Historic, archived document Do not assume content reflects current scientific knowledge, policies, or practices. U.S. DEPARTMENT OF AGRICULTURE. BUREAU OF PLANT INDUSTRY—BULLETIN 10. 96. B. T. GALLOWAY, Chief of Bureau. TOBACCO BREEDING. BY A. D. SHAMEL | AND W. W. COBEY, In CHarGE oF Topacco BREEDING EXPERIMENTS, PLANT BREEDING INVESTIGATIONS. IssuUED MarcH 12, 1907. ‘= CD Any ed PAS AEs WASHINGTON: GOVERNMENT PRINTING OFFICE. IS HOLLG: BUREAU OF PLANT INDUSTRY. Pathologist and Physiologist, and Chief of Bureau, Beverly T. Galloway. Pathologist and Physiologist, and Assistant Chief of Bureau, Albert F. ‘Woods. Laboratory of Plant Pathology, Erwin F. Smith, Pathologist in Charge. investigations of Diseases of Fruits, Merton B. Waite, Pathologist in Charge. Plant Breeding Investigations, Herbert J. Webber, Physiologist in Charge. Plant Life History Investigations, Walter T. Swinele, Physiologist in Charge. Soil Bac teriology and Water Purification Investigations, Karl F, Kellerman, Physiologist in Charge. Beononic ‘Investigations of Tropical and Subtropical Plants, Orator F. Cook, Bionomist in Charge. Drug and Poisonous Plant Investigations and Tea Culture Investigations, Rodney H. True, Physiologist in Charge. Physical Laboratory, Lyman J. Briggs, Physicist in Charge. Taxonomic Investigations, Frederick V. Coville, Botanist in Charge. Farm Management Investigations, W illiam J. Spillman, Agriculturist in Charge. Grain Investigations, Mark A. Carleton, Cerealist in Charge. Arlington Experimental Farm, Lee C. Corbett, Horticulturist in Charge. Sugar Bect Investigations, Charles O. Townsend, Pathologist in Charge. Western Agricultural Extension, Carl S. Scofield, Agriculturist in Charge. Dry Land Agriculture, E. Channing Chilcott, Expert in Charge. Pomological Collections, Gustavus B. Brackett, Pomologist in Charge. eee vestigations in Pomology, William A. Taylor and G. Harold Powell, Pomologists. in Charge. Experimental Gardens and Grounds, Edward M. Byrnes, Superintendent. Seed and Plant Introduction. Investigations, David Fairchild, in Charge. Forage Crop Investigations, Charles V. Piper, Agrostologist in Charge. Seed Laboratory, Edgar Brown, Botanist in Charge. Grain Standardization, John D. Shanahan, Expert in Charge. Mississippi Valley Laboratory, St. Louis, Mo., Hermann von Schrenk, Expert in Charge. Subtropical Laboratory and Garden, Miami, Fla., Ernst A. Bessey, Pathologist in Charge. Plant Introduction Garden, Chico, Cal., Palemon H. Dorsett, Pathologist in Charge. Cotton Culture Farms, Seaman A, Knapp, Lake Charles, La., Special Agent in Charge. Editor, J. EH. Rockwell. Chief Clerk, James EK. Jones. PLANT BREEDING INVESTIGATIONS. SCIENTIFIC STAFF. Herbert J. Webber, Physiologist in Charge. Alkali and Arid Plant Breeding: T. H. Kearney, Physiologist in Charge; L. L. Harter, Scientific Assistant. Corn Breeding: C. P. Hartley, Assistant Physiologist in Charge; E. B. Brown, Scientific Assistant. Cotton Breeding: D. A. Saunders and D. N. Shoemaker, Special Agents in Charge of Cot- ton Laboratories, Waco, Tex.; R. L. Bennett, in Charge of Cooperative Work, Texas Agricultural Experiment Station ; S. M. Bain, in Charge of Cooperative Work, Tennes- see Agricultural Experiment Station ; E. B. Boykin, Special Field Agent; H. A. Allard, Scientific Assistant. Oat and Potato Breeding: J. B. Norton, Assistant Physiologist in Charge. Tobacco Breeding: A. ID. Shamel, Physiologist in Charge; W. W. Cobey, Special Field Agent; W. W. Garner, Scientific Assistant : W. H. Scherffius, Tobacco Expert. 9 ~ 96 EEPIER OF SEKaANSMIEPIAL U. S. DeparTMENT oF AGRICULTURE, BureEAv oF Puantr INpustry, OFFICE OF THE CHIEF, Washington, D. C., September 25, 1906. Str: I have the honor to transmit herewith a paper on “ Tobacco Breeding,” by Messrs. A. D. Shamel and W. W. Cobey, of the Plant Breeding Investigations of this Bureau, and would recommend its publication as Bulletin No. 96 of the Bureau series. The tobacco-breeding work of the Bureau of Plant Industry has now been under way for several years, and results have been obtained which have attracted widespread attention and proved of very great value to practical tobacco growers. This bulletin is intended to summarize the results secured up to date and to place the knowledge obtained before growers in such a form that they will be able to understand it and apply it in the practical work of im- proving their crops. The experiments have shown that tobacco can be improved in many important ways, and the methods by which such improvements can be produced are here described in detail. The illustrations form a very important part of the publication and are necessary to enable the grower to understand clearly the character and improvements discussed in the text of the bulletin. Respectfully, B. T. GaLioway, ; Chief of Bureau. Hon. James WItson, Secretary of Agriculture. 96 B. P. I.—234. POR VOCO hie DING: INTRODUCTION. The growing importance of the tobacco industry may be realized from a brief summary of the estimates of the value of the crop in the United States in the season of 1906. About 796,099 acres of tobacco were grown, producing an average yield of 857.2 pounds to the acre, or a total of 682,428,530 pounds. The average value of the crop was 10.0 cents a pound, or a total of about $68,232,647. While it is almost impossible to comprehend the magnitude of the value of the ‘manufactured products of tobacco, a glance at the total figures may convey some idea of the development of this great and distinctively American industry. In 1900 the total value of the manufactured products of tobacco was $283,076,546. - These products may be divided into three general classes, of which the values were as follows: Cigars and cigarettes, $160,223,152; chewing, smoking, and snuff products $103,754,362; stemmed and rehandled tobacco, $19,099,032. In the manufacture of these products 142,277 people were employed, who earned a total wage of $49,852,484. In addition to the tobacco grown in the United States there was imported into the United States in the year ended June 30, 1906, $4,143,192 worth of tobacco in a manutfac- tured condition and $22,447,514 worth of unmanufactured prod- | ucts, making the total value of the importations during this period $26,590,706. During the same time the exports of manufactured tobacco were valued at $5,410,480, and of unmanufactured tobacco at $28,808,367. In 1891 the tobacco industry furnished almost $50,000,000 revenue to the Federal Government, and the revenue from this source now amounts to about one-eighth of the Government’s total net receipts. ; The United States now grows by far the largest quantity of tobacco produced by any country in the world. While tobacco was grown by the first settlers in the colonies and was one of their principal cash crops, the extensive development of this industry has been a matter of comparatively recent years. The introduction of tobacco into the different sections of the United States, with their widely varying if 96 § TOBACCO BREEDING. conditions of soil and climate, has resulted in the production of types adapted to the soils and conditions of these sections, as well as supply- ing a product for the varied manufactures now. demanded by the consumers of tobacco. Improvements in methods of culture, curing, and fermentation have resulted in the production of tobacco having an increased value, but the most important factor in the development of more valuable tobaccos has been the production of improved varieties by seed selection and breeding. The production of these improved varieties adapted to local soil and climatic conditions has made possible the rapid development of the industry and enabled the United States in a comparatively short time to rank as the foremost tobacco-producing country in the world. The prosperity of the tobacco industry as a whole and of the growers in particular depends on the development of improved varieties of tobacco adapted to the demands of manufacturers and consumers. There is no crop which responds so readily to breeding as tobacco, as has been proved by the experiments of the writers, and it is further true that without careful selection and breeding there is no crop which so quickly deteriorates in yield and quality. The extent of the areas in the United States in which the conditions of soil and climate are suitable to tobacco culture is almost unlimited, so that it seems possible that by giving attention to the production of varieties adapted to those conditions this country can continue to produce an increasingly large yield of valuable tobaccos. The experiments of the writers have shown that it is possible to increase the yield and improve the quality of the crop by seed selec- tion and breeding. The methods of breeding worked out in the course of these experiments are simple and practical and can be carried out by every grower with little or no extra cost in the pro- duction of the crop. The fact that tobacco is perfectly self-fertile and that self-fertilized seed produces more uniform and_ better developed plants than seed resulting from cross-fertilization within the variety makes it possible by the adoption of proper methods of saving seed to make rapid progress in the improvement of the crop. Improvement in the shape, size, and quality of leaves or increase in the number of leaves borne by the individual plants, all of which can be attained by breeding, means increased profits to the erowers and manufacturers, and therefore is of vital interest to al! who are interested in the production, manufacture, and consump- tion of this crop. The production of new varieties of tobacco by hybridization and selection is a most important phase of tobacco breeding. The new hybrids of native New England varieties with standard foreign- erown varieties, combining certain valuable characters of both par- 96 GREAT VARIABILITY OF TOBACCO PLANTS. 9 ents, described in this bulletin, are good illustrations of the use of breeding in the improvement of the tobacco crop. The making and testing of hybrids are matters of experiment and require con- siderable time and expense, but experience has shown that the results justify the necessary expenditure. The production of improved breeds of live stock and varieties of fruits and cereals, in fact, of all crops, might be cited to prove the importance of applying the principles of breeding to the tobacco crop. It is only recently that systematic breeding experiments have been undertaken. It is hoped that the results of the experiments cited in this bulletin will serve as a means of creating general inter- est in this subject and of inducing investigators, breeders, and grow- ers to turn their attention to the further improvement of their crops. Tobacco growers in the sections where these experiments have been carried on have generally adopted the improved methods of bagging carefully selected seed plants and of separating the seed, and they are using the improved varieties of tobacco produced in the course of these investigations. In most of these districts cer- tain men have become interested in the careful and systematic breed- ing of tobacco. THE GREAT VARIABILITY OF TOBACCO PLANTS. Under the intensive system of cultivation necessary for the pro- duction of profitable crops of tobacco, the condition of the soil, the fertilization, and the cultivation are fairly uniform so far as individual fields are concerned. In those tobacco-growing sections where the best grades are produced it is a common practise to grow tobacco year after year on the same field without rotation, instances being known where more than fifty consecutive crops have been pro- duced on the same field. This system of cropping enables the grow- ers to become thoroly familiar with the character ‘of the soil in all sections of the fields, so that any inequality in fertility can be reme- died by the judicious application of manures or commercial fertili- zers, or by methods of cultivation. Notwithstanding these favorable circumstances for the production of uniform plants, a careful study of the plants in these fields reveals a great lack of uniformity as regards all characters. This lack of uniformity is particularly noticeable with respect to the variation in number, size, venation, shape, and habit of growth of the leaves borne by individual plants, the time of ripening of the leaves on the same plant and on different plants, the number and size of the suckers, and the structure and arrangement of the flowers and flower heads. From the practical standpoint, there is no more important problem in tobacco culture than the production of uniform crops. A lack of uniformity in the 96 1 re et 10 TOBACCO BREEDING. crop not only results in a low yield, as a whole, and more especially of the best and most profitable grades of the cured and fermented product, but also increases the cost of sorting out the different types of leaves into their respective grades for market, the expense of which must be borne directly or indirectly by the grower. The principal cause of the lack of uniformity in tobacco is cross- fertilization. In tobacco, as in all other crops, seed resulting from cross-fertilization produces many plants unlike either parent. There- fore such seed is undesirable for the general planting of a crop where uniformity is so important a factor. Where the tobacco seed plants are grown without protection from cross-fertilization some of the flowers are cross-fertilized by insects or other agencies. De- sirable plants may thus be crossed with undesirable plants in the > same field or in adjoining fields, and the plants grown from the seed thus produced are usually extremely variable, some of them resem- bling the desirable plants from which the seed was harvested, others resembling the inferior plants from which the pollen was carried for crossing, while the remainder are of an intermediate type, un- suited to the purpose for which the crop is grown, and therefore causing a loss to the grower. The writers have observed number- less cases in different tobacco-growing sections where several dis- tinct and worthless new types appeared in the fields, the plants of which were grown from carefully selected seed. These undesirable types could only be accounted for by the accidental crossing of the seed plants the year preceding or at some previous time. The cross- ing of individual plants of the same strain, even if both are desirable plants, results in undesirable variations, many of which are appar- ently reversions to earlier and unimproved types of tobacco. In those varieties of tobacco in which the buds are removed long before the flowers open on all of the plants except those saved for seed production, or where early topping is practised, the opportunity for the crossing of the flowers borne by the seed plants with other plants in the same field is almost wholly limited to the seed plants. However, it frequently occurs that late or diseased plants, or pos- sibly sucker branches that have been overlookt, develop flowers which open at just the right time to allow insects to carry the pollen from these flowers to the seed plants and thus effect cross-fertiliza- tion. There is little doubt that many of the plants of irregular and unusual types are produced as a result of this kind of cross-fertiliza- tion. An important cause of variation in tobacco plants is the use of immature seed. Many growers cut off or harvest the seed heads before all of the seed pods have turned brown; hence, before ma- turity. The writers have observed hundreds of instances where the 96 GREAT VARIABILITY OF TOBACCO PLANTS. Vek seed plants have been cut off while many of the flowers were still in bloom. On such seed heads seed pods in all stages of maturity can be found. Some of the pods are fully ripe and contain mature seed, while others have not fully developed. Much of the seed is imma-_ ture and contains little food for the nourishment of the plantlet. These seed heads are frequently thrashed out with a flail or the pods are crushed by the hands in order to shell out the seed. In this way the immature seed is mixt with the ripe seed sown in the seed beds. In the seed beds the immature seed frequently sprouts earlier than the mature seed, and the early seedlings grown from such seed are naturally used for transplanting in the field. Such plants have a great tendency to vary, in some cases being very early, and as a rule having leaves that are small, coarse, and wholly undesirable for any purpose. These weak, immature tobacco seeds, according to careful and extensive observations by the writers, produce plants which are more subject to certain diseases, particularly the mosaic disease, than are plants grown from mature seed. The excess of plant food in the soil where heavy applications of barnyard manure and commercial fertilizers are used is usually thought to produce variations in the plants. This variation is usually shown by an increase in the size of the leaves, which is generally correlated with changes in color, flavor, and other characters. In these cases there is usually a tendency for the type of plant to break up, so that the uniformity of the crop is disturbed. Where it is necessary to use large quantities of fertilizers in the growing of a profitable crop, the inclination to variation induced by this intensive system of cultivation must be controlled by the most rigid selection of seed from the type of plants best adapted for the purpose for which the tobacco is grown. The change of soil and climatic conditions, particularly the taking of tobacco seed from southern or tropical conditions to the north, is a fertile source of variations in tobacco. The fixation of a uniform type in this case requires several years of acclimatization, supple- mented by selection of seed from the desirable plants. In the production of improved varieties of tobacco by breeding, va- riation in type can be secured by crossing, and by continued saving of self-fertilized seeds from plants most nearly reaching the growers’ ideal of perfect plants uniform types can be fixt. Growers will fre- quently find plants that are markedly better than the rest of the plants in the field, so that by selecting these desirable variations a steady im- provement in the yield and quality of the crop can be effected. Vari- ation, therefore, is a basis for selection in an experimental way, but in practise every effort must be put fofth in order to secure uni- 96 12 TOBACCO BREEDING. formity of the plants in the field and thus produce the most profit- able crops. . The variations in tobacco plants may be divided into two general classes—variation in type and individual variability within the type. The causes of the variations in type, or striking variations, include crossing and change of soil and climatic conditions, particularly the change of seed from the Tropics to temperate regions. The causes of individual variations within the type include the fortuitous va- riations or inherent tendency to variability, methods of soil fertil- zation and cultivation, maturity of seed, and various local conditions. With an understanding of these conditions the grower can to a great extent control the degree of variability by methods of saving seed, systems of cultivation, and other practical methods of culture. THE INTRODUCTION AND ACCLIMATIZATION OF VARIETIES. The introduction of the seed of standard foreign-grown varieties of tobacco has been the source of increased wealth and prosperity in certain tobacco districts of the United States. In other regions such importations have resulted in great financial loss to the grow- ers, which in most cases has been due to a lack of knowledge of the effect of the change of soil and climatic conditions on the particular type of tobacco grown. The writers have had an unusual opportu- nity in the course of their work to observe the behavior of crops grown in different tobacco-growing sections from imported seed, and have conducted extensive experiments in taking seed from one dis- trict to another, with a view to securing definite information on this subject for the benefit of the growers. The results of these observations are presented here for the guidance of tobacco growers who desire to use foreign-grown seed or who wish to change their seed. The western Florida and southern Georgia Sumatra tobacco in- dustry is an illustration of the successful introduction of a foreign- grown variety of tobacco. Tobacco growers in this region secured small samples of the seed of the Sumatra variety of tobacco from the island of Sumatra. At first small experimental crops were grown and seed saved from the best plants in these crops. In the course of this experimental cultivation it was noticed that the plants grown under the partial shade of trees in freshly cleared fields produced finer and more desirable leaves for cigar wrappers than the plants grown in the open. This fact led to the erection of an artificial shade over the fields, made of slats laid on a suitable framework. This method of growing tobacco was introduced about 1896 by Mr. D. A. Shaw, of Quincy, Fla. Later, other growers 96 INTRODUCTION AND ACCLIMATIZATION OF VARIETIES. 13 used a coarse cheese cloth as a substitute for the slats. The shade method of growing tobacco in this region has developed rapidly, and at the present time several thousands of acres of tobacco are grown under either slat or cloth shade, and the industry has become established on a profitable and successful basis. During this time considerable attention has been paid to the production of a uniform type of tobacco adapted to the climatic and soil conditions of this section by the saving of seed from carefully selected plants of the Sumatra variety. When Sumatra seed was first introduced into Florida the variety broke up into a number of different types, some of which were de- sirable, while others were undesirable. By reason of the small crops grown from such seed, the loss to the growers from the production of undesirable types of plants was not very great. The growers naturally saved for seed those plants which produced the most de- sirable types of tobacco, and as a result of continued selection of this kind a fairly uniform type of tobacco which was adapted to the local conditions in this section was secured. As the demand for this Florida-grown Sumatra tobacco developed, resulting in in- creased acreages, seed was at hand which was thoroly acclimatized for planting the larger area devoted to this crop. From time to time the tobacco planters in this region have obtained small quanti- ties of seed from Sumatra, but in such cases this seed has been grown in a very limited way in very small fields until it has been acclimatized and uniform types have been secured by seed selection. About the time of the Cuban revolution it became apparent that the supply of Cuban-grown tobacco for the use of cigar manufac- turers in the United States might become limited by reason of the unfavorable conditions for tobacco growing then prevailing in Cuba. Under these circumstances it was thought to be a propitious time to introduce the growing of Cuban tobacco into southern Florida, where the conditions of climate and soil were believed to be similar to those of Cuba. Considerable public interest was aroused in this project, and as a result large quantities of Cuban-grown seed were secured and planted in certain sections of Florida. . The crops raised from this seed proved to be a disappointment to the growers. The change of soil and climatic conditions resulted in the breaking up of the type of the Cuban variety into a large number of sorts, some of which were desirable, while others were undesirable. Many of the plants developed a branching habit of growth, bearing very small, undesirable leaves of poor quality, resulting in a very low yield of an inferior tobacco. One of the main causes of failure was the lack of understanding on the part of the growers of the effect of the change of conditions on the type of tobacco and their 96 laa A ET A RR RS mR ae 14 TOBACCO BREEDING. neglect to appreciate the necessity of securing strains of plants by seed selection of the desirable types adapted to the particular con- ditions of soil and climate in southern Florida. If the acclimatiza- tion of these strains had been accomplished by seed selection in small fields, with little loss to the growers, the strains could have been grown on a more extensive scale with better chances of success. In order to illustrate the necessity for the acclimatization of a variety of tobacco before it 1s grown on an extensive scale, the successful experiments of the Bureau of Soils in the introduction of Cuban tobacco in Texas may be cited. After a previous unsuccess- ful attempt by farmers in Texas to grow Cuban tobacco from freshly imported Cuban seed the Bureau of Soils began systematic experiments in growing small fields of tobacco and saving the seed of the most desirable plants according to the method described — in this bulletin. In these crops certain plants were found which produced leaves possessing the flavor and aroma desired in a high- grade filler tobacco. The seeds from these plants were saved under bag, and their product has been found to possess the desirable char- acters of the parent plants. ~This tobacco has been sold at profitable prices, and the area devoted to the growing of this crop is being gradually extended in order to meet the demands of the manufac- turers for this grade of filler tobacco. In northern Florida the tobacco growers, as a result of their experience with the imported Sumatra seed, experimented in growing, in the open, small fields of a cigar filler tobacco of a variety the seed of which was originally introduced from Cuba. This variety of Florida filler tobacco is now being grown extensively and profitably in that section. The best illustration of the effect of a change of climatic and soil conditions upon the character of a variety of tobacco is found in the experience of tobacco growers in the Connecticut Valley in the planting of Florida-grown Sumatra seed and seed of the Sumatra variety imported from the island of Sumatra. As dis- cust in an article upon the improvement of tobacco by breeding and seed selection in the Yearbook of the Department of Agriculture for 1904,* tobacco growers in the Connecticut Valley in the seasons of 1901 and 1902 grew extensive crops from seed introduced from Florida and Sumatra. In a careful examination of these fields it was found that the change in conditions had resulted in the breaking up of the type of the variety, so that several distinct types of tobacco were found growing in the same fields. Some of these types of plants produced well-rounded leaves, with fine venation and the elasticity, strength, gloss, grain, and other characters necessary in a@Shamel, A. D. Yearbook of the Department of Agriculture, 1904, pp. 435-452. 96 INTRODUCTION AND ACCLIMATIZATION OF VARIETIES. 15 a high-grade cigar wrapper tobacco. However, many of the types of plants produced long, narrow, coarse, pointed leaves, wholly unsuited for cigar wrapper manufacture. In the case of another inferior type it was found that the leaves would not burn, altho the remaining characters were those of a desirable grade of tobacco. It -was impossible to sort out this type, even by the most careful inspection of the crop, and, as a result, when the manufacturers wrapt cigars with leaves of this type and found that the wrapper would not burn, the quality of the entire crop was condemned. Certain other types of plants produced leaves of such thin texture or ight body that when wrapt on cigars and allowed to dry out the wrapper frequently broke, or when the consumer carried the cigars 1n his pocket the wrappers were easily injured. | One of the most striking types of plants produced in the crops grown from this imported seed was the Belgian type, an illustration of — which is shown in figure 1. In this case the plants bore leaves measur- ing from 30 to 46 inches long and only from 5 to 10 inches wide. These leaves, as shown in figure 2, were very pointed in shape, with coarse, angular veins, and as the cured tobacco lacked the appearance Fic. 1—Belgian type of Connecticut Suma- necessarv for a cigar-wrapper to- tra tobacco plant. These long, narrow De : leaves with oblique veins, coarse texture, baeeco its production was a total burn very poorly, and after curing light loss to the growers. The variation green color are almost absolutely worth- : : 2 i less for clgar-wrapper purposes. Thistype In type Was accompanied by differ- appeared in crops grown in the Connecti- ences in time of maturity, so that cee ory tones srowne cue e seed. : the cost of the growing and har- vesting of these crops was greater than in the case of uniform ~ crops. The mixture of types was accompanied by great variation as regards the individual plants of each type. In many cases plants bearing 25 leaves were found growing by the side of plants pro- ducing 10 leaves. Marked variations in size and shape of leaves and in the number of suckers borne by the individual plants were observed, and as a result it was found that the comparatively small number of desirable types could not make up for the loss in the pro- duction of undesirable and worthless grades of tobacco. As a nat- ural consequence of this condition the extensive culture of this vari- ety of tobacco from imported seed has been abandoned, and the grow- - 96 ao oe a ea cn ain 16 TOBACCO BREEDING. ers now have small fields and are selecting those plants producing the most desirable grade of cigar-wrapper tobacco with a view to securing strains which are adapted to the conditions of the Connecti- cut Valley. In the experimental fields of the Bureau of Plant Indus- try, covered with cloth shade, where seed of de- sirable plants has been saved under bag for three seasons, uniform strains, which are absolutely free from the unusual or distinct types observed in the fields from which the original selections of seed were made, have been produced. A few of the growers of the Sumatra variety in the Connecticut Valley introduced Cuban- grown seed and used it for planting their general crops. In these fields the breaking up in type was not so noticeable as in the case of the Suma- tra variety, but the effect of change of conditions in the variety was shown in the production of so-called freak plants. These plants had a branching, or suckering, habit of growth, bearing very small, sharply poimted, coarse leaves that were worthless for cigar-wrapper purposes. In a careful study of the plants in these fields it was found that at least one-third of the entire crop consisted of these freak plants. In one of these fields the writers made selec- tions of seed plants of the most desired type, bearing the size, shape, and general character of leaves adapted for cigar wrappers. This seed was saved under bag, and a similar plan has Fic. 2.—Typical leaf of Belgian type of tobacco, showing the character- istic shape, venation, and other characters of this variety which are been followed up to the present time. In Plate I, figure 1, the original crop raised from freshly imported seed from which the seed wholly unsuited for ci- gar-wrapper manufac- ture. The presence of such types of leaves re- duces the value of the crop and is detrimental _ to the reputation of the variety of tobacco in which they are pro- duced. selections were made is shown. In Plate I, figure 2, is shown a crop grown on the same field after two generations during which the seed was saved under bag. As can be seen from the illustrations, this method of seed selection and bagging has produced a uniform type of to- bacco without the freaks and other undesir- able types of the original crop. It is fortunate that on this farm the planter saved seed in the open for his own use from the same field. The crop grown from such seed was visited during the season of 1905 by the writers, in company with Dr. H. J. Webber and several tobacco growers, and was found to contain a large proportion of freaks; in fact, about the same proportion as the crop grown from 96 STRUCTURE AND ARRANGEMENT OF FLOWERS. 17 freshly imported seed. In other words, the seed saved under bag pro- duced uniform strains adapted to soil and climatic conditions in this section, while seed saved in the open and subject to cross-fertilization with freaks and other undesirable types produced about the same pro- portion of freak plants as the crops grown from the freshly imported seed. The writers during the past season planted in Florida Connecticut- grown seed of the Sumatra variety. It was found that while there was a noticeable change in the shape of leaf and in some minor char- acters in the Florida-grown tobacco, there was no violent breaking up of type or indication of unusual variability. This experiment and other observations have led the writers to believe that the effect of changing seed from the north to the south is not accompanied by such marked changes as when seed is taken from tropical conditions to northern latitudes. In summing up the observations on this subject it can safely be said that it is a dangerous policy to plant large crops of tobacco with imported seed or with seed from a very different section. In most cases it has resulted in failure and caused considerable loss to the growers. The general crops should be planted from seed pro- duced under the same conditions as the crop which is to be grown. If it is necessary to change the seed or desirabie to test imported vari- eties, it should be done on a small scale, followed by a most careful selection of seed plants, and the seed should be saved under bag, safe from cross-fertilization. THE STRUCTURE AND ARRANGEMENT OF FLOWERS. A careful study of the tobacco flower is one of the most essential factors in the beginning of tobacco breeding. Successful results, particularly in the production of new varieties, can seldom be ob- tained until one becomes familar with the structure of the parts of the flower and the manner in which these parts perform their sev- eral functions. A full realization of the ease with which crossing takes place can only be obtained in this way, and, as has been pre- | viously stated, the prevention: of promiscuous cross-pollination is of first importance in the production of a desirable and uniform type of tobacco. _ The tobacco flowers are arranged upon a branching determinate flower head, which appears when the middle leaves are about half grown and continues to develop and produce new flowers during the rest of the life of the plant. Figure 3 is a diagrammatic sketch of a tobacco flower, showing the parts of the flower and the general way in which pollination takes place. The calyx (a) is the outer, green, five-parted, floral envelop at the base of the flower which serves to 15507—No. 96—O7 Mm 2 , 18 TOBACCO BREEDING. protect the flower in the bud. The corolla (6) is the delicately colored floral envelop inclosing the reproductive organs of the flower. Tts color tends to attract insects, which are the principal agents in cross-poliination. Next inside the corolla are the five stamens, which are the male reproductive organs of the flower. Each stamen consists of the filament (7), supporting the anther (7) in which the pollen grains (4) are produced. The central organ is the pistil, or female part of the flower. The terminal enlarged portion (q) is the stigma. The pollen grains (/) adhere to the surface of the stigma (g) and germi- nate, sending an extremely minute tube (7) down thru the central conductive tissue (h) of the style (7). This tube extends into the cavity of the ovary (d) and finds its way into the ovule (7) thru a small duct or micropyle (m), where fertilization takes place. Other ovules (e) are fertilized by other pollen tubes. These ovules develop into seeds after fertilization. The ovary is two- celled, with a fleshy central placenta (c) on which the ovules are borne. The early capsules mature- always before flowering ceases. The shape of the del- cately colored corolla is somewhat tubu- lar, or, perhaps, more nearly lke an elongated funnel. It is comparatively small from the basal end to a point about two-thirds the distance to the ter- minal end of the flower. At this point it enlarges suddenly to more than twice the size of the basal part of the tube (fig.83). It is composed of 5 petals, TES ear een ene which coalesce to form the corolla tube, and separate only at the extreme end. The tobacco flower is symmetrical. The number of sepals and stamens is always the same as the number of petals, but these floral circles do not remain constant, varying rather indefinitely in different strains and even among individuals of the same strain. Trimerous flowers, or flowers with three parts in each flower circle, have been found growing on the same plants with pentamerous flowers, or those having five floral parts. This is the exception, however, and not the rule. The tobacco flower is naturally self-fertile, and plants grown from 96 STRUCTURE AND ARRANGEMENT OF FLOWERS. 18S) self-fertilized seed are always stronger and more vigorous than those from cross-pollinated seed when the crossing is within the variety. In Sprengel’s discourse on the cross and self pollination of plants the statement is found that “nature seems to have wished that no flower should be fertilized by its own pollen.” Later, Darwin stated that “ nature abhors perpetual self-fertilization,” but, unlike Spren- gel, recorded a number of exceptions’ to this rule, and tobacco was among them. The experiments of the writers conclusively substan- tiate the findings of Darwin in this connection. They have found self-pollination in the case of tobacco to be most desirable in all cases. A very interesting phenomenon of growth takes place in the fila- ments of the stamens immediately after the opening of the flowers, which can be taken- as an evidence of the natural self-fertilizing habits of the plants. An examination of the flower just previous to the time of opening will reveal the fact that the pistil is longer and extends up beyond the stamens, but when the anthers open and the stigma becomes receptive a very rapid growth of the filaments takes place, which causes the open pollen sacks to be pushed up past the stigma, and in almost all cases they come in direct contact with the stigma in passing upward. This gives an opportunity for at least a portion of the pollen grains to adhere to the viscous surface of the stigma and for self-fertilization to take place, as shown in figure 3. It is just before this process occurs or while it is in progress that there is danger of, or opportunity for, cross-pollination. The open flower contains a small drop of nectar at the base of the corolla tube, which is sought by honeybees, bumblebees, and humming birds, as well as by many species of minute insects, all of which carry pollen from flower to flower and from plant to plant in their constant search for the honey-like substance secreted in the corolla tube. In passing in and out of the flowers the bodies of the bees and other insects and the beaks of the humming birds become dusted with pollen, which is transported by them to the pistils of the next flower visited. The ovules are as readily fertilized by pollen from the surrounding plants as by the pollen from the flower in which they are produced. This continuous crossing necessarily results in the introduction and intermixture of poor and undesirable varieties in our best strains of tobacco. The observations of the writers plainly show the absolute necessity for protecting the flowers of the seed plants from cross-pollination. Careful experiments have demonstrated that in many instances the stigma of the tobacco flower remains in a receptive condition for three or four days. This condition results in a twofold disadvantage when no protection is used against cross-pollination. In the first place, it affords abundant time and opportunity for complicated 96 ster Ne ern toe 20 TOBACCO BREEDING. crossing, for each flower is visited many times a day by various in- sects and often by humming birds; and, secondly, it brings about con- ditions favorable for the production of seed of weak vitality. Pre- vious experimenters have pointed out the fact that the best seed is not produced as a result of premature or late pollination, either of which is likely to occur in the case of tobacco flowers under natural con- ditions. When fertilized only by pollen of the same flower, the pollination takes place at exactly the right time, or when the stamens push past the receptive stigma, which results in the production of a superior grade of seed. : The readiness with which tobacco flowers can be cross-fertilized greatly facilitates the opportunity for producing new and valuable varieties by artificial cross- ing. In the course of the experiments here recorded it has been found per- fectly possible to combine certain desirable qualities found in different strains and at the same time to eliminate some of the un- desirable characters by pro- ducing hybrids between strains of tobacco. The method of cross-pol- lination used in the experi- ments of the writers is to FG. 4.—Three tobacco flowers at proper stage foremascu- Yemove all capsules, open lation, and the scissors and forceps used in emasculat- flowers and flower buds ing fiowers. The corolla of the central flower has been y opened in order that the anthers may be conveniently from the flower head ex- removed. The flower on the left has been emasculated cept those which are in the preparatory to cross-pollination. | proper stage of develop- ment to open within the following twelve or fifteen hours. In preparing these remaining flowers they must be carefully opened and emasculated by the use of a scalpel, small scissors, and a fine pair of forceps, as illustrated in figure 4. Great care must be taken in re- moving the anthers before they have dehisced, in order to avoid injury to the stigma. The emasculation should be done in the after- noon, after which all of the flowers must be carefully covered with a thin paper bag as a protection against insects or other agencies whereby pollen might be transferred to them. In the forenoon of the following day the emasculated flowers should be ready for polh- nation, but the exact time for applying the pollen must be determined in the case of each individual flower by the appearance of the viscid, sticky fluid on the surface of the stigma. The pollen from the male 96 STRUCTURE AND ARRANGEMENT OF FLOWERS. yA parent plant can be best applied to the stigma of the female with the point of a scalpel or other sharp instrument. When apphed with a brush there is danger of some of the pollen grains adhering to the hairs of the brush after each operation, resulting in considerable mixture of pollen, but where the scalpel is used there is no difficulty in removing all the pollen after each operation. The paper bag must be replaced over the flowers as soon as they have been pol- linated, and must be allowed to remain until the seeds have set and all danger of crossing has past. In crossing it is not essential that both of the parent varieties be grown in the same community. Pollen from tobacco flowers when thoroly dry will keep for several weeks or longer without deteriora- tion. The writers have sent pollen thru the mails a distance of more than a thousand miles with perfectly satisfactory results. When not intended for immediate use, it should be harvested when perfectly dry and carefully taken off the anthers after they have dehisced and become dried out. These dry anthers may be put in small vials, and the pollen kept long enough for all practical purposes, provided the vials are carefully corked and kept dry. The large number of seed produced in a single pod and on a single plant makes it possible to obtain definite results from selection or hybridization in tobacco much more quickly than in the case of most other plants. Careful counts show that from 4,000 to 8,000 seeds are produced in a single pod of normal size, and an estimate of the aver- age number of pods on each plant shows that the ordinary tobacco plant produces from 500,000 to 1,000,000 seeds. In many cases the writers have secured from 1 to 14 ounces of seed from a single plant when the seed has been saved under bag according to the method out- lined in this bulletin. This large number of seeds gives an excellent opportunity for testing each selection or cross on a large scale. The quantity of seed produced varies inversely with the number and size of leaves on the plant. The production of a large number of good- sized leaves 1s almost invariably accompanied by the production of a small quantity of seed. On account of the large quantity of seed produced by a single plant under normal conditions and the fact that the various charac- ters of a tobacco plant are inherited so strikingly and uniformly by its progeny the following year when the seed is saved under bag, pro- tected from cross-fertilization, it is possible for the tobacco grower to secure uniformity with a considerable degree of improvement in type, quality, and yield by one year’s selection. One plant often fur- nishes enough seed for an entire crop, and the plants raised from this. seed always produce a very uniform lot of tobacco when cross-fertili- zation is not allowed to take place. 96 99, TOBACCO BREEDING. THE NECESSITY FOR INBREEDING. Tn the season of 1903 the writers, in company with Dr. H. J. Web- ber, visited the tobacco fields of the Connecticut Valley in response to a request of the growers for assistance in the production of uniform strains of tobacco by breeding and seed selection. During the survey of this region with a view to gaining an idea of the variability of the varieties of tobacco, it was determined to inaugurate a series of ex- periments in a practical way for the investigation of the methods of saving seed. In view of the results of the investigations of Darwin and others on the comparative vigor of growth, seed production, and other char- acters of tobacco plants raised from seed obtained by cross and by self fertilization * the seed of select plants of the different types of tobacco was protected from cross-fertilization by inclosing the flower heads with a hight but strong form of paper bag. Other seed plants were saved without such protection, as is ordinarily done by the tobacco growers. The seed harvested from these seed plants was saved sepa- rately, stored in small glass vials adapted to this purpose, and labeled according to the system now in use by the breeders in the plant breed- ing investigations of the United States Department of Agriculture. The record of the number of leaves, size, thickness, shape, and color of leaves, number of suckers, height of plant, habit of growth of leaves and plants, time of maturity of leaves and seed, and other characters was kept according to the system used by Doctor Webber, modified by the writers for use in keeping a pedigree of tobacco varieties. The seed of the plants finally selected for experimental purposes was sown in ordinary seed beds, separated into many small sections by thin board partitions, each of which was capable of holding 500 seedlings. The seedlings from these separate seed-bed plats were transplanted to separate rows or plats in the experimental field, each row or plat being carefully labeled so that the plants could be traced directly back to the original seed plants. The manuring, or fertili- zation, and preparation of the soil in the experimental field and the transplanting, cultivation, and harvesting of the plants were all conducted with the greatest possible care to give all of the rows or plats equal opportunity for growth. For instance, the seed of all of the plants of a variety was sown the same day, and at the proper time the seedlings of this variety were all transplanted the same day. At the time of harvest the leaves of the individual rows or plats were primed or the plants cut on the same day, and the leaves or plants were hung in the curing shed, so as to get as nearly uniform condi- a Darwin, Charles. Cross and Self Fertilization in the Vegetable Kingdom, pp. 203-215. 96 NECESSITY FOR INBREEDING. ; 23 tions for curing as were consistent with the practical handling of the crop. The leaves of plants selected for seed were harvested sepa- rately and labeled so that the product of each plant could be intelli- gently used in comparative tests of the cured and fermented crop. This labeling process involved considerable extra work and attention in the field, curing sheds, and warehouses, but was absolutely neces- sary for a definite selection of seed plants for the next season’s use, based on the character of the fermented tobacco. The rows or plats of plants grown from seed of individual plants saved under bag, 1. e., self-fertilized seed, showed remarkable uni- formity in type, size, shape, and appearance of leaves, habit of growth, and all other characters, and conformed closely to the type of the parent plants from which the seed was saved. Plate II, figure 1, shows a type of parent plant and Plate I, figure 2, the progeny of this plant, raised from self-fertilized seed. From this illustration it can be seen that the transmitting power of tobacco is most strongly marked and the progeny of plants raised from inbred seed remarkably uniform in all characters, every plant closely resembling the parents. In the hundreds of tests of this char- acter which have been carried on by the writers during the past three seasons, not only in the Connecticut Valley but in Maryland and Florida, the benefits to be derived from using inbred tobacco seed have been confirmed and emphasized. It is true that some plants have the power of transmitting their characters to their progeny more strongly than others, but on the whole every case under obser- vation has offered additional evidence of the value of the practise of saving tobacco seed under bag, free from possible cross-fertilization. The continued saving of self-fertilized seed for three seasons has furnished no evidence of a decrease in the rate of growth or constitutional vigor of tobacco plants as a result of this practise. On the other hand, by reason of the selection of the best plants in the different varieties every season there has been a marked increase in the productiveness and the general vigor of constitution of the varieties under consideration. This conclusion is emphasized by the vigorous and productive strains of Connecticut tobacco shown in Plate III. : Self-fertilization is the closest possible degree of inbreeding, and it is the general impression that this practise is usually associated with a loss of vigor of growth, with a predisposition to disease, and other undesirable results. In tobacco, so far as our experience goes, this does not happen, and the exact opposite of this condition obtains, viz, that inbreeding is beneficial to the general development of the variety. It is unfortunate that it is impossible to present tabular data at 96 24 TOBACCO BREEDING. the present time showing the behavior of plants raised from arti- ficially cross-fertilized seed within the variety in comparison with plants raised from self-fertilized seed. The principal object of this work has been the achievement of practical results, so that the opportunity for scientific observations and experiments has been necessarily limited. However, the writers have had the privilege of making careful observations on the results of saving seed from plants grown under large field tents and comparing the plants raised from such seed with the plants raised from self-fertilized seed. Under these tents there is little opportunity for cross-fertilization with other varieties, except thru the small doors opening into these tents, which are kept closed all of the time when persons or teams are not actually passing thru them, so that there is little hkelihood of bees or other insects passing in and out. The probability is that the cross-fertilization that takes place is wholly between the plants saved for seed in these tents or with other plants under these shades that are in bloom at the proper time for cross-pollination. The comparison of the plants raised from seed saved under these field tents and exposed to cross-fertilization with the surrounding plants and of plants of the same variety raised from seed protected from cross-fertilization by paper bags shows that self-fertilized seed produces more uniform, vigorous, and productive plants than the open-fertilized seed, which is to a greater or less extent cross-fer- tilized between plants of the same variety. It appears that the cross-fertilization of tobacco seed, even tho it occurs between good individuals, has a tendency to seriously break up the type. Along with the variability of type induced by cross- fertilization, it frequently happens that many freak plants resembling the wild species appear; these can only be explained with our present knowledge of the subject as reversions. Such freak plants are not usable for profitable manufacture, and consequently are a source of loss to the growers. The size and weight of seed from the inbred plants are equal to and in most cases greater than the seed saved from open-fertilized plants. In a series of comparative tests of the two kinds of seed in the case of four varieties grown in the Connecticut Valley it was found that the inbred seed was heavier and larger than the cross- fertilized seed. The total quantity of seed harvested from the open-fertilized plants usually exceeded that of the inbred plants. This was due to the fact that in the case of the inbred plants more of the seed-bearing branches were removed than where the plants were allowed to set seed under natural conditions, in order to adapt the seed head for the best possible development under the paper bags. Where an equal number of seed pods was examined for 96 NECESSITY FOR INBREEDING. 25 vield of seed the inbred seed equaled or exceeded in quantity the cross-fertilized seed in the variety. In the case of hybrids or in the crossing of two distinct strains or varieties the yield of seed, as well as rate of growth of the hybrid plants, was greater than thai of the inbred seed and plants. It is true that some of the improved inbred strains produce but little seed compared with unimproved types of the same variety. Inbred strains have been selected for increased yield and number of leaves, which seem to be correlated with lessened seed production. The same correlation holds true where open-fertilized strains have been selected for increased num- ber and total yield of leaves. The rate of germination of the inbred in comparison with the cross-fertilized seed was thought to be slower in some of the experi- ments carried on in the season of 1904. However, further compari- sons have failed to bear out this conclusion, and it is the behef of the writers, based on careful observations on this subject, that the inbred seed sprouts as rapidly as the cross-fertilized seed. It seems probable that in the cases observed by growers in 1904 a difference in moisture content of the rotted apple-tree fiber, the medium used for sprouting, was the cause of the apparent differences in time of sprouting. So far as the writers’ observations go, the inbred seed produces more rapidly growing plants than the open-fertilized seed, and consequently earlier plants for transplanting. There is no doubt, Turther, that the inbred seed produces a larger proportion of seed- lings for transplanting at one time than the open-fertilized seed, which is an important matter to the tobacco grower, who is fre- quently forced to wait for seedlings on account of the lack of uni- formity of plants in beds sown with open-fertilized seed. Darwin’s conclusions on the comparison of tobacco plants raised from inbred and cross-fertilized seed for three years are as follows: ¢ Taking the plants of the three generations altogether, the crossed show no superiority over the self-fertilized, and I can account for this fact only by sup- posing that with this species, which is perfectly self-fertile without insect aid, most of the individuals are in the same condition as those of the same variety of the common pea and of a few other exotic plants which have been self-fer- tilized for many generations. In such cases a cross between two individuals does no good; nor does it in any case, unless the individuals differ in general constitution, either from so-called spontaneous variation or from their pro- genitors having been subjected to different conditions. I believe that this is the true explanation in the present instance, because, as we shall immediately see, the offspring of plants which did not profit at all by being crossed with a plant of the same stock profited to an extraordinary degree by a cross with a slightly different subvariety. @Parwin, Charles. Cross and Self Fertilization in the Vegetable Kingdom, Dp. -210: 96 26 TOBACCO BREEDING. These conclusions of Darwin were based upon greenhouse and gar- den tests, where, of course, it was not possible to study and compare the characters of quality or the value of the tobacco from the inbred and cross-fertilized seed. The observations of the writers upon to- bacco grown in the field under normal conditions in the different tobacco-growing sections bear out the conclusions of Darwin on this subject, and show, further, that the inbred seed produces more profit- able crops of tobacco than the seed resulting from open or cross fer- tilization within the variety. THE IMPROVEMENT OF THE SHAPE OF LEAVES. The shape of the leaves is a very important factor in determining the value of all classes of tobacco, and is of first and particular im- portance in cigar-wrapper varieties. Many varieties which possess some of the most desirable characteristics of high-grade wrappers are totally valueless for this purpose on account of the narrowness of the leaves. From such leaves it is impossible to cut cigar wrappers eco- nomically. The manufacturer of cigars demands a leaf which is wide and well rounded at both ends. This shape admits of the best op- portunity for cutting into wrappers of the desired shape and size with the least possible waste, while the long, pointed leaf will yield very few wrappers, and a very considerable proportion of it must be consigned to the waste pile. The long, pointed leaf is not only undesirable because of its shape, but the texture toward the basal end is poorly adapted for cigar wrappers and the grain is usually unevenly distributed. In such cases a large part of each leaf can be utilized only for binders or low-grade fillers. A striking example of leaves of this character may be found in the case of many strains of Connecticut and of | Pennsylvania Broadleaf varieties. When working these varieties the manufacturer expects to cut wrappers from the middle portion and tips of the leaves only, while the remainder of the leaves, often half or more, must be used as binders or for filler purposes. A wider leaf and one which is more nearly round would yield many more wrappers to the pound and would be proportionately more valuable. In addition to being wide, with well-rounded tips, the best wrap- per leaves must have small, fine veins which are widely separated and which form an obtuse or right angle with the midrib. The veins in narrow leaves extend along down the leaf toward the tip, are coarse, and present a very unattractive appearance when wrapt on cigars, while in the case of wide leaves the veins usually extend out toward the edge of the leaf and are almost perpendicular to the midrib, smaller in size, and wider apart. In some of the strains of 96 IMPROVEMENT OF SHAPE OF LEAVES. Da fl Broadleaf tobacco which have been improved by careful seed selec- tion and breeding the veins are sufficiently wide apart to allow wrappers for cigars of standard size to be cut between the veins. Such wrappers have a very smooth, attractive appearance on cigars, and where they can be cut in this way the waste material from each leaf is exceedingly small. Figure 5 illustrates the superior value of the wide over the narrow form of leaf for cutting wrappers economically. Attention is also called to the character of venation Fic. 5.—Diagrams showing the superiority of broad over narrow leaves for cigar-wrapper purposes: A, broad leaf; wrapper cuts numbered 1, 2, 8, 4,5,6,7,and 8. B, narrow leaf; wrapper cuts numbered la, 2a, 3a, and 4a. Waste indicated by hatching. Not only do the broad, round leaves yield more cigar wrappers, but on account of the venation and other characters they produce wrappers of superior quality. in the two leaves shown. It will be observed that the veins extend out almost directly toward the edge of the leaf from the midrib in the case of the wide leaf (A), while in the case of the narrow leat (B) the veins run upward, and, consequently, when used as wrap- pers injuriously affect the appearance of cigars. The variability of the plants in the field in respect to shape of leaf is found upon close observation to be more striking than the variability in many other characters. The variability of strains of 96 TES LIE OS OO, 28 TOBACCO BREEDING. tobacco grown from the seed of the same variety and under similar conditions as respects shape of leaves is shown in Plate IV. It is very important that this variability be reduced to the minimum, and the writers have found that it is possible to greatly reduce the variability by systematic seed selection. Plants producing leaves which are very long and narrow are frequently found growing be- side others with well-rounded leaves of a desirable length. Plants bearing leaves of the ideal cigar-wrapper shape and those that were totally worthless on account of their shape, as well as many grada- tions between these extremes, have been found growing side by side. In the case of a field of Florida tobacco grown from freshly im- ported Sumatra seed a similar variability was observed in the sum- mer of 1905. The production of leaves of undesirable shape results in a direct loss to the grower and manufacturer alike. The cost of production to the grower is no greater where the leaves are all of uniform size and shape, and the cost of grading is greatly reduced. In nearly all varieties of cigar-wrapper tobacco, most of the leaves are small and narrow near the basal end and this portion of the leaf is seldom wide enough for wrapper purposes. Figure 5, A, shows a leaf wide at both ends, from which wrappers may be cut down to the extreme basal end, thus avoiding the waste which can not be avoided in the form of leaf shown in figure 5, B. It is im- portant to produce the form of leaf shown in figure 5, A, not only because it will yield more wrappers to the pound and _ necessitate very much less waste in cutting, but because more wrappers to the acre may be obtained. The variability in the shape of leaves on the same plant is often very marked, and may be as readily corrected by proper methods of seed selection as the variation among the individual plants in the field. The size and shape of the individual leaves on almost all plants vary more or less, but on some much more than others. An occasional plant will be found on which the leaves are compara- tively the same size and shape from the top to the bottom of the plant, while in other cases there are marked differences in this respect. Where this uniformity is found the top leaves are seldom as thick or heavy as where there is a lack of such uniformity, and therefore a larger percentage of the leaves is adapted for cigar- wrapper purposes. ; By selecting for seed the plants possessing leaves of the most desirable shape from top to bottom and protecting them from cross- fertilization, it is possible to produce a crop which will be uniformly hke the parent plant. It can be plainly seen that this will result in a larger yield to the acre of much more valuable tobacco because of the uniformly well-shaped leaves, best adapted to cigar-wrapper 96 i MODIFICATION OF SIZE OF LEAVES. 29 purposes, and this may be accomplished with no additional expense to the grower. The shape of the leaf in certain types of smoking tobacco largely determines its adaptability to both soil and market conditions. In Maryland it has been found generally true that a broad leaf gives best results on very light sandy soil, and is best adapted to the demands of the German market, while a somewhat narrower or longer leaf is more desirable on heavy clay soil; from this latter type the highest grades of red tobacco are produced to meet the requirements of the markets of France. In the case of Maryland tobaccos the shape of leaf js usually correlated with the number of leaves, there being more leaves to the plant where they are narrow than in cases where they are broad. The writers have observed many exceptions to this rule and have found that by keeping this point in mind when selecting seed plants it 1s possible to find round-leaved plants producing a large number of leaves, and to procure varieties from these plants which will produce a large number of uniform leaves and at the same time a grade of tobacco which will be adapted to the demands of the market for which it 1s grown. It is within the power of the tobacco grower to produce the shape of leaf best adapted to the purposes for which his tobacco is grown, and to continually improve the shape and gradually bring it up to the ideal of a perfectly shaped leaf, by carefully selecting seed plants year after year which produce leaves most nearly approach- ing this ideal. In all cases if uniform types are to be produced cross-pollination must be prevented, in order that the progeny the following year may inherit only the characters of the desirabie parent plants. THE MODIFICATION OF THE SIZE OF LEAVES. The modification and control of the size of tobacco leaves is of almost equal importance to the improvement of their shape, and the size is so intimately associated with the shape that both features ean be dealt with along the same lines in the improvement of tobacco by breeding and seed selection. The purpose for which the crop is grown must always determine the most desirable shape and size of the leaf, and the individual grower must decide for himself what size and shape will be best adapted to the local market demands. The importance of producing a comparatively definite and uniform size of leaf is well recognized by the manufacturers of all classes of tobacco. In the case of the Maryland Smoking tobacco grown for export purposes it is difficult to procure a leaf which is too large for the highest market demands, - especially when it is grown for the French market. In most cases 96 30 TOBACCO BREEDING. the leaves are too small. This defect may be remedied to a con- siderable extent by selecting for seed those plants which have the largest leaves, and at the same time this will result in a material increase in the yield of the crop. When grown for plug wrappers, the size of the leaf is as important and worthy of as careful con- sideration on the part of the grower as when grown for cigar wrap- pers. In the manufacture of certain brands of plug tobacco the entire side of the leaf is used for one wrapper, which method is often preferable to using large leaves which have to be cut into two or more wrappers. Where this system is followed, leaves are de- manded which are comparatively narrow and of sufficient length to cover the standard size of plug with the least possible waste. A eareful study of the size of wrapper desired by the manufac- turer will give the grower a very definite idea of the most desirable size of leaf to produce, and by selecting plants having this style of leaf for seed the grower is enabled to produce uniformly the type of tobacco which will be best adapted to his market conditions. In cigar-wrapper varieties of tobacco the size is of as much impor- tance as the shape of the leaf. A short, wide leaf is always the kind most in demand, and has the advantage of being much less suscepti- ble to injury in the curing barn. The. manufacturers of certain brands of cigars prefer to cut only two wrappers from each leaf, and for this reason demand a very small, round leaf. Most man- ufacturers prefer a leaf sufficiently large for cutting two or more wrappers from each side, for the reason that nearly all classes of clgar-wrapper tobacco may be used more economically in this way. The size of Sumatra leaf most desired at present is about 16 inches long and sufficiently wide to admit of the most economical cutting. When leaves become very much larger than this there is danger of coarse venation, altho this can be very largely controlled by select- ing for seed only those plants which produce leaves that have small, fine veins. The question of venation is very intimately associated with both size and shape of leaf, and a certain correlation seems to exist between these characters. The writers have been able to produce types hav- ing leaves of desirable shape and size in which the venation is fine and in every way desirable. The experiments that have been con- ducted with this end in view prove beyond a doubt that these im- portant characters may be successfully correlated and largely con- trolled by methods of selection and saving seed. In curing tobacco in the barn the size of the leaf has been found to be an important factor. This has been clearly shown in the breeding experiments in the Connecticut Valley during damp curing seasons. One of the objects sought in Connecticut has been to secure a shorter 96 MODIFICATION OF SIZE OF LEAVES. 31 and rounder leaf than is now being produced in the Havana Seed and Broadleaf varieties. Numerous crosses have been made and hybrids produced with this end in view, and considerable progress has been made in securing a rounder leaf in the native varieties by careful seed selection: It has been invariably observed that these round- leaved varieties and strains have suffered much less injury from pole-sweat than the old standard varieties. This difference is attrib- uted to the fact that in the case of the old long-leaf varieties the leaves after wilting hang down close around the stalk and adhere to one another, thus preventing the proper circulation of air when it is most needed for successful curing; while in the shorter, round- leaf types, the leaves stand out from the stalk, do not adhere closely together, and admit air freely to all the leaves on the plant, thus preventing in a large measure the injurious effects of pole-sweat or house-burn. The importance of the size of leaf from this standpoint can not be too strongly emphasized. The loss in the Connecticut ~ Valley, as well as in many other sections of the country, due to pole- sweat often takes away the profit of the crop and is keenly felt by tobacco growers. The best crops are occasionally totally ruined by pole-sweat after they have been grown successfully and put into the barn in good shape. Therefore it can be plainly seen that the pro- duction of shorter, rounder leaved varieties in sections of the country where pole-sweat is disastrous will result in great profit to the tobacco growers and packers. Crops which have been badly injured in the barn are a source of endless trouble, and are very expensive to sort and pack successfully. For cigar-filler purposes a comparatively small, short, and thick leaf is demanded. The small leaves are usually thicker and have better body and a very much better aroma and flavor than large, thin leaves. It has been definitely demonstrated from observations made by experimenters on the island of Cuba and from the observations of the writers made in certain filler districts of the United States that the best and most aromatic fillers are always obtained from plants producing comparatively small leaves. Plants which in a way seem to represent dwarf types or strains almost invariably produce leaves which have a much higher aromatie flavor than can be obtained from plants of the same variety producing larger and finer leaves. In an attempt to improve the aroma of some of our domestic filler types thru breeding and selection the Department of Agriculture is endeavoring to produce new types of Cuban tobacco with very small leaves, with the belief that such types will have a superior aroma and will excel the filler grades which are now being grown in this coun- try. These experiments have not advanced far enough to admit of any very definite conclusions, but they have indicated very clearly 96 32 TOBACCO BREEDING. that it is possible to produce better fillers by originating and per- petuating small-leaved varieties of tobacco. The yield from such types has been comparatively small, but by setting the plants closer together it is believed that there will be very little decrease in the yield to the acre in the production of small leaves uniformly thruout the crop. The great variation in the sizeof leaf which is found in nearly all tobacco fields makes 1t possible to breed up and fix varieties which will produce uniformly the size of leaf most desired to meet special market demands. Plants producing small leaves are found growing along with those producing large leaves when all are, as far as we know, of exactly the same variety and grow under equal and uniform conditions. This variation is undoubtedly due to promiscuous acci- dental cross-pollination which has taken place in preceding genera- tions. This variation in size as in shape of leaf also occurs much more strikingly on some individual plants than on others. Plants may be found in all tobacco fields with leaves of comparatively the same size and shape from the top to the bottom of the stalk, while in the majority of instances they are much smaller near the base and top than the middle of the stalk. By selecting seed plants that pro- duce leaves which are uniformly of the desired size and shape from top to base of plant and by covering the flower heads with light paper bags, leaves very uniform in this respect may be grown the following year from seed saved in this manner. Any fonacee grower will recognize immediately the advantages to be gained by producing types of tobacco in which the leaves on all of the plants are uniform in size and shape and where the leaves on the individual plants are likewise uniform in this respect from the top to the bottom of the plant. The yield of the crop will be mate- rially increased, as will the value of the tobacco, while the cured product from such fields will be much more uniform in the packing house and the cost of handling proportionately reduced. The writers have already secured striking uniformity in some of the best strains of cigar and smoking tobaccos grown from seed which they have se- lected carefully and systematically for three years, and have found a considerable increase in the yield and value of the crop grown from such varieties. A recognition of the importance of producing this uniformity is emphasized by the great number of demands made upon the writers for seed of these improved strains. It is easily within the power of tobacco growers to improve their present strains of tobacco in the shape and size of leaf, as well as in other characters, by selecting for seed the plants which are most nearly perfect in these respects and by saving the seed under bag according to the methods outlined 96 CONTROL OF NUMBER OF LEAVES. 33 in this bulletin, in this way preventing intermixture with undesirable strains by accidental cross-pollination. THE CONTROL OF THE NUMBER OF LEAVES ON INDIVIDUAL PLANTS. The variation in the number of leaves borne by individual plants is just as marked as the variation in size and shape, but the size and shape are not always correlated with the number of leaves. In a gen- eral way it has been the observation of the writers that in cigar- wrapper tobaccos the plants which produce the best-shaped leaves usually produce more than the average number of leaves. The varia- tion in the number of leaves on individual plants grown in the same field may be almost invariably attributed to the lack of systematic seed selection, to crossing, and to the use of a large proportion of light, weak seed in planting. The variation may be correlated with the height of the plants or the length of the internodes, or both. Different strains of the same variety are extremely variable in respect to the number of leaves produced, and until pure strains are devel- oped no very great degree of uniformity in the number of leaves borne by individual plants in the crops may be expected. The pro- duction of strains true to type and uniform in the number of leaves, as well as other characters, is made possible by the careful selection of seed. The control of the number of leaves is directly associated with the yield of the crop, and bears a very important relation to success in the handling and curing processes. The possibility of procuring a larger number of desirable leaves on each plant thru careful selection of seed is no longer doubtful, as is clearly borne out by experiments in tobacco breeding. An increase in the production of leaves borne by individual plants has been effected without any increase in the height of the plants and with no detriment to the quality of the tobacco. The reduction in the height of the plants is especially important in Sumatra tobacco grown under shade. It is difficult to prime or pick the top leaves from plants over 7 or 8 feet high, and it would not be advisable for the grower to produce plants which must be topt above that height. The most convenient height for a tobacco shade is about 9 feet. A tent higher than this would be difficult to build, and would be more lable to damage from severe windstorms; hence the necessity for keeping plants below this height by growing more leaves on each plant or by producing plants bearing shorter inter- nodes. The Sumatra and Cuban varieties have a tall habit of growth, with long internodes, but respond readily to methods of breeding in the production of shorter stalks and shorter distances between the leaves. 15507—No. 96—O7 Mm 3 34 TOBACCO BREEDING. In all the varieties of tobacco which the writers have improved by seed selection and breeding the internodes are short: and the number of leaves proportionally greater in the improved strains. In a care- ful count of the number of leaves to the plant in a good field of. Sumatra tobacco the average was found to be between 19 and 20, while the records made in the breeding plats of strains of tobacco originally grown from the same seed as the general field where these leaves were counted show that the number of leaves was increased by two years’ selection to an average of between 23 and 24. The breed- ing plats and the general field were grown under exactly the same conditions in order to eliminate any influences outside of the results of careful seed selection for the production of a greater number of leaves. ‘The leaves produced on the plants giving an increased num- ber were equal in size and more desirable in shape than those from stalks producing a smaller number. The increase in yield due to the production of a greater number of leaves on individual plants and to shortening the internodes may be secured by systematic seed selection with no additional cost to the grower. Aside from the increase in yield, the quality of the leaf when there is a large number of leaves borne by the stalks is usually better than when the stalks produce but few leaves. This is particularly true in cigar and high-grade smoking’ tobaccos. None of the im- proved types of Sumatra tobacco have leaves sufficiently close together to cause any deterioration in the quality or texture of the leaf during the curing process. In the case of certain types of export and plug tobaccos and in some of the northern-grown cigar-tobacco varieties an increase in number of leaves is not desirable, for the reason that it is conducive to pole-sweat when the crop is being cured. It is further true in the case of these varieties that if the number of leaves is in- creased without shortening the internode, the plants will become too tall for expeditious handling. Therefore, it 1s necessary in certain varieties of tobacco to keep the leaves down to a certain definite num- ber, with a desirable length of internode. 7 It is entirely possible for the grower to control largely the number of leaves by careful seed selection and in this way produce uniformly the plants which give the number desired. Care must always be taken in selecting for a large number of leaves not to increase the number at the expense of leaf uniformity. Only plants having leaves uni- form in size and shape should be selected for seed purposes, and this selection must be kept up with unremittent persistency from year to year in order to hold constant the characters of improved strains of tobacco after they have been produced. A large number of leaves to the plant is almost invariably closely correlated with a much lessened tendency to sucker and with de- 96 PRODUCTION OF NONSUCKERING TYPES. 35 creased seed production. The plant food in such cases goes to the leaves, where it is most needed, and not into the production of suckers and of seed, which would be a loss to the grower. A large growth of leaf greatly retards the growth of suckers, and in some instances types have been produced which were comparatively suckerless—i. e., types which produced only very few and small suckers. These types are desirable not only from the standpoint of an increased leaf produc- tion, but the expense of suckering is in a-_large measure eliminated. The habit of growth of the leaves, whether erect, or at right angles to the stalk, or drooping, greatly influences the number of “sand” or ground leaves obtained from the crop. When the leaves are drooping or pendent on the stalk the tips of a number of the lower ones come in contact with the ground and are often covered with sand or beaten and bruised by heavy rains, and are therefore partially or totally damaged. This loss of the lower leaves of the stalk can be very largely overcome by carefully selecting for seed those plants on which the leaves have an upright or erect habit of growth. This very important point in the habit of growth of the plant is often over- lookt, but can be easily controlled by systematic selection. In view of the fact that the sand leaves are not nearly so valuable as those which have not been injured in this way, it is highly desirable that this habit of growth of the plant be kept constantly in mind when selecting plants for seed purposes. The number of leaves on plants of a drooping habit of growth is sometimes greater than where the leaves grow erect or in an upright position, but where a large number of the lower leaves are badly damaged a larger number of the best grade of wrappers may be ob- tained from plants producing a somewhat smaller number of leaves, but all erect. Individual plants producing a large number of the desirable erect leaves may be found, however, and such plants should be saved for seed under bag in order to propagate the strain the fol- lowing year. THE PRODUCTION OF NONSUCKERING TYPES. The number and size of suckers borne by individual tobacco plants are subject to considerable variation. In making selections of seed plants in many tobacco fields the writers have found plants bearing from 8 to 12 large suckers, and in the same fields other plants pro- ducing only one or two small suckers. In Plate V are shown two plants growing side by side in the field, at about the same stage of maturity, one of which bore five large suckers, while the neighbor- ing plant bore only one small sucker. Instances of this kind are common in most tobacco fields. As can be seen in the illustration the nonsuckering plant has a larger number of more rounded leaves 96 36 TOBACCO BREEDING. than the suckering plant, which condition is usually true in all such- cases. . | | The production of many large suckers is usually correlated with the development of few, heavy, dark, and usually narrow, pointed leaves. This condition is explained on the ground that the large sucker branches take from the plants the elements of plant food which otherwise would be utilized in the development of many broad, round leaves. The possibility of securing nonsuckering types of tobacco was suggested in the course of a series of experiments in the improvement of cigar-wrapper tobaccos. In the selection of seed plants great care was exercised to pick out those bearing the largest number of rounded leaves with fine veins. In the course of the study of the progeny of these plants it was observed that few suckers were produced by the most desirable types of plants. The continued observations on this subject have confirmed the conclusions that there is a correlation between the number, shape, and character of the leaves borne by individual plants and the number and size of suckers produced by these plants. The number and size of the suckers produced by the plants in all tobaccos is an important practical problem from several standpoints. Owing to the dwarfing and otherwise injurious effect of the suckers it is necessary to remove them by breaking them off, or to “* sucker ” the plants, as the process is commonly called. There is great danger of breaking, tearing, or injuring the leaves during the suckering srocess, and this causes much loss in cigar-wrapper varieties, as the injured cigar-wrapper leaf is rendered practically worthless. A care- less laborer frequently causes great loss to the grower during the process of suckering the plants. Owing to the fact that the suckers do not develop on all of the plants at the same time and consequently can not all be removed at once, it is necessary to go over the field several times during the season in order to remove all of them. The cost of suckering is one of the important items of expense in the cultivation of tobacco. Therefore the production of nonsuckering types is an economic problem of great importance, not only by reason of the reduction in the cost of growing the crop, but from the fact that the nonsuckering types usually produce a larger yield of a more desirable quality of tobacco than the suckering types. It has been found possible to produce uniform strains of different varieties of tobacco having but few and small suckers by saving the seed from nonsuckering plants under bag. As an illustration of the possibility of the growers producing such types the experience of one of the writers in the improvement of the Connecticut Broadleaf tobacco may be cited. In these experiments desirable plants were selected for seed in 1903, producing round leaves of fine, silky texture 96 PRODUCTION OF EARLY VARIETIES. 37 and few suckers. ‘The crops raised from this seed were found to produce but few suckers, the progeny of the different plants varying somewhat in this respect. From the strains producing the best type of leaves and bearing the least number and smallest size of suckers nonsuckering plants were again selected and the seed saved under bag in 1904. In the season of 1905 it was found that the progeny of these selections were almost .free from large suckers. In one strain in particular only a few very small suckers, none of which grew more than 4 inches in length, were produced. The plants raised from ordinary seed of the same variety in the same field produced many large suckers, and as usual it was necessary to sucker the crop several times during the season. The remarkable difference in the sucker- ing and nonsuckering habit has become so well fixt in this particular strain that a limited distribution of the seed was made for testing during the season of 1906. It has been suggested that by saving seed from sucker branches strains of tobacco are developed which produce an increasingly large proportion of suckers; in other words, that sucker seed tends to produce suckering types of tobacco. In experiments with plants raised from seed saved from the central flower cluster the writers have observed little or no difference. As a rule, however, it has been found that the seed pods in the central flower cluster contain more large and heavy seed than the pods borne by the sucker branches, so that where seed is not carefully separated in order to secure only heavy seed for planting it is probably the best practise to save seed from pods borne by the central flower cluster of the seed head. THE PRODUCTION OF EARLY VARIETIES. Early maturing varieties of tobacco are of particular importance to northern tobacco-growing districts. Owing to the fact that frost kills the plants it is necessary for northern farmers to grow varieties which will mature between the time of the last frost in the spring and the first frost in the autumn. After the tobacco crops have been harvested and hung in the barns the curing processes are carried on most favorably during warm weather. The length of time required for the completion of the curing varies with the variety grown, the purpose for which the tobacco is to be used, and the weather. Under normal conditions, however, the natural curing period extends from four to eight weeks. It can readily be seen, therefore, that early-maturing varieties are likely to have more favorable conditions for curing than late varieties, as has proved to be the case in the experience of the tobacco growers in northern districts. 96 38 TOBACCO BREEDING. Another fact of importance in this regard is the likelihood of late-maturing varieties being injured in the field by autumnal storms. The earlier the crops can be harvested, the less is the probability of injury by severe rain, wind, or hail storms. In one district of the Connecticut Valley in the season of 1905 a severe hailstorm at about the usual time of harvest completely destroyed all except the early-maturing tobacco, which had been harvested and hung in the curing sheds. This experience is common to other northern tobacco regions and emphasizes the value of early-maturing varieties. The uniformity in time of maturing of the individual plants in the fields is an important practical matter. In those districts where the tobacco crop is harvested by cutting off the plants near the ground all of the plants in a given section of the field must be cut off at one time. The immature plants can not be left to ripen and the early-maturing plants can not be harvested before the rest of the plants in the field. Overripe or underripe tobacco is likely to be of poor quality. In cigar-wrapper varieties the overripe leaves lack elasticity, gloss, and strength. ‘The underripe leaves are likely to have uneven color and are susceptible to injury by various fun- gous and bacterial diseases. It is very important, therefore, that the ‘individual plants in the field ripen uniformly, so that they can be harvested at one time without loss or injury. The lack of uniformity in the maturity of leaves borne at the base, middle, and top of the plants is a cause of loss in value of the crop to the growers. As a rule the bottom or so-called “sand” leaves ripen first, the middle leaves next, and the top leaves last. In the varieties of cigar-wrapper or smoking tobaccos, especially where the entire plant is harvested at one time, the overripe sand leaves and the im- mature top leaves on such variable plants are inferior in value to the middle leaves. As stated, the color of these sand and top leaves is usually poor and undesirable, and there is also generally a corre- sponding inferiority in the texture and quality of these leaves. ). The valve (ce) is connected directly with the seed receptacle (e). The seed receptacle consists of a 1-inch glass tube (¢) about 14 inches in length, cemented in the reducer (d) with plater of Paris. At the extreme bottom of the glass tube e and just above the top of the valve (c) a fine wire gauze is fastened. The object of this gauze is to prevent the seed from falling into the valve trom the receptacle, and there- fore it is necessary to use a wire gauze with very small mesh. An ordinary gas pipe coupling (7), about 3 inches in length and slightly larger than the tube e, is firmly cemented to the top of the tube to serve as a support for the tube g. The tube g is of glass the size of tube e and about 6 feet in length. The apparatus can be sup- ported by a convenient frame, which may be fastened to the wall or set up wherever desired. The bellows and tubing for this apparatus may be procured from any chemical supply house, and the remaining parts from hardware stores. They can be easily put together and the apparatus set up in the proper manner by anyone who wishes to use it. A complete device of this nature should not cost more than $5, a very small sum compared with the benefits to be derived from getting rid of the light and undesirable seed. In the successful operation of this apparatus the following method should be employed: Pour about 1 ounce of the tobacco seed to be separated into the seed receptacle, and by means of the foot bellows pass a current of air of sufficient strength thru the entire apparatus. The strength of the current of air may be regulated by the globe valve so as to blow out the desired proportion of the hght seed. The light seed is blown out thru the top of the tube and the heavier seed falls back into the seed receptacle. The degree of separation may be controlled accu- rately by means of the valve, the length of the tube, and the work- ing of the foot bellows. A much more complete separation may be made by the use of a long tube than where a short one is used. This simple apparatus serves to completely eliminate the evil re- sults associated with the use of light and inferior seed. It is thoroly practical in every way, and delicate enough in its operation to sepa- rate the smallest kinds of seeds according to their individual speci- 96 60 TOBACCO BREEDING. fic gravity. One apparatus is sufficient to separate seed for an en- tire community, a plan which is being followed in some cases. A pound of seed may be separated in less than half an hour. Thus it is seen that the apparatus and cost of operating are very small and not sufficient to prevent any tobacco grower from eliminating all heht and poorly developed seed, in this way not only increasing the yield, but also improving the uniformity and quality of his crop. DISEASE RESISTANCE. In practically all fields producing diseased tobacco plants where the writers have made observations some degree of immunity has been noticed in individual plants which have been found growing among. badly diseased plants on infected soil. These cases of immunity could not be explained on the ground of any differences in treatment, but their resistance to disease was evidently inherent in the individual plants. The same conditions have been found by other investigators and workers in other farm crops, and from these resistant individuals many unmunhe strains have been developed. Among the most notable are the variety of wilt-resistant cotton, improved by Mr. W. A. Orton and Mr. Rivers, and the Iron cowpea, which is resistant to root-knot caused by nematodes, improved by Dr. H. J. Webber and Mr. W. A. Orton. The transmission of this immunity found in indi- vidual plants has made it possible to develop immune strains, and in that way to produce thoroly healthy crops on disease-infected soils. In most cases where immune plants occur, if seed is saved from a large number of such plants some of them will be found to transmit their resistance to the progeny uniformly and thus give rise to the easiest known method for the control of certain plant diseases. In the case of tobacco, the seed of the immune plants must be saved, with precautions to avoid cross-pollination, to insure the best results. In the season of 1903 the writers made selections of plants in several tobacco fields in the Connecticut Valley which showed immunity to the tobacco wilt. These plants stood out very plainly and strikingly in the diseased sections of the field, making a normal growth, and were apparently not affected by the wilt in any way, while plants growing all around them were so badly diseased that they produced no tobacco, and many of them died before maturing seed. Seed was also saved from some of the diseased plants that reached maturity. Two rows were planted the following year on the infected soil, one from seed of a resistant plant and the other from seed of an immune plant, with the results shown in Plate LX, figure 2. In this instance, by reason of the foregoing and other observations, it was found that complete resistance to the wilt was obtained by one year’s selection. 96 DISEASE RESISTANCE. 61 A wilt in tobacco occurs in North Carolina which is evidently a parallel case with the one found in Connecticut, and in all probability could be controlled in the same way. One of the most serious diseases affecting tobacco at present is the root-knot caused by nematode worms. ‘Tobacco seems to be particu- larly susceptible to the attacks of nematodes, and many crops are more or less seriously affected by this enemy. There is no known remedy for this pest that 1s applicable in a practical way to field con- ditions. Soil-sterilization methods are used successfully in plant beds or in greenhouses, but such treatment is rather too expensive to be used on large fields. The most practicable method for the con- trol of this disease seems to be in the way of securing immune strains of tobacco by seed selection and breeding. The writers have se- lected a large number of individual plants that showed immunity in the field, and the seed of these selections will serve as a basis for ex- periments in the production of nematode-resistant types. The com- plete success of other workers in obtaining resistance to nematodes im varieties of sugar beets and cowpeas is good evidence that similar results may be obtained in their efforts to obtain resistance to this enemy in tobacco. The [ron cowpea shows strong resistance to the nematode when planted on badly infected tobacco fields, and for this reason can be highly recommended to tobacco growers for use in this connection. The mosaic disease causes very serious injury in tobacco fields in many parts of this country. The writers believe, from indications observed during the past two years, that it will be possible to develop strong, vigorous strains of tobacco which will be largely resistant to this disease. In the case of some Maryland selections, resistance to the mosaic disease seems to have been transmitted in a large degree to the progeny of certain vigorous strains. In two plats grown side by side under uniform conditions, one from seed of a parent plant affected with mosaic disease, the other from a perfectly healthy one, the following results were obtained: Plat 1, grown from the seed of the mosaic plant, showed 80 per cent of diseased plants in the field; plat 2, grown from the seed of a perfectly healthy plant, showed less than 20 per cent of diseased plants. It maybe impos- sible to entirely eradicate this disease by the production of immune varieties, owing to the peculiar nature of the malady, but these fig- ures, which have been duplicated many times, show very conclusively that by the development of stronger and hardier types of tobacco, especially where heavy seed is used for sowing, it will be possible to gradually reduce the percentage of mosaic plants in ordinary to- bacco fields. 96 62 TOBACCO BREEDING. There are numerous tobacco diseases which the writers believe may be largely eradicated by producing immune strains. It is the inten- tion of the Office of Plant Breeding Investigations to take up work with as many of these diseases as seems practicable and endeavor by selection to produce resistant types wherever it is possible to do so. A NEW VARIETY PRODUCED BY SEED SELECTION. UNCLE SAM SUMATRA. The original plants from which the variety of tobacco known as Uncle Sam Sumatra has been produced by seed selection were grown under shade on the plantation of the Connecticut Tobacco Corporation, near Tariffville, Conn. The first selections were made in the season of 1903 on this planta- tion in a field the plants of which were grown from seed originally brought from Florida. The Florida seed was produced by plants which were grown from seed originally imported from the island of Sumatra. In a careful study of the Connecticut-grown Su- matra crops in 1903 a number of dis- tinct types were discovered, some of which were evidently very undesira- ble, while others were apparently desirable. , x » 2 2 ee ee —_ samen: — meat net nrg om ay rat\ehy fi GRETA Sy he Met ‘ ser nh Lesage Heb ofh foe Be dap fe Aa het! 5H Bike Mh We” Me Paw gh bebe: he phe th HD ay Mb beter Caper Ray Ae . eh. we ® Wopere make emt Miwa Rebs 45 BARBS alae aa eas Wed is se ee WU ATKT ee ; ‘ top be He raroya CN haa % Gel 2A thet) FR: a Dope pe eer Be ee wpconate ys we seth a . tt mr fore peg et ik ig ) apie ; ( phe ey eth hath bey» iran Peet net * . HB fy ae * ni \ Def We Med toh . ite Bide pe be 4 Bitlet agree me a Hefty: aut ‘se ue By) Nek Mpeeeneber Are Sieve a tren pee. tr ici opi i “Bet Ys: Vag obs 5S a he rin ey ie ? i ote * phate RS Par any ‘EET | en ey he fo i ah iat unite ie wie ie it where hh i hs bss reincarch on if bot habs viet ih he “ah by he Bee gute: His iy wes aecrtate rhe ard ee ae be ie 94 pbsee bee Searien: pita teh aynls cit Vs Avy » a asaets ae . » ie pectititicnt a ait jebehs mat gene bens Me er etibeents boobs bent aie hye saat if bth Nps ior nee bere ithenbovdbr gies Mitt We the in bepsaeay ob. ht Ht rire : je eh ie mesh & sai fa cit rye v6 pepe babs Seta eb beb o Stee ina in ie ‘Goede baAP Se rors ib depen te ” a ep if diy) Lada nies ug ’ Nhe visfege ie Seb eg stetitrtitatat cit ope ieee i be P y sire ia teat isha H Haan Tsesbeas « YY Cah ci rr bd w eb bs ioe es Apri ‘i Dt i mvp earns ges it oh arora fteota BEifc a pe ee Renbeboo- hoses boy “} i beh ee e +i bebe Poca TR 4 in ty fatal Lbreae nse Vey Si satel Dt, fad sees oe phates ee Works te AE hil wh< bel Th tit ie be tobe Bete hehe eb y aie Adhd i Toenseree yah £02 Dh eed fet ee er Perris tt bik wh “y % rf URE Wha de Hes ue hon. rn ane a winwrien, ey} i ia, | a oe ear heb ruth b mbhe bey ie iowtiob-B) dons rs ye ‘a Seb a) theh Pe or ¥ Se ot ep Rene ey ucata bs. *: ” th He Kah. Bae Pg bee Bad ay fete ® Me , ath We bhete Heb. . le hed : ES es LU ’ AAA wan se babe oe Leu ee pow ® hte br La sibel 4 heel ” *. we ryt oy eaty ih ‘antes r Rua rer an w ah sete 8, ry 4 p,B beslehebee megs hoe bP. hebben Re yh b Aedes rot Ortiharty ae Shy hee oh | ahd pets > : mee Wethep es * we Me tpn d a ris bed Rist Mote? ahd of Oe St OP a al le WG esl Me Be sibR. 1) W aera eon Ys Sait Nihon “Eye sy rdetev ot be acy rs . Sache bakas we Hedapes att Bef rats apabee Hen ‘ eg separ nt he sre te osieh, a3 Ly i ae . One rt im fy ash Ny? ah iduadoeh ort vagrant » abst ele eh batbes Hehe epee Ps Gober yy he! Hitias Shek ab phot ot sits Resublatse nee eds te et ieee eke wpe aud AR a pene ® for sey befyee if iver ites Hanes ey hst Hopol ort mie Selb? beh Weretateh. © ta Heh oR ae tse b y yy Ae 1 “h t eet abet Th (aot, 6 a Ae hedd Lea ha tee Dor tbay vie heh feels sts be d he sen ieee Hite why pb hbet seer He akwere it Benen ered eet he TE at ey wi 2 Pac ee esi sae ha bone obo wa pe be fone Dy! ' xo be dieine ts " * a ie eh fo >. Hy Wy sigs apt é -~ vane ee t. (elite te Be b: vb J obe a =) ttt pahone " rag ima erty re aie Bebe) Fp jetyy + batheboheh papa iieh, br det emees ash A ipyshee sh ahr beh ehabe ere tr hee Sea hed iH = nb Dive rear she rs ry ' Maha iy ‘ heb + ine batt abeheh k= bef wir Da the be bene bene) My hope tb Piet peebarblend a pease wht bey Tphaehs wee Hf stag ke ere hoe sae Uoteenen rad Me At ser A omer Ch teak heaet hee he La a ee ~ bruh byt € eb one tr Leper : ettel deb ba baPeiet ip ber rane ber: Primed eiersetes bee Ooi | q ry ohh ini ” tee Weanrand eee babe ehagqe hate f Phe bebe A te Dre be ewe Stabe wat eine are bs wee ti its ie eee Porn| Heb Prise ier Pb Pra a) Neg hone aks eer ans heb 4 “phoh s& Wey othe heh Behe > heb ie bebe ve Her ies BP eben t oe ery a “asaya ayy » pen & vhemintéwnestests oh thes feadent Seothe wih +.) P beb te ehibeh® Ze » Nd igaegts reariars eds , eownverar tc Seed opab eho saat ithe & rs s Pa Phe bei we Aber PO he ler Q8be > ~ poh beneiye 2 yekeciebes teat yi how nee she he (4 . ey ane Pee ey nwt Miteks bet yee 4 EMT deh Hed» Bthebled hep regenes We beets Co Neep Be Ht Wade fiw fer owste frie fet is Ape eth wee Ded SW the h W. Soon obscene titi mh Ga fenebeh phew tes war * gabety ahetepete. enw b Soden Dit i dada ciyhearic a ny Gea TCA NY at" Tatutanchebet, WED Node by phir Cas Hh Fold ped: fp obey! fe hohe > he Medi dh baled eh Boe che » bene “pr demek: he b Pathe UAral were oh Later teh ul oy Ae hotel | ‘ uy be tahabs “bey AL: ofa teh tm Yebom Aebebonee Patan Healy bes ‘ batter fe «i *h ok Seofateeite } ri pf teeeb ie cat ob Bae eh rf Rent: » i oe Harr beh bh ofr foyih) fe We ibe hg Se the adeeb eked bek aheabes be O Pllepabee RE BER: fh = oo Rainy pe ee bbe: : Sabo bed oPob het + at * et i he afm ob hace | rests bah he Liebe 5 th NTs = tafe’ one bat Se wl Tose ahh: Aieebehe . ks Wee iBet Berets shh Deh: be thbe od Phe hee ber eibspeite b bhi pibor ety He pet cine baketent MA } s han spetich te © raat web siedpiat : i i CoB rRe eben der eP bs bepkie he ei bs re bee teehee Darth: le br pret he eee tbe Pa i! oat HHH ob ieeyiebee t brie epets Webb eleat ist J) plied sy ort: * by’ feo ee a * Ab Babetisp obese? i pepsi ke pene pes (heel bad Be irt mba pre b pape bern phe ok bpp p Beer t eho Be Lv rhe Mh, . ret a be ot (rage ps adang oe Lieb prk- Bee hes 2 hen . herprave - “} hey took bo peP om ete beh eg ten batehshobsheires ty ber Hobs Obed se bob b . eabebeten-= Bepserbed ele De- ibe Pe brie benebe Size’, H SESE rhea a bs sorbdbabeb eee , i hehe ¥ eb whee be ye Whe Delve ae pet el hei prben See Sofie Prd bent gbyoe Pee Dated 18 sat bibhet oF wire eh. nahep Be} Ob P be bids! we “b z bbs ber aaa take bepeh lobee Diet or soot ten ' dese Lied Naa d wee % ' rv bh bP Bp bnbbee eat vee Saat ea rete =) * . orb rh We \ fie fede be bart RE tibsh bib Bove babe 6: ” yk ot dd els bod eo) rena: © eheb @ hsb b4