It is largely in con- sequence of this fact that plants bear heaviest when they attain approximate maturity. Trees are often thrown into bearing by girdling, heavy pruning, the attacks of borers, and various acci- dental injuries. The gardener knoAVS that if he keeps his plants in vigorous growth by con- stantly potting them on into larger pots, he will get little, or at least very late, bloom. The plant-breeder, therefore, may be able to induce the desired initial variation by attention to this principle. (See discussion of variation in rela- tion to food supply, page 16.) Arthur has re- cently put the principle into this formula : " A decrease in nutrition during the period of growth of an organism, favors the development of the re- productive parts at the expense of the vegetative parts." A most important means of inducing variation SIMULTANEITY OF VARIATION. 117 is the simple change of seed, the philosophical reasons for which are explained on pages 59 and 28. A plant becomes closely fitted or accus- tomed to one set of conditions, and when it is placed in new conditions, it at once makes an effort to adapt itself to them. This adaptation is variation. No doubt the free interchange of seeds between seed-merchants and customers is one of the most fertile causes of the enormous increase in varieties in recent times. When once a novel variety appears, others of a similar kind are likely soon to follow in other places, and some persons have supposed that there is a synchronistic variation in plants, or a tendency for like variations to appear simultaneously in widely separated localities. There is perhaps some remote reason for this belief, because there is, as Darwin expresses it, an accumulative effect of domestication or cultivation, by virtue of which plants which long remain comparatively invariable may within a short time, when cultivation has been continued long enough, vary widely and in many directions; and it is to be expected that even when plants have long since responded to the wishes of the cultivator, an equal amount or accumulation of the force of domestication would tend to produce like effects in different places. But it is probable that by far the greater part of this synchronistic variation is simply an apparent 118 HOW DOMESTIC VARIETIES ORIGINATE. one, for whenever any marked novelty appears the attention of all interested persons is directed to looking for similar variations amongst their own plants. 12. The person who is wishing for new varieties should look critically to all perennial plants, and particularly to trees and shrubs, for bud-varieties or sports. It has already been said (pages 28, 6) that the branches of a tree may vary amongst themselves in the same way in which seedlings vary, and for the same reasons. As a rule, any marked sport is capable of being perpetuated by bud-propagation. The number of bud- varieties now in cultivation is really very large. Many of the cut-leaved and colored or variegated varieties of ornamental plants were originally found upon other trees as sports. The " mixing in the hill " of potatoes is bud-variation. Nectarines are derived from the peach, some of them as sports and some as seedlings. The moss-rose was prob- ably originally a sport from the Provence rose. Greening apple trees often bear russet apples, and russet trees sometimes bear greenings. So far as I know, there are no varieties of annual plants which have originated as sports. The probable reason for this is the fact that the duration of the plant is short and that its constitution is not pro- foundly modified in a single generation by the new circumstances in which it is placed every BUD-VAEIETIES. 119 year. The effects of the conditions in which it lives are recorded in the seeds, and the plant dies without allowing a second season of growth to express the impressions which were received in a former generation. The fact that every branch of an annual plant — as of perennials — is unlike every other branch, is evidence enough that the annual is not unlike the perennial in fundamental constitution ; and there is every reason to believe that if any given annual were to become a peren- nial, it would now and then develop differences sufficiently pronounced to make them worthy the name of sports, the same as hyacinths, bouvardias, trees, and all other perennial plants are apt to do. Bud- varieties may not only come true from buds — as grafts, cuttings and layers, — but they occa- sionally perpetuate themselves by seeds. Now, these seedlings are amenable to selection, just the same as any other seedlings are ; the bud-variety, therefore, may give the initial starting-point for plant-breeding. But, more than this, it is some- times possible to improve and fix the type by bud-selection as well as by seed-selection. Dar- win cites this interesting testimony : " Mr. Salter brings the principle of selection to bear on varie- gated plants propagated by buds, and has thus greatly improved and fixed several varieties. He informs me that at first a branch often produces variegated leaves on one side alone, and that the 120 HOW DOMESTIC VARIETIES ORIGINATE. leaves are marked only with an irregular edging, or with a few lines of white and yellow. To im- prove and fix such varieties, he finds it necessary to encourage the buds at the bases of the most distinctly marked leaves and to propagate from them alone. By following, with perseverance, this plan during three or four successive seasons a dis- tinct and fixed variety can generally be secured." Ernest Walker, a careful gardener at New Albany, Indiana, is of the opinion that the abnormal char- acter of sports often intensifies itself if the sport is allowed to remain upon the parent plant for a considerable time. He has observed this particu- larly in coleus, where color sports are frequent. "In these," he says, "the sport begins with a branch, which may be taken off and propagated as a new variety. If left on the parent, other parts of the plant are apt to show similar variations. Indeed, I think it is not best to be in too great a hurry to remove a sporting branch, for its char- acter seems to tend to become more fixed if it remains on the plant." 13. The starting-point once given, all permanent progress lies in continued selection. This, as I have already pointed out, is really the key to the whole matter. In the greater number of cases, the oper- ator cannot produce the initial variation which he desires, but, by looking carefully amongst many plants, he may find one which shows an indication SELECTION THE KEY-NOTE. 121 of his ideal. This plant must be carefully saved, and all the seeds sown in a place where crossing with other types cannot take place. Of a hun- dred seedlings from this plant, mayhap one or two will still further emphasize the character Avhich is sought. These, again, are saved and all the seeds are sown. So the operation goes on, patiently and persistently, and there is reward at the end. This is the one eternal and funda- mental principle which underlies the amelioration of plants under the touch of man ; and because we know, from experience, that it is so important, Ave are sure, as Darwin was, that selection in nature must be the most potent factor in the progress of the vegetable world. But suppose this suggestion of the new variety does not appear amongst the batch of plants which Ave raise? Then soav again; vary the con- ditions; select the most widely variable types; cross ; at length — if the ideal is true — the sug- gestion will come. "Cultivation, diversification of the conditions of existence, and repeated sow- ings " are the means which Verlot would employ to induce variations. But the skill and the char- acter of the final result lie not so much in the securing of the initial start, as in the subsequent selection. Nature affords starting-points in end- less number, but there are feAv men alert and skil- ful enough to take the hint and improve it. If I 122 HOW DOMESTIC VARIETIES ORIGINATE. want a new tomato, I first endeavor to discover what I want. I decide that I must have one like the Acme in color, but more spherical, with a firmer flesh, and a little earlier. Then I shall raise an acre of Acme tomatoes, and closely allied varieties ; or if I cannot do that, I make arrange- ments to inspect my neighbor's fields. I scruti- nize every plant as the first fruits are ripening. Finally, I find one plant — not one fruit — which is something like the variety which I desire. Very Avell ! Wait two to five years, and you shall see my new variety ! Some of these initial variations possess no ten- dency to reproduce themselves. The seedlings of them may break up into a great diversity of forms, no form representing the parent closely. In such cases, it is generally useless to proceed further with this brood. Another start should be made with another plant. So it is always impor- tant, as we have already seen (Rule 6), to have as many starting-points as possible, to lessen the risk of failure. Whilst it requires nice judgment to se- lect those plants which possess the most important and the most transmissible combination of charac- ters, the greatest skill is nevertheless required to carry forward a correct system of selection. 14. Even when the desired variety is obtained, it must be kept up to the standard by constant attention to selection. That is, there is no real stability in SELECTION TO MAINTAIN PURITY. 123 the forms of plant life. So long as the conditions of existence vary, so long will plants make the effort to adapt themselves to the changes. No two seasons are alike, and no two fields, or even parts of fields, are alike ; and there are no two cultivators who give exactly the same and equal attention to tillage, fertilizing, and the other treatments of plants. All forms or varieties, therefore, tend to "run out" by variation or gradual evolution into other forms ; hut because we keep the same name for all the succeeding generations, we fancy that we still have the same variety. In 1887 I found a single tomato plant in my garden in Michigan, which had several points of superiority over any other of the one hundred and seventy varieties which I was then growing. It came from a packet of German seed of an inferior variety. The tomato was very solid, an unusually long keeper, productive, and attractive in size and appearance. The variation was so promising that I named it in a sketch of tomatoes which I published that year, calling it the Ignotum (that is, unknown), to indicate that the origin of it was no merit of my own. I sent seeds to a few friends for testing. I sowed the seeds for about five hundred plants in 1888 in an isolated patch upon uniform soil. The larger part of the plants were more or less like the parent. A few reverted. A few of the best 324 HOW DOMESTIC VARIETIES ORIGINATE. plants were selected, and the seed saved. I then moved to New York and took the seed with me. This was sown in uniform soil in an iso- lated position in 1889. This crop, probably as a result of the careful selection of the year before and of the change of locality, was remarkably uniform and handsome. Of the 442 plants which I grew that year, none reverted to the little Eiformige Dauer, the German variety from which it had come, but there was some variation in them due to different methods of treatment. I again saved the seeds, and I was now read} 7 to intro- duce the variety. I therefore sold my seed, six pounds, to V. H. Hallock & Son, Queens, New York, who introduced it in 1890. The very next year, 1891, I obtained the Ignotum from fifteen dealers and grew the plants side by side. Of the fifteen lots, eight bore small and poor fruits which were not worth growing and which could not be rec- ognized as Ignotum ! Grown from our own seed, it still held its characters well. Here, then, only a year after its introduction, half the seedsmen were selling a spurious stock. It is possible that some of this variation arose from substitution of other varieties by seedsmen, although I have yet secured no evidence of any unfair dealing. It is possible, also, that the product of some of the samples which I early sent out for testing had found their way into seedsmen's hands. But I am DURATION OF VARIETIES. 125 convinced that very much of this variation was a legitimate result of the various conditions in which the crops of 1890 had been grown, and the varying ideals of those who saved the seeds. I am the more positive of this from the fact that the Ignotum tomato, as I first knew it and bred it, appears to be lost to cultivation, although the name is still used for the legitimate family of descendants from my original stock. All this experience illustrates how quickly varieties pass out by variation and by the unconscious and unlike selection practised by different persons. The duration of any variety is inversely propor- tional to the frequency of its generations. Annual plants, other conditions being the same, run out sooner than perennials, because seed-reproduc- tion — or the generations — intervenes more fre- quently. Trees, on the other hand, carry their variations longer, because the seed-generations — in which departures chiefly take place — are far- ther apart. Of all the so-called fruit plants, the strawberry runs out soonest and the varieties change the oftenest, because a new generation can be brought into fruit-bearing in two years, whilst it may require a decade or more to bring a new generation of apples or chestnuts into bearing. Yet, my reader will remind me that the Wilson strawberry has been and is the leading variety in many places for nearly forty } r ears, to which I 126 HOW DOMESTIC VARIETIES ORIGINATE. reply that tlie Wilson of to-day is not necessarily the same as that introduced by James Wilson, simply because the name is the same. Every dif- ferent soil or treatment tends to produce a different strain or variation in the Wilson strawberry, as it does in any other plant ; and every grower, when setting a new plantation, selects his plants from that part of his field which pleases him best, rather than from those plants which most nearly correspond to the original tj])G of the Wilson. That is, this unconscious selection on the part of the grower takes no account of what the variety was, but only of what it ought to be, and this ideal differs with every person. It is not surpris- ing, therefore, to find strains of Wilson strawberry which are as unlike as many named varieties are ; and it is to be expected that all of the strains now in existence have departed considerably from the original type. This example borrowed from the strawberry is a most important one, because it illustrates how a variety may vary and pass out of existence even though it is propagated wholly asexually, or by buds. There are to-day several different types of Rhode Island Greening apple in cultivation, which have originated from variations produced by envi- ronment and by the different models which propa- gators have had in mind ; and the same is true of many other fruits. AMELIORATION DUE TO SELECTION. 127 All the foregoing remarks demonstrate the importance of constant attention to selection if one desires to maintain the exact type of any variety which he has produced. They explain the value of the "roguing" — or systematic de- struction of all " rogues " or non-typical plants — which is invariably practised by all good seed- growers. But they still more emphatically show that every variety is essentially unstable, and that the only abiding result is constant evolution, the old forms being left behind as the type expands into new and better strains. Varieties to be valu- able, therefore, ought not to be rigidly fixed, and, fortunately, nature has prescribed that they can- not be. Probably every decade sees a complete change in every variety of any annual species which is propagated exclusively from seeds, and every century must see a like change in the tree fruits. These changes are so gradual, and the original basis of comparison fades away so com- pletely, that we generally fail to recognize the evolution. 15. It is evident, therefore, that the most abiding progress in the amelioration of plants must come as a result of the very best cultivation and the most intel- ligent selection and change of seed. Every reflec- tive person must admit that the cultivation of plants — which is the fundamental conception of agriculture — has been and is crude and imperfect, 128 HOW DOMESTIC VARIETIES ORIGINATE. and that there has been no conscious effort on the part of the human race to produce any given final re- sult upon the cultivated flora. Yet, this imperfect cultivation has already modified plants so pro- foundly that we cannot determine the originals of many of them, and we can trace the evolution of but few. The science of rural industry is now fairly well understood in it's essential fundamental principles, and the intelligence of those classes of persons who deal with plants is rapidly enlarging. The opening of the twentieth century will virtu- ally mark a new era for agriculture, and from that time on the onward evolution of plants should proceed confidently and unchecked. Our eyes are too often dazzled by the novelties which suddenly thrust themselves upon us, and we look for some mystic power which shall enable us to produce varieties forthwith at our will. We need not so much varieties with new names as we do a general increase in productiveness and efficiency of the types which we already possess ; and this augmen- tation must come chiefly in the form of a gradual evolution under the stimulus of good care. The man who will accomplish most for the amelioration and unfolding of the forms of plants, is he who fixes his eyes steadily upon the future, and with the inspiration of a long forecast, urges the better- ment of all conditions in which plants grow. DEWBERRY AND BLACKBERRY. 129 III. Specific Examples. The foregoing principles and discussions will become more concrete if a few actual examples of the origination of varieties are given. In order to begin with a very simple case, I will relate the introduction of the varieties of dewberries, for this fruit is yet little cultivated, the varieties are few, and the domestication of it is not yet thirty years old. The Dewberry and Blackberry. The dewberries are native fruits, and it is only within the last ten years that they have become prominent among fruit-growers. The most impor- tant one is the Lucretia. This was found grow- ing w T ild upon a plantation in West Virginia in war time. In 1876, a few of the plants were sent to Ohio, and from this start the present stock has come. It is probable that similar wild varieties are growing to-day in many parts of the country, but they have not chanced to have been seen by per- sons who are interested in cultivating them. It is a form of the common wild dewberry, which grows all over the northeastern states. Just why this particular patch in West Virginia should have been so much better than the general run of the species, nobody knows, but it was undoubtedly the prod- uct of some local environment of soil or position. 130 HOW DOMESTIC VARIETIES ORIGINATE. Early in the seventies, T. C. Bartel, of Huey, Clinton Co., Illinois, observed very excellent dew- berries growing in rows between the lines of stubble in an old cornfield, where the plant had evidently been quick to avail itself of unoccupied land. This was introduced as the Bartel dew- berry, and is now the second in point of promi- nence amongst the cultivated varieties. Other varieties have appeared in much the same way. A fruit-grower in Michigan found an extra good dewberry in a neighboring wood-lot, and intro- duced it under the name of Geer, in compliment to the owner of the place. In Florida, an un- usually good plant of the common wild dewberry of that region was discovered, and introduced by Reasoner Brothers, under the name of Manatee. There are now about twenty named varieties of dewberries in cultivation, as described in our horticultural writings, all of which, so far as I know, are chance plants from the wild. As the dewberries become more widely grown, good seedlings will now and then appear in cul- tivated ground, and these will be named and sold. After a time persons will begin to sow seeds for the purpose of producing new varieties ; and those seedlings which chance to possess un- usual merit will be propagated, and in due time introduced. This is the history of the cultivated blackberries and raspberries which have come EVOLUTION OF THE APPLE 131 from the wild plants in less than half a century. These fruits are now so far developed that we no longer think of looking to the woods and copses for new varieties of promise, but the novelties are mostly chance seedlings from cultivated varieties. A few years ago a friend purchased plants of the Snyder blackberry. When they came into bear- ing he noticed that one plant was better than the rest. It bore larger fruits, and the bearing season was longer. He took suckers from this plant, and from these others were taken, until he now has a large plantation of the novelty, mostly selected from plants which pleased him best. The variety has such distinct merit that I have named it the Mersereau, in honor of the man who recognized and propagated it. He will continue selecting from the best plants, as he propagates year by year, and it may be that in a few years he will have so much improved it that it will no longer be the variety with which he started. The Apple. The original apple is not definitely known, but it was certainly a very small and inferior, crabbed fruit, borne mostly in clusters. When we first find it described by historians, it was still of small value. Pliny said that some kinds were so sour as to take the edge off a knife. But better and 132 HOW DOMESTIC VARIETIES ORIGINATE. better seedlings continued to come up about habi- tations, until, when printed descriptions of fruits began to be made, three or four hundred years ago, there were many named kinds in existence. The size had vastly improved, and Avith this in- crease came the reduction of the number of fruits in the cluster; so that, at the present time, whilst apple flowers are borne in clusters, the fruits are generally borne singly. That is, most of the flowers fail to set fruit, and they complete their mission when they have shed their pollen for the benefit of the one which persists. The American colonists brought with them the staple varieties of the mother countries. But the needs of the new country were unlike those of the old, and the tastes and fashions of the people were changing. So, as seedlings came up about the buildings and along the fences, where the seeds had been scattered, the ones which prom- ised to satisfy the new needs best were saved, and many of the old varieties were allowed to pass away. In 1817, the date of the first American fruit-book, over sixty per cent of the varieties particularly recommended for cultivation in this country were of American origin. In 1845, nearly two hundred varieties of apples were de- scribed as having been fruited in this country, of which over half were of American origin. Be- tween these two dates, introductions of foreign EVOLUTION OF THE APPLE. 133 varieties had been freely made, so that the per- centage of domestic varieties had fallen. But the next thirty years saw a great change. Of 1823 varieties described in 1872, nearly or quite seventy per cent were American, and a still greater proportion of the most prized kinds were of domestic origin. In the older states, the apple had now become so thoroughly accustomed to its environment, and the tastes of the people were so well supplied, that there was no longer much need for the introduction of foreign kinds. It was not so in the Northwest. There the apples of the eastern states did not thrive. The climate was too cold and too dry. Attention was turned to other countries with similar or rigorous cli- mate. In 1870, the Department of Agriculture at Washington imported cions of many varieties of apples from Russia; but these did not satisfy many fruit-growers of the northern states. It was then conceived that the great interior plain of Russia should yield apples adapted to the upper Mississippi valley, whilst those already imported had come from the seaboard territory. Accord- ingly, early in the eighties, Charles Gibb, of the province of Quebec, and Professor Budd, of Iowa, went to Russia to introduce the promising fruits of the central plain. The result has been a most interesting one to the pacific looker-on. There are ardent advocates of the Russian varieties, and 134 HOW DOMESTIC VARIETIES ORIGINATE. there are others who see nothing good in them. There are those avIio believe that all progress must come by securing seedlings from the hardi- est varieties of the eastern states; there are others who would derive everything from the Siberian crabs, and still others who believe that the final result lies in improving the native crabs. There is no end of discussion and cross-purposes. In the meantime, nature is quietly doing the work. Here is a good seedling of some old variety, there a good one from some Russian, and now and then one from the crab stocks. The new varieties are gradually supplanting the old, so quietly that few people are aware of it; and by the time the con- testants are done disputing, it will be found that there are no Russians and no eastern apples, but a brood of northwestern apples which have grown out of the old confusion. All these new apples are simply seedlings, almost all of them chance trees which come up here and there wherever man has allowed nature a bit of ground upon which to make garden as she likes. In 1892, there were 878 varieties of apples offered for sale by American nurserymen, and it is doubtful if one in the whole lot was the result of any attempt on the part of the originator to produce a variety with definite qualities. And Avhat is true of the apple, is about equally true of the other tree fruits. In the small fruits and BEANS. 135 the grapes, where the generations are shorter and the results quicker, more has been done in the way of direct selection of seeds and the crossing of chosen parents; but even here, the methods are mostly haphazard. Beans. Perhaps there are no plants more tractable in the hands of the plant-breeder than the garden beans. Some two or three years ago, a leading eastern seedsman conceived of a new form of bean pod which would at once commend itself to his customers. He was so well convinced of the merits of this prospective variety, that he made a descriptive and " taking " name for it. He then wrote to a noted bean-raiser, describing the pro- posed variety and giving the name. " Can you make it for me ? " he asked. " Yes, I will make you the bean," replied the grower. The seeds- man then announced in his catalogue that he would soon introduce a new bean, and, in order to hold the name, he published it, along with the announcement. Two years later, I visited the bean-grower. " Did you get the bean ? " I asked. "Yes, here it is." Sure enough, he had it, and it answered the requirements very well. Another seedsman would like a round-podded, stringless, green-podded bean. This same man produced 136 HOW DOMESTIC VARIETIES ORIGINATE. it, and J went into a field of fifteen acres of it, where it was growing for seed, and the most fas- tidious person could not have asked for a closer approach to the ideal which the breeder had set before him some four or five years before. How is all this done ? It looks simple enough. The ideal is established first of all. The breeder revolves it in his mind, and eliminates all the impracticable and contradictory elements of it. Then he goes carefully and critically through his bean fields, particularly those varieties which are most like the desired kind, and marks those plants which most nearly approach his ideal The seeds of these are carefully saved, and they are planted in isolated positions. If he finds no promising variations amongst his plantations, then he must start off the variation in some other way. This is usually done by crossing those varieties which are most like the proposed kind. He has got a start ; but now the science and skill begin. Year by year he selects just those plants which please him best and which he judges, from experience, will most surely carry their features over to the offspring. He starts with one plant ; the next year he may have only two. If he has ten or twenty good ones, then the task is an easy one, for the variety has elements of permanence — that is, of hereditability — in it. But he may have no plants the second year. In that case, he begins BEANS. 137 again ; for if the ideal is true, it can be attained. This bean-breeder to whom I have referred, and upon whom many of our best seedsmen rely for new varieties, tells me that he has discarded fully three thousand varieties and forms as profitless. This only means that he is a most astute judge of beans, and that he knows when any type is likely to prove to be a poor breeder. The bean also affords an excellent example of the care which it is generally necessary to exercise to keep any variety true to the type. The person of whom I have spoken, in common with all care- ful seed-growers, searches his field with great pains to discover the "rogues," or those plants which vary perceptibly from the type of the given variety. The rogue may be a variation in size or habit of plant, season of maturity, color or form of pods, productiveness, susceptibility to rust, or other aberrance. In the dwarf or bush beans, which are now most exclusively grown, the most frequent rogue is a climbing or half-climbing plant. This is a reversion to the ancestral type of the bean, which was no doubt a twining plant. This rogue is always destroyed, even though it may be, itself, a good bean. In some cases, the men who perform the roguing are sent along every row of a whole field on their hands and knees, critically examining every plant. The ef- fect of this continual selection is always to push 138 HOW DOMESTIC VARIETIES ORIGINATE. on the variety to greater excellence. The vari- ous " improved " strains of plants are obtained in essentially this fashion. If the grower has been painstaking with his roguing, he soon finds that his seed gives better and more uniform crops than the common stock of the variety. If the improve- ment is marked, he may dignify his strain with a distinct name, and it thereby becomes a new variety. The improvement may be a very im- portant one to a careful bean-grower, and at the same time be so slight as to escape the attention of the general farmer, or even of experimenters who are not particularly skilled in judging the merits of beans. All these examples drawn from the bean are excellent illustrations of the best and most scien- tific plant-breeding, and the same methods — varied to suit the different needs — apply to the ameliora- tion of all other plants. The recent dwarf Lima beans may be cited as examples of accidental or fortuitous varieties, in which the preconstructed ideal of the plant-breeder had no place. Four or five of these beans have attained some prominence. Henderson and Kumerle dwarf Limas were intro- duced in 1889, Burpee in 1890, and Barteldes in 1892 or 1893. The variety which is now called the Henderson was picked up twenty or more years ago by a negro, who found it growing along a roadside in Virginia. It was afterwards grown BEANS. 139 in various gardens, and about 1885 it fell into tlie hands of a seedsman in Richmond. Henderson purchased the stock of it in 1887, grew it in 1888, and offered it to the general public in 1889. The introduction of Henderson's bean attracted the attention of Asa Palmer, of Kennett Square, Pennsylvania, who had also been growing a dwarf Lima. He called upon Burpee, the well-known seedsman of Philadelphia, described his variety, and left four beans for trial. These were planted in the test grounds and were found to be valuable. Mr. Palmer's entire stock was then purchased, — comprising over an acre, which had been carefully inspected during the season — and Burpee Bush Lima was presented to the public in the spring of 1890. Mr. Palmer's dwarf Lima originated in 1883, when his entire crop of Large White (Pole) Limas was destroyed by cut-worms. He went over his field to remove the poles before fitting the land for other uses, but he found one little plant, about ten inches high, which had been cut off about an inch above the ground but which had re-rooted. It bore three pods, each containing one seed. These three seeds were planted in 1884, and two of the plants were dwarf, like the parent. By discarding all plants which had a tendency to climb, in succeeding crops, the Burpee Bush Lima, as we iioav have it, was developed. The Kumerle, Thorburn, or Dreer, Dwarf Lima originated from 140 HOW DOMESTIC VARIETIES ORIGINATE. occasional dwarf forms of tlie Challenger Pole Lima, which J. W. Kumerle, of Newark, New Jersey, found growing in his field. The stock which came from these selected dwarf plants was introduced by Thorburn and Dreer, under their respective names. The singular Barteldes Bush Lima came from Colorado, and is a similar dwarf sport of the old White Spanish or Dutch Runner bean. Barteldes received about a peck of the seed and introduced it sparingly. It attracted very little attention, and as the following season was dry, Barteldes himself failed to get a crop, and the variety was lost to the trade. Carinas. Few plants have shown more remarkable evolu- tions in very recent years than the cannas. At the present time, the Crozy cannas — so named from Crozy, of Lyons, France, who has introduced the greater number of them — are most popular. This type is often called the French Dwarf, or the Flowering Canna, and it is marked by a com- paratively low stature, and very large and showy spreading flowers in many colors, whereas the can- nas of a few years ago were very tall plants, Avith small and late dull red, narrow flowers, and they were grown exclusively for their foliage effects. How has this transformation come about? CANNAS. 141 In the first place, it should be said that there are many species of canna, and about a half dozen of these were well known to gardeners at the opening of the century. About 1830, the cannas began to attract much attention from cultivators, and the original species were soon variously hybrid- ized. Crossed seeds, and seeds from the succes- sive generations of hybrids, introduced a host of new and variable forms. The first distinct fash- ion in cannas seems to have been for tall, late- flowering forms. In 1848, Annee, a cultivator in France, sowed seeds of Canna Nepalensis, a tall oriental species, and there sprung up a race of plants which has since been known as Canna Annan. It is probable that this Canna Nepalensis had become fertilized with other species growing in Annee's collection, very likely with Canna glauca. At any rate, this race of cannas became popular, and was to its time what the French dwarfs are to the present day. The plants were freely introduced into parks, beginning about 1856, but their use began to wane by 1870 or before. Descendants of this type, variously crossed and modified, are now frequently seen in parks and gardens. The beginning of the modern race of dwarf, laro-e-flowered cannas was in 1863, when one of the smaller-flowered Costa Rican species (Canna Warscewiczii) was crossed upon a large-flowered 142 HOW DOMESTIC VARIETIES ORIGINATE. Peruvian species (Canna iridiflora). The off- spring of this union came to be called Canna Ehemanni. This hybrid has been again variously crossed with other species, and modified by culti- vation and selection, until the present composite type is the result. Seeds give new varieties; and any seedling which is worth saving is thereafter multiplied by divisions of the root, and the result- ing plants are introduced to commerce. These various examples are but types of what has been and can be accomplished in a given group of plants. There is nothing mysterious about the subject, so far as the cultivator is concerned. He simply sets his ideal, makes sure that it does not contradict any of the fundamental laws of devel- opment of the plant with which he is to Avork, then patiently and persistently keeps at his task. He must have good judgment, skill, and inspira- tion, but he does not need genius. "In the improvement of plants," writes Henri L. de Vilmorin, "the action of man, much like influences which act on plants in the wild state, only brings about slow and gradual changes, often scarcely noticeable at first. But if the efforts toward the desired end be kept on steadily, the changes will soon become greater and greater, and the last stages of the improvement will become much more rapid than the first ones." LECTURE IV. BORROWED OPINIONS; BEING EXTRACTS FROM THE WRITINGS OP B. VERLOT, E. A. CARRIERE. AND W. O. FOCKE. I. Verlot's Classification of Varieties of Ornamental Plants. Verlot (Sur la Production et la Fixation des Varietes dans les Plantes oV Ornement) distributes the varieties of ornamental plants into sixteen groups. I shall now transcribe these groups, and under each shall give a very brief quotation or abstract of some of his remarks concerning them. 1. Varieties distinguished by diminution of stat- ure, or dwarfing. — Dwarfing is one of the most frequent variations in the vegetable kingdom, but, unlike many similar phenomena in the animal kingdom, these dwarfs are nearly always very fertile. If we question the cultivators upon the subject, they respond that these variations are purely accidental, and that whenever the varia- tions offer any reward they are propagated and distributed. Dwarfing may be brought about by sowing the seeds in the autumn (page 115), and at the same 143 144 BORROWED OPINIONS. time successively transplanting the plants, as they need. Suppose, for example, we soav seeds of Co- reopsis tinctoria in August or September. When the plants have developed leaves, the}?- are trans- planted, leaving sufficient space between them to allow of liberal growth. When the plants begin to touch each other, transplant again, per- haps three or even four times. The plants become strong, vigorous, and stocky ; we encour- age the development of the lowest branches and thereby tend to shorten the leading stem, thus making the individual comparatively dwarf. The seeds saved from plants thus treated during several generations will be more apt to produce dwarf varieties than seeds taken from other plants. The greater part of dwarf varieties appear in those plants which are sown in autumn, and in those, if sown in spring, which are submitted to successive transplantings. Thus, amongst the annual species which we habitually sow in July and September, the following have produced dwarf varieties : — Calceolaria plantaginea. Senecio cruentus. Lychnis (or Agrostemma) Cceli-rosa. Coreopsis (or Calliopsis) tinctoria. Oenothera Drummondii. Helichrysum bracteatum. Leptosiphon densiflorus. veklot's classification. 145 Diantlms Cliinensis. Scabiosa atropurpurea. Schizanthus retusus. Iberis umbellata. Amongst those species which we sow in spring, but frequently transplant, the following have dwarf forms : — Impatiens Balsamina. Callistephus hortensis. Tagetes patula. Tagetes erecta. Tagetes signata. 2. Varieties distinguished by augmentation of stature, or giant forms. — These varieties result from various causes, amongst which are amount and fertility of soil, the employment of newly harvested seeds, and crossing. 3. Hardy varieties. — These are produced by successive selections from the most hardy indi- viduals. Hardy races are also obtained by cross- ing with hardy species or types. Thus, the forms of Rhododendron arboreum are rendered hardier when crossed with R. Catawbiense. 4. Large-flowered varieties. — These variations are always due to a good soil which is rich in humus, and above all by thorough and intelligent cultiva- tion. These are easily fixed, but they pass away 146 BORROWED OPINIONS. insensibly when the conditions under which they were produced are neglected. This return to small flowers is well illustrated in the pansy. [All plants which are grown for the beauty of their flowers tend to increase the size of those organs, because of the vigor which comes of good care, and the selection which necessarily follows. The frequency of the varietal name "grandiflora" is proof of this. As soon as the plant has made «ny perceptible gain in the size of its flowers, some nurseryman adds this adjective to its name, as in Hydrangea paniculata grandiflora. No doubt the name is sometimes bestowed without warrant, for the purpose of selling the plant. Frequently the catalogue-maker drops the proper specific name and uses the varietal adjective as if it were the legitimate name of the plant, as in Gazania grandi- flora, which is properly Gazania rigens, variety grandiflora.] 5, 6. Early and late varieties. — The chief agency affecting the duration of plants is climate, and desired variations in earliness or lateness are obtained by transporting the plants to those cli- mates which produce such effects as we seek, and growing them there for a time ; or we can procure seeds from such climates, if the given plants are already grown there. The age of the seed also has an influence upon the resulting individual : the fresher the seed, the veelot's classification. 147 more rapid its germination, and consequently the more prompt the development of the plant. We have reason to expect that fresh seeds will have a tendency to produce early varieties, and that, on the contrary, old seeds, by germinating more slowly, will produce variations more or less late. [Not only does the age of the seed seem to be important in this connection, but recent experi- ments seem to show that the degree of maturity also modifies the offspring. Seeds which are barely ripe enough to germinate have a tendency to give earlier progeny than those which are fully matured and ripened. See page 103.] The first seeds to mature on any plant may be expected to give early plants, and the subsequent seeds give later plants. [Verlot states that a cold region tends to make the plants later when trans- ferred to it, but this seems to be an error. The fact is generally just the reverse. Plants taken towards the poles or to higher altitudes, become earlier in two ways, by shortening their period of growth, and by vegetating at a less sum- temperature in spring. See page 26.] 7. Odoriferous varieties. — Odor varies greatly, even amongst varieties of one species. The causes of the differences in fragrance are not numerous and they are little understood. Climate, exposure, and the nature of the soil are leading factors. The odor of plants which grow on dry and arid 148 BORROWED OPINIONS. hills is much more penetrating than that of the same species cultivated in humid and shady places. It is possible, even, to entirely change the odor by transporting the plant from one place to another. For example, Satyrium hircinum exhales a most pronounced goat-like odor in the vicinity of Paris and northward, whilst in the east, and particularly in the southern regions, its flowers have an odor which is somewhat like that of vanilla. 8. Varieties ivith colored parts. — Coloration may be either complete or partial, and it may reside in any or all parts of the plant, as follows : — The stems. The leaves. The flowers, j var f/f ed - ( spotted. The fruits. The seeds. Variations in color are the most frequent of all modifications in cultivated plants. These depart- ures may be expected to arise under the influence of continued cultivation and repeated sowings; and the variations must then be selected until they are fixed. 9. Varieties without color, or albinos. — Partial albinism, or variegation, is as frequently observed in spontaneous plants as it is in cultivated ones. It usually occurs in the leaves only, but it is some- verlot's classification. 149 times a feature of the entire plant. A variegated plant does not exist of which we do not know the non- variegated type. These variegated plants appear both from seeds and from bud-variations, and they are most surely propagated in the latter case. It has been said that when the albinism affects the margin of the leaf it is more likely to be transmitted than when it occupies the central part of the blade, but this generalization has many exceptions. Page 157. It is a curious fact that variegation and double- ness of flowers are generally antagonistic, for they do not appear in the same plant. One excludes the other. It is supposed by Morren [and gen- erally accepted] that doubling is the result of excessive vigor and that partial albinism comes of an enfeebling of the vital functions. Variegations sometimes disappear entirely and then, after two or three years, reappear in the same individual. The first leaves of seedlings from variegated plants may be perfectly green, and the seedlings may afterwards take on the varie- gated character. This behavior is well marked in some ferns. Complete albinism, or chlorosis, indicates a pro- found alteration in the tissues, and it is impossi- ble of propagation. This decoloration is most commonly a bud- variation. 10. Double varieties, or those distinguished by 150 BORROWED OPINIONS. the transformation of the sta?7iens and pistils into petaloid organs — There are various degrees of doubling or duplication in flowers. The calyx and corolla alone may be duplicated, in which case the flower is still fertile. Sometimes the sta- mens only are transformed into petal-like organs, and the flower is then fertile if pollen is trans- ferred from another flower. Sometimes all the floral series — calyx, corolla, stamens, pistils — may be duplicated or transformed ; then we have a full (pleine) flower, which is incapable of pro- ducing seeds or of fecundating another flower. [Annual plants, and others not propagated by buds and other asexual parts, which bear full double flowers, must be propagated by seeds taken from flowers which are nearly full double, but which bear a few seeds ; or, sometimes from a nearly single flower which is fertilized by pollen from a nearly full double flower. In these cases, it is unusual for all the seedlings to produce full double flowers.] A rich soil, a cultivation which produces a luxu- riant vegetation, are the conditions which gen- erally produce doubling in flowers. But we can repeat with De Candolle, " That if we are gen- erally ignorant of the causes of the doubling of flowers, Ave also know that if we gather seeds from an individual with semi-double flowers, the plants which result have a greater tendency to pro- verlot's classification. 151 cluce double flowers than seeds taken from simple flowers." Doubling may occur in all plants, whether an- nuals, biennials, herbaceous or woody perennials, and in all of them, when they are fertile, we can finally make them reproduce the character iden- tically. We must always choose for seed-parents the individuals of which the flowers are very double, and exclude with the greatest care the single- flowered plants, which are the most fertile and the progeny of which quickly smother the progeny of the double flowers. 11. Proliferous varieties. — [These are varia- tions which are characterized by growths arising from unusual places, as one flower springing out of another flower, a branch or rosette growing out of a flower, an unusual production of bulbs or young plants from the root, bulbs from leaves or the fronds of ferns, and the like.] These forms are infrequent in cultivated plants and very rare amongst wild plants. They are generally asso- ciated with the fertility of the soil. The pro- liferous form of Papaver somniferum known as Papaver monstruosum, perpetuates itself per- fectly by seeds, but these variations are usually unstable. 12. Varieties ivith conjoined parts (yariStSs par soudures). — We know of a single example of this 152 BORROWED OPINIONS. monstrosity in ornamental plants : it is that of Papaver bracteatnm, in which the corolla has become monopetalous by the growing together of the petals. This monstrosity (described and fig- ured in Revue Horticole) is cultivated by Vilmorin. They can propagate it only by cuttings. They have tried in vain to multiply it by seeds. [Simi- lar forms of other species are known.] 12. Abortive varieties. — This type of monstros- ity, which constitutes one of the most interesting chapters in vegetable teratology, has been ob- served in all parts of the flower. They are mal- formations which have no interest from the point of view of ornament. [The petals, or other or- gans, sometimes almost entirely disappear in this type of variations.] 11. Peloric varieties. — [Peloria is a name ap- plied by Linnaeus to a form of the toad-flax, Lina- ria vulgaris, in which all of the five petals have spurs, while the normal form has only one petal spurred. The term is now applied generically to all similar regularity of structure in normally ir- regular flowers.] The causes which produce tins transformation are not known, but aridity and dry- ness of the soil, and new conditions of vegetation, appear to favor its development. These mon- strosities, at least in linaria, are propagated easily by cuttings or buds, and Willdenow records an experiment in which they came true from seeds. VERLOT'S CLASSIFICATION* 153 15. Chloranthic varieties. — Here are included, in a general way and in lieu of a better name, all those transformations which render the flowers absolutely sterile, and transform them more or less completely into branches or leaf -like organs. They are purely bud- variations, and can be per- petuated only by cuttings, buds, or other asexual parts. 16. Various or polymorphous varieties, compris- ing the following types : — thornless, spineless, Stems \ f astigiate, filiform, weeping, etc. f crisped, T J fasciated, bullate or blistered, laciniate, etc. The various modifications originate both by seed-variation and bud-variation. II. Carriere's Account of Bud-Varieties. The subject of bud-variations or sports never fails to interest the student, and however familiar he may be with these forms he never ceases to 154 BORROWED OPINIONS. wonder at them. I have taken pains, therefore, in addition to what I have already said upon the subject (see pages 28, 117), to translate almost bodily Carriere's account of bud-varieties (in Pro- duction et Fixation des Varietes dans les Veye- taux), because, although written in 1865, it is the most extended list of bud-varieties which I know. The catalogue might be greatly extended by inserting the current varieties in commerce in this country, but the original list is sufficiently full for all purposes of illustration. Carriere's account now follows : — 1. General Remarks upon Bud- Variation. Plants being composed of a certain number of elements disposed in a certain order, and, more- over, these elements, under the influence of organic laws, being able to separate or group themselves in different ways, it follows that the same plant can, upon its different parts, present characters and properties more or less different from those which it normally presents. It is this fact which constitutes that which in practice we call an acci- dent [or bud variety], either of dimorphism 1 [of form] or of dichroism [of color]. 1 French writers use the word accident in the sense in which we use bud-variation. The word dimorphism, used by Car- riere for one of the features of bud- variation, is now applied to carriere's remarks on bud- variation. 155 We refer to bud-variation the phenomenon, whose cause is unknown, which allows a bud on any part of the plant to develop a member whose form and appearance differ from those borne on other parts of the plant. Thus the common beech producing a branch with laciniate leaves, Podocar- pus Koraiana producing a branch whose ramifi- cations are whorled and spreading instead of being scattered, and Avhose leaves are distichous instead of being alternately disposed about the branches as they are normally, are examples of bud-varie- ties. Taken in its most absolute sense and considered in the sum of all its characters, bud- variation, aside from the details which it presents, can be divided into two sections : one which includes all the phe- nomena which are manifested suddenly, as in the case of the fern-leaved beech, the hemp-leaved rose, the English willow -leaved cherry, sour grapes Avith long seeds, etc.; the other includes all slower transformations, as in the case of Rosa Eglanteria, tulips, Iris Xiphium, Viola Rotho- magensis, var. pallida, etc. Strictly, we could establish a third section to include all the trans- formations resulting from the age of the individ- different permanent and characteristic forms of individuals of the same species. It is most commonly observed in the different relative lengths of stamens and pistils. I have substituted other words for it in most places in the text. — L. H. B. 156 BORROWED OPINIONS. ual, which are the consequences of its aclultness. HoAvever, this last series of phenomena is seen only in polymorphous species, which change in appearance, form, and nature when they grow old and especially when they bear fruit ; such are the ivies, Ficus stipulata or scandens, eucalyptus, etc. Horticulture often profits by this peculiar prop- erty of plants ; multiplying separately the parts with the exceptional characters, it obtains individ- uals which present an appearance different from the plants from which they arise. [This dissimi- larity between young and mature individuals of the same species is well marked in some of the Conifera3, as the cedars and retinosporas.] In a general way, then, dimorphism refers to a different form on the same individual, whether the change be complete or partial. Dichroism is exactly analogous to it in essential points, only that it refers to color instead of to form. Thus, Flon's Pink [Dianthus semperflorens of gardens, introduced by M. Flon of Angers], which has red flowers, developing a branch similar to the plant in aspect and form, but bearing white flowers, the ovate -leaved privet and the Japanese fusain [Euonymus Japonicus] producing buds giving rise to variegated leaves, white kidney-beans pro- ducing black ones, and vice versa, are examples of dichroism. Let us say that in bud-variation, less than any- CARRIEKE ON BUD-VARIATION. 157 where else, we can do nothing towards obtaining or producing the variations. Bud-varieties most often spring up spontaneously, so to speak, and in this respect our work is purely passive, consisting in superintending these digressions or accidents in the endeavor to take advantage of them when they are presented. Let us state, also, that in these series of varieties we find a considerable diversity, either in the habit or aspect of the plants or in their foliage or flowers, or sometimes even in their fruits, and that we oftener find vari- egations than among plants which come from seeds. We ought to recall just here, — what we have said on the subject of plants issuing from seeds, — that variegations are the more constant the more completely they circumscribe the organs upon which they occur, whether upon the flowers or the leaves ; also, that when, on a plant whose variega- tions are disposed in stripes or bands, we find a part upon which they are disposed circularly, we can be almost certain that, if we detach and graft or make a cutting of this part, we shall preserve its new character. This phenomenon is very fre- quent in the camellias and especially in the azaleas. The greater part of the varieties of azalea which present these characters have had no other origin. 1 1 To preserve variegations, it is best to resort to graftage, generally speaking, as cuttings tend to produce individuals mure vigorous, and winch therefore tend to return to the green 158 BORPOWED OPINIONS. Certain species are much more disposed than others to produce these bud-varieties, either of dimorphism or dichroism. We give an example from the Chinese chrysanthemum. About 188(3, the horticultural establishment of Fromont re- ceived from England three varieties of this chrysanthemum; one had the flowers red, one variegated, and one white flesh-colored. Planted in the open air, the following year we saw the three varieties on one plant, which seems to show that these three varieties were only sports from a common form. A phenomenon analogous to the preceding, and which, like it, concerns the Chinese chrysanthemum, was shown at the Mu- seum 1 in 1856 upon a variety called Surprise. This, which bore flowers scarcely rose flesh-col- ored, produced, on one of its branches, flowers of a deep rose-lilac. Cuttings having been made, it has preserved all its characters, and to-day it is still one of the most beautiful of the section. We call it Gain du Museum. In 1862, upon this same Gain du Museum, a branch developed which bore flowers perfectly white, of almost the same size and form as those of the type; then upon dif- color, or even to the normal form, if the variety differs also in form. We must select the parts, in perpetuating variegations, in which the variegation is very pronounced, although we must exercise care that the variegation be not too intense, else the offspring will be weak and poor. — Carkieke. (See page 149.) 1 Museum d'llistoire Naturelle de Paris. .CARRIERE ON BUD-VARIATION. 159 ferent branches beside it were found others bear- ing flowers half red, half white. In making cut- tings from these two kinds of branches, we would then obtain from the Surprise still other varieties. Let us look at the variety Sophie. This, which has dirty white flowers very slightly tinged with red, with yellow centre, has produced, by bud- variation, a plant known as Trophee. The latter, which has flowers of a rose-lilac-violet, bears some resemblance to the Gain du Museum. There were also upon the same branch, but on different twigs, flowers similar to those borne by the varie- ties Trophee and Sophie. These new flowers were flat and had narrow and imbricated petals, whilst the Trophee has convex flowers, large and slightly serrate petals. The Madame Richard chrysanthemum, of which the flowers are whitish very lightly bordered with rose, has produced on one of its branches violet flowers stronger than those of the plant from which it sprung; the petals are also larger and more imbricated. In 1863 Ave observed on certain varieties of chrysan- themum the following sports: the variety called Ceclo nulli, with double white flowers very lightly rose, produced a branch which bore flowers much larger and much more spreading than those of Cedo nulli. The Argentine, with small white flowers, pompon-form, gave a branch more vigor- ous than itself, whose spreading, very large flow- 160 BORROWED OPINIONS. ers, of a beautiful yellow, resembled to a certain extent those of the large-flowered chrysanthemum, a fact which tends to show that from the pompons to the large-flowered sorts there is but a step. In 1864 we saw upon a stem of the Vesta (a pompon chrysanthemum which has white flowers) several branches which bore flowers entirely deep yellow. The dimensions, as well as the form of the flowers, were the same. Varieties obtained by bud-variation are very numerous. There is not a genus among those which comprise a number of species which has not produced them. Although we shall mention, farther on, a certain number of these bud-varie- ties, adding some observations, there are some which, in our opinion, are so interesting that, by anticipation, we ought to speak of them here. One of them relates to a kind of pink which is known in commerce as Flon's pink. This Flon's pink, which is closely related to those which we call Spanish pink, Badin pink, etc., has flowers very deep red, almost double, so that it does not produce seeds, and we are obliged to multiply it by cuttings. Nevertheless, it has already given, by bud-variation, several varieties, of which the most remarkable, a very beautiful white, was de- veloped in 1858. Since that time this variety has been maintained with all its characters. Ob- tained by M. Pare, horticulturist at Paris, this CARRIEUE ON ROSE SPORTS. 161 variety has been called Marie Pare, for one of the children of the originator. Other varieties, presenting colors different from that of which we have spoken, have been developed from Flon's pink by M. Pare. [The pinks are fertile in bud- varieties, particularly the carnation. Many of the carnations which are now well known to com- mercial growers first appeared as sports. The Portia, which is a deep self -red, frequently sports, sometimes into almost pure white.] The genus which, probably, has produced the most examples of this nature is the Rose. The examples are so very interesting that we cannot resist the temptation to say something in detail concerning them. We will cite several remark- able examples, commencing with those which have sprung from the Hundred-leaved Rose [Pro- vence rose, Rosa centifolia]. The bud-varieties which have issued from this rose can be arranged in two series : one which includes all individuals which are but little removed from the type, which differ from it only in color or form, either of the flowers or sometimes of the leaves, and comprise the ordinary Hundred-leaved roses; the other series includes individuals possessing the charac- ters of the first series, but which, in addition, are provided with small bracts or glandular hairs which give the name "Moss-rose." Bud-varieties produced by Rosa centifolia : — 162 BORROWED OPINIONS. A. Ordinary Hundred-leaved roses. I. Flowers more or less large. Cabbage-leaved or lettuce-leaved B. cen- tifolia. Celery-leaved. Anemone. Nancy. Peintres. Flore magno, or Foliaceous. Apetalous. Unique white. Unique variegated. II. Flowers small. — Pompons. Burgogne pompon. White pompon. Bordeaux pompon. Kingston pompon. B. Moss-roses. I. Flowers more or less large. Ordinary. Cristata. White-flowered Variegated. Sage-leaved. Unique Provence. Zoe, or Mousseuse partout. II. Flowers small. — Pompons. Pompon. CABEIEBE ON ROSE SPORTS. 163 One must not suppose that all the moss-roses which he meets with to-day in commerce are the result of bud- variation. The larger part, on the contrary, come from seeds. The moss-rose is nearly a race. From seeds taken from the moss- rose, we have obtained a certain number of individuals which have preserved the general characters of the plants from which they came ; they are more or less " mossy." Let us state, however, that this "mossy" character is not pecu- liar to any section of roses, but that we find it in most garden species, as the hybrid remontants, rose-of-four-seasons, etc. The fact of the repro- duction of the " mossiness " of roses by seeds, proves again, what we have asserted several times, that everything in a plant tends to reproduce itself, that the peculiarities, properties, monstrosi- ties even, may become hereditary. The Zoe moss-rose is one of the most remark- able bud-varieties which has been produced by Rosa centifolia. This variety, instead of being " mossy " only upon the peduncle or calyx, as most of the other varieties of this group are, is "mossy" on all its parts, whence the name Mousseuse part out, ["mossy everywhere "]. This variety was produced again in 1864, at M. Ja- main's, horticulturist, Paris, where we followed the development of it. We also learned that at this place, in two beds planted with ordinary 164 BORROWED OPINIONS. moss-roses, beside the Zoe, there were several stems which tended likewise to modify them- selves, some in their leaves, others in their flowers. We must remark that it often happens that certain individuals of bud-varieties return, on some one of their parts, to the type from which they came. Thus, on a moss-rose from R. cen- tifolia, we have seen a branch of the ordinary Hundred-leaved rose. We should observe, how- ever, that most usually the parts which seem to return to the type present, notwithstanding, differ- ences from it. There has been a step in advance, and it is contrary to nature to retrace completely. The Rose du Roi, known by nearly every one, has produced the following six bud-varieties : — 1. Bernard Perpetual. This rose has the branches more slender than those of the parent; its flowers and leaves are also smaller. Its pom- pon flowers are very pretty, with a rose-color very much brighter than that of the Rose du Roi. 2. Long-peduncled Rose du Roi. This has branches much longer than those of the type ; the internodes are more distant, and the peduncles are also longer. It is only a sort of degeneration. 3. Madame Tellier. Very similar to the last, being distinguished only by its flowers, which are less colored, possessing a very bright rose flesh- color. CARRIERE ON ROSE SPORTS. 165 4. Mosrador. This rose differs from Rose du Roi by its stronger flowers, of a more vivid, deeper red ; its branches more colored, permitting it to be distinguished even in winter. Horticulturists do not like this variety, because it is hard to force, and because it passes very quickly to a dirty violet. 5. Capitaine Renard, or Variegated Rose du Roi. This variety differs from Madame Tellier by its flowers being variegated or ribboned with Avhite. It was found at Orleans by M. Desfosse- Thuillier. 6. Ccelina Dubos. Found by M. Dubos, hor- ticulturist at Pierrefitte, near Saint-Denis, upon Rose du Roi. It has the branches more slender and the leaves a little smaller than the parent ; its flowers, very similar in form to those of the type, are white, slightly flesh-colored. The Rose de la Reine has produced two sports : one, Belle Normancle, whose flowers, rose flesh- colored, recall those of Souvenir de la Malmaison; the other, Madame Cambel of Isly, or Triomphe de Valenciennes, which differs from the parent only in its marbled-variegated flowers. The Duchesse de Cambaceres rose, which has uni-colored, deep rose flowers, has produced by bud-variation Belle de Printemps, which has rose flowers marbled with brown. The Baronne Prevost has produced, to our 166 BORHOAVED OPINIONS. knowledge, five varieties, two of which have variegated flowers and one marbled. One of the two variegated varieties, Madame Desiree Girand, was found at the place of M. Desire Girand at Marly, near Valenciennes. It is not vigorous. The second variety, Panachee d'Orleans, which was observed for the first time at Orleans, is very vigorous. Its branches are more slender than those of Baronne Prevost, and the very smooth and shiny bark has few prickles. In short, its branches recall those of Cnisse de Nymphe. It sometimes happens that this variety produces large branches, vigorous and very thorny, but less so than those of Baronne Prevost ; its flowers also resemble the type more closely. It is an intermediate produced by the single matter of vegetation. The Baronne Prevost marbre dif- fers from the type only in its flowers, which are marbled with brown. Another variety, placed in the trade by M. Pierre Oger, horticulturist at Caen, differs from the type only in the color of the flowers, which are very much paler. The fifth sport produced by the Baronne Prevost is more recent. We observed it first in 1864, at Vitry-sur-Seine, in a garden under the care of M. Lachaume. We called it Madame Lachaume. It differs from the type by its branches being a little less thorny, but especially by its inflores- cence, which, long-paniculate, very much branched, CARRIERE ON ROSE SPORTS. 167 recalls that of certain Noisettes. The flower, also, is a little Aveaker than that of the type. But a very remarkable fact is that the hip, instead of being very regularly attenuated at its base and becoming confounded with the peduncle, as in Baronne Prevost, is abruptly and slightly in- flated, then contracted, and inflated again near the summit. The peduncles are also much more slender and longer than those of the parent. The Duchesse d' Orleans, whose flowers are violet-rose, produced by bud-variation, in 1858, a variety known as Soeur des Anges. This variety differs from its parent particularly in the color of the flowers, which is pale flesh-rose, like that of the flowers of Souvenir de la Malmaison. The rose called Quatre-Saisons has produced the following sports : — 1. White Moss, or de Thionville. This was first observed at Thionville about 18-35. It dif- fers from the type by its branches being more slender and supplied with hispid, glandular hairs. Its light green leaves are also softer to the touch and slightly tomentosc. Its flowers are pure white. Sometimes it produces strong branches which bear rose-colored flowers. In this latter condition it is the ordinary Quatre-Saisons, a fact observed by M. Duval of Montmorency, later by M. Victor Verdier, Paris, and recently (1864) at the Museum. 168 BORROWED OPINIONS. 2. Quatre-Saisons pompon. 3. White. The Provence roses have likewise produced a number of bud-varieties. Among the best known are : — Pompon Saint-FranQois. Pompon Saint-Jacques. Camaieu. Panache semi-double. Tricolore de Flandre. The last variety, which appeared in Belgium some years ago, is remarkable for its variegated floAvers ; it is a slender grower, although it comes from a very vigorous variety. It sometimes re- turns to the type. The variety Camaieu is remark- able for its striped flowers, very pretty, and al- most unique in the genus. Its wood is meagre and its leaflets are toothed. In the Damask roses, which are sorts of Quatre- Saisons roses, not remontants, we consider as bud- varieties the three following : — Damask York and Lancaster. Damask with blistered leaves. The ordinary Bengal rose has sported into the Bengale a bois strie [striped-stemmed Bengal]. The branches are often almost completely yellow. A very curious sport of the rose is the plant which we have called Rosier a feuilles de Chanvre BUD-VAUIATION IN THE UOSE. 169 [hemp-leaved rose] . By its flowers and especially by its leaves, this variety differs considerably from Rosa alba, from which it comes. Its leaflets are hooded, long and narrow, and very coarsely den- tate-serrate, sometimes as if gnawed on the edges, strongly nerved, of a dark green, rugose-scabrous. It happens sometimes, also, that its leaves are opposite upon certain branches. The flowers of this variety are smaller than those of Rosa alba, often irregular, and somewhat monstrous, and always sterile. [Probably no plants are so prolific of bud-varieties as the roses. Every gardener of experience has observed the fact. The follow- ing experiences of a single horticulturist (Ernest Walker, New Albany, Indiana), with one rose, illustrate this fact admirably. "I have had a number of sports of the Perle des Jardins rose," he writes me, " in our greenhouses. The first one was a double silvery pink with a short bud, and a very double, somewhat quartered flower. The stock of this I sold, as a new variety, for fifty dollars. The next sport was a white Perle. [The Perle is a golden-yellow rose.] I sold a plant of Perle to a local customer, who afterwards com- plained that it was not true to name, because the flower was white. She took it to be Cornelia Cook. I went to see the rose, and found a Perle rose in everything but color. I secured the plant, and was intending to introduce it, when, within 170 BORROWED OPINIONS. a few months, I heard that Nanz & Neuner, of Louisville, Kentucky, had one, and that a London firm had another ; and later I found that one had originated in Germany. Another sport of Perle was a single rose, like Isabella Sprunt. Another was like a Madame Falcot. At another time a whole branch sported into a form with a long, slender bud (about two inches long and five- eighths inch in diameter), with only two cal} T x lobes, and only two petals, — which were very broad, — in each cycle or series. This sport was really a monstrosity, and I could not propagate it.-] The so-called ornamental plants are not the only ones which present these examples of hetero- morphism. Fruit trees furnish very remarkable examples. We will cite some cases, beginning with those furnished by the cherry called Anglais hatif [Early English]. The most curious sport given us by this cherry is that which we call Cerisier Anglais heterophylle or a feuilles de saule [heterophyllous or willow-leaved English cherry]. This is the history of the sport: Upon a young tree whose parts are normal, we see, sometimes suddenly and without apparent cause, a vigorous bud develop, which bud, instead of producing leaves of the ordinary form, bears those which are very long and narrow, often somewhat falcate, and often irregularly erose. CARRIERS ON CHERRY SPORTS. 171 Grafted, this variety presents very singular pecu- liarities, as follows: so long as it preserves its exceptional characters the plant does not nourish, but as it constantly tends to lose them we observe that when the leaves have almost returned to the normal form the trees flourish and bear. Never- theless, this variety never resumes identically the characters of the type from which it came. Its aspect is always distinct. The tree is never fertile, and its fruit also differs from that of the Early English. The young shoots preserve their accidental character, and each year the leaves which it develops are nearly identical to those which the variety produced when it was first developed. This variety is not the only one which is pre- sented by the Early English. Thus, when the trees are old, it frequently happens that we find on the same individual three kinds of fruits, dis- tinct in their times of maturing. There is, first, the Early English, whose fruits become black; the Late English, whose fruits, of a beautiful deep red, shining as if varnished, ripen later. Finally, we nearly always find another variety, very late, whose fruits, a little smaller, are still entirely green when the other two have been gathered a long time. In these three sports, the differences are shown only on the fruits. The Indule cherry is also only a sport from the Early 172 BORROWED OPINIONS. English. It is distinguished by its foliage and earliness. The Early English cherry is not the only one which furnishes bud- varieties pertaining to the fruits. We find analogous examples in the May Duke, Cherry Duke, and Reine Hor- tense. These varieties, indeed, have produced on different branches of the same individual sub- varieties whose fruits ripened a fortnight later than normally. Grafted, each of these sub- varieties preserves its accidental character. A phenomenon analogous to the preceding ones is shown each year at the Museum upon an or- dinary double-flowered cherry. The tree upon which this anomaly was developed is nearly four- teen inches in diameter, is grafted on the Sainte- Lucie about twenty-seven inches above the ground. Above the junction the stem is naked for about six and a half feet. At this height is a large branch, which every year is covered with extremely double flowers, whilst the flowers of other branches, ex- panding very much later, are scarcely half double, and yield fruits. The Coe violette plum is an example of dichro- ism. It is a bud-variety which was produced on the White-fruited Coe, and which, grafted, is maintained with all the characters which it pre- sented at the time of its appearing. We have very often observed upon the Damas de Tours plum an instance almost the same as the preced- CARRIERE ON SPORTS OF FRUITS. 173 iner. On the same tree there were branches which bore fruits different in form and color, and differing a fortnight in time of maturing. Thus, while the fruits of the type are very large, length- ened, of a deep red color which recalls the Pond Seedling, marked only on one side by a very slight furrow, the fruits of the later sub-variety are a little smaller, and their form is that of the ordinary Reine Claude ; they are of an herba- ceous green, which passes more or less into a very clear red ; the stem, arched, swollen at the base, is inserted in a cavity quite large by the widening of the furrow, whilst the stem of the typical fruits is erect, little or not at all swollen, inserted in a very small cavity placed almost on the surface of the fruit. Another plum, the Prunier Puget, presents the following peculiarities : Upon the same branch it very frequently happens that there are fruits of a violet-red, dotted or striped with red-green. We find some, also, which present all the inter- mediate tints and others which are almost uni- colored. By multiplying them separately, there may be a chance to establish these varieties and to obtain several from one tree. We have seen on a red-fruited currant bush a branch which bore fruits as white as those of the Hollande a fruits blancs [White Dutch]. The fact of the nectarine coining suddenly from 174 BORROWED OPINIONS. a peach can no longer be doubted. Recent ex- amples have come to support the experiments of certain authors, notably Sieulle. Two other similar examples, of which we ought to speak, are furnished by two varieties of Chasse- las grapes, one known as Chasselas panache [Va- riegated Chasselas] and Chasselas Suisse [Swiss Chasselas] . Both appear to have come from a va- riety with black fruits, the color which predomi- nates in them. These are the peculiarities which they present : almost all the bunches bear some fruits more or less variegated or striped, white in Chasselas Suisse, red in Chasselas panache. But it happens frequently that the elements are separated and that we have then, upon different shoots, sometimes upon the same shoot, bunches of grapes of different color, almost entirely white if they belong to the Chasselas Suisse, and red if they belong to the Chasselas panache. One of the varieties is only a modification of the other, which is itself only a modification of some other. The pear Saint-Germain gris, whose deep gray fruits are very different in appearance from those of the ordinary Saint-Germain, is a bud-variety which was produced upon a branch of the latter, and which, multiplied by grafting, is maintained in all its characters. A similar variety was pro- duced on the Messire-Jean, so that at present we possess in the gardens a Messire-Jean gris, and CARRIERS ON SPORTS OF FRUITS. 175 a Messire-Jean jaune [gray and yellow Messire- Jean]. To these examples we will add two other analogous ones, which were recorded in the Bul- letin de VAcademie des Sciences, xxxiv., meeting of May 17th. One, given by M. Dureau de la Malle, refers to the Bon Chretien pear, which produced sometimes typical fruits and at others " of a form entirely different and unknown." The other example, cited by M. Mourriere, professor at Bernay, has reference to an apple which, on the same branches, produced fruits which had the appearance of a Reinette rousse and others which resembled a kind of Reinette du Canada. The latter is smooth, punctated, and often of a bright red upon one side. [The recent experiments of Waite, in this country, respecting the imme- diate influence of pollen, raise the question if some of these minor variations in form of the pear fruit may not have arisen from vagaries of pollination.] The various examples which we have cited are common to a very large number of plants, among which we will cite the banana and sugar-cane. Indeed, although these plants do not produce seeds, we find in each species a large number of varieties which are very distinct in vigor, aspect, habit, and in the banana in form, size, and quality of fruit. All these varieties are produced by bud-variation. These remarks can be applied to 176 BORROWED OPINIONS. other monocotyleclonous plants, as Arundo, Pha- laris, Bamboos, Dracaena, Yucca, etc. [Carriere cites the different shapes and colors of beans in the same pod as examples of bud- variation, but it is a question if these differences are not determined in the seed of the previous year. At all events, since there is only a single year in the life of the bean, we prefer to ascribe variations in it to the generation of the parents from which it has just sprung. There is no pre- vious year's growth of the same individual with which to compare variations and to ascertain if they are bud-departures from the type. Page 118.] 2. List of Bud-varieties. 1 After having sought to present certain examples of bud-variation which, by their importance, seem to be sufficient to fix the attention, we will con- tinue by the enumeration of a certain number of others, without, however, entering into details for each one of them. Sometimes we shall give only the name of the variety. If, however, they pre- sent particular interest, either in a practical or scientific point of view, we shall dwell upon them more at length, considering either their origins or peculiarities. [The garden names of the plants 1 The student should also consult Darwin's "Animals and Plants under Domestication," carriere's list of bud-varieties. 177 are given essentially as they stand in the original, for, as the purpose of this list is to acquaint hor- ticulturists with the nature and frequency of bud- variations, 1 have considered it unnecessaiy to make any particular attempt to revise the nomen- clature. The names are familiar, and, therefore, useful as they stand.] It would have been easy to extend this enumera- tion of examples of bud- variation. We have not thought it necessary because, aside from leading us too far, the real interest of the subject would gain nothing by it. We have, then, thought it our duty to put limits upon a subject which has no limits. Acer eriocarpum, fasciatum. Very remarkable for its much fasciated branches. This variety showed itself at the Museum in 1857 upon a seedling Avhich, during the first two years, presented nothing abnormal. In the third year, when the tree had been cut back, the sport appeared, since which time it has maintained itself Avith all its characters. This variety is to A. eriocarpum what the variety montrosa is to Sam- bucus nigra. Acorus gramineus, variegata. 1 1 When a name is not followed by remarks, the reader is to understand that it represents a known variety and that the name of the species indicates the origin of the variety. — L. H. B. 178 BORROWED OPINIONS. iEsculus rubicuncla, variegata. [iEsculus Hippocastanum, double - flowered. Upon a well-known Horse Chestnut tree in the environs of Geneva the owner, in 1822 or 1823, detected a single branch bearing double flowers. This still continues to bear double flowers and grafts from it do the same. It is thought to be the original of all the doubled-flowered Horse Chestnuts in the world. — A. De Candolle in Acad. Sei., Paris, 1875, quoted by Asa Gray, Sillimans Journ. 3dser. x. 238]. Agathsea amelloides, variegata. Ageratum Mexicanum, nanum. This plant, which is now used to so much ad- vantage for borders, is the product of a branch which developed accidentally from A. Mexicanum. Its heads are almost sessile and a little irregular, borne so close to the leaves as to make the plant undesirable from some points of view. The type plants, on the contrary, which are very much larger, have the heads large and regular and raised on long peduncles. Ageratum Mexicanum, intermedium. This variety, which is a bud-variety of sec- ond degree, that is, a sport from a sport (from the variety nanum), is intermediate. The plants are very floriferous. Their heads are also better carriere's list of bud- varieties. 179 than those of the type, and as they are borne upon longer peduncles, the plants are not only suitable for garden ornament but for cut flowers. The dimensions of this variety are intermediate between the last variety and the specific type. Ageratum Mexicamim, variegatum. This differs little from the type except by its leaves being variegated with, yellowish-white on the margins. Its inflorescence is, however, a little more slender and its heads are smaller. In gen- eral, the plant is " leggy," weak. Almond with variegated leaves. Leaves bordered and made satin-like with white ; vegetation delicate. It sometimes returns to the type. Anemone Japonica, Honorine Jobert. Very vigorous and very beautiful. This vari- ety, of which the flower is white, is a bud-variety from the so-called A. hybrida or A. elegans, which was obtained in England by M. Gordon by crossing A. Japonica with A. vitifolia. The va- riety was produced some years ago at M. Jobert 's, amateur at Verdun. Apricot with variegated leaves. 180 BORROWED OPINIONS. Aralia trifoliata, Cookii. This plant has its leaves, in general, simple, long, and narrow. Arunclo Donax, variegata argentea, and A. Donax, variegata aurea. These varieties differ from the type by the leaves being bordered with white in the first, and with yellow in the second. They are much more delicate than the type. Aspidistra elatior, variegata. Aster bicolor. This plant, which we believe not to be a dis- tinct species but simply a dwarf form, very prob- ably a bud-variety of A. versicolor, produced at the Museum in 1856, upon one of its stems, a vigorous bud which presents all the characters of A. versicolor except that it is a little smaller. This variety, to which we have given the vari- etal name Major, has preserved all its characters Avhen multiplied by root-cuttings, and to-day is still one of the most beautiful perennial plants. Azalea Indica, Dieudonne Spae. Flowers salmon, margined with white. It is a sport from A. formosa, Ivery, which has rose flowers. carriere's list of bud-varieties. 181 Azalea Indica, Beaute de l'Europe. This variety has flowers white at the base, variegated with red. It is a sport from A. deli- cata, which has deep salmon flowers. Azalea Indica, Criterion. Flowers deep rose bordered with white. This is a sport from A. Iveriana, which has flowers white, striped with rose. Azalea Indica, alba rosea. Flowers rose, slightly bordered with white. A bnd- variety from A. Iveriana. Azalea Indica, exquisita grandiflora. Flowers deep rose bordered with white. It is a bnd- variety from A. alba perfecta, which has flowers white, very lightly striped with rose. Buxns Balearica, cncnllata. This bud- variety of B. Balearica differs from its parent by its smaller leaves, which are very strongly convex, and rounded in the middle. Buxus sempervirens, argentea. Buxus sempervirens, aurea. Buxus sempervirens, marginata. All these varieties are distinguished from the type by their leaves being variegated or bordered with either white or yellow. 182 BORROWED OPINIONS. Camellia Japonica, Comte de Paris. This variety, which has strongly striped rose- flesh-colored flowers, is a sport from the Duchesse d'Orleans, which bears white striped flowers. This variety is not only much more vigorous than its parent, but it has the merit of fully expanding its flowers, while the buds of the Duchesse d' Orleans almost always fall before opening. Camellia Japonica, Montironi rosea. This plant, whose flowers are entirely rose, is a sport from the Montironi, which has white, very lightly striped flowers. Camellia Japonica, Giardino Franchetti. Flowers deep rose, bordered with white. It is an offshoot from C. Targioni, which has white flowers lightly striped with rose. Camellia Japonica, Comtesse Woronzoff. This variet}^, which has delicate rose flowers, is a sport from C. centifolia alba, whose flowers are pure white. Camellia Japonica, Giardino Schmitz. Flowers delicate rose-color. It is a bud-variety from the Elisa Centurion, which bears very lightly rose-striped white flowers. carrieke\s list of bud-varieties. 183 Camellia Japonica, Imperatrice Eugenie. Flowers rose-flesh-colored. A bud-variety from Montironi, whose flowers are very striped with rose. Camellia Japonica, Paolina Armari. Flowers deep rose. Bud-variety from INIiss Abby Wilder, which has white, lightly rose-striped flowers. Camellia Japonica, Princesse Aldrovandi. Flowers rose, bordered with white. Sport from Teutonia, which bears flowers white, rose-striped. Camellia Japonica, Bicolor de la Reine. Flowers rose, bordered with white. It is a sport from de la Reine, whose flowers are white, lightly striped with rose. Last year we saw upon a camellia with rose flowers, some branches bearing flowers completely white. Cephalotaxus pedunculata, fastigiata. This variety, which has been described and figured as being a species of Podocarpus (P. Koraiana), is an example of bud-variation. We had proof of this statement at the Museum in 1863. Having made cuttings from a certain num- ber of branches of the so-called Podocarpus, one 184 BORROWED OPINIONS. of them, instead of producing simple and scat- tered, strictly erect brandies bearing scattered leaves, produced wliorled, horizontal branches bearing distichous leaves. The variety fastigiata is to C. pedunculata what Taxus baccata, fastigi- ata, is to T. baccata. Cereus Peruvianus, monstrosus. Sometimes returns to the type. Clematis bicolor or Sieboldii. This plant, of which the flowers, violet inside, are almost double from the transformations of the stamens, is a sport from C. florida, which has single greenish- white flowers. We have several times had occasion to ascertain that such is the origin of this clematis. The variety known as C. bicolor, flore pleno, which we sometimes call Atragene Americana, so remarkable for its enormous greenish-white flowers, is a direct sport from C. bicolor, conse- quently a bud- variety of the second degree from C. florida, a fact which Ave have been able to verify again this year. On a plant of C. bicolor, planted in the open air, there is developed, almost from the base, a branch which bore flowers en- tirely full, monstrous, yellowish-green, so that the two bud- varieties — C. bicolor and its variety flore pleno — were united upon the same individual. caeiiiere's list op bud-varieties. 185 Clematis, Helena monstrosa. This plant is none other than C. Helena which, by bud- variation, is transformed and has become double flowered. This example is analogous to that which is produced by C. bicolor. Cheiranthus Cheiri, variegata flore pleno. Sport from the double-flowered yellow gilly- flower of the walls. Cherry. See page 170. Cornus sanguinea, variegata. Cornus Mas, variegata. Cytisus Adami. Whatever may be the origin of this plant, whether a hybrid, as is generally believed, or a peculiar form, we propose here to say nothing concerning it beyond a verification of its peculiar- ities. It develops very frequently and normally, so to speak, some branches of C. Laburnum and others belonging to C. purpureus. When we graft separately these two kinds of branches, these species remain invariable, although the grafts were taken from C. Adami. Dactylis glomerata, variegata. 186 BORROWED OPINIONS. Echinocactus multiplex, cristata. This variety, instead of having a regular, lengthened, melon-like stem, forms a thick mass which extends itself into little fan -shaped bunches, and instead of longitudinal furrows, large and deep, and separated by protuberances upon which are borne long, very rigid spines (about three - fourths to one and two -fifths inches) ; the variety has only very slight fur- rows or kind of folds disposed transversely to the direction of the fasciation, consequently in a contrary direction to those which are presented by the type, and upon the borders are spurs (about two-fifths inch long) disposed in stars. In a word, the variety is entirely different from its parent. Elseagnus reflexa, variegata argentea and varie- gata aurea. These two varieties differ from the type in having their leaves bordered, respectively, with white or yellow. Elseagnus pungens, variegata. Euonymus Japonica, argentea and aurea. These, especially aurea, return sometimes to the type. caerieke's list of bud-varieties. 187 Euonymus Japonica, flavida. This plant, which developed upon a type plant with green leaves in 1862, is distinguished by its leaves being bordered with yellowish-green, some- times with nearly white. It is vigorous. Euonymus Japonica, fasciata. Very remarkable for its much fasciated branches. This variety appeared at the Museum in 1864, upon a typical E. Japonica. Euonymus Japonica, calamistrata. This sprung from the variety argentea, from which it differs in its more slender parts. Its leaves are smaller and crisped as if erose. E. Japonica has also produced many other bud- varieties, which differ in variegation, or sometimes even by the form of the leaves. It is more than probable that the various varieties which have been introduced recently from Japan are bud- varieties. Fagus sylvatica, fern-leaved. This variety once presented the following peculiarity : Having grafted it upon the com- mon beech, the branches developed from each side of the stem almost distichously. All those upon one side bore leaves similar to those of the common beech, whilst those upon the other side bore only laciniated leaves. 188 BORROWED OPINIONS. Ficus scandens, microphylla. This variety, which we sometimes meet in com- merce under the name of F. buxifolia, is a bud- variety which appeared in 1856 at the Botanic Garden of Orleans upon a plant of F, scandens grown in a greenhouse. Its leaves are very small, someAvhat suborbicular and marked with brown. This variety is preserved in all its characters, both upon the original plant and in all the multi- plications which have been made of it. Fontanesia phyllireoides, variegata. This very pretty variety appeared at the Museum in 1854. Since its appearing, this vari- ety has not varied. Its branches, of a yellowish- green, are slender, and the leaves are deeply bordered with yellowish- white. Fraxinus Americana, variegata. Fraxinus excelsior, jaspidea. This variety is distinguished by its bark being striped or slightly ribboned with yellow. Fraxinus excelsior, variegata. The common ash has produced several sports Avhich are marked by the variegation of their leaves. This one has yellow and white disposecV in bands and bordering the leaves, or sometimes caiirieke's list of bud- varieties. 189 in spots upon all parts of the blade, as upon the leaves of Aucuba, for example ; hence the various names, argentea, aurea, striata, maculata, aucubaj- folia, etc. Gardenia radicans, variegata. In this instance the variation is two-fold. The leaves are bordered, and are also much narrower than in the type. Gillyflower, called Savoyarde, with variegated leaves. This is a sport from the double-flowered brown gillyflower. Grape. See Vine. Hedera. Variegated tree-ivy. This sub- variety is a bud- variety from the so- called tree-ivy (lierre en arbre,) from which it varies only in the yellowish-white variegations of the leaves. What we call tree-ivy is a common ivy, or one of its varieties, arrived at the full-grown state and which then fruits. The branches are large, short, cylindrical, and destitute of climbing roots. The leaves, instead of being lobed, are heart-shaped, more or less lengthened, sometimes very obtusely rounded. As there are several forms of creeping ivy, so there are several sub-varieties of the tree- 190 BORROWED OPINIONS. ivy. They partake of the character of the varieties from which they come and are distinguished by the form and dimensions of the leaves, by the size of the branches, these characters all depending upon the vigor and appearance of the mother varieties. We obtain the tree-ivy either by cuttings or by grafting from adult branches, that is, branches which have been modified by fructification. They then branch and form very pretty bushes. Some- times, especially near the ground or in badly aired places, branches arise supplied with climbing roots, bearing leaves more or less lobed, and which creep and take root as soon as they touch the ground. Here, in the case of the tree-ivy, is an instance of bud-variation due to the maturity of the individual. Hibiscus Syriacus, flore pleno variegata. This variety, whose leaves are variegated with yellowish-white, appeared in 1858 upon a plant with entirely green leaveSc Hibiscus Syriacus, variegata. Remarkable in the variegation of its leaves. Its flowers are similar to those of the last. It is a va- riation directly from the type. It is not vigorous. Hyacinth. The double blue or Globe terrestre is a bud- variety from the double white or Sultan Achmet. CARRIERE'S LIST OF BUD-VAKI E TIES. 191 The double white with blue eye, or Sphaera Muudi, is a sport from the double white. The single red, called Acteur, cultivated for a very long time without varying, has produced by bud-variation at Hemstede, near Haarlem, a variety with double, imbricated red flowers. The hyacinth Ami du Cceur, with single blue flowers, also long culti- vated without varying, lias produced, from the same bulb, two flower stalks, one of which bore flowers dregs-of-wine color, while the other bore flowers of a delicate flesh-colored rose. Hydrangea Hortensia. This sterile plant is a sport from the form called H. Japonica, analogous to those which are pro- duced upon Viburnum Keteleerii and V. Opulns. H} T drangea Japonica, variegata. Differs from the type only by its leaves being bordered with white. Ilex Aquifolium, calamistrata variegata. This variety is a sport from the I. calamis- trata, which is a sport from the common holly. Ilex Aquifolium, ferox aurea, and ferox argentea. Bud-varieties from the variety ferox, from which they differ in the variegation of the leaves, yellow in the first, white in the second. 192 BORROWED OPINIONS. The very numerous varieties of common holly in cultivation are for the most part fixed bud- varieties. Iris spectabilis. This plant, so remarkable for its color, is a bud- variety from Iris Xiphium, from which, however, it is very different. Juniperus communis, variegata. Juniperus excelsa, variegata. Juniperus Virginiana, variegata. Juniperus Virginiana, monstrosa. This variety, which arises from knaurs or burrs, is shown frequently upon the Virginian juniper (red cedar). Lamium album, variegatum. Laurocerasus vulgaris, angustifolia. This plant, which for a long time has gone under the name of Hartogia Caj)ensis, is a bud-variety, a fact which we have been able to ascertain several times. Its leaves are very straight, long, of a clear green, and more strongly toothed than those of the plant from which it comes. It is very con- stant. We have a record of its variation. carriere's list of bud-varieties. 193 Laurocerasus vulgaris, variegata, Laurocerasus Lusitanica, variegata. Ligustrum Japonicum, variegatum. The L. Japonicum appears to be subject to bud- variations, especially in the direction of dichroism. We have already produced from it several varie- ties distinct in the color or disposition of the variegations, the varieties receiving names accord- ing to their character. There is one which differs somewhat in the form of its leaves. Ligustrum ovalifolium, aureum. This variety, which is distinguished by having its leaves bordered or ribboned with yellow, was produced at the Museum in 1861. It comes from a branch which was developed spontaneously upon a type plant. It is unstable. Ligustrum vulgare, variegatum. This variety has leaves variegated with yellow. It occurs quite frequently in the wild state. We have found it several times in the woods. It is not stable. Lilac, common, variegated. Lilac, Persian, laciniate-leaved and white-flowered. Although we are not able to state precisely when these two sports appeared, we have no doubt that 194 BORROWED OPINIONS. they are bud-varieties, as the Persian lilac never gives seeds. The origin of the Persian lilac itself, even, is in much doubt. Mamillaria nivea, dedalea. This variety forms a compact mass whose folds and circumvolutions are disposed in a sort of laby- rinth (whence dedaleci), giving it a little the ap- pearance of calves' pluck. The type from which this variety comes forms a melon-formed cylinder which is slightly swollen at the summit. It bears spines disposed in bundles and from about four- fifths inch to one inch and a fifth in length, rigid, very sharp, surrounded at the base by a series of smaller ones disposed in the form of a star. The variety, on the other hand, aside from its peculiar form, has no spines. It is invested upon all its parts with silky hairs, silvery and as soft as felt to the touch. The parent and offspring have nothing in common in their general form. Mentha rotundifolia, variegata. Molinia coerulea. variegata. Musa paradisiaca, vittata. This is distinguished from the type by the white bands upon its leaves. The variegation is con- spicuous upon the yellow plants. It often disap- pears with time, so that in the old plants we do not often find any trace of it. caiumeue\s list of bud- varieties. 195 Myrtle, common, variegated-leaved. Frequently returns to the type. Opuntia cylindrica, cristata. In exterior characters this plant has nothing in common with the type, which forms a regular, cylindrical column. The variety, on the contrary, is made up of enlarged pieces placed against each other in different ways, much the same as those presented by the various species of opuntia which we call "Semelles du Pape." Orange tree, Turkish. This variety, which is a sport from a kind of Seville orange (probably from the Horned Seville orange), bears at times upon its various branches leaves narrow and irregular (as erose), variegated or rather satin-bordered, white, and, upon other branches, green leaves, large and strongly eared, as well as fruits which recall those of the Horned Seville orange. Orontium Japonicum, variegatum. Osmanthus Fortunei, 1 ovata. This variety is unstable. After having pre- served it for more than a year without varying, it has resumed in large part its primitive character, which is to have leaves long, exceedingly thorny 1 The Olea ilicifolia of commerce. — Carriere. 196 BORROWED OPINIONS. and strongly nerved. Sometimes we find branches bearing leaves of different forms. Osmanthns Aquifolius, variegatus. Differs from the type by the yellowish- white variegations of the leaves. O. Aquifolius, which we can consider as the representative in Japan of our common holly, appears, like the holly, to be very subject to bud- variation. We have no doubt that the varieties recently introduced from Japan originated in this manner. Peach, carnation-flowered (Persica dianthiflora), and many-colored (P. versicolor). These two varieties are sportive forms of P. rosseflora, of which the flowers are very deep red. Like this, the two varieties have double flowers, but of very different colors from those of their parent. The carnation-flowered has flowers of a flesh-colored rose. The many-colored, on the contrary, has white flowers striped or rib- boned with brilliant rose. This last is very much more delicate than P. rosseflora or the carnation- flowered. Peach, willow-leaved red Madeleine. This variety, remarkable for the form of its leaves, which are very long and narrow, plane, carriebb's list of bud-vaiuettes. 197 glossy, very shortly toothed, is the result of a bud-variation from the variety designated by cer- tain horticulturists as Madeleine de Courson (red Madeleine). It appears to us to have great re- semblance to that very anciently known as the willow-leaved. Peach, laciniate-leaved red Madeleine. Leaves very strongly and coarsely toothed or laciniated. Pears, variegated fruited. The following pears have given by bud- varia- tion variegated varieties : Duchesse d'Angouleme, Amanlis, Guenette or Madeleine, Saint-Germain, Bergamotte d'automne, Culotte de Suisse, etc. These bud-varieties are further remarkable in that the variegations extend to the branches and fruit, but not to the leaves, a character which distin- guishes them from the next variety, which is like- wise a bud- variety. Pear, Amanlis, with } T ellow bark and leaves. A very remarkable variety. We could almost say that it is pretty. It was developed upon a stem of Amanlis which presented nothing abnormal. It is very vigorous, and produces a most singular effect, with its parts all yellow except the bark, which is grayish-white. It has not yet fruited. 198 BORROWED OPINIONS. [One of the most marked cases of bud- variation which ever came under my notice was observed a few years ago upon a tree of Onondaga pear. One branch, so placed as to remove all possibility of its being a root-sprout or a graft, bore about a dozen pears which were intensely and uniformly russeted. They were so different in appearance from the pears upon the remainder of the tree that no one would suppose for a moment that they were the same variety. Even the Sheldon does not differ more widely from the Onondaga in appear- ance than did this singular sport.] See page 174 for further notes on the pear. Pelargonium zonale, Manglesii. Distinguished from the species by its white variegated leaves, which are more deeply lobed, and by the weaker branches. It has, in its turn, produced several varieties by bud- variation. The bud-varieties produced by P. zonale and P. inquinans (which are in reality only one) are very numerous. There are among them some varieties of such pronounced characters that, if we ignored their origin, we might consider them species. Pelargonium hederaefolium, variegatum. Phalaris arundinacea, picta and aurea. These two varieties differ from the species in the variegation, which is produced by white in cabbiere's list of bud- varieties. 199 the first case and yellow in the second. They are exactly representative of the phenomenon which is observed in Arundo Donax, as well as in the sugar-cane. Phlox decnssata, white Croix de Saint-Louis. This variety, of which the flowers are entirely white, appeared upon the variety Croix de Saint- Louis in 1863. The parent variety has white- striped rose flowers, and cross-form, whence its name. Phragmites vulgaris, variegata. Leaves bordered or margined with white. Picea excelsa, tabulaaformis. This variety, which attains a height of scarcely more than a foot or two, and which spreads out horizontally so as to form a sort of carpet, is a most remarkable variation which resulted from a knaur or burr upon the stem of a very large spruce. It was produced in the park of Trianon at Versailles. Pinus sylvestris, nana monstrosa. Produced from a knaur from the stem of a large pine. It is dwarf and monstrous, remarka- ble for its long, unequal, crowded leaves, and by its slender, sometimes almost filiform, irreg- ular branches, which are produced in such quan- 200 BORROWED OPINIONS. tity that they sometimes completely conceal the branches and even the trunk. Pinus sylvestris, nana compacta. This variety is also the result of a knaur from a large pine. It attains but a foot or two in height. Its very short and numerous branches have already borne two crops of cones, some nearly ripe, small, though well formed, others much younger, still herbaceous. Pittosporum Tobira, variegatum. Populus Greeca, 1 pendula. We cannot say whence this variety came nor how it was obtained. It has been in cultivation very long. We have a singular experience to record concerning it. In 1858 we grafted fifteen plants of P. nivea with P. Grseca, and, seven of the grafts growing, there was one which produced slender and drooping branches just like the variety pendula of P. Graeca, of commerce. This phe- nomenon is one of the most curious which we know. The tree is planted in the nursery of the Museum at the side of one of its brothers, to which, physically, it bears almost no resemblance, although coming from the same parent plant. 1 Undoubtedly our native large-toothed aspen, Populus grandidentata. — L. II. B. carriere\s list of bud-varieties. 201 Both are pistillate and are covered each year with catkins. Potato. Potatoes furnish many examples of bud-va- riation ["mixing in the hill"]. Many of our cultivated varieties are bud -varieties from the subterranean parts. Every year at digging time, if we wish to keep the varieties true, we are obliged to throw out those which, we say, are "degenerated." This so-called degeneracy << in- stantly tends to remove the products from the starting-point, and has, then, the result of pro- ducing new varieties. Modifications in potatoes may also occur in the manner of vegetation or growth of the under- ground parts. Such is the case in the variety called Pouse-debout [" tubers standing on end "]. This name was given the variety because the tubers, instead of lying horizontally, or nearly so, are placed upright, one against the other, much as small pieces of wood are arranged for the making of charcoal. The Marjolin we consider nothing else than a peculiarity of vegetation. This is proved by the fact that its characters — not blossoming and maturing veiy early — are not constant. It has produced two other varieties by modifications of its underground parts. One variety is the Mar- 202 BORROWED OPINIONS. jolin tardive [Late Marjolin], called also Marjolin de deuxieme Saison, which is sometimes sold in the Paris markets for the Hollande jamie [Yellow Holland]. It is remarkable for the period of its growth, which is more prolonged than that of the type, and it is also covered each year with flowers, while its parent scarcely ever blossoms. The other variety has no resemblance to the Marjolin in form. It is round, and its sunken eyes give it exactly the appearance of the ordinary yellow potato. When we cultivated the Marjolin there was not a year when we did not obtain round ones, although we had planted long ones very true in appearance. A very remarkable example of the modifications furnished by the ordinary yellow potato is the following : In a square planted exclusively with this variety, very true in appearance, Ave gathered a certain number of which the skin was more or less dark ; some had yellow flesh, others white. Planted separately, these bud-varieties have given us potatoes round in form like the parent type, but among which there were found some entirely violet in both exterior and interior, and some had black flesh slightly marbled with white. This modification of color was not the only change. In some cases the quality was very much modified. Thus, instead of being nearly like the yellow potato, the flesh of these varieties was compact, neither good nor bad. carriere's list of bud-varieties. 203 We give two other examples of bud- variation in potatoes, observed by us at the Museum in 18(34 : Half of a plat was planted with the smooth long yellow called Hollande, and half with the regular long red commonly called Vitelotte lisse. The first half yielded tubers similar to those which we had planted. The second half, on the contrary, produced tubers differing from the pa- rent in color, being of a reddish-yellow, although the form remained about the same. The quality, also, did not vary, so that while we confounded them sometimes with the Hollande, we were able to distinguish them readily when cooked, as they remained whole, while the others fell to pieces. On the end of a plat where we had planted fifty of the ordinary round yellow potato, one plant grew until late in the season and gave round potatoes of a deep red. In this same year, 1864, in a square planted entirely to Chardon potato, we observed some plants exactly similar to the others in growth and appearance, but which differed entirely in the color of the flowers, being dull white, a little sulphur-colored, while those of Chardon are violet- rose or rose-violet. The tubers from these white- flowered plants differed from those of the type in being more round and regular and having less pronounced eyes. Aside from these variations, 204 BORROWED OPINIONS. we have found among the Chardon both earlier and later varieties, and this in spite of the fact that we had planted only such tubers as appeared to be entirely true and which for a long time had produced no variations whatever. Here, as in the preceding cases, the modifications were from the tubers, seed not having been soavii. An instance similar to the above is reported by M. Joigneaux in the Journal de la Ferme et des Maisons de Campagne : Nine or ten years ago six beautiful tubers of a long, pale yellow potato were given us. In order to increase the number of hills we divided each tuber into three pieces. We planted them ourselves. The cultivating was also done by us. Some of the potatoes, a very small number, resembled the type, but the larger num- ber were spherical, some yellow like the parent, others deep red. All cultivators know that the smooth or even Vitelottes, whose eyes are few in number and scarcely perceptible, often produce tubers of vari- ous forms and with eyes so much sunken that it is almost impossible to peel them. Once we ob- tained a variety which, besides the many and deep eyes, produced, in considerable quantity, agglom- erations which gave to the whole a monstrous form. They were veritable hydras. Although coming from the Vitelotte, which is a good potato, this variety was very acrid and bad. carriere's list of bud-varieties. 205 All these examples show without doubt how a part of the varieties of potatoes are pro- duced, and proves that they do not all come from seeds. We may convince ourselves of it when, having observed the growth of the plants, we mark the peculiar plants and gather their tubers separately. See, also, page 209. The phenomena presented by potatoes prove that the cause of the appearing of new varieties is not always due, as we generally suppose, to crossing, as fecundation can act only upon the seeds. It is also very rarely that we practise crossing in potatoes, but we can number the varie- ties by the hundred. But it often happens that cuttings made from portions of the top of the plant produce varieties different from the parent. More- over, the existence of numerous varieties of certain plants which we cultivate and which never pro- duce seeds, proves beyond a doubt that there are causes aside from crossing which tend to the pro- duction of new varieties. Primus Mahaleb, variegata. Aside from this variety, which is very pretty with its long and very slender branches and white-variegated leaves, P. Mahaleb lias by bud- variation given several sub- varieties which are distinguished by the form of the leaves, and especially by the color of their variegations. 206 BORROWED OPINIONS. Almost all these varieties are more delicate than the type. Rheum australe, variegatum. Remarkable for the beautiful white variegation of the leaves. Ribes nigrum, variegatum. Leaves variegated with yellowish- white. Ribes rubrum, variegatum. Robinia hispida, arborea and macrophylla. R. hispida, var. arborea of the gardens (R. macrophylla DC), differs from the type by its greater vigor, its branches very much larger, the bark dark, glossy, and smooth, and by the thicker coriaceous leaves, which are glossy as if varnished. R. hispida, var. macrophylla of the gardens, is nearer the type than the last; it differs from it, however, by its greater vigor, and espe- cially in its flowers, which, less abundant and a little more developed, are paler in color. Like the type, of which even the origin is doubtful, these varieties do not produce seeds. The fact of the sportive production of the variet} r biennis upon R. hispida is wholly be- yond doubt. Several times Ave have found the tAvo sorts of twigs growing side by side upon the same branch. It is only necessary to multiply carriere's list of bud-varieties. 207 them separately in order to obtain distinct varie- ties. 1 Robinia Pseudacacia, umbraculifera. This plant, now so commonly used either for ornament, under the common name of Acacia boule, or as a dwarf shrub and considered as a forage plant, and called in consequence Acacia a faucher, comes, according to Turpin, from a knaur which appeared on the stem of a R. Pseudacacia. This fact does not surprise us. It shows us Iioav important these singular excrescences may become. Rose Eglanteria, punicea. This differs from the ordinary yellow-flowered capucine rose (R. Eglanteria) in the color of the flowers, which is an orange-red. In many soils this variety returns more or less quickly to the type. It frequently happens that we may see even upon the same stem a red and a yellow flower. Sometimes we find a flower winch pre- sents these colors separately on opposite sides and of nearly equal extent, or some petals may be half red and half yelloAV. In general, the sport is less vigorous than the type, so that under the influ- 1 Plants of normally thorny species are often found which bear no thorns. There is a so-called variety inermis of the honey-locust, Gleditschia triacanthos. Wild blackberries with smooth canes are occasionally found. In fact, most prickly or thorn-bearing plants vary much in these characters. — L. H. B. 208 BORROWED OPINIONS. ence of a slow modification Ave sometimes see the color gradually disappear, and at the end of a certain time we have a rose with completely yel- low flowers in the place where we planted one with orange-red flowers. For further notes on the rose, see pages 161 to 170. Salix Babylonica, annularis. Very remarkable for the form of its leaves. For a long time we have noticed it showing itself each year upon an old tree. The parts upon which it appeared, being generally weak, pro- duced, instead of long, linear, plane leaves, those which were rolled up on the edges and distorted into rings. It is very constant. We have no examples of its reverting. It is much less vigor- ous than the type. Sambucus nigra, variegata aurea and variegata argentea. These two varieties differ from the species in the variegation, which is made by yellow in the first, white in the second. The second is much less vigorous than the type. Sambucus nigra, monstrosa. Analogous to Euonymus Japonica, f asciata. The flowers are also monstrous, and, up to this time, the seeds which they have produced have always been poor. CARRIERE'S LIST OF BUD- VARIETIES. 209 Solanum Dulcamara* variegatum. Solanum tuberosum, variegatum. Remarkable for its yellow- variegated stems and leaves. During the year previous to the appear- ing of this variety, its parent type presented no unusual characters. See Potato, page 201. Spiraea Ulmaria, variegata. Symphoricarpus vulgaris, variegata. Symphytum officinale, variegatum. Thuyopsis dolibrata, variegata. This variety, of which the leaves are variegated with white, is remarkable for its vigor and its facility for forming heads when multiplied by cuttings. Ulmus campestris, variegata, argentea, aurea picta, etc. Variegated varieties of the common elm are numerous. They are distinguished both by the color and form of the variegations. Viburnum Opulus, sterilis or Boule de neige. Viburnum Opulus, sterilis variegatum. Viburnum Keteleerii, macrocephalum. Analogous to the sterile varieties of V. Opulus. 210 BORROWED OPINIONS. Viburnum Tinus, variegatum. Viola Rothomagensis, pallida. This variety, of which the two superior petals are pale lilac and spotted, while the three others are yellowish-white and lightly striped, was pro- duced by bud- variation at the Museum. We have said elsewhere that the phenomena of bud-variation could be divided into two classes, one group including those variations which appear abruptly, the other those which take place slowly. This violet falls under the latter group. In 1863 we received from the hills of Vernon a certain number of plants of V. Rothomagensis. Planted at the Museum, they preserved nearly all their characters except the villosity, which in large part disappeared the first year. During this year 1863 and the entire year 1864, they produced blue flowers abundantly. In the winter of 1864-65 all the plants perished except one. The remaining plant, instead of being covered with beautiful blue flowers as it had been the two preceding years, produced flowers almost white. 1 Vine (Grape). Bud- variation is comparatively common in the vine. [Frequent cases occur in the American 1 This variation appears, after all, to have been an abrupt one. — L. H. B. CARRIERE ON GRAPE SIM HITS. 211 grapes.] It is well understood in this case, as the vine is one of the oldest of cultivated plants and it is multiplied almost always by cuttings; and as cuttings are made by millions each year a bud-variety soon becomes widely disseminated. It frequently happens that a shoot Avill produce grapes differing in form or color from those which are borne upon other shoots of the same vine. We may add that these variations nearly always pre- sent peculiar qualities also. We will cite ex- amples. Upon a plant of black-fruited Muscat grape we have observed for several years that some shoots produce white grapes. The white seedless Corinth is a bud-variety from a variety which has much larger fruits, with seeds. This is a fact which we have several times observed upon bunches where some fruits were unusually developed and which contained seeds. The white Cornith is analogous to the Chasselas de Demoiselles. A proprietor of large vineyards in the middle of France, the late Cazalis Allut, wrote some years ago as f oIIoavs : — " A stock of Teret produced with me, for sev- eral years, black grapes upon shoots of two of its arms, and gray grapes upon shoots of the other arms. A stock of Epiran gris, trained in cordon, is now about forty feet long. The first twenty 212 BORROWED OPINIONS. feet produces constantly gray grapes and the re- mainder produces white ones. I have in an enclosure a stock of Epiran noir having several arms. The shoots of one of the arms give grapes almost twice as large as those on other parts of the vine." Another viticulturist, M. Henri Bouschet of Montpellier, wrote very recently: — " I had occasion for several years to see in my collection at Lot-et-Garonne, a plant of Prunella gris, which, sometimes upon one stem, sometimes upon two, bore black grapes, while the remainder of the vine bore gray ones. I have noticed for two years in my collection at Calmelte a most curious fact upon three grafts of a Spanish va- riety which came to me from the collection of Luxembourg, where it is called Parrel del Reyno de Lorca and which I have recognized as our Morastel noir. One of these three stocks has borne on one side, to my great surprise, black grapes similar to those of the Morastel, and upon the other side, constantly, white bunches having an appearance very different from an ordinary white Morastel, and presenting a foliage very different in size and form. This odd foliage ap- pears to me to be identical with that of the Oyo de Rey de Morada, of which the bright yellow- ish-green leaves present very shallow rounded dentate lobes, while the leaves of the Morastel are CARRIE 1 J H ON GRAPE SPORTS. 213 deep green with deep divisions, the lobes acute, with teeth detached and terminating in a point." A passage which we find in the Parfait Vigneron (edition of 1811) seems to confirm the opinions which we give concerning bud- variations in grapes : — " A citizen of Vilmorin has observed a stock of Meunier to bear, upon some shoots, leaves and fruits of Maurillon precoce. A citizen of Ju- milhac has seen likewise the Meunier become Maurillon." Therefore the grape called Madeleine Juillet, Maurillon hatif, etc., is only a sport from Meunier, a fact which shows, as we have said before, that the varieties produced by bud-variation may pre- sent qualities different from those presented by the plants from which they come. Upon a plant of Pinot gris there appeared at the Museum in 1863 a shoot whose leaves were much variegated or striped with yellow. It produced a grape very similar to the variety from which it came. It appeared to be much less fer- tile, however. In 1863 we observed two other very remarkable examples. One example concerns the Precoce Malingre, the other the variety designated by the name of Verjus. These examples present contrary results. Thus, while Precoce Malingre has long, oval, scattered fruits, and the bud- variety 214 BORROWED OPINIONS. which appeared upon it had round fruits borne close together and larger than those of the type, the Verjus has slightly oblong or nearly spherical fruits and the bud-variety which was developed upon it had fruits long-oval and attenuated at both ends, and somewhat later than those of the type. The Chasselas gros Coulard is a bud-variety which appears frequently upon the ordinary Chasselas. Its fruits are large and spherical. They often drop. It differs especially from the Chasselas by its stronger shoots with joints much closer together, and by its leaves being less lobed, a little longer and thicker, of a glossy green as if varnished. It differs also from the ordinary Chasselas in its temperament. It needs much heat and also shelter from the influence of the air. It generally succeeds well in forced culture. The Chasselas de Demoiselles, remarkable for its fruits, which are scarcely larger than shot, is a bud-variety from the ordinary Chasselas. This phenomenon appears to be due to the partial abor- tion of the sexual organs and particularly of the anthers, whence results the lack of impregnation of the flowers and the consequent abortion of seeds. Propagated by cuttings, it preserves its characters. A variety with variegated leaves has appeared from the ordinary Chasselas. FOCKE ON THE PRIMARY CROSS. 215 Wigandia Caracassana, variegata. Distinct by its leaves and even its branches being variegated with white. The variety ap- peared in 1862 upon a plant which, placed in the open air at the beginning of the season, presented no unusual characters. III. Focke's Discussion of the Character- istics of Crosses. (Translation of Chapter IV. of Die Pflanzen-Mischlinge. 1 ) There is no absolute distinction between plants of a pure and those of a .hybrid or mixed descent; there are therefore no signs by which one can, under all circumstances and with certainty, rec- ognize the mongrel nature of a certain plant. Hybrids, nevertheless, often sIioav a series of characteristics which indicate with greater or less accuracy their mixed descent. Certain rules can be made in regard to them, of which none, of course, is without exception. I. THE SIMPLE PRIMARY CROSS (AxB).2 1. All individuals formed by the crossing of tivo pure species or races are, if they have been iFocke uses the word mischling (derived from mischen, " to mix ") in about the same sense in which we use the noun cross ■ i.e. it is a generic term for both cross-breeds and hybrids. 2 In these formulas, the letters are used to designate the parents of any cross. In common with usage amongst botanists, 216 BORROWED OPINIONS. produced and grown under the same conditions, exactly like each other, as a rule, or they differ hardly more than specimens of one and the same species are apt to do. This carefully stated proposition, founded upon experience, appears to be sufficiently justified by numerous experiments, but it is, nevertheless, subject to many exceptions. Some students of crosses have so limited its application that they dared only to assert the similarity of the speci- mens obtained from a capsule fertilized by the same plant. At all events, the rule proves itself to a certain degree trustworthy only in those cases in which the similarity of origin and con- dition of growth demanded by the terms of the rule are really present. The question easiest to answer is just the one about which there has been the most violent dis- cussion, namely, that relating to the stronger influence of the one or the other sex upon the offspring. The crosses of the two species or races A and B resemble each other whether A was male or female in the crossing. Experimenters, es- pecially Kolreuter, Gartner, Naudin, and Wichura, have not, on the whole, been able to find any arbitrary signs are used in the text to designate the sex of each parent. The sign $ represents the male, or the parent which furnished the pollen ; 9 stands for the female or seed-hearing parent. — L. H. B. FOCKE ON THE PRIMARY CROSS. 217 difference between A 9 x B $ , and 13 9 x A $ . More than one hundred years had passed since Kolreuter had proved the conformity of Nicotiana rustica 9 x panicnlata $ , and N. paniculata 9 X rnstica $ , when one of the acutest florists of our time, Timbal-Lagrave, was also astonished in the highest degree by a similar experience. All the rules and supposed experiences according to which the florists were to know, from the morphological characteristics of a hybrid, which of the progeni- tors had furnished the pollen for its formation, and which had borne the seed, are entirely ground- less and foolish. Besides, it has been proved by many experiments, that in the vegetable kingdom, with regard to species, as a rule, the form-deter- mining power of the male and female elements in the progeny is entirely equal. As with all other rules touching the crossing of plants, this one of the similarity of the products of reciprocal crossings is not without its excep- tions. It is a matter of course that an observed difference can be ascribed, with any probability, only to a stronger influence of the male or female elements, provided that the experiments are car- ried on in exactly the same manner, and if they always lead to the same result after repeated trials. So far, almost all the recorded experi- ments could be improved in this respect, for they are open to justifiable doubt. The following 218 BORROWED OPINIONS. declarations of the differences in the products of reciprocal crossing seem worthy of notice. a. The influence of the female element in Pe- largonium fulgidum x grandiflorum, P. peltatum X zonale, Epilobium hirsutum x Tournefortii, 1 ex- ceeds that of the male as regards general form. This influence is also shown more strongly in the colors of the flowers of several hybrids of Dig- italis, in some also in the form of the corolla. In Nymphaea rubra x dentata, the seed-leaves resemble those of the female progenitors. b. The female element shows, apparently, an overpowering influence in the power of with- standing cold in the Rhododendron (hybrids of R. arboreum), Lycium, and perhaps also in Cri- niim (hybrids of C. Capense). c. The male element influences the general form especially in Papaver Caucasicum x somni- ferum, and Cypripedium barbatum x villosuin (if constant ?) ; it shows a stronger influence upon the colors of the flower in Petunia. d. Gartner says that he has sometimes observed differences in the fruitfulness and progeny of re- ciprocal hybrid forms, e.g. in Dianthus barbatus X superbus. The experiences of Gartner should, however, hardly suffice to establish the uniformity of this occurrence. The principal differences between A ? and B $ 1 The various examples cited in the text are fully explained in the body of Focke's work, Die Pflanzen-Mischlinge. FOCKE ON THE PRIMARY CROSS. 219 and B 9 and A $ , have been observed by Kolreuter and Gartner in some Digitalis hybrids. That these differences really show themselves each time and in the same manner, is by no means proved. Much oftener, the departures from the regular uniformity of single specimens, observed among hybrids, are entirely independent of the influence of the species in the hybrid formation. Often important differences appear among the seedlings of the same cross which have been treated exactly alike. These differences show themselves in dif- ferent ways : — 1. Single specimens of the cross may show but slight differences among themselves, especially in the color of the blossoms and similar character- istics, which are easily changed ; compare, for example, the hybrids of Verbascum phoeniceum, Salix Caprea x daplmoides. 2. The cross may appear in two different types, each of which represents a different combination of the characteristics of the progenitors. As a rule, one type more closely resembles the one, and the second the other progenitor; the numbers of the two types are often very unequal. Gartner designated the rarer forms " exceptional types " (AuBnahmetypus). For examples, study Cistus, Dianthus, Geum, Oenothera, Lobelia, Verbascum Thapsus x nigrum, Nicotiana quadrivalis x Taba- cum, var. macrophylla. 220 BORROWED OPINIONS. 3. The cross may appear in several different types. Gartner gives some examples of these, although in his cases it is probably a question of three known forms of one polymorphous union. 4. The cross may appear in a typical interme- diate form, and a number of vacillating ones approaching one or the other parent, and among which no distinct type can be distinguished. Such is Medicago falcata x sativa, usually also Melandrium album x rubrum. 5. The cross may have from the beginning very many forms. The experiments which have been made leave it doubtful if in these cases one or more constant types, with a similar combina- tion of characteristics, can be distinguished among the vacillating forms. Study Abutilon, hybrids of Pelargonium glaucum, P. radula x myrrhi- folium, Passiflora, Hieracium, Nepenthes, Narcis- sus. Gartner made the assertion that hybrids between two species are of similar form, whilst crosses of varieties are polymorphous. If, by va- rieties, is understood the unsettled garden forms or garden crosses, then the remark is justifiable ; but if fixed races of pure descent are also included, then it is decidedly wrong. See " Cross-breeds and Hybrids" (section III., page 247). Entirely different results are obtained by com- paring hybrids which, although springing from the same species, were produced and grown in FOCKE ON THE PRIMARY CROSS. 221 different places. Spontaneous hybrids are, as a rule, much more variable than those produced artificially, e.g. Verbascum Lychnitis x Thapsus and V. Lychnitis x nigrum. My hybrids between Digitalis purpurea and D. lutea resembled each other much when I had sown the seeds ; on the other hand, very different forms arose when the seeds had accidentally sown themselves. It may be that in this case there was no real causative connection between the many forms and the man- ner of sowing ; nevertheless it is certain that several growers have very often obtained different results in their crossings from the same species. The uniformity of all the products of the same crossing, which undoubtedly is the rule with the experiments of growers, appears in nature rather to be the exception than the rule. It remains to be ascertained what influence the unequal nour- ishment of the parent species, or of the hybrid seedlings, has upon the variability of the hybrids. 2. The characteristics of the crosses may be differ- ent from the characteristics of the parent species. It is in size and luxuriance (see proposition 3), as ivell as hi their sexual ability (see proposition 4), that they differ most from both parent sjiecies. The manner differs in which the characteristics of the parent species are united in the cross. Generally, a blending or mutual union takes place, but often in such a manner that in one 222 BORROWED OPINIONS. respect one, and in another respect the other parent seems to appear. Sometimes, for example, the cross resembles one parent in the leaves, and the other parent in the flower. Sometimes there appears a variety of the cross, in which the char- acteristics are distributed in the reverse order. Some crosses resemble at first one species, and later the other parent species ; as their leaves show in the spring the one and in the fall the other type (Cistus, Populus); or the colors of the flowers change during the blossoming (Melan- drium album x rubrum, Epilobium roseum x montanum ; compare also Lantana), or in the fall (Nicotiana rustica x Tabacum, Tropseolum, Lobe- lia, etc.), also sometimes in different years (Bletia crispa x cinnabarina, Galium cinereum x verum) . By the union of races (rarely in hybrids in the more restricted sense of the term), one finds, among other things, the characteristics of the par- ent species unmixed beside each other (compare Cucumis Melo, the thorniness of the Datura fruits, color of the flower in Rhododendron Rho- dora x calendulaceum, R. Ponticum x flavum, Anagallis, Linaria vulgaris x purpurea, Calceo- laria, Mimulus, Mirabilis). The colors of the flower often appear in an unexpected and unex- plainable manner : the hybrids of Verbascum phceniceum are very variable in the colors of the flowers, and in other respects quite uniform; FOCKE ON THE PRIMARY CROSS. 223 on the hybrids of Helianthemum have sometimes been found differently colored flowers on the same stalk, at the same time. From the crossing of closely related races, especially varieties in color, there often appear plants which' exactly or very closely resemble the parent types; compare Brassica Rapa, var., Linum, Pisum, Fhaseolus ; Anagallis, Atropa, Datura Stramonium, Salvia Horminum, etc. Usually the influence of the second parent race does not show itself till the second generation; and, in fact, in such a manner that a part of the seed- lings return to that form, either entirely or only in certain respects. Only in Atropa, no return to the (slightly fixed) yellow form has been observed. In some cases, the cross resembles one of the parent forms so much that it could be considered as only a slight variation of the same. Even in crosses between two considerably different species, the predominating influence of one parent species is sometimes shown in a striking degree. The hybrid of Dianthus Armeria x deltoides resem- bles D. deltoides much more than the other parent species ; of D. Caryophyllus x Chinensis resembles D. Caryophyllus ; of Melandrium ru- brum x noctiflorum resembles M. rubrum ; of Ver- bascum Blattaria x nigrum resembles V. nigrum ; of Digitalis lutea x purpurea resembles D. lutea. 224 BORROWED OPINIONS. Sometimes even the primary hybrids show characteristics which are entirely different from either parent species ; especially is this the case, amongst other things, in the colors of the flowers. The most curious is the regularly blue-flowering hybrids of the white Datura ferox with the equally white D. he vis and D. Stramonium, var. Bertolonii. There are many examples of unexpected coloring of the blossoms of hybrids from species of colored flowers, while the crosses by no means always show the shades of color which would be obtained by a mixture of the parental pigments. Noticeable instances are shown, e.g., by Clematis recta x integrifolia, Aqui- legia atropurpurea x Canadensis (and others), Nicotiana suaveolens x glutinosa, Verbascum pulverulentum x thapsiforme, hybrids of V. phceniceum, Anemone patens x vernalis, Be- gonia Dregei x Sutherlandi (and others). In the crosses of races, e.g., of Papaver somniferum and Datura Stramonium, many characteristics appear which belong not to the parent forms, but to other races of the same species. Nicotiana rustica x paniculata sometimes shows the colors of the blossoms of N. Texana, a foreign sub- species of N. rustica. Other characteristics which the hybrids show in a greater degree than the parent forms, are, e.g., the greater stickiness of some hybrids of Nicotiana (rustica x paniculata). FOCKE ON THE PRIMARY CROSS. 225 the apparently greater wealth of honey in N. rustica x paniculata, the stronger nauseous odor of the hybrids of Melandrium viscosum, the supposed increase in the quantity of quinine (?) in hybrids of Cinchona (according to Kuntze). In later generations of hybrid growth, devia- tions from the characteristics of the parent species are much oftener observed. 3. Crosses of different races and species are dis- tinguished from plants of a pure race, as a rule, by the poiver of vegetation. Hybrids between very different species are often very weak, especially when young, so that it is difficult to successfully raise the seedlings. On the other hand, crosses of more closely related species and races are, as a rule, uncommonly luxuriant and strong ; they are distinguished mostly by size, rapidity of growth, early flowering, abundance of flowers, longer life, stronger reproductive power, unusual size of some special organs, and like characteristics. For a closer confirmation of this proposition, it will be to the purpose to refer to some examples. Delicate seedlings are mentioned, e.g., in Nym- phsea alba crossed with foreign species, Hibiscus, Rhododendron Rhodora with other species, R. Sinense with Eurhododendron, Convolvulus, poly- phyllons Salix hybrids, Crinum, Narcissus. The experience that seedlings of hybrid-fertilized seeds are delicate, is frequent. A dwarfish growth has 226 BORROWED OPINIONS. rarely been observed in hybrids, although there are some instances of it. Large growth is, on the other hand, much more common, e.g., Lycium Datura, Isoloma, Mirabilis. Usually the hybrids exceed in height both parent species, or at least their medium height ; compare, e.g., many hybrids of Nicotiana, Verbascum, Digitalis. The vege- tation sometimes takes place exceedingly fast. Klotzsch emphasizes the rapid growth of his hybrids of Ulmus, Alnus, Quercus, and Pinus. The flowering time often comes earlier than in the parent species, e.g., in Papaver dubium x somniferum, some Dianthus hybrids, Rhododen- dron arboreum x Catawbiense, Lycium, Nicotiana rustica x paniculata, Digitalis, Wichura's six-fold Salix hybrid, Gladiolas, Hippeastrum vittatum x Regime, etc., but especially in many hybrids of Verbascum. On the other hand, there are, no doubt, also single hybrids which blossom only after a long period, or not at all, e.g., in the genera Cereus and Rhododendron. Only one example of earlier maturity of seeds, independent of an earlier maturing of the blossoms, is known to me, namely, in Nuphar. Very often and very com- monly an extraordinary abundance of flowers has been observed in hybrids ; compare, e.g., Capsella, Helianthemum, Tropaeolum, Passiflora, Begonia, Rhododendron, Nicotiana (rustica x paniculata, glutinosa x Tabacum, and others), Verbascum, FOCKE ON THE PRIMARY CROSS. 227 Digitalis, many Gesneracese, Mirabilis, Cypripe- dium. The size of the blossoms is often increased in hybrids; in a cross of two species with different sized flowers it is not rare that the flowers of the hy- brid attain the size of those of the larger-flowered parent, or almost that size. Dianthns arenarins x snperbns, Rnbns ceesius x Bellardii, hybrids of Rosa Gallica, Begonia Boliviensis, Isoloma Tyda3um, give examples of uncommonly large flowers. A strong vegetative reproductive power is very common among hybrids ; compare, e.g., Nymphsea, hybrids of Rubus cresius, Nicotiana suaveolens x Tabacum, var. latissima, Linaria striata x vulgaris, Potamogeton. A longer life is especially notice- able in some hybrids of Nicotiana and Digitalis. A greater power of resistance against cold is also noticeable in Nicotiana suaveolens x Tabacum, var. latissima, while Salix viminalis x purpurea is said to be more sensitive to frost than either of the parent species. These facts indicate partly a certain loss of vigor which is inherent to the hybrids on account of their abnormal descent, and partly, on the con- trary, an exceptional vegetative power. This latter fact, which occurs much oftener than the former, has only lately been elucidated to any extent. The important experiments of Knight, Lecoq, and others, had been known for some time, but only through the careful researches of Charles 228 BORROWED OPINIONS. Darwin has the favorable influence of a cross between different individuals and races of one and the same species been clearly shown. The strengthening of the vegetative power in some hybrids is evidently a universal experience which needs no special explanation, on account of the peculiar circumstances of the formation of the hybrids. It was formerly thought that the di- minished sexual fruitfulness is compensated by a greater vegetative luxuriance, a statement the untenableness of which, as Gartner showed, is most plainly shown by the fact that many of the most fruitful crosses (Datura, Mirabilis) are also distinguished by an enormous growth. 4. Hybrids of different species often form in their anthers a smaller number of normal pollen grains, and in their fruits a smaller number of normal seeds, than plants of a pure descent ; sometimes they pro- duce neither pollen nor seeds. In crosses of nearly related races, this weakening of the sexual organs does not occur as a rule. The blossoms of sterile or slightly fruitful hybrids usually remain fresh for a long time. No characteristic of hybrids has received more attention than the diminution of the sexual power which has been observed in them. Even Kolreuter believed that this characteristic permitted the drawing of a sharp line between species and vari- eties. The same thought has prevailed amongst FOCKE ON THE PRIMARY CROSS. 229 botanists to a great extent since then, and even in very recent times Naudin, Decaisne, and Cas- pary have defended the ideas of Kolreuter in a more or less modified form. Knight, Klotzsch, and formerly also Gordon, considered the pollen of hybrids to be entirely impotent, a vieAv which was even then contradicted by the accurate ex- periments of Kolreuter. It has often been wrongly declared that Kolreuter had himself spread the doctrine of the total sterility of hybrids ; this assertion is to be explained only by ignorance or by a misunderstanding of the Latin text. Kol- reuter, in fact, does not speak of sterility, but of lessened fruitfulness, as being a common charac- teristic of hybrids. The fruitfulness of hybrids is considerably dif- ferent, according to the individual genera. For instance, the hybrids of Papaver, Viola, Verbas- cum, and Digitalis show but little fruitfulness ; the hybrids of Anemone, Nicotiana, Mentha, Cri- num, the Cucurbitacere, and Passiflora are much oftener fruitful, while in Aquilegia, Dianthus, Pelargonium, Geum, Epilobium, Fuchsia, Coty- ledon, Begonia, Cirsium, Erica, Rhododendron, Calceolaria, Quercus, Salix, Gladiolus, Cypripe- dium, Hippeastrum, the Gesneracese, and Orchids, the fruitful hybrids are commoner than the bar- ren ones. In the genera Vitis, Prunus, Fragaria, and Pyrus, we use the crosses of closely related 230 BORROWED OPINIONS. species as fruit plants ; in Cereus, hybrids of even widely different species show undiminished fruit- fulness. The sterility of hybrids is sometimes shown by their exhibiting no inclination to blossom, a char- acteristic which has been observed particularly in some hybrids of Rhododendron, Epilobium, Ce- reus, and Hymenocallis. These are, however, rare exceptions, for as a rule hybrids bloom earlier and more profusely than the pure species (see page 226). In hybrids which have flowers of only one sex, the staminate blossoms often fall off while they are yet buds, as in Cucurbitaceas and Begonias (hybrids of B. Froebeli). Sometimes the anthers are arrested in their growth and form hermaphro- dite flowers, as has been observed in some hybrids of Pelargonium and Digitalis (D. lutea x pur- purea form tubiflora Lindl.). The common result of the production of hybrids is a different forma- tion in the pollen grains in the hybrids from that of the parents. Often the anthers of the hybrids are dead and contain no pollen, or they are small and do not open at all. Such deficiency of pollen can be observed, for example, in Rubus Idreus x odoratus, Ribes aureum x sanguineum, Alopecurus geniculatus x pratensis. In other cases, the pollen dust consists of small powdery grains of irregular form and size, which do not swell when moistened, FOCKE ON THE PRIMARY CROSS. 231 and among which are usually found a few well- formed pollen cells which are capable of germina- tion. Often the number of normal pollen grains is greater, and comprises ten or twenty per cent, or more, of the whole number. There are often found, in a greater or less degree, large angular grains capable of swelling, as well as well-formed ones, by the side of those dwarfed or stunted. Among crosses of closely related species, as Melan- drium album x rubrum, there are usually found only few irregularities in the form of the pollen grains. In a hybrid Sinningia, the pollen of the second year of flowering was better than that of the first. In hybrids of undoubted different species, a normal formation of the pollen is seldom seen. The statements on this subject require, for the most part, corroboration, nevertheless I refer to Nymphsea Lotus x rubra, Begonia rubrovenia x xanthina, Isoloma Tydseum x sciadocalyx, Salix purpurea x repens ; almost perfectly formed pollen grains were found in Salix aurita x Caprea, and S. viminalis x repens. On the other hand, it is still more rare that a deficiency of pollen is found in an evident cross of races. Perhaps it could be found oftener if one sought for it. The only certain example of which I know is an Anagallis cross by myself. Whether Raphanus sativus and R. Raphanistrum 232 BORROWED OPINIONS. are to be considered as species or races is doubtful. Nevertheless there are some crosses of very closely related species which appear to be entirely sterile, as, for example, Capsella rubella x Bursa-pas- toris, Viola alba x scotophylla, Papaver dubium x Rhoeas. The sexual capabilities of the female organs are, as a rule, not so much weakened in hybrids as the male. Nevertheless they are greatly diminished. Many hybrids never produce fruit. Even after many experiments, one must not make definite assertions about the absolute sterility of a hybrid : in Rubus cassius x Idreus, for example, one can see several thousand blossoms remain sterile, and nevertheless, now and then, fruit is produced. Compare, further, Digitalis lutea x purpurea, Lobelia fulgens x syphilitica, Crinum Ca- pense x scabrum. A morphologically distinguish- able defect of the ovule of hybrids has seldom been referred to so far, although it has been observed in Cistus, by Bornet. If one wishes to obtain a defi- nite judgment on the female creative power in hybrids, he must fertilize the ovules with pollen of the parent species, which, as a rule, develops more perfect fruit than the pollen of hybrids which has been weakened in its generative power. In some cases, hybrids whose pollen shows but slight potency produce normal fruit with pollen from the parent, as in Luffa. FOCKE OX THE PRIMARY CROSS. 233 Some hybrids drop their entire blossoms un- wilted, with the calyx and flower stalk intact, e.g., Ribes, Nicotiana rustica x paniculata, and other hybrids of Nicotiana. As a rule, the corolla wilts, after a longer period, in a normal manner, or is thrown off as in the parent species ; but no setting of fruit thus takes place, or only a little of very poor character. Sometimes the fruit is well formed externally, but contains no seed. In many cases hybrids set fruit, but to a less extent and with fewer seeds than the parent species. Even in crosses of closely related species, the number of seeds seems to be smaller than in the parent species ; so, for example, according to Gartner, in Melandrium album x rubrum, Lobelia cardinalis x fulgens, and even in undoubted crosses of races of Verbascum. Hybrids of essentially different races rarely show an undiminished fruitfulness, although no noticeable diminution has been proved in Brassica Napus x oleracea, Dianthus Chinensis x pluma- rius, var. Sibiricus, Pelargonium pinnatum x hir- sutum, Abutilon, Medicago, a few Cereuses and Begonias, Hieracium aurantiacum x echioides, Ni- cotiana alata x Langsdorfhi, a few hybrids of Erica, Calceolaria, Isoloma, Veronica, and several Orchidaa. Fruits and seeds in abundance are also found in many other garden hybrids, and in many wild ones, as in Roses, Epilobiums, Fuchsias, Cir- 234 BORROWED OPINIONS. siums, HieraeinmSj Willows, of Lobelia Lowii, etc. In these cases, however, one cannot exactly ascertain if the plants are primary hybrids, or, which is much more probable, if they belong to later generations or arose through derivative hybridization QRiickkreuzungeri) . In order to set seeds, or at least to produce a vigorous progeny, some hybrids require to be fer- tilized by other individuals, even though they are themselves hybrids ; compare, for example, hybrids of Cistus, Begonia, Gladiolus, and Hippeastrum. In some hybrids, only the first flowers produce seeds, as Aquilegia, Dianthus, Silene, Lavatera Thuringiaca x Pseudolbia, Rubus foliosus x Spren- gelii ; in other cases, the first flowers are regu- larly sterile, while the later ones are often fruitful, as in Datura, Nicotiana rustica x paniculata, N. rustica x quadrivalvis, Mirabilis. In longer-lived plants, often all the flowers of the first years are sterile, while later, when the plant has reached a certain age, a few fruits are formed ; this has been noticed, for example, in Rubus Idceus x csesius, R. Bellardii x csesius, Calceolaria integri- folia x plantaginea, Crinum Capense x scabrum. Although, as a rule, the female creative power in hybrids is less weakened than the male, still there are some cases in which the reverse is true ; compare Nympluea Lotus x rubra, Cico- nium x Dibrachya in the genus Pelargonium, FOCKE OK THE PRIMARY CROSS. 235 Lobelia fulgens X syphilitica, Verbascum thapsi- forme x nigrum, Narcissus montamis, etc ; tliese are probably, for the most part, only accidental occurrences. The longer duration of the flowers (especially the pollinized ones) on mairy sterile hybrids, is an occurrence which is analogous to the longer duration of unfertilized or incompletely fertilized flowers. Often in sterile hybrids, especially after dusting with the pollen of the parent or a related species, the fruits swell more or less without per- fect seeds' being formed in them. Externally well-formed but seedless fruits are found in Cacti, Passifloras, Cucurbits, and Orchids. Gartner has studied this characteristic, which is of no special use or value in the culture of hybrids, very care- fully ; however, it offers an important proof of the truth of the statement that the development of the outside coats of the fruits takes place in a normal manner on account of the irritation which the germinating pollen produces, but that it is nevertheless independent of the fertilization of the ovule, and the development of the embryo and of the seed. One can, in general, make the statement that hybrids of closely related races are more fruitful on the average than those of considerably differ- ent species. One can also consider, as a rule, as has been shown above, that closely related species 236 BORROWED OPINIONS. form hybrids among themselves more easily than those considerably different. Both rules are true only to a certain extent. If one should infer from these that hybrids are more fruitful the easier they are formed, then he would go far wrong. There is no definite relation between ease of formation and fruitf ulness of crosses. From a teleological point of view, we formerly saw in the sterility of hybrids a means of keeping " species " separate. What purpose this separa- tion was to serve was not explained, unless it was for the convenience of the systematise Now, on the contrary, we ask if the formation and sepa- ration of species does not really depend upon the fruitfulness of the crosses between the dis- tinctly marked races of the parent types. The noticeable resemblance between illegitimate and hybrid progeny gives us no standpoint for further investigation of the cause of unfruitf ulness. More light may be given by the fact that in the hybrid pteridophytes and mosses the formation of sexless spores is just as much wanting as is the formation of pollen grains in hybrid phanerogams. The obstacle to the regular propagation of hybrids seems to lie in the development of certain cells which have the power to maintain the type of the parent form, it being immaterial if these cells have sexual functions or not. In any case, more facts must be gathered to justify the adoption or pres- FOCKE ON THE PROGENY OF CROSSES. 237 entation of a principle of such scope. This view of the situation can even now be considered as a hypothesis which certainly as yet offers no expla- nation of the limitations of species, but it leads the way to a final solution, because it brings a long list of different but plainly analogous phe- nomena as facts in the animal and vegetable king- doms under one common point of view. 5. Malformations and curious forms are much more common, especially in the flower parts of hybrids, than in individuals of a pure descent. Compare Papaver, Dianthus, Pelargonium, Ni- cotiana, Digitalis. Double flowers appear to be formed especially easily in hybrids. II. THE PROGENY OF CROSSES. Hybrids are more easily and completely fertil- ized by the pollen of the parent forms than by their own. Exceptions to this rule are hardly known (although compare Hieracium echioides x aurantiacum), although no large number of ex- periments have been made in this direction. By their own pollen is to be understood the pol- len of hybrids of the same cross, as well as that of the same individual. When hybrids grow in the vicinity of parent forms, they must naturally often be fertilized by them. In the progeny there will be, therefore, a number of forms between the 238 BORROWED OPINIONS. primary hybrids and the parent species. In sow- ing the seeds of hybrids, one cannot always deter- mine if fertilization from the parent species could have taken place or not. The general statement that the progeny of a hybrid has shown itself to be very variable is therefore of little weight or worth. Sometimes a hybrid is more easily fertil- ized by a third species than by its own pollen. Compare, for example, Nicotiana rustica x panic u- lata, and Linaria purpurea x genistsefolia. 1. Progeny of Crosses with their own Pollen. (Ax B) 9 x (Ax B) $. a. If one protects the fruitful hybrids from the influence of the parent forms or other related spe- cies, then he obtains hybrids of the second genera- tion. It seems to me that the progeny of hybrids appears or acts very differently according to the length of their life. In long-lived plants, the mingling and mutual blending of the two types which have been joined in the hybrid often ap- pears to be more complete, so that the progeny also inherits the characteristic of the new intermediate form or type in an equal or uniform manner. The progenies of one- or two-year hybrids are, as a rule, very different and varied in form ; compare Pisum, Phaseolus, Lactuca, Tragopogon, FOCKE ON THE PROGENY OF CROSSES. 239 Datura, Nicotiana alata x Langsdorffii, etc. Ex- ceptions are found in Brassica, CEnothera, Nico- tiana rustica x paniculata, Verbascum Austriacum x nigrum. The progenies of hybrids of several generations (mehrjahriger) are generally similar, although the cases in which the intermediate type shows itself to be constant appear to be much more frequent. Some hybrids of Aquilegia, Dianthus, Lavatera, Geum, Cereus, Begonia, Cirsium, Hieracium, Pri- mula, Linaria, Veronica, Lamium, and Hippeastrum appear to come very true to seed. The progenies of shrubs and trees are, in the majority of cases, very constant or invariable ; compare iEsculus, Amygdalus, Primus, Erica, Quercus, Salix. Also many hybrids of Fuchsia and Calceolaria are said to be constant. The Rhododendron hybrids are partly true to seed and partly variable. The progeny of the crosses of Vitis, Pyrus, and Crataegus, on the other hand, seem to be very variable. b. The several forms in which some primary hybrids appear seem to be unsettled in the prog- eny. In Dianthus, according to Gartner, the "exception types," when sown, mostly return to the normal mongrel form. The various primary forms of the hybrids of Hieracium, Mendel found to be true to seed. c. C. F. von Gartner and other botanists have 240 BORROWED OPINIONS. made the assertion that the progeny of hybrids becomes, from generation to generation, weaker and less fruitful. It is certain that their growth, which was at first increased, gradually decreases when they fertilize themselves. Gartner's experi- ments, however, were made on a very small scale, so that his hybrids were influenced not only by close in-breeding, but also by the various circum- stances which so often result in the loss of garden plants cultivated only in small numbers. Even Gartner noticed exceptions, as in Aquilegia, Dian- thus barbatus x Chinensis, D. Armeria x del- toides. Crosses of closely related species evidently can be continually propagated or kept up with ease ; compare Brassica, Melandrium, Medicago, Petunia. Many gardeners assert with much con- fidence that many hybrids can be propagated very well for several generations by means of seeds ; compare Erica, Lychnis, Primula Auricula x hir- suta, Datura. Many observations of wild plants seem to confirm this view. The principle has also been laid down that the fruitfulness of hybrids increases again in later generations. It does not, however, appear that this rule can have a very broad application. It is much more probable that single fruitful specimens arise among hybrids, which can easily reproduce themselves under favor- able external conditions by inheriting this pe- culiar ity. Fruitful descendants of hybrids are FOCKE ON THE PROGENY OF CROSSES. 241 probably sometimes the product of derivative hybridization. d. Complete reversions to the parent forms without the aid of pollen from parent forms arise only in crosses of closely related races. Even in such crosses, true reversions only seldom take place, e.g., Phaseolus. e. From the variable progeny of fruitful crosses a few principal types often arise after a few — per- haps three or four — generations. If one protects these new types from crossing, they tend to be- come constant or fixed. Scientific trials or exper- iments which confirm this statement have been made only to a small extent, especially by Lecoq in Mirabilis, by Gordon in Linaria and especially in Datura. These gardeners have produced many new races by the crossing of related species and well-fixed races. Also many wild, fixed, inter- mediate forms may have arisen in this manner ; compare Brassica, Lychnis, Zinnia, Primula, Petu- nia, Nicotiana commutata, Pentstemon, Mentha, Lamium. The new types of the descendants of crosses frequently differ in some characteristic from both parent forms. My Nicotiana rustica x paniculata had in the second and third genera- tions, on the whole, much smaller leaves than either of the parent forms. /. The sterility and inconstancy of the progeny of hybrids have often led botanists to conclusions 242 BORROWED OPINIONS. which are not confirmed by experience. It is entirely wrong, as can be seen by the facts which have been stated, to assert that all crosses would necessarily soon be lost on account of these char- acteristics which have been promiscuously ascribed to them. The unsettled forms arising from crosses are the plastic material out of which not only gardeners form their new varieties, but which material is biologically the more valuable as it furnishes new species in the household of nature. 2. Derivative Hybridization of Crosses tvith the Parent Forms. (i?x5J)9xiJ,(i?x5J)9x5J,l?x(ix5)J. As long as one laid much stress upon the male or female influence which one or the other parent species may have had upon the hybrid, a differ- ence was carefully made between the advancing hybrid forms, or those which more closely resem- bled the male parent, and the degenerating hybrid forms, or those which more closely resem- bled the female parent. But these differences are, according to experiments that have been made, of very subordinate importance, or, per- haps, of none at all. By treating hybrids with parental pollen, one obtains, as a rule, a rather varied progeny. The FOCKE ON DERIVATIVE HYBRIDS. 243 intermediate form between the hybrid and respec- tive parent is apt to be more numerous and more fruitful. Besides this, there are formed a less number of individuals, some of which resemble the hybrid, and some the parent species. Both are apt to be but slightly fruitful. The three-fourths hybrids, (A x B) ? x A $ , are often quite fruitful with their own pollen and appear to give races true to seed easier than the original hybrid ; compare iEgilops speltaBformis. Gartner observed frequently that in later genera- tions the pollen became more regular and the fruitfulness greater, as in Dianthus (Chinensis barbatus) x barbatus, and also in other three- fourths hybrids of Dianthus, Lavatera, and Nicotiana. If one treats the three-fourths hybrid (A x B) $ X A $ again with pollen from A, then he obtains a seven-eighths hybrid, or the third hybridized generation, which, as a rule, is very much like the parent which furnished the seven- eighths part, but it is still apt to show marked differences in form and fruitfulness in some indi- viduals. The last traces of the one original parent species disappears mostly in the fourth, fifth, or even sixth hybridized generation. Kolreuter and Gartner have completed the transition of one parent species into the other in many cases. They found that for a complete 244 BORROWED OPINIONS. change, three to six generations were necessary — as a rule, four to five. Evidently the greater or less duration of the change depends more or less upon surrounding circumstances. Gordon found that Melandrium album x rubrum was like the parent species in the second generation, when fertilized with its own pollen, while Gartner found three to four generations necessary to bring the one into the other by means of parental pollen. As a rule, the products of the fertilization of a parent species with pollen from a hybrid, as A 9 x(AxB)<£, are similar to those of the opposite cross ; nevertheless the statements of observers agree that the variety of forms is apt to be greater when one uses the hybrid as the male element ; compare Dianthus and Salix. There appears, in the products of derivative hybrids of crosses, as among the direct progeny, new characteristics, Avhich are lacking in the parent forms, but are for the most part found in related races or species. 3. Hybrids of Several Species. a. Triple Hybrids. In the first years of his experience, Kolreuter succeeded in uniting three entirely different Nico- tiana species into one hybrid form. The simplest FOCKE ON MULTIPLE HYBRIDS. 245 formulas according to which such a union could take place are : (A x B) 9 x C $ , C 9 x ( A x 13) $ , and (A x B) 9 x (A x C ) $ . Jn the genera Dianthus, Pelargonium, Begonia, Rhododendron, Nicotiana, Achimenes, Calceolaria, Salix, Hippeas- trum, Gladiolus, and a few others, there has been made a number of such unions without any par- ticular difficulty. One must nevertheless deter- mine if he unites three essentially different species, or if two of the factors, or even all three are only closely related to each other. There are similar but evidently different species which, in crossings among themselves, behave almost like races of the same species, as for example : — Melandrium album and rubrum. Vitis vinifera, cordifolia, aestivalis and Labrusca. Lobelia fulgens, splendens, and cardinalis. Rhododendron Ponticum, arboreum and CataAv- biense. R. flaviim, viscosum, nudiflorum, and calendu- laceum. Berberis Aquifolium, and the most closely related species. Hybrids between the crosses of two species of these groups with the third species of the same genus, can no more be called true triple hybrids than crosses of species belonging to some smaller or narrower group of Vitis, Lobelia, and Rhodo- dendron. True triple hybrids which have been 246 BORROWED OPINIONS. formed from three essentially different species are apt to be much varied in form, especially if the male parent species was a hybrid. On the other hand, in those unions which are most easily formed and are made by the formula (A x B) 9 x C $ , the type of C is apt to predominate strongly, as, for instance, Nicotiana (rustica x paniculata) ? X Longsdorfhi $ , Achimenes (grandiflora x Can- dida) ? x longiflora $ , and other Gesneraceas. The hybrids of Erica are said to produce just as uniform a progeny as the pure species. Several Salix hybrids have acted in the same manner. For gardeners, therefore, the triple hybrids in some genera (as in Pelargonium, Begonia, Rhodo- dendron, Achimenes, Isoloma, Cypripedium, Gladi- olus) are very valuable. If they produce seeds, their progeny is very variable. b. Hybrids of Four to Six Species. If one does not count the crossings of very nearly related species (as Vitis, Rhododendron, etc.), these hybrids of four or more parent forms are somewhat rare. We know them especially in the genera Dianthus, Pelargonium, Begonia, Rho- dodendron, Nicotiana, Salix, Hippeastrum, Gladi- olus. The artificial union of different species in a single hybrid form has been carried farthest by Wichura, who united six Salix species. FOCKE ON CROSS-BREEDS AND HYBRIDS. 247 c. Crosses of Plants Grown Together. In some genera, as Pelargonium, Fuchsia, Be- gonia, Rosa, Erica, Rhododendron, Achimenes, Cal- ceolaria, Gladiolus, Hippeastrum, gardeners have crossed species and hybrids in the most manifold manner, intentionally and unintentionally, and have used the most promising forms obtained for further propagation. The progeny of this com- plex crossing is naturally almost always very vari- able. There appear, however, to be exceptions to this rule ; Sweet plainly asserts that one always obtains the same cross from the crossing of some complex Pelargonium hybrids. Such constant complicated hybrids are, according to him, P. (hyb.) involucratum x (hyb.) ignescens and P. (hyb.) Mostynae x (hyb.) ignescens. That the Erica and some Salix hybrids produce a uniform progeny has already been mentioned. III. CROSS-BREEDS AND HYBRIDS. According to usage, we designate unions of two different varieties of one species as cross-breeds, unions of two different species as hybrids. It is necessary, on account of the indefiniteness of the term " variety," to remember that only varieties true to seed, or races and sub-species, can bequeath their characteristics with any degree of certainty ; 248 BORROAVED OPINIONS. inconstant species, which are so often designated as varieties, are not considered in the theory of hybridization. Many writers have taken great pains to find a difference between cross-breeds and hybrids ; they held firmly to the hope that by means of trials in crossing a boundary between species and sub- species could be formed. Gartner, who expresses himself plainly in several parts of his work that the appearance of crosses clearly proves the specific differences of relationships of the parent forms, becomes very reticent as soon as he attempts, on pages 574-582, connectedly to unfold the princi- ples of "variety hybrids." Herbert and Naudin have formed the opinion, after their many experi- ments, that it is impossible to draw the line be- tween cross-breeds and hybrids ; but, nevertheless, later botanists have again tried to find precise dif- ferences between them. The following propositions have been made : — 1. The pollen of cross-breeds is normal : hy- brids have a greater or less number of imperfectly formed grains in their pollen. 2. The fruitfulness of cross-breeds is normal : that of hybrids plainly diminished. 3. Hybrids of two species with differently col- ored blossoms produce flowers of mixed or uni- formly modified colors : plants with irregular, mottled flowers have always been produced by FOCKE ON CROSS-BREEDS AND HYBRIDS. 249 the crossing of varieties. It is the same with the coloring, marking, covering of the fruits, and other characteristics. 4. Cross-breeds have a strong inclination to return to the parent form in later generations. These four propositions are in the main correct, but they offer little help, in a case of doubt, to a right decision as to specific merits. The cross of the red and white Anagallis arvensis would have to be considered as a hybrid on account of its pol- len, and as a cross-breed on account of the appear- ance of flowers of two colors. In Datura, crosses, which in other respects are plainly characterized as hybrids, easily show complete returns to the parent forms. Hybrids whose fruitfulness appears to be in no way diminished have already been men- tioned (page 229). One can, consequently, make the rule, that crosses of closely related races are apt to show the characteristics ascribed to cross- breeds, but it is impossible by that means to es- tablish any sharp line between race crosses and species hybrids. Usually a few other characteristics are ascribed to cross-breeds by which they are distinguished from the hybrids of species. Gartner has asserted that cross-breeds of like descent are even in the first generation very dissimilar, while hybrids of the first generation are always very uniform. This assertion, which is also repeated by others, 250 BORROWED OPINIONS. is entirely wrong. The polymorphism of the hy- brids of the species of Abntilon, Passiflora, Hiera- cium, etc., has already been shown, Avhile, on the other hand, the crosses of races, in the first gener- ation, are usually just as uniform as the real hy- brids. Again, it has sometimes been asserted that the " varieties" of one and the same species, when crossed with another species, always produce the same hybrid forms. Gartner, especially, has laid particular stress upon this supposed behavior of varieties, although he must have known that Kol- reuter had already observed the inheritance of color of the blossom in the races of Mirabilis, Dianthus, and Verbascum, the doubling of flowers in Aquilegia and Dianthus, the carriage and form of the leaf in the races of Nicotiana Tabacum and Hibiscus. The white-blossoming Datura ferox gives with D. Stramonium a white-blooming cross, and with the smooth-fruited race (var. Bertolonii) of the same species, a blue -blossoming cross. Nympheoa Lotus x rubra is different from N. Lotus x dentata. It cannot be in the least doubt- ful that the inheritable characteristics of races and so-called varieties are also bequeathed to their progeny. One will hardly go wrong if he assumes that Gartner came to make this rule about the be- havior of varieties through the behavior of unfixed garden crosses and garden sorts. It is a matter FOCKE ON CROSS-BREEDS AND HYBRIDS. 251 of course that forms which show themselves un- fixed in their normal progeny should produce polymorphous hybrids, and that unfixed variety- signs are apt to disappear entirely in the products of crossing with pure species. The true situation is, in short, as follows : — The nearer the morphological and systematic rela- tionship of the parent forms is, the less the sexual capacity of reproduction in the cross is apt to depart from the normal direction ; the greater the difference between the parent forms, the more, on the average, is the fruitfulness of the cross weak- ened. Exceptions are not rare. The nearer the parent forms are related to each other, the oftener the progeny of crosses show complete returns to the parent forms. Crosses from nearly related parent forms some- times show in their blossoms and fruits the pe- culiar characteristics of the parent forms unmixed beside each other ; this rarely takes place in crosses whose parent forms were considerably different. Most unsymmetrically colored flowers (Mirabilis, Camellia, Mimulus, Petunia, etc.) first originated in the progeny of crosses. LECTURE V. pollination; or how to cross plants. 1. The Structure of the Flower. Pollination is the act of conveying pollen from the anther to the stigma. It is the manual part of the crossing of plants. The word fertilization is often used in a like sense, although erroneously; for it is the office of the pollen, not of the opera- tor, to fertilize or fecundate that part of the flower which is to develop into a seed. The chief requirement in pollinating flowers is to know the parts of the flower itself. The con- Fig. 1. — Bell-tiower. spicuous or showy part of the flower is the envelope, which is endlessly modified in size, form, and color. 252 STRUCTURE OF THE FLOWER. 253 This envelope protects the inner or essential organs, and it also attracts insects, which often perform the labor of pollination. This floral envelope is usu- ally of two series or parts, — an outer and commonly green series known as the calyx, and an inner and generally more showy series known as the corolla. These two se- ries are well shown in the bell-flower, Fig. 1. The calyx, with its re- flexed lobes, is at C, and the large bell-form portion is the corolla. When the calyx is com- posed of separate parts or leaves,, each part is called a sepal; in like manner each separate part of the corolla is a petal. In the lily, Fig. 2, there is no dis- tinction between calyx and corolla ; or, it may be said, the calyx is wanting. These envelopes of the flower are often much disguised. This is particularly true in the orchids, one of which, a lady-slipper, is illustrated in Fig. 3. The sepals are seen at DD. They are apparently only two, but there is reason to believe that the lower sepal Fig. 2. — Flower of white lily. 254 POLLINATION. is really made up of a union of two. The three inner leaves are the petals, the lower one, H, being enlarged into the sac or slipper. The most important organs of the flower, how- ever, to one who wishes to make crosses, are the so-called sexual organs, the stamens and pistils. They can be readily distinguished in the B lily, Fig. 2. The six bodies shown at S are the ends of the stamens, or so- called male organs. These stamens gen- erally have a stalk or stem, known as a filament, and the en- larged tip as the anther. It is in this anther that the pol- len is borne. The pollen is generally made up of very mi- nute yellow or brown- ish grains, although it is sometimes in the form of a more or less glu- tinous or adhesive mass, as in the milk-weeds and orchids. The irritating dust which falls from the corn tassels at the later cultivatings is the pollen. Fig. 3. — Flower of greenhouse cypripedium. THE ESSENTIAL ORGANS. 255 The pistil, or so-called female organ, is shown at OP, Fig. 2. The enlarged portion at O is the ovary, which will develop into the seed-pod. The stigma, or the enlarged and roughened part which receives the pollen, is at P. Between these two parts is the slender style, a portion which is absent in many flowers. The stamens and pistils are known as the essen- tial organs of the flower, for, whilst the calyx and corolla may be entirely absent, either one or both of these organs is present ; and these are the parts which are directly concerned in the reproduction of the species. Like the floral envelopes, these essential organs are often greatly modified, so much so that botanists are sometimes perplexed to distinguish them from each other or from mod- ified forms of the petals or sepals. The particu- lar features of these organs which the plant-breeder must be able to distinguish are the anther and the stigma; for the anther bears the pollen, and the stigma must receive it. In Fig. 1, the stamens are shown at E. In the flower A, which has just expanded, these stamens are rigid and in condition to shed the pollen, but in the flower B, they have shed the pollen and have collapsed. The stigma in this case is divided into three parts, but when the flower first opens, these parts are closed to- gether, H in flower A, so that it is impossible that they receive any pollen from the same floAver ; 256 POLLINATION. when the stamens have withered, however, as in B, the stigma, H, spreads open and is ready to Fig. 4. — Flower of night-blooming cereus. receive any pollen which may be brought to it by insects or other agencies. In this case, the ovary THE ESSENTIAL ORGANS. 257 or young seed-pod, which is in the bottom of the flower, is not shown in the engraving. Some of the particular forms of essential organs are Avell illustrated in the accompanying photo- graphs. In the night-blooming cereus, Fig. 4, the many-rayed stigma is shown just below the Fig. 5. — Flower of the shrubby hibiscus (Hibiscus Syriacus). centre of the mouth of the flower, and the nu- merous stamens are arranged in a circular manner outside of it. The many petals and numerous spreading sepals are also well shown. The hibis- cus, Fig. 5, has a central column with the anthers hanging upon it, and a large stigma raised beyond 258 POLLINATION. Fig. 6. — Bugbane (Cimicifuga raceniosa). them. The wild bugbane, or cimicifuga, is seen in Fig. 6, natural size. Here is a long spike or cluster of flowers. At the top are the unopened buds, in the centre the expanded flowers with the floral envelopes fallen away, — the fringe-like stamens very prominent, — and below are seen the pis- tils, the stamens having fallen. These pistils will now ripen into pods, but the tip-like stigma may still be seen on them. The stamens and the long protruding style, tipped with its stigma, are also shoAvn in the fuchsia, Fig. 15. The essential organs of orchids are cu- riously disguised. They are combined into a sin- gle body. In the lady- slipper, Fig. 3, the lip-like stigma is shown at P. Upon either side, at its STAMINATE AND PISTILLATE FLOWERS. 259 base, is an anther S. Projecting over the stigma is a greenish ladle-like body, T, which is a trans- formed and sterile anther. In all lady-slippers, these organs are essentially the same as in the drawing, although they vary much in size and shape ; but in most other orchids, the two side anthers, S, are wholly wanting, and the terminal organ, T, is a pollen-bearing anther. In numer- ous plants, there are many distinct pistils in each flower. Such is the case in the strawberry, where each little yellow "seed" on the ripened berry represents a pistil ; and the blackberry and the raspberry, where each little grain or drupelet of the fruit stands for the same organ. A flowering raspberry is illustrated natural size in Fig. 7, for the purpose of showing the ring of many anthers near the centre of the flower, inside of which, in the very centre, is a little head of pistils. It frequently occurs that the stamens and pistils are borne in different flowers, rather than together in the same flower as they are in the examples which we have studied. In these cases the flower is said to be staminate, or male or sterile, in one case, and pistillate, female or fertile, in the other case. If these two kinds of flowers are borne together upon the same plant, as in pumpkins, melons, cucumbers, chestnuts, oaks, and begonias, the plant is said to be monoecious; but if the stami- nate and pistillate flowers are on entirely different 260 POLLINATION. plants, as in willows and poplars, the plant is dioe- cious. The two kinds of squash flowers are shown in Fig. 8. The pistillate flower is on the left, and it is at once distinguished by the ovary or little squash beloAV the colored portion, Fig. 7. — Blossom of flowering raspberry (Hub us odoratus). or corolla of the flower. The lobed stigma is seen in the centre. The staminate flower is on the right. It has a longer stem, no ovary, and the anthers are united into a conspicuous cone in the centre. The flowers expand early in the morning. Insects carry pollen to the pistillate flower, which STAMLNATE AND PISTILLATE FLOWERS. 261 then begins to set its fruit, whilst the staminate flower dies. The flowers of the common wild clematis are shown in Fig. 9. Upon the right Fig. 8. — Squash flowers of each sex. are the sterile flowers, which are wholly stami- nate. On the left, the flowers with larger sepals — the petals are absent — have a cone of pistils in 262 POLLINATION. the centre, and a few short and sterile stamens spreading from the base of the cone. These dif- ferent flowers are borne on different plants in this species of clematis, and the plants are therefore practically dioecious, because the stamens of the pistillate flowers generally bear no pollen. A sim- ilar mixed arrangement occurs in some strawber- efe .■*■-■ .« *SrSi- Fig. 9. — Flowers of clematis (Clematis Virginiana). ries, except that there are no purely staminate flowers. There are purely pistillate varieties, others, like the Crescent, with a few nearly or quite abortive stamens at the base'_ of the cone of pistils, and others in which the flowers are per- fect or hermaphrodite, that is, containing the two sexes. COMPOSITOUS FLOWERS. 263 The compositous flowers — like the asters, daisies, goldenrods, sunflowers, dahlias, zinnias, chrysan- themums, and their kin — need to be considered in still a different category. In these plants, the head, or so-called flower, is an aggregation of sev- eral or many small flowers or florets. Each seed in a sunflower head, for example, represents a dis- tinct flower. Sometimes all of these flowers are perfect, — contain the two sexes, — and sometimes they are pistillate or staminate in different parts of the head ; and in some cases the plants are dioecious. In many plants of the composite fam- ily, the flowers near the border of the head are unlike those of the centre or disc, in having a long ray-like corolla ; and these ray-flowers are frequently of different form from the others in the character of the essential organs. Very frequently the ray-flowers are pistillate, whilst the disc-flow- ers are generally hermaphrodite. The anthers, in these plants, are united in a ring closely about the style and below the stigma. The ovary, as we have seen, ripens into the pod, berry, or other fruit ; but it is not able to bear seeds until it is assisted by the pollen. The pollen falls upon the roughish or sticky surface of the stigma, and there germinates or sends a minute tube downwards through the style and finally reaches the ovule, which, when fertilized, rapidly ripens into the seed. The nature of this 264 POLLINATION. fecundation is not germane to the present subject ; but it may be said that only one pollen grain is necessary to the fertilization of a single ovule, but the addition of a superabundance of pollen greatly stimulates the growth of the fleshy or enveloping parts of the fruit. It is important that the person who desires to cross plants should become familar with the stigma when it is "ripe," receptive, or ready to receive the pollen. This condition is gen- erally indicated by the glutinous or sticky or moist condition ol the stigma, or in those stigmas which are not glutinous it is told by the appearing of a distinctly roughened or papillose condition. This receptive condition generally occurs about as soon as the flower opens. If pollen is withheld, the stigma will remain receptive much longer than when fertilization has taken place, — in some flow- ers for two or three days. The pollen is discharged from the anther in various ways, but it most commonly escapes through a chink or crack in the side of the anther. Sometimes it escapes through pores at one end of the anther ; and in other cases there are more elaborate mechanisms to admit of its dis- charge. In most plants, the anthers and stigma in the same flower mature at different times, so that close-fertilization or in-breeding is avoided. This is well illustrated in the bell-flower, Fig. 1. Here the anthers wither and die before the stig- PREPARING THE FLOWERS. 265 matic lobes open. In other cases, the stigma matures first, although this is not the usual con- dition. II. Manipulating the Flowers. We are now familiar with the essential principles in the pollination of flowers. Before a person pro- ceeds to operate upon a flower with which he is unfamiliar, he should carefully study its structure, so as to be able to locate the different organs, and to discover when the pollen and the stigma are ready for the work. The first and last rule in the pollinating of plants is this : Exercise every precaution to prevent any other pollination than that which you design to give. The anthers, therefore, must be removed from the flower before it opens. This removal of the anthers is known as emasculation. Just as soon as this is done, tie up the flower securely in a bag to protect it from foreign pollen which may be brought by wind or insects. As soon as the stigma is ripe, remove the bag and apply the de- sired pollen, placing the bag on the flower again, where it must remain until the seeds begin to form. The stigma may be receptive the day fol- lowing emasculation, or, perhaps, not until a week afterwards. Much depends upon the age of the bud when emasculation takes place. It is gener- 266 POLLINATION. ally best to delay emasculation as long as possible and not have the flower oj^en ; but the operator must be sure that the anthers do not discharge or that insects do not get into the flower before he has emasculated it. The bud at B, in Fig. 3, is Fig. 10. — Tobacco flowers, showing the parts of the flower, a hud ready to be emasculated, and an emasculated subject. nearly ready to emasculate. The older buds on the top of the spike of bugbane, Fig. 6, are ready to operate upon ; and so is the bud seen at the left in Fig. 7. The manner of emasculating the flower varies EMASCULATING. 267 with the operator. It is a common practice to clip off the anthers with a pair of small scissors, or to hook them out with a bent pin or a crochet hook. Others nse tweezers. For myself, how- ever, I do not like any of these methods, because the anthers are apt to drop into the bottom of the corolla, where it is sometimes difficult to rescue them ; and if one uses tweezers, there is always danger that the anthers may be crushed and that some of the pollen may adhere to the instrument and contaminate future crosses. I therefore usu- ally cut the corolla completely off just above the ovary, with a pair of small, long-handled surgeon's scissors (see Fig. 12), removing everything but the pistil. The operation is explained in Fig. 10, which shows the tobacco flower. The flower at the left shows the pin-head stigma in the centre of the throat, and the five anthers surrounding it. The second flower is spread open for the purpose of showing these organs. The third figure is a bud in the right condition for operation. The right-hand figure shows this bud cut around with the points of the scissors, leaving only the pistil. The line at W, in Fig. 2, shows where the flower of the lily might be cut off. The manner of oper- ating upon a compositous flower is shown in the picture of the zinnia, Fig. 11. In this plant the outer florets of the head are pistillate, whilst those of the disc are perfect. It is only necessary, 268 POLLINATION. therefore, to remove the central stamen-bearing flowers before any of them open, and to cover the flower up before any of the pistils near the border Fig. 11. — Zinnia flowers; the upper head ready for emasculation. the lower one showing the operation performed. have protruded themselves. The upper head in Fig. 11 shows the untreated sample, whilst the lower one shows the same with the cone of central EMASCULATING. 269 flowers pulled out. This treated head should now be covered, to await the maturing of the stigmas. In many compositous plants, however, the case is not so simple as this, because all the flowers are perfect. In such cases, nearly all the florets should be removed from the head, and a few remaining ones emasculated in essentially the same manner as described for the tobacco, Fig. 10. Whenever flowers are borne in clusters, nearly all of them should be removed and the attention confined to only two or three of them. One is then more cer- tain of getting seeds to set. In some cases, like the apple cluster, only one or two flowers of any cluster ever set fruit, and the operator should then choose the two or three strongest and most prom- ising buds, and cut all the others off. Flowers which bear no stamens, as the pistillate flowers of squashes, strawberries, and many other plants, of course do not require emasculating. They should be tied up while in bud, however, to prevent the access of any foreign pollen. Indian corn is a case in point. The pistillate flowers are on the ear, each kernel of corn representing a single floAver. The silks are the stigmas. If it is desired to cross corn, therefore, the ear should be covered before any silks are protruded, and the pollen should be applied some days later, when the silks are full grown. The staminate or male flowers are in the tassel. 270 POLLINATION. The pollen should be derived from a flower which has also been protected from wind and in- Fig. 12. — Instruments used in pollinating flowers, natural size. Pin scalpel, scissors, lens. APPLYING THE POLLEN. 271 sects, because foreign pollen may have been dropped upon an anther by an insect visitor and it may be unknowingly transferred by the operator. The pollen-bearing parent needs no operation, of course, but the flower should have been tied up in a bag when it was in bud. The pollen is best obtained by picking off a ripe anther and crush- ing it upon the thumb-nail. Then it is trans- ferred to the stigma by a tiny scalpel made by hammering out the small end of a pin, as shown, full size, at the left in Fig. 12. The stigma should be entirely covered with the pollen, if pos- sible. It is often advised to use a camel's hair Fig. 13. — Ladle for pollinating house tomatoes. brush to transfer the pollen, but much of the pollen sticks amongst the hairs of the brush and is ready to contaminate a future cross ; and where the pollen is scarce it cannot be conserved to advantage by a brush. In some cases the pollen is discharged so freely that the anther may be rubbed upon the stigma, or even shaken over it, but in most instances it will be necessary to actu- ally place the pollen upon the stigma with some hard instrument. When pollinating house-grown melons and cucumbers, the staminate flower is broken off, the corolla stripped back, and the 272 POLLINATION. anther-cone inserted into the pistillate flower, where it is allowed to remain until it dries and falls away. In pollinating house tomatoes, an implement shown in Fig. 13, one-third size, is used. This is simply a watch-glass, T, secured to a handle. When the house is dry, at midday, the watch- glass is held under the flowers, which are tapped, and the pol- len falls into the glass. The glass is then held up under another flower until the stigma rests in the pollen. It should be said, however, that this pol- lination of tomatoes is for the purpose of making the fruit set in the absence of insects, , not to effect a cross. If the c.^J^,^__^!Sj latter purpose were the object sought, the flowers which are Fig. 14. -Bag for cov- t fe ^ d ^ need enng the flowers. to be emasculated. Sometimes it is impossible to secure the pollen at the time the stigma is ready. In some cases of this kind, the intended parents can be grown under glass so as to bring them into bloom at the same time. In other cases, it is necessary to keep the pollen for some time. The length of time that pollen will keep varies with the species and KEEPING THE POLLEN. 273 probably also with the strength and vigor of the plant which bears it. As a rule, it Avill not keep more than a week or two, and, in general, it may Fig. 15. — Fuchsias, showing the stamens and pistils, and a bud ready to be emasculated. be said that the fresher it is the better it may be expected to act, It is best kept in dry and tig] it 274 POLLINATION. paper bags, such as are used for covering the flowers. Something more should be said about the bags which are used for covering the flowers. After having tried every kind which is recommended, I find grocer's manilla bags much the most satis- factory. For most flowers the four-ounce size is the handiest. When the bags are still flat, as Fig. 16 Fuchsia flower emasculated. they come from the packages, a hole is made through the two overlapping folds near the open- ing, and a string is passed through it and then tied at one of the folds, as shown in Fig. 14. The bag is then ready for use. Before it is put on the flower, the lower end of it is dipped in water to soften it so that it can be puckered tightly about the stem and thereby prevent the BAGGING THE FLOWER. 275 entrance of any in- sect. A bag is put upon the seed-bear- ing flower when emasculation is per- formed, and upon the intended pol- len parent when the flower is still in bud. The bag may be removed from the emascu- lated flower from time to time to ex- amine the stigma, and • again when the pollen is ap- plied ; but it should not be taken off permanently until the pod or fruit begins to grow. By way of re- capitulation, let us consider the cross- ing of a fuchsia Fig. 17. — Fuchsia flower tied up after flower. In Fig. 15 emasculation, two flowers are shown in full bloom, with the long style and the eight shorter stamens. The single 276 POLLINATION. bud is just the right age to emasculate. We therefore cut off the two flowers and emasculate the bud, as in Fig. 16. The pollen of another flower is applied and the bag is tied on, as seen in Fig. 17. The best label is a small merchandise tag, and this records the staminate parent and the date. It will be seen that in the operation of emascu- lating the fuchsia flower we cut off the sepals as Fig. 18. — Tomato and quince, showing how the sepals were cut off in emasculating. well as the petals. In some plants the calyx adheres to the full-grown fruit, as on the apple, pear, quince, gooseberry, or persists at the base of the fruit, as in the tomato, pea, raspberry. In these fruits, therefore, the cutting awa} r of the calyx leaves an indelible mark which at once dis- tinguishes the fruits which have been crossed, CROSSING IX FLOW EH LESS PLANTS. 277 even if the labels are lost. In Fig. 18 a tomato and quince are shown which are thus marked. All the foregoing remarks do not apply to the crossing of ferns, lycopods, and the like, because these plants have no flowers; yet cross-fertiliza- tion may take place in them. When the spores s i Fig. 19. — Pollinating kit. of these flowerless plants are sown, a thin green tissue, or prothallus, appears and spreads over the ground. In this tissue the separate sex-organs appear, and after fecundation takes [dace, the fern, as we commonly understand it, springs forth. Thereafter, this fern lives an asexual life and 278 POLLINATION. produces spores year after year ; but it is only in this primitive prothallic stage that fertilization takes place, once in the lifetime of the plant. If these plants are to be crossed, the only procedure open to the gardener is to sow the spores of the intended parents together in the hope that a nat ural mixing may take place. There are various well-authenticated fern hybrids. The pollination of flowers is such a simple Avork that few implements are required for its easy performance. Great care is more important than Fig. 20.— Pollinating kit. any number of tools. Every one who expects to cross plants should provide himself with the three instruments shown in Fig. 12, — a pin scalpel, sharp-pointed scissors, and a large hand-lens. If one contemplates much experimenting in this direction, however, it is economy of time to have some sort of a box in which there are compart- ments for the various necessities. These various compartments suggest at once whatever accesso- ries are wanting, and they hold a sufficient supply IMPLEMENTS USED IN CROSSING. 279 for several hundred operations. There should be a compartment for bags, string, lens, scissors, and pencils, tags, note-book, and the like. Figs. 19 and 20 show a convenient case for an experi- menter, and one which I have used with satisfac- tion for several years. This kit is twelve inches long, nine inches wide, and three inches deep. The chances of success in pollinating are dis- cussed in Lecture II. (page 88). GLOSSARY. 1. The Flower. Anther. — That portion of the stamen which bears the pollen. It is the uppermost portion of the stamen. < 'alyx. — The outer series of floral envelopes, usually green. The various separate parts of the calyx are sepals. Corolla. — The inner series of floral envelopes, usually colored and forming the showy part of the flower. If it is divided into separate parts, these are called petals. Essential organs. — The stamens and pistils. Female. — Said of flowers which have only pistils or the seed-bearing part, or of plants which bear only such flowers ; applied also to the pistils in any flower. Filament. — The stalk or stem of a stamen, bearing the anther. Floral envelopes. — The calyx and corolla. Male. — Said of flowers which bear only stamens, or of plants which have only staminate flowers ; also applied to the stamens or pollen-bearing organs of flowers. Ovary. — The lowest part of the pistil, containing the ovules. It is the most thickened portion of the pistil, and it may stand either below or above the petals. The ovary ripens into the fruit. Ovule. — A body in the ovary which ripens into a seed. Pet'-al. — The separate portions or leaves of the corolla. Pistil. — The seed-bearing organ of the flower. It always comprises two parts, the ovary — which becomes the pod or fruit — and the stigma. Usually there is a 281 282 GLOSSARY. style connecting the two. Often called the fertile or female organ. Pistillate. — Said of a plant or flower which has only pis- tils or female organs. Pollen. — The contents of the anther, capable of fertil- izing the ovules. It is usually composed of minute yellow or brown grains. Se'-pal. — The separate portions or leaves of the calyx. Spore. — The reproductive organ of flowerless plants, by means of which they propagate, as other plants propa- gate by means of the seed. The spore is asexual. Stamen. — The pollen-bearing organ of the flow 7 er. Often called the male or sterile organ. Its essential part is the anther. The stalk, when present, is called the filament. Staminate. — Said of a flower or plant which bears only stamens or male organs. Stigma. — The top end of the pistil, where the pollen lodges and germinates. It is usually a somewhat ex- panded surface, and is roughened, or sticky, or moist when ready to receive the pollen. Style. — The more or less slender portion of the pistil which lies between the stigma and ovary. The pol- len-tubes pass through it in reaching the ovary. 2. Crossing. Bigener; bigeneric-liybrid. — A hybrid between species of different genera. Bigeneric half-breed. — The product of a cross between varieties of species of different genera. Close-fertilization ; self-fertilization. — The action of pollen upon the pistil of the same flower. Close-pollination : self-pollination. — The transfer of pollen to a pistil of the same flower. Cross. — The offspring of any two flowers which have been cross-fertilized. GLOSSARY. 283 Cross-breed; half-breed; mongrel; variety-hybrid. — A cross between varieties of the same species. Cross-fertilization. — The action of pollen upon the pistil of another flower of the same species. Crossing. — The operation or practice of cross-pollinating. ( 'ross-pollination. — The conveyance of pollen to the stigma of another flower. Derivative- or derivation-hybrid; secondary-hybrid. — A hy- brid between hybrids, or between a hybrid and one of its parents. Fertilization ; fecundation ; impregnation. — The action of the pollen upon the ovules. Half-hybrid. — The product of across between a species and a variety of another species. Hybrid. — The offspring of crossed plants of different species. Hybridist)) ; hybridity. — The state, quality, or condition of being a hybrid. Hybridization. — The state or condition of being hybrid- ized, or the process or act of hybridizing. Hybridizing. — The operation or practice of crossing be- tween species. Individual cross. — The offspring of two crossed flowers on the same plant. Individual fertilization. — Fertilization between flowers upon the same plant. Mongrel. — A cross. Mule. — A sterile (seedless) hybrid. Pollination. — The conveyance of pollen from the anther to the stigma (page 252). The term cross is used to denote the offspring of any sexual union between plants, whether of different species or varieties, or even different flowers upon the same plant. It is a general term. And the word is 284 GLOSSARY. also sometimes used to denote the operation of per- forming- or bringing about the sexual union. There are different kinds of crosses. One of these is the hybrid. A hybrid is a cross between two species, as a plum and a peach, or a raspberry and a blackberry. There has lately been some objection urged against this term, because it is often impossible to define the limitations of species, — to tell where one species ends and another begins. And it is a fact that this diffi- culty exists, for plants which some botanists regard as mere varieties others regard as distinct species. But the term hybrid is no more inaccurate than the term species, upon which it rests; and, so long as men talk about species, so long have we an equal right to talk about hybrids. Here, as everywhere, terms are mere conveniences, and they seldom express the whole truth. In common speech the word hybrid is much misused. Crosses between varieties of one species are termed half-breeds or cross-breeds, and those between different flowers upon the same plant are called individual crosses. 3. Classification. Break. — A radical departure from the type. Ordinarily used in the sense of sport, but in its larger meaning it refers to the permanent appearance of apparently new or very pronounced characters in a species. Bud-variation. — Variation or departure from a type through the agency of buds (pages 28, 153). Bud-variety. — A variety resulting from bud-variation. Bud-sport. Family (Order in botany.) — A group of genera and species; as Cupuliferm, the Oak Family, Rosacea?, the Rose Family. Worm. — A minor variety, usually transient, produced by some local environment, GLOSSARY. 285 Genus (plural, genera). — A group or kind comprising a greater or less number of closely related species; as Acer, the maples, Fragaria, the strawberries. Race. — A fixed cultural variety; that is, a cultural va- riety which reproduces itself more or less uniformly from seeds. Seedling. — A plant growing directly from seed, without the intervention of grafts, layers, or cuttings. Seed-variation. — Variation or departure from a type through the agency of seeds. Seed-variety. — A variety resulting from seed- variation. Species (plural, species). — An indefinite term applied to all individuals of a certain kind which come or are supposed to come from a common parentage. A per- ennial succession of normal or natural similar indi- viduals perpetuated by means of seedage. "All the descendants from the same stock." — Gray. Sport. — A variety or variation which appears suddenly and unaccountably, either from seeds or buds; more properly restricted to varieties originating from buds, and so used in this book. Stock. — The parentage of a particular strain or variety. Strain. — A sub-variety, or individuals of a variety, which has been improved and bred under known conditions. Variation. — 1. The act or condition of varying or be- coming modified. 2. A transient variety, more or less incapable of being fixed or rendered permanent. Variety. — A form or series of forms of a species marked by characters of less permanence or less importance than are the species themselves. Wilding. — A wild individual from a cultivated species. INDEX Abortive varieties, 152. Abutilon, crosses, 220, 233, 250. Accident, 154. Acclimatization, 24, 20. Acer, bud-variety, 177. Achimenes, crosses, 245, 246, 247. Acorns, bud-variety, 177. Acquired characters, 14. Adult forms, 156. ^Egilops, crosses, 243. ./Esculus, bud-variety, 178. ^Esculus, crosses, 239. Agatlnea, bud-variety, 178. Ageratum, bud-varieties, 17S. Agrostemma Coeli-rosa, dwarfs, 144. Albinos, 148. Allut, Cazalis, 211. Almond, bud-variety, 179. Alnus, crosses, 226. Alopecurus, crosses, 230. Altitude and plants, 25. Amelioration, gradual, 50. Amygdalus, crosses, 239. Anagallis, crosses, 222, 223, 231. Anemone, crosses, 224, 229. Anemone, varieties, 179. Animal and plant contrasted, 5, 91. Annee, 141. Antagonistic features, 95. Anther, 254. Apple, Wealthy, 10S. Apples, bud-variation in, 1 IS, 175. Apples, hybrid, 06, 79, 111. Apples, races of, 90. Apples, variations in, 3, 27, 37, 99, 131. Apricot, bud-variety, 179. Aquilegia, crosses, 224, 229, 234, 239, 240, 250. Aralia, bud-variety, 180. Arthur, 103, 116. Arundo, variation, 176, 180. Asexual propagation, 7. Aspidistra, sport, 180. Aster, varieties, ISO. Atavism, 106. Atragene, 184. Atropa, crosses, 223. Azalea, bud-varieties, ISO. Bag for covering flowers, 272. Bamboos, variation, 176. Banana, varieties, 175. Bartel, T. C, 130. Barteldes, 140. Bean, bud-variation, 176. Beans, types of, 135. Beet, crosses, 56. Begonia, crosses, 224, 226, 227, 230, 231, 233, 234, 239, 245, 247. Begonia pollinations, 86. Bell-flower, 252. Berberis, crosses, 245. Bigness, variation in, 1^. Blackberry, crosses, 79, 111. Blackberry, introduction of. 129. Bletia, crosses, 222. Bohnhof, 80. Bornet, 232. Bouschet, Henri, 212. Brassica, crosses, Braun, Alexander, 17. Breaking the type, 19, 23, 93 Bruant, 113. Buckwheat, crosses, 56. Budd, Professor, 133. 'J4C, 239, 240, 241. 287 288 INDEX. Bud-variation, 6, 21, 28, 8T, 101, 118, 126, 153. Bugbane, 258. Burpee, 139. Buxus, bud-varieties, 181. Cabbage, crosses, 56. Cacti, crosses, 235. Calceolaria, crosses, 222, 229, 233, 234, 239, 245, 247. Calceolaria plantaginea, dwarfs, 144. Calliopsis tinctoria, dwarfs, 144. Callistephus horteusis, dwarfs, 145. Calyx, 253. Camellia, bud-varieties, 182. Camellia, crosses, 251. Canary-grass, crosses, 57. Cannas, 140. Capsella, crosses, 226, 231. Carman, 79. Carnation, 115. Carriere, 96, 116, 153. Caspary, 229. Cedars, 156. Cephalotaxus, sports, 1S3. Cereus, bud-variety, 184. Cereus, crosses, 226, 230, 233, 239. Cereus, night-blooming, 256. Change of seed, 28, 59, 116. Checking growth, 116. Cheiranthus, sport of, 1S5. Cherry, hybrid, 112. Cherry, sports of, 171. Chloranthic varieties, 153. Chlorosis, 149. Choice of variations, 31. Chrysanthemum carinatum, 100. Chrysanthemum, sports of, 158. Cimicifuga racemosa, 258. Cinchona, crosses, 225. Cirsium, crosses, 229, 233, 239. Cistus, crosses, 219, 222, 231, 234. Clematis, crosses, 224. Clematis, flowers, 262. Clematis, varieties, 1S4. Climate and variation, 24, 114, 146. Coleus, sports in, 120. Coloration, 148. Colors, modified by climate, 25. Conifera\ 156. Contradictory attributes, 9S. Convolvulus pollinations, 85. Coreopsis tinctoria, dwarfs, 144. Cornus, bud-varieties, 185. Corolla, 253. Cotyledon, crosses, 229. Crabs, hybrid, 66, 111. Crata-gus, crosses, 239. Crinum, crosses, 218, 225, 229, 232, 234. Cross-breeds, Focke on, 247. Cross, function of, 50. Cross, primary, 215. Crosses, characteristics of. 68. Crosses, Focke on, 215. Crossing a means, 107. Crossing and change of seed, 59. Crossing, limits of, 44. Crossing, philosophy of, 39. Crossing, rule for, 109. Crozy, 113, 140. Cucumber pollinations, 85. Cucumis, crosses, 222. Cucurbita Pepo, 75, 84. Cucurbitacea j , crosses, 46, 5S, 74, 82, 229, 230, 235. Cultivation, philosophy of, 22. Currant, sports of, 173. (See Kibes.) Cypripedium, 254. Cypripedium, crosses, 218, 227, 229, 246. Cytisus Adami, 185. Dactylis, bud-variety, 185. Darwin, 17, 23, 32, 42, 47, 51, 54, 56, 60, 63, 69, 72, 84, 87, 117, 119, 121, 176, 228. Dating back, 106. Datura, crosses, 222, 223, 224, 226, 228, 234, 239, 240, 249, 250. Decaisne, 229. De Candollo, 150, 178. Derivative crosses, 238. Dewberry crosses, 79, 111. Dewberry, introduction of, 129. Dianthus Chinensis, dwarfs, 145. Dianthus. crosses, 21&, 219, 223, 226, 227, 229, 233, 234, 237, 239, 240, 243, 244, 245, 246, 25a. Dianthus seraperflorens, 156. Dichroism, 154. INDEX. 289 Digitalis, crosses, 21s. 219, 221, '-".'3, 226, 2-21, 229, 230, 232, 237. Dimorphism, 154. Dioecious plants, 260. Divergence of character, 23. Division of labor. 42. Doubleness in hybrids, 237. Doubleness of flowers, 149. Dracaena, variation, 1 7 * ► . Duval, M., 16T. Dwarfing, 25, 114, 143. Early varieties, 140. Echinocactus, sports, 186. Eckford, 113. Egg-plant, crosses. . r >7, 74. Egg-plant pollinations, 65. Egg-plants, variation in, 95. Egypt, plagues of, 40. Ela?agnus, bud-varieties, 186. Emasculation, 265. Envelopes, floral, 252. Environment and variation, 12. Epilobium, crosses, 218, 222, 229, 230, 233. Equilibrium of organisms, 20, 61. Erica, crosses, 229, 2:;::. 239, 24i>, 2 til, 247. Essential organs, 200. Euonymus Japonicus, 156, 1S6. Fa-us, fern-leaved, 187. Fall sowing, 11.".. 143. Ferns, crossing-, 277. Fertility of soil. 18, 22. Finis, forms of, 188. Filament, 254. Fittest, survival of, 32, 39. Fixation of plants, 31. Flavor, modified by climate, 25. Flon, M., 156. Flowerless plants, crossing, 277. Focke, 6S, 81, 108, 215. Fontanesia, sport, 188. Food supply, 1<», 11C. Fortuitous variation, 0. Fragaria, crosses, 229. Fiaxinus, bud-varieties, 1S8. Fromont, 158. Fuchsia, crosses, 229, 233, 239, 247. Fuchsia flowers, 273, 274, 275. Function of the Cross, 50. Fusain, 150. Galium, crosses, 222. Gardenia, bud-variety, 189. Gartner, 216, 21s, 219, 220, 228, 233, 235, 239, 240, 24:'., 244, 248, 249, 250. Gazania rigens, 140. Genera, monotypic, 97. Gesneraeea?, crosses, 227, 229, 246. Geum, crosses, 219, 229,239. Giant forms, 145. Gibb, Charles, 133. Gideon, Peter M., 108. Gillyflower, bud-variety, 1S9. Giraud, Desire, 100. Gladiolus, crosses, 226, 229, 234, 245, 240. Gleditschia triacanthos, 207. Glossary, 2S2. Golf, 103. Gordon, 229, 241,244. Gourd, crosses, 58, 74, 82. Grape, bud-varieties, 174, 210. Grapes, hybrid, 00, 78, 110, 111. Gray, Asa, 33, 178. Greenhouses, produce variation, 115. Halloek, V. II., &Son, 124. I lardy varieties, 145. flartogia Capensis, 192. Iledera, forms of, 189. Ilelianthemum, crosses, 223, 220. Helichrysura bracteatum, dwarfs, 144. Henderson, 138. Herbert, 24S. Hibiscus, bud-varieties, 190. Hibiscus, crosses, 225, 250, Hibiscus Syriacus, 257. Hieracium," crosses, 220, 233, 234, 237, 239, 250. Hippeastrum, crosses, 226, 229, 234, 239, 245, 246, 247. Holly, sports, 191. Horse-chestnut, bud-variety, 178. Husk-tomato, Oo, 85. Hyacinth, forms, 190. Hybrids, characters of, 68, 215. Hybrids, Focke on, 215. 290 INDEX, Hybrids, multiple, 246, 247. Hybrids, rarity of, 53. Hybrids, seven-eighths, 243. Hybrids, three-fourths, 243. Hybrids, triple, 244. Hydrangea, 146, 191. Hymenocallis, crosses, 230. Iberis umbellate, dwarfs, 145. Ignotum tomato, 123. Ilex, bud-varieties, 101. Impatiens Balsamina, dwarfs, 145. In-breeding, 72. Indeterminate varieties, 87. Individuality, causes of, 8. Individuality, fact of, 2. Instruments for pollination, 270. Ipomoeas, colors of seeds, 104. Iris, bud-variety, 192. Isolation of the plant, 22. Isoloma, crosses, 226, 227, 231, 233, 246. Ivy, forms of, 1S9. Jamain, M., 163. Jobert, M., 179. Joigneaux, M., 204. Juniperus, bud-varieties, 192. Klotzsch, 226, 229. Knight, Thomas Andrew, 17, 54, 227, 229. Kohl-rabi, 80. KOlreuter, 54, 73, 216, 217, 219, 228, 229, 243, 244, 250. Kumerle, W. J., 140. Kuntze, 225. Labor, division of, 42, 48. Lachaume, M., 166. Lactuca, crosses, 238. Ladle for pollinating, 271. Lamium, bud-variety, 192. Lamium, crosses, 239, 241. Lantana, crosses, 222. Large-flowered varieties, 145. Late varieties, 146. Latitude and plants, 25. Laurocerasus, sports, 192. Lavatera, crosses, 234, 239, 243. Leanness, 25. Lecoq, 227. Lemoine, 113. Lens for pollinating. 270. Leptosiphon densirlorus, dwarfs, 144. Lettuce, crosses, 56. Ligustrum, sports, 103. Lilac, bud-varieties, 193. Lily, white, 253. Lima beans, 138. Limits of crossing, 44. Linaria, crosses, 222, 227, 238, 239, 241. Lindley, 68. Links, missing, 41, 48. Linnaeus, SI, 152. Linum, crosses, 223. Lobelia, crosses, 219, 222, 232, 233, 234, 235, 245. Luffa, crosses, 232. Lupines, heredity in, 106. Lychnis Coeli-rosa, dwarfs, 144. Lychnis, crosses, 240, 241. Lycium, crosses, 218, 226. Lycopods, crossing, 277. Maize, crosses, 56. Malle, Dureau de la, 175. Mamillaria, sports, 194. Maple, Wier's, 109. Meadow, plants in, 23. Medicago, crosses, 220, 233, 240. Melandrium, crosses, 220, 222, 223, 225, 231, 233, 240, 244, 245. Mendel, 239. Mentha, bud-variety, 194. Mentha, crosses, 229, 241. Mersereau, 131. Mimulus, crosses, 222, 251. Mirabilis, crosses, 222, 226, 227, 22S, 234, 241, 250, 251. Mirabilis pollinations, 85. Missing links, 41. Mixing in the hill, 118, 201. Molinia, bud-variety, 104. Momecious plants, 259. Monotypic genera, 97. Moore, Jacob, 110. Morning-glory, 54. Morong, Dr. Thomas, 00. INDEX. 291 Morren, 149. Mourriere, M., 1T5. Mulberry, Teas', 109. Multiple hybrids, 246, 247. Munson, Professor, 58. Munson, T. V., 79. 111. Musa, bud-variety, 194. Muskmelon pollination, B5. Myrtle, bud-variety, 195. Nanz it Neuner, lTu. Narcissus, crosses, 220, 225, 236. Natal variations, 15. Natural selection, 32, 51. Naudin, 216, 229. 248. Nectarine, origin of, 118, 173. Nepenthes, crosses, 220. Nicotiana, crosses, 72, 217, 219, 222, 224, 225, 226, 227, 229, 233, 234, 237, 238, 239, 241, 243, 244, 245, 246, 250. Nicotiana pollinations, 85, 86. Nuphar, crosses, 226. Nymphsea, crosses, 218, 225, 227, 231, 234, 250. Odoriferous varieties, 147. Oenothera, crosses, 219. 239. (Enothera Drummondii, dwarfs, 144. Oger, Pierre, 166. Olea ilicifolia, 195. Opuntia, bud-variety, 195. Orange, bud-variety, 195. Orchidae, crosses, 229, 233, 235. Orchids, hybrids, 79. Orontiuin, sport, 195. Osmanthus, sport, 195. Ovary, 255, 263. Palmer, Asa, 139. Pansy, 146. Papaver, crosses, 218, 224, 226, 229, 231, 237. Papaver, forms of, 151. Pare, M., 160. Parents, influence of, 81, 217. Passitlora, crosses, 220, 226, 229, 235, 250. Peach, bud-variation in, IIS, 173, 196. Peach, hybrids, 47. Peaches, races of, 91. Pear, bud-varieties. 174, 197. Pears, hybrid, 66, 79, 111. Pears, variation in, 99. Peas, viney, 16. Pelargonium, crosses, 218, 220, 229, 230, 233, 234, 237, 245, 246, 247. Pelargonium, sports in, 198. Peloric varieties, 152. Pentstemon, crosses, 241. Pepino pollinations. S6. Pepino, variation in, 95. Pepper, red, pollination, 85. Peppers, variation in, 96. Persica, 196. Petal, 253. Petunia, crosses, 21S, 240, 241, 251. Petunia pollinations, S5, 86. Phalaris, sports, 198. Phaseolus, crosses, 223, 23S, 241. Phlox, bud-varieties, 199. Phragmites, bud-varieties, 199. Physalis, 60. Physalis, variation in, 96. Picea, bud-variety, 199. Pink, 156, 160. Pinus, bud-varieties, 199. Pinus, crosses, 226. Pistil, 255. Pisum, crosses, 223, 238. Pittosporum, sport, 200. Plant-breeding, 91. Pliny, 131. Plum, hybrids, 47, 112. Plum, sports of, 172. Plums, Japanese, 27. Podocarpus, 155, 183. Pollen, 254, 264. Pollinating kit, 277, 278. Pollination, 252. Pollination, uncertainties of, 83. Polymorphous varieties, 153. Polytypic genera, 97. Populus, bud-variety, 200. Populus, crosses, 222. Position, advantage of r 22. Post-natal variations, 15. Potamogeton, crosses, 227. Potato, 37, 117. Potato and tomato, 95. Potato, bud-varieties, 201, 209. 29: INDEX. Potato, seedlessness, 99. Precocious varieties, 146. Prin ui la, crosses, 239, 240, 241. Progeny of crosses, 237. Proliferous varieties, 150. Propagation, asexual, 7. Pruning, 28. Prunus, bud-variety, 205. Prunus, crosses, 229, 239. Pumpkin, crossing, 4('», 58, 74, 82. Pyrus, crosses, 229, 239. Quercus, crosses, 226, 229, 239. Quince, pollinated, 276. Paces in fruits, 90. Radish pollinations, 85. Eaphanus Kaphanistrum, 116, 231. Raphanus sativus, 231. Raspberry, flowering, 260. Raspberry, hybrids, 79, 111. Representative species, 66. Retinosporas, 156. Rheum, bud-variety, 206. Rhododendron, crosses, 145, 218, 222, 225, 226, 229, 230, 239, 245, 246, 247. Ribes, bud-varieties, 206. Ribes, crosses, 230, 2:'.:;. Robinia, bud-varieties, 206. Rogue, 89, 127. Rosa, 161. Rose, bud-varieties in, US, 161, 207. Roses, crosses, 233, 247. Rubus, crosses, 227, 230, 232, 234. Rubus odoratus, 260. Running out of varieties, 36, 125. Russia, fruits from, 27, 90, "133. Rye, hybrids, 79. Salix, bud-varietv, 208. Salix, crosses, 219, 225, 226, 227, 229, 231, 234, 239, 244, 245, 246, 247. Salter, 119. Salvia, crosses, 223. Sambucus, sports, 208. Satyrium hircinum, 14S. Scabiosa atropurpurea, dwarfs, 145. Scalpel for pollinating, 270. Schizanthus retusus, dwarfs. 145. Scissors for pollinating, 270. Secondary crosses, 238. Seed, change of, 28, 59. Seeds, colors of, 104. Seeds, early, 147. Seeds, immature, 103, 147. Seeds, large and small, 101. Selection and progress, 120, 122, 127 Selection, natural, 32, 51. Self-fertilization, effects of, 54. Senecio cruentus, dwarfs, 144. Sepal. 253. Seven-eighths hybrids, 243. Sex and variation, 11, 43. Silene, crosses, 234. Sinningia, crosses, 231. Solanum, bud-varieties, 209. Solatium, variations in, 95. Spencer, 01. Spiraea, bud-varieties, 209. Sports, 22, 28, 37, 153. Sprengel, 54. Squash, crosses, 58, 74, 82. Squash flowers, 261. Squash, Hubbard, 46. Stamens, 254. Stigma, 255. Strawberry, Wilson, 125. Struggle for life, 20, 29, 39. Sturtevant, 103. Style. 255. Sugar-cane, varieties, 175. Survival of the fittest, 32, 39. Swamping effects of inter-crossing, 46. Sweet, 247. Svmphoricarpus, sport, 209. Symphytum, bud-varieties, 209. Synchronistic variations, 117. Tagetes, dwarfs, 145. Teas, 109. Teleology of hybrids, 236. Thinning. •_':!. Three-fourths hybrids, 243. Thuyopsis, sport, 209. Tillage and food supply, 17, 22. Timbal-Lagrave, 217. Toad-flax, 152. Tobacco flowers, 266. Tobacco pollinations, S6. Tomato and potato, 95. ENDEX. 293 Tomato, crosses, 58. Tomato, Lgltotum, 123. Tomato, pollinated, '-'Til. Tomato pollinations, 85. Tomato, Trophy, .">T. Tomato, variation in, 9S. Tomatoes, breeding, 103. Tragopogon, crosses, 238. Triple hybrids, 244. Tropaeolum, crosses, 222, 226. rimus, bud-variety, 209. rimus, crosses, 226. Variability, variation in, 25. Variation and environment, 12. Variation caused l.y sex, 11, 48. Variation, fortuitous. 9. Variation, philosophy of, 1. Variations, choice of, 31. Variations, fixation of, 31. Variations, natal and post-natal. 15. Variations, origin of, 8, 41. [157. Variegation, perpetuating, 120, I4',t, Varieties, running out, 36, 125. Variety, what is a, 35. Verbascum. cros>cs. 219, 221, 222, 223, 224, 226, 229, 235, 239, 250. Verdier, Victor, 167. Verlot, 121, 143. Veronica, crosses. 283, 239. Viburnum, sports, 209. Vilmorin, 152. Vilmorin, Henri L. de, 100, 142. Vilmorin, Louis Leveque de, 106. Vine, bud-varieties, 1 74, 210. Viola, bud-variety, 210. Viola, crosses, 22!>, 231. Vitis. crosses, 229, 239, 245, 246. Walker. Ernest, 120, 169. Wallace, 47. 60. 67. Watermelon pollination, 85. Wcismann, 13, 14. Wheat, hybrids, 79. Wichura, 216, 246. Wier, 1>. P.., 109. Wigandia, sport, 215. Willdenow, 152. Yucca, variation, 176. Zinnia, crosses, 241. Zinnia, flowers, 268. raorexrr uiJMJtr N. C. State College NEW EDITION. The Horticulturist's Rule-Book. A COMPENDIUM OF USEFUL INFORMATION FOR FRUIT- GROWERS, TRUCK-GARDENERS, FLORISTS, AND OTHERS. By L. H. BAILEY, Professor of Horticulture in the Cornell University. Third Edition, Thoroughly Revised and Recast, with Many Additions. i2mo. 302 pages. Limp Cloth. 75 cents. This volume is the only attempt ever made in this country to codify and condense all the scattered rules, practices, recipes, figures, and histories relating to horticul- tural practice, in its broadest sense. It is much condensed, so that its three hundred pages comprise many thousand facts, the greater part of which the busy man would never possess if he were obliged to search them out in the voluminous literature of recent years. All the approved methods of fighting insects and plant diseases used and discovered by all the experiment stations are set forth in shape for instant refer- ence. This feature alone is worth the making of the book. Amongst the additions to the volume, in the present edition, are the following : A chapter upon " Greenhouse and Window-garden Work and Estimates," comprising full estimates and tables of heating glass-houses, lists of plants for forcing, for cut flowers, for window-gardens, aquaria, and the like, with temperatures at which many plants are grown, directions for making potting-earth and of caring for plants, etc.; a chapter on " Literature," giving classified and priced lists of the leading cur- rent books and periodicals on American horticulture, and directories of officers of whom the bulletins of the various experiment stations may be obtained; lists of self- fertile and self-sterile fruits; a full account of the methods of predicting frosts and of averting their injuries; a discussion of the aims and methods of phenology, or the record of climate in the blooming and leafing of trees; the rules of nomenclature adopted by botanists and by various horticultural societies; score-cards and scales of points for judging various fruits, vegetables, and flowers; a full statement of the metric system, and tables of foreign money. THE MACMILLAN COMPANY, 66 FIFTH AVENUE, NEW YORK. The Rural Science Series. NOW READY. The Soil. By Franklin H. King, Professor of Agricultural Physics, University of Wisconsin. 161110. Cloth. 75 cents. IN PRESS. The Spraying of Plants. By Ernest G. Lodeman, Cornell Uni- versity. IN PREPARATION. The Apple in North America. By L. H. Bailey, Editor of the Series. The Fertility of the Land. By I. P. Roberts, of Cornell University. Milk and its Products. By H. H. Wing. Under the editorship of Professor L. H. Bailey of Cornell University, Mac- millan & Co. purpose issuing a series of books upon agricultural subjects to be known as the Rural Science Series. Professor F. H. King, of the University of Wisconsin, has written upon The Soil, treating the subject from the new attitude which considers it as a scene of life rather than as a mere mechanical or chemical mixture. The physics of the soil are fully considered and the physical effects of fertilizers, drainage, and cultivation are discussed, as well as the adaptation of dif- ferent types of soils to various crops. Professor I. P. Roberts, of Cornell University, will write upon The Fertility of the Land. This volume, while entirely indepen- dent of that of Professor King, will carry the subject directly into the practice of the field, giving a full discussion of the philosophy of plowing, cultivating, and the like, and an account of the best methods of maintaining and increasing the productivity of the land. The editor will contribute a monograph upon the cultivation of The Apple in North America, with a discussion of its evolution and the difficulties which now confront the apple-grower. The Spraying of Plants is treated by E. G. Lodeman of Cornell, in a comprehensive account of the origin and philosophy of the modern means of controlling insect and fungous troubles, and the application of these meth- ods to the leading crops. Some of the other volumes to be arranged for are : Forestry. Grape Culture. Planting Manual. Landscape Gardening. Small Fruits. Plant Life. Rural Economics. Etc.. etc. THE MACMILLAN COMPANY, 66 FIFTH AVENUE, NEW YORK. WORKS BY L. H. BAILEY. Professor of Horticulture in Cornell University. Talks Afield: About Plants and the Science of Plants, pp. 173. Illustrated. Field Notes on Apple Culture, pp. 90. Illustrated. The Horticulturist's Rule-Book : A Compendium of Useful Information for Fruit-Growers, Truck-Gardeners, Florists, and Others. Fourth edition, pp. 312. The Nursery-Book: A Complete Guide to the Multiplication and Pollination of Plants, pp. 304. Illustrated. American Grape Training, pp.95. Illustrated. Annals of Horticulture in North America for the Year 1889 : A Witness of Passing Events and a Record of Progress, pp. 249. Illustrated. Annals of Horticulture for 1890. Annals of Horticulture for 1891. Annals of Horticulture for 1892. Annals of Horticulture for 1893 : Comprising an Account of the Horticulture of the Columbian Exposition, pp. 179. Illustrated. Gray's Field, Forest, and Garden Botany: A Simple Introduc- tion to the Common Plants of the United States East of the 100th Meridian, Both Wild and Cultivated. Revised and extended by L. H. Bailey, Editor of The Rural Science Series of agricultural and horticultural books. Plant-Breeding: Being Five Lectures upon the Amelioration of Domestic Plants, pp. 293. Illustrated. THE MACMILLAN COMPANY, 66 FIFTH AVENUE, NEW YORK. pp. 312. Illustrated. pp. 415. Illustrated, pp. 387. Illustrated ii t d ••/*>