Superior Germ Plasm in Tobacco

By W. W. Garner, Principal Physiologist^ H. A. Allard^ Senior Physiologist, and E. E. Clayton, Senior Pathologist, Bureau of Plant Industry ^

PLANT breeding lias acconiplisliod a great many extraordinary things, but in all modesty ttie tobacco breeder must confess at the outset that he is somewhat baffled by the old problem of how to give Americans a good 5-cent cigar. There are two difficulties in tobacco breeding that condition his w^ork and that account for the present direction of his efforts. The first difficulty is the fact that the thing of greatest importance in tobacco is quality. This is true to a considerable extent for other plant crops also, but \s'ith very few crops is quality as all-important as it is in the case of tobacco. Moreover, in the case of many other crops, quality can be measured, not easily, perhaps, but relatively easily. In wheat, for example, it depends to a large extent on the quantity of protein in the grain and the baking strength of the gluten, and these can be measured in one way or another. The finer elements in the quality of tobacco cannot be measured. They depend mainly on two things—flavor and aroma. There are only two known devices capable of testing flavor and aroma, and they are not machines—the palate and the nose. This is the reason why, when so much else has been reduced to a mechanical basis, w^e still have to have coflee tasters, tea tasters, butter tasters, cheese tasters, and perfume experts. And even these are few and far between. Tliey arc born, not made.

The Present Wor\ of the Tobacco Breeder and Possibilities for the Future

IN OTHER words, these elements in the quality of tobacco are extraordinarily subtle. Not only can they not be measured as yet; by the same token, they are not under the control of the plant breeder. Since he has no notion what factors are responsible for these qual- iThe authors are mdeblcd to Harold 11. Smith, Division of Tobacco and Plant Nutrition, Bureau of Plant lûduaLry, for valuable ayslstaiice in preparing this survey.

785 ities, Le cannot breed for them. The most he can say is that they fall in the class of quantitative factors—that is, the characteristics involved are not sharply divided off from one another, like black from white; they run through a range from the very weak to the very intense, something like a scale of musical notes. The inheritance of characteristics of this sort is ahvays complicated, and it almost ahvays depends on several or many factors acting together and having complex effects in combination. The second difficulty in breeding is that these and other cliaracter- istics of tobacco are enormously influenced by soil and environment. Every living organism is influenced by environment; it starts out wîth certain inherited potentialities, but w^hether they will be realized fully, or how they w-ill be realized, depends on the environment. Tliis is both especially noticeable and especially important in the case of tobacco. A certain tobacco grown in one region and on one kind of soil may be quite different w^hen it is grown in another region on another kind of soil. So true is this that there may be very important differences between tobacco in one field and the same tobacco in the next field. Moreover, it is not known as yet just w^hat accounts for these differences. Certain sections in Cuba, for example, grow the world's finest cigar filler. A farmer in one of these sections may produce a top-notch product; a neighboring farmer, using the same seed and the same cultmâl methods, and having apparently veiy similar soil, may be unable to produce anything but a comparatively low-grade leaf. This is one of the reasons why tobacco growing pcays such uneven returns. Seemingly unimportant difl'erences in environment may make all the difference between a premium price and a very low price. The proportion of high-grade leaf in a crop, rather than the total yield, is likely to determine the value of the crop. Every element in the environment of tobacco has a similar importance—use of fertilizers, cultural methods, and methods of harvesting and curing. Much may be done, through these methods, to change the character of the final product and to affect its quahty for better or wôrse. Thus the growing of tobacco gives plenty of scope for good judgment and expertness. What does all this mean to the plant breeder? (1) Breeding new varities for improved quality of the finished product is not wdthin the scope of his operations. This is not to say that it may not be at some time in the future. Even though quality is not definitely pinned down to certain genetic factors, it may prove possible to find observable traits that are associated with the more elusive characters, and to breed for these, thus getting at quality indirectly. Some leaf characteristics that influence quality can be observed and tested far more readily than taste and aroma—for example, tissue structure, elasticity, combustibihty, and the character of the ash when smoked. Fortunately, there is no great pressure to breed for improved quality, since the varieties now in cultivation in the United States are reasonably satisfactory in both quality and yield. (2) It is possible, however, to breed for other characteristics of great importance, especially to the growler. Size, shape, number, and spacing (length of internodes) of leaves; earliness in maturing; thickness, venation, color, and nicotine content of leaves; suckering habit, which afl'ects production costs, since free-suck crin g plants are

786 more expensive to cultivate; aucl above all, disease resistance—all of these do come within the scope of breeding work. One of the fruitful projects has been the breeding of tobacco plants with a very high nicotine content, for use as a source of a valuable insecticide. At present, most of the effort is concentrated on disease resistance, since there are several diseases that take a heavy toll in tobacco growing. (3) But even in. this work, the two stubborn facts—that quality is of j)rime importance, and that environmental influences have much to do with the character of the final product—must be engraved on the brain of the tobacco breeder so that he never forgets them. It may be quite possible to breed a plant resistant to a certain disease; but if, in the process, desirable qualities are lost or modified adversely, the new plant may be valueless, even though it is completely immune to the disease. Or again, the breeder may produce a plant that is satisfactory in the environment where the work is done, but imsatis- factory in the environment where that type of tobacco is commercially grow n, or vice versa. He must, then, constantly observe three precautions: (1) Use as breeding stock plants that, in addition to the other desired characteristics, have the necessary quality; (2) carry any new strains right through the entire process of curing and manufacturing to make sure that they will meet the test of quality; and (3) produce and test his new strains in a uniform environment as alike as possible to the one in which they are deslined to be grown commercially. ^ This is the background and these are the chief limiting conditions in tobacco-breeding work. A good deal has been accomplished in this field. First, the work that is now in progress will be taken up, and then the development of present types and varieties, which came about as a result of environmental influences and 300 years of selection, will be traced. THE TECHNIQUE OF BREEDING The technique used in inbreeding and crossbreeding is not diflBcult but it requires close attention to details. Owing to the size and simple structure of the flowers, the great number of seeds produced, and the readiness with which the flowers can be handled in the technique required for self-pollination and cross- pollination, tobacco is particularly well suited for breeding operations. A single flower may produce from 4,000 to 8,000 seeds, and the entire flower cluster of a plant may yield a million seeds or more. The pollen itself is readily secured, and when carefully dried and stored in stoppered vials it will retain its active fecundating properties for several weeks, so that it can even be shipped long distances for use. Under proper conditions of storage in cork-stoppered glass vials, the seeds will retain their viability for 10 to 15 years, or longer. The abundance of seeds produced by a single capsule makes it possible cjuickly to secure an experimental progeny numbered in the thousands. An ounce of cleaned seed contains about 300,000 individuals. The tobacco flower (fig. 1) is white or colored. It is borne on a short stem and has a green calyx cut into five more or less pointed divisions known as the calyx lobes. From the calyx emerges the corolla tube, which in some species of tobacco, as in Nicotiaiia longi- florüj is very long and slender; in A^ rustica^ it is short and swollen. The corolla tube usually expands at the top into a colored limb,

787 with five parts or lobes. Within the tube, the stamens and pistils are borne. These are the essential sex organs of the flower. The pistil is the female element, and in the tobacco it comprises three distinct structures. Its swollen base, the ovary, contains the ovules closely arranged upon a fleshy axis called the placenta, which attaches them to the ovary. At maturity the ovary is greatly enlarged and becomes the seed pod, bearing the ripened seeds. ^ From the top ojf the ovary, a long, slender stalk or style rises. This ends in a blunt, more or less two- corolla iobes lobed swelling, the stigma, \èL \ near the opening of the corolla tube. When the flower opens, this stigma is usually more or less glutinous or stickjr so that any pollen reaching it will adhere. Around the stigma and its stiffma supporting stalk or style, filament five separate stamens are closcljr arranged. Each of these in turn consists of a slender stalk or filament bearing an enlarged saclike portion at the top laiown corolla lube as the anther. The anther cali/jc lobes consists of cells which open lengthwise by a slit when they are mature, to liberate the fertilizing pollen. This falls on the stigma, which placenta they surround more or less closely. The filaments of ovarj/ ouule^ the anthers are attached to the corolla tube. The anthers are the male structures of the flow^ers. The tobacco flower is so constructed that the anthers normally pollinate the

FIGURE 1.—Longitudinal section through a tobacco flower of stigma of the same flower— tho White Burley variety, showing the corolla, calyx, sta- that is, the flower is self- mens, pistil, and ov ules borne upon a fleshy axial placenta (drawn by II. A. Allard). fertilized. However, pollen from other flowers, other plants, or even other races, varieties, or species of tobacco growing in the vicinity may fall on the stigma or be carried to it by insects or other pollinating agencies. To be absolutely sure that the flowers are self-fertilized, therefore, the breeder has to keep out any foreign pollen by covering the blossoms with bags, screening, etc. Bagging of the seed heads is also necessary to insure close fertilization. For this purpose, strong paper bags of the satchel-bottom type, capable of resisting wind and rain throughout the season, are recom- mended. For large, un trimmed seed heads, bags of 16- to 20-poimd vohime are best. ^ If the seed heads are trimmed to 25 or 30 capsules, bags of smaller size may be used.

788 The first flowers to open xisually develop the largest capsules and seeds. The outside flowers of the seed head, arising late in the season when the vigor of the plant is waning, are likely to be smallest. Tests have shown that trimming the seed head to 25 to 30 capsules may result in heavier and better seed. If the maximum amount of seed is desired from a plant, however, untrimmed seed heads should be bagged. When a plant is bagged, seed pods and blossoms that have already opened should be removed, as foreign pollen may have been introduced by insects. It is exceedingly important to inspect the flower head very carefully to remove all caterpillars, since these will feed on the flowers and capsules in the bags. Lead arse- nate dust or combinations with paris green maj'also be used. The bag is then placed over the flower head and its lower edge is crumpled snugly around the stem and tied (fig. 2). When large quantities of selfed seed are secured, a superior grade of seed can be obtained by separation of the heavy from the lighter seed and chaff, using a cleaning or blowing deface. Making Crosses This description ap- plies when flowers are to be self-fertihzed. The operation of artificial crossing or hybridization requires other special manipulations. When a desirable plant is selected as the mother of the seed, all open FIGURE 2.—Placing paper ba^ over flower head after it has been trimmed and the upper leaves of tlie ¡ilant have been removed. blossoms and seed pods The mouth of the bag is drawn about the stem and securely tied are carefully cut away, in place. By this means cross-pollination is effectively prevented. and only closed buds, those chosen to receive pollen, are left. These should be the buds that will open tiieir corollas the next day; the large size and the pink color of the tip will readily show which buds are about ready to open. Since self-pollination must be avoided in this case, these flowers to receive foreign pollen must be emasculated; that is, all

789 the anthers or male organs must be removed. This is done by slitting the corolla open with a sharp knife or scalpel and taking the anthers out with a pair of forceps. The hands, scalpel, and forceps should be sterilized with alcohol before the buds are opened, especially if a series of emasculations are being made on different types, to prevent possible contamination with foreign pollen. It is very important that no bud be saved for crossing if it is found that any of the anthers have opened prematurely, as sometimes hap- pens, so that pollen might already have gotten on tlje stigma. And when the flower head is finally ready for bagging, only those flowers that have been emasculated should be ready to open the next day. Any others that look as though they might open should be removed. The treatment of the flowêrs chosen to serve as the mole parent by furnishing pollen is a simple matter, since the flower head is merely bagged after all blossoms that had opened previously are removed. The morning after emasculation, when the stigmas are normally stick}^ and in a receptive condition, the transfer of pollen is made from the male flowers. To do this, the male flower may be picked from the plant and applied to the stigma of the emasculated flowêr so that an open anther cell \vill rub its pollen upon the sticky surface, or all the anthers may be thus used. This male flower may then be thrown away and another male flower removed to pollinate other blossoms. Once pollination has been accomplished, the individual flower stem should receive a small tag giving data as to variety, individual, male parent, etc. These hand-polUnated flowers are then covered with bags again. It is important to see that all buds on the mother plant that give evidence of opening within the next few days are cut away lest their pollen contaminate the cross before final fertilization with the selected foreign pollen has been completed. It is an advantage to make crosses on the first blossoms of the flower head to open, since at this time flowering is in its fullest vigor, and there is less chance that blossoms will be shed or drop off than there is later. The pollination of decidedly immature blossoms that will not open for several days should not be attempted. The transfer of pollen can be done with a small brush, but if difl'erent crosses are made, there is danger of contamination from adhering pollen grains that cannot be seen with the naked eye owing to their microscopic size. If crosses are made between compatible strains or varieties, true fertilization of the ovules soon imparts a stimulus to the ovary, which quickly enlarges and produces fertile seed. In crosses between some species, the foreign pollen may be almost or quite sterile for that flower, and the capsule wdll grow little or not at all, and perhaps ultimately fall off. The pollen of other species, even though it is sterile or nearty so as far as fertilizing the ovules is concerned, may yet lead to a fairly normal growth of the ovary itself, though few or no viable seed are produced. This is a growth influence of the pollen, distinct from fertilization of the ovules, and it is know^n as parthenocarpy.^ Because of the small size of tobacco seed, it is more difficult thiin might be supxôsed to avoid, occasional accidental mixing of seeds in handling and planting operations. A few foreign seeds, or even one, might ruin a whole genetic experiment. In cleaning the seed, therefore, care is required to insure that no individuals remain on the screen 2 From the (Jreek rooi, parlhcno, virgin, or without being fertilized.

790 or ia the blowing* apparatus when a second lot is introduced. In propagating, the seed may be planted in pots or flats in the greenhouse, or, if seedlings are required in large numbers, planting can be made in the usual type of cold frame; but regardless of the method of planting it is highly important that adequate safeguards be employed to prevent transfer of seeds from one planting to another by insects, wind, or other means. Stress has been laid on the marked effects of environment on growth habit, size characteristics, and the commercial quality of tobacco, and on the importance of making sure that the material for study is grown under a uniform environment. It is essential also that such cultural details as spacing of the plants in the field, the height of topping if this operation is carried out, the stage of maturity at which the plant or individual leaf is harvested, and the method and conditions of curing be carefully standardized. Control of these environmental and cultural factors cannot be disregarded if dependable results are to be obtained. AVAILABLE BREEDING STOCKS One other point should be mentioned in connection with the technique of breeding operations—maintaining stocJvS of viable seed of all varieties for use at any time in breeding experiments. At the present time, collections of seed of numerous strains representing the standard varieties, and including of course the new productions, are maintained in the Bureau of Plant Industry and some of the State agricultural experiment stations. Many of these are pure lines. There are also limited collections of foreign varieties, and these are being extended, especially in the direction of increasing stocks for breeding for disease resistance and other special purposes. A partial list of varieties and strains covered in these collections, and their points of special value, is given in table 3 (p. 829). Because of the vast number of strains in use, it would obviously be a huge imdertaking to bring together and maintain seeds of all of them in a single collection. As collections of disease-resistant material are assembled and tested, it is extremely important that provision be made to keep all of this material available for future study. Varieties resistant to disease in A arying degree have been reported by previous investigators, and in some cases it is now impossible to get authentic seed of these strains. In breeding work, it is not practicable to consider at any one time all characters of possible value. Diseases at present considered minor may later assume importance, and new ones may appaer, just as did wildfire and more recenth' mildew. Further, as improved varieties resistant to disease are developed, authentic sources of seed must be established, beciiuse it may be expected that resistant and susceptible strains will be identical in appearance, and in the hands of growers mixtures are likely to occur. It appears that these needs can be met only by continued close cooperation between State and Federal authorities.

INVESTIGATIONS IN PROGRESS ON DISEASE RESISTANCE The increasingly heavy losses from disease in recent years make breeding for greater disease resistance imperative, and major attention at present is being given to this problem. Earher studies indicated clearly both the possibilities and the complexity of the problems.

791 Partly in cooperation and partly on an independent basis, the Bureau of Plant Industry and most of the important tobacco-growing States are conducting investigations in the production of desirable disease- resistant varieties. These projects are listed in table 4. Cooperative studies are now under w^ay betw^een the Bureau of Plant Industry and the States of Connecticut, Massachusetts, Wisconsin, Maryland, Tennessee, West Virginia, North Carohna, South Carolina, and Georgia; and as the work is not yet fully organized, it is expected that additional cooperative arrangements will be made in the near future. The objective of the program here briefly outlined is (1) to provide a sound fundamental background for the entire problem, and (2) with the aid of this information, to carry forward specific disease studies applicable to limited areas. Nicotiana tabacum is a most variable species, and innumerable types and varieties are grown in all parts of the world. The first under- taking has been to secure adequate seed material, and to study the occurrence of disease resistance in this material. Certain varieties are known to be resistant to certain diseases, but whether these are the only resistant types, or the best, is not known. So far, in addition to domestic varieties, 140 foreign collections distributed among the following countries have been studied: Africa, Argentina, Brazil, British Guiana, Dutch Guiana, Colombia, China, Cuba, Honduras, Java, Japan, Mexico, Puerto Rico, Peru, Philippine Islands, Union of Soviet Sociahst Eepublics, Sumatra, Turkey, Uruguay, and Vene- zuela. Other foreign collections are constantly being received and included in these studies. In conducting this work it has been necessary (1) to develop methods for detecting and measuring disease resistance, the latter being especially important, since in few instances does resistance approach immunity, so that it is necessary to determine which of a number of varieties possesses the maximum degree of resistance; (2) to test each collection separately for each disease; and (3) to investigate the mode of inheritance of this resistance. In the following discussion, (1) and (2) will be treated together. As promising lines of resistant plants have been detected by means of the laboratory and greenhouse studies, they have been planted in field plots in the several areas where the diseases are most serious. These field tests are depended on to furnish information as to the value of different degrees of resistance imder field and seedbed conditions, and also to supply information as to the morphological and quahty characteristics of new" varieties. Studies on the occurrence of disease resistance are by no means completed and conclusions as to the best parent stocks for the breeding of new types more resistant to disease than our present varieties may be materially modified by work in progress. It was only a few years ago that discussions of tobacco breeding for disease resistance were practically limited to black root rot. If this is kept in mind, it will be evident in what foUow^s that the possibilities have been rapidly and strikingly extended.

Occurrence of Disease Resistance and Methods of Testing Granville w41t {Bacterium solanacearum) is a bacterial disease—the browai rot that also affects the potato, tomato, eggplant, and pepper. Workers in the Netherlands studied resistance by inoculating plants in the stems with bits of diseased tissue. However, under natural conditions this disease invades the plants by way of the roots, as a 792 result of soil-borne infection. Comparison of soil and stem inoculation shows that several varieties are resistant to soil inoculation and susceptible to stem inoculation. Resistance to soil inoculation indicates resistance to invasion, that is, morphological resistance. Some varieties, however, resist the development of the disease even after the organisms are inside the plants. This resistance is referred to as physiological. Resistance to wilt is modified by age, increasing as the plants become older. The development of the disease is also favored by rapid plant growth and high temperatures (80° F. and above). All domestic varieties tried are highly susceptible to wilt, but this is not the case with foreign varieties tested, about one-third of which

FIGURE 3.—Resistance to Granville wilt. The row in llae foreground was planted with Cash, a standard flue-cured strain, and the row to the left with a resistant selection from a foreign variety, 79A. At the time photographed, 15 percent of the plants had been killed in the resistant row and 92 percent in the Cash row. show slight to moderate resistance. The most resistant strain obtained so far is a selection, No. 79A, from an unknown variety secured in Java. This selection was tested in the field in 1935 under severe wilt conditions and 25 to 30 percent of the plants were dead at the end of the season, while check plantings of a susceptible domestic type were 90 to 95 percent killed. Figure 3 shows the ability of the resistant strain to produce a fair crop under conditions such that the check planting of Cash was almost a complete loss. Strain No. 79A is not a flue-cured type and does not approach immunity in its resistance. Mildew or blue mold {Peronospora tabacina) is a disease due to the type of fungus known as a downy mildew. Satisfactory inoculation is no problem with this disease, because once the organism is introduced into a greenhouse or plant bed, and favorable conditions

38143°—36- -51 X 793 prevail, the fungus naturally spreads so that every plant is infected. The ^ disease is^ very dependent on environmental conditions, and studies on resistance can be^ conducted only when temperatures between 55^ and 70° F. prevail, in combination with high humidity. Resistance increases with age of plants and rapid-growing plants aro most susceptible. All domestic varieties are Completel}^ susceptible as far as known, but a few foreign sorts are slighly to moderately resistant. The most resistant of these is No. 57, a variety obtained from Argentina under the name Chileno Corren tino. This variety gives a fair stand of plants under conditions such that susceptible sorts are completely destroyed. Using older plants, leaf injury is greatl}^ reduced witli the resistant type, and in one seedbed test the susceptible variety showed much stem injury while the stems of No. 57 remained clean. A higher degree of resistance, however, is very desirable. Mosaic. The first virus disease ever transmitted experimentally from a diseased to a healthy plant was tobacco mosaic. It is the one disease studied where, within reasonable limits, the method of inoculation, age, condition of plants, and environment make little difference in the results. However, in studying the progeny from crosses between resistant and susceptible types, simple inoculation and observation is inadequate, because plants showing different degrees of resistance are obtained. By inoculating healthy plants with the juice from each mosaic-resistant plant, however, it is possible to dift'erentiate clearly between (1) resistant plants carrjdng the virus and showing faint leaf symptoms and (2) plants apparently free from virus and showing no leaf symptoms. All domestic varieties are susceptible, and the same is true of all foreign collections tested excepting Ambalema. This variet}^ is highly resistant to ordinary tobacco mosaic, as reported by NoUa and Roque (1933). Figure 4 shows an Ambalema plant growing beside a plant of Maryland Broadleaf, both having been inoculated with mosaic. The Ambalema variety is being extensively used in the breeding work now in progress for control of mosaic. Wildfire {Bacterium tabacurn). In contrast to mosaic, this disease, due to bacteria, is one of the most difficult to study with respect to resistance. Seedling inoculations have been much used, but our results with seedlings were not comparable with those obtained from mature plants of the same varieties under field conditions. Hence, since most wildfire damage occurs in the field, it is necessary to work with field-grown plants. In evaluating wildfire resistance, it is further necessary to study separately (1) resistance to infection and spread of the bacteria througli normal leaf tissues, and (2) resistance of leaf tissues to water-soaking. It is water-soaking that is responsible for the development of the destructive, epidemic type of w^ildfire. To study resistance to water-soaking, leaves of the same maturity grown under uniform field conditions are exposed to a uniform water spi'ay and the time required to produce water-soaking is recorded for both upper and lower leaf surfaces. Results indicate that difi-erenccs in susceptibility to water-soaking dependent on both cultural and heritable factors can readily be measured by this method. Contrary to general belief, most domestic varieties are moderately resistant to wildfire under field conditions. On the other hand, about two-thirds of the foreign varieties are highly susceptible. Further stud^^ will be required before conclusions can be arrived at 794 regarding tlie possibility of increasing the wildfire resistance of our present varieties. So far no type immune to wildfire has been discovered. Blaclifire {Bacterium angulatum). The same complicated situation indicated for wildfire also holds true for blackfire or angular leaf spot, another bacterial disease, and similar methods are being used in studying resistance. Most domestic varieties of tobacco are moderately resistant. Studies in measurement of disease resistance are in progress with root knot (Heterodera marioni), a disease caused by a minute eelworm

FIGURE 4.—Mosaic resistance: A, iWaryland Broadleaf; B, Ambalema. Both plants were inoculated with mosaic and the severe dwarfing produced by the disease is noticeable with the Maryland Broad- leaf; the Ambalema plant did not become infected. or nematode, and with the following fungus diseases: Black root rot (Thielavia basicola), stem rot {Sderotium rolfsii) and black shank (Phytopthora nicotianae). Black root rot. As reported by Johnson (1930), domestic varieties vary from highly susceptible (Judy Pride Burley) to moderately resistant (certain cigar types and resistant hurleys), and foreign varieties range from highly susceptible to practically immune (Xantjd and other Turkish types). Figure 5 shows the marked difl'ercnce in growth on diseased soil of susceptible and moderately resistant

795 varieties. In breeding projects now under way for root rot resistance, Johnson Resistant Standup Burley, resistant cigar leaf varieties, and Turkisli varieties are being used. Root knot. Extensive tests so far indicate that all domestic varieties tried are susceptible with the exception of a strain of Orinoco known as Faucette Special, which shows moderate resistance. Since this is a flue-cured sort adapted to the area where root knot is most severe, it may prove of considerable value. One foreign variety, White Honduras, is also moderately resistant. There is great need for a tobacco that is highly resistant to root knot. Black shank. Reports by Tisdale (11) ^ that as far as known all domestic varieties excepting the Big Cuban and ordinary Cuban are susceptible have been confirmed by further test. Studies of foreign collections have not yet been completed. No. 301, the resistant

FIGURE 5.—The growth of susveptihle ;in(i moiierateh rei^ist mt strains of tobacco on ^oil infested with black root rot. No. 22, Connecticut Broadleaf; iSio 2¿, selection from a cross between the susceptible Maryland Broadleaf and Johnson Resistant Burley, No. 24, Maryland Broadleaf. Nos. 25 and 2fi, other resistant lines. cigar-leaf type developed by Tisdale in Florida, is highly resistant to black shank and is being used as the resistant parent in breeding work now in progress. Figure 6 illustrates the ability of No. 301 to produce a healthy crop of tobacco under conditions such that the susceptible variety was a complete failure. Inheritance of Disease Resistance From the genetic standpoint, resistance to disease has been investigated by Johnson (9) in the case of Thielavia root rot, and by Tisdale {11) in black shank {Phßophthora). Unfortunately, in these diseases resistance appears to be quantitative in nature and controlled by multiple factors. It behaves genetically in a manner similar to that 3 Italic numbers in parentheses refer to selected references, p. 828.

796 observed in studies of size characters. The studies, however, are being continued with the different diseases, and they are being facilitated by the use of definite, controlled methods for measuring resistance. Thus, the method previously discussed for measuring resistance of leaves to water-soaking in the wildfire and blackfire studies gives data capable of careful analysis. Close comparisons of this sort are very difficult to secure under natural field conditions. Again, uniform procedure in inoculating plant populations avoids the compli- cations presented in the field by the presence of susceptible individuals that escape natural infection.^ The present studies on inheritance of resistance have progressed far enough to lead to definite conclusion only in the case of mosaic.

FIGURE 6.—Black shank susceptible and resistant varieties growing (under artificial shade) on infested soil. The diseased row in center is the susceptible Connecticut Round Tip; the healthy rows on either side are the resistant 301. (Courtesy Florida Agricultural E.\periment Station.) Here counts on a large F2 population from a Maryland Broadleaf- Ambalema cross indicate that mosaic resistance is dependent on three pairs of recessive factors. This most favorable circumstance, if confirmed by further work under way, makes possible back-crossing of F2 plants, with the certainty that these will be homozygous for resistance. Some Essentials for Further Development of the Disease'Resistance Program. Breeding for disease resistance offers a promising means of effectively combating some of our most destructive diseases, and is receiving more and more attention. It is worth while, then, to point out briefly some of the major obstacles to be overcome. * See Introduction» p. 129, for discussion of disease inoculation in breeding work. 797 First and most obvious is the need for stocks more highly resistant. The mosaic-resistant Ambalema probably^ approaches most nearly to the ideal sought, as it is practically immune to common mosaic and the number of factors involved is small enough to permit genetic analysis. It is doubtful whether this satisfactory situation exists with any other resistant stocks now available. To aid in meeting this need, the Bureau of Plant Industry is at present collecting material in Mexico, Central America, and South America—regions that are the native home of the species and the centers where the greatest diversity of types is to be found. A second need that has been emphasized by results from previous work, and has been mentioned earlier in this section, is the need for increased attention to quality. Resistant types have been selected chiefly on the basis of visible characters. Sometimes they looked like the desired type but were actually quite different in qualit}^. At present it must be accepted that it is impossible to select for quality in the growing crop, and repeated back-crossing to the parent that possesses the desired quality is the only method that gives promise of solving this problem. The fact that resistant types lacking even slightly in the desired quality will not be commercially acceptable probably constitutes the greatest single obstacle to the success of the breeding ¡program. Another important thing that is at present being largely ignored by tobacco breeders is the need to take into account the reaction of parent stocks to all diseases and not merely to the single disease under consideration. This omission may have serious consequences. Thus it has not been recognized that certain foreign types highly resistant to black root rot are also highly susceptible to wildfire, to which most domestic varieties are moderately resistant. At present wildfire occurs in serious form only in limited areas, but past history indicates that the disease is capable of developing in many other areas, and consequently introduction of varieties highly susceptible to it would not be safe. Tests during 1935 have shown that one recently developed variety resistant to root rot is distinctly less resistant to wildfire than the original standard type. The complexity of this problem may be further illustrated by table 1, which shows the disease resistance of two varieties, Cash and No. 79A, now being used in the production of a wilt-resistant flue-cured type.

TATíLF. 1.— Resislance of Cash and No, 79A to different âiseamn

Black root Ftisariurii Black Variety Wilt Wikifire Blackfiro rot wilt Mosaic shank

Cash. Slight Moderate. _. Moderate. Slight Moderate... None None. 79A_... Moderato. Slight do Moderate. __ Slitiht -._do ])o.

Study of the above analysis shows that the F2 generation from this cross, which is now being tested, might, if studied for resistance to wilt only, give wilt resistance in combination with susceptibility or moderate resistance to black root rot, Fusarium wilt, and wâldfire. On the other hand, by careful testing it might be possible to continue with lines possessing the maximum of resistance to all four diseases. The fact that at the present time wilt is the only disease of the four

798 that is a serious problem in the area concerned would by no means justify ignoring the other diseases, since all are potential hazards. A list of projects now under wa^ on the breeding of disease- resistant varieties of tobacco is given in table 4 (p. 830), and a list of Department of Agriculture and State experiment-station workers on page 830. in considering the future development of this work, it is to be remembered that losses fi'om disease have increased rapidly in recent years, and, since our standard types are highly susceptible to most of the diseases, there is a pressing need for introducing resistance, without sacrificing quality, for practically every type of tobacco and every producing area. General estimates of the losses from tobacco diseases are not available, but two specific examples may be cited. In North Carolina alone root knot and wilt cost the growers at least $5,000,000 annually. In Pennsylvania wildfire reduces the value of the crop in some years as much as oO percent. The mag- nitude of the breeder's task needs no emphasis. Up to now, investigators have attempted to develop varieties resistant to individual diseases, but the logic of combining in one variety resistance to more than one disease cannot be escaped and must be met in the future. Moreover, since in research in disease resistance we are dealing not with isolated needs, but rather with a long-time major program for varietal improvement, it is obviously desirable to coordinate future investigations closely. Much of the actual work must be carried forward in local areas because only under these conditions can progenies be studied for both disease resistance and quality—and improved resistance at the expense of quality is a barren achievement. Coordination of thèse individual projects, however, will make possible the use of only the most desirable parent stocks, and concentration on the best methods. Finally, it is suggested that in the future, new varieties should be subjected to most rigid tests before they are released, since only in this way can the desire of the grower for protection against disease and that of the manufacturer for stable quality be adequate^ safeguarded. OTHER INVESTIGATIONS AND PROBLEMS IN TOBACCO BREEDING Although the work of the tobacco breeder is largely concerned with disease resistance for the present, this does not by any means tell the whole story. A number of other investigations and experiments will be summarized before discussing the outlook for the future. The Kentucky Agricultural Experiment Station is endeavoring to develop a strain of White Burley of improved suckering habit by crossing root-rot resistant Burleys with the variety known as One Sucker, the objective sought being to reduce somewhat the cost of labor in growing the crop. When tobacco is topped or disbudded, as is usually done in practice, shoots or suckers develop in the axils of the leaves and these must be removed from time to time as they appear in order that the leaves may attain full development. At the Florida Agricultural Experiment Station, investigations are in progress to determine what elements of quality, if any, are linked with resistance to the black shank disease. Studies also are being made on the inheritance of leaf size and ninnber of leaves and on the number of factors involved in expression of the disease-resistant character. 799 A genetic study of inberitance of resistance to Thielavia basicola is in progress at the West Virginia Agricultural Experiment Station. In cooperation with the Massachusetts and Connecticut Agricul- tural Experiment Stations, the Bureau of Plant Industry is seeldng to standardize the Havana Seed type of leaf in the Connecticut Valley as far as this may be possible. The effort is to i)roduce one or two strains meeting tbe requirements of both farmer and manufacturer in all essentials, including root rot resistance, yield, and quality, when grown on either infested or on healthy soil. A good deal of experimental w^ork has been done in breeding for nicotine content, both high and low; and results in the case of extremely high nicotine content promise to be of great value in the development of cheaper nicotine sprays to combat insect pests in farming operations.

THE increasingly heavy losses from disease in recent years ma\e breeding for greater disease resistance imperative, and major attention is at present being given to this problem^ The objective of the program is to provide a sound fundamental background for the entire problem, and to carryforward specific disease studies applicable to limited areas. In conducting this wor\ it has been necessary to develop methods for detecting and measuring disease resistance; to test each collection separately for each disease; and to investigate the mode of inheritance of this resistance. Only a few years ago, discussions of tobacco breeding for disease resistance were practically limited to blac}{ rot. If this be \ept in mind, it will be evident that the possibilities have been rapidly and &tri\ingly extended.

Low Nicotine Content Nicotine is the characteristic alkaloid of tobacco to which it owes its habit-forming properties, and relative nicotine content affords the best measure of the strength of a tobacco product. The quantity of alkaloid in the tobacco leaf is influenced by many factors, of which heredity is one of the more important. As far as is "known, the relative content of nicotine is not definitely correlated with any important element of quality in the leaf other than strength, at least within rather wide limits. Because of the marked effects of environment and cultural practices, the nicotine content of tobacco usually depends more on the commercial type and the grade of the leaf than on the variety or strain of seed used. The dark fire-cured and dark air-cured

800 types are high in nicotine, and the chewing grades of these and other types as a class contain more nicotine than the corresponding smoking grades. Among smokers as a whole, there are some who prefer strong cigars, cigarettes, or smoldng mixtures, but apparently the majority favor mild products, and many desire products with a very low content of alkaloid. The problem of reducing the nicotine content of manufactured tobacco products, both by partial removal of the alkaloid from the cured leaf or the finished product and by developing low-nicotine varieties or strains, has recently attracted considerable attention. On the other hand, since tobacco is used chiefly because it contains nicotine, there would seem to be no purpose in producing a tobacco containing only negligible quantities of alkaloid except for use in blending with strong types of leaf. As early as 1907-9 {4) two strains of the Cuban variety characterized by a low nicotine content were obtained by selection in the Bureau of Plant Industry. As grown in east Texas in 1909, 30 individuals of one strain showed a range in percentage content of nicotine from 0.50 to 0.18, with a mean of 0.34, and the corresponding values of the second strain were 0.72, 0.36, and 0.50—whereas other control strains showed values three to five times as high. The low-nicotine strains have been grown repeatedly under various conditious and invariably have shown a greatly reduced content of nicotine as compared with ordinary strains of Cuban and other varieties. These strains have not been grown commercially for the reason that the Cuban variety is a poor yielder and is not used in this coimtry for production of cigar fdler. Recently strains of the Cuban tobacco containing an exceedingly low content of nicotine as w^ell as cigarette strains moderately low in nicotine also have been produced at the Kaiser Wilhelm Institute for Breeding Research at Müncheberg, Germany. It was found that inheritance of the low-nicotine characteristic is remarkably constant, only about 1 percent of the individuals in a population of 4,000 having a high content of nicotine. These strains were obtained by selection based on a rapid method for determining in the field the approximate content of nicotine in a very larger number of individuals. In the Bureau of Plant Industry, United States Department of Agriculture, work is in progress in cooperation with the Pennsylvania Agricultural Experiment Station in developing low-nicotine strains of the cigar-filler type by hybridization of local varieties with a low^- nicotine strain of Cuban, and in cooperation with the Maryland Agricultural Experiment Station in development of similar strains of the Maryland cigarette type by selection in local varieties. Similar work is under way at the Kentucky Agricultural Experiment Station in the production of a low-nicotine strain of White Burley by means of crosses of Burley with a low-nicotine sort obtained from Baur in Germany. Production of strains of tobacco with an extremely low nicotine content is entirely feasible. But according to present commercial standards for cigar leaf, nicotine content is not directly taken into account in judging quality. Before strains very low in nicotine could be commercially useful, manufactin-ers would have to develop a special market for new brands of mild tobacco products.

801 High Nicotine Content High nicotine content is another story. Nicotine has long been a valuable insecticide, and it is used in large quantities for control of certain insect pests of plants, particularly the aphis, and for control of mange or scab on sheep and cattle and certain parasites on poultry. The present supply is obtained almost exclusively from tobacco stems and low-grade leaf. It appears that the retail price is an important limiting factor in the use of nicotine, and recently there has been considerable interest in the possibility of growing high-nicotine tobaccos for direct use as raw material for extraction purposes. Although the large-leaf varieties of ordinary tobacco, Nicotiana tabacum, will produce a high tonnage of dry matter, the content of nicotine is relatively low. Another species, N. rustica, produces a lower tonnage of dry matter but has a high percentage content of nicotine (fig. 7 ). This species suckers badly, which adds materially to the cost of production. There are many varieties of N. rustica which differ greatly in size and to a lesser extent in content of nicotine. From a comparison of a large number of forms from various parts of the world, it has been found that a large-leaf variety known as Brasilia, seed of which was obtained some years ago from Italy, perhaps is the most promising from the standpoint of output of nicotine per acre. With this form, yields of 150 pounds of nicotine per acre have been obtained under favorable condi- FiGUKE 7.—Brasilia variety of Nicotiana rustica. Tliis Is one of the largest sorts of rustica and produces a tions, and in exceptional cases relatively high yield of nicotine. Brasilia appears the yield has approximated 200 to be the most promising rustica now available for production of tobacco for nicotine extraction purposes. pounds. This variety, in common with all others tried, possesses undesirable habits of growth, especially with respect to suckering. The goal sought in the work with A^". rustica has been an average yield approximating 200 pounds or somewhat more per acre under conditions that will avoid high production costs. Because of practical difficulties, particularly that of standardizing their strength and their rapid deterioration, home making of sprays is a rather complex problem. In any case, theproblems in econom- ically growing the raw material remain essentially the same as for commercial extraction purposes. The Bureau of Plant Industry coop- erating with the Pennsylvania Agricultural Experiment Station is endeavoring, through crosses between varieties of N. rustica and crosses of this species with ordinary tobacco and other species, to obtain new forms capable of producing a high acre yield of nicotine and having desirable growth habits, especially with respect to suckering.

802 THE OUTLOOK FOR THE FUTURE For the immediate future, it appears tliat perhaps one of tlie more promising courses for the breeder to take would be to limit his efforts largely to increasing the percentage of high-grade leaf in the crop without modifying quality characteristics in the present varieties, and with only incidental attention to total yield. In the three major cigarette types of leaf, burley, flue-cured, and Maryland, and in the cigar-wrapper and binder types, probably the best opportunity for achieving this objective lies in well-directed efforts toward increasing the number of desirable leaves on the individual plant. As will be pointed out later, significant changes in this direction already have occurred, as illustrated by the replacement of the old broadleaf, drooping sorts of White Burley by the so-called stand-up sorts, the latter yielding an increased percentage of high-quality cigarette leaf. Increase in number of leaves per plant usually is associated with a slight reduction in leaf size and a more ei^ect growth habit. Increas- ing the number of leaves per plant tends to reduce the proportion of "tops," which for the commercial types mentioned do not yield leaf of the grades used for cigarettes or cigar wrapper or binder. Solution of these immediate problems, as well as that of obtaining satisfactory disease-resistant strains and varieties—and, in fact, the great majority of problems in varietal improvement—is subject to difficulties and limitations, some of which have been discussed. The importance of these limitations should not be overlooked. With respect to hybridization, it is the general consensus of opinion among those who have had the most experience, including in particular the Netherland East Indies investigators in this field (12), that anything in the way of promiscuous crossing of sorts differing widely in genetic composition will ordinarily be hardly more than a waste of effort from a practical standpoint. When hj^bridization is necessary, this should be limited as far as possible to local varieties or strains that are closely related genetically. Practically all the variations in desirable characteristics of tobacco, from the standpoint of both the farmer and the manufacturer, are of a quantitative rather than a qualitative nature so far as inheritance is concerned. They are governed by a relatively complex genetic mechanism involving many Mendelian factors. It is difficult to eliminate or control the subtle influences of envii'onment and cultural practices on the important properties of the leaf, especially in dealing with complex crosses. Moreover, as already pointed out, quality cannot be judged until the leaf has been carried tlirough all the processes required to fit it for manufacture. Without necessary precautions in this direction, even continued selection within the variety is not without its dangers. Netherland East Indies investigators have found that different strains may be identical in visible characteristics and constant in their progenies but different in quality characters not recog- nizable in the field. Again, even in crosses between closely related pure lines, these investigators found that selfing for about 8 generations was required to obtain uniformity in leaf shape, while apparently about 20 generations would be necessary to secure uniformity in quality. 803 Despite the serious difficulties involved, the situation is not hope- less. The great genetic variability existing in Nicotiana tabacum affords a wealth of material for securing new combinations in dealing with the problems of satisfactory yield, quality, and disease resistance. Obviously the smaller the number of gene differences involved in a cross, the simpler will be the task of securing the desired recombination in a superior product. Progress w^ill be facilitated, of course, by growing the experimental material in large numbers. Backcrossing combined with selfing and selection is an important aid in obtaining the desired recombinations of multiple factors. Having obtained the Fi of a desirable cross, ordinarily the procedure is to backcross separately to each parent and in each case to select especially for desirable characters of the nonrecurring parent in each subsequent generation. Before the backcrossos have departed too far from the respective nonrecurring parent with regard to its desirable characters, they should be selfed for several generations, the selection being continued as before. By crossing the two approximately pure- breedmg lines ultimately obtained, each largely like the parent employed in backcrossing but also containing desirable characters of the other parent, and selfing the Fi and later generations, it may be possible to secure the desired combination of characters in a pure- breeding line. Time may be saved by simultaneously carrying through several parallel series. With a better understanding of the genetics of tobacco, it may ultimately be possible, as previously suggested, to simplify the problem of selection for quality by utilizing such correlations as may be found to exist between visible growth habit and structure and the less tangible properties influencing quality. Since doubling the chromosome numbers is known to produce more vigorous plants with larger leaves in some of the Solanaccae, this appears to furnish a possible means of improvement, especially in yield; and when no new genes are introduced in the doubling process, qu ality migh t well be maintained. Several methods other than hybridization are available for effecting chromosome doubling—including X-ray treatment, centrifuging, and temperature treatment of maturing germ cells. Although only a limited number of spontaneous mutations have been reported in tobacco, it seems quite possible that through systematic search mutations of the less obvious type could be isolated which would represent significant improvement in yield or in quality. It has been recently suggested that artificially induced mutations such as may be obtained by X-raying may furnish new strains of considerable value, especially since in most cases the new form differs in only a single factor and therefore soon becomes constant. Finally, straight selection as a means of improvement should not be overlooked. To a limited extent growers practice selection, but observation shows that in the main considerable variability exists in the many sorts of each variety in cultivation. If the breeder has clearly in mind the most desirable growth habit and size characteristics for a given variety and commercial leaf type, selection toward the ideal form presents no special problems. However, it must be stressed again that all lines developed should be systematically checked for quality after the leaf has been properly cured and processed.

804 The Story of the Development of the Present^Day Varieties of Tobacco

I IKE many other plnnt crops, tobacco is divided into distinct ^classes or groups, each of which has a different use. These divisions have come about during three centuries of cultivation of the plant and they are as basic in breeding work as they are in production, commerce, and manufacture. In the discussion so far, it seemed best to concentrate on present projects and future possibilities in breeding; but back of this lies much previous work, including the selection practiced by generations of growers and the more recent breeding for improvement by the experiment stations and the Depart- ment of Agricultm^e. In order to make this part of the story clear, the present status and the early historical background of tobacco culture in the United States will be briefly sketched.

TOBACCO A WORLD-WIDE CROP

Growing tobacco is an intensive form of farming, and in the centers of heaviest production it is often the principal cash crop. The average area in tobacco per farm is only about 6 acres. Tlie crop has a relatively high value per acre, but the cost of production also is high, chieliy because of the large amount of labor required. On the average farm only certain areas are capable of producing tobacco of the highest quality, but in most centers of production the acreage of soil well suited to tobacco is usually large enough so that crop rotation can be practiced. This is a matter of importance in the control of diseases and insect pests and in obtaining high-quality leaf. In some sections, however, tobacco has been grown successfully on the same land year after year over a long period, without rotation. Few staple crops ordinarily sliow such a wide range in gross returns per acre, depending on the quality or grade of the product. Although the hazards of commercial tobacco culture are numerous and often serious, under favorable conditions the crop is an attractive one for the average farmer in the recognized producing areas because of the possibility that he may get high returns when he has a high-quality product. For the past 10 years the tobacco crop in the United States has averaged about 1,350,000,000 pounds, grown on 1,740,000 acres and haAaiig an estimated farm value of $214,000,000. This country is the leading world exporter and exports of leaf tobacco have averaged 500,000,000 pounds a year. Imports of foreign types have amounted to 70,000,000 pounds a year, mostly cigar wrapper from Sumatra and Java, cigar filler and wrapper from Cuba, eigar filler from the Philippines, and cigarette leaf from Greece and Turkey. Important quantities of Puerto Rican filler leaf are used in the manufacture of cigars. The principal tobacco-producing States and the approximate average percentage of the total crop produced by each for the past 5 years are: North Carolina, 36; Kentucky, 26; Tejmessce, 8.7; Vir-

805 ginia, 6.7; South Carolina, 5.3; Georgia, 3.6; Pennsylvania, 2.7; Ohio, 2.5; Wisconsin, 2; Maryland, 1.9; Connecticut, 1.6; Indiana, 0.8; Massachusetts, 0.5; Missouri, 0.5; Florida, 0.5; West Virginia, 0.25. North Carolina and Kentucky produce more than 60 percent of the crop, and these States, together with Tennessee, Virginia, South Carolina, and Georgia, account for than more 85 percent. The crop is grown all over the world. Exclusive of China, world production is estimated at approximate^ 4.5 billion pounds, and the output of China probably would raise the total to more than 5 billion pounds. India is outstanding in point of total output, apparently producing about the same amount as the United States. The combined production of tlie two countries accounts for somewhat more than half of the world crop. Although accurate data are lackmg.

SINCE our standard types of tobacco are highly susceptible to most of the diseases^ there is a pressing need for introducing resistance^ without sacrificing quality^ for practically every type and every producing area. General estimates of the losses from tobacco diseases are not available^ but two specific examples may be cited. In 7

China doubtless ranks third in importance, accoimting for 10 percent or more of the total world production. The remaining coimtries producing 100,000,000 pounds or more, in the order of their rank, are the Union of Soviet Socialist Republics, Brazil, Japan, Italy, Greece, and the Philippine Islands. The Netherland East Indies ranks second to the United States as an exporting country, followed in order by Greece, Turkey, Bulgaria, the Philippine Islands, and Cuba. EARLY HISTORY OF TOBACCO CULTURE Tobacco is one of the products given to the world by the natives of the Americas. When the Spaniards landed in Mexico in 1519, they found the people of Yucatan cultivating the strange plant with much care and skill, and using it both for smoking and as snuif (fig. 8). The conquerors were themselves conquered by the gentle weed. 806 Using seed from Yucatan, they began culture of the crop in Santo Domingo about 1531, and it was soon extended to Trinidad. Com- mercial production began in Cuba considerably later, about 1580, and was highly successful from the outset. In Venezuela the Span- iards also established commercial culture at about the same time, at Varinas near Cumana, and the product soon became famous for its excellence. In Brazil, culture by the Europeans apparently began in Bahia about 1600. There are apparently no worth while very early descriptions of the types of tobacco grown by the Indians and the European colonists of the West Indies, Mexico, and South America. However, there is a partial description, dated as early as 1667, of four distinct forms of Nicotiana tabacum in cultivation in the West Indies at that time.

FIGURE S.—TWO species of tobacco cnltivaied exietisivcly by the aborigines when tlie first colonists carae to North America: Nicotiana rustica (A) and N. tabacum (B), sliowiiisr buds, flowers, and green seed pods (drawn by PI. A. Allan!). These are given as (1) Green Pctun or Big Petun, so-called because of the brilliant green color of its leaves; (2) Tongue, described as resembling in shape the tongue of the ox; (3) Amazon, so-called because it was brought from the region of the Amazon River; (4) Varinas, deriving its name from the settlement in Venezuela from which the seed was obtained. Green Petun is described as the most beautiful, with leaves 2 feet long and a foot wide, fleshy, downy, sappy, of a fine green color but shrinking markedly when dried so that large yields were never obtained. The Tongue variety produced leaves even somewhat 807 longer but oiil}^ 7 to 8 iriclics in width, fleshy, oily, strong, less sappy than those of the Green Petun. There was less shrinkage in drying and good yields were obtained. The Amazon produced leaves as long as those of the preceding varieties but broader and more rounded at the end. The principal veins of the leaf tended to form right angles with the midrib, while in the other varieties they followed the contour of the leaf and sloped toward the tip. The leaves were very thick and fleshy, and although apparently they were very sappy, there was relatively little shrinkage in drying and the yield was good. The freshly cured leaf possessed a strong, disagreeable odor and when used caused dizziness and nausea. However, after fermentation the qualit}^ was excellent. The Varinas produced small leaves rarely 6 inches in length, narrow, very pointed, sappy, and shrinking so markedly in drying that it was a poor yielder. The leaf w as mild and highly aromatic so that in mixtures no more than a fourth of it wôuld impart its aroma to the whole. For this reason the Varinas was often adulterated with the less expensive sorts. It will be seen that of the four varieties under cultivation, two were imported from South America and probably one or both of the remaining two were introduced from Mexico. In the northeastern part of the continent, tobacco culture was begun at Jamestown by John Rolfe in 1612. By 1616 culture of the crop had become general in the colony, and in 1619, 20,000 pounds of leaf were shipped to England. Although precise information on the question is lacking, it seems clear that at the outset or shortly there- after, Nicotiaim rustica, the i^lant grown by the natives in this region, was supplanted by N. tabacurn, seed of which probably wâs obtained from South America or the West Indies. Moreover, as early as 1632 it wâs provided by law that only the 'long sortes" be grown, showing that different varieties of the species were under cultivation. There were, in fact, not only two distinct sorts in general use, known as Aronoka (Orinoco) and Swêet Scented, but numerous sub varieties or strains of each of these were recognized and given the names of prominent growers associated with them. It was in this manner that the sort known as Pryor received its name at that time. The Orinoco wâs described as having a sharp pointed leaf and yielding a stronger product, while the Sweet Scented had a rounder leaf, with fine veins, and was mild. Whether the Orinoco and Sweet Scented were closely related to, or identical with, any of the four sorts em- })loyed in the West Indies is not clear. The Orinoco, which ¡pre- sumably was obtained directly or indirectly from the region of the Orinoco River, still survives, at least in name, but the Sweet Scented as such has long since disappeared. Tobacco culture was extended into southern Maryland about 1635, and probably the Virginia sorts were grown at the outset. It is apparent that numerous sorts of Nicotiana tabacurn were available to the aborigines and the early settlers. Unfortunately the eüñy descriptions of the sorts most used commercially in the West Indies and in Virginia are too meager to permit of close comparison with the many forms now employed. Since divers sorts were grown in the same locality, occasional hybridization must have occiu^red, resulting in modification of the original types, and thus there can be no assurance that present-day varieties closely conform to any of the original forms. As will be developed more fully later, most of the standard varieties and their numerous strains now grown in the United States appar- 808 ently have been derived from two original stem varieties, namely, Vir- ginia or Orinoco, and Maryland Broadleaf. In the botanical literature on N, tabacum, Maryland and Virginia are recognized as distinct varieties, and in some instances they have even been classed as distinct species. The terms macrophylla (largo leaf), latissima (very broad), and latifolia (broad leaf) have been used as s,ynonyms for Maryland, although, the descriptions do not closely apply in all respects to the Alary land types now in cultivation.

The Rapid Spread of Cultivation We are accustomed to thinking of communication in the sixteenth century as being extremely slow; but within 25 years after the Span- iards began growing tobacco in Santo Domingo, it was being grown in Europe, and by the time John Rolfe made his first plantings, tobacco culture had apparently spread throughout Europe and the Far East. In Europe the plant was first growTi as an ornamental and for medicinal fjurposes, but it soon came to be used as it is now. On tlie Continent it was consumed in the beginning chiefly in the form of snuif, but in England pipe smoking at once became popular. Tobacco was first grow^n or became know^n in France in 1556, in Portugal in 1558, in Spain in 1559, and in England in 1565. Various names were at first applied to the plant, including AHcotmna, after Jean Nicot, French, ambassador at Lisbon.' Nicot was one of the first to grow tobacco in Portugal, and was largely responsible for making the plant known at the royal court in Paris; and it was the Duo de Guise who proposed the name Nicotiana in 1585. (Actually, tobacco was first grown in Portugal in 1558 by De Goes, a friend of Nicot's, in the Royal Gardens at Lisbon. De Goes gave seed to Nicot, who grew it in 1559.) However, the name tobacco had become established in North America and was able to survive all others for common usage. Early writers differ as to the origin of this name, but it seems to have been applied originally not to tobacco itself but to a type of tube used by the aborigines for inhaling the smoke from the tobacco, or to a cylinder of tobacco prepared for smoking. The tobacco grown in France and Spain was Nicotiana tabacum, froin seed originating in Brazil and Mexico. The species first grown in Portugal and England was iV. rustica y the seed in Portugal coming from Florida, while that used in England came from Virginia. By 1600, or thereabout, tobacco had been introduced into Italy, German}^, Norway, Sweden, Russia, Persia, India, Indochina, Japan, possibly China, and on the west coast of Africa. With few if any exceptions culture of the crop was quickly established in these countries, although it has not survived in all cases. Wlnle both Nicotiana tabacum and A^. rustica were introduced into various foreign countries at the outset, little is Icnown as to the forms of these species originally employed in culture and their relationship to the varieties now in use in these countries. From the small beginning at Jamestown, the production of tobacco in Virginia and Maryland increased rapidly, for it was about the only commodity the colonists could produce to excliangc for the many manufactured products they required from Europe. From the crop of 20,000 pounds in 1618 at Jamestown, exports in 1627, only 9 years later, had increased to 500,000 pounds. ^ In fact, although the foreign market rapidly expanded, production increased at an even greater

38143°—36 :S2 809 rate. The total exports for Maryland and Virginia were 1,500,000 pounds in 1639, but the vahie per pound had declined from nearly 55 cents in 1619 to about 6 cents. At the outbreak of the Revolu- tionary War, exports of tobacco had increased to about 100,000,000 pounds, nearly all of which was produced in Virginia and Maryland. After the close of the Revolution, culture was extended into North Carolina, Kentucky, Tennessee, Ohio, and Missouri, and later into several other States. Domestic manufacture of tobacco first assumed importance after the Revolution and has continued progressively to absorb an increasing portion of the crop, until at present more than half of the total production is utilized for this purpose. THE BACKGROUND OF TOBACCO TYPES Development of the numerous standard varieties now^ in use is very closely linked with expansion of tobacco culture into new territory and its evolution over a long period into a highly specialized industry, based on the production of numerous standardized commercial types of leaf. Today, each of the types produced possesses distinctive characteristics and is designed to meet exacting and relatively fixed trade requirements for specific manufacturing ^ purposes. For example, certain types are used almost exclusively in the manufacture of cigars while other widely different types are especially suited for cigarettes. These ciga*r and cigarette types are essentially non- competitive since they arc not interchangeable for manufacturing purposes. As tobacco growing w as pushed farther and farther into new^ territory, marked differences in soil type and in climate wêre encountered and it was soon learned that these influences gre-atly changed the properties of the tobacco. With increasing specialization in tobacco manufacturing, based at first mainly on production of snuff and pipe- smoking products and subsequently on manufacture of specialized forms of chewing products, cigars, and, finally, cigarettes, it was discovered that the product of some particular area or areas was especially suited to the manufacture, for example, of a particular sort of chewing tobacco, while other areas yielded a desirable cigar wrapper or perhaps a good cigar filler, and so on. As a consequence, production of each principal type of leaf became more or less definitely localized. It was established also that the properties of the leaf may be greatly modified by changes in methods of growing and curing the crop. Thus cultural practices in each, producing area are modified so as to accentuate the natural influences of environment in developing the properties of the leaf desired for the particular type being grown there. Within each principal commercial leaf type there is a wide range of grade or quality, and here again local differences in soil and wêather, and in cultural practices, play an important part. For one type the leaf may be valued for its extreme length and narrow^ness, while in another type a small broad leaf is desired; for one type the leaf should have a clear lemon or orange color, in another it should be almost black; in one case it should be very thin and elastic, in another thin but not elastic, in still another very thick and clastic; in some instances pronounced aroma is wanted, in others only weak aroma; for some types a very strong leaf is desired, but in others a very mild product is preferred. As illustrative of the efl'ects of cultural practices on type or grade, it may be mentioned that, independently of other factors, 810 low topping of tlie plants in the field would be ruinous in the production of cigar wrapper leaf, but high topping wôuld be equally unsatisfactory in producing dark fire-cured tobacco. The principal commercial types or groups of types of leaf tobacco, the areas in which they are chiefly grown, and the uses made of them are indicated in table 2.

TAKLK 2.—The principal commercial types of tohaccOy areas in which they are grown, their chief uses y and the varieties used in their production

Ty[)e of loaf Area in whicli mainly grown Chief use Variel.y of seed

Fired-curcd (U. S. typos Central Virginia, wesitern For export, manufac- Orinoco. 21-24). Kentucky, and north- ture of snuiï and western Tennessee. plug wrappers. Dark air-cured (l.r. S. Immediately east of fire- For chewing i.)lug and Orinoco, One Sucker. types 35, 3(5, 37). cured area of Kentucky export. and Tennessee, and in north-central Virginia. Maryland (U. S. typo32). Southern J\îaryland For cigarettes and ex- Maryland liroadleaf. port. Cigar wrapper, shade- C!onnecticüt Valley and Ci gar wrapper Cuban, Florida 301, grown (U. S. types 61, Quincy, Fla. Ko und Tip. 02). C^iiïar binder (U". S. tvpes Connecticut Valley, south- Cigar binder Connecticut Broadleaf, 51,52,54,55). ern and southwestern Wis- Havana Seed. consin. Cisar filler (T. S. typos Lancaster, Pa., and south- Cigar filler Pennsylvania Seedleaf, 41-44). western Ohio. Ohio Seedleaf, Zim- mer Spanish, Little J>utch. Fliie-curod (C S. types Southern Virginia, northern Cigarettes, pipe and Orinoco. Il-H). and eastern North Caro- chewing tobaccos, lina, eastern South Caro- and for export. lina, southern Georgia, nortiiern Florida. Barley (U.S. typo 31)-_- Central and northern Ken- Cigarettes, pipe and While Piirley. tucky, southeastern Indi- chewing tobaccos. ana, southern Oliio, western West Virginia, eastern and central Tennessee, western North Carolina, western Virginia.

THE PRESENT VARIETIES The immediate origin and the development of the standard A^arieties now used in culture of the hurley and most of the cigar types of tobacco are fairly well known, but only meager information is available on the history of varieties employed in growing the much older dark air- cured, dark fire-cured, and the Maryland or the comparatively young flue-cured types. Undoubtedly the original varieties have undergone considerable modification as a result of accidental crossing and selection on the part of growers over a long period since the beginning of culture in Virginia and Maryland and subsequently in other States. Distribution of present tobacco varieties, including new productions by breeders, coincides essentially with distribution of the standard commercial leaf types, for with very few exceptions these varieties are not interchangeable in i\\Çi different types.

Dark Fire-Cured and Dark Air-Cured Varieties In the dark fire-cured and air-cured types, the cured leaf is dark in color, thick, quite heavy, oily, tough, and high in nicotine content. The fire-cured product possesses a characteristic creosotelike odor derived from the fuel used in curing. Fire-cured types of tobacco have always been chiefly used for export, and at present about two-

811 thirds of the output is exported. The remainder is mainly employed in production of snuff and chewing plug. On the other hand, about 75 percent of the dark air-cured types is used in domestic manufacture, especially for plug chewing tobacco. Of the present-day commercial types, the dark fire-cured and dark air-cured of central Virginia, western Kentucky, and northwestern Tennessee most nearly represent the original type grown at Jamestown. Fire-cured tobacco in the process of curing is subjected to the smoke from slow wood fires on the floor of the curing barn, while air-cured leaf is allowed to cure or dry in ventilated barns without the use of artificial heat except ^ perhaps under special conditions. The rich river-bottom lands of the tidewater section where this type was originally produced were of a light, sandy nature, but as culture was extended into the Piedmont and eventually into Kentucky and Tennessee, it was found that the much heavier loam and clay soils produced a better grade of leaf. Con- sequently tobacco culture eventually was abandoned in the tidewater section. In Virginia various strains of the Orinoco (fig. 9) differing somewhat in size and shape of leaf and other characteristics are chiefly used in growing the dark fire-cured and dark air- cured (so-called sun-cured) types. The present Orinoco apparently represents a survival, doubtless in modified form, of the Orinoco variety FIGURE 9.—Plant of the Orinoco variety, topped in accord- originally grown at James- ance with usual cultural practice tor the flue-cured type of leaf. The Orinoco in its numerous forms is the most town. Among the strains widely cultivated of all domestic varieties, being used generally for production of flue-cured and most dark air- now chiefly grown are Lizard cured and flre-cured types of leaf. Tail, Big Orinoco, Narrow Leaf Orinoco, and Ken- tucky Little Yellow. A typical plant of the Orinoco is shown in figure 9. In Kentucky and Tennessee the principal dark varieties belong mostly to the Pryor group. According to early authorities, the Pryor, as originally grown in the tidewater section of Virginia, was merely a form of Orinoco. Since tobacco culture was carried into Kentucky and Tennessee from Virginia, it seems probable that the present Pryor is a survival of the original Pryor of the latter State, and therefore is closely related to, if not essentially identical with, Orinoco. The latter also is grown to some extent in the dark-tobacco districts of Kentucky and Tennessee. Among the principal strains of Pryor may be mentioned Blue Pryor, Yellow Pryor, Madole, and Yellow Mammoth.

812 A fairly distinctive variety used in the culture of dark air-cured leaf in Kentucky and Tennessee is known as One vSucker, and the type of leaf produced also is known by the same name. The One Sucker produces a very long, narrow, heavy leaf, having a large midrib with which the principal veins form a very acute angle. This variety derives its name from the fact that it is supposed to sucker somewhat less freely than most other varieties. The origin of the One Sucker is uncertain, but the leaf characteristics suggest that it may be a survival of the Tongue, one of the four varieties described in the West Indies in 1667. Up to the present, comparatively little has been done in the improvement of the dark fire-cured and air-cured varieties by breeding. In the Virginia sun-cured district a selection from Narrow Leaf Orinoco, made by the State Agri- A cultural Experiment Station, whichis believed to be resistant to the blackfire disease, has come into extensive use by growers. In tlie sun- cured and in the fire- gÊÛÊI^^Ê^^^^^ÊFMtt' ' îinf cured districts of the J^ÉÍ^^^^Î^^^^^^BHP^^^^ State the Lizard Tail and similar strains of Orinoco have been reported as being among the best available sorts. Maryland Varieties The Maryland is the second oldest type grown in the United States. As now produced in southern Maryland, this tyi>e has distinctive properties which sharply distin- gmsh it from other types FIGURE 10.—Maryland Broadleaf in the early stage of üoweriug. with the exception of This is the parent variety from which the northern Broadleaf or Seedleaf cigar tobacco varieties were derived. White Burley also Burley. The latter and probably originated as a mutation in this variety. the Maryland type show considerable similarity in properties. When cured, the Maryland leaf is of an extraordinarily light, dry, chaffy character, low in nicotine, light reddish brown in color, and relatively thin, and it has a rather weak aroma and possesses excellent burning qualities. The leaf should be relatively large and broad. Formerly Maryland tobacco was chiefly an export type, but at present the better grades are used mainly in the manufacture of blended cigarettes. There is some doubt as to the original source of the Maryland Broadleaf of today (fi.g. 10). It appears to resemble the Sweet Scented of early colonial days in Virginia, as far as may be judged from fragmentary early descriptions of the latter, although as now grown it is not outstandinglv "sweet scented." Since tobacco was

813 grown immediately south of the Potomac in those days, it is li]vel.y that both the Orinoco and the Sweet Scented were introduced into southern MaryUind and that tlie latter was found to give better results in the new area. One early writer states that the Maryland variety came from Mexico, but even so it may have come by w^ay of Virginia. Until quite recently narrow leaf forms also were grown in southern Maryland, but these did not resemble the Orinoco of Virginia. At present, strains designated as Medium Broadleaf, which, produce a somewhat narrower leaf than the ordinary broadleaf strains, are popular among growers. A peculiar form of the Maryland discovered some years ago is of interest because of its exceptional vegetative vigor and the very- large number of leaves it produces imder Maryland conditions. This variety, which appeared as a mutation, is known as Maryland Mam- moth. Normally it does not flower during the growing season. Since flowering and fruiting represent the final stages in the maturing of the tobacco plant, and the development of terminal flowers prevents further upward growth, Maryland Mammoth may be thought of as continuing growth indefinitel}^ because it never reaches maturity from the biological standpoint. It was largely through a systematic study of this peculiar behavior that the phenomenon of photo- periodism or the eft'ect of length of day on the growth and maturing of plants w^as discovered. Under suitable conditions this variety is capable of producing large yields of high-quality leaf, and from a cross with the Standup form of White Burley a ''mammoth'' or non- flowering form of considerable promise for Maryland conditions has been obtained recently. This form shows the stand-up or erect habit of grow^th and produces large yields of high-quality leaf. From a cross of ordinary Maryland with a black root rot-resistant strain of Burley developed at the Wisconsin Agricultural Experiment Station, resistant forms apparently suitable for Maryland conditions also have been obtained. Cigar-Leaf Varieties It was about a hundred years ago that cigar leaf began to be recognized as a distinct commercial type. Actually, cigar tobacco as now grown is divided into three principal types or type groups—cigar wrapper, cigar binder, and cigar filler. Wrapper leaf is grown mainly under artificial shade in the Connecticut Valley and in the vicinity of Quincy, Fla. Binder leaf, which in cigar manufacture is used to hold the shape of the filler or core of the cigar, is produced chiefly in the Connecticut Valley without the use of shade, and. in southern and southwestern Wisconsin. Lancaster County, Pa., and the Miami Valley section of southwestern Ohio are the principal centers of filler-leaf production. Formerly cigar leaf also was produced in important quantities in several other areas. The cigar was a favorite form of smoking among the European settlers in the West Indies almost from the outset, and during the latter portion of the eighteenth century cigars were imported into this country in important quantities, mainly from Cuba. Domestic manufacture of cigars also became of some importance at this time in Philadelphia and soon afterward in Connecticut, both domestic and imported ''Spanish'' leaf, mostly from Cuba and Brazil, being used. At that time tobacco culture in the present cigar-leaf-producing areas 814 was of little importance. About 1825 it began to be recognized that the Connecticut Valley was well adapted to the production of cigar leaf. At first a narrow-leaf variety known as Shoestring was grown, but about 1833 the Maryland Broadleaf was introduced and under the local soil and climatic conditions was found to produce a very large, broad, thin leaf with fine veins, very elastic, having an agree- able taste, a pleasant but not pronounced aroma, and good burning qualities. This introduction had an important influence in further- ing development of the local industry, and Connecticut Broadleaf or Seedleaf, of which there are numerous strams, is still P .. _ a leading variety in local production of binder leaf. Culture of cigar leaf did not become important in Pennsylvania and Ohio till the middle of the last century, an Wisconsin became and important producing State about 20 years later. While other varieties of unknown origin also were used at the outset, Broadleaf or Seedleaf introduced from the Connecticut Valley soon came to be standard for these areas, although more recently the Seedleaf has been largely abandoned in Wis- consin. The various strains of Seedleaf now in use in Pennsylvania and Ohio, however, have come to diner somewhat from the Connec- ticut Seedleaf, and all differ more or less from the parent Maryland Broadleaf. An- other distinctive variety in-

troduced into the Miami Val- FIGURE U.—A form of the Cuban variety. This variety is ley at an early date and still employed in the culture of cigar wrapper under artificial shade in tlie Connecticut Valley and in the production of some importance locally of cigar filler and wrapper in Cuba. is known as Little Dutch. Nothing is known as to the origin of this variety, except that it is said to have been introduced into Ohio from Germany. It possesses very narrow leaves of erect habit of growth, has a peculiar odor in the field, is early in maturing and highly resistant to black root rot, and has a characteristic aroma when smoked. About 1870 an important new introduction was made in all of the four principal cigar-tobacco growing areas. In the Connecticut Valley tliis introduction came to be known as Havana Seed or Con- necticut Havana to distinguish it from the directly imported Cuban. In the Ohio cigar-tobacco district the variety received the name of Zimmer Spanish, the last term of the name indicating its West Indian

815 origin and the first being that of a pioneer grower of tlie variety. Similarly in Wisconsin it was given the name Comstock Spanish. It appears that tliis variety was developed in the Connecticut Valley as a selection from imported Cuban, w^hich frequently has been found to be composed of many distinct strains. Finding it necessaiy to practice seed selection in order to obtain the new variety in pure form, growei^ received the impression that the freshly imported variety underwent an important progressive transformation during the first few years of culture in the new environment, a notion that still is rather widely prevalent. Actually, however, the diversity of forms had a genetic basis and was not due to enviromnent. Havana Seed at once became a very important variety and has remained so. It is a competitor of Scedlcaf in the production of cigar binder in the Connecticut Valley and of cigar filler in Ohio, and it luis practical^ supplanted Seedleaf in the production of binder in Wisconsin. Havana Seed has a larger, longer leaf and more upright growtli habits than ordinary Cuban. It produces a smaller, somewhat narrower, and a distinctly more erect-growing leaf than tlie Seedleaf and yields considerably less per plant. The most recent addition to our commercial types of leaf tobacco is the shade-grown ci^ar-wrapper leaf which first became firmly established about 1906 in the Connecticut Valley and in the Quincy, Fla., district, as a result of extensive experiments conducted by the United States Department of Agriculture and the State experiment stations. Previously the Broadleaf and Havana Seed, grown without shade, had been largely utilized for wrapper. In the early yeai^ of the shade industry the Sumatra variety was tried but proved unsatisfactory, and in the Connecticut Valley it was replaced by selected strains of imported Cuban (iig. 11). The selection and introduction of suitable strains of Cuban through the joint efforts of the Bureau of Plant Industry and the Connecticut x\gricultural Experiment Sta- tion appears to have had an important efl'ect in placing the shade industry on a sound basis. Numerous strains differing somewhat in leaf characteristics are commonly found in the Cuban, but as now grown under shade in the Connecticut Valley the variety is fairly well standardized. The Cuban has smaller, less pointed leaves and longer internodes, and it is taller than Broadleaf or Havana Seed. In the Quincy district the Sumatra was replaced by a variety known as Big Cuban wiiich is said to have originated from a cross between Sumatra and the ordinary Cuban, although the details of its development are not known. Up to the present the cigar-tobacco varieties haA^e received more attention from breeders than the other commercial types of leaf. Beginning about 1900, experiments in the production of improved varieties through hybridization and selection were undertaken in several of the cigar-tobacco areas. However, the difficulties involved in securing the desired recombinations of visible leaf characteristics by these methods without upsetting the various elements of quality in the cured product were not fully understood. While in several instances new forms were obtained which showed much promise in the field, notably the Halladay Havana, from a cross of Havana Seed on Sumatra, nearly all proved to be unsatisfactory to the manufacturer for one reason or another and had to be abandoned. Some years later, however, a new variety known as Round Tip was produced at the Connecticut Agricultural Experiment Station (East

816 and Jones, 1921), in accordance with a clearly defined objective and plan of procedure, which, achieved considerable success and has served to indicate that there are possibilities in the creation of improved varieties when all necessary precautions are observed. Nevertheless, the subsequent history of this variety in culture further emphasizes the difficulties and uncertainties of the problem of developing new varieties that will be acceptable to the trade with respect to all elements of quality. Round Tip was produced from a cross of Connecticut Broadleaf on Sumatra. It combines, at least in some degree, the largo size of leaf, the short internodes, and the high yield of tlie Broadleaf with the round tip, erect growth habit, large number of leaves per plant, and fight color of leaf of the Sumatra, and it is resistant to black root rot. The cured leaf appeared to be quite satisfactory with respect to flavor, elasticity, and combustibifity. In Connecticut this variety gave very^ promising results when grown as a substitute for Havana Seed in the production of so-called ^'primed Havana" Vv-rapper leaf. Subsequently, however, culture of the primed Havana type was abandoned and there appeared to be no further place for Round Tip. Curiously enough, when tried under shade in the Florida area, Round Tip proved to be superior to the Big Cuban and for a time it practicaUy replaced the latter. Unfortunately, the black shank {Fhyto])hthora) disease soon became epidemic and Round Tip proved to be highly susceptible, so that again its culture was greatly reduced. The most recent productions in the cigar-wrapper varieties are black shank-resistant substitutes for Round Tip in the Quincy area. The Florida Agricultural Experiment Station first developed highly resistant strains of Big Cuban and ordinary or Little Cuban through systematic selection. The former, like the parent variety, is rather deficient in quality, while the latter gives low yields. From a cross of these resistant strains a variety designated as 301 was obtained through selection for six generations. This combines higli resistance to black shank with reasonably good quality of leaf, and it is now widely grown. As regards the cigar-binder types, the Wisconsin Agricultural Experiment Station in 1915 released an improved strain of Havana seed which came to be widely grown in the State because of its erect growth habit, uniformity, broad leaves from top to bottom, increased number of leaves per plant, and the high quality of the cured leaf. This strain was obtained by selection in a cross of two fairly distinct strains which had been isolated from Havana Seed introduced from Connecticut. However, the new strain, known as Havana No. 38, is susceptible to root rot and for this reason has not been so extensively grown in recent years. To overcome this difficulty, Havana No. 38 was crossed with a selection from a strain of Havana Seed locally known as Page Com- stock which was resistant but did not possess desirable habits of growth. From this cross^ was developed Havana No. 142, which proved to have greater resistance to root rot than the more resistant parent. It also has desirable growth habits and is a good yielder, although it is somewhat later in maturing than ordinary Havana Seed. AVhile there has been some difference of opinion among man- ufacturers as to the quality of leaf produced by this strain, it has been widel}^ grown in the State and there is no question as to its

817 superiority over susceptible strains of Havana Seed when planted on root rot-infested soil. From the same cross another resistant strain known as Havana No. 211 was later developed. This strain appears to have certain advantages over Havana No. 142 with respect to quality and has given promising results both in Wisconsin and in the Connecticut Valley. In the Pennsylvania cigar-filler area two new sorts of Pennsylvania Seedleaf have been introduced in recent years by the State agricultural experiment station and the Bureau of Plant Industry cooperat- ing, and these sorts have been extensively grown. The Schwarr- Hibshman was developed by selection in a cross of the local Schwarr and Hibshman strains, and the Olson Broadleaf was obtained from a cross of local Seedleaf and Pennsylvania Havana seed. These new sorts give good ^delds, are early maturing and relatively resistant to root rot, and produce leaf of standard quality.

Flue-Cured Varieties Flue-cured tobacco, often spoken of as bright tobacco, owes its name primarily to the unique method emi^loyed in cnring the leaf. Artificial heat is supplied through a system of flues or pipes so that the leaf does not come in contact with smoke from the fuel used, and the curing process is rapid. This has come to be by far the world's most important commercial type of leaf tobacco in point of quantity produced. It is grown mainly in southern Virginia, northern and eastern North Carolina, eastern South Carolina, southern Georgia and northern Florida. Among the outstanding characteristics of the cured leaf are its bright lemon to orange color, its distinctive aroma, and its high content of sugar. Its principal uses are for domestic manufacture of straight and blended cigarettes, pipe- smoking and plug-chewing tobaccos, and for export. The type as now grown is essentially the result of change in soil type from the heavy silt loams of central Virginia to the less fertile light gray soils of the North Carolina border counties and later the Coastal Plain section as culture was extended into these areas, together with subsequent changes in the method of curing and in cultural practices. There wâs no significant change in the varieties of seed used. The lighter soils of the central border counties produced a leaf of much lighter color, sweeter, and finer in texture, wliich quickly found favor in the export market and for the manufacture of chewing tobacco. For leaf of these characteristics it wns found advantageous to use less heat in curing, and about 1825 charcoal as fuel came into use in order to eliminate the eft'ect of smoke on the odor and taste of the leaf. Metal flues for supplying heat in the barn without admitting smoke came into general use after the Civil War. Toward the close of the last century culture of flue-cured tobacco was introduced into eastern North Carolina and South Carolina and there increased very rapidly. Finally, about 15 years ago, culture of this type was further extended into southern Georgia and northern Florida. As already indicated, the Virginia Orinoco in its various forms was used from the outset in the culture of the type now known as flue cured. Apparentl^^^ the present strains of the Orinoco as a whole do not differ greatly from those employed originally. Broadly speak- ing, the larger broad-leaf forms are preferred for the lighter soils of the Coastal Plain and the eastern portion of the Piedmont areas, 818 while the narrow-leaf forms are popular in the western j)ortion of the Piedmont producing area. This is in accord with the fact that, other things being equal, a heavy close-textured soil tends to produce a smaller, narrower but thicker leaf than a light, open soil. Among the more popular older strains are Yellow Orinoco, White Stem Orinoco, Little Orinoco, Gooch, Adcock, Yellow^ Pryor, and Silky Pryor. At present an extremely large number of so-called varieties are used in growdng flue-cured leaf, but these are merely strains, most of them so similar as to be almost indistinguishable from one another. Of those most widety used at present, mention may be made of Cash, Jamaica, Virginia Bright, Bonanza, Yellow Mammoth, and the old White Stem. Comparatively little has been done in the improvement of flue-cured varieties by breeding. In the period 1929-32 the Coker^s Pedigreed Seed Co. in South Carolina released two improved strains of Jamaica and one each of Yellow Mammoth, Bonanza, Virginia Bright, and Cash. All of these strains wêre obtained by straight selection for yield, quality, and disease resistance. ^ They have^ been growm extensively, more particularly the selection of Jamaica know^n as Gold Dollar, and appear to have given generally satisfactory results. At the Oxford, N. C., tobacco station three strains resistant to black root rot have been develo])ed by selection from Jamaica, Paris Wrapper, and an unnamed form supposed to have originated from a cross of Warne on Big Gem. Of the three strains, the last-named shows the highest disease resistance, and it is a good yielder and appears to produce leaf of excellent quality. As yet this strain has been grown only to a limited extent but has given very promising results. Ît is designated as Special No. 400. Pure lines of White Stem, Virginia Bright, Cash, Jamaica, and Bonanza selected for growth habits, yield, and quality were released to growers in the period 1920-27, the two first being especially adapted to the light, sandy soils and the others to heavier soil types. Burley Varieties The commercial type of leaf know^n as Burley represents the only instance in the history of the tobacco industry in w^hich the discovery of a distinctly new variety, by one of those lucky accidents that shape the course of history, has been solely responsible for the estab- lishment of a new, highly important commercial leaf type. More- over, the leaf type bears the same name as the variety. Burley, which is second to the flue-cured type in importance, is grown chiefly in central Kentucky; certain areas of West Virginia, Ohio, and Indiana bordering on the Ohio River; eastern Tennessee and portions of central Tennessee; and the extreme western part of Virginia and North Carolina. At present its principal uses are for the manufacture of blended cigarette and pipe and chewing tobaccos. Tobacco wâs grown commercially in Kentucky and southern Oliio by settlers from Virginia and Maryland for many years before the appearance of the new variety. These settlers of course carried tobacco seed with them from their home States. In northern Ken- tucky and southern Ohio a very light, chafly, reddish or cinnamon- colored leaf, almost free of gum, was produced from a variety or strain know^n as Burley. These characteristics of the leaf, wîich closely resemble those of the southern-Mar^dand product, made it

819 especially suitable for the manufacture of a form of chewing tobacco known as fine cut. The name Burley probably was derived from that of a grower just as in the case of the Pryor in Virginia, previously mentioned. There is little reason to doubt that the original Burley was a form of Maryland Broadleaf. In 1864 George Webb, a farmer in Brown County, Ohio, observed in his tobacco seedbed, which is said to have been planted with Burley seed from Bracken County, Ky., a number of seedlings having a peculiar chlorotic appearance—that is, they were somewhat deficient in green coloring matter. These seedlings were discarded as being diseased, but when they appeared again the following year in a second planting of the original seed, some of them, were grown to maturity. The mature plants, though otherwise normal, showêd an almost pure cream color instead of the normal light green. The leaf cured to a light 3êllowish red color, and the light, chaffy, porous characteristics of the parent Burley were accentuated. Because of the peculiar creamy color of the stalk and the midrib and veins of the leaf, and the very pale greenish yellow color of the leaf web when mature, the new variety was called White Burley to distinguish it from the old Burle}^, w^hich came to be known as Red Burley. The latter has long since lost its identity and the White Burley is now commonly spoken of simply as Burley. The White Burley has never been a leading export type, but it played a very important role in shaping the course of domestic tobacco manufacture. Originally it wâs found to be excellent for the manufacture of fine cut, and consequently its cultui*e increased very rapidly. Beginning about 1875 it came to be recognized that the heavier grades of Burley were especially adapted to the manufacture of strongly sweetened plug because of the remarkable capacity of the leaf to absorb flavoring sauces. This discovery resulted in enormous development in manufacture of the new type of plug and a greatly increased demand for Burley leaf. As in the case of Orinoco and other distinctive varieties, numerous strains of Burley dift'ering in erectness, leaf size, and other characteristics were developed, but during this period broadleaf, drooping forms of Burley, similar in general appearance and growth habits to the Maryland Broadleaf except in color, were in greatest demand for producing a tough filler leaf for plug. AVliite Twist Bud, Lockwood, and Big Silk were popular strains. SubsequentljT^ it developed that Burley leaf, especially that from the lowermost part of the plant w^hich was classed as **trash" and previously had no ready outlet, is well suited for the production of blended pipe-smoking mixtures. Here again it found extensive use, particularly for production of the cut-plug type of pipe tobacco. Finally, as the use of chewing tobacco declined and the consumption of machine-made cigarettes began to increase with phenomenal rapidity, Burley once more demonstrated its general adaptability, for it has come to be a leading type in the manufacture of blended cigarettes. Thus, while the principal use of Burley has repeatedly shifted, in each case because of specific characteristics or properties» consumption has tended to increase over a long period. However, the shift in principal demand from chewing to smoking grades of leaf called for modification in growth characteristics of the variety, in order to secure an increased proportion of smoking grades and a corresponding reduction in chewing grades in the crop. To meet this situation, recourse was had to culture of the so-called

820 stand-up strains with leaves of a more erect habit of growth, somewhat smaller in size but with a greater number per plant, and yielding a higher percentage of light-colored, light-bodied leaf than the broadleaf, drooping forms of Burley. Among the popular strains have been Judy Pride, Halley Special, Pepper, and Kelley. Green coloring matter plaj^s íx vital part in the functioning of the leaves of plants, and perhaps because of its chlorotic characteristic, White Burley is a less rugged and vigorous plant than the typical green sorts. Consequently it does not yield as well when grown on soils of low fertility. Under such adverse conditions the leaf loses some of its desirable properties. Burley also is more severely affected by such diseases as mosaic than are most other varieties. On rich loam soils, however, it grows vigorously and is capable of producing good yields of light yellowish red leaf which is light-bodied, chaffy, and fine-textured, and has excellent burning qualities but weak aroma. These properties of the cured leaf make it especially valuable for blending with otlier types of leaf. Burley is highly susceptible to the black root rot disease, and in recent years considerable attention has been given to the development of resistant strains. In 1919 the Bureau of Plant Industry (U. S. T>, A. Bull. 7Ö5) reported results of experiments demonstrating the superiority of a highly resistant strain, obtained by selection, over the ordinary strains when grown on infested soil. However, the resistant strain was the drooping form of Burley, whereas the stand-up sorts are in much greater demand for production of cigarette leaf. More recently, several resistant strains of the stand-up form have been produced. In the period 1922-24 the Kentucky Agri- cultural Experiment Station released two strains obtained by selection from ordinary forms .and designated as Strain W and Strain 36-12 which were moderately resistant to root rot. Those strains have been largely superseded by White Burlejr No. 5, which was developed from a cross of the W strain on a cigar-tobacco variety and released in 1929. No. 5 is highly resistant, produces a leaf of good quality, and has been widely grown. Foreign Types and Varieties Of the great array of types of leaf grown in foreign countries, only a limited number enter into commerce in important quantities. There are three principal foreign types imported into the United States for use in domestic manufacture—Sumatra and Java cigar wrapper, Cuban cigar filler and wrapper, and so-called T urkish leaf, a cigarette type which is mainly imported directly or indirectly from Turkey and Greece. The Sumatra variety is fairly distinctive, but numerous strains varying somewhat in various characteristics are in use. In the past this variety has been extensively tried in this country but without pronounced success, and it also has been employed in certain cases for hybridization with domestic varieties. Sumatra is relatively tall, with moderately long internodes, and it produces a fairly large number of medium-sized, round-tipped, broad, erect-growing leaves, of somewhat lighter green color than most varieties. It is resistant to black root rot. The Cuban variety has been shown to consist of numerous races which ordinarily breed true to type. These races in some instances differ rather widely^ in number, size, and shape of leaf and other growth characteristics. Selected strains of Cuban are extensively 821 used in the culture of cigar wrapper under shade in the Connecticut Valley (fig. 11). It somewhat resembles Sumatra in general growth chfiracteristics. As grown in Cuba this variety is famed for its excellent aroma, but experience has shown that expression of this characteristic is largely dependent upon the soil and climatic conditions under wdiich the tobacco is grown. In other words, the same tobacco grow^n elsewhere does not give the same results. The so-called Turkish or oriental tobacco of commerce is in reality composed of several types grown in different regions, chiefly in ancient Macedonia and in the general region of Smyrna and along the southern shore of the Black Sea in Asia Minor. The varieties used in producing these types differ decidedly in many respects, and even in the same district there are material differences in the varieties and strains grown. The more important Turkish varieties, though differing among themselves, are characterized as a class by relatively very small leaves which when cured possess a distinctive aroma, are mild, and have good burning qualities. The commercial types of chief importance in this country are those from Macedonia, Smj^'rna, and Samsun. Representative varieties from these regions have been grown for observation in the Bureau of Plant Industry. The Samisun variety is characterized by an ovate (egg-shaped) or almost cordate (heart-shaped) form of leaf, very broad at the base and having a naked petiole or leafstalk, very short internodes, and a large number of leaves per plant. Smyrna produces a relatively large leaf resembling Havana Seed iii shape. A Dubek or Xanthi sort from Mace- donia has an exceptionally broad and well-rounded leaf, distinctly smaller in size than the other two. It w^^s found that w^hen the plants w^ere grown on fertile soil and widely spaced, the Turkish sorts produced surprisingly large leaves The numerous sorts of both Nicotiana iahacum and iV. rustica grown in India have been extensively studied by the Howards (5) and others in that country, and considerable work has been done in their improvement by breeding. These tobaccos, however, are of no special interest in this country. The same is true of the tobaccos grown in Italy, the Union of Soviet Socialist Republics, Germany, Rumania, and other countries in w^hich some research in tobacco breeding has been conducted.

The Botany of the Tobacco Plant and the Studies Made in Genetics

IN ADDITION to the practical breeding wx)rk with tobacco, which develops out of economic needs, there has been a good deal of basic genetic research for which the tobacco plant furnished a convenient subject. This work is important as indicating the genetic basis for breeding programs. Before it is surveyed briefly, it will be well to include enough about the botany of the tobacco plant to make clear its general characteristics and its place in the plant world. BOTANY OF THE TOBACCO PLANT Tobacco belongs to the genus Nicotiana which w^as established by Linnaeus in 1753. Nicotiana is but one of a very large and important

822 family of plants known as the Solanaceae or nightshade family, which includes many of our most useful food and medicinal plants, some of the latter being highly poisonous. Among the members of tlie family are potato, eggplant, tomato, henbane, belladonna, garden peppers, ornamental petunias, salpiglossis, and many others. The greatest development of the entire family, as indicated by the number of genera and sfjecies, has been in the New World, more especially in Central America and South America. In his original classification Linnaeus included the two species of tobacco chiefly cultivated by the North American aborigines and subsequently by the early colonists—Nicotiana tabacura and N, rustica. He was acquainted, however, with very few distinct species, probably less than half a dozen, including the two cominercially useful species mentioned. Since the time of Linnaeus, many others have been discovered and the genus has grown to include a large assemblage of species, varieties, and forms. The genus has been revised but three times up to the present, the oldest systematic revision being that of Lehman in_ 1818. In 1852 Dunal made the second revision of the genus, and in 1899 O. Comes made the third and tlie latest systematic revision. There were about 50 species and as many varieties recognized at the time of Comes. At the present time the genus Nicotiana is greatly in need of careful revision from the standpoint of systematic classification, but this l)ecomes a problem of great difliculty because much of the original hei'barium material is inaccessible and widety scattered in the herbaria of England, Europe, and South America. Although possibly 100 distinct species may comprise the genus Nicotiana, seeds of scarcely more than a third have been obtainable in this country. Nicotiana is primarily a New World genus ; in fact the only Icnown exception is the species A^. suaveolens and related forms occurring in Australia, and perlmps in other islands of the Australasian archipelago. The only species of economic importance are A^. tabacum and A", rustica. The species Nicotiana tabacum has slender tubular flowers with the flower tube greatly exceeding the calyx in length, and usually pink in color in our commercial types, although white and carmine-red forms of the species exist. The lobes at the upper end of the corolla are commonly distinctly separated, and acute or pointed. The leaves, which, are usually sessile, that is, without a stalk, and auricled ^ or partly clasping the stem, Y^TJ greatly in size, the variation depending on the variety and conditions of growth. In some of the Turkish forms as grown commercialh^ the leaves are very small, less than 6 inches long; in some of the larger growing varieties they may reach 2 to 3 feet in length. This species, like A^. rustica,, is covered with a sticky pubescence or hairy growth. The plant usually reaches 4 to G feet in height in the field. The species Nicotiana rustica is characterized b}^ pale yellowish flowers, with a short, thick corolla tube, having rounded lobes. The leaves are thick, broadly ovate (egg-shaped) with a distinct, naked petiole or leafstalk. The entire plant is beset witli a very glutinous pubescence. It usually grows from 2 to 4 feet in height. The flowers of A'^. tabacum and A^. rustica are shown in figure 8. Under favorable conditions, varieties of both of these species may possess a very high content of nicotine.

s Having an ear-shîipcd appendage at the biise of t lie leaf. From Latin aarLs, ear.

823 Several species of tobacco are given a place among the garden ornamentals, including Nicotiana sylvestris and N. alata. Native species of tobacco have not been found east of the Missis- sippi River. West of the Mississippi a number of small native species occur, covering portions of the territory from Texas to southern California and northward to British Columbia.' If the number of species is a criterion, the genus Nicotiana appears to have reached its greatest specialization in South America south of the Equator, and to have spread outward, ultimately to reach Central America, Mexico, and the western United States. Since there are no eastern native species of Nicotiana, it would appear that the drier conditions of the Mexican high- lands and the western area of the United States have been more favorable to the northern distribution and evolution of species than the more humid, originally heavily forested territory east of the Mississippi River and the Appalachian Mountains. Aboriginal man unquestionably has been a factor in the transplantation of at least two species far northward of their original habitat, especially in the eastern United States. This was a cultural introduction based on the use- fulness of the plant for smoldng, chewing, and snuffing. The species Nicotiana rustica, probably native to Mexico, which was widely cultivated by the aborigines, has tended to FIGURE 12.—Prof. E. M. East, an outstanding geneti- persist in Texas and also in the cist whose extensive investigations in the genus Nico- tiana have greatly advanced our knowledge of the East wherever field conditions genetics of tobacco. Professor East also has successfully applied genetic principles to the development have given it some degree of of valuable new strains and varieties, notably the freedom from competition with Connecticut Round Tip. the native woody plants. The species N. tahacum, probably an introduction froin South America, appears to be even more dependent upon open field conditions and is even less able to maintain itself without cultivation. Nicotiana rustica was grown by the American aborigines throughotn the region to the east and immediately west of the Mississippi, from Canada to Mexico. A'^. tahacum was the principal aboriginal tobacco of the West Indies, most of Mexico, Central America, Colombia, Venezuela, the Guianas, and Brazil. Apparently A^. guadrivalvis and its variety multivalvis were cultivated by the Indians of northwestern United States. According to Setchell (1921), A*", attenuata was widely used by the Indians over the Southwest, most of the Plains area, and the North Pacific coast, and A^^. trigonophylla was employed by the Yumas of the Southwest for smoking. « In the western area of the United States, the recognized native species are Nicotiana trigonophijUa Dunal; N. attenuata Torr.; N. repanda Willd; N. clevelandii A. Oray; N. bigeloiii S. Wats.; N. palmeri A. Gray; N. guadrivalvis Pursh. and its variety multivalvis Gray; N, plumbaginifolia Viv. N. glauca Graham, native to Argentina, has become widely established from Texas to southern California and Mexico. 824 PRESENT GENETIC BASIS OF BREEDING PROGRAMS Because of its many natural advantages, the genus Nicofiana was used extensively by hybridizers long before the days of Mendel, and it has continued to receive much attention in recent years by geneticists, notably East at Harvard (fig. 12) and Goodspeecí at the University of California. Inheritance in Species Hybrids The Fi generations of crosses between species furnish favorable material for the study of sterility, since they range from those that are entirely fertile to those that are completely sterile. In an analysis of published data on hybridization experiments, East (1928) found that in 228 combinations tried between 18 species, only 79 of these combinations produced hybrid plants. N. tabacum and N, rustica have been shown to contain 24 pairs of chromosomes, while the range in all known species in the wild is from 9 to 82. Successful crosses have been obtained between tabacum and the following species—grouped according to Goodspeed's classification—each having the number of pairs of chromosomes indicated: Alata, sanderae, and langsdorfii (9); glauca iuid 2>a7í{culaia (12) ; sylvestris {12) \ tomentosa, tomentosvformis (rusbyi) and glutinosa (12); suaveolens (16); rustica (24); bigelovii (24). In some cases viable seed are obtained with considerable difficulty, and in some instances, as in rustica X tabacum, the crosses can be made more readily in one direction than in the other. It appears that usually crosses between species differing in number of chromosomes are more likely to be successful when the species with the higher chromosome number is used as the female parent. Inheritance of certain morphological or form characters has been studied in crosses between species of the alaM group. Reciprocal hybrids of Nicotiana langsdorfii and N, alata were found to be identical and generally intermediate in characteristics, but the blue pollen and green flower color of the former were dominant over the ivory-colored pollen and white flower of the latter. In studies of flower size it was determined that in Fi populations corolla length was as uniform as in the parents when self-fertilized, while the F2 populations were three times as variable. The variabilities of F3 families were always smaller than those of F2 families. It appears that these color and size characteristics, as well as red and purple flower color in A". sanderae, are controlled in the main by a relatively small number of genes, although doubtless there are several less easily detectable genetic factors which modify their expression. Of the alata group of species, Nicotiana alata and N, sanderae are usually self-sterile, while N, langsdorjii is self-fertile. According to East, self-sterility and cross-sterility in these species are dependent upon a limited number of genes functioning in such a way as to cause differential growth of the pollen tube. In general, hybrids between species are intermediate with respect to growth habit and pattern of leaf and flower. These constitute the most obvious differences between the species, and apparently the diflerences can best be explained on the basis of a large number of nondominant Mendelian factors. East (1935) suggests that it is these genes that are of importance in evolution rather than the dominant qualitative factors that are usually studied by geneticists.'' 7 According to this conception, species have evolved through the bnildinji up of genos with different elTects on the pattern of leaf, flower, growth habit, etc., each ôoniplement oí such genes being capable of producing a distinct biotype with the ability to survive natural selfíction. The nualitative factors or genes, which usually aro dominant, are mostly limitations on various physiological processes, involving, for example, such phenomena as dwariism.

:í8143''~,'>>() ").■/. X 825 It lias been suggested by Goodspeed that some, and possibly all, species containing 24 pairs of chromosomes have resulted from hybridization in which the numerical chromosome complement of both parent forms was retained, so that, added together, they gave 24 pairs of chromosomes in the progeny. Evidence has been presented that Nicotiana tahacum originated from a cross between a member of the tomentosa group and N. sylvestris, each having 12 pairs of chromosomes. In the same manner A^. rustica may have resulted from hybrids between members of the glauca group with 12 pairs of chromosomes. Genetic Studies in. Nicotiana Tabacum As has been pointed out previously, varieties of Nicotiana tabacum vary widely in number, size, shape, color, and texture of leaves, length of internodes, and color and size of flower. On the basis of leaf shape and corolla shape, Comes (1905) referred all existing forms to six primary varieties, namely, fruticosa Hook, lancifolia (W.) Comes, virginica (Agdh.) Comes, brasiliensis Comes, havanensis (Lag.) Comes, and macrophylla Schrank. Of these, virginica and brasiliensis were considered as no longer to be found in the pure state. Anas- tasia (1907, 1914) has proposed that the fundamental varieties be reduced to four—havanensis^ brasiliensis, virginica, and purpurea. The Howards {8) attempted to classify the cultivated forms in India according to leaf form and length of internode, and 51 sorts were recognized. However, study of numerous crosses disclosed recombinations which exceeded the limits of the parents and showed that the derivation of a variable cultivated species cannot be traced by a classification based simply on morphology or structure. In the early work in improvement of tobacco varieties beginning about 1900, the impression prevailed that change from one environment to another may cause a breaking up of the type. Careful experiments made later by Hasselbring (6), East and Hayes (1912, 1914), Honing (1915-19), 'David (1925), and others have shown that this is not the case. Although tobacco normally is self-fertilized, cross-pollination is sufficiently common so that seeds produced by self-fertilization may show variations in type because they came from cross-fertilized ancestry. But it is true that expression of inherited potentialities may be modified or even suppressed by environment. It has been suggested by Christofï (1925) that three factors determine flower color in tobacco, and Setchell {10) and his associates (1922) concluded that at least three factors affect leaf-base characters. Lodewijks (1911) investigated an aurca or yellow-color condition observed in tobacco in Sumatra, and Allard (1919) found the aurea character in the stem of Nicotiana rustica to be dependent upon a single Mendelian factor recessive to green stem. Of special interest is the genetic behavior of the chlorotic stem and leaf vein and the yellowish-green color of the leaf lamina which are characteristic of the commercially important White Burley variety. Investigations by Kajanus in Sweden (1924) and Henika at the Wisconsin Agricul- tural Experiment Station (1932) showed that duplicate genes are involved in production of the normal green color in tobacco, and absence of both results in the Burley color. In crosses of the White Burley with normal green varieties, the Fi was green, and segregation in F2 approximated a ratio of 15 green: 1 yellow, the latter segregates breeding true. 826 Extensive studies of hybrids between varieties of Nicotiana iabacum have been made by Hayes (1912), Hayes, East, and Beinhart (7), and others with respect to quantitative characters susceptible of direct measurement in the field, particularly number of leaves per plant, length, breadth, and area of leaf, and height of stem. In general the first generation is intermediate and relatively uniform, while the second generation shows a continuous variety of sizes which cannot be readily separated into classes or ratios. In an experiment with a cross of Connecticut Havana on Sumatra, selection was made in self-fertilized lines for a high number and a low" number of leaves. It was found that any desired average number of leaves from about 12 to 30 could be obtained in combination with various sizes and shapes of leaf. However, in some instances selection for at least 12 years may be necessary to obtain fixed forms (3), These results indicate that inheritance of size is controlled by a large number of quantitative factors which produce cumulative effects. Itesults of an extensive study of crosses between several tobacco vari ties also have been reported by Setchell, Goodspecd, and Clausen (10). In a study of suckering habit, Johnson, at the University of Wisconsin (1919), found this to be a purely quantitative characteristic yielding results similar to those applying to leaf size. Ilefercnce already has been made to variation of nicotine content in tobacco and the development of low-nicotine strains. Hackbarth and V. Sengbusch (1935) suggest that such strains may be obtained in any variety of tobacco if a sufllciently large number of individuals can be examined. They find that a single fundamental factor deter- mines the difference between high-nicotine content and the ''nicotine-free'^ condition. However, since the plants arbitrarily designated as nicotine-free on the basis of a rough method of chemical analysis appear still to contain small quantities of the alkaloid, the genetic situation may not be as simple as indicated.^ Up to the present the tobacco-breeding work pertaining to improvement in quality has had to do chiefly with visible or measurable characteristics, mainly number of leaves produced and their size, shape, and gross structural characteristics. Comparatively little of special moment has been done on inheritance of the various properties or characteristics of the cured leaf which largely determine its commercial value but which unfortunately are not readily measurable, especially in the field. Some of the practical difficulties involved in dealing with these properties have already received consideration. Few mutations have been observed in tobacco, although little is known as to the possible occurrence of mutations involving only slight changes which would not be readily detected by inspection. Appear- ance of the White Burley variety in 1864 constitutes, from a practical standpoint, the outstanding example of mutation, if it is true that this was really the origin of the variety. A frequently recurring mutation is that known as ''gigantism'' or "indeterminate growth,^' referred to in the discussion of Maryland varieties of tobacco. The mammoth typo is peculiar in that it continues to grow without formation of the usual terminal inflorescence when exposed to the long days of summer, but with the short days of winter or in low latitudes it flowers normally. This type of mutation was reported by Hunger in Sumatra tobacco in 1905, and it was subsequently observed in Maryland tobacco, Cuban, Havana Seed, Orinoco, Puerto Kican, and other varieties. In crosses with ordinary varieties of the normal or flowering types and 827 with Nicotiana sijlvestris and N. glutinosa^ the ordinary or flowering- typo usually is dominant, and probably a single gene is involved. The one-fact or ratio of 3:1 is difficidt to obtain, possibly because the character is sensitive to changes in environment. Doubling of chromosome numbers has been shown to produce larger plants in various genera of the family Solanaccae to which tobacco belongs. It occurs in jimsonweed, petunia, black nightshade, and tomato. In tobacco, Clausen and Goodspeed have described a race resulting from the cross glutinosa X tahacum which had the dou- bled chromosome complement and enlarged structural characters. Goodspeed and his associates have published extensive dataon the production of mutations in tobacco by irradiating buds of the terminal inflorescence with X-rays. Seven pure-breeding derivative types and seven other types not breeding true have been obtained from the progeny of a single X-rayed female gamete of N. tahacum (1934). Netherland East Indies investigators also have conducted similar investigations. Existing forms of N, rustica have been referred by Comes (1899) to six stem types on the basis of structural characteristics, and the Howards (Í910) have studied 20 forms in India which were classified on the basis of differences in length of internode and structure of inflorescence. Crosses between the various forms have served primarily to furnish additional evidence of multiple-factor inheritance. REFERENCES ON HISTORY, BREEDING, AND GENETICS (1) Du TERTRE, R. P. 1667. DU PETUX. In Histoire General Des AntiUes, t. 2, pp. 99-101. Paris. (2) EAST, E. M. 1928. THE GENETICS OF TILE GExus NICOTIANA, Bibliographia Genética 4.: [24BJ-320, illus. (3) and JONES, D. F. 1921. ROUND TIP TOBACCO, '^\ PLANT MADE TO ORDER." . , . Jour. Hercditv 12: 50-56, illiis. (4) GARNER, W. W. 1909. SOME OBSERVATIONS ON TOBACCO BREEDING. Amer. Brccders' Assoc. Kept. 5: 299-303, illus. (5) GOODSPEED, T. H., and AVERY, P. 1934. THE CYTOGENETICS OF FOURTEEN TYPES DERIVED PROM A SINGLE X-RAYED SEX CELL OF NICOTIANA TAB AC UM. Jour. GeilcticS 29: 327-353, illus. (6) HASSELBRING, H. 1912. TYPES OP CUBAN TOBACCO. Bot. Gaz. 53: 113-126, illus. (7) HAYES, H. K., EAST, E. M., and BEINHART, E. G. 1913. TOBACCO BREEDING IN CONNECTICUT. Coim. Agr. Expt. Sta. Bull. 176, 68 pp., illus. (8) HOWARD, A., and HOWARD, G. L. C. 1910. STUDIES IN INDIAN TOBACCOS. NO. 2. THE TYPES OF NICOTIANA TABACUM L. Dept. Agr. India, Mem., Bot. Ser. 3:59-176, illus. (9) JOHNSON, J. 1930. BREEDING TOBACCO FOR RESISTANCE TO THIELAVIA ROOT-ROT. U. S. Dept. Agr. Tech. Bull. 175, 20 pp., illus. (10) SETCIIELL, W. A., GOODSPEED, T. H., and CLAUSEN, R. E. 1922. INHERITANCE IN NICOTIANA TABACUM. I. A 3ÎEP0RT ON THE RESULTS OF CROSSING CERTAIN VARIETIES. Calif. UniV. Pubs., Bot. 5: 457-582, illus. (il) TiSDALE, W. B. 1931. DEVELOPMENT OF STRAINS OF CIGAR WRAPPER TOHACCO RESISTANT TO BLACK-SHANK (PHYTOFHTHORA NICOTIANAE BREDA DE HAAN). ria. Agr. Expt. Sta. Bull. 226, 45 pp., illus. (12) ToLLENAAR, D., and MIDDELBURG, H. A. 1930. PRLNCIPLEä AND RESULTS OF RECENT TOBACCO-BREEDING IN THE voRSTENLANDEN. Procfsta. Voi'stcnland. Tabak [Netherland East Indies], Mcded. 63, 88 pp., illus, [In Dutch. English sum- mary, pp. 85-88.1 828 Appendix

TABLE 3.—Breeding stocks possessing characteristics of special value

Variety or strain Commercial leaf type Origin Points of superiority Remarks

Wisconsin Havana 38_._ Cigar binder Wisconsin Agricultural Exi)eriment High yield, high qualitij' Valuable parent stock for yield and qual- Station, 1915. ity. Havana 142_.. _ do Wisconsin Agricultural Experiment Valuable parent stock. Limited acreage station, 1925. in Wisconsin. No. 301 _ Cigar wrapper Florida Agricultural Experiment Major variety in Florida cigar-leaf area; Station, 1930. valuable parent stock. Round Tip Cigar binder and cigar Connecticut Agricultural Experi- Uigh quality, root rot resistant._ Widely grown in Florida prior to appear- wrapper. ment Station, 1921. ance of black shank. Maryland Mammoth Marvland _ Maryland Agricultural Experiment Produces many leaves, indeterminate Valuable breeding stock for increasing Station and U. S. Department of growth habit. number of lea ves. Agriculture, 1906. Standup IVTaryland Mam- do U. S. Department of Agriculture, High yield, high ouality.. _ Superior in quality to ordinary .Maryland moth. 1933. Mammoth. Maryland Standup .Resis- ---. do...- U. S. Department of Agriculture, Resistant to root rot Limited acreage in ^Maryland at present. to tant. 1930. o Johnson Resistant White Burlev Wisconsin Agricultural Experiment Resistant to black root rot Valuable parent stock. Limited acreage Barley. Station, 1922. in Virginia anc Tennessee. Kentucky No. 5 do Kentucky Agricultural Experiment do___ Extensively planted in Kentucky. Station, 1929. Stand-up Burley Mammoth. do U. S. Department of Agriculture, Produces many leaves of upright growth, Only grown experimentally. 1933. resistant to black root rot. Narrowleaf and broadleaf One sucker Kentucky Agricultural Experiment Reduced production of suckers Breeding stocks for suckeriiig habit, strains of One Sucker. Station. Gold Dollar, strain of Ja- Fluo-curcd Coker Pedigreed Seed Co., South High vield. hio-h aualitv AS'idely grown in flue-cured areas. maica (Orinoco). Carolina, 1930. Special No. 400 .___do__. North Carolina Department of Agri- High yield, high quality, resistant to Limited acreage in North Carolina. culture and U. S. Department of black root rot. Agriculture, 1934. Ambalema Puerto Rico Department of Agricul- Highly resistant to mosaic Valuable parent stock. ture and Commerce, 1933. No. 79A U. S. Department of Agriculture, Modcratelv resistant to wilt.- Do 1935. No. ñ7 do _ Moderatelv resistant to iriildew Do Low-nicotine Cuban 408 Cigar filler :_ U. S. Department of Agriculture, Very low in nicotine Not grown (;onnj]ercially. 1909. A^. rustica, var. Brasilia Importation from Italy, 1913 1 1 At present grown only experimentally. TABLE 4.—Breeding projects under luay and plans for future development

Diseases for ].)iseascs for which resist- Type of tül)arco which resist- Location of work Type of tobacco ance Is liocation of work concerned ance is concerned sought sought

Black root ro|._ Connecticut i... Cigar. Boot knot Georgia! Flue-cured. do Cigar (shade). South Carolina! _ ])o. Massachusetts!. Cigar. North Carohnai. Do. Wisconsin! Do. Black shank-._ Florida Cigar, ¡Maryland i Maryland. North Carolina!. Flue-cured. Kentucky Burlcy. Fusariumwilt. Kentucky Burle y. Tennessee! Do. Ciranvillewilt. North Carolina!. Flue-cured. West Virsïiniai-. Do. Mildew Georgia 1 ])o. North Carolina!. Flue-cured. North Carolina!. Do. Viririnia Do. South Carolina!. Do. Kentucky Dark air-cured. Mosaic Maryland ^ Maryland. do Fire-cured. Kentucky Burley. Viruinia __. Do. North Carolina . Flue-cured. W^iscoDsin 1 Cigar.

1 Cooperative project between U. S. Department of Agriculture and the State Indicated.

Personnel Engaged in Tobacco Breeding—United States Department of Agriculture and State Experiment Stations

Dr. P. J. Anderson, Tobacco Substation, Connecticut Agricultural Experiment Station, Windsor, Conn. Dr. D. F. Jones, Connecticut Agricultural Experiment Station, New ITaven, Conn. Dr. James Johnson, Wisconsin Agricultural Experiment Station and United States Department of Agriculture, Madison, Wis. Mr. W. B. Ogden, Wisconsin Agricultural Expeiiment Station, Madison, Wis. Mr. C. V. Kightlinger, United States Department of Agricultiire and Massachu- setts Agricutural Experiment Station, Amlierst, Mass. Mr. Otto Olson, United States Department of Agriculture and PennsyU^ania Agricultural Experiment Station, Ephrata, Pa. Dr. T. C. Mcîlvaine, West Virginia Agricultural Experiment Station and United States Department of Agriculture, Lakin, W. Va. Prof. E. J. Kinne\', Kentucky Agricultural Experiment Station, Lexington, K\-. Dr. W. D. Valleau, ICentucky Agricaltural Experiment Station, Lexington, Ky. Dr. E. M. Johnson, Kentucky Agricultural Experiment Station, Lexington, Ky. Mr. Frank S. Chance, United States Department of Agriculture and Tennessee Agricultural Experiment Station, Greeneville, Tenn. Mr. D. E. Brown, United States Department of Agriculture and Maryland Agricultural Experiment Station, Upper Marlboro, Md. Mr. R. G. Henderson, Virginia Agricultural Experiment Station, Blacksburg, Va. Mr. E. G. Moss, United States De])artment of Agriculture and North Carolina Department of Agriculture, To])acco Experiment Station, Oxford, N. C. Mr. J. F. Bullock, United States Department of Agriculture and North Carolina Department of Agriculture, Tobacco Experiment Station, Oxford, N. C. Mr. W. M. Lunn, United States Department of Agriculture and Pee Dee Ex[)cr- iment Station, Florence, S. C. Mr. J. V. Williamson, Coker's Pedigreed Seed Co., Hartsville, S. C. Mr. R. R. Kincaid, Nortli Florida Experiment Station, Quincy, Fia. Dr. L. 0. Gratz, North Florida Experiment Station, Quincy, Fla. Dr. J. E. McMurtrey, Jr., Tobacco and Plant Nutrition, Bureau of Plant Indus- try, Washington, D. C. Dr. E. E. Clayton. Tobacco and Plant Nutrition, Bureau of Plant Industry, Washington, D. C. Dr. Harold H. Smith, Tobacco and Plant Nutrition, Bureau of Plant industry, Washington, D. C. 830