v\ b 0itti)Vitn'si jWonograpfnt on biological j^uhjttt^ General Editor : G. R. de Beer, M.A., D.Sc. THE CHROMOSOMES METHUEN'S MONOGRAPHS ON BIOLOGICAL SUBJECTS F'cap 8uo., 3*. 6d. net each General Editor : G. R. de Beer, M.A., D.Sc. Fellow of Merton College, Oxford SOCIAL BEHAVIOUR IN INSECTS. By A. D. Imms, M.A., D.Sc, F.R.S. MICROBES AND ULTRAMICROBES. By A. D. Gardner, M.A., D.M., F.R.C.S. MENDELISM AND EVOLUTION. By E. B. Ford, M.A., B.Sc. THE BIOCHEMISTRY OF MUSCLE. By D. M. Needham, M.A., Ph.D. (55. net). RESPIRATION IN PLANTS. By W. Stiles, M.A., Sc.D., F.R.S. , and W. Leach, D.Sc, Ph.D. SEX DETERMINATION. By F. A. E. Crew, M.D., D.Sc, Ph.D. THE SENSES OF INSECTS. By H. Eltrinqham, M.A., D Sc F R S PLANT ECOLOGY. By W. Leach, D.Sc, Ph.D. CYTOLOGICAL TECHNIQUE. By J. R. Baker, M.A., D.Phil. MIMICRY AND ITS GENETIC ASPECTS. By G. D. Hale Carpenter, M.B.E., D.M., and E. B. Ford, M.A., B.Sc THE ECOLOGY OF ANIMALS. By Charles Elton, M.A. CELLULAR RESPIRATION. By N. U. Meldrum, M.A., Ph.D. PLANT CHIMAERAS AND GRAFT HYBRIDS. By W. Neilson Jones. INSECT PHYSIOLOGY. By V. B. Wigglesworth, M.A., M.D. TISSUE CULTURE. By E. N. Willmer, M.A. {is. net.). PLANT VIRUSES. By K. M. Smith, M.A. NEMATODES PARASITIC IN ANIMALS. By G. Lapaqe, M.A., M.D. (45. 6rf. net.). THE CHROMOSOMES. By M. J. D. White. In Preparation THE MEASUREMENT OF LINKAGE IN HEREDITY. By K. Mather, B.Sc. MYCORRHIZA. By J. Ramsbottom, O.B.E. Other volumes to follow ^K s^ THE CHROMOSOMES hy M. J. D. WHITE, M.Sc. LECTURER IN ZOOLOGY, UNIVERSITY COLLEGE, LONDON WITH 20 ILLUSTRATIONS METHUEN & CO., LTD. 36 ESSEX STREET W.C. LONDON First published in 1937 PRINTED IN GREAT BRITAIN INTRODUCTION I HOPE that this book will be of use to many bio- logists who realize that chromosome -cytology has made considerable progress in the last ten years, and that the existing text-book apcounts of mitosis and meiosis are hopelessly inaccurate, but who have no time to read the larger works of Darlington and B^lar, which must remain the standard sources of information on the subject. Chromosome cytology is essentially a practical sub- ject, which can only be thoroughly mastered by a study of actual preparations under the microscope. Unfortunately this study is usually regarded as too difficult to be included in a degree course in biology. It is surprising, however, how much can be seen, even without using an oil-immersion objective, provided that one chooses suitable material with large chromo- somes. There is no doubt that for most purposes the testes of Locusts and Grasshoppers (any species will do) provide the best introductory material. They should be fixed in Flemming's solution and stained in one of the aniline dyes like Gentian Violet. In the course of the past year I have made the ordinary degree students in this department work through material of this kind (sectioned at 25 // so as to obtain whole nuclei). They were able to see all the stages of mitosis and meiosis and even to work out the average number of chiasmata per nucleus in three different species. That it is possible for students to do this in a course involving only one afternoon a week should destroy the myth that cytology is a fantastically difficult subject. In a book of this size it is necessarily not possible vi THE CHROMOSOMES to quote ' chapter and verse ' for each statement. I mention this in apology for a certain amount of dogmatism imposed by limitations of space. The illustrations are essentially diagrammatic and designed to illustrate principles rather than to serve as actual illustrations of ceU-division in particular organisms. In any subject it is impossible to avoid technical terms, but I have endeavoured to reduce them to a minimum, and have explained in the text all-. those which are not self-explanatory. Where a term occurs for the first time and is defined it is printed in italics. In the description of meiosis I have adopted the term bivalent instead of tetrad as being far less likely to cause confusion. Barbarous terms such as ' hetero- typic ', ' homotypic ', ' preheterokinesis ' and ' posta- crosyndesis ' have done more to frighten the general biologist away from chromosome cytology than any- thing else, and those authors who introduced them must bear responsibility for the delayed integration of cytology into general biological knowledge. Only by a rigid avoidance of such terms can we hope to link up the study of the nucleus with colloid chemistry on the one hand and with animal and plant breeding on the other. For those who find certain parts of the book diffi- cult to understand, I would recommend the use of some models, which can be constructed in a few minutes out of soft copper wire or plasticine, using the illustrations as a guide. With the aid of these it should be possible for anyone to understand the details of chiasma -formation and meiosis. Department of Zoology, University College, London, April 1937. CONTENTS CHAP. PAGE INTRODUCTION V I THE RESTING NUCLEUS 1 II THE GENERAL OUTLINE OF MITOSIS 8 Prophase. Prometaphase. Metaphase. First Stage of Anaphase. Second Stage of Anaphase. Telophase in SPECIAL PROBLEMS OF MITOSIS 21 Number, Form and Size of Chromosomes. Genetically Active and Genetically Inert Chromosomes. Salivary Gland Chromosomes in Diptera. Sex-Chromosomes. Polyploidy TV THE GENERAL OUTLINE OF MEIOSIS 47 Place of Meiosis in the Life-Cycle. Lepto- tene. Zygotene. Pachytene. Diplotene. Diakinesis. Prometaphase. Metaphase of First Meiotic Division. Anaphase of First Meiotic Division. Telophase of First Meiotic Division. Second Meiotic Division V SPECIAL PROBLEMS OF MEIOSIS 69 Chiasmata and Crossing-Over. Meiosis in Hybrids. Univalent Chromosomes at Meiosis. Meiosis in the Male Drosophila and in other Diptera. Genetically determined Abnormal- ities of Meiosis. Behaviour of Sex-Chromo- somes at Meiosis. Meiosis in Haploid Organisms. Meiosis in Polyploid Organisms. Meiosis in Complex Heterozygotes VI CHROMOSOMES AND EVOLUTION 89 Species -formation in Angiosperms. Origin of complex Heterozygote Organisms. Origin of new species in Bisexual Organisms. Forma- tion of species in the genus Drosophila. Conclusions vii 5798:J viii THE CHROMOSOMES PAGE GLOSSARY 109 BIBLIOGRAPHY 113 INDEX 125 CHAPTER I THE RESTING NUCLEUS THE term ' resting nucleus ' is unfortunate, since it seems to imply that the metabolic activities of the nucleus are reduced to a minimum when it is not dividing — a view for which there is no evidence. The alternative term ' metabolic nucleus ' is equally un- fortunate in that it suggests that dividing nuclei are physiologically inactive. On the whole it seems best to retain the established, although misleading, term. The resting nucleus is, then, one which is not divid- ing. Usually it remains optically unaltered for long periods — it is not obviously changing either its shape or its appearance. This is, however, not always so ; many resting nuclei increase steadily in size (either by uptake of water or by actual ' growth ') and may alter their shape or appearance in various ways without dividing. The chief structural parts of the resting nucleus are the nuclear membraney the nuclear sap and the chromosomes ; these three constituents are always present — in addition bodies known as nucleoli may also be present. It has been suggested ^^ that the latter are to be regarded as portions of cjrtoplasmic material included in the nucleus, but this is not so ; true portions of the cytoplasm may become accident- ally included in the nucleus, and are then seen to be quite different in appearance from either nucleoli or nuclear sap.^'^ The nuclear membrane has been shown by micro- dissection to be a definite structure with physical properties. If a fine needle is gently pushed against 1 1 2 THE CHROMOSOMES it, it becomes indented at the point where the pressure is applied ; if the pressure is released it regains its former shape ; if the pressure is increased it can eventually be punctured. ^^ The shape of the nucleus undoubtedly depends in part on the properties of the nuclear membrane. Most nuclei are approxi- mately spherical, but many are ovoid. Those of many Wrtebrate leucocytes are in the form of a long strand with periodic enlargements (Fig. la) while those of the secretory cells of many insects are very irregularly branched (Fig. 16).^^^ In all these cases of non-spherical nuclei the nuclear surface is very large relative to its volume, and it has been assumed that this is connected with the process of secretion ; but many secretory cells (such as the salivary gland cells of liiptera) have approximately spherical nuclei. Some unusual nuclei do not consist of a single body at all, but of a number of separate vesicles, each con- taining a single chromosome and a certain amount of nuclear sap inside a separate membrane (Fig. Ic). In one case the sex chromosome is enclosed in a separate membrane from the main nucleus (Fig. Id). The nuclear sap is usually a clear fluid ; its vis- cosity has been determined in one case to be about twice that of the water, ^^ and this is probably typical of most nuclei ; in some, however, the nuclear sap may be a solid gel. The amount of nuclear sap relative to the volume of the chromosomes varies enormously from one type of nucleus to another. Thus in the micronucleus of Ciliates and in the sperm- head nuclei of many animals there is practically no nuclear sap ; on the other hand, the total volume of the young oocyte nuclei of birds (diameter up to 100 fj) may be 200,000 times that of the chromosomes at metaphase. The chromosomes may be, and usually are, in- visible in living nuclei during the resting stage. 2*- 60. 100 jj^ some plant nuclei, however, and also in a few animal nuclei, fine threads can be seen in the THE RESTING NUCLEUS 3 living resting nucleus, which are almost undoubtedly chromosomes, although highly hydrated and almost M.n. Fig. 1. — Unusual types of nuclei ; a in a vertebrate marrow- cell ; 6 in one of the spinning gland cells of Platyphylax (Caddis-fly) ; c in a spermatogonium of the grasshopper AiUarchea ; d in a spermatogonium of the Bush-Cricket Pholidoptera griseoaptera. In c each chromosome lies in a separate nuclear membrane, in d only the sex-chromo- some lies in a separate membrane. C^t. = cytoplasm ; C.v. = chromosomes lying in nuclear vesicles ; X = X - chromosome . invisible, due to their having nearly the same refrac- tive index as the nuclear sap. That the chromosomes do actually persist through the resting stage is 4 THE CHRO^NIOSOMKS certain, since in some cases they become visible at the beginning of one mitosis in the same position as they occupied at the end of the ])receding division. ^^ Also, in some cases special portions of chromosomes persist in a condensed state throughout the resting stage (proch romosomes ) . ^ * It is now becoming increasingly clear that no idea of the structure of tlie resting nucleus can be o])tained from studying fixed j)reparations. The usual text- book figure of a ' network of linen tlireads with granules of chromatin at the ])oints of intersection ' is meaningless save as a descri])tion of a gross artefact which bears only the most remote relation to the living structure. We must resign ourselves to the fiict that the resting nucleus and its chromosomes, due probably to their high water content, are unfix- able. It should be emphasized that this only a])])lies to resting nuclei — thei'c is every reason to believe that fixed preparations of nu( lei in mitosis present a very accurate ])icture of what is taking place in the living cell (see next cha])ter). Most cells in the body of an adult human being hav^e undergoiK^ about 50 mitoses since th(* feitiUzed egg (i.e. an adult man consists of about lO''* cells, making allowance for the erythrocytes, s])erms and other cells which are constantly being d(\stroye(l and replaced). In the case of insects the adult cells have undergone only about 20-30 divisions, and in the case of the Nematoda and Rotifera still fewer. Be- tween each of the^ divisions a resting stage has intervened. These resting stages are, however, of very uneven duration ; on the one hand two divisions may follow on one another without being se})arate(l by any resting stage at all (the end of one division passing directly into the beginning of the next) — or on the other hand the resting stage may last for years as in many adult tissues of vertebrates. Many adult cells may be said to have entered a i)ermanent resting stage, since they will never divide again. In most THE RESTING NUCLEUS 5 cells the resting stage lasts a much longer time than the few hours required to undergo mitosis, but this is not always so. Some nuclei begin to undergo division and then become arrested, remaining in a particular stage of mitosis for the greater part of their life -cycle. Others may be caused to do so by specific chemical agents such as auramine, sodium cacodylate and colchicine. 1^2 xhe phenomenon normally occurs in some Ciliates of the family Opalinidae, where in different species the nuclei become arrested in meta- phase or anaphase and only subsequently resume mitosis after a long static period. ^^^ Moreover, in Vertebrate oocytes with much yolk (Sharks, Amphibia and Birds) the chromosomes may remain in one stage of meiosis for many months. 2^. 1^4 This naturally leads to a consideration of the ques- tion : what is it which causes a nucleus to leave the resting stage and enter on the complicated system of changes which we call mitosis ? In certain tissues such as cleaving embryos and frequently in lobules of the testis, division takes place synchronously in all the cells — that is to say every nucleus will be in exactly the same stage at any given moment. On the other hand in epithelia such as the skin and the gut-lining isolated cells enter on mitosis quite sporadic- ally and independently of the neighbouring cells. In the first case we appear to have a ' tissue- control ' of mitosis and in the second a ' cellular control '. Prob- ably in synchronously dividing tissues the substance which is inducing mitosis can diffuse freely from cell to cell (perhaps as the result of protoplasmic ' bridges ' between the cells), while in the second case it is unable to do so. It must not be assumed that there is a single mitosis-producing agent. Possibly there is, but if so it is as yet undiscovered. But it is certain that a large number of physiological conditions are capable of stimulating cell division and it seems probable that in natural tissues more than one agency may be 6 THE CHROMOSOMES effective. In Fishes and Amphibia prolonged starva- tion followed by a meal may cause a considerable increase in the number of dividing nuclei, while in other cases starvation alone may be sufficient. ^^^ Such diverse stimulants as a peritoneal in j taction of foreign blood serum *^ or even a bacillus culture ^^ have been found to produce the same effect. What the underlying chemical mechanism is in these cases we have no idea. It must be pointed out that a mere increase in the number of dividuig nuclei in a histo- logical section is not sufficient to j)r()ve that mitosis has been stimulated hi resting nuclei ; it may e(jually well result from a slowing down of mitosis. Gross mechanical injury to the nuchuis such as results from the withdrawal of a microdissection needle i)reviously inserted into the* nucleus, will often produce a sudden acceleration of mitosis, ^^ but in some cases it will cause the normal course of mitosis to be reversed, so that nuclei which have already entered on the division cycle go back into the resting stage. 1^* The effects of irradiation with X-rays also vary from one type of tissue to another. In some cases mitosis ceases altogether in a tissue for some days after irradiation ; cells which would have entered on mitosis are retarded in the resting stage. ^^'^ Jn other cases X-rays probably accelerate the onset of mitosis. The whole subject of nuclear pathology is urgently in need of re-interpretation. Thus many nuclei when injured or dying become 'i)ycn(Aic\ that is to say their chromosomes become fused into a single large mass which stahis intensely with dyes like Haematoxylin and anilin derivatives. Pycnosis is probably to be interpreted as a highly modified non-f\mctional mitosis, since in some cases ^^'' ])yc- notic nuclei may begin to divide, but we do not know what actually happens to the chromosomes in a pycnotic nucleus. Some recent authors have omitted the primary dis- THE RESTING NUCLEUS 7 tinction between two forms of nuclear division, mitosis and atnitosis, regarding all kinds of amitosis as merely modified or concealed mitosis. While it is clear that most of the nuclear phenomena in Protozoa which were formerly regarded as amitosis are best considered in this light, there is no doubt that true amitosis occurs in many somatic tissues of insects (fat body, genital ducts, &c.). Here the chromosomes probably divide and separate into their halves in a ' resting ' nucleus and are then passively distributed in approximately, but not exactly, equal numbers to the two daughter nuclei into which the original nucleus divides by elongation into a ' dumb- bell ' shape.ii* CHAPTER II THE GENERAL OUTLINE OF MITOSIS WE have seen in the previous chapter that the stimuli which can cause the onset of mitosis are extremely diverse ; they may be regarded as external agencies releasing an inherent chain of bio- chemical and biophysical events in the nucleus. If the cell is almost ready to divide in any case, stimula- tion by a mitogenetic agent or puncturing the nuclear membrane will produce a more or less normal mitosis ; if the mitotic mechanism is not ' wound up ' then pycnosis results. A study of the gross disturbances which lead to mitosis does not help us much to understand the internal causes normally responsible for the initiation of the whole cycle of events ; neither does it explain how in some cases that cycle may be interrupted at certain stages and then subsequently resumed. It is usual to divide mitosis into four stages, pro- phase, metaphase, anaphase and telophase. For con- venience in description it seems best to subdivide anaphase into two parts and to insert a stage which can be called prometaphase between prophase and metaphase. 1. Prophase At the beginning of prophase the chromosomes become ' fixable ' — that is to say their appearance in fixed material approximates closely to that seen in living cells by the most reliable methods of observa- tion. This ' fixability ' increases throughout pro- phase until at prometaphase and metaphase there is 8 THE GENERAL OUTLINE OF MITOSIS 9 every reason to believe that fixed and stained prepara- tions give us an almost perfect picture of the appear- ance in the living state. In the majority of nuclei (all those which have not got prochromosomes) the fixability is zero during the resting stage — thus the first sign of prophase is the appearance of visible threads (chromosomes) in the nucleus in place of the network which results from the fixation of a resting nucleus. The ' fixability ' of early prophase nuclei varies a good deal, but appears to be correlated with the volume of the nuclear sap in which the chromosomes lie ; thus in the first spermatogonial division of the grasshopper Metrioptera the nucleus is large and the prophase chromosomes very ' fixable ' ; in the suc- ceeding divisions the nucleus gets progressively smaller and the chromosomes less fixable until the sixth division when there is a sudden increase in the size of the nucleus which is accompanied by an increase in fixability ; the seventh and eighth divisions again show small nuclei in which the early prophase chromosomes fix badly. ^^^ The question now arises : what is the physical basis of this property of fixability? There is con- siderable evidence that it depends on the degree of colloidal hydration of the chromosomes. Belar ^^ showed by experiments on the dehydrating power of hypertonic solutions on the nucleus that the meta- phase chromosomes contain less water than any other nuclear constituent and we have reason to believe that the resting-stage chromosomes are highly hydrated, so one must infer that dehydration is one of the processes involved in prophase. In all cases the chromosomes at the very earliest prophase are separate ; there is thus no ' continuous spireme ' as described by some of the earlier cytolo- gists (the oogonial nuclei of the scale insect Icerya purchasi form a possible exception, ^^ but the case needs re-investigation). Further, the individual c Prometaphase e Late anaphase b Mid-prophase B / .B S A EP d Meta phase-ana phase f Telophase Fig. 2. — Diagrams of the main stages of mitosis. Only two pairs of chromosomes A and A', B and B' are shown. Both of these have sub-terminal spindle attachments, those of the B chromosomes being nearer the end. At early prophase the ' relic spirals ' are clearly seen. S.A. == spindle attachments, E.P. = equatorial plane of the spindle. R and L are regions of the chromosomes which are spiralized in a right- or left-handed direction at metaphase. THE GENERAL OUTLINE OF MITOSIS 11 chromosomes are always double from the very begimiing of prophase, with the two threads or chromatids of which they are composed closely approximated throughout their length. Where there are size differences between the chromosomes the ratio of the lengths at early prophase is approxi- mately equal to the ratio of the volumes at meta- phase and it is possible to pick out the pairs of chromosomes in a diploid organism. The appearance of early prophase chromosomes depends to a certain extent on whether the preceding resting stage has been of long or short duration. Where it has been short and the chromosomes are few in number as in the cleavage divisions of the Horse Roundworm, Ascaris megalocep?iala, it is possible to show ^' that the individual chromosomes become visible in the same positions as they had disappeared in at the previous telophase. Again, where the resting stage has been short and the chromosomes are relatively long and * fixable * it can be seen that at the early prophase stages they are coiled into loose spirals (Fig. 2a). Frequently each chromosome is not merely a * spring ', but this spring itself describes a wide open spiral in the nuclear cavity ; in this case the smaller spiral has perhaps 5-20 turns, the larger one lJ-2 turns. It appears that these two spirals are always in the same direc- tion in a single chromosome, left- or right-handed, as the case may be.^* The two chromatids are always closely approximated from one end of the spiral to the other. As prophase proceeds the volume of the chromo- Bomes increases considerably ; there is thus an actual manufacture of new material during prophase. Side by side with this a shortening and thickening of the chromatids takes place. We have thus identified three processes which are involved in the development of prophase : dehydration, growth, and condensation or contraction ; we must now add a 12 THE CHROMOSOMES fourth — ' despiralization '. That is to say that as shortening and thickening take place the spirals of the early stages unwind. ^^ In well-fixed chromosomes it is possible to see from the very beginning of prophase that the staining substance of the chromatids is not continuous from end to end ; it is interrupted at one point at least (in plantr chromosomes usually several points) to form a non-staining gap (Fig. 2). ^ These gaps become more obvious later, and are called constrictions : their position is constant for each chromosome. They are filled by a non-staining substance which is not nuclear sap and which holds the chromatids on either side of it together. Throughout the prophase of mitosis the outlines of the chromatids present a slightly irregular woolly or hairy appearance which is probably an ai^tefact : they do not in general show a series of granules (chromomeres) such as are seen at the meiotic pro- phase ; this may be a real difference and not due to difference in fixability. By the end of prophase the woolly appearance referred to above has almost disappeared and a smooth outline has taken its place. The long threads of the early prophase chromo- somes appear to wind more or less at random through- out the nuclear cavity ; but they never actually come in contact with one another, or indeed approach within a certain minimum distance : there is tlius something which keeps them apart, which is probably in the nature of a generalized electrostatic repulsion distributed over the surface of the chromosome. As prophase advances there is a tendency for the shortened and thickened chromosomes to move to the periphery of the nucleus and to arrange themselves on the inner surface of the nuclear membrane. If a nucleolus or nucleoli are present in the resting stage they usually lose their staining power during prophase and have disappeared completely by pro- THE GENERAL OUTLINE OF MITOSIS 13 metaphase. This is not the case, however, in many of the Protozoa, where what appear to be nucleoli often persist through the entire mitotic cycle.* In many cases nucleoli have been shown to be attached to particular pairs of chromosomes (e.g. the sixth in order of size in Maize) during prophase 40, 105 . apparently the point of attachment is always one of the constrictions referred to above. It has been stated that in these cases nucleolar material contri- butes to the growth of the chromosomes during prophase, but the evidence for this is unconvincing. 2. Prometaphase At the end of prophase the nuclear membrane usually disappears. In many of the Protozoa, and even in some higher forms, however, it persists and the whole process of mitosis is intranuclear.* The term prometaphase designates the period from the dissolution of the nuclear membrane up to the end of the process of spindle -formation. In ' intra- nuclear mitosis ' there is no stage which can be separately distinguished as prometaphase. The mode of origin of the spindle varies con- siderably, but it is probably possible to reduce the essential details to a common plan. In the simplest cases it is formed (probably entirely out of nuclear sap after the dissolution of the membrane) as separate spindle elements corresponding in number to the chromosomes. This is the type of spindle found in the meiosis of some scale-insects '^ and in the female meiotic divisions of Artemia salina^ the Brine Shrimp (Fig. 12).^^ Usually these spindle ele- ments fuse completely to form a single gelatinous body in which the separate elements are no longer visible (Fig. 2c, rf), but in the above cases they remain distinct, and do not even converge towards ' poles ' but end in fan-shaped expansions (Fig. 3a). In these and a number of other cases no trace can be found of centrosomes or asters, which leads one 14 THE CHROMOSOMES to the conclusion that, whatever their relation to the spindle when present, they are at any rate not essential to the division process. Where centro- somes and asters are present (as in the majority, but by no means all the Metazoa) an apparent spindle ifiay form between them outside the nuclear membrane. In this case when the membrane disap- pears this structure (which is a good deal smaller than the final spindle and can be called the central spindle) moves into the middle of the nuclear area. The nuclear sap then apparently undergoes rapid gelation round the original central spindle so as to increase its volume considerably. Thus a compoimd spindle is formed which differs from the previous types only in having a central element not formed of nuclear sap but of extranuclear cytoplasm.^ The development of this central part of the spindle outside the nucleus in some cases has proved very confusing, since it has led to the conception of the chromosomes attaching themselves to a preformed spindle ; actually they are associated with the true spindle elements as soon as the latter are formed, and are probably never attached to the central element. 3. Metaphase At the end of prometaphase (the period in which the spindle is formed) the chromosomes are * at- tached ' — the term is misleading, but has to be retained — to the spindle in the region of the equator, that is to say equidistant from its two ends. The arrangement of the chromosomes at metaphase depends on a number of factors, (1) whether a central spindle element is present or not, (2) the number of the chromosomes and (3) their sizes. Where the chromosomes are very long or a large central element exists, as in the dividing leucocytes of Salamandra,* all the chromosomes are arranged round the periphery of the equator, irrespective of Fig. 3. — Various unusual types of spindles, a = the spindle at the first meiotic division in the scale insect Llaveia houvari '° where the spindle elements are quite separate. 6 = a tripolar spindle at early anaphase, c = the half- spindle formed at the first meiotic division in the fly Sciara cojyrophila.'^^* d = a. late anaphase spindle in a hybrid where the central region has undergone great elongation.*^ In c all the maternal chromosomes (including the two ' limited ' ones L and L') go to the pole, the paternal ones (I^, 11^, III^ and IV^ in order of size) moving in the opposite direction. 16 THE CHROMOSOMES their number ; their points of attachment are ap- proximately equally spaced round the edge, so that if there are 6 chromosomes they will be 60° apart, if there are 24 (as in Salamandra) they will be 15° apart. Here it appears that the spindle elements associated with the chromosomes form a circle round the central element (Fig. 4a). In some organisms Fig. 4. — ' Polar views ' of chromosomes and spindle at meta- phase ; a in the Salamander, where there are 24 large chromosomes which arrange themselves on the periphery, with a large central spindle element in the middle ; h in an organism where there are 16 large chromosomes and 16 microchromosomes which arrange themselves in the centre of the spindle. the two chromatids at this stage are strictly parallel, in others they wind round one another (cf. Fig. 56 and ^). Where there is no central element some of the chromosomes may be entirely embedded in the middle of the spindle. The following table shows the number which usually occupy the central region.^' These arrangements are found when all the chromosomes of the set are about the same size ; THE GENERAL OUTLINE OF MITOSIS 17 TABLE I Total number of chroinosomes Numhei '' in the centre Below 5 6 0-1 7-9 1 10 2 11-13 3 14 4-5 15 5 16 5-6 17-18 6 19 7 where there are considerable size -differences it is always the smaller chromosomes which occupy the centre of the spindle (Fig. 46).* All these types of arrangement can be explained if the generalized repulsion between the surfaces of the chromosomes referred to earlier persists throughout metaphase, keeping the chromosomes with their associated spmdle elements at a certain distance from one another. In the case of the peripheral chromosomes it is clear that they are attached by a single short region to the spindle, so that their long arms float freely in the cytoplasm outside the spindle, perhaps covered by a layer of spindle-substance.^^ The region of attachment corresponds to one of the constrictions seen during prophase ; this constriction, the spindle- attachment, is thus a permanent cell-organ which, although visually similar to the other or secondary constrictions, behaves entirely differently. Although the smaller central chromosomes are usually entirely * There are a few exceptions to this rule. The clearest case is that of the Tree Cricket Oecanthus longicaiida "' where the smaller chromosomes take up a peripheral position, with the larger ones in the centre. The same thing happens in many hybrids such as that between the moths Biston hirtarius and Nyssiazonaria, where the chromosomes of the two parent species are different in size and the large ones of hirtarius take up a central position in the hybrid. ^^ 2 18 THE CHROMOSOMES embedded in the substance of the spindle, they can be seen to have spindle attachments of exactly the same nature as the peripheral ones. Where a chromosome has been broken into two parts as a result of irradiation by X-rays that part which contains the spindle attachment becomes associated with the developing spindle at prometa- phase, while the part lacking a spindle attachment floats freely in the cytoplasm and never becomes attached to the spindle. ^^^' ^^^ There is thus some evidence for regarding the spindle attachment as the only part of the chromosome which plays a part in organizing the gelation of the spindle elements from the original nuclear sap ; perhaps ' spindle - element-organizer ' would be a clumsy but descriptive name for it. The position of the spindle attachment is constant for each individual chromosome, but may vary from one chromosome to another in the set. Thus in Drosophila melanogaster (Fig. 20a) chromosomes I and IV have subterminal spindle attachments, while chromosomes II and III have median attachments. Where the attachment is median the chromosome will have the shape of a V with two limbs of equal length ; where it is submedian the two limbs will be unequal (Fig. 5). It was formerly believed that the spindle attachment was terminal in many cases and a distinction was drawn between ' V-shaped ' and ' rod-shaped ' chromosomes. It is now known that the attachment is never quite terminal ; in other words there are always two limbs to the V, only one may be so short as to be practically below the limit of optical resolution. i*^' i^^' ^^^ In many cases of chromosomes with median or submedian spindle attachments there appears to be a minute granule in the centre of the spindle attach- ment which stains with aniline dyes and Haema- toxylon ; it resembles the minute granule which forms the short limb of ' rod-shaped ' chromosomes. THE GENERAL OUTLINE OF MITOSIS 19 Darlington ^^ regards this as the actual organ of attachment and calls it the ' attachment chromo- mere ' or ' centromere '. I have, however, ^^^ given reasons for believing that it is the non- staining region which is the true attachment -organ ; in many cases the ' centromere ' cannot be seen in the middle of the non-staining region, although it may be below the limit of visibility in these cases. Apparently the spindle attachment, unlike the rest of the chromosome, remains undivided during prophase and only divides at prometaphase ; its two halves then organize a spindle element, above and below the equatorial plane. Up till now we have only been considering ordinary * bipolar ' spindles ; but bipolarity is not an essential feature of the spindle — a fact which eliminates theories of mitosis based on a superficial analogy with electrical or magnetic models. In many cells such as those of cancerous tissues and in Sea -Urchin eggs which have been fertilized several times as a result of poly- spermy, multipolar spindles with a number of equatorial planes intersecting one another are found 1, 10, 143 . there may be as many as 12 poles and 6 equatorial planes. The probable structure of these multipolar spindles is indicated in Fig. 36. Even more interesting than the multipolar spindles are the'unipolar ones (half -spindles) found at meiosis in some insects (Fig. 3c) and described by the ScHRADERS '^^' ^^^ and by Metz, Moses and Hoppe.^^s At present no useful suggestion can be put forward as to how they are formed. So far we have said nothing of the ' spindle fibres ' described in many text-books, but have considered the spindle as a bundle of ' elements ' corresponding in number with the chromosomes, with or without the addition of a central element between the centrosomes. There appears to be little doubt that many of the ' continuous ' or ' interzonal ' fibres described by various workers were in fact fissures 20 Tin: CITROMOSOMES between the separate elements. On the other hand, this explanation eannot be put forward for the ' attachment fibres ' connected with the chromo- somes. These, liowever, have never been seen in living material and are invisible in well-fixed material in the majority of cases ^^' i^' ^^^ ; they are most probably caused by shrinkage of the spindle in the fixing solution in those cases where they can be seen. ^2 xhe term ' spindle fibre ' should thus be dropped, at any rate until much more definite evidence is available ; whatever is the physical basis of the artefact it is not a fibre ; nor is the conception that ' spindle fibres ' arise from ' lines of force ' a useful one. If, as I believe, these structures arise from shrinkage of the spindle forming wrinkles or other longitudinal distortion-artefacts on the surface of the spindle elements it is only natural that these should be associated with any solid object such as the attachment point of a chromosome, which interrupts the substance of the spindle element. Lewis and Lewis ^^^ have shown that it is possible to produce ' spindle fibres ' in living tissue-culture cells by the use of acid media ; the phenomenon is reversible, since on subsequently raising the ^jH the structures disappear. We left the chromatids in prometaphase as a pair of thickened threads lying closely approximated tliroughout their length and only interrupted by the spindle attachment and the secondary con- strictions (if any). Usually they come to lie even closer together at metaphase, so that the visible * split ' between them disappears ; a metaphase chromosome in transverse section thus has the shape of the symbol oo. By most ordinary methods of fixation the metaphase chromatids show no trace of internal structure, but a2)pear as homogeneous cylindrical rods (hi some cases they are slightly club- shaped having a greater diameter at the distal ends than at the spindle attachment). By various special THE GENERAL OUTLINE OF MITOSIS 21 methods of fixation, however (fixing in boiUng water, ^^® squeezing the chromosomes under a cover- glass, ^^ exposing them to fumes of ammonia or strong acids ,^^' ^® it is possible to show that each chromatid has a spiral structure, the apparent cylinder being a spring in which the successive gyres are in contact (Fig. 2d). There is no doubt that this is the true structure in the living state ; all that the special methods of fixation have done is slightly to separate the gyres and thus reveal the spiral. The two chromatids are coiled independently (in other words their gyres do not interlock as happens when two parallel wires are wound round a cylinder) and in the same direction (right- or left- handed) at any one level ; the direction of coiling may change at the spindle attachment, but does not necessarily do so ; there is considerable doubt whether it may change direction elsewhere. 3*' '^^^ ^^^' ^'' The existence of a spiral structure in metaphase chromosomes was discovered as early as 1880 by Baranetsky and there seems no doubt that it is universally present both in plants and in animals. The occurrence of a metaphase spiral explains the contract and condensation process during prophase ; this must now be interpreted as due to the develop- ment of the metaphase spiral. It will be remembered that in nuclei whose resting stage is of short duration, the early prophase chromosomes are also spiralized. We have therefore two kinds of spirals, those of early prophase and those which develop at the end of prophase and are completed by metaphase. The metaphase spirals are, however, not a continuation of the early prophase ones, since the latter have disappeared by mid -prophase ; as a matter of fact the reverse is the case ; that is to say, the early prophase spirals are the remains of the metaphase spirals of the previous division which have persisted through the intervening resting stage, and only finally unwind in the mid-prophase of the next 22 THE CHROMOSOMES division if the resting stage is short (if it is protracted they may completely unwind before the beginning of prophase). Darlington ^^ calls the early pro- phase spirals relic spirals ; if in addition to the relic spirals a larger spiral is present at early prophase, he calls the larger one a super spiral, but both relic and super spirals are due to the same cause, namely unwinding of the previous metaphase spiral. There are three ways in which the metaphase spiral might develop during late prophase and pro- metaphase. There is, unfortunately, no direct evi- dence as to which of these actually occurs, since observations on the origin of the metaphase spiral are very incomplete, the internal structure of the chromatids being difficult to study at this period. According to the first method the chromosome rotates in order to become spiralized (either one end remains fixed and the other rotates or both ends rotate in opposite directions or the spindle attachment remains fixed and the ends rotate). Darlington ^* has given several reasons why this cannot be the mode of origin of the metaphase spirals. According to the second method an internal compen- sating spiral (whose twists are below the limit of resolution of the microscope) develops in the opposite direction to the main one (i.e. right-handed if the main one is left-handed and vice versa) and with the same number of turns as the main one. In order to understand this internal compensating spiral one can carry out a simple experiment with a piece of copper wire : fix the two ends in a pair of vices and wind the middle j)art into a spiral round a metal rod : the compensating spiral will easily be seen. Darlington believes in the existence of this com- pensating spiral (which he calls the molecuktr spiral) and regards it as causing the development of the metaphase spiral. According to the third method (which is possible in a colloidal body like a chromo- some, but not in a piece of copper wire) there is no THE GENERAL OUTLINE OF MITOSIS 23 internal compensating spiral ; a sliding of molecules on one another takes its place. It does not appear possible to decide at present which of the second and third alternatives is actually found in the chromosome. First Stage of Anaphase Metaphase is a period during which almost no appreciable change takes place in the cell : it is nearly always one of the shortest stages of mitosis. At the end of metaphase the halves of the spindle attachments (the latter having divided at prometa- phase) appear to repel one another. At any rate the proximal ends of the chromatids (those, that is to say, which are attached to the spindle) begin to diverge and to move up the sides of the spindle towards the poles (Fig. 2d). From the mode of travel- ling of the spindle attachments and in view of the fact that the spindle itself does not undergo any change of shape at this stage the hypothesis of an active repulsion between the divided spindle attach- ments is the only possible one. Certainly there is no evidence for a ' traction of fibres ' at this stage — the movement of the chromatids is autonomous and depends on the spindle attachments. Why this repulsion-force should not manifest itself earlier is not clear ; perhaps the division of the spindle attach- ments is not finally completed until the end of metaphase. As a result of this movement of the chromatids the attachment regions of the latter move up the spindle towards the poles until in most cases they have travelled about two -thirds of the distance from the equator to the poles. Where the chromosomes are short this means that the spHt halves are now com- pletely separated ; where they are long the distal ends (those farthest away from the spindle attach- ments) wiU be still in contact (Fig. 2e). 24 THE CHROMOSOMES Second Stage of Anaphase When the autonomous movement of the chromatids has come to an end a remarkable change in shape takes place in the spindle (Fig. 2e). Its middle region between the two groups of spindle attachments undergoes elongation so as to complete the separation of the two sets of chromatids (which must now be called chromosomes). The growth and elongation of the middle region of the spindle to form a stem- body is apparently a universal feature of mitosis ^ ; unfortunately we have no idea what it is due to. The stem-body is clearly a solid gel like the rest of the spindle ; under abnormal conditions (e.g. in some hybrids and in cells cultured in hypertonic solutions) it may go on growing until the spindle is forced by lack of space to curl round in the cell (Fig. 3rf). Usually the stem-body shows con- spicuous longitudinal striations which are probably remnants of the divisions between the original spindle elements. Telophase The two groups of ' daughter chromosomes ' never actually reach the poles of the spindle, although as a result of the elongation of the stem-body they may travel farther apart than the original distance be- tween the poles of the metaphase spindle. When the cell has reached the stage represented by Fig. 2e the polar caps of the spindle disappear by a process of gel-solution ; the stem-body, on the other hand, frequently persists for a long time, even after cell division has been completed (Fig. 2/). As the polar caps of the spindle are destroyed a new nuclear membrane is formed round each of the telophase groups of chromosomes. The details of the process whereby the cytoplasm becomes divided into two daughter cells are outside the scope of this book. The changes which take place inside the nuclear THE GENERAL OUTLINE OF MITOSIS 25 membrane of the daughter nuclei during telophase are rather complicated hut may ])riefly he described as a reversal of those which take place during the latter half of prophase — that is to say, de-condensa- tion and de-spiralization. As a result of the latter the chromosomes become elongated and thrown into tight zig-zags inside the nuclear membrane. As they pass into the resting stage they become once more hydrated and lose their ' fixability '. It will be remembered that we described the early prophase chromosomes as longitudinally split or divided into two chromatids. The telophase chromo- somes, on the other hand, are unsplit and thus consist of a single chromatid each. The division of the chromosome in preparation for the next division thus takes place during the resting stage.* The whole process of mitosis usually takes several hours from start to finish. From 2-24 hours is probably the usual range of variation in most organ- isms. In special cases, however (see Chap. I), it may take much longer. Prophase is nearly always the longest stage, prometaphase, metaphase and the two parts of anaphase are all short stages, while telophase is considerably longer, but usually not so long as prophase. To sum up : mitosis consists of a series of cyclical colloidal phenomena, each of which is reversible. The main ones, so far as the chromosomes are con- cerned, are hydration, de-hydration, spiralization (condensation) and de-spiralization (de-condensation). Growth of the chromosome substance is probably an irreversible process and leads to the formation of two chromosome sets from what was originally a single * There may be exceptions to this. Thus it appears that in anaphase and telophase of the second meiotic division in the plant T radescantia the chromosomes are already split in preparation for the next division.'' But in general the division of the chromosomes would appear to take place during the resting stage. ^i'- ^^^^ ^^^ 26 THE CHROMOSOMES set. The actual longitudinal division of the chromo- somes takes place during the resting stage when they are invisible (or at any rate un-fixable), but the split halves (chromatids) remain closely approximated, due to the existence of a force of attraction between them, up to the beginning of anaphase. The forma- tion of the spindle and the behaviour of the chromo- somes at metaphase and anaphase depend on a special part of each chromosome, the spindle attach- ment, which persists throughout the mitotic cycle and is a self-perpetuating cell-organ with peculiar properties. CHAPTER III SPECIAL PROBLEMS OF MITOSIS Number, Form and Size of Chromosomes THE number of chromosomes in the somatic nuclei of an organism is usually the same for all the tissues and for all the individuals of the same species. There are exceptions to both these state- ments (i.e. organisms with different chromosome num- bers in different tissues and species with different chromosome numbers in different * varieties '), but they need not be considered at present. The number of chromosomes in a somatic nucleus is usually even and is referred to as the somatic number. Where there are size -differences between the chromosomes of a somatic set it will usually be found possible to arrange them in pairs (Fig. 5), the two members of each pair being exactly alike in size, in position of spindle attachment and (where they exist) of secondary con- strictions. The complete set of chromosomes is thus made up of two identical haploid sets. Organisms in which this is so are called diploid organisms and the somatic set may be called the diploid set. Sometimes even in diploid organisms one pair of chromosomes are unequal in size (Fig. 20) and sometimes the diploid number is uneven in one sex, there being a chromo- some which does not form a member of a pair in that sex (Fig. 5/, g). In these cases the uneven pair or the odd chromosome are sex -chromosomes. Their behaviour will be considered later. In hybrids between species whose chromosome sets differ it is naturally not possible to arrange them in pairs since the two haploid sets in the hybrid are not identical. 27 '3s!l (I d « S^/ crSt X g f ^ ' Fj(;, ;") SPECIAL PROBLEMS OF MITOSIS 29 The two chromosomes of a pair are said to be homologous, since they contain the same series of genes arranged in the same order. The concept of homology is one which may be appHed to parts of chromosomes as well as to whole ones, since one some- times finds a pair of chromosomes which are homo- logous in some regions but not in others (see below). In some organisms the chromosomes can be grouped, not into pairs, but into threes, fours or groupings of higher numbers. Such organisms (in which the soma- tic number is not diploid) are called polyploids^ those with three of each kind of chromosome being triploidSy those with four tetraploids and so on (pentaploids, hexaploids, heptaploids, octoploids, &c.). The lowest diploid number found in any organism is 2, which occurs in the Roundworm, Ascaris megalo- cephala var. univalens (this species also has a tetra- ploid variety, bivalens with 4 chromosomes in the diploid set).* The highest diploid numbers hitherto recorded are 208 for a Crayfish ^^ and a Crab ^^ and 200 for the Great Water Dock (Rumex hydro- * These numbers for the two varieties of Ascaris megalo- cephala refer only to the germinal tissues. In the somatic tissues a much larger number are found, as a result of frag- mentation of the 2 or 4 originally present in the fertilized egg (see later). Fig. 5. — Somatic chromosome sets of various organisms. a after Making, i®^ b after Darlington,^* d and e after Yamamoto, / and g after Hughes Schrader,'** i after Matsuura and Suto (J. Fac. Set., Hokkaido Imp. Univ. Ser., V, 5, 33), ; after Morgan, 127 the rest original. All figures slightly re- drawn, b, d, c, i are plants, the other organisms are animals. In h the coiling of the chromatids round one another is noticeable, in i the chromatids are parallel. The spindle attachments cannot be seen in c, /, and g, but are visible in all the others. In o, c, /, g, and j no trace of a ' split ' between the chromatids can be seen, in the others it can be seen, either at the ends of the chromosomes, or throughout. 30 THE CHROMOSOMES lapathum).^^* Between these limits nearly all pos- sible numbers are found in at least one species of animal or plant. Diploid numbers between 12 and 32 are common, those above and below these figures being progresgively rarer. Both in animals and plants the commonest diploid number is 24 (Tables VIII and IX). As regards size, the smallest known chromosomes are approximately 0-25 ju in length and about the same breadth at the metaphase of mitosis ^'"^ ; the longest ones are about 25 ju long and 2 // in width. ^ Normally each chromosome possesses only a single spindle attachment. By irradiation with X-rays it is, however, possible to cause two chromosomes to fuse in such a way that a compound chromosome is formed with two spindle attachments (Fig. 6a and 6). Such a chromosome may behave in either of two ways at mitosis ^^^ ; either both spindle attachments in each chromatid may go to the same pole at anaphase or to opposite poles (if those in one chromatid go to opposite poles, naturally those in the other will do likewise). When the spindle attachments in a chromatid go to opposite poles the part of the chroma- tid between them will be stretched out and eventually broken. Apparently each of the two ways of divi- sion happens in 50 per cent of cases ; the number of chromosomes with two spindle attachments is thus progressively reduced in the course of a few divisions, and such chromosomes stand no chance of becoming permanent. One case exists, however, in which normal chromo- somes have more than one spindle attachment each. That is in Ascaris megahcephala where the middle region of the long chromosomes found in the ' germ- line ' appears to contain about sixteen separate spindle attachments (Fig. 6c). Owing to the fact that these are very close together all those in one * Some Protozoa (e.g. Aggregaia) appear to have even more than this. Fig. C. — a and b : diagrams to show the two alternative modes of behaviour of a chromosome with two spindle attachments at anaphase ; c : a diagram of the method of anaphase-separation in one of the germ-line chromo- somes in Ascaris megalocephala ; the numerous spindle attachments are situated so close together that those in one chromatid always go to the same pole. 32 THE CHROMOSOMES chromatid are forced to go to the same pole at ana- phase. ^^^ This Ijappens in the spermatogonia! and oogonial divisions and at the first cleavage division. In the later cleavage divisions, however, a different type of division takes place in all those blastomeres in which are going to give rise to the somatic tissues of the adult — here the central region of the chromo- somes breaks up into a number of much smaller chromosomes, each of which probably has a single spindle attachment. The ends of the chromosomes are left with no spindle attachments and do not become connected with the spindle in any way but degenerate in the cytoplasm. This process results in an organism whose germ-cells contain two or four (according to the variety) large chromosomes, while the somatic cells have a much larger number of small chromosomes (the exact number of which is still in doubt). In another species of Ascaris, A. lumhricoides , the ends of the chromosomes are broken off and degen- erate in the <3ytoplasm in the same way, but there is no fragmentation of the central region. ^^ Two other types of chromosomes may be men- tioned here which have been observed on a number of occasions, but do not appear to have become per- manent in any wild species of organism. The first of these is the branched chromosome. It was formerly believed on genetical evidence that the 'pale ' mutation of Drosophila melanogaster was due to a small piece of the second chromosome having been broken off and attached to the side of the third one. This explanation has^now been abandoned as far as the pale mutation is concerned, but undoubted cases of branched chromosomes have been seen in several organisms. 2' » ^^ In most of these cases, the side- branch was joined on to the main chromosome at the spindle attachment. The other type of chromosome is the ring chromo- some , which has been found on a number of occasions, SPECIAL PROBLEMS OF MITOSIS 33 both in normal cells (in which ring chromosomes have apparently arisen spontaneously) and in X-rayed material. ^3^' ^^^' ^^^ Here the two ends of the chromo- some are fused together so that a closed ring results. In Drosophila melanogaster there is a stock (the * closed-X ' stock) in which the X-chromosome has its proximal and distal ends fused so as to form a ring. ^2^ In these cases the two rings which separate from one another at anaphase may become inter- locked like two links of a chain. This may lead to both chromatids going to the same pole instead of to opposite poles. During prophase and metaphase the chromatids of which the chromosome is composed are held to- gether in a paired condition throughout their length (Fig. 2). It appears (and observations on meiosis confirm this) that they are held together, not merely in a mechanical way (such as would result from a common investing sheath — like two sausages in a skin) but by a force of mutual attraction between the homologous genes of which the chromatids are com- posed. Speculation as to the nature of this force is outside the scope of this book ; but at the moment it appears to be unparalleled in biological systems. Now is this force exhausted when two chromatids are in contact, in the paired condition, or does it extend to other homologous chromatids (in other words is the force exerted merely between pairs of genes, or between threes and fours, &c.) ? In diploid nuclei at prophase and metaphase each chromatid is represented four times and if there was some ' resi- dual ' attraction we should expect homologous chromosomes to lie side by side in close approxima- tion. Usually this is not the case, i.e. there is little or no residual attraction, but it is exactly what does happen in the two-winged flies (Diptera) including Drosophila (Fig. 20) where the homologous chromo- somes lie side by side (but not actually touching) during prophase and metaphase. This state of 3 34 THE CHROMOSOMES aflFairs is called somatic pairing ; there is apparently sufficient residual attraction to cause it lo develop even in triploid nuclei (Fig. 206). Apart from this and a few other similar cases chromosomes always keep at a certain distance from one another through- out the mitotic cycle as a result of the surface repul- sion force already referred to. Genetically Active and Genetically Inert Chromosomes No attempt will be made here to explain how the idea that the genes are arranged in linear order along the chromatids has been proved, since the matter is dealt with fully in all text -books of genetics. Certain special problems must, however, be gone into. It has long been known that the Y- chromosome in Droso- phila melanogaster is genetically almost inactive. Males lacking it (called X-nought males) are viable but sterile. It has been shown that the Y contains two separate genes necessary to ensure fertility in addition to the normal allelomorph of the gene * bobbed ' i^s . apart from these it may contain a few other genes. In Maize there is also a chromo- some (the B-chromosome) which contains no known genes and may be present any number of times in the chromosome set or may be absent altogether without in any way affecting the phenotypic appear- ance of the plant. ^^* These, then, are examples of inert chromosomes. It has recently been shown ^^^ that the part of the X-chromosome in D. melanogaster which is next to the spindle attachment (about one- third of the total length) is almost completely inert (up to the present this region only contains one known gene, namely bobbed, which suggests that this region is homologous with that part of the Y which likewise contains bobbed). In addition there are almost certainly small regions in the centre of chromosomes II and III, on either side of the spindle attachment, which are also inert. The possibility thus arises SPECIAL PROBLEMS OF MITOSIS 36 that in many other organisms some chromosomes or regions of chromosomes may be genetically inert ; for this reason we must not expect the number of linkage groups and the haploid number of chromo- somes to correspond in all cases. Salivary Gland Chromosomes in Diptera In the nuclei of tha salivary glands in the Diptera (and also in the nuclei of the rectal epithelium and the Malpighian tubules, which show an essentially identical structure) the chromosomes appear as enormously enlarged threads whose structure could not for a long time be interpreted in terms of ordinary mitotic chromosomes. The first clue to the analysis came when it was realized that in the salivary gland nuclei the phenomenon of somatic pairing is developed to the point where the homologous chromosomes are completely in contact throughout their length. ^^ As a matter of fact it is often possible by means of crushing under a cover -glass to separate slightly the homologous chromosomes ; when this is done it will be seen that they are spirally wound round one another as in a piece of two-strand rope.^^ Xhe number of these threads thus corresponds to the haploid, and not to the diploid number of chromo- somes, but the two arms of V-shaped chromosomes are represented by separate threads ; thus in the female Drosophila melanogaster with eight chromo- somes in the diploid set, four of which are V-shaped the salivary gland nuclei contain six threads (i.e. one representing the two X- chromosomes, one for each of the arms of the Ilnd and Ilird chromosomes and a short one representing the two IVth chromo- somes). In Drosophila all these six threads are attached at one end to a body called the chromocentre, to which the nucleolus is also attached by a thread, but in many other Diptera (e.g. Bibio and Chironomus) there is no chromocentre, the threads being separate and unconnected. 2. ^5 Apparently the chromocentre 36 THE CHROMOSOMES results in part from a fusing together of the spindle attachments and adjacent regions of all the chromosomes. The nuclei of the salivary gland cells are very large (about 25 /i in diameter in Drosophila) but even so the chromosomes are so long that they are tangled and coiled up so as to pack them inside the nuclear membrane. The usual method of studying these chromosomes consists in fixing and staining them simultaneously with a saturated solution of carmine in 40 per cent acetic acid to which a trace of an iron salt is added just before use. The preparations are then lightly crushed under a cover-glass in order to rupture the nuclear membranes and spread out the chromosomes on the slide. This crushing results in a considerable stretching of the chromosomes ; the X-chromosome of Z). melanogaster maybe 260// long, in acetocarmine preparations, while in the living salivary gland nucleus it is probably only about 150^ in length. Even so the unst retched chromosomes in the salivary gland chromosomes are at least 50 times as long as the chromosomes at ordinary somatic mitoses and 1000 times their volume. Each of the homologous chromosomes which pair to form the threads in the salivary gland nucleus is transversely striated with bands which stain with ordinary nuclear dyes (e.g. Crystal Violet and Haematoxylin). These cross bands are of varying thickness and are separated by non-staining inter- nodes. It is apparently * these internodes which stretch when the nucleus is crushed. The bands correspond exactly in position in the two homologous chromosomes and are always the same for a particular chromosome in different individuals of the same species. The thicker bands as seen in the uncrushed nuclei are apparently made up of several thinner bands which can be separated by crushing and stretch- ing under a cover -glass. The total number of bands in the X-chromosome of D. melanogaster is at least Maternal- band of (^)256 Chromomcrrs Fig 7. — Diagrams illustrating the structure of salivary gland chromosomes, a = a general view of a salivary gland nucleus with the chromosomes coiled within it. The ' bands ' are shown, but not the threads connecting them. The nuclear sap is shown in black, b — the chromosomes of a salivary gland nucleus in the male Drosophila mclatiogaster spread out by crushing the nucleus. The maternal parts of the paired chromosomes are shown in black, the paternal parts white. Chr. = chromocentre (stippled) ; Nuc. = the nucleolus. II L and II R are the two arms of the Ilnd chromosome, III L and III R the two arms of the Illrd chromosome ; the IVth chromo- some and the X and Y chromosomes (the latter very small) are also shown. An inversion (Inv.) is shown in the ' right ' arm of the Ilird chromosome, c = a diagram of a small part of a salivary gland chromosome, composed of 2r)6 threads, 128 paternal and 128 maternal, d = adia- gram illustrating how the chromocentre is made up (in the female of D. melanogaster) by a fusion of the inert parts of all the chromosomes (represented by stippled bands). 38 THE CHROMOSOMES 4,000 if one counts all the finer bands of which the thicker ones appear to be made up.^^ Very frequently (but not always) the bands are in pairs, that is to say two adjacent ones are of exactly the same thickness (Fig. 6). Each band is clearly a disk, that is to say they extend through the thickness of the chromosome ; moreover each disk or band is made up of a number of granules which have more or less completely fused to form a transverse plate ; in Chironomus there are at least 256 of these granules in each band. The granules in one band are connected with those in the next by means of fine longitudinal threads which run through the non- staining intemodes. The two strands of our ' rope ' are thus themselves made up of a number of threads which bear periodic enlargements in the form of granules that tend (at any rate in acetocarmine preparations) to fuse into transverse bands.i9. ^2, 95 This interpretation of the structure of salivary gland chromosomes probably applies to all Diptera ; various other theories as to their structure have been put forward in connexion with genera (Drosophila and Sciara) whose salivary gland chromosomes do not appear to fix very satisfactorily in acetocarmine ; but in larvae of midges (Chironomidae) there can be no doubt that the above is the correct interpretation. The salivary gland chromosomes are thus to be regarded as resting stage or early prophase chromatids stretched out straight which are not wound into a tight spiral as at an ordinary somatic metaphase and which have split longitudinally again and again. All the peculiarities of the salivary gland nuclei can be explained on the basis of these two simple assumptions if one takes into account the well-known somatic pairing phenomenon found in all Diptera. The fact that the salivary gland chromosomes are about fifty times the length of the ordinary metaphase chromo- somes will be understood if one takes a tightly coiled spring and pulls it out straight. The great volume of SPECIAL PROBLEMS OF MITOSIS 39 the salivary gland nuclei (about 65,000// ^) is natural if they are polyploid and the thickness of the chromo- somes results from the fact that they have divided a large number of times. The formation of the chromocentre has recently been shown ^^^ to be due to the fact that the regions immediately adjace ntto the spindle attachment (which arp inert but all 'homologous') have, as a result of their homology, all fused into a single mass (Fig. 7d). The origin of the chromocentre is thus ultimately the result of the somatic pairing pheno- menon. Diptera which lack a chromocentre pre- sumably lack the homologous regions next to the spindle attachments. Sex-Chromosomes All chromosomes other than sex-chromosomes are called autosomes. The sex-determining mechanism of the majority of bisexual animals and plants con- sists of a pair of chromosomes which may be regarded as modified autosomes. In one sex these form an equal pair of homologous chromosomes while in the other the pair is unequal. The sex which possesses the unequal pair is called the heterogametic sex since it produces two kinds of gametes or spores ; the other sex is called the homogametic sex. In most groups of organisms it is the male which is the heterogametic sex (that is to say there are two kinds of sperms or pollen grains and only one kind of egg or megaspore), but in some groups it is the female which is hetero- gametic — that is to say there are two kinds of eggs or megaspores, all the sperms or pollen grains being alike (see Table II). In the Bryophyta (mosses and liverworts) where the sexual stage of the life cycle is haploid the male and female gametophytes each contain one member of a pair of sex-chromosomes. It will be seen that the sex-determining mechanism is merely a special kind of heterozygosity — in respect of a whole chromosome instead of in respect of a 40 THE CHROMOSOMES single gene. The sex-chromosomes are m most cases (perhaps in all) not the only ones bearing sex-deter- mining genes ; probably all the autosomes carry genes which are concerned with the development of characters of one or the other sex ; all that the sex- chromosomes do is to act as a differential mechanism which switches the development of the embryo over to maleness or femaleness from a potentially herma- phrodite condition. According to the usual terminology the chromo- somes forming the equal pair in the homogametic sex are called X- chromosomes. The diploid set in the heterogametic sex contains one X- chromosome and in addition a chromosome bearing a greater or less resemblance to it called the Y-chromosome. We can regard the various types of sex-chromosomes as progressive evolutionary modifications of an original pair of autosomes, these modifications consisting of the XY pair becoming more and more unlike. In the majority of fishes and amphibia there is no visible cytological difference between the X and the Y ; they probably only differ in respect of a few genes. Thus in fishes of the genus Platypoecilus one species has male heterogamety, another female heterogamety.^ In most mammals and in many insects (such as Drosophila- melanogaster) the X and the Y are of very different sizes so that they can easily be dis- tinguished. Here it is probable that only a short region of the Y is homologous to a similar short region in the X. Usually the Y is smaller than the X but in some cases, as in D. melanogaster it is con- siderably longer. In many organisms the Y is very small indeed and in a large number of groups it has been lost altogether. In these the diploid set in the heterogametic sex consists of an odd number of chromosomes (one less than in the homogametic sex). In a number of organisms the X is represented by two separate chromosomes which can be called X* SPECIAL PROBLEMS OF MITOSIS 41 and X^. Thus in the Praying Mantis the chromo- some sets of the two sexes are : Male 13 pairs of autosomes, X^ X^, Y Female 13 „ „ X^, X\ X\ XK In this case it will be seen that the heterogametic sex has an odd number of chromosomes, but a Y is present. ^^ The two ends of the Y are probably homologous to regions in the X^ and X^ respectively (Fig. 86). y Y I ^ »v\\\\\\\\\\\\\\\\\\\^ X sSme^t Differential Segments ^t^ ) Y S.A. Mechanism S.A. M XiXjY T^^/.. Differential Segments ^^9^ Mechanism\(\ SA. X, S.A. X Y Fig. 8. — Diagrams of the sex-chromosomes (above) the male rat, and (below) the male Praying Mantis. The homo- logies of the various regions are indicated diagram- matically. S.A. = spindle attachment. In the insect Perla marginata ^® there is a similar situation but a Y is absent so that the two sets are : Male 10 pairs of autosomes, X^, X^ Female 10 „ „ X^, X^, X^, X2. In some cases it is the Y which is represented by two separate chromosomes, there being only one X. Thus in the dioecious Sorrel Dock, Rumex acetosa, we have : 42 THE CHROMOSOMES Male plant 6 pairs of autosomes, X, Y^, Y^ Female plant 6 ,, ,, X, X. In this case it will be seen that the heterogametic sex possesses one more chromosome than the other instead of one less as in the XX : X type.^^® TABLE II Male Heterogametic and Female Homogametic. Animals All Insects except Crustacea Arachnida Opisthogoneata Annelida Nematoda Echinodermata (as far as known) All Vertebrates except. Female Heterogametic AND Male Homogametic. those with haploid males and the Lepidoptera and Trichoptera ^*^' ®' 161 Some Fishes • Some Reptiles at any rate ^*** All Birds ^". !«' Hexaploid Fragaria ela- tior 101 Plants 1" Rumex section Acetosa Humulus Melandriiim Populus Empetrum Elodea TABLE III Organisms with Sex -Determination by Male Haploid y Insects : Rhynchota Some Aleurodidae I'l Some Coccidae i*^' •* Hymenoptera Probably all i". 172. 134 Coleoptera : Micromalthtis dehilia i** Arachnida : Acarina (Mites) Some at least i«o. i" Rotifera : Aslanchna 1®* (Also possibly in some Thysanoptera) SPECIAL PROBLEMS OF MITOSIS 43 The identification of sex-chromosomes is naturally difficult when the X and Y are cytologically indis- tinguishable. Where they are visibly different it is usually possible to identify them by careful compari- son between the somatic chromosome sets of the two sexes. In the more highly evolved types of sex- determining mechanism the sex-chromosomes, or at any rate parts of them, are often clearly recognizable by the fact that they contract during prophase at a different rate to the autosomes (usually slower). Thus in the spermatogonial mitoses of the Acrididae and Gryllidae (Grasshoppers and Crickets) the greater part of the X has only reached the mid-prophase degree of contraction by the time the autosomes are in metaphase.^^i' ^^^ This process of differential condensation is known as heteropycnosis . In the above case we can speak of negative heteropycnosis ; the opposite condition of positive heteropycnosis (where the sex-chromosome reaches the metaphase degree of contraction when the autosomes are still in early prophase) is often found in the first meiotic division (see Chap. 5).'^' ^^ Heteropycnosis is not confined to sex-chromosomes ; thus in many species of grasshoppers there is one autosome which shows strong positive heteropycnosis in about two -thirds of its length at the first meiotic division. ^^ Moreover in Drosophila melanogaster all the inert regions show heteropycnosis at the somatic divisions. There seems to be some sort of connexion between inertness and heteropycnosis, but it is not possible as yet to state definitely that all heteropyc- notic regions are more or less inert and vice versa. In a number of groups of organisms sex-determina- tion does not depend on a pair of sex-chromosomes, but on whether the eggs are fertilized or develop parthenogenetically. Thus in these cases the males are haploid , the females diploid . This is the case in the Hymenoptera, the Mites (some species at any rate) and in some Scale Insects (Coccidae). In the Hymenoptera 44 THE CHROMOSOMES it is clear from recent work on the parasitic wasp Habrohracon that the haploid- diploid scheme of sex- determination CO -exists with a mechanism involving female heterogamety.^'^ xhe females possess an X- and a Y- chromosome and thus produce two kinds of eggs, X-eggs and Y-eggs. If these develop with- out fertilization they give rise to haploid males which are accordingly of two types, X-males and Y-males. These produce sperms by a modified meiosis which does not involve any reduction in number of chromo- somes. Normally X-eggs are only fertilized by Y sperms and vice versa, giving rise to diploid XY- females ; occasionally, however, X-sperms fertilize X-eggs and Y-sperms Y-eggs. The result of this is to produce extremely rare diploid male individuals. Femaleness in Habrobracon clearly depends on an interaction of genes present in the X- and Y- chromo- somes ; which explains why both haploid and homozy- gous individuals are males. Polyploidy Polyploid cells arise in the first place through a failure of cell division ; that is to say that in a par- ticular cell the chromosomes divide, but the cell fails to do so.^'^' ^'^ If this happens in a diploid organism a tetraploid cell will result. Such a process occurs normally in a small percentage of the meristem cells in the root tips of certain plants 121, 126. ^g^ result patches of tetraploid tissue are produced, surrounded by diploid cells. In the tomato it is possible to pro- duce tetraploid shoots by repeated cutting back.*^^ From such shoots it is possible to produce tetraploid tomato plants. If the process is repeated in a tetraploid plant octoploid cells will be produced. Triploids arise in the main through crossing between diploids and terta- ploids, hexaploids by a doubling of the chromosome set in a triploid and pentaploids by crossing between tetraploids and hexaploids. SPECIAL PROBLEMS OF MITOSIS 46 It has been found necessary to distinguish between two kinds of polyploidy which are called auto -poly- ploidy and alio -polyploidy. An auto -polyploid is an organism with more than two haploid sets of chromo- somes which are all alike — it has arisen by doubling of the chromosomes in an individual which is not a hybrid. In an alio -polyploid, on the other hand, the doubling has taken place in a hybrid, so that the several haploid sets are not all identical, having been derived from different parent species. Polyploidy may occur within a single species, as in the case of the Cruciferous plant Biscutella laevigata where diploid, triploid, tetraploid, pentaploid and hexaploid plants have been found ^^^ or as between different species of the same genus. A good example of the latter phenomenon is the Section Lapathum of the genus Rumex (Table V). Intraspecific poly- ploidy is of course always auto-polyploidy, while the interspecific kind may be either auto- or alio -poly- ploidy. A special t3rpe of polyploidy occurs where one or more chromosomes, but not the whole haploid set, are present more than twice in the complete chromo- some set ; this is called polysomy. The most common t3rpe of polysomy is trisomy where one or more chro- mosomes are present three times, the others being only present twice. The opposite phenomenon where a polyploid lacks one or more chromosomes from one TABLE IV If n = the haploid number of one species and n that of another, then — 2n = diploid number of first species 2n = ,, „ ,, second species n + n = diploid number of hybrid between them 3n and 3n = the auto-triploids 4n and 4m = the auto-tetraploids 2n + 2n = the allo-tetraploid 2n + 1 = a trisomic ) these terms are partially inter- 4n — 1 = an aneuploidi changeable 4n -f 2n and 3n + 3n = different kinds of allo-hexaploids 46 THE CHROMOSOMES haploid set is called aneuploidy : it is obvious that there is no hard and fast distinction between polysomy and aneuploidy (see Table IV). TABLE V Somatic Chbomosome Numbers in the Genus Rumex (section Lapathum) ®*« ''' '^ R. salicifolius ..... 20 R. alpinus . 20 R. obtusifolius (one var.) 20 R. conglomeratus . 20 R. sanguineus 20 R. scutatus . 20 R. puleher (one var.) . 20 R. maritimus 40 R. limosus . 40 R. brittanicus (one var.) 40 R. puleher (another var.) 40 R. domesticus 40 R. obtusifolius (another var. ) 40 R. crispus . 60 R. patientia 60 R. orientalis 60 R. domesticus (another vfi-r.) 60 >» ( >» >* ) 80 R. japonicus . 100 R. hymenosepalus 100 R. andraeanus . 120 R. brittanicus (another var.) . 160 R. aquaticus . 200 R. hydrolapathum . 200 Owing to the fact that in bisexual organisms the sex -determining mechanism depends in general on the segregation of a single pair of chromosomes at meiosis pol3^1oid species and varieties are usually found only in groups where reproduction does not depend on bisexuality (most Angiosperms, partheno- genetic and hermaphrodite animals such as the Pulmonata , Oligochaeta, &c. ) . If polyploidy occurred in bisexual species it would completely upset the sex -determining mechanism. ^^^ Since the vast ma- jority of Angiosperms are monoecious while the bulk of the Metazoa are dioecious polyploidy is far com- moner in the higher plants than in animals. CHAPTER IV THE GENERAL OUTLINE OF MEIOSIS MEIOSIS is the antithesis of fertilization ; in diploid organisms it results in the chromosomes being reduced to the haploid number. If meiosis takes place at the beginning of the life-cycle, just after fertilization (this type of meiosis is called initial or zygotic meiosis and occurs in most of the Sporozoa and in the Charices,^? Basidiomycetes ^^ and Ascomy- cetes ^^^ among plants as well as in some of the lower Algae ^^) the synthesis (to use a philosophical term) of fertilization and meiosis will be a haploid adult organism. If on the other hand meiosis occurs just before fertilization (i.e. during gamete formation) as happens in all the higher animals (Metazoa) then the adult organism will be diploid. In the higher plants with an alternation between the sporophyte and gametophyte generation meiosis takes place during spore-formation, i.e. occupies an intermediate place in the life -cycle ; where the gametophyte generation is the predominant one (as in the mosses and liverworts) the ' adult ' phase of the life -cycle will be haploid ; where it is the sporophyte which is predominant (as in the Pteridophytes and Phanerogams) the ' adult ' phase will be diploid. Thus a moss-plant is haploid and a buttercup diploid, but in both meiosis takes place at the same stage in the life -cycle, during spore - formation. Meiosis has been defined by Darlington ^^ as ' the occurrence of two divisions of a nucleus accompanied by one division of its chromosomes '. The whole process must be regarded as having arisen in the 47 THE GENERAL OUTLINE OF MEIOSIS 49 first place through a profound modification of two mitotic divisions. It is very remarkable that in all essential details meiosis is the same wherever it occurs ; it is consequently possible to give a general account of it which will apply equally well to the gametogenesis of an insect and the sporogenesis of a plant. The first meiotic division always has an elongated prophase ; since this is in many ways different from a mitotic prophase it is necessary to subdivide it into a number of stages which, although they corres- pond to the early, mid- and late prophase stages of mitosis, have different names to indicate the main processes which take place. The names of these stages are, in order, leptotene, zygotene, pachytene, diplotene and diakinesis. After diakinesis (which corresponds to the end of prophase) comes a short prometaphase, followed by the metaphase of the first division (' First Metaphase '). In the following account we shall describe meiosis in a diploid organism ; the meiosis of a polyploid is in some respects more difficult to understand and is best left until the details of the process in a diploid have been explained. As no mention will be made of the cytoplasmic phenomena of meiosis the description will do for either spermatogenesis or oogenesis, macro- or micro-sporogenesis, since there are no constant differences between the nuclear phenomena in the two sexes. Fig. 9. — Diagrams of the main stages of meiosis. Two pairs of chromosomes AA' and BB' are shown, the A and A' chromosomes having submedian spindle attachments, the B and B' chromosomes having sub-terminal ones. Three chiasmata are formed in the AA' bivalent, one in the BB' bivalent. S.A. = spindle attachment, Ch. = chiasma. There is no " terminalization " of chiasmata. In the BB' bivalent a rotation of the arms takes place, in the AA' bivalent there is no rotation. 4 THE GENERAL OUTLINE OF MEIOSIS 61 Leptotene This is the earliest part of the prophase of the first meiotic division — it corresponds to the very beginning of prophase in an ordinary mitosis . The chromosomes in the leptotene stage resemble those of the early prophase of mitosis except for one important feature ; they are not longitudinally divided — in other words each consists of a single chromatid and not of two chromatids held together throughout their length as in the case of mitosis. Another point of difference which is somewhat variable is that the leptotene chromosomes are rather clearly made up of a series of granules (called chromomeres) connected by non- staining intervals ; this may also be the case at mitosis, but it is not usually so obvious. As in the case of the granules in the salivary gland chromo- somes it has naturally been suggested that the chromomeres are actual genes. In various Liliaceae the total number of chromomeres in the whole chromosome set has been counted and found to lie between 1,500 and 2,500 ^ ; of course we do not know the total number of genes in these plants, but in Drosophila the total number (including the hitherto undiscovered ones) has been estimated at 14,000.^^ The leptotene chromosomes are present in the same number as in the somatic tissues ; very often they are not arranged at random inside the nucleus, but preserve the arrangement of the previous tele- phase (with all the spindle attachments together at one side of the nucleus and the chromosomes arranged as in a bunch of flowers ^^ ; in this case they are said to be polarized. Zygotene Ijcptotene is usually a stage of short duration. It is followed by a stage called zygotene in which the homologous chromosomes come together in pairs and become closely approximated throughout their length. 62 THE CHROMOSOMES This process is called pairing or (in the older accounts) synapsis. In the case of each pair of homologous chromosomes the pairing process begins at one or more points and then spreads along the length of the chromosomes (Fig. 96). Where the telophase arrange- ment of the previous division has been retained pairing begins at the spindle attachments — other- wise it may T^egin anywhere. It must be pointed out that the pairing is not merely between homologous chromosomes, but always between strictly homo- logous regions ; this can be seen very clearly where the chromomeres are distinct, since they are of slightly different sizes (Fig. 96). If we call those in one homologous chromosome a, b, c, d, . , . and those in the other a', b', c', d', . . . then a will pair with a' and b with b' and so on. If a short region has become inverted in one homologous chromosome but not in the other (as sometimes happens) then the inverted region will remain as an unimired loop in the middle (Fig. Ha). If a rather longer section is inverted the loop will twist round and pair as in Fig. 116. If a small section is completely missing from one chromosome, then the corresponding section in the homologous chromosome will form a short loop (Fig. llc).^^ It appears, therefore, that the force of attraction is a mutual one between homo- logous chromomeres (or genes) and that it is probably identical with the force which keeps the two chroma- tids of a chromosome together throughout their length in the prophase of mitosis. It seems natural that this force should lead to a pairing of homologous chromo- somes at zygotene, since these have not undergone the usual longitudinal division — they are still unsplit at a stage when they would normally be split ; the force of attraction thus satisfies itself at mitosis by maintaining chromatids together and at zygotene by bringing distinct chromosomes into longitudinal approximation. 1^* Darlington regards the pro- phase of the first meiotic division as ' preqocious ' in «-— •-^^ Paternal Maternal Fig. 11. — Diagrams of chromomeres (or genes) at the two- strand pachytene stage in a bivalent which is hetero- zygous for (o) a small inversion, (6) a large inversion, (c) a deletion. 64 THE CHROMOSOMES comparison with the prophase of an ordinary mitosis, but it is probably better to regard the splitting of the chromosomes as being delayed. Pachytene As a result of pairing the apparent number of chromosome threads (in a diploid organism) has been reduced to half ; if there were 2n chromosomes in leptotene there will be n associations of two chromo- somes at the beginning of pachytene. These associa- tions of pairs of chromosomes are called bivalents. Each bivalent has a split down the middle and thus closely simulates an ordinary mitotic chromosome at Attract/on ■ Pater nat qWX jiVjcl xwG^yz jiK^kl xwA^yz lkG2mn zyV^a^ IkA^ab zyB.,(x^ nmY ^PP i^aGrge/ nmR^op ^oiA-jmn Rings of four, six or more chromosomes are formed at the first meiotic division in a number of other plant genera {Campanula, Pisum, &c.) and from a study of these cases (some of which, notably Pisum, have arisen under experimental conditions) it has been possible to explain the origin of the whole mechanism (Chap. VI). CHAPTER VI CHROMOSOMES AND EVOLUTION TWO different concepts are included in the term evolution and are frequently confused with one another — morphological change and the forma- tion of new species. There has recently been a tendency to regard the species as a category which varies from group to group, so that a * species * of beetle is not equivalent to a ' species ' of primate or conifer. From a morphological point of view this is no doubt true — the extent of the taxonomic diiferences between species do vary from group to group. But from a genetical point of view the species (i.e. a group of individuals all of which nor- mally and regularly breed together except in so far as they may be separated by geographical isolation) is still a satisfactory concept which only breaks down in organisms which reproduce without fertilization. With the exception of a small number of cytoplas- mic characters in plants (and possibly also in animals, but there is no well-established case as yet) all evolu- tionary changes have clearly arisen in the first place as changes in the chromosomes. It was formerly believed that these changes were of two main types : gene-mutations (which were conceived of as sub- microscopic molecular changes) and structural altera' lions in chromosomes (i.e. microscopically visible rearrangement of whole blocks of genes by transloca- tion, inversion, &c.). It has recently been shown, however, 1^- that this distinction is rather one of degree than of kind ; most (possibly all) gene-muta- tions are minute structural alterations involving 89 90 THE CHROMOSOMES the rearrangement of a small section of a chromo- some. The detailed analysis of evolution by cytological and genetical methods has only just begun. Already, however, one thing is clear — the mechanism of evolu- tion (or at any rate the mechanism of species-forma- tion) has not been the same in all groups of organisms. Translocation and inversion of chromosome segments, hybridization, auto- and alio -polyploidy are the raw materials of species formation ; but they have con- tributed to different degrees in different groups of organisms. The mechanism of evolution has varied from group to group and even from genus to genus, so that it is becoming increasingly difficult to formu- late general ' laws ' of evolution, and the universal applicability of such ' laws ' as have been derived from palaeontology and morphology is becoming more and more doubtful. The chief limiting factor in determining the mech- anism of evolution in a group of organisms is the method of reproduction and the mode of sex-deter- mination. At least five main types of life -cycle can be distinguished, each of which probably possesses a highly variable system of evolution and species formation : 1. Clonal or vegetative reproduction (Meiosis and Fertilization absent or non-functional). 2. Sexual reproduction in Hermaphrodite organ- isms with self- or cross-fertilization. 3. Sexual reproduction in Complex Heterozygote organisms. 4. Sexual reproduction in dioecious organisms with a chromosomal sex-determining mechanism. 5. Sexual reproduction in organisms with male haploidy. Species-formation in Angiosperms If one makes a list of the known haploid chromo- some-numbers in the species of higher plants and CHROMOSOMES AND EVOLUTION 91 animals one finds that in the case of the Angiosperms there is an excess of even over odd haploid numbers amounting to over 40 per cent, while in the case of animals the numbers of even and odd haploid num- bers are not significantly unequal (Tables VIII and IX). This figure of 40 per cent gives a lower limit to the number of polyploid species of Angiosperms. At least 40 per cent of all Angiosperm species are tetraploid, hexaploid, &c. There is no other reason why more even than odd haploid numbers should exist and a detailed investigation of the frequency of particular numbers ^^^ bears out the general conclu- sion. One cannot set an upper limit to the extent to which polyploidy has occurred in plants, since aneuploids and other ' derived polyploids ' are not included in the above minimum figure. Neither is it possible to estimate the extent to which all Angio- spt^rms have a more or less remote polyploid ancestry. It is also difficult to judge the relative importance of auto- and allo-polyploidy. The distinction between them is, however, only a relative one, since, on the one hand, allo-polyploids exist which are the result of hybridization between very closely related parent species or varieties and, on the other hand, inde- pendent mutation in the two diploid sets of an auto- tetraploid will give rise to an organism which will, like an allo-tetraploid, have a number of genetic differences between its two diploid chromosome sets. Thus in two ways forms will arise which are essen- tially intermediate between auto- and allo-polyploids in their genetic constitution. Whatever their origin, all even-numbered poly- ploids (tetraploids, hexaploids, &c.) will, as a result of independent mutation in their multiple chromosome sets, tend to evolve towards a new condition of genetical diploidy, in which no gene will be repre- sented more than twice in the chromosome set. There can be no doubt that many plant species represent stages in this process. Mutations are less likely to 92 THE CHROMOSOMES prove lethal in a tetraploid than in a diploid, since two of the four genes (and similarly four of the six in a hexaploid) are more or less superfluous and can consequently mutate without vitally affecting some important process as usually happens when the genes of a diploid mutate. It seems fairly clear that a number of new Angio- sperm genera have arisen as a result of hybridization between fairly widely separated species followed by allo-polyploidy, just as Raphanobrassica (allo-tetra- ploid Raddish- Cabbage hybrid) and Aegilotriticum (allo-polyploid Wheat -Rye hybrid) have arisen under experiment. It is at any rate certain that well- marked species like Aesculus camea, Galeopsis Tet- rahit, Spartina Townshendii, and the American form of Phleum pratense, have arisen in the first place as allo-polyploids.^^' i^^' '2* ^^ On the other hand where several chromosome numbers which are multiples of the lowest one exist within the same species of plant (as in Biscutella laevigata (see Chap. Ill) and Nastur- tium officinale, ^^^ they are clearly due to auto -poly- ploidy. Such auto -polyploid varieties are usually taxonomically distinguishable from the diploid form and from one another by slight morphological dif- ferences and may form the starting-point for new species if, as is usually the case, their hybrids with the diploid form or with one another are more or less sterile. It is probably in this way that the species of the genus Rumex (section Lapathum) have originated (Table V). Although polyploidy has undoubtedly taken place in nearly all the larger plant genera, and been re- sponsible for the origin of a large number of species, there are some genera such as Carex where the known somatic chromosome numbers (18, 30, 32, 38, 52, 54, 56, 62, 64, 68, 70, 72, 74, 76, 80, 82) do not suggest that it has played any important part in the forma- tion of new species. The reason for this is not clear ; perhaps the actual formation of pol3rploid cells is CHROMOSOMES AND EVOLUTION 93 rare in a genus like Carex, or perhaps polyploid individuals are considerably less viable than diploid ones. Origin of Complex Heterozygote Organisms Hypericum punctatum ^^ is the only species of its genus which is a complex heterozygote, all the other St. John's Worts being normal diploids ; its evolu- tionary origin is thus extraordinarily interesting, but cannot be analysed. Bhoeo discolor is the only representative of its genus, so that here again we have no method of determining the origin of the complex - heterozygote mechanism. The genus Oenothera, on the other hand, contains a series of forms ranging from ordinary diploid species to organisms which form a ring of fourteen chromosomes at meiosis (see Chap. V). If we take one of the diploid species of Oenothera which must be regarded as ancestral to those which form rings we can consider two pairs of chromosomes : abcdefghi and mnopqrst abcdefghi mnopqrst If a reciprocal exchange (mutual translocation) takes place between one member of each pair of chromo- somes it will give rise to two new chromosomes : abcpqrst and mnodefghi. The four chromosomes : abcdefghi mnopqrst abcpqrst mnodefghi will now form a ring at meiosis ; but there will not be any ' median segment ' in such a ring ; on the other hand, if a second reciprocal interchange takes place it will establish a median segment if it does not correspond in position with the first. The second interchange may result from normal crossing-over if the first one was interstitial instead of terminal. By a repetition of this process rings of 6, 8, 10 and higher numbers can arise. The ring-forming species of 94 THE CHROMOSOMES Oenothera produce trisomic gametes fairly often as a result of one chromosome in the ring going to the wrong pole at the first meiotic division. They also produce new types of gametes as a result of occasional chiasma-fotmation in the median segments. The * mutants ' of De Vries (on which he based his * Mutationstheorie ') were not really due to genet- ical mutation, but to occasional cross-overs of this type. Origest of New Species in Bisexual Organisms In animals (with the exception of a few partheno- genetic and hermaphrodite groups such as the Pul- monata, Oligochaeta and in some Lepidoptera and Crustacea) polyploidy is entirely absent and conse- quently cannot account for the origin of new species. On the other hand the variation in chromosome number is almost as great in animals as in plants (cf. Tables VIII and IX). It is therefore clear that various mechanisms exist whereby chromosome num- bers can be altered, and it is probable that these are closely connected with the origin of new species in some cases. It used to be supposed ^^' that two or more chromosomes could merely fuse together to form a single one, and alternatively that one chromosome could break into a number of pieces, each of which would behave in future as a separate chromosome. If this were so a study of chromosome numbers would be of little importance. We now know, however, that each chromosome contains a single spindle attachment which is a self-perpetuating body ; new spindle attachments only arise from pre-existing ones.^^^ Moreover, although spindle attachments divide longi- tudinally at mitosis, they do not appear to be trans- versely divisible. It is possible that in some cases V-shaped chromosomes have two spindle attach- ments in the middle only separated by a very short CHROMOSOMES AND EVOLUTION 95 interstitial region ; in this ease the two spindle attachments may be expected to function mechanic- ally as a single unit, as in the case of A scar is megalo- cephala. Breakage of the interstitial region will give two clu'omosomes with quasi-terminal spindle attachments (' rod-shaped ' chromosomes). Con- versely two chromosomes with quasi-terminal spindle attachments may fuse together so as to give a V with two attachments in the middle. Any other kind of fusion will give a chromosome with two widely separated spindle attachments which will break at anaphase (Fig. 6) and any other kind of breakage will give rise to a fragment with no spindle attachment, which consequently cannot form an independent chromosome. Thus aj)art from these two special cases it does not seem possible that * fusion and fragmentation ' have played any part in the evolution of new chromosome numbers. In some groups of animals the chromosome number is fairly constant, while in others it varies from species to species. Thus nearly all Urodeles have a diploid number of 24, while in the Crickets (Gryllodea) thirteen different chromosome numbers have been found, ranging from 12 to 30 (in the female diploid set).^"*- In some cases, where it is due to the pro- cesses (hvscHbed above, the origin of new chromosome numbers is clear. Thus most species of Acrididae (Locusts and Grasshoppers) have 11 pairs of rod- shaped autosomes with quasi-terminal spindle attach- ments, but in the genera Chorthlpjms, Stenobothrus and Stauroderus there are only S pairs, 3 of which are V-shaped, with median or submedian spindle attachments. The South American species Alexias vitticolUs has 9 pairs, two of which are V's.^^^ It is clear that here originally separate chromosomes with subterminal spindle attachments have fused together so as to decrease the number of autosomes. Whether the V-shaped autosomes in these genera have two spindle attachments situated very close together 96 THE CHROMOSOMES is not certain, but appears probable ; the only alternative would be to suppose that one spindle attachment had been lost in each case as a result of a deletion '. In the first case the formation of V's would be a reversible phenomenon, in the second it obviously would not . The only other aberrant chromo- some number in the Acrididae is found in the species Podisma mikado which has only 10 pairs of rod-shaped autosomes. ^^^ In view of the remarkable constancy of the chromosome set in the whole family this is very surprising, particularly as the related species P. sapporoense appears to have the usual number of 11 pairs of autosomes. Whether a whole pair of autosomes (possibly genetically inert ?) has been lost in P. mikado J or whether all of it except the spindle attachment has been translocated to another pair of autosomes and the spindle attachments lost cannot be decided at present. If one examines a number of groups of animals one finds that there is a correlation between variation in size and variation in number of chromosomes which is sufficiently striking to be significant. Thus in groups such as the genus Drosophila and in the Lacertilia there are nearly always a number of very small chromosomes in the set which may be only just above the limit of resolution of the microscope (like the IVth chromosome in D. melanogaster ^ the Vth in D. pseudo-obscura, the Vlth in D. virilis and the microchromosomes of Lacertilia and Birds). Usually, but not always, there is a sharp difference in size between the microchromosomes and the ordinary ones, with no intermediates (as in the genus Drosophila). A particularly interesting case is that of the Urodela ; the majority of these (all species of Triton J Salamandra, Amphiuma, &c.) have 24 large chromosomes with median or submedian spindle attachments (Fig. 4). Where, as in the genera Crypto- branchus, Megalobatrachus and HynobiuSy ^^^' ^^^' '* there are more than 24 the extra ones are small CHROMOSOMES AND EVOLUTION 97 microchromosomes. In view of what has been said above about spindle attachments, it is clear that the microchromosomes must have arisen as duplications of the spindle attachment regions of the larger chromosomes. We have already seen that in some and perhaps in all species of Drosophila the regions round the spindle attachments are genetically inert ; it seems at least possible that in other groups with microchromosomes the same is the case ; if this were so no upset of the ' genie balance ' would result when a spindle attachment region was reduplicated in the chromosome set. Perhaps in groups like the Acrididae, with a remarkable constancy of chromo- some number, the regions round the spindle attach- ments are genetically active. The importance of microchromosomes lies in the fact that, once they have arisen, one or more translocations from other chromosomes may (convert a microchromosome into one of normal size. Duplication of a whole chromosome (polysomy) or even duplication of a relatively large region probably always upsets the ' genie balance ' and leads to the production of an organism which is less viable than the original one. It is obvious that the condition of maximum unbalance is reached when half the total chromosome set has been reduplicated (i.e. when the somatic chromosome number is mid-way between diploidy and triploidy).^^ It seems clear that in actual fact only very small duplications are likely to become established as new additions to the chromosome set. That they really do so is proved conclusively by the work of Bridges on the salivary gland chromosomes.^* He found that a number of small regions which could be recognized by the charac- teristic sequence of bands were repeated or duplicated. There are several possible ways in which this may occur ; if we consider a region involving four bands, p q r s, these may be repeated in the same chromo- some in any of the following four ways : 98 THE CHROMOSOMES ( 1 ) abcdefghijklmnopqrspqrstuvwxyz (2 ) abcdefghijklmnopqrssrqptuvwxyz (3) abcdefpqrsghijklmnopqrstuvwxyz (4) abcdefsrqpghijklmnopqrstuvwxyz The ' reversed duplications ' (types 2 and 4) appear to be commoner than the ' direct ' ones (types 1 and 3), but the data are not as yet sufficient to form an opinion as to whether there is any special significance in this. In addition to duplications within the same chromosome, duplications of small regions probably also occur in other non-homologous chromosomes of the same set. Duplication of the spindle attachment region is obviously a phenomenon of an entirely different nature to duplication of other regions ; if it takes place without the region in question being attached (either terminally or interstitially) to a pre-existing chromosome it will lead directly to the production of a new microchromosome as an addition to the chromosome set. It is this process which may have led to the appearance of microchromosomes in many groups of animals. On the other hand, if it happens in such a way that the duplication is at first attached to one of the original chromosomes (or inserted into it in an interstitial position) a chromo- some with two spindle attachments will have arisen. Such a chromosome will break at about half the mitoses in each cell-generation. There is no reason why it should break at the junction between the duplicated and the non-duplicated region and conse- quently this process, while increasing the number of the chromosomes by one, will not necessarily give rise to a microchromosome, but may equally well give rise to a chromosome of a relatively large size. Bridges' discovery (if it applies to animals in general, and not only to Drosophila melanogaskr) means that so-called diploid species are really only partially diploid, a certain number of chromosome CHROMOSOMES AND EVOLUTION 99 regions being present in the tetraploid condition. It should eventually be possible to ascertain the diploid- tetraploid ratio in different species of Drosophila. The evolutionary significance of duplications is, as Bridges points out, that they offer ' a method of evolutionary increase in length of chromosomes with identical genes which could subsequently mutate separately and diversify their effects '. It has already been pointed out that mutations which may be lethal when homozygous in a diploid are often quite viable when present twice in a tetraploid. Thus mutations will be more likely to establish themselves in the duplicated part of the chromosomes than in the remainder. The diploid-tetraploid ratio may thus be one of the factors determining the rate of evolution in a species or group. Perhaps the shark Scaphano- rhynchus owstoni which has existed unchanged from the Upper Cretaceous to the present day has almost no tetraploid regions in its chromosome set, and perhaps duplications may be only very rarely pro- duced in some groups (or be almost always lethal, which comes to the same thing). Formation of Species in the Genus Drosophila The genus Drosophila has now been investigated sufficiently for it to be possible to make some definite statements as to how new species have arisen (the problem of species-dichotomy as distinct from the problem of species -differentiation). In the first place the chromosome number varies considerably from species to species (Table X). Those species with a high proportion of rod-shaped chromo- somes have the highest numbers, those (like D. ananassae) with all the chromosomes V-shaped have the lowest numbers. But it is clear from the Table that not all the changes in chromosome number can be accounted for by fusion of rods to form V's and vice versa ; even if we assume that all V's have two spindle attachments there is still a variation in m^ VI n D. melanogaster (triploid) XY D. funebris w 1 ^^^n Race A Race B D. pseudo-obscura D. miranda m . VI XY D. virilis X Race A Race B D. montium Fig. 20. — Somatic chromosome sets of various species of Drosophila ; all male except the triploid D. melanogaster. Genetically active parts of chromosomes black, inert parts stippled. Inversion between D. melanogaster and D. simulans and between the A and B races of D. pseudo- obscura labelled 1, 2, 3, 4. . . . In D. montium the extent of the inert regions is not known so that they have not been indicated. The microchromosomes have been drawn as if they had median spindle attachments, but nothing is really known as to the exact location of the spindle attachments. CHROMOSOMES AND EVOLUTION 101 number of spindle attachments of 10 to 16 in the genus. The majority of species have one pair of microchromosomes, but not all ; no species has more than one pair of them. There is so far no means of ascertaining whether the pair of microchromosomes is always homologous throughout those species in which it is found. It is worth while concentrating attention on two groups of species, (1) that which includes D. melano- gaster and D. simulans, (2) the obscura group, which includes a large number of forms of which only D. sub-obscura, D. miranda and the various races of D. pseudo- obscura and D. affinis have been adequately investigated. D. melanogaster and D. simulans have quite clearly not evolved very far beyond the point at which one or the other separated off as an incipient species. The differences between their chromosomes have already been described (page 73). The question arises as to whether the original cause of species- dichotomy was the inversion in the Ilird chromosome, or whether this arose later, when the two forms were already separated by the sterility-barrier in the Fj hybrids. This question cannot be answered at present. It is, however, extremely interesting that the gene -differences between the two species (as revealed by the sequence of ' bands ' in the salivary gland chromosomes) are not distributed at random along the chromosomes, but are concentrated in several quite short regions. These regions are (1) the IV th chromosome, (2) a region in the Ilird chromosome between the inversion and the spindle attachment, (3) the distal end of the X-chromosome. Now, although the frequency of mutation varies from one gene to another, there is no reason to believe that whole regions of the chromosomes are more mutable than others (apart from the difference be- tween active and inert regions). We are thus once more driven to the same conclusion as we already 102 THE CHROMOSOMES reached on entirely different grounds, namely that mutations in some regions are less lethal than in others. One may at least tentatively suggest that the regions where the differences between melanogaster and simulans lie were originally duplications which have become more and more unlike one another in the course of the differentiation of the two species. The chromosome set of the species in the obscura group is entirely different from that of the melano- gaster group (Fig. 20). It does not appear possible to homologize any large region of the chromosomes of pseudo-obscura with a corresponding region in melanogaster or simulans. So many rearrangements of the genetic material have taken place since the evolutionary separation of these two groups of species that no large region of any chromosome has been unaffected by inversions, duplications, trans- locations, &c. The American species pseudo-obscura consists of two main * incipient species ' which have been called Race A and Race B. They can be crossed but their hybrids are sterile *^ (the male hybrids completely so, the female hybrids almost completely). Physiologic- ally Races A and B are thus almost as distinct as melanogaster and simulans ; it is only the absence of definite taxonomic differences which prevents them from being regarded as separate species. They do, however, show difference in rate of development. The main cytological differences between the two races are six inversions (four in the X-chromosome and one each in Chromosomes II and III) ; these inversions are shown in Fig. 20. Each of the two races is sub- divided into sub -races which are also distinguishable by inversions in various chromosomes, but the intra- racial inversions are not identical with the inter-racial ones. The sub-races can also be distinguished by differences in the length of the Y- chromosome. In Race A four types of Y have been discovered. One of these is a chromosome with a subterminal spindle CHROMOSOMES AND EVOLUTION 103 attachment (Fig. 20), another is a smaller chromo- some with a median spindle attachment, the other two are chromosomes with submedian spindle attach- ments. Within Race B three types of Y-chromo- some have been found, all with submedian spindle attachments (one of these may or may not be the same as one of the types found in Race A). It is not possible to state exactly how these different types of Y-chromosome have arisen, but it is clear that rearrangement of sections of the chromosome has taken place, probably involving both duplication of short lengths and actual loss (' deletion ') of certain portions. Since in all races and sub-races the Y appears to be genetically inert the loss of portions of it would not affect the genie balance. Drosophila pseudo-obscura thus appears to be breaking up into a number of ' incipient species '. Two of these are already effectively separated by the barrier of hybrid sterility, and each of these is in its turn splitting up into a number of sub-races, which, while not yet inter -sterile, will probably eventually become so. Ordinary gene -mutation appears to have played little part in the separation of all these forms which lends support to the view that it is more important as a factor in species-differentiation than in species-dichotomy. When we come to D, miranda we find a species closely related to pseudo-obscura, but separated from it by some remarkable cytological differences. ^^ The mode of sex-determination in miranda is unique among the species of Drosophila in that it involves two pairs of chromosomes, the males being X^ Xg Y, the females Xj Xg X^ Xg. The X^ of miranda clearly corresponds to the X of pseudo-obscura, the sequence of bands in both its arms being similar. The X2 chromosomes, of which there are two in the female miranda, but only one in the male, corresponds to the Ilird autosome in pseudo-obscura, and appears to have become involved as part of an entirely novel 104 THE CHROMOSOMES mechanism of sex-determination. There is appar- ently a non-random distribution of the sex-chromo- somes at the first meiotic division in the male so that only two kinds of gametes (X^ Xg and Y) are formed. *i Although all the chromosomes of miranda can be homologized in a general way with those ofpseitdo- obscura, a large number (100 at least) of rearrange- ments in the sequence of the genes have taken place. The majority of these are inversions, but some trans- locations of small regions from one chromosome to another have also been established. It is natural that translocations should be rarer in evolution than inversions, since they lead to the production of hypo- and hyper- diploid individuals, whilst inversions do not upset the ' genie balance '. There are indications that the arrangement of the bands in the salivary gland chromosomes of miranda is more similar to that in the A race of pseudo-obscura than to that in the B race. As in the case of the melanogaster-simulans pair of species the differences between the arrangement of the bands in the microchromosomes are very great. The group of species related to Drosophila affinis includes at least seven species ; they are fairly closely related to the pseudo-obscura forms, but will not interbreed with them ; the chromosomal diflferences have not been fully worked out as yet. The species D. montium contains two definite races (A and B) and there are probably more. The two known races differ in regard to the IVth pair of auto- somes which are V-shaped in Race A and rod-shaped in Race B ; it seems clear that one arm of the IVth chromosome has been completely lost in Race B ; possibly it is genetically inert in Race A.^'' The two races of montium can clearly be regarded as ' incipient species \ To sum up : we can say that the process of species- dichotomy appears to be going on at a considerable rate in Drosophila, so that each species is breaking up into a number of * incipient species * distinguished CHROMOSOMES AND P.VOLrTTON 10r> by differences in the sequence of the genes in the chromosomes, and separtated })y more or less complete Kterility-barriers. Conclusions In tlie liglit of modern cytoloory and genetics Darwin's stat(*ment that ' varieties are incipient species ' must be modificnl to n^ad ' some varieties are incipient species '. We have seen that in all the types of reproductory mechanism the primary origin of new species lies in some accident in the chromo- some set. The occurrence of such an accident is entirely unconnected with natural selection, but if it TABLE VIII Freqi UENCY OF DIFFERENT Haploid Numbers IN Metazc )A" No. of Hapk»i(l Haploid No. of Ilaploid No. of No. Sj)ecie8 No. Species No. Species 1 1 18 . 40 37 . 2 . 10 19 . 24 38 . 3 . 22 20 . 12 40 . 4 . 38 21 . 16 42 . 5 . 31 22 8 49 . 6 . . 105 23 . 13 52 . 7 . 68 24 . 21 56 . 8 . 63 25 . 8 58 . 9 . 44 26 . 5 60 . 10 . 57 27 . 6 62 . 11 . 58 28 . 22 84 . 12 . 128 29 . 13 87 . 13 . 41 30 . 26 98 . 14 . 35 31 . 44 100 . 15 . 19 32 . 6 104 . 2 16 . 46 33 . 1 17 . . 14 Total 34 . of even nu 3 mbers : 644 Total of odd niir Altogeth nbers : or : 426 1,070 106 THE CHROMOSOMES is to give rise to an ' incipient species ' it must not be lethal or upset the genie balance of the organism too profoundly. If, however, the original balance is only slightly upset subsequent mutation and/or recom- bination of existing genes may re-establish a new ' secondary balance '. While the origin of * incipient species ' is thus independent of natural selection, the latter is an important factor in the subsequent evolution of taxonomic differences between the new form and the original one (' species-differentiation '). Gene -mutation remains the only fundamental mech- anism of morphological change in organisms, but it TABLE IX Frequency oi ^ DIFFERENT Haploid Numbers IN Phanerogams ^^ Haploid No. of Haploid No. of Haploid No. of No. Species No. Species No. Species 3 . 5 21 . 64 40 . 5 4 42 22 . 25 41 . 1 5 . 27 23 . 8 42 . 6 6 . 134 24 . 80 45 . 8 7 . 236 25 . 3 46 . 1 8 . 332 26 . 20 48 . 4 9 . 170 27 . 31 50 . 3 10 . 126 28 . 24 51 . 1 11 70 29 . 4 52 . 1 12 . 391 30 . 11 55 . 1 13 . 30 31 . 3 56 . 2 14 . 125 32 . 25 57 . . 1 15 . 27 33 . 3 60 . 2 16 . 153 34 . 3 65 . 1 17 . 48 35 . 3 72 . . 1 18 . 58 36 . 19 100 . . 2 19 . 22 38 . 5 20 . 47 39 . 1 Total of even nu mbers : 1,646 Total of odd nur Altogeth nbers : er : 768 2,414 CHROMOSOMES AND EVOLUTION 107 only rarely acts as the primary origin of a now incipient species. Most so-called diploid organisms (such as the vast majority of animal species) are really only partially diploid and that part of their gene-complex which is tetraploid is possibly less subject to the conserv^ative effect of natural selection and is consequently in more active evolution than the rest of the genes. In bisexual animals, where polyploidy of whole chromo- some sets is excluded by the sex-determining mechanism the reduplication of small segments of TABLE X ^^' ^*' ®^' ®^' ^^^' ^^^ Diploid Chromosome Sets in the genus Drosophila ((^c?) (V = a chromosome with median or submedian spindle attachment, I = one with a quasiterminal attachment, m = a microchromosome, A = the number of spindle attachments on the assumption that V's have only one, B on the assumption that they have two.) Species X Y I Autosomes A 10 B D. affinis V 3 prs. I, 1 pr. m 11 D. ananassae V V 3 prs. V 8 16 D. funebris . I I 4 prs. I, 1 pr. m 12 12 D. hydei V I 4 prs. I, 1 pr. m 12 13 D. melanogaster . I V 2 prs. V, 1 pr. m 8 13 D. miranda . V, I V 2 prs. I, 1 pr. m 8 10 D. montium (Race A) I V 3 prs. V 8 15 D. montium {Race B) I V 2 prs. V, 1 pr. I 8 13 D. pseudo-obscura (Race A) . . . V V 3 prs. V, 1 pr. m 10 12 D. psevdo-obacura (Race B) . . . V V 3 prs. I, 1 pr. m 10 12 D. repleta V I 4 prs. T, 1 pr. m 12 13 D. robiista V V 1 pr. V, 1 pr. I, 1 pr. m 8 12 D. simulans . I V 2 prs. V, 1 pr. m 8 13 D. sulcata V V 2 prs. V, 1 pr. m 8 14 D. virilis I I 4 prs. I, 1 pr. m 12 12 D. willistoni V V 1 pr. V, 1 pr. I 6 10 108 THE CHROMOSOMES chromosomes is thus of great significance, both in the origin of incipient species and in providing new ' raw material ' for evolution. The loss of portions of chromosomes from the set must be assumed to have occurred in evolution to compensate for the acquisition of new ' duplicated ' segments ; otherwise the size of the chromosomes would have increased indefinitely in the course of evolution. Such loss probably takes place in two stages ; the regions in question first become pro- gressively inert as a result of successive mutation and then get lost as a result of a ' deletion ' (deletions of active regions, however small, seem to be nearly always lethal in Drosophila when homozygous). We know next to nothing of the physical basis of inert- ness, but the mutation of an active gene to an inert condition seems to differ from an ordinary mutation in being irreversible. In bisexual animals duplica- tion is the only way whereby the total number of genes can be increased, deletion the only way it can be reduced. GLOSSARY Alio -polyploid : An organism with more than two haploid sets of chromosomes which have been derived from two or more ancestral species, by hybridization. Cf. auto -polyploid . Amitosis : A form of nuclear division in which no spindle mechanism is present, so that the chromosomes are not necessarily equally distributed to the two resulting nuclei. Anaphase : The stage of mitosis w^hich follows on metaphase and precedes telophase. During anaphase the chromo- somes move from the central region of the spindle towards the poles. Aneuploid : An organism in whose somatic chromosome set one or more chromosomes are represented more times than the rest ; consequently an irregular polyploid. Auto-polyploid : An organism with more than two haploid sets of chromosomes which have been deriv^ed from tho same parent species. Autosome : Any chromosome which is not a sex-chromosome. Bivalent : Two homologous (or at any rate partly homologous) chromosomes which have paired at meiosis and are held together, either by a mutual attraction or by chiasmata. {Trivalent : a similar association of three chromo- somes ; quadrivalent, an association of four ; multivalent, an association of more than two chromosomes.) {Unequal bivalent : a bivalent in which one of the constituent chromosomes is longer than the other, so that it has an unpaired region at one end.) Centromere : See Spindle attachment. Chiasma : A visible change of pairing affecting two out of the four chromatids in a bivalent at meiosis ; an out- ward sign that a genetical cross-over has taken place. (Chiasma -frequency : the average number of chias- mata formed in a particular chromosome or in a par- ticular organism under given circumstances.) (Terminal Chiasma : an association of four chroma- tids end-to-end which results from the shifting of a chiasma until it reaches the end of the bivalent or multi- valent — see Fig. 14c.) 109 110 THE CHROMOSOMES Chromatid : A longitudinal half of a chromosome in prophase or metaphase ; the two chromatids of each chromosome separate from each other at the anaphase of mitosis, so that a telophase chromosome consists of one chromatid only. Chromomere : A granule on a prophase chromosome at mitosis or meiosis, or on a salivary gland chromosome. Chromomeres are now believed to be identical with genes. Chromosome : At prophase and metaphase of mitosis, two chromatids and a spindle attachment ; at meiosis half a bivalent. {Branched Chromosome : a chromosome in which the chromatids fork dichotomously, either at the spindle attachment or elsewhere.) {Ring Chromosome : a chromosome in which the two ends are fused together, in such a way that a con- tinuous circle results.) Diakinesis : The last part of the prophase of the first meiotic division, between diplotene and prometaphase. See Fig. 9e. Differential Segment : A segment of a chromosome which is not present in another chromosome that is otherwise homologous. The opposite to pairing segment. Diploid Set of Chromosomes : A group of chromosomes which can be divided into two equal haploid groups. {Diploid organism : an organism which has a diploid set of chromosomes in each of its somatic cells.) Diplotene : The stage in the prophase of the first meiotic division which follows on pachytene and precedes diakinesis. See Fig. 9d. Fixation : The process of killing and coagulating a cell by means of some chemical or physical agency. {Fixable : a nucleus which can be killed and coagu- lated without seriously altering its visible morphology ; unfixable : a nucleus which is seriously altered in structure by the process of fixation.) Heterogametic : Producing gametes of more than one kind, which differ as to the chromosomes which they contaiii. The opposite to homogametic. Heteropycnosis : the property of contracting or condensing at a different rate from the majority of the chromosomes in the nucleus. {Negative heteropycnosis : condensing more slowly than the other chromosomes do.) {Positive heteropycnosis : condensing faster, earlier or more completely than the other chromosomes.) Homogametic : Producing gametes which are all alike as to the chromosomes which they contain. The opposite to heterogametic. GLOSSARY 111 Homologous (as applied to chromosomes) : Containing the same genes in the same sequence. Inert Chromosome : A chromosome all or most of whose genes are physiologically inactive. Interference : The process by which the occurrence of one cross-over or chiasma reduces the probability of another taking place in its immediate neighbourhood. Interkinesis : The resting stage which often occurs between the end of the first meiotic division and the beginning of the second. Interphase : See Interkinesis. Inversion : A section of a chromosome which is reversed in comparison with the usual sequence. Kinetochore : See Spindle Attachment. Leptotene : The earliest part of the prophase of the first meiotic division, before pairing of the chromosomes has taken place. Meiosis : Two modified mitoses in the course of which the chromosomes only divide once. The two divisions are called the first and second meiotic divisions. Metaphase : The stage of mitosis which follows prophase or pro-metaphase and precedes anaphase ; .when the spindle attachments of the chromosomes are lying in approximately one plane, the equatorial plane. Microchromosome : A chromosome which is considerably smaller than the other members of the set, e.g. the IVth pair of chromosomes in Drosophila melanogaster . Multivalent : A group of more than two chromosomes which are held together at meiosis by mutual attraction or by chiasmata. Nuclear Sap : The substance inside the nucleus, in which the chromosomes lie. Pachytene : The middle part of the prophase of the first meiotic division, when the pairing of the chromosomes is complete. Pachytene may be subdivided into 2 -strand pachytene (before the chromosomes have split) and 4-strand pachytene (after they have split). Pairing : The approximation of genetically homologous genes, chromomeres or chromosomes, considered either statically or dynamically. Somatic pairing : the more or less complete pairing of homologous chromosomes which is sometimes found at mitosis. Zygotene pairing : the pairing of chromosomes which takes place at the zygotene stage. Pairing Segment (of a sex-chromosome) : A segment or short portion of a chromosome which undergoes pairing with a corresponding segment in another chromosome. Polyploid : An organism with more than two baploid sets of chromosomes in its somatic cells. 112 THE CHROMOSOMES Polyaomy : A condition in which one or more chromosonies, but not the entire set are present in the polyploid state. Pycnoais : A condition in which all the chromosomes of a nucleus have fused together to form a single mass — occurs only in moribund cells. Quadrivalent : A multivalent composed of four chromosomes. Ring Chromosome : A chromosome in which the two ends have fused together so that it forms a continuous circle. Rotation {of chiaamata) : The relative rotation of the four ' arms ' of a bivalent on either side of a chiasma, which often occurs between early diplotene and diakinesis. Salivary Oland Chromosomes : Chromosomes in the nuclei of the salivary gland cells in Diptera. These chromo- somes have undergone complete somatic pairing ; consequently what is ordinarily called a saUvary gland chromosome is thus really two chromosomes fused side by side. Spindle Attachment : A special region of the chromosome by which the rest of the chromosome is attached to the spindle at metaphase and anaphase. The spindle attachment is a special ' organ ' of the chromosome which can be seen at all stages of mitosis and meiosis, under favourable conditions. It does not divide at the same time as the rest of the chromosome. Spindle Elements : The elements of which the spindle is prob- ably composed and which usually correspond in number to the chromosomes. {Central spindle element : a spindle element which is not related to any of the chromosomes and which forms the * core ' of the spindle surrounded by the other elements which lie parallel to it. Spindle Fibres : Fibres which were supposed to exist in the substance of the spindle, running from pole to pole or from pole to equator. Stem Body : The equatorial part of the spindle which elongates at anaphase and telophase, forming a long strand between the two resulting nuclei. Synchronous Mitosis : The occurrence of a number of cell divisions which take place at exactly the same time in a group of neighbouring cells. Terminalization : Shifting of chia^mata from their original positions towards the end of the bivalent. Tetrad : See Bivalent. Trivalent : A multivalent composed of three chromosomes. Univalent : A chromosome which has not undergone pairing at the zygotene stage or one in which pairing has come to an end due to failure of chiasma-formation. Zygotene : The stage of meiosis which follows on leptotene and precedes pachytene. The stage when pairing of the homologous ohromoisomes takes place. BIBLIOGRAPHY ^ Balzer, F. 1908. t)ber mehrpolige Mitose bei Seeige leiern. Verh. d. phys. med. Oes. Wiirzburg, 39. 2 Bauer, H. 1935. Der Aufbau der Chromosomen aus den Speicheldriisen von Chironomus Thummi Kiefer. Z. Zellf., 23, 280. ^ Behre, K. 1929. Physiologische und zytologische Un- tersuchungen iiber Drosera. Planta, 7, 208. * Belar, K. 1926. Der Formwechsel der Protistenkerne. Erg. Forschr. ZooL, 6, 235. ^ 1927. Beitrage zur Kenntnis des Mechanismus der indirekten Zellteilung. Naturwiss., 36, 725. • 1928. Die Cytologischen Orundlagen der Vererbung. Berlin, Borntraeger. 1929. Beitrage zur Kausalanalyse der Mitose. II. Untersuchungen an den Spermatocyten von Chorthip- pus (Stenobothrus) lineatus Panz. Arch. Entwickm., 118, 359. ® Bellamy, A. W. 1936. Interspecific hybrids in Platy- poeilus : one species ZZ-WZ ; the other XY-XX. Proc. Nat. Ac. Sci., 22, 531. • Belling, J. 1928. The ultimate chromomeres of Lilium and Aloe with regard to the number of genes. Univ. Cal. Publ. BoL, 14, 307. ^^ Bleier, H. 1930. Untersuchungen iiber das Verhalten der verschiedenen Kemkomponenten bei der Reduk- tionsteilung von Bastarden. Cellule, 40, 83. ^* 1931. Zur Kausalanalyse der Kernteilung. Oen- etica, 13, 27. ^2 Bonnevie, K. tlber Chromatin Diminution bei Nema- toden. Jena Z. Naturwiss., 36, 275. *^ Boveri, Th. 1909. Die Blastomerenkerne von Ascaris megalocephala und die Theorie der Chromosomenindi- vidualitat. Arch. Zellf., 3, 181. ^* Bridges, C. B. 1935. Salivary Chromosome maps. J. Hered., 26, 60. ^^ BuLLER, A. H. R. 1931. Researches on Fungi. IV. Observations on the Coprini, together with some investiga- tions on social organization and sex in the Hymenomy- cetes. London. 8 113 114 THE CHROMOSOMES 16 Bytinsky-Salz, H. 1934. Verwandschaftsverhaltnisse zwischen den Arten der Gattungen Celerio und Pergesa nach Untersuchungen iiber die Zytologie und Fertilitat ihrer Bastarde. Biol. Zentr., 54, 300. ^' Cannon, H. G. 1923. On the nature of the centrosomal force. J. Genet., 13, 47. 1* Carlson, J. G. 1936. The intergeneric homology of an atypical euchromosome in several closely related Acridinae (order Orthoptera). J. Morph., 59, 123. *• Carnoy, J. B. 1884. La Biologie Cellulaire (Fasc. 1). Lierre. *® Catcheside, D. G. 1931. Critical evidence of para- synapsis in Oenothera. Proc. Roy. Soc. B., 109, 165. *^ 1 932. The chromosomes of a new haploid Oenothera. Cytologia, 4, 68. ** Chambers, R. 1924. General Cytology (edited by Cowdry). Chicago. ** 1925. !6tudes de microdissection. IV. Les struc- tures mitochondriales et nucleaires dans les cellules germinales males de la Sauterelle. Cellule, 35, 105. ** Chambers, R.,and Renyi, G. S. 1925. The structure of the cells in tissues as revealed by microdissection. I. The physical relationships of the cells in epithelia. Am. J. Anat.y 35, 385. •* Crew, F. A. E. 1933. A case of non-disjunction in the Fowl. Proc. Roy. Soc. Edinburgh, 53, 89. *• Darlington, C. D. 1929. Meiosis in Polyploids. II. J. Genet., 21, 17. ^' 1929. Chromosomal behaviour and structural hy- bridity in the Tradescantiae. J. Genet., 21, 207. ** 1930. A cytological demonstration of genetic Crossing-Over. Proc. Roy, Soc. B., 107, 50. " 1931. The cytological theory of inheritance in Oenothera. J. Genet., 24, 405. *" 1932. The origin and behaviour of Chiasmata. V. Chorthippus elegans. Biol. Bull., 63, 357. ^^ 1932. Recent Advances in Cytology . (First Edition.) London and Philadelphia. ** 1934. Anomolous chromosome pairing in the male Drosophila pseudo-ohscura. Genetics, 19, 95. '* 1935. The time, place and action of crossing-over. J. Genet., 31. 185. ** 1935. The Internal Mechanics of the Chromo- somes. I, II, and III. Proc. Roy. Soc. B., 118, 33. — 1936. The limitation of Crossing-over in Oenothera. J. Genet., 32, 343. SK BIBLIOGRAPHY 116 3* Darlington, C. D. 1936. Recent Advances in Cytology. (Second Edition.) London. *' 1930. Crossing-Over and its Mechanical relation- ships in Chorthipptis and Stauroderus. J. Genet. y 33, 465. ^8 1936. The Internal Mechanics of the Chromo- somes. V. Relational coiling of the chromatids at mitosis. Cytologia, 7, 248. ^* Darlington, C. D., and Dark, S. O. S. The origin and behaviour of chiasmata. II. Stenohothrus parallelus. Cytologia, 3, 169. *** Dearing, W. H. 1934. The material continuity and individuality of the somatic chromosomes of Ambly- stoma tigrinum with special reference to the nucleolus as a chromosomal component. J. Morph., 56, 157. *^ Dobzhansky, Th. 1934. Studies on hybrid sterility. I. Spermatogenesis in pure and hybrid Drosophila pseudo-obscura. Z. Zellf., 21, 169. *2 1935. Drosophila 7nira7ida, anew sipocies. Genetics^ 20, 377. *3 Dobzhansky, Th., and Tan, C. C. 1936. Studies on hybrid sterility. III. A comparison of the gene arrangement in two species, Drosophila pseudo-obscura and Drosophila miranda. Z. ind. Abst. u. Vererbl., 72, 88. ** DouTRELiGNE, J. 1933. Chromosomes et nucleoles dans les noyaux du type euehromocentrique. Cellule, 42, 31. *^ DozoRCHEVA, R. L. 1936. The morphology of chromo- somes in the Ichneumon, Pteromalus puparum. C. R. Acad. Sci., U.R.S.S., 1936 (3). ** Dubinin, N. P., and others. 1936. Occurrence and distribution of chromosome aberrations in nature. Nature, 137, 1035. *' Du Bois, A. M. 1933. Chromosome behaviour during cleavage in the eggs of Sciara coprophila (Diptera) in relation to the problem of sex-determination. Z. Zellf. t 19, 595. *^ DusTiN, A. p. 1921. Deolenchement experimental d'une onde cinetique par injection intraperitoneale de serum. C.R. Soc. Biol., 85, 23. *' Emmens, C. W. Unpublished work. ^^ Ernst, A. 1918. Bastardierung als Ursache der Apogamie im Pflanzenreich. Eine Hypothese zur experimentelle Vererbungs- und Abstammungslehre. Jena. *^ Fasten, N. 1914. Spermatogenesis of the American Crayfish, Cambarus virilis and Cambanis immunis (?) with special reference to synapsis and the chromatoid bodies. J. Morph., 25, 587. 116 THE CHROMOSOMES ** Fern ANDES, A. 1931. Estudos noa cromosomas das Liliaceas e Amarilidaceas. Bol. Soc. Broteriana (Coim- bra) 7, 1. *' Friesen, H. 1934. Kiinstliche Auslosung von Crossing- Over bei Drosop/iiZa-Mannchen. Biol. Zentr., 54, 65. •* Fbolova, S. L. 1936. Struktur der Keme in den Speich- eldriisen von Drosophila sulcata Stert. Bull. Biol, et de Med. Exp., 2, 93. •^ GowEN, J. W. 1933. Meiosis as a genetic character in Drosophila melanogaster. J. exp. Zoo., 65, 83. " GowEN, J. W., and Gay, E. H. 1933. Gene number, kind and size in Drosophila. Genetics, 18, 1. *' Gray, J. 1927. The mechanism of cell division. IV. The effect of gravity on the eggs of Echinus. Brit. J. Exp. Biol, 5, 102. " Greqor, J. W., and Sansome, F. W. 1930. Experi- ments on the genetics of wild populations. II. Phleum pratenae L. and the hybrid between P. pratense L. and P. alpinum L. J. Genet., 22, 373. *• Gross, F. 1935. Die Reifungs- und Furchungsteilungen von Artemia salina im Ziisammenhang mit dem Problem der Kemteilungsmechanismus. Z. Zellf., 23, 522. •® Gross, R. 1917. Beobachtungen und Versuche an lebenden Zellkernen. Arch. Zellf., 14, 279. •^ Haldane, J. B. S. 1 931 . The cytological basis of geneticfiJ interference. Cytologia, 3, 34. •' Harrison, J. W. H., and Doncaster, L. 1914. On hybrids between moths of the Greometrid Sub-Family Bistoninae with an account of the chromosomes in gametogenesis in Lycia (Biston) hirtaria and Ithysia {Nyssa) zonaria and their hybrids. J. Genet., 3, 229. •' Hartmann, M. 1929. Verteilimg, Bestimmung und Vererbung des Geschlechtes bei den Protisten und Thallophyten. Handb. der Vererbungswiss., 2. •* Hearne, E. M., and Huskins, C. L. 1935. Chromosome pairing in Melanoplus femur-rubrum. Cytologia, 6, 123. •^ Heitz, E., and Bauer, H. 1933. Beweise fiir die Chro- mosomennatur der Kernschleifen in Knaulkemen von Bibio hortulanus. Z. Zellf., 17, 67. •• Hoar, C. S. 1927. Chromosome studies in Aesculua. Bot. Gaz., 84, 156. •' 1931. Meiosis in Hypericum punctatum Lam. Bot. Gaz., 92, 396. •* HuGHES-ScHRADER, S. 1927. Origin and differentiation of the male and female germ cells in the hermaphrodite of Icerya purchasi (Coccidae). Z. Zellf., 6, 509. •• 1930. The cytology of several species of iceryine Coccids with sp>ecial reference to parthenogenesis and haploidy. J. Morph., 50, 475. BIBLIOGRAPHY 117 '*• HuGHES-ScHRADER, S. 1931. A study of the chromo- some cycle and the meiotic division figure in Llaveia boiivari, a primitive coccid. Z. Zellf., 13, 742. '^ 1935. The chromosome cycle of Phenacoccus (Coccidae). Biol. Bull, 69, 462. '2 HusKiNS, C. L. 1931. The origin of Spartina Town- sendii. Genctica, 12, 531. '3 HusKiNS, C. L., and Smith, S. G. 1935. Meiotic chromo- some structure in Trillium erecttmi L. Afin. Bot., 49, 119. '* Iriki, Sh. 1932. Studies on amphibian chromosomes. VII. On the chromosomes of Megalohatrachus japon- icus. Sci. Rep. Tokyo Bunrika Daigaku., B.l, 91. '^ IsAWAKi, Y. 1925. Sur le declenchement experimental des ondes de cinese dans le sang de quelques insectes. Ann. de Physiol, et de Physiochem. Biol., 1, 580. '« Janssens, F. a. 1924. La chiasmatypie dans les in- sectes. Cellule, 34, 135. '' Jaretsky, R. 1928. Histologische und karyologische Studien an Polygonaceen. Jahrh. wiss. Bot., 69, 357. '® Jensen, H. W. 1936. Meiosis in Rumex. I. Poly- ploidy and the origin of new species. Cytologia, 7, 1. '^ J0RGENSEN, C. A. 1928. The experimental formation of heteroploid plants in the genus Solanum. J. Genet., 19, 133. ®" Junker, H. 1923. Cytologische Untersuchungen an den Geschlechtsorganen der halbzwitterigen Stein- fliege, Perla niarginata (Panzer). Arch. Zellf., 17. ** Karpetchenko, G. D. 1927. The production of poly- ploid gametes in hybrids. Hereditas, 9, 349. ®^ 1927. Polyploid hybrids of Raphanus sativus L. Brassica oleracea. L. Bull. Appl. Bot., 17, 305. ®^ KiHARA, H. 1931. Genomanalyse bei Triticum und Aegilops. II. Aegilotriticum und Aegilops cylindrica. Cytologia, 2, 106. 8* KiHARA, H., and Ono, T. 1926. Chromosomenzahlen und systematische Gruppierung der i^wmea;- Arten. Z. Zellf., 4, 475. ®*KiKKAWA, H. 1935. An inference as to the constitution of X-chromosome in Drosophila. Proc. Imp. Acad., 11, 62. ®* 1936. ChromosoTneH of Drosophila ananassae. Jap. J. Gen., 12, 65. ®' 1936. Two races of Drosophila montium. Jap. J. Gen., 12, 137. ®^ King, R. L. 1932. Chromosomes of three species of Mantidae. J. Morph., 52, 525. *• Klingstedt, H. 1931. Digametie beim Weibchen der Trichoptere Limnophilus decipiens. Kol. Acta ZooL Fennica, 10, 118 THE CHROMOSOMES ^^ KoEPERiCH, J. 1930. ^^tude comparative du noyau, des chromosomes et de leur relations avec le cytoplasme. Cellule, 39, 307. •1 Roller, P. C. 1932. Further studies in Tradescantia virginiayia var. humilis and Rhoeo discolor. J. Genet., 26, 81. *2 1935. The internal mechanics of the chromosomes. IV. Pairing and coiling in salivary gland nuclei of Drosophila. Proc. Roy. Soc., B, 118, 371. ®3 1936. The genetical and mechanical properties of sex-chromosomes. II. Marsupials. J. Genet., 32, 451. »* KoLLER, p. C, and Darlington, C. D. 1934. The genetical and mechanical properties of the sex-chromo- somes. I. Rattus norvegicus ^. J, Genet., 29, 159. ®* KoLTZOFF, N. K. 1934. The structure of the chromo- somes in the Salivary glands of Drosophila. Science, 80, 312. ^* KuwADA, Y., and Nakamura, K. 1934. Behaviour of chromonemata in mitosis. II. Artificial unravelling of coiled chromonemata. Cytologia, 5, 244. *' KuwADA, Y., and Nakamura, T. 1935. Behaviour of chromonemata in Mitosis. III. Metaphasic and ana- phasic longitudinal split of chromosomes in the homo- type division in pollen mother cells in Tradescantia. Cytologia, 6, 314. •* La-Cour, L. 1935. Technic for studying chromosome structure. Stain Technology, 10, 57. '• Levan, a. 1932. Cytological studies in Allium. II. Chromosome morphological contributions. Hereditas, 16, 257. ^^^ Lewis, W. H., and Lewis, M. R. 1924. Behaviour of cells in tissue cultures. General Cytology. Chicago. ^^^ LiLiENTELD, F. A. 1936. Karyologische und genetische Studien an Fragaria. II. Ist Fragaria elatior sine autopolyploide Pflanze ? Jap. J. Bot., 8, 119. ^®2 LuDFORD, R. J. 1936. The action of toxic substances upon the division of normal and malignant cells in vitro and in vivo. Arch. exp. Zellf., 18, 411. ^^^ McClintock, B. 1932. A correlation of ring-shaped chromosomes with variegation in Zea mays. Proc. Nat. Ac. Sci., 18, 677. 1^* 1933. The association of non-homologous parts of chromosomes in the mid-prophase of meiosis in Zea mays. Z. Zellf., 19, 191. 105 1934. The relation of a particular chromosomal element to the development of the nucleoli in Zea mays. Z. Zellf., 21, 294. HIHLIOCRAPHV 119 I "* Maeha, T. \\)'M). ()i» i\\o cotifiguratioiis of gcmiiii in tho pollen inothor cells of V'irid Juba L. Mrui. Coll. Sci. Ki/otO'B, 5, 12."). ®' Makinc), S. 1932. An unocjual ])air of idiochroniosomos in the Tree Cricket, Occanthus longicaiuUi Mats. J . Fac. »SV/., Hokkaido Imp. Univ., Ser. VI, 2, 1. "8 WY.Vl. Not es on the chromosomes of 7ia/o>-r//- /a L. and Bijo sdchalinciisis. Nikolski. Proc. Imp. Acrid., 8, L>3. "8 1934. The chromosomes in lltpwhins leech/i and H. nebidosus. Trans. Sapporo Nat. Hist. Soc, 13, liol. 1" \\)\\~). Tiie chromosomes of Cryptobranchus alle- gheniensis. J. JMorph., 58, 573. 11 11)3(1. Tho spiral structure of chromosomes in the meiotic division o'i Podisma (Orthoptera). J. Fac. Sci., Hokkaido Imp. Univ., Ser. VI, 5, 29. 12 M ANTON, I. 1935. Some new evidence on the physical nature of plant nuclei from intra-specific polyploids. Proc. Roy. Soc, B, 118, 522. 13 1935. The cytological history of Watercress [Nasturtium officinale R.Br.). Z. ind. Abst. VcrcrbL, 69, 132. 1* Marshall, W. S. 1907. Amitosis in the malpighian tubules of the Walking Stick. Biol. Bull., 14, 89. 15 Mather, K. 1933. Interlocking as a demonstration of the occurrence of genetical crossing-over during chiasma-formation. Am. Nat., 67, 47t5. 16 1935. Meiosis in Lilium. Cytologia, 6, 354. 17 Mather, K., and Stone, L. H. A. 1933. Tho effect of X-radiation upon somatic chromosomes. J. Genet., 28, 1. 1® Matsuura, H. 1935. Chromosome studies on Trillium kamtschaticum. II. The direction of coiling of the chromonoma within the first meiotic chromosomes in the P.M.C. J. Fac. ScL, Hokkaido Imp. Univ., Ser. V, 3, 233. i» Metcalf, M. M. 1923. The Opalinid Ciliate Infusorians. U.S. Nat. Mus. Bull, No. 120. 20 1935. The germ-cell cycle in Phytophaqa destructor Say. Q. J. Micr. Sc, 11, 285. 21 Meurman, O. 1933. Chromosome morphology, somatic doubling and secondary association in Acer platanoidcs L. Hereditas, 18, 145. 22 Metz, C. W. 1914. Chromosome studies in the Diptera. I. A preliminary survey of five different types of chromosome groups in the genus Drosophiia. J. Exp. Zool.y 17, 45. 120 THE CHROMOSOMES 2' Metz, C. W. 1916. Chromosome studies in the Diptera. II. Additional types of chromosome groups in the Droso- philidae. Am. Nat., 50, 587. 2* 1933. Monocentric mitosis with segregation of chromosomes in Sciara and its bearing on the mechan- ism of mitosis. Biol. Bull., 54, 333. 2* Metz, C. W., Moses, M., and Hoppe, E. 1926. Chromo- some behavior and genfetic behavior in Sciara. I. Chromosome behavior in the spermatocyte divisions. Z. ind. Abst. Vererbl., 42, 237. *• MoFFETT, A. A. 1932. Studies on the formation of multinuclear giant pollen-grains in Kniphofia. J. Genet., 25, 315. 2' Morgan, L. V. 1933. A closed X-chromosome in Dro8ophila melanogaster. Genetics, 18, 250. 28 Morgan, T. H., Bridges, C. B. and Sturtevant, A. H. 1925. The genetics of Drosophila. Bihl. Gen., 2, 1. *• MoRGULis, S. 1911. Studies of inanition in its bearing on the problem of growth. Arch. Entwickm., 8, 448. *° MtTLLER, H. J. 1925. Why polyploidy is rarer in anim- als than in plants. Am. Nat., 59, 346. " MuLLER, H. J. and Painter, T. S. 1932. The differen- tiation of the sex chromosomes of Drosophila into genetically active and inert regions. Z. ind. Abst, Vererbl, 62, 316. '2 Muller, H. J., Prokofieva, a., and Raffel, D. 1936. Minute intergenic rearrangement as the cause of apparent gene-inutation. Nature, 135, 253. '* Muntzing, a. 1930. Outlines to a genetic monograph of the genus Galeopsis. Hereditas, 11, 267. ** Nachtsheim, H. 1913. Cytologische Studien iiber die Gesehlechtsbestimmung bei der Honigbiene {Apis mellifica L.). Arch. Zellf., 11, 169. ^^ Navashin, M. 1930. Unbalanced somatic chromo- somal variation in Crepis. Univ. Gal. Publ. Agr. Sci., 6, 95. '• 1932. The dislocation hypothesis of evolution of chromosome numbers. Z. ind. Abst. Vererbl., 63, 224. " Nebel, B. R., and Ruttle, M. L. 1935. Chromosome structure in Tradescantiae. VIII. The direction of coiling in Tradescantia refleoca Raf. as related to the mode of crossing over. Gytologia, 6, 457. ''^ Newton, W. C. F., and Darlington, C. D. 1929. Meio- sis in polyploids. I. J. Genet., 21, 1. '• NiiYAMA, H. 1935. The chromosomes of the edible Crab, Paraliihodes camtschatica (Tilesius). J^ Fac» Sci., Hokkaido Imp. Univ., Ser. VI, 4, 69, BIBLIOGRAPHY 121 1" Oguma, K. 1934. Studies on the Sauropsid chromo- somes. II. The cytological evidence proving female heterogamety in tlie hzard {Lacerta vivipara). Arch. Biol., 45, 27. **^ 1932. On the chromosomes of throe species of Gryllodea. Proc. hup. Acad. (Tokyo), 8, 197. ^*2 Ohmachi, F. a. 1935. A comparative study of chromo- some complements in the Gryllodea in relation to taxonomy. Bull. Mie. Imp. Coll. Agr. and For., No. 5. 1*3 OsTERHOUT, W. J. V. 1897. Vber Entstehung der karyokinetischen Spindcl hci Equisetum. 1** Pariser, K. 1927. Dio- Zytologio und Morphologie der triploiden Intersexe des riickgekreutzten Bastards von Saturnia pavonia und S. pyri. Z. Zellf., 5, 415. 145 Patau, K. 1935. Chromosomenmorphologie bei Droso- phila melanogaster und Drosophila simulans und ihre genetische Bedeutung. Naturwiss, 23, 537. 1*' 1936. Cytologische Untersuchungen an der hap- loidparthenogonetischon Milbe, Pediculoides ventricosus. Newp. Zool. Jahrb. {allg. ZooL), 56, 277. 1*' PoLiTZER, G. 1932. Pathologie der Mitosc. Berlin, Borntraegor. *" Prokofikva, a. 1934. On the chromosome morphology of certain Pisces. Cytologla, 5, 498. 1** Prokofieva-Belgovskaja, a. a. 1935. The structure of the chromocentor. Cytologla, 6, 438. ^^^ Reuter, E. 1930. Beitriigo zu oiner einheitlichen Auf- fassung gewisser Chromosomenfragen. Acta Zool. Fennica, 9, 1. "1 RiLfiY, H. P. 1936. The effect of X-rays on the chromo- somes of Tradescantia glgantea. Cytologla, 7, 131. 1^2 RoBYNS, W. 1929. La figure achromatique sur material frais, dans les divisions somatiques des Phanerogames. Cellule, 34, 365. 153 Rosenberg, O. 1927. Dio semiheterotypische Teilung unh ihre Bedeutung fiir die Entstehung verdoppelter Chromosomenzahlen. Hereditas, 8, 305. 15* Ruckert, J. 1892. Zur Entwicklungsgeschichte des Ovarialcies bei Selachiern. Anat. Anz., 7, 107. 155 Saez, F. a. 1931. Cromosomas multiples en Aleuas vltticolUs. Rev. d. Mus. de la Plata, 33, 189. 156 Sakamt^ra, T. 1927. Fixiorung von Chromosomen mit siedendom Wasser. Bot. (iaz., 41, 59. 157 Sanderson, A. R. 1933. The cytology of partheno- genesis in Tenthredinidae. St. Andrews Univ. Puhl., No. 33. The Hague, M. Nijhoff. 158 Sansome, F. W., and Philp, J. 1932. Recent Advances in Plant Genetics. London, Churchill. 122 THE CHROMOSOMES 1^* Sato, D., and Sinoto, Y. 1935. Chiasma Studies in Plants. III. Chromosome Pairing and Chiasma be- haviour in the male Rumexacetosa with special reference to the tripartite sex-chromosome. Jap. J. Genet. , II, 219. ^" ScHRADER, F. r.)23. Haploidie bei einer Spinnmilbe. Arch. mik. Anat.y 79, 61 J). *^ 1928. Die Geschlechtschromosomen. Berhn, Born- traeger. *^ 1929. Experimental and cytological investigations of the life-cycle of Oossyparia spuria (Coccidae) and their bearing on the problem of haploidy in males. Z. wias. Zool.y 134, 149. ^* 1935. Some notes on the behavior of long chromo- somes. Gytologia, 6, 422. •* Scott, A. C. 1936. Haploidy and aberrant spermato- genesis in a Coleopteran, Micromalthus debilis Le Conte.« J. Morph., 59, 485. •^ Seiler, J. 1920. Greschlechtschromosomenuntersuch- ungen an Psychiden. I. Arch. Zellf.^ 15, 249. '^ 1925. Ergebnisse aus Kreutzungen von Schmetter- ILngsrassen mit verschiedener Chromosomenzahl. Arch. Jul. Klaus Stijtung, 1, 63. •' SoKOLOW, N. N., and Trofimow, I. E. 1933. Individu- alitat der Chromosomen im.d Geschlechtsbestimmung beim Haushuhn. Z. ind. Abst. Vererhl., 65, 327. *® Stern, C. 1927. Ein genetischer und cytologischer Beweis fiir die Vererbung im Y-Chromosom von Droso- phila melanogaster. Z. ind. Abst. Vererbl., 44. ** Stevens, N. M. 1909. Further studies on the chromo- somes of the Coleoptera. J. Exp. ZooL, 6, 101. "* Taylor, W. R. 1931. Chromosome studies on Gasteria. III. Chromosome structure during microsporogenesis and the postmeiotic mitosis. Am. J. Bot., 18, 367. '^Thomsen, M. 1927. Studien iiber dieParthenogenesebei einigon Cocciden und Aleurodiden. Z. ZeUf.y 5, 1. '2 Torvik-Greb, M. 1935. The chromosomes of Habro- bracon. Biol. Bull., 68, 25. '* VoRHiES, C. T. 1908. The development of the nuclei in the spinning gland cells of Platyphylax designatus (Trichoptera). Biol. Bull., 15, 54. '*Wada, B. 1933. Mikrurgische Untersuchungen lebender Zellen in der Teilung. I. Gytologia, 4, 114. '^ 1934. Uber die Entstehung der Vakuolen im Kern. Gytologia, 5, 248. '• Wenrich, D. H. 1916. The spermatogensis of Phryno- tettix magnus with special reference to synapsis and the individuality of the chromosomes. Bull. Mus. Com/p. ZooL Harvard, 60, 67. BIBLIOGRAPHY 123 ^" White, M. J. D: 1932. The chromosomes of the Domes- tic Chicken. J. Genet., 26, 345. ^'* 1933. Tetraploid spermatocytes in a Locust, Schistocerca gregaria. Cytologia, 5, 135. 1^' 1935. Eine neue Form von Tetraploidio nach Rontgenbostrahlung. Naturwiss., 23, 390. ^'" 1935. The effects of X-rays on mitosis in the spermatogonia! divisions of Locusta migratoria L. Froc. Roy. Soc, B, 119, 6L *8^ 1936. Chiasma-locaUsation in Mecoslcthus grossus L. and Metrioptcra brachyptera L. Z. Zellf., 24, 128. *•* 1936. The chromosome cycle of Ascaris niegalo- cephala. Nature, 137, 783. **' Unpublished results. 184 Whitney, D. D. The chromosome cycle in the Rotifer, Asplanchna ampfiora. J. Morph., 47, 415. 185 Wilson, E. B. 1905. Studies on Chromosomes. I. The behavior of the idiochromosomes in Hemiptera. J. Exp. ZooL, 2, 371. **® 1905. The paired microchromosomes, idiochromo- somes and heterotropic chromosomes in Hemiptera. J. Exp. ZooL, 2, 507. *®' 1925. The Cell in Development and Heredity. 1*® 1932. Polyploidy and metaphase patterns. J. Morph., 53, 443. 189 WiNGE, 0. 1935. On haplophase and diplophase in some SaccJuiromycetes. C. R. Lab. Carlsberg, 71, 77. Yamamoto INDEX \ Acaiina, 42 Acrididae, 43, 95 .4 rg ilatritic u rn , 9 2 Ae/iculufi earnca, 92 Aleuas vUticollis, 9') AlcuroJidar, 42 Allo-polyploid}^ 45 AlyJus, 79 Aniitosis, 7 Anaphase, of mitosis, 23 of first meiotic division, 64 of second meiotic division, 67 Anasa tristis, 79 Aneuploidy, 46 Annelida, 42 Arachnida, 42 Archimerus, 79 Artemia salina, 13 Ascaris lumhricoide.s, 32 megalocephala, 11, 29, 30 Ascomycetes, meiosis in, 47 Asplanchna, 42 Asters, 13 Auto-polyploidy, 45 Autosomes, 39 Basidioraycetes, 47 Belar, 9 Bee, 81 Bibio, 35 Biscutella laevigata, 45 Biston hirtarius, 17 Biv^alent chromosomes, at meiosis, 54 Branched chromosomes, 32 Breakage (of chromosomes), 92 Cabbage, 92 Campanula, 88 Celerio euphorbia e, 74 gain, 74 Central spindle, 14 Centrosome, 13, 14 Charices, 47 Chiasma, 56 compensating, 73 diagonal, 72 formation and crossing- over, 69 frequency, 56 (localized), 71 reciprocal, 72 Chironomus, 35 ChorthippuSy 95 Chromatid, 11 interference, 73 Chromocentre, 35, 39 Chromomeres, in salivary chromosomes, 38 at meiosis, 51 Clonal organisms, Coccidae, 42, 43 Coleoptera, 42 Complex heterozygotes, 84, 90, 93 — meiosis in, 84 — origin of, 93 Crossing-over, 69 Crustacea, 42 Darwin, 105 Datura, 84 Deletion, 52, 108 Diakinesis, 62 Differential segment (of sex- chromosome), 41, 80 125 126 THE CHROMOSOMES Diploid chromosome number, 27 Diplotene, 55 Drosophila ajffinis, 101, 107 ananassae, 107 funebriSf 100 melanogaster, 18, 32, 40, 70, 73, 80, 101 miranda, 100, 103 montium, 100, 104, 107 jjseudo-obscura, 100, 101, 102 repleta, 107 simulans, 73, 101 8ub-obscura, 101 sulcata, 107 virilia, 100, 107 Duplication, 97, 108 Echinodermata, 42 Elodea, 42 Empetrurriy 42 Evolution and chromosome numbers, 91, 105, 106 and origin of new species, 99 and structural alterations in chromosomes, 89, 97, 98 Female heterogamety, 42 Fixability, 8, 9 Fragaria elatior, 42 Fragmentation of chromo- somes (in Ascaris), 29, 95 (by X-rays), 18 Fritillaria, 71 Frog, 5 Fusion of chromosomes, 30, 94 Oaleopsis Tetrahit, 92 Gene-mutations (nature of), 89 Genie balance, 97 Gryllodea, 95 Habrobracon, 44 Half-spindles, 16, 19 Haploids, meiosis in, 81 Haploidy, male, 42 Heterogametic sex, 39 Heteropycnosis, 43 Homogametic sex, 39 Humulus, 42 Hybridization in genus Drosophila, 102 Hybrids, meiosis in, 73 Hymenoptera, 42 Hypericum punctatum, 84, 93 Icerya purchasi, 9 Inert chromosomes, 34 Incipient species, 102, 105 Interference, 72 Intranuclear mitosis, 13 Inversions, 74 Lacertilia, 96 Lepidoptera, 42, 94 Leptotene, 51 Lethal deletions, 108 Liliaceae, 51 Lilium, 69 Llaveia bouvari, 15 Localization of chiasmata, 70 Loss of parts of chromo- somes, 108 Lygaeus, 80 Maize, 69, 71 Mantis, 41, 81 Marsupials, sex-chromosomes in, 80 Mecostethus, 70, 71 Melandrium, 42 Melanoplus femur-rubrum, 73 Metaphase, of mitosis, 14 of first meiotic division, 63 of second meiotic division, 67 Metrioptera, 9 Microchromosomes, 96 Micrmnalthus debilis, 42 Multipolar spindles, 19 Multivalents, 82 INDEX 127 Xasturtium officinale, 92 Nematoda, 42 Nuclear membrane, 1 sap, 1 Nucleolus, 1 Nucleus, shape of, 2 Nyssia zonaria, 17 Oecanthus longicauda, 17 Oenothera, G6 Oligochaeta, 94 Opalinidae, 5 Opisthogoneata, 42 Pachytene, 2 -strand, 54 4-strand, 55 Pairmg, at zygotene, 51 somatic, 34 in complex heterozygotes, 85 in salivary gland chromo- somes, 35 between X- and Y-ehromo- somes, 80 segment, 80 Perla marginnta, 41 Phleum, 92 Photinus, 79 Phytophaya destructor, 11 Pisum, 88 Podisma, 96 Polyploidy, 29 Polyploids, meiosis in, 82 Polysf)my, 45 Popuhts, 42 Prochromosomes, 4 Prometaphase, of mitosis, 13 of first meiotio division, 62 Prophase, of mitosis, 8 of first meiotic division, 49 Protenor, 79 Protozoa, 7, 13, 30 Pteridophytes, 47 Pulmonata, 94 Pycnosis, 6 Raddish, 92 Raphanobrassica, 92 Rat, sex-chromosomes in, 79, Rhoeo discolor, 93 Rhynchota, 80 Ring- chromosomes, 32 Rotation of chiasmata, 59 Rotifera, 42 Rumex, 45, 46 Rumex acetosa, 41 Rye, 92 Salamandra, 16 Salivary Gland Chromo- somes, Saturnia pavonia, 74 pyri, 74 Scale insects, 13 Scaphanorhynchus owstoni, 99 Schistocerca, 79 Sciara coprophila, 15, 76 Selective fertilization, 44 Sex-chromosomes, at mitosis, 39 — at meiosis, 78 Spartina torvrisendi, 92 Spindle, 13 attachment, 17 elements, 13 fibres, 19 Spiral structure of chromo- somes, 21, 63 Spireme, 9 Stem body, 24 Stp.nobothras, 69 Synchronous mitosis, 5 Telophase, 24 Terminalization, 59 Thysanoptera, 42 Trade scantia, 25 Trichoptera, 42 Trisomy, 45 Trivalents, 81 Univalents, behaviour of, 75 Quadrivalent, 82 Vicia faba, 56 128 THE CHROMOSOMES Wheat, 92 X-rays, 6, 18 Y-chromosome, at mitosis, 40 X-chromosomes, at mitosis, at meiosis, 79 40 at meiosis, 78 Zygotene, 51 Printed in Great Britain by Butler & Tanner Ltd.. Frome and London