Journal of Heredity 5: 273-280 (1914)
One of Earliest Crops Cultivated by Man—Perhaps Valued at First Only for its
Roots—Doubt as to Time of its Introduction to America—Prehistoric
and Allied Forms—Irregularities of its Behavior in Cultivation.

(Paul Popenoe)

1) Annual Report Smithsonian Inst., p. 481, Washington, D. C., 1903.
2) Beccari, Odoardo, Nelle Foreste di Borneo, p. 611. Firenze, 1902.

THERE seems little reason to doubt that the banana was one of the first foods of man, and that it was one of the first plants cultivated. "Wild bananas and their botanical relatives," says O. F. Cook,1 "are natives of the rocky slopes of mountainous regions of the moist tropics, where shrubs and trees prevent the growth of ordinary herbaceous vegetation." It is probably in a similar region that the first appearance of Man is to be looked for. "Everything leads one to believe," as Beccari2 says, "that the principal cultivated fruits originated in the region where man first acquired a high grade of civilization." Primordial man of the tropics was undoubtedly an agriculturist rather than a live-stock breeder. He lived on the resources most readily furnished him by nature, and among these, few would be more readily available than the banana. It is permissible, then, to suppose that the banana was one of the first fruits which attracted his attention; that he soon brought it under cultivation, and that he at once began to submit it to that long process of improvement which has continued for some hundreds of thousands, perhaps, of years, and is more active today than ever before.

If Man appeared in the Indo-Malayan region, as is widely believed at present, it seems natural to seek for the origin of the banana in the same region; and such a location for it is accepted by most botanists. This primitive banana probably did not differ widely from the wild bananas found today in many parts of the tropics, although none of the latter can be confidently pointed out as representing the ancestral type. Beccari, indeed, considers that all the wild forms known today are merely cultivated forms which have escaped from cultivation at some time in the past. He found in Borneo four new species which grew only in regions deforested by man. Whence were they brought he asked himself, and was obliged to conclude, after a survey of the whole problem, that probably each region develops its own well characterized species of Musa—a conclusion which finds support in the fact that no species yet known has a very wide geographical distribution. At present the genus seems to be dependent on man for its possibilities of development: it can not make its way in the primitive forest, he concludes. It is one of the many crops which have been so changed by man to meet his own needs that they are no longer able to hold their own in the free competition of nature.


The original form of banana must have been of little value as a fruit. Cook has therefore concluded that it was first a root crop, the roots even yet being used by the natives of some regions, while the tender heart was doubtless also an article of food, as it is today in Abyssinia. Cultivated for its roots, the banana began to produce better fruits, by chance, or as a result of asexual propagation, and at a very early day must have become more prized for the latter than for the former.

"The wild varieties are almost wholly seeds," Beccari observes, "but what pulp exists is sweet and agreeable. It therefore only requires some agent to inhibit the growth of seeds and promote that of pulp to produce good bananas. Effective causes are sterility produced by hybridization, and improvement by asexual reproduction." Both of these means may have been used by the prehistoric plant-breeders of the tropics. Cross-pollination between different species would easily take place, and would result in at least partial sterility of the product. These hybrids, asexually propagated either by man or by nature, would retain their sterility, and a "horticultural variety" would be established. Beccari's own idea is that all the bananas of today are, in fact, the results of hybridization of various original wild forms which have now disappeared. This in itself would be sufficient to explain the seedless condition of the fruit of commerce; while the numerous seed-bearing species found wild at present are accounted for by Beccari's hypothesis that they represent the product of one of the normally sterile forms pollinated from some species sufficiently distinct to cause the production of normal seed.


3) Fawcett, William. The Banana. London, 1913.
4) Ber. Bot. Ges. XXX, 686.
5) I owe to A. B. Stout, Director of the Laboratories at the New York Botanic Garden, the following reference: "G. Tischler, Archiv, Zellforschung 5:622-670, 1910. Tischler investigated three races of Musa sapientum and found that the chromosome numbers were respectively 8, 16 and 24, and that the volume of the nuclei was proportionately 1:2:3. He found irregularities in the development of pollen. Some chromosomes lagged behind and formed extra nuclei. Often eight pollen grains are formed from a single-mother cell. I believe this is all the cytological work that has ever been done on any of the bananas."
6) Philippine Agric. Review, V, 383, 1912.

This hypothesis, although somewhat unusual, is given color by recent work in Jamaica, described by Fawcett.3 Experiments in pollenizing the ordinary, sterile varieties, at the Hope Gardens, were unsuccessful until pollen from the distinct but equally sterile red banana (var. rubra) was used: the normally seedless commercial bananas then set a full complement of seed. He quotes a similar observation from A. d'Angremond4: "Most of the pollen of the Jamaican and Apple bananas is sterile, and only a few of the ovules in those plants have an embryo sac. However, dusting the ovaries of these cultivated fruit plants with pollen of Musa basjoo and M. ornata [two wild species] was sufficient to produce seeds."5

Seeds may be produced in an ordinarily sterile variety as a result of environmental conditions, if there is any basis of fact in the story given to O. W. Barrett6 by a Porto Rican native, who advised: "Get a stool of bananas growing rapidly in shallow soil by the addition of artificial fertilizers; let one bunch of fruits set; but before that ripens, cut down all but one of the stems in the clump. The remaining shoot, 'thinking it has but one more chance to perpetuate its kind before being killed,' on account of the tremendous shock to the more or less connected stem bases in the clump, at once produces a small bunch of somewhat abnormal fruits, some of which will contain seeds." "As a matter of fact." Barrett adds, ''it is a usual thing  to find seeds in the commonest of the Philippine bananas, the Saba.''

The origin of the present seedless varieties is explained by many writers as a matter of simple selection, rather than of hybridization. The knowledge which we are gradually acquiring of the results of plant-hybridization, however, makes it seem plausible that some cross was the starting point from which the tropical native began his process of selection. The little knowledge we have of the agricultural skill of primitive man gives abundant reason to believe him intelligent enough to propagate choice strains of his staple crops by offshoots. In the banana Nature herself showed him the way: for in addition to seeds, which must always have been the normal method of reproduction, the banana could propagate itself rapidly by suckers—unless the primitive types were very different from those we know today. Around the base of the plant numerous small suckers are thrown up; these, it is believed, finally separate themselves from the parent, by the formation of a layer of abscission-cells, and roll down hill (when the plant is growing on a slope) until their progress is arrested by some obstacle; then they take root and reproduce their parent form.

The fruit is concealed by huge bracts, part of which have been raised to show the "fingers." This was probably the original habit of the plant, although these bracts have disappeared in the cultivated forms, so commonly seen in fruit stores. The banana here photographed (by the Bureau of Agriculture, Philippine Islands) is known as Virgen; it is possibly Musa glauca, or perhaps a new species not hitherto described; natives of the Philippines propagate it, but only by seed. (Fig. 16.)


To sum up, we find the banana established as an important crop as far back as we can see. Beccari, indeed, does not hesitate to suggest its cultivation in the Pliocene epoch, although there is by no means agreement of paleon­tologists as to whether Man existed as a distinct species in that epoch. In the Miocene, Beccari recalls, we find a wide variety of strange forms; in the Pliocene, we meet forms similar to those which we know today. ''It is only in the Pliocene that we find forms of mammals identical with those of the present. ... It is possible that it was in that epoch that man, clearly established as a species with the characteristics he now shows, had begun to domesticate plants and animals," and if so, we must certainly put the cultivation of the banana in that epoch: first, perhaps, as a root crop, and then as a fruit crop, when man seized and perpetuated in the plant the variations favorable to his needs, which chance inter-specific hybridization may have offered.

From the Indo-Malayan region, according to the generally received account. man must have carried the banana on his migrations, both eastward to the islands of the Pacific Ocean, and perhaps to America; and westward to India, the Mediterranean region, and finally on to America. The plant is admirably adapted for transportation over long distances, because its suckers can be dried and carried without difficulty in that condition for several months, to take root at once when placed in the deep, rich soil which they love.

Of the eastward travels of the banana from the Indo-Malayan region we have little knowledge, but its westward travels are interestingly shown by its names, with an occasional written record.


7) Caius Plinius, Historia Naturalis, XII, 12 (6), Rome, A. D. 67.
8) This can not be taken too literally: yet Stanley (Darkest Africa, 1, 252) mentions specimens of plantains 22 in. long, 2 1/2 in. diameter, nearly 8 in. around.

Pliny7 is commonly held to be the first writer to describe the banana, although his account, at second hand, is inaccurate enough to have caused some doubt whether he was describing the banana or not. ''There is another tree in India," he writes, "of still larger size, and still more remarkable for the size and sweetness of its fruits, upon which the sages (Brahmins) of India live. The leaf of this tree resembles, in shape, the wing of a bird, being three cubits in length and two in width. It puts forth its fruit from the bark, a fruit remarkable for the sweetness of its juice, a single one containing sufficient to satisfy four persons.8 The name of this tree is pala, and of its fruit ariena. They are found in the greatest abundance in the country of the Sydraci, a territory which forms the extreme limit of the expedition of Alexander." The name pala is said still to be found as a vernacular name of the fruit in India, while his remark as to its being the food of the sages has given the specific name to the ordinary cultivated banana—Musa sapientum,  ''the Musa of the wise men.''

9) A number of authorities have given credit to the ridiculous story that the genus was named after Musa, a physician of the Emperor Augustus. This etymology is no better than the one put forth by a certain unnamed Franciscan friar who declared "chiamasi questo gentil frutto Musa, percioche le (nine Grecian) Muse usano tal cibo." As Dottor Nicolo Monardes (Delle Cose che Vengono portate dall' Indie, p. 206. Venetia 1582) justly remarks, "E cosa da muouer le risa." A legend that the banana was the Tree of Knowledge of Good and Evil of the Garden in Eden led to the name of Apple of Paradise or Adam's Fig.
10) C. C. Torrey, professor of Semitic languages at Yale University, kindly aided me in tracing the travels of this word.

As to the generic name Musa,9 we may conjecture that it represents the name by which the fruit was received in India from its more southerly tropical home. It comes to us from the Sanscrit Moca, through the Arabic and Latin; a course that prettily illustrates the gradual dispersal of the fruit itself from India through Persia, Arabia, and Syria to the Mediterranean. The old Persian form, which represented the first transition from the Sanskrit, is not known to us. It was probably taken from Persia by the Aramaic, whose form for it would have been Moza, and the Arabs borrowed it from the latter language, as Mauz or Muz. It was spelled Musa by the Romans, and one or the other of these two forms—Muz and Musa—was the accepted English name until comparatively recent times.10 In 1578, for instance, Lyte writes in the Dodoens (VI, 38, 704) "of musa or mose tree. The Mose tree leaves be so great and large that one may easily wrap a childe in them." Sixteenth century writers commonly call the fruit Apples of Paradise or Adam's Fig. The name banana gradually came into use in that century; it is the vernacular name given to the fruit by a tribe in the African Kongo. De Orta mentions it in 1563, while Hartwell (Pigafetta's Congo (1597) in Coll. Travels (1746) II, 553), says, "Other fruits there are, termed Banana, which we verily think to be the Muses of Egypt and Suria."

11) Oviedo, Hist. Gen. y Nat. de las Indias, Lib. XXX, cap. I, pub. in 1535.

Thus the fruit, carrying with it the name which may have come all the way from its first station in the Indo-Malayan region, reached the Mediterranean and—after the colonization of those islands—the Canaries. From the Grand Canary it was introduced to the New World in 1516, according to the very definite statement of Captain Gonzalo Fernandez de Oviedo y Valdes,11 who heard the story "from many people." He ascribes the introduction to Hispaniola (Santo Domingo) to the "reverendo padre fray Thomas de Berlanga, de la Orden de los Predicadores"; "and from here," he continues, "it has spread to the other villages of the island, and to all the other islands populated by Christians, and has been carried to the mainland; and in every region where it has been established, it has yielded excellent results."

Although the seeds are numerous and fully developed, they are much fewer in proportion to the amount of pulp than in the African species shown in the frontispiece. This fruit is from Cavite, Philippine Islands, where it is colloquially known as Alinsanay. Botanically, it is probably an undescribed species. On Beccari's hypothesis, it is to be regarded as a once cultivated form, that escaped from cultivation long ago and has regained fertile seeds through cross-pollination with some distantly related type. Photograph from the Bureau of Agriculture. P. I. (Fig. 17.)


12) Names given by natives to two of the principal forms of the cultivated Musa. The word platano is Spanish, from the Latin platanus, and is the origin of the English plantain, by which bananas fit only for cooking are generally known. Unnatural as such an etymology may seem, it appears to be a fact that the name is due to some confusion with the oriental plane tree (Latin, platanus), wrongly called sycamore in the western United States.

This circumstantial account has always failed to satisfy a certain number of botanists, whose belief that the banana was found here long before the arrival of Columbus is based partly on tradition, more on the belief that it could never have spread so rapidly in the years following the conquest, as to account for its abundance in the many localities where it is reported by early writers; partly on the large number of distinct varieties to be found in the tropical parts of America, and partly on the finding of leaves resembling those of the banana, in pre-Columbian graves in South America. The first consideration seems to have weighed heavily with von Humboldt, who did not hesitate to declare the fruit a native of America, saying, "It is a constant tradition, in Mexico and on all the mainland, that the platano arton and the Domenico12 were cultivated there long before Europeans arrived." Most of the botanists who have studied the subject have not considered tradition a sufficient ground for judgment: De Candolle contented himself with a verdict for "a prodigious antiquity of cultivation; in consequence, a primitive existence in Asia and a diffusion synchronous with that of the races of mankind, or even earlier."


As to the evidence afforded by the exhumation of leaves, those who uphold the Asiatic origin of the banana contend that knowledge that these leaves were really Musa is lacking, and that they were more probably leaves of some such plant as Heliconia, a South American relative. O. F. Cook has brought the case prominently forward during the last few years by championing the theory of American origin, but the majority of writers on the subject are still on the other side.

13) Annual Rep. U. S. Geol. and Geog. Survey of the Territories, p. 418, Washington, D. C., 1873.

Whether the Musa, as we know it today, was actually cultivated by the natives of the Spanish Main when Columbus found them, there seems reason to believe that it or a closely related plant existed on this continent several millions of years ago. Researches of paleontologists in North America have resulted in the identification of a genus which has been named Musophyllum, and bears extraordinary resemblance to the bananas, although of course there is not sufficient evidence available to decide the exact degree of relationship. The best known of these finds in the deposits of the Eocene epoch are from the vicinity of the Yellowstone National Park. Leo Lesquereux, who described13 Musophyllum complicatum  as a new species in 1873, writes:

"Though the specimens representing this species are very numerous and very large, I could not obtain one showing exactly the size and form of these leaves. They appear either folded around a thick stem, from which they diverge, or on both sides of a thick rachis, extending along it like two wings, two or three centimeters wide on each side. From the fact that large specimens are covered by fragments of these leaves crushed and folded upon one another, without any trace of middle nerves or peduncles, the leaves must have been of great size. Their substance is not very thin. The surface is perplace covered with an epidermis which shows the veins are crossed by veinlets at right angles. When the epidermis is destroyed, this character is not observable, it may, therefore, result of a wrinkling of the epidermis. The species is related to Musa Bilinica Ett., differing, however, by essential characteristics."


14) Monographs of the U. S. Geol. Survey, XXXII, pt. II,
p. 686. Washington, 1899. Plate LXXXIII, fig. 1.

Fifteen years later F. H. Knowlton14 collected the same species on the northeast side of Crescent Hill, opposite a small pond, in the Yellowstone Park, and wrote:

"This species was described by Lesquereux from a 'shale over a thin bed of coal, eight miles southeast of Green River Station, Wyoming', in what he at first regarded as the Washaki group, but which he later decided was the true Green River Group. This locality has not since been visited, and in fact can not now be satisfactorily located. It is more than probable, however, that the former determination of the horizon is correct.

"So far as I know, this is the second time this species has ever been found. It is represented by five or six fairly well preserved specimens, which agree perfectly with  Lesquereux's descriptions and figures.

"On one of the specimens there are a number of thick stems or stipes. They are longitudinally striate, as described by Lesquereux, and bear only fragments of the leaves preserved. In the specimen figured we have a narrow leaf preserved almost entirely. It is about five centimeters broad and seven centimeters long, as preserved, with perfectly entire margins. In still another specimen the stipe, with portions of lamina attached, is fully 20 cm. long. There is no evidence from these specimens of the leaves having been as broad as described in some of the original specimens, but Lesquereux also speaks of narrow leaved forms."

15) Another striking instance is the well known sassafras (S. officinale) of the eastern United States. The only other species of its genus is found in the interior of China.

In addition to this remnant of a remote epoch, we still have in America members of the natural order Scitaminaceae, to which the tribe Museae belongs. The most conspicuous is the Traveler's Palm (Ravenala guianensis), representing an interesting genus which is known to most people only by the other of its two species, the larger Ravenala madagascariensis. The presence of these two closely related species, one confined to the northern part of South America, and the other to the East African island of Madagascar, affords an interesting problem in the geographic distribution of plants,15 and brings vividly to mind the antiquity of the order. Other members of the same order are the arrowroot, turmeric, cardamom and ginger plants and the strikingly beautiful Bird of Paradise flower, Strelitzia reginae, which is often seen in gardens in California and Florida, as well as in its tropical home.


16) Baker J. G., in Ann. of Bot. VII, 204.

The present distribution of the banana, then, appears to be no wider than the distribution of its order throughout its history. At present the genus Musa comprises 32 or more distinct species and at least a hundred subspecies, many of them badly confused, and many of them probably representing only the escape of cultivated varieties, or the result of natural hybridization. The Philippines and the Indian archipelago are richest in forms, followed by Ceylon; America is, by comparison, rather poor in them—a fact that has often been adduced to show that the cultivated banana was not known before its introduction by the Spaniards. The genus is divided into two broad sections: Eumusa, with edible fruits, and Physocaulis, with inedible fruits; the former is, for commercial purposes, divided into bananas and plantains, the latter being larger and coarser fruits which are widely cultivated in the tropics, and eaten only after cooking. The cultivated varieties are innumerable, and in confusion, partly due to their great variability and tendency to bud variation. Dr. George V. Perez, for example, has called attention to a very recent mutation of the Canary Island banana (Musa cavendishii) which reaches double or treble the height of its dwarf parent. The inflorescence is identical, the fruit somewhat larger and better, in his opinion, than that of the parent. It is called by the natives a "male banana." Another sport of a different nature is described from Grenada, in the West Indies, by W. Malins-Smith (Agr. News, Vol. VI): "A few days ago I picked a bunch of 'claret' bananas which contained two hands of green colored fingers and one hand of claret and green fingers There was one finger which was one-half green and one-half claret. The green fingers ripened yellow. The bunch when ripe presented a very curious appearance." There is no doubt but that a study of the plantations in the tropics would reveal an immense number of similar cases. Baker16 mentions that Musa fehi, which grows widely in Tahiti, is seedless at the lower levels of the forests, but bears seeds when found at higher altitudes, say 3,000 or 3,600 feet. The sexual irregularity of the flowers also deserves attention. Altogether there is reason to believe that this genus, which has received the attention of plant breeders ever since man appeared on the earth, can yet furnish a great deal of valuable data throwing light on many of the most important problems of heredity.

Fruit of Musa martretiana, an African species, natural size. It consists of little but seeds. All bananas, before they were cultivated, must have been very much like this—their worthlessness has led O. F. Cook to suppose that the plant was first cultivated for the sake of its roots. Then, when sterile fruit, containing more pulp and only rudimentary seeds, was produced as a result of chance hybridization, primitive man must have seized upon such a fortunate result and perpetuated it by suckers. (Frontispiece).