THE MORPHOLOGY OF THE APPtE

Transcription

1 230 APRIL 1910 CORNELL UNIVERSITY AGRICULTURAL EXPERIMENT STATION THE MORPHOLOGY OF THE APPtE AND OTHER POME FRUITS L. H. MACDANIELS ITHACA, NEW YORK PUBLISHED BY THE UNIVERSITY Received for publication December 22, 1939

2 CONTENTS PAGE Review of literature Evidence supporting the receptacular theory..., Evid.ence supporting the appendicular theory The basic pattern of the pome flower Detailed study of the apple, the pear, and the quince flower... ',,,, " 11 Pome structure in the light of comparative study The morphology of the apple fruit,..,,,,,.,, ',, ' ' ','. ',.. ""'.'.. 23 Discussion...,... '. '., '.,,.,. ' ', '.,...,,...,., ',,. ',. '. 27 Summary... ',. '... '... ::-..,, References...,.,,,,.,.,...,. '...,..., 31

3 THE MORPHOLOGY OF THE APPLE AND OTHER POME FRUITSl L. B. MACDANIELS The morphology of the pome fruits has long been a matter of controversy. The essential facts of the structure of these fruits are well known, out the interpretation of the observed structure has developed along two widely divergent lines. One of these is that the part of the apple flower which makes up the fruit (with the exception of the ovary) is axial, or receptacular, in origin and nature, and is therefore a part of the stem, or axis. The other point of view is that this outer fruit tissue is derived from the fused floral parts ~ that is, the sepals, petals, and stamens - and is therefore appendicular in nature. There is a third possibility that the tissue may be both axial and appendicular, a situation which is found in the fruit of the genus Rosa (Van Tieghem, 1878; Bonne, 1928; Jackson, 1934) and which is claimed also for the pomes by Hillmann (1910).. The difference between the axial and the appendicular interpretation of pome morphology is fundamental. If the fruit of the apple, for example, is basically stem in its make-up, then it is essential to locate the structures characteristic of the stem in the fruit. Kraus (1913) and other investigators have been logical in designating parts of the fruit morphologically as pith, cambium, and cortex. If, on the other hand, the apple fruit is appendicular - made up of fused appendages or leaflike partsthe vascular and other' tissues found should show the characteristics of fused appendages. Furthermore, a correct interpretation of the morphological nature of the parts of the apple fruit is of pract~l value. For example, different parts of this fruit show differences in behavior under storage conditions, particularly as to discoloration and breakdown. This without question depends upon structural and chemical differences, which in turn are related to the basic nature of the parts concerned. If morphology is of any real significance, it must form a logical and consistent basis for the interpretation of plant structures. Therefore it is important that this controversy be settled so as to clarify the interpretation of pome structure,' particularly as related to the behavior of fruit in storage. REVIEW OF LITERATURE The receptacular, or axial, theory of pome structure is widely accepted among horticulturists and taxonomists, and is the one most frequently found in literature at the present time. Linnaeus originally interpreted. the floral tube, or cup, of the Rosaceae as being composed of fused floral parts. This view was supplanted by the receptacular theory, owing in part to the ontogenetical studies of Payer (1857). The receptacular theory became widely held in Europe, particularly in Germany, and was accepted in America as well. Bailey (1919), in a _)". 'Contribution from the Department of Pomology, Cornell University, Ithaca, New yl;fk. 3

4 4 L. H. MACDANIELS technical description.of the genus Pyrus, wrote: "torus urn-shaped and attached to the carpels and finally closing over them." He wrote further: The morphology of the pome is still perhaps a subject of disagreement, although most botanists now consider it to be a hollow torus (receptacle), or hypanthium, or cupula, in which the ovary is imbedded... It was formerly held that the edible part is largely calyx-tube, but various morphological considerations have inclined students to regard it as stem rather than calyx; the term calyx-tube is still retained, however, in descriptive writings. Rydberg (1898) stated that the theory of origin from the calyx tube is the one generally accepted by American taxonomists, but he considered that, with few exceptions, the whole of the gamosepalous part of the calyx, which he terms the hypanthium, must be regarded as belonging to the flower axis. Rendle (1925), following Fo.cke in Engler and Prand, writes: "The floral axis forms a deep cup with the inner wall of which the five or fewer carpels are more or less completely united as well as with each other." Kraus (1913) worked through the problem of the structure of the pome by developmental studies after the manner of Payer (1857), and interpreted its morphology as axial, or toral. In his summary he wrote: A pome is to be regarded as consisting of one to several drupe-like fruits more or less intimately connected with a fleshy torus, on and within which they are borne. In the apple, the carpels are composed of a fleshy exocarp and mesocarp, and a parchment-like endocarp. All floral parts are borne on the torus, none on a so-called fleshy calyx. Black (1916) in general follows Kraus. She states: The fruit of the apple may then be considered as-a reenforced or composite fruit consisting of one to several drupe-like fruits embedded in a fleshy torus and is called a pome. The torus is the receptacle of the flower and by excessive growth produces the flesh of the fruit, in which a well-defined pith and cortical layer can be seen. A survey of a dozen texts of botany shows that nearly all of the authors accept the receptacular interpretation of the pome fruit. Robbins and Rickett (1934) popularize the idea, saying: "When the receptacle surrounding an inferior ovary becomes fleshy, we call the result a pome. An example is the apple... When we eat an apple we eat the 'stem' and throwaway the ovary; when we eat a banana we peel off and discard the 'stem' and eat the ovary." Stanford (1937) offers both theories. Wilson and Haber (1935: ), however, say: "Outside the endocarp is the region of the fleshy exocarp; the mass of the flesh of the apple is composed of the fleshy calyxtube." In one of the most recent and comprehensive texts, Hill, Overholts, and Popp (1936: 253) write: "... the carpels are imbedded in the cuplike receptacle, with the styles projecting... After fertilization, the receptacle, the exocarp, and the mesocarp of the several carpels enlarge extensively and become fleshy, forming the edible portion of the fruit." The above citations indicate the wide acceptance of the idea that the fleshy part of the pome is toral, or axial, in nature. Some investigators, however, particularly those who have given attention to cou:parative

5 THE MORPHOLOGY OF THE ApPLE AND OTHER POME FRUITS 5 morphology and anatomy, hold to the appe;"dicular theory. This interpretation goes back to the fundamental work of Van Tieghem artd his development of the stelar theory. He not only advanced the stem.r theory and emphasized its importance in phylogenetic studies, but he also worked specifically on floral structures (Van Tieghem, 1868); and he stated plainly (1878) that on the basis of his studies he considered the calyx tube of the Rosaceae to consist of the fused floral parts, with the exception of the genus Rosa, in which he considered at least the basal part of the structure of the fruit to be stelar in nature. Gray (1887) considered the fleshy part of the pome to be calyx. Bailey also held this view until about the time of the writing of the Standard Cyclopedia of Horticulture (Bailey, 1916), when he accepted the receptacular nature of the pome fruits. More recently, Eames (1929) has emphasized the importance of the anatomical approach in interpreting the structure of flowers with relation to their phylogeny, and still later (1931) he has given a general statement of the significance of anatomy in determining floral morphology and phylogeny. The Rosaceae, ho'wever, were not specifically interpreted. About this time, Bonne (1928) published an extensive treatise dealing specifically with.jl:ie Rosaceae and giving the comparative vascular structure of many genera. Bonne's conclusions substantiate.tl;.wse of Van Tieghem and are based on a much more comprehensive s~ey of the family. Jackson (1934), after a detailed comparative anatomical study of Rosa and allied genera, confirmed the interpretation of Van Tieghem and Bonne that the fruits of Rosa are in part receptacnlar in nature and in part appendicular. This brief account of the history of the receptacular and appendicular theories of the morphology of the pome, shows that the divergence of opinion is due mainly to different interpretations of the structures observed. The available evidence makes the latter theory seem the more logical of the two, and the moore nearly in line with the interpretation of inferior ovaries in general. EVIDENCE SUPPORTING THE RECEPT ACULAR THEORY To substantiate effectively the appendicular interpretation, it is necessary to present detailed evidence for both interpretations. The receptacular theory is based on a number of lines of evidence, possibly the most obvious and spectacular being found in the abnormalities which occur in pomes and particularly in the <l.{lple. The abnormality most frequently observed is the occurrence of a calyx lobe on the side of the fruit. In some specimens this may be an extra lobe; in others, one of the five sepals is located on the side of the fruit rather than at the apex. In more marked abnormalities these calyx lobes may closely resemble leaves having normal vein structure. In other instances the apple or p'e~r flower, because of some change in its fundamental growth status, -has produced green leaflike structures or even normal leaves instead of petals and stamens, and the center of the axis has continued to develop., as a normal stem. Such an abnormality has been described by Tukey U938) and by Shaw (1939), and is without question much the same phenoj.11enon as is reported occasionally in periodicals devoted to horticultural subjects.

6 6 L. H. MACDANIELS Observers have concluded that if calyx lobes may occur at any point on the apple fruit, and if these structures in some cases appear leaflike, then the structure which apparently bears them must be stem. A second line of evidence for the receptacular theory is found in the ontogeny of the flowers themselves, as was shown by Kraus (1913) and others. According to Kraus' interpretation, the floral parts develop first as protuberances on the growing point, and then, by means of intercalary growtfl. behind these meristems, the developing floral parts are shoved upward together. This development occurs in about the same way in flowers of those genera of Rosaceae in which the floral tube is fused to the ovary (for example, the pomes), as it does in those in which the tube is free (for example, in the genus Prunus). A third line of evidence is based upon the apparent structure of the pome itself. Kraus (1913) pointed out that tissues which are apparently pith in the developing fruit are continuous from the pedicel of the apple fruit into the region surrounding the carpels, and that the pith is thus continuous with and a part of the inner structure of the mature fruit. From the center of the fruit, pith rays are considered to extend outward between the ten primary vascular bundles to the outer part of the fruit, which is considered cortex, and a cambium region is recognized connecting these bundles. This conclusion is justified by the g-eneral appearance of the pome fruit in cross section. Another point, which is emphasized by Rydberg \1898), is that in the Rosaceae, sepals, petals, and stamens are borne on the floral disk just as they are borne on stems, and that the petals in most species are deciduous and the sepals in a few other species' ar.e...shed from the fruit. Such behavior is like that of normal appendages attached to stems. Still another reason in support of the receptacular nature of the pome fruit is that the so-called primary vascular bundles are not simple bundles supplying a single organ, but are complex in that they represent the vascular supply of several floral organs. For example, the so-called petal bundle supplies one petal, three stamens, and two lateral traces to adjacent sepals. If this is true, then the situation in the apple fruit is like the condition in a normal stem with a broken stele, where a given segment of the stele supplies several organs. This review is not a complete statement of the receptacular theory, but it contains at least the more important arguments that favor the toral nature of the pome. The basic criticism of this interpretation is that it leaves out of account an exceedingly important approach to the problem, namely, that of comparative morphology and anatomy. Most of the points made are based qn observations of external morphology. A more detailed consideration of this evidence is given later. EVIDENCE SUPPORTING THE APPENDICULAR THEORY The evidence for the appendicular nature of pome fruits starts with the basic idea that the flower is morphologically a determinate stem with appendages. These appendages are homologous with leaves, at least in the sense that each of the floral parts involves only a part of the axis and develops from the primary meristem in much the same way as does the

7 . THE MORPHOLOGY OF THE ApPLE AND OTHER FOME FRUITS 7 leaf. This' viewpoint is widely accepted and is considered axiomatic in the interpretation of floral structures by their comparative anatomy. There are, of course, a number of workers in the field of fl6ral morphology who do not accept this basic concept, and, in fact, ignore the significance of the stelar theory and anatomical evidence in general. Because the evidence for the appendicular nature of the po me is based upon the validity of the concept of the flower as a determinate stem with appendages, a statement of the premi?es upon which the argument is built is pertinent. A vascular dicotyledonous plant of the type to which the apple belongs is made up of an axis and lateral appendages. The axis consists of two parts, the stem and. the root, both of which may branch freely. The primary. axis is composed of several morphological layers, or zones: the outer layer, or epidermis; beneath this, the cortex; and the central core of vascular tissue, known as the stele. The stem differs from the root fundamentally in a number of ways, important among which is the manner and the order in which branches are formed. In the stem, branches are initiated at or very near the terminal meristem in a definite order, or phyllotaxy, and the vascular traces to branches leave gaps in the stele at the point of departure from it. In the root, branches arise without reference to the terminal meristem and in no regular spiral order, and no gaps are left in the stele. Appendages of the first rank on the stem, which include epidermis, cortex, and a part of the vascular tissue, are known as leaves. Logically, therefore, any apparent lateral structure on the stem 'is a branch if it includes a complete stele; it is a leaf, or at least a leaflike structure or appendage, if it includes only a part of the vascular stele. This concept of the relationship of the stele to its appendages carries no implication as to the phylogenetic origin of the sporogenous tissue or of the stamen or the carpel as such. Quite probably, a stamen never was a leaf in the sense that leaves are now known, and, in the course of evolution, may have been a branch system. The point to be made is that descriptive anatomy, to be useful, must furnish a concept and a terminology that will apply to plants as we find them today, and the concept that a flower is a shortened axis with appendages is so applicable. If floral parts are appendicular and leaflike, then, in general, they behave as leaves do in their development and structure, particularly in their relationship to the stele as shown by the nature of the vascular traces and the gaps in the stele which are left where the vascular supply of the appendages is given off. This concept is reviewed in detail by Eames (1931), and abundant evidence makes the theory acceptable anatomically'. Floral parts are modified phylogenetically into many varied forms, the, fundamental nature of which may not be obvious. The basic structure as regards leaf traces and gaps, however, appears to be, according to the comparative studies of Eames (1931), that the sepal typically has three traces, the petal one, the stamen one, and the carpel three, Three may have been the basic number of traces for all floral parts, but the vasc1llar supply of the petals and the stamens has been reduced in its development, so that the condition stated above applies to plants as they now #xist. The pedicel of the flower is morphologically a stem. This may ha~e an

8 8 L. H. MACDANIELS unbroken vascular cylinder, or it may be made up of discrete bundles arranged about the central pith in a circle. Both types of stele-siphonoitele and dictyostele-appear in the pedicels of the Rosaceae (Bonne, 1928), and even in those of the pomes. In the pedicel of the apple the stele is entire near the base and somewhat broken toward the apex (figure 4, H, page 12). Different varieties apparently show some variation in this regard. In the phylogenetic development of the floral axis, a number of modifications have occurred which complicate the interpretation of the basic vascular system of the flower (Eames, 1931). One of thcse is that in the flower the internodes are greatly shortened, so that the leaf traces supplying the floral parts come off the stele at nearly the same level. Also, according to the principle of cohesion, floral parts may have fused with like parts in the same cycle. This is particularly true of the sepals and the petals. Under the principle of adnation, there is a fusion particularly at the base of the members of different cycles, so that there may be sepals fused with petals, and petals fused with stamens, or all of these parts fused with the carpels into a structure such as is found in the pome. Such cohesion and adnation may be fairly obvious or altogether obscured in the external appearance of the flowers. The real nature of the parts also may be obscured by their becoming fleshy or by the appearance of anomalous structures of various kinds. In species where external fusion between the floral parts has taken place, the vascular anatomy is of the greatest importance in indicating the nature of the fusion, in that the vascular bundles passing into the floral parts do not fuse in phylogenetic developm_e!,1t as quickly as do the fleshy outer layers of the parts themselves. The vascular bundles appear to resist change and are classed as conservative. A brief review of the evidence supporting this hypothesis is given by Kozo-Poljanski (1936). Vascular traces do, of course, fuse, and such fusion may be complete. In general, however, by the study of the comparative anatomy of different species within a genus, some indication of the basic vascular structure can be found which will shed light upon the interpretation of the more obscure relationships of floral parts. Thus, in large families such as the Ericaceae and the Rosaceae, genera are found showing different degrees of fusion of the vascular bundles of the floral parts. In general, the greatest fusion is considered as phylogenetically advanced, and a lack of fusion as the primitive condition. In the Ericaceae there are species which show progressive fusion of the vascular bundles in adnation of the outer floral parts to each other, and of these fused parts to the carpels, in the production of an inferior ovary (Eames, 1931). A similar series has been shown by Jackson (1934) in the Rosaceae, for interpreting the structure of the floral tube, or hypanthium, of Rosa. If the structure of pomes is studied by the same comparative method, their morphology is likewise better understood. In the work of Bonne (1928) and others, the main facts of the vascular anatomy of many genera of the Pomoideae are available. The more detailed anatomy of the apple, the pear, and the quince is here reported. Other rosaceous genera also have been examined in the present study.

9 THE MORPHOLOGY OF THE ApPLE AND OTHER POME FRUITS 9 THE BASIC PATTERN OF THE POME FLOWER In studying the detailed anatomy of the flowers of any group- of allied plants, the basic plan of the flower must be kept in mind. This has already been worked out for the apple by Kraus and Ralston (1916) as shown in figure 1. Their study shows that each of the ten so-called primary vascular bundles seen in the flesh of the apple is complex. Those S 5 de de s p s p FIGURE 1. DIAGRAM OF THE VASCULAR SYSTEM OF THE APPLE FLOWER SPREAD OUT IN ONE PLANE (ADAPTED FROM KRAUS AND RALSTON, 1916) s, Sepal; p, petal; st, stamen; dc, dorsal carpellary bundle. Tbe ventral carpellary bundles are not shown opposite the sepals, referred to as "sepal bundles", supply one dorsal carpellary bundle, one stamen, and the median trace of the sepal. Those opposite the petals, known as "petal bundles", supply the median trace of the petal, one trace to each of three stamens, and one lateral trace to each of the adjacent sepals. These ten primary bundles join the stele near the.. base of the flower. The stelar tissues remaining after the departure of the ten primary bundles, form the ten ventral carpellary bundles which extend upward through the center of the flower, giving off lateral traces to the ovules and terminating in the styles. This basic pattern of the pome flower is clarified in figure 2, which shows the course of the vascular bundles as seen from above the flbwer in a cleared section of some thickness. In this diagram, only th(;t/main vascular bundles are shown. In the center are the ten ventral caq)ellary

10 10 L. H. MACDANIELS bundles. Exterior to these are the five dorsal carpellary bundles. Oppo ~ite each of these is the median sepal bundle which gives off a vascular trace to one stamen. Alternate with the sepal bundles are the five petal bundles, each one made up of three stamen traces, two lateral sepal traces, and the median petal trace. In addition to this main vascular system, the CJQ) 8 (1) o G 0 e.0:....:--v C ~ (f) FIGURE 2. DIAGRAJlI OF THE BASIC VASCULAR SKELETON OF THE APPLE FLOWER s, p, st, de, and vc indicate sepal, petal, stamen, dorsal carpellary, and ventral carpellary bundles, ~~~. bundles of which are of morphological significance, there are other smaller bundles branching from them which form an anastomosing system supplying the fleshy parts of the floral cup (figure 3). In tracing the anatomy of different species in a comparative way, therefore, it is necessary to keep in mind this basic structure, which is essentially the same for all of the pomes and many of the other rosaceous genera as well. Important differences to be sought are the various levels at which the trace bundles leave the stele, and particularly the differences in the degree of fusion of various bundles in adnation. Another signifi-

11 THE MORPHOLOGY OF THE ApPLE AND OTHER POME FRUITS 11 FIGURE 3. I I I VC DIAGRAM OF THE VASCULAR SKELETON OF THE KIEFFER PEAR FLOWER, SHOWING ANASTOMOSING SECONDARY,BUNDLES s, p, st, dc, and vc, as in figure 2. The smaller anastomosing bundles supply the floral tube and are not a part of the basic vascular pattern of the ~pome flower, cant anatomical feature is the amount and position of the vascular tissue remaining at the top of the stele after the ventral carpellary traces have separated from it. This tissue, varies in amount in different species, and even in different horticultural varieties within a species. DETAILED STUDY OF THE APPLE, THE PEAR, AND THE QUINCE FLOWER Diagrams of the vascular bundles in the flower of the apple are shown in figure 4. The longitudinal diagram (figure 4, I) shows the structure of the flower as cut medianly through a sepal and the opposite petal, and may be used in orienting the transverse sections. The pedicel of the flower has the vascular tissue arranged in a somewhat broken ring, or stele, as indicated in figure 4, H. (With further development of secondary tissue during the maturation of the fruit, this ring of vascular tissue is widened and becomes unbroken.) Higher in the pedicel, at a point about 4 millimeters below the beginning of the swelling of the base of the flower, the ring of vascular tissue becomes further divided (figure 4, G). At this point the sepal and the pet<ti bundles are beginning to separate from the remaining stelar tissues. At level F as shown in the figure, the bundles have become more widely s~parated into a larger ring, and by tracing them upward it is possible to distinguish the ventral carpellary bundles and the petal and sepal bundles as indicated in the figure. At level E the ventral carpellary bundl~;s and the residual stelar tissues are located near the center of the axis,,' The dorsal carpellaries are distinct, but have not assumed their positiorli oppo-

12 12 L. H. MACDANIELS A --.. de dh ~ ';c.... B '&r'. 5 \ I!} de:,de Q ~.. p t _... "-l."c: 5 Apple... _---_. p ~w Qi dc,.. ~ rp ~. l ) ~ v p Ii> ~...- <$9 I» 19e!J8' ~@ fj (J~ <;;) ~ <!D- vc ~ o ~ ~... p (;) (3J~~V <& ~~~ -.s E FIGURE 4. F G H DIAGRAMS SHOWING THE VASCULAR SKELETON OF THE APPLE FLOWER The capital letters A, B, C, D, E, and F on the longitudinal.section diagram indicate the level" of the cross section diagrams with the same letters. The small letters s, p, st, dc, and vc, indic",e sepal. petal, stamen, dorsal carpellary, and ventral carpellary hundles, respectively. (See text.or further explanation) site the sepal bundles. In this general region there are fragments of vascular tissue within the ring of the ventral carpellary bundles, as shown photomicrographically in figure 5, A. Kraus and Ralston (1916) considered these bundles a part of the carpellary vascular system. However, this extra vascular tissue, remaining after the ventral carpellary bundles have been given off, may logically be interpreted as consisting of what is left of the stele after the bundles have been given off to the various floral parts. This residual stelar tissue is shown more in detail in photomicrographs of the Kieffer pear flower (figure 7, D, page 15). At level D as shown in figure 4, some of the dorsal carpellary bundles have assumed their position in a line radial with the sepal bundles. The ventral carpellary bundles have each divided to forn.! two bundles which have passed to the edges of each carpel and have become oriented so that the xylem is toward the outside. This inversion of the vascular tissue in the ventral carpellary bundle is of great importance in the interpretation of the nature of the ovary, because it indicates that each locule of the com.pound ovary represents morphologically a modified leaf or carpel with the same basic structure that would be found in a free carpel or

13 THE MORPHOLOGY OF THE ApPLE AND OTHER POME FRUITS 13 follicle.. The detailed structure of two ventral carpellary bundles showing the inversion of xylem and phloem is shown in figure 5, C. The vascular supply of the ovules passing from the ventral carpellary bundles is shown in figure 4, C. Here the dorsal carpellary bundles are all opposite the sepal bundles, which are lined up with the petal bundles in a ring. At level B in figure 4, at the top of the locule, the one stamen bundle has separated from each of the sepal bundles and the petal bundles have given off two or three stamen traces. Between level C and level B, in the flower of the variety Baldwin, the limits of the individual carpels are FIGURE 5. PHOTOMICROGRAPHS OF SECTIONS OF AN APPLE FLOWER AND OF A S::\1ALL APPLE FRUIT A, Section of an apple flower taken below the locules. The ten primary bundles are distinct. The five dorsal ca.rpejla_ry bundles are free from the sepal bundles. The ten ven~ral carpellary lmndles are fused In pairs near the center. The central group of vascular strands IS the residual tissue of the receptacular stele B, Section of a Baldwin apple flower taken medianly through the locules, showing the epidermal tissues which line the central axile sac and extend into thc locules and between the ca. Is C, Section of a small McIntosh apple, showi!1g inversion of the xylem in the ventt:_,,1 carpellary bundles. The protoxylem elements are on the Side of the bundle toward the locule which is above' the phloem is adjacent to the axile sac, the limits of which are below '

14 14 L. H. MACDANIELS plainly marked by the extension of the epidermis of the carpel both between the margins of the same carpel and between those of adjacent carpels (figure 5, B). At level A in figure 4, the five dorsal and ten ventral carpellary bundles are included in the fused styles of the pistil, which is free from the floral tube. All of the starnen bundles are now free. Just above the point where the third stamen bundle leaves the petal bundle, two lateral bundles are given off from the petal-bundle trace. These lateral bundles extend into the sepal on either side. The relationship of these bundles is similar to that in the Kieffer pear flower, shown in figure 6, A. ~ -s --- p A B \ C FIGURE 6. D E K;effer Pea r DIAGRAMS SHOWING THE VASCULAR SKELETON OF THE PEAR FLOWER (See legend to figure 4 for meaning of the capital and small letters.) The letter x indicates the residual vascular tissue of the receptacular stele after it has given off the ventral carpellary bundles The gross morphology of the flower of the Kieffer pear is basically the same as that of the apple. It differs mainly in that the five styles are free at the base instead of fused. The central canal between the styles extends downward nearly to the base of the locules, and the inside surface of the floral tube is more flattened. The vascular anatomy of the flower shows some minor differences (figure 6). Important among these is the fact that the vascular tissue remaining after the departure of the ventral. carpellary traces is less diffuse than in the apple. Instead of consisting of an irregular group of vascular strands, it is made up of five tdngues of vascular tissue which turn inward and terminate just below the base of the stylar canal. These strands of tissue are alternate with the pairs of ventral carpellary bundles, as is indicated in figure 6, C. A more detailed representation of this tissue is given in the photomicrograph, figure 7, D. The vascular anatomy differs further from that of the apple in that the starnen tl"aces leave the petal and sepal bundles at nearly the same point and at a higher level in the floral tube. Another minor difference is that the dorsal carpellary bundle curves inward and downward over the top of the locule, so that a section at level B shows this bundle cut three times (figure 6, B, and figure 7, H).

15 THE MORPHOLOGY OF THE ApPLE AND OTHER POME FRUITS 15 A series of photomicrographs of the Kieffer pear flower is presented in figure 7. These show the appearance of some of the sections from which the diagrams were made, and are particularly useful to illustrate what happens to the end of the receptacular stele. In B of the figure, the vascular tissues remaining after the ten primary vascular bundles have FIGURE 7. PHOTOMICROGRAFHS OF SECTIONS OF THE KIEFFER PEAR FLOWER, TAKEN FROM THE BASE UPWARD (TO BE COMPARED WITH FIGURE 6) A, Section from below level D in figure 6. The ten primary bundles have left the stele B, The central stele has taken a circular form C, The ventral carpellary bundles are fused in pairs. The gaps in the stele are opposite the dorsal carpellary bundles D, Section from level C in figure 6. The residual stelar tissue is turning inward under the. base of the axile sac E, At base of locule above the vascular tissues of the receptacular stele. The ten ventral carpellary bundles are distinct F, At base of axile sac, showing the end of the stelar tissues projecting in the center G, Median section through locule, showing the ventral carpellary bundles after departure of the vascular supply to the ovules H, Section from level B in figure 6. The dorsal carpellary bundles curve downward oyer the top of the lo('ules and ur.ward into the styles, and are cut three times I, Section from evel A in figure 6. The stamen bundles are free from the sepal!lundles, and in most cases from the petal bundles also

16 16 L. H. MACDANIELS departed are arranged in a circular stele. In D the ventral carpellary bundles are differentiated and the remaining vascular tissue is turning inward. In F the beginning of the locules is shown, and the end of the stele in the center, free from the carpels. In I is shown a section through the floral disk at the level where most of the stamens are free from the primary bundles. The external morphology of the quince flower differs from that of the apple in that it has a bowl-shaped floral tube which is larger than the cone-shaped floral tube of the apple. In the quince there is a larger number of ovules per locule, and the separation of the edges of the carpels making up the compound ovary is more distinct (figure 8, C). In the vascular, structure of the quince flower the stamen traces are given off B c FIGURE 8. D '" oiii"-tlc ~~-'-s 0" 1tI. ~ ; : vc-:, ~--. p._."fj &<:IJ. ~ ~.--s DIAGRAMS SHOWING THE VASCULAR SKELETON OF THE QUINCE FLOWER (See legend to figure 4 for meaning of the capital and small letters) E - "D from the sepal and petal bundles at a somewhat lower level than in the apple and the pear, and are shown as distinct from the petal bundle at level C (figure 8, C). Another minor difference is the branching of the dorsal carpellary bundle in the base of the style, to form three pundles (figure 8, A and B). Considering the vascular structure as a whole, the quince would be rated phylogenetically more primitive than either the apple or the pear because of the separation of the stamen traces at a lower level. On the same basis the pear is considered the most advanced. POME STRUCTURE IN THE LIGHT OF COMPARATIVE STUDY A correct interpretation of the fundamental nature of floral parts is to be gained, not from a study of a few species, but from a comparative study of many closely allied forms and their interpretation in the light of the recognized trends in phylogenetic development. As a basis of such interpretation, a view of the development of flowers in the plant kingdom as

17 THE MORPHOLOGY OF THE ApPLE AND OTHER POME FRUITS 17 a whole is necessary. With this background the development within the genus, within allied genera, and within the family, can be interpreted. In general, the phylogenetic development of the flower from simple or primitive types to more complex types is well understood. -'It has been set forth by many investigators, and particularly, in more recent years, by Eames (1931). The flower of.ranunculus illustrates the primitive type (figure 9, A). In such a primitiveoflower all the floral parts are free from one another, with separate vascular bundles leaving the stele at different levels. In Aquilegia, with a similar structure, the end of the stele terminates in weakly developed vascular tissue somewhat above the d~parture of the vascular supply to the carpels. In the evolutionary development of the more complex types, there is an increasing amount of fusion of the vascular supplies to these parts. Further specialization may lead to a reduction in floral parts and to a loss of their vascular supply. Viewed in the light of the whole range of floral structure, the pome occupies an intermediate stage of adnation and cohesion of floral parts and their vascular supply. There is no loss of parts. To derive the pome-flower structure from a simple flower, s,uch as that of Ranunculus, it is necessary to show progressive stages in the fusion of like parts in cohesion and of cycles in adnation. In the Rosaceae, the floral cup or tube is the structure characteristic of the family. On the nature of this tube hinges the interpretation of pome morphology. The interpretation of such a structure as appendicular in nature, assumes the acceptance, not only of the basic idea that a flower is made up fundamentally of appendages borne on an.axis, but also of the principle advanced by Van Tieghem (1868) that the" differentiation of a floral part begins at the point where its vascular supply leaves the stele. If these premises are accepted, the development of the pome as an appendicular structure follows as a natural progression. Its interpretation becomes clear particularly when considered in the light of cohesion and adnation in other families. This interpretation has been set forth at length by Van Tieghem (1878), Hillmann (1910), and Bonne (1928). In interpreting the vascular anatomy of any flower, all vascular structures must be considered; but in many species, and especially those in which the floral parts become fleshy, it must be recognized that vascular bundles are developed which have a physiological rather than a morphological significance. In such physiological modification the bundles branch and increase in number, usually forming a vascular system which ramifies to all parts of the fleshy structure. The secondary vascular structure in the pear flower is shown.in figure 3 (page 11). In mature fruits and other fleshy structures, these physiological bundles may entirely or in part obscure the basic morphological pattern. Thus it is only by the comparative study of many different forms in the early stages of devejopment that the basic anatomical structure is determined. Evidence for the appendicular nature of the floral disk of the Rosaceae is found in a study of the vascular anatomy of many different genera. Such a study shows phylogenetic progression in the cohesion and adnation of floral parts to form the floral disk. External evidence of such fusion within the family to form the floral tube is wanting, as the whole family is characterized by the presence of the floral tube or disk. T!!roughout

18 18 L. H. MACDANIELS this large and diverse family, however, all degrees of vascular fusion are found. Jackson (1934), in the study of Rosa, shows the vascular structure of the flowers of a series of genera which illustrate the progressive phylogenetic adnation of the vascular supply of the stamens with that of the sepals and petals. A similar series is shown in figure 9. In Ranunculus fascicularis (figure 9, A), all the floral parts are free. In Rubus allegheniensis (figure 9, B), the sepals, petals, and stamens have A B c 1< IGURE 9. DIAGRAMS INDICATING THE PROBABLE METHOD OF DEVELOPMENT OF THE FLORAL TUBE OF THE ROSACEAE BY THE FUSION OF FLORAL PARTS AND THEIR VASCULAR SUPPLY A. Ranunculus fascicularis (adapted from Jackson, 1934); B, Rubus allegheniensis; C, Fragaria v1rgtniana The letters s, p, and st have the same signification as in preceding figures. The letter t indicates the floral tube. fused externally to form the floral tube.' The vascular structure of the flower is essentially the same as that of Ranunculus, differing only in that the stamen traces are fused. The recee_tacular stele extends straight through the flower from base to apex in eadi case. In Fragaria virginiana (figure 9, C) there is similar external fusion to form the disk, but the vascular supply of the stamens is fused with that of the sepals and petals. Here, a slight spreading of the stele below the carpels suggests the condition found in Rosa. Examples of the progressive fusion of stamen traces with sepal and petal traces are found in several of the rosaceous tribes. From the standpoint of comparative anatomy, this evidence may be taken to prove conclusively that the floral tube is derived from the fused floral parts and is thus appendicular in its composition. The Rosaceae also give anatomical evidence for the fusion of floral parts in cohesion. Thus, according to Jackson (1934), in Dalibarda and Rubus either the stamen bundles are free-a single trace leading to a single stamen-or several stamen traces are fused at the base into a single trace which branches above to supply several stamens. This condition is reported for Rubus by Hillmann (1910), and has been observed also in the present study. The petal bundles are separate from one another in all genera studied, as are the median sepal traces. The adjacent lateral sepal traces, however, are fused with each other in cohesion and with the petal trace in adnation. Evidence of the separate origin of the lateral sepal traces has not been found in the Rosaceae, as such phylogenetic fusion is apparently complete for the whole family. This fusion is, however, quite in line with what occurs generally in flowers in which the calyx lobes unite to form a tube. To derive the po me-flower structure from a rosaceous flower with a

19 THE MORPHOLOGY OF THE ApPLE AND OTHER POME FRUITS 19 fleshy flo-ra! disk and a superior ovary, it is necessary merely to show the fusion 6f the disk with a 3-5-celled compound ovary. Progressive stages in the fusion of the floral disk with the ovary are obvious irva series of genera such as that shown in the diagrams in figure 10. In Physocarpus (figure 10, A) the floral disk is entirely free from the carpels; the same FIGURE 10. SERIES OF DIAGRAl\1S SHOWING PROGRESSIVE FUSION OF THE FLORAL TUBE IN ROSACEOUS FLOWERS TO THE CARPELS, IN THE PROBABLE DEVELOPMENT OF THE POME FLOWER A, PltysocarpltS opuiifolius; B J Spirea ulmifolia; C, Sorb us Aucuparia; D, Amelllnchier vulgaris; E, Pyrus communis; F, Agrimoff'z,a Eupatoria (A, B, C, D, and F adapted from Bonne, 1928) condition exists in Prunus and some other genera also. In Spiraea (figure 10, B) there is adnation of the floral disk with the ovaries at the base; in Sorbus (figure 10, C) the fusion extends nearly to the top of the compound ovary; in Amelanchier (figure 10, D) the ovary is completely inferior, with the fleshy disk closing in over it; in Pyrus (figure 10, E,~ the ovary is completely submerged. Other genera might be chosen to illustrate each of these stages. The structure of Agrimonia (figure 10, F) is of special interest tn that the floral tube closes over the carpels in much the same way as in:pyrus but is not adnate with them. Fusion of this tube with the carpels' would

20 20 L. H. MACDANIELS give a flower of essentially the same structure as that found in the pomes. Evidence of such phylogenetic fusion is obvious in a series such as that shown in figure 10. In the fusion of the carpels to form the compound ovary, many degrees of cohesion exist in the Rosaceae, from the free achenes of Fragaria and Geum, the associated drupelets of Rubus, and the follicles of Physocarpus, to the fused condition of the pome. Even in the pome the inner edges of the carpels are distinct in the flower, as is indicated by the extension of the inner epidermis in between the carpels in Malus (figure 5, B) and Cydonia (figure 8, C), and by the freedom of the styles in Cydonia and Pyrus. Tbe vascular structure of the carpel of the pome is not complicated, because the two ventral carpellary bundles and the dorsal carpellary bundle are distinct and free. There are, to be sure, anastomosing bundles between the dorsal and the ventral carpellary traces, but these are relatively small, especially in the flower, and appear to have physiological rather than morphological significance. The cohesion in the other cycles is more complex. In the apple, the pear, and the quince, and in pomes generally, the members of one set of five stamens are free from one another but are adnate with the central traces of the opposite sepals. The other fifteen stamens are fused. in groups of three, with the petal bundles. The five single petal traces are free. so far as cohesion is concerned. They are'fused with the lateral traces of the adjacent sepals, however, as well as with the groups of three stamen traces. This complex nature of the petal bundle is confusing if considered by itself, but it can be logically interpreted in the light of comparative studies. In a primitive flower the sepal' normally has three vascular traces. In cohesion the lateral traces of adjacent sepals fuse with each other. This condition is to be found in many families, and in the Labiatae the various degrees of fusion of these lateral sepal traces with each other are used as a basis of species separation. Phylogenetically, therefore, it is probable that in the Rosaceae the lateral traces of adjacent sepals fused, and these in turn fused with the petal trace di[ectly above in adnation. From the evidence found in Rubus and DalibaMa, it is highly probable that the three stamen traces fused with one another in cohesion, and these in turn fused with the opposite petal trace. One of the basic principles of the phylogenetic approach in comparative flower morphology is the fusion of parts that are in close proximity in the same or nearly the same radius. It is thus entirely logical to expect the stamen traces to fuse first with one another. and then with the opposite petal and sepal traces with \vhich they are closely associated. Important evidence for the appendicular interpretation of pome structure is found in the genus Rosa. Here the structure of the fruit is clearly in part receptacular (Bonne, 1928; Jackson. 1934). This is shown by the fact that in the rose flower the receptacle is much broadened before the carpellary bundles are free. These carpellary traces, which are all that remain of the stele, extend downward on the inside of the floral tube from their point of attachment to the sepal, petal, and stamen traces, and finally reach the carpels. The receptacle may thus logically be considered as including the concave fleshy base of the flower up to the point where the

21 THE MORPEOLOGY OF THE ApPLE AND OTHER POME FRUITS 21 carpellary bundles separate from the other vascular tissue. The orientation of the xylem and the phloem is essentially that shown in figure 12, B. An example of how such doubling-back may OCcur phylagenetically is FIGURE 11. PHOTOMICROGRAPHS SHOWING THE VASCULAR STRUCTURE OF RUBUS FLOWERS A, Rttbus allegheniensis. The stamen bundles are free from the sepal and petal bundles. The remaining bundles of the stele extend straight upward into the elongate receptacular knob B, RttbllS odorattts. The receptacle bearing the carpels is flattened. The stelar bundles extend out to the edges of this flattened disk and then recurve sharply to supp1y the carpels

22 22 L. H. MAC DANIELS found in the genu~ Rubus. In a species with a convex receptacular knob, Rubus allegheniensis, the stele extends up into the receptacle and gives off traces to the ovules in regular acropetal succession (figure 11, A). In Rubus odoratus, however, the receptacle is much broadened and flattened (figure 11, B). The bundles supplying the sepals and petals are free from the stamens, and leave the stele at the apex of the pedicel. Above this point the stelar bundles extend outward to the edges of the flattened receptacle, and then turn, through more than a right angle, and extend inward toward the center. From these stelar bundles arises the vascular supply to the carpels. The stamen traces are g~ven off from the stele at its widest point, where the carpellary bundles turn inward. In the pome and mos.t other rosaceous genera, however, the stele is not markedly broadened and the vascular supply to the carpels extends directly into these from the stele. If the pome flower were in part receptacular in nature, a doubling-back of the stelar bundles supplying the carpels would be expected, as is the case with Rosa and with Rubus odoratus. The vascular structure necessary to form a terminal plate or a terminal or an intercalary cup which is receptacular or stelar in its make-up, is clearly set forth by Van Tieghem (1878). The diagrams in figure 12, A, B, and C illustrate the structures which he describes. In figure 12, A, A B c FIGURE 12. DIAGRAMS SHOWING THE VASCULAR STRUC TURE NECESSARY TO FORM A TERMINAL DISK OR CUP THAT IS RECEPTACULAR (PHLOEM TISSUES ARE STIPPLED) A, Terminal disk; B, terminal cup; C, intercalary cup from single meristern; D, intercalary cup from fused phylloclads. The cross (X) marks the position of the terminal stelar meristem

23 THE 1foRPHOLOGY OF THE ApPLE AND OTHER POME FRUITS 23 the s'teie, or receptacle, broadens out, and the vascular bundles curve inward assuming a horizontal PQsition. This is essentially the condition in Rubus odoratus (figure Il, B). In figure 12, B, whicn resembles the structure in Rosa but in more extreme form, the stelar bundles extend upward into the receptacular cup, and then double back sharply downward to supply the carpels. In this downward-extending bundle the xylem is adjacent to the xylem of the upward-extending bundle on the outside of the floral tube, and the phloem faces the inside of the tube, as must be expected in such an inverted bundle. In figure 12, C, the "intercalary cup" of Van Tieghem, the structure is the same as in B except that the stele extends upward from the base of the cup. This would of necessity be the condition in such perigynous flowers as those of Rubus allegheniensis if the floral tube were receptacular. The fact that in this species single bundles leave the stele to supply the floral parts, and the stele itself extends straight up into the receptacle giving off traces to the ovules in acropetal succession without any doubling-back, is strong evidence that the floral tube is not axial. Another anatomical possibility for the formation of a terminal or an intercalary cup that is stelar in its make-up would be a structure composed of laterally fused branches. Such a possibility is shown in figure 12, D. This would be the same as a structure made up of phy1l6clads, in which the principal vascular bundles of the flattened structure form or represent complete steles instead of appendages or leaves. Such vascular structure is found at the base of the immature fig fruit, but this is admittedly a fused inflorescence and not a flower. In the floral disk of the Rosaceae, however, there is no anatomical evidence of such structure. THE MORPHOLOGY OF THE APPLE FRUIT The structure of the apple fruit in transverse section is shown in figure 13. According to the appendicular interpretation, this fruit is made up ~ ft '.. FIGURE 13. DIAGRAMS OF APPLE FRUITS IN TRANSVERSE SECTION, SHOWING GROSS MORPHOLOGY 1 A, Malus pztmila, variety York Imperial; B, Ala/us ioensis.r ft, Floral tube; cl, outer limit of carpel, or Core line; cb, carpellary bundles connecting dorsal with ventral carpel1ary bundles; s, p, de. and vc, sepal, petal, dorsal carpellary, and ventral carpellary bundles, respectively

24 24 L. H. MACDANIELS of a five-carpeled ovary fused with the floral tube. No part of the receptacle is included, except possibly the end of the receptacular stele at the base of the locules. The ovary extends from the center out to the region between the dorsal carpellary bundles and the ring 'of ten primary bundles. In Malus coronaria and some other crab apples, the line of separation between the ovary and the adnate floral tube is marked by an aggregation of stone cells and is a region through which cleavage readily takes place. In some species - for example, Malus ioensis (figure 15, A and D)- the ovary tissues oxidize more rapidly than do the surrounding tissues, and can thus be identified. In some apple varieties - for example, Orenco and Red Sauce - the flesh of many of the fruits may be tinged with red except the central tissues of the ovary (figure 14). FIGURE 14. APPLE, VARIETY RED SAUCE The red pig=ent is distributed through the tissues of the floral tube, but is lacking in the ovary. The core line separates the two tissues The line of demarcation is very sharp and distinct. In horticultural varieties of apples, the tissues of the outer ovary and the inner part of the floral disk may not be sharply differentiated, but are made up of homogeneous ground or fundamental tissue such as is found in any fleshy organ, particularly that associated with the fruit. In this intermediate zone are few vascular bundles. Possibly this lack of vascular tissue is associated with this observed greater susceptibility to breakdown and other storage troubles. The five-carpeled ovary is differentiated into the cartilaginous endocarp, which lines the locules, and the fleshy exocarp, which is fused with the floral tube. In many varieties of apples, a sheet of differentiated parenchyma cells inside the primary bundles forms an irregular ring or a somewhat angular pattern, seen when the apple is cut transversely. These cells differ from the other parenchyma cells in that they are smaller and more elongate (figure 16, A), with somewhat thicker walls; also, there are more inter-

25 THE MORPHOLOGY OF THE.ApPLE AND OTHER POME FRUITS 25 cellular spaces extending along these lines, a structural condition indicating a natural line of c1eav.age. Inside this line the cells of the pulp which make up the exocarp of the ovary are obviously of smaller size than those toward the outside, in the tissue of the floral tube. The pattern formed by this sheet of tissue as cut in cross section is characteristic for anyone variety (figure 15) and has been used in variety identification by Kraus (1916), who published photographs of many cleared sections of the fruit of apple varieties showing variation in pattern. In these cleared sections this tissue appears macroscopically much like the vascular tissue, but it contains no vascular elements. This sheet of tissue is here interpreted as the line of fusion of the ovary and the ad nate floral disk. In some apple varieties this line is regular, as shown in figure 15, E; in others it FIGURE 15. PHOTOGRAPHS OF APPLE FRUITS IN SECTION, SHOWING CORE LINES A and D, Malus ioensis; B, Malus coronaria; C, E, and F, Malus pumila, varieties Ben Davis, York Imperial, and McIntosh, respectively may be irregular, with points extending out between the primary bundles (figure 15, F). In the immature fruit of some varieties there is a zone of cleavage, apparently in this tissue, which makes it possible to lift out the core from the quartered fruit without cutting it further. In some varieties of pears, the limits of the ovary appear to be marked by a concentration of grit cells. This demarcation would be logical, in that in the pear the grit cells are characteristically localized underneath the epidermis of the fruit. The line of fusion of the floral tube and the ovary would mark the fusion of the inner epidermis of the tube and the outer epidermis of the ovary, and the presence of grit cells may be regarded as indicating the position of an epidermis that has disappeared

26 26 L. H. MACDAKIELS in phy 16geny. In Malus ioensis this line of separation is distinctly visible, with parenchyma differing from that of the adjacent fruit pulp and groups of stone cells (figure 16, B). Exterior to the ovary 1S the main bulk of the apple fruit, made up of the fused bases of the sepals, petals, and stamens, or the so-called floral disk or tube. FIGURE 16. PHOTOMICROGRAPHS OF SECTIONS THROUGH CORE LINE OF APPLE FRUITS A, ilia/lls pumila, variety York Imperial; B, Malus ioensis The arrows indicate the position of the core line tissue. s, Stone cells next to the core line; t. parenchyma tissues of the floral tube; 0, tissues of the ovary Obviously, many changes have taken place in the development of the mature apple fruit from the flower. The basic pattern of the vascular structure is well shown by Kraus and Ralston (1916). The greatest enlargement in the apple takes place between the level of the departure of the dorsal carpellary bundles from the stele, and that of the departure of the stamen traces from the sepal and petal bundles. The extensive development of the vascular tissues is of the type characteristic of fleshy tissues generally, and has physiologicall'ather than morphological significance. The floral receptacle can be considered as beginning below the base of the ovary, in the pedicel, at the point where the sepal and petal bundles leave the ventral carpellary bundles and the remaining stelar tissues. It terminates at the base of the locule, where the ventral carpellary bundles leave the remaining stelar tissue and enter the carpel. In many apple varieties, the fleshy part of the fruit does not cover the whole receptacle. In some varieties of pears with a long neck - as, for example, Bosc - the fleshy part of the fruit extends downward along the pedicel for some distance. In such cases there is no clean-cut division line between the cortical tissues of the pedicel and the fleshy neck of the

27 THE MORPHOLOGY OF THE ApPLE AND OTHER POME FRUITS 27 fruit. Here it is well to recall that, according to the stelar theory, the leaf is itself an appendage of stelar origin and the mesophyii is cortical in nature. This is obvious in the continuity of the cortical tissues from the stem into the petioles of leaves. It is not to be expected, therefore, that there would be any clear-cut difference between the cortical tissue of the pedicel, and the fruit tissues where the fruit joins the pedicel. Further, leaves frequently are decurrent on the stem. DISCUSSION In the present study, the evidence favors the interpretation of the floral tube and the fleshy structures of the pome as being made up of modified floral parts rather than of receptacle, or stem. This hypothesis interprets in a satisfactory manner all the conditions found in the floral anatomy of the groups c-oncerned, and integrates that interpretation with the whole field of floral anatomy. The basic criticism of the theory that the floral disk of the Rosaceae and the flesh of the pome are receptacular in nature, is that the evidence upon which the theory is based is largely the result of observation of the gross morphology of a rather restricted group of plants without reference to other groups. The studies dealing with the development of floral structures in ontogeny have undoubtedly been of great value. But external observation of the protuberances as they occur on the meristem in the formation of floral parts, can reveal very little as to the basic nature of those parts. A protuberance on a meristem that is destined to be a sepal or a carpel would appear the same, externally, regardless of whether it were fundamentally stelar or appendicular in nature. Only where the relations of the vascular structures of such a protuberance to the stele are studied, can any indication of its real nature be determined. Vascular study is particularly necessary where floral parts have fused phylogenetically in either cohesion or adnation. The limitations of such ontogenetical studies of the meristem are that they ignore the significance of comparative anatomy and morphology in interpreting the same phenomena. It is generally agreed, among anatomists at least, that the comparative approach is of great importance, especially where matters of phylogeny are concerned. In the opinion of the writer, the evidence given to support the receptacular theory of pome structure can be better interpreted to support the appendicular theory. For example, the occurrence of sepals or of welldeveloped leaves on the sides of the apple is taken to indicate that the apple is stem. This abnormality is just as readily explained, however, on the basis that the apple is made up, in part, of what are morphologically fused leaf bases, or petioles, and that in fruits such as this the blade of the leaf is Qut of its normal position. Misplaced structures of this ~ort are of frequent occurrence in fruits and flowers. The development of normal leaves in the place of any of the perianth parts or the extension of the central part of the flower into a normal stem, as reported by Tukey (1938), offers no difficulty to the appendicular interpretation, and, in fact, may logically be considered to support it. Admittedly, the flower is basically a stem bearing modified appendages. The differentiation of a floral part begins in the same way a~ does the

28 28 L. H. MACDANIELS differentiation of any other leaf or appendage. In the apple the basal whorl of floral parts, or the sepals, are the first to differentiate and the carpels are the last to take form. At least partial differentiation of the floral parts takes place in the summer preceding flowering. In the plants described by Tukey, the flowers in the buds placed on the stocks were only partially differentiated into floral parts, the central apical meristem being undifferentiated. Placing the bud upon a vigorous seedling stock changed the growth status of the bud, so that, instead of continuing to develop as a normal flower cluster, the apical meristem in some of the flowers developed into a vegetative shoot. In the plants observed, all gradations, from vegetative leaf to floral parts, were found in the flowers from the buds placed on the stocks. Such change in growth status is well recognized by horticulturists, and probably may best be attributed to either the presence or the lack of growth-determining substances, or hormones, that control the formation of flowers, or possibly to the nutritional change resulting from placing a flower bud upon a vigorously vegetative stock. If the pome flower is receptacular, anatomical evidence of this should be found in the abnormal flowers reported by Tukey. In the flowers which approached normal structure sufficiently to form a floral tube as indicated by the presence of petals or stamens, the stele should show evidence of forming an intercalary collar or cup after the manner suggested by Van Tieghem (figure 12, C) in which the stelar bundles turn sharply inward and downward to the bottom of the cup before turning upward into the extended vegetative axis. Detailed evidence of the nature of the primary vascular skeleton of these abnormal flowers is not reported. Tukey's diagram of the swollen base. of the vegetative flower, however (figure 4 of the reference cited), shows essentially a normal stele with leaf gaps indicating t\yo-fifths phyllotaxy. There is no evidence of doubling-back or other irregularity. The diagram shows that the vascular traces of the leaves representing floral parts were given off the stele in the normal manner for leaves, and that the whole structure differs from normal shoot growth only in that the leaves are crowded together vertically with short internodes. The structure is not unlike that found in a normal fruit spur of the apple, which has crowded leaves. The anatomical evidence presented therefore supports the interpretation that the floral parts are leaves or leaf-like appendages, and that they are attached to the stele in a normal manner without any indication of the stele being modified as would be essential if a receptacular cup or collar were involved. One of the most important lines of evidence given for the receptacular interpretation of the pome fruit lies in the studies of the differentiation of floral parts. The point is made that the sepals, petals, and stamens begin as protuberances on the meristem, and then are pushed outward and upward by intercalary growth behind the meristem which shoves the parts upward together, as would occur in a stem. However, this would occur only if the floral tube were made up of fused axes, as shown in figure 12, D, and there is no anatomical evidence that this condition exists. In the differentiation of the pome flower the sepals are differentiated first, and are followed in succession by the petals, the stamens, and the carpels. This sequence would be possible if the stelar bundles doubled back upon themselves to form the floral tube, as shown in figure 12, B. There is,

29 THE MORPHOLOGY OF THE ApPLE AND OTHER POME FRUITS 29 however, no anatomical evidence of such doubling-back; and, further, if such doubling-back did occur, it would necessitate the functioning of an intercalary meristem which developed the same bundle, on one side of the floral tube from the base toward the apex, and on the other side from the apex toward the base - truly an anomalous situation. The acropetal succession of formation of floral parts, and the unilateral development of vascular tissues and their orientation in the floral disk, are typical for the development of appendicular structures by a terminal meristem. Further, the interpretation that this floral disk is stem fails to consider the anatomical evidence found in the distinct vascular strands supplying sepals, petals, and stamens, which are present in many species and which indicate the fusion of floral parts. The presence of such separate strands, or traces, is quite meaningless if the floral disk is stelar in nature. The floral disk of the Rosaceae has been considered stelar in nature because upon it are borne the sepals, petals, and stamens, the first two of which are sometimes deciduous. This is not inimical to the appendicular interpretation, because specialized leaves may have articulations and form abscission layers at almost any point. If this is presented as evidence for the stelar interpretation, then the condition in Prunus and some other genera must be explained, in which the floral tube is deciduous at the base of the ovary or at its insertion upon the stele as is normally the condition in leaflike appendages. The reasoning that the complex bundles of the pome flower resemble the complex bundles in a broken stele in that they each supply several organs, might have some weight except for the fact that such reasoning fails to consider the manner in which these structures in the pome flower originated phylogenetically. In other words, this grouping of fused bundles is exactly what would be expected in phylogenetic specialization, according to the established principles of cohesion and adnation. Apparently the crux of the difference in the interpretation of pome structure is the acceptance or rejection of the stelar theory and its implications, especially the hypothesis of Van Tieghem that the differentiation of a floral organ begins where the vascular trace leaves the stele. Certainly this basic approach is gaining rapidly in its acceptance by anatomists, and, coupled with increasing knowledge of the behavior of floral parts in adnation and cohesion as interpreted by studies of comparative anatomy, it has done much to clarify a confused situation. The use of the stelar hypothesis has given a fundamental conception of floral morphology which has universal application and which has n\ade order and meaning out of what seemed chaos - and still is chaos whev the conflicting interpretations of the various ontogenetic studies are considered separately... Recently a number of theories of floral morphology have been advanced, such as those of Troll (1928), Thomas ( ), Thol11pson (1935), Saunders (1937), and Gregoire (1938). The province of this paper does not include criticism of those theories in detail, particularly since the reviews of Bancroft (1935), Kozo-Poljanski (1936)''lArber (1937), and Wilson and Just (1939) have appeared. It is sug-gested, however, that for the most part those theories confuse rather thatf clarify our ideas of floral morphology, and make unnecessarily complex a sit-

30 30 L. H. MACDANIELS uation that may be interpreted in a more straightforward manner. It is, of course, recognized that different groups of flowering plants have developed along widely divergent lines and may even have had different origins. Even so, the concept of the flower as basically an axis with specialized and modified leaflike appendages appears to be capable of the widest application and to fit the observed facts. The exceedingly complicated and diverse structures found in some families offer no difficulty for this concept of the flower, for such specialization is what may be expected in evolution. In the anatomical approach to the interpretation of floral structures, the differentiation of a stamen or a carpel is recognized as a function of the primary apical meristem, or growing point, even as with other leaves or appendages. Whether or not these structures achieve the characteristic form of stamen or carpel and develop sporogenous tissue, is apparently dependent upon the action of growth-determining substances, or hormones, the development of which is determined by environmental conditions such as length of day, temperature, drought, or other factors. A basic criticism of the work of many investigators in the field of floral morphology is that, in their attempt to prove certain theories, these workers have lost sight of the fact that the plant is a functioning organism. Any concept of a plant, to be useful and realistic, must take into consideration all the complex factors that enter into its structure and development. Thus the study of its physiology, anatomy, and ontogeny each must be included before sound 'conclusions or interpretations can be made. The work of Gregoire (1938) is particularly at fault in this regard, in that it holds that the floral parts arc fuqdamcntally different from the vegetative organs. Certainly all can agree that each plant in its life span goes through a cycle - beginning with the fertilization of the egg, then passing through seed, vegetative, and reproductive stages, and ending with the production of gametes which fuse to start the new cycle. The cycle is simplest in the annual plants, in which the vegetative plant dies after the production of a new embryo; but these are not basically different from the perennials. During development from the embryo within the seed to the formation of flowers, it is the same meristem that is continuously functioning, forming branches and leaves. In many plants the transition from normal leaf to floral part is gradual. through a series of bracts with progressive reduction in size and change in form. The change in the meristem from the production of vegetative leaves to the production of an inflorescence ane! floral parts, can be initiated by changes in temperature, length of day, water supply, and other factors. A plant may complete this change at a height of a few inches or many feet. With this concept it is logical to consider the terminal meristem as an exceedingly flexible structure forming both branches and leaflike organswhich latter, regardless of function, are basically the same, at least as far as their ontogenetic and anatomical relationship to the stele are concerned. The claim that the apical meristem is changed in its fundamental nature as soon as it begins to differentiate floral parts, is certainly not substantiated by the anatomical evidence, which now includes a wide range of material in comparative studies.

31 THE MORPHOLOGY OF THE ApPLE AND OTHER PO':'.IE FRUITS 31 SUMMARY A review of the evidence on the interpretation of the morphology of the pome flower as largely receptacular or largely appendicular in nature, and a critical study of a number of pomaceous and other rosaceous genera, support the conclusion that the pome flower is appendicular in its derivation and structur~. The appendicular interpretation is consistent with the comparative anatomy of the whole range of families, and with the nature of the inferior ovary generally. Phylogenetically the pomc flower has arisen from the fusion, by adnation, of the fleshy rosaceous floral tube with the compound ovary. The floral tube, composed of the fused bases of the stamens, petals, and sepals, enlarges greatly as the fruit matures, and forms the larger part of the pome fruit. The so-called "co're line" of the apple is interpreted as representing the line of fusion between the floral tube and the ovary. According to the appendicular interpretation, the fruits of the apple, the pear, and the quince are described as fleshy accessory fruits, made up of a five-carpeled ovary, with cartilaginous endocarp and fleshy exocarp, united with a fleshy floral tube or disk consisting of the fused bases of the sepals, petals, and stamens. Variation in pome structure is chiefly in the following characters: in the number of carpels, which may vary from 2 to 5; in the nature of the endocarp, which may range from bony in Crataegus and some other genera to cartilaginous; and in the degree of completeness with which the ovary is covered by the fleshy floral tube.. REFERENCES ARBER, AGNES. The interpretation of the flower: A study of some aspects of morphological thought. Bio!. rev. 12: BAILEY, L. H. Pyrus. In The standard cyclopedia of horticulture, 5: BANCROFT, HELEN. A review of researches concerning floral morphology. Bot. rev. 1: BLACK, CAROLINE A. The nature of the inflorescence and fruit of Pyrus Malus. New York Bot. Gard. Mem. 6: BONNE, GABRIELLE. Recherches sur Ie pedicelle et la fleur des Rosacees, p EAMES, ARTHUR J. The role of flower anatomy in the determination of angiosperm phylogeny. Internatl. Congo Plant Sci. Proc. 1926, 1: The vascular anatomy of the flower with refutation of th.e theory of carpel polymorphism. Amer. journ. bot. 18: GRAY, ASA. The elements of botany for beginners and for schools, p (Gray's lessons in botany, revised edition.) GREGOIRE, V. La morphogenese et l'autonomie morphologique de l'appareil floral. I. Le carpelle. Cellule 47: ;1 HILL, J. BEN, OVERHOLTS, LEE 0., AND Popp, HENRY W. Fruits,.' seeds, and seedlings. In Botany, a textbook for colleges, p '1936.

32 32 L. H. MACDANIELS HILLMANN, AUG. Vergleichend-anatomische Untersuchungen iiber das Rosaceenhypanth. Bot. Centbl., Beihefte 26: JACKSON, GEMMA. The morphology of the flowers of Rosa and certain closely related genera. Amer. journ. bot. 21: KOZO-POLJ AN SKI, B. On some "third" conceptions in floral morphology. New phytol. 35: , KRAUS, E.]. The pollination of the pomaceous fruits. 1. Gross morphology of the apple. Oregon Agr. CoIl. Exp. Sta. Research bul. 1, part 1: Variation of internal structure of apple varieties. Oregon Agr. ColI. Exp. Sta. Bul. 135: KRAUS, E.]., AND RALSTON, G. S. The pollination of the pomaceous fruits. III. Gross vascular anatomy of the apple. Oregon Agr. CoIl. Exp. Sta. Bul. 138: MACARTHUR, MARY, AND WETMORE, R. H. Developmental studies in the apple fruit in the varieties McIntosh Red and Wagener. I. Vascular anatomy. Journ. pomol. and hort. sci. 17 ( ) : PAYER, J.-B. Traite d'organogenie comparee de la fleur, p RENDLE, ALFRED BARTON. Pomoideae. In The classification of flowering plants, vol. 2, p ROBBINS, WILLIAM J., AND RICKETT, HAROLD W. Kinds of fruits. In Botany, p RYDBERG, PER AXEL. Morphology. In A monograph of the North American Potentilleae..columbia Univ., Dept. Botany. Mem. 2: SAUNDERS, E. R. Floral morphology: a new outlook, with special reference to the interpretation of the gymecenm, 1: SHAW, J. K. Abnormal behavior of newly set Oldenburg buds. Amer. Soc. Hort. Sci. Proc. 36 (1938) : STANFORD, ERNEST ELWOOD. General and economic botany, p THOMAS, H. HAMSHAW. The old morphology and the new. Linn. Soc. London. Proc., 145th session: THOMPSON, ]. McLEAN. The acarpous nature of modern flowering. Sixth Internatl. Bot. Congo Proc. 2: TROLL, WILHELM. Organisation und Gestalt im Bereich der Blute. Gesamtgebiet Wiss. Bot. Monog. 1: TUKEY, H. B. The occurrence of apple blossoms with prolonged central axes and its bearing upon flower morphology. Amer. Soc. Hort. Sci. Proc.35 (1937) : VAN TIEGHEM, PH. Recherches sur la structure du pistil. Ann. sci. nat., Bot., ser. 5, 9: Anatomie de la rose, et en general caracteres anatomiques des axes invagines. Soc. Bot. France. Bu!. 25: WILSON, CARL L., AND HABER, JULIA M. The fruit. In An introduction to plant life, p WILSON, CARL L., AND JUST, THEODORE. The morphology of the flower. Bot. rev. 5: