Other formats

    TEI XML file   ePub eBook file  


    mail icontwitter iconBlogspot iconrss icon

Tuatara: Volume 27, Issue 1, August 1984

Principia Botanica: Croizat's Contribution to Botany

page 26

Principia Botanica: Croizat's Contribution to Botany


Croizat's 1961 book Principia Botanica is reviewed in relation to previous and subsequent work by other botanists. It is shown to be a major contribution to developing effective methods for analysing and solving problems concerning the geographical distribution, morphology and systematics of plants. Croizat's analysis and synthesis of the factors space (geographical distribution), time (phylogeny) and form (morphogeny/symmetry) in evolution, and how this relates to particular groups of plants is discussed in detail.

Keywords: biogeography, botany, carpel, Croizat, homology, leaf, morphogeny, morphology, phylogeny, symmetry.


The main theme permeating Croizat's Principia Botanica (1961) is expressed beautifully in the terza rima from Dante used by Croizat to head the work:

Fai come quei che la cosa per nome

Apprende ben, ma la sua quiditate

Veder non puo, se altri non la prome.

(Paradiso XX, 91-93)

(“You are acting like someone who knows something by name perfectly well, but will never understand its true nature if someone else does not resolve it.”)

In the Principia Botanica Croizat sets himself the task of developing basic concepts which will lead to a deeper knowledge of the quiditate, or true nature, of plant structure. One would receive the impression, judging from most introductory texts, that the problems Croizat deals with, for example “What is a leaf?” and “What is a carpel?”, have been answered adequately years ago. Surely a keen student can easily acquire knowledge of a particular plant organ by turning to the glossary often provided in modern texts and find a simple, easily learned definition? Comparative plant morphology defines the various plant organs essentially in terms of given categories. The plant shoot is considered as being fundamentally composed of two types of elements, leaf-like (phyllome) and stem-like (caulome), and the plant root comprises a third element. These elements are imagined to have been modified in many different ways to result in the variety of form which is the subject matter of comparative morphology. The flower, for example, is explained as being an aggregate of modified leaf-like appendages, some of which bear ovules.

The morphological research programme which effectively began in 1790 with the publication of Goethe's Versuch die Metamorphose der Pflanzen zu erklären (see Arber 1964 for an English translation and commentary) gained momentum rapidly. Over the last two centuries it has led to the accumulation of an immense body of factual knowledge. These facts naturally constitute problems—they must be explained. As outlined above, traditional ideology explains a particular organ as a modification of one of the three pre-existing categories. Morphological problems are presented by organs which do not fit easily into the conceptual framework offered by the accepted categories. In these cases botanists, who must be technically highly skilled, are often inclined to undertake research on the microscopic page 27 structure, in particular the vascular anatomy, of an organ in order to solve the problem posed by the question “To which category does this organ belong?” The vascular anatomical criteria used to establish the homology of particular organs with one or other category were largely developed by van Tieghem (e.g. 1871). But are the categories satisfactory? Or are there fundamental flaws? Croizat suggests that there are, and over the last twenty years his views have been echoed by thoughtful botanists from many different fields.

A fundamental problem with the traditional body of theory is that the categories used in the definition “explaining” the organ are themselves understood poorly, if at all. One example is the definition of “carpel” as a “leaf-like organ bearing ovules”. Apart from such thoroughly un-leaf-like carpels as those of Arachis, the peanut, and Hakea with its interfascicular cambia, the definition is of course useless, unless we have a prior, adequate understanding of what we mean by “leaf”. The man in the street may feel that a leaf is simply a plant part which is flat and green, but for a botanist familiar with even a small part of the range of plant diversity this definition fails very soon. The fact is that we simply do not have sufficient understanding of what consititutes “leaf” to make the simple homology of “carpel” with “leaf-like organ bearing ovules” mean anything useful at all.

Our morphology must be based on a minimum of sound and fundamental concepts, and Croizat argues that this is precisely what is lacking from present day botany, since most authors have been content to simply homologise among themselves completely unanalysed organs.

For many botanists the main problem with the traditional research programme is a practical one—it has simply ceased to provide new ideas which can be used with effect to attack old problems. Thus in 1946 Dormer wrote:

“A striking feature of botany as the science exists at present is the lack of any coherent body of comparative morphological doctrine dealing with the angiosperms”.

Twenty years later Sattler (1966) echoed these sentiments:

“Relatively little progress has been made during the last 100 years in the study of the ‘comparative’ morphology of higher plants.”

This is surely cause for a reassessment of the bases of our knowledge and, even more importantly, our methodology.

Of course, the traditional research programme is by now such an integral part of botanical training and research that it would appear to be extremely difficult to change it. Meeuse (1966:29) has stated:

“Both Takhtajan…and Eames…* whilst adducing corroborative evidence in their argumentation, minimise or misrepresent contradictory findings and often dismiss dissident opinions and alternative theories with a summary self-assurance that verges on contempt. They write with so much self-assumed authority that the inexperienced student may be completely overawed and is likely to accept their apodictic statements as the “last word’ in a particular branch of botany. Adoption of scientific principles by accepting the teachings of an “authority’ was a medieval (“scholastic’) tradition—‘Galenus dixit’—it is certainly not promotive of educating a budding scientific mind towards independent thinking.”

And in a sympathetic review of Meeuse's book Cutter 1966 has stated, rather more diplomatically but to the same general end, that:

“Plant morphologists, in general, tend to be a somewhat conservative page 28 group. One has only to read the reception accorded to the hypothesis put forward in 1908 by W. H. Lang, on alternation of generations to concede that, historically, novel viewpoints in this field have tended to encounter opposition”.

Despite this conservatism and invocation of authority, both hardly conducive to scientific progress, there is every indication that the botanical literature of the past decade reflects a growing awareness of problems raised in the Principia Botanica, and also of the value of many of the suggestions made in Croizat's work towards improving the situation.

The Evolution of the Carnivorous Plants in Space, Time and Form

It will be disconcerting to many botanists that Croizat's analysis of plant form begins with a thorough analysis of the dispersal of the carnivorous plants and their allies. Step by step Croizat methodically establishes that Droseraceae, Nepenthaceae, Sarraceniaceae, Lentibulariaceae and their relatives have, in fact, evolved stressing the same ancient, cardinal biogeographic nodes as angiosperm development itself. This conclusion is totally novel and thus, of course, unexpected, but it is based on a well documented argument which, while being lengthy and complex, is logically reasoned. Rather than agree or disagree with the conclusion on the spot the reader would be best advised to read Chapter 2 of Principia Botanica first, and then decide where he or she stands. The dispersal of the carnivorous plants (for example the striking vicariism displayed by Nepenthaceae with respect to the main massing of carnivorous forms in Australia, and also to the closely allied Dioncophyllaceae in West Africa) indicates that these taxa are not derived independently from other extant angiosperm families, as is commonly thought, but are instead the result of differentiation of a widespread ancestor. The actual tracks featured in their dispersal (for example Usambara (Tanzania)—Madagascar—Drakensbergs (South Africa); Duida-Roraima (Venezuela)—Futadjallong (Guinea)—Cameroon) point to an ancestor as old as angiospermy itself. As Croizat states, these ideas can be related closely to those of Huxley (1888) on the ancestry and evolution of Gentiana. Nevertheless, Croizat's analysis is considerably more extended, and provided early evidence for the geologically hybrid nature of North America (Princ: Fig 8, p. 79 and see p. 11 this issue) (geological heresy in 1961, but see Nur and Ben-Avraham, 1982), as well as providing strikingly novel interpretations of carnivorous plant evolution in space and time.

Maguire and Ashton (1977) have commented that Croizat's work is “studded with prescient insight” but in a review such as the present it is perhaps more important to emphasise Croizat's method rather than apparent “prescience”—his conclusions are due, after all, to effective analysis rather than lucky guesswork!

It is traditionally assumed that the carnivorous plants form an assemblage that is “artificial” in evolutionary terms, resulting from convergent offshoots of various extant taxa such as Saxifragaceae, Scrophulariaceae etc. This assumption is a consequence of essentialistic thinking in morphology whereby leaf, stem and root comprise the original, irreducible categories. Notorious among traditionally minded botanists are carnivorous plants such as Utricularia (Lentibulariaceae) in which it is often impossible to distinguish organs corresponding to any of leaf, stem page 29 or root. The pre-evolutionary morphologists, with their a priori notions of the fundamental nature of the categories leaf, stem and root, felt that the ideal angiosperm would naturally display these organs in their purest, most unsullied form. Thus pre-evolutionary doctrine regards these plants and their organs as degraded—to the extent that in them the pure morphological categories are no longer recognisable. That these plants must cause discomfort is evidenced by the infrequency with which they are discussed and illustrated. The transition from pre-evolutionary botanical thought to the acceptance of an evolutionary interpretation must be recognised as a great conceptual advance. But the necessary revisions of all aspects of botany were not carried out. In fact, as Sattler (1974a) has established, botanical research and teaching is plagued still by the pre-evolutionary essentialistic morphology. Nowhere is this more clearly evident than in the maintenance, albeit disguised, of the dogma of leaf, stem and root as fundamental and primal. This dogma is itself perhaps nowhere more firmly established than in the conventional interpretation of the “derivative” nature of the carnivorous plants.

This pseudoevolutionary interpretation proceeds as follows: Given that in angiosperms leaf, stem and root are “primitive’ (the old concept in its new guise), and given that many carnivorous plants and their allies show no distinction between these categories, then it follows that the carnivorous plants must be derived. But as, Croizat contends, any evolutionist must agree that the first premise of the argument is unwarranted. In fact, Croizat's analysis of the carnivorous plants indicates that many of them, and their allies, e.g. Podostemonaceae, represent an early level of angiosperm evolution at which the modern leaf, stem, root distinction has not yet been attained.

Over recent years evidence has continued to accumulate which points to the consanguinity of the carnivorous plant assemblage. For instance, Marburger (1979) examined the microscopic structure of the stalked and sessile glands of Triphyophyllum peltatum (Dioncophyllaceae). Unlike gross morphology, this is an aspect of anatomy least likely to show convergence if the groups are truly unrelated. Nevertheless Marburger found that the structure of the glands were “remarkably similar” to those of Drosophyllum lusitanicum of the Droseraceae. Likewise, Adams & Smith (1977) have illustrated the many morphological similarities, both gross and microscopic, among the five genera of pitcher plants.

Lentibulariaceae is closely linked with several other carnivorous groups through its very diverse members. The whorled Utricularia tabulata connects the family with the Hippuridaceae. Through the U. “avesicaria” group Lentibulariaceae is allied with the Podostemonaceae. The pitchers and traps of the U. “Dichotoma-monanthos” aggregate, as well as Genlisea and Cephalotaceae establish morphogenetic links with the Nepenthaceae.

At this level of evolution the transition from floral zygomorphy to actinomorphy is easily achieved—for instance through Cladopus, Dicraeia, Inversodicraea, Castelnavia, to Jenmaniella, to Dalzellia, Apinagia, Loncostephus and Tulasneantha all in Podostemonaceae. Thus the zygomorphic flower of Lentibulariaceae is not at all necessarily remote from the actinomorphic flowers of Droseraceae, Nepenthaceae and Cephalotaceae. The flower of Parnassia connects Droseraceae and Saxifragaceae. Through the Sarraceniaceae, Nepenthaceae and Dioncophyllaceae the carnivorous plants are connected rather directly with 20-30% of angiosperm families. For example, Airy Shaw (1951) and page 30 Schmid (1964) provide evidence for the affinities of the Dioncophyllaceae with the huge assemblages around Flacourtiaceae, Guttiferae and Droseraceae. The Podostemonaceae are vegetatively closely bound with the Hydrostachyaceae. Florally the former family can easily be interpreted as the result of the reduction of the unisexual inflorescence of the latter family to a lone ovary with the addition of a few stamens. Together these two families furnish the minimum common denominator for a vast range of angiosperms. For example, the close affinity of Hippuridaceae and Halorrhagidaceae is widely acknowledged—the fusion of four Hippuris carpels is clearly competent to produce a normal halorrhagidaceous flower. Thus Hippuridaceae form a morphogenetic link between Podostemonaceae on the one hand, and Halorrhagidaceae, Gunneraceae, Lythraceae and Onagraceae on the other.

Lentibulariaceae and Droseraceae are bound by Byblidaceae (and Aldrovanda) which also serves to bring together Cheiranthera (Pittosporaceae), the actinomorphic Pittosporaceae and the more or less zygomorphic Ochnaceae.

The conclusion of Croizat's analysis thus point to the carnivorous plants and their allies as representing a level of evolution underlying all of modern angiospermy, and not at all as a derivative group.

It may be pointed out here that unless the botanist is prepared to take Croizat's conclusions on faith, which is hardly what the author intended, he must first gain a knowledge of the diversity of plant form rather wider than that taught in degree courses at most universities. An excellent way of achieving this basic familiarity is through living and studying in the vicinity of tropical rainforest, although this must be supplemented with time in the herbarium, garden and library. Illustrated floras of large tropical regions, epitomised by the Flora Malesiana, are invaluable. One or two volume synopses of angiosperm families are of limited use only, works by Baillon (1871-1888) and Engler & Prantl (1924-) give a much more satisfactory impression. Croizat's sound advice to the student is clear—look hard and long at the plants and the pictures, and ignore written description, to begin with, as much as possible.

Analysis of Floral Morphology and Angiosperm Systematics

After analysing the transection through angiospermy represented by Podostemonaceae—Theaceae—Lythraceae, Croizat continues his analysis of floral morphogeny with an examination of the level of floral evolution attained in the Amentiferae (Fagaceae, Betulaceae, Juglandaceae, Balanopsidaceae, etc.) Although Croizat's interpretation of the evolution of the modern flower is contained in his analysis of the carnivorous plants, the main argument is in Chapter 4 of the Principia. This chapter provides the factual basis for interpretation of Hamamelidaceae as representing morphogenetic transition between the level of floral organisation attained in the Amentiferae on one hand, and Davidiaceae, Garryaceae, Nyssaceae, and Cornaceae on the other. Thus a further transection of angiospermy is examined, and in the process Croizat's novel analysis of floral morphogeny emerges.

The nature of the flower is one of the most interesting and popular problems in botany. The range of floral diversity can be exemplified by considering a Magnolia or monimiaceous flower together with a single tiny flower of a Betula catkin, fig sycone or saururaceous inflorescence (see Fig 1 a and b).

page 31

For various historical reasons the former kind of flower came to be looked upon as “primitive”, and the latter kind as “secondary”, being derived, somehow, from the first. Guedes (1979:325) has succinctly described the process whereby:

“I. W. Bailey became interested in wood, and generations of students had to learn the twists and turns of deriving Nothofagus from Drimys.”

These “twists and turns” are still taught in some universities but a growing number of botanists feel that rather than deriving one modern type of flower from the other we must rationalise the two in terms of evolution from a common ancestor. The actual structure of the ancestor may never be known, but hopefully we can begin by identifying some main morphogenetic tendencies in the evolution of modern forms. These tendencies must be established on a broad basis of factual comparison. Theorising on the basis of study of only a few forms will lead in many cases to unwarranted extrapolation, or simply uninformed guesses.

Fig 1a. A flower of Magnolia. (From Croizat 1964b: Fig 6). b. An inflorescence (since stamens are present on the wall of each carpel) of Houttuynia (Saururaceae). (From Croizat 1964b: Fig 7).

Fig 1a. A flower of Magnolia. (From Croizat 1964b: Fig 6).
b. An inflorescence (since stamens are present on the wall of each carpel) of Houttuynia (Saururaceae). (From Croizat 1964b: Fig 7).

Floral evolution in Magnolia, Monimiaceae, Houttuynia and Betula can be easily rationalised within the concept of an ancestral strobile or cone. This is a matter of visual evidence (see Fig. 1 a, b) and is widely agreed upon. But what are the processes responsible for the reduction of this strobile to form the modern flower in all its diversity? Morphological definition establishes that technically the female sexualised scales in modern Magnolia must be termed carpels, whereas those in modern Betula and Houttuynia must be termed female flowers. But it must be noted that the development of stamens, or even incipient stamens as staminodes, on the carpellary wall is a simple matter in angiosperms (e.g. Yampolsky 1920)—it is simpler at least than deriving the Betula or Houttuynia type of flower from the Magnolia type of flower! This former process instantly relegates the Magnolia flower (Fig. 1a) to technical status of inflorescence (Fig. 1b) and the distinction between flower and inflorescence becomes blurred.

Thus, closely tied to any concept of the flower is the concept of carpel, and the limits between these two concepts.

In 1790 Goethe introduced the idea of the angiosperm carpel as being an organ of “leaf nature” bearing and enclosing ovules (the text-book “sporophyll”). This concept was also put forward by A. P. de Candolle in page 32 his influential text-book (1827), and succeeded in establishing itself as botanical dogma. However, of great relevance is Gregory's (1956) comment that:

“although no satisfactory definition of a leaf is thus possible I shall assume we all know what we shall be talking about”.

It seems clear that the identification of “carpel” with “metamorphosed leaf” is premature at best, and is probably tautological, representing no conceptual progress whatever. Not surprisingly, Satina (1959:146) states.

Fig 2a. Placenta with subtending scale. b. Placenta in part “transfused” within the subtending scale. c. Placenta fully transfused within the subtending scale and creating the illusion of the “carpel” as a unitary “leaf carrying the ovules upon (or toward) its margins”. (From Croizat 1964b: Fig. 12).

Fig 2a. Placenta with subtending scale.
b. Placenta in part “transfused” within the subtending scale.
c. Placenta fully transfused within the subtending scale and creating the illusion of the “carpel” as a unitary “leaf carrying the ovules upon (or toward) its margins”. (From Croizat 1964b: Fig. 12).

“It is obvious that the nature of the carpel is far from being settled and that additional information from different branches of botanical science is necessary for a radical evaluation of its true origin”.

In contrast with Goethe and de Candolle's homology of unanalysed leaf and unanalysed carpel, Croizat provides an analysis of the carpel which leads to fundamentally different conclusions. This analysis leads to the interpretation of gynoecial morphogeny as an interplay of two distinct elements—fertile placenta and sterile scale. The range of morphological possibilities rationalised by this single morphogeny is illustrated in Fig. 2. This analysis is becoming more widely accepted among floral morphologists. For example, Sattler (1974b) discusses the important consequences of accepting Croizat's:

“great achievement of demonstrating that ‘placenta’ and “gynoecial appendage’ (sterile scale) must be separated conceptually to cope with the whole gynoecial diversity including the carpellate condition in which gynoecial appendage and placenta(e) form a compound unit ((Fig. 2c))”.

Thus, the carpel itself does not form a basic gynoecial unit at all. (Croizat's scale + placenta may be compared with Melville's (1962) gonophyll equals leafy blade + epiphyllous fertile branch, but see Croizat 1964a:799 for further comment.) In the past the traditional Goethian interpretation has been forced upon many gynoecia, and

“as a consequence one has to assume arbitrary and imaginary lines of demarcation for which no evidence is available”. (Sattler 1974b:28).

Sattler describes this activity as “pseudoscience” (1974b:23). Likewise page 33 Lorch (1963) emphasises the practical difficulties of the Goethian carpel and concludes a valuable review of the history and inevitable failure of the concept by writing:

“the carrying afield of ostensive definitions into regions where they do not apply, necessarily involves the crossing of the ‘high confidence’ boundary through the surrounding zone of decreasing confidence to the peripheral zone of meaninglessness”.

Much recent work has concentrated on gynoecia in which the ovules are borne not on the “foliar” gynoecial appendages as Goethe's theory predicts, but rather on the axial (cauline) portion of the gynoecium. This is one morphological possibility easily explained with reference to Croizat's analysis of gynoecial morphogeny (see Fig. 2a, b).

For instance, Maze, Dengler and Bohm (1971) and Maze, Bohm and Beil (1972) describe floral development in graminaceous plants in which ovules develop from the floret apical meristem. They interpret the ovules as cauline.

Macdonald and Sattler (1973:1972) conclude that

“the nucellus of Myrica gale, as in a number of other taxa, must be considered terminal. The same conclusion applies to the ovule.”

They also observe (1973:1973) that

“Croizat's (1960 (Princ), 1962 (1964a) and Meeuse's (1966) frameworks offer a more satisfactory explanation of the morphological nature of the floral parts of Myrica gale than does the classical concept”.

(They also note (1973:1966) that if these former concepts are used “M. gale would be interpreted to be primitive and perhaps preangiospermous (rather than considerably derived).” It is clear that these modern concepts of floral morphogeny have important consequences!)

Macdonald (1979:150) has described the wall surrounding the ovule in Myrica californica with the neutral term “circumlocular” and states: “It (the wall) is certainly not carpellary”. Stamens may be located on the wall.

Moeliono (1970:245) concludes an extensive study of the Centrospermae thus:

“I have rejected the concept of the carpel as a sporophyll. I have interpreted the ovary of the Centrospermae as being composed of a sterile abaxial part, formed by one or more congenitally fused phyllomes, and a fertile axial part (placenta,) the ovules of which are axis borne and thus of cauline origin”.

Van Heel (1973:273), describing the flowers and fruits of Scaphocalyx spathacea (Flacourtiaceae), stated that:

“although this pistil in its regularity would seem a most rational system, it cannot be understood under the classical carpel theory”.

Posluszny and Sattler (1976:1146) studied the floral development of Najas flexilis (Najadaceae) and showed that:

“the ovule is initiated in an exactly terminal (basal) position and not laterally on the gynoecial appendage (i.e. the gynoecial wall). Therefore it is not possible to apply the (Goethian) carpel concept to this situation”.

Pauze and Sattler (1979) discussing Ochna atropurpurea state:

“if it is agreed upon that axillary branches arise from the stem (caulome), then the axillary placentation of this species is also of cauline origin. It then follows, as a logical consequence that the gynoecium is acarpellate”.

Sattler and Perlin (1982) conclude work on floral development in Nyctaginaceae by observing that:

“the data of the present study do not support a carpellary interpretation of the gynoecium because the ovule is not borne on an appendage (megasporophyll)… Those who find the conclusion that the Nyctaginaceae page 34 have no carpels too unpalatable might consider a re-definition of the carpel concept…If “carpel” is not defined as a foliar appendage (megasporophyll) which Bears and encloses ovule(s), but instead as an appendage which Encloses ovule(s), then the Nyctaginaceae are definitely carpellate”.

Once it is accepted that the carpels can most easily be interpreted merely as sterile scales which surround and may have fused with the placenta, the distinctions between ovary wall and integument, and thus between ovary and ovule, and between inflorescence and flower begin to break down. Van Heel (1967) agrees:

“In general I can follow Croizat in that there is a tendency for reproductive units to be sheathed by ever more telescoping coverings, leading from ‘ovules to ovary’ and from “flower to inflorescence’ or vice versa…The rating of the reproductive units (of Scyphostegia) as ovules (as opposed to ovaries)…does not mean that we are closer to an understanding of their nature”.

Poszluszny and Sattler (1974) point out that the Potamogeton “flower” is “a structure somewhat intermediate between a flower and inflorescence …” Likewise, Charlton (1972) studying the inflorescence of the juncaginaceous Triglochin maritima suggests that “flower and inflorescence might well not be defined as different organ categories”. Sattler (1973) in his landmark “Organogenesis of Flowers” states.

“such structures which cannot be classified because they show characters of both a ‘flower’ and an “inflorescence’ are considered as flowers in this book. This does not mean that they are homologised with flowers; they are included simply for practical reasons. As there are intermediates between an “ovule’ and a “pistil’ intermediates occur between a “flower’ and an “inflorescence”’.

Botanists in New Zealand are aware of the essentially academic debate concerning the correct categorical pigeon-holing of the flower/inflorescence of Centrolepidaceae. Unfortunately the evolutionary view—that it is an “inflorescence” “caught” evolving into a “flower” (Princ. 516)—has not been considered in discussion.

Croizat's analysis of the angiosperm flower into whorls of scales representing a wide range of morphogenetic potential, plus the ovule-bearing placenta, leads naturally to the idea of the ancestral strobile or cone-like organ which has evolved by reduction, and redistribution of stamens and carpels into the modern flower. One defining feature of the flower is the germination of the pollen on a stigma formed from a portion of the carpellary scale. However, as indicated above, if the Goethian concept of carpel is rejected, then carpellary scales are in principle no different from the ovular integuments. In most angiosperms the stigma is provided by the “ovary wall”. However, in the urticaceous Myriocarpa and Leucosyke the inner seed coat forms, by means of a tubular projection, what “one might almost call ovular stigmas”. (McLean and Ivimey-Cook 1956:1440). In Myrica gale the stigma is provided by what can easily be interpreted as an outer seed coat. The gymnospermous Gnetum and Ephedra produce a thin micropylar tube (tubillus) from their inner seed coat, furnishing a normal, functional stigma. Thus the stigma can be produced by any one of the sets of scales surrounding the nucellus—inner seed coat, outer seed coat, ovary wall—and the boundary between gymnospermy and angiospermy begins to wear very thin indeed. It seems, in fact, to be artificial.

For Croizat, three alternative “morphogenetic gates” to the development of angiospermy from gymnospermy are represented by the flowers of Magnoliaceae and Nymphaeaceae; by the flowers of Monimiaceae and the page 35 inflorescences of Ulmaceae; and by the inflorescences of Betulaceae and flowers/inflorescences of Hamamelidaceae. In these groups the various flowers and inflorescences display clear relationships with, and evolution from, preangiospermous strobiles, although the new structures are identified by different names due to technical, descriptive distinctions. The flower of Betula (c.f. that of Houttuynia, fig. 1b) is formed through the development of the scales immediately surrounding the nucellus to form a stigma. Thus the strobile develops into an inflorescence, the catkin. In contrast, the Magnolia flower (Fig. 1a) is built up from the whole ancestral strobile, not just from the circumnucellar region, resulting in a pan-strobilar flower. Although the distinction between flower (Magnolia) and inflorescence (Houttuynia) may rest simply on the location of stamen development, the structure of both is easily accounted for by evolution from a preangiospermous strobile.

“Peloria” are structures produced, occasionally, at the apex of, for instance, Digitalis inflorescences, composed of the lengthwise condensation of part of the inflorescence. Several of the zygomorphic flowers fuse to form a strikingly new, virtually actinomorphic “flower”. In the Hamamelidaceae Croizat's concept of the role of peloria in floral morphogeny is easily seen. The typically zygomorphic amentiferous (circumnucellar) flowers of Distylium and Rhodoleia pass by a process of pelorisation into the (panstrobilar) flowers of the same genus (as abnormalities) or of Sinowilsonia, Sycopsis, etc.

It should be noted that, superficially at least, Croizat's-ideas on floral evolution are strikingly similar to many of those developed in the excellent, but almost totally overlooked work of A. D. J. Meeuse (e.g. 1975, 1982). Close parallels seem to exist between Croizat's ancestral strobile, panstrobilar flower and circumnucellar flower, and Meeuse's prefloral anthocorm, brachyblastic holoanthocormoid, and anthoid. These conceptual relationships will be discussed in a future article.

Chapter 5 of the Principia Botanica concludes the series of transections of the angiosperm families and develops the systematic implications of Croizat's analysis of floral morphogeny. The first conclusion is naturally that members of the woody Ranales are no closer to the ancestors of flowering plants than are any of a whole front of families on the threshold of angiospermy. As well as many of the woody Ranales and the “gate” families mentioned above, these families include Tetracentraceae, Trochodendraceae, Casuarinaceae, Leitneriaceae, Cercidiphyllaceae, Daphniphyllaceae, Eucommiaceae, Eupteleaceae, Salicaceae, Myricaceae and Tamaricaceae, all uninvolved directly with the woody Ranales.

These ideas find support in much recent work. For example, concluding a study of the wood anatomy of Daphniphyllaceae, Carlquist states: (1982:265)

“Some botanists are surprised to learn how specialised woods of such families as Annonaceae, Hernandiaceae, Calycanthaceae, Lactoridaceae, Myristicaceae, and Piperaceae are. Hamamelidales contain two families with primitively vesselless wood (Tetracentraceae and Trochodendraceae) and other families (Cercidiphyllaceae, Eupteleacae, and Hamamelidaceae itself) with strongly primitive woods…Thus the annonalean (magnolialean, lauralean) radiation was not the only radiation of primitive angiosperms, and our view of early angiosperms could well afford a more inclusive look.”

Leroy (1980) concludes that dicliny in Cercidiphyllum is primitive. Dilcher (1979:324) draws the following conclusions

“from the present fossil record of early angiosperm reproduction:

page 36

(1) Magnoliales should no longer be thought to epitomise the primeval angiosperm flower. The Ranalian Complex may represent one of the early lines of angiosperm evolution, but not singularly the most primitive. (2) The so-called “reduced” flowers of such orders as the Trochodendrales, Cercidiphyllales, Eupteleales, Hamamelidales and Piperales may be considered initially simple rather than reduced from a monoclinous (bisexual) ancestor. Independent lineages of some anemophilous flowers developed early and perhaps separately from entomophilous flowers from a common diclinous (unisexual) ancestral stock.

(3) Modern flowers are the product of a web of evolution…with repeating themes. The homologies of flower parts of various major taxa are complex, difficult and not presently well-understood”.

Another aspect of the origins of angiospermy is discussed by the eminent Russian palaeobotanist Krassilov (1973). He states:

“Another escape from Darwin's “abominable mystery’ (the origin of angiosperms,) is the theory of the upland origin of angiosperms. Their ancestors are then claimed to have been upland plants as well and not preserved in the fossil record…This is the best way to make the problem unsolvable ab initio.”

Krassilov thus supports Croizat's severe criticism of the notion of an upland origin of angiospermy and states further that

“If the origin of angiosperms is still a mystery, the explanation should be found not in the gaps of the geological record but in the deficiencies of our evolutionary concepts.”

The idea that new fossil discoveries are essential for, and lead automatically to the resolution of evolutionary problems was criticised by Croizat and has recently come under attack both with respect to the origin of structures (e.g. Grierson and Bonamo 1979) and to the phylogenetic relationship of taxa (Patterson 1981).

Formerly seldom discussed, the idea of a polyphyletic origin of angiospermy has, since Croizat's work, been considered more and more seriously by many authors. Krassilov (1973, 1977) concludes that “angiospermous characters arose in several lineages of gymnosperms”, and in the latter paper cites Khokhrjakov (1975) to the effect that there is convincing evidence that monocotyledons were derived from gymnosperms. This interpretation of the palaeobotanic evidence is in accord with Philipson (1974:104) who adopts Croizat's concept of the “capacity (of angiosperms) to advance on a broad front”. Meyer (1970) discusses evidence from another source and proposes the “development of angiosperms over a wide area from groups already morphologically distinct”. In a 1977 review Merxmüller stated that “nobody seems to defend…a strictly monophyletic theory of angiosperm origin”. Anatomical studies of floral development by Maze, Bohm and Beil (1972) indicate that “flowers in general are polyphyletic”. Thus Johnson and Briggs' comment that “the monophyletic Magnolian theory is on shakier ground than many have assumed” (1975:90) seems more and more justified.

Analysis of Phyllotaxis and Symmetry

The word symmetry is often used in modern times to mean radial and even, as Weyl (1952) has noted, bilateral symmetry. The science of crystallography places no such restrictions on the idea of symmetry, and if development of general concepts of biological symmetry is desired we must likewise avoid any tendency to think that radial/bilateral symmetry is the only important mode. The study of biological symmetry is obviously of page 37 great importance, but apart from work on phyllotaxis (leaf arrangement) it has been generally overlooked. For this reason biology has nothing to compare with the sophisticated work attained in the description and interpretation of crystal structure (Phillips 1971). This has had a crippling effect on the development of biology and on the flow of ideas between the sister sciences of biology and geology, in much the same way as has the lack of efficient concepts in biogeography.

As one example of a general, fundamental problem of biological symmetry we can examine the question “why do components of organic structure so often possess symmetry based on five”? It is often observed that among flowers symmetry of five is most frequent. It is less well known that sometime around 1510-1516 A.D. Leonardo da Vinci determined that in many plants the sixth leaf stands above the first (Richter 1939), this being perhaps the first reference to what later became well known as 2/5 phyllotaxy (the system consists of repetitions of five leaves in two turns of the axis). This is the most common of all patterns of leaf arrangement. In the animal kingdom symmetry based on five is manifest rather less obviously, but even so recurs with such frequency as to constitute a phonomenon of general interest. That radiolarians furnish many forms with pentagonal symmetry will some as no surprise to those familiar with these beautiful animals. Examples include the Pentasphaeridae, the Pentinastrum group of genera in the Euchitoniidae, and Circorrhegma (Circoporidae) (Campbell 1954). The foraminiferan Pentellina pseudosaxorum exhibits a pattern of growth identical to phyllotaxy in mode 2/5, sectors of growth being separated by a difference of five members (Van Iterson 1907; Croizat 1964a:440). The Priapulida, a group of burrowing marine worms, possesses dental armature arranged on its proboscis in pentagonal whorls (Nichols 1967). Echinoderms, of course, show a striking adherence to pentamerism. Since the time of the earliest known amphibians 360 million years ago, five has been the dominant number of digits in tetrapods, reductions from five (e.g. in horses and birds) have been frequent, but increases occurring hardly at all, and then constituting abnormality.

The “biological rule of five” is discussed only seldom (Nichols, 1967, has discussed it with reference to animals), but Croizat, in Chapters 7 and 8 of the Principia Botanica has subjected it to a thorough analysis.

D'Arcy Thompson (1917) in a very influential work, unfortunately failed to realise that the key difference between the main modes of phyllotaxis is simply one of superposition. For instance in his fig. 327 “leaf” 1 is clearly superposed by “leaves” 14, 9, 6, 4 in Fig. 327a, b, c, d respectively, leading to modes 5/13, 3/8, 2/5 and 1/3. In describing the obviously distinct appearances of the phyllotactic modes Thompson failed to mention this, noting instead that “the mathematical side of this very curious phenomenon I have not attempted to investigate”. This simple oversight led to over half a century of deep confusion, with many authors bent on analysing the mathematics of this evidently very complex subject!

As with the general problem of biological symmetry, the concept of superposition must be primarily biological rather than geometrical. Biological superposition cannot require, as does geometrical superposition, the presence of an exact perpendicular. The analysis of biological symmetry which utilises essentially geometrical premises is similar in many respects to the analysis of plant morphology which begins with the given categories leaf, stem and root or biogeography which begins with casual migrations.

page 38

Utilising the concept of superposition, the symbolism “2/5 phyllotaxy” refers to an important biological reality. Phyllotactic modes of 1/2 (two leaves per turn) and 1/3 (three leaves per turn) represent the morphological consequences of a meristem producing the minimum number of primordia (two and three) in the lowest modes of symmetry—biradial and triradial. The minimum number of systems which can interact is, of course, two. Beginning with the two systems which themselves represent minimal symmetry, Croizat's analysis shows that an immediate result of interaction between 1/2 and 1/3 symmetry is, in fact, 2/5 symmetry, or symmetry in fives. As with 2/5, the other common modes of phyllotaxis, 3/8, 5/13 etc., also produce between two and three leaves per turn—i.e. 1/2 and 1/3 mark the basic symmetries.

Croizat concludes that whenever a system of many primordia evolves by reduction of parts (e.g. by fusion) or by increase in number of parts, the system will tend to the minimal symmetries of 1/2 and 1/3, and their first “sum” 2/5. This tendency is responsible for the establishment, at the level of coelocanthid fishes, of the morphogenetic premise which led for example, to the five fingers of Homo and for the reduction of the ancestral strobile around five sectors of growth which led to modern pentamerous flowers.

The genetic spiral, described by the developmental sequence of leaf primordia, is often assumed to have special significance for phyllotaxis. But by a constant, gradual displacement of primordia, phyllotaxis may pass easily from, say, mode 1/2 to mode 2/5. This is possible simply because both modes have two spirals of growing points (cyclosectors), just as modes 1/3 and 3/8 both have three. Thus the genetic spiral, while being descriptively useful, is interpretatively useless, since all phyllotaxies are composed of more than one cyclosector.

The evolutionary history of echinoderms would furnish excellent material for a study of the inter-relationships of minimal symmetries. Biradial, pentaradial, and possibly archaic triradial forms exist. Current debate is concerned with whether or not this trimerous stage occurred in echinoderm evolution (Philip 1979, Stephenson 1979), but unfortunately none of the students involved have related the problem to general concepts of biological symmetry.

Modern studies of symmetry and ontogeny have rejected the traditional reliance on adaptationist “explanations” of aspects of organic form. For instance Goodwin (1982a, b) interprets developing organisms as:

“entities with an extensive range of morphological potential, describable in terms of probabilistic fields which collapse…into specific morphologies”. (1982b:52).

These concepts parallel those of Croizat on morphogeny vs. morphology very closely indeed, which is interesting as they have emerged from what are usually regarded as distinct areas of study—ontogeny and comparative morphology. For Goodwin the probabilistic field properties are a function of “general organisational principles”. (1982b:53). As a consequence of this fundamental change of emphasis Goodwin has subjected the neo-Darwinist approach to a critical and severe analysis, and concluded:

“Once it is recognised that there are principles of organisation and laws of form in biology, these time-independent properties of the living realm become once again central to the subject…the realisation that genes do not generate biological form leads to a rather different view of the evolutionary process in terms of the potential forms of the organisms and their appearance of the earth.” (1982a: 111-112).

page 39

Thus Goodwin, like Croizat, would place the emphasis on Darwin's “laws of growth”, in contrast with the neo-Darwinist tradition of virtually ignoring them.

Working with rather different subject matter from Goodwin, the Russian palaeobotanist Meyen (1973, 1978) has argued at length for a greater emphasis on the study of “general structural principles” (again corresponding to Darwin's “laws of growth”) independent of phylogenetic considerations.

Baas (1982), discussing the evolution of wood anatomy, criticises “rigid adaptationist interpretations”, advocating the important role of “functionless trends imposed by correlative restraints…”

Interpretation of Leaf and Stipule

The question “What is a leaf?” is certainly one of the great problems of botany. As mentioned above, the current state of despair is well exemplified by the comment:

“Although no satisfactory definition of a leaf is thus possible I shall assume that we all know what we shall be taking about.” (Gregory 1956).

It is clear that the definitional approach, using morphological, especially anatomical, criteria has failed to supply us with efficient concepts involving the true nature of the leaf. If we agree that the assumption “we all know what we shall be talking about” is hopeful but unfounded, we must also agree that it is necessary to do something.

It is Croizat's contention that descriptive anatomical criteria cannot supply an answer to the question “what is a leaf?”, and are, in fact, an inappropriate basis for evolutionary studies. In support is the comment by Schmid (1972:442) that it is Croizat who provides:

“the most pertinent and clear statement I have encountered regarding the problem of the validity of vascular conservatism”.

Croizat asserts that the most useful analysis of “leaf” would be in terms not of unanalysed morphological homologies, but rather of morphogenetic processes. With respect to Croizat's attitude towards interpretation of morphogeny vs. technical description of morphology, it is interesting to note Einstein's view that:

“It is really strange that human beings are normally deaf to the strongest arguments while they are always inclined to overestimate measuring accuracies.” (from a letter, quoted in Feyerabend 1978:58).

Croizat's analysis of leaf form and morphogeny (Princ. caps. 9, 10) includes consideration of many structures which were traditionally regarded (if at all) as unusual, insignificant, and, ultimately, accidental. Arber (1934:312) has summed up this unproductive attitude well:

“Another dictum of formal morphology is that the power of producing lateral shoots is confined to axes. When a leaf does, in fact, bear a shoot-bud, this shoot is described as ‘adventitious’, which means, literally, “accidental’. This is a typical example of the tyranny exercised by words over thought; just because they have themselves labelled these buds “accidental’, botanists feel justified in dismissing them as of no morphological significance.”

Just as “inexplicable”, thus “accidental”, patterns of dispersal (such as trans-tropical-Pacific) are analysed in his Panbiogeography, Croizat does not fail to consider the nature of “inexplicable” and “accidental” morphological facts in the Principia Botanica. The morphology of plants with structure problematical in the light of traditional interpretations has often led students, for example Jong and Burtt (1975) working on Gesneriaceae, to reject the “traditional morphological categories” as page 40 adequate guides for efficient analysis. In Croizat's work these “difficult” plants are not ignored, on the contrary, as with incongruent area/taxa cladograms in biogeography, they are shown to be in themselves useful guidelines in analysis.

The problem of the modern leaf was eliminated in the past, when “leaf” was proposed to be itself an irreducible category, an element, an essentially simple structure. The sociological and philosophical reasons for this proposition are complex—suffice it to say that Croizat begins with no such assumption. Indeed, by means of a thorough analysis of leaf form, Croizat demonstrates that, in fact, the modern angiosperm leaf is an essentially compound body (c.f. his analysis of Wallacean areas of endemism as compound entities). We are all aware of the complexity of leaf organogenesis, attained by a variety of meristems (Jeune 1981) and Croizat's conclusion may initially seem relatively innocuous. Nevertheless, it has important consequences.

Croizat approaches the morphogeny of the modern foliage leaf, stipules, cataphylls, buttress, buds and certain types of thorn (Cactaceae, Euphorbia, Fouquieria) as a general problem. As with the leaf/shoot question, rather than simply providing homologies between unanalysed organs it would seem more productive to inquire about limits between the organs, thus identifying processes of divergence. Croizat's analysis concludes in fundamental agreement with Tyler (1897) that the “lower foliar organs” (stipule, cataphyll etc.) are neither reduced leaves nor additions subsequent to the development of the modern leaf, but are the “primitive foliar organs”. The modern leaf represents a development of, and upon, this primitive leaf. The lateral portions of the primitive leaf, when separated, form the stipules, petiole wings etc. of the modern leaf. The sheathing petiole is a product of the development of the lateral and central basal parts of the primitive leaf—it is essentially distinct from the true petiole of the modern leaf. The leaf buttress represents the immediate continuation of the primitive leaf into the cortex.

These ideas are supported by studies of leaf organogenesis. For example, Cross (1938) has shown that in Viburnum cataphylls are not leaf homologues—cataphyll and leaf ontogeny diverge dramatically at the 80μ stage and leaf growth progresses by means of a distinctive ventral meristem which cataphylls lack. Bruck and Kaplan (1980) have shown that scale-leaves are not homologous with foliage leaves in Muehlenbeckia. In Morus, Cross (1937) found that the stipule development is unlike that of the foliage leaves, but like that of the cataphylls. Again the difference is fundamental, the foliage leaves being the result of the activity of a ventral meristem totally absent in the cataphylls and stipules. Thus the primitive scaliform leaf, represented by stipules and cataphylls, and the modern foliage leaf represent essentially distinct variants, at a very low level, of a single foliar type—the scale of Croizat. In the case of the foliage leaf, this scale or undeveloped phyllome has inherited a specialised “dab of meristem” which, once activated, leads to leaf organogenesis.

Macdonald (1981) stated recently that “The phylogeny of the stipule remains unresolved”, indicating also that little, if any, conceptual progress has been made since Sinnott and Bailey (1914) made similar comments sixty seven years earlier. Macdonald also noted that:

“to conclude that the stipules of Comptonia (Myricaceae) are lobes or outgrowths of the leaf base, while true in an empirical sense, contributes little to our understanding of their phylogeny”.

page 41

However, recent work by Jeune (1981) concludes that, in the light of current knowledge of leaf ontogeny, Croizat's analysis is of special significance in adequately resolving the morphogenetic and phylogenetic nature of the stipule.

Macdonald concluded his perceptive paper by suggesting homology between stipule and prophyll (he unfortunately overlooked Croizat's analysis of “prophyll”, and his identification of it with “stipular sector” in Princ. 718).

Croizat's synthesis of the morphogeny of the modern foliage begins with verticils or whorls of primitive “leaves” (to be regarded more as phyllomes or primordia). As Howard (1974:160) has noted:

“The evolutionary progress from alternate and spirally arranged leaves to opposite or whorled leaves has become established as a dictum in most botanical publications…without any real evidence.”

Howard (1974:163) concludes that “the primitive leaf was probably borne in whorls or even vertically grouped clusters…”

These verticils, each of n “leaves” became suppressed, recombined and dirempted, or pulled apart, along the shoot into local sectors of (1 foliage leaf + 2 stipules (undeveloped leaves)). This process of reduction naturally takes place within the limits imposed by laws of symmetry, for example those outlined above.

What is the explanation for the development of the single leaf as the keystone of reduction and diremption of the ancestral verticil? The stipule/leaf distinction is not adequately defined by the presence of an axillary bud (e.g. Chaenomeles japonica stipules may have axillary buds), but histologically the distinction can be made on the basis of the ventral foliar meristem. The crucial question thus becomes “What is the phylogenetic and morphogenetic nature of this important meristem? What is its origin?” The only real answer given to this question is that of Croizat. The meristem is the product of ancient fusion of axes and primitive “leaves” leading to the modern leaf. The meristem itself represents the incorporation into the primitive “leaf” of the primordium of the ancient axis.

Fundamental to Croizat's analysis is the nature of the axillary buds observed in modern plants. Although some plants may have only one bud per axil, in principle there may be a series of axillary primordia lined up between the petiole and the axis. Today these are responsible for shoot-making of various kinds (floral and vegetative), including the fusion of epiphyllous inflorescences and leaves. In the history of the leaf, one element of these primordia has been competent in fusing an axis with the ancient “leaf” leading to the meristem which characterises the “modern” leaf. The relationship between the primitive scaliform “leaf” and the axis which it subtends and later fuses with is essentially hypocladial, in the sense of Kursner (1954). Hypocladial relationships between rameal and foliar organs are well known in many extant plants, for example, see Fig. 2. The axis which has become reduced and incorporated into the primitive scaliform leaf, leading to the development of the modern, hypocladial leaf could naturally be expected to have left some indications as to its nature— in other words the hypocladial transfusion may well not have been completed in every instance.

Dickinson (1978) in a timely and thorough review of the phenomenon of epiphylly has urged the study of the laws of growth which have determined the development of epiphyllous inflorescences. He has also noted the possibility of ephiphylly having been of “fundamental importance page 42 …to the origins of the angiosperm leaf itself” and suggests that this concept provides significant avenues for future research. The idea of the inherent sexuality of angiosperm leaves can be traced back at least to C. de Candolle (1890), but its phylogenetic and morphogenetic meaning is first explained by Croizat, in terms of the hypocladial nature and history of the modern angiosperm leaf. Epiphylly is, of course, a worry for traditional morphology which regards the leaf as an irreducible category—the consequent problem lies in explaining how and why an inflorescence managed to “get up” onto a leaf. However, with Croizat's analysis of the leaf as essentially complex, the “problem” of epiphylly vanishes—it simply means that the axis involved with the primitive scaliform foliage was, in fact, or potentially, floriferous. Thus the hypocladial relationship between (potentially) floriferous axes and the primitive scaliform foliage is an important law of growth for the modern angiosperm leaf. In a study of epiphylly in Helwingia japonica Dickinson and Sattler (1975) state:

“The main conclusion is that such conditions (epiphylly) suggest that “laws of growth”…probably are as important, or more so, than laws of natural selection in determining plant form.”

The same authors (1974:8) in a study of the epiphyllous inflorescence of the saxifragaceous Phyllonoma integerrima state that

“our observations cannot be incorporated into the rigidly formulated classical theory of the shoot without distortion of the observations themselves, or of our understanding of them.” They also note that (1974:9) “The value of atypical situations like epiphylly is that they point out aspects of the real morphogenetic potential of plants in nature, corresponding to these “Laws of Growth”, that we often overlook.”

This potential may lie at no great depth, for example Stebbins (1965) describes the single “gene” controlling the production of epiphyllous inflorescences in plants of Hordeum trifurcatum (see Princ. 1533).

A question of crucial significance for an understanding of the leaf is posed by the nature of glands. Most recent work concentrates on their present day ecological significance and tends to ignore their morphogeny. But as Schnell (1970:433) has said, regarding extra-floral glands:

“Leur signification…parait a rechercher dans la morphologie et dans la phylogenie de la feuille plutôt que dans une utilité pour la plante…”1

Since Schwendt's important 1907 paper, botanists have interpreted the morphology of plant glands as a result of the reduction of the hemming-in of pre-existent structures (e.g. see Schnell 1969:153-154 for discussion).

Glands, often secretory, are well known from the vegetative parts of many plants, as well as from the flowers. Particularly common are glands found at, or near, the petiole-lamina junction. Persistent meristems of various kinds have also been reported from this locality (e.g. Jong 1973 on Streptocarpus). Glandular teeth are well known on stipules (e.g. the rubiaceous Coprosma) and leaves (e.g. Chloranthaceae). The interpretation of floral and extra-floral glands as loci of hemmed-in tissue has been supported by Schnell (1969) with respect to domatia found in the axils of foliar venation (again Coprosma provides striking examples). He suggests that they indicate “une croissance avortée”. With respect to early theories on the origin of ant domatia by means of natural selection, Philipson (1964) cities Bailey (e.g. 1922, 1923) favourably, to the effect that:

“insects are not concerned with the origin of development of the structures”.

page 43

Thus there are many different structures present in the morphology of modern foliage which represent remnants of hemmed-in growth, and which must all be accounted for in any synthesis of leaf morphogeny. Howard (1974), in an important contribution, discusses aspects of nodal and petiolar form and concludes that these provide further support for Croizat's concept of the leaf and its essentially compound nature.

Interpretation of the Root

The question of the seed-plants' root has been answered even less satisfactorily by descriptive morphology than has been the question of the leaf. Descriptive criteria fail dramatically when faced, for example, by Lentibulariaceae, and, in fact, there are virtually no concepts efficient in producing even a minimum of understanding. In Chapter 11 of the Principia Botanica Croizat begins his discussion of the root with an analysis of the structures found at the junction of hypocotyl and root in the seedlings of various taxa. These structures (known in the literature as “wurzelhals”, “foot”, “peg”, “collet”, etc.) are much more widely distributed throughout angiosperms than was formerly realised, occurring in such taxa as Hippuris, Cucurbitaceae, Triglochin, etc. In Eucalyptus erythrocoris the normally horizontal collet forms a sheathing structure morphogenetically identical to the graminaceous coleorhiza. In the embryo of the seed plants, root and plumule, representing contrastingly polarised centres of development are joined by a transitional zone, the hypocotyl. Beyond the root initials lies the meristem (e.g. the rib meristem of Pseudotsuga—Allen 1946) from which arises the root cap. These three embryonal meristems are all bound within a jacket of ground tissue, Croizat's synthetic concept which comprises a sheath of considerable morphogenetic powers. The jacket is responsible for the development of various structures, for example tubers, napiform taproots, the collet, the haustorium of parasitic plants, the holdfasts and coralloid roots of the Podostemonaceae, the grass coleorhiza, and of course the rather inconspicuous cortical layer surrounding the hypocotyl in other “higher” plants.

The concept of “jacket” is an efficient tool for rationalising many aspects of root morphology from Pseudotsuga, to Triticum, to Utricularia. But a fundamental problem is offered by such plants as Podostemon ceratophyllum which have no distinct root but do possess a root cap. In fact Croizat recognises the problem of root cap as of fundamental significance. He suggests that it is advisable to maintain a general concept of rhizophore for what is currently termed “root”, “rhizome”, “runner”, “hypocotyl”, “pneumatophore” etc. in opposition to one of root cap. Thus the rhizophore is seen essentially as the bearer of the root cap, and structures such as the underground axis/rhizophore plus roots of the form genus Stigmaria, and the corm (rhizophore) plus roots of Isoetes, long considered problematical, are simply interpreted as relatively undeveloped morphologies in the morphogeny leading to angosperm roots. (c.f. Stewart 1947; Sporne 1975:69-70).

Croizat's analysis has not been developed by modern workers on the root, but it is clear that this is the least understood plant organ of all, and basic morphogenetic concepts are urgently needed. Too many recent studies of root form still provide “explanations” of that form in terms only of function and adaptation! Invariably aspects of root morphology apparently restricted to a small number of taxa are regarded as “advanced”, page 44 “derived” etc., homologised with something or other, and forgotten about. A good example is provided by the “macropodous” condition in certain monocotyledonous embryos (see Dahlgren and Clifford 1982: Fig. 88).

Other Botanical Contributions

This review has so far mentioned some of Croizat's work on the fundamentals of botanical philosophy, but it would be unwise to ignore altogether his contributions to other aspects of botany.

Croizat felt it was crucial to be able to place work and ideas in their historical context. In the Principia Botanica he seldom considers this as an end in itself, but the importance he placed on it and the interest it held for him can be seen in various other publications dealing largely with historical questions (1945) and biography (1949a, b).

Most of Croizat's early work deals largely with the taxonomy and nomenclature of Cactaceae and Euphorbiaceae, and is of only minor interest to the botanist who does not have a special interest in and knowledge of these families. In much of this work it is interesting to observe the enthusiasm with which Croizat dealt with the horticultural aspects of his favoured plants (e.g. 1941a). Practical rather than theoretical questions provided the impetus for all of Croizat's work which was not done in a vacuum but is of interest to all laymen and scholars who ask the question “Why is this plant the way it is, where it is?” It goes almost without saying that Croizat was a formidable enemy of ivory-tower scholasticism.

Because of the practical necessity of supplying a name for a plant, Croizat was naturally concerned with the correct application of the rules of botanical nomenclature, and also took an active interest in the development and improvement of the rules themselves (e.g. 1941b, 1953).

Croizat has published several large and important botanical works since the Principia. In Croizat (1964a, b) his ideas on angiosperm evolution in space, time and form are summarised and, to an extent, developed and refined. Croizat (1965, 1967, 1972a, 1973a) provide both an overview and a detailed analysis of Euphorbiaceae, Euphorbieae and Euphorbia and represent conclusions reached after several decades of cultivation and study of these plants. Croizat's work on these taxa is undoubtedly among the most important yet produced.

Croizat (1970) is a critical interpretation of Corner's fascinating but almost totally neglected Durian Theory of the origin of the modern tree. Croizat (1971, 1972b, 1973b) are very important contributions to the study of the leaf and of phyllotaxy, in which a considerable amount of new material is introduced.

Although the most striking aspect of the Principia Botanica is undoubtedly the fundamental nature and originality of the principles developed, the reviewer would be failing in his task if he did not attempt to place Croizat's botany in some sort of historical perspective. As with his biogeography, many of his ideas were hinted at, though scarcely developed by earlier students. Cusset (1982), in what must be regarded as a landmark of botanical historiography, has produced a review of the conceptual bases of plant morphology. In it the affinity of Croizat's concepts, particularly concerning foliation, with those of botanists such as Warming and Trécul is underscored. In the fields of floral morphogeny and high systematics page 45 Croizat's conclusion often reflect more his sympathy with those of Baillon, a botanist neglected by modern students.

Many of the principles established in Croizat's work are becoming accepted, or at least discussed, by the botanical community, but too often the work is cited only by leading researchers examining fundamentals of botanical knowledge. Obviously even the reading of a large work such as the Principia is a major undertaking—much time is required to digest the arguments and check up on examples in field, herbarium and literature. But probably the main reason for the surprisingly small number of papers discussing Croizat's ideas is the fundamentally heterodox nature and the complexity of the analyses. As his ideas become more acceptable to the botanical community more credit will undoubtedly be given Croizat for the great significance of his contribution to the “Beginnings of Botany”.


Adams, R. M. II and G. W. Smith, 1977: An S.E.M. survey of the five pitcher plant genera. Amer. J. Bot. 64:265-272.

Airy Shaw, H. K., 1951: On the Dioncophyllaceae, a remarkable new family of flowering plants. Kew Bull. 3:327-347.

Allen, G. S., 1946: Embryogeny and development of the apical meristems of Pseudotsuga I. Fertilisation and early embryogeny. Amer. J. Bot. 33:666-677.

Arber, A., 1934: The Gramineae. Cambridge University Press.

——, 1946: Goethe's botany. Chron. Bot. 10:67-124.

Bailey, I. W., 1922: Notes on Neotropical ant-plants. I. Cecropia angulata sp. nov. Bot. Gaz. 74:369-391.

——, 1923: Notes on Neotropical ant-plants. II. Tachigalia paniculata Aubl. Bot. Gaz. 75:27-41.

Baas, P., 1982: Systematic, phylogenetic, and ecological wood anatomy: history and perspectives. In New Perspectives in Wood Anatomy (P. Baas, ed.):23-59. Martinus Nijhoff/Dr. Junk, The Hague.

Baillon, H. E., 1871-1888: The Natural History of Plants. 8 vols. Transl. M. Hartog. L. Reeve, London.

Bruck, D. K. and D. R. Kaplan, 1980: Heterophyllic development in Muehlenbeckia (Polygonaceae). Amer. J. Bot. 67:337-346.

Campbell, A. S., 1954: Radiolaria. In Treatise on Invertebrate Paleontology (R. C. Moore, ed.) Part D Protista 3:D11-D163.

Carlquist, S., 1982: Wood anatomy of Daphniphyllaceae: ecological and phylogenetic considerations, review of Pittosporalean families. Brittonia 34:252-266.

Charlton, W. A., 1972: Features of the inflorescence of Triglochin maritima. Can. J. Bot. 59:2108-2115.

Croizat, L., 1941a: On methods and experiments. Cactus and Succulent J. 13:184-185.

——, 1941b: A further comment on stability in nomenclature. Science 93:109-110.

——, 1945: History and nomenclature of the higher units of classification. Bull. Torrey Bot. Club 72:52-75.

——, 1949a: Une biographie peu connue de Hipolito Ruiz. Lilloa 18:295-329.

——, 1949b: Rafinesque: a concrete case. Archivio Bot. Biogeogr. Ital. 24(3rd ser. vol. 8): 169-184. (As “Henricus Quatre”).

——, 1953: On nomenclature: the “type-method”. Taxon 2:105-107, 124-130.

——, 1961: Principia Botanica. 1 vol. bound as 2. Published by the Author, Caracas. (“1960” on title page).

——, 1964a: Space, Time, Form: The Biological Synthesis. Published by the Author, Caracas. (“1962” on title page).

——, 1964b: Thoughts on high systematics, phylogeny and floral morphogeny, with a note on the origin of the Angiospermae. Candollea 19:17-96.

——, 1965: An introduction to the subgeneric classification of Euphorbia L., with stress on the South African and Malagasy species. I. Webbia 20:573-706.

——, 1967: An introduction to the subgeneric classification of Euphorbia L., with stress on the South African and Malagasy species. II. Webbia 22:83-202.

——, 1970: Gigantomachie botanique: la “Théorie du Durian” contre la “Théorie de la Lentille d'Eau”. Adansonia II, 11:47-76.

——, 1971: La filotaxia: sus principios, métodos de investigación y de docencia. Simbolismo. Bol. Acad. Cienc. Fis., Mat. Nat., Caracas 31:53-119.

——, 1972a: An introduction to the subgeneric classification of Euphorbia L., with stress on the South African and Malagasy species. III. Webbia 27:1-221.

Croizat, L., 1972b: En torno al concepto de hoja. Ensayo de botanica analítica y sintética. Bol. Acad. Cienc. Fís., Mat. Nat., Caracas 32:29-207.

——, 1973a: Les Euphorbiacées vues en elles-memes, et dans leurs rapports envers l'angiospermie en général. Mem. Soc. Broteriana 23:5-207.

——, 1973b: Algo mas en torno al concepto de “Hoja” y de “Nudo”: notas morfogeneticas y pegagogicas. Bol. Acad. Cienc. Fis., Mat. Nat. Caracas 33:35-93.

Cross, G. L., 1937: The origin and development of the foliage leaves and stipules of Morusalba, Bull. Torrey Bot. Club 64:145-163.

——, 1938: A comparative histogenetic study of the bud scales and foliage leaves of Viburnum opulus. Amer. J. Bot. 25:246-258.

Cusset G. 1982: The conceptual bases of plant morphology. In Axioms and Principles of Plant Construction (R. Sattler, Ed.):8-86. Junk Nijhoff, The Hague.

Cutter, E. G., 1966: Review of Fundamentals of Phytomorphology by A. D. J. Meeuse, 1966. Quart. Rev. Biol 41:320.

Dahlgren, R. M. T. and H. T. Clifford, 1982: The Monocotyledons: A Comparative Study. (Botanical Systematics v. 2). Academic Press, London.

De Candolle, A. P. 1827: Organographie Végétale, ou description Raisonnée des Organes des Plantes. 2 vols. Déterville, Paris.

De Candolle, C., 1890: Recherches sur les inflorescences épiphylles. Mém. Soc. Phys. Nat. Hist. Geneve. (Vol. Suppl. Centenaire) 6:1-37.

Dickinson, T. A., 1978: Epiphylly in angiosperms. Bot. Rev. 44:181-232.

—— and R. Sattler, 1974: Development of the epiphyllous inflorescence of Phyllonoma integerrima (Turcz.) Loes.: implications for comparative morphology. Bot. J. Linn. Soc. 69:1-13.

—— and ——, 1975: Development of the epiphyllous inflorescence of Helwingia japonica (Helwingiaceae). Amer. J. Bot. 62:962-973.

Dilcher, D. L., 1979: Early angiosperm reproduction: an introductory report. Rev. Palaeobot. Palynol. 27:291-328.

Dormer, K. J., 1946: Anatomy of the primary vascular system in dicotyledonous plants. Nature 158:737-739.

Engler, H. G. A. and K. A. E. Prantl, 1924—: Die Natürlichen Pflanzenfamilien. 2nd. Ed. Verlag von Wilhelm Engelmann, Leipzig, and Duncker and Humblot, Berlin.

Feyerabend, P. K., 1978: Against Method. Verso. London.

Goodwin, B. C., 1982a: Biology without Darwinian spectacles. Biologist 29: 108-112.

——, 1982b: Development and evolution. J. theor. Biol. 97:43-55.

Gregory, F. G., 1956: General aspects of leaf growth. In The Growth of Leaves (F. L. Milthorpe, Ed.):3-17. Butterwortns, London.

Grierson, J. D. and P. M. Bonamo, 1979: Leclerqia complexa: earliest ligulate lycopod (Middle Devonian). Amer. J. Bot. 66:474-576.

Guédès, M., 1979: Magnolioid island plants and angiosperm evolution. In Plants and Islands (D. Bramwell, Ed.): 307-328. Academic Press, New York.

Howard, R. A., 1974: The stem-node-leaf continuum of the Dicotyledoneae. J. Arnold Arboretum 55:125-181.

Huxley, T. H., 1888: The gentian: notes and queries. J. Linn. Soc. Bot. 24:101-124.

Jeune, B., 1981: Modèle empirique du dévelopment des feuilles de Dicotylédones. Bull. Mus. Natn. Hist. Nat., Paris, 4e ser., 3, section B, Adansonia, 4:433-459.

Johnson, L. A. S. and B. C. Briggs, 1975: On the Proteaceae—the evolution and classification of a southern family. Bot. J. Linn. Soc. 70:83-182.

Jong, K., 1973: Streptocarpus (Gesneriaceae) and the phyllomorph concept. Acta Botanica Neerlandica 22:244-245.

—— and B. L. Burtt, 1975: The evolution of morphological novelty exemplified in the growth patterns of some Gesneriaceae. New Phytol. 75:297-311.

Khokhrjakov, A. P., 1975: Somatic evolution of the Monocotyledons. Nauka, Moscow. (in Russian).

Krassilov, V., 1973: Mesozoic plants and the problem of angiosperm ancestry. Lethaia 6:163-178.

——, 1977: The origin of angiosperms. Bot. Rev. 43:143-176.

Kursner, E., 1954: Recherches sur la mode de ramification dans l'inflorescence de la Bourse-à-Pasteur Capsella bursa-pastoris (L.) Medikus. Bull. Scientifique Bourgogne 15:145-177.

Leroy, J. F., 1980: Développment et organogenèse chez le Cercidiphyllum japonicum: un cas semblant unique chez les angiospermes. C. R. Acad. Sc. Paris D 290:679-682.

Lorch, J., 1963: The carpel—a case-history of an idea and a term. Centaurus 8:269-291.

Macdonald, A. D., 1979: Development of the female flower and gynecandrous partial inflorescence of Myrica californica. Canad. J. Bot. 57:141-151.

——, 1981: The stipules of Comptonia peregrina. Canad. J. Bot. 59:694-700

—— and R. Sattler, 1973: Floral development of Myrica gale and the controversy over floral concepts. Canad. J. Bot. 51:1965-1976.

McLean, R. S. and W. R. Ivimey-Cook, 1956: Textbook of Theoretical Botany. Vol. 2. Longmans Green & Co., London, New York, Toronto.

Maguire, B. and P. Ashton, 1977: Pakaraimoideae, Dipterocarpaceae of the Western hemisphere. Part II. Systematic, geographic and phyletic considerations. Taxon 26:343-368.

Marburger, J. E., 1979: Glandular leaf structure of Triphyophyllum peltatum (Dioncophyllaceae) a “fly-paper” insect trapper. Amer. J. Bot. 66:404-411.

Maze, J., N. G. Dengler and L. R. Bohm, 1971: Comparative floret development in Stipa tortilis and Oryzopsis miliacea (Gramineae). Bot. Gazette 132:273-298.

——, L. R. Bohm and C. E. Beil, 1972: Studies on the relationship and evolution of supra-specific taxa utilising developmental data. I. Stipa lemmonii (Gramineae). Canad. J. Bot. 50:2327-2352.

Meeuse, A. D. J., 1966: Fundamentals of Phytomorphology. Ronald, New York.

——, 1975: Angiosperm evolution and emended Anthocorm Theory. Intern. Bio-Sci. Monogr. 1:1-188.

——, 1982: Evolution of the Magnoliophyta: Current and dissident viewpoints. Ann. Rev. Plant Sciences 2:393-442.

Melville, R., 1962: A new theory of the angiosperm flower. I. The gynoecium. Key Bull. 16:1-50.

Merxmüller, H., 1977: Summary lecture. Plant Syst. Evol. Suppl. 1:397-405.

Meyen, S. V., 1973: Plant morphology in its nomothetical aspects. Bot. Rev. 39:205-260.

——, 1978: Nomothetical plant morphology and the nomothetical theory of evolution: the need for cross-pollination. Acta Biotheor. 27:21-36.

Meyer, V. G., 1970: A facultative gymnosperm from an interspecific cotton hybrid. Science 169:886.

Moeliono, B. M., 1970: Cauline or Carpellary Placentation Among Dicotyledons. 2 vols. Van Gorcum, Assen Netherlands.

Nichols, D., 1967: The rule of five in animals. New Scientist 35:546-549.

Nur, A. and Z. Ben-Avraham, 1982: Oceanic plateaus, the fragmentation of continents, and mountain building. J. Geophys. Res. 87 B5:3644-3661.

Patterson, C., 1981: Significance of fossils in determining evolutionary relationships. Ann. Rev. Ecol. Syst. 12:195-223.

Pauze, F. and R. Sattler, 1979: La placentation axillaire chez. Ochna atropurpurea. Canad. J. Bot. 57:100-107.

Philip, G. M., 1979: Carpoids—echinoderms or chordates? Biol. Rev. 54:439-471.

Philipson, W. R., 1964: Morphogenesis affected by associated organisms. Phytomorph. 14:163-185.

——, 1974: Ovular morphology and the major classification of the dicotyledons. Bot. J. Linn. Soc. 68:89-108.

Phillips, F. C., 1971: An Introduction to Crystallography, 4th Ed. Oliver and Boyd, Edinburgh.

Poszluszny, U. and R. Sattler, 1974: Floral development of Potamogeton richardsonii. Amer. J. Bot. 61:209-216. —— and ——, 1976: Floral development of Najas flexilis. Canad. J. Bot. 54:1140-1151.

Richter, J. P., 1939: The Literary Works of Leonardo da Vinci. 2 vols. Oxford University Press, London.

Satina, S., 1959: Chimeras. In Blakeslee: The Genus Datura (A. G. Avery, S. Satina, and J. Rietsema Eds.):132-151. Ronald, New York.

Sattler, R., 1966: Towards a more adequate approach to comparative morphology. Phytomorph. 16:417-429.

——, 1973: Organogenesis of Flowers. A Photographic Text-Atlas. University of Toronto Press, Toronto.

——, 1974a: Essentialism in plant morphology Proc. XIV Int. Congr. Hist. Sci. No. 3:464-467.

——, 1974b: A new approach to gynoecial morphology. Phytomorph. 24:22-34.

—— and L. Perlin, 1982: Floral development of Bougainvillea spectabilis Willd., Boerhavia diffusa L. and Mirabilis jalapa L. (Nyctaginaceae). Bot. J. Linn. Soc. 84:161-182.

Schmid, R., 1964: Die systematische Stellung der Dioncophyllaceen. Bot. Jahrb. 83:1-56.

——, 1972: Floral bundle fusion and vascular conservatism. Taxon 21:429-446.

Schnell, R., 1969: Sur les structures développées aux ramifications des nervures. Mém. Soc. Bot. France 1969:147-156.

——, 1970: Introduction à la Phytogéographie des Pays Tropicaux. Vol. 1. Gauthiers-Villars, Paris.

Schwendt, E., 1907: Zur Kenntnis der extrafloralen Nektarien. Bot. Centralbl. 22:245-286 + 2 plates.

Sinnott, E. W. and I. W. Bailey, 1914: Investigations on the phylogeny of the angiosperms. 3. Nodal anatomy and the morphology of stipules. Amer. J. Bot. 1:441-453.

Sporne, K. R., 1975: The Morphology of Pteridophytes. 4th Ed. Hutchinson, London.

Stebbins, G. L., 1965: From gene to character in higher plants. American Scientist 53:104-126. Stephenson, D. G., 1979: The trimerous stage in echinoderm evolution: an unnecessary hypothesis. J. Paleont. 53:44-48.

Stewart, W. N., 1947: A comparative study of Stigmarian appendages and Isoetes roots. Amer. J. Bot. 334:315-324.

Thompson, D. W., 1917: On Growth and Form. Cambridge University Press.

Tyler, A. A., 1897: The nature and origin of stipules. Annals New York Acad. Sci. 10:1-49.

Van Heel, W. A., 1967: Anatomical and ontogenetic investigations on the morphology of the flowers and the fruit of Scyphostegia borneensis Stapf (Scyphostegiaceae). Blumea 15:107-125. ——, 1973: Flowers and fruits in Flacourtiaceae. I.

Scaphocalyx spathaceae Ridl. Blumea 21:259-279.

Van Iterson, G. Jr., 1907: Mathematische und Mikroscopische—Anatomische Studien über Blattstellungen. Fischer, Jena.

Van Tieghem, P., 1871: Recherches sur la symetrie de structure des plantes vasculaires. Anns. Scs. Nat. (Bot.) Paris 5 sér., 13:5-314.

Weyl, H., 1952: Symmetry. Princeton University Press.

Yampolsky, C. 1920: Sex intergradation in the flowers of Mercurialis annua. Amer. J. Bot. 7:95-101.

1 “An understanding of their significance would seem to require research into the morphology and phylogeny of the leaf, rather than into any usefulness for the plant.”

* Both botanists of the traditional school.