The Cell-Theory: A Restatement, History, and Critique: Part III. The Cell as a Morphological Unit

Journal of Cell Science, Jun 1952


A PDF file should load here. If you do not see its contents the file may be temporarily unavailable at the journal website or you do not have a PDF plug-in installed and enabled in your browser.

Alternatively, you can download the file locally and open with any standalone PDF reader:

The Cell-Theory: A Restatement, History, and Critique: Part III. The Cell as a Morphological Unit

Part III. The Cell as a Morphological Unit 0 1 0 (From the Department of Zoology and Comparative Anatomy 1 A Restatement , History, and Critique A long time elapsed after the discovery of cells before they came to be generally regarded as morphological units. As a first step it was necessary to show that the cell-walls of plants were double and that cells could therefore be separated. The earliest advances in this direction were made by Treviranus (1805) and Link (1807). The idea of a cell was very imperfect, however, so long as attention was concentrated on its wall. The first person who stated clearly that the cell-wall is not a necessary constituent was Leydig (1857). Subsequently the cell came to be regarded as a naked mass of protoplasm with a nucleus, and to this unit the name of protoplast was given. The true nature of the limiting membrane of the protoplast was discovered by Overton (1895). The plasmodesmata or connective strands that sometimes connect cells were probably first seen by Hartig, in sieve-plates (1837). They are best regarded from the point of view of their functions in particular cases. They do not provide evidence for the view that the whole of a multicellular organism is basically a protoplasmic unit. Two or more nuclei in a continuous mass of protoplasm appear to have been seen for the first time in 1802, by Bauer. That an organism may consist wholly of a syncytium was discovered in i860, in the Mycetozoa. The syncytial nature of the Siphonales was not revealed until 1879. The existence of syncytia constitutes an exception to the cell-theory. No wholly syncytial plant or animal reaches a high degree of organization. Natural polyploidy was discovered by Boveri (1887), who was also the first to produce it experimentally (1903). Although many organisms contain some polyploid constituents and others are polyploid throughout their somatic tissues, yet diploid and haploid protoplasts (haplocytes and diplocytes) are the primary components of plants and animals and are still retained as such by most organisms. The haplocyte is more evidently unitary than the diplocyte. Haplocytes and diplocytes are not composed of lesser homologous units, and with the necessary reservations required by the existence of syncytial and polyploid masses of protoplasm, they may therefore be said to be the fundamental morphological units of organisms. CONTENTS I N T R O D U C T I O N PAGE 1 3 8 . 1 5 8 . 1 6 3 1 6 8 1 7 7 1 8 3 . 1 8 6 1 8 6 . 1 8 6 1 8 7 INTRODUCTION I the form of seven propositions. The second of these was as follows: N the first of this series of papers (1948) the cell-theory was restated in Cells have certain definable characters. These characters show that cells (a) are all of essentially the same nature and (b) are units of structure. Part II of the series (1949) was devoted to the statement labelled (a) in this proposition; that is to say, to the fundamental similarity of all cells as revealed by the discovery of protoplasm and the nucleus. We are now concerned with the part of the proposition labelled (b); that is, with the idea of the cell as a morphological unit. The idea of the cell as a functional unit cannot escape mention here, but this aspect of our problem will be more fully considered under the heading of Proposition V. EARLY RESEARCHES BEARING ON THE MORPHOLOGICAL SEPARATENESS OF CELLS The cellular nature of plants is much more obvious than that of animals, and the earliest cytological observations were naturally made mostly on plants. Since it was the cell-wall and not the cell itself that called attention to the existence of cellular structure, attention was concentrated on the wall as a matter of course. The wall that separates two cells appears single on superficial examination, and the early observers were not inclined to regard cellwalls as separate boxes enclosing material within. The idea of a cell as a morphological unit only originated when it was found that the wall separating two cells was in some cases demonstrably double. The history of this advance will now be related. It has been mentioned in Part I that Grew described the parenchyma of plants as 'nothing else but a Mass of Bubbles' (1672, p. 79). In his later work he formed a wrong impression and made his well-known comparison with lace. This false simile had a profound influence that remains with us today in such erroneous names as tissue (and its counterparts in other languages) and histology. We are all accustomed today to speaking of the 'tissues' of the body and may tend to forget that until relatively recent times this word meant nothing else than a textile fabric woven of threads. Until late in the seventeenth century no one could have conceived how such a name could be applied to a part of the body of an organism, and indeed the word was not used in this sense in literary English until about 120 years ago. It is strange to reflect that the modern English usage derives indirectly from a fallacy about the microscopical structure of plants published by Grew in 1682. FIG. I . Part of Grew's drawing of a piece of a branch of the sumach, highly magnified. The supposed fibres, corresponding to the threads of lace, are well shown. (Enlarged from Grew, 1682, plate 40.) Grew considered that all the parenchymatous parts of a plant, including its fruit, consisted of 'Threds or Fibres', variously woven together. In a wellknown passage (1682, pp. 121-2) he compares these fibres to the threads of lace. The comparison is actually to lace while it is being made in a horizontal piece upon a cushion. The pins, inserted vertically into the latter, are imagined to be hollow, and thus represent the vessels of the wood. The lace is supposed to be made in many thousands of layers, one on top of another. The holes between the threads represent the cavities of the cells; and it is understandable that the layers of threads could be added in such a way as to make closed vesicles instead of holes. 'And this', remarks Grew, 'is the true Texture of a Plant: and the general composure, not only of a Branch, but of all other Parts from the Seed to the Seed.' Of Grew's many figures, the one that illustrates best the ideas just expressed is his representation in perspective of a piece of a branch of the sumach, highly magnified. This figure is here reproduced as fig. 1. It shows clearly his belief in the essentially fibrous nature of plant parenchyma. It follows from what Grew writes and also from this illustration, that if there is a unit in plant tissue, that unit is a fibre and not a cell; for the latter is merely a space left here and there by the intertwining of the fibres. For well over a century after the time of Grew, attention continued to be focused mainly on the cell-wall rather than on the contents of the cell. The belief that the wall consisted of microscopically-visible fibres did not persist (though the nomenclature based on that mistake was retained); but another kind of error began to be generally accepted. It was thought that the system of cell-walls was a continuous substance throughout the whole of a plant. Spaces (cells and vessels) were supposed to appear in this continuous substance ; nourishment was thought to flow through them to the really essential constituent, the continuous membrane or cell-wall. The cells were often regarded as freely open to one another. According to this belief in its extreme form, a plant would consist of two substances, a membranous meshwork and the nourishing fluid filling its meshes; each would be completely continuous. There would indeed be cells, or enlarged meshes; but there would be no unit of structure. One of the most obstinate adherents to this view was Brisseau-Mirbel, who wrote (1808, pp. 14 and 128): 'The first idea, the fundamental idea is that all vegetable organization is formed by one and the same membranous tissue, variously modified. This fact is the base of all the others. The contrary idea is a source of errors. . . . Plants are composed of cells, all the parts of which are continuous among themselves; they present only one and the same membranous tissue.' In very early times the contrary opinion began to appear, though the development of two opposing schools of thought came slowly, and more than a century was to elapse after the publication of Grew's Anatomy of Plants before anyone began to think clearly in terms of the cell as a unit. Grew's contemporary, Malpighi, nevertheless inclined towards what may be called the utricular view: he did not regard the cell as merely a space between interlacing fibres. This appears, for instance, in the account of the microscopical structure of the petals of the tulip and other plants, in his Anatome plantarum (1675, pp. 46-47). He remarks that when the IB surface is torn, the outflowing material contains microscopical objects resembling icicles in shape, which had d r a ^ n g r f a r o W b e e n entangled loosely together in the intact petal. Each cells from the petal is formed, he tells us, of utricles arranged in a row. One of the tulip. (Mai- of n i s illustrations of such a row is reproduced here as 288 fi' i645)' P a 6 ^S- 2- There is nothing in such descriptions that would call to mind the lacework of Grew. The utricular view had its supporters in the succeeding century. Duhamel du Monceau (1758) made some suggestive observations. He separated by maceration small pieces of the 'tissu cellulaire' of the branches of the lime tree. Sometimes he succeeded in detaching little oval bodies of fairly regular shape, which he thought might be the 'vesicules' of Malpighi and Grew; but neither his words nor his illustration (fig. 7 on his plate 2) permit us to feel any more certain of this than he did himself. Before the cell could be regarded as a unit it was necessary to show that the wall between two contiguous cells was double and that the cell could therefore be isolated as a separate object. This separation can be achieved because of the relative softness or solubility of the pectose or pectic acid of the middle lamella. The first person who clearly demonstrated that plant-cells are separable units was G. R. Treviranus. Referring to the globules mentioned by Wolff (see Part I of this series), Treviranus writes (1805, p. 233), 'I have nowhere seen these little bladders so clearly as in the buds of Ranunculus Ficaria L. A thin section of this, brought under the magnifying-glass in water, allows itself to be divided by the point of a needle into nothing but little bladders (in lauter Blaschen).' He generalizes thus: 'The first beginning of all organization of the living being is an aggregation of little bladders that have no connexion with one another. From these arise all living bodies, just as they are all dissolved into them again.' This statement, which resembles Oken's speculations as expressed in Die Zeugung (1805), is of course an induction based on insufficient evidence; but Treviranus's conception of the cell as a unit had actual observation behind it. Shortly afterwards Link (1807) made a considerable advance towards the understanding of the separateness of cells. He remarks (p. n ) that most authors suppose the existence of an open communication between them, so that the 'Saft' of one can pass into another. Link denies this. When he put cut twigs in coloured fluids, he never observed the passage of the fluid from one cell to another, except when a cell-wall happened to be damaged. He FIG. 3. Link's drawalso noticed that certain plant cells have red sap, but ing ^a transyerse sec, , , 1 1 1 J T - t l 0 n through the pith of are surrounded by others that are uncoloured. To Datura tatuia> s h o w i n g ohnim,thcisellssubajnedctthiesirowfapllasrtwiceurlearseipnateraretes.t. H' Aistrepmlaacreks ethaceh dcoeul1blaebutlinseo nw,hane"re . . . . . . 1 1 1 / \ other. (Link, o1807, plate where cells adjoin one another , he writes (p. 13), 1 pg 2 \ 'one often notices a double line, as if there were a space between the cells.' He illustrates this by a drawing, here reproduced asfig.3. Two years later Link (1809) was quite definite on this subject. 'Cellular tissue', he writes (p. 1), 'consists of little bladders completely separated from one another; but their membranes [cell-walls] usually lie so close to one another that they appear to constitute only a single partition-wall.' He instances the petioles of the large leaves of Rheum undulatum and other species as suitable objects for exhibiting the separateness of the cell-walls. The petioles of certain ferns (he mentions Scolopendrium vulgare and Adiantium pedatum) provide striking examples. Link continued his investigation (1812) by studying boiled plant tissues, such as kidney-beans and the roots of several garden-plants. He found that not only the cells of the parenchyma but also the elongated bast-cells are separated by this treatment or become separate if gentle pressure is subsequently applied. Link also found completely separate cells, each with its own wall, in many ripe fruits, especially berries. He considered all partitions between cells as originally double; often they remain so and the double wall can be seen, while in other cases it becomes single by subsequent fusion. Meanwhile L. C. Treviranus (1811) had arrived at similar results. He found that in some cases mere sectioning of plant tissue sufficed to make the cells fall apart, and that this separation can be promoted by pulling gently on the tissue. He reached this conclusion (p. 1): 'It is at once evident to every unprejudiced observer that the cellular tissue of plants is an aggregate of semitransparent bladders, which cohere to a certain extent.' He attacked the contrary opinions of Grew and Brisseau-Mirbel. Moldenhawer (1812) tried macerating sections of plant tissue in water. He expressed his results very clearly. He remarks on the double nature of the cell-wall. But maceration [he writes (pp. 81, 86)], if only employed with due caution, also splits the cellular substance into separate, self-contained utricles . . . [the cellular substance] breaks down sooner or later, according to the firmness of the connexion, into single closed utricles that show no trace whatever of injury, which they necessarily would reveal in the form of irregularly-broken, jagged walls, if there were violent rupture of one and the same continuous tissue. . . . Such an aggregate of single cells has nothing in common with a tissue (Gewebe), and the name cell-tissue (Zellgewebe) therefore appears to be less appropriate than that of cellular substance, that is, substance consisting of cell-shaped utricles. Like Link, Moldenhawer mentions cells with differently coloured sap lying close to one another. He explains Grew's 'Threds or Fibres' as mere wrinkles in cell-walls. Dutrochet (1837) introduced the use of concentrated nitric acid, in a tube immersed in boiling water, as a macerating agent. He showed by this means that the cell-wall is double, and that the substance of plants can be separated into its constituent elements or cells. By this time, however, the older view had lost its grip: even BrisseauMirbel admitted his error. In his work on the liver-wort Marchantia he writes of 'the utricular composition of the tissue, which I formerly denied, and of which today I confess the reality' (1835, p. 352). He thus allowed that the cavities of the utricles were not in free communication with one another; but he still adhered to the singleness and continuity of cell-walls. He communicated his paper on the subject to the Academie des Sciences in 1831, but it was not published till four years later. In the meantime his conversion had been complete, and he announced it very frankly in a note appended to the paper. The cellular tissue of Marchantia polymorpha [he writes (p. 363)] did not offer me spaces between cells. These canals, which are nothing else than the spaces the utricles leave between them, and which for this reason M. Treviranus calls intercellular, exist in many plants and are absent in others. Thus one can say that the utricles composing cellular tissue are welded together either completely or incompletely. . . . today, when I have obtained the most direct proof of the utricular composition of the tissue, I understand and I see the spaces, which I neither understood nor saw before, and I retract my objections to the fine discovery of M. Treviranus. These magnanimous words may be said to mark the end of the controversy about the morphological separateness of ordinary plant-cells. THE DISCOVERY OF THE CELL-MEMBRANE A clear picture of the nature of the cell could not be obtained so long as the wall was regarded as an essential part. It was necessary to realize that the wall was sometimes present and sometimes absent, while the cell itself was always bounded by a special membrane, not mechanically separable from the ground-cytoplasm within. This advance could not be made in one step. It was necessary first to discard the cell-wall as unessential to the idea of a cell, which was then looked upon as consisting of 'naked' protoplasm. The discovery of the cell-membrane came much later. It must be remarked at the outset that the early workers generally called the cell-wall the membrane or Membran. This rather confusing usage will appear in various passages quoted in the present paper. Most cells of animals are obviously devoid of a covering corresponding to the cellulose cell-wall of plants, but one can understand why the cytologists of Schwann's time did not recognize this. It was partly because they were swept away by the new idea of the 'Uebereinstimmung' of all cells, whether plant or animal, and this correspondence would be greatly weakened if one of the most characteristic features of plant cells were found to be absent from those of animals. They therefore looked with confidence for a cell-wall in the animal cell, and in many cases found what they were seeking. The free border of intestinal epithelial cells, the vitelline membrane of eggs, and the cortical layer of ciliates are examples of real structures that seemed to them to represent the cell-wall. Such walls, however, were also described where in fact there are none. This was probably due to the appearance of double lines at the edges of cells, caused by the low numerical aperture of the microscopical objectives available at the time. The study of animal embryos might have caused a change of opinion, because the blastomeres, which clearly represent the whole organism in its early stages, are devoid of anything resembling the cell-wall of plants; but if blastomeres were to be regarded as cells, as was being suggested, then a cellwall must be shown to exist, according to the opinion of the day. Reichert (1841), who had the help of du Bois-Reymond, investigated this matter in the developing eggs of amphibians. They isolated uninjured blastomeres in late cleavage-stages and placed them in distilled water under the microscope. According to their account, a surface-membrane was pushed off by endosmosis. They homologized it with the familiar cell-wall and regarded it as evidence for the cellular nature of blastomeres. Ecker reported that he had seen movements in the blastomere of the frog; he thought these were inconsistent with the presence of a cell-wall, and therefore with the cellular nature of blastomeres (see Remak, 1851). It was Ecker's report that brought the great embryologist, Remak, into the controversy, unfortunately on the wrong side. He opposed Ecker's opinion, and claimed to see two firm membranes surrounding each of the upper blastomeres in the eight-cell stage of the frog; he regarded this as evidence that blastomeres were cells (Remak, 1851). Like Reichert, he saw the membrane surrounding the blastomeres swollen by the osmotic absorption of water, and might have realized the existence of a cell-membrane not corresponding to the cell-wall of plants. He was impressed, however, by the detached envelope that he saw round blastomeres that had been treated with various reagents (hydrochloric, sulphuric, and chromic acids, mercuric chloride, and alcohol), and convinced himself that this envelope corresponded to the cell-wall of plant tissues (1855, pp. 135-6, 173-4). He opposed the view that the surface of a blastomere was merely a modified part of the protoplasm. The cell-walls of plants could be exhibited and distinguished from the underlying protoplasm by simple techniques already in common use in his time, and he looked forward eagerly to the discovery of methods that would play the same part in animal cytology, with equal clarity and certainty. The first person who stated in unequivocal terms that the cell-wall is not a necessary constituent of the cell was Leydig. He wrote (1857, p. 9): . . . not all cells are of bladder-like nature; a membrane separable from the contents is not always distinguishable. For the morphological idea of a cell one requires a more or less soft substance, primitively approaching a sphere in shape, and containing a central body called a kernel {nucleus). The cell-substance often hardens to a more or less independent boundary-layer or membrane, and the cell then resolves itself, according to the terminology of scholars, into membrane, cellcontents, and kernel. The idea that a primitive cell is devoid of a wall was recognized by de Bary in the first of his important contributions to the knowledge of Mycetozoa (i860, p. 161). He wrote of the flagellulae that have emerged from spores: 'In the swarmers there is no cell-membrane in the ordinary sense of the term, but there is indeed a nucleus. As has already been shown above, they are to be regarded as skinless or primordial cells. . . . " Max Schultze (i860), who had studied protoplasm in various Protozoa, especially the Foraminifera, and had confirmed some of de Bary's work on Mycetozoa, made the following generalization (p. 299): 'But the less perfectly the surface of the protoplasm is hardened to a membrane, the nearer to the primitive membraneless condition does the cell find itself, a condition in which i t exhibits only a small naked lump of protoplasm with nucleus. . . . " L a t e r i n the same paper (p. 305) he repeats his definition of a cell as 'ein nacktes Protoplasmaklumpchen mit Kern'. Schultze now turned to the study of striated muscle and produced the paper (1861) that is so commonly quoted in textbooks of the history of biology. The writers of these, however, have overlooked the rather peculiar character of Schultze's communication. He had set himself the problem of discovering the nature of the 'Muskelkorperchen' or small masses of protoplasm containing nuclei that occur among or outside the contractile elements of striated muscle. He reached the remarkable conclusion that each is to be regarded as a cell. It follows that in the fully differentiated muscle-fibre the contractile elements are extracellular. He knew that each little mass of protoplasm has no cell-wall surrounding it, and his particular point is that this absence of a wall does not indicate that the Muskelkorperchen is not a cell. This led him on to consider what are the essential characters of a cell. He was not content with the definition that had prevailed in the past: 'a vesicular structure with membrane, contents, and nucleus.' To find out what was essential he turned to blastomeres. 'From what has gone before', he wrote (p. 11), 'the component parts of the blastomeres are nucleus and protoplasm, and our definition of what one has to call a cell assumes the following form: a cell is a little lump of protoplasm, in the interior of which lies a nucleus.' The actual words are: 'Eine Zelle ist ein Kliimpchen Protoplasma, in dessen Innerem ein Kern liegt.' Schultze refers in a footnote to Leydig's definition. In another place (p. 9) he defines cells as 'little sheathless lumps of protoplasm with nucleus'. He says that the protoplasm holds together because it does not mix with water. 'A membrane', he insists, 'is not necessarily connected with the idea of a cell.' He even considers it a sign of degeneration. 'A cell with a membrane differing chemically from protoplasm is like an encysted infusoriumlike an imprisoned monster.' Schultze recognizes that his definition cannot be wholly reconciled with the word cell, which conveys the idea of something provided with a distinct wall. Briicke (1862) agreed with Schultze that the wall is not a necessary attribute of animal cells, but Remak (1862) repeated and amplified his old conclusions. He insisted that the animal cell has at its surface a 'Hulle' or 'Membran', chemically distinct from the ground cytoplasm; and he mentioned once more the separation of this envelope from the underlying protoplasm by chemical agents. He considered that in animals, as in plants, cell-division occurs by the ingrowth of solid septa from the envelope into the protoplasm. Schultze, however, had the support of de Bary in the latter's monograph on Mycetozoa (1864). 'The skin of the cell', wrote de Bary (p. 106), 'is therefore no essential attribute of the cell: it may be formed, but need not.' A notable advance had been made when the cell came to be regarded as a lump of naked protoplasm with a nucleus, even though the existence of the cell-membrane was not yet clearly recognized. The advance demanded a change in nomenclature, for a cell is a box and a lump of protoplasm is not. To no one did the old name seem so absurd as to Sachs. Hooke had named the cells of plants after the cells of the comb in a beehive; and if it were right to call the protoplasmic unit a cell, then, according to Sachs (1892, p. 60), a bee should be called a cell, and the cell of the comb should be called the capsule of the cell! Hanstein had been thinking along these lines long before. He recognized that such degree of independence and functional individuality as the cell possesses reside in its protoplasm. This word, however, conveys no sense of an object, but only of the material of which an object may consist. He therefore coined the name 'Protoplast' for the protoplasmic part of a single cell (1880, p. 169). He applied the name to the vital units of both plants and animals. The unit might secrete a wall, but he recognized that this was far from being necessary; for the protoplasts of many animals had no external covering (p. 217). Thus Hanstein agreed in general with Leydig, Schultze, Briicke, and de Bary, but expressed his ideas more exactly. The word protoplast is useful, and modern cytological writings would benefit from a more frequent use of it. It gradually became apparent that if the vital unit were indeed naked, yet it might at least have a skin. The early comments on this subject are equivocal : we cannot tell whether the authors were referring to the special outer layer or ectoplasm exhibited by certain protoplasts, or to the membrane that is always present. Hanstein himself remarked on the skin-like, firmer, 'aussere Hautschicht' of the protoplasm of the plant-cell (1880, p. 167 and fig. 1); he called the Tonoplast the 'innereHautschicht'. This was a considerable advance beyond the unqualified idea of naked protoplasm, but Hanstein did not undertake the physiological studies that would have been necessary to disclose the real nature of the membrane. The swelling and shrinkage of cells by osmotic pressure might have led to its recognition had not the early work on this subject been confined to the vacuolated plant cell. Pringsheim (1854, p. 51) noticed that solutions of salts, acids, and sugar caused the 'Zellinhalt' to collapse inwards away from the cell-wall. In his famous study of this subject de Vries (1884) denned plasmolysis as the detachment of the living protoplasm from the cell-wall (Zellhaut) through the action of aqueous solutions. He used various plant cells, but chiefly the violet epidermal cells of the lower surface of the leaf of Tradescantia discolor. He realized that the membrane responsible for the osmotic phenomena was at the boundary of the vacuole, and since it was involved in the maintenance of turgor, he named it the Tonoplast (1884; see also 1885, p. 469). Osmotic studies of animal cells would have given new insight into the nature of the boundary of living protoplasm, because the swelling and shrinkage of cells that lacked a vacuole would have directed attention to the true cell-membrane; but such studies were not undertaken at the time, and the view expressed by Leydig and the others prevailed. For more than thirty years it was generally agreed that in its most primitive form the cell consisted of naked ground-cytoplasm (with a nucleus). It required the genius of Overton to recognize a clear distinction between the ground-cytoplasm and cellmembrane, and to bring forward strong evidence that the latter was always present as a covering. He had spent a number of years in a study of the osmotic properties of the living cells of plants and animals. It was known that an 8 per cent, cane-sugar solution caused slight but definite plasmolysis in Spirogyra. Overton (1895) tried a solution of ethyl alcohol of the same osmotic pressure, and found no plasmolysis. He thought it possible that this might be due to easy penetration of the outer part of the protoplasm by alcohol. He extended his observations to many diverse plant-cells, with the same result. He then discovered that a number of other substances (various alcohols, ethers, acetone, aniline, phenol) exerted no plasmolytic effect. He realized that it was not the cellulose cell-wall but the 'Grenzschicht' of the protoplasm FIG. 4. Overtoil's diagram that was responsible for osmotic effects. He found of a typical plant-cell, showthat in general animal cells resemble those of ing the distinction between plants in admitting certain particular kinds of the cellulose cell-wall (cm.) and the cell-membrane (pi. substances and excluding others. On the basis ext.). (Overton, 1895, fig. 1.) of these observations he produced the diagram of the plant-cell here shown as fig. 4. The diagram makes a clear distinction between ground-cytoplasm, cell-membrane, and cell-wall. Hanstein's diagram (1880, fig. 1) is comparable, but it was not based on a clear understanding of the remarkable properties of the cell-membrane. Although Overtoil's diagram marks an important advance, the kind of evidence on which it was based was not quite so satisfactory as an ocular demonstration would have been. A distinction must be drawn between what can be seen and what inferred on indirect evidence. In making his diagram Overton had to decide arbitrarily what thickness he would ascribe to the boundary-layer of the protoplasm. Overton noticed that the substances that enter cells easily (and hence do not cause plasmolysis) are those that are more soluble in ether, fatty oils, and similar substances than in water, while those that have difficulty in entering cells are those that are readily soluble in water but scarcely or not at all in ether and oil. He was thus driven to the conclusion that the outer layer (Grenzschicht) of the cell must be impregnated with a substance that has dissolving properties similar to those of fatty oils. He rejected the possibility that the substance could be a triglyceride, because he found that filamentous algae could be kept without damage in a solution of sodium bicarbonate that would have saponified a fat. He thought cholesterol or a cholesteryl-ester, with perhaps lecithin and sometimes triglycerides in addition, as the most probable composition of the impregnating substance (Overton, 1899). So far, Overton had relied on the absence of plasmolysis as evidence that certain substances had entered cells. He now sought more direct indications by following the behaviour of coloured substances (1900). He studied the capacity of various dyes to enter cells, and related it to their solubility in lipoids. He found in general that basic dyes are soluble in melted cholesterol and in solutions of lecithin in organic solvents, and enter living cells easily, while acid dyes are insoluble and do not enter. The exceptions proved the rule, for methylene blue tannate, though basic, is almost insoluble in the solvents mentioned and is not taken up by living cells, while the acid dyes, methyl orange and tropaeolin, are somewhat soluble and are taken up slowly. Overton thus proved the connexion between lipoid-solubility and the capacity of substances to enter living cells, and strengthened the evidence he had previously obtained for the existence of a special lipoid-containing membrane on the surface of cells. Very various methods have been adopted by later students of the cellmembrane. More refined investigations of permeability have been made; the tension at the surface has been measured; the capacity of various agents to destroy the membrane has been studied; approximations to its thickness have been obtained by indirect methods. The results of these and other experimental studies have been brought together in theoretical diagrams of the molecular and ionic structure of the membrane. The researches on this subject have been admirably reviewed by two active workers in the field (Davson and Danielli, 1943; Danielli, 1951). It suffices for the present purpose to say that although much has been discovered, nothing has occurred to shake the foundations laid by Overton. His work has been of great importance for the cell-theory. Nothing can be a unit that has not a distinguishable boundary. Thanks to Overton we know where that boundary is: we can say with confidence, in many cases, what is part of the protoplast and what is external to it. CONNECTIVE STRANDS BETWEEN CELLS There are several theoretical possibilities as to the nature of the connective strands that are sometimes seen to extend from one cell to another. These are illustrated in fig. 5. In A and B the connexion is made solely through the cell-wall. In c the strand consists of the material of the cell-membrane, while in D and E the ground cytoplasm participates, though the membranes, double (D) or fused (E), still to some extent separate the cells. In F there is direct continuity between the ground cytoplasm of the two cells; and though there may be no cyclosis involving the passage of protoplasm from one cell to the other, yet there is obviously an easy route for the diffusion of molecules and ions. It is often impossible to decide which kind of strand is present in particular cases. The subject is difficult from the technical standpoint: fixed preparations are liable to be misleading. As Weiss has pointed out (1940, p. 35), if a preparation shows connexions between cells, we often cannot be sure that they are not formed of coagulated intercellular matter; while if it does not, there may have been connexions in life that were broken by fixation. Connexions involving the cell-membrane or ground-cytoplasm are often called cell-bridges; but a bridge is something across which there is movement, and it would be begging an important question to give them a general name that implied the transport of material from one cell to another. The noncommittal name of Plasmodesmen (singular Plasmodesma), introduced by Strasburger (1901, pp. 503, 607), is preferable. The Greek plural plasmodesmata will be substituted here for the German, as being better suited for international use. It is convenient to employ it in a wide sense, to cover all the possible arrangements shown diagrammatically in fig. 5, C-F. It should be mentioned that the most active study of plasmodesmata took place during FIG. 5. Theoretical possibilities as to the nature of connective strands between cells. the long period when cells were commonly regarded as 'naked' lumps of protoplasm, before Overton had proved the existence of the cell-membrane: nice distinctions such as those shown in fig. 5, c-F, would not have had much meaning at that time. Plasmodesmata must have been seen first in the sieve-plates of plants, which were discovered by Hartig in 1837 (see Esau, 1939) and later reinvestigated by him (1854). At first their real nature could not be appreciated because biologists had not yet recognized the nature of protoplasm itself (see Part II of the present series of papers). Sachs, however, who studied them in Dahlia (1863), clearly recognized their character, for he mentions that the apertures of the sieve-plates are filled with protoplasm, which he stained with iodine. Modern research shows that two adjacent sieve-tubes are separated by a layer of a substance containing phospholipine, where they abut on one another in the apertures of the sieve-plate; and in some cases there are two layers of the lipoid material, with a very narrow space between (Salmon, 1946, p. 77 and fig. 11). If the lipoid represents a thickened cell-membrane, the plasmodesmata connecting sieve-tubes are of the types represented diagrammatically in fig. 5, D and E. After the discovery of the plasmodesmata of sieve-tubes, many years elapsed before anything of the kind was shown to exist in other plant-cells. It is stated by Goebel (1926, p. 118) that Hofmeister demonstrated connexions between the cells of the endosperm of Phytelephas and Raphia in the course of a lecture given during the winter of 1873-4; but nothing was FIG. 6. Tangl's figure showing plasmodesmata in the endosperm of Strychnos nux-vomica. The protoplasm (including that of the plasmodesmata) has been stained with iodine-iodide. i, inner, and m, outer part of the cell-wall; z, material lying between adjacent cell-walls. This figure (with those accompanying it in Tangl's paper) was the earliest representation of any plasmodesmata other than the strands connecting sieve-tubes. (Tangl, 1879-81, plate 5, fig. 10.) published. Fromman (1879), w n o had c o m e to the conclusion that the protoplasm of the ganglion-cells of the mammalian retina is continuous from cell to cell, turned to plants in search of similar connexions and claimed to find them among the epidermis and parenchyma of the leaves of Rhododendron and Dracaena. Tangl (1879-81) discovered plasmodesmata between the cells of the endosperm of Strychnos nux-vomica and of palms (Areca oleracea and Phoenix dactylifera). He showed by staining with iodine that the protoplasm was continuous between neighbouring cells through narrow canals traversing the cell-membrane (fig. 6), and considered that the cells entered into a unity of a higher order through these connexions. He thought that there was a striking similarity between the plasmodesmata of endosperm and the (supposed) fibres of the remnant of the mitotic spindle (p. 182). Elsberg (1883) copied the use of silver nitrate and gold chloride from animal histology, and by the use of these reagents claimed to find connexions between various cells of flowering plants. Russow (1884), who had already studied sieve-plates (1881) and was acquainted with Tangl's discovery, announced the existence of intercellular (non-nucleated) protoplasm in the medullary rays of Acer, and said that it was connected by strands with cellular protoplasm. He went farther and claimed that 'in every plant throughout its life the whole of the protoplasm stands in continuity' (p. 581). Gardiner (1884) also thought the existence of protoplasmic filaments connecting cells by no means uncommon: the wide perforations in sieve-plates he regarded merely as special cases of a quite usual arrangement. Kienitz-Gerloff (1891a) made a general study of the subject, and reported the existence of plasmodesmata in mosses, ferns, horse-tails, and conifers, as well as in mono- and dicotyledons. He came to the conclusion that the whole of the living material of the higher plants is bound together by connexions, though he admitted (p. 23) that in many cases he could not find them. He examined them carefully in the tissues of Viscum, where they are particularly evident, and showed that they are not related to the spindle, as this disappears completely before they are formed (18916). Kuhla (1900), like Kienitz-Gerloff, found the mistletoe particularly suitable for investigations of the subject. He considered that all living cells were connected by plasmodesmata. Davis (1905), in a useful general discussion of the whole problem, pointed out that plasmodesmata arise in some cases by incomplete cell-division, in others by outgrowth from previously separate cells. Typical plasmodesmata, in his view, arise in the latter way and are to be regarded as differentiations of the cell-membrane only (pp. 224-6). Plasmodesmata of particular interest exist in cycads, where they were discovered by Goroschankin (1883), connecting the 'Corpusculum', as he called it, with the surrounding jacket-cells of the endosperm. It is unfortunate that this particular cell appears not to have a generally recognized name; for it cannot properly be called the ovum until its nucleus has divided to form the ovum-nucleus and the ventral-canal-nucleus. It will here be called the developing ovum. Goroschankin found that the cell-wall surrounding it contains a great number of pits, each provided with a sieve-plate through which the protoplasm of the jacket-cells is in open communication with that of the developing ovum. Smith (1904) subsequently discovered the long 'haustoria' connecting the developing ovum of Zamia, through pores in its wall, with the jacket-cells. There seems to be no doubt that the haustoria of cycads are nutritive plasmodesmata. Among the Protista, connective strands between cells are particularly evident in Volvox, where they were discovered by Cohn (1875). He described the 'Tiipfelkanale' or pit-canals in the cell-wall and the delicate thread-like processes that appear to pass through them from one cell to another; he figures these in V. globator (in his fig. 1 on plate 2). Strangely enough, Cohn himself thought that there was no actual connexion between the cells, as he supposed that the pit-canals were closed (p. 95); this is curious, because in his figure he shows obvious plasmodesmata connecting the cytoplasm of adjacent cells. Modern studies (Janet, 1912, p. 48) show that V. globator provides one of the clearest instances of the kind of connexion that is illustrated diagrammatically in the present paper in fig. 5F. The reality of protoplasmic communication in Volvox was recognized by Biitschli (1883). Meyer (1895, 1896) also regarded the strands as 'Plasmaverbindungen'; he pointed out that they are long and thin in V. aureus, but shorter and much thicker in V. globator. It is particularly suggestive that the growing macrogamete of Volvox is abundantly provided with plasmodesmata connecting it with the surrounding cells (Janet, 1912, p. 91); the arrangement is reminiscent of the developing ovum of cycads. Among the Metazoa, plasmodesmata were first discovered in the skin of mammals. We may overlook Henle's (1841) description of prickles (Stacheln) projecting from the surface of the cells of the chorioid plexus, partly because there is no certainty whether what he saw were cilia or elements of the striated border of the cells, and partly because it is unlikely that these cells are in fact connected by plasmodesmata. The discovery was made by Weber (1858) in suppurating skin and in epithelial cancers. Weber himself, however, regarded the strands as cilia. They were studied by Schultze (1864a) in the lower layers of the epithelium of the mammalian tongue and skin. He made the mistake of thinking that each cell had its own prickles (Stacheln), which interdigitated like the bristles of two brushes pressed together. This opinion he retained in a second communication on the same subject (18646). Schron (1865) considered that the appearances really indicated the existence of canals in a 'Membran' separating skin-cells. Bizzozero (1876), who appears to have waited six years for the publication of his paper, was the first to describe the plasmodesmata of skin correctly as direct connexions between one cell and another. No one has ever produced evidence that they represent connexions of the type shown in fig. 5F. It seems reasonable to regard them as serving to hold the cells together mechanically, while allowing bending movements. Similar connexions have been reported from time to time in very various animal cells. Heitzmann (1873) was an enthusiastic student of plasmodesmata. He claimed to display them by the use of silver and gold impregnations in a wide variety of mammalian tissues, including bone-marrow and even cartilage. Kultschitzky (1888) claimed to find protoplasmic connexions between the smooth muscle cells of the intestine, and looked forward to the discovery of a universal system of such connexions between neighbouring cells in both plants and animals. Connexions between notochordal cells, resembling the prickles of skin-cells, were reported by von Ebner (1896). Reports of direct connexions between nerve-cells were much more plausible because of the undoubted transmission of the nervous impulse from one cell to another. Different opinions were held on this subject from early times, but no strong evidence was forthcoming until the eighties. Using his potassium dichromate and silver nitrate method, Golgi (1883, p. 289) never found a single anastomosis among the ramifications of the main processes of nervecells; but he considered that other processes existed, which subdivided in a complicated manner and anastomosed, so that the nerve-cells lost their individuality and took part in a nervous reticulum. His (1886) was the first to base the contrary view on sound foundations and state it clearly. As a result of his study of human embryos he came to regard the axons of nerve-fibres as outgrowths from separate nerve-cells that push their way between other tissue-constituents; he denied that these outgrowths ever form actual anastomoses (pp. 509, 513). He remarks: 'I present as an established principle, the proposition that every nerve-fibre arises as an offshoot from one single cell. This cell is its embryonic (genetisches), nutritive, and functional centre, and other connexions of the fibre are either only indirect, or have originated secondarily' (p. 513)- Shortly afterwards Forel arrived at the same conclusion from his work on the cavy, but expressed it more tentatively: 'I might presume that all fibre-systems and so-called fibre-nets of the nervous system are nothing else than nerve-processes, [each] always [arising] from a particular ganglioncell' (1887, p. 166). His (1889) went on to describe the outgrowth of the axon from the neuroblast in various vertebrates. The idea of the separateness of nerve-cells has been followed up by many workers, notably Ramon (1934, &c), in modern times. There are still distinguished neurologists who oppose the neurone-theory and view the nervous system as a reticulum, but some of their arguments are not strong (as, e.g., when Boeke (1940, p. 144) attacks the cell-theory, and the neurone-theory as part of it, on the ground that it 'belonged to the mechanistic and analytic mental attitude' of the nineteenth century). The modern literature referring to the chordates has been well summarized by Nonidez (1944), who reaches the conclusion that nerve-cells are not directly continuous with one another at the synapse. The nervous system of coelenterates has long been supposed to provide strong evidence for the reticular theory. The belief that the nerve-fibres of these animals are continuous from cell to cell originated with Korotneff (1876), who studied the layer of nerve-cells and fibres that lies below the external epithelium of the acrorhagi ('bourses marginales') of Actinia. He described the fibres as running without a break from nerve-cell to nerve-cell. He even thought the fibre maintained its individuality within the nerve-cells. Korotneff mentioned only single rows of cells, but subsequent workers believed that the apparent nerve-net of coelenterates was formed by continuous fibres passing uninterruptedly from cell to cell. This opinion was generally accepted for several decades, though Schafer had stated in the most definite manner (1878) that each fibre of the bipolar cells of the sub-umbrellar surface of Aurelia 'is entirely distinct from, and nowhere structurally continuous with, any other fibre'. He knew that the fibres came into close relation with one another, and thought it reasonable to conclude that nervous impulses passed from one to another, but he considered each nerve-cell with its two processes a separate unit. This was confirmed in modern times by Bozler (1927), who treated the fresh tissues of the jellyfish Rhizostoma with reduced methylene blue solution and exhibited the nerve-cells with their processes as separate units (some bipolar, others multipolar); the units make contact with one another but do not anastomose. He denied, therefore, that the nervous system of Rhizostoma is a genuine nerve-net. The results of the physiological study by Pantin (1935, a and b) of the nervous system of the sea-anemone Calliactis are consistent with the belief that here also the nervous system consists of cellular units. There are, of course, cases in which nerve-cells make such evident junctions that an actual syncytium is formed. Thus in cephalopods each of the two first-order giant-cells, which lie in the brain close to the statocyst, sends back a giant axon, and the two axons are connected by a wide bridge as they pass through the palliovisceral ganglion (Young, 1939). Again, the giant fibre that runs along the ventral nerve-cord of the polychaet Myxicola is part of a syncytium, for at least 1,300 nerve-cells are continuous with it, directly or indirectly (Nicol, 1948). These interesting facts, however, throw no light on the nature of ordinary synapses. In the light of existing knowledge it is best to draw the provisional conclusion that the nervous system does not provide us with convincing examples of plasmodesmata, though actual syncytia are found in particular cases. It was found by Berthold that the developing eggs of the nematomorph worm, Gordius, are fixed together like grapes in bunches (see von Siebold, 1843). Reinvestigating the matter, Meissner (1856) reported that the eggs originate in groups of 8-20 in the ovary by bulging outwards from a mothercell, and remain for a long time in organic connexion with one another through their stalks. It is necessary to mention this, because it would have been the first example of plasmodesmata discovered in animals, if true. It appears, however, from the careful work of Vejdovsky (1888, p. 204), that in reality the eggs are only held together in groups by the ovarian epithelium, which becomes very thin in late stages and closely pressed to the surfaces of the eggs. The connexion of animal eggs with external protoplasm was first reported by von Ihering (1877) in the lamellibranch Scrobicularia, six years before Goroschankin's discovery in cycads. According to von Ihering, the follicles of the ovary are lined by a syncytium. In this, an egg develops by the accumulation of protoplasm round a nucleus. The egg then projects into the cavity of the follicle and eventually remains in connexion with the syncytium only by a narrow stalk. Yolk is seen in the syncytium, in the stalk (where it is arranged in regular lines), and in the egg. Korschelt (1886) undertook a detailed study of the morphological relation between nurse-cells and eggs in various insects. Gross (1901) described and figured the long, narrow 'Dotterstrange' that lead from the nutritive end-chamber of the ovarian tubule of Hemiptera to the oocytes; the end-chamber itself is a mass of protoplasm in which nuclei are degenerating or have degenerated. Later (1903) he studied the relation between nurse-cells and oocytes in other insects. In the carabid beetle Harpalus he illustrated a nurse-cell nucleus in passage along a narrow neck connecting the oocyte with its food-supply (plate 11, fig. 124). In the hemipteran Triecphora he described cords like Dotterstrange connecting some of the nurse-cells with the main protoplasmic mass of the end-chamber, which itself supplies the growing oocytes through long Dotterstrange. This arrangement in Hemiptera was confirmed by Mestschershaya (1931), who investigated the permeability of oocytes and endchamber to various substances in solution, and reached the conclusion that the end-chamber is adapted to take up food-substances and pass them to the oocyte. There seems to be no doubt that the Dotterstrange of certain insects are genuine plasmodesmata of the type shown in fig. 5F (see, e.g., Korschelt, It was reported by Hammar (1896) that in the cleavage of Echinus the outermost protoplasmic layer is continuous from one blastomere to the next. Protoplasmic connexions between blastomeres might be thought to indicate a primary condition, characteristic of cells in general, and the subject attracted attention. Flemming (1896) considered that blastomeres are primarily separate but in some cases become secondarily connected by strands. Andrews reported that in the starfish and other echinoderms the blastomeres send out thin protoplasmic processes that join them together (1897), and that some of the cells of the blastulae are connected in this way (1899). Hammar (1897) made a general study of this subject, using as fixative a saturated solution of mercuric chloride in evaporated sea-water, with the deliberate intention of shrinking the cells and thus increasing the intercellular spaces and exhibiting the strands crossing them. He claimed to find protoplasmic connexions between blastomeres in various invertebrates from coelenterates upwards, but it would be unwise to place much reliance on results obtained by a method so particularly liable to produce artificial appearances. In Dendrocoelum, as Fulinksi (1916) showed, not only are the blastomeres not connected by plasmodesmata: they do not even touch one another, but lie separately in a fluid derived from the yolk-syncytium. The belief of Sedgwick in the continuity of protoplasm from cell to cell has been so influential that it is desirable to treat separately the controversy he aroused. In studying the development of Peripatus he noticed that the endoderm cells, previously separate, put out branches that anastomosed (1885). This was the origin of his doubts about the truth of the cell-theory. He was later impressed by the structure of the mesenchyme of elasmobranch embryos, which he described as 'a reticulum of a pale non-staining substance holding nuclei at its nodes' (1894). He regarded the ectoderm and endoderm as 'simply parts of this reticulum in which the meshes are closer and the nuclei more numerous and arranged in layers'. What were taken by others to be sites of cell-proliferation were described by him as places where nuclei multiplied. 'In short, if these facts are generally applicable', he wrote, 'embryonic development can no longer be looked upon as being essentially the formation by fission of a number of units from a single primitive unit, and the coordination and modification of these units into a harmonious whole. But it must rather be regarded as a multiplication of nuclei and a specialization of tracts and vacuoles in a continuous mass of vacuolated protoplasm.' The 'vacuoles' of Sedgwick would be what adherents to the cell-theory would call intercellular spaces not filled by cell-walls or other solid matter. He claimed that nerves were laid down before any trace of nerve-cells could be made out. 'In short, the development of nerves is not an outgrowth of cell-processes from certain central cells, but is a differentiation of a substance which was already in position.' These arguments were answered by Bourne (1895), who took Sedgwick's objection to the cell-theory to be based essentially on embryological evidence. Bourne himself did not accept the idea that a multicellular organism is actually a 'cell-republic', but he insisted that it is an aggregate of elementary parts, and thought that Sedgwick's views would take us back to the Cytoblastem of Schwann. He considered that in the case of spiral cleavage at any rate the blastomeres are not connected by protoplasmic strands. Sedgwick did not leave Bourne's rather mild criticism unanswered. His study of Peripatus had led him to the view that 'the differentiation of the Metazoa had been effected in a continuous multinucleated plasmatic mass, and that the cellular structure had arisen by the special arrangement of the nuclei in reference to the structural changes' (1895). He could not accept the idea of the zygote dividing into blastomeres, and he insisted that both ovum and spermatozoon were individuals, simplified so as to make their fusion possible. Animals for him were generally 'tetramorphic': that is, they exhibited four kinds of individual: male, female, spermatozoon, and egg. He thus allowed individuality to the gametes, while denying it to other cells. Sedgwick's ideas were not novel, for Heitzmann had expressed them long before. In the paper already quoted he wrote (1873, P- : 5 5 ) : "The animal body as a whole is one lump of protoplasm, in which are embedded to a smaller extent isolatedprotoplasmic bodies (wandering bodies, colourless and red bloodcorpuscles) and various other substances that are not alive (gelatinous and mucous substances in the widest sense, together with fat, pigment granules, &c.).' He compared the whole of a higher animal to an Amoeba, and denied that there is any such thing as intercellular substance, even in blood; for him, there was only 'Grundsubstanz' and protoplasm (1873, p. 155). Ten years later, about the time when Sedgwick was first turning his attention to the subject, Heitzmann wrote: 'What have previously been considered as cells prove, in our conception, to be nodal points of a network that traverses the tissues' (1883, p. 57). He applied this generalization to both plants and animals. It was Sedgwick's eminence as a zoologist, however, that gave currency to the new ideas. He spread them not only by his writings but through personal contacts. Dobell, one of the strongest opponents of the cell-theory, was trained in his school. There is no reason to suppose that protoplasm flows freely through every connexion we may find between cells. Schultze made a comment on this subject long ago that is memorable for its good sense and moderation. 'But I dispute', he wrote (1861, p. 26), 'that the individuality of cell-life is encroached upon by the anastomoses, and I dispute that in normal circumstances, with full integrity of the individual cells, the situation can be even quite roughly interpreted as a protoplasmic vessel-system.' In much the same sense Strasburger (1901, p. 595) drew a distinction between the existence of plasmodesmata on the one hand, and the loss of cellular individuality on the other. Plasmodesmata should be considered from the functional point of view. Where a necessity has arisen for bulky materials to stream into a particular cell, protoplasmic connexions have evolved for their passage. We have seen examples in the developing ovum of cycads, in the macrogamete of Volvox, and in the oocytes of various insects. Where it has been particularly important for cells to combine the property of holding together firmly with that of allowing changes of relative position, strands are seen to pass between neighbouring cells. Flemming (1896) remarked on the physiological necessity for junctions in certain epithelia. Indeed, the mammalian skin provides one of the most familiar examples of plasmodesmata. Another good example is the mesenchyme of the embryos of very diverse animals. In this embryonic connective tissue there are no extracellular fibres, and the function of holding together is served by direct connexions between cells. In some cases we cannot assign a function to the plasmodesmata; but in general we find them serving some particular purpose, and not existing as biological necessities indicative of an essential protoplasmic unity of the whole organism. We have no reason to suppose that cells usually possess them. The nomenclature of multinucleate masses of protoplasm is very confused. The word coenocyte is generally employed in botany to mean a whole plant containing several or many nuclei not marked off from one another by cellboundaries, but it is also sometimes used for parts of plants that contain more than one nucleus in a continuous mass of cytoplasm. The word syncytium suffers from having been used in different senses by its originator. Haeckel defined a cell (Cellula) as a lump of protoplasm with a nucleus, and called the expression 'mehrkernige Zelle' a Contradictio in adjecto (1866, vol. 1, pp. 275, 296). In coining the word Syncytium he restricted its meaning specifically to a complex formed by the fusion of previously separate cells (1872, vol. 1, p. 161); he applied it to the dermal epithelium of calcareous sponges, which he believed to be of this nature. Cienkowsky, however, had invented the word Plasmodium for a continuous mass of protoplasm formed by the fusion of previously separate cells (1863a, p. 326); he applied the word to a stage in the life-history of Mycetozoa. He knew that this stage was reached by the fusion of nucleate cells, but considered the plasmodium itself to be devoid of nuclei (1863J, pp. 435-6). Haeckel accepted this, and thus drew a distinction between a syncytium and a plasmodium (1872, vol. 1, p. 161). Later he supposed that Mycetozoa were at first nucleate and later non-nucleate; he expressed this by saying that they were syncytia that became plasmodia (1878, p. 51). Later again he evidently revised his ideas on the nature of a Contradictio in adjecto, for he referred to the Siphonales and other multinucleate organisms that lacked cell-boundaries as polykaryote cells (1894, p. 70). Finally he stated distinctly that a Bryopsis or Caulerpa consists of a single cell with many nuclei, and gave a new definition of his word syncytium: 'The whole body consists of a single colossal cell, which includes many nuclei in its voluminous body' (1898, vol. 2, p. 421), He applied the name to Siphonales, Mycetozoa, Actinosphaerium, and certain Foraminifera, but not to any constituent part of any organism; and he equated syncytia with plasmodia. Gegenbaur applied the term syncytium to striated muscle (1874, p. 26); Huxley introduced it into English, with Haeckel's original meaning (1877, p. 113). Delage and Herouard (1896, p. 41) also restricted the sense of the word to cases in which previously separate cells fuse together. In an attempt to reduce the confusion caused by these various meanings I shall use the word syncytium to mean any obviously continuous mass of protoplasm that contains more than one nucleus, whether that mass constitutes a whole organism or part of an organism, and whether the hi- or multinucleate condition has been reached by aggregation of previously separate cells, or by nuclear division without cell-division, or by both aggregation and nuclear division. I shall restrict the word plasmodium to a syncytium formed by aggregation of previously separate cells, whether subsequent cell-division increases the number of nuclei or not. The word coenocyte appears to be superfluous and will not be used. It may be suggested that if used at all, it should refer to a syncytium that constitutes a whole organism. For the sake of consistency it is necessary to classify as syncytia certain temporary arrangements of nuclei and cytoplasm that are not customarily so regarded. In all cases in which nuclei are formed after mitosis before the cytoplasm has completely divided, a short-lived syncytium may be said to exist. In many (perhaps most) animals the male and female pronuclei do not fuse at fertilization; the two sets of chromosomes only come together at the two-cell stage. We have here an example of a short-lived syncytium with two haploid nuclei. As is well known, the process is carried much farther in some copepods, for double nuclei are sometimes seen up to the 32-blastomere stage, and indications of gonomery may persist even later (Hacker, 1892, 1895; Ruckert, 1895). Cells connected by plasmodesmata should not generally be regarded as constituting a syncytium, because the protoplasm is not obviously continuous. Although, as Fol (1896, p. 211) remarks, there is no sharp distinction between cellular tissue and syncytium, yet doubtful cases are rare. It was pointed out by Pringsheim long ago (i860, pp. 229-30) that the hyphae of certain Saprolegniaceae are constricted at intervals, but not divided right across; usually there is only one nucleus between each constriction and the next. Reinhardt (1892, p. 562) described a similar arrangement in Peziza and claimed that a part of the streaming cytoplasm passed through a central opening in the transverse wall. If so, the 'cells' clearly constitute a syncytium. Syncytia appear to have been first noticed in plant tissues by Bauer in the style of Bletia tankervilliae (Orchidaceae) in 1802. His drawings were not published till much later; the book evidently appeared in -sections that were not separately dated (Bauer, 1830-8). In one of the drawings, here reproduced as fig. 7, the loose tissue of the stigmatic canal after fertilization is shown. FIG. 7. Bauer's drawing of cells in the stigmatic canal of Bletia tankervilliae. One 'cellule' contains two and another three nuclei. The drawing was made in 1802. (Bauer, 1830-8, 1st part ('Fructification'), plate 6, fig. 3.) Bauer noticed that there were from one to three 'specks' in each of the 'cellules'. It seems just possible that the 'cellules' were sections of pollentubes. This drawing by Bauer was known to Brown (1833, p. 711), who mentioned it in the famous paper in which he introduced the word nucleus. Brown confirmed Bauer's finding and remarked that it was the only example known to him of more than one nucleus in a cell. Meyen (1837, p. 208) reported that two or three nuclei often occur in 'langgestreckten Zellen'. Unger (1841, fig. 6 on plate V) gave a representation of a cell in the root of Saccharum, elongating to form part of a vessel; it is in fact a syncytium containing three nuclei. In their botanical textbook Endlicher and Unger (1843, pp. 22-23) remark, like Meyen, that there are often several nuclei in 'langgestreckten Zellen'. Nageli (1844, p. 62) mentioned the existence of more than one nucleus, without cellular partitions, in pollen-grains, in the pollentube, and in the embryo-sack. The more recent literature of syncytia in the vegetative cells of phanerogams has been reviewed by Beer and Arber (1920). Syncytia were recorded in the gill-cartilages of Pelobates and Rana by Schwann (1839, p. 23 and fig. 8 on plate 1). It is clear also that Rathke saw the syncytial stage in the development of the crustacean egg. He expresses himself rather obscurely, but this is only to be expected, as the partial cleavage of a centrolecithal egg had never previously been described. In the passage that follows, which is translated from the Latin, he is referring to the cells seen after cleavage. 'In fact', he writes (1844, p. 8), 'before the cells that I have already mentioned originate, there is formed for each of them, among the structural elements of the yolk [i.e. among the yolk-globules], a particular nucleus, consisting of a vesicle filled with a coagulable liquid. As a result of this, every cell now formed is provided with its own nucleus.' Thus Rathke saw many nuclei in the egg before radiating partitions had appeared to divide it (imperfectly) into blastomeres. Kolliker professed to find syncytia in the embryos of certain invertebrates in which cleavage is in fact total, but stated correctly that more than one nucleus occurs in the giantcells (polykaryocytes) of bone-marrow and in certain nerve-cells (Kolliker, 1852, pp. 20-21 and fig. 7). It is interesting to trace the course of events as naturalists were groping their way towards the discovery that a whole organism might consist of a continuous mass of protoplasm containing many nuclei. This truth was first revealed by the investigators of the Mycetozoa. It was known to de Bary (i860) that the plasmodium is formed by the fusion of nucleate cells and that each of the chambers of the sclerotium may contain more than one nucleus, but he did not describe nuclei in the plasmodial stage itself. Schultze, however, in the same year described the plasmodia of Aethalium septicum as consisting of 'naked lumps of protoplasm, naturally with the nuclei appertaining to them' (i860, p. 301). It is strange that Schultze should have made this discovery, for he undertook no very detailed study of Mycetozoa, and stranger still that some years later Cienkowski (1863, p. 436), as has been mentioned, and de Bary (1864, p. 108) were still of the opinion that the plasmodium lacks nuclei. Meanwhile events had been leading towards the discovery of the syncytial nature of certain Heliozoa. Many years before, Kolliker (1849) had seen the nuclei of Actinosphaerium eichhornii, and described them as 'kern- und zellenartige' bodies (p. 211); but he did not regard them as nuclei. Stein, too, saw them, and remarked that they had 'das Ansehen von Zellenkernen' (1854, pp. 153-4); DUt n e does not pronounce upon the matter. Haeckel thought the objects were probably nuclei, and suggested that bodies within them might be nucleoli (1862, p. 165). Schultze calls them 'zellenartige Korperchen' in one place and 'Kerne' in another; he makes no definite statement as to what they are (1863, pp. 35-36). It was Wallich who first stated definitely that these bodies in Actinosphaerium are nuclei; he uses the word 'nucleus', and defines it (1863, pp. 444, 450; see also his fig. 2 on plate X). Two years later Cienkowski (1865) saw and figured several nuclei in another Helizoon, Nuclearia delicatula, and recognized their nature. (It should be remarked that the authors mentioned above called Actinosphaerium 'Actinophrys'.) In the Radiolaria, also, the numerous nuclei present in certain species were seen long before they were recognized as such. Muller saw them first, in Acanthometra (1859, P- I5)> n e reported the existence of 'many round transparent vesicles' within the central capsule. They were later seen by Haeckel in the Acanthometrida (1862, pp. 141 and 374, and fig. 2 on plate XV), and in the monopylarian Lithomelissa (p. 302); he did not know they were nuclei, though he suggested that some Radiolaria might be multinucleate (p. 165). The numerous nuclei of certain Acanthometrida and of Tridictyopus (Monopylaria) were for the first time recognized as such by Hertwig (1879, pp. 11, 84), who stained them with carmine. His conclusion was accepted by Haeckel (1887, pp. 32-33). The syncytial nature of certain other rhizopods was gradually disclosed by the labours of many investigators. Among the thallophytes, on the contrary, most of the important discoveries were made in a short time by one man. Till towards the end of the seventies many of the lower plants were universally regarded as non-nucleate (see, e.g., Haeckel, 1874, p. 409; Sachs, 1874, p. 273; Strasburger, 1876, pp. 86-88; Haeckel, 1878, p. 53). Sporadic discoveries of syncytia had indeed been made. Pringsheim, for instance, as we have seen (p. 178), had described this condition in the Saprolegniaceae, and de Bary (1862, p. 14) had seen stages in the development of the ascus of Peziza, with two, four, and eight nuclei not separated by cell-walls. The existence of whole groups of syncytial plants was, however, unsuspected. The plants were known, but their nuclei were not; for though they had been seen in some cases (e.g. in Cladophora), they were not regarded as nuclei, simply because there were many of them in each 'cell' (Strasburger, 1876, pp. 86-88 and 324). They were first recognized as such by Schmitz, who revolutionized knowledge of the lower plants by his discoveries. The latter were all published in the journals of local natural history societies. Schmitz used simple methods, staining the nuclei sometimes with alcoholic iodine solution, sometimes with a mixture of haematoxylin, alum, and glycerine after fixation in alcohol or osmium tetroxide solution. He began by studying seaweeds in the Gulf of Athens in 1878, and at Naples in the following year. He reported his first results verbally on 30 November 1878, but there was delay in the printing of this particular paper (Schmitz, 1880a). Meanwhile he had published others. He first announced his discovery of nuclei in Valonia and related forms, and showed that each of the apparent 'cells' is a syncytium; he founded the group Siphonocladiaceae for these plants, which had up till then been variously classified (1879a, 1880a). He noted the formation of uninucleate zoospores. Turning next to the Siphonales, he showed the existence of numerous nuclei in the continuous cytoplasm of these nonseptate forms (18796). In the same paper he showed the syncytial nature of several Phycomycetes and of the internodes of Chara. He also noticed syncytia in parenchyma cells of certain higher plants. He continued his work and showed that among the Rhodophyceae, the species differ in their nuclear arrangements; in one species the 'cell' may be multinucleate, in another closely related form it may be uninucleate (18806). The wholly syncytial plants and animals show that the cell, as a morphological unit, is not a necessary component of organisms. It is possible, however, to exaggerate the importance of this fact. No organism reaches a high degree of complexity without adopting cellular structure. Some of the Siphonales show a limited degree of resemblance to higher plants in external form, and there is thus a suggestion of a much greater degree of differentiation of parts than in fact exists. Pressed specimens, too, tend to look more like higher plants than do these lowly organisms in their natural form. It is doubtful whether Caulerpa and its allies are in fact the most highly differentiated syncytial organisms. Some of the Ciliophora appear more complex. This most aberrant group, however, is excluded from consideration in the present paper, as it will be discussed under Proposition VI. Reasons for regarding the ciliates and their allies as syncytia will be given there. (See also Baker, 1948, b and c.) At a meeting of a scientific society in Wurzburg on 23 November 1878 Sachs demonstrated a series of Siphonales and remarked that these, as well as the Mucorineae, had up till then been regarded as 'einzellige'; he considered that they should rather be called 'nicht cellulare' (Sachs, 1879). It is to be noted that this demonstration took place seven days before Schmitz began to make known his discoveries on syncytial plants. Like everyone else, Sachs considered that the plants he called non-cellular were devoid of nuclei. When their nuclei were discovered, the name stuck to them; and indeed it is not inappropriate. It is unfortunate that the name was also applied by some writers to certain uninucleate protists, which evidently correspond in their structure to single cells. This matter will be discussed with Proposition VI; I have already commented upon it elsewhere (Baker, 1948, b and c). Sachs reverted to the subject of non-cellular plants many years later in two important papers (1892 and 1895). He considered that the accepted terminology of cytology was misleading and should be changed. For him, the cell was the cell-wall, or sometimes the cell-wall with the contents of the cell (1892, p. 62; see also p. 166 of the present paper). He felt that a new word was required. 'Under the name of an Energid', he wrote (1892, p. 57), 'I think to myself of a single cell-nucleus with the protoplasm controlled (beherrschten) by it.' He chose the word energid to indicate that the vital activities reside in the nucleus and cytoplasm; he did not intend 'energy' to be understood here in its physical sense (1895, p. 410). He regarded the energid as a morphological as well as a physiological unit. It might produce a cell-wall or other secreted objects, or it might not. In most cases each cell is inhabited by one energid, but the Siphonales were obviously peculiar in this respect. Sachs had changed his mind; he now called them one-celled plants, but remarked that the cell was produced by numerous energids. The difference between such forms as the Siphonales on the one hand and cellular plants on the other, in Sachs's terminology, was that the neighbouring energids of the former were not sharply marked off from one another (1892, p. 62; 1895, p. 435). Sachs's new word was never widely accepted. Biologists might perhaps do well to reconsider their tacit rejection of it. His expression 'beherrschten' has gained rather than lost in significance since his time. A difficulty is that in syncytia we have generally no means of delimiting the nuclear zones of control, which must be constantly shifting in cases where the cytoplasm is in motion. Again, it is difficult to be certain that a particular part of the cytoplasm is 'controlled' by only one nucleus. Occasionally, however, the zones of control announce very clearly that they exist. A good example is provided by the syncytial zoospore of Vaucheria, in which two flagella are related to each nucleus. Boveri (1905) called the nucleus of the spermatozoon or egg a Hetnikaryon; the fusion-nucleus of a zygote an Amphikaryon; and a nucleus in which the number of chromosomes had doubled without nuclear division a Diplokaryon. Strasburger was using the words diploid and haploid in 1907 (pp. 490, 529) and tetraploid, oktoploid, and polyploid in 1910 (pp. 422, 444), but it seems doubtful whether we have any authoritative statement of their meanings. A quotation from a well-known and excellent textbook will exemplify the doubt. Its author writes: 'The lowest diploid number found in any organism is 2, which occurs in the Roundworm, Ascaris megalocephala var. univalens (this species also has a tetraploid variety, bivalens with 4 chromosomes in the diploid set).' Thus in one sentence we are told that four is both the diploid and the tetraploid number of chromosomes in the variety bivalens. Some authors use the expression diploid as synonymous with the somatic number in sexually produced organisms, and haploid as synonymous with the gametic number, whether or not there happen to be only two sets of different chromosomes in the cells called diploid and one in those called haploid. I shall not follow this usage, but shall employ the word haploid to refer to a single set of different chromosomes and diploid to refer to two such sets, and shall use triploid, polyploid, &c, in conformity with this system. (Thus a gamete-nucleus is characteristically haploid, but may be diploid.) The haploid protoplast, or haplocyte as I shall call it, is a better example of a unit than a diplocyte, with its two sets, just as a box containing the playing-cards of a single suit is more perfectly unitary than one containing two suits or the tetraploid pack; but since the great majority of organisms arise, directly or indirectly, from the fusion of two haplocytes, the most usual morphological unit in both plants and animals is the diplocyte. The degree of duplicity exhibited by this cellular unit is an expression of one of the most fundamental facts of biology. The Uebereinstimmung postulated by Schwann was to some extent upset by the discovery that there are typically both haplocytes and diplocytes in organisms; but much more serious from the point of view of the cell as a morphological unit is the fact that polyploid nuclei also exist. The history of this discovery must now be briefly traced. Guignard (1884, p. 27) suggested that certain plant-cells may contain about twice the usual number of 'batonnets chromatiques', but he was dealing with a normally haploid structure, and the discovery of polyploidy must be ascribed to Boveri (1887). The latter showed that there are two varieties of Ascaris megalocephala, which he called Typus Carnoy and Typus van Beneden, after earlier students of the chromosomes of this nematode. He showed that in Typus Carnoy there are two chromatic elements (chromosomes) in the ripe egg, in Typus van Beneden only one. Thus the somatic cells of the former variety were tetraploid. This case is not quite so simple as it appeared to be, for, as is well known, the chromosomes break into fragments in cells other than those of the germ-track; and when they have done so, the number in Typus van Beneden appears not to be exactly half that in Typus Carnoy (Walton, 1924). It seems allowable, however, in the present state of knowledge, to regard Typus Carnoy as tetraploid before fragmentation. Boveri later (1903, a and b, 1905) shook the eggs of the sea-urchin Strongylocentrotus immediately after fertilization and by this means suppressed the first cleavage, while the chromosomes divided; he thus obtained tetraploid larvae experimentally. A discovery in plants similar to Boveri's in Ascaris was made by Rosenberg (1903), who showed that Drosera rotundifolia has 20 chromosomes in its somatic cells, while D. longifolia has 40. Strasburger (1905) counted the number of chromosomes in the pollen mother-cells of Alchemilla arvensis and A. speciosa, and found that the latter had twice as many as the former. These were the first indications of what is now known to be the very widespread occurrence of polyploidy among higher plants. A complication is introduced by certain species which seem to have become secondarily diploid, with double the usual diploid number of chromosomes, by differentiation of the four sets into two. Difficulties in the sex-determining mechanism prevent most dioecious animals from doubling their chromosome numbers throughout their tissues, but there are indications that certain hermaphrodites are polyploid (see White, 1940). There is no particular barrier against chromosome replication in somatic cells, and it is not unusual for some of these to become polyploid. The classical example of mosaic polyploidy is provided by the honey-bee. Petrunkevitch found long ago (1901, pp. 587-8) that there are only 16 chromosomes in the first division of the nucleus of the drone-egg, but about 64 in cells of the blastoderm of the later embryo. In a celebrated paper Meves showed that while the diploid number (counted in the oogonia of the queen) is 32 and the haploid 16, more than 60 chromosomes are present in the follicle-cells of the testis (1907, pp. 471-2). Nowadays we have simpler methods of detecting polyploidy, in particular cases, than laborious chromosome-counts. We may measure nuclear volumes (Jacobj, 1925), or count either the nucleoli (especially where there is one per haploid set in early prophase (de Moll, 1923, 1928)), or the heterochromatic X-chromosomes (Geitler, 1937), or heterochromatic satellites (Berger, 1941). The most extreme instance of mosaic polyploidy seems to be provided by the pond-skater, Gerris lateralis, in which the degree of replication reaches 1024-ploidy, or even farther, in the salivary glands (Geitter, 1938). There is evident similarity between multinucleate conditions and polyploidy. There is general correspondence between a single mass of cytoplasm containing two diploid nuclei and another containing one tetrapoid nucleus: both may be called tetraplocytes. Particular tissues tend to provide examples of both the binucleate and the uninucleate tetraploid states. Thus in the roots of Pisum sativum treated in life with chloral hydrate, Strasburger (1907, pp. 484-5) found both mitoses with tetraploid chromosome-numbers and binucleate cells. In the tapetal layer of the anthers of certain plants, some cells show polyploid chromosome-numbers at division, while others are bi- or multinucleate (see especially Witkus's study of Spinacia (1945)). In certain parts of young seedlings of Allium cepa, again, a number of tetraploid cells are formed in certain regions; in the same sites binucleate cells are common (Berger and Witkus, 1946). In the mammalian liver there are polyploid cells of various degrees of chromosome-replication; there are also bi- and multinucleate cells (Wilson and Leduc, 1948). Fell and Hughes (1949, p. 366) have shown by the study of living tissue-culture cells of the mouse that polyploid nuclei may arise by mitosis of binucleate cells, a single spindle being formed for the chromosomes of the two nuclei; fusion of diploid nuclei and endomitosis are other mechanisms by which the same end is achieved (Berger, 1937; Geitler, 1939; Wilson and Leduc, 1948). The 'polyenergid' nuclei of certain Protozoa will be mentioned later under the heading of Proposition VI. I have already discussed them elsewhere (Baker, 19486). The existence of polyploidy undoubtedly constitutes an exception to the general rule of the 'Uebereinstimmung' of all cells. One polyploid cell cannot be regarded as homologous with one diploid cell. Boveri (1903a), for instance, found that the protoplasts of tetraploid larvae of Strongylocentrotus were much larger and fewer than those of the diploid form; in the case of the mesenchyme he found that there was about half the normal number. Thus one tetraploid corresponds with two diploid protoplasts. Polyploidy, however, is clearly a secondary condition. Diplocytes and haplocytes are the characteristic primitive morphological units of plants and animals, and are still retained as the elementary components of most organisms. Those organisms that show mosaic polyploidy have haploid germ-cells and are everywhere diploid in the early embryonic stage of the sexually produced form. THE INDIVISIBILITY OF CELLS INTO SMALLER HOMOLOGOUS UNITS Hirsch remarks (1942) that the body of an organism consists of a series of 'partial systems', each of which (till protons and electrons are reached) is built of partial systems of a lower order. Thus the body is made up of organs which are divisible into tissue-units and these into cells; the latter contain Mikronen (mitochondria, Golgi-bodies (lipochondria), various granules and vacuoles, muscle-fibrils, chromosomes, nucleoli), and these are composed of sub-microscopic Submikronen, themselves made up of large molecules; and so on. If an object is composed of parts, all of which are divisible into smaller parts that show what Schwann called 'Uebereinstimmung' with one another, then these smaller parts are clearly the true units of construction. It is important to stress the fact that the Mikronen are not homologous parts: a muscle-fibril, for instance, does not correspond, in any predicable way, to a nucleolus. Further, the Mikronen taken together do not constitute the cell, which consists largely of ground-cytoplasm and nuclear sap. There is no intention here to criticize Hirsch's analysis adversely, but only to point out that it in no way invalidates the cell-theory. The cell is not composed of any lesser homologous units, other than those minute particles that compose all matter, and to these the idea of homology does not properly apply. As Hanstein remarked long ago: 'In the last resort the protoplast, not the molecule or the micelle, is the organic individual' (1880, p. 295). With the reservations that have already been noted, this is true. Adequate critiques have already been given in this paper of the various facets of the cell-theory that are included under the head of the second part of the second Proposition. It remains to make one general comment. Whenever a student wants to 'understand' a complex histological object that is unfamiliar to him (a Pacinian corpuscle will serve as an example) or a researchworker to grasp the minute structure of a previously unknown organ, he proceeds first of all to try to determine where the boundaries of the protoplasts arewhat is cellular and what is intercellular. In other words, he tries to interpret what he sees in terms of the part of the cell-theory that is summarized in Proposition II, in the formulation here adopted. That is the measure of the homage he pays (often unwittingly) to the founders of the cell-theory.

This is a preview of a remote PDF:

JOHN R. BAKER. The Cell-Theory: A Restatement, History, and Critique: Part III. The Cell as a Morphological Unit, Journal of Cell Science, 1952, 157-190,