The Cell in Development and Inheritance

The entire structure, resulting from the foregoing changes, is known as the karyokiiictic or mitotic figure. It may be described as consisting of two distinct parts; namely, i, the chromatic figure, formed by the deeply staining chromosomes ; and, 2, the achromatic figure, consisting of the spindle and asters which, in general, stain but slightly. The fibrous substance of the achromatic figure is gener- G. Metaphase; splitting of the chromosomes {e.p.^. ». The cast-off nucleolus. //. Ana- phase ; the daughter-chromosomes diverging, between them the interzonal-fibres (/./.). or central spindle; centrosomes already doubled in anticipation of the ensuing -division. /. Late anaphase or telophase, showing division of the cell-body, mid-body at the equator of the spindle and bcgm- ning reconstruction of the daughter-nuclei. J. Division completed.

ally known as archoplasm (Boveri, '88), but this term is not applied to the centrosome within the aster. 2. Metaphase. — TYiO, prophases of mitosis are, on the whole, pre- paratory in character. The metaphase, which follows, forms the initial phase of actual division. Each chromosome splits lengthwise into two exactly similar halves, which afterward diverge to opposite poles of the spindle, and here each group of daughter-chromosomes finally gives rise to a daiii^^hter-nucleus (Fig. 26). In some cases the splitting of the chromosomes cannot be seen until they have grouped themselves in the equatt)rial plane of the spindle ; and it is only in this case that the term " metaphase " can be applied to the mitotic figure as a whole. In a large number of cases, however, the splitting may take place at an earlier period in the spireme-stage, or even, in a few cases, in the reticulum of the mother-nucleus (Figs. 54. 55). Such variations do not, however, affect the essential fact that the clnvDiatic nctivork is converted into a thread^ which, ivhether continuous or discontinuous, splits tJiroughojit its entire lengtJi into two exactlj equivalent halves. The splitting of the chromosomes, discovered by Flemming in 1880, is the most significant and funda- mental operation' of cell-division; for by it, as Roux first pointed out {^^}i\ the entire substance of the chromatic network is precisely halved, and the daughter-nuclei receive precisely equivalent portions oj chro- matin from the mother-nucleus. It is very important to observe that the nuclear division always shows this exact quality, whether division of the cell-body be equal or unequal. The minute polar body, for example (p. 238), receives exactly the same amount of chromatin as the Q^^g, though the latter is of gigantic size as compared with the former. On the other hand, the size of the asters varies with that of the daughter-cells (Figs. 58, 175), though not in strict ratio. The fact is one of great significance for the general theory of mitosis, as will appear beyond.

3. Anaphases. — After splitting of the chromosomes, the daughter- chromosomes, arranged in two corresponding groups,^ diverge to oppo- site poles of the spindle, where they become closely crowded in a mass near the centre of the aster. As they diverge, the two groups of daughter-chromosomes are connected by a bundle of achromatic fibres, stretching across the interval between them, and known as the interzonal fibres or connecting fibres? In some cases these differ in a marked degree from the other spindle-fibres ; and they are believed by many observers to have an entirely different origin and function. A view now widely held is that of Hermann, who regards these fibres as belonging to a central spindle, surrounded by a peripheral layer of mantle-fibres to which the chromosomes are attached, and only exposed to view as the chromosomes separate.^ Almost invariably in the division of plant-cells and often in that of animal cells these

1 It was this fact that led Flemming to employ the word mitosis (fiiros, a thread). 2 This stage is termed by Flemming the dyasier, a term which should, however, be aban- doned in order to avoid confusion with the earlier word amphiafter. The latter convenient and appropriate term clearly has priority. 3 Verlnndutigsfasern of German authors ; filaments reunissants of Van Beneden. fibres show during this period a series of deeply staining thickenings in the equatorial plane forming the cell-plate or mid-body. In plant- mitoses this is a very conspicuous structure ( Fig. 34). In animal cells the mid-body is usually less developed and sometimes rudimentary, being represented by only a few granules or even a single one (Fig. 29). Its later history is described below.

4. Telophases. — In the final phases of mitosis, the entire cell divides in two in a plane passing through the equator of the spindle, each of the daughter-cells receiving a group of chromosomes, half of the spindle, and one of the asters with its centrosome. Meanwhile, a daughter-nucleus is reconstructed in each cell from the group of chromosomes it contains. The nature of this process- differs greatly in different kinds of cells. Sometimes, as in the epithelial cells of Amphibia, especially studied by Flemming and Rabl, and in many plant-cells, the daughter-chromosomes become thickened, contorted, and closely crowded to form a daugJiter-sph'enie , closely similar to that of the mother-nucleus (Fig. 29); this becomes surrounded by a mem- brane, the threads give forth branches, and -thus produce a reticular nucleus. A somewhat similar set of changes takes place in the seg- menting eggs of Ascaris (Van Beneden, Boveri). In other cases, as in many segmenting ova, each chromosome gives rise to a hollow vesicle, after which the vesicles fuse together to produce a single nucleus (Fig. 52). When first formed, the daughter-nuclei are of equal size. If, however, division of the cell-body has been unequal, the nuclei become, in the end, correspondingly unequal — a fact which, as Conklin and others have pointed out, proves that the size of the nucleus is controlled by that of the cytoplasmic mass in which it lies.

The fate of the achromatic structures varies considerably, and has been accurately determined in only a few cases. As a rule, the spindle-fibres disappear more or less completely, but a portion of their substance sometimes persists in a modified form i^e.g. the Nebenkern, p. 163). In dividing plant-cells, the cell-plate finally extends across the entire cell and splits into two layers, between which appears the membrane by which the daughter-cells are cut apart.i A nearly similar process occurs in a few animal cells,- but as a rule the cell-plate is here greatly reduced and forms no mem- brane, the cell dividing by constriction through the equatorial plane. Even in this case, however, the division-plane is often indicated before division takes place by a peculiar modification of the cyto- plasm in the equatorial plane outside the spindle (Fig. 30)- This region is sometimes called the cytoplasmic plate, in contradistinction to the spindle-plate, or mid-body proper. In the proi)hases and meta- 1 Cf. Strasburger, '98. 2 Cf. Hoffmann, '98.

phases the astral rays often cross one another in the equatorial region outside the spindle. During the anaphases, however, this crossing disappears, the rays from the two asters now meeting at an angle along the cytoplasmic plate (Fig. 31). Constriction and division of the cell then occur.^ The aster may in some cases entirely disappear, together with the centrosome (as occurs in the mature dgg). In a large number of cases, however, the centrosome persists, lying either outside or more rarely inside the nucleus and dividing into two at a very early period. This'division is clearly a precocious preparation for the ensuing divi- sion of the daughter-cell, and it is a remarkable fact that it occurs as a rule during the early anaphase, before the mother-cell itself has divided. There are apparently, however, some cases in which the centrosome remains undivided during the resting stage and only divides as the process of mitosis begins.

Like the centrosome, the aster or its central portion may persist in a more or less modified form throughout the resting state of the cell, forming a structure generally known as the attraction-sphere. This body often shows a true astral structure with radiating fibres (Figs. 8, 49); but it is sometimes reduced to a regular spherical mass which may represent only a portion of the original aster (Fig. 7). The nature and source of the material from which the mitotic figure arises form a problem that has been almost continuously under discussion since the first discovery of mitosis, and is even now but partially solved. The discussion relates, however, almost solely to the achromatic figure (centrosome, spindle, and asters) ; for every one is agreed that the chromatic figure (chromosomes) is directly derived from the chromatin-network, as described above, so that there is no breach in the continuity of the chromatin from one cell-generation to another. With the achromatic figure the case is widely different. The material of the spindle and asters must be derived from the nucleus, from the cytoplasm, or from both ; and most of the earlier research was devoted to an endeavour to decide between these possibilities. The earhest observers {'71-7S) supposed the achro- matic figure to disappear entirely at the close of cell-division, and most of them (Butschli, Strasburger, Van Beneden, '75) believed it to be re-formed at each succeeding division out of the nuclear substance. The entire mitotic figure was thus conceived as a metamorphosed nucleus. Later researches ('75- 85) gave contradic-

tory and apparently irreconcilable results. Fol ('79) derived the spindle from the nuclear material, the asters from the cytoplasm. Strasburger ('80) asserted that the entire achromatic figure arose Fig. 27. — The prophases of mitosis (heterotypical form) in primary spermatocytes of Salama)idra. [Meves.] A. Early segmented spireme ; two centrosomes outside the nucleus in the remains of the attraction-sphere. B. Longitudinal splitting of the spireme, appearance of the astral rays, disin- tegration of the sphere. C. Early amphiastcr and central spindle. D. Chromosomes in the form of rings, nuclear membrane disappeared, amphiaster enlarging, mantle-fibres developing.

from the cytoplasm, and to that view, in a modified form, he still adheres. Flemming ('82), on the whole, inclined to the opinion that the achromatic figure arose inside the nucleus, yet expressed the opinion that the question of nuclear or cytoplasmic origin was one of minor importance. A long series of later researches on both plants and animals has fully sustained this opinion, showing that the origin of the achromatic figure does in fact differ in different cases. Thus in Infusoria the entire mitotic figure is of intranuclear origin (there are, however, no asters); in echinoderm eggs the spindle is of nuclear, the asters of cytoplasmic, origin ; in the testis-cells and some tissue- cells of the salamander, a complete amphiaster is first formed in the cytoplasm, but to this are afterward added elements probably derived from the linin-network; while in higher plants there is some reason to believe that the entire achromatic figure may be of cytoplasmic origin. Such differences need not surprise us when we reflect that the achromatic part of the nucleus (linin-network, etc.) is probably of the same general nature as the cytoplasm. ^

Many observers have maintained that the material of the astral rays and spindle-fibres is directly derived from the substance of the protoplasmic meshwork, whether nuclear, cytoplasmic, or both ; but its precise origin has long been a subject of debate. This question, critically considered in Chapter VI., will be here only briefly sketched. By Klein {;7%\ Van Beneden ('83), Carnoy ('84, '85), and a large num- ber of later observers, the achromatic fibres, both of spindles and of asters, are regarded as identical with those of a preexisting reticulum which have merely assumed a radiating arrangement about the cen- trosome. The amphiaster has, therefore, no independent existence, but is merely an image, as it were, somewhat like the bipolar figure arising when iron filings are strewn in the field of a horseshoe magnet. Boveri, on the other hand, who has a small but increasing following, maintains that the amphiastral fibres are not identical with those of the preexisting meshwork, but a new formation which, as it were, "crystallizes anew " out of the general protoplasmic substance. The amphiaster is therefore a new and independent structure, arising in, or indirectly from, the preexisting material, but not by a direct mor- phological transformation of that material. This view, which has been advocated by Druner ('94), Braus ('95), Meves ('97, 4, '98), and with which my own later observations ('99) also agree, is more fully discussed at page 318.

In 1887 an important forward step was taken through the inde- pendent discovery by Van Beneden and Boveri that in the ^gg of Ascaris the centrosome does not disappear at the close of mitosis, but remains as a distinct cell-organ lying beside the nucleus in the cyto- 1 In the case of echinoderm eggs, I have found reason ('95, 2) for the conclusion that the spindle- fibres are derived not merely from the linin-substance, but also from the chromatm. Despite some adverse criticism, I have found no reason to change my opinion on this point. The possible significance of such a derivation is indicated elsewhere (p. 302).

plasm. These investigators agreed that the amphiaster is formed under the influence of the centrosome, which by its division creates two new "centres of attraction" about which the astral systems arise, and which form the foci of the entire dividing system. In them are centred the fibrillae of the astral system, toward them the daughter- Fig. 28. - Metaphase and anaphases of mitosis in cells (spermatocytes) of the salamander. [Druner.] E Metaphase. Tlie continuous central spindle-fibres pass from pole to pole of the spmdle Outside them the thm layer of contractile mantle-fibres attached to the d.vKlcd ^^^'-:^'^^''^^'^} which only two are shown! Centrosomes and asters. F. Transverse section ^ -- ^ \ ^^ "^ !?^ J figure showing the ring of chromosomes surrounding the central spmdle. the cut ^^'^^/^^ ' rL'''!"^/ a;peanng as ''dots. G. Anaphase; divergence of the ^»-^g'^^--f-°"--"\^.^- ^j^J^^^ "S„f^,^^^ tral spin5leas the interzonal fibres; contractile fibres ^vn^^^'^-\^or..^-' \ '^Xj^^^^^^^ shown. H. Later anaphase (dyaster of Flemming) ; the central spmdle uly exposed to Mev. mantle-fibres attached to the chromosomes. Immediately aftenvard the cell d.vdes (see P .g. ^) .

chromosomes proceed, and within their respective spheres of influ- ence are formed the resulting daughter-cells. Both Van Beneden and Boveri fully recognized the importance of then" discovery. "We are justified," said Van Beneden, ^ in regarding the attraction-sphere with its central corpuscle as forming a permanent organ, not only of the early blastomeres, but of all cells, and as constituting a cell-organ equal in rank to the nucleus itself ; and we may conclude that every central corpuscle is derived from a preexisting corpuscle, every attraction- sphere from a preexisting sphere, and that division of the sphere precedes that of the cell-nucleus." ^ Boveri expressed himself in similar terms regarding the centrosome in the same year {"^y, 2, p. 153), and the same general result was reached by Vejdovsky nearlv at the same time,- though it was less clearly formulated than bv either Boveri or Van Beneden.

All these observers agreed, therefore, that the achromatic figure arose outside the nucleus, in the cytoplasm ; that the primary impulse to cell-division was given, not by the nucleus, but by the centrosome, and that a new cell-organ had been discovered whose special office Fig. 29. — Final phases (telephases) of mitosis in salamander cells. [Flemming.] /. Epithelial cell from the lung; chromosomes at the poles of the spindle, the cell-body divid- ing; granules of the "mid-body" or '/Auischcnkorpcr ■M the equator of the disappearing spindle. y. Connective tissue-ceil (lung) immediately after division; daughter-nuclei reforming, the cen- trosome just outside of each ; mid-body a single granule in the middle of the remains of the spindle. •

was to preside over cell-division. **The centrosome is an indepen- dent permanent cell-organ, which, exactly like the chromatic elements, is transmitted by division to the daughter-cells. 77ic cciitrosojue rep- resents the dynamic cejitre of cell T "^ That the centrosome does in many cases, especially in embryonic cells, behave in the manner stated by Van Beneden and Boveri seems at present to admit of no doubt ; and it has been shown to occur in

many kinds of adult tissue-cells during their resting state ; for example in pigment-cells, leucocytes, connective tissue-cells, epithelial and endothelial cells, in certain gland-cells and nerve-cells, in the cells of some plant-tissues, and in some of the unicellular plants and ani- mals, such as the diatoms and flagellates and rhizopods. On the other hand. Van Beneden's conception of the attraction-sphere has proved untenable ; for this structure has been clearly shown in some cases to disintegrate and disappear at the close or the beginning of mitosis^

Whether the centrosome theory can be maintained is still in doubt ; In the first place, it has been shown that the primary impulse to cell-division cannot be given by fission of the centrosome, for there are several accurately determined cases in which the chromatin-elements divide independently of the centrosome, and it is now generally agreed that the division of chromatin and centrosome are two parallel events, the nexus between which still remains undetermined.^

Secondly, an increasing number of observers assert the total disap- pearance of the centrosome at the close of mitosis ; while some very convincing observations have been made favouring the view that cen- trosomes may be formed de novo without connection with preexisting ones (pp. 213, 305). Thirdly, a large number of recent observers (including Strasburger and many of his pupils) of mitosis in the flowering plants and pteridophytes agree that in these forms no centrosome exists at any stage of mitosis, the centre of the aster being occupied by a vague reticular mass, and the entire achromatic figure arising by the gradual grouping of fibrous cytoplasmic elements (kinoplasm or filar plasm) about the nuclear elements.^ If we can assume the cor- rectness of these observations, the centrosome-theory must be greatly modified, and the origin of the amphiaster becomes a far more com- plex problem than it appeared under the hypothesis of Van Ik-neden and Boveri. That such is indeed the case is indicated by nothing more strongly than by Boveri's own remarkable recent experiments on cell-division (referred to at page 108).

Comparative study has shown that almost every detail of the pro- cesses described above is subject to variation in different forms of cells. Before considering some of these modifications it may be well to pomt out what we are at present justified in regarding as its essential features. These are : (i) The formation of the chromatic and achro- matic figures; (2) the longitudinal splitting of the chromosomes or spireme-thread ; (3) the transportal of the chromatin-halves to the respective daughter-cells. Each of these three events is endlessly varied in detail ; yet the essential phenomena are everywhere the same, with one important exception relating to the division of the chromo- somes that occurs in the maturation of certain eggs and spermatozoa.^ It maybe stated further that the study of mitosis in some of the lower forms (Protozoa) gives reason to believe that the asters are of second- ary importance as compared with the spindle, and that the formation of spireme and chromosomes is but tributary to the division of the smaller chromatin-masses of which they are made up.

(a) TJic Achromatic Figure. The phenomena involved in the his- tory of the achromatic figure are in general most clearly displayed in embryonic or rapidly dividing cells, especially in egg-cells (Figs. 31, 60), where the asters attain an enormous development, and the centrosomes are especially distinct. In adult tissue-cells the asters are relatively small and difficult of demonstration, the spindle large and distinct ; and this is particularly striking in the cells of higher plants where the asters are but imperfectly developed. Plant-mitoses are characterized by the prominence of the cell-plate (Fig. 34), which is rudimentary or often wanting in animals, a fact correlated no doubt with the greater development of the cell-membrane in plants. With this again is correlated the fact that division of the cell-body in animal cells generally takes place by constriction in the equatorial plane of the spindle ; while in plant-cells the cell is usually cut in two by a cell-wall developed in the substance of the protoplasm and derived in large part from the cell-plate.

In animal cells we may distinguish two general types in the forma- tion of the amphiaster, which are, however, connected by interme- diate gradations. In the first of these, typically illustrated by the division of epithelial and testis-cells in the salamander (Flemming, Hermann, Drijner, Meves), a complete amphiaster is first formed in the cytoplasm outside the nucleus, while the nuclear membrane is still intact. As the latter fades away and the chromosomes appear, some of the astral rays grow into the nuclear space and become attached to the chromosomes, which finally arrange themselves in a ring about the original spindle (Figs. 27, 28). In the completed amphiaster, therefore, we may distinguish the original central spi7idle (Hermann, '91) from the surrounding mantle-fibres, the latter being

attached to the chromosomes, and being, according to Hermann, the principal agents by which the daughter-chromosomes are dragged apart. The mantle-fibres thus form two hollow cones or half-spin- dles, separated at their bases by the chromosomes and completely surrounding the continuous fibres of the central spindle, which come into view as the ''interzonal fibres" during the anaphases (Fig. 28). There is still considerable uncertainty regarding the origin and relation of these two sets of fibres. It is now generally agreed with Van Beneden that the mantle-fibres are essentially a part of the asters, i.e. are simply those astral rays that come into connection with the chromosomes — wholly cytoplasmic in ori- gin (Herma.nn, Driiner, MacFarland), or in part cytoplasmic, in part dif- ferentiated from the linin- network (Flemming, Meves). Driiner ('95), Braus ('95) (salamander), and MacFarland (yPleicro- phyllidia, '97) believe the central spindle to arise secondarily through the union of two opposing groups of astral rays in the area between the centrosomes. On the other hand, Hermann ('91), Flemming ('91), Heidenhain ('94), Kos- tanecki ('97), Van der Stricht ('98), and others believe the central spindle to exist from the first in the form of fibres stretching between the diverging centrosomes ; and Heidenhain believes them to be developed from a special substance, forming a ''primary centrodesmus," which persists in the resting cell, and in which the centrosomes are embedded.^ MacFarland's observa- tions on gasteropod-eggs ('97) indicate that even nearly related torms may differ in the origin of the central spindle, since in Plcurophyllidia it is of secondary origin, as described above, while in Diaiilula it is a primary structure developed from what he describes as the " centro- some," but which, as shown at page 314, is probably to be regarded as

Earlier stage above, showing thickenings along the line of cleavage. Later stage, below, showing spindle-plate and an attraction-sphere surrounding; the centrosomes, and is perhaps comparable to Heidenhain's " centrodesmus." In the second type, ilhistrated in the cleavaf]^e of echinoderm, annehd, niolhiscan, and some other egjrs, a central s])indle may be formed, — sometimes already during the anaphases of the preceding mitosis (Figs. 99, 155), — but afterward disappears, the asters moving

Fig. 31. — The middle phases ot mitosis in the first cleavage of the Ascaris-^gg. [BOVERI.] y-f. Closing prophase, the equatorial plate forming, B. Metaphase; equatorial plate estab- lished and the chromosomes split; b. the equatorial plate, viewed en face, showing the four chro- mosomes, C. Early anaphase; divergence of the daughter-chromosomes (polar body at one side), D. Later anaphase; p.b. second polar body, (For preceding stages see Fig. 90; for later stages Fig. 145.)

to opposite poles of the nucleus. Between these two poles a new spindle is then formed in the nuclear area, while astral rays grow out into the cytoplasm. There is strong evidence that in this case the entire spindle may arise inside the nucleus, i.e. from the sub- stance of the linin-network, as occurs, for example, in the eggs of echinoderms (Fig, 25, E), and in the testis-cells of arthropods. In other cases, however, a part at least of the spindle is of cytoplasmic

origin, since the ends of the spindle begin to form before dissokition of the nuclear membrane, and the latter is pushed inwards in folds by the ingrowing fibres (Figs. 25, C, 99).! In some cases, however, it seems certain that the nuclear membrane fades away before com- pletion of the spindle (first maturation-division of TJialasscma, CJice^- toptcriis), and it is probable that the middle region of the spindle is here formed from the Hnin-network. In most, if not all, mitoses of the second type the chromosomes do not form a ring about the equator of the spindle, but extend in a flat plate completely through

A. Early prophase witli single aster and centrosonie. B. Initial formation of intraniK car spindle. C. Divergence of the daughter-centrosomes. D. Early anaphase ; nuclear nieniiji..ne still intact. its substance. Here, therefore, it is impossible to speak of a " cen- tral spindle." It is nevertheless probable that the spindle-fibres are of two kinds, viz. continuous fibres, which form the interzonal fibres seen during the anaphases, and half-spindle fibres, extending only from the poles to the chromosomes. It is possible that these two kinds of fibres, though having the same origin, respectively corre-

1 Cf. Platner ('86) on Avion and Lepidoptera, Watase ('91) on Loligo, Braus ('95) on Triton, and Griffin ('96, '99) on Thalassema. Erlanger ('97, 5) endeavours to show that in the mitosis of embryonic cells in the cephalopods (^Scpia'), where the inpushing of the mem- brane was previously shown by Watase, the entire spindle arises from the nucleus. spond in function to those of the central spindle and to the mantle- fibres. It seems probable that the difference between the two types of spindle-formation may be due to, or is correlated with, the fact that the nuclear transformation takes place relatively earlier in the "first type. When the nucleus lags behind the spindle-formation the centrosomes take up their position prematurely, as it w^ere, the cen- tral spindle disappearing to make w^ay for the nucleus.

It is in the mitosis of plant-cells that the most remarkable type of achromatic figure has been observed. In some of the lower forms (Alga?) mitosis has been clearly shown to conform nearly to the process observed in animal cells, the amphiaster being provided with very large asters and distinct centrosomes, and its genesis corre- sponding broadly with the second type described above (Figs. 32, 33), though with some interesting modifications of detail.^ Swingle ('97) describes in Stytopocauloii a process closely similar to that seen in many animal cells, the minute but very distinct centrosomes being surrounded by quite typical cytoplasmic asters, passing to opposite poles of the nucleus, and a spindle then developing between them out of the achromatic nuclear substance (Fig. 32). In the flowering plants and pteridophytes, on the other hand, mitosis seems to be of a quite different type, apparently taking place in tJie entire absence of centrosomes. Guignard ('91, i, '92, 2) clearly described and figured typical centrosomes and attraction-spheres both in the ordinary mitosis (Fig. 34) and in the fertilization of the higher plants, giving an account of their behaviour nearly agreeing with the views then prevaiUng among zoologists.