Jour. Washington Acad. Sci. 6: 537-547. 19 Sept 1916.
Morphology and evolution of leaves.
O. F. COOK, Bureau of Plant Industry.

In seedlings of many palms, grasses, and other plants the first leaves are simple, bladeless sheaths, or the blade appears as an appendage or expanded outgrowth from the rim of the sheath. Leaves with larger blades are produced as the plant grows, until the adult degree of specialization is attained. The essential feature of the primitive leaf is the basal sheath, in the form of a cylinder. Each sheath in turn encloses the terminal bud of the shoot, and later encircles the stem.

Three elements of leaf structure—blade, petiole, and stipules—are recognized generally in manuals and textbooks. These are convenient for purposes of description, but for understanding the structure and evolution of leaves it is better to begin with the sheath or with the sheath and the blade, the two elements that appear to have been differentiated in advance of the others. Sheaths have been looked upon as expanded petioles or united stipules, instead of being considered as a primitive element. It has seemed reasonable to suppose that the petiole and the stipules have been derived from the blade, which in most plants is the largest and most important part of the leaf, but a general interpretation in accord with evolutionary facts is needed in the study of structural variations of plants.


In plants like Equisetum, Ephedra, and Casuarina the internodes perform the vegetative functions. Leaves are represented only by sheaths or scales which are mere appendages of the internodes, apparently of little use except for protecting the buds. Palms and many other plants afford examples of internodes and sheaths that have chlorophyll and stomata in the epidermal tissues and share the vegetative functions with the blades of the leaves. Cacti and specialized desert plants of other families have very small or rudimentary leaves, thus reducing transpiration.

It is customary to think of bud scales and similar organs as leaves that have been specialized by reduction, but it may be more correct morphologically to think of foliage leaves as enlarged bud scales that have assumed the vegetative functions formerly discharged by the internodes. The function of the bud scales is older than are the present forms of leaves, and may date back to the stage when the sheaths were simple cylinders, before the development of more specialized forms of leaf structure. It is unnecessary to suppose that the early types of seed plants lacked bud scales.


Other specialized forms of leaves, such as sepals, bracts, and scales of subterranean rootstocks, may also be considered as representing primitive sheaths or bud scales, rather than as reductions from the fully developed type of foliage leaves. From this point of view the foliage leaves appear to be the most specialized. Yet in developing the foliage leaves plants have not lost the ability to produce the simpler organs—sheaths, bud scales, or bracts. The plant body is a succession of different kinds of internodes, or metamers, bearing different kinds of leaves. At one end of the series are the cotyledons or seed-leaves, at the other the carpels or fruit-leaves, with many intermediate stages between the different kinds of foliar and floral organs.

As stamens are often transformed into petals, so we may think of cotyledons and foliage leaves as sterile carpels performing vegetative functions. Some of the species of Sterculia have broad leaflike carpels that persist and remain green long after the seeds have been shed. That the stamens and carpels of the different families of plants are generally more alike than the leaves or other parts of the plant body is more easily understood when we consider the evolution of plants as a process of intercalation of more numerous and more specialized forms of metamers. Plants like the junipers, pines, and eucalypts have two distinct kinds of foliage leaves, showing clearly that a double evolution of these organs has taken place.

1 COOK, O. F. Dimorphic leaves of cotton and allied plants in relation to heredity. U. S. Dept. Agric., Bur. Pl. Ind., Bull. 221. 1911. Heredity and cotton breeding. U. S. Dept. Agric., Bur. Pl. Ind., Bull. 256. 1912.

The succession of different kinds of leaves, the classical example used by Goethe in presenting the idea of evolutionary changes in the forms of the same organ, may not be less significant for the strictly morphological purpose of understanding the origin and homologies of the structural elements of the leaves. The changes that take place in passing through the succession of leaf-forms, during the development of the plant, are also of interest for the study of heredity.1


2 My attention has been called by Mr. H. Pittier to illustrations of the subligule of Trithrinax campestris published by C. De Candolle, Bull. Soc. Bot. Genève, II. 5: 106, pl. 3. 1913.

In the development of foliage leaves the ligule was probably the first structure to be added after the sheath and the blade. The ligule may be described as a thickened rim of the leaf sheath, bearing the weight of the blade or holding the sheath in place around the stem. In Serenoa and related genera of fan-palms the ligule is supplemented by a ligule-like expansion of the rim that supports the bases of the segments on the under side of the leaf. Although these subligules, as they may be called, appear in only a few palms, they are not without general interest as organs that are closely parallel in structure and function with the true ligules.2

The sheaths of most palms are thickened on the side that supports the petiole, and are split on the opposite side, allowing the leaf to diverge more widely from the trunk, but the more primitive condition of closed sheaths with nearly uniform texture is also found, as in Calamus, Desmoncus, and Chamaedorea. Even in the large royal palms and many others the sheaths are not split until the leaves are ready to fall.

In palms like Desmoncus and Pyrenoglyphis, as in many members of the family Polygonaceae, the sheath, instead of being specialized at the mouth to form a ligule, is prolonged far above the insertion of the petiole, forming what is known as an ocrea. A similar prolongation of the sheath beyond the point of attachment with the petiole is found in the bud scales of Magnolia and Ficus. Some writers have taken it for granted that the ocrea, or the entire sheath, has been formed by the union of stipules, but the indications point rather to the formation of stipules by reduction and specialization of the sheaths of the more primitive forms of leaves.

Whether we consider that the blade arose simply as an expansion of the upper portion of the sheath or as an outgrowth from the rim of the sheath, the ligule may be considered as marking a line of separation of the two portions of the primitive leaf that became specialized. Even this indication of definite specialization of parts is lacking in many plants, as among the lilies, orchids, and plantains, where the leaves still appear as simple sheaths with a gradually broadened distal portion to serve as a blade, but no distinct separation of special parts. The idea of leaf blades arising as outgrowths from leaf sheaths appears more probable, or at least less fantastic, when we consider such a case as the leaf of Smilax, where the end of the sheathing base produces a pair of slender tendrils several inches in length, in addition to the blade of the leaf. It may not be without significance that rudimentary blades of abnormal or reduced leaves often assume very slender, tendril-like forms.


3 BERRY, E. W. The origin of stipules in Liriodendron. Bull. Torrey Bot. Club, 28: 493. 1901.

It has been supposed that stipules were developed from basal lobes of leaf blades. This interpretation was suggested by Ward and later adopted by Berry on the basis of evidence drawn from fossils and abnormal leaf-forms of Liriodendron, but these can be understood as intermediate stages between normal leaf-forms and bud scales. It seems unnecessary to suppose that stipules began with Liriodendron, or that the stipules of Liriodendron are not homologous with those of other genera and families.3

The very general occurrence of stipule8 or stipular structures is in itself a fact that must make it difficult to credit the idea of recent development or independent derivation of such organs, either in the family Magnoliaceae or in the many and widely different families of plants that have stipules. The stipules of Liriodendron may be more primitive than those of other Magnoliaceae in retaining more of the vegetative functions of the primitive sheath, but their nearly complete separation from the petiole and from each other may be considered as a specialization, since partially united stipules occur. Union between the stipules and the petiole is shown in many of the reduced leaves or the large floral bud scales that have small petioles and blades. Union between the stipules on the side opposite the petiole is sometimes shown in very large stipules of strong shoots.

Another line of evidence tending to throw doubt on this idea of stipules as specializations of leaf blades might be drawn from bracts and other organs that take the form of sheaths, even in families whose normal foliage leaves are without stipules. Thus, in the genus Hicoria the outer scales of lateral buds have the form of a closed sheath, carinate on each side, somewhat similar to the prophylla of grasses and palms. The large membranous bud scales that precede the leaves on new shoots of Hicoria also represent sheaths and sometimes appear in modified form as stipular wings of the petiole, on abnormal leaves of shoots developed late in the season.

4 SINNOTT, E. W., and BAILEY, I. W. Investigations on the phylogeny of the angiosperms. 3. Nodal anatomy and the morphology of stipules. Amer. Journ. Bot., 1: 441-453, pl. 44. 1914.

Considerations drawn from the study of "nodal anatomy" have been used recently by Sinnott and Bailey as the basis of interpretation of leaf morphology. Stipules, sheaths, and ligules are classed together as "modifications of the base of the petiole," the final conclusion being that "a leaf with two distinct stipules is more ancient in type than one with a sheathing base."4

5 See, TYLER, A. A. The nature and origin of stipules.
Ann. N. Y. Acad. Sci., 10: 1-49, pls. 1-111. 1897.

Stipules have also been considered as "accessory leaves," or as a specialized development of the basal portion of the "true petiole."5 In the present view, stipules, bud scales, and various forms of bracts appear to have been derived not from the blade or the petiole of specialized leaves, but from the primitive sheath. The ligule and even the blade itself are considered as specializations from the distal end of the sheath; but scales and stipules as specializations of the basal portion. Petioles are not all homologous, but are of two kinds, some derived from the sheath, and others from the blade.


6 COOK, O. F. Origin and evolution of angiosperms through apospory.
Proc. Wash. Acad. Sci., 9: 174. 1907.

Taking account of the ligule or the ocrea as marking the end of the primitive sheath makes it possible to distinguish very clearly the two kinds of petioles. Both kinds are represented among the palms. In fan-palms the so-called petiole is below the ligule, whereas the part described as petiole in some of the pinnate-leaved palms certainly is above the ligule or ocrea. In the fan-palms it is plain that the petiole is a narrowed, elongated portion of the leaf sheath, ending at the ligule, whereas the petiole of the Cocaceae and Chamaedoreaceae represents a naked basal portion of the rachis or midrib.6

That the petioles are of two kinds is apparent also from the fact that in the fan-palms the petioles agree in structure and are entirely continuous with the leaf sheath, while in the pinnate-leaved palms there is equally complete agreement and continuity with the rachis. But the leaf structure is not the same in all of the pinnate-leaved families. There appear to have been several independent derivations of pinnate leaves from fan-leaved ancestors. The Geonomaceae are a pinnate-leaved family in which the petiole appears to be a part of the sheath, the same as in fan-palms.

The magnolia family affords another example of double differentiation of the petiole, the lower part of the organ being formed by a thickened segment of the leaf sheath, while the upper part is a narrowed base of the blade, as shown by the decurrent margins which run down to the ligule. In such species as Magnolia virginiana the lower portion of the petiole is marked very distinctly by the scar of the deciduous leaf sheath, with a minute hairy prominence at. the end, which may be considered as a thickened base of the ligular prolongation that forms the apex of the sheathing bud scale.

It is conceivable that a petiole formed originally from the sheath element might become separated from the remainder of the sheath, which would account for the evolution of such organs as the deciduous bud scales of the Artocarpaceae. But all petioles formed from sheaths would remain essentially different morphologically and developmentally from those that were formed originally as basal elongations of the midrib of a simple leaf, or by moving the pinnae farther up the rachis of a compound leaf, as in the pinnate palms.


The recognition of the two kinds of petioles as representing distinct morphological elements makes it necessary to have more convenient ways of designating the two classes of organs to which the word petiole has been applied indiscriminately. The use of such terms as sheath petiole and blade petiole would afford a way of indicating the distinction. Sheath petioles would be understood as those that represent specializations of the primitive sheath, and blade petioles as specializations of the blade or midrib of the leaf.

Other terms that mighty be used are infraligular for the sheath petioles and ultraligular for the petioles, in allusion to the differences of position in relation of the ligule. It might be objected that ligules are confined to a few families, but the terms would still serve to indicate the homologies of the parts to those of plants whose possession of ligules gives the most definite basis for the distinction between the two kinds of petioles.


For general. descriptive purposes and especially for dealing with plants in which the resemblance to the primitive sheath has entirely disappeared, it may be simpler to treat the petiolelike sheath element as a new organ not formally distinguished hitherto. On this basis instead of sheath petiole or infraligular petiole, a single word like foot, in Latin descriptions pes, might be used. This would have the advantage of leaving the term petiole as nearly as possible in its present signification, which would need to be modified only in those cases where the so-called petiole might be found to represent the sheath element. No doubt there are many plants where not only the stipules are lacking, but also the foot.

7 COOK, O. F. Jointed leaves of Amygdalaceae.
Journ. Wash. Acad. Sci., 2: 218-220. 1912.

From this point of view it would be possible in dealing with different families of plants to define the foot in any way that might be most convenient, as a thickened segment of the primitive sheath, as the element of the primitive sheath that supports the blade, as the element that is between the petiole and the internode, or as the element to which the stipules are attached Thus in the peaches, almonds, plums, and other members of the family Amygdalaceae, the petiole is articulated at its base to the foot, to which the stipules are attached. The petiole is deciduous with the blade, but the foot persists for another season and functions as a bud scale. The foot is present also in apples, pears, and roses, but falls with the leaf, there being no joint at the base of the petiole.7 The joint is present in Oxalis and in many leguminous plants.

Objection might be taken on etymological grounds to using the word foot for an element that in many plants is smaller than the petiole, which term means a small foot. It does not seem, however, that this is likely to cause confusion, since the obvious signification of foot is in relation to the lowest, most truly basal portion of the leaf. Absence of the foot is to be considered as a specialization of leaf structure, and it will be interesting to determine the status of the organ in the different families. Even though not present in the foliage leaves, the foot may still be represented in the bud scales, bracts, or other organs which from our present point of view appear to be more primitive and less specialized than the foliage leaves. That the cotyledon is sometimes called the nursing foot, or simply the foot, hardly constitutes an objection to the use of this word, with other leaves, for the element that corresponds to the cotyledons and primitive bladeless sheaths of seedlings.


Pulvini and articulations represent special forms of tissue connected with the ligule or base of the blade in grasses and palms, and found in corresponding positions in other families of plants. The chief function of pulvini is to control the position of the leaf blade, which is accomplished by varying the turgidity or water pressure in the rather loose cells of which the pulvini are composed. A flexible pulvinus is in the nature of a joint.

Many leaves have a basal pulvinus, with a layer of absciss tissue to form an articulation with the internode, when the leaf separates at maturity. In other leaves the joint that provides for the detachment of the leaf is between the foot and the petiole, as in the case of the peaches, plums, cherries, and related plants. In Magnolia virginiana there is a pronounced development of the entire upper surface of the foot as a pulvinus, which has a special function in lifting the winter bud scales in the spring.

8 COOK, O. F. Branching and flowering habits of cacao and patashte.
Contr. U. S. Nat. Herb., 17: 609-625, pls. 44-54. 1916.

The organ that is usually described as a true petiole may prove to be a foot in cases where there are pulvini at both ends, as in cacao. In the patashte tree, a relative of cacao, the petioles of the leaves of the fruiting branches have the structure of pulvini for their whole length.8 In the cotton plant also there are pulvini at both ends of the petiole. Other reasons for considering the cotton petiole as a foot may be found in the fact that while the margins of the leaf blade never show any tendency to become decurrent, the petiole is often united with the margins of enlarged stipules like those that form the involucral bracts.

9 COOK, O. F. Brachysm, a hereditary deformity of cotton and other plants.
Journ. Agric. Research, 3: 387-400, pls. 53-62. 1915.

The suppression or extreme reduction of the blade in the formation of the involucral bracts of the cotton plant might also be considered as an indication that the blade, in spite of its much greater size, still behaves in some respects as an appendage of the sheath. That the general course of evolution has been in the direction of enlarging the blade and reducing the sheath element may be considered as at least a partial explanation of the fact that enlargement of the stipules is almost invariably accompanied by reduction of the blade of the leaf. If. only one stipule is enlarged, a lobe is likely to be wanting on the same side of the blade.9


The leaves of angiosperms show a primary division into two morphological elements, (1) a basal sheath supporting (2) an expanded blade, as represented in the leaves of palms, grasses, and many other plants. The organs that are usually described as petioles are of two kinds, (1) some that appear to have arisen through a narrowing of the base of the blade, and (2) others through a narrowing of the primitive sheath. The name foot is suggested for a specialized portion of the leaf sheath that serves as a petiole. Both the petiole and the foot are represented in many plants, in such families as the Amygdalaceae, Rosaceae, and Magnoliaceae. Stipules, bud scales, bracts, ligules, and pulvini are other specializations of the primitive sheath element, and the blade also appears to have arisen as an outgrowth or expansion of the sheath.