A CYTOLOGICAL AND GENETICAL STUDY OF TRIPLOID MAIZE BAB~A~A MCCL~NT~~K Botany Department, Cornell University, Ithaca, N. Y. Received July 11, 1928. TABLE OF CONTENTS hTRODUCTlON ........................................ Material and methods. ................................. A brief resume of meiotic behavior in diploid individuals. Meiosis in the triploid individual. ....................... Plaamodial sporogenous tissue. .......................... Chromatin extrusion in sporogenous tissue ................ TheF~generation ..................................... Meiosis in Fr individuals. .............................. Meisosis in In+ 1 individuals. ........................ Unusual meiotic phenomena in 2n+ 1 individuals. Meiosis in 2n+2 individuals. ......................... Meiosis in plants with 2n+3 to 2n+ 7 chromosomes. .... Fr triploid plant ..................................... Irregular synapsis in a 2n+5 individual. ............... Genetic investigations .................................. D~scussro~ ........................................... SPRY ............................................. hT?ZRATURE CITED. ................................... ....... ....... ....... ....... ....... ....... ....... ....... ....... ....... ....... ....... ....... ....... ....... ....... ....... ....... ........ ........ ........ ........ ........ ........ . ........ ........ ........ ........ ........ ........ ........ ........ ........ ........ ........ PAGE . 180 . 181 . 182 . 183 . 199 200 201 . 205 . 206 209 . 209 . 212 . 213 . 213 . 214 . 219 . 220 . 221 INTRODUCTION During the course of investigation on the chromosome number and behavior of different genetic strains of maize a triploid individual appeared in an otherwise diploid culture. As has been reported in a preliminary note (RANDOLPH and MCCLINTOCK 1926), this individual was notably more vigorous than its diploid sibs. Since the triploid possessed certain known genetic characters it was desired to follow the genetics along with the cytology in the descendants of this individual. In the preliminary note it was suggested that the triploid probably arose through the fusion of a diploid and a haploid game,te. Since the pub- lication of this note much evidence has accumulated which would support this probability (see pp. 209 and 209; see also figures 17 and 27). The literature on Zea chromosomes is rapidly becoming more extensive. Since Miss FISK (1927) has reviewed most of the existing literature on the subject (RUWADA 1911,1915,1919,1925; LONGLEY 1924,192s; FISK 1925; KIESSELBACH and PETERSEN 1925) little more need be said with regard to CBNErrCS 14: 180 Mr 1929 TRIPLOID MAIZE 181 previous investigations on Zea. To this list we might add the short paper of REEVES 1925, in which it is stated that of 8 varieties of maize examined, 7 showed 10 pairs of chromosomes during meiosis; the eighth, a Black Mexican variety, on the other hand, showed 12 pairs. LONGLEY (1927) and RANDOLPH (1928) in recent investigations have extended considerably our knowledge of the behavior during meiosis and the distribution through gametes of such "extra" chromosomes. It is possibly unjust to restrict the term triploid to those 3n forms which show trivalents during meiosis. However, in the 3n Zea plants showing 10 trivalents, behavior during meiosis can be compared directly &ith other * 3n forms which show trivalents. The literature on such types is growing rapidly. Under the heading of triploids showing trivalents would come some individuals of Morus (OSAWA 1920), Canna (BELLING 1921, 1925), Datura (BELLING et al.), Hemerocallis (BELLING 1925), Hyacinthus (BELLING 1925), Campanula (GALRDNER 1926), Lycopersicum (MANN- LESLEY 1926) Zeax Euchlaena F1 (LONGLEY 1924; Randolph unpublished), Drosophila (METZ 1925; MORGAN 1925 : BRIDGES 1921, 1922, 1925) and Zea. The genetics of triploids has been followed in the case of Datura, Droso- phila and Lycopersicum. For the literature on trisomic inheritance, see discussion under Genetic Investigation. MATERIALANDMETHODS The behavior of chromosomes during meiosis, exclusive of the early prophase stages, has been studied in many diploids, two triploids and SO F1 individuals resulting from direct.and reciprocal crosses made between diploid individuals and the triploid plant (see table 1). Studies were made only on microsporcytes. For the study of chromosome number and behavior during meiosis the iron-aceto-carmine method was found to be very serviceable. In some cases the smears were dried slowly and then mounted in balsam. Photomicro- graphs 14 and 15, in plate 6 and text figure 7 were made from preparations seven months old which had been so treated. In some cases where the stain was too heavy it was extracted in a weak solution of acetic acid, the whole slide being warmed to hasten the extraction. Other sporocyte material and root tips were fixed in a Bouin solution modified according to Allen, mounted in paraffin and sectioned lo-15 microns thick. Heidenhains iron-alum-haematoxylin and Flemming's triple stain were used. GENETICS 14: Mr 1929 182 BARBARA McCLINTOCK The photomicrographs were taken with the aid of a Leitz "Makam" camera, an 8 X periplanatic ocular, and a 1.8mm, N.A. 1.25 Spencer achro- matic objective. A Wratten green filter, number 56, was used to increase definition. A BRIEF RlhJMi OF MEIOTIC BEHAVIOR IN DIPLOID INDIVIDUALS At diakinesis the diploid individuals show characteristically 10 pairs of chromosomes. At the time the nuclear membrane disappears the bivalent nature of the chromosomes becomes relatively indistinct because the chro- mosomes are so contracted and closely appressed. However, after the bi- valent chromosomes line up at the equator and show evidence of commen- cing anaphasic disjunction, characteristic forms are assumed (see bivalent chromosomes in metaphase figures in plates 2-S). The double (dyad) nature of each chromosome of the pair is clearly indicated at early anaphase, frequently before the chromosomes have completely disjoined. It is evident from a study of metaphase I', early anaphase I, and metaphase II that the spindle fiber attachment is either sub-terminal, sub-median or median. In consequence of partial anaphasic separation of members of the dyads the chromosomes appear as double Vs (medianorsub-median attachments, or as almost single Vs (sub-terminal attachments). This picture varies considerably in different anaphases due to the unusual compactness of the dyads in some figures and to their loose, long-armed form in other figures (figure 34). In the latter case one or more arms of a dyad may become much pulled out, remaining near the equator during the continued pole- ward advance of the spindle attachment part of the chromosome. The telophase passes over into a more or less typical interphase wherein the individual chromosomes lose their observable identity. A cell plate, apparently initiated by the phragmoplast, is evident at this time,. From the very earliest prophase of II in which the chromosomes can be recognized as distinct bodies, their form is very characteristically that of an H or an X. Contraction of the chromatic material continues until the chromosomes become relatively small Hs and Xs (figure 31). The nuclear membrane shrinks about the chromosomes with the formation of a multipolar spindle. This, in turn, passes into a bipolar spindle with the disappearance of the nuclear membrane. The shape of the metaphase II chromosomes depends greatly upon the amount of contraction of the chro- mosomes at this time. In some cases the dyads appear as relatively long, double threads. On the other hand, if contraction of the chromosomes 1 The first meiotic mitosis will be designated as I, the second meiotic mitosis as II. TRIPLOID MAIZE 183 continues, many of these dyads appear as double `dumb-bells", the dumb- bell shape probably being due to an accentuated constriction (see SAKA- MURA 1920; GOTOH 1924). Apparently the anaphase II chromosomes derive their varied shapes, such as bent rods in some sporocytes, or dumb- bell-like bodies in others, from the state of contraction of the chromosomes at metaphase II. As a result of division II a quartet of spores is formed. Cell plate formation apparently initiates wall formation as has been shown by REEVES (1928). In most cases meiosis is thus characteristically regular, but several types of behavior having much theoretical interest have been observed. Occa- sionally and sometimes rather frequently in certain diploid cultures a pair of chromosomes showed non-conjugation as described by BELLING (1925) and others. Many of the diakinesis figures within an anther contained distinctly 10 bivalent chromosomes, while other sporocytes within the same anther showed 9'r+2 I.2 Whether there was never any meiotic association of the 2 homologous chromosomes or whether the disassociation of a synapsed pair to form 2 univalents took place in late prophase I is unknown. Neither is it known whether the same homologous pair was concerned each time a figure with 9rr+2r was observed. The cause of this appearance is certainly not a lack of homology of the chro- mosomes in all cases (RANDOLPH 1928). Other irregularities such as non-disjunction of a pair of chromosomes, lagging and splitting of univalents at I, two-nucleated sporocytes, meta- phase I figures with more or less than the expected chromosome number, and metaphases and anaphases with multiples of the diploid chromosome number have been observed in diploid individuals. MEIOSIS IN THE TRIPLOID INDIVIDUAL A brief description of meiosis in a triploid plant has been given (RAN- DOLPH and MCCLINTOCK 1926); therefore certain features stressed in that paper will not be redescribed here. Diakinesis stages in aceto-carmine in the triploid gave far better results than similar stages in fixed and sectioned material. In the former method the synaptic condition, that is, whether trivalent, bivalent, or univalent, of all the chromosomes in a single nucleus usually could be determined (figure l)," whereas, in the section only a few of the chromosomes within z Bivalents will be designated by the exponent Roman numeral II; univalents, exponent I; trivalents, exponent III, etc. s Figures in plates l-5 are numbered from l-47. The photo-micrographs in plate 6 are numbered I-15. To avoid confusion, "plate 6" will always be added to the figure number when reference is being made to the photo-micrographs. (hiETlCS 14: Mr 1929 184 BARBARA McCLINTOCR EXPLANATION OF PLATES Figures in plates l-5 were made with the aid of an Abbe camera lucida. Approximate mag- nifications: Figures i-24,26-29,3246: 1200x. Figures 25,30,3 1: 500x. Figure 47: 700x. Figures in plates 1 and 6 are from aceto-carmine smears and sectioned material. Figures in plates 2-5 from aceto-carmine smears only. The small numbers at left of figures indicate the culture and plant number. PLATE 1. FIGURE 1 .-Diakinesis in microsporocyte of triploid (56J, showing 10 trivalents. .4ceto- carmine. FIGURE 2.-Diakinesisin triploid, showing 9 trivalents, 1 bivalent and 1 univalent. Sectioned material. FIGURE 3.-Metaphase Z in triploid, showing 10 trivalents. Aceto-carmine. FIGURE 4.-Metaphase Z in triploid. 10 trivalents. Sectioned material. FIGUFCE 5.-Metaphase Z in triploid. 9 trivalents, 1 bivalent and 1 univalent; univalent pre- paring to split and separate while trivalents and bivalents disjoin. Sectioned material. FIGURE 6.-Metaphase Z in triploid. 7 trivalents, 3 bivalents and 3 univalents. 2 univalents in the spindle region nearer the lower pole, 1 in the cytoplasm outside the spindle. Sectioned material. FIGURE 7.-Metaphase Z in triploid. 8 trivalents, 1 bivalent and 1 univalent in main spindle; 1 trivalent in separate spindle running perpendicular to plane of section. See photograph of same: plate 6, figure 5. FIGURE 8.-Anaphase Z in triploid. 15 chromosomes going to each pole. Aceto-carmine. FIGURE 9.-Anaphase Z in triploid. 14 dyads at each pole; 2 lagging and separating in center of figure. Sectioned material. FIGURE IO.-Anaphase Z in triploid. 12 dyads in upper pole, 11 in lower pole and 7 separating dyads in center of spindle. See photograph of same plate 6, figure 6. Sectioned material. MCCLINTOCK, B., ~IPLOID MAIZE PLATE 1 6 GENETICS 14: 185 MI 1929 186 BARBARA &lcCLINTOCK PLATE 2 Figures from 2n+l Fr plants. FIGURE Il.-Metaphase I showing 9r1+1*r'. FIGURE 12.-Metaphase I from same plant showing 10r*+l'. The univalent is lying in the equatorial plate. FIGURE 13.-Same as figure 12, only the univalent is lying toward the upper pole. FIGURE 14, 1.5, and 16.-Metaphase I trivalents taken from different 9"+1'" figures arranged to show the various appearances of the same trivalent in different cells and the tri- valents from different 2n+ 1 plants. FIGURE If.-Metaphase I in a 2n+ 1 plant, showing the double number of chromosomes (20"+2') in one figure. FIGURE 18.-Metaphase I spindle with four times the normal number of chromosomes: 4orr+4r. FIGURE 19.-Three connecting metasphase I spindles in a plasmodial mass. FIGURE 20.-Anaphase Z figure in a 2n+ 1 plant, showing fragmentation of chromosomes in progress. MCCLINTOCK, B., TRIPLOID MAIZE PLATE 2 88 3 Gammcs 14: 187 Mr 1929 188 BARBARA McCLINTOCK PLATE 3 Figures from 2n+2 Fi plants. FIGURE Il.-Diakinesis: 91r+11*1+1r. FIGURE 22.-Late metaphase I: 811+2 rK1. The members of the trivalent to the right have nearly disjoined. FIGURE 23.-Metaphase I: 10"+2'. FIGURE 24.-Anaphase I: 10 dyads in upper pole, 11 toward lower pole, 1 dyad with separat- ing halves at center of spindle. 8 FIGURE 25.-Microspore quartet with unequal spores. FICUPE 26.-Metaphase I: 7rt+2tt*. This figure contains 1 bivalent less than most of the sporocytes in this plant. FIGURE 27.-Metaphase I: 411+6'v+2 v1. 2 large sexivalents above and below plate region; counting from the right in region of plate, the second, third, fourth, sixth, seventh and eighth. groups are quadrivalents, the remaining 4 chromosomes are bivalents. FIGURE 28.-Anaphase I in same plant with the double number of chromosomes. FIGURE 29.-Anaphase ZZ: 12 monads at each pole. FIGURE 30.-Microspore quartet resulting from Z and II. Note chromatic bodies in cell plate region. FIGURE 31.-Prophase II: 10 dyads in cell to left, 12 dyads in cell to right. PLATE 3 MCCLII-WXK, B., TRIPLOID MAIZE GENETICS 14: 189 MI 1929 190 BARBARA McCLIN 1'OCK PLATE 4 Figures from 2n+3 to 3n Fr individuals. FIGUBE 32.-Metaphase I: 7'r+3"*. Fxcurms 33.-Metaphase I: S1x+2111+11. Frcna~ 34.-Anaphase I in 2n+3 individual: 10 dyads toward upper pole, 11 towards lower pole, 2 splitting at center of spindle. FICU~E 35.-Metaphase I: 611+4111. FIGURE 36.-Metaphase I: 811-l-1111+41. FIGURE 37.-Metaphase I: 611+4111. See photograph of same, figure 14, plate 6. FIGURE 38.-Metaphase I: 31r+71r1. Members about to disjoin in 2 trivalents at left of figure. FIGURE 39.-Metaphase I: 611+4111. FIGURE 40.-Metaphase I: 101". MCCLINTOCK, B., TRIPLOID MAIZE PLATE 4 GENETICS 14: 191 MI 1929 192 BARBARA McCLINTOCK PLATE 5 All figures are from plant 942,2n+S. FIGURE 41.-Metaphase I: 5*1+5*11. FIGURE 42.-Metaphase I: 9I*+7*. FIGURE 43.-Diakfnesis: 41*+3"*+8*. FIGURE 44.-Division I: 25'. FIGURE 45.-Anaphase I: 10 univalents at upper pole; 11 at lower pole; l"-l-2' at center of spindle. FIGURE 46.-Early anaphase I: l"+ 1***+2Or. FIGUFCE 47.-Showing production of 6 unequal spores from a single microsporocyte. PLATE 5 MCCLINTOCK, B., TRIPLOID MAri% GENETICS 14: 193 Mr 1929 194 BARB.4R.A RIcCLINTOCK PLATE 6 Figures 1-13 are of sectioned and haematoxylin stained material. Figures 14 and 15 are of ceto-carmine smears, Xx.gnification, approximatelv 080x. FIGURE I.-Diakinesis in triploid (56,). One trivalent visible. FIGURE 2.-Late metaphase I in t:iploid showine disjunction of 3 members of a trivalent. FIGURE 3.-Same sporocyte as in figure 2, showing another trivalent. FIGURE 4.-Metaphxse Z in sporocyte of triploid, showing 3 trivalents in same focus. FIGURE .5.-Metaphase Z with 2 spindles. Small spindle to right of cell runs perpendicular to the plane of the section and contains one trivalent. See drawing of same in plate 1, figure 7. FIGURE 6.-Anaphase Z showing lagging and separation of halves of split univalents. Same as figure 10, plate 1. FIGURE f.-Prophase II in triploid. Spindles are forming about both dwarf and major nuclei. FIGURE 8.-Telophase of I; two large cells and one small cell. The nuclei of the two large cells are not in focus. FIGURE 9.-Teiophase of II resulting from a tripolar spindle. FIGURE IO.-Formation of "dwarf" nuclei from prophase Z nucleus. Arrow 1 points to 3 small nuclei in row. Arrow 2 points to a larger nucleus beside major nucleus. FIGURE Il.-Prophase Z. A chromatic thread apparently projecting through the nuclear membrane into the cytoplasm. FICURIX 12.-Early anaphase ZZ in triploid. Sporocyte possessing but one spindle contain- ing the double number of chromosomes. FIGURE 13.-Metaphase ZZ in triploid showing several spindles, variously oriented, with few chromosomes in each. FIGURE 14.-Metaphase 2 in 2n+4 individual (953) showing 6rr+4**`. Same as figure 3i, plate 4. FIGURE 15.-Metaphase Z in adjacent sporocyte to above showing 10*`+4t. M&LINTOCK, B., TRIPLOID XAIZE PLATE 6 3 __--.-- 12@ ., GENETICS 14: 195 Ilr 1929 TRIPLOID MAIZE 197 a nucleus could be easily interpreted as a general rule. Occasionally all the chromosomes in a single nucleus could be well interpreted (figure 2, plate 1; figure 1, plate 6). In many cases the chromosomes at diakinesis are arranged in ten groups of three each (figure 1). A definite and typical association of any 3 homo- logues, giving a constant, morphologically recognizable trivalent, apparently does not exist. This is more clearly evident in F1 individuals in which the single extra chromosome forms one trivalent, the different figures of which are directly comparable (see plate 2, figures 14, 15 and 16). We see here that the same trivalent both in diakinesis and metaphase I assumes various forms which are undoubtedly due to the relative closeness and type of approximation of the three members. Figure 2 illustrates an apparent tendency on the part of one of the homologous chromosomes to disassociate from the other two. In diakinesis all stages of approximation of the third chromosome are seen: a group of three equally spaced chromosomes; a group which would definitely be recognized as a bivalent with an attached univalent; a bivalent with the univalent rather distantly attached by a fine thread; and finally, an appar- ently complete disassociation of the univalent from the bivalent. It is obvious, therefore, that although distinct univalents are evident at diakinesis, this by no means indicates that synapsis among the three homologues had not taken place at an earlier prophase stage. FIGURE 1 .-Polar view of somatic metaphase from root tip of triploid plant. 3n=30. As has been stated, the proportion of trivalents, bivalents and univalents varies in adjacent sporocytes. The presence of 9 trivalents, 1 bivalent and 1 univalent was most usual, those with 2,3 and 4 univalents being less frequent. We observed in no case more than 4 univalents, that is, 6"`+4" +4'. It is not at all inconceivable, from what is known of the F1 indi- viduals, that greater disassociation took place but was unobserved in the material examined. Again, at metaphase 1 we find, very frequently, lOII*(figures 3 and 4). However, as in diakinesis, there are commonly present fewer trivalents with a consequent increase in bivalents and univalents (figures 5 and 6). GENE~CS 14: Mr 1929 198 BARBARA McCLINTOCK Sometimes no equatorial plate.was formed, the metaphase I chromosomes being distributed along the length of the spindle (figures 4 and 6). In position and behavior the univalents vary considerably. Sometimes they appear in or near the plate. Here they are easily distinguishable because of their smooth contour and their round or oblong shape. How- ever, very occasionally, a univalent lying in the plate exhibits distinct spindle fiber attachments (figure 5). Here the univalent is probably preparing to divide at the same time that the trivalents and bivalents are about to disjoin. This is in contrast to the usual appearance of the separa- tion of the members of a dyad in I which takes place in lagging chromo- somes after disjunction of the members of the trivalents and bivalents. Sometimes the univalents lie within the spindle area some distance from the plate on either side. It is not uncommon to find one or more univalents lying in the cytoplasm apart from the spindle (figure 6). Occasionally a sporocyte was observed with one trivalent lying in a separate spindle distinctly apart from the main metaphase I spindle (figure 7, and photograph of same, plate 6, figure 5). In other sporocytes the minor spindle was connected with the major spindle. The distribution of the chromosomes in anaphase I appears to be a random one. In many anaphase I figures the chromosomes passed to the poles without irregularities (figure 8). However, in many other figures marked irregularities were obvious, such as lagging of dyads, lagging and splitting of dyads (figures 9, 10; figure 6, plate 6) and splitting and partial separation of a dyad which had commenced to pass to one pole. Other irregularities, such as the fragmentation of part of a dyad as it passed to the pole, were numerous in some sporocytes of certain anthers. The extent of fragmentation within a sporocyte varied from a part of a single dyad to practically the whole complement. In consequence of the many irregularities observed in diakinesis, meta- phase I and anaphase I, the appearance of the sporocytes in telophase and interkinesis is diverse. In some we find two nuclei of approximately equal size between which a cell plate has formed. In figure 8, plate 6 is shown the formation of three distinct nucleated protoplasmic masses separated by cell plates formed in division 1 (nuclei of two larger protoplasmic masses not in focus). In other cases one or more small nuclei are not set off by cell plates from the major nuclei. It is conceivable that the condition depicted in figure 8, plate 6 resulted from a two-spindle sporocyte as in figure 5, plate 6, in which a minor spindle connects at right angles with the major spindle. In the other cases, however, the small nuclei may have TRIPLOID MAIZE 199 originated from univalents which lagged during anaphase I. Some sporo- cytes showed a great many nuclei grading in size from large to very minute. The very small nuclei probably represent reorganized chromatin fragments, since breaking up of the chromosomes during anaphase I is so obvious in some cells. During interkinesis the dyad and monad chromosomes become invisibIe as such and the nucleus takes on a metabolic appearance. As the prophase II progresses the characteristic X-, H- or rod (monad)-shaped chromosomes become evident. Frequently, in those cells which possess small nuclei besides a large one, a similar prophase II transformation appears to be going on within these small nuclei. Such is the case shown in figure 7, plate 6. The small nucleus, about which a spindle is forming, contains a dyad chromosome. Consequently more than two metaphase 11 spindles may occur within a single sporocyte (figure 13, plate 6). In both aceto- carmine and sectioned material sporocytes were seen with all the chromo- somes in a single metaphase II spindle (figure 12, plate 6). In general, the second meiotic mitosis appears much freer from irregu- larities than the first. There is very little fragmentation and compara- tively Iittle Iagging of chromosomes. In many cases the wall Iaid down during I disappears soon after its formation. In consequence, the spindles of II lie in a common mass of cytoplasm. Rather infrequently tripolar spindles were observed which had been formed from a partial fusion of the two metaphase II spindles (figure 9, plate 6). In consequence of the irregularities attending division I and II a single sporocyte can produce from two to many spores, the nuclei of which con- tain various amounts of chromatin. These spores, in turn, may be uni- nucleate or, as is frequently the case, multinucleate, the nuclei being varia- able in size and probably in chromosome content. The pollen grains exhibit such size relationships as would be expected from the above des- cription. PLASMODL4L SPOROGENOUS TISSUE In the triploid plant and its FL descendants the sporocytes within some anther loculi, instead of existing as well denned individual cells, showed various degrees of cytoplasmic fusion. All conditions from a partial union of a few adjacent cells to a compIete loss of identity of the sporocytes were observed, the result in the latter case being a plasmodium which tiled the entire anther cavity. In some plasmodia two or more nuclei were so close together that only a very thin layer of cytoplasm appeared to exist between them. In several instances two adjacent nuclei appeared to be fusing. GENrnCS 14: MI 1929 200 BARBARA McCLINTOCK The nuclei in the plasmodium continue their prophase meiotic develop- ment, forming well defined diakinesis and metaphase I figures (text figure 2). Probably as the result of the closeness of nuclei in these plasmodial masses, metaphase I spindles connect in various ways. It is conceivable that some of the metaphase I spindles exhibiting a double number of chromosomes have arisen in this manner. It is also conceivable that other metaphase I spindles possessing the double number of chromosomes, the homologous members of which, however, show a close synaptic attraction, have arisen by an earlier prophase fusion of nuclei, or possibly even by a pre-meiotic fusion. FIGURE 2.-Photomicrograph of plasmodium in anther showing 3 metaphase I spindles and one nucleus in late diakinesis. As a result of the first meiotic division the plasmodium becomes divided into smaller masses separated by w,ell defined cell plates. CHROYATIN EXTRUSION IN SPOROGENOUS TISSUE The breaking up of chromatin to form many small nuclei which become distributed in the cytoplasm was seen to occur in several stages of meiosis. Sporocytes in prophase I exhibited numerous figures of "cytomixis" as described by many authors. Another st,riking prophase I phenomenon was the presence of small supernumerary nuclei distributed in the cytoplasm or closely associated with the main nucleus (figure 10, plate 6). It is pas- TRIPLOID.MAIZE 201 sible, in some cases, that the small nuclei were formed from the main nucleus by budding (see SAKAMURA 1920). In other cases chromatin threads which have been seen extending through the nuclear membrane into the cytoplasm (figure 11, plate 6) may have given rise to some of these diminutive nuclei. Extensive chromosome fragmentation occurring during anaphase 1 was observed in a number of sporocytes. Many small nuclei were observed in some sporocytes which obviously had been formed from reorganized chromosome fragments. THE F1 GENERATION Because of the observed cytological behavior in the triploid individual, it was of great interest to follow the distribution of the chromosomes through future generations. Also, as little was known concerning the gene- etic constitution of the three chromosomes of an homologous group, crosses were made to determine this to some extent. In almost every case where crosses were made the chromosome number and meiotic behavior of the plants used had been determined (see table 1). This was obviously neces- sary, as LONGLEY (1927) and RANDOLPH (1928) have shown that in many genetic cultures one or more extra chromosomes are present which are car- ried on to future generations. Further indication that the extra chromosome in the F1 individuals originated from the gametes in the triploid was seen in the contrast between the appearance and behavior of these F1 individuals with extra chromo- somes and in those extra chromosome individuals within genetic strains of maize. The extra chromosomes (when more than one extra chromosome is present it seems to be due to a reduplication of the same chromosome) run- ning through some genetic strains of maize and found principally in Black Mexican sweet corn do not affect size, vigor or fertility to any noticeabIe degree. They are carried through in the eggs or the pollen. The extra chromosomes in the F1 plants of the cross triploid X diploid, on the other hand, affect size, vigor and fertility and are carried only to a small extent through the pollen. In a 2n+ 1 plant of Black Mexican the chromo- somes do not, so far as I am aware, form 9Ir+ lxlI, but always lO"+l'. This is quite in contrast with the 2n+ 1 F1 plants from the crossdiploid X triploid, which form most frequently 9r1+1r1r. On the basis of studies of chromosome morphology made by Doctor RANDOLPH in extra chromosome plants of Black Mexican sweet corn and in some genetic strains of maize, it is believed that this extra chromosome is not a whole duplicate of any one member of the haploid set of ten chromosomes. On the other hand, the GENE~CS 14: Mr 1929 202 BARBARA McCLINTOCK extra chromosome in the 2n+ 1 F1 individuals of the cross diploid X triploid is a'true duplicate of one member of the haploid set. The same chromo- some is very likely involved in all the usual 2n + 1 plants of Black Mexican sweet corn and genetic strains of maize. The extra chromosome in the several 2n+ 1 F1 individuals of the cross diploid X triploid is not always the same, but represents different members of the haploid set. Table 1 was constructed to show what crosses were made and the general results obtained. All the available material resulting from these crosses has not yet been investigated, but a sufficient amount has been examined to give an indication of what, in general, has taken place. As it was desired to obtain the best results with the least amount of pollen, care was taken to collect the pollen soon after shedding. In every cross sufficient pollen was used to pollinate a whole ear in an ordinary maize plant. As the triploidplant had several good tillers, much pollen was available even though many anthers never opened. The fact that only a few kernels developed on an ear as a result of these crosses could not, then, be attributed to a lack of sufficient pollen. The occurrence of a selection of gametes containing certain chromosome complements (in the cases where triploid pollen was used in crossing on to diploid plants) was markedly indicated by the few kernels on an ear and the chromosome number shown by the individuals coming from these kernels (table 5). Where there are few kernels on an ear the lo-chromo- some carrying gametes apparently have been successful, whereas the extra- chromosome carrying gametes probably have not functioned in fertiliza- tion. In other crosses where many kernels were found on an ear (362X 56*, table 5) not only the IO-chromosome carrying gametes functioned but also many of the lO+l and 10+2-chromosome gametes. Why it is that in all cases where a sufficient amount of pollen was used, there appeared such a dissimilar selectivity toward certain chromosome-carrying gametes among the different plants, is not known. When the cross triploid P X diploid C? is made, the results are strikingly different (table 5). H ere, apparently, the eggs containing various chromo- some numbers function. However, they do not function perfectly, as relatively few kernels developed on an ear. Since microsporogenesis is known to produce such varied types of cells with regard to chromosome complement,it is conceivable that megaspores in many ovules are incapable of developing a functioning gametophyte. The plants which were least vigorous contained the highest number of extra chromosomes. These grew from small underdeveloped kernels which had been carefully treated to induce germination. As there were many of them on the ear, it is probable B.,nIruwnun TUSE 1 %+ W' 54x33. IO"' ABpr t, jY S6,XEl4-811~ 55, kernels formed in main mdtillertassels. 10"' W" - 2,X%1 ? A b PI lQX.%r a C R S.W, 1.r b PI 12" 171X56, ACr S. b PI 10" 2011x561 A PI. ? --- 2114x56 ACR fir b PI 10" &.X56% 10" obPtCR 2411X56, AcRbPz 10" 251x56, ACr I. 261X56* ACr r. y l'l,,X%, AbPt LY 2&X56, AbPc f. 3014X561 A& r.j 13" 16` 10" 10" ? JlfX56r ? il. 33Ix56s I l@' ABpt 1.i S. Y 34nXs6r Anther ear ? 36,XSdt AcRsbPrv 10" 38.X 56, ACrgbbr l@l 4411x561 ? g', 51,X56, 6. P ? ElL533~56s El4-NliX56r Rl9LXS6. R251rX.56, X250,X56, R251, X 56, tiller * See note in table 2 --- 10" 34 I 1 _- -- -- == 10"' 6 --_ 10 --- ? 22 W-s LO"' 0 --- (0"' 0 --- ion' 17 --- 1O'U 0 --- 1om 1 --- 1P 9 --- 10"' 6 --- lam 2 --- I@" 22 --- lam 0 --. ion' 0 --_ 10" 1 --- W' 0 --. -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- - 10"' 7 --- 10" 3 --. loI* 49 --. lam 18 --. 10" 0 -- IO'U 10 --- lo'= 15 --- 10"' 6 --- 10" I --. 1P 13 --. 10"' 3 --. 10"' 2 --. 1om 1 1 purple sugary 3 colorless sugary 2 underdeveloped 8 purple starchy 68 26 colorless stmby 69 4 colorless starchy 6 poorly developed 9 purple starchy 64 4 colorless starchy 65 2 colorless sugary 66 7 underdeveloped 67 - 9 purple starchy Y 8 colodesr starchy Y 1 purple starchy :. 4 purple starchy Y 77 5 colorless starchy Y 4 purple starchy 1 colorless starchy fE 1 outside pollbation 2 wlorkss, underdeveloped 9 purple starchy 73 13 colorless starchy 74 1 purple underdeveloped 3 yellow starchy 4 hrdly infected, broken kernels 70 3 yellow starchy - 19 purple sugary 71 30 colorless sugary 72 - 4 purple starchy 75 14 colorless starchy 76 - - 4 purple starchy Y 6 colorless starchy Y Red p&carp, starchy 5 purple, 10 colorles9 dtminte-like 1 teosinte-like 13 purple starchy Y 3 yellow sugary 2 purple sugary 1 colarlens sugary Y __ .- .- -- -- -- -- == -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -. -- -- -- -- -- -- - --- ,. --- 88 89 20 10 --- 90 91 8 6 __-- __ _. _. -. -. ._ __ -_ _- __ -- -- . . --- --- --- --- ,. --- . . ., --- --- ,. ._ ____- ., --- % 97 6 0 __-- 94 31 11 95 --- __-- 7 3 --- ,. --- --- 81 1 1 __-- 82 83 8 6 --- 84 4 4 --- . --- 8.5 86 13 8 --- . -__- -~- ., --- 93 3 2 I- /) 2n sn+1 zn+a ants _--- _--- 1 3 1 ---- . . ---- . 1 1 1 =m-- . ---- ---__ . . _--- -- -- 1 . . ---- 6 ., ---__ 3 ,. ---- . ---- 6 1 _--- - -- __ -. __ ci -. -. -. _- -_ _. _- _. _. -_ _. -. -. -. - -_ _. -. -. -. -. -. - _._ ---- ---- t. _--- _--- 2 ___-- . . --__~ 3 s 2 -- -- 6 . . __--- ---- . -______ _____- ___-~ _-____ -__-__ ----- . _~-- - 2n-H 2nts 2+ 44. 2tr+7 *n+* *n+o 3n ----- . . . . -__- -- 3 1 1 . ---__ --- . . --- ---- . . . - 14 -- m- ., . . ,. . . . . --- --- ) ) 1 . 1 . . ) ( . . ------ _, . . . . ____- --- -1 I- . . . . ------- . ---- -- . . ., ,. . . . . . ., ,. . . . ---- ---- --- --- . . . . . . . . . . . . __------ __------ , , ------- ------- . , . . . , . . 1 1 -__----- -__----- . . . . ------- ------- __----- - t ., __ ..- . __--- --- -__-__ --- I_-~---- ____--_____- I_~-- ~-- ., ~------ __--__---- . . . ------- ------- . ---- --- ., ~--~ ~-- ii= h+l . . . . . . . -- c e TRIPLOID MAIZE GENETWS 14: Mr 1929 204 BARBARA McCLINTOCR FIGURE 3.-Photograph of F, plants qown in the field resulting fmm the cross triploid 0 X diploid 8 arranged according to chromosome complement. Plant no. 1, 941; no. 2,95; no. 3 9.5; no. 4, 9.53: no. 5,944; no. 6, 9&; no. 7, %. that if all could have been induced to germinate a wider-range of chromo- some number in the FL would have been found. Thus the-limiting factor in FIGURE 4.-Photograph of F, plants from reciprocal cross (diploid 0 X triploid 3) to that shown in text figure 3. Bath individuals were diploid. Plant no. 8,932; no. 9,931. some cases is probably not the inability of the female gametophyte to function, but rather the capacity of the chromosome complement within the embryo to participate in growth and development. TRIPLOID MAIZE 205 The marked effect of the presence of extra chromosome on the appear- ance of the plant is graphically represented by text figures 3, 4, 5 and 6. These photographs were made after the plants had matured ears. In con- sequence the leaves are dried and curled. However, the relative sizes of the plants containing varying chromosome numbers are preserved. Two plantings of F, kernels were made, one in the spring of 1926 in the greenhouse and one in the summer of 1926 in the field. For a summary of germination results see table 3. Origin of FL kern& 562 tassel 562 X336 562 selfed 2lrrX56a 23s X562 24,8x562 261 X562 3014x562 33s X562 362 x562 38~1 X562 TABLE 3 Rest& of germinalion. (IREENBOUBI Kernels Mature planted Pbb 7 3 lO(3 died) t 4 2 2 2 2 5(2 died)t 3 l(died)t 0 lO(5 died)t 3 .S(l died)t 4 - _. _. _. _. _. -. _. -. -. -. .- - 21* 7 6 0 1 1 6 4 2 2 8 5 1 0 2 2 a 10 7 f 3 I 2 *9 well developed kernels gave rise to 4 mature individuals, one with 8rr+2ux, one with 7"+3*rr and two with 6r+4rrr. 12 underdeveloped kernels gave rise to 3 mature individuals with chromosome numbers 6r1+4rJ1, S"+Sxl* and 3rr+7rrr respectively. t Plants began development but died from "damping off" in seedling stage. MEIOSIS IN Fr INDI+IDUALS The behavior of the chromosomes during meiosis was studied in fifty F1 individuals. When diploid individuals were pollinated with pollen from the triploid, twenty of the thirty-seven plants examined showed ten pairs of GENETICS 14: Mr 1929 206 BARBARA McCLINTOCK chromosomes at diakinesis and metaphase I. The meiotic behavior in these plants was similar in all observed details to that in ordinary diploid individuals. In size and vigor they surpassed those F1 individuals which possessed one or more extra chromosomes (compare plant No. 14, text figure 6 with its sib, plant No. 10, text figure 5; plants 15, 16 and 18, text figure 6 with their sibs, plants 11, 12.and 13, text figure 5). FIGURE 5.-Diploid individuals resulting fmm cross diploid 0 X triploid @ . Plant no. 10, 861; no. 11, 882; no. 12,8&; no. 13,891. Meiosis in Zn+l individuals Out of fifty F1 individuals examined, eight showed the presence of one extra chromosome: two out of thirteen in the cross triploid 9 Xdiploid 3 and six out of thirty-seven in the reciprocal cross, diploid 0 x triploid CF. Cytological examination indicated that different chromosomes of the haploid set we involved in the several 2n+l plants. However, the 2n+l plants did not show characteristic differences that could be ascribed to certain chromosomes of the set. In describing meiosis in these plants reference will be made, at times, to figures from plants which contained more than one extra chromosome. Since similar conditions existed in all the extra chromosome plants, one TRIPLOID MAIZE 207 figure will suffice for illustration. This has been done to avoid repetition of figures illustrating similar phenomena. There were fewer meiotic irregularities in these individuals than in those containing more than one extra chromosome. As in all F1 extra chromo- some plants, a variation in synaptic expression of the extra chromosome was noticed. In the Zn+l plants 9rI+1"' (figure 11, plate 2) was found about twice as frequently as lOI'+ 1' (figures 12 and 13) in both diakinesis and metaphase I (table 4). The appearance and position of the univalent during metaphase 1 are rather varied. The univalent may be found in any region of the spindle. It occasionally is not included in the spindle but FIGURE &-Extra chromosome individuals resulting from cross diploid 0 X triploid 3. These are sibs of those plants shown in text figure 5. Plant no. 14,862; no. 15,881; no. 16,8&; no. 17, 88a; no. 18, 891. lies in the cytoplasm. When a univalent lies in or near the plate it may split simultaneously with the disjunction of the bivalents, or it may lag until the bivalents have completely disjoined and have passed toward either pole, when, in turn, it may split., the halves separating and advancing toward opposite poles. These may or may not be included in the reorgan- izing nuclei (figures 9, 10, 24, 28 and 34). The univalent may remain unsplit at the equator and subsequently be left in the cytoplasm as a chromatic mass. When a trivalent is present, on the other hand, dis- junction of the three homologues and the passage of two toward one pole and one toward the other pole seem to be quite regular. Interkinesis most frequently shows two well organized nuclei lying in cytoplasmic masses separated by a cell plate. Occasionally, in the cyto- Grmrrcs 14: MI 1929 208 BARBARA McCLINTOCK plasm of one or both cells, there lie one or more chromatic bodies distin- guishable as dyad or monad. Besides the two major nuclei one or more very small nuclei may be found in the cytoplasm of one or' both cells. TABLE 4 Frequency of appearance of different synafitic combinations in F, extra chromosomes individuals. 881 8% 884 941 942 9x+ p1 lo"+ 1' 9"+ 1111 10**+ 1' 911+ 1x11 lo"-+- 1' 811+2X11 9x1+ l"`+ 1' lo"+21 ~11+2111 9II+ l"`+ 1' lo"+2 511+5111 6"+4"`+ 1' 7"+3"`+2' S1'+2'"+3' 9"+ 1"`+4' 10"+5' 9II+ 7' 411+3"`+ 81 3"+2"`+ 13' 2"f l"`+ 18' l'q- 1"`+20' 25= - -- _- _- _- _- _- Dls. 8 4 1 Dk. = diakinesis; MI =metaphase 1. MI PLmT -- 28 10 5 7 943 40 12 -- 7 4 - 95, 8 7- 2 - 952 20 8- 2 7 6 3 953 1 2- 0 1 1 944 1 1 3 BniIPSlB 311+7"1 4"+6'"+ 1' jII+j111+21 6"+4*"+3' 7"+3"`+4' 8"+ 2*11+5' 9'I+ l'"f6' lo"+7 711+3111 8"+2"`+ 1' 9'r+1111+2' 1O"+31 not recorded @I+4111 7"+3"`+1' 8'1+2"`+2' 9"+ l"`+ 3' lo"+4 611+4111 7"+3"`+1 8"+2111+21 9"+ l"`f3' 10"+4' - I- _- _- _- - DK. - -- -- -- -- -- - 4 a 8 5 2 2 33 41 27 6 6 11 10 3 1 18 12 4 Since there were many 10-11 anaphase I chromosome distributions, this same relation appeared in pairs of prophase II nuclei. The second meiotic mitosis proceeded with little irregularity. Occasionally a dyad chromosome lagged in anaphase II with or without subsequent separation of its mem- bers; or, as the anaphase advanced, a monad ceased to keep pace and was TRIPLOID MAIZE 209 left out of the telophase nucleus. Sometimes such a lagging chromosome, whether a dyad or a monad, became included in the forming wall (figure 30). Some cells possessed a second small spindle, oriented at various angles or connected with the major spindle. In consequence of such meiotic behavior we most frequently encountered complete quartets of spores, each spore containing one well formed nucleus. Occasionally other small nuclei or chromatic bodies were observed in the cytoplasm. Only seldom did a sporocyte give rise to more than four spores. Unusual meiotic phenomena in h-l-1 individuals Some anther loculi in both greenhouse and field plants were tiled with a plasmodial mass including either prophase I nuclei, like or unlike in size, or metaphase I figures, free or connecting with each other at various angles (figure 19). The chromosome number in the various figures was not always the same. Some had more than the normal number for the plant in question and some less (upper spindle, figure 19). In one case the number amounted to four times that expected, the spindle being perfectly formed and the chromosomes regularly arranged. There appeared, in a few of these plasmodial sporogenous masses, nuclei containing amounts of chromatin not in accordance with their various sizes. A large nucleus might contain but a single chromatic thread lying against the nuclear membrane, nearly the whole volume, of the nucleus being karyolymph. An adjacent nucleus, small in volume might be densely packed with a chromatin spireme. In the same plasmodium all states between these two existed. In figure 17 is shown a metaphase I spindle containing 20"+2', double the expected number; this was seen in a well rounded sporocyte. Figure 18 shows a well formed spindle containing 4Orr+4', or four times the expected number. Fragmentation appeared in these and other F1 plants. It was most evident during anaphase 1 (figure 20). The fragments vary in size from a mere speck of chromatin to a large portion of a dyad. Appearances indicate that some of these fragments have the ability to divide. This is observable in the chromosome fragments at the center of the spindle (figure 20). Meiosis in 2n + 2 individuals Out of fifty F1 individuals examined, six possessed two extra chromo- somes belonging to two different homologous groups. In the direct cross triploid Q Xdiploid 8 four out of thirteen showed two extra chromosomes; in the reciprocal cross, two out of thirty-seven. These plants were smaller and less vigorous than their sibs with one extra chromosome. F 561 tassel z s 54 X336 54 selfed Total 2114X.562 238 X54 261 X54 2418X562 3014X!& 33s x.54 36% x54 388 X54 Total Grand Total TABLE 5 (Abridgment of Table 1) NUYBER cm NUYFmR 0. NUYBlB 01 2n IERrnRLB IlrtNaLB PLANTS 2n+1 2n+2 2nt3 ON 4x3 PLANT'BD rnYEI.oP~D 1011 QII+lIII ,31+2111 7143" ---- 22 7 3 1 1 1 ---- 34 31 11* 1 3 1 ---- 6 6 0 ---- 44 14 2 4 2 -- -w 1 1 1 1 ---~ 9 8 6 6 ____~ ---___ 22 13 8 6 1 ---- 6 4 4 3 -______ 1 1 0 --_____ 3 3 2 2 49 20 10 3 5 2 -___- 18 8 6 6 -__ -__ 58 37 27 6 2 -- ___- 102 51 27 8 6 2 - II CT. _ -. _. -. -. -. -. -. -. - 2n+4 2n+s 2n+il 2n+7 2n+s 6'1+4111 @+5111 3 x+7111 ----- ----- 3 1 1 -- --__ ----- 3 1 1 s- - ___- --- -- --- -__- -- -- --A -__ -___- -__- -- ___-__~ -~ -__- --- 3 1 1 - -. -. -. -. =: _. _. _. . -. -. _ . - 2n+Q 3n 10111 1 1 1 -_ -- -. -_ =r -- -_ -_ -- -_ - 2n+2t 1111 -- 1 -- 1 1 o Chromosome number of one plant not obtained. t See note in table 2. TRIPLOID MAIZE 211 Meiosis in these plants was similar in most details to that of the Zn+l individuals with the added complication of the one extra chromosome; thus making the observed irregularities more numerous than in the 2n+ 1 plants. The univalents acted similarly but independently of one another. During diakinesis and metaphase 1 (figure 22, plate 3) were seen most frequently Srr +2 III, less frequently 9rr+irrr+lr (figure 21) and least frequently 10" +2I (see table 4). Because of the extra chromosome, anaphases were markedly more irregular than in the 2n+l plants. Here we frequently found 11 chromo- somes passing to each pole, although 10-12 distributions were numerous. The lagging of one or two chromosomes, the lagging and splitting of one or two, or combinations of these two types were rather numerous. Figure 24 illustrates the lagging and splitting of one dyad in a spindle where ten chromosomes have gone to one pole and eleven to the opposite pole. Appearances in teIophase I and prophase 11 were similar to those in the 2n + 1 plants. Figure 31, a prophase II, in which the chromosomes in the left cell are farther advanced than those in the right, indicates that a 10-12 distribution in anaphase I must have taken place. Such a prophase as that in the right half of figure 31 might be followed by an anaphase IJ distri- bution like that illustrated in figure 29. More frequently, however, the anaphases of II were less regular, being marked by the lagging of one or two chromosomes which eventually become excluded from the reorganiz- ing nuclei, In consequence, the spores frequently appeared much as in figure 30. Sometimes an irregular spore quartet, as illustrated in figure 25, was formed. One greenhouse pIant (6&) exhibited more striking irregularities in meiotic phenomena than any other individual examined. Many anthers possessed one continuous plasmodium in which lay nuclei ranging in size from very minute to those two and three times the volume of the normal nucleus. These, in turn, probably gave rise to plasmodial sporogenous masses in which lay spindles of all sizes containing from very few chromo- somes to more than four times the normal number for the plant. In one huge sporocyte there were two parallel metaphase 1 spindles, in each of which lay four times the normal number of chromosomes. The chromo- somes were associated as bivalents, trivalents, quadrivalents, quintivalents and sexivalents. Unfortunately, the slide was damaged before a drawing was made. In this plant some sporocytes exhibited metaphase 1 figures with twice the normal number of chromosomes. In all previous cases in other plants where tetraploid metaphase I spindles were observed, the chromosomes GENETICS 14: Mr 1929 212 BARBARA McCLINTOCK of the four sets gave no synaptic evidence of homology (see figures 17 and 18). Here, however, the homologies were frequently expressed by the synaptic union of four chromosomes to form a quadrivalent or of six to form a sexivalent (it is to be remembered that this plant showed, normally, 8"+2"`). Not all the chromosomes of the four sets necessarily formed such synaptic combinations, for two homologous bivalents sometimes appeared distinctly separated. Such a case is illustrated in figure 27, which was drawn from a remarkably clear figure with 2v1+61v+411. Figure 28 represents an anaphase I distribution of chromosomes in a tetraploid sporocyte including a lagging bivalent, a lagging and splitting dyad and two lagging but separating monads. Figure 26 shows a metaphase I spindle from this same plant which contains 1 bivalent less than normal, that is, 711+2"I. Meiosis in plants with 2~~4-3 to 2n+7 chromosomes A reference to table 5 will show that individuals with more than two extra chromosomes appeared only in the cross triploid 9 Xdiploid c? (with the exception of a new triploid). There were two plants with 711+3111, three with 6rr+4111 , one with 511+5111, and one with 311+7111. The pres- ence of extra chromosomes was suggested in the small size and decreased vigor of these F1 individuals before meiosis was observed (see text figure 3). Since synapsis was so varied, it was of interest to determine the per- centage of bivalents, trivalents and univalents in the various diakinesis and metaphase I figures of each plant. Reference to table 4 will show that there was a general tendency on the part of the univalents to be associated with the bivalents to form trivalents at diakinesis and metaphase 1 in a good percentage of the cases. Figure 32,plate 4, shows 71r+3r11, while figure 33 represents a metaphase 1 with 811+211f+11. In figure 36 we observe 811+1T11+41, indicating that there has been non-synapsis or early disjunction of one of the bivalents ordinarily present as such. A metaphase I figure in plant 952 with four extra chromosomes is shown in figure 39; here there are clearly 611+4111. Plant 953 with four extra chromosomes is represented by figure 37, in which we find 6"+4"`. When it was determined that photography could be of service in describing the results, this same figure was photographed (figure 14, plate 6). In an adjacent sporocyte(text figure 7) we find 7'I+3"' +l', while in still another sporocyte we see 10rr+41 (figure 15, plate 6). A metaphase I with 311+71r1 is shown in figure 38,which was drawn from a sporocyte of a 2n + 7 individual. Anaphases in these plants were very similar to those already described TRIPLOID MAIZE 213 for other F, extra-chromosome individuals, with the added complication of more univalents and consequently more irregular behavior. In an ana- phase I (figure 34) of plant 9S1 (2n+3) two lagging dyads show clearly the split in preparation for II, while the components of two other dyads have already commenced to separate. FICVRE 7.-Photomicmgraph of microsporocyte of 2nf4 individual showing 7 bivalents, 3 trivalents and 1 univalent. See photographs of adjacent sporocytes, figures 14 and 15, plate 6. From an aceto-carmine smear preparation. Since irregularity increases in proportion to the number of extra chromo- somes present, the variation in spore formation increases accordingly. Giant spores, dyads, quartets and polyspory with and without extra, small nuclei or chromatin masses were repeatedly observed. F, lriploid plant In the greenhouse plants there appeared a new triploid from the cross diploid 0 X triploid 8. Since meiosis in the new triploid was essentially similar to that of the triploid already described, no further discussion will be given. Figure 40 represents'a metaphase 1 in which there are 10"`. Irrregdur synapsis in (I 212-G indinidual Since plant 94? (2n+5) showed an almost complete series of non- synapses of homologous chromosomes during diakinesis and metaphase 1, ranging from jr*+3 `II1 to 25*, it has been considered separately. In many of the sporocytes of an anther we found the synaptic condition to be very similar to that already described for other extra-chromosome F1 indivi- GENETICS 14: hlr 1929 214 BARBARA McCLIXTOCK duals, that is, 5"+S11r, 6"+4"`+ 1' to 10rl+S1 (figure 41). Besides this, non-synapsis extended to the bivalents themselves, forming, for instance, 9"+7' (figure42) or411+3111+81 (figure43). The appearance of the diakinesis figures (figure 13) points toward a possible early disjunction as an explanation, since the members of many bivalents and trivalents were specially far apart, remaining connected only by long, fine threads. Anaphases vary in appearance according to the number of univalents present. When there are few univalents the figures are quite normal. When there are many univalents there is no well organized equatorial plate (figure 45, 1"+23'; figure 46, 1"+1"1+20'; figure 44, 25') the chromosomes being distributed throughout a greater part of the spindle. We have, at present, no satisfactory explanation of this appearance. It is difficult to explain why such wide variations in synaptic expression occur in sporocytes of the same anther or of other anthers at the same stage of development. In the summer of 1926, under field conditions, RANDOLPH (1928) examined a number of inbred strains which were showing this non-synaptic phenomenon in many sporocytes. Kormally in these individuals there are 10" at metaphase I and meiosis is regular. However, in those diploid plants which showed some sporocytes with varying degrees of non-synapsis the univalents thus formed behaved much as they did in the extra-chromosome F1 individuals and in hybrids in general. They lag- ged, or lagged and split, giving appearances of irregularity in a num- ber of sporocytes. Other sporocytes in the same anther were perfectly normal in appearance and behavior. Non-synapsis at diakinesis, whether it represents a true non-synapsis or merely an early disjunction, is caused in some cases by conditions other than lack of homology of the chromosomes present. The F1 extra-chromosome individuals described in this paper have been selfed and crossed for further cytological investigations to be reported later. GENETIC INVESTIGATIONS The kernels of culture 56 in which the triploid plant arose were planted for use as a sugary tunicate stock. In consequence, no particular attention was paid to the color of the kernelsother than to note that a segregation for purple and white aleurone had occurred on the ear from which they had come. Since the culture was also segregating for the plant color factor A, little could be concluded directly about the constitution of the triploid itself. It was known to be phenotypically dilute sun red, tunicate and sugary. TRTPLOID MAIZE L13 Since three sets of chromosomes were present, it was desired to know what the genetic constitution of the triploid might be with respect to certain genetic factors, and how these might be carried to future genera- tions. For this reason pollen of the triploid was placed upon silks of several genetic "testers" (see table 1) whose mirosporocytes had been examined to detect the possible presence of extra chromosomes or irregular behavior. One ear of the triploid plant was used for the cross .5&><335, the recipro- cal cross (33j~ 562) being made also. A second ear was selfed and a third ear on a tiller was used in the cross triploidx tetraploid (perennial teosinte). It will be seen by an examination of table 1 that very few kernels devel- oped as a result of most of the crosses. If me disregard the possible selective germination of the F1 kernels carrying extra chromosomes, it will appear on examination of table 5 that in most crosses the male gametes from the triploid which functioned possessed ten chromosomes. In the crosses 211,X562, 23*x56? and 33,X56* (tabIe 5) very few kernels developed on the ears. ,411 the plants that grew to maturity from these crosses possessed but ten pairs of chromosomes. In certain crosses, namely, 261~56n and 362~562 (table 5) where relatively more kernels developed on an ear, the F, plants coming from these kernels showed not only diploids but some 2nfl and 2n+2 individuals. This suggests that in the former case only the pollen grains carrying the haploid set of chromosomes functioned, while in the latter case not only the haploid pollen functioned but also the haploid+l and some of the haploidf2 pollen, therefore producing more kernels on the ears. Before beginning a discussion of the results obtained it would be well to outline briefly the expected diploid x triploid ratios, assuming random assortment of the extra chromosomes in the triploid with no selection of pollen or of F, chromosome combination. It is known from the obser- vations of meiosis in the triploid that pollen carrying various chromosome numbers was formed. Assuming, for descriptive purposes, a factor A and its allelomorph a, we can say that if the triploid were homozygous for A (An-d), that is, triplex (BLAKESLEE, BELLISG and FARNHAM 1920), the F1 would be, phenotypically, A. If the triploid had the constitution AA a (duplex) the expected gametic ratio would be 2A : 1AA : 2Aa: la, and when crossed to ct it would give, phenotypically, 5A : la. However, if the triploid were Aaa (simplex) the gametic ratio would be 1A: 2Aa: 2~: laa, and the backcross phenotype ratio 1.4: la. If the triploid were homozygous recessive (nulliplex) the backcross would be totally recessive, phenotypically. GEXEIICS 14: l\lr 1929 216 BARBARA llcCLINTOCK If, now, there is a decided selectior against the gametes carrying extra chromosomes, the functional gametic ratio becomes modified accordingly. In the case of AAa (duplex) eliminatyon of the extra chromosome carrying gametes would bring the functioning gametic ratio to 2A : la. In the case of Aaa (simplex) such a selection would give a functioning gametic ratio of 1A : 2~. Gametic selection in the reciprocal crosses of the diploid X triploid has been discussed in connection with the cytology of the F, individuals. We have seen that meiosis in diploid individuals usually results in the production of haploid spores and in consequence haploid gametes contain- ing ten chromosomes. Meiosis in the triploid, however, results in the pro- duction of microspores containing various chromosome numbers. When pollen from the triploid is used in crossing on to silks of a diploid individual the pollen grains that seem to be functionally successful contain for the most part 10 chromosomes, or occasionally lO+l or lOf2. Hence, from such a cross we should expect a duplex or simplex factorial condition in the triploid to appear-in the backcross in the phenotypic ratios of 2A : la or 1A : 2u respectively. In the cross triploid 0 X diploid 3, on the other hand, we have seen that there has been no such marked elimination of the extra chromosome carrying gametes of the triploid. In consequence the ratios should be expected to show some approach to those indicated in the preceding paragraph. Although not all the FI kernels have been grown and the genetic data are few, the results so far obtained and outlined below tend to confirm this expectation. The triploid was triple recessive or nulliplex for the plant color factors b and PI (Emerson 1921); since it was dilute sun red and gave only dilute sun red individuals when crossed to plants which were homozygous A@,. Likewise it was nulliplex for the factor s,, responsible for sugary endosperm (see table 1, 362X562 and 562 selfed) and the endosperm factor y which is the allelomorph of I' responsible for yellow endosperm (see tablel, 362X 562) (CORRENS 1901, EAST and HAYES 1911). The evidence points toward the fact that the triploid was simplex for the factors C and R, two complementary aleurone color factors (EMERSOK 1918). Considering the factor C, which with A and R seems to be respon- sible for the presence of colored aleurone, the following ratio of F1 kernel color was noted when the triploid was crossed to an AAccRR or so-called C-tester (table 6). Since in these crosses there was a selection against pollen carrying the higher numbers of chromosomes expected on the assumption of random assortment, the F1 ratio, according to our interpretation, should approach TRIPLOID MAIZE 217 TABLE 6 mom AAccRRXtriploid 241~X5& 362 XX& Total a 1: 2 if the triploid was simplex for the factor C. Although the data are few, this ratio is approximated. Certainly it does not approach the duplex 2: 1 ratio. With reference to the factor R, the simplex condition in the tripIoid was suggested by the results of the foIlowing crosses which were made to R testers (AACCrr) (table 7). TMLE 7 CROBB AACCrrXtriploid 17iX5& 251x562 261x562 38eX 56-z Total r, *LrnRONE COLOR CCdOd Colorelas 9 8 2 9 13 4 14 22 37 Since the functioning male gametes are predominantly carrying only ten chromosomes (table S), the results recorded in table 7 can best be inter- preted as a 1: 2 modified simplex ratio. Fortunately, the triploid possessed the dominant plant factor tunicate (5?,J COLLINS 1917). Direct and reciprocal crosses were made to the double recessive (kly). Since there was no obvious selection against gametes carrying higher chromosome numbers in the cross triploid 9 Xdiploidc?, the unmodified tripioid ratios would be expected to appear in the FI of this cross, that is, if the triploid were simplex for T,,, the ratio would be a 1: 1, if duplex, 5: 1. In the reciprocal cross, since selection against the higher chromosome gametes occurs, the modified ratio should appear, that is, if simplex a 1:2, if duplex, a 2: 1 ratio. The following are the results obtained in the direct and reciprocal crosses (tables 8 and 9). GENETICS: hlr 1919 218 BARBARA McCLINTOCK TABLE 8 Fl trip1oid 9 Xdipzoid (t"t") #. 562 X336 I 8 I 2 TABLE 9 Fl diploid (tutu) 0 Xlriploid 3. CR088 llmN!*TE 2114x562 0 23s XS62 2 2418X 562 2 261 x562 7 33s x562 1 361 x562 7 388 X562 4 Totals 23 _. .I. NON-mNIChm 1 3 2 1 1 3 2 - 13 The results tabulated above afford evidence for a duplex condition of T, (TUT&) in the triploid. Although the number of plants is small, the data in the two crosses in no way conflict. On the contrary, the two crosses act as decided checks upon one another and upon the interpretation as- sume.d to explain modified ratios. Where no modified ratio is expected, that is, where no decided gametic selection has occurred, the duplex 5 : 1 ratio is approached (table 8). In the reciprocal cross, moreover, a duplex modified ratio of 2 : 1 is approximated where gametic selection is known to have occurred (table 9). The plant anthocyan color factor A was present in the triploid as either simplex or duplex. Unfortunately one cross (169~562, table 1) made to test this factor produced no kernels. The second cross (238X562, table 1) produced but 9 kernels, 4 colored: 5 colorless. The results merely indicate that the triploid plant was heterozygous for A, the numbers being too low to decide between the simplex and the duplex condition. From what has been stated concerning the inheritance of the factors mentioned above, we are able to reconstruct, to some extent, the constitu- tion of the triploid with regard to these factors as follows: A?a b66 p$@l TuTutu sus,s, Ccc Rrr. TRIPLOID MAIZE 219 DISCUSSION Among investigations on plants several cases of trisomic inheritance have been observed. At present "trisomic inheritance" has two meanings in the literature, both of which, however, depend upon the presence of three instead of two homologous chromosomes. In the first category the character followed in inheritance depends simply upon the presence of the extra chromosome for its expression. The 2n+ 1 chromosomal complement produces a particular recognizable charac- ter in a plant as contrasted with the 2n condition, and this character may be followed without special reference to the genie constitution of the extra chromosome as compared with that of its homologues. If a 2n+l indivi- dual is crossed on to a 2n individual the phenotypic ratio in F1 depends solely upon the relative number of individuals which do or do not possess this extra chromosome. Thus, in this type of trisomic inheritance it is the distribution of the extra chromosome that is followed through succes- sive generations. This kind of inheritance, depending upon an extra chromosome for the expression of a mutant character, has been extensively investigated (BLAKESLEE et al.) in the Datura mutant types Globe (2n+l), Poin- settia (2nf 1) etc., each of these types being the result of a duplication of a different chromosome of the set. Other examples are the Nicotiana mutant "enlarged" flower (Znf 1) (CLAUSEN and GOODSPEED 1924), the Matthiola mutants Large, Slender, Smooth, Crenate, etc, (FROST and MANN 1924; FROST, 1927), and 2n+ 1 types in Lycopersicum (LESLEY 1928). In Zea any one particular chromosome when it is present in triplicate in Znfl individuals has not produced, so far as we have noticed, any one particular recognizable character. The presence of extra chromosomes was recognizable, phenotypically, only in decreased size and vigor. In the second category would come the inheritance of characters which depend upon a factor located in an extra chromosome in a triploid or a 2n+l individual. It is essentially Mendelian. The ratios obtained in inheritance of the first category are expressions merely of the relative number of 2n and 2n+l individuals. Those in the second category follow the accepted laws of Mendelian inheritance on a trisomic rather than a disomic basis. In most cases the formulation of an expected ratio would depend upon the assumption of a random assortment at meiosis of the three homologous chromosomes containing the factors in question, and a per- fectly random mating. In all cases such expected ratios must be modified by any selective action against certain extra-chromosomal gametic types. GENE~CS 14: Yr 1929 220 BARBARA itlcCLINTOCK This genie type of trisomic inheritance has been studied in the case of purple and white flower factors in the Poinsettia chromosome of Datura. Recently other gene characters (B-white, swollen, tricarpel) on other chromosomes have been studied (BLAKESLEE 1927; BUCHHOLZ and BLAKESLEE 1927; GAGER and BLAKESLEE 1927). In these cases both types of trisomic inheritance are exhibited together. Trisomic inheritance in Zea, as reported in the present study, is of the second type as indicated by the inheritance of the factors T,,, C and R. LESLEY (1926) has found similar trisomic gene inheritance in tomato. Among animals the inheritance in chromosomal types other than the diploid has been rather extensively investigated in Drosophila (BRIDGES 1916,1921,1922; L. V. MORGAN 1922,192s; ANDERSON 1925; BRIDGES and ANDERSON 1925). With regard to Zea it should be kept in mind that the data are necessa- rily few owing (1) to the limited amount of pollen produced bythe triploid and thus the limited number of crosses that could be made, (2) to the great sterility of this pollen and consequently the few kernels that resulted from each cross, and (3) to the inability to recognize chromosomal types other than by a cytological examination. Hence, the amount of material grown was limited strictly to that whichcould be studied by one observer in a certain period of time. SUMMARY 1. A triploid Zea mays individual was found in a culture which was otherwise diploid. 2. Meiosis in the microsporocytes of this individual is described. The basic chromosome number in Zea is believed to be ten. At diakinesis and metaphase of the first meiotic mitosis the triploid frequently showed ten trivalents. There was a tendency for members of the extra set of chromo- somes to be disassociated from their homologues at diakinesis and meta- phase, thus forming nine trivalents, one bivalent and one univalent or eight trivalents, two bivalents and two univalents etc. 3. Reciprocal crosses were made with diploid Zea. 4. Observations on fiftyF1individuals of the reciprocal crosses of diploid X triploid were made. It was found that there had been a decided selection against extra chromosome carrying male gametes and a less obvious selec- tion against extra chromosome carrying eggs. 5. The presence of extra chromosomes in Zea plants, other than in a polyploid series, is associated with a decrease in size and vigor. TRIPLOID MAIZE 221 6. Ratios for several factors (T,, R and C) were found to accord with the theoretical expectancy on the basis of trisomic inheritance. ACKNOWLEDGMENT I wish to acknowledge my indebtedness to Doctor L. W. SHARP and Doctor A. C. FRASER and to express my sincere appreciation for their thoughtful criticism and constant encouragement during the progress of this investigation. LITERATURE CITED ANDERSON, E. G., 1925 Crossing over in a case of attached X chromosomes in Drosophila mefanogaster. Genetics 10: 403417. BELLING, J., 1921 The behavior of homologous chromosomes in a triploid Canna. Proc. Nat. Acad. Sci. 7: 197-201. 1925 Homologous and similar chromosomes in diploid and triploid Hyacinths. Genetics 10: 59-71. 1925 Chromosomes of Canna and of Hemerocallis. Jour. Hered. 16: 46.5466. BELLING, J. AND BLAKESLEE, A. F., 1922 The assortment of chromosomes in triploid Daturas. Am. Nat. 56: 339-346. BLAKESLEE, A. F., 1921 The Globe, a simple trisomic mutant in Datura. Proc. Nat. Acad. Sci. 7: 148-152. 1921 The Globe mutant in the Jimson weed (Dduru stramonizrm). Genetics 6: 241-264. 1927 The chromosomal constitution of Nubbin, a compound (2n+l) type in Datura Proc. Nat. Acad. Sci. 13: 79-85. BLAKESLEE, A.F., BELLING, J. AND FARNHAII, M. E., 1920 Chromosomal duplication and Men- delian phenomena in Datura mutants. Sci. 52: 388-390. BLAKESLEE, A. F. AND FARNHAM, M. E., 1923 Trisomic inheritance in the Poinsettia mutant of Datura. Am. Nat. 57: 481495. BFSDCES, C. B., 1916 Nondisjunction as proof of the chromosomal theory of heredity. Gene- tics 1: l-52,107-163. 1921 Triploid intersexes in Drosophila mdanogaster. Sci. 54: 252-254. 1922 The origin of variations in sexual and sex-limited characters, Am. Nat. 56: 51-63. BRIDGES C. B. AND ANDERSON, E. G., 1925 Crossing over in the X chromosome of triploid fe- males of Drosophila melanogaster. Genetics 10: 418-441. BUCHHOLZ, J, T. AND BLAKESLEE, A. F., 1922 Studies of the pollen tubes and abortive ovules of the Globe mutant of Datura. Sci. 55: 597-599. 1927 Abnormalities in pollenytube growth in Datura due to the gene "Tricarpel." Proc. Nat. Acad. Sci. 13: 242-249. CLAUSEN, R. E. AND GO~DSPEED, T. H., 1924 Inheritance in Nicofiana t&cum. IV. The trisomic character "Enlarged." Genetics 9: 181-197. COLLINS, G. N., 1917 Hybrids of Zea ramosa and Zea tunicata. Jour. Agr. Res. 9: 383-395. CORRENS, C., 1901 Bastarde zwischen Maisrassen mit besonderer Berticksichtigung der Xenien. Bibliotheca Botanica 53: 1-161. EAST, E. M. AND HAYES, H. K., 1911 Inheritance in maize. Conn. Agr. Exp. Sta. Bull. 167: l-142. EMERSON, R. A., 1918 A 6fth pair of factors, Aa, for aleurone color in maize, and its relation to the Cc and Rr pairs. Cornell Univ. Agr. Exp. Sta. 1Mem. 16: 231-289. 1921 The genetic relations of plant colors in maize. Cornell Univ. Agr. Exp. Sta. Mem. 39: l-156. 222 BARBARA McCLINTOCK FISK, E. L., 1925 The chromosomes of Zea mays. Proc. Nat. Acad. Sci. 11: 352-356. 1927 The chromosomes of Zea mays. Am. Jour. Bot. 14: 53-75. FROST, H. B., 1926 Bud variation and chimeras in Matthiola incana R. Br. Jour. Agr. Res. 33: 41-46. 1927 Chromosome-mutant types in stocks (Matthiola incana R. Br.) I. Characters due to extra chromosomes. Jour. Hered. 18: 475486. FROST, H. B. AND MANN, M. C., 1924 Mutant forms of Matthiola resulting from non-dis- junction. Am. Nat. 58: 569-572. GAGER. C. S. AND BLAKESLEE, A. F., 1927 Chromosome and gene mutations in Datura fol- lowing exposure to radium rays. Proc. Nat. Acad. Sci. 13: 75-79. GAIRDNER, J., 1926 Campanula persicifolia and its tetraploid form, "Telham Beauty." Jour. Gen. 16: 341-351. GOTOH, K., 1924 Uber die Chromosomenzahl von Secale cereale, L. Bot. Mag. (Tokyo) 38: 135-151. KIESSELBACH, T. A. AND PETERSEN, N. F., 1925 The chromosome number of maize. Genetics 10: 80-85. K~VADA, Y., 1911 Maiosis in the pollen mother cells of Zea mays L. Bot. Mag. (Tokyo) 25: 163181. 1915 Ueber die Chromosomenzahl von Zea mays L. Bot. Mag. (Tokyo) 29: 83-89. 1919 Die Chromosomenzahl von Zea nays L. Jour. Col. Sci., Imp, Univ., Tokyo 39: 148 PP. 192.5 On the number of chromosomes in maize. Bot. Mag. (Tokyo) 39: 227-234. LESLEY, J. W., 1926 The genetics of Lycoperskum esculentum Mill. I. The trisomic inheritance of "Dwarf." Genetics 11: 352-3.54. 1928 A cytological and genetical study of progenies of triploid tomatoes. Genetics 13: l-43. LONGLEY, A. E., 1924 Chromosomes in maize and maize relatives. Jour. Agr. Res. 28: 673 682. 1925 Segregation of carbohydrates in maize pollen. Sci. 61: 542-543. 1927 Supernumerary chromosomes in Zea mays. Jour. Agr. Res. 35: 769-784. MANN-LESLEY, M., 1926 Maturation in diploid and triploid tomatoes. Genetics 11: 267- 279. METZ, C. W., 1925 Prophase chromosome behavior in triploid individuals of Drosophila melano- go&r. Genetics 10: 345-350. MORGAN, L. V., 1922 Non-crisscross inheritance in Drosophila mdanogastw. Biol. Bull. 42: 267-274. 1925 Polyploidy in Drosophila melanogaster with two attached X chromosomes. Genetics 10: 148-178. OSAWA, I., 1920 Cytological and experimental studies in Morus, with special reference to triploid mutants. Bull. Imp. Sericul. Exp. Sta. 1: 317-369. RANDOLPH, L. F., 1928 Chromosome numbers in Zea mays L. Cornell Univ. Agr. Exp. Sm. Mem. 117. RANDOLPH, L. F. AND MCCLINTOC~, B., 1926 Polyploidy in Zea mays L. Am. Nat. 60: 99-102. REEVES, R. G., 1925 Chromosome studies of Zea mays L. Iowa Academy of Science 32: 171- 17.5. 1928 Partition wall formation in the pollen mother cells of Zea mays. Am. Jour. Bot. 15: 114-122. SAKAMURA, T., 1920 Experimentelle Studien tiber Zell- und Kernteilung mit besondere Riich- sicht auf Form, Grosse und Zahl der Chromosomen. Jour. Coll. Sci. Imp. Univ. Tokyo 39: 1-221.