CHROMOSOME CONSTITUTIONS OF MEXICAN AND GUATEMALAN RACES OF MAIZE Barbara McClintock During the past year attention was de- voted mainly to study of the chromosome constitutions of plants of various races of maize grown in Mexico, Central America, and the Antilles. The purpose of the investigation was to determine whether differences in chromosome constitution among plants of a race and among plants of different races would reveal something about the origin of races and the degrees of their genetic relationship. Evidence that such information might be useful for this purpose had been derived from a pre- liminary study conducted with some of the well defined races of maize of Vene- zuela, Ecuador, Bolivia, and Chile, re- ported in Year Book 58. The investigation to be reported here was undertaken at the invitation of the Rockefeller Foundation and in collaboration with Dr. E. J. Well- hausen, who is in charge of the maize program within the Foundation. Dr. Wellhausen and his collaborators had made an intensive study of maize grown in Mexico and Central America. The types grown within this large area are exceedingly diverse with respect to morphological and physiological charac- ters expressed in different parts of the plant and in the ear and kernel. Through a comparative study of these many types, Dr. Wellhausen’s group devised a system of classification that considered the prob- able genetic relationships. A number of classes were proposed, each incorporating a distinctive set of plant and ear characters. Many of the maize types could be placed in one or another of these classes, either fitting closely into the scheme of a class or conforming to it in many though not all characters. Each such idealized class was given the status of a “race” of maize, and an identifying name. The maize that cor- responded most closely to the scheme of a class was considered the purest example of the racial type. Some of the races so classified are re- stricted in distribution, each being en- countered only in a specific locality. Others are grown over a Wide territory. Within one region, several different races may be grown in close proximity. By careful selec- tion of seed, these several races may be maintained year after year, even in the same field, without suffering marked al- teration of type as a consequence of hy- bridization. It is also apparent that races extensively grown in one locality have been introduced into new localities, often distantly removed, where their racial char- acters have either been maintained by care- ful seed selection or been altered by hy- bridization with maize of other types grown in the new localities. 462 CARNEGIE INSTITUTION OF WASHINGTON It seemed evident that some of the pres- ent-day races of maize must owe their origin to just such hybridization, some probably occurring centuries ago and some in more recent times. It was thought that analysis of chromosome constitutions in plants of such races, and in the suspected parents, might confirm their hybrid ori- gins. The past year’s study yielded some very good evidence in support of these conjectures; examples will be considered later. The plants whose chromosomes were to be examined came from kernels produced by plants belonging to different races. The seed representing each race was obtained from a locality where the plants exhibit the combination of characters considered most representative of the race, as defined above. The selections were as follows: 25 races from Mexico; 17 from Guatemala; 1 each from Honduras, Nicaragua, and Costa Rica; 6 from Cuba; and 1 each from Haiti, Martinique, and Trinidad. The names and geographical sources of the Mexican and Central American races are given in table 16. All the selected races were grown in the summer of 1959 at the agricultural experi- ment station at Chapingo, Mexico, oper- ated by the Rockefeller Foundation. As material for examination of chromosomes at the meiotic prophase, young tassels were removed from six plants grown from seed of each selection. These were placed in an alcohol-acetic acid fixative, which was replaced with 70 per cent alcohol after 48 hours; they were then stored in a deep- freeze unit until needed for chromosome examination. Temporary smear prepara- tions of microsporocytes at pachytene, diplotene, and diakinesis stages, stained with propiocarmine, were examined. Per- manent preparations were not made. It was impossible for one investigator to attend to them within the period of time allotted to the survey; and, as there was no advance assurance that the study would reveal anything of importance, technical assistance was not obtained. Experience showed, however, that it is advisable to have permanent preparations. It would have been instructive to make direct com- parisons of the chromosome constitutions of certain plants that were examined at very different times. The main object in examining the chromosomes was to determine the pres- ence or absence of knobs at the knob-form- ing regions in each of the 10 chromosomes that compose the haploid set in maize. (Knobs are deep-staining, heterochromatic components of the chromosomes.) If a knob was found at any one of these loca- tions, it was then necessary to determine its size, shape, and stainability, and also to find out whether both homologues carried a knob at that location. If they did, the next step was to discover whether the two knobs were alike, or in what ways they differed. If B-type chromosomes were present, the number was ascertained. Note was taken also of any conspicuous altera- tion in chromosome organization, such as a rearrangement of easily recognizable type, an enlarged chromomere at a par- ticular location, or other unusual modifica- tion. In many plants, the study was compli- cated by the presence of a large number of knobs, all or most of which were fused with one another in many cells at the pachytene stage. It was often necessary to examine hundreds of cells before finding one in which the type of knob present at a particular location could be identified. In the strain from Trinidad, fusion of knobs was so extensive and persistent that no ac- curate determination of the type of knob at each knob-forming region could be made in any of its plants. In a few other strains, similar difficulty was occasionally encountered, but only in some plants and with regard to certain knobs. In chromo- somes of most plants, it was possible to learn what type of knob had been pro duced by each knob-forming region. Knob size and shape were recorded from cells in the late pachytene stage, but it w° DEPARTMENT OF GENETICS 463 TABLE 16. The Races of Maize Whose Chromosomes Were Examined, and the Place Where Each Was Collected Name of Race A. Mexico: Arrocillo Amarillo Bolita Cacahuacintle Celaya Chalquefio Chapalote Comiteco Cénico Cénico Nortefio Harinoso de Ocho Jala Maifz Dulce Nal-Tel Olotillo Olotén Olotén Palomero Toluquefio Pepitilla Reventador Tabloncillo Tehua Tuxpefio Vandefo . Zapalote Chico Zapalote Grande Guatemala: Comiteco Dzit-Bacal Imbricado Imbricado Nal-Tel Armarillo, Tierra Baja Nal-Tel Blanco, Tierra Alta Nal-Tel Blanco, Tierra Baja Nal-Tel de Ocho Nal-Tel de Ocho Olotén Quichefio Precéz Quichefio Tardio Salpor San Marcefio Serrano Tepecintle Tuxpefio C. Honduras, Nicaragua, and Costa Rica: Salvadorefio de Costa Rica Salvadorefio de Honduras Salvadorefio de Nicaragua Where Collected Zaragoya, Puebla Etla, Oaxaca Toluca, México Irapuato, Guanajuato Chalco, México El Fuerte, Sinaloa Comitdn, Chiapas Toluca, México EI Sotelo, Guanajuato El Fuerte, Sinaloa Jala, Nayarit El Sotelo, Guanajuato Mérida, Yucatin Tuxtla Gutiérrez, Chiapas Las Casas, Chiapas Ojo de Agua, Chiapas near Toluca, México near Iguala, Guerrero Ejido de Tuxpan, Nayarit near Tequila, Jalisco near Comitdn, Chiapas Gutiérrez Zamora, Veracruz Escintla, Chiapas Tehuantepec, Oaxaca Colonia Lizaros Cardenes, Chiapas San Raymundo, Guatemala San Antonio Monjas, Jalapa Tecpdn G., Chimaltenango Tecpan G., Chimaltenango (another collection) Ipala, Chiquimula Chequiral, Quetzeltenango Ladrillo Barrio, Baja Verapaz Comalapa, Chimaltenango Santiago Buena Vista, San Marcos San Martin, Quetzeltenango Uspantan, El Quiché Cantén Sasiguan, Solol4 Cheluj, Quetzeltenango Concepcién, Huehuetenango San Sebastian, Huehuetenango Panzos, Alta Verapaz Sayaxche, Petén Santa Cruz, Guanacaste, Costa Rica Nueva Ocotepeque, Honduras Juicalpa, Nicaragua 464 CARNEGIE INSTITUTION OF WASHINGTON necessary to observe the early diplotene stage also in order to discover whether a particular knob was present in only one homologue or in both, It is not always possible to make such a determination at the late pachytene stage. A truly objective method of determining knob size and shape is not possible. The size and shape of a knob at late pachytene depend on several factors. The first of these is the capacity of a knob-forming region to produce a knob of a particular size and shape, but secondary factors may modify the expression of this capacity. In some plants, the chromosomes at late pachytene are very long, and each of the knobs is extended, often being long and relatively slender. In other plants, the late pachytene chromosomes are much shorter, and the knobs are correspondingly con- densed and widened. Growing conditions may affect the composition of the chromo- somes, and this influence also is reflected in length of chromosome, size of chromo- meres, and size of knobs. When plants are growing poorly, the chromosomes, includ- ing the knobs, are sometimes thin and may stain weakly. Despite such factors that alter the form of knobs, extensive ex- perience in observing chromosomes and their knobs at pachytene and later meiotic stages makes it possible to judge conditions in a particular plant and their effect in modifying knob appearance. To minimize bias in determining knob constitution in plants of the same selection, the six plants from each selection were not examined in sequence. Results being obtained with material from one plant were not compared with the records for sister plants until after the analysis of the plant’s knob constitution had been com- pleted. Although the method of estimat- ing knob type was purely subjective, a considerable degree of consistency was evi- dently obtained. This fact was not fully appreciated until all the data had been col- lected and the geographical distribution of each knob type plotted on maps. It was then apparent that there must have been a large measure of consistency in the evalua- tions. Otherwise, a number of previously unsuspected correlations between a re- corded knob type and its geographical dis- tribution would not have appeared on the maps. The chromosomes of maize growing in different parts of the world have been examined by many investigators. Their findings have demonstrated a marked de- gree of conservatism in the linear organiza- tion of the chromosomes. There is no evidence of perpetuation of a reciprocal translocation. Perpetuation of readily de- tectable inversions also must be rare. Only one such inversion is known to have per- sisted. The inverted segment is composed of the distal two-thirds of the short arm of chromosome 8. This inversion is readily detected in plants that are heterozygous for it. Such plants were found in four of the races examined this past season. Notably, all these races came from central Mexico. It had previously been observed in plants of one particular race now grow- ing in Bolivia. Since the inverted segment was the same in all plants, it may be suspected that this one race in Bolivia re- ceived a chromosome 8 from maize intro- duced from central Mexico. Other knobs present in chromosomes of this race also support the supposition. One other type of chromosomal modi- fication was found to be widely distributed. It is carried by chromosome 10 and in the past has been designated “abnormal chromosome 10.” “Abnormal chromosome 10” is readily identifiable because it has an extensively modified segment at the distal end of its long arm, This segment greatly increases the length of the arm, and close to its distal end there is a very large knob. “Abnormal chromosome 10” appears 11 maize plants grown in the southwest part of the United States, in various localities in Mexico, and in Central America, and was found in some plants of a Cuban race- It had been encountered in Ecuador an Peru, but only in plants from sev¢r@ isolated localities. Its very wide distribu- tion in North America and its apparently limited distribution in western South America suggest that it may have been introduced into South America in recent times, The components of organization of the maize chromosomes that do show marked grades of difference in different plants of a race or in plants of different races are the knobs, the nucleolus organizer, and some chromomeres whose locations are quite specific. The nucleolus organizer and these chromomeres may be much enlarged in plants of some races. The different types of expression shown by any of these com- ponents of the chromosome complement are genetically controlled; each type is heritable. Thus, they are useful in tracing relationships between plants of different races growing either in the same region or in widely separated regions. In the present study, attention was concentrated on the knobs and on one particular chromomere. Differences in appearance of other specific chromomeres were very effective for the purpose of this study, but the fact was not fully appreciated until after many plants had been examined. Therefore the notes on these other chromomeres are incom- plete. The same is true of differences in appearance of the nucleolus organizer, whose size may range from very small in plants of some races to extremely large in those of other races. Knobs are produced by knob-forming regions at particular locations in each of the 10 chromosomes of maize. The regions are short, and they are compound in the sense that each component of a knob- forming region acts independently in pro- ducing knob substance. Differences in knob-forming capacity of components within a region are responsible not only for the length of a knob but also for dif- ferences in width along its length. The combined result is a knob having a par- ticular size and shape. With regard to length alone, very great differences may be expressed in one knob-forming region. As DEPARTMENT OF GENETICS 465 an illustration we may take the knob- forming region at the end of the short arm of chromosome 9. Its capacity to pro- duce knob substance ranges from ap- parently none through an ascending series that results in the production of a tiny knob, a small knob, a medium-sized knob, a large knob, or a very large knob. A very large knob may extend from the end of the short arm approximately one-third the length of the arm as measured at the pachytene stage. Such a knob, since it is not only long but also very wide, is truly tremendous, Although there are these large differ- ences in knob-producing capacity at a par- ticular location, any one expression of capacity is constant, in that it continues to produce the same type of knob genera- tion after generation. This being so, a question arises as to the origin of the very different kinds of expression. Because of the types of knobs observed, and their in- dividual distributions in the examined races in both North and South America, I am led to consider the possibility that cultivated maize may have had several independent origins, from plants whose knob-forming regions had distinctly dif- ferent capacities for producing knob sub- stance. One cultivated type may have originated from plants in which all the knob-forming regions had such limited capacities that the derived maize had no detectable knobs or only a small knob in one or several of the knob-forming regions. Another type may have originated from plants whose knob-forming regions were less limited in capacity, so that the chromo- somes of the cultivated plants had either small or medium-sized knobs. Still an- other may have originated from plants whose knob-forming regions were able to produce large amounts of knob substance; in this cultivated type large knobs would be present. On this basis it may be rea- soned that much of the maize now being cultivated in western Mexico, the west coastal regions of Central America, and 466 CARNEGIE INSTITUTION OF WASHINGTON northern Venezuela was derived from original types in which most of the knob- forming regions had well developed ca- pacities for producing knob substance; and that the maize now being grown in west central Guatemala, on the contrary, was de- rived from an original type whose knob- forming regions were very limited in ca- pacity. If the study of knobs in present- day maize were extended, it might be possible to infer the nature of the knob complexes in several other early types of cultivated maize. The origin of cultivated maize is a mystery. No form of wild maize is known from which cultivated maize could have arisen directly; and none of the theories proposed to account for the origin of cultivated maize has received general ac- ceptance. Therefore, the hypothesis out- lined above regarding possible independ- ent origins of cultivated maize cannot be supported directly, even though much evidence to suggest it has developed from the study of knob constitutions in pres- ent-day maize. The types of knob com- plexes in different strains of the two closely related wild genera Euchlaena and Tripsa- cum should be explored thoroughly in order to determine whether or not they reflect differences in capacity of the knob- forming regions comparable to those dis- cussed above, and, if so, whether or not the different complexes are geographically localized. If this should prove to be true, it would support the above-proposed theory, not necessarily in its present form but in some form that might aid in in- terpreting the origin of cultivated maize and in elucidating the derivation of the different knob complexes so clearly de- picted in present-day maize. Since the over-all organization of the chromosomes is much alike in Euchlaena and maize, and since the two genera may be crossed read- ily, interchange of segments of chromo- somes, including knob-forming regions, undoubtedly has occurred between them. Therefore, the knob constitutions of Euchlaena and maize growing in the same region should be compared, in a search for some evidence of the part Euchlaena may have played in introducing compo- nents of particular knob complexes into maize. The constancy of the knob-forming ca- pacities of particular knob-forming regions was spectacularly revealed in the prelimi- nary study of races of maize of western South America. In races from the high Andean valleys of Ecuador, Bolivia, and Chile, previously under the control of the Inca Empire, one particular knob complex was present in plants of nearly all races examined. In the examined races from the high Andean valleys of Venezuela, beyond the control of the Inca Empire, a very dif- ferent complex of knobs was present. It is well known that in the past each Indian tribe carefully preserved its own types of maize; and this may account for the ex- tension of one particular complex of knobs, the Inca-Andean complex, throughout such a vast territory. Within this territory, however, a few exceptional races were found whose chromosomes had types of knobs other than those belonging to the Inca-Andean complex. Investigators had concluded, on the basis of morphological characters, that the exceptional races con- tained foreign germplasm, and that in two of them it had probably been introduced from Mexico. The types of knobs in plants of these two races support their deduction. On the basis of evidence from the recent study with Mexican and Central Amer: can maize, it is suspected that in one of the two races, Pisinkalla of Bolivia, the foreign germplasm was derived from maize 0! central Mexico, although this germplasm is now much diluted with that of indige- nous Inca-Andean maize. The types ©! knobs found in the other race, Conguil of Ecuador, suggest that its foreign germ: plasm may have been derived from maize introduced from the southern part of the state of Chiapas, Mexico, although here. too, dilution with Inca-Andean germplasm is evident. The study of knob constitutions in the lowlands of these same South American countries revealed that Inca-Andean maize had also contributed to the development of new races in those regions. Along the eastern as well as the western slopes of the Andes there is evidence of extensive mix- ing of the Inca-Andean germplasm with that of indigenous or introduced maize of the lower lands. In some of the low- land areas, such as in northern Bolivia, the Inca-Andean germplasm predominates, whereas in the southeastern lowlands of Bolivia the Inca-Andean germplasm is much diluted by that from other sources, one of which appears to be the same that contributed to maize now being grown in the Antilles. In the northwest coastal region and adjacent inland valleys of Ecuador, the Inca-Andean germplasm has become mixed with germplasm that seems to be the same as that of maize now growing in parts of Central America. Far- ther south, germplasm from other sources has been mixed with that of Inca-Andean origin. In Chile, some of the germplasm apparently stems from Central America and Mexico, but a still unknown source also has contributed to the mixed germ- plasms of the races growing in the western and coastal parts of Chile. It is now quite evident that a knowledge of knob constitutions can reveal how for- eign maize introduced into a given ter- ritory has contributed to the origin of new races, and sometimes it is also pos- sible to infer the source of the introduced maize. The regions where foreign germ- plasm has been extensively introduced are those in which the human population and culture are mixed, and probably in these regions many introductions have occurred in recent times. Thus, migration of plants with particular germplasms along certain pathways, and hybridization between them in regions of contact, as well as for- eign introductions and subsequent hybridi- zations with indigenous maize, must have contributed to the origin and spread of many present-day races of maize. DEPARTMENT OF GENETICS 467 The method of sampling in the earlier study of South American maize was quite different from that employed in the recent study. In the earlier work, the examined plants of any particular race were derived not from one locality but from several dif- ferent localities in a country where the race was grown. Thus there was a wider sampling of plants within any given area. In my opinion this method should be followed in the future, for it is more effective in revealing the migration of a particular knob (and the chromosome seg- ment carrying it) throughout a region, and its penetration into different races growing in the region. By this method it was learned that some segments of chromo- some pass through racial boundaries and filter along certain paths. Knowledge of such migration and filtration is important for an appreciation of the genetic contri- bution of particular segments of chromo- some to the characters shown by races within a region. The method also revealed the extensive migration of the B-type chromosome throughout a certain area, passing through racial “boundaries” as if no such boundaries existed. Since the B-type chromosome does not alter genetic expression of racial characters, it is easy to imagine how it might infiltrate all races within a territory. There would be no selection against it by the methods em- ployed to preserve a racial type, that is, to preserve intact a given set of characters considered desirable by the grower. In the Central Mesa of Mexico, which has a long cultural history, maize with several distinctive knob complexes must have been introduced early, for compo- nents of each of them are distributed among plants of different races now grow- ing throughout a wide area in central Mexico. It appears that early hybridization between plants having these distinctly dif- ferent original knob complexes contributed to the origin of some of the races that have been established in this region for many centuries. Archaeological evidence and 468 CARNEGIE INSTITUTION OF WASHINGTON tradition attest to the antiquity of such races. One of the knob complexes of this area seems to have been derived from an original type of maize in which knobs were absent; it may be designated the “no-knob” complex. The present-day race, Cacahuacintle, may have much of that original germplasm. Plants of this race which were examined had very few knobs, and when a knob was present it was usually heterozygous; that is, there was a knob in one chromosome but none at the same location in the homologue. The production of relatively large knobs at most of the knob-forming chromosomal regions characterized another original knob complex entering into maize of this area, Plants with this complex also con- tributed to the origin of races of maize now growing in the west central and northwest parts of Mexico. A third com- plex resembled that now predominating in a restricted area of the central highlands of Guatemala. It was characterized by the production of relatively small knobs at many of its knob-forming regions and no knobs at others. It may be designated the “small-knob” complex. In none of the examined races of central Mexico does this small-knob complex predominate, but components of it are distributed among many of them. The same statement ap- plies to what may be the remnants of a complex composed of extremely large knobs. Exceedingly large knobs produced at each of three different knob-forming regions were occasionally encountered in plants grown in this area. “Tr appears, then, that maize having distinctly different knob complexes en- tered into the formation of present-day races of the Central Mesa of Mexico. The contribution of each complex to the origin of the different races cannot be stated pre- cisely, as the sampling in the area has been much too restricted. In the sample of the Palomero Toluquefio race, however, com- ponents of the large-knob complex pre- dominated, although the germplasm that contributed them was somewhat diluted by others carrying the no-knob complex, the small-knob complex, and the very- large-knob complex. Plants of the race Cacahuacintle, growing in the same re- gion, had very few knobs, as was men- tioned earlier. These two races are con- sidered by Wellhausen and his collabora- tors to be rather direct descendants of two ancient races. They conjecture that the two original races gave rise, by hybridiza- tion, to some of the most productive maize that has been grown in the Central Mesa for many centuries. The knob constitution of one such conjectural hybrid, the present- day race Cénico, conforms with this as- sumption. Its plants have some of the large knobs that are present in Palomero Toluquefio, but their germplasm is much diluted by that of the no-knob complex. In fact, components of all the above-de- scribed knob complexes were found to be present to various degrees in plants of each of the five examined races of maize of the Central Mesa. Only one sample of maize now growing on the east central coast of Mexico was examined, of the race Tuxpefio. Therefore no conclusions can be drawn about the different types of germplasm that may be present in this general area. Nevertheless. it was clear from the knob constitutions of these plants that they had some germplasm that was either absent or not very prevalent in the examined plants of central Mexico but was found to be extensively distributed throughout Cuba. Judged by knob consti- tutions, the maize of Cuba gives little evi- dence of close affinity with that of either central or western Mexico. On the basis of morphological characters alone, it is as- sumed to have germplasm in common with maize of the east coast of South America. The knob constitutions of that maize have not been examined; but, 0? the assumption that Cuban maize con tains much germplasm derived from east: ern South America, it may be conjectured that plants of the race Tuxpefio, growing in the state of Veracruz, also have some of that germplasm. The three examined races of maize grown in the Bajio, north of the Central Mesa of Mexico—Cénico Nartefio, Maiz Dulce, and Celaya—appear to share germ- plasm with races that grow to the west and north of that region. In general, knob constitutions in plants of these races are much like those of the races Chapalote and Harinoso de Ocho from the state of Sinaloa, the race Reventador from the state of Nayarit, and the race Tabloncillo from the state of Jalisco, although some minor differences characterize the knob constitutions in the different races. Rela- tively large knobs, of similar type at any one region, are produced by many of the knob-forming regions in plants of these races; and, for the most part, the expres- sions of homologous knob-forming regions are alike. This relationship is in contrast to the high degree of heterozygosity of knob expression seen in the races grown in the Central Mesa. Nevertheless, neither in the races of the Bajfo nor in those of the northwest and west central coast ‘was any plant found that showed complete homozygosity of knob formation. Some dilution of the predominant knob com- plex was observed in each plant. A small amount of dilution by the small-knob com- plex and the no-knob complex was evi- dent in plants of the Bajfo, and a small amount of dilution by the no-knob com- plex, or by one having very few knobs, was apparent in plants of the northwest and of the west central coastal region. In the state of Nayarit one race, Jala, whose distribution is restricted almost ex- clusively to the small Jala valley, is con- sidered by Wellhausen and his collabora- tors to have originated through hybridiza- tion between maize introduced into this valley and maize that was being grown there. The foreign maize is assumed to have come from the southern part of the state of Chiapas, Mexico, or from an adja- cent region in Guatemala. Some of the DEPARTMENT OF GENETICS 469 knobs in plants from western Mexico and from southern Chiapas are alike, and many of the knobs in the Jala plants are of the type commonly found in samples from both regions. In the Jala plants, however, there is one particular knob of a very distinctive type, which was otherwise en- countered only in plants from southern Mexico, southwest Guatemala, Honduras, Nicaragua, and Costa Rica, where this type of knob is the one commonly ex- pressed by the particular knob-forming region, The presence of this knob in a place so distantly removed from the area where it regularly appears supports the above-stated assumption about the origin of the race Jala. Of all the areas examined in the recent study, southern Mexico and Guatemala proved the most interesting. This was so partly because of the more extensive sampling in that region, which made it possible to follow more closely the migra- tion of components of different knob complexes. At least four recognizably dif- ferent knob complexes have contributed to the races of maize now growing there. One of them is characterized by the pro- duction of very large knobs at many of the knob-forming regions. It is the pre- dominant complex in maize growing along the southwest coast, from the state of Oaxaca in Mexico to Costa Rica, the southernmost region from which a sample was examined. The northward flow of components of this complex into maize of the southern highlands of Guate- mala could be traced, as well as the gradual dilution of the complex by components of the other knob complexes that pre- dominate in the interior of Guatemala. Two distinctly different complexes pre- dominate in the interior. One appears to be the Inca-Andean complex. No plants were found that had only this complex, but its components are most highly con- centrated in maize of the west central highlands. Components appear also in maize grown to the east of this region and 470 CARNEGIE INSTITUTION OF WASHINGTON in maize grown to the north, in the con- tiguous state of Chiapas, Mexico; but in both these areas the complex is very much diluted, and other complexes predominate. The predominant knob complex in maize of the central highlands of Guatemala is one in which only a small or medium- sized knob is produced at many of the knob-forming regions. Components of this complex have migrated into maize of adjacent regions, and also into the one race from Honduras that was examined. As might be expected, components were found in maize of southern Chiapas. In all these contiguous areas, however, the complex was much diluted, As was men- tioned earlier, components of this com- plex were also found in central Mexico. One race growing in Guatemala, Imbri- cado, has a very limited distribution. It is considered to be related to the race Palomero Toluquefio of the Central Mesa of Mexico, discussed earlier, and the knob constitutions of its plants support this as- sumption. Two very distinctive knobs were found that otherwise were encoun- tered only in plants from central Mexico, one of them only in the race Palomero Toluquefio and other races growing in proximity to it. The evidence suggests that this race may well have arisen by hybridization of indigenous maize with maize introduced into Guatemala from the Central Mesa of Mexico. Only one sampling of maize from the northeastern lowlands of Guatemala was examined, of the race Tuxpefio. The knob constitutions of plants of this selection were similar to those of Tuxpefio plants from Veracruz, Mexico, discussed earlier. Knob constitutions were examined in six selections of the race Nal-Tel, five of them native to Guatemala and one native to the coastal region of Yucatan, Mexico. The characters displayed by plants of this race are considered by Wellhausen and his collaborators to be primitive. Plants of all six selections exhibited these assumedly primitive characters, but the individual selections differed in some other respects. Examination of chromosomes from these different selections yielded no clear evidence of the presence of one par- ticular type of knob constitution that might be considered characteristic of the race as a whole. Instead, the constitutions in five of the six selections reflected the predominant knob complexes found in plants of other races growing in the same general region. The sixth selection, that from Mérida, Yucatan, was so far re- moved geographically from other maize examined that no conclusions could be drawn about the relation of its knob con- stitution to those of other races growing in the same vicinity. Although consistency in type of knob constitution was expressed among the plants of each selection of the Nal-Tel race, very great differences were observed among the different selections. Clearly, the predominant characters that have caused these plants to be assigned to the same race need to be explored by genetic techniques. Racial classification of plants growing in widely different regions, based on similarity of morphological char- acters and presupposing a high degree of similarity in genetic constitution, is not well supported by cytological evidence in this instance, although it is so supported in some of the other cases. General Conclusions Examination of chromosome constitu- tions in maize now growing in Mexico and Guatemala has revealed an extensive mixing of different germplasms in many regions of this large area. The mixing has resulted through the flow of particular germplasms along certain paths and through direct introductions of maize having one type of germplasm into locali- ties where other types were predominant. Some of the introductions must have 0¢- curred long ago, for.components of the introduced germplasm have been spread over a wide territory. Others probably o«- curred in relatively recent times, because the influence of the introduced germplasm is still confined to a restricted locality. In the Central Mesa of Mexico, as was mentioned earlier, three. or more original types of germplasm are extensively mixed. In Guatemala, not much direct introduc- tion of foreign maize seems to have oc- curred. An early introduction of Inca- Andean maize into one part of the western highlands is suspected, as well as a much later one from central Mexico to a locality in the south. Most of the mixture of germplasm in Guatemala appears to have resulted from an inflow of germplasms from surrounding territories. In the high- lands, the flow seems to have come from adjoining regions to the south. The flow of germplasm carrying components of the small-knob complex away from Guate- mala has apparently been much more re- stricted. Probably the spread of this germ- plasm came about either when maize carry- ing it was introduced directly into a distant region or when maize of an adja- cent area, in which components of the germplasm had previously been incorpo- rated, was taken to a new locality. Flow of particular germplasms along certain paths is well demonstrated by the geographic distributions of types of knobs that are distinctive in some way, either because they have very special size and shape or because they are of a type that ap- pears infrequently at a certain location in the chromosome complement. Plotting the distribution of such knobs on geo- graphic maps, according to the geographic locations of the selections, has produced some striking illustrations of flow. In ad- dition, as was mentioned earlier, these distinctive knobs have sometimes made it possible to infer the probable source of maize that has been introduced into a different locality. The distribution of the B-type chromo- some, in the areas examined in both North and South America, is instructive in sev- eral ways. This chromosome is widely distributed in Indian maize of the central DEPARTMENT OF GENETICS 471 and southwest United States, and also among Mexican maize. It was found to be highly concentrated in some areas of the Andes, and in some lowland regions in Chile. It was not found in the examined plants of the Antilles and was present in plants of only one examined race of Guate- mala—Tuxpefio, which is native to the northeastern lowlands. The B-type chromosome is an accessory chromosome containing no genetic components that alter in any yet recognizable way either the appearance or the physiology of a plant. Plants may contain one or more B-type chromosomes, whose presence is revealed only by cytological examination. The chromosome exhibited the same, very dis- tinctive, linear organization in all plants carrying it, thus paralleling the pronounced constancy shown by the normal chromo- somes of the complement. Its wide distri- bution is remarkable. As was explained earlier, however, its flow throughout a population of plants in a given territory could be rapid, and need not be accompa- nied by a similar flow of genetic com- ponents from the normal set of chromo- somes. In fact, its independent flow sug- gests the methods that must have been used in the past to preserve intact the characters of a particular race within a given territory where many different races were grown. Before concluding, it may be in order to comment regarding the concept of race, lest some misunderstanding arise. The term race has been applied throughout this report in conformity with its usage by Wellhausen and others. It must be obvious from the preceding discussions that the appearance of particular plant, ear, and kernel characters in specific com- binations in a large number of plants within a given territory does not carry the implication of established homozygos- ity of the genetic components responsible for these characters. In fact, some of the examined samples, all selected because they most clearly expressed the combina- 472 tion of morphological characters chosen to define the race, were highly heterozygous. Even within the same race—for example, the race Nal-Tel discussed earlier—differ- ent selections were found to be hetero- zygous to various degrees and with regard to different chromosomal components. Nevertheless, it is both useful and neces- sary to classify present-day maize into distinctive races, on the basis of well de- fined sets of morphological and physiologi- cal characters kept together through propa- CARNEGIE INSTITUTION OF WASHINGTON gative methods. By means of such classifi- cation, it has been possible to apprehend degrees of genetic relationship among plants of different types growing either in one area or in widely separated areas, For a clearer understanding of these relation- ships, a precise genetic and cytological study of certain races should be under- taken, to determine and compare the types of genetic components that have con- tributed to the establishment and success of each race. BIBLIOGRAPHY Balbinder, Elias. See Rudner, R. Cocito, C., and A. D. Hershey. Transfer of DNA-glucose from parental to offspring phage T2. Biochim. et Biophys. Acta, 37, 543-544, 1960. Demerec, M. Frequency of deletions among spontaneous and induced mutations in Sal- monella. Proc. Natl. Acad. Sci. U. S., 46, 1075-1079, 1960. Demerec, M., and J. Sams. Induction of muta- tions in individual genes of Escherichia colt by low X-radiation. Proc. Symposium on immediate and low-level effects of ionizing radiations, Venice, 1959 (ed., A. A. Buzzati- Traverso), suppl. to Intern. J]. Radiation Biol., London, 1960, pp. 283-291. Demerec, M. See also Miyake, T. Gay, H. In Chapter V, Morphological organi- sation of nucleus and cytoplasm, In Biologi- cal Organisation, Cellular and Subcellular, edited by C. H. Waddington, Pergamon Press, London, pp. 110-135, 1959. Gay, H. Nuclear control of the cell. Sci. Ameri- can, 202, 126-136, 1960. Gay, H. See also Kaufmann, B. P. Hashimoto, K. Streptomycin resistance in Esch- erichia coli analyzed by transduction. Ge- netics, 45, 49-62, 1960. Hershey, A.D. The production of recombinants in phage crosses. Cold Spring Harbor Sym- posia Quant. Biol., 23, 19-46, 1958. Hershey, A. D. See also Cocito, C.; Koch, G. Kaufmann, B. P. Genetic effects of roentgen rays. J. Am. Dental Assoc., 59, 1155-1168, 1959. Kaufmann, B. P. Varying patterns of chromo- somal fine structure. In The Cell Nucleus, edited by J. S. Mitchell, Butterworth, Lon. ’ don, pp. 251-263, 1960. Kaufmann, B. P., H. Gay, and M. R. McDonald. Organizational patterns within chromo- somes. Intern. Rev. Cytol., 9, 77-127, 1960. Koch, G., and A. D. Hershey. Synthesis of phage-precursor protein in bacteria infected with T2. J. Molecular Biol., 1, 260-276, 1959, McDonald, M. R. See Kaufmann, B. P. Miyake, T. Fertility factor in Salmonella typhi- murium. Nature, 184, 657-658, 1959. Miyake, T., and M. Demerec. Proline mutants of Salmonella typhimurium. Genetics, 4, 755-762, 1960. Rudner, Rivka, and Elias Balbinder. Reversions induced by base analogues in Salmonella typhimurium, Nature, 186, 180, 1960. Thomas, R. Effects of chloramphenicol on ge- netic replication in bacteriophage A. Virol- ogy, 9, 275-289, 1959.