.^^i^XfT^fi^ XT-d^^ JOURNAL OF AGRICULTURAL RESEARCH Volume XIX APRIL I— SEPTEMBER 15, 1920 PUBLISHED BY AUTHORITY OF THE SECRETARY OF AGRICULTURE WITH THE COOPERATION OF THE ASSOCIATION OF LAND-GRANT COLLEGES WASHINGTON, D. C. xT EDITORIAL COMMITTEE OF THE UNITED STATES DEPARTMENT OF AGRICULTURE AND THE ASSOCIATION OF LAND-GRANT COLLEGES FOR THE DEPARTMENT KARL F. KELLERMAN, Chairman Physiologist and Associate Chief, Bureau of Plant Industry EDWIN W. ALLEN Chief, Office of Experiment Stations CHARLES L. MARLATT Entomologist and Assistant Chief, Bureau of Entomology FOR THE ASSOCIATION J. G. LIPMAN Dean, State College of Agriculture, and Director, New Jersey Agricultural Experi- ment Station, Rutgers College W. A. RILEY Entomologist and Chief, Division of Ento- mology and Economic Zoology, Agricul- tural Experiment Statton of the University of Minnesota R. L. WATTS Dean, School of Agriculture, and Director, Agricultural Experiment Station, The Pennsylvania State College All correspondence regarding articles from the Department of Agriculture should be addressed to Karl F. Kellerman, Journal of Agricultural Research, Washington, D. C. All correspondence regarding articles from State Experiment Stations should be addressed to J. G. Lipman, New Jersey Agricultural Experiment Station, New Bninswick, N. J. n w CONTENTS Page A Teosinte-Maize Hybrid. G. N. Collins and J. H. KbmpTon. . . i Banana Root-Borer. G. F. Moznette 39 Effect of Calcium Sulphate on the Solubility of Soils. M. M. McCooL and C. E. Millar 47 Further Studies on the Influence of Humidity upon the Strength and Elasticity of Wool Fiber. J. I. Hardy 55 Composition and Density of the Native Vegetation in the Vicinity of the Northern Great Plains Field Station. J. T. Sarvis 63 Effect of Reaction of Solution on Germination of Seed and on Growth of Seedlings. Robert M. Salter and T. C. McIl- VAINE 73 Philippine Downy Mildew of Maize. William H. Weston, Jr . . 97 Effect of Drugs on Milk and Fat Production. Frank A. Hays and Merton G. Thomas 123 Artificial and Insect Transmission of Sugar-Cane Mosaic. E. W. BrandES 131 Halo-Blight of Oats. Charlotte Elliott 139 Influence of Fermentation on the Starch Content of Experimental Silage. Arthur W. Dox and Lester Yoder 173 Effect of Premature Freezing on Composition of Wheat. M. J. Blish 181 Behavior of the Citrus-Canker Organism in Soil. H. Atherton Lee 189 Decline of Pseudomonas citri in the Soil. H. R. Fulton 207 Variation of Individual Pigs in Economy of Gain. R. C. Ashby and A. W. Malcomson 225 Production of Conidia in Gibberella saubinetii. James G. Dick- son and Helen Johann 235 Effect of Manure-Sulphur Composts upon the Availability of the Potassium of Greensand. A. G. McCall and A. M. Smith .... 239 Rust in Seed Wheat and Its Relation to Seedling Infection. Charles W. Hungerford 257 Practical Universality of Field Heterogeneity as a Factor Influ- encing Plot Yields. J. Arthur Harris 279 Transmission of the Mosaic Disease of Irish Potatoes. E. S. Schultz and Donald Folsom 315 Relative Susceptibility to Citrus-Canker of Different Species and Hybrids of the Genus Citrus, Including the Wild Relatives. George L. Peltier and William J. Frederich 339 (ni) IV Journal of Agricultural Research voi. xex Page Presoak Method of Seed Treatment: A Means of Preventing Seed Injury Due to Chemical Disinfectants and of Increasing Ger- micidal Efficiency. Harry Braun 363 Daily Development of Kernels of Hannchen Barley from Flower- ing to Maturity at Aberdeen, Idaho. Harry V. Harlan 393 Development of Barley Kernels in Normal and Clipped Spikes and the Limitations of Awnless and Hooded Varieties. Harry V. Harlan and Stephen Anthony 43 1 Investigations in the Ripening and Storage of Bartlett Pears. J. R. Magness 473 Further Data on the Orange Rusts of Rubus. L. O. Kunkel- • • • 501 Germ-Free Filtrates as Antigens in the Complement-Fixation Test. William S. Gochenour 513 Mosaic Disease of Corn. E. W. Brandes 517 Genetics of Rust Resistance in Crosses of Varieties of Triticum vulgare with Varieties of T. durum and T. dicoccum. H. K. Hayes, John H. Parker, and Carl Kurtzweil 523 Line-Selection Work with Potatoes. O. B. Whipple 543 Occurrence of the Fixed Intermediate, Hordeum intermedium haxtoni, in Crosses between H. vulgare pallidum and H. dis- tichon palmella. Harry V. Harlan and H. K. Hayes 575 Index 593 ERRATA AND AUTHORS' EMENDATIONS Page 97, line 20, "were" should read "would be. " Page no, Table I, column i, line 13, "24" should read "23." Page 136, line 2, "Pseudococcus boninensis" should read " Pseudococcus honinsis." Page 202, following line 16, add " C. trifoliaia seedlings, however, on testing are readily susceptible to the disease." Page 232, Table XVI, footnote, ">•=— 0.452 ±0.068" should read "r=— o.i66±o.o83." Page 236, line r, "The spores, both conidia and ascospores, behaved alike in germination" should read "The conidia from both conidia and ascospores behaved alike in germination." Page 242, line 22, "Al2(SO<)3 0.18 H2O" should read "Al2(S04)3 • 18 H2O" and "FeS04 0.7 HsO" should read "FeSOi . 7 H2O. " The same correction should be made in Tables III, IV, V, and VII. Page 278, Plate 47 and legend. "Internal hilar sorus shown at i" should be omitted. The marking x is incorrectly placed on the plate. There are two internal hilar sori shown in the lower right-hand portion of the plate. Page 362, legend for Plate 59, line 10, "Infected leaves from twigs" should read "Infected leaves and twigs." IIvIvUSTRATIONS A Teosinte-MaizE Hybrid Text Figures Page 1. Height: frequency distribution of plants in Fo 13 2. Total leaves: frequency distribution of plants in F2 13 3. Height of sucker: frequency distribution of plants in Fj 14 4. Sucker index : frequency distribution of plants in F, 14 5. Circumference index: frequency distribution of plants in Fo 16 6. Nodes without branches: frequency distribution of plants in Fj 15 7. Nodes above: frequency distribution of plants in Fj 17 8. Nodes above on third: frequency distribution of plants in F2 18 9. Nodes on third: frequency distribution of plants in Fj 18 10. Primarj^ branches: frequency distribution of plants in F2 19 11. Secondary branches: frequency distribution of plants in Fo 19 12. Secondary index: frequency distribution of plants in F2 20 13. Tassel branches on third: frequency distribution of plants in F3 20 14. Male branch index: frequency distribution of plants in F2 21 15. Male secondaries: frequency distribution of plants in Fj 22 16. Double male alicoles: frequency distribution of plants in F2 23 17. Mixed alicoles: frequency distribution of plants in Fo 23 18. Single female alicoles: frequency distribution of plants in F2 24 19. Double female alicoles: frequency distribution of plants in F2 24 20. Alicole index: frequency distribution of plants in Fo 25 21. Nodes silking on third: frequency distribution of plants in F2 25 22. Nodes silking index: frequency distribution of plants in F2 26 23. Prophyllary spikes: frequency distribution of plants in Fg 26 24. Length of prophyllary: frequency distribution of plants in F2 27 25. Rows in central spike: frequency distribution of plants in Fo 27 26. Rows of alicoles: frequency distribution of plants in Fj 28 27. Position of best spike: frequency distribution of plants in Fj 29 28. Number of alicoles: frequency distribution of plants in Fj 29 29. Number of suckers: frequency distribution of plants in F2 30 30. Branch silking first: frequency distribution of plants in F2 31 31. Days to pollen: frequency distribution of plants in F2 31 32. Pollen to silk: frequency distribution of plants in F2 32 ^2, • Length of intemode on third : frequency distribution of plants in F2 32 Plates 1. A. — General view of F2 plants of teosinte-maize hybrid. B. — Fj plants of teosinte-maize hybrid, showing diversity in size and season 38 2. Teosinte-maize hybrid: A. — F2 plant No. 36. B. — F2 plant No. 49 38 3. Teosinte-maize hybrid: Pistillate inflorescence of Fg plant No. 36, shown in Plate 2, A 38 4. Pistillate inflorescence of plant No. 49, shown in Plate 2 , B 38 5. Pistillate inflorescences from plant No. 94, illustrating an intermediate type of inflorescence 3^ V VI Journal of Agricultural Research voi.xix Page 6. Teosinte-maize hybrid: A. — F, plant No. 31, showing compact growth characteristic of many of the plants. B. — F2 plant No. 113, showing stiff, erect leaves. C. — Fj plant, grown at Lanham, Md 38 7. Teoshite-maize hybrid: Pistillate inflorescences of the Fj plant shown in plant 6, C 38 Banana Root-Borer Text Figure I. Cosmopolites sordidus: Section of sheath with egg m ji7m in compartment . 44 Plates 8. Banana root-borer (Cosmopolites sordidiis): Adult 46 9. Egg and larva of banana root-borer: A. — Egg. B. — Larva, side view. C. — Head of larva, dorsal view. D. — Head of larva, face view. E. — Head of larva, side view. F. — Dorsal view of seventh, eighth, and ninth abdominal segments. G. — Posterior view of segments 7 to 10. . . . 46 10. Pupa and adult of banana root-borer: A. — Ventral view of pupa. B. — Lateral view of head and thorax of pupa. C. — Dorsal view of pupa ... . 46 11. A. — Yoimg, healthy banana plant bulb with lateral roots. B. — Young banana plant cut into, showing work of the larvae of the banana root- borer 46 Further Studies on the Influence op Humidity upon the Strength AND Elasticity of Wool Fiber Text Figures 1. Graphs showing the effect of humidity upon the breaking strength of wool fiber 56 2. Graphs showing the effect of humidity upon the elasticity of wool fiber. . . 57 3. Graphs showing the relation between the diameter and the tensile strength of wool fiber 60 4. Graphs showing the effect of humidity upon the breaking strength and elasticity of wool fiber 62 Composition and Density of the Native Vegetation in the Vicinity op THE Northern Great Plains Field Station Text Figures 1. Diagram of grass mat: A, from side; B, from above, a, Basal cover; b, foliage cover 68 2. Meter quadrat in 30-acre pasture mapped in detail in 1915 6g Plates 12. General view of native vegetation near Mandan, N. Dak., showing com- position and density 72 13. A. — View across area of Andropogon furcatus. B. — Close view of Aristida longiseta bunches 72 14. A. — Close view, from above, of meter quadrat in 30-acre pasture. B. — Meter quadrat in 30-acre pasture 72 Apr, i-sept. IS, 1921 Illustratiofis VII EifPEcT OF Reaction of Solution on Germination of Seeds and on Growth op Seedlings Text Figures Page 1. A, graph showing the relation of reaction to the contents of HjCgHjOy and NaOH employed in the cultures of series A; B, graph showing the change of reaction found after 4 days' growth of wheat seedlings in series A. . . 76 2. A, graph showing the change in reaction obtained by electrometric titration in series B; B, graph showing the change of reaction found after 4 days' growth of wheat seedlings in cultiu-es of series B 77 3. Graph showing the relation of growth of wheat seedlings to reaction in series A 79 4. Graphs showing the relation of growth of wheat, soybean, com, and alfalfa seedlings to reaction in series B 81 5. Graphs showing the relative weights of sprouts produced by seeds of wheat, com, soybeans, alfalfa, and red clover in 7-day germination period at various reactions 89 Plate 15. A. — Method of growing wheat seedlings. B. — Appearance of wheat seed- lings in series A at time of harvesting. C. — Appearance of wheat seedlings in series B at time of harvesting 96 Philippine Downy Mildew op Maize Text Figures X. Graph showing variation in length of conidia of Sclerospora philippinensis . . 11 1 2. Graph showing variation in diameter of conidia of Sclerospora philippinensis . 11 1 3. Diagram showing ratios of length to vsddth of conidia of Sclerospora philippi- nensis, arranged in classes and indicating limits of variation and mode. . 11 1 Plates A. Young maize plant, showing the effects of a very early attack of the downy mildew on a small, early maturing variety, Manobo Yellow 122 B. Young maize plants, showing the effects of later attack of the downy mil- dew on a large, late-maturing variety, Guam White Dent 122 16. A. — Portion of a field of Moro White maize, showing heavy loss from the downy mildew. B. — View near the edge of a field of Guam White Dent maize, showing the ravages of the downy mildew 122 17. A. — A frequently encountered type of downy mildew effect. B. — A serious case of downy mildew injury, one of hundreds in a badly attacked maize field 122 18. A. — A common result of downy mildew attack. B. — A maize plant seri- ously injiu-ed by the downy mildew stands in front. C. — One hill in a maize plot which lost heavily from attacks by the downy mildew 122 19. A. — A case of abnormal growth of a maize plant as the result of an attack by downy mildew. B.— A maize plant which, when nearly mature, became infected by the downy mildew through a small sucker previously devel- oped 122 20. A. — A deformed and partly sterile ear complex produced by a maize plant as a result of downy mildew infection. B. — A maize ear developed abnormally as a result of the downy mildew. C. — Ear of a maize plant infected by the downy mildew 122 VIII Journal of Agricultural Research voi. xix Page 21. A. — A near view of the thick down of conidiophores which has been pro- duced on the upper stirface of a badly diseased maize leaf. B. — Upper surface of a badly infected maize leaf from a maturing plant. C. — Upper surface of the middle portion of a maize leaf from a very young plant which has only recently developed the markings of the disease 122 22. A. — View of a row of Egyptian sorghum showing tall, green, healthy plants at tlie left and at the right a dwarfed, yellowish white plant whitch is in- fected by the downy mildew. B. — Near view of this diseased sorghum plant. C. — A comparative view of healthy teosinte (right), and teosinte seriously infected with the downy mildew (left) 122 23. A. — Portion of the typical crooked, irregular mycelum with nximerous haustoria which is found in the mesophyll of badly infected leaves, here freed from the host tissue by maceration. B.— Longitudinal section cut from the center of a maize stem 8 inches from the ground. C. — Portion of the mycelium freed by maceration from tissue of the midrib at the base of a badly infected leaf. D. — Hypha cut in cross section as it lies be- tween three adjacent mesophyll cells of the host. E. — Transverse sec- tion from a badly infected portion of a maize leaf, sho-wing the abundant mycelium running between the cells of the bundle sheath and forming in the substomatal air chamber the branches (a) that grow out through the stoma to form the conidiophores. F. — Portion of a hypha lying between adjacent mesophyll cells, one of which has formed a many- layered wall around the haustorium invading it. G. — Portion of a hypha similar to that shown in F but with the haustorium unhindered in its invasion of the host cell. H. — Bit of mycelium such as is shown in A but more highly magnified to show the haustoria 122 24. A. — Slender, sparingly branched conidiophore bearing comparatively few conidia. B. — Tip of branch with two conidia in situ. C. — Stout, much- branched, mature conidiophore bearing 38 spores. D. — Upper portion of a nearly mature conidiophore with one secondary branch which has failed to branch further and has terminated in a single conidium only. E. — Small, stunted, sparingly branched conidiophore produced on maize during the light dew of the hot, dry season. F. — Basal cell with two thick crosswalls; from maize. G. — An unusual basal cell with two septa and an abnormally large footlike base. H, J, L. — Typical basal cells of conidiophores. I. — Upper portion of an underdeveloped coni- diophore bearing three spores on sterigmata arising directly from the top of the main axis. K. — Tip of an ultimate branch with two sterigmata- bearing conidia. M. — Basal cell of a conidiophore from teosinte with septum formation progressing by the centripetal extension of a cellulose- pectose ring 122 25. A. — Conidiophore from sorghum, partly matured and bearing few conidia. B. — Conidiophore from teosinte, nearly mature, with extensive system of branches bearing many conidia. C. — Typical conidia from sorghum. D. — Typical conidia from teosinte. E. — Typical conidia from teosinte which have germinated in dew on the leaf smface. F. — Conidium from teosinte germinating by an extensive branched hypha when main- tained in dew at 7° C. G. — Conidium from teosinte germinating while still attached to its sterigma. H. — Typical conidia from maize, showing common variations in shape and size. I. — Two conidia fi'om maize just beginning to germinate in rain water. J. — Two conidia from maize germinating in sterilized brook water maintained at 8° C. K. — Coni- Apr. i-sept. IS. 1921 Illustrations ix Page dium from maize germinating in dew on the leaf surface. L. — Conidium from maize giving rise to extensive branching hyphae in a dilute decoc- tion of young maize kernels 122 Effect of Drugs on Milk and Fat Production Text Figures 1. Graph showing effect of tonic mixture No. i on butter-fat and milk yield. . 127 2. Graph showing effect of air-slaked lime on butter-fat and milk yield 127 3. Graph showing effect of Fowler's solution of arsenic on butter-fat and milk yield 128 4. Graph showing effect of powdered gentian on butter-fat and milk yield 128 5. Graph showing effect of the German tonic mixture on butter-fat and milk yield 129 6. Graph showing effect of physostigmine sulphate on butter-fat and milk yield 129 7. Graph showing effect of sodium bicarbonate on butter-fat and milk yield. 130 8. Graph showing effect of ginger on butter-fat and milk yield 130 Halo-Bught of Oats Plates C. Halo lesions on flag leaves of Wisconsin No. 14 oats 172 26. Typical isolated halo lesions 173 27. Halo lesions on Wisconsin No. 14 oats produced by spraying with a water suspension of the stock organism May 26, 1917 172 28. Infection from imtreated 1916 seed of Wisconsin No. 124 oats 172 29. Spikelets of Wisconsin No. 14 oats: A. — Left and center spikelets show natural infection with halo-blight. B. — Upper spikelet shows typical isolated halo lesion near base 172 30. A. — Two per cent -I-5 glucose Difco peptone-beef bouillon agar slant of No. 36. B. — Two per cent potato-dextrose agar slants 172 3 1 . A. — ^Three-day colony of stock on 2 per cent dextrose-potato agar. B . — Two- day colonies of stock on -fio beef-peptone agar. C. — Five-day colony of stock on potato-dextrose agar. D. — Five-day colony of No. 36 on -fio beef -peptone agar. E.— Three-day colony on 2 per cent glucose Difco peptone beef bouillon agar. F. — Seven-day colony of stock on -f 15 beef- peptone agar. G. — Margin of 3-day colony of stock on -\-i^ gelatin 172 32. A. — Five-day colonies of stock on potato-dextrose agar. B. — ^Three-day colony of stock on potato-dextrose agar 172 33. Isolations from sections of halo lesion 172 34. A. — No. 36 from 24-hour potato-dextrose agar slant; carbol fuchsin stain. B. Stock from 2 4-hom- potato-dextrose agar; Ribbert's capsule stain. C. — Stock from 4-day potato-dextrose agar; carbol fuchsin stain, showing long chains. D. — Stock from 3 -day potato-dextrose agar; Ribbert 's cap- sule stain. E. — Stock from i-day -f 15 beef-peptone agar; Van Ermengem stain. F.^No. 36 from i-day -I-5 beef-peptone agar; Caesar-Gil stain. . 172 35. Sections of oat leaves through halo lesions, showing bacteria in the tissues. 172 Behavior of the Citrus-Canker Organism in the Sou, Plates 36. A. — Citrus-cankers on mature wood of trunk of Citrus aurantifolia. B. — Citnxs-cankers on mature wood of branches of Citrus aurantifolia 206 418°— 21 2 X Journal of Agricultural Research voi. xix Page 37 . A. — Results of inoculations with Pseudomonas citri upon roots of sweet orange {Citrus sinensis). B. — Skeletonized leaves of Ellen grapefruit recovered from buried soil 206 Production of Conidia in GibberElla saubinetii Text Figure I. Conidial production in Gibberella saubinetii (Mont.) Sacc: A, Ascospores from cornstalk, germinated in distilled water, producing conidia in three days; B, D, typical conidia and conidiophore from a 28-hour-old hanging drop culture from a conidium from A; C, germinating conidia from a 52-hour-old plate culture; E, conidiophore and germinating conidia from a 47-hour-old colony in a Van Tiegham cell 236 Effect op Manure-Sulphur Composts upon the Availability of the Potassium of GreEnsand Text Figure I. Diagrams showing relation of the water-soluble acidity to the water-soluble potassium at different time periods for different greensand composts .... 250 Rust in Seed Wheat and Its Relation to Seedling Infection Text Figure I. Diagram of air- washing apparatus for isolated room used for growing rust- infected seed 268 Plates 38. Heads of Kubanka durum wheat heavily infected with stemrust 277 39. A. — Portion of one of the heads shown in Plate 38. B. — Wheat kernels showing typical stemrust infection 278 40. Exterior view of isolated room in the pathological greenhouse at the Uni- versity of Wisconsin, showing (a) the exterior portion of air- washing apparatus used to wash all air drawn into the room, (6) the canvas curtain used for shading on warm days, and (c) the sprinkling attachment used to throw spray of water over the roof to aid in keeping the room cool 278 41. A. — Photograph of wheat grown in fiats in isolated room in greenhouse at the University of Wisconsin. B. — Same plants as A, when well headed. . . . 278 42. Longitudinal section through hilar portion of an immatiu"e wheat kernel, showing sorus of stemrust 278 43. Longitudinal section through the hilum of a wheat kernel infected with stemrust, showing imusually large internal sori extending nearly across the kernel 278 44. Cross section of a mature wheat kernel infected with stemrust, showing telia in the ventral groove 278 45. Enlarged portion of section shown in Plate 44, showing telia on surface of ventral groove 278 46. Longitudinal section of embryo of germinated wheat kernel showing large internal rust at x in hilar tissue at base of embr>'onic tissue 278 47. Longitudinal section of the embryo fiuther advanced in development than that shown in Plate 50 278 48. Longitudinal section through young secondary root of wheat embryo, showing presence of intracellular mycelium 278 Apr. i-sept. IS. 1921 Illustrations XI Practical Universauty ok Field Heterogeneity as a Factor Influ- encing Plot Yields Text Figures Page 1. Montgomery's diagram of 5.5 by 5.5 foot plots of Turkey wheat, showing variations in the percentage of nitrogen in the grain 283 2. Diagram showing yield of alfalfa in first cutting, 19 13, on the Huntley ex- perimental tract 286 3. Diagram showing yield of alfalfa in second cutting, 19 13, on the Huntley experimental tract 287 4. Diagram showing yield of alfalfa in first cutting, 1914, on the Huntley ex- perimental tract 288 5. Diagram showing yield of alfalfa in second cutting, 19 14, on the Htmtley ex- perimental tract 289 6. Diagram showing yield of alfalfa in third cutting, 1914, on the Htmtley ex- perimental tract 290 7. Diagram showing yield of unhusked rice in Coombs and Grantham's 54 plots j4hy yi chain square 296 8. Diagram showing yield of ear com, 19 15, on the Huntley experimental tract . 297 9. Diagramshowingyieldof ear com, 19 16, on the Huntley experimental tract. 298 10. Diagram showing location of sample areas examined for soil moisture in a field at San Antonio Experimental Farm 302 Transmission of the Mosaic Disease gb Irish Potatoes Plates 49. Vines of Green Mountain variety inoculated with juice from healthy foliage of the same variety 338 50. Vines of Bliss Triumph variety inoculated with juice from healthy foliage of Irish Cobbler variety 338 51. Vines of Irish Cobbler variety inoculated with juice from healthy foliage of the same variety 338 52. Vines of Green Mountain variety inoculated with juice from mosaic foilage of the same variety 338 53. Vines of Green Mountain variety inoculated with juice from mosaic foilage of Bliss Triumph variety 338 54. Vines of Green Moimtain variety inoculated with juice from mosaic foliage of Irish Cobbler variety 338 55. Vines of Bliss Triumph variety inoculated with juice from mosaic foliage of Irish Cobbler variety 338 56. Vines of Irish Cobbler variety inoculated with juice from mosaic foliage of the same variety showing bad stage of mosaic 338 Relative Susceptibility to Citrus-Canker of Different Species and Hybrids of the Genus Citrus, Including the Wild Relatives Plates 57. Leaf of Casimiroa edulis with naturally occurring spots from the greenhouse inoculations 362 58. Infected leaves fi'om plants of Chaetospermum glutinosum in the greenhouse experiments, showing the types of canker spots produced 362 XII Journal of Agricultural Research voi. xix Page 59. A. — Leaves of Atalantia citrioides and A. ceylonica (center) from plants in the greenhouse experiments, showing the canker spots typically produced on these plants. B. — Compound leaf of Hesperthusa crenulata from isolation field, with naturally occurring canker spots on two of the leaves. C. — Leaves of Microcitrus Garrowayi from plants in the greenhouse experiments, with different types of canker spots. D, E. — Infected leaves and twigs of Eremocitrus glauca from greenhouse plants, showing the large flat spots on the leaves and the rather corky spots on the twig and thorn 362 60. A. — Typically infected leaf of Fortunella margarita. B. — Old leaf of Citrus grandis, with raised, compact, oily, unruptured spots. C. — Fortunella Hindsii, with ruptured corky spots. D. — Citrus sp., Kansu of Yuzu Orange. E. — Citrus aurantifolia, showing typical infection 36a 61. Upper and lower leaf surfaces of a Citrus hystrix leaf with a heavy natural canker infection 36a 62. Typically infected leaves of Citrus grandis from field plants showing ex- treme susceptibility 36a 63. A. — Leaves and twigs of faustrime from greenhouse experiment with typical spots. B. — Types of spots foxmd on Citrus nobilis (King of Siam, Naran- jita, and tangerine). C. — Leaf of tlie citrangequat from greenhouse ex- periment. D. — Citrumelo leaf with typical canker spots. E. — Upper and lower surface of a naturally infected leaf of Citrus mitis in the field . . 36a 64. Naturally infected leaves of Citrus nobilis var. unshiu from the field, show- ing various types of spots produced 36a 65. Some of the hybrids of Porcirus trifoliata, showing vigor, type of growth, leaf characters, and relative susceptibility to citrus-canker, arranged in order of their susceptibility 36a 66. A. — Limelos in the greenhouse inoculation experiments, showing type of growth, leaf characters, and susceptibility to citrus-canker. B. — Lime- quats in the greenhouse inoculation experiments, showing type of growth, character of leaves, and susceptibility to citrus-canker 36a 67. A. — Siamelos in the greenhouse inoculation experiments, showing type of growth, leaf characters, and susceptibility to citrus-canker. B. — Com- parison of type of growth, leaf characters, and susceptibility to citrus- canker in clemelo, satsumelo, and tangelo in the greenhouse inoculation experiments 3^3 68. A. — Results of the greenhouse inoculations with some of the false hybrids. B. — ^Tangelos in the greenhouse inoculation experiments, showing type of growth, leaf character, and susceptibility to citrus-canker 36a Presoak Method of Seed Treatment: A Means of Preventing Seed Injury Due to Chemical Disinfectants and of Increasing Germi- cidal Efficiency Text Figures 1. Graph showing effect of formalin i to 400 treatments with and without pre- soaking 37° 2. Graph showing effect of formalin i to 200 treatments with and without pre- soaking 37* 3. Graph showing effect of formalin i to 320, with and without presoaking 374 4. Graph showing effect of copper sulphate i to 80, with and witliout presoaking, on wheat and barley seed germination 375 5. Graph showing effect of formalin i to 320, with and without presoaking, on wheat seed germination under field conditions 377 Apr. i-sept. IS. 1921 Illustrations xiii Page 6. Graph showing effect of formalin and copper-sulphate presoak treatments of >^-bushel wheat seed lots 378 7. Graph showing effect of formalin i to 320 treatments for various periods, with and without presoaking 385 8. Graph showing effect of formalin i to 320 and i to 200 on germination of com, barley, and oats with and without presoaking 386 g. Curve showing rate of absorption of water by dry wheat seeds 387 Plates 69. Relative injury to wheat-seed germination caused by short and long for- malin treatments 393 70. Eft'ect of formalin i to 400 treatment for 6 hours, with and without 6-hour presoak. i. Fultz wheat. 2. Poole wheat 392 71. Effect of formalin i to 400 treatment for 6 hotus, with and without 6-hour presoak. i. Fulcaster wheat. 2. Turkey wheat 392 72. Stimulating effect of the presoak method of treatment with formalin i to « 400 392 73. Effect of formalin i to 200 treatment for 6 hours, with and without 6-hour presoak 392 74. Effect of formalin i to 320 treatment for 6 hours, with and without 6-hour presoak 392 75. I. Fife wheat: A, B, control, 83 per cent germination; C, D, seeds treated with formalin i to 320 for 3 hours, 62 per cent germination; E, F, seeds presoaked 6 hours, then formalin i to 320 for 6 hours, 90 per cent germi- nation. 2. Poole wheat: A, B, control, 85 per cent germination; CD. seeds treated with formalin i to 320 for 6 hours, 55 per cent germination; E, F, seeds presoaked 6 hours, then formalin i to 320 for 6 hours, 88 per cent germination. 3. Effect of soaking in water throughout presoak period, compared with procedure of keeping moist 6 hours: A, B, Fife wheat; Q, D, Poole wheat; A, C, seeds soaked in water 5 hours, then treated with formalin i to 320 for 7 hours; B, D, seeds soaked in water 10 minutes, drained, and kept moist 6 hours, then treated with formalin I to 320 for 6 hours 392 76. Effect of formalin and copper sulphate on wheat and of copper sulphate on barley, with and without presoaking 393 77. Effect of presoak method used with copper-sulphate treatment of wheat. . 393 78. Effect of presoak method used with formalin i to 320 on wheat under field conditions 392 79. Effect of presoak method used with fortnalin i to 400 on blackchaff bacteria 393 80. Effect of presoak method used with formalin i to 400 on blackchaff bacteria 39a 81. Effect of presoak method on barley and oats 393 82. Effect of presoak method used on barley, oats, and com 393 Daily Development of Kernels of Hannchen Barley jrom Flowering TO Maturity at Aberdeen, Idaho Text Figures 1. Graph showing length, lateral diameter, and dorsoventral diameter of barley kernels for the 25 days following flowering 395 2. Graph showing percentage of moisture per kernel from date of flowering. . . 397 3. Graph showing maximum and mean daily temperatures recorded at Aber- deen, Idaho, from July 8 to August 4, 1916 (broken line), and from July 15 to August 10, 19 18 (solid line) 398 XIV Journal of Agricultural Research voi. xix Page 4. Graph showing lateral and dorsoventral diameters of the ovary tip as com- pared with length, lateral diameter, and dorsoventral diameter of the kernel for the 26 days following flowering in 1916 408 5. Graph showing average length of kernels 6, 7, and 8 (solid line), average lateral diameter (dotted line), and average dorsoventral diameter (broken line) from plot i in 1917 410 6. Graph showing average length of barley kernels, including ov,af y tip, from flowering to near maximum development in plot i in 1917 411 7. Graph showing average lateral diameter of barley kernels from flowering until near maximum development, plot i, 1917 41a 8. Graph sliowing dorsoventral diameters of barley kernels from, flowering until near maximum development 413 9. Graph showing wet weight of individual kernels 5, 8, and 10, by days, from date of flowering to near maturity in 1917 418 10. Graph showing average wet weights of kernels from flowering to mattu-ity in plot I in 1917 419 1 1 . Graph showing dry matter per kernel from date of flowering to near maturity in 1916 (dotted line) and in 1917 (solid line) 420 12. Graph showing dry matter per kernel at 12-houx intervals from flowering to matiurity 42 1 13. Graph showing dry- weight gain of kernels 6, 7, and 8 in 12-hour periods. . 422 14. Graph showing percentage of moisture in morning and evening samples of Hannchen barley in 1917 42a 15. Graph showing nitrogen per kernel from date of flowering La 1916 (broken line) and in 1917 (solid line) 424 16. Graph showing ash per kernel from date of flowering in 1916 (broken line) and in 1917 (solid line) and the percentage of ash in 1917 (dotted line). . 426 17. Periods of development of the barley kernel as indicated by records during three years at Aberdeen, Idaho 427 Plates 83. A. — Fertilized ovary. B. — Kernel i day old. C. — Kernel 2 days old. D. — Kernel 3 days old 430 84. A. — Kernel 4 days old. B. — Kernel 5 days old. C. — Kernel 6 days old. D. — Kernel at later stage of development 430 85. Kernel 5 days after fertilization 430 86. Kernel 6 days after fertilization 430 87. Kernel 9 days after fertilization 430 88. Kernel 14 days after flowering 430 8g. Kernel 20 days after fertilization, at which time growth was nearly com- pleted 430 90. Kernel 25 days after fertilization, growth completed 430 91. Section of a nearly mature kernel, showing cells next the furrow 430 Development op Barley Kernels in Normal and Clipped Spikes and THE Limitations of Awnless and Hood&d Varieties Text Figures I. Graph showing growth in length, lateral diameter, and dorsoventral diameter of kernels of Manchuria barley in normal and clipped spikes 448 2. Graph showing wet weight of kernels of Manchuria barley from normal and clipped spikes 449 Apr. i-sept. 15. 1921 Illustrations xv Page 3. Graph showing dry matter in kernels of Manchuria barley from normal and clipped spikes 449 4. Graph showing total ash in kernels of Manchuria barley from normal and clipped spikes 450 5. Graph showing total nitrogen in kernels of Manchuria barley from normal and clipped spikes 451 6. Graph showing water in kernels of Manchuria barley from normal and clipped spikes 451 7. Graph showing growth in length, lateral diameter, and dorsoventral diam- eter of kernels of Hannchen barley in normal and clipped spikes 463 8. Graph showing wet weight of kernels of Hannchen barley from normal and clipped spikes 464 9. Graph showing dry matter in kernels of Hannchen barley from normal and clipped spikes 465 10. Graph showing percentage of ash in the kernels, rachises, paleae, and awns of normal spikes of Hannchen barley and in the kernels, rachises, and paleae of clipped spikes 467 ^i. Graph showing total nitrogen in kernels of Hannchen barley from normal and clipped spikes 468 12. Graph showing water in kernels of Hannchen barley from normal and clipped spikes 469 13. Graph showing relation of length of awn to weight of clipped kernels and undipped spikelets on a 2-rowed barley gro\vn at Arlington Farm, Va. . 470 Investigations in the; Ripening and Storage of Bartlett Pears Text Figures 1. Sugars in Bartlett pears from Sacramento, Calif 484 2. Sugars in Bartlett pears from Suisun, Calif 485 3. Sugars in Bartlett pears from Medford, Greg 486 4. Sugars in Bartlett pears from Yakima, Wash 487 5. Acids in Bartlett pears from Sacramento, Calif 490 6. Acids in Bartlett pears from Suisun, Calif 491 7. Acids in Bartlett pears from Medford, Greg 491 8. Acids in Bartlett pears from Yakima, Wash 492 Further Data on the Orange-Rusts of Rubus Plates D. I. — Infected black raspberry leaf covered with the caeomas of Gymnoconia inlerstitialis . 2. — Blackberry leaf infected with the short-cycled orange-rust 512 92. Manner of germination of the spores of the two orange-rusts 512 93. Short-cycled orange-rust 512 94. Gymnoconia interstitialis 512 Mosaic Disease of Corn Plates 95. Mosaic disease of com 522 96. Mosaic disease of com : Effect of early infection on the ear 522 XVI Journal of Agricultural Research voi. xix Genetics of Rust Resistance in Crosses of Varieties of Triticum vul- GARE WITH Varieties of T. durum and T. dicoccum Plates Page 97. A. — Pollen grains of Marquis wheat B. — Fj Marquis XKubanka (C I 2094), showing sterile grains C. — Pollen grains of Kubanka (C I 2094). D. — Stems of Kubanka (C I 2094) grown under rust-epidemic conditions. E. — Fj Kubanka (C I 2094) X Marquis, showing normal uredinia. F. — Marquis, the susceptible parent. H. — Fj emmer (Minnesota ii65)XMar- quis, showing no normal uredinia. I. — Minnesota 1165, the resistant emmer parent 542 98. A, B, C. — Face and side views, respectively, of heads of lumillo (C I 1736), Fj lumillo X Marquis, and Marquis. D, E, F. — Face and side views, respectively, of heads of emmer, Minnesota 1165, Fj emmer X Marquis, and Marquis. G, H, I. — Kernels of Marquis, Fi emmer X Marquis, and emmer 542 99. Representative heads of F3 families of the cross between dtuiim and Marquis. 542 100. A. — F3 family of a cross between emmer (Minnesota 1165), and Marquis, showing face and side view. B. — Heads of an F3 family which resem- bled common wheat. C. — Heads of different plants of an F3 family of a cross between emmer (C I 1524) and Marquis. D.— Represents a very frequent sort of segregation obtained in the F3 generation 542 loi . Heads of resistant and susceptible wheat obtained in the F3 generation from the cross between Marquis and durum 542 102. Greenhouse experiment 542 C)ccurrence of the Fixed Intermediate, Hordeum intermedium haxtoni, IN Crosses between H. vulgare pallidum and H. distichon palmella Plates 103. Individual heads, representing the three phenotypic progeny classes in which the lateral florets bear a^vns 591 104. Individual heads, representing the phenotypic progeny classes in which the lemmas of the lateral florets are rounded and awnless 591 105. A. — Awn-pointed individual, heterozygous for regressive 6-rowed and 2-rowed characters. B. — Short-awned individual, heterozygous for re- gressive 6-rowed and 2-rowed characters. C. — Long-awned individual, heterozygous for regressive 6-rowed and 2-rowed characters. D. — Three spikes of Mansfield barley from the same plant, showing the variations of fertility in the lateral florets 591 106. A. — Infertile spike of potentially fertile Hordeum intermedium. B. — Fertile spike of H. intermedium. C. — Var. atterbergii, probably a sterile intermedium 591 Vol. XIX Ar>RIL 1, 1920 No. 1 JOURNAL OP AGRICULTURAL RESEARCH CONXE^NTS Pag* A Teosinte-Maize Hybrid ...... i G. N. COLLINS and J. H. KEMPTON (Cotitribotion trom Buieaa ot Plant IndiMtty } Banana Root-Borer ..-.-..-39 G. F. MOZNETTE (Contribtttlaii from Bttreao ot Batomologsr) PUBUSHED BY AUTHORITY OF THE SECRETARY OF AGRICULTURE, WITH THE COOPERATION OF THE ASSOCIATION OF LAND.GRANT COLLEGES WASHINGTON, B. G. WAaHiHOTOM : oovcuhmcnt printino opfioc : ing EDITORIAL COMMITTEE OF THE UNITED STATES DEPARTMENT OF AGRICULTURE AND THE ASSOCIATION OF LAND-GRANT COLLEGES FOR THE DEPARTMENT KARL F. KELLERMAN, Chairman Physiologist and Associate Chief, Bureau of Plant Industry EDWIN W. ALLEN Chief, Office of Experiment Stations CHi^LES L. MARLATT Entomologist and Assistant Chiey, Bureau of Entomology FOR THE ASSOCIATION J. G. LIPMAN Dean, State College of Agriculture, and Director, New Jersey Agricultural Experi- ment Station, Rutgers College W. A. RILEY Entomologist and Chief, Division of Ento- mology and Economic Zoology, Agricul- tural Experiment Station of the University of Minnesota R. L. WATTS, Dean, School of Agriculture, and Director, Agricultural Experiment Station, The Pennsylvania Slate College All correspondence regarding articles from the Department of Agriculture should be addressed to Karl F. Kellerman, Joxunal of Agricultural Research, Washington, D. C. All correspondence regarding articles from State Experiment Stations should be addressed to J. G. Lipman, New Jersey Agricultural Experiment Station, New Brunswick, N. J. NEW yok:- o)-bi 1°^ •as 0 w 0 . 0 0 « 79 4 9 12 s 3 2 2 2 I I I 8s 34 3 2 I 3 6 II 4 8 9 10 8 7 9 I 7 2 4 2 3 2 4 7 10 9 II 11 7 9 8 10 I 13 2 S I 2 3 2 56 10 10 13 6 13 10 4 z I 47 6i 10 3 I I I 8 5 22 29 24 20 7 12 2 1 I 4 4 3 5 II 9 10 II 12 14 9 S II 5 S 3 I 37 36 6 1 1 I 4S 8 S 3 6 S 25 4 21 9 II 9 8 S 3 1 4 I 9 3 I I I I 4 II 6 10 7 19 3 II 6 8 S 2 7 3 6 3 I S I 2 4 12 7 II 8 21 la 14 7 10 5 4 I 16 17 18 I 4 I 2 3 I I I 36 37 I 38 I I 2 I 30 33 3 I I 34 36 I 38 I I I 46 I 53 Number 123 14. I 4.0 122 22.7 S-i 123 16.2 3-47 120 1.28 2.21 I2S r.36 .62 122 1.78 .81 123 7.0 1.92 I2S 16.8 4-1 12s 10.9 II. 8 112 6.13 4.07 123 2.04 2-39 Mean S-S2 1.67 Standard deviation "Figures indicate number of plants exhibiting each character to the extent shown in the first column. For discussion of units of measurement see p. 7-8. dcs ing ex. Prophyl- Iiry spikes. Length j of pru- phyllary. Rows in central spike. Position of best spike. Number of alicoles. Number of suckers. Branch silking first. Days to pollen. Pollen to silk. I,ength ui inter- node on third. 3.46 . 12 . 02 •23 ■ 07 •17 •38 . 22 0. 07 . 01 D- .01 D- .12 .00 — .17 0. 16 D— .04 . 20 .01 D .02 — .17 D 0. OS - . 14 D .06 D— .01 D- .03 . 02 —0. 20 - .18 D .03 D .31 . 21 D- .IS D 0.07 - .18 D .17 D .09 . 12 D— .12 0. 05 D- . 10 . 21 • 17 — .19 — .31 0. 02 •33 •03 D- .36 0.47 •79 .14 D- .42 — ^17 D .36 D 0. II D .02 D-0.33 D- .66 D .18 •31 .48 D .04 — .29 — -13 — . 10 — .29 — .09 — .06 — .19 . 20 .26 •OS . 10 •03 •73 . 20 .16 •IS — . IS — . 02 D .OS - .30 - .26 - ^37 — -23 - .26 . II D- .06 D— .04 D .04 — .02 — •OS . 12 .14 . 20 •03 .05 . 00 D— .09 .04 — .00 .02 D- .16 D .12 — .07 — -30 — . 18 — . 02 — . 20 — .17 D .24 D .10 — .08 — ^25 — .19 .01 D— .ox .04 D- .OS .01 •32 •32 •59 •57 •17 D .04 D- .2S D- .36 - .03 •OS D- .15 .02 •13 •23 .19 . 12 • 38 — .04 D .08 . 12 D .18 D .06 D .16 •14 .02 •33 D— .29 •03 D— .40 D— .04 •51 D .04 . 10 D- .13 D- .09 D- .04 D .07 D- .09 . 10 . 12 D- .10 D .13 D- .10 D .22 — .29 — -35 •23 — . 12 D .03 - .08 . 00 .04 - .OS D .13 - .IS - -ss. • 31 - -37 .00 .00 •30 - .29 . II .00 . 00 •IS — -IS •17 . 00 . 00 D .01 •05 .09 D— . 21 •35 .00 •39 .09 .40 - -03 — .04 D .47 — .25 D .05 D .09 .01 •03 D- .12 .07 .07 .06 D .18 D-.04 .09 •39 .40 .04 ' -25 .09 .03 .07 .06 .09 •59 •59 D .08 — •OI - .56 D .31 D .OS T> .IS .08 . 10 •14 •05 .08 D- .19 — . 18 D .06 . 10 • 29 D- .08 - -56 D .05 .08 .14 .08 - .18 . 10 — .01 D .31 D .13 . 10 •OS D- .19 D .06 •29 D- .OS •37 D .02 — . 01 — .09 D- .18 D .06 D- .OS .01 D .09 — .34 - .09 .40 D .02 •37 .01 — . il~- D .05 — .29 .01 D .23 D .02 D .09 — . XX 7o2^ D- .09 - ^23 .16 — . 01 - -34 D .OS .02 - .09 - .09 - .29 D— .09 •31 D- .s6 D- .18 .40 .01 - .23 D .06 D .02 D .23 .16 D- .21 — .14 D- .56 D- .21 ■^^^^^ ^^~^~^-^ Table HI. — Correlation of characters in Fj of ti ..„«„..„.„.. Hcisht. leaves. sucker. s^^ Cireum- branches a'Si^ Nodes on third. o?&. SSS. Second- Second- Tassel branches i£' ^ Double au"les. Sinijle female alicolcs. Allcole .Snl ilidS' Prophyl- Length 'Sik'e' ofbcsT .pike. "3 " urn I. silkinB K° todit' LeoEth bt - — e;<»^ ^ D-.,. -O.ISJ D .04 -0... -a., D .« .43 0.S4 D- .j6 ":" ■E III D-o a8 -■^ °T "4 D^o.., D-o. 04 dI.:: 0..8 D.03 -0... D 00. D- .j6 d':: T lull o.6p D « : S: z ° 2 : :: ■'I 3« ii 11 ' hi f ckcr D 4B ^ "^ "" ;; "' I - z D z D- D- z D 03 S ck "ndcx .3. 41^ °-'» ::: :» D ;n dI:: d ..a .48 C'r umtercncc index Nodes w'lhout branehcs D .04 D- .„ D .„ D .10 -.49 -.09 D .04 - .»o ao .":: D D I .08 D- ... >j :::^ It [:-;i : I :::: - .3= D~ °° - T I z .00 .00 — .ao .18 :::: „= 63 - j» : :: : z : E z - !» - :! D- .06 D .04 nI^h^'o',;™.';™"""' Primary braachcs :p "5^ i "s ll 36 ■~ D ..8 : !^ : 5 - .0. :i :r^^ ■" :!! D- .06 l\ I .00 .«, D-.30 D ... "'.Z D- D- I ■« E - I "-_ I ^- °f °: I D- °o! Z ;; D .0, D-..S D-.J6 ^. . . . -•■■ D- ... .00 ■» -.., -.03 TuMdbranche. on third .lO D .,0 D .,« •39 :T^Tv-^ D-.09 D-.08 D-..0 OS D-..S D^.ii ' "l D- 7. -'Z D~ « d" 11 d-'k D-' r D-'I! D-!^ ■^T^ ^ n-'.Z D- 30 D-.o, >9 38 D z 0 « D ti 0. D-..9 D- " Malesccondflric* D .« D-.j. .47 -.3 D-.,4 -.0. D .04 .!* ■~ - «■ - I D- .0] - ~ = II d" E To ": .83 .83 D .58 ll D .38 D D- I9' d" E r ;: - I - i - 38 - " - r ■i D-io' Alleolc index D- .oj -.09 -■.4 .0, D-.08 ..9 ^'^^ ^.^^ .09 _ D-..6 °i D-.o. D- r ■°° -'Z _ ^ — 38 04 D .6j ;^ D-.06 D-.o. D-:!o dI :!! « .09 ... D-.,6 D .., dI:" "^ M :>,.:". » .9 ~ z °_ " ° z 0. D-.i. z D .,8 D-.o. Prophylioiy spitccs D- z D-.o. D- o! .00 - M I » -.'" 04 - !oo .00 D- .09 !" 36 D- .13 "".Z dI:ii D .0; D ..J l-'Z 00 ;» ir -^ " ° t d" !6 JJ z 08 ^05 D- 08 d":!! ]» Rowsin central spike ° 'i .: ° ::: D- D- D 3« 4S D- .03 D .36 D- i: "' :: .::E '4 -;;; — .04 D- I ■i :;5 - -38 :E Z ° °8 °":: -~ ss D- :; D- z ? °r i^- r Sa^ 14 dI ■I*- ;- D .06 jj Number o( suck D- )i Duyaol pollcD. . ~z -'■Z T .0. D ::; "■Z -..4 D- .s6 •" -.., -..3 D-.36 D-... .0, D-.., ' ■ ■ '■""-^ April I, 1920 A Teosinte-Maize Hybrid II Table IV. — Distribution of individuals in Fj of teosinte-maize hybrid with respect to varioxis characters — Continued Units of measuremement. a c^ 4 ei as C « .ss l-t i a. "o si S «°8 8^; .2 D be fcS S3 |3 •3 a i_ bo ;2; .9^ "" ti s a a st Q CO S ci i1 •" *^ a og-;^ 23 23 12 12 iti II 8 S I 2 3 3 13 22 23 iS 29 II 6 S 3 I a 3 43 43 II 6 3 3 IS 27 25 15 13 II 10 3 3 I 2 I I I 5 4 4 3 8 5 3 S I 3 8 7 6 6 8 2 I 6 III 10 I I I I 3 I I 2 8 2 3 II 8 10 8 10 9 11 9 7 I S 3 S 2 3 I 3 3 7 9 s 15 15 23 14 19 5 62 6 4 S S 14 s IS 8 6 8 2 II . . s 9 9 7 16 6 16 14 8 4 2 S 17 9 6 13 14 I 8 9 7 4 15 6 16 3 S 3 17 18 19 2 20 4 ai I 3 2 I 32 23 34 25 2 4 26 I I 37 I 38 I I 3 39 31 3 I 'I 33 2 35 36 37 Z 38 39 40 I 2 42 I 46 122 18.3 9-7 Number 114 3-04 121 13-4 7.69 125 6.62 1.87 123 2-13 • 38 120 2.47 2.14 17.9 6.17 127 II. 7 3-36 106 1.89 •44 125 III. 9 21-5 119 10.9 4.19 Mean " Figures indicate number of plants exhibiting each character to the extent shown in the first column. For discussion of units of measurement see p. 7-8. >> First date recorded 71 days after planting and subsequently at lo-day periods. DISCUSSION OF CHARACTERS AND THEIR CORRELATIONS HEIGHT GROUP HEIGHT Confining the measurement of height to the main stalk does not always give a fair idea of the size of the plant, since there were many individuals in which the suckers greatly exceeded the main stalk. (See distribution of sucker index, Table V.) 12 Journal of Agricultural Research Vol. XIX, No. I Table V. — Distribution of individuals in Fj of teosinte-maize hybrid with respect to characters recorded as indices ii ■3 |8 "n " eg 0« aj'o s as "3 .He rr ti fcS 0 3 V. 3 c« a' ^ Sa "3 0 ■3 •a Co, 3 li sa 1 •Ho c . a M ii s .5, 1'" Per ct. 0- 4 5- M IS- 24 as- 34 35- 44 4S- S4 55- 64 105 2 3 6 S 3 no 6 I 3 3 34 27 13 S s 4 7 s 6 S 13 18 4 6 4 8 7 9 6 14 23 34 36 25 12 5 S 3 7 4 4 3 19 41- 50 51- 60 61- 70 71- So 81- 90 91-100 lOI-IIO III-IIO I2I-I3C 131-140 141-150 151-160 I6I-I70 171-180 I 2 I I n 20 30 21 19 7 2 I I Perct. 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 4^ 43 44 45 46 47 48 49 SO 51 52 53 54 55 56 57 58 61 62 63 69 78 83 91 3 3 2 5 4 8 7 5 5 6 S 4 4 S 3 4 Percent. 0 I- IC 11- 20 21- 30 31- 40 41- SO 51- 60 61- 70 71- 80 81- 90 91-100 lOl-IIO 111-121 121-130 131-140 141-150 151-160 161-170 171-180 181-I9O 191-200 2 1 13 12 17 12 12 12 9 8 I 7 4 3 r 2 a Perct. c- 9 ic-19 20-29 3C-39 40-49 50-59 60-69 70-79 80-89 90-99 100 I 2 2(11 23 48 15 5 4 I I 3 8 It IS 27 21 3S 65- 74 75-84 85- 94 ■ 221-230 I 451-460 I I 3 I 321-330 I I 2 4 3 X 4 3 I I 3 I I I I lie 45.2 12.4 No.. 123 A. CC 123 2. 36 7-96 123 32-4 46- S 123 61. 0 68.0 123 34- c 41-3 119 121. 5 37-7 124 70. 0 49.0 Xi8 41-5 8.8 M... 80 6 ff II. Q 18.8 1 o Figures srive number of plants exhibiting each character to the extent shown in first column in this sec- tion of the table. The average height of Tom Thumb maize plants at Chula Vista was 6 dcm. and that of Florida teosinte 23 dcm. The Fj plants averaged 17 dcm. The mean of the Fj plants was 14. The range was from 2 to 23 dcm. The distribution (fig. i) was as nearly normal as could be expected from the number of individuals involved. There is, furthermore, no in- dication of skewness, the mode and the mean practically coinciding. Although the parental varieties differ greatly in height, the parental species overlap. Indeed the taller varieties of maize probably exceed the tallest teosinte in height. Height is positively correlated with all of the four tassel measurements, and the correlations are significantly higher than was found in a progeny April I, 1920 A Teosinte-Maize Hybrid 13 of Tom Thumb where two of these character's were recorded. Thus there is evidence of coherence between height and the character of the tassel. Disherence with male secondaries is indicated by a correlation of - 0.28 ± 0.07. The negative correlation of - 0.46 ± 0.06 with nodes silking index would also seem a clear example of dis- herence. The correlation of 0.47 with days to pollen indicates coherence of this character with height. In both parent populations this correlation was negative, but under most circumstances the late plants of a maize variety are taller than the early plants. The negative correlation of —0.33 with length of intemode on third is in the direction of a dis- herence, though this is probably associated with the negative correlation between height and sucker index, which is to some extent physical. Anything which tended to interfere with the growth of the main culm would doubtless stimu- late the development of all the branches. TOTAt, LEAVES The total number of leaves on the main culm in Tom Thumb is usually 1 1 , in Florida teosinte about 37. The mean in the Fj hybrid plants was 23, with a range from 9 to z^. The distribution (fig. 2) is normal, and the variability as measured by the coefii- cient of variation is the low- est recorded for any char- acter. The larger varieties of maize equal or exceed teo- sinte in number of leaves just as they do in height. In both maize and teosinte total number of leaves is a character very little affected 30r /fS/GAfT //VO£C/M£TePS Fig. I.— Height: frequency dis- tribution of plants in Fj. Class value, 2 dan. - -1 ■n -1 20 \ /O Q \ii ^TT^^T^nV^W^V^ n Fig. a.-Total leaves: frequency distribution of plants in F2. by changes in the environ- ciass value, one leaf. ment. In maize there is usually an intra-varietal correlation of about 0.3 between total leaves and height. Corresponding data for teosinte are not available, but the coefficient of 0.69 in the hybrid material affords some evidence of coherence between these characters. The correlations of total leaves with other characters are similar to those of height, with the exception that there is no evidence of disherence with nodes silking index. There is also coherence with branch silking first. 14 Journal of Agricultural Research Vol. XIX, No. I HEIGHT OK SUCKER Measurements were taken from the ground to the tip of the tassel of the tallest sucker or tiller and recorded in decimeters. Tom Thumb almost never produces a sucker. In Florida teosinte there are usually numerous suckers of practically the same height as the main culm. The parent varieties are thus widely separated, but there are varieties of maize with suckers taller than any recorded in teosinte. The mean of the F2 hybrid plants was 16.2, ranging from 6 to 27, mth a practically normal distribution (fig. 3). The only character outside the group showing a Significant correlation with height of sucker is secondary branches. The correlation is in the direction of a coherence. 3C? "i^/o -| j- -1 □: TT SUCKER INDEX //£yG//r W 0£C/Af£7F/lS Pig. 3, — Height of sucker: fre- quency distribution of plants in Fa. Class value, 2 drm. 30 /o This character was determined by dividing the height of the tallest sucker by the height of the plant and multiplying by 100. It is thus the height of the tall- est sucker expressed as a percentage of the height of the main culm. This measurement was taken as the best single expression of the tendency to pro- duce tall suckers. Since Tom Thumb almost never produces suckers, the index is practically zero for the male parent of the hybrid. In Florida teosinte the index is usually about 100. In one population of 87, the mean was 99.4, with a range from 90 to no. In the Fj hybrid plants the mean was 117, with a range from 50 to 460. The distribution (fig. 4) was unimodal and symmetrical with the ex- ception of a few stragglers probably represent- ing plants with abnormal main culms. The coherences outside the group are with. male secondaries, mixed alicoles, and length of intemode on third. The disherences are with three members of the height group, nodes on third branch, two of the tassel measurements, position of best spike, branch silking first, and days to pollen. There is thus more direct evidence of disherence than of coherence with this character. It should be remembered, however, that the negative correlation of sucker index with height is in a sense physical, since the -rir "hr PS/? cs/\'r Fig. 4. — Suckerindex: frequency distribution of plants in F2. Class value, 10 per cent. One plant at 230 and one at 460. April I, 1920 A Teosinte-Maize Hybrid 15 one is a function of the other. The other disherences may follow as secondary relations due to this correlation with height. CIRCUMFERENCE INDEX In a population of 87, the circumference of the culm of Florida teosinte averaged 61 mm. Under similar conditions the circumference of Tom Thumb was approximately 35 mm. The mean of the F2 hybrid plants was 56 mm. Since circumference is so closely associated with the general size of the plant, the circumference measurement was recorded as a percentage of the height of the plant, and the measure- ment is termed a circumference index. While in direct m-easurement the culms of teosinte are thicker than those of Tom Thumb, teosinte is much more slender. In circum- ference index a high value is therefore a variation toward the maize parent. The mean index of Florida teosinte was 2.7, that of Tom Thumb about 6.0. The mean of the hybrid plants was 4.5, with a normal dis- tribution (fig. 5). Circumference shows one significant and independent coherence, that with pollen to silk, and a disherence with male secondaries. 30 I I " -- Tin r NODES WITHOUT BRANCHES t\, Sj (*) V V<> '<^ k) Vi N n' ^ Si « A>S/?C£A/r Fig. 5. — Circumference index: fre- quency distribution of plants in F2. Class value, s per cent. This character is the number of nodes be- tween the lowest branch and the uppermost sucker or the surface of the ground. In teosinte, branches are normally developed in the axils of all leaves on the main culm, except the uppermost. The tendency to suppress branches at the nodes just above the ground appears, however, when the plants are grown under unfavorable conditions. In a planting of Florida teosinte at Chula Vista in 1918 the average number of nodes ^vithout branches was 7.6. In maize there are always a number of nodes without branches between the uppermost sucker and the lowest ear. In Tom Thumb where no suckers are developed, the number can not be definitely determined, since the surface of the ground can not be located with accuracy. But since the average total number of leaves in Tom Thumb is 11 and there is an average of 3 nodes above the single ear and about 5 nodes below the surface of the ground, the medn number of nodes without branches is about 3. i6 Journal of Agricultural Research Vol. XIX, No. I The mean number of nodes without branches in the Fj hybrid plants was 1.05. The distribution (fig. 6) was far from normal, and there is some indication of two modes. Seventy-nine of the individuals were at zero. Of the remaining 41 plants the largest number, 12, had 3 nodes without branches, with a fairly uniform distribution ranging from i to 9. The significant correlations outside the group were coherences with both of the characters in the male branch group, number of suckers and length of internode on third. The disherences are with secondary branches and days to pollen. SO ^O \ \ /o Iddx NODES above; group NODES ABOVE In teosinte of all varieties there is almost with- fj out exception one node above the uppermost ^^^H 1 branch. In maize the number varies from 8 or 9 to 2 or 3; only in rare and abnormal specimens is it reduced to one. The limits as observed in the Tom Thumb variety are 3 and 5, with the mean at 3.4. This character, while not so con- stant as total leaves, is less subject to environ- mental influences than most of the characters recorded. There is some question of the propriety of con- sidering the number of nodes above the ear in maize as strictly homologous with the number of nodes above the uppermost branch of teosinte. In maize the uppermost branch, or ear, is nor- mally the best developed, while in teosinte the most fruitful branch is usually the third or fourth from the top. See Tables I and II. If the uppermost branch in teosinte is not homologous with the uppermost branch or ear in maize, the complete absence of any trace of a bud in the axils of the leaves above the upper ear in maize calls for some explanation. It is difficult to believe that branches in the axils of the upper leaves of maize could have be^n so completely suppressed as to leave neither a trace nor a tendency to reappear as an abnormality. It appears more reasonable to assume that in maize additional nodes have been intercalated or that these sterile nodes in maize, instead of i;epre- senting a change from the condition found in teosinte, have been derived from a distinct ancestor. In the first generation there were two plants with one node above and three with two. The range in the second generation was from one to four, with one possibly abnormal plant with none. The distribution Fig. 6. — Nodes without branches : frequency dis- tribution of plants in Fj. Class value, one node. April I, 1920 A Teosinte-Maize Hybrid 17 (fig. 7) is decidedly skew, more than half the plants having one, but there is no indication of bimodality. In maize there is always an intravarietal correlation between nodes above and total number of leaves and other characters that are expres- sions of size. Since Tom Thumb maize is smaller and has a much smaller number of leaves than teosinte, coherence with these size characters would not be masked by physiological correlations. It is therefore interesting to note that the tassel characters, which in pure strains of maize are positively correlated with nodes above, are here negatively correlated, affording clear evidence of coherence. There are also significant correlations in the direction of co- ^^ herence v/ith the male branch characters and node I silking first. There are no significant disherences. -5^ NODES ABOVE ON THIRD SO I < /o In all varieties of teosinte the number of nodes above the uppermost secondary on the third branch is one, as on the main culm. In maize the value will depend on what is considered the homologue of the third branch from the top in teosinte. Taken strictly, the upper branch in maize is the ear, and the third branch from the top, v/hen such exists, would be an earlike branch that in some types vv^ould partake somewhat of the nature of a sucker. If sufficiently suckerlike, the number of nodes above the uppermost secondary of such a branch would correspond to those of the main stalk, that is, the range v/ould be from 3 to 8. If, how- ever, the ear of maize be assumed to correspond to some branch below the uppermost in teosinte, those above the ear having been suppressed, the number of nodes above the uppermost secondary would be much greater, for in this case the branch would be an ear and the secondary branches would be the secondary ears which almost invariably are borne in the axil of the lowest husk. In any case the number would be larger in maize than in teosinte. This character was recorded for three of the F^ plants. In two of these the number was i; in the other it was 2. The average number in the Fj hybrid plants was i .78, with no indication of bimodality. The distribution (fig. 8) is much less skew than for the nodes above on the m^hi^cjim, the mode being at 2. In its correlations, this character is similar to nodes above on the main stalk, 164175°— 20 2 ffl Or/VOO£S Fig. 7. — Nodes above: frequency distribu- tion of plants in F2. Class value, onenode. Journal of Agricultural Research Vol. XIX, No. I NODES ON THIRD eo n<5/ 47 30 This character is instructive chiefly as a means of throwing light on the homologies between the branches of teosinte and maize and as a means of calculating the average length of internodes on the third branch described below. If each husk on an ear of maize represents a node, the third branch from the top, which would still be earlike, would have even in Tom Thumb from 6 to 8 nodes. In other varieties this number would be even larger. On the other hand, if the leaves from which the husks are derived have been subdivided, thus increasing the apparent number of nodes, the number of nodes on this branch of the hybrids might be expected to agree pretty closely with the number in teosinte, which varies from 2 to 5- The modal number in the Fg hybrid plants was 6, the mean was 7.15 , with a range from 4 to 14. There was no indication of bimodality (fig. 9). There k 2c\ I ! -1 ^^^^ ^° indication in the hybrid plants that leaves were subdivided, each leaf being borne on a well-defined internode. The increased number of nodes over that of teosinte goes to support the idea that each of the husks on an ear of maize represents an internode of the branch. In common with the other charac- ters of this group, the correlations with secondary branches and pro- phyllary spikes would seem signifi- cant coherences. In these correla- tions a high value of one character is correlated with a low value of the other, and any general tendency to vigor would reduce the correlation. The correlations with position of best spike and node silking first are of doubtful significance, there being an obvious physiological connection between these characters and the number of nodes in the third branch. All the significant disherent correlations are of a nature that suggests a physiological explanation. I o tu X C\) O) ^}.lo (o OFfl/OOES- Fig. 8. — Nodes above on third : frequency distribution of plants in F2. Class value, one node. 30 \ \ ft I QnJ AfUM3Sff OFA/ooes Fig. 9.— Nodes on third: frequency distribution of plants in F2. Class value, one node. April I, 1920 A Teosinte-Maize Hybrid 19 TASSEL GROUP SO ^20 \ I r PRIMARY BRANCHES Primary tassel branches are much more numerous in teoslnte than in any but the very largest varieties of maize. In the Tom Thumb variety the maximum number observed was 9, and this falls far below the number in any normal teosinte plant. The mean number for Tom Thumb and Florida teosinte grown under similar conditions was 4.6 and 12.5, respectively. In the Fi plants the number ranged from 16 to 20. In the second generation the mean was 16.7 with a range from 5 to 29. The distribution (fig. 10) was symmetrical and unimodal. There are two significant independent co- herences, one wath characters of the height group and the other with days to pollen. There are also two disherences, one with number of single female alicoles, the other with length of internode on third. The apparent disherence with sucker index is probably associated with the nega- tive correlation of sucker index with the other height characters. rr o\2 -. 'wiii SECONDARY BRANCHES Teosinte has a much larger number of second- ary tassel branches than maize. The specific may overlap, but the Tom Thumb variety Fig. 10. — Primary branches: fre- quency distribution of plants in F2. Class value, two branches. 20 /O ranges of the parents seldom develops sec- ondary branches, while in Florida teosinte the mean number was 10.3. In the F2 hybrid plants the mean was 12.6, with a range from o to 46. The distribution (fig. 11) is very skew, the mode being near 8, but there is little evidence of more than one mode. This character shows more evidence of coherence than does the character primary branches. It is closely correlated with three of the measurements of height. There is the same negative correlation with sucker index; and in addition the positive correlation with nodes without branches, which Is in the direction of a disherence, is here above 0.25. klflta D-d ?f^ ^^^ Fig. II. — Secondary branches: frequency distri- bution of plantsin F2. Class value, two branches 20 Journal of Agricultural Research Vol. XIX. No. I A most striking example of coherence is the negative correlation of secondary branches with all three of the nodes above group. The other coherences are v.ith male branches, branch silking first, and days to pollen. The only clear evidence of disherence is with male secondaries and length of internode on third. 1 FiG. 12. — Secondary index: fre- quency distribution of plants in F". Class value, 20 per cent. One plant at 340. SECONDARY INDEX This character, which is the number of secondary branches expressed as a percentage of the total branches, distinguishes more sharply between maize and teosinte than does the direct measurement of either primary or secondary branches. Second- ary tassel branches are relatively as well as abso- lutely much more numerous in teosinte than in maize. In teosinte they equal or exceed the number of primary branches, while in maize the number of secondaries equals the number of the primaries only in some of the large tropical varieties. In the F2 hybrid plants the mean was 70, with a very skew distribution (fig. 12) but with no evi- dence of more than one mode. The correlations are similar to those with the direct measurements of tassel branches, except the additional coherences with number of alicoles and rows in central spike. TASSEL BRANCHES ON TIirRD In teosinte the modal number of tassel branches on the third branch from the top is two. When teosinte is grown under rather unfavorable conditions where the number of branches is reduced, there Is evidence of a bimoda] distribution, in that plants with two branches or none are more numerous than plants with a single branch. In maize the number is zero, since all branches from the upper nodes of maize are normally un- branched. In the F2 hybrid plants the mean was 6.1. The distribution (fig. 13) was skew, with slight indication of two modes. Although closely correlated with the tassel characters of the main stalk, this character shows no significant correlations outside the group except with number of nodes on third. This is in the direction of a disherence; but the relation is doubtless physiological, since both characters would be similarly affected by changes in general vigor. Fig. 13.— Tassel braucheson third: frequency distrib- ution of plantsinFi.CIass value, two branches. April 1, 1920 A Teosinte-Maize Hybrid 21 The five tassel characters form a closely correlated group. With few exceptions all the members of the group show similar relations with other characters. With the exception of secondary index, all are direct measurements that might be expected to increase with increased vigor; and were it not for this character the correlations with the direct measure- ments in the height group might be considered physiological. The same may be said of the nodes above group. The disherent correlation of tassel branches on third with nodes on third branch is also physiological, since a highly developed third branch would naturally have a larger number of tassel branches. The clearest evidences of coherence are the correlations of secondary index with rows in central spike and that between secondary branches and branch silking first. Disherence is indicated by the negative correla- tions between all the tassel characters and the two characters male secondaries and length of internode on third. MALK BRANCH GROUP MALE BRANCH INDEX P 30 \^ ^W^ 4ff This character Vv-as calculated by dividing the num- ber of branches terminating in staminate inflorescences, excluding suckers, by the total number of leaves and multiplying by 100. It is thus the number of male branches expressed as a percentage of total leaves or internodes of the main culm. In normal maize none of the branches above the suckers bear staminate flowers, although staminate tips and perfect flowered spikelets are common abnormal- ities. In teosinte all primary branches normally end in a staminate inflorescence. There is, then, no over- lapping of either of the varieties or species with respect to this character. The F2 hybrid plants ranged from o to 71 with the mean at 37. The distribution (fig. 14) is practically symmetrical and clearly unimodal. There are four significant correlations, all in the direc- tion of coherences. They are with nodes without branches, nodes above, alicole index, and branch silking first. Except in the correlation with alicole index, a physiological explanation is suggested. MALE SECONDARIES As a measure of this character, all secondary branches on the third branch from the top of the plant that bore staminate spikelets were counted. Fig. 14. — Male branch index: frequency dis- tribution of plants in F2. Class value, lo per cent. 22 Journal of Agricultural Research Vol. XIX, No. I SO ■20 In normal maize there would be no secondary branches bearing stami- nate spikelets. In Florida teosinte the number is usually 2 or 3. In the F2 hybrid plants the mean number was 2. The range was from o to 8. Nearly half the plants had no staminate secondaries and there is almost no indication of a second mode (fig. 1 5) . Nearly all the significant correlations not readily assignable to physiological relations are disherent. Thus height, total leaves, and circumference index in the height group, secondary branches and secondary index in the tassel group, branch silking first, and days to pollen all show disherent correlations. Many of these are related, since they would be similarly affected by changes in vigor, but it is difficult to understand why increased vigor should result in a smaller number of male secondaries; and the nega- tive correlation with secondary index is difficult to understand as other than genetic. The absence of correlation between male second- aries and male branch index, which are placed in the same group because both are measures of the tendency to produce staminate spikelets, is in itself an indication of disherence. \/o CHARACTERS OF THE PISTILLATE INFLO- RESCENCE ALICOLE GROUP To discuss the characters of the pistillate inflo- rescence of the hybrids between maize and teosinte, a short preliminary description is necessary. In maize both staminate and pistillate spikelets are borne in pairs. In the pistillate inflorescence each pair of spikelets occupies a pit or alveolus In the staminate inflorescences there is only a faint suggestion of an alveolus. In teosinte the arrangement of the staminate spikelets is like that in maize; but in the pistillate inflorescence the spikelets are borne singly, each occupying a highly specialized alveolus. In hybrids of maize and teosinte, all permutations of the above arrangements occur, and to facilitate description the term alicole is used for the spikelet or spikelets arising from a single alveolus or having a common origin. Thus an alicole may consist of one or more staminate spikelets, one or more pistillate spikelets, or both pistillate and staminate spikelets.^ J For a more complete discussion of tlie pistillate inflorescence of teosinte and maize hybrids see Collins, G. N. structurs of thb maize ear as indicated in zsa-euchlaena hybrids. In Jour. Agr. Research, v. 17, no. 3, p. 127-135, i fig., pi. 16-18. 1919. ■yse m - Fig. 15. — Male secondaries: frequency distribution of plants in F2. Clas s value, one branch. April 1, 1920 A Teosinte-Maize Hybrid 23 D"" /c In normal maize the number of rows of alicoles is always half the num- ber of rows of grains. In the hybrid plants, however, 4-rowed spikes may consist of either two rows of alicoles, each with two seeds, or four rows of alicoles, each with a single seed. Plants exhibiting the range of variation with respect to the pistillate inflorescence are shown in Plates 2 to 5. As a basis of comparing the pistillate inflorescences of the hybrid plants, the best-developed spike on the third branch from the top of the plant was chosen and the number and nature of the alicoles were recorded. To eliminate as far as possible differences associated with the size of the spike, the number of alicoles of the classes single male, double male, single female, double female, and mixed (one male and one female) was ex- pressed as a percentage of the total number of alicoles in the spike. DOUBLE MALE ALICOLES Neither maize nor teosinte normally produces male spikelets in the pistillate inflorescences. In the Fj hybrids, however, out of 123 plants in which the nature of the pistillate inflorescences was determined, 18 had some alicoles with two staminate spikelets, in 2 plants the number being as high as 50 per cent (fig. 16), 20 I td ^s/? C£/vr Fig. 16. — Double male alicoles: frequency distribution of plants in Fa Class value, 10 percent. f/O D //o \20 I Ld MIXED ALICOLES Mixed alicoles are not a character of either maize or teosinte. There were, however, 13 Fg hybrid plants with mixed alicoles, the highest percentage being 40 (fig. 17). SINGLE FEMALE ALICOLES Single female alicoles are a universal character of teosinte, while in maize no variety is known in which the seeds are not paired. Single female alicoles may occur in rare instances on a part of an ear of maize where the number of rows is reduced toward the tip. In the F2 hybrid plants, although there was practi- cally a continuous series from o to 100 per cent, there Fig. 17. — Mixed ali- coles: frequency distribution of piantsinFs. Class wcrc distiuct indications of a tendency to segregate into value. 10 percent, ^j^^ ^^^ parental forms, there being two modes, one at o, the other at 100 (fig. 18). The numbers at these two modes were 34 and 12, indicating that the maize character is dominant. 24 Journal of Agricultural Research Vol. XIX, No. I In this connection it should be recalled that in the Fj plants the alicoles of the pistillate inflorescence all bore two spikelets. •52^ SO I- ^5^ DOUBLE FEMALE ALICOLES Double female alicoles may be considered alle- lomorphic to single alicoles, but owing to the occurrence of plants with small percentages of double male and mixed alicoles the percentages are not exact reciprocals. There is, however, the same bimodality (fig. 19), the numbers indicating the dominant nature of this character. alicolE index 'X SO Fig. 18. — Single female ali- coles: frequency distri- bution of plants in Fs. Class value, lo per cent. With the idea that mixed and male alicoles were in the nature of abnormalities, the number of sin- gle female alicoles was expressed as a percentage of the combined single and double female alicoles. There were 36 plants with no single female ali- coles and 19 with no double female alicoles. If the individuals are separated into two groups at the low point in the bimodal curve, which is 50 per cent, the numbers are 83 below this point and 37 above (fig. 20). The tendency for either the single alicoles of teosinte or the double alicoles of maize to predominate is the nearest approach to Mendelian behavior among the characters recorded. The measurements of the alicole group form such a closely related series that their correlations may be discussed together. Significant coherences are shown with both characters of the male branch group and with number of alicoles, rows in the central spike, and number of suckers. The only significant disherence is between single female ali- coles and primary branches. Some of the coherences may be of a physiologi- cal nature, but the almost complete absence of any evidence of disherence with this group of charac- ters which most nearly approaches an alternative method of inheritance should perhaps be noted. I Fig. 19. — Double female ali- coles: frequency distribu- tion of plants in F2. class value, 10 percent. April I, 1920 A Teosinte- Maize Hybrid 25 NODES SILKING GROUP so \ \ I 1^ \ I y ESrDEPENDENT CHARACTERS POSITION OF BEST SPIKE In maize the pistillate spike is terminal on the branch. In teosinte there are usually a number of spikes of nearly equal size, the prophyllary branch usually producing spikes as large as any on the branch. In the F2 hybrid plants this character was de- termined on the third branch. The nodes were numbered from the base of the branch, the pro- phyllary branch being recorded as o. The range was from o to 9, with the mode at 3. The mean was 2.22. The distribution (fig. 27) was decidedly skew, but there was little evidence of more than one mode. In -its relation to other characters, this character is very irregular. The large number of disherent corre- lations may indicate that the terminal position of the pistillate inflorescence in maize is not inherited as a tendency for the lateral pistillate inflorescences to be located near the top of the branch. When secondary ears are developed in maize they are always near the base of the branch, and the expression of this tendency in inheritance may be the explanation of the apparently disherent correlations. NUMBER OP AUCOLES The number of alicoles in a well-developed spike of Florida teosinte is 7. In Tom Thumb maize the number is seldom less than 100. The number recorded in a normal ear of maize is 50, in a Peruvian variety from the region of I^aike Titicaca. This is, therefore, one of the characters in which there is no approach to overlapping in the parental species. The Fi plants had spikes with from 11 to 18 ailicoles. Fig. 27. — Position of best spike: frequency distribu- tion of plants in Fs. Class value, one node. Am Fig. 28. — Number of alicoles: fre- quency distribution of plants lowCSt in F2. Class value, three ali- coles. In the secpnd generation the range was from 7 to 40. The mean was 17.85 with nearly symmetrical distribution (fig. 28), the mode being at 16. 30 Journal of Agricultural Research Vol. XIX, No. I The significant coherences with characters in the alicole group aflford perhaps the most direct evidence that has appeared that the characters of the pistillate inflorescence tend to be inherited as a unit. The correlation with rows in the central spike is perhaps physiological. There are no significant disherences. NUMBER or' SUCKERS Florida teosinte is characterized by a large number of suckers or branches that arise from below or near the ground. In a population of Florida teosinte at Chula Vista, grown in 191 7, the average number of suckers was 14. Tom Thumb never produces suckers on normal plants, and no variety of maize has been studied that produces as many suckers as teosinte. The expression of this character is so dependent on environ- mental conditions, however, that statements regarding the range in maize would have little value. The most vigorous F^ plant produced II suckers. In the second generation the range was from o to 32, with the mode at 13 and the mean at 11.7. There is no evidence of more than one mode (fig. 29). There are, in all, three significant correlations with this character, nodes without branches, single female alicoles, and double female alicoles — all of them coherences. The first of these is obviously physiological, since a large number of suckers and a small number of vacant nodes are both expressions of a tendency to produce branches. The other two are practically Affer- ent expressions of the same character and indi- cate a coherence. BRANCH SILKING FIRST ^'O p I In recording this character the primary branches were counted from the top. In maize the uppermost branch is the first to silk, except Fig. 29.— Number of suckers: fre- \^ j-^re instances where the second ear may silk quency distribution of plants in , , . , e ^^ c ^ t ^ • ±. F2. Class value, two suckers, a day or two lu advancc of the iirst. in teosmte One plant at 32. ^-jjg fourth or fifth branch is usually the first to silk. This character therefore distinguishes sharply between the parents with respect to both the variety and the species. The F2 hybrid plants ranged from i to 5, with equal numbers at i and 2 . The mean was 1.9, the distribution (fig. 30) was skew and unimodal. With the height group there are two significant correlations, one a coher- ence with total leaves, the other a disherence with sucker index. This disherence doubtless results from the negative correlation between total leaves and sucker index. The partial correlation of node silking first April 1, 1920 A Teosinte-Maize Hybrid 31 with either total leaves or sucker index, with the other character constant, is less than three times the probable error. There are also significant correlations with all the characters of the nodes above group. These correlations are in a sense physical, since the value representing the node silking first must always be greater than the nodes above. In the male branch group there is a significant coherence with male branch index and a disherence with male branches on third. In addition there are significant coherences with secondary branches, position of best spike, and days to pollen. DAYS TO POLLEN 527 ^o y I s^ ^^^ I ^ ^ /€? nl D Although profoundly influenced by the environment, the length of time before pollen is shed is the best measure of the length of season required for develop- ment. Under similar conditions there are few varieties of maize that require so long a time to mature as Florida teosinte, and Tom Thumb is one of the earliest varieties of maize. The period for Florida teosinte under condi- tions similar to those where the hybrid plants were grown was 162 days, and for Tom Thumb 43 days. The average time for the Fj was 98 days. The F^ plants averaged 112 days, with a single mode at 96 days (fig- 31)- l^he earliest plant flowered U in 71 days, and the latest required ^ ^O i — H-i 1 1 65 days from the date of planting. With characters of the height group there are two significant coher- ences, height and total leaves, and two significant disherences, sucker index and nodes without branches. ^ The correlation with height is an especially strik- ing coherence, since the positive correlation is 0.47 while the same correlations in both teosinte and Tom Thumb are negative, being 0.46 and o.ii, respec- tively. Days to pollen and total leaves in teosinte have a correlation of 0.14, a correlation significantly Fig. 31— Days to pollen: fre- lower than the 0.79 of the hybrids. quency distribution of ^, ,. , ,. .,, , . , plants in Fj. Class value. The negative Correlation with sucker index ap- 10 days. pears to result from the negative correlation of sucker index with total leaves. The coherence with nodes above on third is barely significant and may be physiological. There are significant coherences with three of the four tassel measurements, and in Tom Thumb the three tassel measure- ments recorded are all negatively correlated with days to pollen. Fig. 30. — Branch silking first: fre- quency distribu- tion of plants in Fj. Class value, one branch. 32 Journal of Agricultural Research Vol. XIX, No. I There are also significant coherences with number of ahcoles and node silking first. The disherent correlations with male secondaries and length of intemode on third appear to be genetic. That with length of internode on third is the highest coefficient with 1 days to pollen. ^^ /^ POLLEN TO SILK ^ Maize is normally proterandrous. There are, ^ however, proterogynous strains of maize, and pro- terogynous individuals in almost any strain are not uncommon. Tom Thumb is normally proter- androus by about lo days, Florida teosinte ap- pears to be normally proterogynous. It has always been so in our experiments; and an examination of the fields at Clarcona, Fla., in 1914, showed the plants to be silking from 7 to 10 days before pollen. Durango teosinte, on the other hand, under most conditions is proterandrous. In both maize and teosinte this character is es- pecially susceptible to environmental influence. The Fj plants were decidedly proterandrous at both Lanham and Chula Vista. None of the Fg plants w^ere proterogynous, the proterandry rang- ing from o to 53 days, with the mean at 18.3. The distribution (fig. 32) was symmetrical and unimodal. There are but two significant correlations with this character, both coherences. These are with circumference index and position of best spike. \)^o The latter is in one sense physiological. • 1^ Fig. 32. — Pollen to silk: fre- quency distribution of plants in Fa. Class value, five days. 30 /O 1 > 1 LENGTH OP INTERNODE ON THIRD This character was determined by dividing the length of the third branch by the number of in- temodes. The branches from the upper nodes of a maize plant are much shortened. An accurate measure is impossible on account of the difficulty of accurately determining the number of nodes. In Tom Thumb it would, however, be somewhat less than i cm., and in normal maize plants of any variety it would scarcely exceed 3 cm. In a nor- mally developed teosinte plant the internodes of the third branch will average about 10 cm. This character was not recorded in the first genera- tion. In the F2 plants the mean was 10.9 cm. The range was from 2 to 22, with practically a normal distribution (fig. 33). Fig. 33. — Length of internode on third: frequency distri- bution of plantsinFj. Class value, 2 cm. April 1. 1920 A Teosinte-Maize Hybrid 33 This character might be expected to be closely related to number of nodes on third, since in a mathematical sense it is a function of that character. However, the correlation between length of internode on the third branch and nodes on third is —0.05. Length of internode on third shows a larger number of significant correlations than does any other character. A high expression of the character might be expected to be associated with increased general vigor, but many of the correlations are in the opposite direction. There is distinctly more evidence of disherence than of coherence. In fact, three of the four significant coherences may be physiological, while most of the disherences are not to be explained in this way. Especially significant are the pronounced negative correlations with height and total leaves. Only slightly less striking are the negative corre- lations with two of the tassel characters. DISCUSSION OF CORRELATIONS It would be very difficult, if not impossible, to determine with accuracy the number of independent correlations. The interrelation of the charac- ters is of a most intricate nature; and even if the data warranted the calculation of the partial correlations of each pair with all other characters constant, the facts would still be very inadequately represented. Cor- relations take no account of causation or the sequence in which characters are determined. It is clear that the values of some characters are directly influenced by others, the relation being causal in nature. Thus the number of total leaves acts as a limiting factor to the number of branches ending in male flowers, and the correlation of any character with number of branches ending in male flowers may to some extent follow as a secondary relation to the correlation between the character in question and total leaves. On the other hand, it is obviously absurd to reason that branches ending in male flowers may influence the total leaves; and to correct the cor- relation with total leaves by making branches ending in male flowers constant might represent a mathematical relation, but the determination would have no biological significance. An attempt was made to determine whether the more striking disherent correlations might result from the correlations of other interrelated char- acters. Thus height and the index of nodes silking on the third branch, which showed a disherent correlation of —0.46, were found to be mu- tually correlated with four other characters to an extent that would materially influence the correlation in question. The partial correlation of height and index of nodes silking on the third branch with all of the 164175°— 20 3 34 Journal of Agricultural Research voi. xix, no. i four correlated characters constant was found to be — 0.69. Such rela- tions must stand, therefore, as disherences so far as the recorded data are concerned. A study of the correlations shows that wilhin wide physiological limits there are no incompatible combinations. On the other hand, all the characters are in a sense interrelated. Having in mind the theory that ascribes the determinants of characters to definite locations on the chromosomes, the authors examined the correlations to determine whether there were groups of characters between which there were no significant correlations. No such grouping was apparent, and it was possible to arrange the entire series so that they formed a single group with no corre- lation lower than ±0.31. If the results of this experiment are interpreted in terms of the theory mentioned above, it follows from the blended character of the inheritance that practically all the characters result from the combined action of numerous factors. The failure of the characters to fall into groups the members of which are genetically correlated further indicates that the factors for the individual characters must be distributed in different chromosomes. CORRELATION AMONG DESIRABLE CHARACTERS Among the characters measured, a certain few are indicative of desir- able characteristics from the standpoint of a forage plant. The more important of those are (i) total leaves, indicative of the luxuriant foliage of the teosinte, (2) circumference index, a small circumfer- ence in proportion to the height indicating the slender, edible stalks of the teosinte, (3) nodes silking on third branch, indicating the pro- fuse production of seed of the teosinte, (4) number of suckers, indicat- ing the abundant production of forage of the teosinte, (5) male branch index, indicative of the numerous branches of teosinte, (6) number of alicoles in the best spike, indicating the large pistillate inflorescences of maize, (7) rows in the central spike, indicating the many-rowed inflo- rescences of maize, and (8) days to pollen, a low value indicating the short season of maize. The interrelation of these selected characters is shown in Table VI. Of the 27 combinations of these characters there are 9 in which both of the desired characters are possessed by teosinte, 3 in which both are possessed by maize, and 15 where it is desired to combine teosinte and maize characters. April I, 1920 A Teosinte-Maize Hybrid 35 Table VI. — Correlation of characters desirable in a forage plant"' Characters considered. Circum- ference index. Nodes silking on third. Number of suck- ers. Male branch index. Number of ali- coles. Rows in cen- tral spike. Days to pol- len. Total leaves -0.31 0.05 . 00 — 0. 10 - • 19 . 01 — 0. 07 - .19 • 14 -0.18 .12 .05 - .11 .Oti -0.14 - .03 - .03 .02 - .04 •37 0.79 Circumference index — .17 Nodes silking on third. . .07 Number of suckers. . . . - .09 Male branch index .03 Number of alicoles - .29 - .09 Rows in central spike <* Figures in bold-face type indicate coefficients of correlation between the characters where a combina- tion of teosinte and maize characteristics is desired. Of the 15 character pairs where new combinations are desired, there is only one significant correlation. This is days to pollen and total leaves. In this one instance the relation is in a sense physical, sirce there is obviously a physical limit to the number of leaves that can be developed in a very short season. The indications from this comparison are, there- fore, that coherence presents few obstacles to the securing of desired combinations. (PI. 2; 6, A, B.) Another view of the comparative independence of the characters may be gained by an examination of the plants that were most like maize or teosinte with respect to some of the more important characters. Table VII is provided to make this possible. Bach pair of columns gives the measurements for two plants, one of which was the most like maize and the other the most like teosinte with respect to the character named at the head of the column. Table VII. -Comparison of individual plants, showing the extreme variations toward maize and teosinte, respectively o- 1 Cir- Num- Num- Height. Total leaves. Height of sucker. Suck- er index. cum- fer- ence index. Male bran- ches.'' ber of suck- ers. ber of ali- coles. Days to pol- len. Rows of ali- coles. ^ X -^ ^ Characters considered. CTl a r^ a t^ a 00 4-» a S 1 a 0 n 00 *•* "3. h* a '« 4-1 n 0 a g "a 4J 3 "5, g ■5. a "a a a "a a "a a "a a -n a 0) a V 0, V u « a V a 01 a m 0. t) Q. CI 0, u 3 (I) ^ a O'l q > S ■t 0 ■3 0 .a C3! 0 "a 0 '3 0 "rt 0 •S 0 .a S 'nt < s H S H s H 14 23 6 H 4 IS. 8 H 17 23 17 % H 9 19 0 14 23 tA H 17 9 23 33 14 19 j-I Height 14 4 7t 14 14 'I Total leaves 3i 21 ,^8 -6 29 7 38 21 24 Height of sucker Tft 20 17 21 27 tS t6 T9 11 Sucker index ri? 460 Qn I-?n ItXI 140 •SO 4ftn 7^0 100 roo 0 no ITO no 140 on Tin no Circumference index 4-'; .l-S 4.4 4.0 4.0 9.1 2.4 4.4 l-S 4-1 ■;.■; S.o 1'S 4.0 S.o Male branches 6 K 9 14 5 8 5 4 S IS 5 4 9 14 tS 9 14 II n S 0 6 9 20 6 5 8 5 1 7 IS Number of suckers 12 21 7 II 2 7 23 18 IS Number of alicoles 18 iq 17 17 14 21 iS 21 IQ 21 21 29 14 17 9 40 7 12 17 30 9 Days to pollen 103 2 T6r .03 103 2 142 3 129 2-3 138 2 131 2 AV IAS 85 6 Tift 2.1 2-3 3 2 2 2 2 2-4 3 4 2 2-3 ?, 1 Each pair of columns gives the measurements of two plants, one of which was most like maize and the other most like teosinte with respect to the character given at the head of the columns. The value of the character for whii-h the plant was selected is given in bold-face type. For description of units of meas- urement, see p. 7-8. ^ The number of primary branches that terminate in a stamiuate panicle, exclusive of suckers. 36 Journal of Agricultural Research voi. xix, no. i It may seem that, except for the character chosen, the values for the most part depart little from the mean values. For example, under total leaves the most maize-like plant which had 13 leaves was partic- ularly maize-like in no other character. It was even below the average in number of alicoles in the best spike and had almost the maximum number of suckers. On the other hand, the plant with the greatest number of leaves had also the greatest number of male branches but was decidedly maize-like with respect to number of suckers and number of alicoles. CONCLUvSIONS The genetic relations of the principal characters of maize and teosinte were investigated in a cross between a small variety of pop com and Florida teosinte, a large forage grass generically distinct from maize. The Fj plants showed characters which, for the most part, were interme- diate between those of the parents. The F2 plants were also intermediate, with a greatly extended range of variation. Thirty-three of the characters that differentiate the parents were chosen and recorded for each of the 127 Fj plants. The distribution of these characters with one or two exceptions showed little or no evidence of alternative or Mendelian inheritance. With respect to the individual characters, the extreme variants approached, and in some instance exceeded, those of the parents; but none of the plants possessed any large number of the characters of either maize or teosinte. The results showed the greatest freedom of recombination. All com- binations of characters appeared that might reasonably be expected with so limited a number of individuals. There were many instances of coherence or partial coupling, but there was an almost equal number of instances where characters derived from different parents showed a tendency to combine more frequently than would be expected as the result of chance. In such a complicated series it was found impossible, how- ever, to distinguish primary from secondary correlations. While there appeared to be no incompatible combinations, there were, on the other hand, no completely independent characters. Every character recorded showed significant correlation with one or more other characters; and these in turn were correlated with still others, with the result that all the characters were interrelated and formed a single group. It is possible, in fact, to arrange all the characters in such a way that they form a single group in which there is no coefficient of correlation lower than ±0.31. The nearest approach to Mendelian inheritance was shown by the arrangement of the spikelets in the pistillate inflorescence (fig. 18, 19, 20). In maize the female spikelets are borne in pairs (double female alicoles) ; in teosinte the female spikelets are borne singly (single female April 1,1920 A Teosinte-Maize Hybrid 37 alicoles). Dominance of the maize character was complete in the first generation. In the second generation the segregation was not complete, there being many plants with both single and double female alicoles; but the number of individuals in which double female alicoles predominated was approximately three times the number in which there were more single female alicoles. It was found that the characters of the pistillate inflorescence were subdivided in transmission to a remarkable degree. Thus the maize ear, instead of behaving as a unit, was subdivided into a large number of separately inherited units, such as number of rows, closely crowded seeds, and shortened peduncles, all of which were inherited more or less independently. Number of rows was still further resolved into paired or single spikelets and the number of rows of alicoles in which they were borne. A surprisingly large number of the plants combined the abundant production of suckers characteristic of the teosinte parent with the sturdy, upright character of maize and resulted in very leafy, compact plants of a type that should prove valuable for forage purposes. (See PI. 6, A.) It remains to be seen whether the new combinations can be maintained and made to breed true. The results of previous experiments with maize hybrids would indicate that selection for a few generations will fix any desired combination. PLATE 1 A. — General view of Fg plants of teosinte-maize hybrid. B. — F2 plants of teosinte-maize hybrid, showing diversity in size and season. (38) A Teosinte-Maize Hybrid Plate I Journal of Agricultural Research Vol. XIX, No. 1 A Teosinte-Maiz8 Hybrid Plate 2 Journal of Agricultural Research Vol. XIX, No. 1 PLATE 2 Teosinte-maize hybrid: A.— F2 plant No. 36. This plant bore the most maize-like pistillate inflorescence that appeared in the second generation. B.— F2 plant No. 49. The pistillate inflorescences of this plant were among those most nearly resembling teosinte. PLATE 3 Teosinte-maize hybrid: Pistillate inflorescence of Fj plant No. 36, shown in Plate 2, A. Natural size. A Teosinte-Maize Hybrid Plate 3 ^1 1 '^^■T^F ,4i. i^Ji* IEHk ^■ ^^^ ^dfeUiw ' jf^ n ^Si J^ 1 m ^^H^^v^^^^^H S& C m ^ ji*S ~ »i.^ W 0m w m ^^^^T^Wr 0^ ^ggy Ht'^^^^^^H g^^at ^'^r^^ai K>^*^H wSm K'^vfl K^.^H ^>^^H ^mr-c-^M K^v^^H ii>-^.':^H ▼ ^ "V Journal of Agricultural Research Vol. XIX, No. 1 A Teosinte-Maize Hybrid Plate 4 Journal of Agricultural Research Vol. XIX, No. 1 PLATE 4 Pistillate inflorescence of plant No. 49, shown in Plate 2, B. PLATE 5 Pistillate inflorescences from plant No. 94, illustrating an intermediate type of inflorescence. The arrangement of the alicoles is much like that of teosinte, but 90 per cent of the alicoles are double female. A Teosinte-Maize Hybrid Plate 5 Journal of Agricultural Research Vol. XIX, No. 1 A Teosinte-Maize Hybrid PLATE 6 Journal of Agricultural Research Vol. XIX. No. 1 PXATE 6 Teosinte-maize hybrid: A. — F2 plant No. 31, showing compact growth characteristic of many of the plants. Althongh only 14 dcm. high, this plant had 30 leaves on the main culm, nearly equaling teosinte in this respect. The plant resembled maize in having no spikes developed in the axil of the prophyllum. B. — F2 plant No. 113, showing stiff, erect leaves. This plant resembled teosinte in being very late in maturing, yet it was among the most maize-like with respect to circumference index. C. — Fj plant, grown at Lanham, Md. PLATE 7 Teosinte-maize hybrid: Pistillate inflorescences of the Fj plant shown in Plant 6, C. ATeosinte-Maize Hybrid Plate 7 Journal of Agricultural Research Vol. XIX, No. 1 BANANA ROOT-BORER By G. F. Moznette; ^ Entomological Inspector, Tropical and Subtropical Fruit Insect Investigations, Bureau of Entomology, United States Department of Agriculture INTRODUCTION The existence in Florida of a root-weevil peculiar to the banana was brought to the writer's attention in December, 191 7, by the receipt of some specimens from a grower near Larkins, in Dade County, Fla., who advised the writer of serious damage to his banana plants. The insect was determined by Dr. W. Dwight Pierce at Washington, D. C, as the banana root-borer. Cosmopolites sordidus Germar, a dangerous banana pest prevalent in almost every section where bananas are grown for commercial purposes. Since this species and all plants infested with it had been declared to be public nuisances in Florida, the State Plant Board at Gainesville, Fla., was immediately notified, and eradication and inspection work was begun. It was during the eradication and inspec- tion work that the writer, cooperating with members of the State Plant Board, was enabled to make a number of observations on the habits of this species; and it was thought well to publish the following data to aid others who may find this pest of the banana in the State of .Florida or wherever bananas are grown. A national quarantine was placed on this species April i, 1918. This quarantine forbids the importation into the United States from foreign coantries where the banana root-borer exists of all species and varieties of banana plants (Musa spp.) or portions thereof, except for experimental and scientific purposes. The spread of the insect from one country to another is probably accomplished by the transportation of infested suckers for planting (jj, p. 33-34) f and its spread within any locality most likely follows the killing out of infested stools, after which the adults travel in search of fresh supplies of food plants. Within a locality they could also be spread by the transportation of infested suckers or young plants for propagation. HISTORY AND DISTRIBUTION The adult (PI. 8) was described as Calandra sordida by Germar (6) in 1824. The genus Cosmopolites was established for this species by Chev- rolat {3) in 1885. E. Fleutiaux (5) recorded it from Madagascar in 1903, stating that it was a serious enemy of the banana on that island. In 1908 C. H. Knowles (9) mentioned carbon disulphid as a means of control in the ' Technical descriptions of the stages of the weevil by W. Dwight Pierce. - Reference is made by number (italic) to " Literature cited," p. 46. Journal of Agricultural Research, Vol. XTX, No. i Washington, D. C. Apr. i, 1020 tw (39) Key No. K.-S4 40 Journal of Agricultural Research voi. xix, no. i Fiji Islands. In 1912 H. A. Ballon (j, p. 112) reported the species as doing serious damage to bananas in the Lesser Antilles. During 1914 T. Fletcher (4, p. 342-343, fig. 201) published records of this species from southern India as existing in the regions of Malabar, Caimbatore, Godavari, and Ganjam. In the same year Frank P. Jepson (8), then working with the species in the Fiji Islands, where it is serious, made a mission to Java in quest of the natural enemies of the species and brought into the Fiji Islands some predatory beetles. He was suc- cessful in introducing some histerid beetles which were keeping the borers down in Java. Later in 191 6 Ballon {2) reported this insect as widely distributed in the Tropics, it being found in Jamaica, Guadeloupe, Dominica, Martinique, and Trinidad in the West Indies; Brazil in South America; and the Philippines, Fiji, Borneo, Sumatra, India, Queensland, and the Straits Settlements in the East. Besides the localities cited, Frank P. Jepson {8) in 1914 recorded addi- tional places where it is found: Java, Ceylon, New Guinea, Malacca, Saigon, China, Raratonga, Reunion, Sikhim, North Bengal, Pequ, Tenasserim, Andaman Islands, Sambak, and the Seychelles. In Florida investigations showed that the infested plantings at Larkins had all been made four years previous to the discovery of the weevils, with plants procured from a nursery in the northern part of Florida which had, in turn, secured the plants from a nursery in southern Florida. In March, 191 8, the weevils were found at the nursery in southern Florida, and every effort was made to exterminate them. It may be that many shipments of infested plants were made from this source, and it is very important that every occurrence of this pest be located and eradicated. Since the insect attacks sugar cane also it is not improbable that its presence would seriously interfere in the future with the development of large sugar and sirup industries. It is not known how this insect found its way into Florida, but no doubt it came in with sprouts or young plants introduced for propagation. HOST PLANTS According to published records there does not seem to be a great variety of host plants. Cosmopolites sordidtis apparently having confined itself thus far almost entirely to the banana, attacking all varieties. The borer has been reported, however, as attacking sugar cane. In Fiji, Jepson (7) states that the borer does not appear to display more partiality for one variety of banana than another. CHARACTER OF THE INJURY The young suckers attacked by the borers wither and die in a very short space of time. This is due to the feeding and tunneling of the grubs or larvae between the lateral roots and the bulb (PI. 11, B), thus cutting off the flow of sap to the plant. The banana plant has no central April 1, 1930 Banana Root Borer 41 tap root, but is supported by numerous lateral roots (PL 11, A). An indi- cation that a young plant is infested is the withering and drying of the curled roll of unopened leaves or growing part of the plant. The root, upon examination, is found to be riddled with the larvae of this insect and when cut open discloses the borer in situ. The adult weevils are abundant in the soil about the root and also are found under loose fiber surrounding the base of the stem, at the crown. They also congregate in the cavities caused by the larvae at the base of the bulb of the banana plant. In the planting at Larkins, Fla., where the infestation was first found, the writer collected 55 adults at the base of one plant and as many as 60 larvae and pupae in the bulb. The older plants infested ap- peared tall and spindling and no doubt succeeded in growing as much as they did by the presence of numerous lateral roots surrounding the bulbs of the plants and because the attacks of the insects had been gradual. Most of the bananas in the planting were old and so riddled by the lan.^ae as to be readily felled. After feeding thoroughly on a plant the weevils abandon it for another. TECHNICAL DESCRIPTIONS OF THE SPECIES The following descriptions by Dr. W. Dwight Pierce are based upon specimens collected at Larkins, Fla., January 19, 1918. The fine draw- ings accompanying the descriptions were made under Dr. Pierce's super- vision by Ml. Harry Bradford and by Dr. Adam Boving. EGG The egg is elongate oval, about 2 mm. in length, rounded at one end and more or less pointed at the other, and pure white in color. LARVA (Ply. 9, B-G) The larva is characteristically calandrid in form (PI. 9, B), having the eighth and ninth segments transformed into a sort of pygidial plate bearing very large elongated spiracles on the eighth segment (PI. 9, F, G). The other abdominal spiracles are all very minute and indistinct. The mesothoracic spiracles are very large. The length of a full-grown larva is at least 13 mm. (The writer has not had a live specimen to measure.) The body is white and the head shield dark reddish brown. The head is quite prominent. The head shield is broadly, elongately emarginate behind (PI. 9, C). From the center of the emargination on the median line the epicranial suture passes forward, separating the epicranium into two parts (PI. 9, C). This sutiue is strongly marked with black on its posterior half and is white from thence forward to the frons, behind which it divides and forms two frontal sutiu-es (PI. 9, D). The frons (PI. 9, D) is subtriangular, emarginate at anterior angles for the antennje, and emarginate along the epistoma for attachment of the clypeus. The median line is faintly indicated by a dark line in the basal half. The frons has two pairs of large setae and two pairs of tiny setae ; the three posterior pairs, the last of which is the small- est though the first is also small, form a triangle, the first and last pairs being almost equidistant. The anterior or epistomal pair of setae are large and are attached opposite the basal angles of the clypeus and some little distance from the antennal fosscE. 42 Journal of Agricultural Research voi. xix. no. i The epicranial areas are located on each side of the epicranial suture (PI. 9, C-E). A pair of light lines depart from the frontal sutures and pass backward almost as far as the light median line of the epicranium, corresponding to adfrontal sutures which sometimes occur in the Rliynchophora. Each lobe of the epicranium bears setae as follows: One at each terminus of the rudimentary adfrontal suture; a small one opposite the middle of the frontal suture, and a longer one behind this almost equi- distant from the epicranial suture ; a long hair opposite the basal third of the frontal suture; one opposite the middle of the pleurostoma; one near the hypostomal angle of the mandible; one opposite the basal third of the hypostoma; one on disk behind this; and four tiny ones on the disk near the median basal angle of the lobe. The antenna is a fleshy 2 -jointed appendage located at the lateral angle of the frons (PI. 9, D); the first joint is broad and short and bears one or more tiny hairs; thesecond joint is slender, finger-like, but short. The mandibles (PI. 9, D, E) are very dark brown, bidentate, with median and basal hairs. The clypeus (PI. 9, D) is attached in front of the frons and is basally marginated with dark brown, but otherwise light in color. It bears four tiny hairs on the epistomal margin. The lab rum (PI. 9, D) is not so broad, is rounded in front, has a row of four setse in front of the middle, and is mar- gined with setae. The maxillae (PI. 9, D, E) are elongate, terminated by a 2-jointed palpus and a setose lacinia. They are provided with four setae, two near palpus and one near base. The stipes labii (PI. 9, D, E) is triangular cordate, rather acutely angulate at base, bearing 2-jointed palpi at basal angles with a discal pair of setse and with several pairs of basal setae. The body is glabrous except for the usual hairs found on each segment (PL 9, B). The prothorax is not divided dorsally on the anterior margin, which corresponds to the praescutum. There are six pairs of setae, the last of which occiu-s in the region of the alar lobe. Behind these on the scutal area are four pairs of hairs, the last of which oc- curs on the alar lobe. The mesothoracic spiracle occurs on a large lobe which causes an emargination of the prothorax and lies very close to the head. It is very elongate with a longitudinal slit. The mesothorax and metathorax dorsally consist of a spindle- shaped praescutum with a single pair of setae and the scutellum, extending from alar lobe to alar lobe and bearing only two pairs of hairs in the region of the alar lobe . The epipleurum of the mesothorax and metathorax bears a single hair. Each hypopleural lobe bears two setae. The sternum of the thorax consists of a median area or eustemum and two lateral lobes more or less connected medianly behind the sternum. The median portion is the stemellum and the lateral portions are the parasternal plates. Each thoracic sternum bears one pair of hairs, and each parastemum bears three pairs of hairs. The first seven abdominal segments are normal, and each bears a very minute spira- cle. In a fully matured specimen these segments grow larger to the fourth or fifth segment and then decrease in size. The seventh segment is the smallest of the normal segments. Dorsally each segment is transversely divided into foiu* parts, praescutum, scutum, scutellum, and postscutellum. Each praescutum bears one pair of setae and each scutellum bears a small lateral pair. Each epipleural lobe bears two pairs of seta; and each hypopleural lobe is apparently longitudinally divided into two parts, the lower of which bears a single seta. Ventrally, each segment has two transverse lobes, the front one being the eustemum with the presternum depressed in front of it and the parastemum and lobe at each side. The second transverse area is transversely depressed and frontally consists of stemellum and poststemellum. There are no setae on the sternum of the abdomen. The eighth segment is dorsally greatly modified with a single pair of hairs on the praescutum and a single pair on the scutellar area, and with very elongated spiracles quite prominent (PI. 9, F, G). Just outside of the spiracles on the epipleural lobe are two strong setae. The dorsal face of the eighth segment is declivous (PI. 9, B) ; the dorsum of the ninth segment is flattened and bears four pairs of setae, as shown in the figure (PI. 9, F). The April 1. 1920 Banana Root Borer 43 dorsum of the ninth segment extends xmderneath, so that it is apical to the entire tenth segment. The tenth segment is completely ventral and very small. The tip of the abdomen showing the position of the tenth segment is illustrated in Plate 9, G. PUPA (PL. 10) Elongate, about 12 mm. long, white. This pupa is characteristically calandrid in the possession of very large thoracic spiracles located on a prominent lobe at the base of the prothorax (PI. 10, B). The beak is very irregularly margined with numerous transverse depressions (PI. 10, A). There are four pairs of large setae and two pairs of tiny setae on the head and beak. The foiu: larger pairs of setae are borne on tubercles, one on the head and three on the beak. The two pairs of tiny setae are located medianly to the two basal pairson the beak, as shown in the drawing. The prothorax (PI. 10, C) is rather elongate subquadrate, rounded in front, with basal angles rounded, and bears six pairs of setigerous tubercles, of which the apical pair are the largest. There are two antero-lateral, two postero-lateral, and one antero-median pairs of setae. The meso- thorax has one pair of scutellar setae. The first six abdominal segments are normal, and each bears three pairs of scutellar setEe. The first six abdominal spiracles are larger and more prominent than the larval spiracles. The seventh and eighth spiracles are minute. The first two ventral segments are very much crowded. The seventh, eighth, ninth, and tenth segments are greatly modified both above and below. Dor- sally the seventh segment is elongate, apically it is tuberculate, and it has two pairs of setigerous tubercles, one pair being on the larger apical tubercles. From a lateral view, it is seen that the seventh segment is dorsally the terminal segment , but ventrally it is surpassed by the other segments. In other words, it is laterally emarginate for the reception of the other segments, each of which includes the succeeding segment. The ninth segment is provided with a pair of very long, chitinous processes, corresponding to the cerci, at the side of which are two setigerous tubercles. Ventrally (PI. 10, A) the mesothorax is smallest, prothorax next, and metathorax next. The mesostemum is protuberant, the metastemum elongate and flattened. The coxae are spherical ; the femora are setigerous at the apex. The wing pads extend only to about the apex of the fouxlh abdominal segment. ADULT (PL. 8) Length 11 mm.; breadth at base of elytra 4 mm. Head small, spherical; beak separated from head by constriction, swollen in basal third, finely punctate in basal half; moderately curved, slender and cylindrical and smooth in apical half. Scrobes located in basal third beneath the swelling, oval, more approximate behind than in front. Gular suture extending almost entire length of venter of beak and head. Antennae geniculate, scape almost as long as funicle. Funicle 6-jointed, first joint moniliform, succeeding joints more closely appressed, last joint very closely appressed to club. Club 2-jointed, basal joint occupying two- thirds of the length, shining, with a few minute hairs; apical joint spongy, short, aad rounded at apex. Other funicular joints bearing a few tiny hairs. Eyes finely granulate, elongate oval, transversely contiguous beneath, anteriorly margined. Prothorax very long; moder- ately evenly punctate, with an irregular smooth median line indicated on disk; constricted near apex, apex tubular; narrowest at apex, roundingly broadening to about the middle; sides almost parallel from middle. Scutellum small, subquadrate, moderately short, with slight humeral angles. Stride moderately impressed, pimctate. Intervals of irregular width, the first, third, and fifth being slightly wider than the alternate intervals, minutely punctate. Pygidium almost vertical, spongy, pubescent, with setigerous punctures. Undersides more sparsely pimctate. Sternum flattened. Procoxae and mcsocoxae cylindrical, metacoxaeoval, trochanters small, femora laterally 164175°— 20 4 44 Journal of Agricultural Research ,'oh XIX, No. I compressed and curved, ventrally inflated at middle, emarginate beyond this and bilobed at apex, thus forming a groove for the tibise. Tibiae moderately straight, grooved beneath and provided with a row of setae on each side of the groove, apically curved downwards, terminating in a strong hook. Tarsi 4-jointed, first longer than broad, widest at apex, second about as long as broad, third about as long as first but broader at apex, emarginate for reception of fourth. Fourth elongate, curved, sub- cylindrical, armed with two curved, divergent claws. Intercoxal piece broad, angu- late. First two abdominal segments connate at middle. Third and fourth segments about as long as second. Fifth segment longer, turned downward. - LIFE HISTORY The female beetle having been fertilized enters between a leaf sheath and the stem and selects a spot for the deposition of an c^gg. The beetle then prepares a small cavity by means of the powerful mandibles located at the tip of the rostrum or beak. After having completed the cavity the beetle reverses its position and with the aid of the ovipositor deposits a single e^gg in the prepared place (fig. i). On February 9, 191 8, many eggs were observed which were laid apparently a short time previously in the tissues, usually in the small com- partments in the sheaths or stem. A few eggs were even found laid loosely in the slightly decayed leaf sheaths close to the healthy fleshy banana bulb, from which place they entered the bulb. The eggs, for the most part, are deposited singly in the sheaths near the crown at the surface of the soil. On hatching, the ^gg does not completely collapse. The larv' ae eat their way in all directions in the bulb, and one can easily trace a channel as it gradually grows wider, terminating in a pouch near the outer surface in which the larva pu- sJ^idus: se^a" 'of pates on reaching maturity. The records for oviposi- i^?omp^rtin«it."* ^'^ tion, hatching, and pupation are given in Table I. _J jmm "1 Table I. — Egg and larval records of Cosmopolites sordidus, igiS V.5Z No. deposited. hatcned. I^arva pupated. I Feb. lo ...do Feb. ic ...do... .. Mar. 2 Mar. 3 Do. Mar. 2 Mar. 3 Mar. 4 Mar. 3 Mar. 6 Mar. 5 Mar. 6 2 ...do ...do 4 ...do Feb. 16 ...do e ...do 6. . . .do Feb. 17 . . .do 7 ...do 8 ...do ...do 0 . . .do . .do . 10. . . ... ..do . .do April 1, 1920 Banana Root Borer 45 From a few experiments the egg period was found to last from 5 to 7 days. From the character of the channels of the grubs it is the opinion of the writer that the eggs are deposited in the outer sheaths or between the outer sheath and the stem, the grubs working their way into the body of the bulb or trunk. The work of the larva is particularly destructive, since they girdle the plant in the immediate vicinity of the lateral roots put out from the bulb of the plant (PL 11, A), thus cutting off the passage of the sap. The larvae not only work fre- quently in this region just described but may be found tunneling into the main trunk as far as the heartwood. The larvae usually work below ground, but in a number of instances the writer has found them in the trunk as high as 2 feet above ground. The larval stage was found to last over a period of from 15 to 20 days. Due to the scarcity of material and to the fact that all infestations were gradually destroyed and cleaned up, the writer was unable to make further records on the seasonal habits of the species. The larvae upon attaining maturity construct an oval space at the end of the burrows, usually vv^ell toward the outer layers, where the larval head is cast, and where the lar\-a pupates. The pupae are naked. Jepson found in Fiji that a period of from 5 to 8 days from the time of pupation elapses before the emergence of the adult. The adults bear wings and are very sluggish. When disturbed they Avill "play 'possum" for a considerable length of time. The adults are gregarious and w^ere found in clusters in cavities and depressions in the outer sheaths of the banana close to the surface of the ground and also below the surface. The length of life of the adult is not known. The writer has kept them in captivity without food for two months. Jepson in Fiji has kept the beetles in captivity about 14 weeks without food, and in the state of nature they undoubtedly will live longer. In all probability the banana root-borer continues to breed all the year round, provided that the food supply is plentiful. The beetles are nocturnal, only coming up from the soil at night for their activities above ground. CONTROL Since bananas are grovv^n year after year on the same land and are pro- duced from suckers springing from the parent plant, a plantation usually forms a breeding ground and nursery for these insects. The borer's mode of life renders it a difficult pest to control, as Knowles and Jepson (10) noted in Fiji. The egg, larval, and pupal periods are passed in or on the bulb of the banana or plantain. The adults apparently do not move far from the place where they have lived and developed so long as suitable food is available to attract the egg-laying females. In Java Cosmopolites sordidus is preyed upon and kept down by the larvae of a histerid beetle and by those of some beetle of the family of Hydrophilidae. Jepson intro- duced these species into Fiji, where the banana root-borer is a serious 46 Journal of A gricultural Research voi. xix, Nj. i pest. Where banana plants are found infested in Florida and elsewhere in the States they should be destroyed immediately, and traps should be laid by using strips of healthy banana trunks. In Florida strips of banana plants proved more successful as a trap than did young plants on an infested piece of ground. As the beetles congregate under and about these strips they should be burned and the process repeated until the beetles are eradicated. It is very important that the traps be renewed, since the beetles are capable of living a considerable time without food. LITERATURE CITED i) Ballou, H. a. 1912. INSECT PESTS OP THE LESSER ANTILLES. Imp. Dept. Agr. West Indies, Pamphlet Ser. 71, 210 p., 185 fig. 2) 1916. THE BANANA WEEVIL. In AgT. News [Barbados], v. 15, no. 364, p. 123. 3) Chevrolat, a. 1885. CALANDRIDES. In Ann. Soc. Ent. France, s. 6, v. 5, p. 275-292. 4) Fletcher, T. Bainbridge. 1914. SOME SOUTH INDIAN INSECTS ... 565 p., 440 fig., 50 col. pi. Madras. 5) Fleutiaux, E. 1903. LES insecTES. In Agr. Prat. Pays Chauds, ann. 2, no. 10, p. 495-502; no. 12, p. 745-760. 6) Germar, E. F. 1824. INSECTORUM species novae AUT MINUS COGNITAE DESCRIPTIONIBUS illustrataE. v. I (Coleoptera), 624 p., 2 pi. Halae. 7) Jepson, Frank P. 1911. REPORT on economic ENTOMOLOGY. Fiji Dept. Agr. Council Paper 25, 89 p.. 7 pi. I914. a mission to JAVA IN QUEST OF NATURAL ENEMIES FOR A COLEOPTEROUS PEST OF BANANAS (COSMOPOLITES SORDIDA, CIIEVR.). Dept. Agr. Fiji, Bul. 7, 18 p., 3 pi. 9) Knowles, C. H. 1909. INSECT PESTS. [the BANANA WEEVIL. (SPHENOPHORUS SORDIDUS)]. In Legisl. Council, Fiji, Agriculture (Report on, for the year 1908), p. 20. Council Paper 27. (10) and Jepson, Frank P. 1912. the banan.a. in Fiji. Dept. Agr. Fiji, Bul. 4, 17 p., 3 pi. (11) Pierce, W. Dwight. 1917. A MANU.\L OF DANGEROUS INSECTS LIKELY TO BE INTRODUCED IN THE UNITED STATES THROUGH IMPORTATIONS. 256 p., 49 pi. Washington, D. C. Published by U. S. Dept. Agr. Off. Sec. PLATE 8 Banana root-borer {Cosmopolites sordidiis): Adult. Banana Root-Borer Plate 8 Journal of Agricultural Research Vol. XIX, No. 1 Banana Root-Borer Plate 9 Journal of Agricultural Research Vol. XIX, No. 1 PLATE 9 . Egg and larva of banana root-borer: A.— Egg. B. — Larva, side view. C. — Head of larva, dorsal view. D. — Head of larva, face view. E. — Head of larva, side view. F.— Dorsal view of seventh, eighth, and ninth abdominal segments. G. — Posterior view of segments 7 to 10. PLATE lo Pupa and adult of banana root-borer: A. — Ventral view of pupa. B, —Lateral view of head and thorax of pupa. C. — Dorsal view of pupa. Banana Root-Borer Plate K Journal of Agricultural Research Vol. XIX. No. 1 Janana Root-Borer Plate II Journal of Agricultural Research Vol. XIX, No. 1 PLATE II A. — Young healthy banana plant bulb with lateral roots. B. — Young banana plant cut into, showing work of the larvae of the banana root- borer. Illustration shows how lateral roots become severed by grubs working near roots. 164175°— 20 5 Vol. XIX APRIL 15, 1920 No. 2 JOURNAL OP AGRICULTURAL RESEARCH CONXENXS Pag* Effect of Calcium Sulphate on the Solubility of Soils - 47 M. M. McCOOL and C. E. MILLAR (Cootrlbution fiom Michigan Agricultural Experiment Station) Further Studies on the Influence of Humidity Upon the Strength and Elasticity of Wool Fiber - - - - 55 J. L HARDY (Contribution from Wyoming Agricultural Experiment Station) Composition and Density of the Native Vegetation in the Vicinity of the Northern Great Plains Field Station - 63 J. T. SARVIS (Contribution from Bureau of Plant Industry) Effect of Reaction of Solution on Germination of Seed and on Growth of Seedlings ------ 73 ROBERT M. SALTER and T. C. McILVAINE (Contribution from West Virginia Agricultural Experiment Station) PUBUSHED BY AUTHORITY OF THE SECRETARY OF AGRICULTURE. WITH THE CO(M»ERATION OF THE ASSOQATION OF LAND-GRANT COLLEGES WASHINOXON, D. C. WAIHINarON i OOVCRNMKNT PRINTINa OPriOK ! KM EDITORIAL COMMITTEE OF THE UNITED STATES DEPARTMENT OF AGRICULTURE AND THE ASSOCIATION OF LAND-GRANT COLLEGES FOR THE DEPARTMENT FOR THE ASSOCIATION KARL F. KELLERMAN, Chairman Physiologist and Asiociate Chief, Bureau of Plant Ipdustry EDWIN W. ALLEN Chief, Office of Experiment Stations CHARLES L. MARLATT Entomologist and Assistant Chief, Bureau of Entomology J. G. LIPMAN Dean, State College of Agriculture, and Director, New Jersey Agricultural Experi- ment Station, Rutgers College W. A. RILEY Entomologist and Chief, Division of Ento- mology and Economic Zoology, Agricul- tural Experiment Station of the University of Minnesota R. L. WATTS Dean, School of Agriculture, and Director, Agricultural Experiment Station, The Pennsylvania State College All correspondence regarding articles from the Department of Agriculture should be addressed to Karl F. Kellerman, Journal of Agricultural Research, Washington, D. C. All correspondence regarding articles from State Experiment Stations should be addressed to J. G. Lipman, New Jersey Agricultiu"al Experiment Station, New Brimswick, N. J. JOMALOFACMaiLTlALlSEARCH Vol. XIX Washington, D. C, April 15, 1920 No. 2 EFFECT OF CALCIUM SULPHATE ON THE SOLUBILITY OF SOILS By M. M. McCoOL, Professor of Soils, and C. E. Millar, Associate Professor of Soils, Michigan Agricultural Experiment Station Additional information on the rate of formation of soluble salts in soils as affected by different factors is desirable. One phase of the sub- ject of special interest is the immediate and residuary effects of fertil- izing materials on soils. It seems that aside from its theoretical interest such information should be of assistance in accounting for results that are obtained from the use of certain substances under field conditions. We have interviewed several of the earlier settlers in southern Michigan and have been informed by them that calcium sulphate was used rather freely by some farmers during the earlier stages of the State's agricul- tural development. The general impression of those whom we inter- viewed is that the application of calcium sulphate resulted favorably for a time, increasing the yields of small grains and clover, but later on failed to bring the desired results ; hence this substance came to be looked upon as a soil stimulant. Some farmers are using it on wheat and clover, although the amount so consumed is relatively small. According to early reports, which are to be considered in a later publication, similar conditions existed in the agricultural regions of some of the eastern States. Because of the experiences of the early agriculturists, the increasing interest in the fertilizing value of calcium sulphate, and the widespread use of acid phosphate, which contains appreciable amounts of the sul- phate, it was considered advisable to investigate the effect of the sul- phate both alone and in junction with calcium phosphate on the formation of soluble salts in soils, as well as the effect on the carbon-dioxid production. The freezing-point method was used to determine the former, and the titration method the latter. METHOD OF PROCEDURE In determining the effect of the chemicals on the rate of formation of soluble salts, 200 gm. of the soils in question were brought into contact Journal of Agricultural Research, Vol. XIX, No. 2 Washington, D. C. Apr. is, 1920 tv Key No. Mich.-io (47) 48 Journal of Agricultural Research voi. xix, No. a with 500 cc. each of distilled water and the substance in solution, and allowed to stand 24 hours with several thorough agitations. At the close of this period the mass was transferred to filter paper in large fun- nels. In some cases the soluble salts were reduced to a very low con- centration by washing with distilled water, while in others the soils were removed, after drainage had ceased, but otherwise treated in the same way as in the former instances. The rate of formation of soluble substances in treated and untreated soils was determined under two sets of moisture conditions. The one which is here called low water content approx- imated the so-called optimum condition for plant growth; and the other, which is here called high water content, was secured by allowing i part of soil to 0.7 part of water and provided sufficient moisture to saturate the soil and leave a small column of about yi inch above it. Soils of the following description were used in all the experiments : Soil I, a silt loam, light phase, containing a large amount of organic matter. Soil 2, a heavy sand, rather low in organic content. Soil 3, a fine sandy loam with a medium supply of organic matter. Soil 4, a very fine sand containing a small amount of organic material. Soil 5, a very heavy silt loam with a very high content of organic matter. Soil 6, a silt loam well supplied with organic material. EXPERIMENTAL RESULTS The first series of experiments to be reported is the one in which the soils were treated with calcium sulphate, drained, and made up to the high water content, or i part of soil to 0.7 part distilled water. Treated and untreated portions of each of the soils studied were placed in jelly glasses, which were tightly covered and let stand in the laboratory. At about 4-day intervals they were thoroughly aerated by stirring, the cov- ers being removed for one-half hour or more. The soils employed were air-dry and had been stored in the laboratory about 160 days. The results are set forth in Table I. Table I .^Effect of calcium sulphate on the solubility of unwashed soils held at high water content for various periods Soil No. Condition of soil. Treated . . Untreated Treated. .. Untreated Treated . . Untreated Treated. .. Untraeted Treated. .. Untreated Treated. .. Untreated Freezing-point depressions. After : days. C. 042 003 044 000 043 002 04s 000 046 003 045 008 After 4 days. °C. o. 040 • 005 •045 . 004 .051 . 004 .050 . 002 .048 . 002 . 042 . 012 After 6 days. "C. 0.055 .GOS . 050 • o°5 • 054 . 006 .050 . 004 .051 . 004 .050 . 012 After 8 days. C. 058 on 055 007 057 007 053 006 055 006 051 or8 After 10 days. "C. 0.063 .013 •059 . 009 .058 .008 •055 . 010 .058 . 012 .052 . 024 After 30 days. 091 024 080 on 057 014 138 017 075 018 089 032 Apr. IS. 1920 Effect of Calcium Sulphate on Solubility of Soils 49 The effect of the calcium sulphate on the rate of formation of soluble salts in the soils investigated is appreciable. According to the data set forth in Table I, as well as other data not recorded, the reaction is rather gradual and prolonged. Of course the initial concentration of the solu- tions of the treated soils was high, and it is possible that this influenced the rate of changes which afterwards took place in the mass. It was considered advisable to wash the soils until the concentration of the solution in the soils was at a very low point. This was done, and the series of tests with washed soils was carried on at the same time and under the same conditions as the previous one. The results obtained are presented in Table II. An examination of this table shows that the residuary effect of the calcium sulphate on the rate of formation of soluble substances in the soils is remarkable. The changes in the concentration of the soil solution did not all take place at once but continued for a num- ber of days. Table II. — Effect of calchim sulphate on the solubility of washed soils held at high water content for various periods Soil No. Condition of soil. Freezing-point depressions. After 2 days. After 4 days. After 6 days. After 8 days. After 10 days. After 3c days. Treated. .. Untreated Treated. .. Untreated Treated... Untreated Treated. .. Untreated Treated... Untreated Treated. . . Untreated "C. o. on .003 . 002 . 000 . 000 . 002 . 000 . 000 . 001 .003 . 000 .008 "C. 0.015 .COS .005 . 004 . 004 . 004 .005 . 002 .015 . 002 .008 . 012 "C. o. 030 .008 . 012 .005 . 010 . 006 .018 . 004 .028 . 004 .013 . 012 "C. D. 044 . on . 024 . 007 . 016 . 007 . 022 . 006 . 042 . 006 . 014 .018 "C. 3. 071 .013 •057 . 009 .028 . 009 .058 . 010 .052 . 012 . 024 .024 °c. o. lOI . 026 ■ -073 . on .066 . 014 . 184 . 017 . 070 .018 .098 .032 A clay loam soil was treated with the calcium-sulphate solution, washed, and let stand 30 days at the high water content, again washed until the freezing-point lowering of the solution in the soil was 0.005° C, and again let stand 30 days. At the end of this period the freezing- point lowering of the control or untreated soil was 0.040°, and that of the treated soil was 0.102°. The residuary effect of the treatment is quite persistent. Another series was run in which the water content of the washed soils was lower, or approximately the so-called optimum point. The con- centration of the solution in the soil was not determined until the end of a 30-day period. At that time the freezing-point lowerings of the soils were great and not strikingly different from those of the high water series. The results of this experiment are given in Table III. 50 Journal of Agricultural Research voi. xix, no. a Table III. — Effect of calcium sulphate on the solubility of washed soils held at low water content for jo days Soil No. Condition of soil. Freezing- point depressions. I 2 3 4 5 6 Treated T. 0.103 .015 .085 . 010 . Ill .013 .163 . 007 .099 .023 . 096 .023 i Untreated Treated Untreated Treated Untreated Treated Untreated Treated Untreated Treated Untreated Inasmuch as acid phosphate contains both calcium sulphate and calcium phosphate, a series was run in which the soils were treated with a saturated solution of calcium sulphate, a N/io calcium phos- phate, and also a combination of the two. After treatment the soils were washed as in the series described above and let stand at the high water content 30 days. At the close of the period the concentration of the soil solution was determined by the freezing-point method. The results are given in Table IV. Table IV. — Effect of calcium sulphate and calcium, phosphate alone and in com-bination on the solubility of soils after jo days Kind of soil and treatment. Sandy loam: Treated with calcium sulphate Treated with calcium sulphate and calcium phosphate Treated with calcium phosphate Untreated Silt loam: Treated with calcium sulphate Treated with calcium sulphate and calcium phosphate Treated with calcium phosphate Untreated Freezing- point depressions. 'C. 3- 134 .094 .028 •035 . 096 .084 .032 . 042 A glance at the data composing Table IV reveals that the calcium sulphate in the presence of the calcium phosphate is somewhat less active in changing the rate of solubility of these soils than it is when used alone. Moreover, where the calcium phosphate alone is added to the soils the solubility is somewhat lessened. This is in accord with the results re- ported by Bouyoucos.^ > BorrYOUcos, George J. rate and extent op soltjeility op seas under dipperbnt treatments AND CONDITIONS. Mich. Agr. Exp. Sta. Tech. Bui. 44, 49 p. 1919. Apr. IS, 1920 Effect of Calcium Sulphate on Solubility of Soils 51 The results cited above immediately raise the question as to whether the great increase in concentration of the soil solution resulting from treatment with calcium sulphate is due to a stimulation of biological activities or to chemical reactions. To throw some light on this question experiments were undertaken in which the rate of production of carbon dioxid was measured. The method of procedure was as follows: The soils were allowed to stand over night in a saturated solution of calcium sulphate. They were then filtered and washed with distilled water until the concentration of the soil solution, when the soils were just saturated, was only a few parts per million. The soils were then allowed to dry. After thorough mixing, 60 gm. were weighed into 4-ounce bottles. The desired amount of water was then added and the bottles stoppered with rubber stoppers fitted with tubing so arranged that a current of air could be readily drawn through the soil. The bottles were stored in the dark at room temperature, and every 10 days the carbon dioxid was swept out by means of a current of air free from this substance, and the amount of carbon dioxid was determined. Samples of untreated soil were prepared in a similar manner and the carbon dioxid determined as outlined. Tables V and VI show the milligrams of carbon dioxid produced during the lo-day periods and also the total production for the 30-day period at the water , contents used. Table v.- -Effect of calcium sulphate on the production of carbon dioxid at high water content Soil No. Treatment. Carbon dioxid produced in — 10 days. 20 days. 30 days. Mgm. Mgm. Mgm. 3-74 9-24 7.26 8.14 9. 02 5-72 5- 06 7.48 7. 26 7. 26 8.80 7.04 5-94 9.90 8.14 7.92 11.88 9.68 5-28 7.04 5-94 3-74 7.04 7.04 9. 02 12. 76 13.20 5-94 10.34 8.80 9.24 12. 76 11.88 II. 00 13.42 10. 56 Total carbon dioxid produced. Calcium sulphate No treatment Calcium sulphate No treatment Calcium sulphate No treatment Calcium sulphate No treatment Calcium sulphate No treatment Calcium sulphate No treatment Mgm. 20. 24 22.88 19. 80 23. 10 23.98 29. 48 18.26 17.82 34.98 25.08 33- 88 34-98 52 Journal of Agricultural Research voi. xix. No. a Table VI. — Effect of calcium sulphate on the production of carbon dioxid at low "water content Soil No. Treatment. Calcium sulphate No treatment Calcium sulphate No treatment Calcium sulphate No treatment. .. ., Calcium sulphate No treatment. .. . Calcium sulphate No treatment. .. . Calcium sulphate No treatment .... Carbon dioxid produced in — ) days. Mgtn. 3-96 8. 14 4.40 7.81 2. 64 8.58 3-30 5-5° 7. 26 9.90 6.38 10.78 20 days. Mgm. 3-52 6. 16 3-30 5.06 2. 20 6.16 2. 64 3-52 6. 16 7.48 5-72 7. 26 30 days. Mgm. 2.86 4.84 3-3° 4.40 1. 70 5.28 2. 42 2.86 5.28 6.60 4. 62 5.28 Total carbon dioxid produced. Mgm. 10.34 19.14 II. 00 17.27 6.54 20. 02 8.36 11.88 18.70 23.98 16. 72 23-32 At the high water content the production of carbon dioxid for the first lo-day period was depressed slightly in four soils by the treatment with sulphate, but in two soils it was stimulated. During the second period three of the untreated samples of soil still showed a slightly greater rate of production of carbon dioxid than the corresponding treated samples, and one of the treated samples of soil produced some- what more of this material than the untreated. The remaining soils showed very slight differences in the production of carbon dioxid. During the third period there were more variations, two untreated samples producing more gas than the corresponding treated samples and three treated samples showing more activity than the untreated. The total production of carbon dioxid for the 30 days was greater for the untreated samples in four cases and less in one, and one soil showed practically no difference. Without exception the untreated samples maintained at low water content showed a greater production of carbon dioxid for each period than the corresponding treated samples. In some instances the differ- ence was so small as to be negligible, while in others it was very great. In every case the total production for 30 days was decidedly greater for the untreated samples. It would appear from the data presented that the biological activities do not account for the changes in the solubility of the soils when treated with calcium sulphate, if the carbon-dioxid production may be taken as a measure. On the whole, there was a slight depression of such activi- ties, especially when the samples were maintained at the low water con- tent. This is somewhat at variance with the results reported by Fred and Hart,^ who found an increased production of carbon dioxid from soil > Fred. E. B.. and Hart. E. B. the coMPARATrvB effect of phosphates and sulphates on son, BACTERIA, wis. Agr. Exp. Sta. Research Bui. 35. P. 35-66, 6 fig. 1915. Apr. IS. 1920 Effect of Calcium Sulphate on Solubility of Soils 53 containing 0.25 and 0.5 per cent calcium sulphate. It should be borne in mind, however, that the method of treating the samples was quite different from that in the experiment here reported. Several investi- gators have also reported a slight stimulation in ammonia production as a result of treatment with small amounts of calcium sulphate. In none of these experiments, however, were the soils thoroughly washed after treatment with the sulphate, and consequently it does not seem to be justifiable to make direct comparisons with our results. At the expiration of 30 days the concentration of the soil solution of the samples maintained at the high water content was determined by thoroughly stirring the sample, withdrawing a portion to a freezing- point tube, and making the determination in the usual manner. Suffi- cient water was added to the samples maintained at the low moisture content to bring them up to that of the corresponding samples maintained at the high water content. The results of these determinations, together with the parts per million of soluble material, are presented in Tables VII and VIII. Table VII. — Effect of calcium sulphate on the solubility of soils held at high water con- tent for JO days Soil No. Treatment. Freezing- point depressions. Soluble material. Calcium sulphate No treatment . . . Calcium sulphate No treatment . . . Calcium sulphate No treatment . . . Calcium sulphate No treatment . . . Calcium sulphate No treatment . . . Calcium sulphate No treatment . . . "C. o. 101 .026 • 073 • oil .066 . 014 .184 .017 .070 .063 .098 . 042 p. m. 2,525 650 1,825 275 1,650 350 4,600 425 1,750 1,575 2,450 1,050 Table VIII. — Effect of calcium sulphate on the solubility of soils held at low water con- tent for 20 days Soil No. Treatment. Calcium sulphate No treatment . . . Calcium sulphate No treatment . . . Calcium sulphate No treatment . . . Calcium sulphate No treatment . . . Calcium sulphate No treatment . . . Calcium sulphate No treatment . . . Freezing- point depressions. •c. O. 103 .015 .085 . 010 • III .013 .163 . 007 .099 .023 . 096 .023 Soluble material. P. p. m. 2,575 375 2,125 250 2,775 325 4i075 175 2,475 575 2,400 575 54 Journal of Agricultural Research voi. xix, no. a The total quantity of soluble material formed during the 30 days does not coincide with the amount of the carbon dioxid produced. The data show the treated samples to contain many times the amount of soluble material found in the corresponding untreated samples. There is one exception to this in the case of soil 5 at the high water content, where the treated sample contained only 175 parts per million more of soluble material than the untreated. It must be concluded, therefore, that the increase in soluble material takes place without the evolution of increased amounts of carbon dioxid and therefore is presumably due to other than biological agencies. SUMMARY AND CONCLUSIONS Six different soils were treated with a saturated solution of calcium sulphate. In one series of experiments the mass was transferred to filter paper, permitted to drain, and then transferred to containers and the rate of formation of soluble substances determined by means of the freezing-point method. The treatment was found to have increased the solubility of the soil to an appreciable extent. In another series the amount of soluble material was reduced to a minimum by washing with distilled water, and the residuary effects of the treatment on the solubility were likewise determined. The calcium- sulphate treatment was found to have resulted in a very large increase in the rate of formation of soluble substances. The effects were great even when the soils were washed the second time. Obviously the treat- ment results in changes in the composition of the soil mass — in other words, a soil of different properties is formed. It seems that it is possible to alter the composition of the soil solution and that whether such change will have any effect on plant growth or not or whether the effect will be favorable or unfavorable will depend upon the nature of the soil and of the substances added. Moreover, it is probable that this phase of the subject has not received sufficient attention in connection with our field experiments. Two soils of somewhat different texture and organic content were treated with a saturated solution of calcium sulphate, a. N/io solution of calcium phosphate, and a combination of the two. The soils were washed, and the rate of formation of soluble salts was determined. The calcium sulphate markedly increased the solubility in each soil, while the calcium phosphate decreased the rate of formation of soluble sub- stances. When calcium phosphate was used in conjunction with cal- cium sulphate, it counteracted the effects of the latter to some extent. If the carbon dioxid produced, as determined by the methods used, is taken as a measurement of the biological activities, the increase in the rate of formation of soluble substances brought about by the calcium- sulphate treatment is due mainly to other causes. FURTHER STUDIES ON THE INFLUENCE OF HUMIDITY UPON THE STRENGTH AND ELASTICITY OF WOOL FIBERS By J. I. Hardy Associate Wool Specialist, Wyoming Agricultural Experiment Station INTRODUCTION In a previous issue of the Journal the author published a preliminary report^ of his work on the influence of humidity upon the strength and elasticity of wool fiber. An attempt was made to obtain a better method of testing wool in order that wool from sheep under various con- ditions of breeding, feeding, and range management might be satisfac- torily tested. A study was also made upon the strength and elasticity of wool in an unsecured state under various conditions of humidity. A review of the literature was given in the earlier report and will not be repeated at this time. EXPERIMENTAL WORK After the work referred to above had been completed, further studies were begun upon scoured wool. As in the previous work, all samples were tested with a McKenzie fiber-testing machine. Wherever diameters are reported they are the results of measurements with a micrometer caliper unless othen^ase stated. This micrometer had a ratchet stop and was graduated to read in hundredths of a millimeter. The micrometer was used in the lower jaw of the testing machine and had a small hand lens held stationary before it. With this arrangement it was possible to interpolate the readings to o.ooi mm. The diameters of the fibers were read at four different points. The smallest of these figures was in each case used in computing the tensile strength of the wool fiber. Samples 991, 994, 996, and 997 had been extracted with ether and washed with hot water and tested at each of five relative humidities, 40, 50, 60, 70, and 80 per cent, when the operator was suddenly called into military service. The results of this work are given in Tables I and II. Table I. — Tensile strength of wool fiber at five different humidities Sample No. Number of fibers. At relative humidity of- 40 per cent. 56 per cent. 60 per cent. 70 per cent. 80 per cent 991. 994- 996. 997- 100 100 100 100 Mgm. 279. 22 274- 77 295. 64 215- 34 Average . 266. 24 Mgm. 299. 47 280. 50 302. 00 210. 48 Mgm. 289. 85 279- 73 281.47 201. 87 Mgm. 264. 29 255- 22 281. 40 200. 67 Mgm. 258. 02 269. 59 271.83 196. 56 273. II 263. 24 250-39 249. 00 1 Approved for publication in the Journal of Agricultural Research by the Director of the Agricultural Experiment Station of the University of Wyoming. 2 Hardy, J. I. influence of humidity upon the strength and the elastioty of wool fiber. In Jour. Agr. Research, v. 14, no. 8, p. 285-296, 2 fig., pl. 48. 1918 Literature cite s-o Fig. I. — Graphs showing the effect of humidity upon the breaking strength of wool fiber. Table I shows an average increase in the tensile strength of scoured wool as the humidity is raised from 40 to 50 per cent, and a decrease as the humidity is raised from 50 to 80 per cent. In Table II the average percentage of elasticity is shown to increase as the humidity is raised from 40 to 80 per cent. A new operator was put upon the work in order to obtain more data under the same conditions and additional data on fibers of a smaller diameter. The diameter of the fibers of sample 991 averaged 0.016 mm., ■""■ — ^v,.,^^ <^l^if?p^(p^ ;5>-=: s.,^^ ■■ c > ^^ \ ^ \ /t c" ^^ /I Apr. 15, 1920 Influence of Humidity upon Wool Fiber 57 while samples 994, 996, and 997 had an average diameter of 0.026, 0.029, and 0.025 mm., respectively. There was one sample of wool with an average diameter of fibers less than 0.02 mm., and there were three sam- ples with the average diameter above that figure. The new set of samples chosen, A, B, C, and D, consisted of four samples with average diameters of 0.012, 0.018,0.017, and 0.031 mm., respectively. Three of these samples were under 0.02 mm. in diameter, and one was larger. The range in average diameter of the fibers tested is from 0.012 to 0.031 mm. Fibers were tested from small locks of scoured wool from samples A, B, C, and D until 200 fibers were tested at each of -/-^ 2^ -5^ SO GO 70 <90 Fig. 2.— Graphs showing the effect of humidity upon the elasticity of wool fiber. five humidities, as shown in Table III. It will be noted that the break- ing strength of the fibers decreases quite uniformly as the humidity increases. Sample D shows a decrease in its tensile strength as the humidity increases up to 80 per cent, when there is a very slight increase. In A, B, and C the tensile strength seems to fluctuate up and down with no particular uniformity. These values for tensile strength were much more variable than those for the breaking strength. Several hundred additional fibers were tested on A, B, C, and D at humidities of 40 and 50 per cent, since the greatest variability seemed to occur at these two points. Graphs showing the values obtained on these samples of scoured wool for breaking strength and elasticity are shown in figures i and 2. 58 Journal of Agricultural Research Vol. XIX. No. 2 Table III. — Diameter, breaking strength, and tensile strength of scoured -wool fibers at five different humidities At relative humidity of 40 per cent. At relative humidity of 50 per cent. Sample No. Diam- eter in thou- sandths of a mm. (average of 100). Break- ing strength (average of 100). Average. Tensile strength per thousandth of a sq. mm. Diam- eter in thou- sandths of a mm. (average of 100). Break- ing strength (average of 100). Average. Tensile strength per thousandth of a sq. mm. Average of 100. Average of 200. Average of 100. Average of 200. A 1 II. 9 I 13-2 1 20. 04 I 20.34 / 19-89 I 17-44 / 30.78 I 30.48 Dgm, 48-59 48-52 86.62 92.04 88.03 82.52 208. 02 211.32 Dgm. 48-56 "89-33' Mgm. 433-97 359-33 274-37 284.60 283. 29 320.63 281.50 280.38 Mgm. 396- 65 10.79 11.54 17-86 18.83 17.40 17.40 32-57 31-36 Dgm. 47.67 43.10 82.81 89.52 82.89 79-53 221.86 207. 22 Dgm. 45-39 Mgm. 521-67 412.08 330.37 319-36 349-05 334- 54 266. 54 268. 74 Mgm. 466.88 B 279.49 86.17 324-87 C 85.28 301.96 81.21 341-80 D 209. 68 280. 94 214.54 267.64 At relative humidity of 60 per cent Sample No. A B C. D Diam- eter in thou- sandths of a mm. (average of 100). 11.84 12.75 16.44 16.55 18.37 30.65 Break- ing strength (average of loo). DgTK. 44-89 44.91 72-85 71-74 81.02 197.02 Average. Dgm. 44.90 72.30 81.02 197.02 Tensile strength per thousandth of a sq. mm. Average of 100. Mgm. 407. 40 352-07 343-30 329-51 305.68 266. 27 Average of 200. Mgm. 379-74 336.41 305-68 266.27 At relative humidity of 70 per cent. Diam- eter in thou- sandths of a mm. (average of 100). 10.80 11.09 18.85 16.85 16.57 17.08 31.26 .30.35 Break- ing strength (average of 100). Dgm. 41.09 39.53 85.86 74.08 74.62 79-23 199.67 193.31 Average, Dgm. 40.31 79-97 76.93 196. 49 Tensile strength per thousandth of a sq. mm. Average of 100. Mgm. 455-25 409.27 324.14 334.41 347- 76 345-78 260.07 267. 20 Average of 200. Mgm. 432.26 329-28 346-77 263. 64 Sample No. At relative humidity of 80 per cent. Diameter in thou- sandths of a mm. (average of 100). Breaking strength (average of 100). Average. Tensile strength per thousandth of a sq. mm. Average of 100. Average of 200. A B C. D 11.97 10.45 17. 26 18.64 13-62 14.65 30.02 28.69 Dgm. 45.41 40.48 73.79 86.08 51.46 56.10 183-75 175-83 Dgm. 42.95 79-94 "53.78' 179.79 Mgm. 403.8s 471.98 315-37 314- 72 353-21 333-42 259- 60 272. 24 Mgm. 437.92 315-05 343-32 265.92 The heavy line shows the average values obtained for all the results secured at each humidity. The average breaking strengths of these samples of scoured wool decrease as the humidity increases, while the elasticity shows an increase with an increase in humidity. The wide variations in the values for tensile strength as compared with similar values for breaking strength led the writer to compare the tensile strengths of fibers of dififerent diameters in locks of wool A, B, and D. Apr. 15, 1920 Influence of Humidity upon Wool Fiber 59 Graphs showing the variation in the tensile strengths of three different samples of wool are shown in figure 3 in the curves A-A, B-B, and D-D. The fibers tested in these curves range from 0.008 to 0.038 mm. in diam- eter. The number of fibers tested at each humidity varies considerably. In some cases only 30 or 40 were tested, while in other cases as many as 250 of a given diameter were tested. Sample A of curve A-A ranges in fineness from 0.008 to 0.018 mm. The tensile strength decreases from 667 to 260 mgm. per thousandth of a square millimeter at the lowest point. Sample B shows a decrease from 466 mgm. at o.oi mm. to 315 mgm. at 0.022 mm. The cur\^e of sample B follows that of sample A very closely from a diameter of o.oi mm. to one of 0.018 mm. and rises slightly from a diameter of 0.018 mm. to one of 0.022 mm. Sample D decreases from 320 mgm. at 0.023 rnni- to 232 mgm. at 0.038 mm. These curves show that the tensile strength of wool decreases with an increase in diameter. The drop is most abrupt with the sample of fine wool. The coarsest sample has the most gradual drop in its diameter and tensile strength curves. If the breaking strength of wool varied directly as the area of cross section, the curv^e would follow the line E-E. If the breaking strength varied as the diameter or circumference, the ten- sile strength curve would follow the line F-E. The curve for the tensile strength of sample D follows the line D-D and lies between these two lines E-E and F-F. This fact seems to indicate that the breaking strength of medium and coarse wool varies with some power of the diam- eter which lies somewhere between the first and second. For fine wool like sample A, a curve showing the strength of the wool very closely follows a curve plotted with i , or any constant, and the first power of the diameter. This fact indicates that the breaking strength of fine wool does not vary directly with the area of the cross section but with a value which is very close to the first power of the diameter. Curve C-C shows the relation between the tensile strengths and diameters of wool fibers obtained from data published by Hill.^ In the present ex- periment, 1 ,000 fibers were broken to obtain the points in this curve, and each diameter was measured after breaking as nearly as possible at the point of breakage. This curve also follows very closely the curve F-F. By inspecting the graphs it is easy to see that the widest variations in the curve F-F plotted from j^are found at the smallest diameters. As this curve approaches the larger diameters it tends to become rather flat. In the first three samples of Table III there is a large variation between the largest and smallest tensile strengths of the wool fibers of those sam- ples. When fibers are tested with such a wide variation in their tensile strength as is found in locks of fine wool, it is necessary that these fibers be carefully mixed in order to get satisfactory results. There is a ten- dency for an operator to pull the largest fibers in fine wool, while with ' Hill, J. A. studies on the .strength and elasticity op the wool fiber, i. the probable error OP THE MEAN. /jiWyo. Agr. Exp. Sta. 21st Ann. Rpt., 1910-11, suppl., 139 p. 191 1. 6o Journal of Agricultural Research Vol. XIX, No. 2 ^ <\ F 1 — \ \ i ; / / / / '^ 1 1 • / \ \ / / \ ; / /* \ \ ; 1 i / i 1 1 1 ; i \ \ ; > / / / / • / / ': ,/ * : / 1 : / 1 \ J : f 1 i i 7 ./ f \ .•V ,< 7 i / 0/ > ^ 'm '9 ..'/ ^ X \ / ■^l V ^ ^ 0 Q o ^ ^ c<5 05 (\j c^ Apr. IS, 1920 Influence of Humidity upon Wool Fiber 61 coarser samples there is not such a tendency. This fact and the fact that larger fibers can be more accurately measured with a micrometer caliper make it possible to get satisfactory results for tensile strengths with sam- ples of coarser wools. Then again the coarser wools have breaking strengths which vary more closely with the areas of the cross section of the wool than do the breaking strengths of fine wool samples, as is shown in F-F of figure 3. The coarse wools may be measured from the original lock, and their breaking and tensile strengths may be determined quite satisfactorily. Sometimes it is necessary to make the closest possible comparison of the effects of various conditions or chemical reagents on a given grade of wool, as in the case at hand. The writer desired to determine the effects of various humidities upon a uniformly mixed sample of wool. Single fibers were drawn from sample B and placed consecutively in six different groups, numbered i to 6, with their ends extending from one piece of adhesive tape to another which was laid parallel to it and about 2^ inches from it. Always beginning with No. i, these fibers were placed one at a time in each of these six groups until 100 fibers, or the desired number, were in each of the six small locks. By making five series of these groups and subjecting the same numbers of each group to the same test, it is possible to get some very satisfactory comparisons. Although it is very tedious work, these fibers may be picked out by hand at the rate of 200 an hour. Five small locks, each containing 120 fibers, were tested in the scoured condition at humidities of 40, 50, 60, 70, and 80 per cent and saturated. Similar locks were scoured with ether and hot water and tested under the same conditions. The satu- rated fibers were kept between moist filter papers until tested. Table IV and figure 4 show the results of this experiment. Table IV. — Elasticity and breaking strength of scoured and unsecured wool from sample B [Average of 600 fibers] Percentage of humidity. Scoured wool. Elasticity. Breaking strength. Unsecured wool. Elasticity. Breaking strength. 40 50 60 70 80 Saturated Per cent. 25. 80 30.76 33-96 37.08 40. 08 33-76 Dgm. 65. 14 64. II 64. 64 59-53 60. 10 59.16 Per cent. 26. 40 31.48 34-72 38.00 43-64 34.60 Dgm. 69. 06 68.97 67. 01 63.80 60. 44 63.42 The curve for unscourcd wool shows that the breaking strength de- creases as the relative humidity changes from 40 to 80 per cent and in- creases when the wool becomes saturated. In scoured wool the curve is more irregular. There is a definite drop as the humidity changes from 62 Journal of Agricultural Research voi. xix. No. a 40 to 80 per cent, although the curve makes almost a straight line from 70 per cent up to the point of saturation. The elasticity curves for scoured and unscoured wool are nearly parallel, rising as the humidity changes from 40 to 80 per cent and falling from this point to that of saturation. SUMMARY (i) The tensile strength of wool increases with the decrease in the diameter of the wool fiber. (2) Fine wool has a breaking strength varying more closely with the first than with the second power of the diameter. 4-4 42 \40 34 / \ /1 A yy \\ >^ /^ \v A ^ N^ ^' A y^ / / "32 \30 \2e %-^ 24 40 s5o SO yo <9o s^mem^ Fig. 4.— Graphs showing the effect of humidity upon the breaking strength and elasticity of wool fiber. (3) Coarse wool has a breaking strength varying with a figure which lies somewhere between the first and second powers of the diameter. (4) It is necessary to mix samples of fine wool carefully before testing in a testing machine if the best results are to be obtained. (5) The breaking strength and tensile strength of both scoured and unscoured wool decrease with an increase in relative humidity from 40 to 80 per cent and show a tendency to increase from this point to that of saturation. (6) The elasticity of scoured and unscoured wool increases with an increase in relative humidity from 40 to 80 per cent and decreases from this point to that of saturation. COMPOSITION AND DENSITY OF THE NATIVE VEGETA- TION IN THE VICINITY OF THE NORTHERN GREAT PLAINS FIELD STATION By J. T. Sarvis Assistant in Dry-Land Agriculture, Bureau of Plant Industry, United States Department of Agriculture INTRODUCTION The grazing industry in the Northern Great Plains area is intimately concerned with the composition and density of the native vegetation. This paper deals with the native vegetation as it exists at present in the section under consideration. While parts of the discussion will apply in general to the Great Plains area, it pertains to western North Dakota and in particular to the territory adjacent to the Missouri River on the west near Mandan. This point lies practically on the one hundred and first meridian and just south of the forty-seventh parallel, north latitude. The Bureau of Plant Industry has one of a number of field stations located here under the direction of the OfRce of Dry-Land Agriculture. One of the lines of investigation in connection with this station is a grazing ex- periment in cooperation with the North Dakota State Experiment Sta- tion. This investigation is primarily concerned with determining the carrying capacity of the range in that section and working out a grazing system adapted to conditions in the Great Plains. In connection with this work it is necessary to make detailed studies of the native vegeta- tion in order to observe any changes that may occur in the structure of the plant cover. These studies have furnished the material of this paper.* TOPOGRAPHY AND SOIL The topography of the area around Mandan varies from rolling to nearly level. The land is cut by numerous ravines and coulees, which drain into the Heart and Missouri Rivers. The altitude of the field sta- tion is approximately i ,700 feet above sea level. The following description of the soil of this area is quoted from "The Story of the Prairies" by Willard (9),^ formerly geologist at the North Dakota Agricultural College: A belt having an indefinite edge to the westward lies along the west side of the Missouri River, which belt represents the western limits of the glaciated area of North Dakota, and of the Continent of North America. This "belt" of land along the west ' The annual reports by the author of the cooperative grazing experiment at Mandan have been frequently- referred to and used in the preparation of this paper. These reports are on file in the Office of Dry-Land Agriculture, the North Dakota Agricultural College, and the Mandan Field Station. 2 Reference is made by number (italic) to "Literature cited," p. 71-72. Journal of Agricultural Research, Vol. XIX, No. a Washington, D. C. Apr. 15. 1930 tx Key No. G-188 (63) 164176°— 20 2 64 Journal of Agricultural Research voi.xix. No. j side of the river shows by the character of the soils and the rocks that lie upon or near the surface that the great continental glacier was once here. Toward the west the belt fades out and becomes indistinguishable from the land farther west over which the ice did not pass, but the eastern part of the belt is sufficiently modified as to the soils and the landscape features to be readily recognized. The soils, therefore, in the belt bordering the Missoiu'i River on the west constitute a transition type from the glacial soils of the eastern portion of the State to the non- glaciated or residual soils of the southwestern portion of the State. CLIMATE The United States Weather Bureau Station at Bismarck has made con- tinuous meteorological observations since 1875. Bismarck is located on the east side of the Missouri River, only about 5 miles distant from Man- dan. Observations were begun at the Mandan Field Station during 191 3. From 1875 to 1914, inclusive, or 40 years, the mean annual precipitation was 17.41 inches. The greatest annual amount during this period was 30.92 inches in 1876, while the lowest was 11.03 inches in 1899. During 191 7 the record at the Mandan Field Station was 10.31 inches. The mean seasonal precipitation from April i to July 31, inclusive, was 9.91 inches during the 40- year period. The month of maximum precipitation is June, with a mean of over 3.5 inches, and the month of minimum pre- cipitation is February, with less than 0.5 inch. The temperature is extreme in both winter and summer. The lowest recorded to date was 45° F. below zero in January, 1916, while the high- est was 107° above zero in July, 1910 and 1917. The average dates of killing frosts in spring and autumn are about May 15 and September 15, respectively, but frosts have occurred as late as June 7 and as early as August 23. The average frost-free period is 128 days. The prevailing wind direction is from the northwest. The average wind movement near the ground is about 6 miles per hour. PLANT FORMATION According to a map of "Plant Formations of the United States," by Shantz and Zon,* this region would come within the "short-grass forma- tion." However, Dr. F. E. Clements, who visited the field station during the summer of 191 7, is of the opinion that it would be more properly placed in the "long-grass" or "prairie formation," because of the long grasses and other plants which are typical of a prairie formation. From actual determinations in the field the percentages of short-grass and long-grass cover have been found to be nearly equal, so that the for- mation could be put in either class, according to the viewpoint of the observer. If the secondary plant layer is considered as the determining factor, the region falls in the long-grass formation. The vegetation in this particular area might be considered as in a transition zone, since the dominating species are typical of both formations. 'Shantz, H. L., and Zon, R. plant formations of the united states. Paper presented before the Ecological Society of America at its annual meeting in New York in 1916. The map will appear in the Agricultural Atlas. Apr. IS, igao Nolvve Vegetotiofi of Northern Great Plains 65 The dominating species are Bouteloua gracilis {B. oligostachya) and Stipa comata, which form a distinct association. This is an association composed of Bouteloua gracilis, which is typical of the short-grass forma- tion, and Stipa comata, which is a typical long-grass species. This asso- ciation is dominated by the Bouteloua. Sarvis ^ has described in a paper other sections of western North Dakota which show the same dominating species. COMPOSITION OF THE VEGETATION In Plate 1 2 is illustrated the general character of the vegetation on the prairie in the Mandan region. In 191 5, when this photograph was taken, the season was very favorable, and all plants reached a maximum devel- opment. The composition of the vegetation is thus very clearly illus- trated. In the following list of plants the arrangement of species is in the order of abundance. The order of the primary and secondary species is subject to slight modifications as the studies are extended. The order of the dominant species was determined by measurements from quadrat maps and in the field. The order of the primary species, other than grasses, was determined by count. The secondary species are listed in the estimated order of their abundance. DOMINANT SPECIES Bouteloua gracilis Stipa comata Artemisia gnaphalodes Koeleria cristata Solidago pulcherrima Agropyron smithii Artemisia dracunculoides Psoralea argophylla Andropogon scoparius Muhlenbergia cuspidata Lacinaria punctata Calamovilfa longifolia Agropyron caninum Bouteloua curtipendula Comandra pallida Carex filifolia Carex heliophila PRIMARy SPEaES A rtemisia frigida Stipa viridula Eschinacea angustifolia Aristida longiseta Polygala alba Stipa spartea Ratibida columnaris 6BCONDAKY SPEQES Aster multiflorus Petalostemon purpureum Petalostemon candidum Lactuca pulchella Vicia sparsifolia Agropyron tenerum. The grasses, other than the dominant species, are in the estimated order of abundance. It is difficult to make individual counts of them, since they usually occur in bunches. If bunches or mats were considered as single plants and enumerated as such the number would have no signifi- cance when compared with that of other plants which usually occur as individuals. 'Sarvis, J. T. native grasses of western north Dakota. Paper presented before the Ecological Society of America at its annual meeting in New York in 1916. 66 Journal of Agricultural Research voi. xix, No. 2 When the vegetation is considered from the standpoint of grazing, only a very few species are important factors in the total amount of forage annually produced. Sampson (6) has discussed this point more fully. In this region, Bouteloua gracilis and Siipa comata are the most important species, both on account of their total forage production and their value as grazing grasses. The value of a given species for grazing purposes depends upon (i) its abundance, (2) whether it is relished by stock, (3) its length of growing season, (4) its ability to withstand trampling and to recover readily from grazing, and (5) its adaptation to drought conditions. According to these requirements, Bouteloxia gracilis would take first rank and Stipa comata would be second in importance. A plant may be of importance in relation to grazing because of its abundance, whether it is or is not of grazing value. If it is a valuable grazing species it is of primary importance, and if it is of minor grazing value it is of importance because it occupies ground surface that might otherwise support a more valuable species. On the other hand, a species may be greatly relished by stock, as Andropogon furcatus at Mandan, but occur in such limited areas that it is unimportant in the total amount of forage annually produced. In pastures where this grass occurs it is cropped close to the ground throughout the season, as illustrated in Plate 13, A. Bouteloua gracilis is grazed with avidity at all times of the year. It cures well on the ground without great loss of its nutritive value, and late in the fall cattle eat it in preference to any other grass. Although Stipa comata has the disadvantage, for a short period, of its sharp needles, it is so much more abundant than other species, except B. gracilis, that it enters largely into the feed of grazing animals. It is the first grass to produce green shoots in the spring, and it usually produces more growth late in the fall than do other species. A grass that is similar in appearance and often confused with Bouteloua, gracilis is Bulhilis dactyioides , or buffalo grass. It has a better reputa- tion for grazing and is more widely known by a popular name than any other single species of grass in the Great Plains. However, out of several thousand acres of native vegetation surrounding the field station, there are less than 5 acres of the true buffalo grass. On a trip over western North Dakota in the summer of 191 6, the author found this grass in only a few small areas. Blue grama {Bouteloua gracilis) is and always has been called buffalo grass by the people in the Great Plains area. This misnomer has been and is so universal that it is difficult to obtain reliable information concerning the abundance and importance of buffalo and blue grama grasses for grazing in the early history of the range. However, at present the true buffalo grass occurs only in small amounts in this region and in western North Dakota, where it is evident it never Apr. IS. 1930 Native Vegetation of Northern Great Plains 67 was as abundant as in western South Dakota. Pound and Clements {4) said in regard to buflfalo grass: The buffalo-grass was, until recently, supposed to have once covered the greater portion of Nebraska; its disappearance has, as a matter of sentiment, been connected with that of the buffalo. The patches of buffalo-grass, which are foimd scattered here and there over the State, are to be regarded as intrusions rather than stragglers left by a retreating species. Griffiths (2) says in regard to Bulhilis dactyloides : Bouteloua gracilis, especially when not in head, is very similar and frequently mis- taken for it. On this account the true buffalo grass is very much overestimated in importance, because there are so many things included with it in the popular mind. Much of the credit given this species is due to the gramas, which in age especially look much like it. On the other hand, the species is an important one throughout its range. In southwestern South Dakota, at the Ardmore Field Station, where a grazing experiment is now being conducted, the important grazing grasses are Bulhilis dactyloides, Bouteloua gracilis, and Agropyron smithii. This association is dominated by the Bulbilis. It often happens that a species that is of little grazing value in one section is of value in another area. For example, Aristida longiseta is of little grazing value at Mandan, since it is the last plant that cattle will take even when the pasturage is short, as illustrated in Plate 13, B. However, in other sections, Griffiths, Bidwell, and Goodrich (2) report this species as being of considerable value. Some species are indicators of overgrazing, as Artemisia frigida at Mandan. In pastures where this plant occurs in abundance it usually will be found that the area has been overstocked for several seasons. In the vegetation of this area no poisonous plants are abundant enough to be harmful. However, in areas farther west in North Dakota, the common "loco weed" (Oxytropis lamberti) is abundant and causes serious losses of stock in certain seasons. All the plants mentioned in the list on page 65 enter more or less into the feed of grazing animals, but, as noted, only a few species produce a considerable percentage of the total forage. One of the reasons for this fact is the inability of many plants to produce more than a limited second growth after they have once been removed by grazing. DENSITY OF VEGETATION In a consideration of plant density in relation to grazing problems it is desirable and necessary to make clear and concise distinctions between frequently recurring terms. Plant density should refer to the "stand" or thickness of plants upon the ground surface. The ground surface is the total area of land under consideration, whether vegetated or unvegetated. Bare ground should be understood to refer to the un- vegetated portion of the ground surface or the spaces in the cover between 68 Journal of Agricultural Research voi. xix. no. 2 individual plants or between mats and bunches of species which grov/ in that manner. The term cover (8), or ground cover (5), is frequently and conveniently used in connection with discussions of vegetation. However, when the term cover is applied in connection with grazing investigations it should be defined, for it may mean one of two things: (i) basal cover, or the ground surface limits of living vegetation, or (2) the foliage cover, which is the plant layers above the basal cover. When the foliage cover is removed, as by close grazing or clipping, the basal cover remains. Plant layers as described by Clements (i) are vertical zones based on the height of plants. On the prairie around Mandan two layers are impor- tant— the ground layer, as Bouteloua and Carex, and the secondary layer, as Stipa and Psoralea. Species that grov/ in mats or in bunches are most accurately expressed in terms of basal cover. F'or example, Bouteloua basal cover would refer to the amount of ground surface actually covered by Bouteloua if the foliage were removed by grazing or clipping. In such species it is possible to make the determinations with almost mathematical preci- sion. Species that occur as individuals are best expressed in terms of their abundance per unit area. Shantz (7) says in regard to this point: Those species which form mats can not be well represented in numbers per square meter, and on this accoimt the percentage of surface covered is given instead. The foregoing statements in regard to basal and foliage cover are very clearly illustrated in Plate 14. In 191 5 the foliage cover was very heavy because growth conditions were favorable and the ^IH^ area had not been grazed. An estimate of the total cover iu^jl based upon the amount of foliage cover could easily have pT^i^ been made at that time. But in 191 6 on the same area, i i^L ! with the foliage cover removed, there would have been '( '(fj\ no basis for comparison with the 191 5 condition. This \.^ illustrates the undesirability of utilizing the foliage cover. Fig. 1— Diagram Under all Conditions, as a basis for estimating the pos- of grass mat: sibiHtics of foragc production and the consequent carry- si from above.' iug Capacity. A clear distinction between basal cover o. Basal cover; g^^^j foliage covcr is, therefore, necessary and important. 0, foliage cover. . . r i • « The two illustrations of Plate 14 picture the same area, but one illustrates a heavy foliage cover and the other only the basal cover. However, the potential ability of the area to produce under similar conditions as heavy a foliage cover as in 191 5 is unchanged. Figure I illustrates the difference between the basal cover and the foliage cover. The limit of basal growth is a, while the limit of foliage growth is b. In a given case the surface area of the foliage cover is greater than that of the basal cover, yet the amount of forage is the same. The basal cover is more permanent than the foliage cover, since the latter may be readily removed by grazing. The quadrat map (fig. 2) in the 30-acre pasture, which was mapped in 191 5 and remapped in 191 6, shows, Apr.' 'is, 1920 Native Vegetation of Northern Great Plains 69 with the exception of a few annual species, the basal cover to be prac- tically the same in both years. If the maps had been drawn on the basis of the foliage cover, there would have been a great difference between the 1 91 5 and 1 91 6 maps. The photographs illustrate this difference more clearly than would be possible by quadrat maps. But if the maps are drawn on a basis of the basal cover, various maps of a given quadrat would show actual changes as they occur from grazing. This is really Fig. 2. — Meter quadrat in 30-acre pasture mapped in detail in 1915. Cross hatching represents BouUloua gracilis; vertical hatching, Stipa comata. The presence of other species is indicated by dots and out- lined areas. the important point in relation to grazing systems. If grazing has been severe, the basal cover is likely to be changed rapidly, but under normal conditions it should change gradually. This is especially true in such regions as Mandan, where most of the vegetation is made up of perennial species. Sampson (5) says in regard to increase of ground cover: The increase in actual stand or ground covered was due almost entirely to the enlargement of the tufts, and text figiu-es 5 and 6 show that even under season-long protection the bunch-grasses and other valuable plants do not increase rapidly by this means. 70 Journal of Agricultural Research voi. xix, no. 2 Since the carrying capacity of the range is largely dependent upon the density of the vegetation, it is obvious that this factor should be carefully determined. If density is determined on the basis of the foliage cover, even when this is possible, the carrying capacity is likely to be placed too high, because of favorable growth conditions or an accumula- tion of previous growth, and overgrazing will result. In normal seasons the amount of forage a given area of ground surface can produce is largely determined by its basal cover. Therefore, the basis for an estimate of the amount of ground surface covered by vegetation should be founded upon the basal cover. The foliage cover is the important consideration for immediate grazing, but the basal cover more nearly determines the future possibilities of a given area of land for grazing purposes. AMOUNT OF BASAL COVER AT MANDAN From quadrat maps drawn to show bare and covered ground surface the total basal cover has been determined. The maps show about 60 per cent vegetated and 40 per cent bare ground. From quadrat maps, such as that in figure 2, made in the various pastures, the percentages of basal cover of Bouteloua and Stipa were determined. These are approxi- mately 20 and 10 per cent, respectively. These determinations were all made from the maps by means of a planimeter. Shantz (7) has made a number of estimates on the amount of cover in a series of quadrats in the mesa region near Pikes Peak. He has ex- pressed the amounts in percentages in each case. The same method is followed in the present studies. This is a most convenient system, especially when it is desired to express a given species in terms of amount of total cover. Sampson (5) expresses the "density of vegeta- tion" in terms of tenths, using 10 as complete ground cover. In order to avoid confusion, the amounts of cover as used in coimection with the Mandan grazing experiment are expressed in percentages. From the amounts of basal cover of Bouteloua gracilis and Stipa comata it is readily seen how important they are from the standpoint of grazing in this sec on. Griffiths, Bidwell, and Goodrich (2) have dis- cussed the value ci these grasses for forage. From clipping experiments at Mandan in 191 7, in connection with the grazing studies, the Bouteloua was found to have produced from 40 to 50 per cent and Stipa from 15 to 20 per cent of the total forage for the season. When the quadrats were clipped, the vegetation was separated into six parts, as follows : Bouteloua gracilis, Stipa comata, Aristida longiseta, other grasses, Carex fUifolia and C. heliophila, and other plants. Columns are also reserved for the sum of B. gracilis and 5. comata and for the total weight of all grasses and of all species. From these data it is possible to determine the relation of one species or group to another or to the total weight of all species. The various amounts were recorded in grams, weighed both green and air- dried. From these data it appears evident that the ground layer is the important one from the standpoint of grazing in this section. Apr. IS. 1920 Nati've Vegetation of Northern Great Plains 71 The abundance of a given species often appears greater than is deter- mined by actual counts per unit area. Pound and Clements (4) have fully discussed this point. From Plate 12 it would appear that Psora- lea argophylla is the most abundant species. However, by a number of actual counts per unit area it was found to be fourth in abundance of plants other than grasses and sedges. SUMMARY (i) The data and conclusions presented in this paper have been obtained in connection with a grazing experiment at the Bureau of Plant Industry Field Station near Mandan, N. Dak. This experiment is designed to determine the carrying capacity of the native vegetation and the effects upon it of different intensities and methods of grazing. (2) The vegetation is composed of a large number of species, only a few of which produce a considerable amount of the total forage. The dominating species are Boutelovu gracilis and Stipa comata. (3) The density of the vegetation is determined by the thickness of plants upon the ground surface and not by the foliage growth. The term cover used in connection with density may mean basal cover or foliage cover. The former remains after the latter has been removed by close grazing or clipping. (4) The total basal cover of all species in the Mandan region is approx- imately 60 per cent of the ground surface. Boutelovu gracilis has a basal cover of about 20 per cent and Stipa comata nearly 10 per cent of the ground surface. (5) Clipping data of different day periods showed that Boutelotta gracilis had produced from 40 to 50 per cent and Stipa comata from 15 to 20 per cent of the total forage. The remainder was made up of a number of other species. LITERATURE CITED (i) CuEMENTS, Frederic E. 1916. PLANT succession; an analysis of the development of vegetation. xiii, 512 p., illus., pi. Washington, D. C. (Carnegie Inst. Washington, Pub. 242.) (2) Griffiths, David, Bidwell, George L., and Goodrich, Charles E. 1915. N.\TIVE pasture grasses OF THE UNITED STATES. U. S. Dept. Agr. Bui. 201, 52 p., 9 pi. (3) Pound, Roscoe, and Clements, Frederic E. 1898. A METHOD OP determining THE ABUNDANCE OF SECONDARY SPECIES. In Minn. Bot. Studies, s. 2, pt. i, p. 19-24. (4) 1900. THE PHYTOGEOGRAPHY OF NEBRASKA, ed. 2, 442 p., 4 maps. Lincoln, Nebr. (Univ. Nebr. Bot. Survey, no. 8.) Bibliography, p. 22-30. (5) Sampson, Arthur W. 1914. NATURAL REVEGETATION OF RANGE LANDS BASED UPON GROWTH REQUIRE- MENTS AND LIFE HISTORY OF THE VEGETATION. In Jour. Agr. Research, V. 3, no. 2, p. 93-148, 6 fig., pi. 12-23. 72 Journal of Agricultural Research voi. xix, No. a (6) Sampson, Arthur W. i917. important range plants: their life history and forage value. U. S. Dept. Agr. Bui. 545, 63 p., 56 pi. (7) Shantz, H. Iv. 1906. A STUDY OP the vegetation OF THE MESA REGION EAST OF PIKES PEAK. In Bot. Gaz., v. 42, no. i, p. 16-47, fig- i~7; ^o- Z> P- 179-207, fig. 10-13. (8) I9II. NATURAL VEGETATION AS AN INDICATOR OF THE CAPABILITIES OF LAND FOR CROP PRODUCTION IN THE GREAT PLAINS AREA. U. S. Dept. AgT. Bur. Plant Indus. Bui. 201, 100 p., 23 fig., 6 pi. (9) WiLLARD, Daniel E. 1908. THE STORY OP THE PRAIRIES; OR, THE LANDSCAPE GEOLOGY OF NORTH DAKOTA, ed. 5, 377 p., illus., pi., maps. Chicago. PLATE 12 General view of native vegetation near Mandan, N. Dak., showing composition and density. The following species are evident in the photograph: Psoralea argophylla, Echinacea angustifolia, Artemisia frigida, Bouteloua gracilis, Stipa comata, S. viridula, and Ratibida columnaris. Native Vegetation of Northern Great Plains Plate 12 Journal of Agricultural Research Vol. XIX, No. 2 Native Vegetation of Northern Great Plains Plate 13 Journal of Agricultural Research Vol. XIX, No. 2 PLATE 13 A. — View across area of Andropogon furcatus. This grass is closely grazed, as it is greatly relished by cattle. Mandan, N. Dak., Nov. 2, 1917. B. — Close view of Arisiida longiseta bunches. All other vegetation has been re- moved by cattle close to the bunches. Mandan, N. Dak., Nov. 2, 1917. PLATE 14 A. — Close view, from above, of meter quadrat in 30-acre pasture. This is the same area shown in B but was taken in 1916 after the foliage cover had been removed by grazing. Only basal cover remains. Mandan, N. Dak., Oct. 10, 1916. B. — Meter quadrat in 30-acre pasture. This shows the cover as it appeared before grazing. Mandan, N. Dak., July 28, 1915. Native Vegetation of Nortfiern Great Plains Plate 14 Journal of Agricultural Research Vol. XIX, No. 2 EFFECT OF REACTION OF SOLUTION ON GERMINA- TION OF SEEDS AND ON GROWTH OF SEEDLINGS By Robert M. Salter, Soil Chemist, and T. C. McIlvaine, Assistant Agronomist, West Virginia Agricultural Experiment Station INTRODUCTION Recent investigations have emphasized the importance of the intensity factor of soil acidity. The growth of plants is more logically associated with hydrogen-ion concentration than with total acidity as measured by a soil's capacity to neutralize or absorb bases. However, other factors than the direct physiological influence of the hydrogen or hydroxyl ion upon the plant itself are undoubtedly operative in producing the effects attributed to soil reaction, these factors being either conditioned by the reaction or associated with it. Thus, indirect efifects upon plant growth would be produced by: (i) The extent to which the soil's reaction is favorable for the development of soil organisms, more particularly those responsible for nitrogen transformation and nitrogen accumulation; (2) changes in the solubilities of soil constituents as affected by reaction, this applying not only to essential elements such as calcium, magnesium, potassium, and phosphorus but also to those having toxic properties, such as aluminium, manganese, and ferrous iron where increases in con- centration would be expected with increase in acidity; and (3) changes produced in physical properties of soils attendant upon changes in reaction. Although the mass of data on the relation of soil acidity to plant growth is already large, few well-defined attempts have been made to separate the individual factors concerned and study them under condi- tions permitting the control or elimination of other factors. The present investigation was undertaken with the aim of studying the direct physi- ological influence of reaction as measured by hydrogen-ion concentration upon plant growth. Solution culture was resorted to in order to control or eliminate other factors as far as possible. EXPERIMENTAL METHODS In the work herein reported, wheat, corn, soybean, and alfalfa seed- lings were grown in a series of solution cultures having as far as possible a constant nutrient composition and osmotic concentration and varying in reaction from a hydrogen-ion concentration of approximately i X 10— ' to I X 10- « or 2 Ph to 8 Ph-^ 1 In this report the Ph values of Sorensen will be used to state the reaction of the solutions, the value Pa being the negative common logarithm of the actual numerical concentration of hydrogen ions. Thus a con- centration of hydrogen ions of iXio— ^ would correspond to a Ph value of 5. Journal of Agricultural Research, Vol. XIX, No. 3 Washington, D. C. Apr. 15, 1920 ty Key No. W. Va.-i (73) 74 Journal of Agricultural Research Vol. XIX, No. 2 NUTRIENT COMPOSITION OF SOLUTIONS The need for a basic nutrient culture of favorable physiological balance was recognized. The attempt was at first made to adjust Shive's solu- tions No, R5C2 and R3C3 (24) / which he found best suited to the growth of wheat seedlings, to the various reactions desired for the work by additions of the requisite amounts of an acid or base. However, because of the extensive precipitation of phosphates of calcium and magnesium in the more alkaline members of such series, these solutions were found unsuited to the work at hand. Two series of solutions were eventually employed which varied some- what in composition from Shive's best solutions. The maximum partial ionic concentrations in volume equivalents for the two solutions used are given in Table I, the composition of Shive's solutions being included for purposes of comparison. Table I. — Maximum ionic concentrations of solutions [Expressed as gram-equivalents per liter] Kind of solution. Na+. K+. KCa-H-. MMg++. NO3-. KSOl-. H2P04-. H3C6H507-. C1-. Series A.. Series B.. Shive's' fo.oioo to .0200 [ .0000 to I .0360 0.0360 • .0180 .0180 .0108 0.0050 .0050 .0104 .0156 0.0050 .0050 .C^OO O.OIOO .0100 OTOA 0 . 0050 .0130 .O'^OO 0.0180 .0180 0180 fo.oioo to \ .0000 io.0050 .0050 R^C^ .0400 .oi?6 .0400 .0108 The salts, acids, and base used and their volume-molecular concentra- tions were as follows : Series A. — Dipotassium phosphate (K2HPO4), 0.0180 m.; sodium nitrate (NaNOj), O.OIOO tn.; calcium chlorid (CaClj), 0.0025 m.; magnesium sulphate (MgSO^), 0.0025 m.; sodium hydroxid (NaOH), o.cooo to o.oioo m.; and citric acid (H3C8H5O7), o.oioo to 0.0000 m. Series B. — Potassium sulphate (K2SO4), 0.0040 m.; potassium nitrate (KNO3), O.OIOO m.; CaClj, 0.0025 m.; MgS04, 0.0025 m., phosphoric acid (H3PO4), 0.0180 m., sodium hydroxid (NaOH), 0.0000 to 0.0360 m. To each 500 cc. of culture solution there were added 5 drops of a ferric phosphate solution containing 0.25 gm. of FeP04 per 100 cc. VARIATION OP REACTION The ideal method of adjusting the reaction in such a series of cultures 1 Reference is made by number (italic) to " Literature cited," p. 93-95. Apr. 15, 1920 Effect of Solution Reaction on Germination and Growth 75 would be one which would permit a variation in unit steps over the desired range and at the same time produce solutions of sufficient stability to prevent small changes in the total amount of acid or base from seriously affecting the reaction. In other words, the solution should have a " buffer " nature. In the titration of strong acids with strong bases, a point is reached, as neutrality is approached, at which further additions of small increments of base produce voxy rapid decreases in the hydrogen-ion concentration. This corresponds to a rapid rise in the voltage curve obtained in the electrometric titration of such solutions. Any solution selected within this region of rapid change is unsuited to work requiring constancy of reaction, particularly when subject to possible small changes in total acidity. With acids and bases of low dissociation this difficulty is not so marked, changes of reaction being much less abrupt under similar conditions.^ Such solutions are commonly said to possess a buffer nature and are well adapted to work similar to that herein reported. In series A the reaction was varied by adding HgCg and NaOH to the successive cultures in amounts equivalent to the following volume- molecular concentrations : Culture No. 3 4 5 6 7 H3C0K5O7. NaOH. M. M. 0. 0100 0. 0000 . 0080 . 0020 . 0060 . 0040 . 0040 . oc6o .0030 . 0070 . 0020 . 0080 . 0000 . CIOO The reaction curve as determined by the hydrogen electrode for this series is shown in figure i, A.^ It mil be noted that this solution pos- sesses sufficient buffer action to prevent any rapid changes in reaction with change in total content of acid and base. ' For a more complete discussion of this subject see Hillebrand {lo). 2 The measurenients of hydrogen-ion concentration were made by means of the gas chain and hydrogen electrode, using the potentiometer system and measuring electromotive force to o.oooi volt. For electro- metric titrations a special cell equipped with mechanical stirring device was designed. 164176''— 20 3 76 Journal of Agricultural Research Vol. XIX, No. 3 3*~ /QO O eo CO 40 20 o cc 0.//\/ C/fr/c Acid per Liter 20 40 60 60 /OO CC. 0./ A/ NaOH per Lifer 3 -4 5 6 Cu/ft/re A/ umber Fig. 1. — A, graph showing the relation of reaction to the contents of HaCsHsO? and NaOH employed in the cultures of series A; B , graph showing the change of reaction found after 4 days' growth of wheat seed- lings iu series A. Apr. 15. 1920 Effect of Solution Reaction on Germination and Growth 7 7 In series B the reax;tiou was varied by adding to all cultures suflficient H3PO4 to make the solution 0.0180 molecular and then NaOH in the following volume-molecular concentrations : CS ^-0.5 Culture No. NaOH. I M. 0. 0000 .0144 .0174 .0181 . 0198 .0288 .0360 2 •2 4. e 6 /o /3 20 e^s 30 cc. Norma f NaOH per Liter JJ 40 2 34 3 Culture /dumber Fig. 2. — A, graph showing the change in reaction obtained by electrometric titration in series B; B, graph showing the change of reaction found after 4 days' growth of wheat seedlings in cultures of series B. The electrometric titration curve for this solution (fig. 2, A) was used as a basis for determining the amounts of NaOH necessary to produce a series of seven cultures ranging from about 2 Pq to about 8 Pg and increasing in approximately equal steps of i Pg. The curv'e shows a rather abrupt rise at a point representing the complete neutralization of one hydrogen ion of the H3PO4 molecule. As will be shown later, the solutions chosen upon the steep part of the cur\'e Avere less stable in reaction than those chosen upon the more nearly horizontal parts of the cur\^e. 78 Journal of Agricultural Research voi. xix.no. a OSMOTIC CONCENTRATION The osmotic concentrations of the solutions were not determined, because the data on the electrolytic dissociation of the component acids and salts under the variety of reactions used is not available and the authors did not have access to the necessary apparatus for making cryo- scopic determininations. However, the relatively small change in total volume- molecular concentration within either series would indicate that little, if any, difference in growth within a given series should probably be attributed to the osmotic factor. WATER EMPLOYED All cultures v/ere made from distilled water which had been rendered nontoxic by treating with carbon black as first recommended by Livingston {14). TECHNIC OF GERMINATION AND GROWTH OF SEEDLINGS The seeds of wheat, soybeans, and corn were germinated by supporting them upon a paraffined wire gauze which was floated by means of corks so that it was just even with the surface of nontoxic distilled water con- tained in a porcelain enameled pan. The seedlings were transferred to the various cultures when the plumules had attained a length of from 4 to 5 cm. The alfalfa seeds were germinated upon pads of filter paper in Petri dishes and transferred to the cultures after the seedling had attained a length of about 4 cm. The wheat and alfalfa seedlings were grown in Non-Sol and Pyrex beakers holding 250 cc. of culture solution and were supported upon per- forated caps of paraffined cheesecloth according to the method of Haas (7). The corn and soybean seedlings were grown in 8-ounce jars of flint glass and supported with corks according to the method of Tottingham {26). All beakers and jars were covered with black paper to exclude light. The solutions were renewed on all cultures every fourth day, and the glassware was thoroughly cleansed and sterilized before being used again. The reactions of the solutions used for growing wheat seedlings in both series were determined both before and after the 4-day periods. It was found that the successive solutions made up for a given reaction varied from each other by negligible amounts, so the solutions used for the growth of soybean, corn, and alfalfa seedlings were tested only at irregular intervals. EXPERIMENTAL DATA AND DISCUSSION OF RESULTS SERIES A Wheat seedlings were grown for a period of 1 6 days in solutions having the composition given for series A. Growi:h was determined by taking the green weight of roots and tops, exclusive of seeds. Twelve seedlings were grown in each culture, and all seven cultures of the series were dupli- cated. The duplicate cultures agreed closely in all cases and are there- Apr. 15, 1920 Effect of Solution Reaction on Germination and Growth 79 fore not reported separately. The green weights obtained for tops, roots, and entire plants, exclusive of seeds, are given in Table II. The average reaction of each culture at the beginning and at the end of the 4-day pe- riods and for the entire 1 6 days is also included in the table. The relative total green weights, based upon the highest, taken as 100, are shown in figure 3 plotted against the average Ph of the solutions, and the appear- ance of the seedlings at time of har\"esting is shown in Plate 15, A, B. "^ s e 7 a React/on as P^ Fig. 3. — Graph showing the relation of growth of wheat seedlings to reaction in series A. Table II. — Average reaction of cultures in series A and green weights of plants grown for period of 1 6 days Culture No. Average reaction of culture. Before growth. I 2 3 4 5 6 7 3.18 3-44 4. 22 5. 26 6. IS 6.61 7.28 After growth. Pn- 3-29 3.80 5.10 7. 02 7.21 7.24 7-34 Entire period. Green weight of 10 plants. Tops. Gm. 0. 410 I- 345 1.440 •570 •737 1. 087 1-375 Roots. Gm. o. 049 .079 . 204 .049 .091 .171 ■33^ Entire seedling. Gm. 0.459 1-425 1.644 . 619 .828 1.258 I. 706 8o Journal of Agricultural Research Vol. XIX, No. a A brief consideration of the results obtained in this series shows them to be abnormal, since one would scarcely expect the decided drop in growth in cultures 4, 5, and 6 if reaction were the only factor concerned. The fact that there developed a decided opalescent or colloidal appearance in these cultures in about 24 hours after their renewal, together with the fact that there was a large decrease in acidity during the 4 days ' growth of seedlings indicated that they were infected with some bacterial organ- ism which evidently used the citric acid present as a source of energy. Microscopic examination of these solutions showed this to be the case, and it was at once surmised that the depressant effect of these solutions upon the growth of wheat seedlings was probably due to the assimilation of the nitrates by these bacteria. This hypothesis was substantiated by a determination of nitrates in all seven cultures at the end of a 4-day period. The relative total green weights of seedlings, based upon the highest taken as 100, the relative nitrate content, based upon the highest taken as 100, and the relative decrease in acidity of the solutions, based upon the greatest decrease taken as 100, are shown in Table III. The relation of the change in reaction taking place in the 4-day period to the original reaction of the solution is shown graphically in figure i,B. Table III- — Comparative total green weights, nitrate content, and acidity of cultures of series A at end of 4-day period Solution No. I. 2. 3- 4- 5- 6. 7- Relative Relative Relative yield (green amount of decrease in weights of nitrates at acidity (in- whole 'end of 4-day crease in plants). period. Ph). Gm. Gm. 26. 9 84.0 6.2 83-5 92.8 22. 2 96.4 78.0 50.0 36.3 6.0 100. 00 48.5 7.8 60.3 73-7 24. 0 35-8 100. 0 100. 0 3-4 The data show that depression in growth in cultures 4, 5, and 6 is associated with low amounts of nitrates left in solution and with large decrease in acidity. It seems safe, therefore, to conclude that the bacteria present were responsible for the abnormal effects obtained in this series. It should be noted that although there was more citric acid available to the bacteria in culture No. 3 than in No, 4, there was actually much smaller assimilation of nitrates in the former culture, while the wheat growth in No. 3 was almost equal to that in the best member of the series. Apparently the acidity of this culture has suppressed the growth of the nitrate-assimilating bacteria but has not had a correspondingly unfavor- able effect on the growth of wheat seedlings. Since there was little dif- ference in the amounts of nitrates present in cultures 1,2, and 3 it seems Apr. 15, 1920 Effect of Solution Reaction on Germination and Growth 8 1 probable that the depression in growth found in cultures i and 2 was due to the physiological effect of their reaction upon the wheat seedlings. The results obtained from this series do not give accurate data con- cerning the effect of reaction upon the growth of wheat seedlings over the entire range investigated. It seemed well, however, to include them in this report on account of their bearing upon a large amount of investigative work showing the ability of bacteria and fungi to compete with higher plants for inorganic nitrogen if supplied with a proper source of energy and carbon in the form of organic matter. This power of micro- organisms has been demonstrated by numerous investigators under both solution and soil-culture methods. For a more complete discussion and an extensive bibliography on this subject the reader is referred to the publication of Doryland (4). ^' 0^' , ^ / ^ \x "^ , . / / ^ \ \x \ / / V— -X;;:^ ^=^' _ /A t3 [TV IX * V. ^ - /'' • / § , '// ■•? Who^-h Soybean / Com Alfa/fa 1 /oo 90 ^ SO \30 20 /O O 2 3 4^6 7 G Reaction as P^ Fg. 4. — Graphs showing the relation of growth of wheat, soybean, com, and alfalfa seedlings to reaction in series B. SERIES B On account of the difficulties arising from bacterial infection when citric acid was employed in the cultures, further work was confined to solutions having the composition given for series B. Wheat, soybeans, corn, and alfalfa seedlings were grown, all cultures being duplicated in the wheat, corn, and alfalfa series and quadruplicated in the soybean The numbers of seedlings grown in each culture were as follows: series. wheat, 12; soybean, 6; corn, 4; alfalfa, 20. The following periods of 82 Journal of Agricultural Research Vol. XIX, No. 3 growth were maintained: wheat, i8 days; soybeans, i6 days; corn, 8 days; alfalfa, 20 days. In Table IV are given the green weights of seedlings at time of harvesting and the average reaction of each culture as shown by the determinations of hydrogen-ion concen- tration made at the beginning and end of the 4-day periods on the cultures of the wheat series only. In figure 4 the relative total green weights, based upon the highest weight taken as 100 in each instance, are shown plotted against the average Ph of the cultures. Plate 15, C, shows the appear- ance of the wheat plants at the time of harvesting. Table IV. — Average reactions of cidhires of series B and green weights of seedlings at time of harvesting WHEAT Culture No. 3 4 5 6 7 Reaction. Ph- 2. 17 2. 96 4. II 5-16 5-94 6.97 7.71 Green weight of lo plants. Tops. Gm. o. 230 1-730 2. 548 3- 581 3. 620 2. 421 2. 103 Roots. Gm. "■ o. 067 • 143 • 149 • 372 • 356 • 324 . 141 Entire plants exclusive of seeds. Gm. O. 297 873 697 953 976 745 244 SOYBEANS Culture No. Reaction. '' Green weight of 10 plants (entire). Ph- 2.17 2. 96 4. II 5-16 5-94 6.97 7.71 Gm. "4-93 7. 62 15-76 18.54 18.74 14-13 12.87 e ... .J 6 . . o Seedlings dead at time of harvesting. * Because of the uniformity of reaction of successive cultures made up to to represent a given reaction and the relatively small changes in reaction produced by growth of seedlings it is assumed tbat the average reactions found in the wheat series apply to the cultures of the soybean, corn, and alfalfa series. Occasional determinations on cultures of the latter series showed this to be true. Apr. 15, i9:!o Efject of SoluHofi Reaction on Germination and Growth 83 Table IV. — Average reactions of cultures of series B and fjreen Heights of seedlings at time of harvesting — Continued Culture Ko. Reaction, a Green weight of 10 plants, exclusive of seeds. I Pe- 2. IT Gvt. 2. 96 5. 66 4. II 8.98 5- 16 13.30 5. 94 j 10. 14 6. 97 10. 03 7. 71 lO- A7 X A 6 7 ^' 1 ALFALFA Culture Kg. Reaction." Green weight of 10 plants (entire). I . Ph- 2.17 2. 96 4. II 5-x6 5-94 6.97 7.71 Gm. (^) (^) 0.317 •397 .496 •435 .310 'I A e 6 7 a Because of the uniformity of reaction of successive cultures made up to represent a given reaction and the relatively small changes in reaction produced by growth of seedlings, it is assumed that the average reactions found in the wheat series apply to the cultures of the soybean, corn, and alfalfa series. Occasional determinations on cultures of the latter series showed this to be true. '' Seedlings dead at time of harvesting. Before discussing the foregoing data mention should be made of the fact that while in practically all cases duplicate cultures agreed closely, there was occasionally considerable variation between the individual plants in a single culture of soybeans and corn, while in alfalfa there was considerable mortality among the plants in all cultures of the series. For this reason in drawing conclusions from the foregoing data the authors prefer to consider the work \\-ith soybeans, com, and alfalfa as somewhat preliminary in nature. This does not apply to the wheat series, where no significant variations were found between the plants in the cultures representing a given reaction. The effects of acids and alkalies upon seedings gro\\Ti in solution cul- ture have been quite extensively investigated by Kahlenberg and True (jj), Heald (9), Cameron and Breazeale (2), Hartwell and Pember (t Exp. Sta. Rec, V. 29, no. I, p. 26. 1913. Original article in Bui. Soc. Nat. Agr. France, t. 72, p. 916-922. 1912. Not seen. (24) Shive, John W. I915. A STUDY OP PHYSIOLOGICAL BALANCE IN NUTRIENT MEDIA. In Physiol. Researches, v. i, no. 7, p. 327-397, 15 fig. Literature cited, p. 396-397. (25) I918. TOXICITY OP MONOBASIC PHOSPHATES TOWARDS SOYBEANS GROWN IN SOIL AND SOLUTION-CULTURES. In Soil Sci., V. 5, no. 2, p. 87-122, 5 fig. References, p. 121-122. (26) ToTTiNGHAM, William E. 1914. A QUANTITATIVE CHEMICAL AND PHYSIOLOGICAL STUDY OF NUTRIENT SOLUTIONS FOR PLANT CULTURES, hi Physiol. Researches, v. i, no. 4, p. 133-245, 15 fig. Literatiure cited, p. 242-245. PLATE 15 A. — Method of growing wheat seedlings. (Paper covers removed from beakers.) B. — Appearance of wheat seedlings in series A at time of harvesting. C. — Appearance of wheat seedlings in series B at time of harvesting. (96) Effect of Solution Reaction on Germination and Growth Plate 15 12 3 4 5 6 7 :W I I M 1 1 Journal of Agricultural Research Vol. XIX, No. 2 Vol. XIX IvIAY 1, 1920 No. 3 JOURNAL OF AGRICULTURAL RESEARCH CONXKNTS Pag* Philippine Downy Mildew of Maize - - - - 97 WILLIAM H. WESTON, Jr. (Contnibotlon from Bureau of Plant Industry) Effect of Drugs on Milk and Fat Production - - - 123 FRANK A. HAYS and MERTON G. THOMAS (Contribution from Delaware Agricultural Experiment Station) Artificial and Insect Transmission of Sugar-Cane Mosaic - 131 E. W. ERANDES (Contribution from Bureau of Plant Industry) PUBUSHED BY AUTHORITY OF THE SECRETARY OF AGRICULTURBt WITH THE COOPERATION OF THE ASSOCUTION OF LAND-GRANT COLLEGES WASHINGTON, D. C. WASHINaTOH I OOVCRNMeMT rMINTINa OmOl I Itl* EDITORIAL COMMITTEE OF THE UNITED STATES DEPARTMENT OF AGRICULTURE AND THE ASSOCUTION OF LAND-GRANT COLLEGES FOR THE DEPARTMENT KARL F. K^LI/ERMAN, Chairman J. Physiologist and Associate Chief, Bureau of Plant Industry EDWIN W. ALLEN Chief, OMce of Experiment Stations CHARLES L. MARLATT Entomologist and Assistant Chief, Bureau ofEntomoloiff FOR THE ASSOCIATION G. LIPMAN Dean, State College of Agriculture, and Director, New Jersey Agricultural Experi- ment Station, Rutgers College. W. A. RILEY Entomologist and Chief, Division of Ento- mology and Economic Zoology, Agricul- fural Experiment Station of the UnrversUy of Minnesota R. L. WATTS, Dean, School of Agriculture, and Diredor, Agricultural Experiment Station, The Pennsylvania State College. All correspondence regarding articles from the Department of Agriculture should be addressed to Karl F. Kellerman, Journal of Agricultural Research, Washington, D. C. All correspondence regarding articles from State Experiment Stations should be addressed to J. G. Lipman, New Jersey Agricultural Experiment Station, New Brunswick, N. J. t.'fcJW YORK 7--* JOINAL OF AGRICETDRAI ISEARCH Vol. XIX Washington, D. C, May i, 1920 No. 3 PHILIPPINE DOWNY MILDEW OF MAIZE By William H. Weston, Jr.^ Pathologist in Charge of Downy Mildew Investigations, Office of Cereal Investigations, Bureau of Plant Industry, U^iited States Department of Agriculture During the past 20 years there have been reported from the Orient several downy mildew diseases of maize, sugar cane, and other economic grasses, caused by members of the genus Sclerospora of the Peronospo- raceae. The most recently noted of these has been found in the Philip- pine Islands, where it causes very serious damage to maize, a crop which in area under cultivation is second only to rice. In 1916, a brief note by Prof. Baker (7)^, of the College of Agriculture, first mentioned the occurrence and destructive power of the disease. In 191 8, a short de- scription of it with drawings of the causal fungus was published by Reinking {17). No further information concerning this dangerous dis- ease has been published, but it is known that it occasions heavy and constant losses in the maize crop of this the richest of our oriental pos- sessions and represents a grave potential menace to this extremely valuable crop of our own country. The danger of the introduction of this disease to the cornfields of America was felt to be sufficiently grave to warrant a full investigation in the Philippines and elsewhere in the Orient. By such research it was expected to determine the distribution and life history of this organ- ism and to devise methods of control. With these data in hand, the chances of promptly checking the disease were greatly increased should it gain a foothold in the United States at any time in the future. In the meantime, a quarantine was established against the importation of corn from the Orient. It was the privilege of the writer to be detailed to this investigation, and since April, 191 8, he has been at work on it in the Philippines. The following paper deals with the generalfeaturesofthedisease and with the characteristics of the causal Sclerospora and its systematic position * The writer wishes to express his thanks to Dean Baker, Prof. Reinking, Prof. Elayda, and others at the College of Agriculture of the University of the Philippines for so generously furnishing laboratory facilities, land, and other assistance; to Mr. S. Apostol, of the Philippine Bureau of Agriculture, for infor- mation on the distribution of the disease; and to Dr. E. D. Merrill, of the Bureau of Science, for many courtesies which have aided the progress of this investigation. ' Reference is made by number (italic) to " Literature cited," p. 121-122. Journal of Agricultural Research, Vol. XIX, No. 3 Washington, D. C. May i, 1930 t2 Key No. G-189 (97) ^8 Journal of Agricultural Research VoLxdcno-s or relationship to other downy mildews destructive to cereal, forage, and sugar-cane crops in the Orient. DISTRIBUTION Broadly speaking, the disease is distributed throughout the Philip- pine Islands. Through the personal observation of the writer and through information given by Dr. Reinking, of the College of Agriculture, by his students, and by members of the Bureau of Agriculture, the dis- ease is known to exist in the Cotobato Valley of the Island of Mindanao at the south, in the Islands of Cebu and Occidental Negros, and in the provinces of Batangas, Laguna, Rizal, Cavite, Bulacan, Tarlac, Pam- panga, Nueva Ecija, Pangasinan, Ilocos Norte, la Union, and Isabela in Luzon at the north. In some of these localities the disease appears to have been present for more than lo years, but as yet not enough is known to warrant a discussion of its probable origin. DESTRUCTIVENESS The disease is unusually destructive. It is impossible for one accus- tomed only to the comparatively light losses occasioned by the maize diseases of the United States com belt to form any conception of the epidemic intensity of the attacks of this downy mildew under favorable conditions, or of the terrible destruction which it occasions (PI. i6). Of the aggregate loss to the $8,820,000 maize crop of the Philippines no estimate can be made, because farmers do not recognize the trouble as a disease but regard it as the result of excessive rain or other unfavorable conditions and accept it with fatalistic resignation. In Laguna and Batangas, however, where maize is a major crop and where the writer has studied the disease in the native fields, losses of 40 to 60 per cent are frequent, and in some cases as high as 82 per cent of infection has been counted. In the experimental and acclimatization plots at the College of Agriculture, where the growing of unacclimatized varieties and the constant presence of actively infecting plants combine to make the con- ditions especially favorable for infection, the losses ordinarily are high. In several beds of United States sweetcorn, planted during the rainy season, every plant was killed before producing any seed. The severity of the disease in the individual corn plant varies with con- ditions from the extreme stunting and weakening of the plant resulting in death about one month after planting to the less virulent attacks in spite of which the plant shows a fair growth and ultimately produces a small, more or less poorly formed ear. Even in the few lightly affected cases the grain production is not nearly normal, and in most cases com- plete barrenness or premature destruction occurs, so the aggregate loss in the average field attacked by the mildew is large. In some localities corn growing has been abandoned for the culture of upland rice because of the ravages of the disease. Moreover, this loss can not be offset in part May 1, 1920 Philippine Downy Mildew of Maize 99 by using diseased plants for fodder, for cattle appear to dislike the taste and will not eat the infected plants unless they are mixed with a liberal proportion of the healthy. One of the most serious features of the attack by the downy mildew is that the infected plants are rendered susceptible to the attacks of a num- ber of secondary parasitic organisms which contribute to the destruction of the weakened plants. In the rainy season there frequently occur de- structive rots of the stem, ear, and shank, with which at least two species of Pythium and bacteria appear to be associated, while a species of Hel- minthosporium, which is only occasionally severe on healthy plants, is usually very destructive to plants weakened by downy mildew. SYMPTOMS The effect of the disease on the corn plant varies greatly with such con- ditions as the age of the plant when infected, the means by which infection takes place, the varietal nature and individual condition of the host, and the environmental conditions which accompany and follow infection. As a result, no small, clearly defined group of symptoms can be described which will entirely cover the effects of the disease on the host. In general, however, the disease may be said to manifest itself by the loss of chlorophyll in more or less sharply defined areas of the leaf, by the production of a whitish down of conidiophores principally on the chlo- rotic area, and by a more or less extensive alteration in the form or the normal growth of the plant. The change in color is the most striking and obvious symptom. Since, however, somewhat similar changes in color and form may result from other causes, the characteristic downiness is the surest indication of , the disease. The effects of the disease may appear at any time from the putting out of the third or fourth leaf to the formation and maturing of the tassel and ear, but in any case the tissue of the host is thoroughly invaded by the mycelium before any external signs appear. When appearing early in the development of the plant the symptoms are as follows: The second, third, or perhaps the fourth leaf, when nearly developed, shows at the base two or three rather narrow, longitudinal stripes of a pale yellow to whitish color (PI. A) with the exception of which the leaf is quite the normal green. However, the two or three leaves already partly developed above this, and all the leaves which subse- quently appear, are almost completely whitish or pale yellow. Moreover, these leaves never attain the normal shape and size but remain much narrower and become rigid, so that they ascend stiffly instead of bending in the normal flexible manner (Pi. A). The growth of the stem is also checked, so that the plant becomes more or less dwarfed. As the growth of the leaf sheaths is not decreased proportionately, they often deeply overlap to form a cover which may inclose and even project beyond the stunted tassel (Pi. 18, A). The root system also is usually affected so that lOO Journal of Agricultural Research voI.xix.No.s it does not develop properly but becomes stunted and functionally inadequate. The subsequent fate of an infected plant varies with conditions. In the rainy season it almost invariably succumbs rapidly to secondary in- fections by species of Pythium, Helminthosporium, or Fusarium. In the dry season, however, although it usually turns brown, withers, and soon dies, such a plant may struggle along to the tasseling state and may even produce a stunted ear with occasionally a few grains. When the disease appears later in the development of the corn plant the symptoms are as follows : The first leaf to show any signs of the disease, which may be the fourth or fifth or even up to the eighth, will have at the base pale stripes similar to, but more extensive and broader than, those described for plants which show the disease early (Pi. B). All subse- quent leaves show a somewhat similar striping but in a progressively more marked degree, the markings on each successive leaf being more extensive than those on its predecessor and running nearer the tip, while the last leaves are striped throughout their entire length. The shape of the stripes varies greatly. On the lower leaves they are usually merged at the base into a solid yellowish white area from which irregular elongations run up into the normal green towards the tip of the leaf (PI. B). On the middle leaves the solid yellowish w^hite area at the base is somewhat smaller in extent, but the prolongations from it run more nearly to the tip, while on the upper leaves these discolored stripes extend from the base to the tip of the leaf but are more broken and irregu- lar and even merge laterally and anastomose so that a marbled or mottled appearance is given to the otherwise green leaf. The shape and size of these leaves, however, is very little altered, and they usually have the breadth and flexibility which characterize leaves of the normal plant. At times, however, the midribs become brittle from the invasion of the fungus mycelium, break where they join the sheath, and hang straight down along the stem (PI. 17). The growth and structure of the stem are often normal, and the root system is strong and well developed. It is in the reproductive structures of these later- infested plants that the injurious effect of the disease is shown especially. The tassel, although usually appearing at the normal time and often seemingly unharmed structurally, may show decreased production of pollen and frequently is extremely malformed (Pi. 17, A). The ear also is even more seriously affected. Even a mediocre ear is a very rare occurrence (in i out of 150 diseased plants), while customarily the ear is more or less completely sterile and malformed (PI. 20). This malformation of the reproductive structures is of frequent and regular occurrence in maize infected by the Philippine downy mildew. In plants attacked at all ages by the disease there is induced a great variety of the most remarkable malformations and monstrosities of the ear and tassel. These show a wide range of the fasciations, phyllodies. May 1, 1920 Philippine Downy Mildew of Maize loi reduplications, virescences, and other abnormalities of the various catego- ries of monstrous growths that are recognized in teratology. Less frequently also the vegetative parts of the infected plants show abnor- malities induced by the disease, fasciations and torsions of the stem (PI. 17, B) and shank (PI. 19, A) being most common. These abnormal- ities, of course, are frequently induced by other diseases and by unfavor- able conditions of the environment, but their occurrence in connection with the downy mildew is so common as to form an accessory symptom of diagnostic value. One other marked effect of the disease is the delaying of ear produc- tion. Normal plants in a plot invariably will bear well-developed ears in the "milk" or "glazing" stage before the diseased plants have devel- oped ears to the "silking" stage. It should be noted that the loss of chlorophyll and the consequent yellowish or whitish color of the marked areas, which is so characteristic a symptom of the disease, is by no means permanent but serves particu- larly to point out the earlier stages of the attack. As the diseased plant matures, however, and the fungus begins to terminate its period of spore production, the marked areas become more and more green, the contrast between the normal green and the paler portions of the leaf becoming less and less distinct until, finally, in plants less heavily at- tacked, the marked areas may so far regain their green color as to be almost indistinguishable from the normal. All these plants which show the disease at a late date do not necessa- rily undergo rapid destructioft as in the cases of early attack. On the contrary, although the plants are more susceptible to the secondary infections than are their healthy companions, they may mature along with them, drying and withering at a date only slightly earlier than normal. In some cases the infected plants seem stimulated by the downy mildew to prolonged activity and show persistent and excessive growth of husks, or of bracts in the deformed tassels, after adjacent plants are withered and dry (PI. 19. A). The susceptibility to infection is greatest in the young seedling and decreases markedly as the plant develops, so that by the time it has tasselled and is forming ears its tissue is, as a rule, too mature and resistant to permit infection. If, however, as is frequently the case in some varie- ties, the main plant sends out secondary shoots or suckers, these may rapidly become infected (PI. 19, ^), and through them the infection may spread to the main plant even though it is so far matured as to have its kernels hardening. When attacked in this way, the mature plant shows symptoms different from any of those described above. The lower leaves are inconspicu- ously marked throughout their length with narrow, pale, yellow-green to rusty green stripes, which are not continouous but are irregularly broken and interrupted. On the middle leaves, as a rule, the markings J02 Journal of Agricultural Research voi. xix. N0.3 are similar in character but occupy principally the more distal part of the leaves, while the upper leaves are either entirely unmarked or have the striping confined to a small part of the leaf tip. Since most of the parts of the plant are matured, they show no change in form as a result of the infection, but the ear, if not mature, may elongate slightly and project from the husks at the tip (Pi. 19, B). It is to be noted that a plant thus attacked, in contrast to those previously described which are infected early, is marked least extensively and conspicuously on the upper leaves and most extensively on the lower leaves, has ears little if at all altered, and bears no conidiophores on the marked areas. The production of conidiophores on the diseased plant is, of course, a symptom valuable in recognizing the disease (Pi. 21). Unfortunately, however, the process of conidiophore formation takes place almost exclusively at night and is controlled largely by conditions of the environ- ment. The details of this relationship will be given later. It need only be said here that a plant may be attacked heavily by the fungus, the mycelium of which invades its tissues throughout, and may show the changes of color and growth which are characteristic of the disease and yet never form conidiophores and conidia unless external conditions are favorable. A comparison of the symptoms of the Philippine downy mildew with those which characterize the downy mildew of maize in other countries shows many similarities. In the closely related Javan mildew of maize, Palm (75) has recognized three distinct sets of symptoms. Of these, the symptoms of type A correspond in general to the description given above for plants attacked early in life, while the symptoms of type B correspond to those of plants attacked later. Type C, however, is characterized by narrow, incon- spicuous stripes of a dark brown color running, the -full length of the lowest leaves and decreasing in extent on successively younger leaves until the last marked leaves show these stripes at the tip only, while the still later leaves are of the normal green throughout. No specimens corresponding exactly to the description and illustra- tions of Palm's type C have been seen in the study of the Philippine maize mildew. Occasional stripings of this sort have been observed on the lower leaves of plants whose upper leaves showed the general or restricted discoloration already described. Maturing plants infected through suckers have shown inconspicuous, d!P.rk orange-colored stripings, exten- sive on the upper leaves and decreasing in area on the lower. In no case, however, has an immature plant been seen with these dark markings of the leaves decreasing oft successively younger leaves until the latest are untouched. For the Philippine maize mildew it does not seem justifiable to attempt to make such hard and fast categories as the types A, B, and C of Palm, although the symptoms shown by many plants can be more or less May 1, 1920 Philippine Downy Mildew of Maize 103 roughly grouped under the types described above. The discolorations, growth changes, and other efifects of the disease all differ markedly in accordance with time of infection, the varietal and individual character of the plant attacked, and the conditions of the environment. Hence there are encountered not only such diseased specimens as can be in- cluded conveniently in the three types recognized by Palm but also some plants which show symptoms intermediate between the types and others which show various combinations of these symptoms. The occurrence of such sharply marked categories of symptoms as those described by Palm might with some justice be suspected to be the manifestation of different biologic strains of the causal fungus. In the Philippine maize-mildew, however, cross inoculations with spores from infected plants corresponding to Palm's symptom types, as well as biometric studies of the spores and conidiophores from these plants, dis- prove this assumption. Moreover, a series of experiments in which sev- eral varieties of maize were inoculated in various ways at different ages and subjected to different environmental conditions, although not en- tirely completed, has shown that the changes of color and growth pro- duced in the plant by the disease differ with variations in these factors. All the evidence of field observations Also supports this conclusion. In general, then, while the symptoms of the Philippine maize-mildew resem- ble those of the Javan, they appear to be much more varied and less easily grouped into sharply defined categories. In the related Formosan downy mildew, Miyake (74) has described in detail only the symptoms shown by attacked sugar cane, which is the host most severely affected. He states that in maize the stripes are not particularly pronounced and the plant is not noticeably hindered in growth, for it ripens and shows only a slight decrease in yield. While this description would fit occasional plants attacked by the Philippine maize-mildew, it by no means depicts adequately the injurious effects in even the average case and would seem to indicate that the Formosan mildew is far less destructive to maize than is the Philippine. Upon comparing the maize downy mildew of the Philippines with that of British India described by Butler (4), it is to be noted that the symp- toms of the latter resemble in general those seen in the Philippines, al- though in the Indian disease more emphasis is laid on the checking of the intemode growth and the consequent stunted appearance of the at- tacked plants than seems to be warranted from observation in the Philippines. Moreover, although the maize-mildew of British India has been present in that country since 191 1, it has,in marked contrast to the Philippine disease, caused only slight sporadic injury. HOSTS Under field conditions throughout the Philippine Islands maize is the only crop, on which the downy mildew occurs with sufficient severity to attract attention or to occasion appreciable loss. In the trial plots at I04 Journal of Agricultural Research voi. xix.no. 3 the College of Agriculture, however, where many kinds of cereal, forage, and cane crops were grown under conditions favoring infection, the downy mildew was found also to attack teosinte {Euchlaena luxurians Schrad.) and sorghum (Andropogon sorghum (Linn.) Brot.). With teosinte, the percentage of infection and resulting loss is not quite so great as with maize, and the symptoms are less pronounced (PI. 22, C), since the attacked individuals, especially those showing the disease late in their development, are much less conspicuously marked and are very seldom appreciably deformed, while the conidiophores are more scattered and more scantily produced. As might be expected, hybrids resulting from the crossing of maize atid teosinte are also susceptible to the disease, the degree of suscep- tibility and the effect on the plants attacked being intermediate between those shown by the two ancestors. In sorghum the percentage of infection is very low, and the few plants infected are easily overlooked, because they turn pale when still very young (PI. 22, B), bear but few conidiophores, and wither and die after a brief period of weak, stunted growth. No cases of individuals more con- spicuously marked or deformed, in which the disease appeared later, were ever seen; and the loss was limited to the destruction of the few attacked plants. Cross-inoculation experiments and the biometric study of spores and conidiophores show that the same causal fungus is involved in all these cases. In view of this condition, it would naturally be suspected that other members of the Maydeae and Andropogoneae might also prove sus- ceptible to the disease. So far, however, in spite of extensive search, no such Sclerospora, characterized by a conspicuous and rapidly spreading conidial stage, has been found in this region under natural conditions on the many wild grasses related to maize. However, the writer has found on Saccharum spontaneum L., a very common wild grass here, a Sclero- spora which although of very frequent and widespread occurrence produces only the characteristic thick-walled resting spores. Further description of this Sclerospora will be given in a later paper, but it should be said at this point that this oogonial form on wild grass does not appear to be connected with the conidial form growing on cultivated maize, sorghum, and teosinte. Moreover, inoculations such as were successful in the case of maize, sorghum, and teosinte have so far failed to accomplish the transfer of the disease to other related Gramineae — namely, Coix lachryma-jobi L., Philippine, United States, and Hawaiian strains; Coix ma yuen, Philip- pine and United States strains ; several varieties of sugar cane {Saccharum officinarum L.), uba or Japanese cane (Saccharum sp.), and the native grasses, cogon (Imperata cylindracea L.), anias {Andropogon sorghum, var, halepense L.), and aguingay {Rottboellia exaltata L.)- In view of the May 1, 1920 Philippine Downy Mildew of Maize 105 difficulties in securing artificial infection, these negative results are by no means conclusive, although the successful infection of maize, sorghum, and teosinte under the same conditions would seem to indicate that these other relatives are far less susceptible. These inoculations will be detailed more fully in a later paper, but it should be said here that they were made from, about 2 a. m. until dawn because the spore production was found to take place at this time. It seems highly probable that spore production is nocturnal in the other related downy mildews of the Orient as well and that the uncertain results of inoculations with them has been due to the failure to use fresh spores. It is of interest to compare these results with those obtained for the other related downy mildews of the Orient. In Formosa, Miyake (14) successfully transferred Sderospora sacchari T. Miy. from sugar cane to maize and teosinte and vice versa, but was unable to infect rice, sorghum, wheat, or millet. In India, although their infection experiments were unsuccessful, Butler (j) and Kulkami (lo) note the occurrence on teo- sinte of the conidial stage of a Sderospora which they suspect may be identical with that of maize {Sderospora maydis (Rac.) Butler.) In Java, no extensive attempts were made by either Rutgers {ig) or Palm {15) to obtain artificial inoculation of other hosts with the Javan downy mildew of maize {Sderospora javanica Palm). They state, how- ever, that under conditions favoring infection in the field, neither sugar cane nor the common wild alang-alang grass {Imperata sp.) was found infected and that, although teosinte itself is immune, the hybrid between teosinte and maize is, if anything, more susceptible than the variety of maize from which it is derived. CAUSAL ORGANISM The fungus which causes this extremely destructive disease of maize in the Philippines belongs to the Peronosporaceous genus Sderospora, as Baker (i) and Reinking (77) already have reported. It should be noted, however, that it shows especially close relationship, not to the type spe- cies Sderospora graminicola (Sacc.) Schroet., which is distinguished by the germination of the conidia by zoospores and the abundant production of oospores, but to those other oriental members of the genus which are characterized by the germination of the conidia by tubes and the partial or complete lack of oospores. However, setting aside the question of the affinities of the fungus for a later discussion, its characteristics will now be considered. MYCELIUM As a rule, as soon as the maize plant shows any external indication of the disease, the mycelium is found to be quite generally distributed throughout the host tissue, the root being the only main organ which is not extensively invaded. Thisin vasion is most marked in the vegetative io6 Journal of Agricultural Research voi. xix.no. 3 parts of the plant but to a lesser degree affects the male and female inflorescences also. Since the mycelium is relatively inconspicuous, its course throughout the host tissue is followed with difficulty. However, by means of trans- verse and longitudinal sections cut in various thicknesses and stained with iron-alum-haematoxylin and eosin or with gentian violet or methyl blue it was possible to trace the relation of the hyphae to the host tissue. Moreover, by subjecting such sections to the processes of maceration, clearing, and subsequent staining used by Mangin (13) and Berlese {2) the host tissue was readily dissociated and cleared sufficiently to permit the examination of large sections of the mycelium. By these methods material was studied from all parts of plants in various stages of infec- tion, and the nature of the hyphae, their relation to the host tissue, and their location and abundance in different parts of the host were ascer- tained. The hyphae are most abundant in the discolored areas of the infected leaves but may be found throughout the plant in unmarked parts of the leaves, in the branch tips of the apparently unaffected tassel, and at the base of the seemingly healthy stem some feet below the first discolored leaf. In the leaf sheaths, leaves, and such modified foliar structures as the husks and glumes, the mycelium is most abundant among the cells of the bundle sheaths and in the mesophyll tissue (PI. 23, E), but occa- sional hyphae are found in the fundamental tissue and even among the elements of the bundles themselves. In the stem, ear shanks, cob, and tassel rachis, the mycelium follows the bundles, running for the most part parallel to them among the cells of the bundle sheath (PI. 23, B) and less frequently sending out hyphae more extensively into the surrounding fundamental tissue. In badly infected ears the mycelium usually runs out from the cob along the funicvlus of attachment into the undeveloped parts of the abortive kernels, and occasional hyphae are encountered even in the chaff, seed coats, and endosperm of the apparently healthy kernels, though not in the embryo itself. Wherever found, the hyphae are almost invariably intercellular in position, occupying even the smallest spaces between the cells, and even forcing adjacent cells apart as they grow between them. Occasionally hyphae were seen which apparently passed within the cells, but the interference of the host tissue was such that their position could not be ascertained with entire certainty. Since the size and shape of the hyphae are determined to a large extent by the nature of the intercellular spaces which they occupy, there is very little regularity in these characteristics in most cases. When separated by maceration, the hyphae are seen to be of two general types — namely, the long, slender, occasionally branching hyphae which lie alongside the vascular bundles in the stem and leaves' May 1. 1920 Philippine Downy Mildew of Maize 107 and the lobed, contorted, irregularly branched, gnarled, and crooked hyphae which run in and out among the mesophyll cells of the leaves (PI. 23, A). The first kind seem to serve for communication from one part of the host to the other and can be followed for considerable distances even in longitudinal section (Pi. 23, B). The second kind appear to act as a means of establishing connection with the mesophyll cells, especially with the bundle sheath, in order to derive nutriment therefrom, since they are found in every possible crevice in the most intimate contact with the host cells (PI. 23, A, E). Haustoria are produced by both types of hyphae but are best developed or most pronounced on the crooked assimilatory hyphae among the mesophyll cells in the leaf. In shape the haustoria are simple, papillate to tubular (PI. 23, F, G), as a rule, but they may be somewhat lobed (PI. 23, H). In no case, however, were such markedly digitate haustoria seen as those figured by Rutgers (19, PL 6) for the Javan Sclerospora. The haustoria penetrate portions of the host cell wall, against which the hyphae are closely appressed, and project into the lumen. Not only the cells of the mesophyll, bundle sheath, and pith are penetrated, but also occasional cells of the epidermis (Pi. 23, E, c) and even the xylem (PI. 23, E, h). In any case, the haustoria accomplish the penetration of the host cell without occasioning its collapse, although the wall often is wrinkled and the turgidity of the cell decreased, apparently through the extraction of its contents by the parasite. The chloroplasts of the parasitized cells are gradually destroyed through the action of the fungus, with the result that the badly infected areas lose their green hue and assume the pale yellow or whitish color symptomatic of the disease. Occasionally the host cell surrounds the haustoria of the parasite wnth a thick wall (PI. 23, F), as if in protective response to the injurious stimulus of the fungus, a condition observed also by Butler (5) in Sclerospora graminicola (Sacc.) Schroet. on Pennisetum. The hyphae are hyaline, rarely if ever septate, thin-walled, with gran- ular content, and vary greatly in size, 8 /x being perhaps the most common diameter. The haustoria are similar in structure and usually about 8')u long by 2 ju in diameter. In the larger air chambers which underlie the stomata, the mycelium develops somewhat irregular clusters of stout branches (PI. 23, E, a), from which, under favorable conditions, the conidiophore initials arise and grow out through the stomata to produce the conidiophores. CONIDIOPHORES Conidiophores may be said, in general, to be produced on any part of the plant save the roots. They occur on the main stem, on the leaves, leaf sheaths, and car husks, and on the main axis, branches, and glumes io8 Journal of Agricultural Research voi. xix, N0.3 of the tassel. Most commonly, however, the conidia appear on the leaves and leaf sheaths, where they occupy principally the conspicuous mottled and discolored areas which have been described. On whatever part of the plant they may be found, the conidiophores emerge at night, provided there is present a thin layer of dew, rain, or mist. Damp air alone does not seem to permit their formation. Under favorable conditions the process of conidiophore emergence and conidia production begins about midnight and may continue a few hours after dawn, provided the weather is favorably rainy. When seen at night in the luxuriance of their growth, the innumerable conidiophores projecting slightly from the thin film of moisture on the leaves form a very distinct grayish white down, which is by no means even suggested by the dry, matted fragments which remain when the hot morning sun has dried the surface of the leaves (Pi. 21, A). This process of conidiophore development and conidia production has never been described, and, since it shows several points of interest, it will be presented in detail in a subsequent paper. In general, however, it occurs as follows : From the stomata of the infected portion one or more club-shaped hyphae grow out. These elongate, and under favorable conditions the paired protrusions finally bud out from their tips and become the stout primary branches. From the tips of these in turn bud out the begin- nings of the secondary, and from these, at length, the tertiary branches, each of which usually terminates in one or two tapering sterigmata. Since the initial protrusions which develop into the branches arise almost invariably in pairs, the structure of the mature conidiophore is character- istically dichotomous, instances of the suppression or delayed formation of a branch being, on the whole, rather rare (PI. 24, D). Finally, from the tip of each sterigma there buds one conidium as a spherical protrusion which enlarges and lengthens until it attains the elongate oval or rounded oblong shape of the mature spore and is separated from the sterigma tip by a cross wall. When fully formed the conidiophore appears as in Plate 24, C, and con- sists essentially of a main axis which begins with an elongate basal cell and broadens gradually until it divides into the two to four stout main branches. From each of these extend two to four smaller secondary branches, each of which in turn bears two to four tertiary branches that terminate severally in one or two tapering sterigmata, each bearing at its tip a conidium. Although under favorable conditions the conidio- phores are of the large, well-developed type just described, they fre- quently show such variations in structure as the omission of the second and third series of branches and a general reduction in branches, sterig- mata, and consequently number of conidia (PI. 24, E). On vigorous conidiophores 32 to 96 conidia may be borne, while on poorly developed ones there may be as few as 8 or even 3. In size also there is great varia- May 1, 1920 Philippihe Downy Mildew of Maize 109 tion, the total length of the conidiophore even in abundant dew varying from 260 to 400 ju, although most commonly it is about 340 ju, while in scanty dew, such as occurs in the hot season, lengths of 160 to 200 ju are generally encountered. In either case, however, the greatest width, just below the branches, is from 15 to 26 ju. The sterigmata are con- sistently about 10 /i in length, with a diameter at the base of about 6 fj.. The basal cell is invariably present in the mature conidiophore, forming a structural feature which should be emphasized as distinctive (PI. 24, H, J, L). This cell reaches its greatest width at the septum which sepa- rates it from the rest of the main axis and tapers gradually downward throughout its length, terminating in a rounded, slightly swollen foot which is connected by a slender hypha with the internal mycelium through the stomatal pore. The greatest width of the basal cell is usually about 12 ju, but the length varies from the customary extremes in heavy dew (60 to 120 n) to 30 or even 20 /x in a scanty film of moisture (PI. 24, E). When fully mature the conidia are most commonly elongate, ellipsoid, elongate ovoid, or rounded cylindric in shape, are thin-walled and hyaline, and have a more or less finely granular content. The tip is broadly rounded and lacks any papilla or other modification, while the base shows an apiculus, a slight thickening and protrusion of the wall at the point of attachment to the sterigma. Wide variations in the shape of the conidia are common, examples being found of all of the types from subspherical, pyriform, or even lemon-shaped to the extremely elongate types which are shown in Plate 25, C-I/. A method has been devised by Rosenbaum {18) for expressing quanti- tatively the shapes encountered in a study of large numbers of conidia of Phytophthora. This method, which consists in classifying and plot- ting the ratios of length to width, is of value in that it gives a quantitative idea of the relative predominance of certain shapes of conidia in a species and furnishes a reliable basis for comparison with others. Unfortunately, however, this method can make no distinction between conidia which are ovate or obovate, pyriform, or obpyriform, elHpsoid, allantoid, or cylin- drical, provided their length and greatest diameter be the same. There- fore, while the ratios of length to width of 400 conidia of the PhiHppine Sclerospora of maize are presented here in tabular and diagrammatic form for comparison with other species, a clearer idea of the variations in shape is probably to be obtained from the figures in Plate 25. The size of the conidia also varies greatly. When large numbers are examined, examples are found with such widely different dimensions as to include those given for several other species. It is difficult, therefore, to give a correct impression of the size of the conidia by means of the extreme dimensions within which they vary, or even by means of the average dimensions. However, the method of grouping together large numbers of representative conidia into a series of measurement classes and plotting curves to show their frequency of occurrence has been used no Journal of Agricultural Research Vol. XIX. No. 3 successfully in describing the size of similarly variable bodies, first by Rosenbaum {i8) for Phytophthora and more recently by Gaumann {6) for Peronospora. This method seems especially valuable in the case of such variable structures as the conidia of the Peronosporaceae, since by means of it data gathered from large numbers of individuals may be so presented that the range of variation in size which is encountered, as well as the size class which predominates in the species, is at once appar- ent. Also it furnishes a most accurate method for comparing the sizes of such bodies in different species. Table I- — Measurements and ratios of length to width of 400 conidia of Sclerospora philippinensis arranged in classes. Niunber Length classes. Number Width classes. ; Number Ratio of length of conidia of conidia of conidia to width in 400. in 400. in 400. classes. M- /«. I 17 to 18. 9 I II to 12. 9 I I. 05 to I. 14 I 19 to 20 9 0 I. 15 to I. 24 2 21 to 2 2 9 8 13 to 14. 9 3 I. 25 to 1.34 I 23 to 24 9 6 1.35 to 1.44 4 25 to 26 9 41 15 to 16. 9 9 I. 45 to I. 54 10 27 to 28 9 30 I. 55 to I. 64 35 29 to 30 9 160 17 to 18. 9 53 I. 65 to I. 74 68 31 to 32 9 59 I. 75 to I. 84 75 33 to 34 9 148 19 to 20. 9 70 I. 85 to I. 94 64 35 to 36 9 59 I. 95 to 2. 04 55 37 to 38 9 41 21 to 22. 9 40 2. 05 to 2. 14 30 39 to 40 9 32 2. 15 to 2. 24 24 41 to 42 9 I 23 to 24. 9 21 2. 25 to 2. 34 21 43 to 44 9 9 2- 35 to 2.44* 7 45 to 46 9 2 2. 45 to 2. 54 2 47 to 48 9 3 . 2. 55 to 2. 64 0 49 to 50 9 0 2. 65 to 2. 74 I 51 to 52 9 3 2. 75 to 2. 84 For these reasons this method seems well adapted to depict the size of the conidia of the Sclerospora of Philippine maize. Accordingly the measurements of 400 conidia are given in tabular form and are also plotted as curves (fig. 1,2). The ratio of length to diameter in classes is given in figure 3. These show clearly that while spores are encountered with such widely differing dimensions as 18 /x long by 12 /x in diameter, and 51 by 23 fx, the size which predominates is 34 /i by 18 /x, and by far the greater number of spores encountered are from 27 to 39 /x in length by 1 7 to 2 1 /x in diameter. Although these measurements are of conidia produced on maize, they have been compared and found to agree with similar measurements of conidia from teosinte and sorghum. On com- paring like tabulations of dimensions of fresh conidia with those from material mounted in glycerin or dried, the wTiter finds constant slight differences, particularly in width. Therefore, these 400 measurements were made on four occasions at the beginning of the period of maximum conidia production (2 to 3 a. m.) from fresh material mounted in dew May 1, 1920 Philippine Downy Mildew of Maize III s vs 'A «» T, '§■3 11 2rt ■5.S as o < 0 o u i : S ^ S a }5 S S «Si iS Si ^ ^^«§5l5^!^« n Ni viaiNOD do HxoNai n m viaiNoo ao aaxawvia 112 Journal of A gricultural Research voi. xix. No. 3 or rain water, and they probably furnish an expression of the conidia dimensions and a means of comparison with other species which is as accurate as it is possible to obtain. It is interesting to note that occasional monstrous conidia were seen, resembling somewhat those described by Miyake {14) for Sderospora sacchari. Since these have the same structure and history as those of more usual size and appear to represent merely the upper extreme of the widely varving conidia dimensions, they are regarded by the writer as of no special significance. Germination of the fresh conidia takes place readily in dew, in rain water, in water from clear brooks, in dilute nutrient solutions of various kinds, and on similar solutions solidified with i per cent agar. When once the conidia are dried, however, they will no longer germinate under any conditions. On the moist surfaces of newly infected plants in the field, large numbers of conidia may be found germinating vigorously at any time from about 3 a. m. until dawn, but after the rapidly drying effect of the early sun has been felt for one or two hours there can be found on the same plants only shrivelled spores incapable of further development. Germination is preceded by a swelling and consequent alteration in size and shape of the conidium and invariably proceeds by the protrusion of one or more germ tubes (Pi. 25, C-L). This may take place from any part of the spore, and the hyphae thus produced may simply elongate (PI. 25, C, J) or develop variously into extensively branching systems (PI. 25, E, F, L). Occasionally the hyphae of germi- nation grow up into the air for a short distance and produce at their tips an ovoid swelling that might perhaps be interpreted as an abortive attempt at a secondary conidium formation such as has been found in other Peronosporaceae. In no case was the production of zoospores by the conidia observed, although repeated attempts were made to induce this method of germination. In spite of the ease with which the conidia produce germ tubes, all attempts to induce continued independent development of the mycelium in artificial media have been unsuccessful, the growth seemingly ceasing when the nutriment of the spore is exhausted. In view of the tropical habitat of this Sderospora it is of interest to note that the conidia ger- minate readily when maintained at a temperature as low as 6.5° C, even though the temperature at which they most commonly germinate is from 20 to 24°. In spite of extensive search, none of the resting or resistant bodies customarily encountered in this or other genera of the Peronosporaceae, such as chlamydospores and parthenogenetic or normal oogonia, have ever been found to be associated with the conidial stage of this fungus. Every effort has been made to find such structures. The progress of the disease has been observed in individual plants from the time of their infection by the fungus to their ultimate disintegration in many varie- May 1. 1920 Philippine Downy Mildew of Maize 113 ties representing the several types of maize and teosinte and in sorghums^ under all the various conditions of the wet, dry, and transitional periods of the year. Furthermore, infected plants have been subjected to various changes in temperature, moisture, light, aeration, and soil, in the attempt to induce the formation of such structures. So far, all efforts have been in vain, and, although facilities were not available for any such experi- ment as subjecting the infected plants to long-continued cold or total freezing such as might occur in our own com belt, still the experiments which were made seem to indicate that the formation of resting bodies by the fungus in maize occurs very rarely if at all under the conditions naturally encountered in the Philippine Islands. The possibility that the conidial stage may be restricted to maize while the production of oogonia takes place on some other host invites con- sideration. As has been mentioned above, the writer has found a Scle- rospora attacking a common field grass, Saccharum spontaneum, in this region; but whether this fungus, of which only the oogonial stage has been seen, is in any way connected with the conidia-bearing Sclerospora on maize remains to be determined. IDENTITY OF THE CAUSAL FUNGUS The important question of the identity of this Philippine Sclerospora necessitates a comparison with the other members of the genus. Since our knowledge of the Philippine form is at present confined to its conidial phase solely, no comparison is possible between it and those species of which only the oogonial Stage has been recorded, such as the remarkable Sclerospora magnusiana Sor. (20) of Equisetum from Russia, the rare Sclerospora farlowii Griff. (7) of Chloris from western North America, the recently described Sclerospora miscanthi T. Miy. {14) of Miscanthus from Japan, or even the more common Sclerospora macros por a Sacc. {8), which is widely distributed on a large number of grasses and even has been found on the tassels of maize in Italy. Likewise, the type species Sclerospora graminicola (Sacc.) Schroet., although known from all over the world on a wide range of wild and cultivated grasses and even recorded on maize in Argentina {21), can not be directly compared, because the conidial stage, even though known, is rare and is character- ized by the germination of the conidia by zoospores and by the invariable predominance of the typical oospores. A far closer relationship is shown between the Philippine form and those Oriental species which occur on maize or related gramineous hosts and are characterized by the partial or complete lack of an oogonial stage, with the concomitant predominance of the conidial phase, which is distinguished further by the germination of the conidia by tubes. Of these there have been enumerated the following: Sclerospora javanica (Rac.) Palm, of Java (originally described by Raciborski as Peronospora maydis); Sclerospora maydis (Rac.) But., of India; and 164177—20-^2 114 Journal of A gricultural Research voi. xix, No. 3 Sclerospora sacchari T. Miy., described by Miyake from Formosa but reported by Lyon (11, 12) also in the Fiji Isalnds and Queensland. It has been assumed by Baker (r) and Reinking (17) that the Phil- ippine Sclerospora of maize is identical with Sclerospora maydis (Rac.) But. of India, and this has been generally accepted by other investi- gators. Since no detailed description of the species with critical meas- urements has been published, and the single conidiophore and few spores figured by Reinking are hardly enough on which to base a decision, it seems necessary to corroborate the identification of the fungus. A comparison with Sclerospora maydis (Rac.) But. of India and also with the other related species mentioned above is accordingly in order. Such a comparison must necessarily consider the field characters of the disease, such as its effect on the plants attacked, its severity, and its fatality to the various hosts, as well as the specific peculiarities of the causal organism itself. Of these the characteristic structure and dimen- sions of the organism itself are most valuable, since the field characters show, on the one hand, a general similarity in all these fungi and, on the other hand, vary so widely under different conditions as to be con- fusing even in one species. The Sclerospora causing the Philippine disease is known in its conidial phases only, and a comparison of this form with other species must be based on this stage. Such a comparison is confronted by many diffi- culties. In the first place, the characters most valuable from the system- atic point of view have been found by the writer to vary greatly under different conditions and at different stages in the development of the Philippine form, and they probably do so in the other forms also. For instance, the very important characters of the size and shape of the conidia and the structure and dimensions of the conidiophores vary greatly at different stages of development and under different conditions. The conidia begin as small spherical outgrowths from the sterigmata tips, and in their development become larger and more elongate, passing through ellipsoid (PI. 24, A), oval (PI. 25, A), and even pyriform stages before they eventually assume the elongate ovoid ellipsoid or rounded cylindrical shape of complete maturity (PI. 24, C). They are then separated from the sterigma tip by the septum. This characteristic shape is transient, however, for, after they are free from the conidiophores, the spores show an almost immediate imbibition of moisture, which results in a marked increase in size and in a more rotund shape, due to the greater bulging of the side walls. Moreover, the apiculus which marks the point at which the spore was attached to the sterigma is modified, by the swelling of the spore, to a low, rounded curve. Since the partially developed spores of various shapes and sizes may be detached from the sterigmata and still retain their contents and germinability, and since marked changes from the shape and size of the May I, igao Philippine Do-wny Mildew of Maize 1 1 5 mature spores normally follow when it is free, it is obvious that a mount of spores usually comprises a motley collection of shapes and sizes, only a comparatively small number of which represent the characteristics of the normal and mature spore. Moreover, aside from these variations which mark the normal devel- opment of the spore, there are also changes in size and shape resulting from abnormal conditions such as the sudden checking of development by unusual drying of the necessary layer of moisture on the leaf. The size and structural characteristics of the conidiophores also vary markedly with attendant environmental conditions. The normal order is for primary, secondary, and tertiary branches to form before the sterigmata develop and begin to bud out the spores. If the gradual drying of the film of moisture on the leaf surface begins to check this process before its completion, however, sterigmata formation and spore production ensue prematurely, arid conidia may be borne on the sec- ondary or primary branches of the conidiophore (PI. 24, I), or even on the apex of the main axis itself. Similarly, the growth of the basal cell and main axis may be curtailed (PI. 24, E). Obviously, as a result of these changes, the height of the conidiophore shows a corresponding alteration. Finally, after it has lost its conidia, the conidiophore shrivels and is dried to an almost unrecognizable mummy by the morning sun. Since it appears highly probable that similar variations in size and structure occur also in the other oriental mildews, it is difficult to make any adequate comparison from the data available. To permit accurate comparison one should have descriptions and illustrations of material, or the material itself, collected under the optimum conditions, which in the case of the Philippine downy mildew occur on cool nights with heavy dew or persistent rain from 2 to 4 a. m. In the light of this fact, Miyake's {14) data are valuable, as he recog- nized that conidiophores and conidia were produced at night, and his drawings show that he illustrated excellent material. Most investi- gators, however, failed to realize this, and their material, as their descrip- tions and drawings show, was inadequate and scanty. When one compares the available data, inadequate though they be, the following points are apparent. The Philippine and Javan Sclerosporas are alike in that the conidial phase is the only one yet known. The conidiophores of the former closely resemble those of Sclerospora javanica Palm both in size and structural characteristics, such as the basal cell, the main axis, the branch system, and the ultimate sterigmata. On the contrary, the conidia of the two forms are noticeably different. In the Javan fungus they are oblong rotund in shape and measure 19 to 26 /z in length by 15 to 20 /x in diameter, while in the Philippine mildew they are elongate ellipsoid, elongate oval, or rounded cylindric and markedly longer, most of the conidia encountered measuring about 34 ^i in length 1 1 6 Journal of A gricultural Research vot. xix. No. 3 by 17 /x in diameter, and comparatively few showing the shortness which marks the Javan form. Moreover, although the field characters of the two diseases are very similar, the Javan Sclerospora presents an addi- tional point of difference in that it does not attack teosinte, although teosinte-maize hybrids are, if anything, even more susceptible to it than maize itself {19). To Sclerospora sacchari T. Miy., of Formosa, the Philippine maize mildew shows a very close resemblance in the size, the form, and even the minor structural characteristics of the conidiophores. Also, the conidia of the two forms are evidently quite similar, since the illustra- tions and the description (ellipsoid or oblong with rounded apex, 25 to 41 /x long by 15 to 23 )U in diameter) of the Formosan conidia are appli- cable to those of the Philippine species also. A marked difference be- tween the two, however, is shown in their virulence on various hosts, for, while Sclerospora sacchari grows on both maize and teosinte as does the Philippine Sclerospora, still the former attacks sugar cane of many varieties, including those grown most commonly in the Philippine Islands, with violent intensity, while the latter, so far as is known, does not infect that crop at all. In the Philippines, in regions heavily infected with the maize-mildew, sugar-cane fields comprising many varieties grown under widely varying conditions and situated adjacent to the badly infected maize, and even containing some maize plants growing among and in contact with the young cane, have been under frequent observa- tion during all stages of their development for over a year, and yet no case of infection with the downy mildew of maize has ever been seen. Moreover, inoculation experiments such as were successful with maize, teosinte, and sorghum have so far failed to cause infection of the Philip- pine Sclerospora of maize on sugar-cane varieties found susceptible to the Formosan disease. Furthermore, the oogonial stage which has been reported for Sclerospora sacchari T. Miy. forms an additional point wherein it differs from the Philippine fungus, although it should be noted that the oogonia, which have been found only once and are not figured, have not been proved to be connected with the conidial stage of Sclero- spora sacchari. On comparing the Philippine downy mildew of maize with the British Indian species {Sclerospora maydis (Rac.) But.), with which it has been regarded as identical, a close resemblance indeed is apparent. The conidia especially are similar in both shape and size in so far as one can judge from the data available; the lack of any other type of spore is another point of agreement. In considering the conidiophores of the former, however, it should be noted that the description, dimensions, and illustrations indicate that the material was imperfect, for if one may judge from the Philippine fungus, the size and the abruptly ending base of the conidiophore signify that the main axis had been broken off just above the basal cell. Any accurate comparison, therefore, is difficult. The May 1. 1920 Philippine Downy Mildew of Maize 117 sterigmata, however, are comparable, and it is clear that those of the British Indian fungus are markedly larger (15 to 20 m long) than those of the Philippine species. Moreover, the field characteristics are noticeably different. Although Butler reported the first attack of the disease at Pusa in 191 2 and empha- sized the probability of its spreading to other fields of the region, his latest report (5) indicates that it has continued to be only slightly and restrictedly destructive, an effect markedly in contrast to the rapid spread and serious damage of the Philippine fungus. Also, Butler's description emphasizes the stunting of the growth and resultant bunchy appearance of the plant as a characteristic feature of the disease in India, while in the Philippines this is but one and certainly not the most striking effect of the disease. While the matter is necessarily unsettled because of lack of adequate description of the British Indian form, certain points would seem to indicate that the maize-mildew of India is a different physiological variety and probably a different species from that of the Philippines. These points are the differences in the causal fungi and the symptoms, and especially the lack of virulence shown by the Indian disease and its failure to spread through Bengal where "maize is a crop of considerable importance" and where the conditions of climate and culture are little if at all different from those of some infested regions of the Philippines. In any case, however, it should be noted that the name Sclerospora maydis (Rac.) But. is not strictly a tenable one, for it was applied to the British Indian maize-mildew by Butler (5, p. 15) on the assumption that it was identical with the Javan. Butler {4, p. 2j 5-276) concluded from his comparison with the diagrammatic drawings and incomplete descrip- tions of Raciborski {16) that the downy mildew of maize in British India — was found to be identical with the one which causes great damage to this crop in Java, and that — its cause is a fungus named Peronospora maydis by Raciborski. The more recent and extensive work of Palm {15), however, has shown clearly that the Javan fungus, although indeed a Sclerospora, is a dis- tinct species, one which Palm names Sclerospora javanica. This leaves Sclerospora maydis (Rac.) But. as the name of the British Indian maize- mildew. Therefore, because the points of difference already considered seem to indicate that the downy mildew of maize in the Philippines is not identical with the one in British India, and because the name Sclerospora maydis (Rac.) But., given to the latter, is technically untenable, it seems nec- essary to distinguish the Philippine downy mildew of maize. Hence it is ii8 Journal of Agricultural Research voi. xix.no. 3 given the name of Sclerospora philippinensis , n. sp., with the diagnosis as follows : Sclerospora philippinensis, n. sp.' Sclerospora Maydis, Reinking, 1918, in Philippine Jour. Sci., s. A, v. 13, no. 5, fig. 39, pi. 20, fig. 1-2, not Butler. Forming linear or irregular whitish yellow to pale spots, often entirely discoloring the leaves and more or less deforming the host. Mycelial hyphae growing intercellularly In all parts except the root, branched, slender, usually about 8 ju in diameter, but irregularly constricted and inflated, haustoria simple, vesiculiform to subdigitate, small, about 8 ix long and 2 ^ in diameter. Conidiophores always produced in night dew and growing out of the stomata, erect, 150 to 400 n long, 15 to 26 /z thick, bearing a basal cell in the lower part, dichoto- mously branched two to four times above, branches robust, sterigmata conoid to subulate, 10 m long, slightly curved. Conidia elongate ellipsoid, elongate ovoid, or rovmded cylindrical, varying in size, usually 27 to 39 n long by 17 to 21 /z broad, hyaline, with thin episporium, minutely granular within, slightly rounded at the apex, provided with a minute apiculusatthe base, always germinating by a tube. Oospores not yet seen. Material of the type has been deposited in the pathologic collections of the Bureau of Plant Industry, Washington, D. C, in the Cryptogamic Herbarium at Harvard University, Cambridge, Mass., and in the herba- rium of the Bureau of Science, Manila, P. I. So far as at present known there exist in the Orient the following Sclerosporas which are of primary importance, since they cause serious diseases of maize. Sclerospora javanica Palm, known on maize and maize -teosinte hybrids in Java, Madoerah, and Sumatra. Sclerospora maydis (Rac.) But., known on maize and teosinte in Bengal, British India. Sclerospora sacchari T. Miy., kno\\'n on maize, sugar cane, and teosinte in Formosa, and on sugar cane in Queensland and the Fiji Islands. Sclerospora philippinensis, n. sp., known on maize, teosinte, and sorghum in the Philippine Islands. All these species are very similar in their effects and show close rela- tionship in structure and development. All are characterized by the ' Sclerospora philippinensis, sp. nov. Maculas lineares vel irregulares, albido-flavas vel pallidas effiiciens, saepe totum folium discolorans, et matricem plus minusve deformans. Hyphismycelicis inter cellulas in totas partes praeter radicem crescentibus, ramosis, tenuibus, plerumque circa 8 m in diametrum, sed irregulariter constrictis inflatisque, cum haustoriis simplicibus, vesiculiformibus subdigitatisve, minutis, circa 8 m longis et 2 m in diametrum omatis. Conidiophoriis semper in rore noctumo productis, e stomatibus egredientibus, erectis, 150-400 m longis, 1S-26M crassis, in parte inferiore cellulas basilares gerentibus, supeme 2-4 dichotomo-ramosis, ramis robustie cuxa sterigmatibus conoideo-subulatis, 10 11 longis, leviter curvatis. Conidiis elongato-ellipsoideis, elongato-ovoideis vel rotundato-cylindraceis, variis dimensione, plerumque 27-39 M longis et 17-21 M latis, hyalinis, episporio tenue, intus minute granulosis, apice leviter rotundatis; basi cum apiculo minute munitis, semper per tubum germinantibus. Oosporis nondum visis. Hab. in foliis, vaginis, glumis. bracteis. culmis, et infiorescentiis praecipue Zeae maydis, rarius Euch- laenae luxurianlis et Andropogonis sorghi per omnes partes in insulis Philippinis. May 1, 1920 Philippine Downy Mildew of Maize 119 predominance of the conidial stage, no oospores having been found con- nected with any save Scleras pora sacchari, with which, indeed, the rela- tionship is not very well established. Furthermore in all these species the conidiophores are large and prominent with a differentiated basal cell, stout main axis, and extensive dichotomous system of branches comprising large primary, secondary, tertiary, and even quaternary branches. The germination of the conidia also is invariably by means of hyphae. In contrast to these species the cosmopolitan Sclerospora graminicola (Sacc.) Schroet., the type on which the genus was established, is char- acterized by the predominance of the oogonial stage, the conidial phase being comparatively rare; by its smaller inconspicuous conidiophores, which lack a differentiated basal cell and give rise to few short primary or at times secondary branches only; and by the regular germination of the "conidia" by zoospores. Such marked and essential differences certainly appear to indicate that these oriental species should be separated from the type as a different genus; but, in the opinion of the writer, such a step can not be made with justice until more is known of the conidial stage of Sclerospora graminicola and of the oogonial stage of the oriental forms. Moreover, whether Sclerospora graminicola var. andropogonis-sorghi Kulk. should be included with the oriental group by virtue of its well- developed conidiophores and the germination of the conidia by hyphae, as Ito (9) suggests, also depends on further knowledge of the points just mentioned. When one considers the great variations in effect on the host and even in such essential features as the characteristics of the conidiophores and conidia, which have been found by the writer to occur in Sclerospora philippinensis under different conditions of the environment at different stages of its "development and on various hosts, one can not avoid a suspicion that these oriental forms may in reality be a single species. It is not inconceivable that such may be the case and that the variation in effect on the host, the susceptibility of different plants in different places, and the variations in structure of the causal organism may all be due to environmental conditions of the regions in which they are found. Obviously to settle these important points conclusively there is need of extensive cross-inoculation experiments and of comparative studies, using optimum material and methods which emphasize impor- tant characters quantitatively as well as qualitatively. The problems of the origin of these destructive Sclerosporas of maize and of their geographic distribution, their appearance in the Orient where maize has only been introduced since about 1496, and their absence as yet from the Western Hemisphere where maize originated, are all too involved for consideration at present. 120 Journal of Agricultural Research voi.xix. N0.3 In any case, however, the increasing attention which these dangerous downy mildews of maize have demanded by their destructive activity throughout the Orient in recent years must necessarily arouse the appre- hension of all who are concerned with the valuable com and sugar-cane interests of the United States. SUMMARY (i) For several years there has been known to occur in the Philippine Islands a destructive downy mildew of maize, which not only causes serious losses in that region but also threatens our own valuable corn crop, should it reach the United States. Prior to investigations by the writer no extensive study of this disease has been made. This paper presents certain results in regard to the distribution, severity, and char- acteristics of the disease and the nature and relationships of the causal fungus. (2) The disease occurs throughout the Philippine Islands, where it evidently has been established for some years. (3) It is extremely destructive. Under favorable conditions whole fields are destroyed, and in some districts it has even forced the natives to abandon com culture entirely. (4) Representative varieties of all types of maize are highly suscepti- ble, and teosinte, maize-teosinte hybrids, and sorghum are attacked, but with less virulence. Inoculation experiments on a number of re- lated plants, both wild and cultivated, gave negative results. (5) Symptoms of the disease may appear from the time the plants are seedlings with three or four leaves to the time the tassels and silk are developed. In general, infected plants show a yellowing of the leaves in more or less restricted striped areas, a whitish down of conidiophores, principally on the leaves, abnormalities in growth of the vegetative parts, and abortive development of the ear, resulting in partial or complete sterility. These effects of the disease are described and illustrated. (6) The causal fungus belongs to the genus Sclerospora of the Perono- sporaceae and is characterized by the predominance of its conidial stage, the lack of oospores, so far as known, and the invariable germina- tion of its conidia by hyphae. In these respects it differs from the type species Sclerospora graminicola (Sacc.) Schroet., which is distinguished by its evanescent conidial stage, its predominating oospores, and the ger- mination of its "conidia" by zoospores. The Philippine species shows close relationship to the following recently described oriental species, all of which attack maize : Sclerospora javanica Palm, of Java, Sclerospora maydis (Rac.) But., of British India, and Sclerospora sacchari T. Miy., of Formosa, Queensland, and the Fiji Islands. The Philippine Sclerospora appears to be a new species and is described as Sclerospora philippinensis , n. sp. May 1. 1920 Philippine Downy Mildew of Maize 121 (7) Maize plants usually are infected as very young seedlings, and less often as they mature. In any case, however, when the symptoms appear, the mycelium of the fungus already has invaded the host tissue exten- sively. The mycelium may be found in practically every part of the maize plant with the exception of the root, but is most abundant among the bundle sheath cells of the leaf. (8) The conidiophores are produced in vast numbers but only at night when a thin layer of dew or rain is on the leaf surface. They vary greatly in size and development according to the depth and persistence of this layer. These variations are described and figured. (9) Since the conidia also show wide variation in size and shape, an attempt is made to give a quantitative idea of this by tables and graphs of the measurements of 400 specimens. When fresh, the conidia ger- minate readily in water and various culture media at temperatures rang- ing from 6.5° to 25° C, and invariably by hyphae. Once they have become dried the conidia no longer germinate; hence their distribution and the infection of new plants occurs almost always before dawn. (10) In spite of extensive search, no oospores or other resting bodies have yet been found to be produced by this Sclerospora. It apparently maintains itself by transmission from plant to plant; The writer has found the oospore stage of a new Sclerospora on Saccharum spontaneum L., a common wild grass of the Philippines. Whether this is in any way connected with the conidial stage on maize remains to be determined. LITERATURE CITED (i) Baker, C. F. 1916. ADDITIONAL NOTES ON PHiLrppiNE PLANT DISEASES. In Philippine Agr. and Forester, v. 5, no. 3, p. 73-78. (2) Berlese, E. J. 1898-1904. SAGGIO DI UNA MONOGRAFIA DELLE PERONOSPORACEE (1898-1903). In Riv. Patol. Veg., s. 2, v. 6, p. 79-110, pi. 7-10; v. 7, p. 19-37, 1898; V. 9, p. 1-126, illus., 1901; V. 10, p. 185-298, illus., 1904. Reprinted. (3) Butler, E. J. 1907. SOME DISEASES OF CEREALS CAUSED BY SCLEROSPORA GRAMINICOLA. In Mem. Dept. Agr. India Bot. Ser., v. 2, no. i, 24 p., 5 pi. (4) 1913. THE DOWNY MILDEW OF MAIZE (sCLEROSPORA MAYDIS (rAC.) BUTL.). Ill Mem. Dept. Agr. India Bot. Ser., v. 5, no. 5, p. 275-280, pi. 8-9 (i col.). (5) 1918. FUNGI AND DISEASES IN PLANTS . . . 547 p., illus., pi. Calcutta. Bib- liography, p. 518-531. (6) Gaumann, Ernst. 1918. UBER DIS SPEZIALISATION OER PERONOSPORA AUF EINIGEN SCROPHULARIA- CEEN. /« Ann. Mycol., v. 16, no. K- P- 189-199, 6 fig. Zitierte Liter- atur, p. 199. (7) Griffiths, D. 190/. CONCERNING SOME WEST AMERICAN FUNGI, /n Bul. Torrey Bot. Club, v. 34, no. 4, p. 207-211. (Continued article.) 122 Journal of Agricultural Research voi. xix. no. 3 (8) Ippouto, G. d', and Tra verso, G. B. 1903. LA SCLEROSPORA MACROSPORA SACC. PARASSITA DELLE INFIOREScaNZE DI ZEA MAYS tiN. In Staz. Sper. Agrar. Ital., v. 36, p. 975-996, pi. 9-11, (9) Ito, Seiya. I913. KLEINE NOTIZEN UEBER PARASITISCHE PILZE JAPANS. In Bot. Mag. Tokyo, V. 27, no. 323, p. 217-223. (10) KULKARNI, G. S. I9I3. OBSERVATIONS ON THE DOWNY MILDEW (SCLEROSPORA GRAMINICOLA (sACC.) SCHROET.) OP BAjRi AND jowAR. In Mem. Dept. Agr. India Bot. Ser., v. 5, no. 5, p. 268-273, pi. 6-7 (i col.). (11) Lyon, H. L. 1911. notes ON THE SUGAR INDUSTRY OF FIJI. In Hawaiian Planters' Rec, V. 4. no. 6, p. 318-339, 6 fig. (12) 1915. THE AUSTRALIAN LEAF STRIPE DISEASE OF SUGAR CANE- In Hawaiian Planters' Rec, v. 12, no. 4, p. 257-265, 2 fig. (13) Mangin, L. 1890. SUR LA STRUCTURE DES p6ronospor6es. In Compt. Rend. Acad. Sci. [Paris], t. Ill, no. 24, p. 923-926. (14) MiYAKE, Tsutome. I9II. on a fungus disease op SUGAR CANE CAUSED BY A NEW PARASITIC FUNGUS, SCLEROSPORA SACCHARi T. MiY. In Rpt. Sugar Exp. Sta. Govt. Formosa, Div. Path. Bui. i, 61 p., 9 pi. In Japanese. (15) Palm, Bj. 1918. ONDERZOEKiNGEN OVER DE OMo LijER VAN DE MAIS. (With an English summary.) In Meded. Lab. Plantenziekten [Batavia], no. 32, 78 p., 7 pi. (16) Raciborski, M. 1897. LiJER, EINE GEFAHRLICHE MAISKRANKHEIT. In Ber. Deut. Bot. Gesell., Jahrg. 15, Heft 8, p. 475-478, 4 %• (17) Reinking, O. a. 1918. PHILIPPINE ECONOMIC -PLANT DISEASES. In Philippine Jour. Sci., s. A.i v. 13, no. 4, p. 165-216, fig. 1-20; no. 5, p. 217-274, fig. 21-42, 22 pi. (18) Rosenbaum, J. 1917. STUDIES OF THE GENUS PHYTOPHTHORA. In Jour. Agr. ReseaTch, v. 8, no. 7, p. 233-276, 13 fig., pi. 71-77. Literature cited, p. 273-276. (19) Rutgers, A. A. L. 1916. DE LijER-ziEKTE DER MAIS. In Meded. Lab. Plantenziekten, no. 22, 30 p., 7 pi. (3 col.). Cover-title : De Peronospora-Ziekte der Mais. (20) SOROKINE, N. 1889. MATERIAUX POUR LA FLORE CRYPTOGAMIQUE DE L'aSIE CENTRALE. In Rev. Mycol., ann. 11, no. 43, p. 136-152, pi. 90. (21) Spegazzini, C. 1909. MYCETES argentinensis. ser. IV. In An. Mus. Nac. Buenos Aires, s. 3, t. 12, p. 257-458. (Continued article.) Philippine Downy Mildew of Maize Plate a ijrnal of Agricultural Research Vol. XIX. No. 3 PLATE A» Young maize plant, showing the effects of a very early attack of the downy mildew on a small, early maturing variety, Manobo Yellow. Notice the dwarfing and the pale appearance of the plant as a whole, the narrowness and stiffness of the leaves, and the narrow striping of the later leaves throughout their length. The plant was 32 days old when photographed and had first shown the disease two weeks after emerging from the soil. One-fourth natural size. ' In the preparation of Plates A and B the diseased plants were photographed and enlargements were colored to correspond as closely as possible to the living specimens. Prepared by L. S. Weston. PLATE B Young maize plants, showing the effects of later attack of the downy mildew on a large, late-maturing variety, Guam White Dent. The two plants at the right are diseased; the one at the left is healthy. Notice the characteristic markings on the larger diseased plant — the whitish yellow sheath of the lowest affected leaf, the short narrow stripes at the base of the next leaf, and on the later leaves the whitening of the entire breadth at the base and the extension of broad stripes increasingly far into the normal green of each successive leaf tip. The leaves are nearly as broad and flexible as in normal plants, and their growth is little checked, if at all. The plants are 31 days old and developed the symptoms of the disease 25 days after emerging from the soil. One-seventh natural size. Philippine Downy Mildew of Maize Plate B Journal of Agricultural Research Vol. XIX, No. 3 fKniCoBiilUTiore PLATE i6' A. — Portion of a field of Moro White maize, showing heavy loss from the downy mildew. At the left, near the scale, is seen the only healthy plant that remains in this part of the field. Near it are several pale, stunted plants which are already withering, while in the background may be seen other plants less seriously attacked. B. — View near the edge of a field of Guam White Dent maize, showing the ravages of the downy mildew. 'The tall, dark-leaved plant near the scale and two others a little farther back are the only healthy plants seen. Notice the stimted, withering specimens in the foreground, and at the right the seriously affected individual with stiffly ascending, striped leaves. ' On the scale which appears in this and the following photographs each black division equals 5 cm. Photsgraphs by W. H. Weston. Philippine Downy Mildew of Maize Plate 16 Journal of Agricultural Hcsearch Vol. XIX, No. 3 Philippine Downy Mildew of IVIaize Plate 17 ^^•^ .^//^ --^e-.-j/n,^ 00 Journal of Agricultural Research Vol. XIX, No. 3 PLATE 17 A. — A frequently encountered type of downy mildew effect. The maize plant is sterile, with no ear borne in the normal place but with a couple of small, abortive ears growing at the base of the tassel. The leaf sheaths are whitish yellow, and conspic- uous stripes of the same color occupy a large proportion of the leaves. These are stiff and brittle, the young ones at the top of the plant ascending at an unnatural angle and the older ones breaking and hanging down stiffly. B. — A serious case of downy mildew injury, one of hundreds in a badly attacked maize field. The conspicuous striping of the leaves and the crooked stalk at once attract attention. The small ear is sterile. At the base of the plant can be seen another one badly dwarfed by the disease. PLATE i8 A. — A common result of downy mildew attack. In both maize plants shown the growth of the intemodes has been checked so that the leaf sheaths overlap and the unexpanded tassel is still partly surrounded by them. The striping of the leaves and their stiff, brittle character are easily seen. Both plants were entirely barren. B. — A maize plant seriously injured by the downy mildew stands in front. Its sttmted habit and striped leaves are striking evidences of the disease. Of the two abnormal ears, the one at the right was entirely sterile while the one from which the husks have been removed bore a few viable seeds. In the same hill, behind, is a healthy plant, only the lower part of which is shov\"n. C. — One hill in a maize plot which lost heavily from attacks by the downy mildew. The diseased plant at the left, although nearly as tall as the healthy companion at the right, is less strong and has a poorly developed ear which is only partly inclosed in husks and bears very few kernels. Philippine Downy IVlildewof Maize Plate 18 Journal of Agricultural Research Vol. XIX, No. 3 Philippine Downy Mildew of Maize Plate 19 Journal of Agricultural Researcin Vol. XIX, No. 3 PLATE 19 A. — A case of abnormal growth of a maize plant as the result of an attack by downy mildew. The shank h?„s elongated enormously, and an excessive development of the husks has taken place. Only a small, completely sterile ear has formed. B. — A maize plant which, when nearly mature, became infected by the downy mildew through a small sucker previously developed. The sucker is obviously in- fected, but the large plant, aside from faint leaf stripings which escape the camera, shows the effect of the disease only in its unexpanded tassel and protruding ear tip. 164177°— 20 3 PLATE 20 A. — A deformed and partly sterile ear complex produced by a maize plant as a result of downy mildew infection. Notice the branching and elongation of the shank, the abnormal development and arrangement of many of the kernels, and the inclosing of a few kernels in tunicate bracts. The husks have been removed. This specimen is from a Yellow Dent variety that normally has one large and well-developed ear. B. — A maize ear developed abnormally as a result of the downy mildew. The husks, beyond which the upper third of the ear protruded, have been partly removed. Save for two or three at the base with partly developed kernels, all the florets were sterile, green in color, and bract-like in texture. Healthy plants of this Yellow Dent variety bear large ears well covered over by husks. C. — Ear of a maize plant infected by the do\^^ly mildew. Only a few viable seeds have been formed, the remainder of the florets being poorly developed and sterile. Notice the conspicuous stripes on the leaves. Before the husks were removed the tip of the ear protruded beyond them. Normally this White Flint variety bears long, well-filled ears. Philippine Downy IVIildew of Maize Plate 20 o Journal of Agricultural Researcli Vol. XIX, No. 3 Philippine Downy Mildew of Maize Plate 21 Journal of Agricultural Research Vol. XIX, No. 3 PLATE 21 A. — A near view of the thick down of conidiophores which has been produced on the upper surface of a badly diseased maize leaf. This gives an idea of the vast num- bers of conidiophores which are borne on even a small area. The regions where they are formed most abundantly correspond in general to the stripings of the leaf. The layer of dew in which the conidiophores were produced has just dried. X 2. B. — Upper surface of a badly infected maize leaf from a maturing plant. Conidi- ophore production is in this case restricted to the yellowish white stripes like those shown in the colored plate. X iX- C. — Upper surface of the middle portion of a maize leaf from a very young plant which has only recently developed the markings of the disease. The stripes are seen to be covered with conidiophores even up to the ends. X iK- PLATE 22 A. — View of a rowof Egyptian sorghum showing tall, green, healthy plants at the left and at the right a dwarfed, yellowish white plant which is infected by the downy mildew. B. — Near view of this diseased sorghum plant. Notice the slender, stunted habit, the pallor and faint stripings of the leaves. C. — A comparative view of healthy teosinte (right), and teosinte seriously infected with the downy mildew (left). The healthy plant has many suckers and is large and vigorous with broad, flexible, dark green leaves and well-developed inflorescences. The diseased plant has no suckers, is stimted and weak, and bears slender, stiff, brittle leaves, which are pale in color and inconspicuously striped, and poorly developed tassels containing a few abortive seeds at the base. Philippine Downy Mildew of iWaize Plate 22 Journal of Agricultural Research Vol. XIX, No. 3 Philippine Downy Mildew of IVlaize Plate 23 Journal of Agricultural Researcli Vol. XIX, No. 3 PLATE 23 ' A. — Portion of the typical crooked, irregular mycelum with numerous haustoria which is found in the mesophyll of badly infected leaves, here freed from the host tissue by maceration. X 375- B. — Longitudinal section cut from the center of a maize stem 8 inches from the ground. The plant, over 5 feet in height, was just putting out its tassel and had recently shown markings of the disease on its four uppermost leaves. A strand of the mycelium can be seen running alongside the bundle between cellsof the biindle sheath which are penetrated by numerous haustoria. X 375- C. — Portion of the mycelium freed by maceration from tissue of the midrib at the base of a badly infected leaf. X 375- D. — Hypha cut in cross section as it lies between three adjacent mesophyll cells of the host. The penetration of a characteristic haustorium into one of the host cells is shown. X 850. E. — Transverse section from a badly infected portion of a maize leaf, showing the abundant mycelium running between the cells of the bundle sheath and forming in the substomatal air chamber the branches (a) that grow out through the stoma to form the conidiophores. The haustoria are seen penetrating not only the mesophyll cells but also a cell of the xylem (b) and the epidermis (c). X 375- F. — Portion of a hypha lying between adjacent mesophyll cells, one of which has formed a many-layered wall around the haustorium invading it. X 850. G. — Portion of a hypha similar to that shown in F but with the haustorium unhin- dered in its invasion of the host cell. X 850. H. — Bit of mycelium such as is shown in A but more highly magnified to show the haustoria. X 850. • The drawings were made with the aid of a camera lucida and are all from preserved material of maiz e. PLATE 24 » A. — Slender, sparingly branched conidiophore bearing comparatively few conidia. It is only partially matured, as can be seen from the small size and rotund shape of the conidia and from the incomplete development of the septum. From maize during heavy dew. X 375- B. — Tip of branch with two conidia in situ. Treated with osmic acid and stained, thus differentiating the two sterigmata as more hyaline than the branch tip. X 750. C. — Stout, much-branched, mature conidiophore bearing 38 spores. From maize during heavy dew. X 375- D. — Upper portion of a nearly mature conidiophore with one secondary branch which has failed to branch further and has terminated in a single conidium only. X375- E. — Small, stunted, sparingly branched conidiophore produced on maize during the light dew of the hot, dry season. Note the poorly formed cell and the small size and restricted development of the conidiophore as a whole in comparison with those formed in heavy dew, as shown in A and C. X 375- F. — Basal cell with two thick crosswalls; From maize. X 375- G. — An unusual basal cell with two septa and an abnormally large footlike base. X375- H, J, L. — Typical basal cells of conidiophores. X 375- I. — Upper portion of an imderdeveloped conidiophore bearing three spores on sterigmata arising directly from the top of the main axis. X 375- K. — Tip of an ultimate branch with two sterigmata bearing conidia. The right conidium is shown as if in optical section, the left in surface view. X 850. M. — Basal cell of a conidiophore from teosinte with septum formation progressing by the centripetal extension of a cellulose-pectose ring. The footlike projection at the base is abnormally large. X 375. ' The drawings were made with the aid of a camera lucida and are from fresh material, with the exception of B, G, K, and M. G, I, and M are from material on teosinte; all other figur^^ are from material on maize' Philippine Downy Mildew of Maize Plate 24 liniXcm-M K_^^ Journal of Agricultural Research Vol. XIX, No. 3 Philippine Downy Mildew of Maize Plate 25 Journal of Agricultural Research Vol. XIX, No. 3 PLATE 25 » A. — QDnidiophore from sorghum, partly matured and bearing few conidia. Com- pare with Plate 24, A. X 375. B. — Conidiophore from teosinte, nearly mature, with extensive system of branches bearing many conidia. Compare with Plate 24, C. X 375- C. — Typical conidia from sorghum. Three are germinating in dew by means of relatively simple hyphae. X 375- D. — Typical conidia from teosinte. X 375- E. — Typical conidia from teosinte which have germinated in dew on the leaf surface. X 375- F. — Conidium from teosinte germinating by an extensive branched hypha when maintained in dew at7°C. X375. G. — Conidium from teosinte germinating while still attached to its sterigma. X 500. H. — Typical conidia from maize, showing common variations in shape and size. X37S- I. — Two conidia from maize just begixming to germinate in rain water. X 375- J. — Two conidia from maize germinating in sterilized brook water maintained at 8°C. X37S- K. — Conidium from maize germinating in dew on the leaf surface. X 375- L. — Conidium from maize giving rise to extensive branching hyphae in a dilute decoction of young maize kernels. X 375. ' The drawings were made with the aid ol a camera lucida and are all from fresh material with the exception of A and G. EFFECT OF DRUGS ON MILK AND FAT PRODUCTION By FRA^fK A. Hays, Associate, and Merton G. Thomas, Assistant in Animal Husbandry, Delaware Agricultural Experiment Station The opinion that milk production and butter-fat yield can be influ- enced by the use of drugs is widespread among dairymen. Many have their own opinions on this question, and some prominent feeders have been accused 6f "drugging" test cows. Three of our advanced registry associations now prohibit the use of all drugs during the official test period. Marshall^ states that some drugs and feeds are said to increase the milk and butter-fat yield. Friedberger and Frohner^ inform us that a number of galactogogues have always been recommended together with a liberal supply of feed largely fluid in character. They mention prepa- rations of antimony, sulphur, chlorate of potash, fennel, juniper berries, caraway seed, aniseed, dill, and common salt. These writers recommend the "milk powder" used as drug No. 5 in the experiment reported below\ The value of an experimental test of different drugs lies not in the fact that it might make possible some abnormal test records in the hands of the unscrupulous but in the fact that it will furnish some information on the relation of feed components to the complex physiological processes of milk secretion. PREVIOUS WORK Henderson ^ reports the effects of using six different drugs as galacto- goges. Each drug was used on 10 cows, and the period of treatment was either two days or one week, with a control period of equal length either before or after treatment. Henderson summarizes his results as follows: 1. With, sodium bi-carbonate the cows increased the milk yield but neither the fat production nor the per cent, of fat in the milk. 2. With ginger the cows increased the per cent, of fat in the milk but decreased the milk yield and total fat production. 3. With pilocarpine hydrochlor injected into the cowshypodermically in most cases the cows increased both the per cent, of fat in the milk and total milk production. 4. With malt extract the cows in most cases appeared to increase the milk and butter fat production, but it had no effect upon the per cent, of fat in the milk. 5. Neither gentian nor powdered nux vomica had any effect either on the milk production or on the quality of the milk. 6. When grain alcohol was applied to the udder just previous to milking, no effect on the milk production or per cent, of fat in the milk was noted. * Marshall, Francis H. A. thg physiology op reproduction ... p. 566. London, 1910. ' Friedbergrr. Franz, and Frohner, Eugen. veterinary pathology. Translated by M. H. Hayes, ed. 6, V. I. p. 396-397. London, Chicago, 190S. ' Henderson, Harry Oram, a study of forced feeding and methods usbd in advanced rbgistrv feeding. In Penn. Agr. Exp. Sta. Ann. Rpt. 1913/16, p. 393-419- 1918. Journal of Agricultural Research, Vol. XIX. No. 3 Washington, D. C. May i. 1930 Ub Key No. Del.-3 (123) 124 Journal of Agricultural Research voi.xix, No. 3 McCandlish^ reports two series of trials with galactogoges. In the first series one cow was used and in the second there were three. The experimental period covered two days in each series, with a control period of two to four days follov/ing. Results as given by McCandlish may be summarized as follows: 1. On tlie whole, alcohol depressed rather than stimulated miljc and butter fat production . 2. Castor oil decreased the percentage of fat in milk, btit the changes in milk yield were not appreciable. 3. Pitutarin treatment resulted in decreased milk and butter-fat yield. 4. Administration of pilocarpine and physostigmine resulted in an increased fat yield in the first series. One of the cows in the second series showed an increased fat yield, while the other two showed a decrease in milk yield. 5. The effect of aloes was greatly reduced milk yield and a fat ;yaeld somewhat reduced, but the averages show little change. 6. A mixture of epsom salts, common salt, and nux vomica showed only slight effect on milk and fat yield. THE EXPERIMENT 2 The experiment was begun April 14, 191 9, and closed July 11, 191 9. The objects of the experiments were: 1. To determine the effect of various drugs on the butter-fat test of milking cows. 2. To study the effect on the total fat yield of producing cows. 3. To determine whether drugs have an effect on the health or on total milk production. METHOD Four cows of mature age were chosen as experimental animals. No. i was a grade Holstein, No. 2 was a pure-bred Holstein, and No. 3 and 4 were pure-bred Guernseys. The interval of experimentation with each drug was five days. A control period of five days preceded all experi- mental periods, except the first five days of the experiment. Each of the four cows received a different drug for a 5-day period. This was followed by a 5-day control period during which no drugs were given. At the end of this period the drugs were shifted so that each cow re- ceived a different drug from the one previously given. Thus the control and experimental periods alternated, and the order in which the drugs were given was so arranged that each cow received each of the eight drugs for a 5-day period. The cows experimented upon were milked twice daily, the weight of milk was recorded, and composite samples of the milk from each cow were tested for butter fat daily. Drug mixture No. i was recommended to us by a prominent dairy- man. The mixture No. 5 is one recommended by Friedberger and ' McCandlish, Andrew C. the possibility of increazing milk and butterfat production by THE ADMINISTRATION OF DRUGS. In Jour. Dairy Sci., V. I, no. 6, p. 475-486. 1918. 2 Credit is due Dr. C. C. Palmer for administering drug No. 6 hypodermically. May 1, 1920 Effect of Drugs on Milk and Fat Production 125 Frohner.^ All the drugs except No. 6 were given mixed with the grain feed twice daily. DRUGS USED 1. Food tonic consisting of 100 pounds oil meal, 5 pounds saltpeter, 5 pounds epsom salts, 5 pounds gentian, 5 pounds fenugrek, 8 pounds powdered charcoal, and 5 pounds sulphur, fed at the rate of 2 ounces daily per cow in two feeds. 2. Air-slaked lime, fed at the rate of 2 ounces daily per cow in two feeds. 3. Fowler's solution of arsenic, fed at the rate of 2 fluid ounces daily per cow in two feeds. 4. Gentian fed at the rate of 2 ounces daily per cow in two feeds. 5. Tonic mixture consisting of the following: 3 ounces black sulphid of antimony; iX ounces sulphur; 5 ounces each of fennel, caraway, and juniper berries, i pound common salt, fed at the rate of 2 ounces daily per cow in two feeds. 6. One gr. physotigmine sulphate injected hypodermically daily per cow, yi grain in two doses. 7. Sodium bicarbonate, fed at the rate of 2 ounces daily per cow in two feeds. 8. Ginger, fed at the rate of 2 ounces daily per cow in two feeds. EXPERIMENTAL RESULTS Figures i to 8 present graphically the individual milk and butter-fat yield of each cow. A solid line is used for the control period and a dotted line for the experimental period. Figure I shows the results of the tonic mixture. There was a slight in- crease in fat for the pure-bred Holstein and for one of the Guernseys, but the other cows showed no perceptible change. The milk yield was slightly increased for one Guernsey and slightly decreased for the other three cows. Figure 2 shows that air-slaked lime increased the fat yield in two cases and the milk yield in two cases. Figure 3 shows that when Fowler's solution of arsenic was used, two cows increased in fat production and three in milk production. Figure 4 indicates that powdered gentian has a tendency to increase fat yield slightly but has little effect on milk production. Figure 5 shows that the German tonic mixture did not increase either fat or milk production. Figure 6 seems to indicate that physostigmine sulphate has a depressing effect on both milk and fat yield. Figure 7 unfortunately shows the fat record for only three cows. There is no indication of any appreciable effect of sodium bicarbonate on production > FRreoBERGER. Franz, and Prohner, Eugen. op. at. 126 Journal of Agricultural Research Vol. XIX, No. 3 Figure 8 indicates that cows fed ginger begin to decline in production about the second or third day. Table I gives a summary of results, showing the average of the four cows in total milk and total butter fat and the average test as obtained by dividing the fat yield by the milk yield given in the table. Table I- — Effect of drugs on milk yield, fat test, and fat yield during 5-day period Drugs used. Food tonic No. i Air-slaked lime Fowler's solution of ar- senic , Gentian German tonic mixture . . . Physostigmine sulphate. . Sodium bicarbonate Ginger Average total milk. Control. Pounds. 92. 2 67-3 107-3 II3-5 108. S 108.7 93- I 9=-5 Treated Pounds. 93-7 81.9 109.0 108.2 io8.o 99.9 90.8 93-8 Gain or loss Pounds. -f- -fi4-6 4- I - 5 Fat test. Control. Per ct. 4-37 4. 28 Treated. Per ct. 4-57 5.00 3-86 3-98 3-73 3-6o 4-43 4.32 Gain or loss. Control. Per ct. +0. 20 + .14 + .04 Average total butter fat. Pounds. 3-998 3-509 3- 995 4-459 4. 205 4. 226 4-132 3-968 Treated. Pounds. 4.285 4- 095 4> 210 4.302 4-049 3. 706 4- 020 4-055 Gain or loss. Pounds. +0. 287 + -s86 -I- -215 - -157 - -156 — .520 — . 112 -f -087 Drugs I, 2, 3, and 8 slightly increased the milk yield, but this increase is insignificant except when air-slaked lime was fed. The increase of 21 .7 per cent for the air-slaked lime group we think is significant. Gentian and physostigmine sulphate seem to depress the milk yield, and the Ger- man tonic mixture No. 5, and sodium bicarbonate have but little efiFect. The fat test was appreciably increased by tonic No. i , by lime, and by Fowler's solution. There was significant decline in fat test shown by the groups fed the German tonic mixture and physostigmine sulphate. Average total butter-fat production was probably significantly in- creased by air-slaked lime. Food tonic No. i and Fowler's solution show increase of 0.28 and 0.21 pound, respectively, in fat for the 5-day period. A decline of 0.52 pound is shown by the physostigmine sul- phate group. The decline in other groups is not considered significant. No difficulty was encountered in getting the cows to take any of the drugs, and no effect on their physical condition was observed. SUMMARY (i) A study of individual records and average records does not indicate that drugs have a very pronounced effect on the production of the dairy cow. (2) Air-slaked lime fed in 2-ounce doses daily may possibly increase milk production and total fat yield. (3) No other drug or mixture tested proved to be of value to increase production. (4) Results do not indicate that the difference in character of milk of Holstein and Guernsey cows has any relation to their manner of reaction to drugs. May 1, 1920 Effect of Drugs on Milk and Fat Prodtiction 127 ci?u^/ya/ ^n^ii/yya^ cviv/ra>f coiv/yo.'^ /s ^ V ^7>/rkh> ]r^yy^/r/ri : X-. -Z /cy j^ ^x/. 2^^ .vS* it ^^^^it^l ^^ S. -,^ ^ -. ^'^^'^ ^^"^^ ^\Z \t - ■ \y -_■ - — ■^'■■- ...--, r ^1_U_L i J . „ . ,.. . , ■^Cy 1 I 1 r y/£'^ ^ -"^ /\ \--\ _rt? -y-p' \ ^__ ^V>- V ~ • * ^. • -s »y ^* -1 -^ ■"■ •" " •-** *• /o Fig. I. — Graph showing effect of tonic mixture No. i.on butter-fat and milk yield. // ^ COW /voy colv/ya^ coiv/yos cow/ya^ 15 ■" -5/ /T- T-i -/• •> f /f" ^ U« /■rf -«' /I ^ "^ 1 ~J» , , »« T 1 iV #« - ,»\ ^ ,J • •"~1 • • _^ / \ / ' ■> 1. K^ / p: 15 Z\ \ \ v. / hJ- \/ \/ \' JL u. 1 /^/i/r y/siD L X t :::::::::tE:----:::: ::::::::::::: t 7 ^"^ ' 7 ^00^^ — -^^ :_:E^::::::::::::: :""T": /S- /o /^^'fSG^ss/o /^^^se:^ss/o /^S'^s6•p'<^^/o /^^'^j-cp'iffs/o Fic. 2. — Graph showing effect of air-slaked lime on butter-fat and milk yield. I2J Journal of Agricultural Research voi. xix. no. 3 /J* /./ /o cvk///o./ cvi^z/yo^* a>k///Q,y cvwa^o.^ /f/yT'TT'j^ >c^^ y/xv/^ £jCy/ yC/r /iM/^ /V£c£y ^t-\t- ' —^ _j^ 'i_ s. -^ ^ >. ,' ^» > /•^7» ^s?* ^^-'^^^ T ^?-v^ ="'^i'^:tr^^ ^ 5^ ± J5^ . s ^ ss JO £5 -a? ys /wz/r y/£W 1 . \ • a y - - s / . ■> / N ** •' N y > ^ *w'N "'.. <■ _. _ _ yJU>^jre^^«3S/0 /eS'^Se^<93/0 /£J>^^6-^<:93/0 /^^'^^e^^^s/o /2^ys Fig. 3. — Graph showing effect of Fowler's solution of arsenic on butter-fat and milk yield. /:? cviiZ/yo./ cvu/y/o.^' C "0 k / /yo.^^ co\////a'^ \ ^^/rr^Tp/^^/^j^iT/:/? 1 0 * '\X ^ k* \ s %. •, ** ^. ' ^ « • ,' ^ -1 1 m4 « % .. V, J \ \ -^r^ S tgr* ' 0,^ ^5 Af///r j^/TZ^ t * ^ 0 -^ > t^ 1 s i t ** y"i 1 1 ' y\ ya > y ■' *" "* '* ^ S _ _ y^J''}fJ^eF'.. ^^ ^-'• %^ A^r ^ % ''•'i^^^s;' 5 '.? ■*«4y Af/ZA-Z^^ZZ? — J>CP \ ^ ^ •« * /o 4- ^ ^ .** s ... • * / •« -i * '• _ J u _ /£'S'^S6-p'<99/0 /£'J'-^Se/='<3S/0 /K>S^^e;^<9S/0 /^J''^^G<'S3/0 Fig. s. — Graph showing eSect of the Germau tonic mixture on butter-fat and milk yield. _ c^u/z/o./ coi\^// ' k \ * %, * \ / n % t *« 0 »!•" \ / s f^ .J ' >\ •• '».l . /if/z/r )^zzzp JO p^- £'0 /o .•\ /^^-^SeFtSS/O /eS'^J'6P'<33/0 /^S^Se/^S9/'0 /£>S'^S6-/'S3/0 Fig. 7.-Graph showing effect of sodium bicarbonate on butter-fat and milk yield. copiz/^./ cvi^^/yia^ coivaos coi^/za^ J ', k 1 Jt/yirrx^o ^yir *^i<7'/> 1 _ _ ';^ 0 -i ^9, ^\ P'- :l =4= = = --s^^^^.^- K?—^--^. — \-^^--'-'-\ J -5^z:!i ■^iiiiiiiiii =1^^ iiiii-iifiiii::: -? XE /^ lIlKl Af/i/ry/sio /o /^s^^e;=^e9/o /^s^f^^&^e^/o /^j'^se/^<3&/o /,?^-^Sf5-;^a&/Q z?/jy^ Fig. 8.— Graph showing effect of ginger on butter-fat and milk yield. ARTIFICIAL AND INSECT TRANSMISSION OF SUGAR- CANE MOSAIC By E. W. Brandes Pathologist, Office of Sugar-Plant Investigations, Bureau of Plant Industry, United States Department of Agriculture The infectious nature of sugar-cane mosaic can hardly be questioned in the light of field observations bearing out this point made in Georgia and Florida last year and in Porto Rico during the preceding two years (/) ^ Records of well-controlled inoculation experiments, how- ever, have been wholly lacking. A number of investigators, beginning with the Dutch workers in Java, have attempted to produce the disease by artificial inoculation and by the use of suspected insect carriers; but in all cases results have been negative or inconclusive. Where success has been reported the experiments were carried on under unsatisfactory conditions, and the results were repudiated by contemporaneous workers who attempted to repeat the experiments. Kamerling (j) in 1902 reports that he secured infection by inoculating healthy plants with sap from diseased plants. He says (in translation) : So far as the kind of disease is concerned, we are dealing with a disease analogous to the notorious mosaic of tobacco, that is, with an infectious disease, which, how- ever, in all probability is not caused by a parasitic organism. As is the case with tobacco mosaic, the disease has been successfully transmitted by inoculating healthy plants with juice pressed out of diseased plants. (Footnote: My inoculations with juice of diseased cane were performed in the same way as the inoculation tests of Beijerinck with juice of tobacco plants affected with mosaic.) These inoculation tests, however, throw little light on the manner of origin and of dissemination in nature. One very great difficulty in carrying out tests on the way in which the disease oviy tnates and is disseminated in nature is in securing cuttings that do not have a predis- position toward the disease. From the best possible selected Moga cuttings a certain number of check plants in my pot cultures showed stripe disease ; and I have had a similar experience with specially selected cuttings from Van Delden in Soekaboemi, which in Koeningen produced a crop practically free from disease. This vague reference to his experiments and his admission of disease in the control plants was not very convincing and was discredited by later Dutch investigators. Kobus (4), van der Stok (6), and Wilbrink and Ledeboer (7) were unable to produce the disease by using the method of Kamerling. Wilbrink and Ledeboer say (in translation) : So sudden a severe outbreak as Kobus already observed gives rise to the suspicion that we are perhaps dealing with an infectious disease, as is the case, for example, with the mosaic disease of tobacco, analogous to the stripe disease in very many respects. Dr. Kamerling states in the Annual Report of the Experiment Station of Kagok for ' Reference is made by number (italic) to " Literature cited," p. 138. Journal of Agricultural Research, . Vol. XIX, No. 3 Washington, D. C. May i, 1920 uf Key No. G-190 (i3i) 164177°— 20 4 132 Journal of Agricultural Research voi. xix. no. 3 1902 that he succeeded in inoculating healthy plants with the disease by injecting sap from diseased plants. We have repeated these inoculation experiments as far as we have been able to obtain data about them, but without success. Neither have we been able to find any other indication that the disease is contagious. They conclude with Kobus and van der Stok that the mosaic is an expression of bud variation. No reference is made to successful inocu- lation experiments in the numerous papers on mosaic in the Hawaiian Sugar Planters' Record for 1911-1919. Stevenson (5) reports hundreds of inoculations of many cane varieties by various methods during 191 7 and 1918, all with negative results. Prof. F. S. Earle, in an unpublished paper, calls attention to a method of inoculating with juice expressed under oil to prevent oxidation. Some of the plants he inoculated be- came diseased, but the experiment was inconclusive and open to the criticism that it was carried on v/ithout control plants in a field where cases of the disease were appearing naturally. Various writers have called attention to the possibility of insect carriers of the mosaic disease, but no published proof has appeared, and the statements have been based on analogy with other apparently similar diseases and on field observations. The failure of all efforts to obtain uniform or dependable results with either artificial methods of inoculation or with insects has been one of the conspicuous peculiarities in the behavior of sugar-cane mosaic. In all inoculation work in plant pathology it is necessary to secure a very high percentage of infection in inoculated plants where control plants are not absolutely protected from extraneous infection. In diseases like cane mosaic, where, for reasons which we are not in a position to discuss at present, the percent- age of infection resulting from experimental inoculation is not high, it is not only necessary that all experimental plants be apparently healthy but also that they be of known healthy parentage for at least one gen- eration back and preferably more. Further than this, the experiments should be performed under absolutely controlled conditions. The pre- vention of contamination of experimental plants with diseased material by direct or indirect contact must be absolute. Special precautions must be taken to prevent the admittance to treated plants of insects or any other animals other than the ones being experimented with. The writer became convinced, after observations and experiments with the mosaic disease dating from, the summer of 191 6, that more reliance can be placed on the results of experiments performed in some region far removed from any chance of accidental infection. It was owing to these considerations that the experiments recorded here were performed at a distance from the seat of any natural infection, because the required conditions would be practically impossible to obtain where the disease is prevalent. The first experiments were conducted in a quarantine greenhouse near Garrett Park, Md. Later experiments were made in several green- May 1, 1920 Artificial and Insect Transmission of Mosaic 133 houses at Washington. The insects used were those at hand which were known to feed on sugar cane. Provision has been made by coop- eration with the Bureau of Entomology to collect im'ormation leading to the identification of the particular insect or insects responsible for secondary infections in the infested cane regions. ]Mr. George N. Wol- cott, of the Bureau of Entomology, is at present working on that phase of the problem in Porto Rico. EXPERIMENTS AT GARRETT PARK, MD.i Seed pieces from diseased parent stock were received from time to time during 19 18 and 1919 and planted in the greenhouse, which was screened with physician's cloth so that insects could not escape. On August 10, 1918, a shipment of diseased Crystalina cane from Ensenada, P. R., was planted. Yellow Bantam sweetcorn and Sugar Drip, Early Amber, and Japanese Ribbon sorghum were planted August 13, 1918, in the same greenhouse. On September 24, 1918, a shipment of dis- eased Rayada cane from Rio Piedras, P. R., was planted. Diseased seed pieces of Morado, Yellow Caledonia, Crystalina, and Rayada varieties from Arecibo, P. R., were planted on April 24, 1919. Similar pieces of Selangore, D.-117, and Rayada from Mayaguez, P. R., were planted on April 25, 1919. Lastly a shipment from Yauco, P. R., con- taining diseased seed pieces of G. C.-701, G. C.-1486, B.-3922, B.-6450, and P. R.-260 were planted May i, 1919. Through the kindness of Dr. Erwin E. Smith, cuttings of Lahaina cane were secured from plants which had been growing in one of his green- houses at Washington for more than six years and showed absolutely no signs of mosaic. These cuttings were planted in pots in a third greenhouse at Washington on December 10, 191 8. All the cane, diseased and healthy, sprouted and grew well. All cuttings from diseased par- ents produced mottled sprouts, without exception, and all cuttings from Dr. Smith's healthy cane produced in great contrast healthy plants with leaves of uniform dark green color. Experiment i. — This was a preliminary experiment to determine whether infection could take place by natural means, merely by exposing healthy plants in the same greenhouse with diseased plants. On May 10, 1 91 9, 5 healthy cane plants, 5 months old, in pots were taken from the greenhouse in Washington and placed in the quarantine greenhouse at Garrett Park, Md., in such a way that the leaves did not come in contact with the leaves of diseased plants. At that time the corn aphis (Aphis maidisY was abundant on the sorghum. The wild grasses, a few clumps of which came up as weeds in the greenhouse, were infested with red spiders {Tetranychus hinaculatus) . Both these insects were seen occasion- ally in the cane. A small leafhopper was also seen but was not captured ' Thanks are due Dr. Caroline Rumbold, who was in charge of this work during the writer's absence on trips to the Tropics. '' Identified by Dr. A. C. Raker, of the Bureau of EntomoIo£y, United States Department of As,rictilture. 134 Journal of Agricultural Research voi.xix, no. 3 and consequently was not determined. On June 3, 1919, all five of the Lahaina cane plants from Dr. Smith's greenhouse showed unmistakable incipient signs of mosaic. Two weeks later all were well-developed cases. Experiment 2. — On July 3, 191 9, 15 healthy cane plants of the Lahaina variety, 7 months old, were removed from the greenhouse in Washington to the "pesthouse" at Garrett Park. Five were placed within the house unprotected as before, and 5 were placed in each of two insect-proof cages. On July 22,4 of the exposed plants showed incipient signs of mosaic. On August 2 the remaining plant showed evidence of being infected, and a week later all the exposed plants exhibited well- advanced leaf symptoms. At this time the 10 control plants in cages were perfectly normal and continued so until they were used in another experiment two months later. Experiment 3. — Seeds of sweetcorn (Yellow Bantam variety) and sorghum (Sugar Drip, Early Amber, and Japanese Ribbon) were planted on August 13, 1 918, in the Garrett Park quarantine greenhouse. They germinated and grew slowly during the winter, then more rapidly in the spring. A number of volunteer grasses that came up as weeds in the greenhouse were allowed to mature. All these plants soon became heavily infested with corn aphis. Sorghum seed from the same lot was planted in a greenhouse at Washington. On May 7, 1919, a few mottled leaves appeared on the sorghum plants at Garrett Park. Examination of the wild grasses revealed the typical streaking and mottling in practically every stool of crabgrass {Syntherisma sanguinalis) , foxtail (Chaeiochloa lutescens) and Panicum dichotomiflorum. Other wild grasses in the greenhouse were normal. At this time the sorghum control plants in the Washington greenhouse and the wild grasses of the same species outside the greenhouse at Garrett Park showed no signs of mosaic, nor did they show any evidence of mosaic during the remainder of the summer. Experiment 4. — On August 7, 1919, about 50 adult individuals of the sharp-headed grain leafhopper {Draeculacephala molipesy col- lected two days previously on mosaic-diseased sugar cane at Audubon Park, New Orleans, La., were placed in a cage with 5 healthy cane plants at the Garrett Park greenhouse. The leafhoppers immediately began feeding on the healthy cane. No infection was evident after two months. EXPERIMENTS AT WASHINGTON During September, 191 9, nearly all experiments were transferred to greenhouses especially prepared to receive them at Washington. Venti- lators of the 2-story greenhouse, formerly used by Dr. Smith for bananas, were screened with physician's cloth ; and the diseased cane plants of all varieties were removed to it from Garrett Park. A greenhouse in another range, separated by a roadway from the first, was screened; and ' Identified by Mr. T. E. Holloway, Bureau of Entomology, United States Department of Agriculture. Mayi, I920 Artificial and Insect Transmission of Mosaic 135 300 healthy Lahaina cane plants, from cuttings supplied by Dr. Smith, were placed therein. These plants were from the same source as the ones previously mentioned. The second greenhouse was divided into halves by a tight glass partition. One compartment was used for propagating healthy stock, and the other compartment was used for artificial inocu- lation experiments. Both compartments were kept free from insects by frequent fumigation. In the banana house, or "pesthouse" fumiga- tion was not practiced on account of cage experiments with insects. The greatest precautions were taken to prevent accidental infection of plants in the house where healthy stock was growing. This house was invariably the first one visited by the gardener for routine work such as watering, and both houses were kept padlocked at all times. Probably because of this care no single case of mosaic has appeared there or on control plants in either house in any of the experiments. INOCUL.-^TlOiNS WITH INSECTS Experiment i. — On October 8, 1919, 10 individuals of Aphis maidis were transferred with a camel's-hair brush from mosaic sorghum to each of four young healthy cane plants in separate cages. A fifth cage was reserved for two healthy plants as controls. On October 28 all four plants showed incipient signs of mosaic. On November 18 they were all unmistakable, well-advanced cases. The two control plants remained healthy. Experiment 2. — On February 2, 1920, 12 to 15 individuals of Aphis maidis were lifted from mosaic sorghum and placed on each of three healthy cane plants in separate cages. Two healthy cane plants were placed in a fourth cage for controls. On February 28 two of the treated plants showed signs of mosaic and on March 5 were typical cases. The two control plants remained healthy. Experiment 3. — -On February 2, 1920, one mosaic sorghum plant in- fested with Aphis maidis was placed in a cage with a healthy cane plant in such a way that the leaves of the two plants intermingled. On March 21 the cane plant showed unmistakable signs of infection. Experiment 4. — February 2, 1920, 10 individuals of Aphis maidis were lifted from a diseased cane plant of variety G. C.-701 and placed on a healthy cane plant in a cage. No infection was apparent on March 15. Experiment 5. — On October 8, 191 9, 15 specimens of Draectdacephala molipes were placed in each of five cages containing one healthy and one mosaic cane plants. On January 5, 1920, approximately three months later, there was no evidence of infection. Experiment 6. — On January 5, 1920, 15 specimens of Draectdacephala molipes were placed in each of two cages containing two mosaic sorghum plants and two healthy cane plants. On March 1 1 there was no sign of infection on any of the cane plants. 136 Journal of Agricultural Research Vol. XIX, No. 3 Experiment 7. — On November 20, 1919, two mosaic cane plants of the Rayada variety, infested with the sugar-cane mealy bug {Pseudococcus honinensis (Kuw.),^ vvere placed in each of two cages, together with two healthy cane plants of the Lahaina variety. A few of the mealy bugs were transferred from diseased plants to all healthy plants. Ants were assiduously tending the mealy bugs. On March 11, 1920, all healthy plants were badly infested with mealy bugs but there was no mosaic infection. ARTIFICIAI, INOCULATIONS Virus was obtained for artificial inoculation by two methods. Cell sap from young leaves, designated as virus No. i, was obtained by grinding the young, tightly rolled leaves of diseased Rayada cane in a food chopper and straining through several thicknesses of cheesecloth. It was used undiluted for inoculating immediately after being prepared. Virus No. 2 consisted of cane juice from the youngest joints, including the growing point. To prevent oxidation this was pressed out under a mineral oil (Nujol) in a specially designed press (2). This also was used undiluted as soon as it was prepared. Inoculations were made in the compartment of the fumigated greenhouse separated from all diseased material and protected by every means from accidental infection. The results of these inoculations are given in Tables I and II. In addition to the control plants injured with a sterile needle, there were about 100 other healthy plants of the Lahaina variety in the com- partment. No case of mosaic developed among these plants. Table I. — Effect of artificial inoculation of Lahaina cane with triturated young leaves {virus No. l) [Plants inoculated Jan. 8, 1920] Number of plants. Treatment. Results. 10 Virus rubbed on unbroken surface of young leaves with fingers. All healthy Mar. 21. 10 Youngest leaves inoculated by numerous needle pricks. Control plants pricked with sterile needle One mosaic Mar. 21. 5 All healthy Mar. 21. 10 Epidermal layer of young leaf cells scarified with sharp needle dipped in virus. Do. 5 Control plants scarified with sterile needle Do. 10 Yotuig leaves scarified as above and virus rubbed in vigorously with the fingers. Do. 10 Inoculated by injecting K cc. of virus into grow- Two mosaic Feb. 14; eight ing point with hypodermic syringe. healthy Mar. 21. 5 Control plants punctured at growing point with sterile needle. All healthy Mar. 2 1 . ' Identified by Mr. Harold Morrison, of the Bureau of Entomology, United States Department of Agri- culture. May 1, 1920 Artificial and Insect Transmission of Mosaic 137 Table II. — Effect of artificial inoculation of Lahaina cane with juice from cane unoxi' dized {virus No. 2) [Plants inoculated Jan. 7, 1920] Number of plants. Treatment. Results. 10 Virus rubbed on unbroken surface of young leaves with fingers. All healthy Mar. 21. 10 Youngest leaves inoculated by numerous needle pricks. Do. 5 10 Control plants pricked with sterile needle Do. Epidermal layer of young leaf cells scarified with Do. sharp needle dipped in virus. 5 Control plants scarified with sterile needle Do. 10 Young leaves scarified as above and virus rubbed in vigorously with lingers. Do. 10 Inoculated by injecting K cc. of virus into grow- ing point with hypodermic syringe. Eight mosaic Feb. 6 to 14. 5 Control plants punctured at growing point with sterile needle. All healthy Mar. 21, r DISCUSSION From the foregoing results it may be inferred that the sugar-cane mosaic virus is highly infectious only when exacting demands in the matter of favorable conditions are satisfied. Erratic spreading under natural conditions in the field also indicates the necessity for special con- ditions, which are not as yet known. It is considered as proved, however, that the cell sap of diseased plants is infectious when introduced in the proper manner and that the disease can be transmitted by insects. Just what insects are responsible for dissemination in the cane regions remains to be proved. The failure of the sharp-headed grain leaf hopper to transmit the disease under the conditions of these experiments is surprising. This insect is very common on cane in Louisiana, and as a result of field observations suspicion was directed toward it from the first. Other leaf hoppers are now being tested. The successful experi- ments with the corn aphis is of great interest scientifically, but it is not believed that transmission of mosaic is restricted to this insect or to other aphids more abundant on cane. Aphis maidis, however, has been re- ported on sugar cane from practically every sugar-cane region in the world. That cane mosaic is analogous with other mosaic diseases is brought out by a number of facts, aside from the visible signs of the disease. As in many other mosaics, the infectious material does not seem to be highly specialized, but may attack other plants of the same family. The cell sap of infected plants contains some organism, not visible by ordinary means, which is capable of inducing the disease when injected into healthy plants. Leaves which are mature at the time of inoculation never show any signs of mosaic. This fact, typical of all mosaics, has been brought \ X 138 Journal of Agricultural Research voi. xix.No.3 out in all inoculation experiments with sugar cane. The disease can be transmitted by certain sucking insects. There is no known period of saprogenesis in the existence of the virus. Seed transmission of the virus is one of the phenomena concerning which divergent results have been recorded for the various mosaic diseases. This point has not been definitely settled for sugar-cane mosaic, but mosaic sorghum plants failed to produce mosaic progeny in two experiments. LITERATURE CITED (i) Brandes, E. W. 1919. THE MOSAIC DISEASE OF SUGAR CANE AND OTHER GRASSES. In U. S. Dept. Agr. Bui. 829, 26 p., 5 fig., I col. pi. (2) Clark, W. Blair. 1917. A SAMPLING PRESS. In JottT. Indus. and Engin. Chem., v. 9, no. 8, p. 788. (3) Kamerling, Z. 1903. DE GELE strEpEnziekte der bladerEn. In Verslag 1902 Proefsta. Suikerriet West Java "Kakok" to Pekalongan, p. 76-81. (4) KoBus, J. D. 1908. vergelijkende proeven omtrent gele-strepenziekte. In Arch. Java-Suikerindus., jaarg. 16, p. 350-374. Reprinted as Meded. Pro- efsta. Javasuikerindus., no. 12, p. 319-342. (5) Stevenson, John A. 1919. the mottling or yellow stripe disease op sugar-cane. In Jour. Dept. Agr. and Labor Porto Rico, v. 4 (i, e. 3), no. 3, p. 3-76, illns., 3 col. pi. Literature cited, p. 73-76. (6) Stok, J. E. VAN DER. 1907. VERSCHIJNSELEN VAN TUSSCHENRASVARIABILITElT BIJ HET SUIKERMET. (prOEVE EENER VERKLARING DER GELE-STREPENZIEKTE EN DER SERE- HZiEKTE.) In Arch. Java Suikerindus., jaarg. 15, p. 581-601. Re- printed as Meded. Proefsta. Oost-Java, s. 4, no. 36, p. 457-477. (7) WiLBRiNK, G., and LedeboER, F. 1910. bijdrage tot de kennis DER GELE STREPENZIEKTE. In Arch. Suiker- indus. Nederl. Indie, jaarg. 18, p. 465-518. Reprinted as Meded. Proefsta. Java-Suikerindus., deel 2, no. 39, p. 443-495, 5 pi. (4 col.). Vol. XIX TvIAY 15, 1920 No. ^ JOURNAL OF AGRICULTURAIv RESEARCH CONTKNTS Page Halo-Blight of Oats 139 CHARLOTTE ELLIOTT ( Contribution from Bureau of Plant Industry ) Influence of Fennentation on the Starch Content of Ex- perimental Silage - - - - - - - -173 ARTHUR W. DOX and LESTER YODER (Contribution from Iowa Agricultural Experiment Station) Effect of Premature Freezing on Composition of Wheat - 181 M. J. BLISH ( Contribotion from Montana Agricultural Experiment Station) PUBUSHED BY AUTHORITY OF THE SECRETARY OF AGRICULTURE, WITH THE COOPERATION OF THE ASSOCUTION OF LAND-GRANT COUEGES lATASHINGTON, D. C. WMHINOTON : OOVBIINMtMT PMINTINQ OFPIOB ; IHO EDITORIAL COMMITTEE OF THE UNITED STATES DEPARTMENT OF AGRICULTURE AND THE ASSOCIATION OF LAND-GRANT COLLEGES FOR THE DEPARTMENT FOR THE ASSOCIATION KARL F. KELLERMAN, Chairman Physiologist and Associate Chief, Bureau of Plant Industry EDWIN W. ALIyEN Chief Office of Experiment Stations CHARLES L. MARLATT Entomologist and Assistant Chief, Bureau ofEnUymology J. G. LIPMAN Dean, State College of Agriculture, and Director, New Jersey Agricultural Experi- ment Station, Rutgers College W. A. RILEY Entomologist and Chief, Division of EnfO' mology and Economic Zoology, Agricul- tural Experiment Station of the University of Minnesota R. L. WATTS Dean, School of Agriculture, attd Directori Agricultural Experiment Station, The Pennsylvania State College All correspondence regarding articles from the Department of Agriculture should be addressed to Karl F. Kellerman, Journal of Agricultural Research, Washington, D. C. All correspondence regarding articles from State Experiment Stations should be addressed to J. G. Lipman, New Jersey Agricultural Experiment Station, New Bnmswick, N. J. LIBRARY NEW YOt, 6t pq 1 z a 246 Journal of Agricultural Research voi. xix, no. 6 Attention is called to the fact that, although both greensands showed a high lime requirement when tested by the Veitch method, neither of them gave evidence of more than a trace of acidity in the water extract. The addition of sulphur to the greensand in the proportion of 3 parts green- sand to I part sulphur caused a gradual accumulation of water-soluble acidity, because of the slow oxidation of the sulphur. Composts 3 and 10, in which both sulphur and manure were mixed with the greensand, show a slight and gradual accumulation of water-soluble acidity up to the end of the fifth week, after which there is a very rapid rise for three weeks. For the remainder of the period the acidity fluctuates at a high and practically constant level. When one-half of the manure is replaced by an equal quantity of soil, as in composts 4 and 1 1 , the acidity is greatly reduced, the maximum for the Maryland greensand being reached at the end of the 12-week period and for the New Jersey greensand after 15 weeks. When the manure was entirely replaced by soil, the acidity increased gradually throughout the entire period, as shown by composts 5 and 1 2 ; but the amount developed was only about one-third as much as when equal weights of soil and manure were used. This indicates rather strongly that in composts made up with a greensand deficient in calcium carbonate the rate of development and the amount of acidity depend very largely on the amount of organic matter present. A further comparison of composts 5 and 12 with 2 and 9 seems to sub- stantiate this conclusion, in that the soil used contained a small amount of organic matter. The acidity titrated did not, of course, at any time consist entirely of free sulphuric acid. As sulphofication progressed and the amounts of free sulphuric acid and sulphates increased, an increasing amount of acid silicates was obtained in the water extract and was precipitated upon titration with the alkali. Careful inspection of several titrations, made after the maximum acidity had been attained, seemed to indicate that from 45 to 55 per cent of the acidity titrated was due to free sulphuric acid, the remainder of the acidity being due to acid silicates and other acid salts. Under the conditions of our experiment the addition of ferrous sulphate and aluminum sulphate when used at the rate recommended by McLean (ii) for sulphur-floats composts has had no appreciable effect, as may be seen by a comparison of composts 6 and 13 with 5 and 12. The addition of ID gm. of calcium carbonate to the sulphur-manure-soil compost had a marked stimulating effect, beginning about the third week in the New Jersey greensand compost and two weeks later in the Maryland greensand compost. In the former the stimulating action persisted up to the end of the experiment, while in the latter the effect of the calcium carbonate had entirely disappeared at the end of 1 2 weeks. A cause for this difference is found when the lime requirement of the New Jersey greensand is June 15, 1920 Effect of Manure-Sulphur Composts on Greensand 247 compared with that of the Maryland greensand. As was previously mentioned, the lime requirement of the New Jersey material is 4,200 pounds of calcium carbonate, while for the Maryland greensand the requirement is only 3,400 pounds. The results recorded in Table III would appear to justify the conclusion that an initial acidity correspond- ing to a lime requirement of 3,400 pounds of calcium carbonate exerts a slightly depressing effect upon sulphofication, and that an acidity corresponding to a lime requirement of 4,200 pounds of calcium car- bonate is less favorable. Ames and Boltz (/) found that calcium carbonate when added in excess of the lime requirements exercised a de- pressing effect upon the oxidation of sulphur in their soil-sulphur compost. When they reduced the application to half, the oxidation of sulphur increased but was less than when no carbonates were added. SOLUBLE SULPHATES A comparison of the results recorded in Table IV with those given in Table III shows that the accumulation of water-soluble sulphates parallels very closely the development of acidity. It will be observed that the sulphur trioxid determinations fluctuate somewhat after having attained a maximum at the end of about 12 weeks. These fluctuations are probably due to variations in the mois- ture content and the temperature of the composts, since such variations are known to have an effect upon colloidal silicates, which in turn might exercise, through adsorption, an appreciable effect upon the soluble sul- phur trioxid obtained in the water extraction. A calculation shows that at the end of our 23-week period, approximately 15 per cent of the total sulphur used in composts 3 and 10 had been oxidized, while for the composts in which one-half of the manure had been replaced by soil about 1 1 per cent of the total sulphur had been oxidized. These figures show that the amount of sulphur used was in excess of the amount necessary to secure the most economical results. SOLUBLE POTASSIUM The amount of water-soluble potassium in each compost at stated intervals is given in Table V. A comparison of these figures with those given in Tables III and IV brings out the fact that with the increase in acidity and the accumulation of sulphur trioxid there is a corresponding increase in the amount of potassium in the water extract. The potassium, however, continues to increase for some weeks after the acidity and sulphur trioxid have reached a maximum. It seems necessary for a certain degree of acidity to be developed before any appreciable amount of potassium is made water soluble, the larger amounts of acidity and soluble sulphate break- ing down the greensand more rapidly. 248 , Journal of Agricultural Research Vol. XIX, No. (■ s ^ •^^ m ^s,?? 00 l^ s -.to 1^ N M VO t 0 S"? ^ 1 M w CM tn w c5 V) N 6 0 en \r, N 0 W M 00 M 0 O* r- 0^ too Wl f^ a M i> 0 fO 1 M r^ M 10 -o 6 0 w. ro «) ■* Tl" t^ -^ t^ r* 0 ^ ^Jat NO a r* ■^ w 0\ M un 10 Ht 00 CI r^ M V H 0 c5 fTi d -o- OS N r^ PO06 J^ R* ti > ■I 0 '^ -0 00 9 ,0 m -- 1 3 . r«^ 0 0 w 00 -^ 00 vO Ht 0 0 OlOO " 4j 1 r- CA ao) 0 10 "T >o 0 ■0 M >" 1 i o> tn M M a> t- 0 0 o\ 0 CO •>l->0 ■^ CO ip ^ •* Sr ^ On-O Ov 10 0 0 ^ moo 0 00 CN Ol 0 Ml) ^ ,:. (5 ^ 06 d t^ rj 0 Ol c'l w di CO c< ■* H 0 " CO IN r; a »0 « 00 Ol « 00 0 IN 00 r^ p-t 0 w 1 w 1006 A w ONOO ^ ^ t HivO 00 N IN 01 •a 1 tn ^.25 5 S,S, Ov o\ 0 w^oo r^ -^ r^oo 0 •* Ol ;3 M -i-vo^ « d ►i CO M 06 1 " ° M m " " « N « JO 05 CO a 0 M M 0 0 >000 CO CO CO IN 01 % ■3 m-^ un/-,0 t 0 1- >0 CO CO Ol 10 »0 M 1 " jj jj^g 06 ■* 0 -t%% S"^ s- M M *0 l-l Q0 ^ 00 0 ,^ r^ w 0 t^ "■ 2 ^ 3 1 m 0 '-' ■>t 01 0.00 lOiO 00 3 M « f^ ^ 't c> CO M di ■^o 0 ^ ^ 1 M OVVO N 00 o HI HI 00 in ^O trt o» **^ ■*o vO OlOO "t IN >n r- 0 ti ^ 7^ •* S 0 10 0 >o 10 00 -0 1/1 10 ■* 00 i CO cJ 1 M 06 06 f^ d '^ A H t^»d i »o6 >o VO ■* m ^ '* M lO to 00 CT, r^ >il 00 0 ^ Tj-00 0 0 w 0 00 C4 ^^ 000 0 r^ Ovu, VO 00 IN CO 01 "1 6 M •«■ t~- ^ C4 t^ ■ IN 4 « 4 CO " a •* 10 IN 10 „-a ro r~ 0 N 01 M 10 w r~.vo Ol M 01 IN t> ir^ r^ t>. ■* •* « 01 1^ M CO 00 0 01 g 6 w « t^ M C4 CO • »iO j M C4 id ^ « " . 00 l-*oo r» r^ r» 0 « 00 ^ CO r- 0 M ^t vo 0 -0 "1 0 0 *t OlOO »o d M ^N M M CO ■ ■ ■^ N 4 & i 8 si § 1 0 0 6 6 ^_. 6 'la 0 0 0 ii 0 ° 0 ^ «l la 8 1 do « 6 6 5 oi^ ii •ss 1 ii ■r-rO .113 ii 0 i 0 1 ii E§8j^8| 0 "" "" 3 g 1^ <« "5 888 ° ITu ^ S 1- 888 add "3 3 0 u 3 3 3 3 N a 3 0 w 3 u a Sail 0 a'a'a -J3J3 u w J3 Z «jXiJ=J3 -.5.5 u V.-S is S g-tfa ^ Q. 3 d CI A A 3 a aS S a3 §3-3 625,3 a ^mmm WM tn ^m^fi WW CO i S 0 W M f*) ■«f i/IO r~ 00 Ol 0 « « CO Tt "0 "0 ui & 1 « b m 1 S 5 i I junei3. I920 Effect of Manure-Sulphur Composts on Greensand 249 '^ 2-a ^ 0000 w O O r^sO O CO O ^ ^ in *-, H Ht d ^ "N 6 rh rh ■ N -O O ^00 or- Oi 0\ fO O i^ <*) * 00 t-00 r^-O < 00 O \0 r- o o »i-( rr ^ r^ r--CO QC t-^ 00 00 »O00 o M 0\ O NO M d ' 0(3 -^ M iH ^ a» o\ < M O GO I d ' d 1 N i^ foo 00 f* rr -^ m rt 6 ' t^ ri^ OvvO IN, \n ' 00 -^ o o o -^ o o 3 3 a a ea K-s-S-S 00 .3 S S'2 r^ si B 8 8 ° ° a 2 0.0 °8°d W I- U 1^ tH U 3 3 3 6 •OS 33^ 35 I 07 152- 77 •34 3-36 7 Alkaline loi. 35 3 50 535- 55 1.78 33-74 8 •OS •OS 42 2. 14 . 16 .48 9 •OS S- IS 98 30. 66 .26 1.08 10 ■05 i34^ 15 4 84 672. 47 3-09 37- II Maryland greensand II ■OS 116.85 2 53 568. 02 2.54 28. 16 12 ■OS 41. 00 87 178. 12 •30 I. 21 13 ■05 36. 20 80 158.08 .28 .82 I 14 Alkaline 112. 15 4 21 577- 65 2.30 26.53 Attention is called to the fact that the potassium liberated from the New Jersey greensand is much greater than that recovered from the Maryland greensand. This is to be expected, since the former had an initial potassium content of 5.88 per cent, while the latter contained only 1.42 per cent of potassium, as shown in Table I. It will be seen that the largest amount of potassium was extracted from compost 3, containing the New Jersey greensand, and the second largest amount from compost 10, which is the corresponding mixture made with Mary- land greensand. The fact that both of these composts have twice the amount of manure contained in No. 4, 7, 11, and 14 would indicate that 252 Journal of Agricultural Research Vol. XIX, No. 6 comparatively large amounts of organic matter favor sulphofication and the liberation of potassium under the conditions of this experiment. These results are not in accord with those reported by McLean (//), who, working with sulphur-floats-soil composts, came to the conclusion that a compost is more efficient in the producing of available phosphorus in the absence of large amounts of organic material. In Table VII the total potassium present in each compost, the water- soluble potassium at the start, and the maximum water-soluble potassium present at any one time during the period of 23 weeks are computed on the basis of the initial weights of the composts. Table VII. — Total potassium made water-soluble (dry basis) Com- post No. 9 10 13 14 Material added to i.soo gm. greensand. None Sulphtir 500 gm Sulphtir 500 gm. ; manure 500 gm Sulphur 500 gm.; mantu-e 250 gm.; soil 250 gm Sulphur 500 gm. ; soil 500 gm Sulphiu- 500 gm.; soil 500 gm.; 0.02 per cent AU( 804)3 o-'^ HjO; 0.02 percent FeS04 o'.^ H2O Sulphur 500 gm.; soil 250 gm.; manure 250 gm.; CaCOg 10 gm None Sulphur 500 gm Sulphiu- 500 gm.; manure 500 gm Sulphtu" 500 gm.; manure 250 gm.; soil 250 gm Sulphur 500 gm. ; soil 500 gm Sulphiu- 500 gm.; soil 500 gm.; 0.02 per cent AU(S04)3 0.18 HgO; 0.02 per cent FeS04o"7 H2O Sulphur 500 gm.; soil 250 gm.; manure 250 gm.; CaCOa 10 gm Total number grams potas- sium in compost. 83-38 83-38 85.68 86.58 87.48 87.48 86.58 20. 97 20. 97 23.27 24. 17 25.07 25.07 24. 17 Water- soluble potas- sium at start (percent- age of total). 0.037 •055 .832 .408 .094 ■494 . 112 .243 3.22 2.57 •29s •275 2-33 Maximum water- soluble potassium present. O. 070 •27s 15.28 7.87 .812 .812 9.46 . 070 .251 9. 62 6.88 .389 .310 6.49 Percent- age of total. C. 084 17-83 9. 10 .928 .928 10.93 •333 I. 20 41-34 28.50 I- 55 1.24 26.85 Reference to the last two columns of Table VII will show that, while the actual amount of soluble potassium which formed in the composts containing the Maryland greensand was much smaller than that which formed in the composts containing the New Jersey greensand, the per- centage of the total potassium made water-soluble in the former was much greater than in the latter. One of the causes for this difference is to be found in Table II, which shows the mechanical analyses of the two greensands. The individual particles are much smaller in the Maryland than in the New Jersey greensand, thus exposing a much greater surface to the solvent action of the acids. Also, the glauconite particles of the juneis, I920 Effect of Manure-Sulph2ir Composts on Greensand 253 former were softer than those of the latter and seemed to be more soluble, as is shown by composts i and 8 in Table V. These figures show that although the New Jersey greensand contains more than four times as much potassium as the Maryland greensand, the amount of water-soluble potassium is the same. In considering Table VII it is pertinent to ask to what extent the manure has contributed to the total amount of potassium recovered in the water extract. To answer this question Table VIII has been pre- pared upon the assumption that all the potassium in the manure was made soluble and was recovered in the water extract. Table VIII. — Relation of potassium content of the manure to the water-soluble potas- siutn obtained Compost No. Total soluble potassium obtained from compost. Total potassium in m^anure. Maximum amount of potassitmi from manure. » 3 4 7 10 II 14 Gm. 15.28 7.87 9.46 9. 62 6.88 6.49 Gm. 2.30 1-15 I- 15 2.30 I- 15 I- 15 Per cent. 15-05 14. 62 12. 16 23.91 16. 72 17.72 a The percentages in this colvmin are based on the assimiption that all the potassium in the iflHp Tvater-«;nhihli» made water-soluble manure waso From the last column of Table VIII it will be seen that even on this basis it is possible in only one case to account for more than 17 per cent of the potassium as coming from the manure. It is evident, therefore, that from 80 to 90 per cent of the potassium found in the water extract must have come from the greensand or from the soil and greensand. Referring again to the manure composts in Table VII, it will be seen that the total amount of potassium recovered by water extracts from these composts varies from 9.1 per cent to as much as 41.3 per cent of the total initial amount present. It is important to consider the relation between the oxidation of sul- phur and the liberation of potassium. This relation is a converging ratio, which was rather wide during the period of greatest oxidation of sulphur and diminished rapidly as the potassium was released. While it was not expected that this ratio would be resolved to a constant figure for all of the composts, because of the different materials used, in each series the composts containing manure do show a rather uniform relation be- tween these processes. On the basis of the initial weights of the composts. Table IX shows the maximum number of grams of sulphur oxi- dized and of water-soluble potassium obtained, and their ratio, as deter- mined from the water extracts. 254 Journal of Agricultural Research Vol. XEX, No. 6 Table IX. — Relation between number of grams of sulphur oxidized and num,ber of grams of potassium m,ade water-soluble Ratio of .grams sul- Compost No. Sulphur oxidized. Potassium made water-soluble. phur oxidized to grams water- soluble potassium. Gm. Gm 3 77-43 15.28 5-07 4 46.65 7.87 5-92 7 52-79 9.46 5-58 lO 70- 15 9. 62 7-29 II 55-51 6.88 8.07 14 56- 52 6.49 8. 70:1 In the New Jersey greensand composts, approximately 5X gm. of sul- phur were oxidized for each gram of potassium made water soluble. For the Maryland greensand composts, the ratio is approximately 8 to i. The ratio varies with the materials used, the high-potassium greensand having a lower ratio than the low-potassium greensand, and the com- posts containing 20 per cent manure having a lower ratio than those con- taining 10 per cent manure. For the composts in which soil was substituted for all the manure the figures are not shown, but the ratio is much wider, the amount of sulphur oxidized not being sufficient to make water soluble any large amount of potassium. The results of this investigation would indicate that the composting of greensand, or of soil rich in postassium, with sulphur and manure may prove to be a practical and efficient method for obtaining available po- tassium from comparatively insoluble materials. SUMMARY Two greensands, one containing 5.88 per cent of potassium and the other 1.42 per cent, were used in studying the effect of sulphofication upon the solubility of the potassium. The outstanding results of the investigation are summarized in the following paragraphs. (i) In composts consisting of greensand, manure, and soil in different proportions, an appreciable amount of the potassium of the greensand was made water-soluble through sulphofication. (2) The composts containing the largest proportion of manure de- veloped the highest degree of acidity, oxidized the greatest amount of sulphur, and produced the largest quantity of water-soluble potassium. (3) The composts in which soil was substituted for a part of the manure developed less acidity, oxidized less sulphur, and produced a smaller amount of soluble potassium. (4) When all the manure was replaced by soil, the rate of sulphofication was so slow that at the end of 23 weeks only a very small amount of acidity had developed and very little potassium had been made soluble. June IS, i9:»o Effect of ManureStdphuY Composts on Greensand 255 (5) When no organic matter was added, the amounts of acidity and soluble sulphates were no greater than might be accounted for by the natural oxidation of the sulphur. (6) The addition of small amounts of ferrous and aluminum sulphates failed to stimulate sulphofication. (7) Calcium carbonate added to the sulphur-manure-soil compost pro- duced a stimulating effect during the early part of the period but failed to increase the acidity, soluble sulphates, or potassium above the maximum reached by the corresponding compost in which no calcium carbonate was used. (8) More water-soluble potassium was formed in the composts con- taining the high-potassium greensand, but a larger percentage of the total potassium present was liberated in the composts containing the low-potassium greensand. (9) In the composts containing manure, the total amounts of potas- sium recovered in the water extracts varied from 9.1 per cent to a maxi- mum of 41.3 per cent of the total initial amount present. (10) Our results indicate that the composting of greensand, or of soil rich in potassium, with sulphur and manure may prove to be a prac- tical and efficient method for obtaining available potassium from com- paratively insoluble materials. LITERATURE CITED (i) Ames, J. W., and Boltz, G. E. 1919. EFFECT OF SULPHOFICATION AND NITRIFICATION ON POTASSIUM AND OTHER SOIL CONSTITUENTS. In Soil Sci., V. 7, no. 3, p. 183-195. References, P- 195- (2) — — — and Richmond, T. E. 1918. EFFECT OF SULPHOFICATION AND NITRIFICATION ON ROCK PHOSPHATE. In Soil Sci., V. 6, no. 5, p. 351-364. References, p. 364. (3) Blair, A. W. 1916. THE AGRICULTURAL VALUE OF GREENSAND MARL. N. J. Agr. Exp. Sta. Circ. 61, 13 p., I fig. (4) Brown, P. E., and GwaNN, A. R. I917. EFFECT OF SULPHUR AND MANURE ON AV.\ILABILITY OP ROCK PHOSPHATE IN SOIL. Iowa Agr. Exp. Sta. Research Bui. 43, p. 369-389, 4 fig. Bibli- ography, p. 389. (s) and Warner, H. W. 1917. THE PRODUCTION OF AVAILABLE PHOSPHORUS FROM ROCK PHOSPHATE BY COMPOSTING WITH SULPHUR AND MANURE. In Soil Sci., V. 4, no. 4, p. 269-282, 3 fig. References, p. 282. (6) Cook, George H. 1868. GEOLOGY OP NEW JERSEY . . . xxiv, 899 p., illus., I fold. col. map. Newark, N. J. (7) HiLGARD, E. W. 1906. SOILS, THEIR FORMATION, PROPERTIES, COMPOSITION, AND RELATIONS TO CLIMATE AND PLANT GROWTH IN THE HUMID AND ARID REGIONS. XXVii, 593 p., illus. New York, London. 175344°— 20 2 256 Journal of Agricultural Research vo1.xix,no.6 (8) Ldpman, Jacob G., and Blair, A. W. 1918. VEGETATION EXPERIMENTS ON THE AVAILABILITY OP PHOSPHORUS AND POTASSIUM COMPOUNDS. In N. J. AgT. Exp. Sta. 38th Ann. Rpt. [i9i6]/i7, p. 353-368. (9) and McLean, Harry C. 1918. EXPERIMENTS WITH SULPHUR-PHOSPHATE COMPOSTS CONDUCTED UNDER FIELD CONDITIONS. In Soil Sci., v. 5, no. 3, p. 243-250. (10) and Lint, H. Clay. 1916. Sulphur oxidation in soils and its effect on the availability of MINERAL PHOSPHATES. In Soil Sci., V. 2, no. 6, p. 499-538. S fig- Literature cited, p. 535-538. (11) McLean, Harry C. 1918. THE OXIDATION of SULFUR BY MICROORGANISMS IN ITS RELATION TO THE AVAILABILITY OP PHOSPHATES. In Soil Sci., v. 5, no. 4, p. 251-290. References, p. 287-290. (12) Patterson, H. J. 1906. RESULTS of EXPERIMENTS ON THE LIMING OF SOILS. Md. Agr. Exp. Sta. Bui. no, 56 p. (13) IkuE, Rodney, and Geise, Fred W. I918. EXPERIMENTS ON THE VALUE OF GREENSAND AS A SOURCE OP POTASSIUM FOR PLANT CULTURE. In Jour. AgT. Research, v. 15, no. 9, p. 483-492 pi. 33-34. RUST IN SEED WHEAT AND ITS RELATION TO SEEDLING INFECTION^ By Charles W. Hungerford Assistant Pathologist, Office of Cereal Investigations, Bureau of Plant Industry, United States Department of Agriculture INTRODUCTION The fact that the mycelium of rust fungi in some cases may enter the seed and seed parts of various plants and produce spore bodies there has been known for many years and has been referred to by various writers. Differences of opinion have existed, however, as to the importance of this phenomenon in the dissemination of the rust concerned. Aside from the occurrence of rust in and upon these plant organs, other facts have seemed to indicate that rust might be transmitted by means of seed. A number of cases are on record where the uredinial and telial stages of various rusts have suddenly appeared in regions where the aecial host was unknown. Lagerheim {ijY found Puccinia coronata Cda. on oats in Ecuador, and since no species of Rhamnus known to bear the aecia of this rust occur there he concluded that the rust was probably introduced by means of oats brought from Europe. He also reported stemrust doing great damage in Ecuador, although barberry bushes were not present there. According to McAlpine {i8, p. 60), P. graminis is common in Australia, while only a very few hedges of barberry exist and the aecial stage of this rust has never been found occurring naturally upon that continent. Bolley and Pritchard (5, p. 647) quote McAlpine as saying that he is convinced that certain grass seeds secured by him from the United States Department of Agriculture introduced certain rusts into Australia. Among these he named P. coronata Cda. on the grass Beck- mannia emcaeformis and P. montanensis Ell. on wild rye {Elymus cand- densis). Numerous other similar instances could be cited. The widespread occurrence of rust epidemics has not been satisfac- torily explained, to some pathologists at least, by our present knowledge of the overwintering of the uredinial stage or by our present knowledge of the importance of infection of wheat by aeciospores from the barberry. These conditions have caused a number of writers to attempt to explain sporadic attacks of rust by a theory of seed transmission. The idea is ' The investigations reported in this paper were carried on at Madison, Wis., under the direction of the Ofifice of Cereal Investigations, United States Department of Agriculture, Washington, D. C. The writer wishes especially to thank Dr. L. R. Jones and Dr. A. G. Johnson, of the Department of Plant Pathology of the University of Wisconsin, and Dr. H. B. Humphrey, of the Office of Cereal Investigations, United States Department of Agriculture, for helpful suggestions and criticisms during the progress of the work and in the preparation of the manuscript. * Reference is made by number (italic) to " Literature cited," p. 275-277. Journal of Agricultural Research, Vol. XIX, No. 6 Washington, D. C. June 15, 1920 ul (257) KeyNo. G-I9S 258 Journal of Agricultural Research voi. xix, no. 6 not new that certain fungous parasites may exist in the vegetative state in the seeds of their hosts and be thus transmitted from one generation to another. Even before this was established for certain of the cereal smuts, various workers had endeavored to show this condition for the cereal rusts. The discovery that certain smuts were systemic in their infection gave impetus to further research along this line. The purpose of the investigations reported here was to determine whether or not Puccinia graminis triiici Erkiss. and Henn. can be trans- mitted to the seedling by being carried over with the seed grain. OCCURRENCE OF RUST IN SEEDS AND SEED PARTS OF VARIOUS PLANTS The earliest report that the v/riter has been able to find that stemrust may attack the seed and seed parts of grain was made by W. G. Smith in 1885 (2j). He found telia of Puccinia graminis in the pericarp of oat kernels and figured teliospores within the oat grains lying inside the aleurone layer and between that and the endosperm. His drawings and notes, however, leave much to be desired. In 1886 {24) the same author figured aecia embedded in the fruits of the barberry. Maddox {18, p. 20) noted rust infection upon — the young haU-grown grain . . . before it had started to go out of the milk stage. He does not state to which rust he refers. Pritchard {22, p. 151) in 191 1 figured stemrust upon wheat kernels and stated that telia and fragments of mycelium were found in abundance in the pericarp of wheat kernels and that seed infection occurs very frequently even in rust-free years. Other reports have been made of P. graminis upon the caryopses of wheat, oats, barley, and various grasses, and the writer has observed this condition upon all of the above-mentioned hosts. Puccinia glumarum (Schm.) Erikss. and Henn. is also known to occur commonly upon the caryopses of wild and cultivated Gramineae. Beau- verie (i, 2) has recently reported at length upon this phenomenon and states that if the seed is hulled the sori are produced upon the interior of the glumule, while if the seed is naked they are formed in the pericarp. He found this rust more or less abundant in the caryopses of Triticum vulgare, Hordeum vulgare, Brachypodium pinnatum, Agropyron caninum, and Bromus mollis. He also reports finding P. simplex on barley kernels and P. coronata agropyri ^ on Agropyron re pens. Blaringhem (j, p. 86) found somewhat the same conditions reported by Beauverie. Eriksson and Henning (jo, p. igg, pi. 7, 9)) fully describe and give excellent figures of whole kernels and cross sections of kernels infected with P. glumarum.^ Various other rusts have been reported as occurring upon seeds and seed parts of various plants. Carleton (6, p. 28-2g) has reported the occur- 1 It is not clear what rust is referred to by this name. » These authors cite several former observations of P. glumarum upon kernels of wheat, the earliest of which was by Schmidt {lo, p. 454) ia 1819. June 15, 1920 Relation of Rust in Seed Wheat to Seedling Infection 259 rence of Euphorbia rust (Uromyces euphorbiae C. and P.) upon seeds of Euphorbia dentaia. Various writers have noted P. malvacearum Mont. on hollyhock seeds. Otlier examples of a similar nature could be given. The discussion of the practical importance of this occurrence in relation to subsequent infection of seedlings will be taken up in a later paragraph. ABUNDANCE OF KERNEL INFECTION IN WHEAT In order to learn to what extent seed wheat may become infected with Puccinia graminis tritici a large number of wheat samples were ex- amined by the writer. These samples were secured from various sources and from the crops of the two years 1915 and 191 6. During the fall and winter of 1915-16 samples of wheat were secured from various points in North and South Dakota, from western Minnesota, from grain commis- sion firms in Minneapolis, and from wheat grown in the rust nursery at the University Farm, St. Paul, Minn. In all, several hundred samples of wheat were examined, all of which came from fields known to be badly rusted or from localities where it was known that rust epidemics had occurred. During the fall of 191 6 a large number of samples of wheat were obtained from the same regions as in the previous year. In those regions there occurred that year an unusually severe rust epidemic. It would seem, therefore, that under these conditions there would be as large an amount of seed infection as ever occurs. It was found at once that the task of determining the percentage of infection was not so easy as it at first appeared. In some cases the kernels were found to be but slightly infected, having only one sorus on the hilum, or germ end. In such cases it was impossible to see that these were infected at all except by means of examination under con- siderable magnification. In other cases the general appearance of the kernels seemed to indicate to the unaided eye that there was rust infection, but upon examination under the microscope no rust was found. Indica- tions were that such discolorations were caused by some other agency. Altemaria and Helminthosporium species were often found to be asso- ciated. In general, it was found impossible to tell in every case whether or not a kernel was infected by rust except by microscopic examination. However, in many cases, especially after some experience, many of the rust-infected kernels could be easily detected by the unaided eye. The large majority of the rust-infected kernels, when mature, were found to bear only telia,^ which appeared as gUstening black specks on the hilum, or the germ end, or a short distance down the groove of the kernel. Sometimes sori were noted a short distance from the hilum with no surface connections between these and the ones at the hilum (see PI. 39, B). Upon sectioning similar kernels, however, the mycelial connec- tions were found. If the hilar end of an infected kernel is scraped with a sharp knife or scalpel, teliospores in abundance can be secured. ' Uredinia were noted on immature kernels at various times. 26o Journal of Agricultural Research voLxix.no. 6 In order to learn the percentage of infection and also to be absolutely certain that all seed used in experimental work was infected with rust, the following method of selecting rusted kernels was employed. The samples of wheat to be examined were spread out in a shallow dish where the light was good, and the discolored kernels were taken out one by one by means of small forceps. A common 5-inch reading glass usually was used to facilitate making the selections. These discolored kernels were then placed one by one under a low-power binocular microscope where it could be easily determined at a glance if any rust sori occurred on their germ ends. As will be shown later, some infected kernels may have been missed, for sometimes the sori on the germ ends are broken off with the flowering glumes in thrashing. Bolley and Pritchard (5, p. 646) state that — in some samples of wheat in the rust-infected crop of 1904 as high as 30 per cent of all grains harvested showed such rust infection. Pritchard {22, p. 153) also states that in 1910, a rust-free year, wheat from elevators at Brookings, S. Dak., showed some rusted kernels in every sample and many in some varieties, especially Bluestem. The writer's observations do not agree with this. In all the hundreds of samples examined the largest percentage of kernels found in any one sample showing rust sori was only about i per cent of the total. Many samples were examined in which no infected kernels could be found. In fact, even in 1916, a very bad rust year, the varieties having kernel infection were the exception rather than the rule. Moreover, varieties of the durum wheat were the ones which most often were found to be infected. This was the case in both years and seems to be consist- ently so. One sample of mixed wheat from Reeder, N. Dak., collected in 1 91 6, contained about i per cent of infected kernels. Although the sample contained Marquis, durum, and Bluestem in the mixture, only durum kernels were found infected. This has been found to be the case in many mixed samples examined. Only in a few cases have any number of infected kernels of other varieties been found. This may be due to the fact that the spike of durum wheat is so compact that it dries very slowly after rains or heavy dews and these moist conditions favor infec- tion by rust. It is a well-known fact that durum varieties are very susceptible to Fusarium scab, possibly for the same reason. To illustrate this point the following observ^ation is of interest. The writer noted in 1916 at Dickinson, N. Dak., that all the durum wheats were more or less badly rusted on the heads. (See PI. 38.) This was especially true of the Kubanka strain known as selection No. 8, C. I. No. 4063 (PI. 38). A large number of heads of this variety were collected which were Hterally covered with stemrust sori (PI. 39, A). Mr. Ralph Smith, of the Office of Cereal Investigations, stationed at Dickinson, kindly furnished the writer some of the seed of this variety when the plots were thrashed. This seed was all examined carefully, and it was June IS, 1920 Relahon of Rust in Seed Wheat to Seedling Infection 261 found that only about one kernel in a thousand showed any evidence of rust infection. METHOD OF KERNEL INFECTION There are two possible methods by which kernel infection takes place, i^irst, the kernel itself may become infected by urediniospores lodging upon its surface under the glumes; or, second, the infection may spread from sori produced upon the inclosing glumes, upon the rachis or the rachilla. Since there probably are no stomatal openings upon the kernel itself and since uredinial infection takes place only through the stomata, the first possibility seems to be eliminated. Cobb (7) reports finding urediniospores of stemrust in abundance in the brush of the kernel of a large number of varieties of wheat, even after the wheat was thor- oughly cleaned. He also reports finding stomata near the brush end and concludes that infection of the kernel may take place at this point. He found sori common on wheat kernels but does not say anything with regard to their location. The writer has never found sori of stemrust produced near the brush end of wheat kernels nor has he been able to find stomata upon wheat kernels at any time in their development. As previously stated the writer has found rust sori on wheat kernels at or near the germ end. This would indicate that infection takes place by the spread of rust mycelium to the caryopsis from infection which had previously taken place at the base of the glumes or on the rachilla. Indeed, our experi- ments have confirmed this. When kernels were examined in the wheat head and were found to be infected, it was found that one or more of the flowering glumes always bore sori; and frequently several sori on the rachis, rachilla, and glumes were found to be confluent and extending over to the base of the kernel. The tissue of the hilar region of the kernel is similar in its structure to leaf tissue, and therefore infection of this region might be expected. In samples of thrashed grain kernels with adhering pieces of glumes often had rust sori extending from the base of the glume to the kernel itself. The glumes seemed to be held thus by the fungus (PI. 39, B). That infection may spread from the glumes to the kernel hilum was shown as a result of artificial inoculation experiments. These were carried out as follows. Artificial inoculations of wheat heads with urediniospores of stemrust were made in the greenhouse during the winter of 191 5-1 6. The first set of inoculations was made when the kernels were less than half grown. Urediniospores were dusted in abundance inside the glumes, and the heads were sprayed with distilled water, inclosed in large test tubes, and kept for two days. Wet cotton was kept in the bottoms of the tubes and the mouths were plugged with cotton, thus giving the conditions necessary for infection. The first attempt was a failure, either because too many spores were used or 262 Journal of Agricultural Research Vol. XIX, No. 6 because the kernels were not developed far enough to survive the inva- sion of the parasite, and the infection was so great that none of the kernels developed. The glumes and rachis at the base of the spikelet in each case were covered with sori :o days after inoculation and the inner surfaces of the glumes were filled with urediniospores. These results appear to confirm Johnson's {15) observations regarding the effect of rust infection upon floret sterility in wheat. He found floret sterility increased 20.03 per cent when wheat heads were sprayed with a water suspension of a mixture of urediniospores of Puccinia grami- nis and P. iriticina. His conclusions were that when the rust attacks the ovary early enough it prevents its development, and other semiparasitic fungi complete the process of destruction, while if it attacks the embryo after it is fertilized and has begun to enlarge, a rusted kernel results. Table I shows the outcome of a second set of inoculations. Kubanka wheat (C. I. No. 1440) was used for these experiments. The glumes were opened, and a very few spores were placed at the base of the inside of the glumes with a fine platinum needle. The heads were then sprayed with distilled water and inclosed in a test tube as before. Every spikelet in each head, with the exception of the smallest ones at the tip, was thus inoculated. Table I. — Results of artificial inoculation of wheat ovaries at different stages of develop- ment Host No. Condition of ovaries. Date of inoculation. Num- ber of heads inocu- lated. Date thrashed. Num. ber of infected kernels. Num- ber of healthy kernels. 5 6 Ovaries two-thirds grown. do Nov. 15, 1915 Nov. 18, 1915 do 2 I I I 10 4 3 Jan. 11, 1916 do do do do do do 3 2 0 0 0 3 15 8 I 7 8 Ovaries size of pinhead . . 0 do do I 9 10 Ovaries somewhat larger than above. Kernels two-thirds grown . . do do Nov. 15, 191 5 Dec. 5, 1915 2 5 8 It will be noted from Table I that in no case was kernel infection obtained when inoculations were made while the ovary was very small. On the other hand, when the inoculations were delayed until the kernels had attained about two-thirds of their normal size at maturity, the ker- nels were able to continue development, and a high percentage of rusted ones resulted. It would seem, therefore, that the amount of kernel infection each year does not depend alone upon the amount of rust occurring upon the heads of the wheat but also upon the time when this infection takes place and whether the kernels are at the right stage of development to become infected. The weather conditions where the kernels are at the right stage of development are also a very important factor. juneis,i92o Relation of Rust in Seed Wheat to Seedling Infection 263 EFFECT OF KERNEL INFECTION UPON GERMINATION Large numbers of rust-infected wheat kernels were germinated and grown to various stages of development for the purpose of making histo- logical studies. Parallel series of unrusted kernels from the same seed lot were germinated and grown for comparison. In these series it was noted that the rusted and unrusted seed gave practically identical percentages of germination. RUST TRANSMISSION WITH SEED GRAIN HISTORICAL DISCUSSION From the vast amount of work which has been done upon this problem it is possible to separate three main theories. Briefly stated, these theories are as follows : (i) Mycoplasm theory of Eriksson ; (2) dormant mycelium in the seed carrying infection to the seedling; and (3) seed- borne spores causing infection of the seedling. MYCOPLASM THEORY OF ERIKSSON Eriksson (9) in 1897 announced his well-known mycoplasm theory. He states that in the summer of 1893, upon microscopical examinations of sections of very young sori of yellow-rust (Pticcinia glumarum) upon wheat leaves, he found adjacent to these sori, besides the usual cell elements, peculiar, elongated, mostly faintly curved, plasmatic corpuscles. He concluded {p. jqj, translation) that — these plasma corpuscles, at first freely swimming in the protoplasm, constitute a phase of the fungus, the primary phase, wherein the fungus by its independent appearance makes itself visible. The fungus has for weeks, months, possibly even years, previously led a latent existence in an invisible form and alongside the pro- toplasm of the host plant, forming a kind of mycoplasm-symbiosis between host and parasite. Although Eriksson describes this mycoplasm in detail and figures it in various stages of development, very few later writers have accepted his evidence as being in any way conclusive. While it is not the present purpose to give a detailed criticism of the theory, yet, in the judgment of the writer, it seems that Eriksson's experimental evidence does not establish his contention in regard to the existence of the so-called myco- plasm. Nothing similar has been encountered in any of the hundreds of sections which the writer has made. More will be said later of this experimental evidence upon which Eriksson based his conclusions. Ward (25, p. j^s) sums the matter up very well when he states that Eriksson merely — inverts all the stages of the fungous attack on the cell, and supposes the last stage to be the first and that this error and misrepresentation of the microscopic appearance account for the whole wearisome persistence in an inherently improbable hypothesis. Detailed criticisms of Eriksson's theory are given by Bolley (4), Zukal (26), Ward (25), and Massee (20). Others could be added to this list, 264 Journal of Agricultural Research vo1.xix,no.6 but it is sufficient to say that no pathologist of note has for any length of time accepted this explanation of rust dissemination, DORMANT MYCELIUM THEORY There has been more support, and probably more ground for support, for the theory that the mycelium of rusts may live over in the seed or seed parts of the plant in a dormant state and then infect the young seedlings at the time of germination. A number of writers have sug- gested this possibility, among whom W. G. Smith {24) was probably the first. He says: If apparently healthy leaves of com are taken, and apparently healthy leaves of Barberry, and these leaves are microscopically examined, fungus mycelium will be commonly found inside the leaves. Neither is the mycelium confined to the leaves, for it invades the seeds of both plants, and these seeds are frequently planted with the mycelium in their tissues. A diseased progeny is the result. Zukal {26), in 1899, published observations which seemed to indicate to him that rust was transmitted by mycelium in seed grain. He con- cluded that rust mycelium might live over in the wheat kernels because the rust appeared so early on the young seedlings. He found septate mycelium at the base of the sheath, in the culms, and at the nodes in the parenchyma cells just under the epidermis. He concluded that the mycelium lived over in the seed and in the spring grew through the scutellum into the embryo and developed with the plant. Pritchard (22, p. 152), in 191 1, found mycelium in the roots, in both central cylinder and epidermis, in the stem, and between the leaf sheaths in plants grown from rusted wheat kernels. This myceUum resembled rust mycelium which he found at the base of the sori upon the germ end of the kernel of wheat from which the plants were grown. He states that the mycelium was abundant in the young stem, filling the intercellular spaces and freely penetiating cell walls as well. More will be said later in regard to Pritchard's work. SEED-BORNE SPORES THEORY Massee (20) secured evidence which seemed to indicate to him that seed-borne urediniospores or urediniospores in the soil might cause infection of young wheat plants. More recently Blaringhem {3) and Beauverie (j) have published extensive observations which they have made. They conclude that Puccinia glumarum may be transmitted by urediniospores borne in the pericarp of the seed. As stated above, they found uredinia in abundance in the pericarp of various grains and grasses and concluded that these spores, so protected, may retain their viability until the germination of the seed, when they become free from the sori through the rupturing of the pericarp and may infect the young plant at this time. Their conclusions, in the writer's judgment, are based upon insufficient experimental evidence, and, although the theory is interesting in itself, certainly it should be supported by more careful experiments. June IS, 1920 Relation of Rust in Seed Wheat to Seedling Infection 265 EXPERIMENTS OP VARIOUS WORKERS A number of workers have grown plants from rusted seeds of various kinds 'under various degrees of isolation. The results of these experi- ments are rather variable. The writer has assembled the results and the methods used in several of these experiments in Table II, which includes the experiments of nine men conducted at different times in different countries. None of these writers claimed to have secured normal conditions for the growth of the host plants, and in no case was any record taken of the atmospheric conditions inside the devices used to secure isolation. Table II. — Summary of results obtained by other investigators in experiments on seed transmission of rusts Experimenter. Year. Place of experiment. Rust involved. Means of isolation. Kind of seed used. Results. Eriksson (9)... Klebahn {16).. 189 2-1898 1S99 1898 J 898 1903-1907 1905 1894 Sweden .... Germany... Austria .... do P. glumarum, P. graminis. P. graminis, P. glumarum. P. glum.aTum do Ventilated glass frames. Glass cages Barley, wheat do Few posi- tive. Zukal(2(5) Isolated garden . . Glass inclosures. Glass cages Wheat do do Negative. Do Hayman (12).. India North Da- kota. England . . . Russia do.. . P. glumarum, P. triticina. Bolley (4) .... .. ..do ... do.... P. glumarum P. coronata, P. glumarum. do Bell jars . do.... Isolated cages . . . Glass cages Oats, barley. . Oats, rye Negative. Do 1902-1906 « Reference is made to Linhart's work by Zukal (26); original work not published. * Referred to by Jaczewski {14); original not seen" Eriksson carried on experiments for seven years and secured only a very few infections upon plants grown inside his "isolation frame." This frame was made of glass with wooden corner posts and an iron roof. Ventilation was secured by drawing air through a cotton filter. At best a cotton filter is not very satisfactory, and it is to be noted that Eriks- son secured his positive results after the cages had been used three or four years. Grove {ii, p. 45-47) makes an interesting comment upon Eriksson's work. He says {p. 45) — on some of his "protected" plants aphides also made their appearance, yet this does not seem to have suggested to him [Eriksson] that the zooplasm of the aphides must also have been latent in the seed . If the aphides got in, so would fimgus spores, since it has been proved that uredospores are carried by them and other insects. Klebahn repeated Eriksson's experiments and found one plant infec- ted with Puccinia graminis in his glass cages. He explains this {16) by the fact that this infection did not appear until a few days after he had been working with P. graminis near this cage. The time which had elapsed was about the normal incubation period for this rust. It seems very likely, therefore, that the one infection noted originated from spores accidentally introduced. 266 Journal of Agricultural Research voLxix.no. 6 Hayman {12) repeated his experiments for five years and grew 195 plants to maturity. The conditions inside his cages were abnormal at all times, although an effort was made to control conditions by means of a blacksmith bellows and cotton filters. Two pustules of rust ap- peared in the fifth year, but the author himself was not satisfied with this result as is evidenced by the fact that he states that the tar used to coat the inside of the cages had oozed through the cracks in the cage in which the plant was found to be infected. Massee, the only other worker who secured positive results, used bell jars placed upon cotton wool with a cotton plug in the opening at the top. He sowed wheat inside these jars, which was known to be shriv- eled by Puccinia glumarum, and as controls he sowed plump seed of the same variety. Sixty per cent of the infected seed germinated, and when the plants were 3 inches high rust appeared in each pot. When the plants were 5 inches high 26 per cent of them were rusted. Of the plump seed sown under the same conditions 96 per cent germinated, and all remained perfectly free from rust. These results are striking, and the problem with this rust is highly deserving of further investigation. Pritchard {21) grew 60 wheat plants from rusted seed in glass cages in the open and later repeated the experiment in the greenhouse. No rust appeared on any of the plants. He states that the plants headed and blossomed but no kernels developed because the temperature and moisture conditions were abnormal. He also refers to an experiment where wheat sown at different dates was inoculated with both aeciospores and urediniospores of stemrust. Rust did not appear abundantly, however, until the wheat began to head, when each sowing became thoroughly rusted. He states that it is possible to attribute this pecu- liar behavior to infection through the seed with a long subsequent incuba- tion period in the growing plant. It seems to the writer that this con- clusion is entirely unwarranted, since it is well known that infection with stemrust is much more easily obtained and more noticeable during the heading period of the plant when stemrust does such great damage by attacking the neck of the stalk. This is a period of rapid growth of the plant and a period when urediniospores are usually present in abund- ance in the air. If climatological conditions are favorable — that is, if high relative humidity and comparatively low temperatures prevail dur- ing this period — a severe rust epidemic is almost sure to follow if the infection material is present. A study of the climatological conditions during the last of June and the first of July in the spring-wheat belt shows that these conditions existed in the years when rust epidemics were severe and did not exist in the years when rust was not prevalent. These conditions are sufficient to explain any such peculiar behavior as Pritchard refers to and also help to explain rust epidemics in the spring- wheat region. June IS, 1920 Relation of Rust in Seed Wheat to Seedling Infection 267 EXPERIMENTAL DATA It is seen from the foregoing review that a number of workers have grown rust-infected seed grain under various degrees of isolation and with more or less conflicting results. The evidence of this kind as to the trans- mission of Puccinia graminis by means of seed seems to be largely nega- tive. Nevertheless some positive results have been reported. The one conclusive way to prove the contention that rust can be carried on seed grain must be to produce the disease upon plants grown under controlled conditions from seed known to be infected with the rust. While histo- logical evidence is valuable from the standpoint of interpretation, yet no amount of such work by itself is fully convincing in connecting seed in- fection wdth the appearance of the disease upon the leaves unless plants can be grown from infected seed under controlled conditions and the dis- ease be produced upon these plants. The writer's experimental investi- gations were along three lines: (i) Greenhouse experiments in which rusted kernels of wheat in large num.bers w^ere sown under isolated con- ditions and the resulting plants watched for infection; (2) field experi- ments in which rusted wheat kernels were sown in the fields and watched to learn if infection occurred upon the resulting plants sooner than upon plants grown from clean seed; and (3) histological investigations in which rusted wheat kernels were germinated under various conditions and the resulting seedlings examined histologically for spread of rust infection from the kernel to the seedling. greenhouse; experiments The writer determined to test this matter thoroughly by growing a large number of wheat plants from kernels known to bear sori of stemrust, under conditions of isolation and at the same time under conditions normal for the development of the host. In order to meet these require- ments a room in the pathological greenhouses at the University of Wis- consin was equipped as shown in Plate 40. The room was examined carefully and every crack and opening sealed. Double doors were con- structed with a space between, which could be sprayed each time before the room was entered. An adjustable shade was placed upon the roof in such a way that a spray of water could be thrown upon the glass underneath the shade to aid in cooling the room, and a system of forced circulation of washed air was installed, as shown in figure i. Thermo- graph and hygrograph records were kept at all times when the plants were growing, and it was found easily possible to control the temperature and humidity within normal limits for growth of wheat plants. Plants grown in this house were entirely normal in appearance and produced plump kernels in every head. The accompanying photographs (PI. 41) taken at different times during the period when the experiments were in progress, show the normal, healthy condition of the plants. In order to test the efficacy of this air-washing apparatus, about a pint 268 Journal of Agricultural Research Vol. XIX. No. 6 of smut spores were thrown up into the opening at i in figure i , and an attempt was made to catch any which went through the drum upon moistened sterile cotton held at h. This cotton was then washed and the water carefully examined with the microscope. No spores could be found upon this cotton, although the experiment was repeated several times. Every time the room was entered the space between the double doors was thoroughly sprayed, and a rubber coat which was kept hanging in this ante- chamber was put on. Although wheat was grown in the adjacent houses and became badly infected with mildew {Erysiphe gra- minis) none appeared on that grown inside of the isolated room. Neither did any aphids, which were plentiful at various times in other rooms in the green- house, make their way into the isolated room. The soil used in these experiments was in every case sterilized, and only boiled water -/ V I was used for watering ^ /* until after the lake ^S from which the water Fig. I.— Diagram of air-washing apparatus for isolated room used for SUpply WaS deriveu growing rust- infected seed: (a) Hose connection; (6) spray nozzle; (c) ^vg^g frOZCU OVCr. galvanized iron cylinder; ((/) greenhouse gutter into which the water f1 + f from spray drained; (e) connection pipe from cylinder to blower; rOUr aluerent iOtS OI (/) electric blower; {g) floor of greenhouse; {h) mouth of the blower j-ysted SCCd WCrC grOWn where air entered the room; (i) air intake. . t-a j. i.- •„ 4-u:„ at different times in this house. Bach lot was sown in flats 12 inches wide, 24 inches long, and 6 inches deep. These experiments will now be considered in the order in which they were performed. Experiment i. — Seed for this experiment was selected from lots of wheat obtained from the following sources: Four varieties of durum from the cereal- disease plots at Madison, Wis.; one lot of Marquis from Maynard, Iowa; one mixed lot of wheat from Leith, N. Dak.; one lot of durum from Brookings, S. Dak.; one mixed lot of unknown source from a June 15, 1920 Relation of Rust in Seed Wheat to Seedling Infection 269 grain elevator in Minneapolis, Minn. ; one lot of durum from Hagen, N. Dak.; and one mixed lot from Fargo, N. Dak. From all of these wheats rusted seed was selected and sown on November 8, 191 5, in the isolated room. Seven hundred and six plants were obtained from this seed and grown to maturity. No rust appeared upon any of the plants at any time. On the primary leaf of two different plants lesions appeared from which cultures of Helminthosporium sp. were obtained. No other infec- tion of any kind appeared upon any of these plants. Plate 41, B, shows three flats of plants from this experiment just after the plants were well headed. Experiment 2. — Experiment i was carried on during the winter months, and it was thought advisable, therefore, to duplicate the work in the spring and sow the seed at the time spring wheat normally would be sown in the field. The same precautions were taken as in experiment i, and the same room was used. Seed was secured from the following sources: Three lots of mixed seed of unknown source from Minneapolis, Minn. ; two lots of mixed seed of unknown origin from Minneapolis, Minn.; three lots of durum from the rust nursery, University Farm, St. Paul, Minn.; one lot of durum from Clark, S. Dak.; two lots of durum from the cereal -disease plots at Madison, Wis.; one lot of Mar- quis from Maynard, Iowa; one lot of durum from Leith, S. Dak.; one lot of mixed seed from Armour, S. Dak. Rusted kernels from these sources were sown on March 19, 1916, and 730 plants emerged and were grown to maturity. No rust appeared on any of these plants at any time. The experiment was discontinued when the wheat became mature. Experiment 3. — The experiment was repeated during the winter of 1 91 6-1 7, when 760 plants were grown to maturity under the same con- ditions as outlined above. Seed for this experiment was obtained from various places in North and South Dakota and Minnesota. No rust appeared upon these plants at any time. The experiment was concluded when the plants were mature. Experiment 4. — It was thought possible that soil temperatures at the time of the germination of the seed might affect the ability of the fungus to penetrate the young embryo and that the temperature in the isolated room might have been too high for successful infection at the time of germination. In order to simulate more closely natural condi- tions of germination and growth of the plants, infected wheat kernels were germinated in soil in an Altmann incubator at different temperatures as indicated in Table III. When the seedlings were about iK inches long, they were carefully transferred to pots of sterilized soil and grown in the isolated room until the plants were mature. Twenty-five kernels of wheat were used for each temperature indicated. No rust appeared upon these plants at any time. 270 Journal of Agricultural Research Vol. XIX. No 6 Table III. — Temperatures at which infected wheat kernels were germinated Date of germina- tion. Dec 12, igi6 Do Do Do Do Do Number of plants 21 20 20 23 24 24 Temperatures. -2° C. alternated with 15° C 7° C. alternated with 15° C 12° C. continuously 2° C. alternated with 21° C. 10° C. alternated with 18° C 15° C. continuously Date of transfer to greenhouse. Dec. 30, 1916. Dec. 26, 1916. Do. Do. Do. Do. Experiment 5. — A number of writers have suggested the possibility of rust infection taking place from spores on the surface of the seed. To test this possibility, several fiats of wheat were sown with seed that had been literally covered with viable urediniospores of stemrust. Preston wheat (C. I. No. 3081) was used for this experiment. In all, about 200 plants were grown. No rust infection appeared upon any of them at any time. FIELD EXPERIMENTS Experiment i. — In the spring of 191 6 rusted wheat from various sources was sown in the field along with clean seed and rusted seed treated with the modified hot-water treatment. These plots were examined every few days from the first appearance of rust infection. After June 27 the plants were examined every other day. Table IV gives the methods employed and results obtained in the experiment. The groups of plots numbered i to 4, 5 to 9, 10 to 13, and 14 to 18 were grown in different locations on the University Farm at Madison, Wis. Stemrust was noted upon Hordeum juhatum near two of the plots on July 3, 12 days before it appeared upon the wheat in these plots.^ Infec- tion also had been common upon adjacent barberries for some time previously. It will be noted that the plants grown from badly rusted samples of seed did not develop rust any earlier or any more severely than those grown from clean seed or from rusted seed which had been treated with the modified hot-water treatment. Recently the writer has had opportunity to consult the notes on an unpublished experiment somewhat similar to field experiment No. i, as described above. The work was done by E. C. Johnson, at that time Pa- thologist in Charge of Cereal Disease Investigations in the Bureau of Plant Industry, and carried on at the University Farm, St. Paul, Minn., in 191 2. The experiment is described and results are given in Mr. Johnson's report, a copy of which is on file in the Office of Cereal Investigations, Department of Agriculture, Washington. D. C. 1 By inoculating wheat plants in the greenhouse this was found to be Puccinia grammis trilici. June IS, 1920 Relation of Rust in Seed Wheat to Seedling Infection 271 •^ •>-. ^ "7^ £S: 4).- Ot3 os3 ij V & a| B| fi o 175344°— 2( + +, + + I H 22s a2^ OOSOOOSOOOO •3 -a-s-H-a^-g •q-q-o-3--a3-a-33-g 3 ^.^^.3^^. "O.^ 3 aa'O.a'^'a'a o o o >> o o •O'O'O rt 'O'd _oo>.ooo>,ooo>, oo-^oodinoooo 5£5^-§-§ .fHt)fHtlHP o o o >. o o 'O'O'O c^ 'U'O oo>.ooo>,oooo 'CO nJ'O'O'a rtTJ'O'O'a ■•*, • ■ ^, 5 ■■•! : -iS -a >. 2=3 y^ jio'a^i-'a-sia'aP'Ou'aiiiii' ai^xg, :ga :§ -x -aaa :Q ^cfiSS T^ t^ 00O^OHn(O^ lovo t^oo 272 Journal of Agricultural Research voi. xix. no. 6 Nine different varieties of wheat seed were sown, and the plants were examined for rust every four or five days. Leafrust appeared on all the plots on June 5, and stemrust appea-red from July 17 to July 29. Johnson sums up the results as follows: Rusted durum, Fife, and Bluestem kernels produced plants showing no earlier or more severe development of rust than adjacent plants from clean, uninfected seed. Experiment 2. — On April 12, 191 6, rusted kernels of wheat were sown in separate flats in the greenhouse. About 25 kernels were used from each of the following varieties: Allora (C. I. No. 1698), Kubanka (C. I. No. 1440), and Marquis (C. I. No. 3641). These flats were trans- ferred to the pathological garden May 1 1 , and were at that time in the fifth or sixth leaf. They were headed about June 22, and stemrust did not develop upon them until July 13, when a few leaves of the Marquis wheat, which still remained green, bore sori of Puccinia graminis. It will be noted by reference to Table IV that this was about the date upon which stemrust developed upon wheat in the field plots and was indeed about the date when stemrust appeared upon all the wheat in the vicinity. The season was very backward, and rust did not make its appearance nearly so early as usual. HISTOLOGY OK SEEDS AND SEEDLINGS Histology of seed. — The general appearance of the exterior of wheat kernels infected with stemrust has been previously described. In order to examine the interior of these kernels two methods were found to be fairly satisfactory: One, in which the grains of wheat were boiled in water and then sectioned on the freezing microtome ; the other, a modifi- cation of the glycerin method described by Howard (jj). This latter method was found to be satisfactory, and good sections of mature wheat kernels were obtained. After sectioning, Pianeze stain was used with good results. When sections of infected kernels were examined with a microscope it was found that not all the sori appeared upon the surface. In some instances the entire hilar region of the kernel was found to be filled with sori, of which from i to 12 were found in a single kernel. These sori often were found facing inward against the aleurone layer which was very much distorted by the pressure (PI. 42). Other sori were found, nearly spherical in form, entirely embedded in the pericarp tissue. There seemed to be no regular arrangement, although the sori were often ar- ranged in a circle around the hilum. This is what would be expected, for many of them undoubtedly were coimected with Infection on the rachilla before the kernel was broken away from the point of attachment. Plate 43 is a longitudinal section through the hilum of an infected kernel and shows the hilum nearly cut off by a large sorus, which probably was formed from several sori that had become confluent. Plate 44 is a cross June IS, 1920 Relation of Rust in Seed Wheat to Seedling Infection 273 section of a mature wheat kernel with telia upon the ventral surface. Plate 45 is an enlarged portion of the same. Internal rust sori of wheat kernels were noted and described also by Pritchard {22) . More recently CoUey {8) has listed 1 1 reports of internal rust sori upon various hosts. He concludes that these are rather com- mon teratological phenomena having no especial morphological signifi- cance and can be expected to occur whenever the point at which the sorus begins to form is located beneath a layer of tissue which is too resistant for the sorus to break through. Plates 46 and 47 also show internal sori. Histology of seedlings. — Rusted kernels of wheat were germinated under various conditions and for various lengths of time. These were fixed, sectioned, and examined for spread of infection from myceHum or spores embedded in the tissues. Various materials were used for fixing these young seedlings, but it was found that Juel's fixative penetrated the embryonic parts better than any other which was tried, although Flem- ming's medium fixative gave fairly satisfactory results. After section- ing, either triple stain with excess of Orange G or Pianeze stain was found to be satisfactory for differentiating host and fungus tissue. Infected seed was germinated under the following conditions. Seed from lot I was germinated in compartments of an Altmann incubator kept at 2°, 12°, and 17° C, respectively. Part of these were fixed when the plumule was about ^ inch long, and the rest when the first leaf was just beginning to unfold. Seed from lot 2 was germinated in com- partments of the Altmarm incubator at temperatures of 2° alternated with 17° and 11° alternated with 21°. The experiments with lots i and 2 were conducted twice — once in November, 191 5, and again in April, 191 6, after the infected seed had been kept in a cool place over winter. Lot 3 was sown in pots which were placed in small chambers in the green- house where the soil temperature was kept between 11° and 15° by the use of ice. When the plants were about 3 or 4 inches tall they were fixed, and a portion of each was sectioned and examined. Lot 4 was germinated and buried out of doors in the ground at seeding time in the spring. The plants were treated as were those in lot 3. Hundreds of sections were prepared from the material described above. In no case was there any positive evidence of spread of infection from the infected seed to the young plant. Plates 46 and 47 illustrate this fact. Plate 46 represents a longitu- dinal section through a wheat embryo in a very early stage of develop- ment. There is no indication of any spread of rust mycelium from the sori seen in the infected hilar region at x. Plate 47 also represents a longitudinal section of a wheat embryo. In this case development has progressed considerably further than that shown on Plate 46. There is, however, absolutely not the slightest indication of spread of rust myce- lium from the large sorus shown at x. 274 Journal of Agricultural Research voi. xjx, no. 6 From all appearances the rust mycelium was dead in the sori of the germinated kernels shown in Plates 46 and 47. The same was true of the rust mycelium in wheat kernels that had been stored for some time. All such mycelium was devoid of normal protoplasmic content. This fact together with the apparent inability of this mycelium to spread to the developing seedling indicates clearly to the writer that this mycelium was dead. In fact, only in fresh kernels which were not fully matured were any living rust myceHa found. Numerous efforts were made also to germinate the teliospores found in sori upon the hilar portions of wheat kernels, but all were unsuccessful. Hyphae of other organisms were present in abundance everywhere in the pericarp of many kernels and in some cases were found to penetrate the embryo. These hyphae were much larger and of an appearance different from the rust hyphae found at the base of the sori in the hilar portions of the kernels, as previously described. They penetrated directly through the cell walls of the host and broke down the cell struc- ture to a much greater extent than rust infection was found to do. Plate 48 shows an oblique longitudinal section of a secondary root of a wheat seedling being invaded by this type of parasite. This was probably some species of Helminthosporium, for typical Helminthosporium spores were found on the germ end of the kernel from which this section was made. Mycelium of the same type was found in the root, stem, and sheath of a number of seedlings which were grown from kernels of wheat having a distinct browning of the hilar ends somewhat similar to the general appearance of rust-infected kernels. It seems not entirely unlikely, therefore, that the apparently similar myceUum referred to by Pritchard {21) may have been of this type, especially since he states that the mycelium he noted also was intracellular. The writer did not find any " palmella-like " developments from the teliospores, as described by Pritchard. However, no seed over i year old was used, and since Pritchard used seed 5 years old this may to some extent account for the difference. SUMMARY (i) Uredinia and telia of Puccinia graminis triiici Erikss. and Henn. have been found embedded in the pericarp on the hilar end of kernels of wheat and sometimes along the ventral groove as far up as the middle of the kernel. Infected kernels have black hilar ends, and groups of teHa appear as shining black specks under either the hand lens or the binocular microscope. (2) Only a small percentage of infection was found by examination of the hundreds of samples of wheat from the crops of 191 5 and 191 6. A little over i per cent was the largest quantity found in any sample. The durum wheats were found most commonly infected. junei5,i92o Relation of Rust in Seed Wheat to Seedling Infection 275 (3) Infection undoubtedly spreads to the kernel from original infection on the rachis, rachilla, or glumes. (4) The germinating power of the seed apparently is not impaired by this rust infection. (5) When rusted kernels of wheat were sown in the field, no earlier or more severe rust infection occurred on the resulting plants than on those grown in adjacent plots which were sown either with clean seed or with rust-infected seed which had been treated with the modified hot-water treatment. (6) More than 2,500 plants were grown from rusted seed in a specially constructed room in the pathological greenhouse at the University of Wisconsin, and no rust infection appeared upon any of them at any time. The conditions of growth of these plants were normal, and they produced plump grain. (7) No spread of infection from the pericarp to the young plant was found by examination histologically, although infected seed were germi- nated under various conditions, simulating as nearly as possible natural conditions in the field. (8) No infection appeared upon plants grown from seed which had been covered with viable urediniospores of stemrust before sowing. (9) The results of the experimental work here reported indicate that stemrust is not transmitted from one wheat crop to the next by means of infected seed grain. Further, in the writer's judgment, the occur- rence of stemrust sori in the pericarp of the caryopses of grains and grasses has no especial significance, but the infection spreads to these tissues just as it does from an infection point in any of the vegetative parts of the plant. LITERATURE CITED (i) BeauvERIE, Jean. 1913. FREQUENCE DES GERMES DE ROUILLE DANS l'iNTiSriEUR DES SEMENCES DE GRAMIN16ES. Compt. Rend. Acad. Sci. [Paris], t. 157, no. 18, p. 787-790. (2) I914. LES GERMES DE ROUILLES DANS l'iNT^RIEUR DES SEMENCES DE GRAM- in6es. In Rev. Gen. Bot., v. 25 (bis), p. 11-27, illus. (3) Blaringhem, T. I914. SUR la PROPAGATION DES ROUILLES DE C^R^ALES EN SUEDE ET EN FRANCE. Bul. Soc. Bot. France, t. 61, no. 1/3, p. 86-94. (4) BOLLEY, H. L. 1898. EINIGE BEMERKUNGEN UBER DIE SYMBIOTISCHE MYKOPLASMATHEORIE BEi DEN GETREIDEROST. In Centbl. Bakt. [etc.], Abt. 2, Bd. 4, p. 887-896, I fig. (5) and Pritchard, F. J. 1906. RUST PROBLEMS; FACTS, OBSERVATIONS AND THEORIES. POSSIBLE MEANS OF CONTROL. N. Dak. Agt. Exp. Sta. Bul. 68, p. 607-672, 30 fig- (6) Carleton, M. a. 1904. INVESTIGATIONS OP RUSTS. U. S. Dept. AgT. Bur. Plant Indus. Bul. 63, 32 p., 2 col. pi. 276 Journal of Agricultural Research voi. xix, No.6 (7) Cobb, N. A. 1902. COMPARATIVE OBSERVATIONS ON THE BRUSH OF ABOUT FIFTY VARIETIES OF WHEAT. In Agr. Gaz. N. S. Wales, v. 13, no. 6, p. 647-649, 3 fig. (8) COLLEY, R. H. 1917. DISCOVERY OP INTERNAL TELIA PRODUCED BY SPECIES OP CRO- NARTiUM. In Jour. Agr. Research, v. 8, no. 9, p. 329-332, pi. 88. Lit- erature cited, p. 332. (9) Eriksson, Jakob. 1897. DER HEUTIGE STAND DER GETrEidErosTFrage. In Ber. Deut. Bot. Gesell., Bd. 15, Heft 3, p. 183-194. Reprinted. (10) and Henning, Ernst. 1896. DIE GETREIDEROSTE ... 463 p., 5 fig., 13 col. pi. Stockholm. Literaturverzeichnis, p. 446-457. (11) Grove, W. B. 1913. THE BRITISH RUST PUNGI ... 412 p., iUus. Cambridge, [Eng.]. Bibliography, p. 393-397. (12) Hayman, J. M. 1907. RUST IN WHEAT. In Rpt. Cawnpore Agr. Sta. United Prov. [India], 1907, p. 54-57- (13) Howard, B. J. 1903. SECTIONING OF WHEAT KERNELS. In JouT. Appl. Micros, and Lab. Methods, v. 6, no. 9, p. 2498-2499, i fig. (14) Jaczewski, a. von. 1910. STUDIEN tJBER DAS VERHALTEN DES SCHWARZROSTES DES GETREIDES IN RUSSLAND. In Ztschr. Pflanzenkrank., Bd. 20, Heft 6, p. 321- 359. 8 fig. (15) Johnson, Edw. C. 19 11. FLORET STERILITY OF WHEATS IN THE SOUTHWEST. In Phytopathology, V. I, no. I, p. 18-27. (16) Klebahn, H. 1899. Ein beitrag zur GETREIDEROSTFRAGE- In Ztschr. Pflanzenkrank., Bd. 8, Heft 6, p. 321-342, illus., pi. 6. (17) Lagerheim, G. 1893. UEBER das VORKOMMEN VON europaischen uredineen auf der hochebene von QUITO. In Bot. Centbl., Bd. 54, No. 11, p. 324-327. (18) McAxPiNE, D. 1906. RUSTS OF AUSTRALIA ... 349 p., 5 pi. (partly col.) Melbourne. Lit- erature, p. 213-221. (19) Maddox, Frank. 1897. NOTES AND RESULTS ON AGRICULTURAL EXPERIMENTS CARRIED OUT UNDER THE AUSPICES OP THE COUNCIL OF AGRICULTURE OP TASMANIA, AT EASTPIELD, NEWNHAM. 9op.,illus. Launceston. (20) MassEE, Geo. 1899. THE cereal rust PROBLEM. — DOES ERIKSSON 'S MYCOPLASMA EXIST IN NATURE? In Nat. Sci., v. 15, no. 93, p. 337-346. References, p. 346. (21) PriTchard, Frederick J. I911. A preliminary report ON THE YEARLY ORIGIN AND DISSEMINATION OP PucciNiA GRAMiNis. In Bot. Gaz., v. 52, no. 3, p. 169-192, pi. 4. Literature cited, p. 190-192. (22) 191I. THE WINTERING OF PUCCINIA GRAMINIS TRITICI E. & H. AND THE INFEC- TION OF WHEAT THROUGH THE SEED. In Phytopathology, v. i, no. 5, p. 150-154, 2 fig., pi. 22. June 15, 1920 Relation of Rust in Seed Wheat to Seedling Infection 277 (23) Smith, W. G. 1885. CORN MILDEW. In Gard. Chron., n. s. v. 24, no. 608, p. 245, fig. 53. (24) 1886. CORN MILDEW AND BARBERRY BLIGHT. In Gard. Chron., n. s. V. 25, no. 636, p. 309-310, fig. 58-60. (25) Ward, H. Marshall. 1903. "on THE HISTOLOGY OF UREDO DISPERSA, ERIKSS., AND THE ' MYCO- plasm' hypothesis." (Abstract.) In Proc. Roy. Soc. London, v. 71, no. 473. P- 353-354- (26) ZUKAL, Hugo. 1899. UNTERSUCHUNGEN UBER DIE ROSTPILZKRANKHElTEN DES GETREIDES IN OESTERREiCH-UNGARN. (i. REihe). In Sitzber. K. Akad. Wiss. [Vienna], Math. Naturw. Kl., Abt. i, Bd. 108, Heft 6/7, p. 543-562. PLATE 38 Heads of Kubanka durum wheat heavily infected with sterarust. Collected at Dickinson, N. Dak., in 1916. (27«) Relation of Rust in Seed Wheat to Seedling Infection Plate 38 Journal of Agricultural Research Vol. XIX, No. 6 Relation of Rust in Seed Wheat to Seedling Infection Plate 39 Journal of Agricultural Research Vol. XIX, No. 6 PLATE 39 A. — Portion of one of the heads shown in Plate 38. X3.6. B. — Wheat kernels showing typical stemrust infection. Abundant infection oc- curs at the base of the attached paleae on the upper row of kernels. In the lower row rust sori occur at the hilar end and along the ventral groove. X6. PLATE 40 Exterior view of isolated room in the pathological greenhouse at the University of Wisconsin, showing (a) the exterior portion of air- washing apparatus used to wash all air drawn into the room, (b) the canvas curtain used for shading on warm days, and (c) the sprinkling attachment used to throw spray of water over the roof to aid in keeping the room cool. (See fig. i and description of apparatus in text.) A. — Greenhouse with canvas curtain rolled up. B. — Greenhouse with canvas ctu-tain rolled down. Relation of Rust in Seed Wheat to Seedling Infection Plate 40 Journal of Agricultural Research Vol. XIX, No. 6 Relation of Rust in Seed Wheat to Seedling Infection Plate 41 Journal of Agricultural Researcli Vol. XIX, No. 6 PLATE 41 A. — Photograph of wheat grown in flats in isolated room in greenhouse at the Uni- versity of Wisconsin. The healthy, vigorous growth of the plants indicates that normal growing conditions prevailed in the greenhouse. B. — Same plants as A, when well headed. Plump kernels of wheat were harvested from all these plants. PLATE 42 Longitudinal section through hilar portion of an immature wheat kernel, showing sorus of stemrust. Abundance of living rust mycelium is shown at the base of the sorus. Note (at left) the alem^one layer which has been forced inward. No evi- dence of mycelial penetration into alexirone layer of cells. X245. Relation of Rust in Seed Wheat to Seedling; infection Plate 42 Journal of Agricultural Researcfi Vol. XIX, No. 6 Relation of Rust in Seed Wiieat to Seedling Infection Plate 43 Journal of Agricultural Research Vol. XiX, No. 6 PLATE 43 L,ongitudinal section through the hilum of a wheat kernel infected with stemrust, showing unusually large internal sori extending nearly across the kernel. Both external and internal sori are shown. No evidence of invasion of aleurone cells was found. X85. PLATE 44 Cross section of a matiire wheat kernel infected with stemrust, showing telia in the ventral groove. Note the normal appearance of the cells of the aleurone layer immediately beneath the sori. X30. Relation of Rust in Seed Wlneatto Seedling Infection Plate 44 Journal of Agricultural Research Vol. XIX, No. 6 Relation of Rust in Seed Wheat to Seedling Infection Plate 45 Journal of Agricultural Researcii Vol. XIX, No. 6 PLATE 45 Enlarged portion of section shown in Plate 44, showing telia on surface of ventral groove. No evidence of penetration into aleurone cells exists. X278. PLATE 46 Longitudinal section of embryo of germinated wheat kernel showing large internal rust at X in hilar tissue at base of embryonic tissue. Hundreds of such sections were examined without evidence of spread of rust infection to the embryo. X65. Relation of Rust In Seed Wheat to Seedling Infection Plate 46 Journal of Agricultural Research Vol. XIX. No. 6 Relation of Rust in Seed Wheat to Seedling Infection Plate 47 Journal of Agricultural Research Vol. XIX, No. 6 PLATE 47 Longitudinal section of the embryo further advanced in development than that shown in Plate 50. Internal hilar sorus shown at x. No evidence of infection of embryonic tissues. X67. 175344°— 20 i PLATE 48 Longitudinal section through young secondary root of wheat embryo, showing presence of intracellular mycelium. The fungus here is probably a species of Hel- minthosporium. This mycelium is larger and more vacuolated and breaks down the cells of the host much more completely than does the rust mycelium. (See PI. 42 for comparison.) X255. Relation of Rust in Seed Wiieatto Seedling Infection Plate 48 Journal of Agricultural Research Vol. XIX, No. 6 Vol. XIX JULY 1, 1920 No. 1 JOURNAL OF AGRICULTURAL RESEARCH CONXKNTS Page Practical Universality of Field Heterogeneity as a Factor Influencing Plot Yields - - 279 J. ARTHUR HARRIS ( Contribution from Bureau of Plant Industry) Transmission of the Mosaic Disease of Irisli Potatoes - 315 E. S. SCHULTZ and DONALD FOLSOM ( Contribution from Bureau of Plant Industry and Maine Agricultural Experiment Station ) PUBUSHED BY AUTHORITY OF THE SECRETARY OF AGRICULTURE, WITH THE COOPERATION OF THE ASSOQATION OF LAND-GRANT COLLEGES WASHINGTON, D. C. WA»HIMaTON ; OOVEHNMEHT PBINTINQ OFFICE : l»»» EDITORIAL COMMITTEE OF THE UNITED STATES DEPARTMENT OF AGRICULTURE AND THE ASSOCIATION OF LAND-GRANT COLLEGES FOR THE DEPARTMENT KA.RL F. KELLERMAN, Chairman J. Physiologist and A ssociate Chief, Bureau of Plant Industry EDWIN W. ALLEN Chief, Office of Experiment Stations CHARLES L. MARLATT Entomologist and Assistant Chief, Bureau of Entomology FOR THE ASSOCIATION G. LIPMAN Dean, Stale College o. Agriculture, and Director, New Jersey A gricultural Experi- ment Station, Rutgers College W. A. RILEY Entomologist and Chief, Division of Ento- mology and Economic Zoology, Agricul- tural Experiment Station of the University of Minnesota R. L. WATTS Dean, School of Agriculture, and Director, Agricuitural Experiment Station, The Pennsylvania State College All correspondence regarding articles from the Department of Agriculture should be ■addressed to Karl F. Kellerman, Journal of Agricultural Research, Washington, D. C. All correspondence regarding articles from State Experiment Stations should be addressed to J. G. Lipman, New Jersey Agricultural Experiment Station, New Brunswick, N.J. JOM£ OF AGRICETDRAL ISEARCH Vol,. XIX Washington, D. C, Jui.y i, 1920 No. 7 PRACTICAL UNIVERSALITY OF FIELD HETEROGENEITY AS A FACTOR INFLUENCING PLOT YIELDS By J. Arthur Harris Collaborator, Office of Western Irrigation Agriculture , Bureau of Plant Industry, United States Department of Agriculture INTRODUCTION With the development of a more intensive agriculture there must be a wider use and a progressive refinement of the method of plot tests in agronomic experimentation. Betterment of the method of plot tests must be sought along two lines, (i) the perfection of biological technic and (2) the more extensive use of the modern higher statistical methods in the analysis of the results. In 1 91 8 Mr. C. S. Scofield, in charge of the Office of Western Irrigation Agriculture, and Prof. E. C. Chilcott, in charge of the Office of Dry-Land Agriculture, asked the writer to undertake an investigation of the statistical phases of the problem of the accuracy of plot tests. The present paper deals with one aspect only of the general problem, that of the lack of uniformity of the experimental field. This is both the most potent cause of variation in plot yields and the chief difficulty in their interpretation. Many of the careful writers on field experimentation have noted the existence of soil heterogeneity. Few have, however, sufficiently recog- nized and none have adequately emphasized the importance of this factor. The problem of field heterogeneity is twofold. First, some measure of the amount of its influence upon crop yields must be obtained. Sec- ond, some means of avoiding or of correcting for its influence must, if possible, be secured. An exact measure of the influence of field heterogeneity, and not merely a vague notion that it may influence experimental results, is the first and most fundamental step in the closer analysis of the factors determining the variability of plot yields. If the application of such a criterion to results obtained by practised agriculturalists from fields selected for their uniformity shows no evidence of heterogeneity, plot ^ tests may be carried out along conventional lines with confidence that Journal of Agricultural Research, Vol. XIX, No. 7 Washingfton, D. C. July i. lyao urn Key No. G-196 —J (279) 28o Journal of Agricultural Research voi. xix, No. ? with reasonable precautions reliable results will be obtained. If, on the other hand, the application of such a criterion shows a high degree of irregularity in fields selected for their uniformity by experienced agri- culturalists, it is evident that very special precautions must be taken to obtain trustworthy results. Some quantitative measure, and some probable error of this measure, of the amount of irregularity of the soil of a field, as shown by actual capacity for crop production, and not merely a demonstration of its existence is, therefore, required. The purpose of this paper is to show by the analysis of the actual yields of test plots reported by agricultural experts that the securing of fields suitable for a direct comparison of yields is, practically speaking, an impossibility. The results show that unless special precautions are taken irregularities in the field may have greater influence upon the numerical results of an experiment than the factors in crop production which the investigator is seeking to compare. The results of this study may seem to be altogether negative — destruc- tive rather than constructive. The unbiased student must, however, admit that a full evaluation of all the sources of error is an essential prerequisite to constructive work. Furthermore, large expenditures of public funds are being devoted to fertilizer tests, variety tests, and rota- tion experiments. It is preeminently worth while to ascertain to what extent results derived from methods now in use may be considered reliable. Subsequent papers will treat other phases of the problem. FORMULAE A criterion of field homogeneity (or heterogeneity) to be of the greatest value should be universally applicable, be comparable from species to species, character to character, or experiment to experiment, and be easy to calculate. In 1 91 5 the suggestion was made (5)^ that we may proceed as follows: Suppose a field divided into N small plots, all sown to the same variety of plants. lyCt p be the yield of an individual plot. The variability of p may be due purely and simply to chance, since the individuals of any variety are variable and the size of the plots is small, or it may be due in part to the diversity of conditions of the soil. If irregularities in the experimental field are so large as to influence the yield of areas larger than single plots,^ they will tend to bring about a similarity of adJQining plots, some groups tending to yield higher than the average, others lower. Now let the yields of these units be grouped into m larger plots, C„, each of n continguous ultimate units, p. The correlation between the 1 Reference is made by number (italic) to " Literature cited, " p. 313-314. 2 Irregularities of soil influencing the plants of only a single small plot may in most work be left out of account, since they are of the kind to which differences between individuals are to a considerable extent due and are common to all the plots of a field. juiyi, I920 Universality of Field Heterogeneity 281 p 's of the same combination plot, C^, will furnish a measure (on the scale of o to ± I ) of the heterogeneity of the field as expressed in capacity for crop production. If this correlation be sensibly o (under conditions such that spurious correlation is not introduced), the irregularities of the field are not so great as to influence in the same direction the yields of neighboring small plots. As heterogeneity becomes greater the cor- relation will also increase. The value of the coefficient obtained will depend somewhat upon the nature of the characters measured, some- what upon the species grown, somewhat upon the size of the ultimate and combination plots, and to some degree upon the form of the combina- tion plots. Knowledge of the values of the correlations to be expected must be obtained empirically. Let 5 indicate summation for all the ultimate or combination plots of the field under consideration, as may be indicated by C„ or p. Let p be the average yield of the ultimate plots and o-p their variability, and let n be constant throughout the m combination plots. Using the for- mulae of an earlier memoir (j) in a notation which is as much sim- plified as possible for the special purposes of this discussion, {[S{C^')-S{p')] lm[nin-i)]] -p' \''2 PiP, This formula assumes the combination plots to be of uniform size — that is, to contain each the same number, n, of ultimate plots. It may be desirable or necessary to have some of the combination plots smaller than the others. Such cases are frequently met in practical work. For example, the wheat field of Mercer and Hall is laid out in a 20 by 25 fold manner. This permits only 2 by 5, 4 by 5, or 5 by 5 combinations of the same size throughout. One of Montgomery's experiments with wheat covered an area of 16 by 14 plots which may be combined in only 2 by 2 or 4 by 2 fold groupings to obtain equal areas suitable for calculation. In each of these cases other groupings are desirable. The formulae are quite applicable to such cases; the arithmetical routine is merely a little longer. The formula is as above, but p and o-p are obtained by a (w-i)-fold weighting of the plots,^ where n is the variable number of ultimate plots in the combination plot to which any p may be assigned — that is, p = S[(n-i)p]/S[n(n-i)], S[{n-i)p^] /S[in-i)p]Y 2_ P S[n{n-i)] /S[in-i)p]V \S[n{n-i)]J 1 That is, each ultimate plot is multiplied by the number less one of the plots in the combination plot to which it is assigned. 282 Journal of Agricultural Research vo1.xix,no. ? Ample illustration of the arithmetical routine has been given in the original paper. The formulae employed assume the symmetry of the correlation sur- face. It has been shown elsewhere (4) that spurious values of the cor- relation coefficient may arise in such cases. Since both p^p^ and i^ is clear that the limiting value of the spurious correlation will be o. Thus it is possible that heterogeneity exists even when rp p =0, but a field can not be considered homogeneous if rp p has a value which is statistically significant in comparison with its probable error. Practically, little difficulty will arise from this source, and it can usually be easily avoided by the exercise of a little care in the selection of the proper grouping in doubtful cases. According to the foregoing conception the relationship between the yield of associated plots is expressed on the universally comparable scale of r, ranging from o to ±1. When symmetrical tables are used — that is, when each plot is used once as a first and once as a second member of the associated pair — Pi = p2, 0 as a'3 b'3 bs . . . , where a and a' are the two halves of the same tuber and b and b' are two halves of another tuber. Thus halves a and b were grown adjoining missing hills and were subject to competition on one side only, whereas halves a' and b' were subject to competition from two adjacent plants. * Since a and a' are halves of the same tuber and b and b' are halves of another, the correlations raa', rbb' might be due to a specific physiological influence of the characters of the tuber upon both plants developing from the corresponding half tubers rather than to an influence of differences in soil conditions. We have, therefore, determined the correlations between the plants occupying the same relative position in the quadruplet but derived from different parent tubers, that is rab, ra'b'. Hence rab represents the correla- tion between the two outside tubers and ra'b' the correlation between the two inside tubers of the quad- ruplet. As a control on the results the correlations between one outside and one inside plant have been determined. These are rab' and rba'. July 1. 1920 Universality of Field Heterogeneity 285 The data given by Stewart are number of tubers and total weight of tubers per plant. These two characters permit the determinations of the average weight per tuber. When all the pairs are omitted which have been omitted by Stewart^ or have been designated as affected by leaf roll, there remain 139 quad- ruplets. Determining the correlations between the yield of the two plants derived from different tubers but exposed to the same conditions for growth, we have the following correlations: For number of tubers per hill — rab =o.3i8±o.o5i, r/E,= 6.19. ^ab = -isSi .056, r/E^^^ 2.46. ra'b = -ssoi .054, r/E^= 4.26. ra'b' = .220 ± .054, r/E,= 4.04. For total weight of tubers per hill — rab = 0.457 ± 0.045, r/Er= 10.10. rab = •3i2± .052,r/Er= 6.00. rab = •427± .047, r/E^= 9.09. ra'b = •290± .052, r/E,= 5.53. For average weight of tubers — rab = 0.237 ± 0.054, r/Er= 4-39- rab' = •i04± •057»»'/£^r= 1-82. rab = •054± -057. ^-/^r^ -QS- ra'b' = •ii7± .056, r/Er= 2.07. The correlations are positive throughout and generally statistically significant with regard to their probable errors. They show, therefore, that this experimental plot was heterogeneous to an extent that influ. enced in a very measurable degree the number of tubers, the total weight of tubers, and the average weight of tubers of neighboring hills. For all four measures of interdependence the coefficients are lowest for average weight of tubers and highest for total weight of tubers, while the correlations for number of tubers produced are intermediate in value. The values of rab are consistently higher than those for ra'b', notwith- standing the fact that a' and b' are more closely associated than a and 6. The measures of interrelationship between the yields of pairs of plants, one of which occupies an inside and the other an outside position in the quadruplet, are sometimes intermediate between Vab and ra'b' and sometimes less than ra'b'- On the assumption that the correlation is due solely to environmental influence one would expect the highest ' Records have been abstracted from Stewart's Table I. Prof. Stewart has kindly furnished some additional information in regard to certain entries in this table. 286 Journal of Agricultural Research Vol. XIX, No. 7 correlation between the most closely associated plants — that is ra'b' > rob- Apparently the reverse condition, ra'v < rah, is due to some influence of the open space adjoining a and h, which allows the fuller development of those plants and in consequence renders them more representative of the extremely localized soil influences to which they are subjected.^ 3. TIMOTHY HAY The records of plot yields of timothy hay published by Holtsmark and Larsen (r?) have been shown elsewhere (5) to present a correlation between the yield of ultimate plots, combined in a 2 by 2 fold manner, of y = o.6ii +0.027, r/Ef. = 22.4. Clearly the field was highly heterogeneous. 4. — ALFALFA HAY Records of the yields of a series of 46 plots on the Huntley Experiment Farm, Montana, may be used to test further the influence of heterogeneity on the yields of alfalfa hay. Data were kindly placed at my disposal by Mr. C. S. Scofield. Alfalfa should be of especial interest in the present discussion since it is a deep- rooted perennial herb, whereas all other herbaceous crops investigated have been annuals, or at most biennials. In field B of this experimental farm there are two series, II and III, each of 23 plots. The 46 plots form a solid block which has been planted each year to one crop just as if it were an ordinary field. The two series of plots are separated from each other only by a temporary irriga- tion ditch. Each plot is 23^ feet wide, 317 feet long, and contains approximately 0.17 acre. These plots have in certain cases been harvested in subplots of 0.085 ^cre when the division has been into halves, of 0.0,567 acre when the division has been III II b a b a 230 305 290 305 180 290 240 290 200 310 300 340 210 265 28s 355 2CX5 260 300 32s 225 285 280 345 215 28s 27s 365 220 235 270 285 255 23s 285 285 210 230 280 260 240 245 300 28s 23s 235 265 265 230 270 270 295 210 260 270 285 225 260 315 340 225 235 320 330 220 240 275 315 230 200 285 35° 255 225 295 340 265 25s 310 295 235 225 320 305 250 280 310 315 240 265 310 280 Fig. '2. — Diagram showing yield of al- falfa in first cutting, 1913, on the Huntley experimental tract. The yield is expressed in pounds per half plot. 1 Possibly competition between closely associated a' and 6' plants tends to make the yield of one lew when that of the other is high* July I, 1920 Universality of Field Heterogeneity 287 into thirds, and of 0,0425 acre when the division has been into quarters of plots. In the spring of 191 2 the whole field was uniformly seeded to alfalfa; only one crop was harvested, and yields were recorded for the entire Ill II b a b a - 70 9S I2S 13s 135 155 135 175 no 7S 8S 160 145 125 I2S 165 80 90 125 no 165 ISS 150 160 100 65 130 130 145 180 145 180 "5 95 no 125 135 165 100 140 "5 "5 135 ^25 125 185 130 155 no 95 120 "5 ■145 175 100 155 120 90 100 "5 140 150 100 180 100 90 80 105 125 150 45 150 95 95 105 120 125 140 60 145 IIS 80 95 100 120 140 65 no IIS 90 90 105 125 145 120 60 no 100 no 130 120 140 no "5 IIS 85 120 165 130 ISO 100 130 loS 105 100 145 130 150 145 140 150 95 100 95 100 150 no IIS 13s "5 90 105 95 no 100 130 155 125 120 100 65 130 "5 115 145 130 145 95 120 120 100 115 170 135 iSS 105 95 135 95 "5 I3S 125 15s 95 no 120 IIS no 140 "5 160 120 no 145 "5 130 150 100 120 160 85 15° 105 85 Fig. 3 .-Diagram showing yield of alfalfa in second cutting. 1913. on the Huntley experimental tract The yield is expressed in pounds per quarter plot. plots only. In 191 3 and 1914 three cuttings were made. The first cutting was harvested in half plots. The second cutting of 191 3 and the first and second cuttings of 1914 were harvested in quarter plots. The Journal of Agricultural Research Vol. XIX. No. 7 third cutting of 191 3 was lost because of a heavy wind which mixed the plot yields at harvest time, so that it was implossible to secure III II b a b a 85 85 130 120 130 150 140 165 105 100 105 120 135 150 140 185 100 80 105 no 120 150 170 165 105 no 95 130 165 155 150 170 100 100 105 130 120 140 145 ^85 100 105 100 125 120 17s 195 155 90 100 100 120 155 155 "5 200 90 100 105 120 85 155 145 170 120 95 90 120 115 140 170 165 85 95 75 no 155 130 105 155 75 95 85 105 85 130 125 240 60 no 90 100 120 140 160 135 75 100 75 140 95 120 120 130 55 100 75 140 120 130 125 165 75 95 85 125 120 130 140 145 85 100 60 "5 125 120 140 160 85 105 100 105 120 13s 135 150 115 100 65 115 "5 140 155 130 115 125 85 125 150 125 140 130 85 135 95 120 135 ^3S 135 135 105 120 105 105 130 140 165 145 100 "5 125 135 140 160 170 140 100 115 140 120 135 120 "5 120 Fig. 4— Diagram showing yield of alfalfa in first cutting, 1914, on the Huntley experimental tract. The yield is expressed in pounds per quarter plot. accurate weights on any of the plots. The third cutting for 1914 was harvested in subplots one-third the size of the original plots. The actual yield of these subdivisions is indicated in figure 2 ^ for the first cutting and figure 3 for the second cutting in 1913 and in figure 4 1 Diagrams are set in type instead of being drawn to scale. July 1, 1920 UniTjersality of Field Heterogeneity 289 for the first cutting, figure 5 for the second cutting, and figure 6 for the third cutting in 191 4. III II b a b a 100 no 135 125 120 145 145 140 80 85 no 120 130 145 175 15s 70 no 140 115 170 155 195 170 70 140 115 125 160 190 145 i6s 85 125 1 85 125 180 190 155 175 55 125 95 100 igo 175 185 185 65 105 115 IIS 225 155 200 195 65 no 95 no 190 190 180 165 70 105 100 135 140 155 15s 165 no 120 60 100 no 120 100 175 100 no 85 125 95 125 70 140 95 120 120 95 75 100 145 105 no .135 125 135 100 75 125 145 130 120 95 150 135 85 90 170 IIS 115 100 140 115 125 105 170 130 130 80 IIS 95 no 95 140 135 115 65 no no 85 90 150 no 115 80 120 120 130 95 180 145 1 160 75 135 120 125 105 140 140 135 80 125 105 145 155 100 135 135 90 120 IIS 155 140 125 120 155 no 130 130 130 135 130 90 1 160 no 115 1 120 130 120 75 F,o 5 -Diagram showing yield of alfalfa in second cutting. x^M, on the Huntley experimental tract. ^ ^ The yield is expressed in pounds per quarter plot. For the yield of alfalfa on quarter plots for the second cutting in 1913 and the first and second cuttings for 1914 and in third plots for the third cutting for 191 4 the correlations are 1913, second cutting, t- = o.i82 ±0.048, r/Er= 3-79- 290 Journal of Agricultural Research Vol. XIX, No. 7 1914, first cutting, r = o. 432 ±0.040, rlEj.= 10.7. 1914, second cutting, r= .449± .040, r/Ej.= ii.:^. 1 914, third cutting, r= .31 1± .052, r/Er= 5.99. Ill II X y z X y z 230 190 22s 160 240 180 220 170 130 220 220 165 215 150 130 200 205 190 17s 150 115 205 190 215 17s 15s 125 205 220 170 153 155 105 17s 160 175 190 130 125 160 175 165 15s 145 IIS 170 16s 165 170 105 no 160 15s 160 140 120 100 150 120 180 155 90 140 95 160 145 125 125 120 125 165 155 210 100 125 145 160 150 17s 140 no 180 165 140 155 145 155 180 195 165 140 115 155 165 185 125 150 125 155 170 170 120 "5 120 150 170 150 ^35 160 150 165 150 165 150 140 165 14c 150 165 160 155 155 15s 165 195 150 150 175 170 175 160 185 185 150 140 90 155 135 Fig. 6. — Diagram showing yield of alfalfa in third cutting, 1914, on the Huntley experimental tract. The yield is expressed in pounds per third plot. It will be noted that the results are in very close agreement indeed for 1 91 4. The second cutting for 191 3 differs significantly from the others, but no explanation can be suggested. July 1, 1920 Universality of Field Heterogeneity 291 Grouping all yields in two comparable subplots, we find 1 91 3, first cutting, r= 0.407 ± 0.059, y/£',= 6.93. 1913, second cutting, r= .343± .062, rlEf= 5.52. 1 914, first cutting, r= .6o2± .045, y/£'y= 13.4. 1914, second cutting, r= .657± .040, r/Ej.= 16.4. We note that all the correlations are higher for a 2 -fold division than for a 4-fold division. The coefficients for the second cutting of 191 3 are again lower than the otner values. The foregoing results are based upon weightings of single cuttings only. It is now desirable to determine the correlations for yield of first and second cuttings combined. If the combined yield be considered in quarter plots as ultimate units in 1 914 we find r==o.5i7±o.036, r/E^^ 14.2. Combining to obtain total yield in half plots in both 191 3 and 1914, we have the following correlations between the yields of the two half plots: For 1 91 3, r = 0.387 ±0.060, rlEj.= 6.46. For 1914, r= .709± .035, r/Ej.= 20.2. 5. — STRAW AND GRAIN IN WHEAT The data of the Rothamsted wheat plots,^ analyzed in an earlier paper (5, p. 436-440, 443-444), show the following correlations when the 500 plots are grouped in 2 by 2 fold manner for the first 22 files and in a 2 by 3 fold manner for the twenty-third to the twenty-fifth file: For yield of grain, r=o.336±o.027, r/Ej.= i2.5. For yield of straw, r= .483^ .023, rlEr=20.g. 6. STRAW AND GRAIN IN RAGI, ELEUSINE CORACANA Lehmann {12) has given a series of data derived from the yields of grain and straw of ragi cultivated on the dry-land tract of the Experi- mental Farm at Hebbel, near Bangalore, Mysore State. The plots used were of i/io-acre area. The land was previously owned by several raiyats who have naturally treated it somewhat differently in regard to manuring and cultivation. The various pieces used as garden lands are of course in much better condition than those used for ordi- nary dry crops. This causes considerable temporary differences to exist in some of the plots in addition to probably slight permanent differences. {12, 6th Rpt., p. 2.) . From these conditions one would expect a high degree of heteroge- neity in the series of plots. The data permit the testing of the possibility of a decrease in heterogeneity due to uniformity of crop and treatment for three years. ' For data see Mercer and Hall {15, p. iiq); also Map B of Harris (5). 292 Journal of Agricultural Research Vol. XIX, No. 7 These data are, furthermore, of particular interest since they consist of the records of yields for three successive years of the same crop on a series of unirrigated plots in a region where crop production is subject to many uncertainties because of inadequate rainfall. Fortunately for our present purposes the meteorological conditions during the three years covered by this experiment were very different from year to year. The values of the most significant factor, the July to October rainfall, are given in Table I. This shows that the rainfall in 1906 was practically twice as heavy as in either of the other two years.^ Table I. — Rainfall at Hehbel, near Bangalore, Mysore State, India July August September. October. . . . Total Inches. 1.77 6-75 1.47 5-76 15-75 1906 Inches. 7.09 9.98 5-50 8.51 31.08 Inches. 4.17 1.50 5.66 12. 14 Average of 10 years. Inches. 3-04 4-32 8.14 5-97 21.47 Maps of the fields are given in the sixth annual report for 1904- 1905. Further descriptive detail is given in the seventh, eighth, and ninth reports for 1905-1908. The yield of grain and straw in plots of i/io acre grown in 1905 is given in the seventh report. The eighth report gives detail of the crop of 1906 but does not contain the yields, which are summarized for the years 1905, 1906, and 1907 in Tables I and II of the ninth report. Unfortunately the yields of a considerable number of the plots have had to be omitted from maps I and II of Lehmann's report. In com- bining in a 2 by 2 fold manner it is necessary either to disregard all com- bination plots in which there are not four ultimate plots or to weight properly in using those containing 2 or 3 plots only. The course followed has been to group the plots by fours and to determine the correlation by the formulae for a variable number of plots when all of the ultimate plots were not planted. The following table shows the correlation between the yield of grain, of straw, and of grain and straw: 1906 Grain Straw Total yield . 3- 735 ±o- 031 .424± .055 •4i5± -055 3. 138 ±0.065 . i64± . 065 . 145 ± -065 o. 7i6±o. 032 .573± -045 .636± . 040 1 A discussion of the grow'th of these crops in relation to the distribution of the rainfall appears in Leh- mann's ninth report {i2, p. 2-7). July I, 192° Universality of Field Heterogeneity 293 The results are of unusual interest. In 1905 and 1907 the correlation between yields of grain are unusually high, falling only slightly below three-fourths of perfect correlation. The correlations for yields of straw and for both grain and straw are of medium value in those two years. In 1906, however, the correlations for all the characters are of a very low order; and any one of them taken alone might not be considered signifi- cant in comparison with its probable error, which has been calculated on the basis of 103 plots, the number actually involved in the calcula- tions. Apparently the unusual moisture conditions of 1906 tended to oblit- erate the differences in the field to which the individuality of adjoining plots was due. That the unusual weather had a profound influence on the yield of the plots is shown by Table II, in which the means, standard deviations, and coefficients of variation for the yield of the individual plots are set forth.i Table II. — Means, standard deviations, and coefficients of variation for the yield of ragi at Hehbel, near Bangalore, Mysore State, India [Yield expressed in pounds per i/io -acre plot] Year. 1905 1906 1907 Grain. 192. 8 136.6 165.0 Stand- ard devi- ation. 31-5 47.1 48.3 Coeffi- cient of vari- ation. 16. 3 34-5 29-3 360.8 191. 6 295-4 Stand- ard devi- ation. 82.0 80.2 Coeffi- cient of vari- ation. 41. 2 42.8 27. I Total yield. 553-5 328.1 460. 4 Stand- ard devi- ation. 190.3 127.4 126. 9 Coeffi- cient of vari- ation. 34-4 38.8 27. 6 The means show that yield of both grain and straw was much lower in the abnormally wet year than in either of the others. The standard deviations are of course largely influenced by the actual magnitudes of the yields and are, in consequence, difficult of interpretation. The rela- tive variabilities, as measured by the coefficients of variation, are more orderly. They show that for grain, straw, and total yield the variability of the individual plot yields is greater in the wet year. Thus the influence of the wet season has not been to make the yield of all the plots alike. It has tended to decrease yield and to increase relative variability from plot to plot. But at the same time it has tended to screen certain factors which in drier years have a marked influence on the individuality of the plots. Further analysis is not desirable without more detailed information concerning the plots. From the information at hand it seems quite 1 These constants are obtained by weighting in an (n-i)-fold manner, since this was the method followed in obtaining the constants for the heterogeneity coefficient. 294 Journal of Agricultural Research voi. xix.no. 7 clear that the innate differences in different parts of the field do not in some seasons exert their full influence upon crop yield because of the weight of other factors. The practical conclusion to be drawn from this result is that an experimental field which might be demonstrated to be sensibly uniform for one crop plant or for one season might not prove to be so for another crop or in a different season. 7. KHERSON oaTs Kiesselbach {10, 11) has given records of yield for 207 i/30-acre plots of Kherson oats. He says: These plats were planted upon a seemingly uniform field for the ptirpose of studying variation in plat yield as a source of experimental error. The entire field had been cropped uniformly to silage com for a period of eight years. It had been plowed each year and was also plowed in preparation for the oats in 1916. The oats were drilled diaring two successive days in plats 16 rods by 66 inches . The plats were separated by a space of 16 inches between outside drill rows. A wide discard border of oats was grown around the outer edge of the field, so that all plats should have a similar exposure. lyove {13) has shown the existence of heterogeneity in this field. Grouping the entries of Kiesselbach 's Table 27 in a 3 by i fold manner the heterogeneity coefficient is found to be ^ = 0-495 ±0-035, ?'/Er=i4- For data on a second test of the influence of heterogeneity on the yields of experimental plantings of oats we turn to a small experiment by Montgomery (77), who has given the yields of thrashed grain in grams from 100 consecutive rows of Kherson oats {17, p. 35, Table XIII) each 12.5 feet in length. The plat chosen for this test was quite uniform and the appearance of the plat at harvest was very satisfactory. Combining by twos, we find for the correlation between adjacent rows ^=0-339±0-o6o, r/E^=5.65. 8. GRAIN AND NITROGEN CONTENT IN WHEAT Montgomery {17, p. 37, fig. 10) has given the yield of grain in grams on 224 blocks each 5.5 feet square. Combining in a 2 by 2 fold manner we deduce r= 0.391 ±0.038, rlEj.= 10.2. Again, Montgomery {17, p. 21-22, fig. 7) has given the values of nitrogen content from 224 Turkey wheat plots of the same size. These values are quoted in figure i of this paper. The correlation between the plots is found to be r = 0.020 ± 0.045 » ^l^r = 0-44- July I, 1920 Universality of Field Heterogeneity 295 Finally, Montgomery {16) has given data for both yield of grain and nitrogen content on 224 plots of wheat grown at the University of Nebraska in 191 1 . The plot {yj by 88 feet) had been sown continuously to Turkey winter wheat for three years. The plat was of about average uniformity and fertility. When grouped in a 2 by 2 fold manner these plots of wheat have been shown (5, p. 440-441, map C) to give the following correlations: For yield of grain, r= 0.603 ±0.029, rlEj.= 2i. For percentage of nitrogen, r= .ii5± .044, r/E,.= 2.59. Yield of grain per plot is clearly influenced by irregularities of the experimental field, notwithstanding the fact that the plots are only 5.5 by 5.5 feet in area. The correlation- for percentage of nitrogen is not certainly significant. 9. — HOPS Stockberger {20) has given a series of yields for 30 rows of hops which he believes to be quite typical of many thousands of acres in the Sacra- mento Valley in California. The yields of these rows cover the period of 1909 to 1 914. Combining the rows by twos and determining the correlation between the yield of the adjacent rows of the 15 pairs for each of the years, we obtain the following constants : Year. Coirelation. rlEr. lOOQ 0.444 ±0.099 .695± .064 .o6i± . 123 .326zfc . no .6o6± .078 .386± .105 4- 5° 10.91 8.50 2.97 7-79 3-69 IQIO igil 1QI2 IQI5 IQI4 Average .419 5.06 Without exception the coefficients are positive in sign. In general they are fairly large and indicate a substantial degree of heterogeneity in this limited area. Probably the heterogeneity would have been shown to be greater had it been possible to work with yields from the sections of the long rows instead of with the rows as a whole. 10. UNHUSKED RICE Coombs and Grantham (2) give the yield in gantangs of a series of 54 square plots K by X chain in dimension. These plots are arranged in 18 ranks and 3 files. They were har- vested from a field of standing rice on which — the crop was extremely regular, as judged before the cutting, and it had not been subjected to any attack of borer or any devastation of rats or birds. 177285° — 20 2 296 Journal of Agricultural Research volxix.no. 7 The yields of the original plots are shown in figure 7. These may be combined in a 2 by i fold manner to give a correlation of f = 0.344 ±0.081, >'/£^r=4-25- These rice yields taken from a field described as "extremely regular" show that as a matter of fact the field is heterogeneous and that this irregularity influences in a measurable degree the yields of the plots. 13.6 12. 0 II. 4 14.6 14. 0 12. 2 14.8 14. 4 12. 0 13.0 12. 4 12.8 15-0 12. 0 12. 0 13-4 13-8 14. 0 14. 2 12. 2 13.0 14. 0 12. 0 12. 8 14.0 12. 0 13-4 14.0 14. c 12. 4 15.0 14.0 12. 6 14.8 14. 0 12. 4 1 14- 0 14. 0 12. 0 14. 4 13.6 12. 4 12.6 13.0 12. 0 12. 2 14.0 12.8 II. 6 12. 0 II. 8 12.4 14. 0 12.4 Fig. 7.— Diagram showing yield of unhusked rice on Coombs and Grantham's 54 plots J^ by J4 chain square. The yield is expressed in gantangs per plot. II. EAR CORN Smith {18) has published a series of corn yields for three years on plots of Vio acre. The yields are given in his original paper. He has kindly supplied the map showing the relative positions of these plots, which are arranged thus : loi, 201, . 102, 202, . 601 602 120, 220, . . . 620 July I, 1920 Universality of Field Heterogeneity 297 Combining yields in a 2 by i fold manner, we find for the correlation between the yields of adjacent Vio-^-cre plots For 1895, r= +0.830 ±0.019, ^/■£^r = 43-4- Fori896,r=+ .8i5± .021, f/Er= 39.6. For 1897, r=+ .6o6± .039 r/Er= 15.5. It is evident that the field was rather highly heterogeneous. Ill II b a b a T-2,Z 132 138 142 136 132 148 140 141 141 132 138 145 ^iS 162 156 135 109 125 135 ^Z3 116 147 130 132 153 131 131 130 123 155 150 132 137 135 140 137 112 131 129 135 132 135 131 134 126 126 135 131 128 121 125 126 115 122 136 135 125 128 131 121 "5 129 137 133 125 125 130 131 124 129 131 137 124 117 131 127 125 129 132 130 117 119 127 132 129 122 141 134 122 115 125 133 123 119 132 129 122 120 132 130 125 136 137 123 118 125 130 124 124 123 136 129 126 134 129 122 126 127 136 134 124 120 121 126 130 132 136 128 125 115 115 122 123 140 135 128 121 no no 116 115 125 123 127 124 119 107 114 116 no "5 134 112 121 123 122 126 116 125 145 148 133 125 132 127 126 134 149 154 i6s 160 162 144 137 130 168 169 165 152 158 169 143 108 Fig. 8. — Diagram showing yield of car com, 1915, expressed in pounds per quarter plot. on the Huntle/ experimental tract. The \'itJd is 298 Journal of Agricultural Research Vol. XIX, No. 7 For a second test of the influence of field heterogeneity on the yield of ear corn we turn to the Huntley data. Ill II b a b a 78 94 104 128 no 121 132 150 73 81 104 118 116 116 140 142 66 77 84 no "3 102 128 138 66 73 80 99 "5 113 128 139 77 79 79 103 116 118 126 127 71 73 86 82 100 no 108 132 76 59 86 90 no 117 in 151 94 65 86 100 102 105 118 116 98 75 80 100 in lOI 104 118 88 76 74 99 X08 92 102 113 91 82 69 80 100 97 lOI lOI 97 87 83 90 103 92 88 96 75 81 80 107 96 78 96 106 67 76 73 117 95 70 90 117 98 85 74 103 98 84 100 116 III 88 76 97 97 92 no no 108 88 73 84 84 86 104 IIS "5 97 66 89 100 87 98 123 104 120 86 ICO 94 94 97 119 110 106 92 99 96 100 83 104 118 no 100 98 "4 108 "3 120 108 100 loS no 93 99 117 104 108 98 95 100 103 99 114 98 Fig. 9.~Diagram showing yield of ear corn, 1916, on the Huntley experimental tract. The yield is expressed in pounds per quarter plot. In 191 5 and 191 6 corn was grown on the Huntley experimental plots, described above, and was harvested in quarter plots. The yields for the two series are shown in figure 8 for 191 5 and in figure 9 for 191 6. These records are of special interest in view of the fact that these are irrigated July 1, 1920 Universality of Field Heterogeneity 299 fields, whereas the data provided by Smith are based on com grown without irrigation. Retaining the original division into quarter plots, we deduce for the correlation between the subplots For 1915, r = o.498±o.o37, r/Er=i3.4. Fori9i6, r= •436± .040, r/Er= 10.8. The results for the two years can not, with due regard to their probable errors, be considered to differ significantly. They indicate a degree of heterogeneity in these Huntley plots quite comparable with that of fields planted to various crops by other observers. If the quarter plots be combined by adjacent twos and the correlation between the half plots be determined, we find For 1915, r = o.494±o.o53, r/Er = 9-29- For 1 91 6, r = .o43i± .057, r/Er=7.53. The measure of heterogeneity has been only slightly lowered by divid- ing the plots into halves instead of into quarters. INFLUENCE OF SUBSTRATUM HETEROGENEITY ON YIELD OF OR- CHARD CROPS In the preceding illustrations the crops considered have been her- baceous plants which are generally fairly superficial in their relation to the soil and most of which complete their development in one or two seasons. It seems of particular interest to extend the studies, as Batch- elor and Reed (7) have done, to the yield of large individual plants, such as orchard trees. For the purpose we employ the splendid series of data of Batchelor and Reed. They say of their various groves (i, p. 251) : The fruit plantations herein discussed, to judge by the surface soil, size, and con- dition of the trees, as well as their apparent fruitfulness, appeal to the observer as uncommonly uniform. All the orchards studied are situated in semiarid regions and are artificially irrigated during the summer months. This fact is believed to be a distinct advantage for the piurpose of reducing the variability of one year's yield compared with another, since it insures a fairly uniform water supply for the soil and reduces one of the variants inevitable in nonirrigated localities. In the case of the Arlington navel oranges grouped in 8-tree plots as the ultimate unit the authors (i, p. 264) report a correlation between plots of r = 0.533 ±0.085 when the plots are combined by fours. It has seemed desirable to test the homogeneity of the soil in each of the orchards studied by them. In determining the following coefficients the individual tree has in each case been the ultimate unit.* Consider first the relationship between the yields of adjacent trees of two navel orange groves. 1 Yields are reported in pounds per tree of ungraded product. 300 Journal of Agricultural Research voi.xix.no. ? Grouping the yield of the i ,000 trees at Arlington, shown in figure i of Batchelor and Reed, in a 2 by 2 fold manner we find r = 0.5 17 ±0.01 6, r/Er = 33.i. A navel orange grove of 495 trees at Antelope Heights, mapped as figure 2 by Batchelor and Reed, when combined in a 3 by 3 fold manner gives ^ = o-375± 0.026, rjEr= i4-4- Grouping the 240 Valencia orange trees of the grove shown in figure 3 of Batchelor and Reed in a 2 by 2 fold maimer, we find for the correlation between yields r = 0.306 ±0.039, r/Er = 7.75. For the yield in pounds per tree of Eureka lemons as shown in figure 4 of the authors cited, we find for a 2 by 2 fold grouping r = 0.448 ±0.028, r/Ej.= 15.8. This last result is of particular interest, since Batchelor and Reed say of this plantation — This grove presents the most uniform appearance of any under consideration. The land is practically level, and the soil is apparently uniform in texture. The records show a grouping of several low-yielding trees; yet a field observation gives one the impression that the grove as a whole is remarkably uniform. Notwithstanding this apparent homogeneity there is a heterogeneity coefficient of over 0.4. Taking the yields of seedling walnuts in pounds per tree as given in figure 5 of Batchelor and Reed and grouping in a 2 by 2 fold manner, we find r = o. 232 ±0.038, r/Ej. = 6.o(). Finally, if the yields in pounds per tree of the Jonathan apple trees mapped by Batchelor and Reed in their figure 6 be treated in a 2 by 2 fold grouping, the coefficient is r = 0.2 14 ±0.043, r/E;. = 4.97. Without exception these groves show material values of the hetero- geneity coefficients which are statistically significant in comparison with their probable errors throughout. PHYSICAL AND CHEMICAL BASIS OF THE HETEROGENEITY OF EXPERIMENTAL FIELDS In foregoing sections it has been shown that when tracts of land are judged by their capacity for crop production the yields are such as to indicate that heterogeneity is a practically universal characteristic of the juiyi, I920 Universality of Field Heterogeneity 301 fields which may be used for fertilizer tests, variety trials, or any other experimental purpose involving plot yields. In the vast majority of cases the heterogeneity is so great as to leave open to question conclu- sions drawn from experiments not carried out with all biological precau- tions and interpreted with due regard to probable errors. While the actual demonstration of differences in crop yields from one portion of the field to another is the result of final importance from the agronomic standpoint, and while it furnishes all but conclusive evidence that this heterogeneity in yield is due to irregularities in the soil itself, it seems desirable to show that such heterogeneity does actually obtain in the physical and chemical properties of the soil which are determining factors in plant growth. The desirability of determining the extent to which heterogeneity, in the sense to which the term is used here, obtains in the physical and chemical properties of the soil of experimental fields is emphasized by the following sentences from one of the pioneer papers (21) on the variability of soil samples. A number of papers have appeared dealing with the variation in the weight of the crop produced over different parts of an apparently uniform field. Such varia- tions reflect the variability of the soil, serving simply as a substratum for the growth of plants, but it is evident that the variations between such measurements as those given do not depend upon the soil as the only variable factor. At the outset we must recognize that many factors may determine differences in yield. Even if one could secure a tract initially uniform in soil and exposure it is not always possible to be sure that it has all been in the same crop in preceding years. Previous cultures may influence tilth and soil composition by organic remains, by infection with disease-producing organisms, or by differences in the demand of various crops for certain of the plant foods.* Such sources of hetero- geneity are not readily detected by the eye or by physical or chemical analysis. Even if the experimenter secures a field of sensibly uniform texture, chemical composition, and previous cultural treatment, the uniformity may be readily destroyed in planting or tillage. Rain may interrupt the ploughing, thus exposing the soil of the different portions of the field to air and light for different lengths of time and affecting the physical condition very profoundly. Such sources of error are par- ticularly great in the planting of large experiments. Thus the sources of field heterogeneity can never be fully determined in any case, although individual factors may be demonstrated. To determine whether an experimental field is heterogeneous with respect to physical or chemical factors, actual measurements of these factors should be made over the field and the heterogeneity coefficient applied. As a first illustration we take a series of soil-moisture ' These are factors of particular importance in rotation experiments. 302 Journal of Agricultural Research Vol. XIX, No. 7 determinations uniformly distributed over a plot on a field at the San Antonio Experimental Farm of the Office of Western Irrigation Agriculture. Hastings (6) has given a condensed account of the soil conditions of the San Antonio region. A map of the experimental farm by Hastings (7, p. 2) shows the location of field C3 in which this plot of borings was located* and gives meteorological conditions prevaihng in 191 5, the year in which the borings were made. Mr. C. S. Scofield kindly informs me that field C3 had been uniformly treated for some time previously and was in apparently uniform con- dition. It is nearly level but with a gradual slope to the south and east. The soil has the superficial appearance of uniformity, but we know from experience that the subsoil, which is usually characterized by a high lim^ content, is in some I 2 3 4 5 6 7 8 9 10 II 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 2,i 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 tg 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 Fig. 10. — Diagram showing location of sample areas examined for soil moisture in a field at the San Antonio Experimental Farm. places much closer to the surface than in others. However, from a general agronomic standpoint, this field would be regarded as extremely imiform, and observation of it diuing the growing season would tend to confirm this view. Borings were made 6 feet in depth and were sampled at every foot.^ Figure 10 shows the form of this field. In order to reduce the 100 sample areas to 2 by 2 fold combinations we have discarded the right file and a portion of one rank, retaining only those which can be grouped into fours as indicated by the cross lines. The percentages of moisture content of these 100 sample areas appear in Table III.=* ' The northern border of the sampled area is a line 60 feet south of the north line of the field and parallel to it. 2 The samples were all taken between March 31 and April 9. During this period there was no rain. Between March 15 and April 10 there were only two rains, one on March 17 of 0.2 inch, the other on March 29 of o.oi inch. Neither of these was sufficient to affect the soil moisture conditions, since in this region a precipitation of less than 0.25 inch scarcely penetrates the surface-soil mulch. Thus moisture changes during the course of the work can hardly influence the results. 2 The 1 2 sample areas which were omitted because of impossibility of combining by fours are starred (*). July I, 1920 Universality of Field Heterogeneity 303 Table III. — Moisture content of 100 sample areas of a field at the San Antonio Experi- mental Fartn [Expressed in percentages] Sample area No. First foot. Second foot. Third foot. Fourth foot. Fifth foot. Sixth foot. I 20. 2 19. I 17-5 13- I 9-7 8.9 2 23-7 21. 6 19.8 16.8 15.0 15-9 3 20. 9 20.5 19.4 16. I 15.0 15- I 4 21.5 20.3 18.3 16. 0 IS- 4 14. 2 5 23-3 22. 4 20.6 19-3 16. I 15.8 6 25.0 24.1 19.7 17.6 16. s 15-3 7 22.8 23.0 20.8 17.0 14.8 14.8 8 24. 6 24-3 20. 7 18.5 IS- 8 14.8 9 25.6 25-3 25-3 25-5 23-7 18.7 10 22. 9 25.8 26. 0 26. 2 23-5 18.6 II 28.0 30-4 30.6 29.8 26.8 21-5 12 25.2 25-7 24- s 26.8 24.0 21. 2 13* 22. I 22. 0 2a I 20. I 16. 0 14.9 14 20. 2 19.7 17.0 14- S II. 4 9. I 15 22. I 21.3 18. 1 14. 6 13.8 12.7 16 25.1 21. 2 20. 0 16.3 iS-S 14. I 17 21.8 21. 0 19. 2 i6.6 IS- 1 14.9 18 23-4 22. 4 20. 0 16. I IS- 7 15-4 19 20.5 2a 8 19. 6 IS- 6 13- S 12. 5 20 24. 0 22. 0 19-5 15-1 II- S 9.4 21 20. 4 20. 7 18.7 13.0 8.1 8.2 22 24-3 24, 0 21. 0 23-7 21.3 15-2 23 21.3 22. 2 21.8 21.7 20.5 16.7 24 24-3 25-7 24. 0 22. 4 18.3 14.9 25 ■ 23- 6 23. 2 24.4 24- 5 22. 2 18.4 304 Journal of Agricultural Research voi. xix.no. ? Table III. — Moisture content of loo sample areas of afield at the San Antonio Experi- mental Farm — Continued [Expressed in percentages! Sample area No. First foot. Second foot. Third foot. Fourth foot. Fifth foot. Sixth foot. 26* 24. 2 23-7 23.0 20. 7 19. I 17.8 27 21. I 19.7 18.7 14.7 14-5 17.7 28 21. 2 19. 6 18.4 17.6 15-2 15.0 29 21. 2 20.5 19. 6 18.9 17-5 17.1 30 22. 9 22. 0 19.9 17-5 15.0 14.8 31 21. 0 20. 7 19. 6 16. 2 14. 0 16. 4 32 23-4 21.6 19-3 18.6 16.8 IS- 9 2>i 22. 2 21.8 20. I 16.6 14. 0 14. I 34 23-9 22. 7 20. 4 17.0 14. 6 14.2 35 21. 6 20. 9 19. 2 16.8 IS- 3 16.3 36 21.4 21.6 20. 6 20. 0 18.4 16.8 37 25-3 25.6 25.6 24.9 22. 2 17.9 38 26. 7 29. 2 27. 0 25-9 23.0 19. I 39* 26. 2 29.8 30-4 28.6 26. I 21.5 40 21.8 20. 0 19-3 15-7 15-7 16.3 41 19.9 19.4 19. 0 15-3 14.8 14-9 42 21. 6 20. 0 18.2 14. I IS- 5 15.0 43 19. 6 21.7 19. 0 14.7 14. 2 13-9 44 21. 6 21.7 19.4 15.6 15-3 IS- 4 45 21. 6 20.5 19-3 16.3 7-5 14.2 46 22. 6 21.3 12. 2 16. 2 14.4 14-3 47 21. 0 22. 0 19-3 15-9 14.7 15.0 48 22. 0 21.5 19.8 19.8 14.7 15-2 49 22. 7 22. I 20. 0 19-5 16. 2 16. 2 50 21. 9 23-3 21. 0 19. I 16. 2 16.5 July I, 1920 Umversahty of Field Heterogeneity 305 Table III. — Moisture content of 100 sample areas of a field at ike San Antonio Experi- mental Farm — Continued [Expressed in percentages] Sample area No. SI 52* S3 54 55 56 57 58 59 60 61 62 63 64 65* 66 67 68 69 70 71 72 73 74 75 First foot. 20. O 29. 6 20. 6 21. 2 19-3 21. 2 22. I 22. 7 21. 2 23.2 19.4 22. 6 21.3 21. 7 22.5 20. 6 18.9 23.4 21. 2 21. 4 21. o 22.8 21. 9 22.8 21.3 Second foot. 20.3 28.4 19.8 20. 7 20. O 20.8 21. O 21. 6 21. O 22.5 21.3 21.3 20. 6 20. I 21. o 21. 2 19. 2 14-5 20. 4 20. I 21. I 21.4 21. 6 21.8 22. 8 Third foot. 19. o 27-3 18.5 18.8 18.9 18.9 19-5 19.7 19.7 20. 7 19.7 18.6 19-3 18.9 20. 2 17.9 18.2 19. o 18.8 18.4 18.9 20. o 20. 2 20.3 22. I Fourth foot. 17. 6 22. o 15-7 15- I 16.3 16.3 IS- 7 18.3 17.4 19. I 17.8 18.6 17- S 15.8 16. 7 17. I 15.0 17.7 17. o 17. o IS- 6 16.5 16. 2 18.0 31. 6 Fifth foot. 15-7 13.2 15-9 14-3 14- I 14. I IS- 1 14.7 15-2 16. s 16. 9 iS-o 17-3 iS-S 17. o IS- 8 14.4 15- 5 14. o IS- 8 14-5 15- S 14.4 15- S 17.7 Sixth foot. 17. 2 16. 2 IS- 6 14- 5 14.9 14.9 IS- 7 IS- 8 16. 4 16. 5 17. 2 17- I 17.9 16.6 20.8 15.0 15-0 IS- 5 13-9 16. o IS- 3 14-3 17-9 17.9 IS- I 3o6 Journal of Agricultural Research Vol. xrx. No. 7 Table III. — Moisture content of loo sample areas of a field at the San Antonio Experi- mental Farm — Continued [Expressed in percentages] Sample area No. First foot. Second foot. Third foot. Fourth foot. Fifth foot. Sixth foot. 76 21.5 22. 0 19.7 17-3 17. 2 16. 9 77 21.4 21. 0 19.7 15.8 15.6 18.0 78* 22. 4 21. 0 19. 2 17.0 16. 2 IS- 6 79 18.5 18.8 18.4 17.0 IS- 2 IS- I 80 20.3 19. 6 18.5 IS- 2 15.0 IS- 8 81 20.3 20. 2 18.8 16.5 14. 6 iS-S 82 21.5 21.7 18.6 IS- 8 14.8 14. I 83 20. 0 20. 4 18.7 IS- 7 16.3 IS- 4 84 20.3 20. 0 18.9 17- S 14.7 14.7 85 22. 4 21.8 21.4 17. I IS- 9 14.8 86 23.2 22. 0 19. 6 16. 0 IS- 7 15.0 87* 21.8 21. 6 20.8 19. I 17.2 16. s 88* 23-7 21.8 20. 2 16. 4 16. 9 16. 4 89* 28. 0 21. 6 20. 2 18.3 17.0 18.5 90" 23.2 21.7 19.1 16.3 16.3 16.6 91* 22.3 22. 9 21.7 19-3 18.5 18.6 92 20. 2 19.7 18.2 17.4 14.7 15.0 93 19. 0 19-3 18. 5 16. I 15-4 15-9 94 22. 0 20. 4 18.3 16. 0 14.9 14. 0 95 21- 5 19. 7 18.8 14.9 14.9 14-5 96 20.8 20.3 18.7 16.3 14.4 15-4 97 20. I 19-5 19. I 17.9 15-0 16.3 98 22. 6 20.3 19. 4 15-3 15.0 15-3 99 20. 4 20. 3 18.6 16. 4 14. 6 14-5 100* 22. 6 21.6 19.4 17-5 16.3 15.0 To determine whether the distribution of soil moisture in these plots is such that it might bring about a correlation between the yields of adjacent plots due to heterogeneity in regard to this physical factor in July 1, 1920 Unwersality of Field Heterogeneity 307 the field we have merely to determine the correlations between the percentages of water content of associated plots. These are Depth. Correlation. rIEr. Firstfoot 0. 3i7±o. 065 .S29± -052 .542± .051 . 704± • 036 .6o7± .045 .484± -055 4.9 10. 2 10. 7 19.4 13-4 8.8 Second foot Third foot Fourth foot Fifth foot Sixth foot The correlations are of a very substantial order, ranging from 0.317 to 0.704. Notwithstanding the fact that there are only 88 stations upon which the probable errors are based, the constants may in every case be considered significant in comparison with their probable errors. Thus, notwithstanding the fact that we are dealing with a field only 150 by less than 264 feet,^ there is a marked and statistically significant heterogeneity in respect to so important a factor in plant growth as soil moisture at each level in the upper 6 feet of soil. This result seems of very real importance in its relation to the practical phases of plot -test work. It shows beyond all dispute that at least under soil conditions such as are found at the San Antonio Experi- mental Farm, substratum heterogeneity may be very great at levels of the soil which are ordinarily left entirely out of account in the selec- tion of fields which are to be used for plot tests but which are not below the extensions of the roots of the deeper-penetrating crops and not too deep to serve as reserves of soil moisture for the higher layers of the soil in the case of crops which draw their water from more superficial levels. It is of some interest to detennine whether the correlations at one level in the field may be looked upon as sensibly higher than those at other levels. We have, therefore, determined the differences between the correlations at the different depths. These are given with their probable errors, and in relation to their probable errors, in Table IV. In the table the positive signs indicate higher correlations at lower levels. Of the 10 possible comparisons between the correlations of the first 5 feet, all but one show greater heterogeneity at the lower levels. The sixth foot seems to be somewhat more homogeneous than the second to the fifth foot. A number of the differences are apparently significant in comparison with their probable errors. Thus there is apparently a real dift'erence in the amount of heterogeneity of this field at different levels. Heterogeneity is least at the surface and greatest at a depth of 4 feet. The significance of this result will perhaps be apparent at once. A field might be reasonably uniform for the surface foot of soil and hence ^ The total length is 264 feet, but this is reduced by discarding the right file. 3o8 Journal of Agricultural Research Vol. XIX, No. 7 fairly well suited to the testing of shallow-rooted crops. Below this it might show a higher degree of heterogeneity. Possibly this heterogeneity of lower-lying strata is the explanation of the large correlations obtained for the yields of neighboring trees in groves planted on apparently uniform soil. Table IV. — Differences and criteria of truftworihincss of differences in the correlation of adjacent plots in soil moisture determinations at various levels Second foot. Third foot. Fourth foot. Fifth foot. Sixth foot. Depth . r. rIEr. r. rlEr. - rIEr. r rlEr. r. rIEr. First foot + 0. 212 ± .083 2.56 +0. 226 ± .082 + .013 ± -073 2.74 .18 -f 0. 387 ± -074 + -ns ± .063 + .161 ± .062 5- 22 2. 76 2.58 +0. 291 ± .079 + .078 ± .069 + .065 ± .068 — . 096 ± .058 3-68 I. 14 .96 1.66 +0. 167 ± .085 — -045 ± .076 — -059 ± -074 — . 220 ± .066 — .124 ± .071 1.97 Third foot .60 •79 Fifth foot 3-34 1.74 We can pursue this question of the relationship between the water content of the plots somewhat further. If the factors which determine the similarity in the moisture contents of the combination plots affect more than a single layer, we should expect a correlation between the contents of the first and second foot, and so on, in the same boring. The possible correlations have been worked out for the first foot and the remaining layers and are as follows : Depth. Correlation. rlEr. First and second feet +0. 748 ± 0. 032 ■i- • 669 ± . 040 4- . 648 ± . 042 -f- . 578 ± . 048 + -353 ± -063 23-59 16.84 15- 53 12. 06 5.62 First and fifth feet There is a statistically significant and even high correlation between the water content of succcessive levels in the same boring. When we turn to the problem of chemical heterogeneity, we find that while a number of soil chemists have noted the desirability of consider- ing the variability of the soil in taking samples, the available data suit- able for testing the degree of heterogeneity of experimental fields are not extensive. July 1, 1920 Universality of Field Heterogeneity 309 Kaserer's series of determinations (9) is not sufficiently large or prop- erly distributed over the field to make desirable an attempt to measure heterogeneity. Fortunately Way nick and Sharp {22) have given four excellent series, two for nitrogen and two for carbon, derived from two California fields. Their samples were taken over a total area of a little more than i .3 acres on two fields of very difiFerent character — a silty clay loam at Davis and a blow sand at Oakley. The fields were both selected for their apparent uniformity, both being nearly level with no change in the soil mass from one part of the field to another great enough to be detected by the usual field methods. Both fields were practically free from vegetation when selected, and before the samplings were made in March, 1918, all extraneous material had been carefully removed. Altogether they took 80 samples distributed at 30-foot intervals over the entire area. These samples were arranged in an 8 by 10 fold manner. The original data are given in their Tables 3 and 4. Arranging these in the order of the map of the borings given in their figure i and combining in a 2 by 2 fold manner, we derive the following heterogeneity coefiicients : For the silty clay loam at Davis — For carbon, r = 0.417 ± 0.063, ^/^r = 6.67. For nitrogen, r = .498 ± .057, r/E^ = 8.75. For the blow sand at Oakley — For carbon, r = 0.317 ± 0.068, rIEf = 4.65. For nitrogen, r = .230 ± .072, r/E^ = 3.20. All these values are statistically significant in comparison with their probable errors. Although the total number of samples is rather small, they indicate in each case a distinct heterogeneity for these important constituents of the soil. Apparently the two fields differ in their hetero- geneity, the coefficients for both carbon and nitrogen being distinctly lower on the blow sand at Oakley than on the silty clay loam at Davis- The average carbon content at Oakley is only 0.444 3-s compared with 1. 109 at Davis, while the nitrogen at Oakley is 0.033 as compared with o.ioi at Davis. Probably greater heterogeneity would be expected on general physical considerations on the silt loam than on the blow sand. The analysis may profitably be carried one step farther. If these fields are heterogeneous in respect to the soil constituents here under consideration, one might anticipate a correlation between the carbon and the nitrogen content of the samples distributed over these fields. The results are For the Davis loam, r„c = 0.785 ±0.029, rjE^^^j. For the Oakley blow sand, r„c= -744^: •034> rlEj. = 22. 3IO Journal of Agricultural Research voi xix, No. 7 Both constants are large. They show that the field is not merely hetero- geneous but that portions which are high in nitrogen are high also in carbon and vice versa. Waynick {21) has given a series of 81 determinations of nitrification in samples of soil drawn from a field on the University of California farm at Davis. The field had been planted to corn in 1914, to Sudan grass in 191 5, and to grain sorghum in 191 6. In 1917 it had lain fallow and was without vegetation when the samples were taken October 20. The particular area chosen was apparently as uniform as one could well find , being level, of uniform texture and color, and free from small local depression of any kind. These samples were taken on eight radii of a circle 100 feet in diameter. The samples were separated by a radial distance of 5 feet. Disregarding the one central sample, we may group the remainder by twos in order to determine whether there is a correlation between adjacent samples. The coefficients thus obtained will, of course, not be comparable with those deduced for cases in which the yields or soil samples were uniformly dis- tributed over the field. They will, however, serve to indicate whether or not this field is heterogeneous in the sense that differences prevailed sufficiently large to influence the properties of adjacent samples in a manner to make them more similar than pairs of samples taken at random over the field. His samples were drawn in two series — the first from the superficial 6 inches, the second from the deeper-lying level, 6 to 24 inches. Waynick's Table i gives the residual nitrate in soil as sampled. From it we deduce For the upper 6 inches, r = o.404±o.o63, r/E;.= 6.4. For the subsoil, r= •596± .049,^/^^=12.2. Table 2 gives the nitrate produced from the soil's own nitrogen after 28 days' incubation. We deduce For the upper 6 inches, r = 0.065 ±0.075, r/E^^ 0.86. For the subsoil, r= .o59± .o75,r/£r= .79. Table 3 shows the nitrate produced from 0.2 gm. of ammonium sul- phate in 100 gm. of soil. The correlation coefficients are For the upper 6 inches, r = 0.298 ±0.069, ^/^r = 4-34- For the subsoil, r= .351 ± .066, r/Er= 5.31. Finally, Table 4 shows the nitrate produced from 0.2 gm. of blood in 100 gm. of soil. The results in this case are For the upper 6 inches, r = o.i20±o.o74, r/Ef= 1.62. For the subsoil, r= .297± .069, r/Er = 4-32. July 1, 1920 Universality of Field Heterogeneity 311 The coefficients show that for both the upper and lower soil layers there is a correlation of about medium value between adjacent samples for the residual nitrate in the soil. These coefficients are unquestion- ably significant in comparison with their probable errors. While the coefficients for nitrogen produced from soil nitrogen after incubation are both positive in sign, neither can be considered statis- tically trustworthy in comparison with its probable error. When nitrogen is added to the soil, in the form of either ammonium sulphate or of blood, the correlations between the nitrogen produced on incubation are larger. All are positive in sign, and three of the four may be reason- ably considered statistically significant. Thus it is clear that this plot, only 100 feet in diameter, shows distinct heterogeneity in residual nitrate and in the amount of nitrification occurring on incubation after the addition of nitrogen. SUMMARY AND CONCLUSIONS The purpose of this paper, which is one of a series on the statistical phases of the problem of plot tests, is to show the extent to which the heterogeneity of experimental fields may influence plot yields. By heterogeneity we understand differences in capacity for crop production throughout the field of such a magnitude as to influence in like maimer, but not necessarily to like degree, the yield of adjacent small plots. Thus, variability of plot yields does not necessarily indicate the heterogeneity of the fields upon which tests are made but may be due to other factors. Heterogeneity is measured by a coefficient which shows the degree of correlations between the yields of associated ultimate plots, grouped in combination plots. This coefficient has been determined for a relatively large series of experimental fields widely distributed throughout the world and planted to a considerable variety of crops, for which a number of diff"erent kinds of yields have been measured. The results show that in every field the irregularities of the substratum have been sufficient to influence, and often profoundly, the experimental results. It might be objected that by chance, or otherwise, the illustrations are not typical of what ordinarily occurs in plot cultures. But the series considered practically exhaust the available data for such pur- poses. Furthermore the records are in large part drawn from the writings of those who are recognized authorities in agricultural experi- mentation and who have given their assurance of the suitability of the fields upon which the tests were made. For example, Mercer and Hall (75) state the purpose of their research to be — to estimate the variations in the yield of various sized plots of ordinary' field crops which had been subjected to no special treatment and appealed to the eye sensibly uniform. 177285°— 20 3 312 Journal of Agricultural Research voi. xix.no. 7 Their mangolds — looked a uniform ajid fairly heavy crop for the season and soil, while in their wheat field — a very uniform area was selected. The data of Larsen were drawn from an experiment — auf einer dem Auge sehr gleichmassig erscheinenden, 3 Jahre alten Timotheegraswiese. Montgomery's data were secured from a plot of land only yj by 88 feet in size, which had been sown continuously to Turkey wheat for three years — and was of about average uniformity and fertility. Coombs and Grantham selected a field on which — the crop was extremely regular as judged before the cutting and it had not been subjected to any attack of borer or any devastation of rats or birds. Lyon's potato field was selected from — a piece of apparently uniform land. Mr. C. S. Scofield kindly informs us that the Huntley tract was selected for apparent uniformity and that prior to the calculation of the constants presented in this paper there was no reason, from general observation, to suspect irregularities in the field. Batchelor and Reed have assured me that their orchards are to all appearances uncommonly uniform. Kiesselbach emphasizes the apparent uniformity of his oat field. Nothing could more emphasize the need of a scientific criterion for substratum homogeneity than the fact that correlations between the yields of adjacent plots ranging from r= +0.020 to r= +0.830 can be deduced from the data of fields which have passed the trained eyes of agricultural experimenters as satisfactorily uniform. A second phase of this investigation has been to ascertain whether the physical or chemical requisites for plant growth are so distributed over experimental fields that they may be reasonably looked upon as the source of the demonstrated heterogeneity in yield. The heterogeneity coefficients for percentage of water content for the upper 6 feet on the Experimental Farm of the Office of Western Irrigation Agriculture at San Antonio, Tex., range from +0.32 to +0.70 and are statistically significant for each of the 6 upper feet of soil. Hetero- geneity is least at the surface and greatest at a depth of 4 feet. The surface layer of soil might, therefore, be apparently uniform in water content while underlying layers might differ greatly from one part of the field to another. This may be the explanation of the correlation between the yields of adjacent trees in groves planted in an apparently uniform locality. Analysis of the data of Waynick and Sharp shows that there is a correlation of from +0.23 to +0.50 between adjacent borings for so important soil constituents as nitrogen and carbon. The correlation juiyi, I920 Universality of Field Heterogeneity 313 between nitrogen content and carbon content of samples from two differ- ent soils is of the order + 0.75. It is interesting to note that these coefficients for water content and for chemical composition of the soil are of about the same order as those found for crop yields. While these results do not prove that the hetero- geneity of experimental fields in their capacity for crop production is directly due to these and other physical and chemical factors, there can be little doubt that this is actually the case. The references here made to the existence of significant heterogeneity in fields passed by agricultural experts as satisfactorily uniform must not be interpreted as a criticism of the work of these investigators. There is, indeed, every evidence of care and thoroughness. The result merely illustrates the inadequacy of personal judgment concerning the uniformity in physical characters or in crop-producing capacity of fields under consideration for experimental work. The demonstration that the fields upon which plot tests have been carried out in the past are practically without exception so heteregeneous as to influence profoundly the yields of the plots emphasizes the necessity for greater care in agronomic technic and more extensive use of the statistical method in the analysis of the data of plot trials if they are to be of value in the solution of agricultural problems. To other phases of the problem we shall return in subsequent papers. LITERATURE CITED (i) Batchelor, L. D., and Reed, H. S. 1918. RELATIONSHIP OF THE VARIABILITY OP •i^ELDS OP PRUlT TREES TO THE ACCURACY OF FIELD TRIALS. In Jour. Agr. Research, v. 12, no. 5, p. 245-283, II fig. Literature cited, p. 282-283. (2) Coombs, G. E., and Grantham, J. 1916. FIELD EXPERIMENTS AND THE INTERPRETATION OF THEIR RESULTS. In Agr. Bui. Fed. Malay States, v. 4, no. 7, p. 206-216, i fig. (3) Harris, J. Arthur. I913. ON THE CALCULATION OP INTRA-CLASS AND INTER-CLASS COEFFICIENTS OP CORRELATION FROM CLASS MOMENTS WHEN THE NUMBER OP POSSIBLE COMBINATIONS IS LARGE. In Biometrika, v. 9, pt. 3/4, p. 446-472. (4) 1914. ON SPURIOUS VALUES OF INTRA-CLASS CORRELATION COEFFICIENTS ARIS- ING FROM DISORDERLY DIFFERENTIATION WITHIN THE CLASSES. In Biometrika, V. 10, pt. 2'3, p. 412-416. (5) 1915. ON A CRITERION OF SUBSTRATUM HOMOGENEITY (OR HETEROGENEITY) IN FIELD EXPERIMENTS. In Amer. Nat., v. 49, no. 583, p. 430-454. (6) Hastings, S. H. 1916. the work of the san antonio experiment farm in i915. u. s. Dept. Agr. Bur. Plant Indus. West. Irrig. Agr. 10 (Misc. Pub.), 17 p., I fig- (7) and Blair, R. E. 191 5. HORTICULTURAL EXPERIMENTS AT THE SAN ANTONIO FIELD STATION, SOUTHERN TEXAS. U. S. Dept. Agr. Bul. 162, 26 p., 8 fig. 314 Journal of Agricultural Research voi. xix.no. 7 (8) HoivTSMARK, G., and LarsEn, B. R. 1906. USER DIE FEHtER, WELCHE BEI FELDVERSUCHEN DURCH DIE UNGLEI- CHARTiGKEiT DES BODENS bedingt werdEN. In Landw. Vers. Stat., Bd. 65, Heft 1/2, p. 1-22. (9) KasERER, Hermann. 1910. BElTRAG ZUR FRAGE DER FESTEI.I.UNG DES NAHRSTOFFGEHALTES EINER ackerparzelIvE. In Ztschr. Landw. Versuchsw. Oesterr., Jahrg. 13, Heft 8, p. 742-747, I fig. (10) KlESSELBACH, T. A. 1918. STUDIES CONCERNING THE ELIMINATION OF EXPERIMENTAL ERROR IN COMPARATIVE CROP TESTS. Nebr. Agr. Exp. Sta. Research Bui. 13, 95 p., 20 fig. (11) 1919. EXPERIMENTAL ERROR IN FIELD TRIALS. In Jour. Amer. Soc. Agron., V. II, no. 6, p. 235-241. (12) IvEHMANN, A. 1901-1909. SECOND TO NINTH ANNUAL REPORTS OF THE AGRICULTURAL CHEMIST. Department of agriculture, Mysore State [India]. 1900/01-1907/08. (13) Love, H. H. 1919. THE EXPERIMENTAL ERROR IN FIELD TRIALS. In Jour. Amer. Soc. Agron., V. 11, no. 5, p. 212-216. (14) Lyon, T. L. 1912. some experiments to estimate error in field plat tests. in proc. Amer. Soc. Agron., v. 3, p. 89-114, 5 fig. (15) Mercer, W. B., and Hall, A. D. 1911. THE EXPERIMENTAL ERROR OF FIELD TRIALS. In Jour. Agr. Sci., V. 4, pt. 2, p. 107-132, 10 fig. (16) Montgomery, E. G. 1912. variation in yield and method op arranging plots to secure com- PARATIVE RESULTS. In Nebr. Agr. Sta. 25th Ann. Rpt., [i9ii]/i2, p. 164-180, 4 fig. (17) 1913. EXPERIMENTS IN WHEAT BREEDING: EXPERIMENTAL ERROR IN THE NURSERY AND VARIATION IN NITROGEN AND YIELD. U. S. Dept. Agr. Bur. Plant Indus. Bui. 269, 61 p., 22 fig., 4 pi. (18) Smith, Louie H. 191O. PLOT ARRANGEMENT FOR VARIETY EXPERIMENTS WITH CORN. In PrOC. Amer. Soc. Agron., v. i, 1907/09, p. 84-89. (19) Stewart, F. C. 1919. missing hills in potato fields: their effect upon the yield. n. y. Agr. Exp. Sta. Bui. 459, p. 45-69. 3 fig- (20) StockbergEr, W. W. 1916. relative precision of formulae for calculating NORMAL PLOT YIELDS. In Jour. Amer. Soc. Agron., v. 8, no. 3, p. 167-175. (21) VVaynick, D. D. I918. VARIABILITY IN SOILS AND ITS SIGNIFICANCE TO PAST AND FUTURE SOIL INVESTIGATIONS. I. A STATISTICAL STUDY OF NITRIFICATION IN SOILS. Univ. Cal. Pub. Agr. Sci., v. 3, no. 9, p. 243-270, 2 fig. (22) and Sharp, L. T. IQI9. VARIABILITY IN SOILS AND ITS SIGNIFICANCE TO PAST AND FUTURE SOIL INVESTIGATIONS. II. VARIATIONS IN NITROGEN AND CARBON IN FIELD SOILS AND THEIR RELATION TO THE ACCURACY OF FIELD TRIALS. Univ. Cal. Pub. Agr. Sci., v. 4, no. 5, p. 121-139, 1 fig. TRANSMISSION OF THE MOSAIC DISEASE OF IRISH POTATOES* By E. S. ScHULTz, Pathologist, Office of Cotton, Truck, and Forage Crop Disease Investi- gations, Bureau of Plant Industry, United States Department' of Agriculture, and Donald Folsom, Assistant Plant Pathologist, Maine Agricultural Experiment Station INTRODUCTION In a previous publication ^ evidence was presented that mosaic of the Irish potato is a transmissible disease. In view of the fact that a large number of the experiments establishing the transmissibility of this disease were conducted in the greenhouse, it was considered advisable to confirm those results under field conditions. Furthermore, in connection with these experiments in the field additional contributions to our knowledge of mosaic of potatoes were secured. It will be the purpose of the following pages to present these results, which, unless otherwise indicated, have been obtained in northern Maine. TUBER TRANSMISSION MODIFICATION OP SEVERITY FROM YEAR TO YEAR It is well known ^ that mosaic of Irish potatoes (Solanum tuberosum L) is transmitted from one generation of plants to another through the tubers. It has been shown ^ that there may be great variation in the severity of the symptoms shown by the progeny of a given stock, strain, hill, or tuber. Progeny of hills which appeared healthy during 191 8 while growing in plots w^hich contained some mosaic hills and which were situated near all-mosaic plots were grown and observed during the season of 191 9. Most were of the Green Mountain, some of the Bliss Triumph, and a few of the Irish Cobbler variety. Each of the various lots contained some mosaic hills, the percentage varying from 12 to 76. Altogether there were over 4,000 hills, of which 1,200, or 30 per cent, were mosaic. In view both of results reported previously ^ and of the abundance of aphids in 191 8, it seems that these mosaic hills represent cases of tuber transmission following aphid transmission occurring so late in the season of 1 91 8 that no symptoms were apparent. The severity of the symptoms ' Conducted as one of the cooperative projects between the OflBce of Cotton, Truck, and Forage Crop Disease Investigations of the Bureau of Plant Industry, United States Department of Agriculture, and the Department of Plant Pathology of the Maine Agricultural Experiment Station. 2 ScHXJLTz, E. S., Folsom, Donald, Hildebrjvndt, F. JI., and Hawkins, Lon A. investigations on THE MOSAIC DISEASE OF THE IRISH POTATO. In Jour. Agr. Research, v. 17, no. 6, p. 247-273, pi. A-B, 25-30. 1919. Literature cited, p. 272-273. Journal of Agricultural Research, Vol. XIX, No. 7 Washington, D. C. July i, 1920 un Key No. G-197 (315) 3i6 Journal of Agricultural Research Vol. XIX, No. 7 shown by the diseased hills of any lot averaged either "slight plus," "medium," or "medium plus," although it was usually "slight" for many hills and often "bad" for some. "Slight" indicates characteristic mottling sufficiently rare to require careful search; "slight plus" means that mottling is' readily apparent but is unaccompanied by wrinkling, "medium" represents both conspicuous mottling and some wrinkling, becoming "medium plus " with marked ruffling and more or less dwarfing; "bad" stands for extreme ruffling and dwarfing which may sometimes cause the mottling to be obscured. Another similar series of small lots was grown in a second plot. In these the percentage of mosaic hills varied from 4 to 63, being 40 per cent for the 800 hills altogether. The severity of the symptoms shown by the diseased hills was about the same as for the lots in the first plot. In addition to the healthy hill selections described above, stocks were grown from hills that showed mosaic in 191 8. These contained 1,100 hills, of which only 5 had not yet shown mottling by July 30. These 5, of which 4 came from one tuber, were not observed later. It is possible that this healthy tuber, supposedly from an Irish Cobbler hill with bad mosaic, was formed by a long rhizome of a neighboring healthy hill, such as is seen occasionally, and was included with the tubers of the mosaic hill in spite of the precautions usually taken. The severity of the .symptoms in the mosaic stocks is indicated in Table I. Table I. — Comparison of mosaic stocks in igi8 and IQIQ Variety. Number of hills. Severity of symptoms, 1918. Severity of symptoms, 1919. 1918. 1919. Variation between hills. Average. Green Mountain 25 34 50 18 20 17 204 269 400 112 77 65 Slight Slight to slight plus.... Slight to bad Slight. Bliss Triumph Green Mountain Bliss Triumph do Medium Bad Do. Slight plus to bad Medium plus to bad . . . do Medium. Medium . ..do plus. Bad. Irish Cobbler ... do do Do. Evidently there was, in the stocks described in Table I, a tendency for the disease to change very little in severity as a result of transmission through the tubers from 1918 to 191 9. Two larger plots, one Green Mountain and one Bliss Triumph, were planted with stock from plots entirely mosaic in 1918. While the per- centages of mottled plants on July 9 were, respectively, 67 and 89, all plants were mottled by the last of July. Although in magnitude the plants and yield were inferior to those of comparatively healthy lots, the appearance of the plants and of the plot as a whole was no worse than for the same stock during the three previous seasons. July 1, 1920 Transmission of Mosaic Disease of Irish Potatoes 317 The foregoing results indicate that mosaic in northern Maine does not necessarily change much from year to year in any diseased stock after the first appearance of the effects of infection. The conditions which determine the severity of the initial symptoms are not yet understood. RELATION TO NUMBER OF TUBERS IN A HILL The tubers from 10 Bliss Triumph hills and 130 Green Mountain hills, healthy in 191 8 but grown near to diseased hills, were all planted uncut in hill lots in 191 9. In Table II these hill lots are classified according to the number of tubers in a hill, and the percentage of tubers of each class that transmitted the disease is given. Table II. — Relation of the number of tubers in a hill to mosaic transmission Number of tubers per hill . . Number of tubers planted . . Percentage of tubers mosaic 2 3 4 5 6 7 8 9 10 II 14 27 80 QO 180 161 136 81 20 33 86 60 32 53 3« 46 46 41 30 36 41 There is a high percentage for the classes with two or three tubers to a hill, but otherwise no consistent relation obtains between number of tubers and percentage of mosaic. The results are not modified appreciably if the Bliss Triumph hill lots are disregarded. It thus seems that the increase of mosaic could be reduced by the selection of hills according to yield only if the hills with very low yields were discarded. RELATION TO RELATIVE SIZE OP TUBERS In connection with the problem of control, the question has arisen whether the selection of tubers according to size would have any effect in regard to the increase of mosaic. Consequently each of the 140 hill lots which are considered in the preceding section was planted in the order of decreasing apparent size of the tubers. With regard to mosaic 69 were mixed — that is, with both mosaic and apparently healthy plants in the same hill lot. In Table III the tubers of mixed lots are classified according to their relative rank. No. i being the largest. In addition, the percentage of tubers of each class that transmitted the disease is indicated. Table III. — Relation of the relative size of tubers in a hill to 'mosaic tra?ismission I 2 3 4 5 6 7 8 9 10 69 69 67 65 .S4 43 29 14 8 3 ^^ 48 42 46 44 42 3« 36 3« 33 Rank of tuber in size j i Number of tubers planted j 69 Percentage of tubers mosaic 1 67 The percentage is high for the group of tubers consisting of the largest ones in the hills and tends to decrease, being 48 per cent for No. 2, 45 per cent on the average for No. 3 to 6, and 36 per cent on the average for No. 7 to 10. 31 8 Journal of Agricultural Research voi. xix, No. ? Another way in which to interpret the results is to consider all tubers of a hill lot as occupying equal parts of a line and to determine the "center of disease," which is the point on the two sides of which there are equal numbers of diseased and, if also possible, of healthy tubers. This center of disease was found, for the 69 hill lots described above, to be on the average closer to the large-tuber end of the hill-lot line, 44 per cent of the line being between the two. That is, there was a greater tendency to show mosaic as the relative size of the tuber was greater. However, this tendency is not marked enough to make it seem desirable to experiment further by selecting tubers according to absolute weight or size. Of 357 hill lots planted in another plot, only the 2 to 6 largest tubers of each were planted, in order of decreasing apparent size. On July 22 to 26, 98 of the hill lots were mixed — that is, partly affected with mosaic. The results are similar to those given in Table III, the per- centages being 57, 44, 48, and 35, respectively, for groups i, 2, 3, and 4. The average center of disease is 46 per cent of the distance from the large-tuber end of the hill-lot line. Before this, on July 2 to 14, only 42 hill lots were mixed; and later, on August 22 to 25, a number of hill lots were either dead or too mature to show mosaic distinctly. RElvATlON TO POSITION OF SEED PIECE IN THE TUBER On July 29, 1 91 8, 18 tuber units were observed which had been planted with quartered tubers and were mixed. Of the hills from stem- end quarters, 45 per cent were mosaic, while 62 per cent of those from bud-end quarters were diseased. Likewise there were 24 mixed tuber units of six plants each. Of the hills from stem-end sixths, middle-part sixths, and bud-end sixths, mosaic hills constituted, respectively, 43, 54, and 61 per cent. No attempt was made to sterilize the knife used to cut the tubers. In 1 91 9 each tuber was cut by means of one of several knives used in rotation and kept, when unused, with blades immersed in 4 per cent formaldehyde solution. Observations made June 28 to July 14 dis- closed 44 tuber units, out of 1,109 observ^ed, to be mixed. In these, 48 per cent of the plants from stem-end quarters and 51 per cent of those from bud-end quarters were mosaic. This slight difference had become more marked at the time of the next observation on July 22 to 26, when 84 tuber units, out of 1,348 observed, were mixed. At that time 28 per cent of the plants from stem-end quarters were mosaic, while 61 per cent of those from bud-end quarters were diseased. This difference was reduced slightly when it was found on August 22 to 25 that 20 more of the tuber units were mixed. The preponderance of mosaic in bud-end hills is of no value in the problem of control because of the small percentage of tuber units that are rriixed. Its cause is not understood. July I, I&20 Transmission of Mosaic Disease of Irish Potatoes 319 CONCLUSIONS REGARDING TUBER TRANSMISSION Tubers from mosaic hills may be expected to transmit mosaic. In addition, at least part of those from apparently healthy hills growing near diseased plants will transmit the disease; and they tend to do so more when the parent hill contains only two or three tubers, when the relative size of the tuber in the parent hill is greater, and when the seed piece is nearer the bud end. However, hill selection results in dis- carding the hills with few tubers. The relation of relative size to mosaic transmission is not sufficiently marked or consistent to justify attempting tuber selection for the elimination of mosaic. TRANSMISSION BY GRAFTING TUBER GRAFTS Grafting was attempted with a few tubers by bringing into contact the freshly cut surface of half a mosaic tuber and half a tuber from an ap- parently healthy hill. In 14 cases the nongrafted half of the supposedly healthy tuber remained healthy, and in 3 of these 14 cases the corre- sponding grafted half produced mosaic shoots. The three cases of apparent transmission were the only ones of the attempted grafts which established organic union. The failure of transmission in the 1 1 other cases indicates that mere proximity in a hill was not sufficient for trans- mission. Furthermore, the small number of successful grafts apparently was due to the fact that relatively old tubers were used. DISEASED SCIONS UPON HEALTHY STOCKS Since transmission by grafting had been somewhat effective both in the field with insects uncontrolled and in the greenhouse with insects controlled,* the same method was finally used in the field with insects excluded by means of cages. Three tuber units were used, each con- sisting of three hills. The untreated plants, the first hill of each unit, remained healthy until dug. In each other hill two or three stalks, from 14 to 17 inches high, were cut down and split, mosaic scions inserted, and contact established with the help of cord and adhesive tape. Soon after the dates of grafting, June 28 and July 2, 191 9, the scions died because of shading in the cages; but the branches of the stocks made good growth, and by July 28 a branch in each of two grafted hills was mottled. By August 9 a number of shoots in each cage were mottled and were tagged. At the time of han.'est, August 26, these were found to belong to the grafted hills. Healthy stalks also came from these hills but were ungrafted, one even coming from the same seed-piece eye as a grafted stalk. As the Irish Cobbler variety had not been used for this kind of grafting, six mosaic scions were grafted upon uncaged stalks when the latter were ' ScHULTz, E. S., FoLSOM, Donald. Hildebrandt, F. M., and Hawkins, Lon A. op. err. 320 Journal of Agricultural Research voi. xix. no. 7 6 inches high, on June 25, 1919. One scion died immediately, and the hill remained entirely healthy. In the other cases branches from the grafted stalks showed mosaic dwarfing with wrinkling and streak necrosis and some slight mottling in the leaves, while the nongrafted ones re- mained healthy. TRANSMISSION WITH PLANT JUICE STOCKS TREATED IN 1918 Although several methods of artificial inoculation performed in 191 8 apparently had no effect/ the high percentages of mosaic shown by some of the 191 9 progeny of the treated plants indicate that certain methods were effective. Of 76 plants, progeny of control plants treated with water, 24 per cent were diseased, probably because of aphid transmission in 1 91 8; and of 463 plants, progeny of inoculated plants, 38 per cent were diseased, most of them probably because of aphid transmission. Of one lot of 53 hills, 77 per cent were mosaic. Those developed from progeny of plants which in 191 8 were inoculated by means of capillary glass tubes inserted into the petioles immediately after these capillary tubes were taken from a similar position on diseased vines. All of another lot of 28 hills were mosaic. These were progeny of plants whose stems were split and partly immersed for several days in the juice ex- pressed by crushing the tubers of mosaic plants. These two methods may be regarded as promising effective transmission if used in more extensive trials. STOCKS TREATED IN 1919 In view of the fact that mosaic of potato was transmitted by trans- ferring juice from diseased plants to the rubbed and crushed leaves of healthy plants first under greenhouse conditions,^ it was considered advisable to confirm these results with a larger number of plants and under field conditions. Consequently, during the season of 191 9 a series of similar inoculation experiments was conducted in field experimental plots, both in the open and under insect cages. INOCULATIONS WITHIN THE SAME VARIETY IN THE OPEN The first inoculation was made when the plants had reached a height of from 3 to 8 inches. The juice was expressed from the vines in a grinder and was separated at once from the pulp by straining through cheese- cloth. At each treatment the undiluted juice was applied to the leaves after they had been bruised with the fingers. At each inoculation the controls were treated with juice from healthy vines before the plants to be treated with juice from mosaic vines were operated upon. One set of 1 ScHULTz, E. S., FoLSOM, Donald, Hildebrandt, F. M., and Hawkins, Lon A. op. cit. July 1, 1920 Transmission of Mosaic Disease of Irish Potatoes 321 instruments was used for the controls and another for the virulent juice. In these experiments the Green Mountain, Bliss Triumph, and Irish Cobbler varieties were used. In each case the juice was taken from vines of the same variety. The plants of the Green Mountain and IBliss Triumph varieties used in this experiment developed from progeny which in 191 8 showed from 11 to 15 per cent of mosaic, eliminated in three roguings. In view of the fact that, with the exception of the Irish Cobbler variety, these were planted by using four seed pieces from a tuber, it was possible to inoculate two of the hills in a tuber unit and have two additional hills of the same tuber unit remaining as uninoculated controls. In each tuber unit the plants in the second and third hills were inoculated — that is, a hill from a stem-end quarter and one from a bud-end quarter. In Table IV are given the results of these inoculations. From Table IV it is apparent that plants not infected in 191 8 if treated with juice from healthy vines remained healthy to the end of the season. (See PI. 49-51.) As indicated, the exceptions to this result, where some tuber units produced plants which became mottled with mosaic after being treated with juice from healthy plants, were due to the fact that such units had become infected in 191 8 in the field but did not present any evidence of infection at the time of the first treatment in 1 91 9. Plants inoculated with juice from mosaic-diseased vines showed the first mosaic mottling upon the newly developed leaves July 14. At this time aphids were just beginning to appear at the rate of a few individuals to a plant, so that those agents of dissemination can be disregarded as a factor in transmission in these open-field inoculations. It will be noted that with virulent juice a certain number of tuber units showed mottling throughout within a few days after the first inoculation, indicating that the tubers had become infected in 191 8. In the remaining inoculated hills every hill, with the exception of one of Bliss Triumph, showed dis- tinct mosaic mottling, while the untreated hills of these same units re- mained healthy to the end of the season. In addition to the mosaic mottling, distinct spotting and streaking of the leaves, petioles, and stems obtained by July 25, so that at this time some of the lower leaves began to die. Furthermore, a marked ruftling and dwarfing of the leaves also became apparent, so that many of the plants appeared like those in the medium plus or bad stage, indicating that in a single season plants may develop an aggravated form of this disease if inoculated properly. (Pi. 52.) 322 Journal of Agricultural Research Vol. XIX, No. 7 ^ ^ °°°88S^8 00 Si ^ a S S 2-^ 0 o 0) a 3.2 .a " S3 S U O li . aia&2| 0 O 0 " N " '^ '2 o — a VO t^CO M !H N P< •2 t z s i-g 3 o 0 a '' fi Tt 'O 0 0 CO CO o »H l-( rt C* « W «.fl o g H3 a • u T t^ O : 00 00 • 0) O o 1 TTJ fiilli "2 O O 00 w l-t : o o : .g 5.Q 1 a '^ Wi lom ^ loo ■ 01 B . • MM . •2-Qi2 ill :2i^^ • 2 ■ >> , H ..^ 5J a • • :-§:§ 2 5 . : • al -a 1 to : t) : w ..a "o N : > 1:^ o ; "■ ■^ ■ a : o ■ ^ .« o " o " o " 1 ojtJ ajT) OJ'O 1 a ; a : a : a a . ; . a ■ a , ;!-> :!-> •«J3 :.9 : ; ti 0.S- . rij3 . 3 K g to J3 J3 ja.o.o.o o .-H ii f— 1 rt m rt rt rt aj cS t/i t/: t/J t« o aj o O o O o ffiSffiSSSS •d •2 3 0 : ■ '.9 • : 2 "a Si : : rt.a : .9 >■ 1 ' ill e^ V «.a S oS'£ OPQh juiyi, I920 Transmission of Mosaic Disease of Irish Potatoes 323 Application of Juice from Plants Showing the Bad Stage of Mosaic In order to determine whether juice taken from plants badly diseased with mosaic and introduced into healthy plants would induce bad mosaic symptoms in the latter, plants of the Green Mountain, Bliss Triumph, and Irish Cobbler varieties were inoculated in the same manner as those mentioned in Table IV. Three applications at weekly intervals were made upon plants of the same variety as that from which the juices were expressed. The height of the vines at the time of the first inocula- tion varied from 2 to 8 inches. Plants of five Green Mountain hills, three Bliss Triumph hills, and five Irish Cobbler hills were treated. At the same time also two Green Mountain, three Bliss Triumph, and two Irish Cobbler hills were treated with but a single inoculation. On July 28, 16 days after the first treatment, the first mosaic mottling was noted upon the inoculated Irish Cobbler vines. By August 1 5 every inoculated plant, regardless of variety, showed distinct mosaic mottling as well as streaking and ruffiingof the leaves as in the bad stage of mosaic; and by August 28 most of the leaves on the lower half of the stems were dead. The plants subjected to but a single inoculation showed symp- toms similar to those given three successive treatments, indicating that a single treatment may be sufficient to induce the disease (PI. 56). INOCULATIONS WITHIN THE SAME VARIETY UNDER INSECT CAGES Early Repeated Application Juice from crushed mosaic plants (not necessarily mottled at the time of the first inoculation but from stock all mosaic in 191 8) was applied to the bruised leaves of two hills in each of three caged tuber units on June 13, 20, and 27, and on July 5. As a control, the third hill in each cage was left untreated; also juice from apparently healthy plants was applied to the bruised leaves of two hills of each of three other caged tuber units on the same dates. In all these cases the plants were from i to 6 inches high at the first treatment. On July 9 the topmost leaves of the treated hills in the former three units began to show mottling, which was slight to medium by July 15. On July 30 mosaic branches in these units were tagged and were found at digging time, August 26, to belong to the treated hills, which had no healthy stalks. The tuber units upon which the juice from healthy plants was used remained green and healthy until digging time, while those died which became mosaic. Late Application On July 14 two hills in each of two caged tuber units were treated with juice from mosaic plants in the same manner as those described in the two preceding sections. Before August 20 the upper leaves of the treated hills became mottled and streaked. 324 Journal of Agricultural Research voi. xix.no. 7 INOCULATIONS FROM ONE VARIETY TO ANOTHER IN THE OPEN Early Repeated Application of Juice In order to determine whether the juice of a mosaic plant of one variety could induce the disease when introduced into a plant of a different variety of potato, intervarietal inoculations were made under open field conditions. The procedure of inoculation practiced in this connection was similar to that followed with the inoculations indicated in Table TV. In this experiment the control plants always were treated before mosaic juice was used, and a separate set of unstruments was em- ployed for each distinct variety and for juice from each source. Green Mountain, Bliss Triumph, and Irish Cobbler varieties were used. These were subjected to four successive treatments at weekly intervals, as indicated in Table V. The results given in Table V show that mosaic juice from one variety of potato may produce the disease when introduced into the plants of another variety. In these inoculations the effect upon the treated plants was fully as severe as that obtained when juice was introduced into plants of the same variety, as explained in connection with Table IV. In fact, in many cases the inoculated plants behaved like those in the late or bad stage of the disease. (Pi. 53-55.) From Table V it is apparent that a large percentage of the plants had become infected in 191 8. In view of the fact that such tuber units did not show the mosaic mottling at the time of the first inoculation, when the plants varied in height from 2 to 8 inches, it was impossible to restrict inoculation to healthy units. However, in this connection it is inter- esting to note that the hills infected in 191 8 and inoculated in this ex- periment showed the disease like the plants in the bad' stage whenever the uninoculated control hills in the same tuber units showed but slight or medium infection, so that apparently inoculation with juice increased the severity of the infection which had resulted from transmission in the field the previous season. Since a considerable number of the plants in this experiment appar- ently had become infected in 191 8, the evident objection might be offered that, in the course of the inoculation, infectious juice was carried from diseased to healthy plants of the same variety and thus caused infection. This objection can be eliminated. Inoculations always were commenced at the same end of the plot and row, and hence the respective tuber units were operated upon in the same consecutive order. In all cases, with the exception of Bliss Triumph inoculations with mosaic Irish Cobbler juices, the inoculated hills of the tuber unit treated first in each of the different varieties became diseased while the uninoculated hills of this unit remained healthy during the course of the experiment. Further- more, a number of mosaic tuber units, apparently infected in 191 8, were among the controls, or the units treated with juice from healthy plants. July I, 1920 Transmission of Mosaic Disease of Irish Potatoes 325 eu a 0 S ?■§ 000000 5;0 o 00 u 9. u o p _ 3d §3 "0-2 O O O O O O r^ •^'O -^ ' <=, H2 B-a H 3 a Si! 3>-|.o |a-o"-- Hi O ^■C ^f"' O-M TtO '^ ' •^0^<2 0 0'^J-0000 OOOO^'C^OOOO Tto '^O O O '^O o^ o OOOONC^OOOO SiNOOOCNOO" eg a >- P8 S'-S ■I H C. >- ti >- • B a id 9 '^ 1 >,cpwO£c £ U U U U I •5:5 0-2 otj ^ — -C — IJ— rt nl nj cs I OQ> CJ -WOOOO ' oH cu o*!: cH ocj : 3 .52 Si 3 .g ) H ^ O P3 « 326 Journal of Agricultural Research voi.xix, no. 7 In no case did any healthy units become infected even though they happened to be treated immediately after a diseased plant had been operated upon. This indicates that infection does not carry very readily from one plant to another by merely rubbing the leaves of one plant and subsequently practicing the same operation upon a neighboring plant. Late Application On July 12, 1 91 9, six healthy Green Mountain hills representing three dififerent tuber units were inoculated with juice from mosaic Irish Cob- bler vines. A second application was made upon these same plants a week later, when the vines were in blossom. On August 15 distinct mottling was in evidence on the upper leaves of the vines in each of the six treated hills, and by August 22 some of the leaves were dying in spots and streaks as in the bad stage of mosaic. Inoculations similar to the foregoing were made July 20 upon the vines of four hills in as many separate tuber units of the Irish Cobbler variety with juices from mosaic-dwarf Green Mountain vines. The plants at the time of the first inoculation had just finished blossoming. By August 20 slight mottling was noted upon the upper leaves of the inoculated vines and also slight streaking of the leaves as in bad mosaic stages. The results in these experiments indicate that plants can be inoculated successfully at the time of blossoming and later, as well as earlier in their development. Also, as stated previously in connection with insect transmission, even though mottling may not be in evidence in the season when infection occurs, nevertheless such plants will not fail to show distinct mottling under favorable environmental conditions during the following season. INSECT TRANSMISSION greenhouse; experiment with aphids Green Mountain tubers furnished by C. I, Gilbert were used at Orono with aphids in a greenhouse experiment because they were expected to be disease-free. This stock was used later in two plots. One consisted of 70 tuber units, of which only i was diseased early, evidently as the result of infection in 191 8. The other, grown and observed in southern Maine by Dr. W. J. Morse, consisted of 1,357 hiUs, of which less than three-fourths of i per cent were mosaic. In the greenhouse experiment 10 tubers were each cut lengthwise with a flamed knife into four sets and planted on March 17. Half the plants from each tuber were inclosed with insect cages, into each of which about 150 individuals of the common green peach aphis, or spinach aphis {Myzus persicae Sulz.) from mosaic potato plants were introduced on April 13 to 16, when the plants were from 2 to 9 inches high. To 15 plants aphids were introduced on leaves on a stick thrust into the soil so that they dispersed without contact between the diseased leaves and the treated plant. To 5 plants they juiyi, I920 Transmission of Mosaic Disease of Irish Potatoes 327 were introduced on terminal-shoot buds in a flask laid upon the soil. The aphids were killed by nicotine fumigation on April 21. All plants appeared healthy when observed by one of the writers on April 2 1 , when from 10 to 25 inches high. Between April 21 and June 2 ^ mosaic symptoms appeared on all of the 15 plants to which the aphids were introduced on sticks. Of the 5 plants to which the aphids were intro- duced in the flask only i became mottled, on July 8. When intro- duced in the flask many aphids had been injured or killed by water condensing on the interior of the flask following transpiration by the bud. Nineteen of the 20 untreated plants remained healthy; i showed slight symptoms on July 9. This plant was the only one found on or before April 28 with uncontrolled aphids upon it — possibly from a mosaic plant or a plant treated with virulent aphids. It was again found to be infested on May 19 and 26. In the case of the 15 plants treated with the stick method of introducing aphids, the percentage showing infection and the average length of the period between treat- ment and the appearance of the symptoms were greater than in the case of plants treated similarly in a previous greenhouse experiment,^ probably because in the later trial the plants elongated to heights of from 44 to 72 inches and thus offered for a longer period a chance for the initial display of mottling in the young leaves. FIELD EXPERIMENTS WITH CAGES EFFECT OF THE USE OF CAGES IN 1918 Although the cages for the control of insects in 1918 did not inhibit completely the dispersal of aphids, nevertheless their use materially checked transmission of mosaic. The effect of these cages upon transmission of mosaic is indicated in Table VI. Table VI. — Effect of cages on transmission of mosaic Variety. Green Motmtain Do Do Bliss Triumph. . Do Number Number of hiUs of tubers selected selected in for 1919 1918. planting. 9 32 3 6 22 50 31 66 20 54 Treatment in 1918. Uncaged Caged with mosaic hill Caged Uncaged ". . . . Caged Percent- age of mosaic 49 100 o 35 o The number of hills reported in Table VI includes only a small per- centage, a representative lot, of the total number planted in 1918. How- ever, each hill indicated was grown under a separate cage. While these • Observations after May i were made weekly by Viola L. Morris, laboratory assistant, and finally by Dr. W. J. Morse, neither having any information regarding the previous treatment of any plant. 'SCHtn[,Tz, E. S., FoLsoM, Donald, Hildebrandt, F. M., and Hawkins, Lon A. op. err. 177285°— 20 4 328 Journal of Agricultural Research voi.xix.No.7 results might be interpreted as suggesting that some insect besides aphids was a deciding factor, it is possible for the aphids observed in the cages late in the 191 8 season to have come from a very few which did not carry mosaic, and as yet no other insect is known to transmit mosaic of potato. EXPERIMENTS WITH APHIDS Small colonies of the peach aphis were brought from Orono May i on radish and mosaic potato plants. Both increased while feeding on these plants. On June 7, when the vines were from i to 4 inches high, 9 caged plants of 3 tuber units were treated with aphids from radish plants, about 150 to each hill. These 3 units were regarded as controls, since the aphids had lived for a number of generations on radish plants and were supposed to be free from a mosaic virus. Three caged plants of a fourth tuber unit were treated with aphids from a mosaic potato plant; 2 plants were left in each hill and the aphids, about 100 to a hill, were introduced on leaves on a stick thrust into the soil near each hill. On June 30 and July 5, when aphids were very numerous, these 12 plants were sprayed with a solution of soap and nicotine sulphate. The plants used came from the Gilbert stock, already described as exceptionally healthy. On July 28 the fourth tuber unit was slightly mosaic in some branches of i hill, and by August 9 it was dead, as the result of excessive aphid infestation. The 3 controls remained healthy until dug on August 26. On June 17, nine half-tuber sets from stock caged in 191 8 were planted under three cages. On June 28, when the vines were from i to 3 inches high, the plants were treated with aphids from mosaic plants; several hundred aphids were introduced by each hill with the stick method de- scribed above. They were sprayed on July 5 and 8. On August 9 one hill showed some mosaic. When dug on August 26, this hill was all mosaic, while two other hills — one in the same cage — were each mosaic in the upper leaves of one stalk. The untreated plants from the other nine half tubers were grown in the field and remained healthy through- out the season. Four tuber units of the Gilbert stock, comprising 12 hills, were treated on July 12, when the plants were large enough to press against the tops of the cages. The first unit was treated with hundreds of aphids from radish plants and the others with aphids from mosaic potato plants. In the latter case several thousand aphids were left on the diseased leaves and stems in a flower-pot saucer set at the base of each of the first and third hills, whence they dispersed within a few days. The first tuber unit remained healthy throughout the season. The other three were still healthy on August 9, but when dug on August 26 two hills were mosaic, each in the upper leaves of one branch of a stalk. juiyi, I920 Transmission of Mosaic Disease of Irish Potatoes 329 Experiment with flea beetles Three caged tuber units (9 hills) of the Gilbert stock were treated with flea beetles {Epitrix cucumeris Harris) on June 13, 191 9, when a few inches high. The middle hill of each unit was covered with a cylindrical cage set inside the larger cubical one; the other two hills were treated with several hundred flea beetles. These insects were collected from small potato vines which developed from 100 per cent mosaic stock. On June 20 the cylindrical cages were removed and most of the flea beetles, which had damaged the plants considerably, were driven out of . the cages or killed by hand. On June 16 two more similar tuber units were treated likewise. All the hills remained healthy until dug on August 27. As controls, four similar tuber units were treated in the same way, except that the beetles were taken from plots of mostly healthy potatoes or, in one unit, from bushes near the potato field. All the hills remained healthy until dug on August 26. EXPERIMENT WITH COLORADO POTATO BEETLES Five caged tuber units (15 plants) of the Gilbert stock were treated with Colorado potato beetles (Lepiinotarsa decemlineata Say.) on July 3, 1 91 9, when they reached nearly to the tops of the cages. The in- sects were gathered with brush and pan from plants in all-mosaic plots when from 2 days old to two-thirds full grown. Two stalks were left in a hill, and the first and third hills in every cage were treated with over 100 of the larvae each. These were shaken from the gathering pan upon a cloth and were either rolled upon the leaves or left on the cloth while it was laid on the plant. Within 24 hours the plants had been damaged rather severely. They were sprayed with an arsenical poison, which soon caused the death of the larvae. All the plants re- mained healthy until dug on August 26 and 27. Three similar tuber units were treated likewise on July 7, except that the larvae were obtained from plants in plots almost disease-free. These also remained healthy until dug on August 27. FIELD OBSERVATIONS WITHOUT CAGES GREENHOUSE STOCKS Tubers from the 53 plants used in the first aphid experiment performed in the greenhouse at Orono^ were planted whole. All of the 37 tubers from plants which became mosaic after the introduction of aphids from mosaic potatoes produced diseased hills, except 2 which came from a plant with 3 out of 7 stalks apparently healthy. The 2 healthy tubers were probably produced by the 3 healthy stalks. All of the 10 tubers from plants which remained apparently healthy until har- vested, although they were fed upon by aphids from mosaic plants, were 'ScHUiTz, H. S., Foi^OM, Donald, Hilderbrandt, F. M., and Hawkins, Lon A. op. aT. p. 35-30. 330 Journal of Agricultural Research voi. xix, no. 7 mosaic. None of the 38 tubers from caged untreated plants or of the 15 from plants fed upon by aphids from a healthy potato plant were mosaic. Of the 34 tubers from uncaged and untreated plants i was mosaic; it came from a half -tuber hill that early showed ruffling and chlorosis along the veins but no typical mosaic mottling such as was shown, in addition to these incomplete symptoms, by the correspond- ing half-tuber hill after treatment with virulent aphids. Of the 2>7 tubers from plants fed upon by aphids from radish plants 4, or 11 per cent, were mosaic; these 4 came from 2 plants recorded as having been fumigated to eliminate a few aphids which were found on them and which were of unknown origin, possibly from neighboring diseased plants. These results agree essentially with those which were secured previously with the first generation of the same stocks and which were described to prove the possibility of transmission by aphids. They also indicate that (i) mosaic mottling may be restricted to the parts of the leaf along the veins, (2) a plant with three stalks healthy and four mosaic may produce three mosaic tubers and two healthy ones, thus explaining the partial infection of hill lots, (3) plants treated with virulent aphids may appear healthy but produce progeny that are all mosaic, as shown previously by the writers,^ and (4) apparently healthy plants inspected often for aphids and fumigated to eliminate these insects as soon as they are dis- covered may produce progeny of which a small percentage is mosaic. In connection with the experiment just considered it was necessary to treat a number of control plants by laying a mosaic leaf upon each. These were kept in a dififerent greenhouse room where aphids were more abundant, and they were never caged. Of 45 tubers from these, and also of 25 tubers from similar plants with no leaf laid on, 20 per cent were mosaic, all coming from plants recorded as being fumigated to eliminate uncontrolled aphids found upon them. PROXIMITY STUDIES WITH PLOTS In 1 91 8, plots I, 2, and 3 were each rogued of mosaic hills three times. Stocks from the first two, Green Mountain and Bliss Triumph, respec- tively, each showed mosaic in 20 per cent of the hills in 191 9, while that from No. 3, Green Mountain, next to No. 4, a Green Mountain plot with 45 per cent of the hills diseased, showed mosaic in 30 per cent of the hills. In 1 91 9, each of the stocks was rogued several times and grew between similar stocks. All these plots in both years were each X acre in area. The greater percentage of mosaic in 191 9 in stock from plot 3 can be explained best by the greater proximity in 191 8 to a half-mosaic plot and by consideration of the apparently greater ease of dispersal of aphids, which were numerous in 191 8, from the half -mosaic plot to No. 3. Plots I, 2, and 3 were planted with stocks A, B, and C, respectively, described in Table VII. 1 ScHUiTZ, E. S., Foi,soM, Donald, Hildebrandt, F. M., and Hawkins, Lon A. op. ai. July 1, 1920 Transmission of Mosaic Disease of Irish Potatoes 331 INTERSEASONAL INCREASE It was very apparent that aphids, which seemed as abundant as the flakes in an ordinary snowstorm when they were migrating in the late summer, were unusually numerous in 191 8. Consequently it is of interest to compare the relative interseasonal increase in mosaic in the same stocks from 1917 to 1918 with that from 191 8 to 191 9. It has been demonstrated in several greenhouse experiments already discussed that aphids may transmit mosaic without the symptoms being shown until the progeny of the inoculated plants is grown the following season. It is considered that they may do likewise in the field during the latter half of the summer, which is usually the only time when they are abundant on potatoes in northern Maine, although there are more species in the field than were used in the various experiments. Previous experiments seemed to indicate that the percentage of mosaic in susceptible varieties could be materially reduced by roguing the diseased plants from the plots as soon as the mottling appeared upon the vines. However, before it was fully demonstrated that insects were capable of transmitting mosaic the plots usually were arranged in such a manner that insect transfer could take place very readily. In view of this situation, it is possible to note the effects of those agents of transmission upon the performance of a few of the plots, each including X acre, which were rogued during the last three seasons. Table VII records the observations on these plots. Table VII. — Relation of aphids to increase of mosaic from season to season Variety. Green Mountain . . . Bliss Triumph Green Motmtain . . . Stock. Location. Next to 100 per cent mosaic stock. do do Per- cent- age of mo- saic. Treatmpnt. Rogued three times, do Rogued once Niunber of aphids. Few. Do. Do. Variety. Stock. 1918. Location. Per- cent- age of mo- saic. Treatment. Number of aphids. Per- cent- age of mo- saic. Green Mountain . . I A Bliss Triumph I B Green Mountain . . ' C Six rows from 45 per cent mosaic stock. Nine rows from 45 per cent mosaic stock. Next to 4S per cent mosaic stock. Rogued three times. ....do do Very abundant do do 332 Journal of Agricultural Research voi. xix, no. 7 It will be noted in Table VII that in 191 7 certain factors seemed to be more favorable for the spread of mosaic than in 191 8 — namely, higher percentage of diseased hills (rogued) in the plots, greater proximity of unrogued mosaic stock, and higher percentage of mosaic in the nearest unrogued diseased plot. However, there was less spread in 191 7 than in 1 91 8, as shown by the lower percentage of mosaic in 191 8 than in 191 9, in correlation with the greater abundance of aphids in 191 8. Fur- thermore, these observations indicate how difficult the problem is of pro- ducing perfectly mosaic-free stocks from susceptible varieties wherever these agents of transmission exist. EFFECT OF VARIATION IN THE TIME OF HARVESTING IN I918 It was expected that if aphids were a deciding factor in mosaic trans- mission the lots of tubers harvested at progressively later dates during their increase in numbers would show an increasing percentage of mosaic. Seventy-eight healthy hills (66 Green Mountain and the rest Bliss Tri- umph or Irish Cobbler) were selected in 191 8 in a plot containing many small lots all with more or less mosaic. Aphids became noticeable on potatoes the last part of July and increased in numbers so that they were very numerous about the middle of August and more excessively abundant as the end of the month was approached. Tubers about an inch in diameter were harvested on August 8 but did not keep with the methods used. Another set of tubers was harvested on August 15 and a third on August 26, one tuber being removed from every hill on each date. On September 12 the remaining tubers — 321 in all — were har- vested. The tubers were planted uncut in 191 9 and transmitted 6, 14, and 50 per cent of mosaic, respectively, for the three lots. Ap- parently some of the infection occurred before August 15, but most of it was too late to affect many of the tubers harvested by August 26. This difference can be explained best by the great increase of aphids during August, together with the results obtained in the experiments on aphid transmission. TEST OF THE SEED-CUTTING KNIFE In 1 91 9 stock was available from 191 8 all-mosaic plots and rogued plots. One hundred tubers from the former were divided by three paral- lel transverse cuts so that no two cut surfaces joined in a seed piece, while 100 tubers from the latter were quartered by a transverse and a longitudinal cut so that each seed piece had two cut surfaces joining at a right angle. The same knife w^as used, cutting alternately tubers from the two lots. The 800 sets were left mixed in the same sack for over a day and planted by hand at 15-inch intervals in two rows. Another mixture was prepared in the same way with 200 tubers from the same two barrels, but in this case the pieces from the all-mosaic lot were sorted July 1, 1920 Transmission of Mosaic Disease of Irish Potatoes 333 out and discarded and only the others were planted. The latter occupied the third row, and the fourth row was used for a control lot prepared similarly except that no all-mosaic stock was used. Upon examination of the four rows on July 23 the control row was found to contain 85 mosaic hills, the third row 72, and the first two 475 — that is, 75 excluding the 400 from all-mosaic stock. No change in the number of mosaic hills was found on August 18. A X-acre plot of the rogued stock was planted elsewhere and contained 80 mosaic hills in each 400. Evidently the fur- nishing of conditions apparently optimum for knife transmission had no effect upon the mosaic percentage. It was thought in 191 8 that the partial infection of tuber units might be due to knife traosmission. As stated before (p. 318), in 191 9 when tuber units were planted three knives were used in rotation, each one being immersed in a 4 per cent formaldehyde solution when not in use. However, the partial infection of tuber units and hill lots was as common as before. TESTS OF EFFECTS OF CONTACT GREENHOUSE EXPERIMENTS As has been reported,^ out of nine healthy plants kept in contact with mosaic plants in a greenhouse one showed mosaic, but not until after a few uncontrolled aphids, possibly from mosaic plants, were discovered upon it. At about the same time, March 13, 191 9, each of 12 tubers was split into three sets and planted in small pots. The plants from 4 tubers became mottled by April i when from 3 to 13 inches tall. The other 24 were transplanted about April i into large pots, 2 from each tuber into steam-sterilized soil and the third into soil containing a mosaic plant. The transfer was made by knocking off the bottom of the small pot and setting it into a hole formed by a small empty pot put in when the mosaic set was planted. The method used permitted the mingling of the roots of the two plants while it kept the two sets of tubers mostly apart and facilitated harvesting them separately. The vines of the two plants were twisted and tied together. All of the 24 plants remained healthy until July 9. ^ They had ceased to elongate by this time and soon afterwards were dug. The tubers were not planted, because of the abundance of aphids on the plants in July. FIELD EXPERIMENTS WITH INSECT CAGES Nine tubers of the Gilbert stock were planted halved in 191 9, each two sets being separated by a mosaic set and all three caged. On July 30 three of the mosaic hills were dead or nearly so. The Gilbert hills all remained healthy until August 9 and when dug on August 27 ' ScHULTz, E. S., FoisoM, Donald, Hildebr,\ndt, F. M., and Hawkins, Lon A. op. ot. 2 Observations after May i were made weekly by Viola L. Morris, Laboratory Assistant, and finally by Dr. W. J. Morse. 334 Journal of Agricultural Research voi. xjx, No. 7 were entirely healthy except for mosaic mottling in the few uppermost leaves of several branches of a stalk in one hill. These leaves appeared young. They had evidently been pushed hard against the inside of the cage and had a very few aphid skins and aphids clinging to them. They may have been infected as the result of contact before aphids entered the cage, by aphids on the outside of the cloth against which the leaves were pressed, or by aphids that came from mosaic plants in the next row and that entered through a small hole that was found to have been made accidentally in the cloth. FIELD EXPERIMENTS WITHOUT CAGES As was pointed out in a previous section regarding the test of the seed-cutting knife, the mixing of all-mosaic stock and rogued stock in two rows was not followed by a higher mosaic percentage for the rogued stock than was shown by it in a control row. The negative results in this case do not disprove the possibility of infection occurring too late to be evident during the current season — that is, after the roots and vines have become intertwined. In 1 91 8 five Green Mountain hill lots were found to be partly mosaic. The healthy hills were harvested separately, were classified according to their proximity in the row to a mosaic hill, and the tubers were planted uncut in 191 9. Twenty-eight tubers were progeny of plants each of which grew between two mosaic hills, and 54 per cent of them were mosaic. Eighty-nine were progeny of hills each of which was be- tween a mosaic hill and a healthy hill, and of these 63 per cent were mosaic. On the other hand, 40 per cent of the 220 tubers from hills each of which grew between two healthy hills were diseased. If these 2.20 tubers are arranged in five groups. Hi, H2, H3, H4, and H5, according to the increasing number of healthy hills between the parent and the nearest mosaic plant in the row, the groups contained, respec- tively, 75, 53, 41, 33, and 18 tubers, with 56, 24, 54, 24, and 17 per cent of them diseased. Since being next to a mosaic plant in the same row seemed to increase the chance of infection as much as 54 or 63 per cent is greater than 40 per cent, it evidently is a contributing factor in mosaic transmission; but judging from the varying percentages of infection among the classes of plants which were not next to mosaic hills in the same row, it probably aids in the spread of the disease only by aiding aphid transmission. A slightly different type of experiment consisted in comparing the progeny of three small i-row Green Mountain lots, of from 100 to 200 hills each, from which the mosaic hills (respectively, 6, 16, and 30 per cent) were removed on August i, 191 8, with two similar lots from which the mosaic hills (respectively, 6 and 18 per cent), together with each healthy hill next in the row to a mosaic one, were removed July 1, 1920 Transmission of Mosaic Disease of Irish Potatoes 335 August I. In spite of the differences in contact with diseased hills, the progeny of the two lots were 27 and 35 per cent mosaic, respectively, and the progeny of the three lots were from 25 to 35 per cent mosaic. Aphid dispersal from neighboring mosaic plots was easy, and it appar- ently nullified any effect that the difference in contact might have had. TEST OF SOIL HARBORING GREENHOUSE EXPERIMENT At harvesting time in 191 8 one tuber was taken from each healthy hill in two hill-selected lots. At Orono on January 14, 191 9, these tubers were split with a flamed knife, and one set was planted in steam-sterilized soil and the other in soil from which a mosaic plant had been removed on December 30 or January 13. Nineteen pairs of half tubers were used, and the plants from 7 pairs were mosaic by February 22, when from i to 20 inches high. The plants from the other 12 pairs reached their maxi- mum height about March 5 and remained healthy until dug in April. The second generation of the 12 healthy pairs was grown and found to be entirely free from mosaic. It is clear that there was no transmission by the soil in which mosaic plants had just been grown, all mosaic that was shown evidently being transmitted by the tubers. FIELD EXPERIMENTS The greenhouse experiment described in the previous paragraph was not concerned with certain factors in the possible soil-harboring of mosaic in fields — namely, old stalks, volunteer potato plants, and insects. There is no doubt, when the proofs of transmission by aphids are remembered, but that volunteer mosaic plants may contribute to the infection of healthy stocks planted where mosaic stocks were grown the preceding season if they are not discovered and removed before the appearance of aphids. Even if they are, other factors might cause the infection of healthy plants. To test this supposition, three rows of Green Mountain stock from a plot rogued in 191 8 were planted across the location of a 191 8 20 per cent diseased Green Mountain plot and a wholly diseased one. Each mosaic hill was dug and the seed piece examined. If volunteers are disregarded, 28 per cent of the 142 hills grown upon the ground of the all-diseased plot were mosaic as were 28 per cent of the 481 plants grown upon the ground which had produced the 20 per cent mosaic plots. This evidently was from infection the previous season, since 27 per cent of the hills were mosaic by July 15. A similar but more extensive test consisted in planting 19 rows of the same stock across the ground which had produced 14 of the 191 8 plots. Similar examination of the mosaic plants on July 30 showed 22 per cent 336 Journal of Agricultural Research Vol. XIX, No. 7 of the 4,466 hills to be mosaic. Although this stock was retarded in its development by being frozen nearly to the ground on June 23, only i per cent of the hills developed mosaic between July 30 and August 18. The nature of the various 191 8 plots and the percentages of mosaic on the same ground in 191 9 are given in Table VIII. Table VIII. — Nature of igi8 plots and percentage of mosaic hills in the parts of the igig plot grown upon the same ground Sec- tion No. Variety. Green Mountain. Bliss Triumph . . Green Moimtain. ....do do Roxbury Wilson. Bliss Triumph . . Green Mountain. do Irish Cobbler. . . . do do Miscellaneous. .. do II per cent mosaic. . :55 per cent mosaic. . 13 per cent mosaic. . 45 per cent mosaic. . 46 per cent mosaic. . 10 per cent mosaic. . 100 per cent mosaic . do 1 1 per cent mosaic . . No leafroU All leafroll . No leafroll Leafroll and mosaic. do Total number of hills. 424 432 454 422 375 281 350 458 140 143 105 169 140 573 Percent- age of mosaic hills from seed pieces. 24 23 22 26 18 23 23 22 28 "22 22 15 23 24 It will be noted that there are few marked deviations from the per- centage for the whole plot, which was 23 per cent. These consist of one deviation upward and one downward for the ground occupied by two half- mosaic plots (4 and 5) and of the same for two comparatively mosaic- free plots (9 and 12) and therefore are without significance in regard to soil-harboring of the disease. SUMMARY (i) Transmission of potato mosaic by means of tubers, grafting, plant juice, and aphids was eflfected under various conditions, including those essentially of the field with insects controlled. (2) Infection was obtained with intervarietal transfer of juice. (3) Transmission was attempted, but without success so far as could be ascertained in the same season, by means of flea beetles, Colorado potato beetles, the seed-cutting knife, and contact of seed pieces, of roots, and of vines. (4) Preliminary observations indicate that infection does not result from growth in soil that produced mosaic potato plants the previous season. July 1. 1920 Transmission of Mosaic Disease of Irish Potatoes 337 (5) It appears impossible either for infected plants to recover or, so long as diseased stock is not far off and insect carriers exist, to assure the maintenance of health of susceptible varieties by roguing plots or by selecting hills, tubers, or seed pieces. (6) Isolation of plants by means of insect cages, as well as elimination of insects in the greenhouse, have maintained stocks disease-free, indi- cating that control of aphids and possibly of some other kinds of insects as well, is the most important means of checking the spread of potato mosaic among susceptible varieties. PLATE 49 Vines of Green Mountain variety inoculated with juice from healthy foliage of the same variety. No mottling and no ruffling of leaves. (33S) Transmission of Mosaic Disease of Irish Potatoes Plate 49 Journal of Agricultural Research Vol. XIX, No. 7 Transmission of Mosaic Disease of irish Potatoes Plate 50 Journal of Agricultural Research Vol. XIX, No. 7 PLATE 50 Vines of Bliss Triumph, variety inoculated with juice from healthy foliage of Irish Cobbler variety. No mosaic mottling. PLATE 51 Vines of Irish Cobbler variety inoculated with juice from healthy foliage of the same variety. No mosaic. Transmission of Mosaic Disease of Irish Potatoes Plate 51 Journal of Agricultural Research Vol. XIX, No. 7 Transmission of Mosaic Disease of Irish Potatoes Plate 52 Journal of Agricultural Research Vol. XIX, No. 7 PLATE 52 Vines of Green Motintain variety inoculated with juice from mosaic foliage of the same variety. Distinct mosaic mottling and ruffling of young leaves on top of stalks. For control see Plate 49. PLAT€ 53 Vines of Green Mountain variety inoculated with juice from mosaic foliage of Bliss Triumph variety. Distinct mottling and ruffling of upper leaves and early dying of lower leaves. Condition of control plants same as vines in Plate 49. Transmission of Mosaic Disease of Irish Potatoes Plate 53 Journal of Agricultural Research Vol. XIX, No. 7 Transmission of IVlosaic Disease of Irish Potatoes Plate 54 Journal of Agricultural Research Vol. XIX, No. 7 PLATE 54 Vines of Green Mountain variety inoculated with juice from mosaic foliage of Irish Cobbler variety. Distinct mottling and ruffling of upper leaves, early dying of lower leaves. Condition of control plants same as vines in Plate 49. 177285°— 20 5 PLATE 55 Vines of Bliss Triumph variety inoculated with juice from mosaic foliage of Irish Cobbler variety. Mosaic mottling of upper leaves, early dying of lower leaves. For control see Plate 50. Transmission of Mosaic Disease of Irish Potatoes Plate 55 Journal of Agricultural Research Vol. XIX. No. 7 Transmission of IVIosaic Disease of Irish Potatoes PLATE 56 Journal of Agricultural Research Vol. XIX. No. 7 PLATE 56 Vines of Irish Cobbler variety inoculated with juice from mosaic foliage of the same variety showing bad stage of mosaic. Distinct mottling of yotmg leaves and early dying of foliage. For control see Plate 51. ADDITIONAL COPIES OF THIS PUBLICATION MAY BE PROCURED FROM THE SUPERINTENDENT OF DOCUMENTS GOVERNMENT PRINTING OFFICE WASHINGTON, D. C. AT 25 CENTS PER COPY Subscription Price, $1.50 Pee Year A Vol. XIX JUL,Y 15, 1920 No. 8 JOURNAL OF AGRICULTURAL RESEARCH CONTKNXS Page Relative Susceptibility to Citrus-Canker of Different Spe- cies and Hybrids of the Genus Citrus, Including the Wild Relatives --------339 GEORGE L. PELTIER and WILLIAM J. FREDERICH (Contribution from Alabama Agricultural Experiment Station and Bureau of Plant Industry) Presoak Method of Seed Treatment: A Means of Pre- venting Seed Injury Due to Chemical Disinfectants and of Increasing Germicidal Efficiency - - - - S63 HARRY BRAUN (Contribution from Bureau of Plant Industry) PUBLISHED BY AUTHORITY OF THE SECRETARY OF AGRICULTURBl WITH THE COOPERATION OF THE ASSOCIATION OF LAND-GRANT COLLEGES WASHINGTON, D. C. WAtHINQTON : QOVERNMENT MUNTINO OmOE : 1 110 EDITORIAL COMMITTEE OF THE UNITED STATES DEPARTMENT OF AGRICULTURE AND THE ASSOCIATION OF LAND-GRANT COLLEGES FOR THE DEPARTMENT KARL F. KELLERMAN, Chairman Pkysiqlogist and Associate Chief, Bureau of Plant Industry EDWIN W. ALLEN Chief, Office of Experiment Stations CHARLES L. MARLATT Entomologist and Assistant Chief, Bureau of Entomology FOR THE ASSOCIATION J. G. LIPMAN Dean, State College of Agriculture; and Director, New Jersey Agricultural Experi- ment Station, Rutgers College W. A. RILEY Entomologist and Chief, Division of Ento- mology and Economic Zoology, Agricul' tural Experiment Station of the University of Minnesota R. L. WATTS, Dean, School of Agriculture, and Director, Agricultural Experiment Station, The Pennsylvania State College All correspondence regarding articles from the Department of Agriculture should be addressed to Karl F. Kellerman, Journal of Agricultural Research, Washington, D. C. All correspondence regarding articles from State Experiment Stations should be addressed to J. G. Lipman, New Jersey Agricultural Experiment Station, New Brunswick, N. J. BOTANICAL, GAi^DEri JOIM£ OF AGMmiAL ISEARCH Vol. XIX Washington, D. C, July 15, 1920 No. 8 RELATIVE SUSCEPTIBILITY TO CITRUS-CANKER OF DIFFERENT SPECIES AND HYBRIDS OF THE GENUS CITRUS, INCLUDING THE Vi^ILD RELATIVES ' By George L. Peltier, Plant Pathologist, Alabama Agricultural Experiment Station, and Agent, Bureau of Plant Indiistry, United States Department of Agriculture, and William J. Frederick, Assistant Pathologist, Bureau of Plant Industry, United States Department of Agriculture ' INTRODUCTION In a preliminary report {6y the senior author briefly described the results obtained under greenhouse conditions for a period of six months on the susceptibility and resistance to citrus-canker of a number of plants including some of the wild relatives, Citrus fruits, and hybrids of the genus Citrus. Since that time the plants reported on have been under close observation; a third experiment has been started, and many inoculations have been made in the isolation field in southern Alabama during the summers of 1917, 1918, and 1919. Many more plants have been successfully inoculated; others have proved to be extremely sus- ceptible; while some of those tested still show considerable resistance. The results obtained up to November i, 191 9, are described in this report. EXPERIMENTAL METHODS In the greenhouse, the methods used and the conditions governing the inoculations described in the preliminary report were closely fol- lowed. The same strain of the organism was used and was applied in the ' Published with the approval of the Director of the Alabama Agricultural Experiment Station. The paper is based upon cooperative investigations between the Office of Crop Physiology and Breeding Investi- gations, Bureau of Plant Industry, United States Department of Agriculture, and the Department of Plant Pathology, Alabama Agricultural Experiment Station. ' The writers wish to acknowledge their indebtedness to Dr. K. F. Kellerman, Associate Chief, Bureau of Plant Industry, United States Department of Agriculture; to Mr. W. T. Swingle, Physiologist in Charge, Crop Physiology and Breeding Investigations; to Mr. T. R. Robinson, Physiologist of the same office, for their cooperation and many helpful suggestions; and to Dr. O. F. E. Winberg, of the Alabama State Board of Horticulture, for valuable assistance and suggestions in the field work. ' Reference is made by number (italic) to "Literature cited," pp. 361-362. ]pumal of Agricultural Research, Vol. XIX. No. 8 Washington, D. C. July 15, 1920 Uo Key No. Ala.-6 (339) 340 Journal of Agricultural Research voi. xix.no. s same manner — that is, infusions of 48-hour-old cultures of Psendomonas citri Hasse in beef bouillon were sprayed on the foliage of the plants by means of an atomizer. In no cases were punctures made, but wounds and scratches were present on some of the leaves. The plants all received identical treatments and were under approximately the same conditions. The plants in experiment I were inoculated August 27 and September 12, 1 91 7, and those in experiment II were inoculated October 23 and November 9. On March 23, 1918, the plants were all trimmed or cut back to force new growth, and some new numbers were placed in the screen cases together with the others. All of these were then inoculated in the usual manner. On the same date, the plants in experiment III were inoculated and have remained in the glass cases. Thus, after initial infection was obtained, natural inoculations of the remainder of the plants were counted on entirely. Therefore, natural infection took place on the greater part of the plants reported on in the following pages, especially on the more resistant plants. In the isolation field, plantings were made in 1917, 1918, and 191 9. Some of the nonhardy relatives and Citrus fruits were killed by hard winters of 191 7-1 8 and 191 8-19, but the majority of the plants sur- vived. Inoculations in the field were started in September, 191 7, by Mr. D. C. Neal and were later continued by Mr. J. Matz up to November I, 1918, when the plants were banked for the winter. During the 1919 season the inoculations and observations in the field were made by the junior author.^ Some of the plants were inoculated only once or twice, while others were sprayed with the inoculum regularly every week throughout the season. The time chosen for inoculation varied, but as a rule the late afternoon was chosen. A large number of natural infections took place after canker had been established on some of the plants. The seasons of 191 7 and 191 8 were normal so far as climatic conditions were concerned. However, during the 191 9 season the temperature was high, and together v.dth an excessive and frequent rainfall it afforded not only ideal conditions for plant growth but also for the most rapid infection and development of canker. Unless otherwise stated, all the plants reported on were in a good growing condition, and the organism was reisolated from the doubtful canker spots, especially in the case of the wild relatives. It must also be borne in mind that tlie plants used were for the most part small seedlings or nursery stock. Thus, the size of the plants and the condi- tions under which the inoculations were carried out made them more susceptible to canker. Plants reported here as susceptible would proba- bly show more resistance under orchard conditions. No doubt maxi- mum susceptibility was obtained with the plants experimented upon. ' The senior author is solely responsible for all conclusions drawn from the results of the three years' work. July 15, 1920 Relative Susceptibility to Citrus-Canker 341 SUSCEPTIBILITY OF NONRUTACEOUS PLANTS Melia azedarach L., China berry, in field, 1919. Inoculations were attempted on this plant in the field for the reason that it is the native host of the Citrus white fly. Needle prick and spray inoculations in the field under the most favorable conditions for the development of canker were negative. Lee and Merrill (5) have reported successful inoculations of the stem and petioles of Lansium domesticum with Pseudomonas citri. This plant belongs to the same family as the China berry. SUSCEPTIBILITY OF WILD RELATIVES OF THE GENUS CITRUS RUTACEOUS PLANTS NOT CLOSELY RELATED TO THE GENUS CITRUS Xanthoxylum sp. (CPB 11269, seedling), III.' So far no canker spots have been found on the plants. Casimiroa edulis Lav. and Lex. White sapote (CPB 7923, seedlings), I, 11,2 -^ field, 1919.^ At the October, 1918, readings a few nontypical spots were observed on several of the young leaves of the plant in experiment I. They occurred only at the woimds and scratches, no spots being found on the xmbroken surface of the leaves. New spots have appeared from time to time, but in all cases they have occurred at wounds and remained unruptured. The spots (PI. 57) are small (about 0.5 mm. in diameter), light colored, slightly raised, compact, and unruptiu-ed. They do not have an oily outline. No yellow zone is present. No positive infections were obtained in the field. As there are three varieties of the white sapote being grown in California and Florida for its fruit, it is of interest to note that it can be successfully inoculated imder green- house conditions, although it does show considerable resistance to citrus-canker. Glycosmis pentaphylla DC. (CPB 2905, seedlings), II, 111,2 in field, 1918. This is one of the few relatives tested which has, so far, remained immime to canker in both the greenhouse and field. Claucena lansium Skeels. Wampi (CPB 7936, seedlings), 1, 11,2 i„ fi^jfj^ j^^^ ^^^ ^^^g^ A few small, nontypical, oily spots appeared on the leaves of the plants in both experiments. The spots are typical of those found on the wild relatives. Repeated inoculations in the field gave negative results. Chalcas exotica Millsp. {Murraea exotica L.)- Orange jessamine (CPB 797SA, seedlings), I, II, in field, 1917 and 1918. Diuing July, 1918, a few nontypical spots were observed on the young leaves of the plants in experiment II. The spots resemble those on Casimiroa edulis in general appearance, except that they are somewhat larger and of a more oily character. A few new spots have devel- oped since that time. The plants are only very weakly positive, and the period of incubation is rather long. The spots in all cases were at wounds and imruptured. In no case were positive results obtained in the field in spite of repeated inoculations. RUTACEOUS PLANTS BELONGING TO TRIBE CITREAE SuBTKiBE Aegunau (Hardshell Fruits) Aegle marmelos Correa. Bael fruit (CPB 7983, seedlings), ly II, in field, 1918. Diu-ing the summer months these plants made a splendid growth and produced an abundance of new foliage. At the July, 1918, reading several small spots typical of ' Roman numerals refer to the number of the inoculation experiment in the greenhouse. ' Included in experiments of March 21, 1918. ' Date of planting in the isolation field. 342 Journal of Agricultural Research voi. xix, No. g those found on Casimiroa edulis were observed, occurring at wounds and scratches. Spots have appeared occasionally since that time, but in every case they were observed along scratches and wounds and remained unruptxu-ed. All inoculations in the field were negative even at wounds. Aeglopsis Chevalieri Swingle. (CPB 7633 and 7772, seedlings and cuttings), II and I,^ in field, 1918. The plants, although producing an abundance of new growth, have remained free from canker, in both the greenhouse and field. Chaetospermum glutinosum (Blanco) Swingle {Limonia glutinosa, Blanco), Tabog (CPB 7799 and 7814, seedlings), I, II and I,^ 11,^ III, in field, 1917 and 1918. An abundance of new foliage was produced by all the plants, and thus they have been in an excellent condition for infection. All five of the plants have developed canker. The spots first appeared in April, 1918, three weeks after the last inoculation. The spots were at first small and nontypical (PI. 58 A), but as they increased in num- bers they became more and more typical. At the last reading the percentage of in- fected leaves ranged from 10 to 30, and from a few small, oily, unruptured spots (PI. 58, B) to many medium-sized, ruptured spots (Pi. 58, C). No spots have been found on the twigs. The small , unruptured spots generally appeared at wounds or scratches and resemble those described for Casimiroa edulis. The more normal spots are medium-sized, of a brick color, almost flat, compact, and slightly corky. They do not break through the upper surface but appear as a flat, discolored spot. The oily outline is very indis- tinct around the unbroken spots, and the yellow zone is absent. Vigorous colonies of Pseudomonas citri were isolated from these ruptured spots. Unfortunately these plants are very susceptible to low temperatures and have been killed in the field each season, so that no thorough test of their susceptibility has been made. However, judging from the susceptibility shown in the greenhouse, they should be successfully inoculated in the field under favorable conditions. No posi- tive results have been obtained so far, although the plants were repeatedly inoculated during the summers of 1917 and 1918. Balsamocitrus Dawei Stapf. (CPB 2920, on Aeglopsis Chevalieri), III, in field, 1917 and 1918. This is a large tree found in the forests of east central Africa at an altitude of 2,000 to 3,000 feet. The plant, although making a rapid growth, has remained free from canker in the field and the greenhouse. SUBTRIBE FeRONINAE (HaRDSHEI,!. FrUITS) Feronia limonia (Corr.) Swingle (F. elephantum Corr.). Wood-apple (CPB 2763, seedlings), I, II, in field, 1917 and 1918. A few oily, unruptured spots were observed at the July, 1918, reading. The spots have become more numerous and are scattered over the new foliage, especially at wounds, but remain small and unruptured. They are typical in all respects to those described for Casimiroa edulis. No positive results have been obtained in the field. Peromelia lucida Swingle. Kavista Batu (CPB 7882, seedlings), I, II, in field, 1917 and 1918. Some small, very slow-growing, oily spots are scattered over the new foliage. They resemble in all respects those found on Feronia limonia. Repeated inoculations in the field have been negative. J Included in experiments of March 21, 1918. July 15, 1920 Relative Susceptibility to Citrus-Canker 343 SUBTWBE LaVANGINAE Hesperthusa crenulata Roem. Naibel (CPB 2759, seedlings), II, III, in field, 1917 and igi8. Because the spots produced were so nontypical the susceptibility of these plants was doubted until cankers developed on the twigs and branches. The spots (PI. 59, B) are small and nontypical, although 90 per cent of the new leaves on the plants are infected. On the twigs they are rather numerous, flat, very oily, and apparently ruptured. In the field the few spots formed on the twigs have remained unruptured, very oily, and slightly raised. On the leaves the spots are nontypical, few, and in some cases slightly ruptured. Triphasia trifolia P. Wilson. Lime berry (CPB 2689A and 7780, seedlings), I, II, and II, in field, 1918. The plants have remained free from canker in both the field and greenhouse. Severinia buxifolia Ten. (CPB 2760, cuttings and seedlings), I, II, in field, 1917, 1918, and 1919. Like Triphasia, the plants are apparently immune. SUBTRIBE CiTRINAE Citropsis Schweinfurthii Swingle and M. Kellerman. African cherry orange (CPB 11260, seedlings), I,' II, infield, 1917 and 1918. Several small spots have developed along a wound on one leaf, while several small, scattering, and imruptured spots were found on a few young leaves. The spots are typical of those found on Casimiroa edulis. No positive results have been obtained in the field. Atalantia citrioides Pierre. (CPB 7534, cuttings), I, II (2 plants), III, in field, 1918. Canker spots first appeared on the plants in experiment I in May, 1918: Since that time all plants have become infected, the spots being well distributed over the new foliage. The spots (PI. 59, A) are small to medium, of a brick color, round, flat, and sometimes breaking out in a corky mass. Only a slight depression is visible on the upper surface. The oily outline is very distinct, and no yellow margin is present. The spots are some- what similar to those described for Chaetospermum , and the plants are almost as sus- ceptible. During 1918 no positive results were obtained on the few plants in the field. Atalantia ceylonica Oliver {Rissoa ceylonica, Am.). (CPB 11225, seedling), III. A few oily spots (PI. 59, A, center) have been produced on all the leaves of the plant. They are present also on the twigs. The spots are identical with those described on A . citrioides, although the plant is slightly more susceptible. Poncirus trifoliata (L.) Raf. {Citrus trifoliata L.). Trifoliate orange (Seedlings, Ala- bama), I, II, III, infield, 1917. All the plants (PI. 65, A) included in the experiments have proved to be extremely susceptible in both the field and the greenhouse. In experiment III the plant was killed outright by the heavy canker infection. Leaves, thorns, twigs, branches, and even the old wood were attacked. As a rule, all the spots on the leaves are small to medium sized and very numerous, while on the stem they are large, girdling, and corky. Poncirus trifoliata is extremely susceptible and therefore will always be a menace to complete eradication of canker in Alabama, especially since it has been found J Ircluded in experiments of March 21, 1918. 344 Journal of Agricultural Research voi. xix. no. s that canker may lie dormant in the bark tissues of the old wood and overwinter for a period of six months (7). Eremocitrus glauca (Lindl.) Swingle (Triphasia glauca Lindl.; Atalantia glatica Benth.). Australian desert kumquat (CPB 7239 and 7397, seedlings), I, II, and III, in field, 1917. All plants have shown infection, varying with their condition. Canker (PI. 59, D) has been observed on the leaves, tlioms, twigs, and old wood. A considerable degree of susceptibility is sho\^'n; and, under favorable conditions, the species should be successfully inoculated in the field, although such attempts have proved negative so far. Fortunella margarita (Lour.) Swingle {Citrus margarita Lour.). Oval kumquat (CPB 7597, seedlings), I, II, III, in field, 1918. Fortunella crassifolia Swingle. Meiwa kumquat (CPB 11047, seedlings), I, II (2 plants), III, in field, 1917 and 1918. Fortunella japonica (Thunb.) Swingle {Citrus japonica Thunb.). Round kumquat (CPB 11301, seedlings), I, III, in field, 1918. Fortunella Hindsii (Oliver) Swingle (5c/eroy/jiw Hindsii Champ., Atalantia Hindsii Oliver). Hongkong wild kumquat (CPB 11046C and 11046A, seedlings), I, II, III, and I,' in field, 1917 and 1918. All four species of kumquats have been successfully inoculated, although in all cases with some difficulty. From the results so far obtained no one of the first three species appears to be more susceptible than the other, the amount of infection depending on the growing condition of the individual. Judging from the type and number of spots (PI. 60, C) Fortunella Hindsii is the most susceptible in that the spots are ruptured and corky. As a rule, the spots on the other three species are characterized by being small, slow-growing, scattering, very dark, compact, and unruptured (PI. 60, A). A few slightly ruptured, corky spots have been found on the plants, but usually at wounds. Three plants of Fortunella Hindsii have been successfully inoculated in the field. A few minute infections were obtained on Fortunella tnargariia and Fortunella japon- ica during August, 1919. Plants of the oval kumquat, budded on Poncirus trifoliata, were inoculated in the field every week during the growing season of 19 18 with negative results. Microcitrus australasica (Muell.) Swingle {Citrus australasica, Muell.). Finger lime (CPB 7600 and 7600B, cuttings and seedlings), I, II, III, and II, in field, 1917. Microcitrus australasica var. sanguinea Swingle. (CPB 7775B, cutting), II. Microcitrus Garrowayi (Bail.) Swingle {Citrus Garrowayi, Bail.). Garroway's finger lime (CPB 11008, cuttings), I, II, III. Microcitrus australis (Planch.) Swingle {Citrus australis, Planch.). Dooja (CPB 7307 and 7427, cuttings and seedlings), I, II, in field, 1917. The last two species have proved to be quite easily infected with canker, but it was not until quite recently that Microcitrus aiistralasica and its variety sanguinea became in- fected. Here infection is limited to a few scattering, small, slow-growing, dark, oily spots with an occasional spot on the thorns and twigs. On M. australis and M. Garrowayi from 30 to 90 per cent of the leaves have tiny, scattering, compact spots (PI. 59, C), which do not penetrate through the leaf. Thorn, twig, and stem infec- tions are also severe, the spots being ruptured and of a girdling type, resembling ' Included in experiments of March 21, 1918. July IS, 19J0 Relative Susceptibility to Citrus-Canker 345 somewhat the loose, corky spots on Poncirus trifoUata. During the 19 19 season M. australis was severely infected in the field, leaves, twigs, and branches being attacked. This species has shown almost as much susceptibility as P. trifoUata. Some leaf and stem cankers have also developed on M. australasica. However, it is much more resistant than M. australis. In Table I the data on the susceptibility of the wild relatives of Citrus obtained by Lee {4) are listed for comparison with those reported on by the senior author. Lee worked in the open under field conditions at the Lamao Experiment Station, P. I., while the senior author, using the same type of plants, carried on his inoculation experiments in the green- houses at Auburn, Ala. None of the field results are included in the table, since with few exceptions they were of a negative nature. Table I. -Findings of Lee and Peltier on the susceptibility of the Citrus relatives to citrus-canker Genus and species. Lee's results. Peltier's results. Remarks. RuTACEOus Plants not Closely Related to THE Genus Citrus. Negative Not tested Not tested Negative Not tested do Do Leaves and stem positive. Stem only. Do do do .... do Not tested Leaves only. Leaves and stem positive. Immune. Lee, petioles and stem weakly positive; Peltier, leaves weakly positive. Do Not tested Negative Glycosmis pentaphylla Not tested do ... do Tribe Citreae. Subtribe Aeglinae. Negative Not tested Positive Leaves only. Negative Susceptible. Immune. Do. Lee, leaves and stem positive; Peltier, leaves only. Lee, stems only positive; Peltier, leaves only. Susceptible. Balsamocilrus gabonensis . . . . Balsamocitrus Daviei Subtribe Feroninac. Negative Not tested Not tested Negative do do Subtribe Lavanginae. Positive Negative Not tested Negative Positive . ...do Negative Leaves and stems positive. Immune. Lee. leaves and stems positive; Peltier, leaves only positive. Lee, leaves and stems positive; Peltier, leaves very easily infected. Leaves and stems easily infected. Leaves and stems only weakly positive. Extremely susceptible. Susceptible. lata. Severinia buxifolia Subtribe Citrinac. Negative do Not tested do Not tested Poncirus tn/oliata Not tested Positive Eretnocilriis glauca .. do Fortunella viargarita Not tested . do Positive .... do Do FoTtunella crasst/olta Not tested do Do .. do.... Susceptible. Somewhat susceptible. Do do do . Not tested do MicTocitrus Garrowayi do do Susceptible. Do. Positive do 346 Journal of Agricultural Research voi. xix, no. s Not only the conditions governing the inoculations but even the methods used were widely different in the two experiments. Lee (4) describes his method of inoculating as follows: In making the inoculations an infusion of the organism was painted upon the leaf blade, midrib, petiole, or stem, as the case might be, with a small camel 's-hair brush, and then the tissue was punct\u-ed through the coating of infusion with a needle. The inoculated twig was maintained in a moist condition by wrapping it in paraffin paper, including with the twig also a small piece of moistened cotton. The senior author of the present paper, on the other hand, used infu- sions of the canker organism, which were sprayed directly on the plants in the screened cases by means of an atomizer. In no case were punctures resorted to, although wounds and scratches were present at all times on some of the leaves. It should also be noted that natural infections took place from the more susceptible plants to the majority of the wild relatives. (See dates of inoculations, p. 340.) Natural infections could be counted on in the greenhouse cases, because the plants were set close together and intermixed, and in addition a thorough syringing with a strong water pressure was applied whenever the plants were watered. No infections occurred on any of the rather remote wild relatives of Citrus until several weeks after the last inoculation. This may be due either to the resistance of the plants and the subsequent period of accommodation of the organ- ism to the host or to the rather extended period of incubation. The last statement appears to be more nearly correct and is substantiated by Lee's results. Of the rutaceous plants not closely related to the genus Citrus positive infections have been obtained on Casimiroa edidis, Chalcas exotica, and Claticena lansium. Lee has successfully inoculated the last two plants and, in addition, Evodia ridleyei, E. latifolia, Melicope triphylla, and Toddalia asiatica. Both of us have failed to produce any infection on Xanthoxylum spp., while so far Glycosmis pentaphylla has remained immune. Jehle (j, 2) reports successful needle-prick inoculations on Xantho- xylum fagara (L.) Sarg. and X. dava-hercules (L.) Sarg. He also obtained watery swelling on Chalcas (Murraea) exotica. In all cases, a few non- typical, unruptured spots have been produced at wounds or scratches only. (PI. 59.) Of the plants infected, Chalcas exotica responds the least, and the period of incubation is long. Lee likewise obtained only a very weak reaction. No doubt other plants widely removed from the genus Citrus will be successfully inoculated under certain conditions, although it is extremely doubtful if any of the plants in this group will prove susceptible enough to warrant any attention except to be of interest from a scientific stand- point. In the subtribe Aeglinae, of the tribe Citreae, Chaetospermum gluti- nosum is susceptible enough to rank with some of the Citrus fruits in jtiiy 15. 1920 Relative Susceptibility to Citrus-Canker 347 susceptibility. This confirms the observations of Lee (4) in the Philip- pines, where he has found Chaetospermum generally infected under field conditions. The spots (PI. 58, C) produced on this plant are ruptured, corky, and more or less typical of those found on the plants in the genus Citrus. They occur on the leaves in the absence of wounds. C. gluti- nosum is the most distantly removed relative so far found which is quite susceptible and produces canker spots typical of those found on Citrus spp. It is rather peculiar that the other plants tested in this subtribe are immune or nearly so, especially Balsamocitrus and Aeglopsis. A few small, nontypical, unruptured spots have been found on Aegle marmelos, but only at wounds. Thus Aegle can be classed with the first group discussed in its resistance to canker. Lee (4) failed to obtain any in- fection on Aegle. This is the only plant of the whole group tested by us where Lee's results and mine failed to check. On both Feronia limonia and Feroniella Ixicida positive infections have been obtained. While the spots are typical of those described for the rutaceous group, they can develop in the absence of wounds on the leaves. Of the plants tested in the subtribe Lavanginae, Triphasia irifolia and Severinia buxifolia have remained free from canker. Lee (4) has likewise failed to obtain infection on these plants after repeated trials. No doubt, both species will prove immune to canker. Hesperthusa crenulata, on the other hand, is quite susceptible, in fact, almost as much so as Chaetospermum. While both leaves and twigs are attacked in the absence of wounds, the spots (PI. 59, B) do not resemble those found on any other host. They are flat, and though they rupture, there is no evidence of the corky tissue so typical of the canker spots on Citrus. Citropsis Schweinfurthii, in the subtribe Citrinae, ranks with the rutaceous plants in susceptibility, in that infections occur only at wounds and the spots are small, nontypical, and unruptured. Atalaniia citri- oides and A. ceylonica have proved quite susceptible. The spots (PI. 59, A) are medium-sized, ruptured, corky, and resemble somewhat those found on Chaetospermum. Lee reports A. ciirioides and A. disticha rather resistant. Poncirus trifoliata is without doubt the most susceptible of the wild relatives. Somewhat less susceptible is Eremociirus glauca. Canker spots (PI. 59, D) have appeared on the leaves, thorns, branches, and stems of this plant. The spots are small but ruptured and corky, while on the branches they are of a girdling type, resembling the stem cankers on P. trifoliata, except in size. Equally susceptible and with the same type of canker spots are Microcitrus australis and M. Garrowayi. M. australasica and its variety sanguinea are more resistant, although spots of the same type occur on the leaves, thorns, and twigs. Of the kumquats, Fortunella Hindsii is susceptible. Lee (4) reports that canker occurs naturally on the wild plants on the mountains near Hongkong. The canker spots (PI. 60, C) on F. Hindsii are ruptured, 348 Journal of Agricultural Research voi. xix. no. s raised, and corky, resembling those found on Citrus. The other three species of kumquats tested are equally resistant. While infection has been more or less general in the greenhouse on the young foliage the spots with but few exceptions have remained unruptured. Thus, outside the subtribe Citrinae, only two susceptible plants, Hes- perthusa crenulata and Chaetospermum glutinosum, have been found under greenhouse conditions. The rest of the plants which were successfully inoculated all produced nontypical, unruptured spots at wounds. To this group can be added Citropsis Schweinfurthii. The plants reported free from canker will probably remain immune, while other plants not tested may prove susceptible when inoculated. The remaining plants in the subtribe Citrinae have all been successfully inoculated. In the field successful inoculation both natural and artificial have been produced on Hesperthusa crenulata, Poncirus trifoliata, Fortunella Hindsii, F. margarita, F. japonica, Microcitrus australasica, and M. austrahs. Of these P. trifoliata and M. australis are very susceptible. No doubt under favorable conditions Atalantia citrioides, A. ceylonica, Eremocitrus glauca, and Chaetospermum glutinosum can be successfully inoculated in the field. However, none of them will prove as susceptible as P. tri- foliata. Thus, only the relatives most susceptible under greenhouse conditions have been successfully inoculated in the field. So far as the menace of citrus-canker to the Citrus industry in the United States is concerned, with the exception of Poncirus trifoliata, none of the relatives, native or introduced, discussed above are susceptible enough to warrant further attention. The index of susceptibility to citrus-canker of these plants should be based not on the ability to successfully produce canker infections through needle pricks or wounds, under abnormal conditions, but rather on the ability of the organism to gain entrance into the tissues through natural openings of the leaves in the absence of both artificial and natural wounds. Therefore, the senior author believes that even though he has been able to inoculate a large number of the wild relatives the results have no bear- ing on the eradication program. It is purely of scientific interest to know that P. citri is not limited to the genus Citrus but can produce, under certain conditions, infections on a wide range of plants in the family Rutaceae to which Citrus belongs. The senior author has devoted considerable attention to a study of the various types of spots produced on the various hosts, hoping to be able to correlate the type of spot with resistance. In brief, the spots as observ^ed on the relatives can be classed as follows: Small, slow growing, nontypical, unruptured spots (PI. 57) occurring only at wounds on rutaceous plants; same type of spots, but occurring on the leaves in the absence of wounds (on Feronia limonia); more typical spots (PI. 60, A) which are unruptured except at wounds (on Fortunella margarita); and typical, ruptured, corky spots (PI. 63, D) (on Poncirus trifoliata.) July 15, 1920 Relative Susceptibility to Citrus-Canker 349 SUSCEPTIBILITY OF CITRUS FRUITS Citrus hystrix DC. (CPB 7872 and 2881, seedlings), I, II, III, in field, 1917 and 1918. "Cabayao" (CPB 7831, seedlings), I, II. Two types of these plants have been tested. One group has pointed leaves while the second has rounded ends. Very little infection has been found on the plants with the pointed leaves, either in the field or the greenhouse (PI. 61). However, 70 to 100 per cent of the leaves having rounded tips were infected with small to large and scattering to many spots. Some defoliation occurred. Rather large spots of a girdling type are common on thorns, twigs, branches, and old wood. Lee (j) finds that of the numbers tested by him in the Philippines seven were severely infected, three moderately so, one slightly, while canker was not observed on fotu-, and one proved resistant. As the group is obscure, although a large one, some forms may be found resistant to canker. However, the majority, especially those with rounded leaves, will prove to be almost as susceptible as grapefruit. Citrus medica L. Citron of commerce (CPB 7768 and 7836, cuttings and seedlings), I, II, and III. "Sidro" (CPB 7816, seedlings), II. "Nana" (CPB 11281, seedlings), II, III. "Odorata" (CPB 11294, seedlings), II, III. "Etrog" (CPB 11178, seed- ling), i.m. Of the citrons tested, the "Etrog" proved to be the most susceptible. All the leaves were infected and some defoliation occmred. Twig and stem infections were also present. A few twig and stem spots were found on CPB 7768, 7836, and 11281. On the other numbers canker was limited to the foliage, the percentage of leaf infec- tion varying from 30 to 100, with little defoliation. The spots were, for the most part, small and scattering and very distinct in charac- ter. No doubt the texture of the leaves has a direct influence on the type of spot pro- duced and also on the susceptibility of the leaves. The citrons, as a whole, while rather easily infected, are not as susceptible as grapefruit but are more so than Satsuma, {Citrus nohilis var. unshiu Swingle). Lee (j) tested 14 numbers in the Philippines and found i resistant, 5 on which canker was not observed, 5 with medium infection, and 3 severely infected. He is of the opinion that some of the citrons may be considered as canker-resistant. Citrus sp. Small lemon (CPB 7833, seedlings), I,i II. Sweet lemon (CPB 1158, seedlings), I, II. "Davo lemon" (CPB 7837, seedlings), II, III. Limon real 18 (CPB 7819, seedling), II. The plants of the lemon group so far tested have all proved more susceptible than the citrons. All the plants in the experiments have few to many large twig and stem spots, while 50 to 100 per cent of the leaves are infected. Canker also caused some defoliation of the plants. Two t>-pes of spots are produced on the foliage. Where the texture of the leaf re- sembles that of the citron the same kind of spot is produced, except that it is larger. On the plants with smooth leaves the spots resemble those found on grapefruit (PI. 62). In the scale of susceptibility, the lemons so far tested rank just below grapefruit. Lee's results (j) also show that the lemons are fairly susceptible under Philippine conditions. Citrus sp. Ichang lemon (CPB 11291 and 11204, seedlings), I, II, III, and I, 11,^ III, in field, 1918. The Ichang lemon was not considered under the lemon group because it appears to be a natural hybrid, possibly betweeh lemon and pummelo. The plants are very susceptible, for from 30 to 100 per cent of the leaves are infected and several plants have severe twig and stem infections. All three plants in the isolation field were reported infected during September, 1918, and August, 1919. However, all spots were localized on the leaves. ' Included in experiments of March 21, 1918. 350 Journal of Agricultural Research voi. xix, no. s The spots resemble those on grapefruit leaves (PI. 62), but the plants rank with the lemons in susceptibility to canker. Citrus aurantifolia (Auct.) Swingle (C. Umetta Auct., not Risso). Sour lime (CPB 7338, seedlings), I, II. Only a small percentage of the leaves are infected with canker. The spots are also very small and scattering. No twig or stem infections have ever developed. The spots (PI. 60, E) resemble those on citron to some extent. However, they are smaller, more compact, less corky, and darker in color. While the plants are rather easily infected, the spots increase slowly in size and are few in number. The sour lime is much more resistant than either the citrons or lemons, almost approaching Satsuma in resistance. Lee (j) reports that the limes, with the possible exception of the "Tahiti," are very susceptible in the Philippines. Citrus grandis (L.) Osbeck (C. decumana L.). Grapefruit (CPB 11170, 7834 and Duncan (Alabama), seedlings), I, II; I, II; and II, III. Grapefruit (Duncan budded on Poncirus trifoliata), I, II. Chinese ptimmelo (CPB iio6r, seedlings), I, III. Hirado Buntan(?) pummelo (CPB 7993, seedlings), I, II. Indian pum- melo (CPB 11166 and 2968A, seedlings), I, III, and I, III. Siamese pummelo (CPB 11201 and 6111, seedlings and on C. Schweinfurthii), I, III, and II. With possibly two exceptions, all the grapefruit plants tested in the greenhouse have proved to be extremely susceptible. Approximately 100 per cent leaf infection occurred, with considerable defoliation. Twig and stem infections were also severe, the spots being large and of a giMling type. Several shoots have been killed by the girdling spots. The "Hirado Btintan," reported very susceptible in the preliminary report, has stood up very well, and at the November, 1918, reading only 5 to 10 per cent of the leaves were infected, with few or no spots on the twigs. The Siamese pummelo, especially the number budded on Citropsis Schweinfurthii, shows some resistance to canker. In the field, severe infections have been obtained on grapefruit seedlings (1919), grapefruit (Duncan) budded on Poncirus trifoliata (1917), Sullivan grapefruit (CPB iiooi and 11054) (1918), Mark's Chinese pummelo (CPB 11061, 11217F, and 11217G) (1918), Roeding's Indian pummelo (CPB 2 968 A and 11166) (1918), Florida Shaddock (CPB 11255) (1918), Orangedale Chinese pummelo (CPB 11212 U) (1918), French Martin's Chinese pummelo (CPB 11213 J) (1919), and pummelo (CPB 11219 I) (1918). (See PI. 62.) Only slight leaf infections have been obtained on the Hirado Buntan pummelo (CPB 7993 and 11021) and recently on the Siamese pummelo (CPB 11201 and 61 11) although these plants have been in the isolation field for the past two seasons and surrounded by badly infected plants. Mr. Swingle reports that in Japan the Hirado Buntan is quite resistant, whereas Lee (j) states that the Siamese pummelo is the only variety of Citrus grandis tested by him which gives any promise of being resistant. Citrus sinensis Osbeck (C. Aurantium Loiu*. and Auct. not Linn.). Grenadine orange (CPB 7773, seedlings), I, ^ III. Parson Brown orange (CPB 11324, seedlings), I,^ III. "Naranja" (CPB 7929, seedlings), II, in field, 1918. Orange (CPB 66A seedlings), I, ' III. With the exception of a few small, scattering spots on the twigs of two plants, canker is limited to the foliage of the plants in this group. Apparently the Parson Brown orange is the most susceptible, followed by CPB 66A. The "Naranja " and Grenadine oranges are somewhat more resistant in that only a small percentage of the leaves are infected, the spots are fewer and smaller, and no twig infections are present. ' Incluaed in experiments of March 21, 1918. July IS, 1920 Relative Susceptibility to Citrus-Canker 351 All the plants of this group tested in the isolation field have been successfully inoculated. The following numbers were represented: CPB 11196 Narute, CPB 1 1 164 Temple, CPB 11028 South Carolina, CPB 11 198 Japanese No. i, and CPB 11199 Japanese No. 2. The spots in all cases resembled those found on grapefruit. In susceptibility this group ranks just above the citrons in that twig and stem infections are of more general occurrence. Lee (j) has obser\'ed that the Mediter- ranean varieties are less susceptible than the others. This fact has also been pointed out by other investigators. Citrus nobilis Lour. King of Siam (^CPB 2105, seedlings), I, III. "Naranjita" (?) (CPB 7830, seedlings), II, III, var. dcliciosa Swingle. Tangarine (CPB 11195, seed- lings), I, III. Cleopatra tangerine (CPB 11338, seedlings), I, II, III, var. unskiu Swingle. Satsuma (on Poncirus irifoliata Alabama), II, III. Twig infection, consisting of small, unruptured, scattering spots, is limited to one piant (Naranjita) of this group. The spots (PI. 63, B, at left) on the leaves are small to medium-sized, and, as a rule, rather scattering. The King of Siam orange is apparently the most susceptible. The Satsuma plants are the most resistant. The spots (PI. 63, B) found on the plants of the Citrus nobilis group are very characteristic, resembling to some extent those produced on kumquats. They are of medium size, dark, raised, compact, mostly unruptured, with a distinctly oily outline and some yellow zone. Ruptured spots are only slightly corky. Recent investigations by Tanaka (9) and Scott {8) show that there are a number of distinct strains of Satsuma in Japan, of which three have been found in Alabama. Experiments are now under way to determine the relative susceptibility and resistance of these strains under field conditions. Successful inoculations have been made in the field on Satsuma (PI. 64) and the Cleopatra tangerine. However, these plants are not easily attacked, canker being limited to the foliage. All the plants tested in this group are very resistant. The writers believe that under field conditions suitable for Satsuma culttu-e, and with no interplanting of susceptible varieties, this orange can be grown with little or no loss from canker even when the disease is prevalent. From the results so far obtained all the plants of the Citrus nobilis group can be placed in this class. In the investigations on susceptibility and resistance any variety showing as much resistance to canker as the Satsuma has been classed as promising. Citrus mitis Blanco. Calamondin orange (CPB 11265, 44305, and 7065 seedlings), I, 11,1 11^ iii^ and 1,1 11,1 iii^ in f^^i^^ j^j^^ j^^g, and 1919. Scattering stem infections and some defoliation have occurred on two of the seven plants tested. From 20 to 90 per cent of the foliage of the other plants have small to large, scattering spots. The spots (PI. 63, E) are altogether characteristic, and for the most part are unrup- tured. They are round, raised, compact, and oily, somewhat like the spots described for kumquat. When ruptured the spots are flat and have very little cork. In the field canker is limited to the foliage, and the plants are more resistant than Satsuma. Lee (j) also finds that in the open Citrus mitis is truly resistant, and he thinks that it is apparently more so than Satsuma. Citrus sp. Kansu (Yuzu) orange (CPB 11242, seedlings), I,' II, III, in field, 1918 and 1919. The Kansu orange, collected by Mr. Frank N. Meyer in north China several years ago, is considered by Dr. T. Tanaka^ to be the same as the well-known "Yuzu "used in Japan for many years as a stock. Under both field and greenhouse conditions the plants have proved resistant. Apparently the leaves are quite easily infected, but the spots rarely increase in size, > Included in experiments of March ai, 1918. ' The data are uiipublished. 352 Journal of Agricultural Research voi. xix, no. s although they commonly rupture (PI. 60, D). The spots do not penetrate to the upper surface. The plants are much more resistant than Poncirus trifoUata, which the writers understand has to a large extent replaced Yuzu as stock for Satsuma in Japan. Other conditions being equal, Yuzu is to be preferred to P. trifoUata from the stand- point of canker susceptibility. Citrus sp. Natsu-mikan (CPB 11184, seedlings), I, II, III, in field, 1918 and 1919. In some ways this plant resembles the hybrids between the grapefruit and loose- skinned oranges, such as the tangerine, known in this country as tangelos. All plants of the Natsu-mikan in the greenhouse and field have been rather severely infected. Some twig and stem infections have been found, and from 50 to 100 per cent of the leaves have medium to large and scattering to many spots. The spots, although larger and more corky, resemble those found on Satsuma. Some defoliation has taken place, due to canker. If the Natsu-mikan is closely related to the mandarin orange it is very much more susceptible than any of the plants so far studied in this group. Lee (3) reports the Natsu-mikan as susceptible in the Philippines. Citrus excelsa Wester. (CPB 11280, seedlings), I, III. From 90 to 100 per cent of the foliage of the two plants is infected with many large spots. Some few spots on the twigs are also present. Because of the citron- like texture of the leaves, the spots resemble those on the citron, except in size. Apparently it is not quite as susceptible as grapefruit. In the Citrus fruits, where so many species and varieties were tested with more or less varying results, it is extremely hard to classify the susceptibility of these plants, especially where so many factors must be taken into consideration. Probably the most important and vexing factor is the physiological condition of the plant. In looking over the notes taken approximately each month on the plants in the experiments, it is found that there are certain cycles of canker infection which coin- cide with the growth periods of the plants. Thus, one or two observations on inoculated plants in the greenhouse or on those exposed to natural infection in the field are not sufficient to detennine accurately the exact susceptibility or resistance of a plant. Some of the points to be reckoned with under the factor of the physiological condition of the plant are the rate of growth, not only of the plant but of the leaves themselves, age and size of the plant and leaves, leaf texture, and rate of maturation of the leaves. All these have an important relation to canker suscepti- bility and resistance. Leaf texture with its various ramifications probably plays an impor- tant role in determining resistance in many cases. This can be best illustrated by comparing an infected kumquat leaf (PI. 60, A) with an old grapefruit leaf (PI. 60, B). The leaves are apparently very similar in texture, and a close study of the spots produced on the two shows that they are identical. In other words, while an ordinary grapefruit leaf is still thin and light green in color, it is very susceptible, large corky spots being produced. However, if an old leaf is taken which has apparently the same texture as a leaf of the kumquat, it is as hard to infect as the kumquat, and small, rounded, glistening spots are formed. July 15, 1920 Relative Susceptibility to Citrus-Canker 353 When no twig or stem infection occurs and only small, scattering, unruptured spots appear on the leaves, the plants show enough resistance to be classed as resistant. An intermediate group can be formed where the spots are larger, ruptured, and more general on the leaves, with occasional twig and stem infections. Plants placed in this group might be found promising under certain conditions. Plants which show large, ruptured spots on the leaves, severe enough to cause defoliation, and large girdling cankers on the twigs and stem should be classed as extremely susceptible. With these remarks in mind, the writers will attempt to rank the Citrus fruits, provisionally, in groups according to their suscepti- bility to citrus-canker. The plants of the grapefruit and pummelo group are extremely sus- ceptible. However, the Siamese and possibly the Hirado Buntan pum- melos give promise of showing some resistance to canker. Of the numbers tested belonging to Citrus hysirix, those with rounded leaves are as susceptible as grapefruit. The plants with pointed leaves are apparently more resistant to canker. More study is needed to determine whether this will hold for all forms of Citrus hysirix in the Philippines. All numbers of the lemons tested, including the Ichang lemon, show about equal susceptibility, which is slightly less than that of grapefruit. The plants of the sweet-orange group vary somewhat in susceptibility. Leaf infections are severe, and twig and stem cankers are common. As a whole, they are not as susceptible as the lemons. Citrus excelsa and the Natsu-mikan are equally as susceptible as the plants of the sweet orange group. Since only one number of the limes was tested, the position which the limes should take in the scale of susceptibility is doubtful. The sour lime tested proved to be somewhat resistant. However, Lee finds that with one exception the limes are susceptible. While the citrons tested are easily infected, the spots are small, in- creasing very slowly in size. Twig infection occurs only occasionally and is the exception rather than the rule. Citrus mitis, at least seedling plants such as were used in the inoculation experiments in the greenhouse, while showing some resistance are more susceptible than in the field. Leaf infections are scattering, and twig cankers are rarely produced. The Kansu (Yuzu) orange so far has proved decidedly resistant. No twig cankers have occurred, and only small, scattering spots have devel- oped on the foliage. All numbers of the Citrus nobilis group tested have proved to be decid- edly resistant, and, no doubt, all of these plants, if not mixed with sus- ceptible varieties, could be grown under canker conditions. That does not mean that it would be economical or at all advisable to allow canker to persist even in unmixed plantings of Satsuma or other varieties of C. nobilis. 354 Journal of Agricultural Research voi. xix. no. s SUSCEPTIBILITY OF CITRUS HYBRIDS Faustrime* {Citrus aurantifolia, West Indian lime, X Microcitrus australasica) . (CPB 49819, 49823, and 49835, cuttings), II; II and III. Faustrimon [Citrus limonia, lemon, X Microcitrus australasica). (CPB 49841, 49843, and 49844, cuttings), II; II and III. Faustrimedin (O'/rwj miiis, calamondin, X Microcitrus azistralasica). (CPB 47431, cuttings), I, III, in field, 1918. From 30 to 90 per cent of the foliage of these plants is infected with scattering spots (PI. 63, A). Some defoliation from canker has taken place. Thorn, twig, and stem cankers are common (PI. 63, A). The spots are similar to those found on Micro- citrus australasica, except that they are larger and more ruptured. The spots on the twigs and stem are of a girdling nature. The last number has been tested in the field with positive results on the foliage only. These hybrids are more susceptible than M. australasica. Citr&nge (Poncirtis trifoliata, X Citrus sinensis) . Seedlings. Two or more plants of each number of the following citranges have been given a thorough test in both the greenhouse and the field: Colman (CPB 7896 and 772 AC), Rusk (CPB 7956, 1 1030, 7895, and 716), Rustic (CPB 7934 A), Sanford (CPB 7963 and 1423 AB), Savage (CPB 7961 and 1423 AB), Willits (CPB 7897 B, 7960, and 777 AB), Etonia (CPB 749 AB), and citranges (CPB 1425 AB, 1416, 43931, 43480, and 43434). The percentage of leaf infection has varied from 10 to 100 per cent, depending on the condition of the plant. The majority showed over 50 per cent of infected leaves. Defoliation of the leaves due to canker was common. Large girdling spots have appeared on the stems of most of the plants. The spots (PI. 63, D) on the leaves and twigs are similar to those produced on Poncirus trifoliata. While some variations have occurred in the susceptibility of the different numbers, the results as a whole show that all the citranges (PI. 65, B) are equally as susceptible as Poncirus trifoliata (PI. 65, B). Thus none of the citranges tested are of any promise in the search for a resistant stock. Citnunelo {Citrus grandis, Bowen grapefruit, X Poncirus trifoliata). (CPB 4493, 4554, 4564, and 4475, seedlings), I, II; I, II (2 plants); I and I, III. Almost 100 per cent leaf infection, with some defoliation, occurred on all the plants. The spots (PI. 63, D) are large, scattering to many, and resemble those produced on Poncirus trifoliata except in size. Girdling spots of various sizes occurred on most of the plants, not only on the twigs and branches but even on the old wood. The citrumelos (PI. 65, C) are even more susceptible than Poncirus trifoliata and, therefore, are of no economic value from the standpoint of their behavior to citrus- canker. Citradia {Poncirus trifoliata, X Citrus aurantium, sour orange). (CPB 50850, seed- lings), I, II. While from 40 to 80 per cent of the leaves have been infected with small, scattering, typical spots, no spots have been produced on the twigs or branches. Apparently, the citradia (PI. 65, D) is a slower grower than the rest of the Poncirus trifoliata hybrids. The susceptibility of these plants, however, is sufficient to bar them from fvulher tests. Citrandarin {Citrus nobilis X Poncirus trifoliata). Seedlings. In the greenhouse, plants of the following numbers have been given a thorough trial. CPB 40210, 40303, 40315, 40368 B, and 48529. All of these numbers are hybrids * The hybrids were supplied by ^Ir. Walter T. Swingle, who informs me that they were labeled with the aboratory names, for the most part still unpublished. Citrange, limequat, and tangelo have been pub- lished, but citnmshu, cicitrange, citrumelo, citradia, citrandarin, faustrime, faustrimon, faustrimedin, citrangedin, citrangarin, citranguma, citrangequat, limelo, bigaraldin, orangelo, orangequat, clemelo, siamelc^ satsumelo, siamor, calarin, and calashu are tentative laboratory names that may not be used in the reports which taay later be issued concerning hybrids. Many hybrids which in this paper are given separate names will in later reports be group>ed under some one name. July IS. I9J0 Relative Susceptibility to Citrus-Canker 355 of the King of Siam orange with Poncirus trifoliata. In the field, CPB numbers 40175 A, 49720, 49721, 49722, 49724, and 49726 (cross between King of Siam orange and Poncirus trifoliata), 49624, 49625, 49629, 49644, 49663, and 49644 (cross between Clemen- tine orange and P. trifoliata), 49686, 49688, 49695, 49699, and 49712 (cross between Oneco tangerine and P. trifoliata), 49732, 49735, 49737. 49746, and 49748 (cross be- tween a tangerine and P. trifoliata) were tested. Some individual variation in susceptibility due to the condition of the plants oc- curred in the greenhouse. However, all plants proved susceptible. From 30 to 100 per cent leaf infection, with some defoliation, was observed on the majority of the plants. Some scattering twig infections occurred on all but one number. Rather large, girdling spots on the old wood were found on several of the plants. In the isolation field, all the plants have been successfully infected. An abun- dance of spots occurred not only on the leaves but on the twigs, branches, and old wood. No differences were noted in the susceptibility of the plants having different Citrus nobilis varieties as one parent. The Poncirus trifoliata blood predominates, in that all the leaves of the above numbers are like this plant and all have the same leaf texture. All the citrandarins (PI. 65, E) are about as susceptible as their parent, P. trifoliata. Citrunshu {Citrus nobilis var. unshiii, Satsuma, X Poncirus trifoliata). Seedlings and on P. trifoliata. These plants are very similar to the citrandarins, and their behavior towards citrus- canker is likewise the same. Of the nine numbers (CPB 51102, 49607, 49608, 49611, 49615, 49616, 49619, 49620, and 49623) tested in the field, all proved equally suscep- tible. Leaf infection was common, and some stem cankers were present. They are more resistant than the citrandarins, although further tests may show them to be as susceptible. The type of spot produced on all these plants is identical to those on Poncirus trifoliata. Cicitrange {Poncirus trifoliata X Colman citrange, and P. trifoliata X Sanford cit- range). (CPB 48290 and 48316A, seedlings), I, II, and I, II, III. These plants have shown considerable susceptibility to canker throughout the course of the inoculation experiments; in fact, one plant was killed by canker, while the others have been severely attacked. Without question, the cicitranges (PI. 65, F) are equally as susceptible as Poncirus trifoliata. Citrangedin (a citrange X Citrus mitis, calamondin). Seedlings and on Poncirus trifoliata. All plants in the greenhouse (CPB 48045) and isolation field (CPB 50485, 50486, 50493, 50495, 50500, and 50501) have been successfully infected. The spots are rather small and scattering on the leaves. Few twig and stem cankers have been observed. The spots are not typical of those produced on the citranges but resemble more those on Citrus mitis. The fact that these plants are more resistant to canker and that the spots themselves are not similar to those on the citranges can be traced back primarily to a difference in the leaf texture of the two hybrids. The citrangedins are more susceptible than C. mitis, but they are more resistant than the citranges. While the leaves still retain their trifoliate character, the size, shape, and texture of the leaflets are different. They are also a darker graen, and apparently mattu-e faster than the leaves of the citranges. Citrangarin (Sanford citrange X Citrus nobilis var. deliciosa, Oneco tangerine). Seedlings. A plant (CPB 48776) of this hybrid was tested in the isolation field and has been successfully infected with a few scattering spots resembling those on Satsuma. While the leaves of this plant are trifoliate, they have a texture similar to that of the 177286°— 20 2 356 Journal of Agricultural Research voi. xix. No. a tangerine. It is interesting to compare the susceptibility of the citrangarin and the citrandarin. The latter was found to be as susceptible as Poncirus trifoliata and was very similar in character. The citrangarin, on the other hand, while it retains the trifoliate character, has a leaf texture more like the second parent and is much more resistant. Citranguma {Citrus nobilis var. unshiu, Satsuma, X Morton citrange). Seedlings. The citranguma (PL 65, G) is possibly slightly more susceptible than the Satsuma. Leaf infections have been secured in both the greenhouse (CPB 48055, and 48055A) and field (CPB 49773). An occasional spot has been found on the smaller twigs of these plants in the greenhouse and field. The leaf texture and type of spot are similar to those found on the Satsuma. How- ever, the leaves do not reach maturity so soon. There is a decided tendency in the citranguma plants for the leaves to revert from the trifoliate to a single leaf. This is especially noticeable on the new growth. Citrangequat (Willits citrange X Fortunella margarita, oval kumquat). Seedlings. The citrangequat, without question, is the most promising hybrid so far tested. No natural infections have ever been obtained in the field (CPB 48010E and 48010F). Under greenhouse conditions (CPB 48010 and 48010D) only several tender leaves have been infected with tiny, compact, unruptured spots (PI. 63, C). So far, the citrangequat has shown more resistance than any of its parents. These plants (PI. 65, PI) make a splendid, rapid, straight growth. The rate of growth is more rapid than that of Poncirus trifoliata, and the plants are much better adapted for budding purposes. The trifoliate leaves are rather small and retain considerable of the leaf texture of the kumquat. The maturation of the leaves is also as rapid as in the kumquat. The leaves, especially those on the new growth, have a tendency to revert to the signle leaf of the kumquat. Limequat (Citrus aurantifolia, West Indian lime, X Fortunella japonica, round kumquat). Seedlings. All plants in the greenhouse (CPB 48787A, 48787B, 49787E, 49792E, and 48798E) and isolation field (CPB 48792E) have been infected. Leaf, twig, thorn, branch, and stem spots are common. In fact the limequat is more susceptible than either parent. Several plants have died in the greenhouse experiments from overwatering, although plants next to them have thrived. On the whole, the limequat plants (PI. 66, B) worked with have not been strong nor altogether healthy. The rate of growth of these plants has also been slow. It may be that strong, healthy plants growing under ideal conditions would sho;v more resistance to canker. However, judging from the results obtained with the plants available, the limequat is somewhat susceptible. Limelo {Citrus aurantifolia, West Indian lime, X C. grandis, sour pummelo). (CPB 40502, 40526A, and 40567B, seedlings), I, II; I, III, and I, II. All limelos (PI. 66, A) tested have proved to be equally as susceptible as grape- fruit, so that they are of no practical importance from their ability to resist canker under orchard conditions. Bigaraldin {Citrus aurantium, sour orange, X C. miiis, calamondin). On Poncirus trifoliata. Only one plant (CPB 50352) of this hybrid was included in the field. It was suc- cessfully infected and no doubt will prove as susceptible as the sour orange. Orangelo {Citrus grandis, Bowen grapefruit, X C. sinensis. Lamb summer orange). Seedlings. July IS, 1920 Relative Susceptibility to Citrus-Canker 357 All three plants (CPB 1668A) in the isolation field have been snccessfully infected. Spots are limited to the leaves. No twig or stem cankers have developed. Judging from the type of spot produced and the leaf texture, this hybrid will prove rather susceptible. Orangequat {Citrus sinensis, Hartley mandarin, X ForhmMa margarita, oval kumquat). On Poncims trifoliata. The one plant (CPB 50312) in the isolation field was easily infected. A few twig and stem cankers have appeared. Judging from the leaf texture and type of spot, it may prove susceptible. Clemelo {Citrus nobilis, Clementine orange, X C . grandis , gTapeimit) . On Povcirus trifoliata. Both the plants in the greenhouse (CPB 49006, 49012, 49013. 49025, and 4903S) and those in the field (CPB 49029, 49032 , and 49049) have been easily infected. From 50 to 100 per cent leaf infection with some defoliation has occurred. Most of the plants in the greenhouse have large, girdling stem cankers. The spots in all cases resemble those on grapefruit. The clemelos (PI. 67, B, at left) are as susceptible as grapefruit and therefore can not be considered promising. Siamelo {Ciirtis nobilis, King of Siam orange, X C. grandis, grapefruit). Seedlings and on C. grandis. The behavior of the siamelos (PI. 67, A) (CPB 47255B, 51588, 51598, 59852, S2G07N2, 52013S9, and 5202iA5)in the greenhouse, and 50320, 52007N2, 52007B3, and 51947 in the field, toward canker is about like that of the clemelos. However, they are not quite as susceptible, in that less twig and stem infections occurred. Satsumelo {Citrus nobilis var. tinshiti, Satsuma, X C. grandis, grapefruit.) On Pon- cirus trifoliata and C. grandis. Many numbers of this hybrid have b^en tested in the field (CPB 50304) and green- house (CPB 50304, 52009A2, 52009G 5, 52011, and 52011G 4). Thus far, the satsu- melos react to citrus-canker in the same manner and extent as the siamelos (PI. 67, B, center) . Siamor {Citrus nobilis, King of Siam orange, X C. sinensis. (CPB 52029E 2, seed- ling), III. The plant included in the greenhouse experiment has proved to be extremely sus- ceptible to canker. Heavy defoliation and severe stem cankers occiured soon after inoculation. Leaf texture and the type of spot are identical with those of grapefruit. The plant is likewise as susceptible. Tangelo {Citrus nobilis var. deliciosa, tangerine, X C- grandis, Florida grapefruit). Seedlings, and on C. grandis. Besides the Thornton (CPB Iv7i5, and 11034 greenhouse and L714B field) and Samp- son (CPB Iy789A and 7664 greenhouse, and 7664 and 11037 field) tangelos, which are now being grown to a small extent in Florida, about a dozen numbers (CPB 1230, 1257A, L16, 7161, 7675C, 1191, 1348A, 1262B, 40971A, 52018C2, and 52018E2 in the greenhouse and CPB 52016E4 in the field) were tested. Plants of all numbers have been infected. From 20 to 100 per cent leaf infection has occurred in the greenhouse. Twigs and stem infections have not been general but are rather the exception than the rule. In the field, canker has been limited to the foliage. The spots in all cases resemble those on grapefruit. Careful observations have shown that those plants (PI. 68, B) with leaves resembling the tangerine are more resistant than the numbers with grape- fruit leaves. For the present the tangelos can be considered somewhat promising but not quite so resistant as the Satsuma. 358 Journal of Agricultural Research voi. xix, no. s Calarin {Citrus mitis, calamondin, X C. nobilis var. deliciosa, tangerine). On Poncirus trifoliata. The one plant (CPB 50314) tested proved to be easily infected in the field. How- ever, no twig or stem spots have been observed. Further trials are necessary before the susceptibility of the plant can be definitely judged. CaX&shw {Citrus mitis , calamondin, X C. nobilis var. unshiu, Satsuma.) On Pon- cirus trifoliata. The plant (CPB 50309) in the isolation field has reacted to canker in about the same degree as the calarin. Among the numerous crosses of Citrus fruits made by Mr. Swingle Poncirus trifoliata was used as one parent in the hope of obtaining hy- brids more resistant to low temperatures than the ordinary Citrus fruits. Thus, P. trifoliata has been crossed with sweet orange, sour orange, grapefruit, King of Siam orange, tangerine, Clementine orange, Satsuma, and citrange. While all of the resulting hybrids have proved to be quite resistant to low temperatures they are equally as susceptible to citrus-canker as P. trifoliata; in fact, the citrumelos are even more so. The hybrids all resemble P. trifoliata in size, shape, and texture of leaves, so that the more or less resistant mandarin group has had no influence on the character or resistance of the hybrid. Thus, it can be safely predicted that all crosses with P. trifoliata will yield a hardy hybrid resembling P. trifoliata and equally susceptible. On the other hand, the citranges, while equally as susceptible as Poncirus trifoliata, when crossed with calamondin, tangerine, Satsuma, and kumquat, yield hybrids which are hardy and at the same time resistant to citrus-canker. In fact, the citrangequat is even more resist- ant than the kumquats themselves, in spite of the fact that it is a rather rank and rapid grower. The citrangarins and citranguma, while not as resistant as the citrangequat, are decidedly more resistant than the citrandarin and citrunshu, the corresponding hybrids with P. trifoliata. These hybrids, while still retaining the trifoliate character of Ponci- rus trifoliata, are more like the resistant parent in texture of the leaves. In the case of the citranguma and citrangequats there is also a tendency for the leaves to revert to a single leaf. Thus, any plant resistant to canker when crossed with the citranges will be hardy and resistant, or even more so than the original resistant parent, as shown by the be- havior of the citrangequat, even though the citranges are equally as susceptible to canker as P. trifoliata. The limequat and the orangequat, while more resistant than the lime or orange, are not as resistant to canker as the kumquat. The leaves resemble those of lime and orange, respectively, in size and shape, but the leaf texture resembles more that of the kumquat. The citrangequat is much more resistant to canker that the two hybrids named above, in spite of the fact that the citrange is much more susceptible than either the lime or orange. July IS. 1920 Relative Susceptibility to Citrus-Canker 359 The influence of pummelo on the second parent varies. As is to be expected, pummelo crossed with lime and orange produced hybrids which are as susceptible as grapefruit. Hybrids between pummelo and Clem- entine orange, King of Siam orange, tangerine, and Satsuma are suscep- tible to some extent, variation in susceptibility depending more or less on whether the leaf is of the grapefruit or the mandarin type. The calarin, hybrid between calamondin and tangerine, and the calas- shu, hybrid between calamondin and Satsuma, so far give promise of being as resistant as their parents. FALSE HYBRIDS A rather large number of "false hybrids" or nucellar bud sports, the the results of Mr. Swingle's crosses with varieties of Chinese pummelo and American grapefruit were tested in both the field and the greenhouse. The plants were vigorous growers and produced abundance of new growth. Most of these false hybrids resemble the Chinese pummelo. All proved extremely susceptible, in fact, even more so than grapefruit (PI. 68, A). Leaf infections were so severe as to cause defoliation. The spots on the twigs and stems were large and of a girdling nature. A number of twigs and several plants were killed outright by complete girdling. The spots in all cases resembled those found on grapefruit. Thus, all the false hybrids tested are extremely susceptible in both the field and the greenhouse, and no one of them gives any promise of canker resistance. SUMMARY (i) The investigations on the susceptibility and resistance to Pseu- domonas citri Hasse of the wild relatives. Citrus fruits, and hybrids of the genus Citrus reported on in a preliminary paper have been contin- ued in both the field and greenhouse. Many more numbers have been successfully inoculated, others have proved to be extremely susceptible, while some still show considerable resistance. (2) The successful inoculation of a large number of wild relatives m the greenhouse shows that Pseudomonas citri has a wide range of hosts and is not limited to the genus Citrus. (a) Of the rutaceous plants not closely related to Citrus, positive in- fections have been obtained on Casimiroa edulis, Chalcas exotica, and Claucena lansium. Xanthoxylum sp., and Glycosmis pentaphylla have re- mained immune. In all cases a few nontypical, unruptured spots have been produced, but only at wounds or scratches on the leaves. Chalcas^ exotica responded the least, and the period of incubation was long. ih) Of the tribe Citreae, subtribe Aeglinae, Chaetospermum glutinosutn is the most distantly removed relative, so far found, that is quite suscep- tible and on which canker spots are more like those found on Citrus. Aegle martnelos has been only slightly infected, while Balsamocitrus Dawei and Aeglopsis Chevalieri have remained immune. 360 Journal of Agricultural Research Voi. xix, No. s (c) Both Feronia limonia and Feroniella lucida of the subtribe Fero- ninae have been successfully inoculated. While the spots are typical of those described for the rutaceous group, they can develop in the absence of wounds. {d) Of the plants tested in the subtribe Lavanginae, Hesperthusa ci'enulata, while producing very non typical spots, is quite susceptible in that infection can take place on the leaves and twigs. Triphasia trifolia and Severinia buxifolia have remained immune. (e) All plants of the subtribe Citrinae which have been tested have been infected. Citr apsis Schweinfurthii is weakly positive; Atalaniia citrioides and A. ceylonica are easily infected; Eremocitrus glauca. Micro- citrus australasica, M. australasica var. sanguinea, M. australis and M. Garrowayi, plants native to Australia, are rather susceptible in that leaves, twigs, thorns, and branches are attacked; all four species of kum- quat have been successfully infected; Fortunella margarito., F. japonica, and F. eras sif alia exhibited considerable resistance, while F. Hindsii is very susceptible; and Poncirus trifoliata is extremely susceptible. (/) Although working under entirely different conditions and using different methods of inoculating plants, Lee's results with the wild rela- tives check with those obtained by the senior author in the greenhouse, with one exception. ig) In the field, only the wild relatives which were most susceptible under greenhouse conditions have been successfully inoculated. Of these Poncirus trifoliata and Microcitrus australis have proved to be susceptible, while M. australasica and Fortunella Hindsii are somewhat susceptible. Hesperthusa crenulata reacts to caxiker in about the same degree in the field as in the greenhouse. (h) So far as the menace of citrus-canker to the Citrus industry of the United States is concerned, with the exception of Poncirus trifoliata, none of the wild relatives, native or introduced, now growing in the Citrus districts are susceptible enough to have any bearing on the eradi- cation program. (3) Little or no change in the susceptibility or resistance to citrus- canker has been noted among the Citrus fruits from that previously indi- cated. All plants tested have been successfully inoculated. (a) The plants of the grapefruit and pummelo group are extremely susceptible, with the exception of the Hirado Bun tan and Siamese pum- melos. (6) Of the numbers tested belonging to Citrus hystrix, those with rounded leaves are as susceptible as grapefruit. The plants with pointed leaves are apparently more resistant. (c) All numbers of lemons tested, including the Ichang lemon, show about equal susceptibility, which is slightly less than that for grapefruit. (d) As a whole, the plants of the sweet-orange group are slightly less susceptible than the lemons. Citrus excelsa and the Natsu-mikan can be classed with the sweet orange in susceptibility. ' juiyis, 1920 Relative Susceptibility to Citrus-Canker 361 (e) The sour lime tested proved to be somewhat resistant. (/) While the citrons are easily infected, the spots are small and increase slowly in size. Twig infection is not common. {g) Citrus mitis, while showing some resistance, is more susceptible than has been reported from the Philippines. Qi) The Kansu (Yuzu) orange has proved somewhat resistant. No twig infection has occurred, and only scattering spots have developed on the foliage. {i) All numbers of Citrus nobilis and its varieties have proved to be rather resistant to canker. (4) All hybrids are attacked by citrus-canker in varying degrees. (a) The citranges, citrumelos, citradias, citrandarins, citrunshus, and cicitranges, all having Poncirus trifoliata as one parent, are extremely susceptible. Apparently all crosses with P. trifoliata will yield hardy but susceptible hybrids. (6) The citrangedins, citrangarins, citrangumas, and citrangequats, with susceptible citranges as one parent, are not only hardy, but decidedly resistant to canker; in fact, the citrangequat is practically immune in spite of the fact that it is a rapid grower. (c) The limequats and orangequats are somewhat susceptible. {d) lyimelos and orangelos are as susceptible as grapefruit, while cle- melos, siamelos, satsumelos, and tangelos are not so resistant as the mandarin oranges. {e) The calarins and calashus are as resistant as either parent, while siamors and bigaraldins are susceptible. (5) All false hybrids are extremely susceptible. (6) Leaf texture is apparently an important factor in influencing resistance to Pseudomonas citri by its host plants. This phase deserves further investigation. LITERATURE CITED (i) JEHLE, R. A. 1917. SUSCEPTIBIUTY OF NON-CITRUS PLANTS TO BACTERIUM CITRI. In PhytO. pathology, v. 7, no. 5, p. 339-344- 3 ^g- (2) 1918. SUSCEPTIBILITY OF ZANTHOXYLUM CLAVA-HERCULES TO BACTERIUM CITRI. In Phytopathology v. 8, no. i, p. 34-35. (3) Lee. H. a. i918. further data on the citrus canker affection of the citrus species AND VARIETIES AT LAMAO. In Philippine Agr. Rev,, v. 11, no. 3, p. 200-206, pi. 9-15. (4) 1918. FURTHER DATA ON THE SUSCEPTIBILITY OF RUTACEOUS PLANTS TO CITRUS- CANKER. In Jour. Agr. Research, v. 15, no. 12, p. 661-665, pl- 60-63. (s) and Merrill, R. I. 1919. The susceptibility of a NON-RUTACEOUS host TO CITRUS CANKER. In Science, n. s. v. 49, no. 1273, p. 499-500. 362 Journal of Agricultural Research voi. xix. no. s (6) Peltier, George 1,. I918. SUSCEPTIBILITY AND RESISTANCE TO CITRUS-CANKER OF THE WILD RELA- TIVES, CITRUS FRUITS, AND HYBRIDS OF THE GENUS CITRUS. In Jour. Agr. Research, v. 14, no. 9, p. 337-357, pi. 50-53. (7) and Neal, D. C. 19 18. OVERWINTERING OF THE CITRUS-CANKER ORGANISM IN THE BARK TISSUE OF HARDY CITRUS HYBRIDS. In JotiT. AgT. Research, v. 14, no. 11, p. 523-524, pi. 58. (8) Scott, L. B. 1918. varieties of the satsuma orange group in the united states. U. S. Dept. Agr. Bur. Plant Indus. Hort. and Pom. Invest. Cir. i, 7 p. (9) Tanaka, Tyozaburo. 1918. VARIETIES OF THE SATSUMA ORANGE GROUP IN JAPAN. U. S. Dept. Agr. Bur. Plant Indus. Crop Physiol, and Breeding Invest. Circ. 5, 10 p., 2fi&- PLATE 57 Leaf of Casimiroa edulis, with naturally occurring spots from the greenhouse inocu- lations. Note that the spots are small, unruptured, and occur only along scratches. Natural size. Relative Susceptibility to Citrus-Canl Journal of Agricultural Research Vol. XIX, No. 8 PLATE 64 Naturally infected leaves of Citrus nobilis var. unshiu from the field, showing va- rious types of spots produced. As a rule the spots on the leaf to the left are found most frequently. All leaves represent rather severe infections. X H- 177286°— 20 3 PLATE 6s Some of the hybrids of Porcirus trifoliata, showing vigor, type of growth, leaf characters, and relative susceptibility to citrus-canker, arranged in order of their susceptibility: A. — P. trifoliata. B. — Rusk citrange. C. — Citruraelo. D.— Citradia. E . — Citrandarin . F. — Cicitrange. G. — CitrangHma. H. — Citrangequat. Relative Susceptibility to Citrus-Canker Plate 65 Journal of Agricultural Research Vol. XIX, No. 8 Relative Susceptibilily to Citrus-Canker Plate 66 Journai of Agricuitural Researchi Vol. XIX, No. 8 PLATE 66 A. — Limelos in the greenhouse inoculation experiments, showing type of growth, leaf characters, and susceptibility to citrus-canker. B. — Limequats in the greenhouse inoculation experiments, showing type of growth, character of leaves, and susceptibility to citrus-canker. The large, broad leaf forms at the left are more susceptible than the narrower leaf forms at the right. PLATE 67 A. — Siamelos in the greenhouse inoculation experiments, showing type of growth, leaf characters, and susceptibility to citrus-canker. B. — Comparison of type of grovvlh, leaf characters, and susceptibility to citrus- canker in clemelo, satsumelo, and tangelo in the greenhouse inoculation experiments. Relative Susceptibility to Citrus-Canker PLATE 67 Journal of Agricultural Research Vol. XIX, No. 8 Relative Susceptibility to Citrus-Canker Plate 68 %fe: Mi*r w Journal of Agricultural Research Vol. XIX. No. 8 PLATE 68 A. — Results of the greenhouse inoculations with some of the false hybrids. Note defoliation and heavy stem infection. B. — Tangelos in the greenhouse inoculation experiments, showing type of growth, leaf character, and susceptibility to citrus-canker. PRESOAK METHOD OF SEED TREATMENT: A MEANS OF PREVENTING SEED INJURY DUE TO CHEMICAL DISINFECTANTS AND OF INCREASING GERMICIDAL EFFICIENCY By Harry Braun ' Scientific Assiitant, Laboratory of Plant Pathology, Bureau of Plant Industry, United States Department of Agriculture INTRODUCTION The widespread use of formalin, copper sulphate, and other germicides in seed treatment for the control of seed-transmitted diseases is generally attended by decreased and retarded seed germination. Pathogens on the seed coats, present as dry bacteria, fungus spores, or dormant myce- lium, are usually in a resting stage and as such require the use of dis- infectants in a fairly strong concentration — i to 80 for copper sulphate (CuSOJ and i to 200 or i to 320 for formalin (CH,0) — which often act very detrimentally on the germinating seed. Even much weaker solu- tions— I to 200 for copper sulphate and i to 400 for formalin — exhibit retarding and killing effects when used on wheat. The use of lime after copper sulphate, while beneficial to some extent, does not entirely prevent seed injury, nor has the detrimental effect of formalin been so far fully counteracted. The economic importance of the annual loss of grain due to seed treatment is such that during the recent war it occasioned an elaborate series of tests of standard grain disinfectants by the War Emergency Board of the American Phytopathological Society. In fact (j)2— the difficulty of avoiding injury to the seed from treatment that is too severe or from improper drying after treatment has undoubtedly had more influence in preventing the general spread of the practice of disinfecting seed grain than has the cost of ma- terials or the difficulty of the treatment itself. In the course of investigations on the blackchaff bacterial disease of wheat (5, 6, 7, 8), under the direction of Dr. Erwin F. Smith, a new treatment (/) of seed wheat with formalin and copper sulphate has been discovered whereby seed injury due to these disinfectants is either en- tirely eliminated or is reduced to a negUgible minimum, while at the same time the bacteria are rendered more susceptible to the action of the dis- infectant. The result has been accomplished by a correlation of two 1 The author wishes to acknowledge his indebtedness to Dr. Erwin F. Smith for helpful criticism and advice throughout the course of this investigation. ' Reference is made by number (italic) to " Literature cited," p. 392. Journal of Agricultural Research, Voi. XIX. No. 8 Washington, D. C. Ju'v '5. 1920 up Key No. G-198 (363) 364 Journal of Agricultural Research voi. xix.no.s fundamental principles of bacteriology and physico-chemistry : First, the established fact that microorganisms in an active vegetative condi- tion or just resuming activity are more susceptible to destructive agents than when in a dry or dormant state; second, the law governing the diffusion of dissolved substances whereby a solvent has a diluting effect on any solute diffusing into it from a stronger solution. EXPERIMENTAL METHODS Numerous experiments with the use of dry heat (90° to 110° C. for various periods) for the control of the blackchaff disease had indicated that it was attended either with serious seed injury or else with incom- plete control of the bacteria so as to render it unsatisfactory for field use. Exposure to 105° C. for one hour killed all the bacteria, but it also killed a very considerable part of the seeds, while in every case exposure to temperatures between 90° and 100° C. failed to kill all the bacteria. Earlier experiments with mercuric chlorid (1:1,000) and with copper sulphate (1:1,000) begun by Dr. Smith were abandoned because many seeds were killed by the mercuric chlorid and not quite all the bacteria were killed by the copper sulphate. A number of experiments to de- termine whether the formalin treatment for cereals might also be appli- cable as a means of controlling the blackchaff disease showed marked injury to the seeds but gave a fairly satisfactory control. The lesser amount of seed injury growing out of long exposures as compared with short exposures is what led to the discovery of the treatment described in the present paper. Two parallel series of experiments were carried out — one, to determine the effects on the blackchaff organism of various treatments, the other to observe the effects of the same treatments on the germination of wheat. Nine of the most widely grown wheat varieties were used in the latter series: Turkey, Fultz, Marquis, Bluestem, China, Preston, Poole, Fife, and Fulcaster, obtained through the courtesy of the Office of Cereal Investigations. Treatments for the seed-germination tests were made as follows, except as otherwise stated. Seeds counted out in sets of 100 were placed in loose cheesecloth bags and soaked thoroughly for 10 minutes in the solution to be tested. The surplus liquid was then drained off, and the seeds were placed in covered moist chambers containing several layers of filter paper previously rinsed with the same solution. After definite periods of time the seeds were removed, spread out to dry over- night, and planted the next day in flats or pots in the greenhouse. Un- treated seeds were also planted as controls. Later in the course of the work treated seeds were planted outdoors at the Arlington Experimental Farm. July IS. 1920 Presoak Method of Seed Treatment 365 The effect of the various treatments on the blackchaff organism was determined by the following method, devised by Dr. Smith. Wheat seeds in lots of 100 or more were placed in double envelopes of filter paper and sterilized by dry heat at 150° to 160° C. for three hours, in order to kill all internal and surface organisms. After they had cooled, the envelopes were opened with aseptic precautions and the seeds were thoroughly coated with blackchaff bacteria taken from 2- to 4-day-old nutrient agar or potato cultures and used as a heavily clouded bacterial suspension in sterile tubes of tap water. In this way five isolations of the blackchaff organism from as many States were tested. The seeds, after soaking in the bacterial suspension for 20 minutes, were replaced in the envelopes and allowed to dry overnight. By this method each kernel was coated with a dry film of live bacteria such as would occur on badly infected seed under natural conditions. The next day the inoculated and dried seeds were dropped into sterile test tubes containing the disinfectant to be tested. The liquid was drained off after it had acted 10 minutes. The tubes containing the seeds, which now had a thin layer of the solution around each kernel, were placed in moist cham- bers previously rinsed with the same solution. After definite periods the seeds, still moist, were replaced in the sterile envelopes to dry over- night. The next day they were transferred to nutrient agar previously determined to be suitable for the organism, in poured plates, 10 seeds per plate. Each seed was handled with forceps which had been dipped in alcohol and flamed. Control seeds which had been inoculated but not subsequently treated were also planted on this agar. After at least nine days the final records were made. The controls usually developed a typical blackchaff colony around each kernel. Treated seeds, if all the bacteria thereon were killed by the solution used, remained sterile unless contaminated by other organisms or slowly produced blackchaff colo- nies if the disinfectant had not been fully effective. This method is a good index of the effect on the bacteria of the various treatments studied. The seeds are sterilized by dry heat externally and internally without leaving any antiseptic residue such as might be left by chemical sterili- zation. The treatment under consideration is performed upon dried but live bacteria which are found on the seed coat exactly as they would occur in field practice, except that ordinarily they would be less viable and there would be fewer of them. The subsequent exposure on agar plates to optimum conditions for bacterial growth reveals the effect of the treatment, in that the bacteria on the seeds, if uninjured by the treatment, are enabled to develop characteristic colonies, the slowness of their development being a very good index of the proportion killed. Over 5,500 seeds were treated in this manner in the course of the investi- gation. 366 Journal of Agricultural Research Vol. XIX. No. 8 EXPERIMENTAL STUDIES EFFECT OF FORMALIN TREATMENT ON BACTERIUM TRANSLUCENS VAR. UNDULOSUM SMITH, JONES, AND REDDY As the bacterial blackchaff disease has often been found together with the covered smut of wheat in western fields, experiments were first per- formed to determine whether the formalin treatment for smut would at the same time control the blackchaff disease. Following the procedure above outlined, sterilized wheat seeds were inoculated with virulent isolations of the blackchaff organism and then treated for various periods with formalin i to 200 and i to 400 (i part of 36.6 per cent formalin to 200 or 400 parts sterile tap water) and finally dried and planted on agar plates. The results of treatments of over 3,000 wheat seeds in four experiments are summarized in Table I. Table I. — Effect of formalin treatment on blackchaff bacteria on wheat seeds Treatment. Percentage Total developing Percentage number of typical remammg seeds used. blackchaff colonies. sterile. 294 0.0 82.3 1,000 . I 92. 6 1,000 . 0 96.7 597 .0 84.7 320 82.6 16. 2 Percentage contami- nated with fungi or bacteria other than blackchaff. Formalin i : 200 overnight Formalin i : 400 for 3 hours Formalin i : 400 for 6 hours Formalin i : 400 for 12 hours Controls inoculated but not treated 17.7 7-3 2,- 2, The data show that the blackchaff bacteria, dried on wheat seeds as under natural conditions, can be held under control by the formalin I to 400 treatment, especially when exposed for six hours or longer. Since all the formalin used in treating the seeds had evaporated during the overnight drying, no residual solution could have been left on the seeds in the plates to prevent bacterial growth. Neither did 24 to 48 hours' drying after inoculation kill the organisms, as was shown by growth in the controls, which acted as an index of the viability of the dried organisms as well as of the suitability for bacterial growth on the part of the particular lot of media used. The conclusion is evident that absence of blackchaff bacterial growth around treated seeds was due only to the effect of the treatment. GREENHOUSE EXPERIMENTS WITH FORMALIN AND COPPER SULPHATE EPKECT OP FORMALIN TREATMENT ON GERMINATION OP WHEAT SEED Parallel with the experiments made to determine the effect on the bacteria, a series was carried out to determine the effect of the two formalin solutions, as used above, on the germination of wheat seed. The I to 200 strength formalin was not tried after the second test because July 15, 1920 Presoak Method of Seed Treatment 367 it appeared too injurious to germination to be of any practical value in the field. A total of 6,300 seeds of three varieties was treated in this series and planted in flats in the greenhouse. Records of germination, counting all seedlings above ground on the seventh day after planting for experiment I and those above ground on the ninth day for experi- ments II and III, are given in Table II. Table II. — Effect of formalin treatment on germ,ination of wheat seed Treatment. Control Formalin i : 400 for 3 hours Formalin i : 400 for 6 hours Formalin 1:400 for 12 hours Formalin 1:200 for 3 hours Formalin 1:200 for 6 hours Formalin 1:200 for 12 hoiu-s China. Average percentage of germination. Exp. Exp I. II. Exp III. 63 40 45 46 68 48 46 58 27 35 43 Bluestem. Average percentage of germination. Exp I. Exp. II. 53 26 34 33 19 17 16 Exp III. 49 27 31 37 14 15 13 Turkey. Average percentage of germination. Exp I. 56 42 48 50 23 17 34 Exp. II. Exp III. Aver- age. 65 SI 54 57 27 25 35 The preceding table shows grave injury to germination where the stronger solution was used. There was also marked retardation of germination. Formalin i to 400, while much less harmful, caused an appreciable decrease in germination as compared with the controls. It was observed, however, that the 12-hour treatment (1:400) invariably produced less retardation and loss in germination than the 3-hour treat- ment. This was repeatedly evident for each variety (PI. 69). A fourth test for Turkey wheat (part of experiment IV), using 400 seeds treated with formalin i to 400 for i hour, 400 seeds treated 12 hours, and 100 seeds untreated, gave on the sixth day after planting 45 per cent germination for the i-hour treatment, 59 per cent for the 12-hour treatment, and 62 per cent for the controls. These results, so contrary to what might have been expected, led to the experiments to be described. A search of the literature after these results had been obtained disclosed a similar condition in a number of cases not commented on by the authors — that is, less injury from long exposures than from short ones. Such a condition is found in an analysis which I have made of the data presented by a subcommittee of the War Emergency Board of Plant Pathologists (4) on the effect of formalin i to 320 acting for various periods on different cereal seeds. With wheat, in 18 tests out of 25 368 Journal of Agricultural Research voi.xix.no.s the short 2-hour treatment shows more injury than a longer treat- ment; with barley, 14 out of 19 tests show more injury from the short treatment; and so with oats in 29 out of 41 tests and with rye in 2 out of 3 tests. In some cases there is a steady decrease of injury as the treat- ment period lengthens. A study of the tables given by Stuart (p) in his paper on the effect of formalin on oat germination shows almost invariably greater injury to germination caused by 2-hour treatment than by 4-hour treatment of seeds, and a similar effect on the final yield of grain and straw — a fact not commented on by that writer. These facts led to the conjecture that the formaldehyde content in the seeds at the end of 3 hours was really stronger than in the 12-hour treated seeds. Such a condition might be explained by the hypothesis that the dry seeds absorb the formaldehyde itself more rapidly than they do the water and that by diffusion later this is diluted to a strength more like the original solution through continued absorption of water by the cell walls and cells. Theoretically, therefore, if the 3-hour treatment could be made in such a way that the final solution content of the seeds would be com- parable in amount and dilution with that finally present in 12-hour treated seeds, the effects on germination should also be similar. By impregnating the cell v.^alls and cells of dry seeds with water and then treating with formalin for three hours, it appeared as if this result might be attained; for, in accordance with the lavs^s of diffusion of dissolved substances, the formalin should be diluted as it diffused into the water- saturated seed tissues. Acting on this hypothesis, wheat seeds were first soaked in tap water for 10 minutes, drained, and kept moist for 9 hours, then soaked thorough- ly in formalin i to 400 for 10 minutes, drained, and kept moist for 3 hours in order that the water absorbed during the first 9 hours might weaken the full strength solution diffusing into the cells during the next 3 hours and might result in less injury than was caused by the 3-hour treatment of dry seeds. Following this, numerous experiments were made, varying the length of time during which the seeds were kept moist but starting out always with a preliminary short plunge into water. This method — short exposure under water followed by varying lengths of exposure to moist air (covered) — has been designated throughout this paper as the "presoak" method of seed treatment. In the same way, whenever exposure to formalin is mentioned it means always that the seeds were plunged into the formalin water solution for a short period only (usually 10 minutes) and then kept moist (covered) for the designated number of hours. The seeds of two varieties so treated were planted in the greenhouse with 12-hour treated seeds and controls. Table III records the results observed. July 15, 1920 Presoak Method of Seed Treatment 369 Table III. — Effect of g hours' presoaking followed by formalin i : 400 treatment Treatment. Percentage of germination. China. Blue- stem. Exp. V. Exp. VI. Aver- age. Mar- quis, Exp. VI. Control Formalin 1:400 for 10 minutes, covered 12 hours Formalin 1:400 for 10 minutes, covered 3 hours Water for 10 minutes, covered 9 hours, formalin 1:400 for 10 minutes, covered 3 hours 70 60 78 60 67 72 46 40 35 45 50 43 23 42 42 29 44 59 56 49 58 Table III shows for the presoaked seeds not only a decrease in injury over the 3-hour treated seeds as previously observed but also a decided increase in germination over the seeds treated 12 hours with formalin. The latter show a loss of 10 to 18 per cent, the former a loss of only 3 to 6 per cent. At the same time the presoaked seeds produced larger and more vigorous plants than the controls. Very clearly the presoak treatment followed by 3 hours of formalin treatment showed a distinct advantage over the usual formalin treatment. As this appeared to be a promising method of reducing formalin injury, a thorough test was then made with all the varieties of wheat at hand. It appeared possible that for some varieties a 9-hour presoak might begin germination before treatment, thus rendering the seeds very susceptible to injury, whereas, a formalin-treatment period longer than 3 hours appeared desirable for control purposes. For these rea- sons, the presoak period was reduced to 6 hours, followed by a 6-hour treatment with formalin. EFFECT OF 6 hours' PRESOAKING FOLLOWED BY FORMALIN I TO 40O TREATMENT ON GERMINATION OF WHEAT SEED The presoak treatment in every instance was given by soaking the seeds lo minutes in water, then draining off the surplus water and keep- ing the seeds in moist chambers 6 hours, allowing the seeds to absorb the surface moisture film during this period. After shaking off all possible surface water, the next step was to soak the seeds thoroughly in a i to 400 formalin solution, stirring and rinsing up and down to bring the solution in contact with each kernel. After 10 minutes in this solution the seeds were removed, drained, and kept moist for 6 hours in moist chambers previously rinsed with the same solution. After treatment the seeds were dried overnight and planted in the greenhouse, 100 seeds per pot, in duplicate or triplicate. 370 Journal of Agricultural Research Vol. XIX, No. 8 PERCENTAGE OF GERMINATION 30 40 50 60 70 80 Fig i.-Graph showing effect of formaUn i to 400 treatments with and without presoakmg: A. control, untreated- B seeds soaked in formalin i to 400 for 10 minutes, drained, and kept moist (covered) 12 hours, dried overnight, and planted; C, seeds soaked in water 10 minutes, dramed, and kept moist (covered) 6 hours, soaked in formalin i to 400 for 10 minutes, drained, and kept moist (covered) 6 hours, dried overnight, and planted. Records of germination were taken on the sixth day alter planting in the greenhouse. July 15, 1920 Presoak Method of Seed Treatment 371 Using a 6-hour presoak followed immediately by a 6-hour treatment with formalin i to 400, 6 experiments were carried on in the greenhouse with 23,700 seeds of nine varieties, the test for each variety being re- peated several times. Table IV and figure i show the data obtained on the sixth day after planting in each test, this date being chosen as showing not only the relative percentage of germination but particularly revealing any retardation or acceleration. At the same time, the effect of this method on the blackchaff bacteria was determined, as discussed later. Table IV. — Effect of 6 hours' presoaking followed by formalin i : 400 treatment on germination of wheat seed Average percentage of germination. Treatment. Fife. Bluestem. Preston. Exp. V. Exp. VI. Exp. VII. Aver- age. Exp. V. Exp. VI. Exp. VII. Aver- age. Exp. V. Exp. VI. Exp. VII. Aver- age. Control 89 75 77 74 66 71 75 75 86 79 72 78 83 61 74 79 54 64 82 66 81 60 79 53 66 75 51 64 77 63 82 77 56 Formalin 1:400 for 10 minutes, covered 12 hours Water for 10 minutes, covered 6 hours, forma- lin 1 : 400 for 10 minutes, covered 6 hours 82 73 71 Treatment. Control Formalin 1:400 for 10 minutes, covered 12 hours Water for 10 minutes, covered 6 hours, formalin i: 400 for lomin- utes, covered 6 hours Average percentage of germination. Marquis. Exp. V. Exp. VI. 82 54 61 Exp. VII. Aver- age. 82 62 73 China. Exp. V. 52 44 50 Exp. VI. 38 34 44 Aver- age. 45 39 47 Turkey. Exp. V. 60 60 71 Exp. VI. 61 52 50 Aver- age. 61 56 61 Treatment. Control Formalin i : 400 for 10 minutes, covered 6 hours Formalin i : 400 for 10 minutes, covered 12 hours Water for 10 minutes, covered 6 hours, forma- lin I : 400 for 10 mintites, covered 6 hours . Average percentage of germination. Poole. Fultz. • Exp. VIII. 72 Exp. IX. 80 Exp. X. Aver- age. Exp. VIII. Exp. IX. Exp. X. 92 81 55 83 74 67 85 70 74 57 70 57 65 82 79 75 42 66 62 76 89 89 88 60 78 76 71 61 57 71 372 Journal of Agriculhiral Research Vol. XIX, No. 8 Table IV. — Effect of 6 hours' presoaking followed by fonnalin I : 400 treatment on germination of wheat seed — Continued Fulcaster. Turkey. Treatment. Exp. vm. Exp. IX. 84 67 72 80 Exp. X. Aver- age. Exp. VIII. Exp. IX. Exp. X. Aver- age. Control 82 62 6S 77 79 60 69 83 82 63 70 80 4i 056 35 46 74 64 67 71 64 48 61 ^5. 59 56 54 57 Formalin i : 400 for 10 minutes, covered 6 hours Formalin 1:400 for 10 minutes, covered 12 hours Water for 10 minutes, covered 6 hours, forma- lin I : 400 for 10 minutes, covered 6 hours . . . 1 Result from one pot only, the other having been overturned. For each variety of wheat used the result is the same — a marked de- crease in retardation and injury to germination where the presoak method PERCENTAGE OF GERMINATION 20 30 40 50 60 70 80 Fig. 2. — Graph showing efiect of formalin i to 200 treatments with and without presoaking: A, control, untreated; B, seeds soaked in formalin i to 200 for 10 minutes, drained, and kept moist (covered) 6 hours, dried ovemisht, and planted; C, seeds soaked in Vt-ater 10 minutes, drained, and kept moist (covered) 6 hours, then soaked in formalin i to 200 for 10 minutes, drained, and kept moist (covered) 6 hours, dried overnight, and planted. Records of germination were taken on the sixth day after planting in the green- house. of treatment was used, as compared with the 6- or 12-hour formalin treat- ment without presoaking. In the case of the three varieties most sus- ceptible to formalin — -Bluestem, Preston, and Marquis — germination of the presoak-treated seeds is within 6 to 9 per cent of the controls, while there is a reduction of 20 to 21 per cent in the seeds treated without presoaking. The other six varieties show practically all injury elimi- nated by the presoak treatment. The relative appearance of controls, treated plants, and presoak-treated plants on the sixth day is shown in Plates 70 and 71 . A very noticeable stimulation in vigor was observed in July IS. 1920 Presoak Method of Seed Treatment 373 all of the presoak-treated seeds as compared with the controls. This is brought out also in Plate 72. EFFECT OF 6 HOURS' PRESOAKING FOLLOWED BY FORMAILN I TO 200 TREATMENT ON GERMINATION OP WHEAT SEED The beneficial effect of the presoak method is strikingly shown in an experiment where a much stronger solution of formalin was used. Seeds of three varieties were treated for 10 minutes with formalin i to 200 and were then kept moist (covered) for 6 hours, dried overnight, and planted in the greenhouse. Another set of seeds received the same treatment but were first soaked in water 10 minutes, drained, and kept moist 6 hours before receiving the formalin treatment. This strength of I to 200 had previously been found to cause a very considerable injury to germination. Table V and figure 2 show the percentage of germina- tion on the sixth day. Table V. — Effect 0/ 6 hours' presoaking followed by formalin i : 200 treatment on germi- nation of wheat seed Treatment. Average percentage of germi- nation. Poole. Turkey. Fultz. Control 80 46 71 74 42 65 83 43 70 Formalin i : 200 for 6 hours • Water for 6 hours, formalin i : 200 for 6 hours Here the 6-hour formalin i to 200 treatment reduced germination 32 to 40 per cent below that of the controls, while the reduction was only 9 to 13 per cent where the same treatment was preceded by 6 hours' water presoak (PI. 73.) EFFECT OF 6 hours' PRESOAKING FOLLOWED BY FORMALIN I TO 32O TREATMENT ON GERMINATION OF WHEAT SEED Formalin i to 320, or one pound of formalin to 40 gallons of water, was next used, since that is the strength now recommended for the cereal smuts. Besides the presoak procedure so far followed, a test was made (experiment XI) of the effect of an actual soaking in water for 5 hours, followed by thorough draining for a few minutes, then 10 minutes' soaking in the formalin i to 320, then covering for 7 hours before drying and planting. Experiment XII was conducted to test the effect of an actual soaking in water for 4 hours, followed by thorough draining for a few minutes, then soaking in formalin i to 320 for 10 minutes, drain- ing, covering 6 hours, drying, and planting. The results obtained are shown in Table VI and figure 3. 177286°— 20 4 374 Journal of Agricultural Research vo1.xix,no.8 Table VI. — Effect of 6 hours' presoaking followed by formalin i : 320 treatment on germi- nation of wheat seed under greenhouse conditions Treatment. Control Formalin i : 320 for 10 minutes, drained, covered 6 hours Formalin i : 320 for 10 minutes, drained, covered 12 hours • ■ ■ Water for 10 minutes, covered 6 hours, formalin I : 320 for 10 minutes, drained, covered 6 hours. . Actual soaking in water 5 hours, formalin i : 320 for 10 minutes, covered 7 hours Actual soaking in water 4 hours, formalin i : 320 for 10 minutes, covered 6 hours Average percentage of germination. Poole. Exp. XI. 94 66 67 90 88 Exp. XII. 85 55 71 79 Aver- age. 90 61 69 79 Fife. E-Kp. XI. 79 59 58 80 77 Exp. xn. 83 62 64 90 85 Aver- age. 81 61 61 85 77 85 10 PERCENTAGE OF GERMINATION 20 30 40 50 60 70 80 Fig. 3. — Graph showing effect of formalin i to 320 with and without presoaking: A, control, untreated; B, seeds soaked in formalin i to 320 for 10 minutes, drained, kept moist (covered) 6 hours, dried overnight, and planted; C, seeds soaked in water lo minutes, drained, kept moist (covered) 6 hours, soaked in formalin I to 320 for 10 minutes, drained, kept moist (covered) 6 hours, dried overnight, and planted. Records of germination were taken on the sixth day after planting in the greenhouse. A marked retardation and diminished germination is shown by both formalin treatments without presoaking. The 6-hour formaUn treatment preceded by 6 hours' presoaking yielded plants similar to the controls in percentage of germination, and they were very evidently stimulated, as shown in Plates 74, 75, and 76. Actual soaking in water did not appear to be so beneficial to the vigor of the seedlings as the procedure of merely keeping the seeds moist for 6 hours before treating with formalin. (See PI. 75, fig. 3.) July 15, 1920 Presoak Method of Seed Treatment 375 EFFECT OF PRESOAKING WHEN USED WITH COPPER SULPHATE FOR WHEAT AND BARLEY SEED The striking reduction in formalin injury to seed germination when the presoak method was used led to trials of this method in conjunction with copper sulphate. Six hundred wheat seeds of Fife and Fulcaster varieties were soaked in a very strong copper-sulphate solution (i :8o, or I pound to 10 gallons of water) for 10 minutes, drained 20 minutes, dipped for a moment in milk of lime, and dried. A like number of seeds received the same treatment except that they were kept moist for 8 hours after PERCENTAGE OF GERMINATION 30 40 50 60 70 80 90 100 FULTZ ^TENNESSEE WINTER ■ZZI (BARLEY) Fig. 4. — Graph showing effect of copper sulphate i to 80, with and without presoaking, on wheat and barley seed germination: A, control, untreated; B, seeds soaked in copper sulphate i to 80 for 10 minutes, drained, kept moist 20 minutes, then limed, dried overnight, and planted ;C, seeds first soaked in water 10 minutes, drained, and kept moist (covered) for 8 hours, then treated as in B; D, seeds first soaked in water 10 minutes, drained, kept moist (covered) for 6 hours, then limed, dried overnight and planted. Records of germination were taken on the fifth and seventh days after planting in the greenhouse. soaking lo minutes in tap water. Four hundred seeds were used as con- trols. All seeds were planted in the greenhouse after drying overnight. The experiment was later repeated, using barley also and wheat seeds of Fultz and Marquis varieties, kept moist in the manner described for 6 hours before treatment. The photographs (PI. 76, jy) and figure 4 show the results obtained. In these experiments the injury produced by the copper-sulphate treatment was prevented by the use of the 6- or 8-hour presoak. The 6-hour presoak appears preferable, because a longer period, by starting 376 Journal of Agricultural Research Vol. XIX, No. 8 seed germination, may render the seed unusually susceptible to the sub- sequent copper-sulphate treatment and thus defeat its purpose. A marked increase in the percentage of germination was observed in the presoak-treated seeds over the controls. This was probably due not only to the lack of injury in the former but to the residual effect of the copper sulphate and lime, which, by preventing seed infection through soil organisms, enabled more seeds to germinate. There was also a marked stimulating effect on the growth of the seedlings. The use of the presoak method of treatment also reduces copper- sulphate injury in barley, as shown in Plate 76, figure 3. The fact that the presoak method can reduce seed injury from formalin and copper sul- phate, two disinfectants of widely different chemical nature, suggests the possibility of its use in conjunction with mercuric chlorid also, another commonly used seed disinfectant. FIELD EXPERIMENTS WITH FORMALIN AND COPPER SULPHATE EPPECT OF 6 hours' PRESOAKING FOLLOWED BY FORMALIN I TO 320 TREATMENT ON GERMINATION OP WHEAT SEED UNDER FIELD CONDITIONS That formalin injury to germination can be greatly decreased under field conditions when the presoak method is used is shown by the follow- ing experiment. Using seven wheat varieties, 16,800 seeds were planted on a uniform level plot at the Arlington Experimental Farm. For each variety 1 2 rows of 200 seeds each were distributed as follows : four rows of controls, four rows of seeds treated with formalin i to 320 for 6 hours, and four rows of seeds similarly treated but presoaked 6 hours. The results were striking. In each variety the central four rows, which received the usual treatment recommended for smut, showed a marked decrease in germination (27 to 53 per cent, averaging 38 per cent for the seven plots) while in each case the four rows receiving the presoak formalin treatment scarcely differed in appearance from the controls (PI. 78). Table VII and figure 5 give the data obtained. Table VII. — Effect of 6 hours' presoaking followed by formalin 1:320 treatment on germination of wheat seed under field conditions Average percentage of germination. Treatment. Fulcaster • Fife. Turkey. Poole. China. Bluestem . Marquis. loth day. 20th day. loth day. 20th day. loth day. 20th day. loth day. 20th day. loth day. 20th day. loth day. 20th day. loth day. 20th day. Control 60.0 321 52s 66.3 36.8 56.4 58.4 310 56.2 59-6 41.8 S8.3 40.6 24.2 40.8 S0.4 28.6 SI-2 60.2 20.4 53-5 643 29.8 58.3 26.1 14.S 30.8 i2.6 21. 1 3S-4 62.2 32-S 43-6 63.0 445 53.6 so. 2 24.2 46.8 53-9 39- 1 Si-6 Formalin 1:320 for 10 min- utes, covered 6 hours Presoaked in water, 10 min- utes, kept moist 6 hours, formalin 1:320 for 10 min- utes, covered 6 hours July IS, 1920 Presoak Method of Seed Treatment 377 The results obtained in the field fully corroborate the greenhouse experiments as to the beneficial effect of the presoak method and show that in actual field practice wheat seed injury caused by the formalin treatment recommended for covered smut can be practically eliminated by allowing the seeds to absorb water in the manner prescribed for six hours PERCENTAGE OF GERMINATION 10 20 30 40 50 60 70 :b= FIFE ApHPfHHMI TURKEY sp— BM POOLE B^BH^^H FULCASTER ^H^^ CHINA b|— [^ ^ A IHBBHBBD B^BBBH^H BLUESTEM ■ I MARQUIS cJHHHHlil AH^^tt^H b|— BHB BtaHHHBiB AVERAGE 1 cIbbhbh Fig. s- — Graph showing effect of formalin i to 320, with and without presoaking, on wheat seed germination under field conditions: A, control, untreated; B, seeds soaked in formalin i to 32ofor lominutes, drained, kept moist (covered) 6 hours, dried overnight , and planted; C, seeds soaked in water 10 minutes, drained, kept moist (covered) 6 hours, soaked in fonnalin i to 320 for 10 minutes, drained, kept moist (covered) 6 hours, dried overnight, and planted. before receiving this treatment. The effect of tne presoak method of treatment in also eliminating retardation of germination is an important factor in preventing the attack of soil fungi on seeds or very young seed- lings unduly delayed in germination. Plate 78 shows the appearance of the field plots on the sixteenth day after planting. 378 Journal of Agricultural Research Vol. XIX, No. 8 EFFECT OF SIX HOURS' PRESOAKING FOLLOWED BY FORMALIN AND COPPER-SULPHATE TREATMENTS ON HALF-BUSHEL LOTS OF WHEAT SEED The next series of experiments was made to determine (i) the effect of the presoak method of treatment, using wheat seed in half -bushel lots 10 PERCENTAGE OF GERMINATION 20 30 40 60 60 ?0 80 CHINA DIETZ ■^^^^9^^ CURRELL Fio 6 -Graph showing effect of fonnalin and copper-sulphate presoak treatments of J4-bushel wheat seed lots- A control, untreated; B. seeds soaked in fonnalin i to 3^0 for 10 minutes, drained, covered 6 hours, dried and planted; C, seeds soaked in water 10 minutes, drained, covered 6 hours, then treated with for- maUn as in B, and seeds from upper one-fourth planted; D, seeds from central part of same lot asC; E average germination of C and D; F. seeds soaked in copper sulphate i to 80 for V:. hour, followed by milk of lime and dried; G, seeds soaked in water 10 minutes, drained, covered 6 hours, then treated with copper sulphate as in F, and seeds from upper one-fourth planted; H. seeds from central part of same lot as G ; I , average germination of G and H . In the Currell variety the same procedure was used except that the seeds were sprinkled instead of soaked. as in practical usage; (2) the relative effects of soaking and sprinkling; (3) the result of a possible lack of aeration and accumulation of carbon dioxid in the center of the presoak-treated mass of seeds. juiyi5. I920 Presoak Method of Seed Treatment 379 A half bushel each of Dietz and China wheats in bushel bags were soaked in water 10 minutes, drained, and covered 6 hours, then soaked in formalin i to 320 for 10 minutes, drained, and covered 6 hours. A similar treatment was made on like quantities of seed, using copper sul- phate I to 80 for K hour after the 6 hours' presoaking, followed by milk of lime. The upper and central parts of each half bushel were dried, and each of the eight lots thus obtained was planted in 10 rows of 300 seeds each. Suitable controls and seeds treated without presoaking were also planted. At the same time, a bushel of Currell wheat, piled up on canvas, was sprinkled with water and covered for 6 hours. Half of this was then sprinkled with fonnaUn i to 320 and covered for 6 hours. The other half was sprinkled \vith copper sulphate i to 80, covered X hour, and Umed. Seeds from top and center were dried and planted, along with controls and seeds sprinkled without presoaking. The results obtained a month after planting are recorded in figure 6. The presoak-treated seeds again showed a marked improvement in ger- mination over seeds treated without presoaking. Seeds from the center of the bag are apparently affected to some extent, probably through lack of aeration and the accumulation of carbon dioxid ; but the average ger- mination of the presoak-treated seeds is better than that of the seeds treated without presoaking. The sprinkling method at first sight appears to possess a distinct ad- vantage over the soaking method. In the former, compared to controls, there is a marked increase in germination of seeds sprinkled first with water and after six hours with the disinfectant. This is most probably due to the incompleteness of the sprinkling method, since the disin- fectant can not reach each kernel as in the soaking method. Hence a large number of seeds, affected only by the water vapor of the prelim- inary sprinkling, receive the stimulation due merely to the absorption of water and drying before planting. EFFECT OF PRESOAK METHOD ON BACTERIUM TRANSLUCENS VAR. UNDU- LOSUM EFFECT OF SIX hours' PRESOAKING FOLLOWED BY FORMALIN I TO 400 TREATMENT ON BLACKCHAFF BACTERIA ON SEEDS PLANTED ON NUTRIENT AGAR A series of experiments was carried out parallel with the germination experiments, using 2,360 seeds, to determine whether the 6-hour formalin treatment when preceded by a 6-hour presoak would destroy or prevent the growth of the blackchaff bacteria. Heat-sterilized wheat seeds were heavily inoculated, using four virulent isolations of the blackchaff organism, and dried overnight as before described. The next day the seeds were placed in tubes of sterile tap water, which was drained off after 10 minutes. The tubes containing the seeds were placed in moist 38o Journal of Agricultural Research Vol. XIX. No. g chambers for 6 hours to maintain a thin film of moisture around each kernel throughout this period. The subsequent treament with a formaUn solution, I to 400, in sterile tap water was made in the same manner — that is, by pouring the formahn solution on the seeds, draining if off after 10 minutes and placing the tubes in moist chambers rinsed with a formalin I to 400 solution. After 6 hours' treatment the seeds were replaced in sterile envelopes, dried overnight, and planted on agar plates. Control seeds, inoculated but not treated, were also planted. Observations were made after 9 to 15 days and are recorded in Table VIII. Table VIII. — Effect of 6 hours' presoaking followed byformalin i : 400 treatment on black- chaff bacteria on seeds placed on nutrient agar Experiment No. Treatment. Number of seeds used. Percent- age de- veloping typical black- chaff colonies. Percent- age ster- ile. Percent- age con- tami- nated with fimgi or bacteria other than black- chaff. I. Inoculated seeds presoaked 6 hours, for- malin 1 1400 for 10 minutes, covered 6 hours 400 160 400 200 800 400 I, 600 760 0. 0 72. 0 . 0 77.0 . 2 98.7 . I 86.6 65-5 19-3 87-3 2. 0 97-7 1-3 87.0 5-9 ^4- "; II. Controls inoculated but not treated Inoculated seeds presoaked 6 hours, for- malin 1:400 for 10 minutes, covered 6 hours 8.7 12.7 21. 0 2. 1 III. Controls inoculated but not treated Inoculated seeds presoaked 6 hotus, for- malin 1:400 for 10 minutes, covered 6 hovirs Summary . Controls inoculated but not treated Inoculated seeds presoaked 6 haurs, for- malin 1:400 for 10 minutes, covered 6 hours . . .0 12. 9 7-5 Controls inoculated but not treated Only 2 out of 1,600 inoculated seeds treated by the presoak method developed typical blackchaff colonies. The controls, which had been inoculated and dried two days, showed 86.6 per cent of the kernels developing the typical colonies when planted on the nutrient agar, thus demonstrating that the absence of growth in the treated seeds was due not to drying of the bacteria but to the treatment as practiced (PI. 79, 80). The presoaking, then, while limiting to a striking degree retardation of seed growth and loss due to failure to germinate, does not reduce the effectiveness of the subsequent formalin treatment as a means of treating diseased seed. In fact, it tends to increase its efficiency in this respect, as will be brought out in the discussion. July IS, 1920 Presoak Method of Seed Treatment 381 EFFECT OF SIX HOURS PRESOAKING FOLLOWED BY FORMALIN I TO 320 AND COPPER SULPHATE I TO 80 ON ARTIFICIALLY INFECTED WHEAT SEEDS PLANTED IN THE SOIL UNDER FIELD CONDITIONS Several field tests were made v/ith seeds of Currell, China, and Dietz varieties artificially infected with blackchaff bacteiia, dried, and treated with formalin i to 320 and copper sulphate i to 80, after a presoaking of six hours. The percentages of infection in the young seedlings two to three weeks after planting are given in Table IX. Table IX. — Effect of presoak method of treatment on inoculated wheat seeds under field conditions Experiment of Aug. 9. Experiment of Aug. 22. Experiment of Sept. 7. Treatment. Variety of wheat used. Isola- tion num- ber of Bad. trans- lucens vai. undti- losum. Per- cent- age of infec- tion of seed- lings. Variety of wheat used. Isola- tion num- ber of Bad. trans- lucens var. undu- losum. Per- cent- age of infec- tion of seed- lings. Variety of wheat used. Isola- tion num- ber of Bad. irans- lucens var. undu- losum. Per- cent- age of infec- tion of seed- lings. Seeds soaked in bacterial suspen- [Currell. sion 20 minutes, dried, plantedXChina. . at same time as treated seeds : 1 Dietz. . <»8so C318 '^394 850 318 394 850 318 394 92.6 Currell. go. I 1 China . . 95. 1 1 Dietz . . 0271-A 394 850 271-A 394 850 271-A 394 8co 85- S 91.0 80.3 •7 •4 1 Currell. [6 213 China. . 271-A 68.9 82.0 below. Seeds soaked in bacterial suspen- sion 20 minutes, dried over- night, soaked in water 10 min- utes, covered 6 hours, soaked in formalin i;32o for 10 min- utes, covered 6 hours, dried, and planted. Seeds soaked in bacterial suspen- sion 20 minutes, dried over- night, soaked in water 10 min- utes, covered 10 hours, soaked in copper sulphate 1:80 for Vi Currell. China . . Dietz . . Currell. China.. Dietr . . 1. 0 •5 •3 .8 I- 5 ■9 Currell. China. . Dietz . . Currell. China . . Dietz . . Currell. China. . 213 271-A .0 •S •9 ■3 CurreU. China . . 213 271-A .8 .6 hour, then in milk of lime 1:80 a moment, dried, and planted. " From Kansas. '' From Colorado. <■ From Montana. Bacterial infection was prevented to a very marked degree in the treated seeds. The controls, infected with five isolations of Bacterium translticens var. u^ndulosum from different localities, showed from 69 to 95 per cent infection, considerably more than would usually occur in naturally diseased seed, owing to the heavy artificial inoculation and brief drying period before planting. Infection of these seeds, heavily inoculated as they were, v/as reduced from o to 1.5 per cent by the pre- soak method, used with both formalin and copper sulphate. The appli- cation of this method, then, to the control of blackchaff on the farm is evident. The only doubt that can be entertained is in cases where the bacteria have penetrated the seed coats. Fortunately, in most such cases at least, the seeds are more or less shriveled and of light weight so that they may be screened out in advance of treatment. 382 Journal of Agricultural Research Vol. XIX, No. 8 RESULTS OF PRESOAK TREATMENTS ON NATURALLY INFECTED WINTER WHEAT PLANTED IN THE WHEAT FIELDS OF IOWA AND KANSAS The first extensive field trial of this method was made in 191 9 at three places in the middle western wheat belt — Ames, Iowa, and Hays and Abilene, Kans.,* where Kharkoff and Kanred from infected fields, screened and unscreened, was treated and drilled in after two to nine days' drying. The treatments used were (i) presoak copper-sulphate treatment, in which seeds were soaked 10 minutes in water, covered 6 hours, soaked K hour in copper sulphate i to 80, limed, dried, and planted; (2) presoak formalin treatment, in which the seeds were soaked 10 minutes in water, covered 6 hours, soaked 10 minutes in formaUn I to 320, covered 6 hours, dried, and planted. Notes on the amount of infection on the seedlings were first made four to seven weeks after planting, since infection at this time would represent mostly primary infections due to diseased seed, before general dissemination from infec- tion centers could set in. The results are summarized in Table X. Table X. — Preliminary results of presoak treatments of infected winter wheat in the Middle West, igiQ Locality. Wheat variety. Treatment. Size of plot. Date treated, 1919. Date planted, 1919- Date ob- served, 1919. Num- ber of plants exam- ined. Num- ber of plants in- fected. Per- cent- age of infec- tion. Kanred «. ..do ..do Kharkoff. Kanred. . Kharkoff. Kanred . . ..do ..do No treatment, un- screened. Presoak copper sulphate treat- ment,unscreened Presoak formalin treatment, un- screened No t r e a tment, screened. do Sept. 27 ...do.. . ...do... Sept. 34 ...do.. . Nov. 7 ...do... ...do... Nov. 10 ...do.. . ...do... ...do... Nov. 9 ...do... 3" 360 343 318 330 3" 331 300 360 38 3 0 17 33 0 X 33 eo 8.9 Ames, Iowa . . ...do.. ,. ...do.. .. Sept. 18 ...do.. . .6 . 0 7.8 ...do 6... Hays, Kans . . Presoak formalin treatment screened. ..do ...do.. .. ..do. Sept. 2a ...do.. ...do.. . ...do... Oct. 6 to 10 ...do... . 0 •3 II. s .0 No t r e a tment, screened. Presoak formalin treatment screened. 45 acres.. Abilene, Kans isacres.. Oct. I o A very susceptible variety. * Part of main field. e One diseased plant was found; but, judged by its advanced stage of growth, it was a volunteer and was therefore not from the treated seeds. The seedlings at the time of observation bore two to five leaves, with infection visible on the first leaf of diseased plants verified by micro- scopic examination. In untreated areas, from 7.8 to 11. 5 per cent infec- tion was present; in treated areas, from o to 0.6 per cent, as shown in ' The author is indebted to Dr. I. E. Melhus, at Ames, and to Mr. Swanson, at Hays, for cooperation and assistance at these localities. July IS, 1920 Presoak Method of Seed Treatment 383 the table, indicating so far a satisfactory degree of control under actual field conditions through the use of the presoak method as formulated. Plants collected from the western experimental plots March 27 to April 3, 1920, were examined microscopically for the presence of oozing bacteria in suspected blackchaff lesions and in dead leaves. Platings were made later from similar lesions where abundant oozing bacteria were found, and these developed the typical blackchaff colonies. Table XI summarizes the results obtained. Table XI. — Condition of western experimental plots in the spring of ig20 Num- Locality. Treatment. Date ob- served, 1920. Num- ber of plants ber with bacteria oozing Per- centage of in- exam- ined. from cut sec- tions. fection. Kanred wheat, untreated, 45 Mar. 27 212 46 21.7 acres. Presoak formalin treated plot,o ...do 185 9 4.8 15 acres; plants collected from Abilene, Kans half of plot farthest from im- treated area. Same treated plot; plants col- ...do 196 12 6.1 lected from other half, near . untreated area. Kharkoff wheat, untreated Mar. 30 141 24 17.0 Kharkoff wheat, presoak forma- ...do 231 7 30 Hays, Kans lin treated plot. a Kanred wheat, untreated ...do 102 49 25. 5 Kanred wheat, presoak formalin ...do 7 218 9 4.1 treated plot.a Kanred wheat, untreated Apr. 3 285 S3 18.5 Kanred wheat, presoak formalin ...do... . 221 6 2.7 Ames, Iowa treated plot.* Kanred wheat, presoak cooper- ...do 236 5 2. I sulphate treated plot.& "Seeds presoaked 6 hours, then treated with formalin i to 330 for 10 minutes, drained, covered 6 hours and dried. * Seeds presoaked 6 hours, then treated with copper sulphate i to 80 for half hour (soaked), and dried after dipping a moment in milk of lime. A marked increase in blackchaff was observed in the untreated plots, evidently due to wind and rain spreading the disease during the resump- tion of growth. The spreading effect was especially noted in the Abilene plots where the treated area adjoins the untreated. Other treated areas at Hays and Ames, more isolated, show from 2 to 4 per cent of infection, as compared with 17 to 25.5 per cent in untreated plots. Observations made at Hays and Abilene, Kans., in the latter part of May, 1920, are summarized in Table XII. At Abilene the plants were I to 2 feet tall, at Hays over 2 feet high. Heads had not yet emerged. Microscopic examination and confirmation of diagnoses by platings were 384 Journal of Agricultural Research Vol. XIX, No. 8 made as previously indicated. Typical yellow colonies of the blackchaff organism, concentrically striated by oblique light, were readily obtained in poured plates from blackchaff leaf lesions, which at this stage appeared characteristically as brown, water-soaked linear areas, narrow and ex- tending for various lengths along the edges or centers of the second, third, or fourth leaves from the top, with clouds of oozing bacteria in cut sections. Septoria was also found in the oldest leaves, distinguished by wider lesions and characteristic black dots of pycnidia. Table XII. — Condition of experimental plots in May, ig20 Locality and date. Abilene, Kans., May 15, 1920. Hays, Kans., May 22, 1920. Plot. Number of plants examined IKanred, untreated Kanred, western area, presoak formalin treated. IKanred, untreated Kanred, presoak formalin treated Kharkoff, untreated Kharkoff, presoak formalin treated. .. . 312 386 293 356 266 298 Number with bacteria oozing from cut sections. 87 26 33 72 Percent- age of fection. 27.8 6.7 33- 9- 27. 7- There is an evident increase in the amount of secondary infection during April and May. It was also observed that most of the lesions on leaves from treated plots were small, 2 to 8 mm. long, consisting of from I to 3 spots on the second or third leaf from the top, and were evidently fairly recent infections. Leaves from control plots showed similar lesions but also a larger proportion of more advanced lesions up to 30 mm. long on the older leaves. No lesions were observed in the young heads, which were still inclosed in the sheath. EFFECT OP MODIFYING THE TREATMENT PERIODS UNDER FIELD CONDITIONS A shortening of the entire treatment period appeared desirable after field experience with the method so far described, mainly for the purpose of facilitating drying after treatment. Three greenhouse germination experiments were made with Currell wheat seed, using a shorter presoak time and a longer (varying) soak in formalin i to 320, followed by very short periods during which the seeds were kept moist, the entire process covering various periods from 5X to 8 hours as outlined in figure 7. Formalin treatments involving a soaking longer than previously used — that is, of 15 to 30 minutes — followed by immediate drying or in- volving a short moist period of i to 3 hours decreased the germination considerably. The same treatments preceded by a 5-hour presoak (in one case a 6-hour presoak) resulted in no injury whatever to germination and in fact caused distinct acceleration. The effect of the various periods upon germination was determined for infected wheat seed also. July 15, 1920 Presoak Method of Seed Treatment 385 Currell wheat seed was inoculated with a bacterial suspension of Bac- terium translucens var. undulosum, isolation No. 850, treated with forma- lin for the periods given in Table XIII, planted outdoors, and the percent- age of infected seedlings determined 27 days after planting. PERCENTAGE OF GERMINATION JO 20 30 40 50 60 70 80 Fio. 7. — Graph showing effect of formalin i to 320 treatments for various periods, with and without pre- soaking: A, control, tmtreated; B, seeds soaked in formalin i to 320 for 15 minutes drained, covered 2 hours, dried overnight, and planted ;C, seeds soaked in water 10 minutes, covered 5 hours, then treated with formalin as in B; D , seeds soaked in formalin i to 320 for 15 minutes, drained, covered 3 hours, and dried overnight; E, seeds soaked in water 10 minutes, covered 5 hours, then treated with formalin as in D ; F, seeds soaked in formalin i to 320 for 30 minutes, drained, covered i hour, and dried overnight; G, seeds soaked in water 10 minutes, covered 5 hours, then treated with formahn as in F; H, seeds soaked in formalin i to 320 for 30 minutes, drained, and dried overnight; I, seeds soaked in water 10 minutes, covered 5 hours, then treated with formalin as in H; J, seeds soaked in water 10 minutes, covered 6 hours, then treated wit h formalin as in H. Records of germination were made on the sixth day after planting and are the averages of three experiments. Table XIII. 'Effect of shortened presoak method of treatment on infected wheat seed under field conditions Treatment. Inoculated seeds dried and planted Vi'ithout further treat- ment Inoculated seeds soaked in water 10 minutes, drained, covered 5 hours, soaked in formalin i : 320 for 15 minutes, covered 2 hours, dried, and planted Inoculated seeds soaked in water 10 minutes, drained, covered 5 hours, soaked in formalin 1:320 for 15 minutes, covered 3 hours, dried, and planted Inoculated seeds soaked in water 10 minutes, drained, covered 5 hours, soaked in formalin i : 320 for 30 minutes, covered i hour, dried, and planted Inoculated seeds soaked in water 10 minutes, drained, covered 5 hours, soaked in formalin i : 320 for 30 minutes, dried , and planted Inoculated seeds soaked in water 10 minutes, drained, covered 6 hours, soaked in formalin i : 320 for 30 minutes, dried, and planted Number of plants exam- ined. 215 196 231 211 246 Number of plants infected. 40 Percent- age of infection after 27 days. 386 Journal of Agricultural Research voi. xix, no. s The best control in this experiment was obtained with a presoak of 5 hours, followed by 30 minutes' formalin soak, then covering i hour. Such a process, requiring 6^ hours in all, would be particularl)^ desirable because of the ease in subsequently drying the seeds by spreading them in the sun on the day of treatment, Further field repetition of this experiment, which was suspended at this time by the advent of winter, PERCENTAGE OF GERMINATION 40 50 60 70 80 90 Fig. 8.— Graph showing effect of formalin i to 320 and i to 200 on germination of corn, barley, and oats with and without presoaking: A, controls ; B, seeds soaked in formalin i to 320 for 10 minutes, drained, kept moist (covered) 6 hours, dried overnight, and planted; C, seeds soaked in water lominutes, drained, and kept moist (covered) 6 hours, then soaked in formalin i to 320 for 10 minutes, drained, and kept moist (covered) 6 hours, dried overnight, and planted; D and G, controls; E and H, seeds soaked in formalin I to 20ofor lominutes, drained, kept moist (covered) 4 hours, dried overnight, and planted; F, seeds soaked in water 10 minutes, drained, kept moist (covered) 10 hours, then soaked in formalin i to 200 for 10 minutes, drained, kept moist (covered) 4 hours, dried overnight, and planted; I, seeds soaked in water lohours, drained thoroughly a few minutes, then soaked in formalin i to 200 for 10 minutes, drained, kept moist (covered) 4 hours, dried overnight, and planted. Records of the germination were made on the fifth and seventh days after planting in the greenhouse.. is necessary, however, before definite recommendations on this modifi- cation can be made. EFFECT OF THE PRESOAK METHOD ON OTHER CEREAI^S The uniform results obtained on nine different varieties of wheat by the presoak method of treatment with formalin and copper sulphate and a consideration of the underlying principles governing its salutary action, as will be discussed later, suggested that it might be generalized for the treatment of all seed-transmitted diseases of economic importance amen- July 15, tgio Presoak Method of Seed Treatment 387 able to control by formalin and copper sulphate. So far, this method has been tested on the germination of oats, barley, and maize with results similar to those obtained for wheat. Copper-sulphate injury can be prevented for Tennessee winter barley as shown above. The results thus far obtained with the presoak method of treatment, using formahn I to 320 on oats and barley and formalin i to 200 on maize, are given in figure 8. Oats and barley were soaked in water 10 minutes, drained, and kept moist 6 hours, then soaked in formalin i to 320 for 10 minutes, and covered 6 hours. Maize, which absorbs water much more slowly than wheat, oats, or barley and is also less susceptible to formalin injury, was given a lo-hour presoak — that is, 10 minutes in water, draining and covering for 10 hours, followed by 4 hours' formalin i to 200 treatment. so g40 S . as $ ID z olO HOURS 1 2 3 4 5 6 7 8 9 10 11 12 Fig. 9. — Curve showing rate of absorption of water by dry wheat seeds. In another case it was given a lo-hour actual soaking in water, followed by the treatment. The presoak method is evidently applicable to these cereals also, as a means of preventing seed injury due to disinfectants (Pi. 81, 82); and the possibility of its application for other kinds of seeds is obvious. GENERAL DISCUSSION As a result of these experiments several facts stand out clearly. First, in all cases, with each variety of wheat, barley, oats, and maize tested, the presoak method minimized or eliminated the injury to seed germina- tion due to the use of formalin and copper sulphate. Second, as shown in the illustrations, a marked stimulation of growth was usually produced. Third, the presoak method proved fully efficient as a means of destroying or preventing the growth of the bacteria of the blackchaff disease borne on the seed and can undoubtedly be apphed for the prevention of other diseases. Fourth, the method is simple and adapted to field conditions, since any farmer can apply it. The cause of the first-mentioned eflfect of the presoak treatment may be partly accounted for upon an analysis of figure 9. This represents the rate of absorption of water by 10 gm. each of dry seeds of two wheat V\3\ R =^^^^-~" " ^ ^ ^ " f\i\ v^ / r / 388 Journal of Agricultural Research voi. xix. no. s varieties, soaked lo minutes in water, drained, and kept moist for 12 hours, with periodical weighings after blotting ofif all surface water each time. The excess weight above 10 gm. represents the amount of water absorbed. The curve rises rapidly in the first 3 hours, then slows down somewhat to a more gradual rise. At the end of 6 hours, about 30 per cent by weight of water has been absorbed; during the next 6 hours, about 10 per cent more is absorbed. The 6 hours' presoaking, as prac- ticed, consequently impregnates the cell walls and cells of the seed with water, increasing the size and adding about 30 per cent by weight to the seeds. The next 6 hours' treatment with water containing formalin in solution adds only one-third as much more; and the formalin solution as it diffuses into the seeds is consequently greatly diluted by the amount of water already present in the tissues. Moreover, the amount of formalin solution which can enter the tissues in the 6 hours after presoaking is only one-third of what enters during a 6-hour formalin treatment without presoaking. Should the subsequent formalin treat- ment last much longer than 6 hours, an equilibrium would finally be estab- lished between the strength of solution within the seeds and that on the surface, resulting in both cases in a solution weaker than the original, in accordance with the laws of diffusion of dissolved substances. Removing the presoaked seeds after 6 hours' formalin treatment leaves them with a solution content considerably more dilute than that finally present in air-dry seeds directly treated with the full-strength solution; conse- quently the weakened solution within the presoaked seeds resulted in a very marked decrease in seed injury, as observed throughout the experiments. As for the stimulation observed in presoaked seeds, this may be due partly or wholly to the well-known stimulating effect of a toxic agent in minimum dose, such as would finally be present in the presoaked seeds. In considering the effect of the presoak method on the blackchaff bacteria on the seed coat, the dominant factor involved is the established principle that microorganisms in an active vegetative condition are more susceptible to the action of destructive agents than when dormant. Presoaking the seeds, and consequently the bacteria on them, for a period of six hours at room temperature causes the bacteria to begin to resume vegetative activity before seed germination commences, because the moisture and temperature conditions are ample for bacterial growth and division to begin during this period. Subsequently exposed to the di- rect action of the fonnalin solution applied full strength to the surface of the seeds, the bacteria are naturally much more susceptible to destruc- tion in this active condition. As a result, the disinfectant must act with greater efficiency than in the usual treatment, where it acts on dried and dormant bacteria. The six hours' presoak, on the other hand, is not sufficient to cause wheat-seed germination, which would produce a condition extremely susceptible to formalin injury. July IS, 1920 Presoak Method of Seed Treatment 389 The^ method of treatment discussed has, therefore, a two-fold advan- tage. On the one hand, wheat-seed injury due to the use of formalin and copper sulphate is eliminated or reduced to a minimum. On the other hand, the blackchaff organisms on the seed coats are rendered particu- larly sensitive to the action of the disinfectant by being previously brought into a vegetative condition. The same physiological principles discussed above should hold true for the general problem of seed treatment for various seed-borne patho- gens. The consistency of the results obtained by this method with nine varieties of wheat and with other cereals, using formalin and copper sulphate, indicates the possibility of the use of the presoak method with other kinds of seeds as a means of minimizing or preventing seed disin- fectant injury. Similarly, other pathogenic organisms, bacteria, or fungus spores, may be stimulated by the presoak method into increased susceptibility to the disinfectant. The presoak method of seed treatment with chemical disinfectants may be formulated for general purposes as consisting of two parts: First, the presoak period, in which seeds are soaked in water for 10 min- utes, drained, and kept covered and moist for a definite period of time, which is 6 hours for wheat, barley, and oats and 10 to 18 hours for maize — in no case sufficient to begin seed germination; second, the disinfectant- treatment period immediately following, in which the disinfectant is applied exactly as now practiced. The relative time of the presoak and subsequent treatment for other diseases, probably varying with each kind of seed and pathogen, is dependent on the following factors: (i) Susceptibility of the kind of seed used to the disinfectant. (2) Susceptibility of the pathogen to the disinfectant. (3) Rate ®f absorption of water by the seeds. (4) Time at which seed germination begins. (5) Time at which vegetative activity of the pathogen begins. A proper balance of these factors must be obtained, such that the optimum seed germination and the optimum germicidal efficiency are secured, as reported for the blackchaff disease of wheat. The length of the presoak period should not exceed half or two-thirds of the period necessary for seed germination to begin, since germination before treatment with the disinfectant would result in extreme sensi- tiveness to injury. On the other hand, the pathogen, especially if bac- terial in nature, usually has a much shorter germination period, which should come within the limit of the time of presoak and thus render it susceptible long before the seed has begun to germinate. The period necessary for the absorption of about 30 per cent by weight of water appears to be sufficient, and in the case of the cereals so far tried seems to counteract disinfectant injury. In wheat, oats, and barley this is five to six hours. The length of time necessarj' for other kinds of seeds to absorb about 30 per cent of water is suggested as the presoak period when not conflicting with the other factors involved. 177286°— 20 5 390 Journal of Agricultural Research voi. xix. no.s Actual soaking in water throughout the presoak period does not appear to be so favorable for wheat-seed treatment as the procedure of soaking lo minutes in water and merely keeping moist for 6 hours. For use in farm practice this method does not involve any radical change in present procedure other than to keep the seeds moist for a definite time before treating. In controlling the blackchaff disease of wheat, seeds should first be screened to remove shriveled grain. Then the seeds in sacks or bags, in quantities of not more than ^ bushel each, can be soaked early in the morning in water for lo minutes, drained, and set away in the bags while moist. Six hours later, at about noon, the seeds should be thoroughly soaked for lo minutes in a formalin solution of I pound to 40 or 50 gallons of water, drained, and left in the bags for 6 hours. In the evening the seeds should be spread out to dry overnight and are ready for planting the next morning. The use of formaldehyde vapor recently proposed by Thomas {10) for seed treatment, while eminently suitable for the disinfection of small seed lots which are not to be planted immediately, is open to the serious objection of lack of penetration throughout the seed mass and is not so well adapted as the presoak method for the treatment of seeds in large masses in farm practice. His experiments indicate that the vapor, while efficient on surface seeds, does not reach seeds at a depth of K inch, so that these remain as badly contaminated as untreated controls. In the presoak method, every seed is surrounded by a film of the disinfectant acting on the pathogens which previously have been brought into a vege- tative condition by the long exposure to moisture at room temperature. The presoak method used with copper sulphate, if efficient for con- trolling the cereal smuts, ^ would be particularly adapted for the grain sections of the Northwest. Extensive soil infection in this area renders the use of copper sulphate preferable to formalin because of its residual germicidal effect; and, as here shown, copper-sulphate injury may be prevented by a 6-hour presoak. The general application of the presoak method, extremely simple in itself, to the formalin and copper-sulphate treatments of the cereal diseases amenable to control by seed disinfection should, if the results here recorded are confirmed for other diseases by subsequent careful 1 A paper by Heald {2), first brought to the writer's attention in Nov. 14, 1919, when these experiments were completed and the manuscript was prepared for the press, shows some interesting data on a somewhat similar method used for treatment of barley smut. Heald soaked barley seeds in water 4 hours, covered them 8 hours longer, then treated them with formalin i to 288 for 10 minutes and then kept them covered 2 hours. No statement as to the manner of arrivins at the use of this procedure is made. His figures indi- cate for this treatment ( i) less injury to germination than for any other formalin treatment which he used, (2) effective control of barley smut — 0.93 per cent smut in a plot treated in this manner and 0.73 per cent in a somewhat sim.ilarly treated copper-sulphate plot compared to an average of 33 02 per cent smut in three untreated plats. This corroborates for barley smut the work reported on blackchaff with the presoak method. Heald does not appear to have followed up his work, which was clearly a rule of thumb, nor did he recommend this particular method for general use with other methods. He made no allowance for loss in number of seeds per bushel through swelling, otherwise he must have obtained results which would have indicated to him clearly the importance of the method, since he must then have obtained larger yields than by any other method which he used. Moreover, my method differs from Heald 's in that it gives only a short plunge'n water rather than a long one, and this is an important difference. juiyis, i9-'o Presoak Method of Seed Treatment 391 research, result in a saving of a large percentage of seeds destroyed by the usual treatments or delayed in germination and thus longer exposed to the attack of soil fungi, giving at the same time a more efficient germi- cidal action on the pathogens involved. SUMMARY (i) The use of formalin and copper sulphate as now practiced usually causes retardation and injury to seed germination. (2) Greenhouse and field experiments here reported have shown that this detrimental effect can be eliminated for standard varieties of wheat by allowing the seeds to absorb water for six hours before submitting them to the treatment with formalin or copper sulphate. Soaking for a short period (10 minutes) and covering for 6 hours, here designated the presoak method, is better than leaving in water for 6 hours. Similar results were obtained in experiments with barley, oats, and com. (3) The saturation of the seed cells and cell walls with water during the presoak period appears to be the factor counteracting the injurious efifect on seed germination by diluting the disinfectant beyond the point of injury as it diffuses into the tissues and also by considerably decreas- ing the amount of water plus disinfectant solution which may enter the tissues after presoaking as compared to what may enter without any presoaking. (4) Actual stimulation of germination has been observed repeatedly in presoak-treated seeds, a factor wliich by shortening germination min- imizes the danger of exposure to the attack of soil organisms during this susceptible period. (5) The bacterial blackchaff disease of wheat can be controlled with- out any injury to seed germination by a 6-hour presoak of surface- infected seeds in water followed by a 6-hour treatment with formahn i to 400 in the manner prescribed. (6) In practice, wheat seeds after being screened should be soaked with water for 10 minutes at about 6 o'clock in the morning, drained, covered, and set away moist till noon, then soaked with formalin i to 400 for 10 minutes, drained, covered, and set away moist till 6 o'clock in the evening, when they should be spread out to dry overnight to be ready for planting the next day. (7) In planting, an allowance must always be made for the fact that ' there are fewer treated seeds in a bushel than there are of dry untreated ones. In general, it is recommended to sow about 25 per cent more bulk than is usual of the dry grain, otherwise fewer seeds will be actually planted and the yield will be reduced correspondingly. (8) The use of the presoak method tends to increase the efficiency of the disinfectant, in that the presoaking stimulates dormant bacteria and possibly fungi into vegetative activity, thereby rendering them extremely susceptible to the subsequent action of the disinfectant. (9) The general use of the presoak method of treatment in farm practice for other diseases involves no radical change in present procedure, 392 Journal of Agricultural Research voi. xix, no. s the only deviation being to keep the seeds moist for a definite period before giving them the disinfectant treatment. (id) In applying the principles here utilized to other kinds of seeds, the determination of the lengths of the two parts of this method — (i) the presoak period, (2) the subsequent disinfectant treatment period — must be governed by the following factors : (a) the rate of absorption of water by the seeds, (b) the susceptibility of the seeds and pathogens to the disinfectant, and (c) the respective periods necessary for the beginning of seed germination and of vegetative activity of the pathogen. In no case must the presoak period be continued until seed germination begins. The length of time necessary for the seeds to absorb about 30 per cent of their weight of water is suggested as the length of the presoak period when not conflicting with the other factors involved. (11) The presoak method of treatment, as here formulated, is pro- posed as a basis for the reinvestigation of practical seed treatment for all seed-transmitted diseases of economic importance amenable to control by formalin and copper sulphate as a means of eliminating seed injury and at the same time increasing germicidal efficiency. LITERATURE CITED (i) Braun, Harry. I919. PRESOAKING AS A MEANS OF PREVENTING SEED INJURY DUE TO DISINFECT- ANTS AND OF INCREASING GERMICIDAL EFFICIENCY. In Science, n. S. V. 49, no. 1275, p. 544-545- (2) Heald, F. D. 1908. SEED TREATMENT FOR THE SMUTS OF WINTER BARLEY. In Neb. Agf. Expt. Sta. 2ist. Ann. Rpt. [igoyJ/oS, p. 45-53, illus. (3) Humphrey, H. B., and Potter, A. A. I918. CEREAL SMUTS AND THE DISINFECTION OF SEED GRAIN. U. S. Dept. AgT. Farmers' Bui. 939, 28 p., 16 fig. (4) Johnson, A. G., Stakman, E. C, and Potter, Alden A. 1918. TABULAR COMPILATIONS AND PRELIMINARY REVIEW OF RESULTS ON SEED INJURY TESTS WITH STANDARD DISINFECTANTS. Prepared in Off. Cereal Invest. Bur. Plant Indus. U. S. Dept. Agr. for the War Emergency Bd. of Amer. Plant Path. Mimeographed. (5) Smith, Erwin F. 1917. BLACK CHAFF OF WHEAT. In U. S. Dept. Agr. Bur. Plant. Indus. Plant Disesase Survey, Plant Disease Bui. v. i, no. 2, p. 40. (6) 1917. A NEW DISEASE OF WHEAT. In Jour. Agr. Research, v. 10, no. i, p. 51-54, pi. 4-S. (7) 1918. BLACK CHAFF OF WHEAT. In U. S. Dept. Agr. Bur. Plant Indus. Plant Disease Survey, Plant Disease Bui. v. 2, no. 6, p. 98-99. (8) JoNES, L. R., and Reddy, C. S. 1919. THE BLACK CHAFF OF WHEAT. In Science, n. s. v. 50, no. 1280, p. 48. (9) Stuart, William. 19OI. formalin AS A PREVENTATIVE OF OAT SMUT. Ind. Agr. Exp. Sta. Bui. 87, 26 p. (10) Thomas, C. C. 1919. SEED DISINFECTION BY FORMALDEHYDE VAPOR. In Jour. Agr. Research, V. 17, no. I, p. 33-39, illus. PLATE 69 Relative injury to wheat-seed germination caused by short and long formalin treat- ments: A. — Effect of formalin i to 400 and i to 200 treatment for 3, 6, and 12 hours on China variety : Row I, formalin i to 400 for 3 hours, 56 per cent germination; Row 2, formalin i to 400 for 6 hours, 53 per cent germination; Row 3, formalin i to 400 for 12 hours, 74 per cent germination; Row 4, control, 78 per cent germination; Row 5, formalin i to 200 for 3 hours, 39 per cent germination; Row 6, formalin i to 200 for 6 hours, 37 per cent germination; Row 7, formalin i to 200 for 12 hours, 57 per cent germination. B . — Effect of formalin i to 400 and i to 200 treatment for 3,6, and 1 2 hours on Turkey variety : Row I, formalin t to 400 for 3 hours, 42 per cent germination; Row 2, formalin i to 400 for 6 hours, 48 per cer.t germination; Row 3, formalin i to 400 for 12 hours, 50 per cent germination; Row 4, control, 56 per cent germination ; Row 5, formalin i to 200 for 3 hours, 23 per cent germination; Row 6, formalin i to 200 for 6 hours, 17 per cent germination; Row 7, formalin i to 200 for 12 hours, 34 per cent germination. Note increased vigor and germination of 12-hour treated seeds compared with 3-hotir or 6-hour treated seeds. Presoak Method of Seed Treatment PLATE 69 UJ < o Journal of Agricultural Research Vol. XIX, No. 8 Presoak Method of Seed Treatment Plate 70 Journal of Agricultural Research Vol. XIX, No. PLATE 70 Effect of formalin i to 400 treatment for 6 hours, with and without 6-hour presoak: 1. Fultz wheat: A, control, 76 per cent germination; B, seeds presoaked 6 hours, then formalin i to 400 for 6 hotirs, 79 per cent germination; C, seeds treated with form- alin I to 400 for 6 hours, not presoaked, 57 per cent germination, plants dwarfed. 2. Poole wheat: A, control, 90 per cent germination; B, seeds presoaked 6 hotirs, then formalin i to 400 for 6 hours, 89 per cent germination; C, seeds treated with form- alin I to 400 for 6 hours not presoaked, 67 per cent germination, plants dwarfed. PLATE 71 Effect of formalin i to 400 treatment for 6 hours, with and without 6-hour presoak: 1. Fulcaster wheat; A, control, 71 per cent germination; B, seeds presoaked 6 hours, then formalin i to 400 for 6 hours, 83 per cent germination ; C, seeds treated with form- alin I to 400 for 6 hours, not presoaked, 54 per cent germination, plants dwarfed. 2. Turkey wheat: A, control, 61 per cent germination; B, seeds presoaked 6 hours then formalin i to 400 for 6 hours, 68 per cent germination; C, seeds treated with form- alin I to 400 for 6 hours without presoaking, 50 per cent germination, plants dwarfed. Presoak Method of Seed Treatment Plate 71 d3i9V0ind^ A^Myni — Journal of Agricultural Research Vol. XIX, No. 8 Presoak Method of Seed Treatment Plate 72 aiisvomj'* A'3>idnj]iB U?) -^Ji. Journal of Agricultural Research Vol. XIX, No. 8 PLATE 72 Stimulating effect of the presoak method of treatment with formalin i to 400. (Repe- tition of experiments shown in PI. 70, 71): 1. Fulcaster wheat : A, C, seeds presoaked 6 hours, then formalin i to 400 for 6 hours, 84 per cent germination, plants stimulated; B, control, 87 per cent germination. 2. Turkey wheat: A, C, seeds presoaked 6 hours, then formalin i to 400 for 6 hours, 64 per cent germination; B, control, 67 per cent germination. Note stimulation in presoak -treated plants. 3. Poole wheat: A, C, seeds presoaked 6 hoiu-s, then formalin i to 400 for 6 hours, 88 per cent germination; B, control, 94 per cent germination. Note increased vigor and stimulation in presoak-treated plants. PLATE 73 Effect of formalin i to 200 treatment for 6 hours, with and without 6-hour presoak: 1. Fultz wheat: A, seeds treated with formalin i to 200 for 6 hours without presoak, 43 per cent germination; B, control, 86 per cent germination; C, seeds presoaked 6 hours, then formalin i to 200 for 6 hours, 70 per cent germination. 2. Poole wheat: A, seeds treated with formalin i to 200 for 6 hours without presoak, 46 per cent germination; B, control, 81 per cent germination; C, seeds presoaked 6 hours, then formalin i to 200 for 6 hours, 71 per cent germination. Note stimulation in presoak-treated plants. Presoak Method of Seed Treatment Plate 73 Journal of Agricultural Research Vol. XIX, No. 8 Presoak Method of Seed Treatment Plate 74 Journal of Agricultural Research Vol. XIX, No. 8 PLATE 74 Effect of formalin i to 320 treatment for 6 hours, with and without 6-hour presoak: 1. Fife wheat: A, control, 79 per cent germination; B, seeds treated with formalin I to 320 for 6 hours, 53 per cent germination; C, seeds presoaked 6 hours, then forma- lin I to 320 for 6 hours, 82 per cent germination. 2. Poole wheat: A, control, 88 per cent germination; B, seeds treated with forma- lin I to 320 for 6 hours, 70 per cent germination; C, seeds presoaked 6 hours, then for- malin I to 320 for 6 hours, 86 per cent germination. PLATE 75 I. Fife wheat: A, B, control, 83 per cent germination; C, D, seeds treated with formalin i to 320 for 3 hours, 62 per cent germination; E, F, seeds presoaked 6 hours, then formalin i to 320 for 6 hours, 90 per cent germination. 3. Poole wheat: A, B, control, 85 per cent germination; C, D, seeds treated with formalin i to 320 for 6 hours, 55 per cent germination; E, F, seeds presoaked 6 hours, then formalin i to 320 for 6 hours, 88 per cent germination. 3. Effect of soaking in water throughout presoak period, compared with procedure of keeping moist 6 hoiu-s: A, B, Fife wheat; C, D, Poole wheat; A, C, seeds soaked in water 5 hoxu-s, then treated with formalin i to 320 for 7 hours; B, D, seeds soaked in water 10 minutes, drained, and kept moist 6 hours, then treated with formalin i to 320 for 6 hours. Note the greater stimulation in B and D. Presoak Method of Seed Treatment Plate 75 Journal of Agricultural Research Vol. XIX, No. 8 Presoak Method of Seed Treatment Plate 76 9inDdVfc!|J sinodvw Journal of Agricultural Research Vol. XIX. No. 8 PLATE 76 Effect of formalin and copper sulphate on wheat and of copper sulphate on barley, with and without presoaking: 1. Marquis wheat: A, B, control, 76 per cent germination; C, D, seeds treated with formalin i to 320 for 6 hours, 57 per cent germination; E, F, seeds pre"soaked 6 hours, then formalin i to 320 for 6 hours, 77 per cent germination. 2. Marquis wheat: A, B, control, 84 per cent germination; C, D, seeds treated with copper sulphate i to 80 for ^.^ hoiu", and limed, 65 percent germination; E, F, seeds presoaked 6 hours, then copper sulphate i to 80 for % hour, and limed, 79 per cent germination. 3. Tennessee winter barley: A, B, control, 89 per cent germination; C, D, seeds treated with copper sulphate i to 80 for yi hour, and limed, 72 per cent germination; E, F, seeds presoaked 6 hours, then copper sulphate i to 80 for % hour, and limed, 83 per cent germination. PLATE 77 Effect of presoak method used with copper-sulphate treatment of wheat: 1. Fulcaster wheat: A, B, control, 8i per cent germination; C, D, seeds treated with copper sulphate i to 80 for yi hour, and limed, 57 per cent germination; E, F, seeds presoaked 8 hoiu-s, then copper sulphate i to 80 for >< hour, and limed, 82 per cent germination. 2. Fife wheat: A, B, control, 79 per cent germination; C, D, seeds treated with copper sulphate i to 80 for >< hour, and lim.ed, 52 per cent germination; E, F, seeds presoaked 8 hoiu-s, then copper sulphate i to 80 for yi hour, and limed, 81 per cent germination. Presoak Method of Seed Treatment Plate 77 Journal of Agricultural Research Vol. XIX, No. 8 Presoak Method of Seed Treatment PLATE 78 Journal of Agricultural Research Vol. XIX, No. 8 PLATE 78 Effect of presoak method used with formalin i to 320 on wheat tinder field conditions: I. Poole; 2. China; 3. Bluesteni; 4. Fulcaster; 5. Marquis; 6. Fife. A.— Eight hundred seeds in four rows, soaked in water 10 minutes, kept moist 6 hours, soaked in formalin 10 minutes, kept moist 6 hours, dried overnight. B.— Eight htmdred seeds in four rows, soaked in formalin i to 320 for 10 m.inutes, kept moist 6 hours, dried overnight. C. — Eight hundred seeds in four rows, control. PLATE 79 Effect of presoak method used with formalin i to 400 on blackchaff bacteria: A. — Controls, dry-heat sterilized wheat seeds inoculated with blackchaff isolation No. 377 from South Dakota, dried, and planted on agar without further treatment. B. — Sterilized wheat seeds inoculated with isolation No. 377, dried overnight, then soaked in sterile tap water 10 minutes and kept moist 6 hours, then soaked in for- malin I to 400 for 10 minutes and kept moist 6 hours, dried, and planted. Note bacterial growth around untreated seeds and absence of growth in presoaked formalin-treated seeds. Two Petri dishes photographed out of a set of 120 dishes in experiment III on tenth day. Presoak Method of Seed Treatment Plate 79 Journal of Agricultural Research Vol. XIX. No. 8 Presoak Method of Seed Treatment Plate 80 Journal of Agricultural Research Vol. XIX, No. 8 PLATE 80 Effect of presoak method used with formalin i to 400 on blackchaff bacteria: A. — Controls, dry -heat sterilized wheat seeds inoculated with blackchaff isolation No. 373 from North Dakota, dried, and planted on agar without further treatment. B. — Sterilized wheat seeds inoculated with isolation No. 373, dried overnight, then soaked in sterile tap water 10 minutes, kept moist 6 hours, then soaked in formalin I to 400 for 10 minutes, drained, kept moist 6 hours, dried, and planted. Note bacterial growth aroimd untreated seeds and absence of growth in presoaked formalin-treated seeds. Two Petri dishes photographed out of a set of 120 dishes in experiment III on tenth day. 177286°— 20 6 PLATE 8i Effect of presoak method on barley and oats : 1. Chevalier barley: A, control, 92 per cent germination; B, seeds treated with formalin i to 320 for 6 hours, 58 per cent germination ; C, seeds presoaked 6 hours, then formalin i to 320 for 6 hours, 91 per cent germination. 2. Burt oats: A, control, 94 per cent germination; B, seeds treated with formalin I to 320 for 6 hours, 66 per cent germination; C, seeds presoaked 6 hours, then treated with formalin i to 320 for 6 hours, 94 per cent germination. Presoak Method of Seed Treatment Plate 81 Journal of Agricultural Research Vol. XIX, No. 8 Presoak Method of Seed Treatment Plate 82 Journal of Agricultural Research Vol. XIX. No. 8 PLATE 82 Effect of presoak method used on barley, oats, and com: 1. Chevalier barley: A, B, control, 92 per cert germination; C, D, seeds treated with formalin i to 320 for 10 minutes, then kept moist (covered) 6 hours, 62 per cent germination; E, F, seeds presoaked 6 hours, then formalin i to 320 for 6 hours, g2 per cent germination. Photographed on seventh day. 2. Burt oats: A, B, control, 96 per cent germination; C, D, seeds treated with for- malin I to 320 for 6 hours, 71 per cent germination; E, F, seeds presoaked 6 hoiu-s, then formalin i to 320 for 6 hours, 92 per cent germination. Photographed on seventh day. 3. Bantam Evergreen com: A, B, control, 94 per cent germination; C, D, seeds treated with formalin i to 200 for 4 hours, 90 per cent germination; E, F, seeds actu- ally soaked in water 10 hoiu-s, then treated with formalin i to 200 for 4 hours, 92 per cent germination. Count taken on twelfth day, photographed on sixteenth day. The presoaked seeds are stimulated. ADDITIONAL COPIES OF THIS PUBUCATION MAY BE PROCURED FROM THE SUPERINTENDENT OF DOCUMENTS GOVERNMENT PRINnNG OFFICE WASHINGTON, D. C. AT 25 CENTS PER COPY SuBscEiPTiON Price, $1.50 Per Year Vol. XIX AUGUSX 2, 1920 No, 9 JOURNAL OF AGRICULTURAL RESEARCH CONXE^NXS Page Daily Development of Kernels of Hannchen Barley from Flowering to Maturity at Aberdeen, Idaho - - 393 HARRY V. HARLAIf ( Contrilmtloa trom Bureau of Plant Indostxy and IdalM Agricultuial Ezperiment Station ) Development of Barley Kernels in Normal and Clipped Spikes and the Limitations of Awnless and Hooded Varieties --------- 431 HARRY V. HARLAN and STEPHEN ANTHONY ( Contribution from Bureau of Plant Induatiy and Idabo Agricultural Experiment Station) PDBUSHED BY AUTHORITY OF THE SECRETARY OF AGRICULTURE, WITH THE COOPERATION OF THE ASSOOATION OF LAND-GRANT COLLEGES WASHINGTON. D. C. WMNINQTON I •OVKaNMCNT PfUMTINO OfVWI t IMO EDITORUL COMMITTEE OF THE UNITED STATES DEPARTMENT OF AGRICULTURE AND THE ASSOCIATION OF LAND-GRANT COLLEGES FOR THE DEPARTMENT KARL F. KELLERMAN, Chairman Physiologist and Associate Chief, Bureau of Plant Industry EDWIN W. ALLEN Chief, Office of Experiment Stations CHARLES L. MARLATT Entomologist and Assistant Chief, Bureau of Entomology FOR THE ASSOCIATION J. G. LIPMAN Dean, State College of Agriculture, and Director, New Jersey Agricultural Experi- ment Station , Rutgers College W. A. RILEY Entomologist and Chief, Division of Ento- mology and Economic Zoology, Agricul- tural Experiment Station of the University of Minnesota R. L. WATTS Dean, School of Agriculture, and Director, Agricultural Experiment Stalicm, The Pennsylvania State College All correspondence regarding articles from the Department of Agriculture should be addressed to Karl F. Kellerman, Journal erf Agricultural Research, Washington, D. C. All correspondence regarding articles from State Experiment Stations should be addressed to J. G. Lipman, New Jersey Agricultural Experiment Station, New Brunswick, N. J. ■ciiN JOIMALOFAGElirniAlRESEARCH Vol. XIX Washington, D. C, August 2, 1920 No. 9 DAILY DEVELOPMENT OF KERNELS OF HANNCHEN BARLEY FROM FLOWERING TO MATURITY AT ABER- DEEN, IDAHO 1 By Harry V. Harlan Agronomist in Charge of Barley Investigations, Office of Cereal Investigations, Bureau 0/ Plant Industry, United States Department of Agriculture INTRODUCTION Several years ago the author made a few elementary experiments on the function of the awn in barley. In these studies the awns were cHpped from some spikes and not from others. The effect on the development of the kernels was so striking that in 191 5 a more elaborate experi- ment was carried out by the author and Stephen Anthony. The develop- ment of kernels on normal and clipped spikes was determined from flowering to maturity. The method of study proved so satisfactory that it led to other investigations in which it offered the same possibility of application. The development of barley on dry land and on irrigated land, the response to irrigation water, and the differences in varietal behavior have all been studied by this means. The last study has been undertaken since the resignation of Mr. Anthony. In these studies, kernel growth has been used as an index of effect. Yield and size of mature kernels, while probably a safe summary of the effects of varia- tions of treatment or differences of types, do not throw much light on the time when the effect occurred, or always on the reasons therefor. This group of studies has been carried on with the idea that variations from a basic growth curve showing the inception, duration, and degree of response would be much more illuminating than the single observation of the final result. A number of studies have been completed, and it is the intention to publish the results of the special projects from time to time. The results represent a normal growth curve. It is intended that this curve shall form the basis of comparison in the later studies and that it shall 1 These studies were made on the Aberdeen Substation, Aberdeen, Idaho, in connection with cereal experiments conducted cooperatively by the Idaho Agricultural Experiment Station and the Office of Cereal Investigations, Bureau of Plant Industry, United States Department of Agriculture. Journal of Agricultural Research, Vol. XIX, No. 9 Washington, D. C. Aug. a. 1930 «a Key No. G-199 (393) 394 Journal of Agricultural Research voi. xix.No.g be a connecting link between the various studies. In this paper the results with the Hannchen variety are given in full for one year at Aberdeen, Idaho. The Hannchen variety was chosen for this basic statement because it has been used more extensively than any other variety. The growth at Aberdeen is selected both because of the fact that most of the studies made and projected are located there and because of the remarkable uniformity of growth of plants at that place from season to season. In three different years the period from flowering to maturity has extended over exactly 26 days. This uniformity, coupled with accurate sampling, has made it possible to take samples at intervals as short as 24 hours or even less and still show consistent growth. In no previous studies on cereal crops, either here or abroad, have samples been taken more frequently than at 3-day intervals, yet it is readily seen in figure i that most of the growth in length is com- pleted in a period of three days. The measurements of kernel dimen- sions are an important index of development which seems to have been generally ignored. HISTORICAL REVIEW The published data on kernel development have little relation to the various lines of investigation at Aberdeen, Idaho. Dififerences of location and variety make anything but general comparisons difficult in this con- nection. The studies of kernel development previously reported in the cereals have been the outcome of a wide range of experiments and are too numer- ous to be reviewed in detail. Kudelka (4),^ Lermer and Holzner (5), and many others have published on the origin and development of tissues in the caryopsis as a whole, or even in the entire plant. Some investigations have been specifically devoted to tissues of the pericarp. Johannsen (5), Brenchley (/), Schjerning (6, 7), and numerous others have investigated the chemical changes of growth and maturation. The work of these in- vestigators is referred to later. Their experiments were carried on under relatively unfavorable conditions. The contrast is remarkable between the humid climates of Denmark and England, with their frequent storms and days of low activity, and the arid climate of Aberdeen. Schjerning had a difference of 1 2 days in the maturity of his plots in two succeeding years. The detail of the experiment at Aberdeen is more nearly like those of Brenchley (/), Schjerning (6, 7), and Johannsen (5) than those of the other investigators. It differs from these in a reduction of the period between samples and in the extensive study of the physical indices of length and diameter of kernel. The chemical and morphological phases are not comparable with those of Schjerning and Johannsen. 2 Reference is made by number (italic) to "I,iterature cited," p. 429. Aug. 2, 1920 Daily Development of Kernels of Hannchen Barley 395 EXPERIMENTAL METHODS Such success as has been obtained in reducing the interval of sampling is due in large part to the accuracy of tagging spikes at the same stage of development. This, in turn, rests on an observation made several years as 30 a.o 7.0 €:s e.o ^.0 3.S 3.0 ^.o /.s /.o 1 1 / \ \ 0* .-■ N^ .'• ^, *-• .'■ ,, \ / V < ■^ y s < \ \ .^^ 1 / / V, '"1 f / s \ y N / — j f \\ n ' — 1 1 1 - ( ( 1 1 1 < 1 1 j V- ei »^ t.. ?- 4< \ ^ ^ y — — - ;j^ 1 1 ^; / y 1 1 y (^/ 0 '^ vr. f^ iSV 1. =^ ■^ •-> .. - l\ '')> / -- ^ ^ ■'.■>' / / ■'/ - "^^ // y y -' J £k«KP .^JfOrf' / ^ ^ ^ s^ s ?' S 9 /o u /^ /S /^ ''S' yv /^ /a / ^o ^/ i'^ ^s ^* ^^ ^-ciinn^ww Fig. I. — Graph showing length, lateral diameter, and dorsoventral diameter of barley kernels for the as days following flowering. The broken lines give the data for 1916, the soHd lines the data for 1917. before the work was started. In agronomic notes taken upon cereal varieties, the time of heading and the time of ripening have always been considered to be important statements of development. Of these, the time of heading is thought to be especially valuable, because drouth and other climatic factors that greatly influence maturity have usually affected the plant but little up to this stage. 396 Journal of Agricultural Research voi. xix, no. 9 While time of heading in barley is doubtless significant, it is very difficult to determine. A barley spike may be visible two or three days before it is fully exserted from the sheath. In some varieties the spikes are never completely exserted. In a study of this difficulty it was noticed that the emergence of the awns offered opportunity for a tangible obser- vation. Upon trial it was found to be a very accurate index of the stage of the development of the spike. With the observation as a basis, spikes tagged as uniform before flowering were of so nearly the same stage of development that, despite individual fluctuations, growth in as short periods as 12 hours was evident in the data for many days; and almost until maturity the individual variations in samples of only two spikes did not obscure the growth in 24-hour periods. The accuracy of the method and the spectacular uniformity of Idaho seasons is well shown in figure 2, where the percentages of moisture in kernels in the seasons of 1 91 6 and 191 7 essentially coincide throughout the entire period of growth. Three or four days after the tips of the awns are visible on the earliest culms a large number of culms are to be found with tips visible. At this time the plots are carefully inspected and the requisite number of culms is marked. The marking is done by tying a piece of wool yarn about the culm. Culms are selected in which the awns are protruding X to J^ inch above the sheath of the uppermost leaf. A sufficient num- ber of culms is tagged to insure against accident. As soon as the spikes are partially exserted a sample is taken. This sample and the one on the following day usually have several florets which have not yet been fertilized. The samples taken in the first few days consist of three spikes in order to secure a greater quantity of material, but later the number is reduced to two per sample. In most cases only one sample is taken each day, but in the cases furnishing the data reported in Table I two samples were taken, one in the morning and one in the evening. The samples are taken in the field by cutting the culms near the ground. These culms and spikes are wrapped in a moist towel and taken to the laboratory. As a protection against evaporation in the laboratory the spikelets are removed one at a time, the remainder of the spike being left in the towel. To secure the data rapidly and satisfactorily two men work on the same sample. The kernels are taken from the florets by the operator of the calipers, who measures the length, lateral diameter, and dorsoventral diameter in tenths of millimeters and records these measurements. The kernels are then passed to the operator of the balance and weighed to tenths of milligrams. Only the kernels on a single side of each spike are measured and weighed individually. The kernels of the other side of the spike are added to those measured indi- vidually and weighed to obtain a larger sample. These are placed in small vials and dried in a water-jacket oven. The vials are then corked, and the material is preserved for later analysis. Aug. 2, 1920 Daily Development of Kernels of Hannchen Barley 397 EXPERIMENTAL MATERIAL Most of the data presented herein were obtained in 191 7, but many of the graphs contain curves of the data for 191 6 as well. The curves for 1 91 6 are added merely for comparison and to give an idea of the ao - ^ V, N ^i( s. s, H \ s s; s >w \ * N, >ik s ^ * \ \ V \ \ \ \ _ _ L / 3 s ^ s^ e- y a 9 /o ^/ je /j /■* /s /o^ j^^/& /s ^ro^x .ss^j j«* ^s ^£OMir,t>MV /9/T^O^K /W/W/^/iy /^ ^eo.e/ ^■^^■3^■^ J'^^v ^:f.^^ ^si^o j/'y^'^ 3 -^ ^ "">?».,•* ^ y»&^i/£t' ^ /o // /•*" /s /<* /W /6 /7 /& /■■9 ^o^y' re ^3^* e^ j0 so aif9^f>i>vAzaf^>iinv /j'3-^^G'yff& /& /^ /^ /(^ yt^y^ /O' yT' y^ y(ff.£tyjyj?^ £^J.it*^SJ^e■ ,yiyy.f^ys'y yy yj' y^ y-^y^/fffy^y,ff ysJVjy^^^^.af^s.^s^y',e»j^jo3y 'T''^ s ■*» ^yc/^.)^ysy7- ys yg'y^ yffy.»^ia>^y^^^s2*^^S4?G^7,^e^3Jos/ y> ^ Unfertilized. 400 Journal of Agricultural Research Vol. XIX, No. 9 Table I. — Normal growth of Hannchen barley in J 2-hour periods from flowering to maturity, at Aberdeen, Idaho, in ipi^ — Continued 6 a. m. 7 p. m. Ker- nel No. from base of spike. Weight of ker- nel on spike — Length of kernel on spike — Lateral diameter of kernel on spike — Dorso- ventral diameter of kernel on spike — Weight of ker- nel on spike — Length of kernel on spike — Lateral diameter of kernel on spike— Dorso- ventral diameter of kerne on spike — A. B. A. B. A. B. A. B. A. B. A. B. A. B. A. B. Gm. Gm. Mm. Mm. Mm. Mm. Mm. Mm. Gm. Gm. Mm. Mm. Mm. Mm. Mm. Mm. 2 3 4 5 6 7 8 9 lO II 13 13 O. 0022 •002S •0033 •0033 .0038 .0036 .0038 .0038 .0029 • 0031 • 0025 .0018 0. 001 3 • 0023 .0030 .0034 • 0037 .0039 .0043 .0048 .0043 .0036 . 0029 2.4 a. 7 2.9 30 3- I 3-5 3-2 3-4 31 30 3.8 3.1 2. 1 3.6 3- a 3-3 3- S 3-7 3-8 4-3 4-1 3-7 3-4 I. 1 I. 2 1-3 1-4 I- 5 IS 1.4 l-S 1-3 1.4 1-3 1-3 xo 1-3 1-3 1-4 1.4 1.4 1-4 1-5 1-3 1-3 1-3 0-7 •7 .8 .8 •9 •9 •9 •9 .8 .8 .8 .6 c6 .7 .8 .8 .8 .8 •9 •9 .8 .8 •7 0. 0029 .0045 .0055 .0058 .0061 .0069 .0076 . 0063 . 0071 .0067 • 0057 0.0033 .0051 .0052 .0056 . 0064 .0071 .0067 .0068 .0069 . 0062 .0030 .0038 3.8 3-6 4-4 4.6 4-9 5.0 5.6 4-7 5-7 5-4 4.8 3-S 3-9 4-2 4-8 5.0 5-S 55 S-6 5-5 51 4-7 3-9 1-4 1-3 1-4 1.6 IS 1-4 1-4 1.6 IS 1-4 1-4 I-S 0.8 .8 •9 • 9 •9 •9 •9 .8 •9 •9 .8 0.8 .8 ■ 9 •9 •9 •9 •9 •9 ■9 .8 .8 •7 JULY 18 3 0 0046 0.003S 3-8 3-6 i-S 1-3 0.9 0.8 0. 0048 0 0054 4.0 4-S S 5 0 7 0.8 3 0076 .0047 5-2 4.0 1.6 1-4 •9 •9 •0073 0075 .5-6 S-6 S 6 8 -9 4 0088 .0056 6.3 4.6 1.6 1-4 •9 •9 .0079 0086 5-9 6.1 6 6 8 -9 S 0094 .0060 6.4 4-7 1.6 1-4 •9 •9 .0083 0089 6-3 6.9 6 7 9 •9 6 009s .0067 6.8 5-4 I-S 1.6 •9 .9 .0082 0095 6.3 6.7 6 6 9 1.0 7 OIOO •007s 6.9 5-9 1.6 1-5 •9 •9 .0098 0096 7-0 6.8 6 7 I 0 1.0 8 0097 .0079 7-1 6.1 1.6 1.6 I.O •9 .0102 0108 7.0 7-5 7 6 I 0 1. 1 9 0093 .0076 6.8 5-9 1.6 1-5 •9 .9 .0104 0093 7-4 6.9 7 7 I 0 1.0 10 0089 .0074 6.8 5-9 1.6 1-5 •9 .8 .0096 0090 7-3 6.6 6 6 I 0 -9 II 0089 .0073 6.S 5-9 I-S I-S .8 .8 .0083 0086 6-7 6-5 6 6 9 •9 12 0068 .0064 6.0 5-S I-S I-S .8 .8 .0075 0068 6-3 5-7 I 6 I 6 .8 .8 0.0023 .0070 2.8 1-3 •5 0.7 .8 2 0.0050 S-5 4.8 I-S 0.8 0. OIOO 0.0084 7.0 6 2 1-7 1-7 0 0.9 3 .0091 .0090 6.4 6.6 1-7 1.6 •9 •9 .OI2S .0109 8.0 7 3 1.8 1-7 a 1.0 4 .0100 .0108 6.7 7-7 1-7 1-7 1.0 1. 1 .0142 .0123 8-7 8 I 1.8 1.8 a 1.0 5 .0114 .0128 7-9 8-3 1-7 1-7 I. 3 .0151 .0131 9.0 8 3 1-9 1-7 3 t. a 6 • 0123 • 0132 8-3 8.3 1.8 1.8 1-3 .0160 .0136 9-1 8 4 1-9 1.8 3 1-3 7 .0128 8.0 1.8 .0170 .0149 9-3 8 8 3.0 1-9 3 x-3 8 .0129 .0151 8.3 8-S 1.8 1.8 1-3 .0166 .0146 9-3 8 7 1-9 1-9 4 1-3 9 .0127 .0156 8.0 8.6 1.8 2.0 1.4 .0154 .0160 8.8 8 9 1-9 1-9 3 1-3 10 .0124 •0153 8.2 8-7 1.8 2.0 1-3 • 0157 .0150 8.8 8 8 1.8 1-9 a I. a II .0116 .0141 7.8 8.2 1-7 1-9 1-4 .0147 .0138 8-5 8 8 1.8 1.8 a I. a 12 .0102 .0132 7-7 8.0 1-7 1.8 I. 0 I. 2 .0108 7 8 1.6 I.I 13 .0082 .0115 6-5 7-5 1.6 1.8 1.0 1. 1 JULY 20 I 0.0060 5.6 1.6 0.8 0. 0046 4-9 1-4 0.6 2 0.0098 •0133 7-1 8.0 1.8 1-9 I. 0 I. 2 0.0099 .0132 7.0 8.3 1-7 1.8 1.0 3 .0107 • 0163 7-3 9.0 1.8 1-9 I. I 1-3 .0107 .0145 7-1 8.8 1.8 1.9 1.0 4 .0138 .0176 9.0 9.0 1-9 I.O I. 2 1-4 .0136 .0178 8.2 9-4 1.9 3.0 I. 3 5 .0150 .0178 9.0 9.0 1.9 2.0 1-3 1-4 .0151 .0192 9.0 9-6 1-9 3. 1 I. a 6 .0170 .0188 9-5 9.1 2.0 2.1 1-3 I-S .0162 .0200 8.8 9-9 2.0 a. a I. a 7 .0177 .0198 9.1 9-7 2.0 2. 2 1-3 1.6 .0165 .0204 8.8 9-7 2.0 3. a 1-3 8 .0174 .0208 9-S 9-7 1-9 2. I 1-3 I-S .0196 9.6 2.1 9 .0159 .0200 9-1 9.6 2.0 2. 2 1-3 I-S .0164 .0194 9.0 9-1 3.0 2. 1 1.4 10 ■ 0165 .0176 9.0 9-3 2.0 2. I 1-3 I-S • 0152 .01 68 8.7 9-3 1.8 3.0 1-3 II • oiso .0165 8.8 8.8 1-9 2.0 1-3 1-4 •0133 • 0151 8.4 8.6 1.8 1-9 t. 3 13 .0126 .0125 8.1 8.3 1.8 1.8 I. 3 1.3 .0121 8.0 1-7 I. a Aug. 2, 1930 Daily Development of Kernels of Hannchen Barley 401 Table I. — Normal growth of Hannchen barley in l2-hour periods from flowering to maturity, at Aberdeen, Idaho, in igiy — Continued 6 a. m. 7 p.m. Ker- nel No. from base of spike. Weight of ker- nel on spike— Length of kernel on spike — Lateral diameter of kernel on spike — Dorso- ventral diameter of kernel on spike — Weight of ker- nel on spike— Length of kernel on spike — Lateral diameter of kernel on spike— Dorso- ventral diameter of kernel on spike — A. B. A. B. A. B. A. B. A. B. A. B. A. B. A. B. Gm. Gm. Mm. Mm. Mm. Mm. Mm. Mm. Gm. Gm. 0.0133 .0179 • 0209 •0225 • 0250 .03S7 .0356 .0375 •o»33 •0334 •0337 .0193 Mm. Mm. Mm. Mm. Mm. Mm. a 3 4 5 6 7 8 19 10 n 0.0093 .0139 .0170 .0181 .0183 • 019s .0199 .0195 .0170 .0158 0.C134 .0165 .0184 .0303 •OI9S .0209 .0216 .0306 .0186 .0169 6.8 8-3 9.1 9.0 9-3 9-4 9-3 95 8.9 8.8 7-9 8.8 9-3 9.6 9.6 10. 0 10. 0 9.6 9.4 1-7 1.8 2.0 3.0 3. I 3.0 3. I 3.0 1.9 1.9 1.9 1.9 1.9 3. 1 3. I a.o a. I 3.0 1.9 3-0 I.O I. 3 1-3 X.4 1-4 I- 5 i-S l-S »-5 x-3 1-3 1-3 1-5 1-5 1.4 I-S 1.6 1-5 1.6 1-4 O.OIS7 .0199 • 03I0 .0227 •0330 .03l8 .0348 .0316 .03I3 .0187 .0168 "z'.'i 9-4 10. 0 10. 0 to. I 9.8 10. 0 10. I 9-3 9-2 8.6 8.x 9-3 9-S 9-7 10.4 10. 3 10. I 10. 0 9.8 9.1 9-6 8.6 3. 0 2. 3 2. I 2. I 3. 2 2. 2 a- 3 a- 3 2-3 2. 2 2. 0 1.9 a. I a. 3 3. 3 a- 3 3.4 a- 4 a-S a- 3 3.4 »-4 2. 3 1-4 I- 5 1-5 1.6 1.6 I- 5 1-7 15 I-S I- 5 1.4 I. a i-S 1-4 1.6 1-7 1.6 1-7 1.8 1-7 1.8 13 i-S 8.9 9-S 3.0 2-3 a 0-0190 0.0200 9.2 9.6 2. 2 2-3 1-4 i-S 0. 0190 .0205 9-1 2. 2 1.4 1-5 3 .0248 .0225 9.9 10. I 3-3 3.4 1.6 1.6 • 0332 •0253 9.9 10. 6 2.4 2.4 1.6 I- 7 4 .0272 .0283 10. 2 10. s 2-S 2-S 1-7 1.8 .0270 • 0364 10. 2 10. >; 2.6 2-S 1.6 1.6 5 .0291 .0294 10. 2 10.3 2.6 2.6 1.9 1.8 .0278 .0289 10. 2 10. s 2-S a. 8 1.8 1.9 6 .0279 .0293 10. I 10. 1 2.6 2-S 1-9 1.8 .0284 .0289 lo. 4 10.4 2.6 3.6 1.8 1.9 7 .0276 .0289 9.8 10. 2 2.7 2.6 1.8 1.8 .0280 .0284 10. 0 10. s 3.7 2.6 1.8 1.8 8 .0283 .0282 95 10. 1 2-7 2-S 1.9 1.8 .0271 .0287 g.6 10. 0 3.6 2.8 I- 7 1.8 9 .0256 .0306 9-S 10. 1 2.6 2.6 1.6 1.9 .0282 .0282 9.6 10. 2 3.6 2-7 I. 8 t.8 10 .0248 .0277 9-3 9.9 2-5 2.6 1.8 1.8 .0266 .0270 9-4 9.9 2.6 2-S I- 7 1-9 II .0303 .0257 8.8 9.8 2.4 2.6 i-S 1-7 .0241 •02S3 9-4 9-7 2-5 2-S I- 7 1-7 13 .0237 9.6 2-S 1-7 .0217 • 0258 8.9 10. 0 2-3 a. 6 1.6 1-7 13 .0195 90 2-3 I-S .0220 8.8 a-S 1-7 0.0084 • 0223 .0270 7-4 10. 0 10. 0 1-7 2-3 2- 7 i>a 0. 0265 .0321 9-9 lo- ,', 2-7 3-8 I- 7 1-8 3 0. 0283 0-02I9 10. 0 9-3 2.8 2-3 1-8 1-7 x-8 4 -0333 .0270 10. 3 10. I 2-7 2-6 1-9 1.8 •033 s •0316 10. I 10. 2 3-0 2- 7 2- 0 1.9 5 -0363 .0294 10.3 10.5 3-0 2. 7 2.0 1-9 •03 59 .0249 10.3 9-S 3-0 2- 7 1-9 1.9 6 -0373 • 0316 10.3 10.5 3-1 2.8 2. 2 2. 0 •0359 .0308 10. 6 ro. 2 3-0 2.8 2- 0 1-9 7 -0340 • 0334 9-9 10. s 3-0 2.9 1-9 1-9 •0352 .0320 10. 7 10. I 2.8 2.8 2- 0 a-o 8 -0JS4 •0334 10. I 10.3 2-9 2-9 2- 0 2. 0 •0372 -0334 10- 6 lO- 0 3-0 2.8 2. I 1-9 9 •0339 .0312 9-9 9-7 2.8 2.8 2- I 1-9 •0343 .0298 9.9 9.9 3-0 2.8 2.0 1.8 10 .0294 .0300 9-4 10.3 2.8 2.8 1-8 1-9 .0348 .0319 10. 0 10. I 3- I 2-9 2.0 1.8 II . 0292 .0309 9-3 9-7 2.8 2.8 1-9 2. 0 •03 IS .0291 q.6 9-S 2.9 2.8 1.9 1.8 12 • 0246 .0275 8.6 9-S 2.6 2.6 1.8 1.8 .0299 .0230 9-7 9-1 2.9 2.6 2. 0 1.8 Jt TLY 2i i I a 0.0142 0-02I2 8.4 9-0 2-3 2-4 1-4 I-S 0.0173 'is' 2. 0 "1.6 3 .0285 .0302 9-9 to. 3 2-3 2.8 1-7 1-8 0- 0291 .0301 9-6 9-4 2-7 2.8 1-8 1.0 4 -0323 .0328 9.8 10. 2 2.9 2.8 1.8 1.9 -0335 •0333 10.3 10. 0 30 3.8 3. 0 a. I S -0346 .0328 10.4 10.3 2.9 3.8 1-9 1.8 •0361 .0372 10. 0 10. 3 3-0 3- I 3. 0 3.0 6 -0359 •034s 10.3 to. 2 30 3-9 1.9 1-9 •0371 •0383 10. 3 10.4 3-0 3-0 a. 0 3.0 7 •0357 •0332 10. 2 10.3 3.9 2-9 1.8 3.0 •03 S3 .0387 10. 0 10. 2 2-9 3-0 a-o 3.3 8 -035s •0342 9.9 to. 2 2.9 3-8 a.o 2.0 •0379 •0381 10. I 10. 3 3-0 3- 2 3. 0 3.0 9 .0328 .0340 9.8 10.4 3-0 3-8 1.8 2. 0 .0322 •0398 10. I 10. 0 2-9 3- I X-9 3.0 to .0342 •0333 9- I 9.9 30 2-9 1-9 3. I •0338 •0384 9.6 10. I 2.8 3- 1 r. 9 3.0 II .0319 .0328 9-4 to- 2 3-0 2-9 t. 6 a. 0 •0344 •0363 9-3 Q-8 3-0 30 3.0 3.x la .0276 •0304 9.0 9- 7 2.9 2.8 1.8 a. 0 .0278 •0351 9-3 9.6 3.8 3.9 3. 0 1.9 ti .0284 8.8 a. 8 1-9 -032s 9-9 2.8 1.8 402 Journal of Agricultural Research Vol. XIX, No. 9 Table I. — Normal growth of Hannchen barley in 12-hour periods from flowering to maturity, at Aberdeen, Idaho, in igiy — Continued 6 a. m. 7 p.m. Ker- nel No. from base of spike. Weight of ker- nel on spike — Length of kernel on spike — Lateral diameter of kernel on spike — Dorso- ventral diameter of kernel on spike — Weight of ker- nel on spike — Length of kernel on spike — Lateral diameter of kernel on spike — Dorso- ventral diameter of kernel on spike— A. B. A. B. A. B. A. B. A. B. A. B. A. B. A. B. Gm. Cm. Mm. Mm. Mm. Mm. Mm. Mm. Gm. Gm. Mm. Mm. Mm. Mm. Mm. Mm. 3 3 4 5 6 7 8 9 10 II 13 13 O.OI35 .0356 .0386 .0449 • 044s .0440 •03S7 .0438 •0349 .0388 •0343 0.0093 -0273 .0308 .0366 •0377 .0367 •0394 • 0384 .0378 •0374 • 0350 •0332 7-S 9.9 10.5 10.3 10.3 10.3 9.8 9.9 9.8 9.4 9.1 7-4 9.2 9.8 10.4 10.3 10. I 10. 6 10. 2 10. 0 10. 2 9.8 9.3 1-7 3-0 3- I 3-i 3-3 3-2 3-0 3-4 2.9 3-3 3-0 IS 2.6 3.8 3- I 3-0 30 3- I 3-0 2.9 3-0 2.9 31 1.4 2. 0 2. 2 2-3 2. 2 2. 2 1.9 2-3 2. I 2. 2 3. 0 1. 0 1.8 1.9 3.0 3. 1 2. 1 1.9 3. 1 3. 0 3. I 2. I 2. I 0. 0134 • 0318 .0391 .0387 .0388 ■ 0390 • 0400 .0384 .0358 •0334 .0314 0. 0307 •03S9 .0387 •0397 .0386 .0409 .0390 •0371 •0333 7.8 9.6 10. 2 10. 0 10. 1 lO. 0 9.9 9-S 9.4 9. I 9.0 '9-8' 10.5 10.4 10. 4 10. 0 10. 2 9-7 9.9 9-3 2-3 3-0 3-2 3-2 Z-i 3-3 l-i 3-2 3-0 3-0 2.9 3-0 3-2 3-2 3-2 3-2 3-1 3-1 3-1 i-S 1.8 3. I 3. 2 2-3 3. I 2-3 2-3 3. I 3. I 3.0 3. 0 2.0 2. I 3. 0 2.0 2. I 3. 0 3. I 3. I JULY 26 3 0.0271 0. 030s 9- S 9.4 2-4 3.8 1-7 1.9 0.0206 0.0174 8-3 8..? 2-S 3.4 1.6 1.4 3 • 0346 • 0320 9.8 9S 3.8 2.6 1.8 1.8 • 0406 •0399 10. I 10. 0 3-3 3-4 2. 0 3.0 4 .0407 • 039s 10.4 10. I 3-2 3-0 3. I 2.0 .0445 .0420 10.3 9.9 3-4 3-3 3.0 3. 3 S .0419 .0418 10.5 9.9 3-2 31 3. I 2. 2 .0445 .0451 10.4 9.8 3-3 3-3 2.4 3. 0 6 • 0403 .0408 10. I 9-7 3-1 3-2 3. I 2. I .0445 .0450 10. 4 10. 3 3-3 3-3 2- 2 3. I 7 •0359 • 0398 10. 2 9.9 2.8 3-2 2. I 3. 2 .0406 •0457 9.8 9.4 3-2 3-3 3. I 2-3 8 • 0400 .0400 lo. 0 10. 0 .1-1 3-0 3. 3 2. 2 •0436 •0443 10. 2 10. I 3-4 3-3 2. 3 3.4 9 .0366 .0378 9-7 9.3 3-0 31 2. I 2. 2 .0428 .0416 10. 4 9.8 3-3 3-4 3. 3 2-3 10 .0358 • 037s 10. 0 95 3-0 3- I 1.8 2. 0 .0409 .0422 10. 3 9.6 3-2 3-4 3. I 2-3 II •03ZS .0269 95 *■% 3-2 2.6 2. 0 1.9 .0360 •0397 9.4 9.4 3-2 3-3 3. I 3. 0 •0336 8.9 3.9 JXJI,Y 27 3 0.0194 0. 0387 8.3 lO. 3 3-S- •3-2 I- 5 3. 1 0. 0346 9.6 3-0 3- 3 3 .0405 .0479 10. 0 10. 6 3- I 3-2 1.9 2. 3 .0431 0.0406 9.8 9.9 3-1 3-3 3. 3 3. 3 4 •0417 . 0489 lO. 4 10. 6 3-5 3-S 3. I 3. 3 • 0479 .0436 10. 7 10.3 3-4 3-3 3. I 3. 3 S .0499 .0481 10. 6 10.5 3-4 3-4 2-3 2-4 .0467 .0498 10.8 lO. I 3-3 3-4 3.0 2-3 6 .0495 •0483 10.4 10.4 3-4 i-S 2-3 3-4 • 0483 .0514 10. s 10.7 3-2 3-i 2-3 2-4 7 .0501 .0483 10.3 10. 4 3-4 3-S 2. 2 3. 3 .0497 .0491 10.5 9-5 3-4 3-2 2-5 2-3 8 .0498 •046s lo. 2 10. I 3' 4 3-4 2-3 3. I .0498 .0499 10. 4 10. I i-3 3-S 3. 3 2-S 9 .0487 •044s lO- 3 10. I i-i 3-i 2-3 2-3 .0465 .0481 10.3 10. 3 3-3 i-S 3. 3 2-4 10 . 0480 .0440 9.9 9.9 3-4 3-2 2-S 2-4 .0479 • 043 s 10. 0 10. I 3-S 3-2 2-4 2- 2 IX •0443 .0430 9.8 9- S i3 3-4 2-S 2-3 .0455 .0429 10. 0 9-S 3-4 3-3 2-4 2- 2 13 •0435 .0408 9. 2 95 3-3 3-0 2-3 2.4 •0433 .0324 9.8 8-7 3-4 2- 7 2-3 2- I 13 •0330 8.6 3-0 3. 1 • 0392 9-3 3-2 2-3 14 • 0390 9.1 3-3 3. I JULY 28 3 0- 0289 8-5 3-0 1.9 0.0417 0.0359 10- 0 9-9 3-4 3-2 3. I 1.9 3 0.0352 .0491 9-6 10. 1 3-i 3-S 1.9 3. 3 .0477 .0484 9-8 10. s 3-S 3-4 2-4 2-3 4 .0417 .0512 10. I 10.3 3-2 3-S 2.0 2-3 •0S3I •0533 10.7 10. 3 3-S 3-S 2-S 2.3 5 .0471 -0523 9-9 lo. s 3S 35 3. 3 2-3 .0556 •OS44 10. 3 10. s 3-7 3-7 3.6 2-S 6 .0477 .0578 10. I 10. 6 3-4 3-7 2-3 a-S .0519 .0549 10. 3 10.4 3-7 3-S 2-S 2-4 7 .0500 •053 s 10. 0 10.4 3-S 3-7 3- 3 2-S .0540 •0529 10. 3 10. 3 3-6 3-6 2-S 2-4 8 -0503 .0560 10. 3 10. s 3-6 3-7 3. 3 3.6 .0510 .0521 10. 0 10. 2 3-4 3-4 2-S 2-S 9 .0479 .0544 10. 4 10. 4 3-4 3-7 3. 3 2-S .0467 .0487 10.3 10.3 3-3 3-S 2-3 2-4 10 .0494 .0526 9-9 10. I 3-S 3-f 3. 3 3.4 -0479 .0492 10. 0 10. X 3-S 3-S 3-4 2-S II .0505 .0529 9-9 9- 7 3-S 3-b 3-3 2-S .0467 •0493 9.8 9-7 3-3 3-6 2-3 2.6 13 .0446 .0487 9- 7 10. 0 3-4 3-6 2. I 2S .0426 .0456 10. 0 9.8 3-4 3-4 3.3 2-S 13 -0427 .0445 9.6 9.6 3-4 3-4 3. 3 3. 3 .0406 9-2 3-3 3-3 14 •0395 -0390 9.4 9.0 3-3 3-2 2-3 3. 3 Aug. 2, 1920 Daily Development of Kernels of Hannchen Barley 403 Table I. — Norntal growth of Hannchen barley in 12-hour periods from flowering to maturity, at Aberdeen, Idaho, in igiy — Continued 6 a. m. 7 p.m. Ker- nel No. from base of spike. Weight of ker- nel on spike— Length of kernel on spike — Lateral diameter of kernel on spike— Dorso- ventral diameter of kernel on spike — Weight of ker- nel on spike^ Length of kernel on spike — Lateral diameter of kernel on spike— Dorso- ventral diameter of kernel on spike— A. B. A. B. A. B. A. B. A. B. A. B. A. B. A. B. Gm. Gm. Mm. Mm. Mm. Mm,. Mm. Mm. Gm. Gm. Mm. Mm. Mm.. Mm.. Mm. Mm. 3 3 4 S 6 7 8 9 10 II 13 0. 0228 .0492 •0513 .0550 .0550 .0504 .0492 .0487 .0498 .0465 .0406 0.0385 .0485 .0502 .0496 .05x1 .0497 .0506 .0485 .0446 .040s 3-S 9.6 9-3 10. 4 9.6 10. 0 10. 1 9.9 9.4 9.6 9.0 9-7 10.3 10.3 10. 2 9.8 10.3 9.8 10. 0 9.6 9.0 2.7 3-4 3-7 3-7 3-8 3-4 3-4 3-6 3-7 3-5 3-3 3-3 3-5 3-6 3-6 3-7 3-4 3-S is 3-4 3-i 1.6 2-3 2.4 2-S 2S 2. 2 2-3 2.6 2-S 2-3 2.4 2.0 3. 1 2.4 2.4 2-5 2-3 2-5 2-S 2.4 2.4 0-0377 .0507 •0532 •0537 •0538 .05x4 .0521 .0520 .0465 .0448 0.0434 .0482 • 0537 • 0532 .0522 •0513 .0522 .0492 .0502 .04x8 9.8 10. 0 10. I 10. I X0.3 9- 7 10. X 9.9 9.4 9.2 9.8 9.8 10.3 10. 0 xo. 3 XO. X xo. X 10. 4 9- 7 8.7 3-0 3-6 3-6 3-5 3-5 3-4 3-6 3-4 3-4 3-4 3-S 3-S 3-S 3-6 11 3-6 3-S 3-6 3-4 2.0 2-S 3-5 2-S 2-S 2-5 2.6 2.6 3.4 2.4 2-3 2.S 2-S 2-S 2-4 2-4 3.6 2-S 2.6 2.4 JULY 30 I 1 0.0478 0. 0430 9.8 10. 2 3-S 3-3 2.4 2-3 0. 0474 0. 0426 9.0 8.8 3-3 3-4 a- 3 2-3 3 .0509 •0517 10. I 10. 2 3-7 3-7 2-3 2-S .0565 .05x7 9.6 9.2 3-7 3-(> 2-S 2S 4 •OSS7 .0550 IX. 0 10. s 3-7 3-6 2-4 3.4 .0588 • 0553 9-7 9.2 3-7 3-8 2. S 2-S S •OS 74 •OS 59 10.8 10.6 3-7 3-6 3.4 2-S .0600 .0589 9.8 9-S 3-7 3-7 3-5 2-6 6 .OS 79 .0541 10.7 II. 0 3-7 3-6 2-S 2-S .0620 .0560 9.8 9-2 3-8 3-7 2.6 2-7 7 8 •0551 .0498 •OS3S xo. 2 10.7 3-6 3-6 2-S 2-4 .0600 • 0577 10. 0 9.1 3-7 3-8 2.6 2-7 3-4 3-S 2-S 2-3 .0627 ■OS34 9.9 3-8 3-S 2.7 3.8 9 .0520 .0499 10. 0 10.3 3-4 3-S 2-S 2-S .0600 .0531 9.8 8.6 3-7 3-7 2-7 2-7 10 .0487 .0485 10. 0 10.3 3-4 3-4 2-S 2.4 • 0531 .0500 9-0 8.6 3-7 3-S 2-S 2-6 II •049s 9-S 3-8 2-S • 0579 .0455 9-4 8-3 3-7 3-4 3.6 2-S 11 •0399 9.6 3-4 2-3 • 0543 9-1 3-4 2.6 13 .0478 8-5 3-S 2-5 JULY 31 6 a. m. 6.30 p. m. 3 3 3 4 S 6 7 8 9 10 II 12 0.0409 .05x1 •0559 •0565 • 0561 .0552 .0525 .0521 .0487 .0488 .0388 0. 0443 •0503 •OS 79 •0575 .0582 •0550 •0575 .0581 •OSS 7 •0552 •0534 . 0469 .0361 7-9 9-2 9.0 9-0 9-1 9-0 8-7 8.9 8.7 8.4 7-9 8.9 9-1 9.8 9-7 9-7 9-7 9.6 9-3 9-1 9.2 9- I 8-8 8.0 3-3 3-8 3-8 3-7 3-7 3-8 3-7 3-S 3-S 3-S 3-3 3-4 3-S 3-7 3-7 3-7 3-S 3-S 3-S 3-5 2-S 3-S 3-3 3-0 2-4 2-S 3-6 2-S 2-6 2.6 2.6 2.6 2.6 2.5 2.3 3.4 3.4 3.5 2-5 3-5 3-5 2.6 2.6 2.6 2.6 2.6 2-5 2.3 0. 0463 .0545 .06x5 .0600 .0588 •0593 .0574 • 0537 • 0533 .0510 0.0374 • 0493 •OSS4 • 0565 • 0552 •05S3 •OS S3 • 0515 .0512 .0450 .0370 7-5 8.1 9-5 9.2 9.2 9-S 9-3 8.7 8-5 8-7 8-0 9.0 8.3 9.0 9.0 9.1 8.9 8.5 8-7 8.2 7-6 3-S 3-7 3-7 3-7 3-7 3-7 3-7 3-6 3-S 3-7 3-3 3S 3-7 3-7 3-6 3-7 3-7 3-4 3-6 3-5 3.0 2.4 2.6 2.8 2- 7 2- 7 3. 7 2. 7 2. S 2.6 2.6 2-3 2-S 3-7 2-7 ,.6 2-S 2-S 2-S J. 6 2-4 2-4 M AU GUST I I 3 0.0351 0.0451 ■8.'4' 8.8 3-2 3-5 2.0 2-3 0.0331 8.0 3.0 1-9 3 •0S15 •0554 9.2 9-S 3-6 3-7 2-4 2.3 0.0500 •0525 9.1 9.0 3-3 3-7 3.3 2-3 4 •0577 -0625 9-8 9.6 3-S 3-7 2.5 2.6 .05C6 •OS97 9.0 9.6 3-6 ,V6 3.6 2. 7 S .0614 .0611 9-5 9-4 3-7 3-8 2.6 2.6 •0575 .0604 9-2 9-3 3-8 3-8 3.6 3. 7 6 .0588 •0599 9-5 9.8 3-8 3-8 2-5 2.6 .0598 .0589 9-3 Q.6 3-9 3-8 2.6 2. 7 7 • 0583 .0615 9-2 9-3 3-7 3-9 2.6 2.6 • 0598 .0589 9-4 9.0 3-7 3-8 2-5 3.6 8 .0581 .0594 9-4 9-4 3-6 3-7 2.6 2.5 •0578 .0596 9-0 9.1 3-7 3-8 3.6 2. 7 9 .0589 .0589 9-3 9-4 3-8 3-7 2.6 3-6 .0571 .0589 9-3 9-3 3-7 3-8 2-7 2.5 10 .0563 •OSS 7 9-4 9-2 3-7 2.7 3.6 3.6 .0571 .0558 9.0 9.0 3-7 ,3-6 2.7 2-S II -0550 • 0521 9-1 8-7 3-6 3-6 2.6 2.4 .0536 .0556 9.0 9-0 3-7 3-7 2.6 2- S 12 .0527 •0473 9.0 8.4 3-5 3-4 2-5 a- 5 .0500 .0505 8.6 8.6 3-3 3-5 2.6 2-4 13 -0514 8.6 3-5 2-5 •0454 •0443 8.3 8.4 3-4 3-2 2.4 2-4 .0450 8.4 3-3 2.4 1 404 Journal of Agricultural Research Vol. XIX. No. 9 Table I. — Normal growth of Hannchen barley in 12-hour periods from flowering to maturity, at Aberdeen, Idaho, in igij — Continued 6 a. m. 7 p.m. Ker- nel No. from base of spike. Weight of ker- nel on spike— Length of kernel on spilce — Lateral diameter of kernel on spike — Dorso- ventral diameter of kernel on spike — Weight of ker- nel on spike — Length of kernel on spike — Lateral diameter of kernel on spike — Dorso- ventral diameter of kernel on spike — A. B. A. B. A. B. A. B. A. B. A. B. A. B. A. B. Gw. Gm. Mm. Mm. Mm. Mm,. Mm. Mm. Gm,. Gm. Mm. Mm. Mm. Mm. Mm. Mm. 0. 0301 .0500 .0617 .0611 .0620 •OS 79 .0601 •OS77 •059s •0574 .0476 .0489 0.0528 .0596 .0620 .0644 .0600 .0610 .0592 • 0572 •0515 .0521 .0429 7-7 8.9 9.6 9.6 9.8 9.4 8.8 9. I 9.0 9.2 8.7 8-5 9. I 9.4 95 9.6 9-3 9.4 9.2 9.0 8.8 8.7 8.0 2.8 3-6 3-8 3-8 3-8 3-7 3-7 3-7 3-8 3-5 3-5 3-5 3-4 3-8 3-8 3-8 3-8 3-6 3-7 3-8 i-S 3-5 2.0 2.4 2-5 2.6 2.6 2-5 2-5 2.6 2.8 2-S 2-S 2-5 0. 0404 .0483 • 0543 • 0571 .0566 .0586 .0561 .0548 .0500 .0464 • 0387 8.8 9.4 91 9-0 9.0 91 8-7 8.1 7.8 8 I 8-7 9-0 9.1 9.1 9.0 9. I 8.7 8-6 8.3 7-5 3-3 3-6 3-9 3-7 3-7 3-6 3-6 3-5 3 5 3-2 33 3-6 3-7 3-8 3-6 3-8 3-7 3-S 3-5 3 5 3-2 2-S 2-7 2.6 2-S 2.S 2.6 2-5 2-5 2 5 2-4 3 4 5 6 7 8 9 10 II 12 13 2.4 2.6 2.7 2.7 2.6 2.6 2.7 2.7 2-5 2.7 2.4 0.0426 • 0546 • 0553 .0566 • 0546 •0558 •0547 .0508 .0465 •0371 3-3 2-S 2.6 2.6 2-7 2.8 2-7 2-S 2-S 2. a 6 a.m. 6.45 p. m. 2 3 4 5 6 7 8 9 10 II 12 13 14 0. 0363 .0510 .0560 -0573 .0621 .os8o -OS 73 .0602 .0600 -053X -0531 •053a -0452 0. 0510 -0572 -0595 .0608 .0602 -0579 •0583 -OS77 •0545 • 0504 .0507 8.0 9.2 9-5 9.2 9.2 9.0 9.1 8.8 8-7 8.7 8.7 8-5 8.2 S.S 9.0 9.6 9-2 9-3 9.2 9.0 8.9 8.7 8.4 8.8 3-3 3-6 3-5 3-6 3-7 3-6 3-6 3-6 3-7 3-5 3-6 3-6 3-2 3-5 3-8 3-8 3-8 3-8 3-5 3-5 3-8 3-7 3-7 3-2 2- 2 2-3 2-4 2.6 2-7 2.6 2-S 2-S 2-7 2.6 2.6 2.7 2.6 2-S 2.6 2.6 2-7 2-S 2-S 2.6 2.6 2.6 2.6 2-3 0.0521 • 0639 .0644 • 0650 •0651 • 0633 • 0651 .0629 -0599 .0591 • 0536 0.042s . 0548 -0565 .0558 • 0551 • 0554 • 0540 • 0513 .0476 • 0437 9.2 9.1 9.4 9.6 9-4 9-3 9-3 9-2 9.1 8.8 8.4 7-9 9-8 9.0 8.8 9.0 8.7 8-5 8.0 7.8 7-6 3-3 3-8 4.0 3-7 3-8 3-8 3-9 3-8 3-7 3-7 3-5 3-3 3-7 3-6 3-6 3-6 3-7 3-6 3-5 3-3 3-4 2-S 2-7 2.7 2.6 2-S 2.8 2.8 2.6 2-7 2.7 2-7 2-S 2-7 2.6 2-S 2-5 2.6 2-7 2.8 2-S 2-S 5-45 a.m. 6.4s p. m. 2 3 4 5 6 7 8 9 10 II 12 13 14 0. 0454 •0537 •OS94 .0605 •0577 .0580 •0539 .0556 •0539 •0475 .0400 0.0409 •0550 .0600 • oiige .0626 .0617 .0592 •OSS 7 •0565 -0550 .0480 8.0 8.7 9-1 9-3 9-2 9- I 9.0 8.8 90 8.2 7-8 8-3 8.9 9-3 9.6 9.6 9.6 9.0 9.1 9.0 9.0 8.3 3-4 3-6 3-7 3-7 3-6 3-5 3-4 3-7 3-7 3-5 3-2 3-4 3-7 3-8 3-7 3-7 3-8 3-8 3-5 3-5 3-5 3-4 2-4 2.4 2-7 2.8 2.6 2.6 2-S 2-7 2.8 2.6 2.4 2-3 2.6 2-5 2.6 2-7 2-9 2.6 2-7 2.8 2-7 2.6 0. 0300 .0567 .0656 .0629 .0629 .0660 .0614 .0614 .0546 • 0546 •0532 .0448 0. 0386 .0623 .066S .0675 .0645 • 0655 .0660 -0633 .0617 • 0596 .0582 .0548 • 0483 7-8 9.0 9-3 9-6 9-3 9.2 9.4 9-1 9.0 8.9 8-5 7-9 8.2 9-3 9-7 9.8 9-6 9-4 9-4 9-1 9-1 9.0 9.0 8.5 8-3 2.8 3-7 3-9 3-7 3-8 3-7 3-7 3-7 3S 3-7 3-5 3-3 3-1 3-7 3-8 3-7 3-7 3-7 3-8 3-7 3-8 3-7 3-7 3-5 3-4 2. 0 3.6 2.7 2.8 2.8 2.9 2.8 2.8 2- 7 2-S 2.6 2.6 a. I a- 7 a. 6 2.7 2.7 2.7 a^8 2.7 2-7 a. 8 2r7 3.8 2.6 Aug. 2. 1920 Daily Development of Kernels of Hannchen Barley 405 Table I. — Normal growth of Hannchen barley in 12-hour periods from flowering to maturity, at Aberdeen, Idaho, in igij — Continued AUGUST S 6 a. m. 1 6. 30 p.m. 1 Ker- nel No. from base of spike. Weight of ker- nel on spike — Length of kernel on spike — Lateral diameter of kernel on spike — Dorso- dLmetirl Weight of ker- of kernel' nelonspJ^e- on spike — Length of kernel on spike — Lateral diameter of kernel on spike — DorSO- ventral diameter of kernel on spike — A. B. A. B. A. B. A. B. A. B. A. B. A. B. A. B. Gtn, Gm. Mm. Mm. Mm. Mm. Mm. Mm. Gm. Gm. Mm. Mm. Mm. Mm. Mm. Mm. a 3 4 S 6 7 8 9 10 II 0.0253 .0590 • 0631 .0637 .0671 .0627 .0609 •0593 .0S64 .0507 o- 0461 •0543 .0601 .0559 •058s .0562 •053 s •0527 .0503 .0446 7.3 91 8.9 9-3 9.4 8-7 9-1 9.0 8.S 8-3 8.1 8.8 8.8 9-2 8.9 8.7 8-7 8-5 8.4 7-7 3.6 3-8 3-8 3-8 3-8 3-9 3-7 3-7 3-6 3-4 3-4 3-6 3-7 l-S 3-7 3-7 i-i 3-6 35 3-4 1.9 2-7 2.8 2.7 2.8 2-7 2.8 2.7 2.7 2-7 2.4:0. 0300 2.5 .0538 2.8 .0597 2.7 1 -0635 2.7 . 0600 3.6 .0639 3. 6 . 0598 3.6 ! .0587 3.6 ! .0574 2. 6 . 0539 0. 0538 .0592 .0596 .0614 .0614 .0588 .0551 .0542 .0466 .0441 7-0 8.9 8.7 9-3 8.7 8.8 8.6 8.8 8-7 8.S 8.3 8.1 9. 3 9.0 9. I 90 8.9 8.2 8-5 8.0 7-6 3-0 3-5 3-6 3-9 3-8 3-7 3-7 3-8 3-7 3-4 3-4 ii 3-S 3-7 3-S 3- 7 3-7 3-7 3-4 3-S 3-4 3-4 3-0 »-3 a- 5 2.6 3.6 3.8 a- 7 3.6 3.7 3.6 a-S a-S 3'S 3.6 3.7 3.S a. 7 a. 7 3-6 a- 7 a-S a. 4 6 a. m. 6.40 p. m. 0.0231 •OS34 7- a 8-7 3.6 1.8 3 0.0528 0.0560 8.9 8.9 3-7 3-7 a-S 3.6 0.0513 8.7 3-S 3S a- 5 a-S 3 .0639 .0606 9-4 8.9 3-8 .3-8 3.7 3.8 .0617 .0578 9- a 8.9 3-6 3-6 3.8 4 .0634 . 0615 9.4 9.0 3-8 3-7 3.8 3.7 .0648 .0580 9- a 8.3 3-7 3-S 3.3 ■ 5 .0657 .062a 9.6 9-4 3-8 3-8 3.8 3.8 .0624 .0566 9-3 8.9 3-6 3-S a- 7 6 .0675 •0634 9-7 91 3-9 .3-8 3.9 a. 7 .0634 •0553 91 8.6 3-S 3-S a- 7 7 .0628 .060a 91 8.9 .1-8 ,3-6 a- 7 a- 7 .0613 .0561 8-7 8-8 3-7 3-S a- 7 8 .0626 .0566 9-4 8.9 3-8 3-6 3.9 3.6 .0589 .0512 9.0 3.3 3-S 3-4 3.6 9 .0633 .0550 9.0 8.4 3-7 3-S a. 9 3.6 .0576 .0476 8.3 8.1 3-S 3-3 3.6 xo .0580 •0528 8.6 8.3 3-6 3-5 3.8 3.6 .0509 .0479 8.5 8.1 3-4 3-3 a. 6 11 .0576 ■0513 8.7 «.a 3-6 3-S 3.6 3-7 .0491 •0330 8.3 7.6 3-4 a. 9 3.6 a. 3 12 •0534 .0487 «-S 8. a 3-S 3-S a. 7 3.6 .0444 8.1 3- a 3.4 6 a. m. 1 7 p. m. 1 1 I a 0.0571 0. 0568 8.6' S.'i 3-7 .3-6 a-S 3.6 0.0541 0.0523 '8.' 7' S-V 3-7 3-S 2.6 a.S 3 ■ 0616 .0641 9-S 9.0 3-7 3-9 a- 5 3.8 .0567 .0593 9.0 8.9 3-6 1-8 a. 8 a. 3 4 .0657 .0646 9.4 9.1 3-7 ,3-8 3.6 3.8 •0579 .0619 g. 0 90 3-7 3-8 a- 7 a.S S .0699 .0623 9.6 9.4 .3-8 3-6 3-0 2.6 • 0561 .0626 8-5 3.9 3-S 3- 7 3.6 a. 7 6 .0672 •0623 9.4 9-3 3-6 3-7 2.9 3.7 .0605 .0618 9-3 9.0 3-7 3-7 3.8 a- 7 7 .0674 •0623 9.3 8.8 3-7 3-7 3.8 3.8 .0589 .0610 8.9 8.9 3-7 3- 7 a. 7 3.8 8 .0639 .0627 9.0 8.7 3-7 3-6 2.8 3.9 •0575 .0600 8.7 90 3-S 3-7 3.6 3.3 9 .0593 •0585 91 8.8 3-S 3-7 2.6 3.6 .0558 .0597 8.7 8.S 3-S 3-7 a. 8 a. 7 10 .0626 .0570 8.7 8-S 3-8 3-6 3.7 3.8 •0559 -OS 73 8-S 8-5 3-S 3-S 3.8 3.6 II .0591 •0533 8.5 8.5 3-7 3-6 3.6 a- 7 .0480 .0516 8.0 8-S 3-4 3- a 3.6 3.6 12 .0576 .046s 8.4 8-3 3-5 3-4 a. 7 a-S .0432 .0488 8.2 8.0 3a 3-4 a-S a.S 13 .0480 8.2 3-4 3.6 •0399 8.0 3-1 3-3 4o6 Journal of Agricultural Research Vol. XIX. No. 9 Table I. — Normal growth of Hannchen barley in 12-hour periods frofn flowering to maturity, at Aberdeen, Idaho, in igiy — Continued 6 a. m. 6.40 p. m. Ker- nel No. .. from , base ^ of spike. V^eight of ker- lel on spike — Length of kernel on spike — Lateral diameter of kernel on spike— Dorso- ventral diameter of kernel on spike. Weight of ker- nel on spike— Length of kernel on spike— Lateral diameter of kernel on spike — Dorso- ventral diameter of kernel on spike — A. B. A. B. A. B. A. B. A. B. A. B. A. B. A. B. Gm. Gm. Mm. Mm. Mm. Mm. Mm. Mm. Gm. Gm. Mm. Mm. Mm. Mm. Mm. Mm. I 0 0363 0. 0243 7- 7 6.7 3- 0 2. 7 2. 1 1.9 2 059s •0532 8.6 8.0 3-6 3-6 2- 7 2-S 0.0440 7-8 .3-3 3-5 3 0613 .0618 8.8 8.7 3-6 3-8 2.6 2.8 0. 0502 • 0571 8.4 8.4 3-S 3-5 2-7 3-6 4 0655 .0634 9-3 8.9 .3-8 3-7 2.9 2.7 •OS78 .0609 8.6 90 3-6 3-6 2.7 2.7 5 0682 .0647 9. 2 9.3 3-6 3-8 2-7 2.8 .0590 .0570 9.0 8.8 3-6 3-6 2.7 2.6 6 0659 .0624 90 8-7 .3-7 3-7 2.7 2.8 • 0569 .0576 8.8 8.8 3-S 3-'? 2.6 2.6 7 0659 .0626 9.0 8.8 .3-6 3-6 2.8 2.8 .0581 •OS54 8.8 8.9 3 6 3-S 2.7 3.6 8 0634 .0615 8.9 8.6 3-6 3-7 2.7 2.8 .0565 ■0553 8.6 8..'; 3-5 3-5 2.6 3-7 9 0642 .0604 8.6 8.4 3- 7 3-7 2.8 2- 7 .0508 •0539 8.6 8.6 3-4 3-4 3-5 3.6 10 0607 • 0590 8.2 8.4 3-6 3-7 2.8 2.8 • 0511 .0507 8.4 8.4 3-5 3-S 2.6 3-7 II 0585 .0500 8.6 8.0 3-5 3-5 2-7 2.4 .0423 • 0412 7-8 7-6 3-» 3-1 J. 4 3-S 8.1 7- 7 0420 7-7 3-1 2-S The course of growth is more apparent in Table II, where the results have been summarized in 24-hour periods. That is, the weights and measurements of all the kernels on each spike of the two samples have been averaged and these averages placed opposite the dates. No summary was made for the data in 12 -hour periods, because the daily changes of dimension after the first few days were too slight to be shown in so short a period. Growth in length in the early stages, however, is per- fectly apparent in 12 hours. Table II. — Average wet weight, length, lateral diameter, and dorsoventral diameter of kernels of Hannchen barley in 24-hour periods from flowering to maturity at Aberdeen, Idaho, in igiy Date. Wet weight. Length. Lateral diameter. Dorso- ventral diameter. Date. Wet weight. Length. Lateral diameter. Dorso- ventral diameter. Mgm. Mm. Mm. Mm. Mgm, Mm. Mm. Mm. July IS 1.9 2. 27 I. 26 July 28 48. 2 10. 02 3-47 2-33 16 3-0 2.88 1.32 0.74 29 49.0 9.84 3-50 2. 42 17 4-S 3-93 1-39 .82 30 53- 0 9.99 3-59 2.50 18 8.0 6. 05 1-57 .90 31 52.2 8.83 3- .0 2-54 19 12.7 7.91 1.79 I. 16 Aug. I 54-8 9. 10 3.62 2.51 20 15.8 8.78 1.94 1.30 2 53-3 8.87 3-59 2-54 21 19.2 9-32 2. 10 1.48 3 .■;.=;• 8 8.87 3.61 2-59 22 25-7 9. 81 2-51 I. 71 4 56.3 8.96 3.60 2. 64 23 31.0 9.96 2.80 I. 90 5 55- 9 8.66 3.60 2.62 24 32-7 9-83 2.86 I. 90 6 ^6.6 8.78 3-57 2. 67 25 36.0 9.80 3. 02 2.05 7 t;8. 0 8.77 3.60 2.68 26 38.2 9-77 3-09 2. 06 8 56.1 8- 53 3-53 2.64 27 44.6 9.98 2>-Z<^ 2.25 Aug. 2, 1920 Daily Development of Kernels of Hannchen Barley 407 The significance of the data in Table II is, perhaps, more easily seen in figure i. The most surprising feature shown by this figure is the remarkably rapid growth in length following fertilization. In the two days from the second to the fourth after fertilization, half the growth in length occurs. The insufficiency of 3-day intervals in sampling at this stage is obvious. Distinct growth is shown in 12-hour periods, and it is probable that consistent increase would be revealed in 6-hour periods. The kernel reached its maximum length by the end of 7 days in each year. After the peak of length is reached, there is a gradual decrease to maturity. This is discussed later in connection with figure 4. The lateral diameter exhibits its most rapid increase as soon as the rate of the growth in length diminishes. This increase continues until about the fifteenth day, after which the lateral diameter remains more or less stationary. The dorsoventral diameter, on the other hand, continues to increase almost until maturity. The increase is somewhat less than in the lateral diameter, there being a greater divergence in the growth curves at the end of the growing period than at the beginning. The effect of the better irrigation in 191 6 is apparent throughout the period of growth. There is a possibility that the 191 6 samples are a few hours farther advanced throughout the series because of differences in temperature or other factors at flowering time. While growth itself is not so easily affected, fertilization is often hastened or delayed many hours by conditions in the environment. During the early growth of the kernel the ovary tip undergoes a sym- pathetic development. When the kernel is first developing, the growth is largely in the pericarp. Some of the tissues surrounding the embryo sac and the ovary walls of the same region develop rapidly and are to be found in the ripened caryopsis. For some reason, the tissues above the embryo sac are temporarily stimulated, forming a body at the end of the kernel, which is referred to here as the ovary tip. This growth, which may be seen in Plates 83 and 84, is of importance because it introduces an error in the measurement of length. After the growth of the first few days, this organ remains stationary in size for a while and finally is largely resorbed. In figure i it will be seen that it was possible to measure the kernel proper without this tip by the fifteenth day after flowering. The records of lengths until that time included the ovary tip. The lateral and dorsoventral diameters of the ovary tip are shown in figure 4. It is probable that the length of the kernel proper increased somewhat after it had apparently reached its maximum by invading the tissues of the ovary tip. This tissue is probably partially responsible for the differ- ence of measurements in the first 15 days of the two years. A second factor in the error lies in the softness of the structure at the base of the kernel. In the early stages of growth it is exceedingly difficult to place the caliper bar at exactly the right point, and in 1917 the kernels may have been measured more closely than in 191 6. The difference of the 4o8 Journal of Agricultural Research Vol. XIX, No. 9 two years, however, is less than 0.5 mm., so that the data coincide far beyond any reasonable expectation. The reason that errors in this connection are suggested is that it is not plausible that the differences in soil or water would affect the kernels by /4? / \ «* .•-■ '% ^*' '*», \ \ 1 1 1 1 1 / / > -c »* -5* ^^* 1 1 1 1 T ~ / 1 1 1 1 1 P* " ... -'' - --■ ^« "%, ■'■ **^ -'' »'' "■ 'Z ^ /' \. / / / / / ^. — _x ^ - - - -^ - ./ — — — "^ i7 3« } > > .< r - ^y !^ c^?- V. ac n* C «« fV ';^ y^ ^ > ^ s Z- a< \sc '/4 TV/ "T^ f^ ^su ■«^ c f«' py 7'^ ^ Fig. 4. — Graph showing lateral and dorsoventral diameters of the ovary tip as compared with length, lateral diameter, and dorseventral diameter of the kernel for the 26 days following flowering in 1916. this time, since there obviously is no insufficiency of nutrients for this primary growth. These data have an incidental bearing on the value of observations on the length and diameter of kernels. Such observations are used frequently to identify varieties. In 1914, the author (2) stated Aug. 2. 1920 Daily Development of Kernels of Hannchen Barley 409 that length was more dependable than lateral diameter and that lateral diameter was more dependable than dorsoventral diameter in the descrip- tion of types. The same obser^'^ations have been made, presumably, by many others. The growth curves confirm this opinion. The length is quickly attained and should vary little with season. The lateral diameter reaches its maximum more slowly than the length, but much sooner than the dorsoventral diameter, which is dependent upon conditions throughout the growing season for the fulfillment of its maximum pos- sibilities. As has been inferred before, the kernels at the base and the tip of the spike are more variable than those near the center. With the increase of the number of kernels on the spike those at the extremes are likely to suffer from competition. On any spike, if nutrition at any time becomes insufficient, the basal and the apical kernels are the first to be affected. Averages which include these kernels show greater fluctuations than those from which they are excluded. This variation was overcome partially by including in the averages no basal kernels which weighed less than half as much as the kernel next above. Since this does not entirely overcome the difficulty the average length, lateral diameter, and dorsoventral diameter of kernels 6, 7, and 8 are plotted in figure 5 as an illustration of the behavior of more typical kernels. It will be seen that the daily fluctuations are much reduced. EFFECT OF POSITION OF KERNEL ON GROWTH There are two main factors that affect the relative size of kernels. These are age and the position of the kernel on the spike. The relative importance of these factors varies with the stage of growth. The age of the kernel depends on the time of flowering. The florets of a spike are not all fertilized on the same day. The earliest flowers usually are those located about two-thirds the distance from the base of the spike to the tip. The last to fertilize are the extreme basal and apical florets. The largest florets are found one-third the length of the spike from the base. Presumably, the kernels found in these florets receive more nourishment than those at the tip, especially toward the end of the growing period. The length of each kernel on one side of the spike is shown by days in figure 6. The growth is practically completed in these eight days. As will be seen, florets 8, 9, and 10 are the first to fertilize and to begin growth. By the third day these three kernels have reached their greatest relative advancement. After the second day there is a gradual shift in the peak of the curve as the basal kernels approach the others in total length. By the fourth day kernels 8, 9, and 10 are no longer prominent, and on the fifth day the curv^e is extremely regular. By the eighth day the length growth is complete and the longest kernels are the fifth and sixth. The curve of the eighth is thought to be more typical than that of the seventh. 183718°— 20 2 4IO Journal of Agricultural Research voi. xix. No. 9 The data for the lateral diameter are shown in the same way in figure 7. The same progress of size from the eighth, ninth, and tenth kernels toward the base is to be seen as was found in the length. Here also, the fifth and sixth kernels have exceeded the upper ones by the eighth day /y/jr / ^ N / ■^ s rf"* / / so / f / s ^ y N e.s / k> ^. / ^.0 AS S.C ' \ 1 ...■ ... -• >... ... ... ... '.. ,,. *.^ ,.•' "•• y / "' ,^ ^ - ■- - ,. /■ -^ -" ^ y -'' / X /.S" "• / x"' y _<- ■- "^ ^c/^.>^ /ff /■? /s- /^ ^t>^/^^ ^s.e^^j'^s^ ^^^ 104 -216 ■375 -380 -095 -495 -679 -352 -656 •738 .668 ■ 230 .062 .486 - 622 •707 ■154 •725 ■552 -913 ■ 510 .116 ■430 -300 -518 •707 -961 -938 •352 -360 -70s .861 -644 -781 - 204 -541 .282 -987 •314 .807 •145 •875 -044 •453 -643 • 562 ■322 •547 Nitro- gen per ker- nel. Ash per ker- nel. Mgm. Mgm. 0.254 0-346 .281 -400 ■323 •432 •345 • 486 -339 -467 ■379 .486 •375 .461 •453 -519 -458 -435 •493 -530 • 461 .566 -548 .526 •475 -607 •552 .606 .464 .526 .501 -538 •509 .582 .632 •519 .488 -539 .626 -628 •570 •536 •577 •573 ^655 •549 •637 .581 -752 -634 .626 •744 .665 •541 The material in Table III is summarized in Table IV. In the first part of Table IV the spikes of each sample are combined so as to give the average growth in 12-hour periods. In the second part of the table the 4i6 Journal of Agricultural Research Vol. XIX, No. 9 morning and evening averages are united to give the average growth in 24-hour periods. While many points of interest are perfectly apparent in the table, it is more convenient to discuss these data under the headings of the separate constituents, where the results are represented graphically. Table IV. — Average percentage of dry matter and water per kernel in Hannchen barley, percentage of nitrogen and ash in dry matter, and actual total weight, weight of dry m,atter, water, nitrogen, and ash at 12-hour and 24-hour periods at Aberdeen, Idaho, in igij 12-HOUR PERIODS Time. July 15 16 16 17 17 18 18 19 19 20 20 21 21 22 22 23 23 24 24 25 25 26 26 27 27 28 28 29 29 30 30 31 Aug. I, I, 2, 2, 3. 3. 4. 4. 5. 5. 6, 6, 7. 7. 8, 8, m. m. m. m. m. m. m. m. m. m. m. m. m. m. m. m. m. m. m. m. m. m. m. m. m. m. m. m. m. m. m. m. m. m. . m. m. . m. m. . m. m. , m. Dry matter. Water. Per cl. Per a. 84. I 85-5 82.6 81.6 80. 2 80.8 79.1 78.8 77.2 77-4 77.8 77.1 74-7 74.1 74-3 72. 2 70.8 70. I 68. I 67.9 64.9 65. o 62.6 63-4 61. 9 62.5 58.6 60. o 57-6 57-8 55- o 55-9 53-2 52-9 50-9 52.6 50.1 49.9 49.6 5°- 7 47-8 48.5 45-7 47.2 42.7 45-3 42. o 43-6 42. I Nitro- gen in dry- matter. Per ct. 2. 27 3-14 2-33 2.33 2. 18 1-95 2.08 2. 17 I. 96 1. 92 2. 04 1.87 2. 04 2. 02 2. 09 1.86 1.86 2. 19 2. 02 2. 00 2. 09 2. 14 1.89 2.08 1.82 1.97 1.85 2. 04 1.85 2. 18 1.86 1. 91 2.03 1.97 2. 27 Ash in dry matter. Per ct. 7.91 8.96 5.86 6. 14 4. 02 4. 20 3.80 4-63 3-54 3. 21 3-79 4. 18 3-09 3-07 2.77 2.88 2-93 2. 62 2. 72 2. 6r 2.84 2.41 2. 42 2.49 2.43 2. 10 2.28 2. 41 2. 22 2. 27 2. 25 1.94 2. 12 2. 10 2. 00 1.97 1.97 1-93 1.94 1.83 1-95 1-93 I- 93 1.98 1-77 Wet weight. Mgm. 9 5 5 2 8 5 4 6 Dry matter. Mgin. 0-3 .4 .6 .6 I. I 1.4 1.8 2-5 3-2 3-6 3-5 4.1 5-2 9.1 9.4 10. 9 11. 7 12.5 12.8 14.9 16. 2 17. I 17.8 20. 2 19-3 21. o 21. 7 23-9 22.8 24.7 25-9 26. 26, 25' 27. 28. 26. 30' 29. 29 31' 31 32' ?,:,■ 3°- Water. Mgm. 1.6 2. I 2.9 2.6 4.6 6. I 6.6 9.1 10. 7 12. 2 12.3 13.6 IS- 4 10. I 22.3 22. o 22. o 23-3 24. 6 23-3 23.8 24.9 28. I 27.8 29. 6 28.6 29. o 28. 5 29.8 30. 6 28.9 28. I 29. I 27.7 29. 2 25 27 5 5 28.0 27.7 27.7 27. 6 25.1 27. 23- 27. 23- 25- 22 Nitro- gen. Mgm. o. 01 02 03 02 04 76 Aug. 2. I920 Daily Development of Kernels of Hannchen Barley 417 Tabl^ IV. — Average percentage of dry -matter and water per kernel in Hannchen barley, percentage of nitrogen and ash in dry matter, ayid actual total weight, weight of dry matter, water, nitrogen, and ash in 12-hour and 24-hour periods at Aberdeen, Idaho' in igij — Continued 24-HOUR PERIODS Time. Dry matter. Water. Nitro- gen in dry matter. Ash in dry matter. Wet weight. Dry matter. Water. Nitro- gen. Pcrct. Per ct. Per ct. Per ct. Mgm. Mgm. Mgm. Mgm. 15- 9 84. I 3-75 1.9 0-3 1.6 0. 01 16. 0 84 0 4-32 7.91 3-0 •5 2-5 .02 19. I 80 9 3.06 7.41 4-5 •9 3-6 •03 20. I 79 9 6. 14 8.0 1.6 6.4 22. 0 7« 0 4. II 12.7 2.8 9.9 22. 4 77 6 2. 27 4.22 15- 8 3-5 12.3 .08 24. I 75 9 3- 14 3-54 19.2 4.6 14-5 •13 25.8 74 2 2-33 3-50 25-7 6.6 19. I •15 28. 5 71 5 2. 07 3- 64 31-0 8.8 22. 2 .18 30-9 69 I 2- 13 2. 92 32-7 10. I 22. 6 . 22 Z3-(> 66 4 1.94 2. 91 36.0 12. I 24. 0 •23 36.2 63 8 I. 96 2. 67 38.2 13-9 24.4 •27 37-4 62 6 2.03 2-73 44.6 16.6 28.0 •34 39-5 60 5 I. 98 2.42 48.2 19. 0 29. I •37 41.2 .S« 8 2.03 2.46 49.0 20. 2 28.8 • 41 43-6 56 4 2. 01 2. 19 53- 0 22.8 30.2 .46 45-5 54 5 2. 12 2.32 52.2 23-7 28.5 •50 48. I 51 9 1.99 2. 26 54- « 26.3 28.4 •52 48.7 51 3 I. 90 2.03 53-3 25- 9 27.4 •49 50-3 49 7 1-95 2.05 .5.5- « 28. I 27.8 •55 50.8 49 2 2. 02 1.97 50. 3 28.6 27.7 •.S« 52.9 47 I 1.89 1.94 55-9 29-5 26. 4 •56 55- I 44 9 2. 00 1.89 56.6 31-1 25-5 .62 56- 4 43 6 2.27 1-93 58.0 32.6 25-4 •75 57-2 42 8 2.28 1.88 56.1 32.0 24.1 .76 Ash. July 15- 16. 17- 18. 19. 20. 21. 22. 24. 25- 26. 27. 28. 29. 30- Aug. I.. 2. . 3-- 4-- S-- 6.. 7-- 8.. Mgm. CHANGES IN WET WEIGHT PER KERNEL The trend of the wet weight is indicated in Table III and is summarized in Table IV. The course of development is more apparent in figure 9, where the growth of kernels 5,8, and 10 is represented graphically. The most rapid increase occurs in the first 16 days. After this time the loss of water is almost equal to the increase in dry matter. The fifth, eighth, and tenth kernels represent different sections of the spike. The order of weight is reversed during the period of growth. The tenth kernel was the first of the three to be fertilized, and it reaches a constant weight some time before the fifth kernel does. The shift of wet weight is much more evident in figure 10, where the weights of all kernels are shown. The trend of development in the wet weight is quite parallel to that of the length, lateral diameter, and dorso- ventral diameter shown in figures 6, 7, and 8. The shift here proceeds toward the base until the fourth kernel is the heaviest, and it is only toward the last that the fifth and sixth kernels become the highest in weight. The wet weight, owing to the difference of moisture content 4i8 Journal of Agricultural Research Vol. XIX. No. 9 between the kernels at the base of the spike and those at the tip, is not an accurate indication of the storage of nutrient material. The curve of wet weight is quite similar to that published by Brenchley (j). The losses after maturity found by Brenchley were not evident at Aberdeen, becaase the sampling was not carried to the same point. Brenchley in- cluded the glumes in the weights, while at Aberdeen these were removed. As the glumes can not be removed after maturity, their removal shortens the period of study. On the other hand, the glumes themselves change TO CO ^o -^o oK? ^O /£? / s 3 ■^ ^ e' -z s s yo y/ /£ /s jv /^ ya- yr ya y» ^o£/ ^^^ss^£^ /^zoH'^ypz/u^ / ^ / ^* / / / b / / \ 1 '/ "v y h ■) / ^ / / ^ / ? ^ 5? L_ V Fig. 9. — Graph showing wet weight of individual kernels 5, 8, and 10, by days, from date of flowering to near maturity in 1917. materially in character between flowering and maturity, and their elimi- nation removes one source of error. INCREASE IN DRY MATTER The daily growth of the kernel is summarized in the daily increment of dry matter. While there are gradual changes in the percentages of the various substances for days of the same week, the added constituents bear a more or less uniform relation to each other. The sum of the daily additions is the increase in dry matter. This increase has been so uniform at Aberdeen as to indicate that the plants were working very nearly at Aug.2, I930 Daily Development of Kernels of Hannchen Barley 419 their highest capacity. In figure 1 1 are given the dry-matter contents of kernels in 1916 and 1917. For the first 17 days after flowering the curves of the two years practically coincide. After the seventeenth day the rate of deposit decreases in 1917 but is maintained for several days in 191 6. This is due, probably, to lack of sufficient water after this date in 1917, the effect of which is noticeable in all the results reported. The gain is surprisingly uniform for the most part. In each season, the curve is essentially a straight line from the sixth until the eighteenth ^ 3 ■*^ s s y s 9 /o // /^ /3 •■^ Fig. 10. — Graph showing average wet weights of kernels from flowering to maturity in plot i in 1917. Numerals at ends of lines indicate days from flowering. day. This is interesting in its relation to the general laws of plant growth. In a developing plant, where the new tissue added becomes immediately productive of nutrient material for growth, the increase is accelerated in geometrical ratio. The curve of growth, in this case, can be reduced to a straight line in plotting by the use of logarithmic paper. In the case of kernel growth, by the fourth or fifth day after flowering, the maximum leaf and sheath surface is exposed. The plant food metabolized is diverted to the storage tissues of the kernel, and, since the productive tissues remain constant in amount, the curve of kernel growth is a straight line. 420 Journal of Agricultural Research Vol. XIX, No. 9 The uniformity of the Aberdeen seasons and the accuracy of the method of sampling used is nowhere more evident than in figure 12. In this figure the dry matter per kernel is plotted in 12-hour periods. For the first 14 days neither the error of sampling nor the differences in rate of growth of individual spikes, separately or together, exceeds the growth /YS/Z 30. 0 ^ao < 4 f ' ^ ^^ /O.O S.O ^e/^r/s/? /s /s /7 /ff /i> ^V4?/^^^3,^^^J'^s^7^<9^s303/ /'\^ ^ ■*«■ v5" e- ^S s Fig. II.— Graph showing dry matter per kernel from date of flowering to near maturity in 1916 (dotted line) and in 1917 (solid line). in 12 hours. There is an apparent reversal of the curve in the fifth and seventh days after flowering, but the larger of these losses is less than 0.2 mgm., and in each case is due to the abnormalities of a few kernels on the spike. When this curve is plotted from the data of the more representative sixth, seventh, and eighth kernels, these irregularities dis- appear. It is only when the fourteenth day is reached that fluctuations Aug.3. J920 Daily Development of Kernels of Han7ichen Barley 421 become common. After this date results are not consistent in such short periods as 12 hours. The original purpose of the 12 -hour interval was to discover, if possible, whether or not growth occurred during the night. For this reason, the periods are not quite equal. The day period consisted of about 13 hours at the beginning. As the days grew shorter this was reduced slightly. This period was thought to include the hours of effective sunlight. In figure 13 the gains and losses for the day and night periods are indicated graphically. The disadvantage of such presentation lies in the magnifi- cation of the fluctuations. For instance, the night sample of July 26 shows a decrease of 120 points, not because it is smaller than the sample of July Is y 7 SX \ / /^ J ^ r ^ r A / / / n / ^ 1 L > -< 1 - / J 1 ^ i r r /7 /'<9 1 t br 1 7 ^3 ^7 XL t ^9 r so 3/ t J" tr J" t (5- T T Fig. ij. — Graph showing dry matter per kernel at 12-hour intervals from flowering to maturity. 25, but because it is the same size and, therefore, no gain is registered. For the first lo days after flowering the day and night gains appear to be nearly equal. From this time until maturity the day gain is obviously greater. The author has no interpretation to suggest, but there are two facts which may be noted. The night gains are most prominent before starch infiltration has become very active. The temperatures, after the first 10 days, are lower, the first night without gain being recorded on July 26, when the mean temperature first falls to 70° F. It is not known whether these facts have any essential relation to the results obtained or not. During the latter part of the growth period, the variation of individual spikes makes the results inconclusive. The significant features of the data on dry-matter content are (i) the long period of daily gains following the completion of length growth, 422 Journal of Agricultural Research Vol. XIX, No. 9 which results in a straight line through a considerable portion of the curve when plotted, and (2) the unusual uniformity of increase which permits the taking of samples which show growth in 12-hour periods for two weeks / ^ J ^ ^ G 7 & s /o // /^ A? /'^ /s /s /r /& /s> ^c?^/ ^■e^si'-^ /^oi^^'fxy/vs- Fig. 13. — Graph showing dry-weight gain of kernels 6, 7, and 8 in 12-hour periods. Gain during the day is shown by the broken line and that during the night by the soUd line. after flowering. The curve of growth as found by Brenchley was quite similar to that shown in figure ii. The same straight line is apparent during the period of rapid starch infiltration, despite the fact that she took samples only every third day. The results seem to agree in a general Aug. a, 1920 Daily Development of Kernels of Hannchen Barley 423 way with those of Schjerning, but inasmuch as his samples were less frequently taken, close comparison is not readily made. CHANGES IN WATER CONTENT The percentage of water in the kernel is highest at flowering time, when over 80 per cent of the caryopsis is water. From flowering until maturity the percentage of water constantly decreases. At maturity the water content has fallen to about 40 per cent. The decrease in per- centage is very uniform, as may be seen in figure 2. The curves of 1916 s>o e& a-^ A? 80 70 %7^ %70 o — — ■)?* nj — " — IZZZZZZZZZZSifZZZZZZZZZZI _^ o ^^ Sg A." ^_ — JL_» ^ ^ ^ ■^ ^ o^ 7 ^ s> /'o // /^ /^s A* /s /G' ^^/ff ys) ^i?^y ^^^^^■^^.s^^^^l"^ Fig. 14. — Graph showing percentage of moisture in morning and evening samples of Hannchen barley in 1917. The average for the day is indicated by the line. The average morning determinations are indicated by circles, and evening determinations by crosses. and 1 91 7 are essentially identical. As previously remarked, the coinci- dence of these curves is evidence of the exceptional opportunity afforded at Aberdeen for comparative studies in development. The loss of water in percentage is much more rapid than in the results obtained by Brenchley. At Aberdeen the rate is almost 2 per cent a day. At Rothamsted the rate during infiltration was in the neighborhood of i per cent a day, although the rate was higher than this at times. The effect of evaporation during the day was noticeable. The morning sample usually showed a gain in percentage of moisture over that of the night before. The loss of water during the day was rapid, evidently exceeding the normal loss, due to the incident of growth. This extra- normal loss and its recovery are shown in figure 14. 4^4 Journal of Agricultural Research Vol. XIX, No. 9 INCREASE IN NITROGEN CONTSNT The nitrogen determinations are the least satisfactory of the studies made. The samples were so small that microchemical methods had to .70 . 1 » I h > • /-■- 1' f 1 / \ \ 1 i n i '/ r li / ■/ 1 J ^ \- n •'/' / / ^ 'y 1 1 > y / ,eo .SO A^O ,30 .4^0 ./o /*2r^/#fi5»sy y''^ ^ ■^ ^ c J^s <» s/i/jLyysys- ^ ye> yy^AS y^ y^ /»■ y ,3/ /^ j Fig. is- — Graph showing nitrogen per kernel from date of flowering in 1916 (broken line) and in 1917 (solid line). be used in the early stages of growth. While the material itself was probably fairly uniform, the determinations are not delicate enough to show a uniform progression in percentage. When the percentages are computed on the dry weight to obtain the milligrams of nitrogen per Aug. 2, 1920 Daily Development of Kernels of Hannchen Barley 425 grain, the results are much more uniform. In figure 15 is given the total nitrogen per kernel in both 1916 and 1917. The curves are essentially identical. The divergence after the seventeenth day is due to the lesser gain in dry matter after that time in 191 7. The divergence in nitrogen content is about the same as in the dry matter shown in figure 1 1 . The results obtained agree very closely with those of Schjeming in Denmark and Brenchley in England. INCREASE IN ASH The percentage of ash in the kernel decreases uniformly from flowering to maturity. At the time of fertilization the percentage is high, and for 48 hours after flowering it is more than 7 per cent. The decrease in percentage from that time is not due to loss of ash, as may be seen in figure 16, but to the more rapid increase of other materials. In other experiments, to be reported later, it has been found that the ash content is in fairly close relation to the amount of water available for the use of the plant. The curves of the ash content of 191 6 and 191 7 again indicate that the irrigation of 1917 was insufficient. The greater growth of 191 6 has been mentioned previously; and while a part of it may have been due to better soil in 191 6 a part was certainly due to the more generous irrigation of that year, coupled with the fact that the soil used in 191 6 absorbed water somewhat more readily than that used in 1917. PERIODS OF DEVELOPMENT The Hannchen barley at Aberdeen exhibits a development which is very uniform from year to year. This development, while steadily pro- gressive from flowering to maturity, varies considerably in its nature. The first five days after fertilization are marked by an extremely rapid growth in length. The kernel has reached its maximum in this respect by the seventh day. About the time the growth in length ceases the rapid gain in dry matter begins and continues for about two weeks. Thus the fifth or sixth day marks a change in the character of growth. About the ninth or tenth day a sticky substance is formed in the outer layers of the caryopsis, which causes the glumes to adhere thereafter to the developing kernel. The nature of this substance has not been included in this study, but its origin is evidently in the caryopsis and not in the glumes. This has been demonstrated in the making of hybrids. In this process the upper part of the florets is removed. At maturity the tips of the projecting kernels are often found stuck fast to the paper in which the spike was wrapped. The appearance of this adhesive sub- stance on the ninth or tenth day would seem to mark a second stage of development. Since the inner tissues -of the kernel are very soft, it is difficult, from this time until the kernel has somewhat hardened, to remove the glumes without tearing the kernel. This hardening occurs 183718°— 20 3 426 Journal of Agricultural Research Vol. XIX, No. 9 about the fifteenth or sixteenth day at Aberdeen. It is accompanied by several other phenomena as well. The lemma begins to lose its color in the center of the dorsal surface. The awns of the Hannchen variety. .SO .SO .70 .&0 .*r^ .AT^W ^T .ir? J>J» JO J8''7"> J ■* .^ « Ji»^ ys/s ^c/^f 9 /o y/ /^ y^ /^ Af yc /t- •» y-9 £o ^/ ^£ ss s* ^j- jw jf7 .ss £9 .so ■9/ / ^ s ^ /\ 1 / \/* » a\^ i V 1 r -^r- ^ t A> \^ / \ 1 1 « « ^1 1 / / 1 1^ / N s 1 / } / / ^ f \ / s y / / / / \ f 1 1 1 y* > i H f ,-- 1 1 \ \ \ 1 / 1 1 \ 1 / // '^ >a r ^ / f* -y c> -'■ / \ 7 / / } ■/' r / / f y / — — / Fig. 16. — Graph showing ash per kernel from date of flowering in 1916 (broken line) and in 1917 (soUd line) and the percentage of ash in 191 7 (dotted line). which are more or less deciduous, drop off in large numoers. The tissues of that part of the ovary above the embryo sac have been resorbed until it is possible at this stage to measure the kernel without including this Aug. 2, 1920 Daily Development of Kernels of Hannchen Barley 427 structure. The fifteenth or sixteenth day marks what probably is the most important change in the course of development. Among the inter- nal changes, this date coincides with the maximum water content of the kernel and the end of the period of most rapid increase in dry matter and ash. Schjeming found a drop in the soluble nitrogen present in the kernel at about this time. From the fifteenth or sixteenth day until maturity the changes are gradual and all in the same direction, differing only in degree. The only point now apparent, which might mark a change of nutrition, is to be found in those varieties which develop anthocyanin colors in the -^ Ot*'£'/?/A/C- f^r/p^,^s- r/^ ^/\>'^ iffy^s/T ^^ ayp^rys /T,: Of^^^fX'/yi^s S^s>p/^^ sy^x-z^a/p.^.rs'a^ /m^/'p^^^r^^^ /S^/p^jtks- /^cvtir/p^/^a s'- ■^}v^7yy G/^vft^r^ ^^0£/'7^ cty-TX^^^TTZ? c/r/^/(?r>',' G>POH^r-/Y ^s-stirA^jry^^^y ca^/^^^'z- /HF?fi^/^z T'oc/ff'^'^yi^s: on^A>y ry^ /ip^sa^.s^^y t^^/^xrry^ ^**vz2ve>«<»#tv>^^^ ^/ftsfv/^cAf'G' ^o S'7'/r:yr/ri^s-n no n^^- c^^k■a'^yt*'7',VA*fVis>v/Siin S'T'o.ev^/O. ,'?»*«f:r C42<»«s>/'»v>-zz av^n /.^^C/f^^s"^ 27 Fig. 17.— Periods of development of the barley kernel as indicated by records during three years at Aberdeen, Idaho. external layers a few days before maturity. This is probably a very minor phase of metabolism, and at present it is not known to be asso- ciated with any vital phase of growth or maturation. The various external indices of internal changes are shown in figure 17. MORPHOLOGICAL CHANGES Microscopical examination of the kernel was made to determine the progress of the internal modifications that must accompany develop- ment. The starch infiltration is shown in various stages in Plates 85 to 91. Starch was found on the fourth day. This had increased per- ceptibly on the sixth day. The starch grains up to this time seemed 428 Journal of Agricultural Research voi. xix, N0.9 to be of a very much lower density than normal barley starch. They did not stain readily and were indefinite in outline. Rapid infiltration of starch began about the time that rapid growth of length ceased. By the ninth day after flowering the starch grains were of very uniform appearance. From this time the development was more irregular, not all the grains continuing to increase in size. By the fourteenth day small grains were apparent among the larger ones, as though new starch grains were forming. These small grains are found in the cells from this time until maturity. The fifteenth and sixteenth days represent a period when the awns are likely to drop off. The dropping of the awns, apparently, coincides with the completion of a stage of starch infiltra- tion. From this time on, although the rate of starch deposit holds up fairly well, the accumulation is made by the continued development of only a part of the large grains and the packing of the interstices between the larger grains with smaller ones, rather than a uniform development of all grains as at first. The first starch was found in the older cells in the middle of the flanks. It is probable that new cells are added about the periphera of the endosperm and especially near the furrow for some time. It is unlikely that new cells are added to the periphera after the fifteenth day from flowering. The new cells added near the furrow develop in a way entirely comparable to the first cells of the starch endosperm. Such cells are shown in Plate 91. After the first two weeks the trans- portation of food material to the sides remote from the furrow may not be so readily accomplished. Here the cells last formed may remain nearly free from starch at maturity, although the development of the cell walls demonstrates that the cells are not young. SUMMARY This paper presents data showing the growth of the Hannchen variety of barley from flowering to maturity, taken at 12-hour inter\-als. In the early stages of development, measurable growth occurs during 12-hour intervals, and during 24-hour intervals until near maturity. The period from flowering to maturity in three successive years at Aber- deen has been 26 days. Measurements were taken of the length, lateral diameter, and dorso- ventral diameter of the kernel. The growth immediately after flower- ing is so rapid that the increase in length is readily measurable at 12-hour intervals. The length growth is completed by the seventh day, and as soon as the rate of growth in length decreases the lateral diameter shows its most rapid increase. The dorsoventral diameter continues to increase almost until maturity. The increase in dry matter in the kernel is very uniform throughout the period of growth. The percent- age of water decreases uniformly from flowering to maturity. During Aug. 2, 1920 Daily Development of Kernels of Hannchen Barley 429 growth the carbohydrates increase most rapidly and the ash least rapidly. There are several well-marked steps in development. About the fifth or sixth day after flowering the growth in length is checked and a rapid gain in dry matter begins. About the ninth or tenth day a sticky sub- stance is secreted, which causes the glumes to adhere to the kernel. About the fifteenth or sixteenth day the kernel toughens, the lemma begins to lose color in the dorsal surface, some of the awns drop off, and the kernel has reached its maximum water content. Maturation occurs gradually. The cells about the furrow continue active longer than elsewhere. The actual date when growth ceases, even where the external conditions are unusually uniform, as they are at Aberdeen, must depend on the temperature and humidity at the time of ripening. LITERATURE CITED (i) BrEnchley, Winifred E. 1912. THE DEVELOPMENT OP THE GRAIN OF BARLEY. In Ann. Bot. , V. 26, no. 103, p. 903—928, 22 fig. (2) Harlan, Harry V. 1914. SOME DISTINCTIONS IN OUR CULTIVATED BARLEYS WITH REFERENCE TO THEIR USE IN PLANT BREEDING. U. S. Dept. Agr. Bul. 137, 38 p., 16 fig. Literature cited, p. 37-38. (3) JOHANNSEN, W. L. 1884. OM FR0HVIDEN OG DENS UDViKLiNG Hos BYG. In Meddel. Carlsbcfg Lab., bd. 2, hefte 3, p. 103-134, 3 pL French resume, p. 60-77. (4) KuDELKA, Felix. 1875. UEBER DIE ENTWICKELUNG UND DEN BAU DER FRUCHT- UND SAMEN- SCHALE UNSERER CEREALiEN. In Landw. Jahrb., Bd. 4, p. 461-478, pi. 5-6. Also reprinted. (5) Lermer, and Holzner, George^ 1888. BEiTRAGE ZUR KENNTNis DER GErste. io6 p., 51 pi. Munchcn. (6) SCHJERNING, H. 1906. ON THE PROTEIN SUBSTANCES OF BARLEY, IN THE GRAIN ITSELF AND DURING THE BREWING PROCESSES. In Compt. Rend. Trav. Lab. Carlsberg, v. 6, livr. 4, p. 229-307, 2 tab. (7) I914. ON THE PROTEID SUBSTANCES OF BARLEY, IN THE GRAIN ITSELF AND DURING THE BREWING PROCESSES. In Compt. Rend. Trav. Lab. Carlsberg, v. ii, livr. 2, p. 45-145- Literature cited, p. 146-147. PLATE 83 A. — Fertilized ovary. B. — Kernel i day old. C. — Kernel 2 days old. D. — Kernel 3 days old. (430) Daily Development of Kernels of Hannchen Barley Plate 83 ■ [E^S! H 1 ■ 1 ^^K* 'JP*^'' 1 i ^1 B ^ ^^^^^^^^^^^^H 1 ^^F ^ ^^ *> ■ l^ill !r i 4 m Be i Journal of Agricultural Research Vol. XIX, No. 9 Daily Development of Kernels of Hannchen Barley Plate 84 Journal of Agricultural Research Vol. XIX, No. 9 PLATE 84 A. — Kernel 4 days old. B. — Kernel 5 days old. C. — Kernel 6 days old. D. — Kernel at later stage of development. PLATE 8s Kernel 5 days after fertilization. Starch grains are apparent. Daily Development of Kernels of Hannchen Barley Plate 85 Journal of Agricultural Research Vol. XIX, No. 9 Daily Development of Kernels of Hannchen Barley PLATE 86 , A '^.•.' ^1* r\. \ >#* \" Journal of Agricultural Research Vol. XIX, No. 9 PLATE 86 Kernel 6 days after fertilization. Starch grains have increased greatly in numbers. PLATE 87 Kernel 9 days after fertilization. The cells are well filled with starch grains of uniform size. Daily Development of Kernels of Hannchen Barley Plate 87 4r f < Journal of Agricultural Research Vol. XIX, No. 9 Daily Development of Kernels of Hannchen Barley Plate €-->^; .-% ^^ l^-- K 'j»> Journal of Agricultural Research Vol. XIX, No. 9 PLATE 88 Kernel 14 days after flowering. Both large and small starch grains are present. PLATE 89 Kernel 20 days after fertilization, at which time growth was nearly completed. Daily Development of Kernels of Hannchen Barley Plate 89 V y Jl- y Journal of Agricultural Research Vol. XIX, No. 9 Daily Development of Kernels of Hannchen Barley Plate 90 ^;^ ^^* 5^^^-v '% V i*l/--- ^ifev3&V W^ •^21^- ^^C^^\ W'lA Journal of Agricultural Research Vol. XIX, No. 9 PLATE 90 Kernel 25 days after fertilization, growth completed. PLATE 91 Section of a nearly mattire kernel, showing cells next the furrow. These cells were formed more recently than those of the main starch endosperm. Daily Development of Kernels of Hannchen Barley Plate 91 Journal of Agricultural Research Vol. XIX, No. 9 DEVELOPMENT OF BARLEY KERNELS IN NORMAL AND CLIPPED SPIKES AND THE LIMITATIONS OF A.WNLESS AND HOODED VARIETIES^ By Harry V. Harlan, Agronomist in Charge of Barley Investigatiotts, and Stephen AutnONV , formerly Technologist in Barley Investigations, Office of Cereal Investi- gations, Bureau of Plant Industry, United States Department of Agriculture INTRODUCTION The studies reported in this paper were made in an effort to obtain some light on one of the farm problems in barley production. Among the farmers of the United States there is a strong prejudice against the growing of barley because the long, rough awns make the crop disagreeable to handle. The beards in barley straw and hay often cause sore mouths in stock. Barley straw and barley hay are also undesirable for feeding to sheep because the awns work into the wool. It is only the high acre yield in pounds of feed that has maintained the acreage of this crop, but this acreage is far below that which would be devoted to barlev if the awns were lacking. Certain types of barley are free from the harsh awns. One of these, the Nepal, produces hoods in the place of awns. This variety, under various names, has been more frequently introduced and more widely tested than any other. Many hybrids have been made and distributed. That they have failed to measure up to the expectations is evident in the annual inquiry of seedsmen as to where they can secure seed of "bald" barley. The only apparent explanation of the failure of the Nepal barley is the lack of awns. The field records of the Office of Cereal Investigations, extending over many years, indicate that the Nepal compares favorably with other varieties only in the high altitudes and in dry years in the northern part of the Great Plains area. As a rule, varieties of this type have yielded less than the awned sorts and have shattered badly. It is evident that the awn is an organ that is functional under most conditions, and especially in those sections where humid weather prevails at ripening time. Zoebl and Mikosch ^ in 1892 showed that the awn of barley was an organ of transpiration. Schmid ^ in 1898 and Perlitus * in 1903 elabo- rated the experiments of Zoebl and Mikosch. All agreed that the awn was an organ of transpiration and all showed the effect of its removal on both the rate of transpiration and the kernel. • These studies were made on the Aberdeen Substation, Aberdeen, Idaho, in connection with cereal experiments conducted cooperatively by the Idaho Agricultural Experiment Station and the Office of Cereal Investigations, Bureau of Plant Industry, United States Department of Agriculture. 2 Zoebl, A., and Mikosch, C. die function der grannen der gerstenahre. In Sitzber. K. Akad, Wiss. [Vienna], Math. Naturw. Cl., Bd. lor, Abt. i. Heft 9/10, p. 1033-1060. 1892. 'Schmid, B. BAtJ und funktionen der grannen unserer getreidearten. In Hot. Centbl., Bd. 76, p. 39. 75. 119. 218, 305-307- 1898. * Perlitus, Ludwig. einfluss der begrannung auf die wasserverdunstxjng der aehren und DIE kornqu/xiTAT. 77 p., 3 pi. Breslau, 1903. Journal of Agricultural Research, Vol. XIX, No. 9 Washington, D. C. Aug. 2, 1920 ur (431) Key No. G-200 432 Journal of Agricultural Research voi. xix, No.g EXPERIMENTAL METHODS AND MATERIAL The first experiment by the present writer was a very elementary one made in Minnesota in 191 1. It included yield only. Plants from which the awns were clipped produced only 75 per cent of the yield of normal plants. In this and the later experiments, sufficient spikes of the same age were tagged on the same day. The method used was that described in an earlier paper. ^ The awns on half the spikes were removed even with the top of the upper most sheath as fast as they appeared. In clipping the awns it was necessary to examine the heads each day for three or four days. It is apparent that mechanical injuries might result from this opera- tion which would affect the later growth of the spike. For this reason it was thought desirable to trace the growth throughout the period from flowering to maturity. The number tagged was sufficient for a sample of two or three spikes per day from both the clipped and the normal plants for a period of 30 days. In a preliminary experiment at Arlington Farm, Va., in May, 1915, it was found that the taking of daily samples was practicable. In July, 1915, a complete experiment was conducted with the Manchuria barley at University Faim, St. Paul, Minn., and in the summer of 191 6 a similar experiment was conducted at Aberdeen, Idaho, with Hannchen barley, The weights, lengths, and diameters of the kernels of the samples were obtained daily. The kernels were later analyzed to determine the nitrogen and ash. The results from the two varieties will be presented separately. EFFECT OF REMOVING THE AWNS FROM MANCHURIA BARLEY IN MINNESOTA The Manchuria is a 6-rowed variety of barley. It is awned, a vigorous grower, and adapted to fairly humid climatic conditions. It cannot be grown in the arid districts with success. The first sample at Minnesota was taken on July i, 191 5; and samples were taken daily until August 7, with the exception of five days. This was the only study conducted in Minnesota. Table I shows the data obtained at St. Paul, Minn., in a humid district, with the variety of barley best adapted to that district. The samples taken in Minnesota differ from those taken at Aberdeen in that they consist of a single spike each day. The weights and measure- ments of the individual lateral and central kernels on one side of the spike were taken under each of the headings "weight," "length," etc., in Table I. The first column contains the weight and measurement of a lateral kernel, the second contains those of the central kernel, and the third contains those of the remaining lateral kernels at the same rachis node. The kernels were studied in order from the base of the spike upward. In the first line under each date are the data from the first fertile florets at the base of the spike. In the second line are the observ^a- tions on the florets at the node above. The last line contains the data on the last fertile florets at the tip of the spike. 1 Harlan, Harry V. daily development of kernels op hannchen barley from flowering To maturity at ABERDEEN, IDAHO. In Jouf. A.gr. Research, v. 19, no. 9, p. 393-43°. 1920. Aug. 2, 1930 Development of Barley Kernels in Clipped Spikes 433 183718°— 20- h) v^M a ►4£j< a o-w^ a hIiu^ a hJo t^ t^ t^OO 00 t*o • vO \0 t^ t^ t*00 «^ ^^ r CO Oioo OS H OiOO • 0\n OsOtOtn 0*0 ■ o»o**o o»t^o 1 O O 0» r-co ,S, 8 8 8 8 8 8 8 8 •O vO 00 t^CO CO 00 00 iJ^^ ■ooo o ao ovo ar-j t^ t^ o\ t>. OS 0*00 00 t^oo p'l CI WTO -O '^ O «^ w »'^ Tf\0 t^ OS f* t*00 M « 0\ »tH u-i^sttn»oMOO *^ O N t^oo Oi O r*0 ^00 oqqooqqSoq 0000000000 TO r^oO 00 ''too HI O t-ico 0»0 O r^ior^c §"^00000000 00000000 *o t^ooo r*r*r*r-oooo -O OiOO o t^oo 00 00 o» t>. o O «^oo 00 >0 00 r* r* r« O r*M N ot^c< ^f*^0»» Tt «^ Ot M « COO 00 <*> -tf W M fp 434 Journal of Agricultural Research voi. xix,no.9 1 00 8 w n < Clipped spikes. 1^ Lat- eral ker- nel. i : 8 2 i'v ■S2 «'3 8 •• V ■•3 2 CS Hi '«4 2 S"" ji t* t* t- r* 0^ P\ i^'O ' 1^ 1 rt2«a • t««00NO t»00\O *» t» i' ti 8 i4 B 2 S'5 )4 V.M n • on«vimoeoo> Sl^l • n * m moo 10 0 0 «»>0 uim « ,S> 8 8 8 8 8 8 8 8 82N ,^888888888 rt 2 i"^ ,5888888888 ^6 .•a 0. o in B « 2 fs hIkj^ a 8 :§ 8 2 ii'3 o»^J< a 8 a 2 fe'3 S :§ V ■■3 2 i4 "5 2 fca hj vM a • 0 «^ ** ** *^*o ** "^oo 1^ ^ 82 OS'S oi^M a ^ 00 t» Oi 0 t«00 Ol «»00 00 l'--^ 5III • v> t^oo CO 00 >o t» Oioo ■>» 1' ^l.si •n0w«Mi»)0>t»0 82 N 0-^ a •notoioMMOooO'fl- ^.......w.m « 2 iu ■OvOMHOMHOlfO'V •s Is 2 S"? >4 4)^ a ^88888888 8^ 6 S2a'S o*"^ a 0.0017 .0020 .0023 • 0023 .0018 .0018 .0018 .0018 .0016 . 0007 ilH 0 8 6 8 8 8 I- 8 88 8 8 1- ■ t*00 00 00 t* 00 Q% r«>w 00 o o 6 • • • • « M >0 <^C0 OtOO o 00 OMMMMMM (•^•'3'O'O^^WOO'O Ot>0 000 *on>0 too * r* O t-i o\ r* ** O ' ^M NOG ti^O t^O'O'- 88888?888 6 6 to t^oO Ov O 00 O^ Ol t^ t^ 6 " ' w CO r«io ^O O r* lA V) ro t^ 6 CO r«oo«o M ov Otoo o c^ 5mmmhhh * 'cO t^ Oi'O «* Ot t* t* OtOO 6 CO ♦O ^ W fit* « Ov n a n <^ n C4 mC/J t^QO O * « M f»5 W> (*) '*00 00'O O^ OkO "O ^ 8 8 8888 8 88to wiO»0 O t-* 0»0»f^» 88?888888« 6 ^00 DO 00 00 00 >o O fO Aug. 2. 1920 Development of Barley Kernels in Clipped Spikes ' 435 0 0 0 0 Pt w 0. OvvO MWHWHCTHWO mOO«OihOO.O> m >o uTOvo i« •<(• « J^ 00 t^OO t^u^lo^« 00 m in>o ^ tn'n m ^ to + ***-*.^«« m r»o *^oo «^ ''^oo »^ * .0 ■* *-9- ■«•*■* « oosj 0064 0069 0074 0072 0072 0046 0030 0021 6 00 « 1/100 N N 0\ -O " t^OOOOO OlOO f^t^rO 8S8888888 6 OO^.loWOiOOO'O ^ m r^ r^O t^^c fo M 888888888 6 O>00 M « « « 0 0 Ov 6 • ^ ^ ►; ^ ^ ^ • «<*5C1M«MIHMOv 0 0 0 *H CI 0 >-< 0*00 «n lA Vi*0 »0 »/l ^ rO « lO 00 0 ov oioo 00 vo in to VO »0 t~ !>■ >0 N X5 >0 H 10 f^so (*i 1^ r^ ^ OM'^WMt^r^OOO 10 io\0 >0>0 i« w ■* « M Tt «»0 0 0- 0 00 W ioOOvO M r- | tn « ^ 0 *^ **l\0 t^oo 1^ »o fl M fr)^fO^**5fOio« w M w t^ r-00 OO'Oi-ioOMOiri'^ .^■»-«-*«*.5)*^«« 01 « 0 Ounmooot^O mH « ^.0-i «>• r-o m m ir, 0 •^sO 0 10 VI ^ ^ 10 (r> - 00 f«^M w 0 toooooo fOO oOw^wmOmO lO»0^»/l»OVO'1-«ia-'^N o»O\O»i^ooM00'' OmnNN««onc^r^^oo OiOOioOOOOOOOOOi •VWU1100 Wl/imiOCOH 0\ 10 lO-O o\ ^ t^ « fl 0 ^c i>)vOvOOnO*oi6 <^r-ioci MvowM(«^0»^r^«sor^ *r, 10*0 OvOvO-O ■^i>i*^H 0068 0072 0098 OIOS OIII 0091 0085 0088 0079 0049 0014 0 Q t^ M ro 0 0 wivo « t^ r* O'-'MCifOrO''0'a- ^00 « 00000000088 6 0066 0084 0109 0106 0107 0109 0120 0082 0088 0049 0013 0 «MM««««0\0\ 0>l^C1t0tlf<>»5>^N0l O^MHWWMWH • 00«N«C «""""«-«" ■ moo cooo oiM a*^«toi 00000 00000 0*0v<^ ^00 0 r- 1^00 00 00 t» i« «00 « M I» 0. Ov WVO M50O VC 1/J •*.oio-*-4-ioiow)« ,- t^-*->tinO\"t»<^Oti- '^ ^ 1/1*0 0 10 100 10 f*5 PH ^oioiiro ot»M »^ to *-» + OT.oio-«-*« H s 0066 0083 0099 0084 0084 0103 OIOI 0047 0037 6 ^ •-. i/ivo r^ 0* --t 1/1 « Oi *- 888oo8ooo8i 6 00 0 « 0 i^co -0 P« o* 0 «8 8ir8l88 8 8 6 436 Journal of Agricultural Research voi. xix, no. 9 CO ^ "T^ n < i a (U 0. .& 0 o'-S Q Lat- eral ker- nel. ..; 0 t<5«<*-<)--r«M0 1 S g SI'S 1 "" i4 v^ fl « to N ■* W : ^0 4) ■•3 1 i4 »4«^ a 00 \0 00 0 0 OvOO VI §"2 i'Z a " 5" ^«P.C«CH« a 2 v'v ^ r^ r- O\oo 0 0 ■ • t- d ^ ** 0 00 N -^O 00 0 0 SIN • vO -^O w 00 M 0 00 0 j^ t^ t^ r^oo t^oo 06 «^ t^tfl rt 2 Sj'3 Hto-;^ a 1" >n r- 0 0 00 • • w 1 loso t^ r* t^ • . i>iy^ 1 M 2 aj'«J kiSj^ a Gm. 0. 0078 .0088 .0085 . 0106 •0134 .0134 . 0119 .0099 . 0070 s g2i« o-^^ a r^O " 0 (N Osr..^oci (^OOOOOOOO oW 6 rt'2 SI'S Hti a •00 0 foM fneorno* OiOO « 2 v'v 00000.0 OM a >A ^M a xji^M a ■ 0 ■* ■* 00 ■* H lO OiO M ^ xo A r^ t^QO 06 06 0 vo ^ « 2 ia ►4 4'^ a ^O\- 00000 0"^ 8 0 n CO 1/1 lyi Tto 0 \ntn 00 V. 1^0 r*o r.-oo 0 t^ -too . ^loiniot^i^iA^ooo r*r^MMwfoto*o«<3 0» HM«MWNMn«M M 0 lOTt'tfOfOfOfOONf 0 NNNNNMMCTWMW '. 0«M«.00 00 W OS o\ O* f^ (*5 !t^oot^Tr««>o«MV) . t^ t^ r^oo 00 00 00 00 vo M 008S 0108 0132 0170 0126 0202 0198 019s 0160 0093 0019 0 Ov N t^O r^ w 00 M 1/1 r^ fo OOOOOOOOOOO 0 bi M rovO fO 0 f* 0 On Ov « ■>, 0000000080 0 6 Tt-O I^ r^ I^ Tj- 10 ro M vo vO ^3 00 00 t^ t^^O CO '^ f 0 r^o 10 r- 1/1 f M mmmmmhmmhM cOMcOCO-tMMM.- c^mnnnnc^pSmM ttVlflOcOfOCOO 0 WNNWCnt/3 I^vO 10 COOO 10 « COOO 0 0000 0»0\OvC*OvOi*^co 0> to 0 *^ t^OO »0 M t^oo 06 dvoo 06 t^ t-- t~-c/5 CO M 1^ t »^ 0 0 1^00 t/1t^t^ONO "liotO OOOOOOOO 0"^ - 6 0208 0213 0225 0241 02 so 0230 0229 0225 0158 003s 6 OnO t m 0 0 1- CO 0 COM 'i2°2'";j;^^. OOOOOOOO 0 6 Aug. 2, 1920 Development of Barley Kernels in Clipped Spikes 437 ^O lAOO t.* *^ Ov t^cO 0^ t^ 00 t^O 0*000000'000 O\00 0^0 ^sO *0 t^ t^ ^^-O t^O t^ 00 t») «5 TT >n« >o «5 ^ ■>»■ M 0 __,^ r^ Oi r.00 0 t- v^ m V, v\ u^>o oit"a.ooNOi^MM to r^ 00 o> CO 00 o\ o> Oi 0. Oi t 0006 0 (7,0.000 0 0 6,«vO wt^Oi'.^wO'^oOOO -0 t^ 00 00 0> 0> 0^ 00 00 0. 00 ■» oooooooooo 0 6 * ■ 00 t^vO V) 0 » N M r^oo ■* Oioo 000000000080 6 oooooooooo 8 6 00 r* t^ r>-oo r>. t» ^ t^'O r*00 t^O NO 'tf- 0 ^ r^ t*oo ** ^ <*• ^ t^ 0 0 00 fo 0, 00 06 d 61 Woo Ovv6i/3 MMMO00NO>r. 0. o, 6 6 o> 6 d o.t/2 06 06 00 61 0. d d 00 w 0. 0179 .0192 ■ 0202 • 0223 S . 0214 . 0207 • 0132 s >n o. m •* 0. ir, 0 M >a.no 00 0 ov c^ tt - oooffooooW 6 00000000'^ 6 frit^MOO MOO 00 0 •O oaiMMOM-oooo M«Mt,HNM«N« ^« O.0000 MOOOOOO >o u^ 1^ 1^00 0» 0* 0» 0 t^ ^ P.«M«««C ^ -^ t^oo o,cJvO\ddoido6 Oi ro-O >0 Oi >0 0 >n-0 >0 Ki&6\66\66666 w^lor^NOO »OU^O\ f*500 060606 dec o>o>6sc.o6 r*i^r-o OoO 0 ^M « 00000000' 00 6 io01^0,00«Or>5M OOOOOOOOOO 6 ■ ■ -t iT)^ OOOO^O t^oO Ov covo r^M wo N f^unt^ OOOOOOOOOO 6 r^Ooooooooo MOO ^ MClMMMMCIHt-t NO I^I^I^OOOO C\0 f.^ « t^ t^oO 1/1 >«00 i- \0 ov 0 »^ 0\ »^ r^O CI NNt^NNNMMn Tj- « iri ui t^ r^O ^ 0 """"""""" «^M00O,O0«M (fjoo Ovoo ci d o> 0. li r» 0\ t~oo 0 >/i fi 0 ■«■ OOOiOvOvOCOOO PlOOriMUjr,.*,-, 00 0,00 On 0 0. 0, 0« t^ s 0. 0230 • 0234 • 0237 .0267 .0259 .0252 .0277 • 0150 10 u^ ^0 r^ 0 o, W100 .«r N M M CO o> f-^oo to OOOOOOOOO vn >o f) ■* 0 ■<»■ covO M On c^ f^ ^0 t/^NO 'J- to OOOOOOOOO « 0 n m CI 00 00 M « " " •^ •^ "WW M ^ f*)NO N 0 0 00 M « N C " " ON On 0. 0 0 On OnOO H M M « " " M hW t^ *^ 0 CI „ 0 00 « « A« n •^ c;ww in «). -J-NO ^ « CI CO I<5 f^ -J-n -n CO .^«w 0 0 M « N C On 0 0 0 0 2 ^'^ H .n 0 0 0 r* WIH, 0606 On On On On On Ot/^ 00 -^ O M 00 0» fi <*> to O f^ 1*1 M 00 o o oooooWc/5 On conO NO f^ M ? 0 0 35 8n2^S, 0 0 0 0 0 w 6 00? ci c M 0 M 0000 oW 0 On H M •* CO 0 00 0 MCI, c n « M cc/3a5 0 M n n M 1*) ,»■ *n On « « « « n n n c Mc/j t^ 0 0 M ^0 r^ M NO iH M Ci « « C M C Mt« »o 0 OS On CI OCO 00 « 1*1 PI c. ^ CI CI CI t/J r/J °o? CO . m 1*) CO en CI 0 M 00 CO CO CO i»5 CO N W 0 On 0 « MOO r-r-w. CO CI CO CO CO d CI CI CI (/J CO NO WNO .* t» r»oo 06 On On On d. On On OnW W vO 0 CI NO NO >onO On O: 0. 0 0 00000 ow CO .If « NO 0 CO CI 00 M 00 On On On 0 On On On OnW CO 00 00 0 CO CO w **) lO 000 CO ■«■ On conO NO 00 r^ On 0 6 0 M r~ CO On CI CO ro ^ 000 CI 00 N^ 0 0 On ^ ^ 4 ^ ^ fOro 0 0 0 0 0 o^J 6 On 0 00 NO CI ro 000 CI CI CO NO ON CO 0000 0 0'^ 438 Journal of Agricultural Research Vol. XIX, No. 9 o I ^ 'Si. ^ a 1 2^ li Lat- eral ker- nel. • O O. 0 «««««« ^ j«««^««^«,;^ gvOiHtoOtOOOOOO o at s 2 •3 (U a ►4 ■S2S-3 i4 4>^ a gO,w«rr,r,r.^0Q0 ^ 8 2 £"3 g «, -^ -^ m\n r^ 0 •»« ■*« « 2 iu p-JfJ^ a . .00 w - .or,^ " " ^ =0 .g M d ^ «2 ol'S h4 v,^ a ^ 00 00 Oi & 6t O^Q A d [/^ §2 ol'S •oouiwoior. uioo 0 t^ ^^o6o6(^d^666666^o6 « 2 fS i4 v^ a • vo o M r^ ^o lo o\ lo t^ 5o6A(7icJ.(j\dAd\6v o fit •s rt 2 S" hJu^ a OOMMtnroO'^'n-t ^ in O -^oo O 0 ^O fi .^OOOOOOOOO ^ ^6 §2 Si's 0-Sj4 a *^00 c* 00 00 00 .^O O 00 o ^^ooooooooooo ° ■^ rt 2 Su i4a:^ a OO-OOOiOt^OoOO* 00 .looooooooo O 6 .1 0. o 2^ P M 2 S" hJt!^ a •mO««ih«i-5O00 1""""""""- w Qi^M a g T??" " " ??" a 2 S"2 i4 ^'^ a **1 M fn HI .ft 0< ^"" "«"""- W 1^ '•0 1 3 i4«^ a ■M»|.i--tM.o t- n « |.......... « 2tn»n^O« wvo ^ooaaosovawo oi c/3 ^ 2*!« OvJSa • looo O 00 w »n "(f 0 00 ^,2!i;-3 au^ a j5o6a.»o\«jio N ' 1 i ? f « f ^ 00 0 " M N 0 0 0 W M « W « « « « "W 0 t» .»■ t 0 " « " M « N « » o\ Q^'ji TTO r^oO t^ N 1/1 0 « 1 i^dodood- otfl 0 M 0 « -o 0 « ^ 0(d\A(7i9t6 Oi to oiM tfl 1000 0 M 00 U1 XI 0 in ifi » r!.oo >0 0 000000 0 6 ^- r*5 (V) -* 0 tnsO a* » ?« 6 1 t/; »A 0 vO n IT) t^so 0 »^nO 00 o> 0 ^ 0000000 0 6 0M0«fO«O»M »7J .^. M Tf 0 1- w « CO 0 Ov 0 0 -1 0 ►> 0 OCfl 10 CO 0 to 0.0 0 to o> Oi 0 0. OV Oi o\ o.tn c - CO COOO C< 0 ct M ^ CO 0 c< w r^ 0 CO CO CO .« .<)• CO .tr 0000000 3'fi ° 1 c c r t 0 CI 11 t~ M in '00 00000 0 ?tn 6 1 a; 0 irt 1^ 0 00 to » 0 CI 0 0 00 CI CO CO 0000000 ?w Aug. a, 1920 Development of Barley Kernels in Clipped Spikes 439 ^o o o "O o r- I -t ^O ^ '^ -^sO -^f ' >0 O* O 00 r*co r^ f* \ I vO ■^ t^« *o fn f^ fO fO fOW 0606 dvdtdvototao\ v© Ot O -^ '*'«»■ 1 o*aoo 00000 oiQO a ot ds otC/} c^ 6s 000000000 O 00 o fo r^ Oi M f^o 00 •-• ^ ^ u^ 10 r-O *0 000000000 OiM'Or^ooO wr* 00000 O^ o o n .0 o 00 »> V, 0 »^ .« M " " " " " " " " «W 0 M •* .„ ro V, „ M 0 " " " " " " " " «M 0 « « VI VO t^ -*• n 0 m m to to to to to to toM « 00 CO r~ r* r^ r->0 V, ") n to to to 10 to to toW 0 ^ ^ 1/1 VO „ to to 0 to tn •0 to to to to to toOT 0 ,r •* t-o M 0 » to 0, » AAA 0 0 0 " " n „ •«• ■♦o ,- r> » „ a 0 0 0 0 0 0 0 OW r* •n r* •0 to t^ to 0 •» CO 01 » 0\ ON a o> 0 Acn ■♦•o «o no „ 0 00 5 9 t^ S'w 0 0 0 0 0 0 0 •» ■0 OVO 00 0 ■» f'- 0 0 0 0 0 0 0 0 Jw 0 „ !* r* ^ .0 „ „ 0 ? 0 o^-* OtO tO^tO^tovi-ftoto Mn««fi«ci««« ^^ Or.•Ov>^^t^T^lo■» t-o cncctinc^ccnc"/! f1 i-i toin-ffototot^ 0 0 « ^»o00 r^-ooo ^ -^ W c^t^li^t^tototot^t^ ro'Z> 0 fOQO »oo*Ooooo or^t^ tOtotototO^OO tototo to^ PI -t Vtl io loo t^ ^ to M W lototototototototo toM n r^ r* 0 r-oo « oi t^ t>.co 000606AAA6AAA At/1 ioO>on in n to« n n 00 COsoeooo AAAAAA Am >oO >oOtot~OioioOio jjj^ttojooajooi-ioo oooooooooo o ^ 6 O 1-" t^oO \r, to toOO w to 1^ w o lOiotoOiiot^Oiio-^t 0 O 0000000*0 o''" 6 s 0398 0356 0381 0421 0457 0439 0472 0408 0385 0333 S 0 PI 0 to ■«• -"f VO« 10 to M """""""''""'" loO o> «^*o »^00 t^ »>• Oi nn«nnnp.«n toC/3 to « ■♦ •* lOO W to to M C«t,n«CC«<,ctn OOto««ioa0>O«« lototototototototo toCrt 000000(5.0{>»^0^0 tOto + totO'*tototo toW n PI vO 1/100 00 t^ lovo « tOtOP^totototOtot;) t^lCfl no0OPlOit»»«t~o Aoo AAAAA6 A AcA ■♦0 •♦■»«vO'O00 0 PI AAAAAooo-ocn 00 0> to Pi >0 f^ i^NO r^ PI 0600 AAAAAAA Aw S'3S???S'??o'w — 6 0 0 ti to to «^ -ifvO 10 ti 10 0 10 i» 10 tovo e^ po to •f 10 10 10 lovo vO loso r^,n 000000000 0'" 6 t-fPI«0O voOOOO 1/1 •"■♦MiO-etoioai^ o?ooooo??o"^ 0 to 10 to 10 pt) m pn PI " " " " " " " " "en 1/. Ov VO Tt r-O r^ 00 VI " - " " " " " " p;w Ov r-. 00 to to 10 tH to p, " " " " " " " " M 0 ^ to m ,, t^O 10 PI PI to to .0 " to PO t;,t/i ,^ p. ■o t^ Oi t^ „ 00 r- to to n to PI PO •» to toW « p, 00 in ■«• 10 p, VI M to ptj " PO PO PO to n torn t^oo <^ 00 M 0 ■♦00 » r-oo 00 00 A A Aoo 06 w •» r-00 0 •» 0 M •* P, 00 t^oo 0> Ov 0 0 0 OW " " M m 00 01 ■* M „ „ „ 00 '' «^oo A A Aoo aot/i ^ 10 t^ ^ ^ ^ 5 ?W 0 0 00 0 r-» to 0 00 0 00 S-m 0 0 0 0 0 0 „ 0 ^ ■0 ^ 10 p-1 !h ? ^ ? p^ to ?w 0 0 0 0 0 n to .0 .0 to t^ to M 00 ** '^ " " " " " " -M r.ao 10 »-^ t^ *^ .0 vO „ " " " " " " " " t^M to .0 n 0 ■» 0 ■* .0 0. " " " " " " " " «W M » •0 .0 10 r~ 10 ■0 0 to PO to p^ PO n to to t^W w „ 00 0 0 t~ »00 •♦ to "> to to ■* to n to loW ■» p, Ti- t^« >o t^ 0 P, PO to to PO PI to to PO poCfi » H p. 10 0 r^>o 10 V, 00 AAA 0 A A Aoo 73 p. 0 .0 01 p, 0 00 VI 0 o> 01 Oi Ao 6 0 0 Ot/2 M 0 to .0 10 r-oo 00 „ O' » o> Ol 01 » A A Ax ^ to ^ ^ r>- t^ r^ ?c« 0 0 ?! „ too pn r>- 10 PI i 0 0 0 ^ ^ -S -^ ?w 0 s 10 10 to 10 0 0 0 o"'* 440 Journal of Agricultural Research Vol. XIX. No. 9 to ^ o '^i. S •^ •^ < i '5 i "3 Lat- eral ker- nel. e O ■* m t^o O.C0 r~oo M 2 S" 1-} v.:^ fl jj w >« lo-o t» •* 0 00 a (4 "a « 2 fc'S g PI t» a oioo o oi o 00 >S <0»^<^<<5'^*W)'* tij^i a 2 «"« • i/lvO o t» t» »» moo i rt 2 vH • "*-OOMt/1000000 k~t/Jo6 6»)t^<»> £Wo6o6 d>« dvdcoooW 1 ^2o-3 • 00'O05t^O*'O'.« xj-i ir, -^ ,? o o o o o 0 o o 0"^ 6 n 2 Ss o.o>op«ooo>nw (^W ? J J J J JSgw 6 'a i o "3 II 1^ « 2 S'S ►j£j4 a |.P,p;.p;.p;.... a'S i-- S 2 K « • vj r» r»oo »^oo r* t^^ 1 «««««««« «wr rt 2 ft! ►4»;.b< a ^ a M -V-O 0 « Wi "1 PO 0 a s '•3 "3 .-J ra 2 «3'3 J4noooo 0 OiOoo oioo ^ r^m^.it'ri-it'hdo\666-MC/3t/5 "a 2 STJ H4a<.i4 a • r^ a 0 •* t'O 00 ui >« "5 ^o6o5 dcjidddovo. CT.t/1 ♦J BO hjSj«i a P*QO P< »/)0 P^ ^ PI »*5 Pi j| 0 o"???© 0 0 0 oW 0 S 2 SI'S o*^M a 0 to f^ Pi OvOO 1/100 w ^- M 0 poi^toeocoM ^^ooooooooo"^''-' 6 . r^CC OOMMPiOvOrO s. 't 'T OS t^oo 0. OS 0 0 »/J (JOOOOOOOOOO 6 ^00 00 t^oo 0 0 00 " " " " " """WM "w sOMP-PiOsOOOOOO 0 p«ni^ .ow 00 0 0 0 CO mco sOOOoOOsOOOmi^ t^ •h n n n 'h ■i -k f^^f^ mW •♦OOtlMwOsOiPi 0 ii-tfl «~« M 0 10 t/3t/3 i^ 1^ •»t/3 ^ nt/i mw PiplOPlP«00P«c^ n 06 6. 6. d< ds os ds tJst/JW d>OT ^t^MPiPiNPII/lrO 0 oOOiOtOOsOOOO oW 0 «»• 0 OsOO wen waaW » (>W WW 1^ rOO P^sO P-lsO «^ t~ 0 M « kO M P-JO 0000 0 "S 0 ofS 6 1^ Jw 0 "tr^M ^loipiMOO ■*»o 'to so 0 vo 0 t^ ooooooooo fw 6 00 t^ ■* 0 t^ " ■il-OO Jl t^ WW S'o'oWS o« t»00 Oi>0 t» p<) """"""""" w O0000000«~t^t» p»ponpiPOiOCT«« w wi mo «~ U100 0 m w 0 *^ r* 0 t* Oi t^oo •^ i^ po •* >^ -^ -^ i^W w 't r^ i-i (.T) to too Oi P< w 0000 OiOiOiOiOiOiOi i-Pi0000PiOt~i-iM w didAdi666i-i6 t^ Pi li^O 0 0 Pi Oi 06 d <7i oi di 6 di diW w 00 1.1 M 10 t^Qo w5 0 00 ioo *^oo 'too po 0 -^ »*5tttiOiOiOiO^ OOOOOOOOO w 0 PP)tO»Pi iOPi iOPi fil 0»0 p^ r^ iooo Pi Pi 0 ■•tiOOOOOOO io ooooooooo w 0 00X00 'to P^Pi P^ « Pi 0 ioOO OiO 0 ■it -t -it ipi 't io 't iof/2 00000000 6 w Aug. 2. 1920 Development of Barley Kernels in Clipped Spikes 441 ". "iT.«-o loi/ii/i^r """""""" "tn U^OCOO 00 O\Q0 0»^ «r,c^««««r.tO .a-Tfi/-, iAi«r,>oi/^w """""""" ""^ w t^ *t io\0 in .* n. **) r#5<*3f*i»*5r*)f^«*3'^ fi)C/3 •♦OoOOCOiOi-w W5*n>*w)iO«^ <«)W ^t^vo^ r*oOoo ^w^ rriwji^tfiwit^K)'^ riC/J OC060606 O^Ol»6^ 6.C/3 060606 6tOi6\ototo6c/3 NOw-Nr^MOOOvO oo?ooooo 0 W 6 vOQC ^0*f^^oO»000 0000 00 00 0"^ 6 0. - r-oo M r. 00 >»■ 0 ^«00^00■/-)')■^^(^ 000000000''3 6 n U-, M t~>c t^vO » 0 """"""""""^ *r>^C oc r^^c oc 00 vO x^ «««c,««Nri«ai « .^ ./, .« ^ u.,0 .«. !.««<;<;«« «to .O in a triOO 0 ^0 .* .jwiw)-^!-; W)w>t^ ^t/3 T r r^ r**o iH 0 0 e^ t^ .lW)<^>n.«-.«-.9.'^ w,t/2 f*^ mvc vc XAO r^ r*) ^ ^ ^ ^ „ ^ „ i^Cfl 0 Orif*i«^0<5't. 0 s a o> ov 6v 6^ 6 d o>(/J 0 to « f*) »A r^ 0 00 0. 0 6. 6 a 6 6 -^ 6> » 6. 6\ a ovW 0 M ro f^ 0 00 0» r^\0 3 3 0 0000 0 w 6 ■ " 1 i oc .^ »n f^ ^ M i^\0 0000000 o^J 6 ■ 1 fi 0 1*5 « ..J. r^ moo 00000000 "■' fO »*> f*) i*> «*>»») ^ K)^ *^oo o O t^ O o r* «^ **) ^ r^ »^ "i" -^ rijW 06060606 OkONOtooCQ 00 o» o* d> ©1 ^ o\ ost/5 'tW ^ O O »^ f^ 060606C/2 do^o>o6t/i r*i ^ f*j ^ rj- i/^ 10 fO-- 00000000^ ^ tvO O O* ■»T -^ <*) 0000000 o ^ 00 •- <*) ©■ "^ t^ fn 000^000 o^ ■ r* r*oo « ^ I oc 'O O\00 ©(-O r» »o « « « N w c/2 « « 75 « 00 r» On r^ VI ^ i^co 00 o 00 o 00 00 r^ t*) fJ) ^ r^ f^ ro t o>tO o* OiCO « f^ fo o "O « 00 00 -^ O 5 « r* « 000000 o^^ 00000000^ « « 0>>C O 00 00 o» o* "*: w. 0 >o 00 00000 ■'■' 00* 442 Journal of Agricultural Research voi. xix, N0.9 c/) ^ Q '»*> -« •^ » < a « la Lat- eral ker- nel. . vn 0 r- 0. " ""T? 1 1 a a a 0 1"^"""""""^ 1 go M W ^ - M P« VO l?""'^'^" rt^.lw j4 i-j^ a &'»""T'?Tf-^ ?T,„ ^..« «««.«.« v 0 s •3 i4 4*^ a ^•^Ot^f^O'^Ot^O'^ ^^^fh^^h^^-itfh^ S 2 Si's ^ t- 0 0 M 0 .^ 0 00 li'^"^^'' ■♦ -i n^ t« 2 S'S • -^OO 00 ^O W « NO 00 vO ti a .5 •00 mt^noooom £ r.t/30600 OiOOQO OlOOW 0*=.i4 a HjS^a • 00 0 ■* »«o w M pn a» ,^ r.06060606 dsONaoot/J '53 ct 2 St hJ*-^ a .* 0 c« 0 0 0 0 0 0 0 «« ^6 0-j< a Gm. 0. 0486 .0584 .0643 .0688 .0703 Dec. .0747 • 0753 .0568 S M 2 u" (4 vM a Gm. 0. 0389 .0458 .0483 .0452 .0442 .0366 .0528 .0514 .0484 S ■5. 1 II * 2 Q rt 2 ilJ i-l v^ a ■ Tfooooooo 000 t»M ^.....^...cn cars il,-^ S 2 4* « u«^ a _; -OOxOOinn ^t/i m <^ n 'lO w 00 0 *^ 1^ tJ5n4 a 2 S"? hJ**^ a ,^000000 6. 6. 6i 6> 6. p-t/5 U-ls^ a • fo 0 » M K500 0 >« a 2 *''3 ^S^a ■ t, - « - « 00 .0 -0 - ^ooo6oi»oi Ov ON Ov 0 otn •♦ r^ t^ 0 n 11 Ov „ 01 oco 6>(/i ■>»■ w1 6 o. t> Q (j 2 2"! 'N00r-O0'-''0'0 M???-8 ?o^ ot« 0 . OS ^ 0 000 0 00 0 000000 0 ^ 6 1 >o r* t^oo 00 « 0 w . . . . . . .(flCfl t^ « 0 « 0 0 « 0 0 " " . f*i**)<^*)«n.C/) n t-o r- 0 t~oo m n | " " . . rn « « « «cn ■0 m n r^ »oo 00 « 0 ro « to rrn «^C/3C« t~ 0 p. M W M M 0 0 n ■» w) ■«■ ■^ ■* ■* ■«■ wjtyj oooooo 0 0 t^ r^ n f^(*)(*)»*)'^^»n«n f'^t/J •* tfio: 1/1 0 00 ■* 0 0 06 00 06 dioo d>t/3(/3 0 n t- >H 00 « 00 00 00 iaao d>6o6co Ov « ci ir, 0, 0 00 « "i r- 06 06 06 06 aoooooot/} 6 0 --2 M ^ " 0 0 0 0 0 o^t^ v88 0 0 0 0 r< t« f " ro 000000 O^ 0 ? ? I" ? 1^ ? ? J'" 6 • Aug. 2. 1920 Development of Barley Kernels in Clipped Spikes 443 0 ■«■ -"C-O ■^'O 10 0 M « « <; n « n «cfiwt« •*«) aoooo 1- t~ 0 1^00 0 « ■* i^»o ■>»•►. '. n C/^^ncinnnn ^ «^ ooi-iar-<^'*»^f^^»*J« fO 'o'^ 0 m ■ ■ •* W<*)(*)to*o**>fn« ■ <*>C/1 (»M'4'nt~i->'«0 •^OOOOOOOO OiOiOOWWW oooo 616 osooidto den f*l c^oo »oco f*) r^ w 0 o'o'Jo'o 0 o'-^WCfi d w>r-M\0 0\0 r-t-'Pi 0 oJoooooo o"?"^ 0 Woooooo'o^ 0 '^ d '^ • m r~»0 MOW ao 0 0 •*-«5«>n ^ o» o\ « i-t t-" r^ «r)0 r^ o* 0, 0 00 »» 00 0 w (*i »*1 (n-O ■* - .« t» 0< » a V) 00 cotfl CO CO CO CO cotO coOT ONOr^O\00 010 06 06 06 OO 0 dvC/1 dioo CO r^ r.. CO CO w r-.00 ^ 0600 dvdid i>(>dWC/5 M t- <« M 0 CO OV 00 :/} 00 06 d\ di den 06 cn 0. 0387 -0477 .0480 .0598 .0608 .0660 s -0584 .0463 s -t-»->«ooo w^og d d M c^ ^ t^oo c* 0 >■• r^ ««« M 0 't fi -.)• -to t^oo 0> 1-1 OT c; t; ^ p; t; « c^cnt/J t-i M c'ir..^..o^o^O^^- '^'^'^""^"'^'^'^^ p^uir^OONi-iOMO n'^n-in-i'-i-i-i^ t-t -t m r.. t>.oo 0 w 1^ <^ n 'h 'h m -icntn -Ococo«0««»i0\ t^oo 06 0606 d> d> dtoocn ■■ro t- CO •* t» « is 0 t^oooo O' 6^ o> o> 6 den 0 (S iri -^ 0 VI Oi Woo 06 06 06 woooocncn «ooooN<-iowa,« OO-tOvOi-luiiNB.-). 000? 00 0 "8 oOT d lor^r^coOsoooOQO irt 0 t^ CO 10 ro 0 t ««««««««« cCT 0 -too t^oo 0 00 r^ Wcococococo-i-co cotn en 10 r^oo c« 0 cs PC 1^ M r-. »Ococo*-»-*-ct-*-* cocn 0 CO VI .«0 VI r-00 NO 0 fOlOcococococococo coC^ 1 VINO 0. Oi n -cf-O -NO 00 VI OT ^,oc^c;ic;>c;>c^c;>OTOT M mO r»NO 00 0 0 t^ m nn'^nri^^-i^n-^^ O00f<00-*vi<^ro« ennWC^lNC^C^C^C^JCOOT WPIOOOOOPIC^ en 00 06 didioo 6\ 6\ a^^tn. 00 0 conO OnQO 0 « « 00 oodNdidtdNONddd dien •ONOcomOwvjm en «--o6 00 06 di On di dv oxn S 0. 0258 -0330 .0401 • 0516 .0462 -0503 .0560 . 0489 s s V) CT 0 VI * -t 0 •«•« «f VIO CO t--ON^»^CiOOO ???S'o'o'^^o?OT 0 NO " NO Sio M 0 X v> OToooooooo ffOT d •<*■ --t 0 On 0 M »00 r* « « ,o« cocow r, «en t^MM^ncoOi-400 M colOcococococoCTCn VI t^O 00 0 0 On t~ 1-1 «ti«CTcoconM«cn « -t r^ 0 00 r- » r»o coc^co-*cocococo coCn V)00 M CO CO CO 0 1-1 00 COCO-*-*-**'*'* cocn * r^ ■*no 0 00 r^o 0 cocococo*cococn coCn On r^ 0 CO 0 conO 0 CO r-eo o> 6> & 6> 6, 6^ 6\Vi 00 n 00 v^no 0 viO v> 06 dioo On On d d d den wNOwi-'OPtviOr^ 06000006 ONdNdidNoocn 5" 0 o'S"S~8'8 0 oOT 0 0467 0580 063s 0780 0741 0804 0692 0719 0617 S 0 NO-Oc-<»0-*r.v,o 05t~~l-.-OONON»• >/1 lO t^CO M ^ IH .^ a < W a 0^ 00»in«>X5*« g w t O 0\ »>0 ro HI « f to Vl'^NNmNNNWjnrOM a .2 •S "(3 « « 2 S" g N'*iniOt»MOM<^0 CO vi''^ .o«)>omm4't"i->J-'« S2 ol'S g I- oooO'>t«' « 2 S" ^ fOVlt^»0«^'^MMeO J5 bo g ■ vo too « 00 vO vO « O >0 ^t/Joooooooooo oidvddv 6.C/5 n 2 SU l4 i'J'i ^ • o vo 00 0 O 00 mo « >o .^ W 00 00 00 oi dvoo d^d^o- d\W 'S >4 U.M fl J lOOO t'O ^ 0\ r- (*l ^ w CO 6 g-iij-s 6 6 a o li ©•3 9 ^ iJ XI ■ OK^i^c-l'flt'^fO • ooo*^««*iOo\0 a 2 •3 "3 « •N0\r^0000r*O» ^ ^ mn-9 ■i -i- 't ^ mVi »f i^'i-t-tt'tt'i- tt/2 • t»00 ■* M « O>00 00 to a l-t •Ono^o^-^n-^on ^06060606 6\6io.A ovC/3 g 2 S*'' •00MmO««00KOM .^^o66id\6\d6oi66^ :§ »M MM • V110M)«Mt»OvO ^ CO 06 WOO 61 oi 6\o6 6\t/3 •a « 2 S'S )-? 01.!^ fl Gm. 0.0350 .0587 • 0590 .0626 .0636 .0691 .0653 •OS77 .0606 s 4( v. 11 4J U*'.^ fl . »0 ^vO 10 t^ PO M VIOO « 0 NO 00 »o XOOO <*^ t 0 6 HlS-«fl Oiii^mONOOOi £. w t r^oO 0 M !>. « ^5 00000006''' 6 ■^ to [^ mo " ? ? fO f* « mW >o 0 M ro PO «co 0 « t « N •* ^ M m V, 00 -0 " to «5 " -0-5.0 « «to M « m r^ f «•>)•■* too " -0-^^^ -5-= »o .^ nM 00 0 M M « 0 M-O " ^ «5 <0 <0 •otO •<»• to V, to. r* 00 00 00 d d\ d d r>(0 >0 00 00 (N too 0 rOOO t CO 00 CO d 6> A 6 0 ddco 00 « 00 c 0 M w too »^o6o6 61 d dcOoo d o nnnfonnnn ri CO 0«o »» Oi Oi t too 0 >o <^tO")W)<0-5 -o t t t tCO CO tooo r* 0 00 to M 0 00 00 t=.o6 doo d d doocO «^tO«00 i^ 0 t^ t lAOO M 00 0 N- t o>o tt>o»/^tt»0»0 t« 10 "tOOt^ioiOMO 000000000 "J »AO 0\ Wl to N O O i-«tMtMn to»^»^. looooofOCO 00000000'"' Aug. 2, 1920 Development of Barley Kernels in Clipped Spikes 445 « « « « N M « y^ N j^ 00 o> 01 0. >o OvOO N ri M « « « <;cnt« ^M MNtl-O-MiO-"}- 00 rf, « ,;, n ^ A ^w «w ,h n ^ A 'h h^ -^^ 0 M M to •* 0 r- <^ to w) .^ to to WW n^ 'I M vn o> N coco d.d.o>Oi6c/3dM M 00 m to ■<* r^ r^ 06 «=.o6 06 06 06 W W 06 W l^vioMOvOn 00 g'ffo'JS'tS'w'gw 6 vO 0 N r^ r. M n IS'o'^?^~8t«S-M d 00 0 ■* 0 N N n 0 5'o'S'i'iww ^W. •000 0 « M woo « c; ^^n ^ ^ ,i ^ .A ^^ ^ ^^ 0 0 tottot^iotoM OOsO 0 fl -tJ-tot^t^N n^tototototototoy^ « « T m r*5 r>-oo w 0> fhththth^rhrht^ t^^ 0000OMO\Wf-lt-ii-( Wcif^ifOCIrOfO*^ »oC/2 mvo CO ro N 0 00 0 m t^t^t^oooo dod aodc/l M 0 M -^W t'.!>>000 CO 0^ <^ d\ &, 6\ 6^ 6^ o^tfx 0 Oi 0 t PO ■TOO t^ T CO t^ooodooodcoooooCO T 2" '■^ "^ <^ « a TOO 00000000*0^ d 00 2^0 \o U-) N o\ m 0 d to t>. cj vo r^ too to P* >oioOl^O>p,ciOsiO oooo'SS'oo 0 ='5 d ^ »rt»0 1000 0 ON w w N « N « t;c/3 mW »o r^co Oi 0 0\ 0 ui " « ci t; ,;, ci c/j ti ri J/3 W to lO^O loo 10 ..t M """""««" «w 1- to CN Tt 10 .» 00 tfl rh ^ n ^ ^ '^m •^(fi N too 00 00 r^ Tt to to t^) c^ -;> to t^iw t^) totn 00 0 to ■* >oo -0 N 0 PI totot^jjot^t^to toW Oi Ci toQO ^0 00 tflt^ooooooooooOT t^ta 0 vo "O »o r* r^ r^ doood d. d ocfl d dW VO M to 1-0 0 00 M 10 ^-00 00 oioo doo doot/j 0 00 00 r4 CO n r^ Wo'o'oooot«oW d 0. 0343 .0403 -0452 .0556 .0606 .0583 s -OS37 .0454 S «*) moo 0 0 r* 0 rt T sO ■-• 0\ ■-■ TOO r- 0 m « rnf»lTTTTT fOf/l 000000000 ■'•* d 0 M - --t »^ -"t to to « " • N « « « «„3 ^uiihO\io t^io»^ o\ to M 10 Tj- >0 to to " « t^ 0\ • « to C< M 0 N C ; ^ ^ ^ ^ ^w CI tOOO 00 « 10 M ^ VO 0. •^ M . M t-vO Tt f^ 10 •"■OOq 06 06 06 06 06 CO t/1 0 0 d • 0 r^ r^QO M ^ 'J r^ Tf to 10 Q ff ? ? 0 o'^ 0. 0328 •0379 .0230 • 0232 .0391 s .0500 .0436 .0476 s Mtp .OO-^-Oflt-M ooQooo'0'0 s^ d -a f*^0 sO r^ 00 r-"0 1 moo 00 0 r^ 0 0 r^ N to to73 too 0 ov 10 0 0 ■* to to to to to -^ ^ totn OCO to-^Tj-f^viiO to c< to to to to to tot/5 r^ 0 0 10 looo CO r-coooooOTooooootfJ CO >o I.* 10 t-»0 0 0 *--co <^ 6^ 9' 6^ 6^ dc/2 0 ^ t^ »n 000 0 to 0606 06 CO 00 06 d dW 0. 0287 .0360 .0419 . 0416 S .0498 .0476 •0429 s 0 !o loM ? 0 "» S> to ^ t 10 »0 t>.0 Tf ,/\ 00000000 ''J oooOtOMOO-tr^ OOO < "3 5 2 1 0 2 V a 1 1 1 2 2 V > < "2 5 2 2 CI CD i 0 Mgm. Mgm. Mgm. Mgm.. Mm. Mm. Mm. Mm. Mm. Mm, Mm. Mm. Mm. Mm. Mm. Mm. July I 0.9 I. I I. 2 0.9 I- 2 1. 0 1. 9 2. 1 1.8 1. 9 0. 7 0- 7 0. 7 0. 7 1-3 1.4 2. I 2. 0 3* 3 2. 0 2. 1 . 7 .8 7 7 3 S I. ^ T.R I. 2 2. 0 2. I I. 9 2. 0 . 7 .8 7 7 1-7 2.7 1.8 2. 2 2-5 2. 2 3-3 .8 I.O 9 9 '0.6 0 s' 0.6 0.7 7 6.3 10. 2 S-7 7-4 4.2 S-2 3-8 4.4 I- 5 1-7 I 4 I 5 1.0 I 1.0 1.0 8 6.7 8.2 5-9 6.9 4-3 4.6 4-0 4.3 I- 5 1.6 I 5 I 5 1.0 I 1.0 1.0 9 7.0 9-3 7-1 7-8 4-3 5-0 4-4 4.6 1.6 1-7 I 7 I 7 1. 1 2 1. 1 1. 1 lo 10.7 14.7 10.3 II. 9 5-9 7-2 5-8 6.3 1.8 2. I I 8 I 9 I. 2 3 1. 1 t. 3 II 17-3 20. 1 16. 2 17-9 8.2 8-5 7-8 8.2 3-3 2. 2 2 2 2 2 1-5 5 1-5 I- 5 13 31. I 27.0 19-3 22. 5 8.9 9-7 8.6 9-1 3-3 2.6 2 2 2 4 1.6 7 i-S 1.6 13 32.7 39-9 23.8 3S-S 8.9 9-7 8-9 9-3 2-4 3-7 2 S 2 5 1-7 8 1-7 1-7 U 31. I 40.7 33-3 35- 0 9.4 10.3 9-5 9-7 3-9 3-1 2 9 3 0 2. 0 2 2 2. I 3.1 IS 41. I 49-3 38.1 43.8 9-5 10. 2 9-5 9-7 3-3 3-5 3 2 3 3 2.2 2 3 2. I 3. 3 l6 39' 2 48.0 36.5 41.2 9-6 10. 0 9-7 9-8 3-3 3-6 3 2 3 3 2. 2 2 4 2. I 2. 3 17 39-3 SI- I 40. 6 43-6 9.4 10. 4 9-6 9.8 3-4 3-6 3 4 3 5 2. 2 2 5 2.3 3-3 19 44-3 S8-S 42.4 48.4 9.4 10. I 9-5 9-7 3-5 3-8 3 4 3 6 3.4 2 8 2.3 3.S 30 45-5 S6.6 44.1 48.7 9-4 10. 0 9-3 9.6 3-5 3-7 3 4 3 5 3.4 2 6 2-3 2-4 31 41.4 S4-S 42.0 46. 0 9-3 10. 0 9-3 9-5 3-4 3-8 3 5 3 6 3-4 2 7 2-4 2-5 33 48.9 61.3 48.8 S3-0 9-4 10. I 9-3 9.6 3-8 40 3 7 3 8 2.6 2 9 2.6 3.7 33 SI. 8 60.2 SO-S S4-2 9-3 9-8 9-3 9-5 3-8 3-9 3 7 3 8 2.6 2 8 2-5 3.6 M 44.0 S4-3 45-6 47-9 9.4 10. 0 9-3 9-6 3-5 3-8 3 6 3 6 3-4 2 7 3-4 2.5 36 41.9 59-3 SI- I 50.8 8.9 9-7 9-1 9-3 3-4 3-8 3 7 3 6 2-3 2 7 2.6 2.5 37 SO. 5 67-3 SI- 7 56.5 9-2 9-9 8.9 9-3 3-7 4-1 3 7 3 8 2.6 2 9 2.7 3.7 38 SO. 0 61. 0 48.3 S3- I 8-5 9-3 8.6 8.8 3-6 3-9 3 7 3 7 2.6 2 9 2.6 2.7 39 54-3 72.2 53-2 S9-9 8.8 9-5 8.8 9.0 3-7 4-2 3 7 3 9 2.8 3 3 2.8 3-0 30 46.0 63.8 51-6 53-8 8.8 9.6 8.9 9-1 3-5 4.0 3 7 3 7 2.6 3 0 3-7 3.8 31 49-7 67.1 S3- 3 56.6 8.8 9-7 9- I 9-2 3-6 4.0 3 7 3 8 2-7 3 I 2.8 2.9 Aug. 3 59-3 74-9 S9-I 64.4 8.9 9-7 8.9 9-3 3-8 4- 3 3 8 3 9 3-0 3 3 3-0 3-1 4 S5-8 73- 0 49-8 59-5 8.8 9-7 8-8 9-1 3-8 4-3 3 7 3 9 2.8 3 3 3-7 2.9 5 S7-S 77.1 57- 3 63-9 8.8 9-9 8-7 9-1 3-8 4-3 3 7 3 9 3-0 3 3 2-9 3-1 6 39-4 49.3 38.3 42-3 8.5 9-4 8-3 8.7 3-3 3-5 3 3 3 3 2.4 2 7 3.5 '■S 7 43-1 52.4 41.2 45-6 8.7 9-2 8.4 8.8 3-3 3-7 3-4 3-5 2.6 2.8 2.6 3.7 ' Harlan, Harry V., op. ai. Aug. a, 1920 Development of Barley Kernels in Clipped Spikes 447 Table II. — Average net weight, length, lateral diameter, and dorsoventral diameter of kernels from, normal and clipped spikes of Manchuria barley at St. Paul, Minn., in igi$ — Continued CLIPPED SPIKES Wet weight. Length. Lateral diameter. Dorsoventral diameter. Date. "3 E 2 "3 1 a t3 "3 g .!4 "3 2 > < M 1 ►4 "3 2 S u "3 E 1 > < 1 2 1 a "3 1 2 1 "3 E 1 5 1 2 a 5 1 2 V a 2 > < July 2 M,m. i-S 1-4 2. 1 5-6 4.0 8-S 10. 8 14-8 20. 2 28.3 29- 0 32- 4 35-6 39- S 39- S 35-8 40.2 S6.3 41- 0 45-2 40.8 4S-7 41.8 39-8 48.4 34-6 57- S 29.0 37-7 34-2 31-3 Mgm. Mgm. i-S I- 5 2.0 5-6 3-9 7-8 10. I 13-6 20.8 28.6 29-5 35-8 35-9 40-7 39-7 42.5 42.2 53-5 35-9 42-3 43-9 48.2 36.9 37-1 49-4 44.0 56-7 33-6 40- 7 38.8 32.2 Mgm. 1.8 1-7 2-4 6-4 4-5 9-2 12. 2 15-6 21.8 31.8 32-7 35-5 38-2 44-5 43-3 42.6 45-0 57-5 44-1 48-2 47.2 52-9 43-6 42-6 52-2 44.2 60.6 33-9 45-1 41.6 33-6 Mm. 3. 2 2. I 2.4 3-8 3-3 5-2 6-4 7-3 8-3 9-1 9.0 9.1 9-3 9-3 9-2 8-6 8-8 9-4 8.9 8.9 8.8 8.6 8.9 8.6 8-5 8-7 9.1 8.6 8.4 8.3 8.4 Mm. 2-5 2.4 2.6 4-5 3-7 6-1 7-7 8-0 9-0 10. 0 10. 2 9.6 10- 0 10.3 9-8 9-3 9-7 9-7 9-7 9-7 9-5 9-3 9-4 9-4 9-2 9-7 10- 0 9.4 9-5 9-3 9-2 Mm. 2.3 2. 1 2-3 3-8 3-3 4-8 6-2 7.0 8-5 9-1 8-9 9-3 9-3 9-3 9-4 8.8 9-1 9-2 8-5 8.9 8.8 8.7 8.6 8-5 8.6 8.8 9-1 8-7 8.4 8-3 8.4 Mm. 2-3 2. 2 2.4 4.0 3-4 5-4 6-8 7-4 8-6 9.4 9-4 9-3 9-5. 9.6 9-5 8-9 9.2 9.4 9.0 9-2 9-0 8.9 9.0 8.8 8.8 9-1 9-4 8.9 8.8 8.6 8.7 Mm. 0.8 ■7 I.I 1-4 1-3 1.6 1.8 2.0 2-3 2-7 2.8 3-0 3-2 3-4 3-4 3-3 3-5 3-8 3-4 3-6 3-5 3-6 3-4 3-3 3-6 3-1 3-8 3-0 3-3 30 3-0 Mm. Afw. 0.7 • 7 1. 1 1.4 1-3 1-6 1-8 1.9 2-3 2-8 2-9 3-1 3-2 3-2 3-5 3-5 3-6 3.8 3-3 3-5 3-5 3-7 3-3 3-2 3-6 3-4 3-8 3-1 3-3 3-2 3-1 Mm. 0.8 Mm. Mm. Mm. Mm. 2 3 7 S II 15 18 24 38 39 38 43 S3 50 49 52 62 S5 57 S6 64 52 50 S8 54 67 39 56 51 37 I 2 9 6 4 8 4 5 6 5 4 2 3 7 4 7 7 S 2 8 7 I 8 7 I 5 I 9 7 2 I I I I 2 2 2 3 3 3 3 3 3 3 3 4 3 3 3 4 3 3 3 3 4 3 3 3 3 8 2 6 4 8 3 2 5 I 3 2 4 7 7 7 8 0 8 8 7 0 6 5 8 7 0 3 7 5 3 I I I I 2 2 2 2 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 7 I 5 3 7 0 0 4 9 0 I 3 4 5 5 6 9 5 6 6 8 4 3 7 4 9 I 4 2 I 5 7 8 9 10 11 12 13 14 15 16 17 19 20 21 22 23 24 26 27 28 29 »o . 31 Aue- 2 4 s 6 7 0.8 x.o •9 I.O I. 2 1.4 1.6 1.8 1-9 1-9 2.0 2-4 2-3 1. I 2-3 2.9 2-3 2-5 2-3 2-5 2-3 2-4 2-7 2. 2 2.9 2.0 2-4 2-3 2-3 0.8 I. I 1.0 1. 2 1-4 1-5 1-7 2. I 2. 2 2. I 2-3 2.6 2-5 2-5 2-6 2-9 2.6 2.8 2.6 3-0 2.6 2.7 2.9 2-7 3-X 2-3 2-9 2-7 2-4 0.8 I-O -9 1.0 1. 2 1-3 1.6 1.8 2.0 2. 1 2.0 2. 2 2-3 2-3 2-4 2.8 2-3 2-5 2-5 2.6 2. 2 2.4 2-7 2-S 2.9 a. 2 2-5 2-4 2-3 0-8 1.0 •9 1. 1 1-3 1.4 1.6 1.9 3.0 2.0 3.1 3.4 3.4 2-3 2.4 2-9 2.4 3.6 2-S 2-7 2-4 2-5 3.8 2-5 3-0 3.3 3.6 2.S 2-3 In both the tables and the figures it is apparent that fertiUzation did not occur until about July 5; therefore, the measurements and weights before that date are of the ovary. Six-rowed barleys do not flower so uniformly as 2-rowed barleys. The central florets flower before the lateral ones. For this reason the curve of growth is less abrupt at the beginning than is the case with the 2-rowed varieties. Even with the prolonged period of flowering the length increases very rapidly after fertilization, as may be seen in figure i . The effect of the clipping is evident in both figure i and figure 2. Although there is little difference in length near maturity, the lateral diameters and the dorsoventral diameters of the kernels from clipped spikes are noticeably smaller than those of normal spikes. The difference is even more conspicuous in the wet weights per kernel. The question of mechanical injury from clipping is answered by a study of the growth by days. There is no such injury. For two weeks after clipping, the kernels in the clipped spikes develop as rapidly in size and weight as do those in 448 Journal of Agricultural Research Vol. XIX, No. 9 normal spikes. The graphs of wet weights per kernel coincide essentially until the fourteenth day after the experiment started. If there were a mechanical injury it would probably most seriously affect the kernel immediately after the injury. After the fourteenth day the kernels in the normal spikes increase more rapidly in weight and size than do those in the clipped spikes. On only two days after July 14 do the clipped spikes exceed the normal ones, and these excesses are unquestionably due to the error of sampling which comes from the use of a single spike for this purpose. The difference in rate of development begins to be noticeable about the time that the growth in length is completed. This coincides roughly with the beginning of the period of rapid starch infiltration. Whether /r/y. ■^ ~1 t ^ — — _ 7" ~j ~ ~ " so ^ >> ■' \ »^ , « •V, .«. \ ^ ^ *. ,^ _ / 1 -1 ' -■' ^^««/^v»zv so ^v5■ r, - a. '/* K ' A/ '/H ■* '/. 1 / SJ -'/ / 4^ », ' _ _ _ — i - 4 t- _ _ _ _ _ _ _ i ^ ^ « iT _ j= _ _ _ _ _ Jtj - - - i if- - ^ k • 0 ■* bS *l - - ^ =s *! -^ ■r' r r' »: 'I ^ ?• 0 - — ££ f / 7. A/ VA h «: «< s _. «f ^ y ^ C^ ./ *» *■ .s ^HrA779«Z\ w ;- -^ •^ ^ f"' •• *•" 0. ■// f. z> >■ s f ^/^y'jtrrsx' IS /.O as ac > •« « ■A W ^ « ' 1_ J L J u L _ _ L L L L Fig. I. — Graph showing growth iu length, lateral diameter, and dorsoveutral diameter of kernels of Man- churia barley iu normal and clipped spikes. this indicates a loss, in the clipped spikes, of the photosynthetic products of the awn, as well as lower transpiration, as indicated by Zoebl and Mikosch, is not shown by the data. At first the difference is more apparent in the weights than in the dimensions. After the twenty- seventh day from the beginning of the experiment the lateral diameter of the kernels from the clipped spikes begins to decrease. This is probably due to the rate of loss of water in maturation, which here exceeds the rate of deposit of dry matter. In the normal spikes the two changes about balance each other so far as their effect on the lateral diameter is concerned. The dorsoveutral diameter continues to increase until full maturity in the normal spikes, while it is slightly less than main- tained by the clipped spikes in the latter days. At the very last the ker- nels in the clipped spikes ripen faster than those in the normal spikes. This is apparent in both figure 2 and figure 3. ^^r- ^ug. 2 , 1920 Development of Barley Kernels ' in Clipped Spikes 449 CO - <\/ i tSS - J / s. / \ / / / / I \ \ tS'O 1 1 1 V y / \ r \ / t \ \ \ -^- K^S / / \ z' 1 1 ) < / 1 1 1 1 \ 1 <^o ~ h 4 A ■-• - '». r \ 1 V' *ss - 4* '/?/ ■M. ri f * - C^ /A ■^/S ? M 90 )r \ 1 \ / 1 «i \ \ /* 1 / 1 1/ Jj 1 1 \i / \ / < 1 1 1 "y / > } I 1 / / { /' / / \ y / C^//='^£Z? / r y / /,' .4 '/ 1 J 1 / y /. ^* asj ^^ i-m. ^ ^ '■r ■>- ^' Fig. 3. — Graph showing dry matter in kernels of Manchuria barley from uormal aud clipped spikes. 183718°— 20 5 450 Journal of Agricultural Research Vol. XIX. No. 9 The nitrogen, ash, and water were determined in all samples. Inas- much as the glumes were removed, the difference between the total of these substances and the dry weight would approximate the sum of the carbohydrates and fats. This calculation has not been made. Its trend would be similar to that of the dry weight. The results of the analyses N o. a ^ I 1*^ -^^ 5 "^, ^ ?c r 'x._, \ / ^ \. \ N ^v. \ *^^\ ^ 1 ^ \ • \i J^ ^1 \ i\ 1 / are given in Table III. The cause of the addition or loss of each sub- stance determined is evident in the tables. Comparisons, however, are much more easily made in figures 3, 4, 5, and 6. The graph of the dry weight is quite similar to that of the wet weight. In each case there is a marked reduction of the rate of growth of the kernels from clipped spikes in the latter half of the period of growth. Aug. 2, I9JO Development of Barley Kernels in Clipped Spikes 451 /^/T?. 0:30 V ~ 1 / \ ^ — L 1 0.70 0j60 / V ^^ V. / f-^ t y \ \ ( \ /^ , y' / v' 1 \ \ \ / / / 0^,50 0f90 O.SO 0./C? ' i \ > t 1 / — — ^'1 l\l 1 1 '\ / .^ / r i> / > ^ \ s i 22 - z^ ■^ ^ ^ oa/^ — — / ^^ 3 ^ s 6 7 a 9 /G// /^/s/1f/s/e/7/3/P£t?^/^^^s^^^s^s^7^3^9sas/ / p 3 -^ s G 7 Fig. 5. — Graph showing total nitrogen in kernels of Manchuria barley from normal and clipped spikes. //;{^'^^<^ /^e s> /o// /2 /J /e^/SAs/7/a /9£v^r ^2^3^^^s^s^7^8^sJ 1.9 05 04 7 16.68 83-32 3-61 7-09 6.4 I I 5-3 04 08 8 17.40 82.60 6.53 4-5 8 3-7 ■■• OS 9 17-51 82.49 2.68 5-55 9.2 I 6 7.6 . 04 09 10 19. II 80.89 2-79 4-30 12. 2 2 3 9.9 06 10 II 21. 04 78.96 3.02 4-44 1.5-6 3 3 12.3 10 IS 12 24.89 75-11 2-57 3.80 21.8 5 4 16. 4 14 21 13 27.48 72-52 1.77 3-44 31-8 8 7 23.1 15 30 14 28.41 71-59 I. 69 3-48 32-7 9 3 23-4 16 32 15 30.48 69.52 2. 46 3.12 35-5 10 8 24- 7 27 34 16 33-61 66.39 2.23 3- 19 38.2 12 8 25-4 29 41 17 36.26 63-74 2.14 2.98 44-5 16 I 28.4 34 48 19 35-64 64-36 2.05 2.89 43-3 15 4 27-9 32 45 20 41. 10 58.90 I. 90 3.01 42. 6 17 5 25.1 33 S3 21 42.30 57-70 1-95 2.83 45- 0 19 0 26. 0 37 54 22 44-93 55-07 2. 16 2. 50 57-5 25 8 31-7 56 65 23 44-77 55-23 2-33 2. 62 44- I 19 7 24-4 46 52 24 46.94 53-06 2. 04 2. 70 48.2 22 6 25.6 46 61 Aug. 3, 1920 Development of Barley Kernels in Clipped Spikes 453 Table III. — Average percentage and weight per kernel of dry tnatter, water, nitrogen, and ash in kernels from normal and clipped spikes of Manchuria barley at St. Paul, Minn., in igi^ — Continued CLIPPED SPIKES Date. Dry matter. Water. Nitrogen. Ash. Wet ■weight per kernel. Dry Weight per kernel. Water per kernel. Nitro- gen per kernel. Ash per kernel. Per cent. Per cent. Per cent. Per cent. Mgm. Mgm. Mgm. Mgm. Mgm. luly 26 48.61 51-39 2. 04 2. 60 47.2 22. 9 24-3 0.47 0.60 27 51-70 48.30 1.88 2.48 52.9 27-3 25-6 5.1 .68 28 52.60 47.40 2. 27 2. 76 43-6 22. 9 20. 7 52 •63 29 58.56 41.44 2. 17 2.48 42. 6 24.9 17.7 54 .62 .30 54-39 45.61 2.51 2.38 52.2 28.4 23.8 71 .68 31 56-49 43.51 2.46 2-59 44.2 25.0 19. 2 02 •65 Aug. 2 58-30 41.70 2.17 2. 07 60.6 35-3 25-3 77 •73 4 52.44 47.56 2.51 2.64 33-9 17.8 16. I 45 •47 5 64.99 35.01 2. II 2.02 45-1 29-3 15-8 02 •59 6 66. 90 33-10 3.36 I. 90 41.6 27.8 13.8 66 •53 7 81.18 18.82 2-^S 2. 12 33-6 27-3 6.3 59 •58 The deposition of ash, on the other hand, is maintained in the kernels of dipped spikes for a much longer period. It is only in the final days of maturation that the total ash per kernel of the normal spikes exceeds that of the kernels of the cHpped spikes. In Table III it will be seen that in percentage of ash the case is reversed. The kernels of the clipped spike have an appreciably higher percentage of ash. That the total is higher in the kernels of normal spikes is due to the greater weight of those kernels. In the experiment with Hannchen barley at Aberdeen several other determinations of ash were made, and a discussion of the significance of the ash content is better made after the results from that variety have been presented. The nitrogen content per kernel is shown graphically in figure 5. During more than half the period of growth there is little difference in the rate of the deposit of nitrogenous materials in the spikes. From July 23 to July 29 there is apparently a more active deposit in the normal spikes. The graphs become confused as the kernels ripen. As a whole, there is not much difference between the two. As there is a definite difference in the dry weight, the deposit of carbohydrates must be decidedly greater in the normal spikes during the last half of the growing period. The water per kernel is a good index of development. In normal development the water rapidly increases after fertilization and quickly attains its maximum. It then remains stabilized, or nearly so, as long as growth is efficiently maintained. When growth is checked or matura- tion begins, the water content drops slowly until complete ripeness occurs. After complete ripeness it drops still more rapidly for two or three days. It will be seen in figure 6 that the water content of the kernels from clipped spikes is about equal to that of the kernels from normal spikes 454 Journal of Agricultural Research Vol. XIX, No. 9 until July 25. After that date the kernels from clipped spikes exhibit a rapid loss of water which becomes accelerated about August 2. In general, the diflFerences in the development of thekernels from normal and clipped spikes are largely evident in the tables and figures. Certain observations and deductions seem justified. The discussion of the significance of the results at Minnesota, however, has been placed with that of results at Aberdeen, Idaho. Table IV. — Growth of kernels of Hannchen barley in owned and clipped spikes at A ber- deen, Idaho, in IQ16 JULY 8 Normal spikes. Clipped spikes. Weight. Length. Lateral diameter. Dorso- ventral diameter. Weight. Length. Lateral diameter. Dorso- ventral diameter. A. B. A. B. A. B. A. B. A. B. A. B. A. B A. B. Gm. 0. 0007 . 001 1 . 0012 . 0012 .OOIJ . 0014 . 0014 .0014 .0017 .0017 .0013 Gm. 0. 0007 . 0010 . 0012 .0013 . 0017 . 0017 . 0017 . 0017 .0017 . 0017 .0017 . 0014 . 0012 .0009 .0008 Mm. 2.4 2. 2 2-3 2-3 2.4 2. I 2-3 2-5 2-3 2-3 2-4 2. 2 2-4 2. 2 1.9 Mm. 1.9 1.8 2.3 2.0 2. 2 2-4 2. 2 2. I 2.4 2.0 2-5 2-3 2.4 2. I 1.8 Mm. Mm. 0.6 •9 X. 0 •9 1. 1 I. I I. 0 I. 0 1. 0 I. 2 I. 0 X. 0 I.O .8 •7 Mm. Mm. Gm. Gm. Mm. Mm. Mm. Mm. Mm. Mm. 0 I 8 6 7 9 8 8 8 9 I 9 8 6 7 6 .0008 . eoo9 0. 0019 2. I 2.6 0.7 I. 2 0. s 0 7 0. 001 5 0. oors 2. 2 2-4 a9 0-7 0. 7 0. 7 • 0025 • 0030 2-5 2.4 I. 2 1-3 -9 I 0 . 0026 . 0027 2-7 2-9 X- 2 2 » 7 .0031 .0039 2.7 3-4 1.4 IS 1.0 I 0 •0033 ■ 0032 2.6 2-7 1-4 2 0 .0039 .0045 3-1 3-S IS X.6 1. 0 I 0 .0035 .0035 2.8 3-3 1-4 4 0 .0041 .0047 3-3 3-5 1.4 I-S 1.0 I 0 .0037 .0038 3-2 3-2 1-3 4 0 9 .0048 .0047 3-7 3-7 1.4 1.4 I. 0 I I .0043 .0043 3-4 3-S I-S 3 9 9 .0049 .0051 3-7 4.0 1.6 I-S I. I I 0 .0047 . 0042 3-4 3-4 I-S 4 0 X 0 ■ OOS9 •0053 4-3 4-5 I-S 1.4 1. 0 I 0 - 0052 .0047 3- 7 3- 7 1.4 4 0 9 .0058 .0054 4-2 4.2 i-S i-S X. I I 0 .0049 .0044 3S 3-S 1-4 S 0 9 ■ 0052 .0054 3-9 4.2 1.4 i-S X. 0 I 0 . C048 .0044 4-2 3- 7 1.4 4 8 9 .0048 .0047 4.0 .V« I-S 1.4 X. 0 9 • 0045 . 0041 3-7 3-4 1-3 4 0 I 0 .0045 .0039 4.1 3-9 1.4 I-S X. 0 8 - 0036 -003s 3-0 2.8 I. 2 3 0 7 .0034 .0037 3-3 3-6 1.4 1-3 .9 I 0 • 0032 .0032 3-1 3- I 1-4 4 9 9 • 0019 • 002s .0017 2-3 2.8 1.8 I. I 1-3 I. 0 .8 8 7 . 0017 -0025 2. 2 2.8 2.8 I. 0 2 6 9 7 6 .0015 1-9 1 JULY II 0.0026 0.004S 2.6 3-8 1-4 I-S 0.8 0.9 0. 0038 0. 0014 3-4 X.8 1.4 0.8 0. 9 0.6 -004S .0068 4-1 4.8 1-4 1-7 1. 0 I. 0 . 0051 . 0042 4-2 3-7 1-4 1.6 0 -9 . 0062 .0074 4.6 S-2 I-S I-S X. 0 -9 . 0060 -OOSS 4.6 4- I x.6 1.6 0 -9 . 0069 .0081 S-i S- 7 1.6 x.6 X. 2 I. 0 .0073 .0063 S-3 4.9 I-S I-S 8 x-o -0073 . 0084 5-4 ■;-8 I-S I- 7 X. X I. 0 .0077 . 0076 S-6 S-S I- 7 1.6 0 X- 0 .0075 .0086 S-4 5- 7 I- 7 x.6 X. X I. 0 .0078 .oo8x S-5 S-4 x.8 x.6 0 1.0 .0081 .0094 ■;-8 6.^ 1-7 I- 7 I. 0 I. 0 .0081 .0080 6-3 S-6 1.6 x.6 0 I- 0 .0084 .0094 S-8 6.4 1-6 I- 7 I. I I- X . 0084 .0082 6.x 6.1 1-7 x.6 0 X- 0 .0079 . 0089 S-8 6.4 1-7 I- 7 I. 0 I. 0 .0083 . 0082 6.x 6.2 1.6 x.6 0 X. 0 .0074 .0087 ,.8 6.4 1.6 I- 7 -9 I. 0 .0083 .0082 6-S S-9 x.6 1-7 0 X. 0 .0068 .0084 S-I 6. 4 1.6 X. 8 ■9 I. 0 .0074 .0079 5-9 S-9 x.6 I-S 0 -9 .0058 .0074 4-4 5-7 I-S x.6 I- 0 X. 0 . 0069 . 0071 S- 7 SS I-S I-S 0 1.0 .0042 . 0066 4.0 S-2 I-S 1.6 -9 I. 0 .0054 .0057 S-o 4-S 1.4 1-4 9 X. 0 -0032 .ooss 3-1 4-9 1.4 1-4 •9 X. 0 .0041 .0048 41 4-3 1.4 1-3 •9 .0039 3-7 1-4 •9 • 0036 3- 7 1-3 •9 Aug. 2, J9«o Development of Barley Kernels in Clipped Spikes 455 Table IV. — Growth of kernels of Hannchen barley in owned and clipped spikes at Aber- deen, Idaho, in igi6 — Continued Normal spikes. Clipped spikes. Weight. Length. Lateral diameter. Dorso- ventral diameter. Weight. Length. Lateral diameter. Dorso- ventral diameter. A. B. A. B. A. B. A. B. A. B. A. B. A. B. A. B. 0 Gm. 0020 0044 00s 5 0074 0074 0081 ocSs 009a 0092 00Q2 0087 ooSs 0077 0068 0050 Gm. 0.0037 .006s .0073 .0086 .0090 .0096 .0098 .0098 .0099 .0098 .0097 .0085 .0028 .0062 .0044 Mm. a. 8 3-7 4-3 5-2 S-3 5-6 6.0 6.3 6.S 6.6 6.9 6.3 6.2 5-7 4-3 Mm. 3-2 4.6 4.9 6.0 S-9 6.4 6-5 6-5 6.7 7.0 6.6 6-3 S. 5-3 4- a Mm. I. 2 1.4 I- 7 1.6 1.6 1.8 1.8 1-7 1.8 1.8 I- 7 1-8 1-7 1.6 I-S Mm. 1.4 I- 5 1-7 1-7 1.8 1.8 1.8 1-7 1-7 1-7 1.8 1-7 1.4 1.4 Mm. 0.7 •9 I. 0 I. 0 I. 0 I.O I.O 1.0 I.O I. 0 I.O I.O I.O I.O I.O Mm. I. 0 I.O I.O •9 I. I I. 0 I. 0 I.O I.O 1. 1 I.O I-O I.O •9 Gm. 0.0023 .0058 .0075 .0090 .0088 .0098 .0096 .0097 .0100 .0090 .0089 .0080 .0069 • 0052 Gm. 0.0042 .0055 .0074 .0083 .0091 .0099 .0098 . 0102 .0100 .0104 .0097 .0082 .0073 .0062 .0050 Mm. 2.6 4.6 5-6 6-S 6.7 6. 9 6.9 7-3 7-3 7.0 6.9 6.4 S-8 4.9 Mm. 3-4 4-S 5-3 S-7 6.6 6.9 6.9 71 6.9 7-4 71 6.1 S-S 5-5 4-7 Mm. I. 2 l-S 1-7 1.6 1-7 1-7 1.8 1-7 1-7 1.8 1-7 1.6 1.6 1.4 Mm. i-S 1-7 1.6 1-7 1-7 1-7 1-7 1.8 1.8 1-7 I-S 1.6 1.6 I-S 1-5 Mm. 0.8 •9 •9 •9 •9 -9 I.O I.O I.O I.O I.O I-O .8 ■ I Mm,. 0.8 I-O •9 I. I I-O I.O I.O 1. 1 1. 1 I- 1 I- 1 I-O •9 .8 -8 0. 0102 0.0071 6.8 S-4 1-9 1-7 I- 1 I-O 0.0037 0.0064 4.8 5-3 1.6 I-S 0.7 0.9 .0130 .0114 8.2 7-S 1-9 1.8 I- 2 I- I • 0103 .0094 7-2 6.6 1-8 1-7 I-O I-O .0146 .0121 8.6 7-S 1-9 1.8 I. 2 1. 1 .0140 .0116 8-3 7-S 1-9 1-7 1-3 1-2 -0153 -OI4S 9-2 8-5 2.0 1-9 1.4 I. 2 -0155 .0125 8.8 8-3 2.0 1-7 1-4 II . 0160 .0146 9-3 8.9 2. 0 1-9 1-4 I- 2 .0169 .0141 9-2 8.S 2.0 1-9 I-S 1-2 -0I7S ■OIS7 9.6 8.8 2. 2 1-9 I-S 1-4 .0184 .0145 9-S 8-8 2. I 2.0 I-S 1-4 .0187 .0:68 10. 0 9-1 2. 2 2.0 1-4 1-4 .0184 • 0144 9-4 8-7 2. I 1.8 I-S 1-3 .0188 . 0162 10. 0 9-3 2- I 2- 2 1-4 1-4 .0179 .0149 90 8.8 2.0 1-9 1-4 1-3 .0184 . 0172 >9 9-6 2. 2 2. 0 I-S I-S .0178 .OI4S 9-2 8.4 2. 2 1-9 1-5 1-4 •0173 .0167 9-S 9-3 2. 2 2.0 1-4 1-4 .0196 .0147 90 8-s 2-3 1-9 I-S I. 2 -0175 .•0149 9.6 9.1 2. I 1-9 I-S 1-4 .0160 .0149 9-0 8..; 2.0 19 1-4 1-3 -0158 • 0137 9-3 8.7 2-1 1.9 1-4 1-4 .0138 •0137 8.7 8-S 1-9 1-9 1.4 1-3 -013s .0129 8-6 8.S 1-9 1.8 1-3 I- 2 .0126 .0127 8-S 8-3 1.9 1-9 I. 2 1-3 . 0126 • 0103 «-S 7-3 1-9 1-7 1-4 r. I .0086 .0106 7.0 7-7 1.6 1-7 I. I 1.2 . 0105 .0085 7-7 6.9 2.0 I-S I. I I.O .0082 6.7 1.6 .8 .0076 6.3 1-7 -9 JULY 14 0.0069 0. 0104 5-3 7-3 1-7 1.9 0.9 I. I 0. 0094 0.0098 7-3 7-3 1-7 1.8 I.O I. 2 .0117 -0153 7-6 8.4 1.8 2 I I. I 1-3 •0139 .0108 8.6 9-1 1.8 2. 1 1-4 1.4 - 0I5I .0161 8.7 9.0 2. I 2 I 1-3 1-4 . 0167 .0200 9-3 9-4 1.9 2. 2 1.6 I-S .0154 .0179 9.0 9- I 2. 0 2 I 1-3 I- 5 •0173 .0207 9-6 9.8 2.0 2.0 l-S I-S .0177 .0196 9-4 9-4 2. I 2 3 1-4 1-6 .0191 . 0220 9.6 9-9 2.1 2. 2 1-5 1.6 .0182 .0199 9-S 9-6 2. 2 2 I 1-4 1.6 .0189 .0227 9-7 10. 2 2.0 2-3 1.6 1.6 .oiSo -0203 9-7 9-7 2. 0 2 I 1.4 1-5 . 0196 .0244 9.9 10.4 2-3 2-3 I-S 1-7 .0185 .0208 9.4 9-5 2. I 2 I I-S 1-7 .0195 .0250 9-9 10.5 2-3 2-4 I-S 1-7 .0186 .0208 10. 0 9.9 2. 2 I 9 I-S 1-6 .0194 .0247 9-6 10.4 2. 2 2-5 1.6 1.6 .0171 .0209 9-4 10. 0 2. I 2 3 I-S I-S .0193 .0242 9-6 9.9 2-3 2-5 1-5 1-7 .0159 .0192 9-S 9-6 1.9 2 3 1-4 1.6 .0164 -0236 9.4 10.3 1-9 2.6 1-4 1.6 .0134 .0179 8.8 9-3 1-9 2 2 1-3 I-S .0146 .0222 8-8 10. 0 1-9 2-3 1-3 1-7 . Olio -0152 8.0 8-7 1-7 2 0 I. I 1-4 -0138 • 0205 8.9 lo- 0 1-8 2-3 1-4 I-S .0094 .0128 7-S 8-5 1.8 I 9 I. I I. 2 .0100 .0171 7-8 9- 5 1-7 2. I I. 2 1.4 . 0090 7-3 I 6 I. 2 .0143 8-S 2. 0 1-4 456 Journal of Agricultural Research Vol. XIX. No. 9 Table IV. — Growth of kernels of Hannchen barley in owned and clipped spikes at Aber- deen, Idaho, in igi6 — Continued Normal spikes. Clipped spikes. Weight. Length. Lateral diameter. Dorso ventral diameter. Weight. Length. Lateral diameter. Dorso- ventral diameter. A. B. A. B. A. B. A. B. A. B. A. B. A. B. A. B. Gm. Gtn. Mm. Mm. Mm. Mm. Mm. Mm. Gvt. Gm. Mm. Mm. Mm. Mm. Mm. Mm. 0.0082 0.004s 6. 4 5-0 1.8 I-S 0. 9 0.7 0. 0041 0. 0058 4.8 5-5 1.4 I-S 0.7 0.8 .0158 . 0017 «-7 8 8 1.9 2. 1 1-3 4 .0144 .0145 8.8 8.8 1.9 1.9 4 3 .0193 .0213 9.6 9 5 2. 1 2. 2 1.4 6 .0207 .0144 9-S 9.0 2-3 2. 2 S S .0215 .0248 10. 0 9 8 2. 1 2. I i-S 6 .0251 .0212 10. 1 9-3 2-3 2. 2 8 S .0230 . 0261 10. 0 10 2 2-3 2-3 1.6 6 •02S3 .0215 10. I 9-S 2. 2 2-3 7 S .0261 •026s 10. 4 10 0 2-3 2. 0 1.4 7 -0263 .0230 10. 0 9-S 2.4 2.4 8 s .025S .0284 10. I 10 4 2.4 2.4 1.6 9 .0264 .0246 10. 0 9-7 2-S 2-3 7 8 •02S3 .0282 10. I 10 4 2-3 2.6 1.6 8 •0275 -0243 lo- 0 9.6 2-S 2.4 7 6 .0259 .0269 10. 0 10 0 2-5 2-5 1.6 7 .0267 .0248 10. 0 10. 0 2.6 2-3 9 7 .0250 .0266 10.3 10 0 2. 2 2.0 1-7 6 .0260 •0235 9-9 9-3 2-S 2-3 7 S .0229 •0253 10. 0 10 0 2. 0 3. 2 1-5 S -0250 .0210 9-7 9-3 2. 2 2.4 S S .0212 •0231 9-5 9 7 2.0 2. 2 I-S 8 • 0237 .0180 9-S 9-3 2. 2 2.0 5 5 .0183 .0213 9-S 9 0 2.0 2.0 I-S 7 .0206 .0134 9-1 8-S 2. 2 1.9 S 4 .0172 .0169 9-3 9 0 2. 0 2. I I-S 5 • 017s 9-2 2.0 S .0141 •013 s 8-7 8.2 1.6 1.8 1.4 1-3 .0119 7-7 1.8 1-3 ). 0167 0. 0293 9.2 10.3 2-3 2-7 1-4 2. 0 0.0223 0.022s 9-S 8-3 3.6 1.9 I-S ■0334 -0323 10. 2 10. S 2.8 2-9 2. 2 2. 2 .0340 .0254 10.4 ro. 0 3-0 2-S 2.0 .0370 -0363 10.6 10.6 3-9 3-0 2. 1 2. I .0366 .0298 10. 0 10. 2 3-0 2.8 2.0 . 041 1 .0396 10.9 II. 2 3-1 3-0 2-3 2. I .0398 •0339 10.4 10. 0 3-2 3-0 2. 2 . 0404 . 0400 II. I 10. 6 3-1 3-1 2-3 2. 2 .0384 •03SI 10. 0 10. 0 3-2 3-0 2. 2 -0436 .0413 11- 0 10. 2 3-3 3-0 2. 2 2. I .0388 .0368 9.8 10. s 3-2 3-2 2.0 .0436 .0423 II. 2 10.3 3-2 3-3 2-3 3. 2 •038s •0362 9-7 9.2 3-2 3-2 2. I .0442 .0406 II. I 10. s 3-2 3-2 2-3 2. 2 •0393 •0344 9-8 9-4 3-2 3-0 2. 2 .0447 .0413 10.6 10-2 3-4 3-2 2.3 2. 2 •038s •0349 9-7 9-8 3-2 3-0 2.2 .0422 .0392 10.5 10.3 3-3 3-0 2.4 2.0 .0384 -03IS 9-8 9-S 3-4 3-0 2-3 .0405 .0392 10.6 10. 0 3-3 3-1 2-3 3.0 •0339 .0299 9-7 9-1 3-0 3-0 2. I .0387 -037s 10.4 9.8 3-3 3-0 2.2 3. I .0316 .0277 8.9 8.9 3-1 2.9 •2.0 .0365 -0343 10. 0 9-S 3-1 3-0 2.2 2. 2 .0298 .0245 9/0 8.6 2.8 2.7 1-9 .0314 •0320 9-7 9.8 2.8 2-9 2.0 2. I •025s .0166 9- I 7-9 2-S 2-S 1.6 .0260 .0287 9-S 9.0 2.6 2.8 1.8 2.0 .0201 8.2 2-S 1.6 1.4 1.8 1.8 2. I 2.0 2. I 2. I 2. 2 2.0 1-9 2.0 1.9 1.9 1.6 JULY 18 0. 02S3 0.0279 9-3 9-3 2-7 2.9 1.6 1.8 0. 0283 0. OI9I 10. 0 8.7 2.8 2-3 1.8 1.6 .0390 •0395 10. s 10.7 3-2 3-0 3.1 3. 2 •0326 •0317 10.3 10.2 2.8 2-7 1.8 1-7 .0405 -0450 10.4 10.3 3-3 3-4 3. I 3. 2 .0368 .0369 10.0 10.7 3-1 3-0 2. 1 2.0 .0450 . 0466 lo. 2 10. 0 3-4 3-3 3.2 2-3 .0386 . 0410 10.3 10.4 3-2 3-3 2. 0 2. 2 -0439 .0478 II. 0 10. s 3-2 3-4 2.2 2-4 .0417 •0430 10. I 10. 2 3-1 3-2 2-3 2. 2 -04SS .0508 10.7 10.4 3-3 3-6 2-3 2-S .0423 -0434 10. 2 10. 1 3-0 3-3 2. 2 2. 3 .0465 -0504 10. I 10.7 3-S 3-5 2.2 2.6 .0418 .0427 10. 0 9-3 3-2 3-3 2. 2 2. 2 .0458 .0500 10. s 10.8 3-4 3-6 2-3 2.4 .0408 .O.HI 9.6 10. 2 3-3 3-2 2. 2 2. 2 -0439 .0431 9-S 10. 2 3-4 3-6 2.0 2.4 .0388 •0454 9.8 10. 2 3-2 3-3 2- 2 2-S .0430 .0468 9-7 10. 1 3-4 3-4 2. I 2.4 .0390 .0415 9-7 9-7 3-2 3-3 2-3 2-4 .0419 .0431 10. 1 9-5 3-3 3-5 2-3 2.4 •0334 .0402 9.2 10. I 3-2 3-2 2.0 2. I -0383 .0416 9-7 9-1 3-3 3-4 2-3 2-4 •0305 .0387 9.2 9.6 3-0 3-1 2.0 2-3 -0324 .0388 9-3 9-3 2.8 3-3 1-9 2. 2 .0271 ■03S3 8-7 9-7 2.8 3-2 1-7 2. 2 •0311 .0350 8.6 9.2 2.9 3-1 2.0 2. 2 -0197 -0330 .0279 7.8 9.2 8.4 2-S 3-0 2.8 I- 3 2. I 1.9 Aug. 2. 1920 Development of Barley Kernels in Clipped Spikes 457 Table IV. — Growth of kernels of Hannchen barley in awjied and clipped spikes at Aber- deen, Idaho, in igi6 — Continued Normal spikes. Clipped spikes. Weight. Length. Lateral diameter. Dorso- ventral diameter. Weight. Length. Lateral diameter. Dorso- ventral diameter. A. B. A. B. A. B. A. B. A. B. A. B. A. B. A. B. Gm. 0. 0205 •0391 • 0396 •044s •0457 •0473 .0465 .0462 .0467 .0405 .0448 .0400 • 0381 .0356 .0278 Gm. 0-0337 .0427 .0475 • 0511 .0481 .0489 .0486 .0488 .0483 •0459 •0457 •0438 •037s •034s Mm. 9-2 10. 1 10. 4 9.9 10. 4 10.3 10.3 9.6 10. 2 10. 2 9-7 9-6 9.6 9-2 9.0 Mm. 9- 7 10. 6 10.5 10.8 10.7 10. 5 10. 4 10.8 10. 0 10. 0 9-7 10. 1 9-2 9-6 Mm. 2-S 3-2 3-3 3-3 3-4 3-4 3-4 3-5 3-4 3-4 3-5 3-2 3-2 3-3 i-o Mm. 3-1 3-3 3-4 3-5 3-7 3-5 3-6 3-S 3-5 3-4 3-4 3-4 3-3 3-2 Mm. 1-7 2. 1 2. 1 2-3 2-3 2-3 2-S 2-5 2-4 2- 2 2-3 2-4 2-3 2-3 2- 2 Mm. 2. I 2. 2 2-S 2-3 2-S 2-3 2-S 2-4 2-S 2- 2 2-S 2-3 2-3 2. 1 Gm. 0-OI37 .0324 .0380 .0408 -0397 .0429 .0431 .0409 -0382 .0391 •0373 •0343 • 0311 .0226 Gm. 0.02S3 -0374 .0409 -0422 .0441 .0425 •0435 •0443 -0434 .0402 .0388 •0363 .0328 .0264 .019S Mm. 7-5 9.8 10. I 10. 1 10. 1 10. 4 9.6 9-6 9-7 9.6 9.2 8.7 8.9 8.2 Mm. 9-S 9-S 10. 2 10. 4 10. 2 10.4 9-7 9-7 9.6 9.9 9. 6 9.6 8.6 8.7 8.0 Mm. 2- 2 2-8 3-2 3-4 3-4 3-3 3-3 3-2 3-3 3-4 3-4 3-2 30 2.8 Mm. 2-9 3-2 3-3 3-3 3-4 3-3 3-4 3' 4 3-4 3-3 3-4 3-3 3-2 3-0 2.6 Mm. i-S 2- 0 2. 2 2-3 2.3 2.4 2-4 2-4 2- 2 2-4 2. I 2. I 2. I 1.8 Mm. 1-7 2.0 2-3 2. 1 2-3 2- I 2. 3 2-3 2.4 2. 2 2- 2 2-1 2. 3 1-9 1-6 O.OII7 0. 0356 7-6 10.7 2. 1 3-1 1-3 2. 2 0- 0395 0. 0359 10.8 9.8 3-2 3-1 2-0 2-3 .0419 .0421 10.3 II- 0 3-3 3-4 2-3 2-3 •0435 .0407 10.4 10. 2 3-3 3-3 2-3 2-3 .0477 .0422 10.3 10.2 3-6 3-2 2-S 2. 2 •031S •0437 9-9 10. s 3-0 3-2 x-6 2-4 .049a .0447 10.6 II. 0 3-0 3-3 2-S 2-3 .0510 .0442 10.8 10-8 3-5 3-1 2-S 2-3 .0521 .0467 10.7 II. 0 3-8 3-2 2-4 2-3 .0478 .0460 10. 4 10.4 3-5 3-4 2- I 2-4 .0500 •04SS 10.8 10.9 3-6 3-3 2.4 2-3 .0491 .0456 10.5 10-4 3-S 3-3 2-S 2- I .0505 .0440 10.7 10. 6 3-7 3-2 2-4 2- 2 •0433 .0450 10-4 10. 1 3-4 3-3 2-3 2-5 .0514 .0458 10-4 10.3 3-6 3-3 2-S 2-S .0470 .0430 10- 2 10. 1 3-S 3-3 2- 2 2-2 .0492 .0444 10. 0 10. s 3-S 3-3 2-S 2-3 .0442 .0400 10.2 9-3 3-5 3-0 2.4 2- I .0493 •0443 10.4 10. 0 3-t> 3-3 2.S 2-3 .0410 -0403 10-2 9.8 3-4 3-0 2- 2 2.0 •047s .0419 10. 0 10.4 3-S 3-3 2.4 2-3 .0392 -03S4 9-9 9-7 3-2 2-9 2-3 2- I •04s 5 . 0409 10- 0 9-7 3-4 3-2 2.4 2-2 .0363 •0332 9-8 9-7 3-2 3-1 2- 2 2-! .0411 •0363 9-7 9-7 3-S 3-2 2-3 2-0 -0320 .0293 9-3 8.9 3-2 2-9 2-0 2.0 .0382 •0331 9-0 9.8 3-3 2.S 2-3 2.0 •0247 .0199 9.0 8-5 2.9 2.7 1.8 1-7 .0326 • 0280 9-3 9.1 2-8 2.8 3.0 2.0 July 21 o- 0200 0- 0228 8.9 9-3 3-4 3-6 1-7 1-8 0.0257 0- 0355 9-4 9.8 3-8 3 2 1-9 1-9 .0446 .0420 10. 6 10-3 3-4 3-2 3 2 3- 3 .0427 •0433 10. 0 10. 2 3-5 3-S 2-4 2.3 .0522 .0492 10. 7 II- 0 3-S 3-5 3 3 2-3 . 0462 . 041 1 10.6 10.4 3-4 3-5 3-3 2- 3 • 0538 .0505 10.8 II- 0 3-7 3-6 3 4 3- 3 .0465 .0488 10. 2 10-5 3-3 3-S 2-3 2-3 .0596 .0527 10.6 10. 6 3-9 .3-6 2 4 2-5 . 0508 .0489 10-3 10.8 3-5 3-7 2-S 2-3 .0487 •0505 10-7 10. 6 3-7 3-4 3 4 2-5 .0492 • 0473 9-4 10- I 3-7 3-5 3-4 2-3 .0342 .0517 10.7 II. 0 .3-8 .3-6 3 S 2-5 • 0477 • 0463 10- 0 9-8 3-5 3-S 2-4 2-3 ■0555 •0517 10. 7 10.6 3-8 3-4 3 6 3-4 • 0456 .0461 10.3 10.3 3-5 3-5 2-3 2-3 •0555 .0514 10.8 10.3 3-8 3-7 2 5 2-5 • 0454 • 0455 10.0 9-7 3-5 3-4 2-3 2- 3 .0509 -0490 10. 4 10.3 .3-8 3-5 2 5 3-4 ■ 0439 .0429 9-5 9-9 3-S 3-S 3-3 2-4 -0489 .0481 10.3 10- s 3-7 3-5 2 6 2-4 .0416 . 0416 10. 0 9.8 3-5 3-1 3. 1 2-3 •047s .0442 10. I 10-3 3-7 3-5 2 5 3-4 • 0387 -0419 9-0 9-5 3- I 3-3 3. 1 2.3 -0434 -0412 10. I 9.0 3-0 3-4 3 4 2-3 -0359 •0356 9.0 9-3 3-2 3-3 2-3 2-1 • 0412 -0373 10- 0 9-7 3-4 3-3 2 2 2. 2 • 0333 ■0350 9. I 9-0 3-2 3-2 2. 2 2.0 •0339 .0289 95 9. I 3-2 3- I i-3 2. 0 ■ 0212 8.4 2- 7 1.9 458 Journal of Agricultural Research Vol. XIX, No. 9 Table IV. — Growth of kernels of Hannchen barley in owned and clipped spikes at Aber- deen, Idaho, in igi6 — Continued July 22 Normal s pikes. Clipped spikes. Weight. Length. Lateral diameter. Dorso- ventral diameter. Weight. Length. Lateral diameter. Dorso- ventral diameter. A. B. A. B. A. B. A. B. A. B. A. B. A. B. A. B. Gm. Gm. Mm. Mm. Mm. Mm. Mm. Mm,. Gin. Gm. Mm. Mm. Mm. Mm. Mm. Mm. o. 0300 0. 0164 q.6 8.0 2.8 2-5 1.8 1.8 0. 0203 0. 0238 8.2 9.0 3-S 2-5 i-S 1-7 .0558 •0433 10 8 10. 5 ,3- 8 3-S 2.6 2.6 •0437 .0382 10. 0 10. 0 3-5 3-2 3. 0 1-9 .0580 •0493 10 9 10.4 3-8 i-i 2.6 2.6 ■0544 .0464 10. 2 9.8 3-7 3-4 2-3 2. 1 •0593 • 0526 10 8 10.7 3-f) 3-6 2.4 2.4 •OS7S •0473 10.4 10. I 3-6 3-S 2-3 2- I . 0600 .0494 10 8 9.8 4.0 3-4 2.6 2.6 •057S .0524 10.8 10. s 3-7 3-5 2-S 2-S .0570 .0519 10 ■; 10. 7 3- 7 .30 2-3 2-3 • 0550 .0503 10.4 9S 3-8 3-S 2-5 2-3 • 0595 ■0539 10 s lo. 0 3-9 2% 2.6 2.6 •0563 .0546 10. 0 10. 0 3-9 3-7 2-5 2-4 •0599 •0493 10 7 10. 0 .3-8 3-6 3-7 2-7 •0557 • 0523 10. 0 10. 4 3-6 3-6 2-S 2-S . 0605 • 0516 10 s 9- 7 3-9 35 2-7 2.7 • 0557 • 0514 9-5 9- 7 3-6 3-7 2-S 2-S ■ 0583 .0S04 10 5 10.5 3-9 .3- 6 2.8 2.8 •0495 • 0503 9.9 10. 0 3-6 3-S 2-3 2-S .0566 .0487 10 S 10. I .3.8 Z-'i 2.6 2.6 .0478 .0483 9-5 10. 2 35 3-S 2.3 2-S •0538 .0486 10 0 10. 0 3-5 3-S 2.4 2.4 . 0460 . 0462 10. 0 9.0 3-4 3-5 2. I 2-3 .0505 .0464 10 0 9.8 .3-6 3-6 ■ 2.4 2.4 .0432 . 0428 9.0 9.4 3-7 3-5 2-3 2. 3 • 0479 .0413 9 2 9-5 .3-6 3-2 2-5 2-5 .0388 .0405 9-3 9. I 30 3-4 2. I 2. 2 .0419 •0385 9.0 9. 0 3-5 3-1 2.4 2.4 .0281 • 0325 9-7 8.3 2.8 3-1 1-9 I. 0 0. 0491 0.0513 10.3 10. 2 3- 7 3-6 2.4 2-S 0. 0294 0. 0489 9-3 9-6 3-0 3-6 1-7 2-4 0591 •OS77 10.5 10.7 3- 7 .3-6 2.6 2-S -0524 .0561 10. s 10.6 3-7 3-7 2-3 2-S 0596 .0606 10.5 10.7 3-9 ,3-8 2- 7 2-7 .0576 -0534 10.5 10. s 3-9 3-7 2-4 2-3 0633 •0593 10. 5 10.8 3-9 .3-8 2-9 2-S .0588 •059s 10. 6 10.7 3-9 3-8 2-S 2-S 0607 . 0629 10.5 10.8 3-9 3-9 2-S 2-7 .0632 .0587 10.7 10.5 4.0 3-7 a. 6 2.4 064s .0658 10.5 10.8 4-2 4.0 2.6 2.8 -058s .0562 10. 5 10.3 3-9 3-7 2-6 2.6 0625 .0659 10.7 IC.4 .3-8 4.0 2-7 2-8 .0602 .0612 10.3 10. I 3-9 3-8 2-7 2-7 0580 .0690 10.3 10.8 3-9 4.1 2.6 2-9 •OS5S -057s 10. 4 10. 0 3-7 3-7 2-S 2.8 0605 .0648 10. s 10. 6 .3-8 4-0 2.6 2-7 • 05IS •0557 10. 0 9-9 3-S 3-7 2-S 2-S 0608 • 0634 10. s 10.4 ,3-8 .3-8 2.7 2.8 -0544 -OS44 10. 2 9.6 3-6 3-7 2-5 2.6 0531 •0532 10. 2 10.4 .3-8 3-7 2-S 2-7 .0528 .0508 9.8 9-1 3-6 3-6 2-4 2-S 0534 •0559 10. 2 10. 2 3-7 3-7 2-S 2-S .0490 . 0440 9-S 9-7 3-6 3-S 2-3 2-4 0470 ■OS54 9-S 9-6 3-S 3-7 2.4 2.6 •04S3 -0453 9.6 8.8 3-4 3-4 2-4 2-S 0454 ■0527 .0450 .0388 9-1 10.0 9-3 8-7 3-4 3-6 3-4 3-3 2-S 2-7 2.S 2-3 •0373 9.1 3-t .'■'.. 1 0.0300 0. 0552 8.7 10.4 2-9 3-7 1-9 2.4 0- 0480 0.023 s 10. 0 8-S 3-6 2.6 2-4 I-S -0583 • 0624 10. 4 10.9 3-7 3-9 2. <; 2- 7 .0600 .0465 10.8 9-S 3-7 3-6 2-S 2-3 .0605 .0603 10. 9 10. s ,3-8 3-7 2.6 2.6 .0624 -048s II. 0 10. 0 3-8 3-S 2-7 a-S .0650 •0633 II. 0 10.7 3-9 4.0 2- 7 2.6 . 0604 -0517 10.7 9.4 3-6 3-7 2.6 2-3 .0608 .0636 10.3 II. 0 ,3-8 3-9 2- 7 2.6 .0660 -0531 10.6 10. 5 4.0 3-6 2-7 2.4 . 0650 .0619 10.8 10.5 ,3-8 3-7 2-7 2-S -0653 -0528 II. 0 9-S 3-9 3-7 2-7 «-3 .0639 • 0613 10. 6 10.8 3-7 .3-8 2.8 2-7 -063s - 0502 II. 0 10. 2 3-6 3-7 2-S «-4 .0628 .0587 10. 6 10.3 ,3-8 3-7 2.6 2.6 .0634 • 0515 10,8 10. 2 3-7 3-4 2-3 2.4 . 0629 .0608 10.4 10.6 .3-8 .3-8 2.8 2-7 .0633 -05IS 10. 0 9.9 3-6 3-4 2.8 2.4 ■ 0615 -0590 10- 4 10.3 4.0 .3-8 2-5 2-7 • 0593 .0500 10.3 10. 0 3-9 3-7 2.6 2-4 •OS 77 -0573 10. 2 10.3 .3-6 .3-8 2-7 2-7 - 0585 -0483 10. s 9.6 3-S 3-6 2-S 2-S .0568 .0569 10. I 9-7 .3-8 3-S 2.6 2.6 .0578 . 0482 10. 2 9-5 3-7 3-4 2-7 2-4 • 0548 -oss6 9-7 9-6 3-7 3-7 2.6 2-S . 0569 .0430 10. 0 9-3 3-6 3-0 2-6 2. 3 • 0526 .0500 9-7 9-S 3-4 3-S 2.8 2.6 ■ 0483 .0418 9-7 9-3 3-5 3-0 2-S 2-3 -047s - 0392' . 0461 .0364 9.0 9.0 9. 2 9.0 3-6 3-3 3-1 3-3 2-5 2-3 2.4 . 0400 • 0335 9-0 8.6 3-1 3-1 2. I 2.0 Aug. 2, 1920 Development of Barley Kernels in Clipped Spikes 459 Table IV. — Growth of kernels of Hannchen barley in awned and clipped spikes at Aber- deen, Idaho, in iqi6 — Continued JULY 26 Normal spikes. Clipped spikes. Weight. Length. Lateral diameter. Dorso- ventral diameter. Weight. Length. Lateral diameter. Dorso- ventral diameter. A. B. A. B. A. B. A. B. A. B. A. B. A. B. A. B. 0 Gm. 0304 0569 0619 0660 0708 0669 0650 0696 068 1 0654 0640 0607 0607 0546 050a Gm. 0. 0382 .0588 .0640 .0700 .0659 .0672 .0673 .0671 .0645 • 0634 .0626 •OSS 7 .0540 .048s Mm. 9.2 9-7 9-7 10.7 10- 7 10.7 10- 2 10.8 10. 2 10.4 10.4 10- 0 lO-O 9-S 9-S Mm. 9- a 10.8 10.8 10.8 10. s 10. 9 10. 7 11. 0 10.3 10.3 lo. 4 9-7 9.6 8.9 Mm. 3-0 3-5 3-9 3-8 4.0 3-8 3-7 3-9 3-8 3-7 3-8 3-9 3-7 3-S 3-S Mm. 3-3 3-8 i-S 3-9 4.0 4-0 3-9 3-9 3-8 3-6 3-8 3-7 3-6 3-S Mm. 1.8 a. 6 a. 4 3.7 2.8 3.8 3.6 3.7 3.6 3.6 3.7 9-S a-s a-S »-S Mm. a. 0 2-5 3-S 3.6 3.6 2.8 3.8 3.6 3.6 2-S 3.6 a- 5 3.6 3.6 Gm. 0. 0394 -OS43 .0614 .0670 .0655 .0658 • 0651 .066s .0630 •o6s3 .0602 .os88 •OS05 .04s6 .0333 Gm. 0. 0368 .0614 .0642 . 0669 .0652 .0624 .o6s7 • 0643 .0600 .0600 .os68 •0530 .0486 •03S3 Mm. 9.0 10.8 10. s 10. 7 10. 6 10. s 10.7 10. s 10.3 10.3 10. 0 10.0 9-S 9-0 8.0 Mm. 9.0 lO-O 10.5 10.8 10. s 10. s 10. s 10. 3 10.0 lO-O 10.0 9-S 9-3 8.7 Mm. 3-3 3-7 3-7 3-8 3-7 3-7 3-7 3-9 3-8 3-7 3-8 3-6 3-* 3-S 3-3 Mm. 3-0 3-5 3-S 3-7 3-7 3-8 3-7 3-9 3-7 3-8 3-4 3-S 3-S 3- a Mm. 3. 0 a- 3 3.7 2-S 2.6 3.8 3.8 3.8 3.6 a- 7 a- 7 a- 7 a-S 2-3 3.0 Mm. 1.8 3-S 2.7 2.7 2-S 2-S 2.8 2-7 2-S 2.6 2-S 2-S 2-S 1-9 0. OS4I 0.0427 10. 0 9-3 3-S 3-0 3-S 3. I 0. 0275 0-0357 8-S 9-S 2-S 3-0 2. 0 2.0 -0538 .0614 10.3 10. 3 3-7 3-2 3-S 3.4 • 0458 .0618 9.8 10. 0 3-S 3-9 2- 2 2.6 . 0604 •0653 10. s 10. s 3-5 .V6 2-S 2-S • 0538 .0663 lO-O 10. S 3-S 3-6 2-S 2-8 . 0620 .0685 10. s 10- s .V8 3-8 3.6 2-7 .0581 . 0622 10. 1 10.0 3-7 .3-8 2-3 2-7 . 0609 -065s lO- 0 10. s 3-» 3-» 2-S 2-4 • 0583 .o6s6 10. 2 10-3 3-S 3-8 2-S 2.8 .0592 . 0664 lO- 0 10. 2 3-6 3-8 2-S 2.6 • 0552 .0637 10. 2 10.3 3-2 3-9 2-S 2-7 • 0583 .o6s6 q.6 lo- 4 .V6 3-8 3.6 3.7 •0S67 .0656 9.4 9-6 3-4 3-7 a- 4 2.6 . 0600 . 0623 9.6 lo- 0 3-6 3-7 3-S 2.6 •0579 .0615 9.8 10. 3 3-6 3-7 a. 7 2-7 -0554 . 0629 9-3 10. 0 3-S 3-5 »-5 2-S •0S38 .0619 9.8 9.8 3-4 3-7 3.6 2.6 • 0562 .0627 9.6 9-S 3-7 3-9 3.6 3.6 • 0S16 •OS 79 9.8 ICO 3-4 3-6 2-3 2-7 • 0530 .0603 9.2 9-0 3-S 3-0 a-s 3.7 .0487 • OS64 9.0 9-2 3-S 3-4 2-S 2-3 .0488 • 0589 9.2 9-S 3-4 3-7 2-3 2-7 .0486 .OS18 9.6 9.0 3-S 3-3 3.6 2-4 -04s 7 -OS4S 9.0 9.3 3-2 3-S 2-3 2-4 . 0416 .0471 9-0 9-S 3-3 3-S 3. 2 2-4 -047s 8.8 3-3 a. 4 .0224 •0343 8-3 9-2 3.7 3-2 1.8 3. I JULY 28 0- 0345 0.0190 8-S 7-3 3-2 2-S 1-8 1.6 0.0362 0.0233 9-1 7-7 3-4 a- 7 a. I t.8 OS89 -OS 74 10- s 10.0 3-6 3-S 2-5 2.7 .0530 •OS52 9-8 lO-O 3-7 3-4 2-3 2.8 0640 .0637 10. 8 lo- 0 4-0 3-8 3.6 3.7 . 0602 .0619 10- 0 10. s 3-8 .1-8 2.7 2-5 0652 .0631 II. 0 10.6 .3-8 .1-8 2.6 3.8 . 0614 .0603 10-0 10. s 3-7 3-7 2-S 3.6 0688 .0671 lo- 7 10- 0 4.0 4.0 2.7 2.8 .0642 • 0632 10. 7 10. s .1-8 3-7 2-7 2-S 0667 .0658 10. 6 II- 0 3-8 3-8 3.8 2-7 • 0625 .0638 10. 0 10.0 3-8 3-7 2-7 3.8 0657 .0688 10. 0 10- s 3-8 3-7 3.9 3.8 .0619 . 0660 9-9 10. 0 3-8 3-S 3.8 3.8 0625 .0683 10.5 XO. 2 3-8 4-0 2- 7 3.8 . 0604 . o6i6 9.8 10. 2 3-8 3-7 3.8 2-7 0608 • 0695 9.6 10.4 3-8 4.0 2.7 3.9 .0585 .0600 10. 0 lo- 0 3-7 3-S 3.8 2.7 0618 .0657 10. 0 10. 0 3-7 3-9 2.6 3.8 .0578 • 058s 9-3 10. 3 3-8 3-7 2-5 3.6 0578 .0620 9.8 9-S 3-S 3-7 3-7 3.7 .0560 .0360 9-7 10. I 3-7 3-7 2-7 3.6 0542 .0602 10. 0 10. 0 3-7 .1-8 3.6 3.8 .0512 .0372 9-0 10.0 3-S 3-7 2-3 3.8 0S02 •057s 9- 7 9 7 3-7 3-7 2-S 3.6 .0504 • 0528 9.0 9-4 3-4 3-S 3-4 3.6 0452 . o';o2 9.2 9.4 3-5 ,1-6 2-S 2.6 .0441 .0483 8.7 90 3-4 3-3 2.4 3.6 .0489 9.2 3-6 2-S • 0279 .0427 8-0 8.S 2-8 3-4 1-9 2-S 460 Journal of Agricultural Research Vol. XIX, No. 9 Table IV. — Growth of kernels of Hannchen barley in owned and clipped spikes at Aber- deen, Idaho, in igi6 — Continued Normal spikes. Clipped spikes. Weight. Length. Lateral diameter. Doiso- ventral diameter. Weight. Length. Lateral diameter; Dorso- ventral diameter. A. B. A. B. A. B. A. B. A. B. A. B. A. B. A. B. Gm. o- 0360 .0686 .0727 .0806 • 0750 .0785 •0752 .0765 .0764 •0713 .0724 . 0650 .0634 .0580 •0537 Gm. 0. 04S4 .0685 •0739 .0719 •0739 .0683 . 0692 .0667 .0648 .0641 •0574 •0597 .0540 • 0460 Mm. 9.0 10.3 II. 0 II. 4 lo. 4 10.8 10. 0 10. I 10. I lo- 6 10. 0 9.2 lO-O 9-S 9-7 Mm. 9.2 10.5 lo- 0 10. 7 10. 0 11. 0 10. 0 10. 0 10. 0 10. 2 9.6 9-S 8-7 8-S Mm. 3-3 3-9 4.0 4.2 4.1 4.2 4.2 4.0 4.1 3-8 4.0 4.0 3-8 3-8 3-6 Mvi. 3-i 3-7 3-9 4.0 4.0 3-8 4.0 3-9 3-8 3-8 3-7 3-7 3-7 3-S Mm.. 2. 1 2.7 2.8 2.9 2.8 3-0 2.8 2.9 2.9 2.8 2.8 2.8 2.8 2.7 2.S Mm.. 2-3 2.7 2.8 2.8 2.9 2.8 2.8 2.7 2.8 2.7 2.6 2.7 2.6 2.4 Gm. 0.0523 . 0650 .0682 .0718 .0709 .0688 .0697 .0690 .0670 .0651 . 0622 . 0629 •0571 •0550 •0525 Gm. o- 0500 .0616 •0633 . 0640 •0637 . 0629 . 0610 . 0621 .0572 .0540 .0548 .0487 •045s •0374 Mm. 9.8 10. 2 lo- 4 10. 4 lo- 0 lo. 0 10. s 10. I 9-5 9.2 9.8 9-5 9.6 9.4 9.0 8.6 Mm. 10. 0 10. 0 9.9 10. 0 10. 5 10. 0 10. 2 10. 2 9-3 9-S 9.0 8.8 9-0 9.0 Mm.. 3-S 3-8 3-7 6-9 4.0 3-8 4.0 3-9 3-8 3-7 3-7 3-7 3-6 3-7 3-S 3-S Mm. 3-S 3-7 3-8 3-7 3-7 3-8 3-8 3-8 3-7 3-S 3-6 3-S 3-4 3-3 Mm. 2-3 2.7 2.7 2-7 2-7 2.7 2.8 2-7 2.6 2.6 2.7 2.7 2.6 2.7 2.6 2.3 Mm. 2-3 2.6 2.6 2.6 2-5 2.7 2-S 2.6 2.6 2-5 2.6 2-5 2.4 3. 3 0.0218 0.0513 7-9 9-S 2.8 3-S 1-7 2-3 0. 0250 0.0451 8.4 9-S 2.8 3-4 1.8 2-3 . 0600 •0735 9.9 II. 7 3-8 4.0 2.7 2.8 •0592 .0566 9.8 10 0 3-7 3-7 2-5 2-4 .0689 •0785 10.5 ii-S 3-9 4.0 2.7 2.9 .0625 .0620 10. 0 10 3 3-7 3-S 2-S 2.S .0677 .0726 10. s II. 0 3-9 4.0 2.8 2.8 .0624 .0672 10. I 9 8 3-8 3-8 2.6 2.6 •0735 .0740 10.6 10. 0 3-9 4.0 2.8 2-7 •0633 .0643 10.3 10 I 3-7 3-7 2.6 2.7 •0743 .0660 II. 0 g.8 4.0 3-7 2.9 2.6 ■0651 .0642 10.3 10 4 3-8 3-7 2.8 2.6 .0704 .0687 II. 0 10.5 3-9 3-9 2.9 2.8 .0636 .0625 10.2 9 7 3-7 3-7 2.6 2-7 .0692 .0725 10.4 10.8 4.0 3-8 3.8 2.8 .0&2 2 .0631 10. 2 10 0 3-7 3-7 2.6 2.7 .0656 .0689 10.4 10.5 3-8 3-7 2-7 2.8 .0590 .0608 9-7 10 2 3-7 3-7 2.6 2.6 .0689 .0679 10. 0 10.5 3-9 3-8 2.8 2.8 . O5S8 .0590 9.9 10 0 3-8 3-7 2.6 2.7 .0630 .0685 10. 0 10.8 3-8 3-8 2.7 2.8 .0588 -0554 9-5 9 6 3-7 3-7 2-7 2.6 .0627 .0668 10. 0 10. 5 3-7 3-9 2.7 2.8 .0570 .0564 9-3 9 S 3-7 3-6 2-S 2-S .0582 .0649 9-7 10. 2 3-7 3-7 2.6 2.7 .0511 .0468 9-4 10 0 3-6 3-S 2-5 2-S •0517 .0605 9-S 9-b 3-S .V6 2-5 2.6 -0493 -0459 9-3 9 2 3-S 3-4 2-S 2-4 .0518 •0550 8.8 9-2 3-5 3-S 2.6 2.b .0422 -0429 8.6 8 9 3-4 3-3 2.4 2-4 .0430 .0388 8.6 8.0 3-4 3-1 2-S "' .0241 8.0 2-7 2.0 0.0447 0. 0494 9-S 9-5 3-2 3-4 2.4 2.4 0. 0488 0. 0330 9.6 8.3 3-5 3-1 2-5 2. I -065s .0641 10. 0 10. 0 3-8 3-8 2 8 2-7 .0644 .0589 10.3 10. 0 3-7 3-7 2.7 2.6 .0664 . 0692 10. 2 10. 4 3-8 3-9 2 8 2.7 .0634 .0636 10.3 10.7 3-7 3-8 2.7 2.7 .0694 .0682 10.7 10.5 4.0 3-9 2 7 2.7 .0648 -0653 10.5 9.8 3-7 3-9 2.7 2.7 .0704 . 0660 10.7 10.8 3-9 .3-8 2 6 2.8 .0059 .0663 10.5 10. 1 3-7 3-9 2.6 2.8 -0713 .0632 10.4 10.4 3-8 3-7 2 7 2-7 .0650 .0666 10.3 10. 0 3-7 3-8 2.7 2.8 .0705 . 0650 10.5 10. 0 3-9 3-8 2 9 2.7 .0627 .0611 10.4 10. 0 3-7 3-7 2.8 2.6 .0687 -0675 10. 0 10.4 3-9 3-9 2 8 2.6 ■ 0592 .0623 9-8 9-9 3-7 3-7 2-7 2.6 .0678 .0661 10. 0 10. 2 3-9 ,3-8 2 9 2-7 .0588 .0612 10. I 9.9 3-7 3-7 2-7 2.7 . 0650 .0655 10.5 10. 0 ,3-8 3-8 2 9 2.8 • 0578 .0568 10. 2 10. 0 3-7 3-6 2.7 2.6 .0639 .0651 10. 0 9-9 3-8 3-8 2 8 2.8 .0519 -0540 9.6 9-3 3-7 3-6 2.6 2.6 -0590 .0625 9-9 10. 1 .3-8 3-8 2 8 2.8 -0539 .0500 9.0 9-6 3-7 3-S 2.7 2.6 .0627 -0554 9-7 9-S 3-8 3-S 2 8 2.6 .0511 .0464 9.0 9-2 3-7 3-4 2-S 2-S • 0571 .058S 9-S 9-3 3-7 3-7 2 8 2.8 .0500 . 0461 9.0 9-8 3-6 3-S 2-S 2-S .0490 .0516 9-S 8.6 3-6 3-6 2 7 2.7 -0443 .0370 9.1 8-S 3-4 3-2 2-S 2-3 .0376 .0424 8-S 8.3 3-3 3-4 2.4 2.6 .0342 9-S 3-2 2.4 Aug. 2, 1920 Development of Barley Kernels in Clipped Spikes 461 Table IV. — Growth of kernels of Hannchen barley in owned and clipped spikes at A ber- deen, Idaho, in igi6 — Continued AUGUST 2 Normal spikes. Clipped spikes. Weight. Length. Lateral diameter. Dorso- ventral diameter. Weight. Length. Lateral diameter.! Dorso- ventral diameter. A. B. A. B. A. B. A. B. A. B. A. B. A. B. A. B. Gm. 0.0561 .0605 .0606 .0604 .0564 .0542 •0545 .C541 .0512 .0430 .0500 .0403 .0310 Gin. 0. 0404 .0600 .0617 .0631 .0654 •0633 .0605 .0585 .0484 .0518 .0485 •0455 •0373 • 0250 Mm. 9.8 10. 0 10. s 9-7 10.3 9-7 9.8 10. 2 9-5 9-S 9.2 8.8 8-S Mm. 9.0 9.6 10.7 10. s 10.3 10. s 10. I 10. 0 10. 0 9-4 9.0 8.9 8.6 7-5 Mm. 3-7 3-7 3-8 3-7 3.6 3-6 3-S 3-5 3-S 3-3 3-4 3-2 2-9 Mm. 3-2 3-6 3-7 3-7 3-8 3-7 3-7 3-7 3-2 3-4 3-4 3-4 3-2 2.S Mm. 2.6 2.6 2.7 2.6 2.6 2.S 2.6 2.6 2.7 2-S 2.6 2.4 2-S Mm,. 2-3 2.7 2.6 2.7 2.8 2.8 2.8 2.8 2.S 2.7 2.7 2.6 2-5 2.0 Gm. 0. 0439 .0614 .0632 .0612 .0598 . 0600 -0557 .0525 .0482 .0460 .0418 .0380 •0334 .0218 Gm. 0.0232 -0577 .0614 .0625 .0580 .0564 .0562 .0314 .0459 .0420 .0440 • 0363 •0339 Mm. 9.1 10. 0 10.3 10. 0 9-5 9-3 9.6 9.5 8.8 9.1 8.7 8.2 8.0 7.0 Mm. 7-7 10. 0 II. 0 10. 0 10.3 10.3 XO. I 9-7 9.2 9-5 9.0 9-2 8-S Mm. 3-2 3-7 3.7 3-5 3-S 3-6 3-6 3-S 3-4 3-4 3-4 3-2 3-2 2.7 Mm. 2.7 3.7 3.7 3.7 i-S 3-S 3-S 3-4 3-4 3-3 3-4 3-1 3-1 Mm. 2.3 2.6 2.7 2.7 2-7 2.6 2.6 2.8 2.7 2.7 2.6 2.S 2.5 2.3 Mm. 1.6 2.5 2.6 2.7 2.6 2-5 2.6 2.6 2.S 2.5 2.S 2. I 2-3 AUGUST 3 0.027s 0. 0484 7 6 9.4 2.8 3-4 I. 9 2.4 0. 0360 0.0418 8.7 8 6 3-2 3-2 2. I 2. .0636 .0639 10 0 10.3 3-7 3-7 2.7 2.7 .0561 .0311 10. 0 9 6 3.6 3.6 2.5 2. .0634 .0650 10 I 10. 0 3-7 3-7 2.7 2.7 -0550 -0332 10. 1 10 I 3-S 3-4 2.6 2. .0648 .0644 10 I 10.7 3-7 3-6 2.8 2.6 .0580 .0342 10.4 10 I 3-S 3-S 2.6 2. .0627 .0630 10 0 10.5 3-7 3-S 2.7 2.7 -0559 -0534 9.8 9 8 3-S 3-4 2.7 2. .0606 .0645 9 8 10. 6 .3-6 3-7 2.7 2.7 .0308 .0440 9.1 9 8 3-4 3-2 2.7 2. .0583 .0645 9 4 10.3 3-S 3-7 2.6 2.7 .0476 .0460 9-9 9 7 3-4 3-3 2.6 2. .0508 .0614 9 5 10. 0 3-6 3-6 2.6 2.7 -0485 -044S 9-7 9 2 3-4 3-3 2.6 2. .0508 .0540 9 S 10. 1 3-4 3-S 2.6 2.6 • 0434 .0425 9-3 9 0 3-3 3-2 2.6 2. • 0433 ■0527 9 0 10.7 3-3 3-4 2.6 2.6 .0419 .0407 9.0 9 S 3-3 3-2 2.6 2. .0418 .0505 8 8 9.8 3-2 3-S 2.S 2.6 .0400 • 0385 9-3 8 2 3-2 3-2 2.S 2. .0428 .0482 9 2 9-4 3-4 3-S 2.6 2.6 • 0374 .0362 8.4 8 4 3-2 3-2 .2-S 2. • 039^ .0402 9 0 8.6 3-4 3-3 2-S 2.4 • 0315 • 0333 8.8 8 I 3-0 3-0 2. 2 2. •o?3i .0390 8 9 8.8 3-1 3-4 2-3 2.5 .0287 .0302 8.2 8 3 3-0 3-0 2.2 2. 8.1 3-2 2-7 6 « 2.6 I. •0233 7.0 2.0 EFFECT OF REMOVING THE AWNS FROM HANNCHEN BARLEY AT ABERDEEN, IDAHO Both the material and the conditions were more favorable for satis- factory investigations at Aberdeen than in Minnesota. The Hannchen is a 2 -rowed, awned variety of barley that grows very well under irrigation. The lateral florets are infertile, and this removes the complication of prolonged flowering and the great range of variation which is present when the small lateral kernels are developing. The normal development of Hannchen barley has been discussed in an earlier paper.^ The growth under irrigation in Idaho is much more uniform than that in Minnesota. There are few cloudy days and fewer days in which the humidity is at all high. Storms which break the culms are very rare, and diseases Vv-hich affect the leaves or culms are entirely negligible. 1 Harlan, Harry V. op. cit. 462 Journal of Agricultural Research Vol. XIX, No. 9 The samples at Aberdeen consisted of at least two spikes. Just after flowering, when the kernels were small, three spikes were used. In Table IV only two of these are reported because the inclusion of the third makes the table even more cumbersome. In this table the steady growth of the kernel is apparent. Even when not averaged, the maxi- mum kernel weights during the early part of the period constitute a very uniform series. The difference between the clipped and undipped spikes becomes increasingly apparent as growth progresses. The average weights and measurements in Table V are more easily studied than are the unsummarized data in Table IV. Table V gives the average by days. In some instances abnormal kernels have been thrown out, because they introduce variations that may as well be excluded. The kernels from the clipped spikes often exceed those of the normal spikes in weight and dimensions during the first week after flowering. As was the case in Minnesota, the normal spikes soon out- strip the clipped ones and maintain their advantage until maturity. The comparative development is illustrated in figures 7 and 8. Table V. — Average -wet -weight, length, lateral diameter, and dorsoventral diameter of kernels from, normal and clipped spikes of Hannchen barley from flowering to m,aturity, at Aberdeen, Idaho, in igi6 UNCLIPPED SPIKES Date. Wet weight. Length. Lateral diam- eter. Dorso- ventral diam- eter. Date. Wet weight. Length. Lateral diam- eter. Dorso- ventral diam- eter. Mgm. Mm. Mm. Mm. Mgm. Mm. Mm. Mm. July 10 3-8 3-35 1-37 0-93 July 22 51-2 10. 17 3^58 2.52 II 5-8 4-53 1-53 •97 24 57^3 10. 27 2>-n 2. 61 12 7-4 5-58 1.65 I. 00 25 $b- ^ 10. 16 3-^7 2-57 13 14.9 8.74 1.98 T-2,Z 26 60.6 10. 22 3-72 2. 56 14 16. 0 8.87 2.03 I. 38 27 58^3 9-79 3^58 2^5i IS 21. 9 9-57 2. 14 ^•57 28 60. 2 10.05 3^74 2-75 17 37-1 10.31 3-03 2.13 29 65.8 10. 00 3.86 2. 72 18 41.7 9.99 3- 29 2. 21 31 64 4 10. 16 3^76 2. 70 19 42.3 10. 01 3-34 2.30 Aug. I 61. s 9.92 3-75 2. 72 20 43-5 10. 27 3-34 2.30 2 51^9 9. 61 3^47 2-59 21 46. 9 IO-3S 3- 51 2.36 3 52.2 9-57 3^47 2.58 CLIPPED SPIKES July 10 3-3 3. 00 1.27 0.85 July 22 47-1 9. 80 3^47 2. 27 II 6 7 5- -^3 I-5S •97 24 52-9 9^95 3-64 2.45 12 7 8 5- 81 i^.^8 .92 2,S 5^-3 9.99 3-5° 2.41 13 13 8 8.28 1.89 1.27 26 57^4 9.98 3-59 2.50 14 18 T, 9. 41 2. 12 1.48 27 52.6 9.66 3-45 2.44 15 21 6 9.46 2.24 1-57 28 55^6 9.72 3.60 2.58 17 31 9 9.49 2. 91 1^95 29 59^6 9.71 3.68 2^57 18 36 4 9.71 3-05 2. 07 31 56-4 9-73 3^6i 2-54 19 37 I 9^56 3. 22 2-15 Aug. I 55-6 9-75 3.62 2. 60 20 39 8 10. 00 3.21 2.17 2 49^7 9^38 3-42 2- 55 21 41 9 9^83 3^36 2.23 3 43^5 9.17 3^27 2. 46 Aug. 9, 1920 Development of Barley Kernels in Clipped Spikes 463 The graphs of the growth in length essentially coincide for six days after flowering. For some reason not apparent, the kernels in the normal spikes reach a greater length than those of the clipped spikes. This greater length is still in evidence at maturity. The difference is only X /^/7P. /aoo s.so SiOO &SO / \ / ^ - " "V V rf"" ^ "^ ^ ^^ , / / ,^ / / t 'v - -' '" ^ /' — - V ^ _ 1 1 / \ 'n z S/ VG 7> / y 1 p^^o- &50 €.00 _,| SdO r / / 41.00 3.^0 t^ ■^ V ^ < ^ .. - - ^ y-r -r> 1 1 /, ,• — /. k ^ 7//f/V£7-S. 1 1 1 1 3.00 / K ^ s- "~ri 1 1 1 _. .^ ^, zyo/Pso-ys-zVpyF^z. \ 200 Y ^ / / - ^ y '- ^ I 9/. 4/ V^ 72 f>? •> ^ A / f-- /J2? / / ^ / /OO 7^ ws f— a 9 /o// AP/s/^ /£/s/7/0/S£o^/^^^Ii^^s^s27^<9^ssa3/ / 2 3^s 6 7 a 9 /oaf/3 Fig. 7. — Graph showing growth in length, lateral diameter, and dorsoventral diameter of kernels of Hann- chen barley in normal and clipped spikes. mm., but it occurred in both Minnesota and Idaho. A part of the differ- ence seems to have been due to the greater water content of the normal kernels, for the graphs of kernel lengths approach each other again at maturity. At Aberdeen, the course of the development of the lateral diameter is much like that of the dorsoventral diameter. Seven or eight days after 464 Journal of Agricultural Research Vol. XIX, No. 9 flowering, the diameters of the normal kernels are larger than those from clipped spikes, and they then continue larger for the remainder of the period of development. In Minnesota, there is little difference between the kernels of the two classes of spikes until near maturity. As maturation approaches, the normal kernels are found to be uniformly greater in diameter than are those from the clipped spikes. The graph of the wet weight is much more uniform at Aberdeen than at St. Paul. As in Minnesota, there is no difference between the clipped and normal spikes in the first few days. The period of equality is shorter 65 €0 S5 SO ^0 ^ 1 / ■^ \ \ / ^ ^ y *v ^x ., \ t / f 4 / -- / \ / \ ^ / / \ \ J - y / t J5 JO 2S 20 /S i |/ .^ / , i 1 > ft > / I 1 r ' 1^ 'A ?/ 7Z D iO s . ^. .J .. .. .. . A > 1 X y f 8 9 /o///^/3/4/5/6/7/8/92o^/^^^5^4^s^6^7^B29J03// 23 Fig. 8. — Graph showing wet weight of kernels of Hannchen barley from normal and clipped spikes. at Aberdeen, however. After July 15 the average wet weight of the kernels from clipped spikes here never equals the wet weight of the kernels from normal spikes. The wet weight -includes a variable amount of water, which increases during the first half of the growing period and decreases during the second half. I^or this reason the curv^e of the wet weight differs greatly from the cur\^e of the dry weight. The dry weights are shown in figure 9. In Minnesota, the trend of increase in dry weight was quite uniform, as was shown in figure 3. In Idaho, the graph of the dry weight is almost a straight line. It would seem that in both the normal and clipped Aug. 2, 19:0 Development of Barley Kernels in Clipped Spikes 465 spikes the rate of growth was very nearly at its maximum. If this is true, the maximum of the cUpped is less than that of the normal spike, for after July 15 the dry matter per kernel is always less. The percentage and weight per kernel of the dry matter are given in detail in Table VI. This table also includes the data on water, nitrogen, ^S7 /c? /\J_ I J:l _ JL..... UJt. 11 ^ ^y Fig. 9. — Graph showing dry matter in kernels of Hannchen barley from normal and clipped spikes. and ash. These figures were obtained from the analyses of the samples reported in Tables IV and V. INIost of the data on the percentages of the various substances have not been included in the figures. An inspection of the tables shows a surprisingly uniform decrease in the percentage of water and, of course, an equally uniform increase in the percentage of 183718°— 20 6 466 Journal of Agricultural Research Vol. XIX, No. 9 dry matter. The difference between the percentage of materials present in the kernels of normal and clipped spikes is necessarily in direct rela- tion to the actual quantities. Table VI. — Average percentage and weight per kernel of dry matter, water, nitrogen, and ash in kernels from normal and clipped spikes of Hannchen barley at Aberdeen, Idaho, in igi6 NORMAL SPIKES Date. Dry matter. ^ Water. I itrogen n dry Qatter. Ash in dry matter. Wet weight per kernel. Dry- weight per kernel. Water per kernel. Nitro- gen Ash per kerc kernel. per el. Per cent. Per cent. P er cent. Per cent. Mgm. Mgm. Mgm. Mgm. Mg m. July 8 20. 48 18. 18 79- 52 • - 81.82 7-46 3- 40 I. A O- "i I. I 0 02 lO 7-33 3 8 ■J 7 3-1 0.05 02 II 19.23 80.77 4. 61 4.21 5 8 I I 4-7 05 05 12 19. 10 80. 90 4. 16 5-47 7 4 I 4 6.0 06 08 13 22.03 77-97 3-56 4-33 14 9 3 3 II. 6 12 14 14 22. 76 77.24 3-15 3-74 16 0 3 6 12.4 II 13 IS 23-87 76. 13 2.94 3-69 21 9 5 2 16.7 IS 19 17 28. 41 71-59 2-33 3.06 37 I 10 5 26.6 24 32 i8 31. 06 68.94 2. 01 3- 21 41 7 13 0 28.7 26 42 19 32-47 67-53 I. 91 3-52 42 3 13 7 28.6 26 48 20 33-21 66. 79 1.80 3-45 43 5 14 4 29. I 26 5° 21 36. 21 63-79 I. 92 2.87 46 9 17 0 29.9 32, 49 22 38.92 61. 08 1-93 2.63 51 2 19 9 3-^-2> 38 52 24 42.38 57.62 1.97 2. 50 57 3 24 3 33-0 48 61 25 47-59 52-41 2. 02 2. 56 56 5 26 9 29. 6 54 69 26 43-96 56. 04 2. 06 2-45 60 6 26 6 34- 0 55 65 27 51-37 48.63 1.83 2. 60 58 3 29 9 28.4 55 78 28 50.69 49-31 2. 06 2-35 60 2 30 5 29.7 63 72 * 29 50-99 49- 01 2.03 2.32 65 8 32, 6 32. 2 68 78 31 56. 01 43-99 2-33 2.41 64 4 36 I 28.3 84 87 Aug. I 60. 96 39-04 2.17 2. 20 61 5 37 5 24. 0 81 83 2 65.81 34- 19 2. 07 2.30 51 9 34 2 17.7 71 79 3 71.91 28. 09 2.25 2. 25 52.2 37-5 14.7 84 84 c LIPPED spn CES July 10 17.8 82. 2 .5- 43 6.63 3-3 0.6 2.7 0 03 0 04 II 19. 0 81. 0 4. 14 4.17 6 7 1-3 5-4 OS 05 12 28.0 72.0 3-73 4. 80 7 8 2. 2 5-6 08 II 13 21. 6 78.4 3-52 4. 28 13 8 30 10.8 II 13 14 23- S 76.5 3- 04 4. 08 18 3 4-3 14. 0 13 18 15 24. 6 75-4 2.56 3-76 21 6 5-3 16.3 14 20 17 29. 2 70. 8 2-59 2-3(^ 31 9 9-3 22. 6 24 31 18 30. I 69-9 2. 02 3-18 36 4 II. 0 25-4 22 35 19 32-4 67.6 2. 00 2.89 37 I 12. 0 25-1 24 35 20 35-3 64.7 2.03 3.00 39 8 14. 0 25-8 28 42 21 39-4 60.6 2.03 2-93 41 9 16. 5 25-4 33 48 22 38.5 61. 5 I. 90 3.00 47 I 18. I 29. 0 34 54 24 43-3 56.7 1-97 2. 69 52 9 23.0 29-9 45 62 25 47- I 52-9 2. 06 2-54 52 3 24. 6 27-7 49 62 26 47.2 52.8 2. 06 2.80 57 4 27. I 30-3 56 76 27 51-2 48.8 2. 10 2.45 52 6 26. 9 25-7 56 66 28 52.8 47-2 2-15 2-39 55 6 29-4 26. 2 63 70 29 52.6 47- 4 2- 15 2. 48 59 6 3^-3 28.3 67 78 31 58-4 41. 6 2. 06 2.08 56 4 32-9 23-5 68 68 Aug. I 62. 5 37-5 2.17 2.27 55 6 34-8 20. 8 76 79 2 67-9 32. I 2. 24 2.31 49 7 33-7 16. 0 75 78 3 77.6 22. 4 1.98 2.26 43-5 33-8 9-7 1 67 76 Aug. 2. I920 Development of Barley Kernels in Clipped Spikes 467 The ash was determined in more organs at Aberdeen than at St. Paul. The percentage of ash in the rachis, paleae, and awns is shown in figure 10, as well as the ash in the kernel. The analysis of the other structures throws much light on the problem. The awn contains a surprising amount of ash. At flowering time 10 per cent of its dry weight is ash, while at maturity 33 per cent of the dry weight is ash. kXs' ~ r r ~ J4 ^ .-. -* J2 J/ JO 29 re er y ■^ ^< y '4 — — ^ J / t / ( / / 25 2^ 23 /a I'" /^ /s /£ // /O s 7- 6 i j / ?^ W i 4 ./ / >v. 1 / / / i / / / /?AC/f/S ^ / ^ ^ .' > f ?- ^ ^ f y «r' -"^ ^' / _, ^ _ 4 y , /< X / ^ (^ / > > ^ *~/^/iliE-A - ' >^ /" »^' •/? V ^ ^ — 1 ^ -- ^ -^ /? 4C -A /^ ■^ ^ * -- -- — ^ ^ Ki _ — <; — ^ ^ ^ ;^ ■-« »* SJ '^ ^m 1; ^ kd Sa Ed m. /O /m /3 f4 /S /6 /7 /8 /9 202/ 22 Z32^2S352rS329JOZ/- / 23-4^S6Z6S Fig. 10. — Graph showing percentage of ash in the kernels, rachises, paleae, and awns of normal spikes of Hannchen barley and in the kernels, rachises, and paleae of clipped spikes. The total amount of ash present is considerable. The percentage of ash in the kernels of the clipped spikes is about the same as in those of the normal spikes. The paleae of the clipped spikes contain more ash than those of the normal spikes. It is in the rachis that the greatest and most significant difference occurs. The rachises of the clipped spikes contain 25 per cent more ash than the rachises of the normal spikes. It would seem that much 468 Journal of Agricultural Research Vol. XIX, No. 9 of the mineral content that usually goes into the awn remains in the rachis of the clipped spike. These rachises were found to be brittle, while the normal ones were not. Both in Minnesota and in Idaho the clipped spikes had a tendency to shatter, while the normal spikes exhibit y^/^- oeo an O.ffo / s i 1 / / \ \ i /■ f ""' ' \ 1 1 - 1 A a 9/ \ "is 1, 1 / f i I 1 c *A '>A '£ O / < • 1 1 G20 / 1 0/0 } ?< / f / _ f 8 9/0// IZ /J H- /S/6 irJ8f9Z0 2l22Z32'fZS2627ZaZ9X)J/-/ 2 3 Fig. II.— Graph showing total nitrogen in kernels of Hannchen barley from normal and clipped spikes. no such tendency. The divergence in ash content is surprisingly large and widens consistently throughout the period of growth. The increase in nitrogen per kernel in Idaho is similar to that found in Minnesota. The amount of nitrogen in the kernels from clipped spikes is almost as large as that in the kernels from normal spikes. The average is slightly less, but as a whole the content of nitrogen is nearly equal in the two, as may be seen in figure 1 1. The difference in water content shown in figure 12 is more striking at Aberdeen than at Minnesota. After July 15 the kernels from clipped Aug. 2, 1920 Development of Barley Kernels in Clipped Spikes 469 spikes never contain as much water as those from normal spikes. This is in full accord with the results obtained at St. Paul, but the greater uniformity of the development at Aberdeen emphasizes the difference of behavior by removing the confusion of abnormal samples. In a preliminary experiment conducted at Arlington Farm, Va., the relation of the length of the awn to the weight of kernel was studied. The awns increase in length from the base of the spike for about one- third the distance to the tip. The spikelets on the upper two-thirds of JLftr /ii/G. Fig. ij.— Graph showing water in kernels of Hannchen barley from normal and clipped spikes. the spike exhibit a gradual decrease in awn length, the shorter awn occurring on the apical spikelet. Figure 13 shows a composite spike resulting from the average of the data obtained. In this case the node numbers include both sides of the spike and are alternate. The weights used are the average of the kernels at two adjacent nodes. It will be seen in the figure that the greatest difference in weight results from the removal of the longest awns. The removal of the short awns near the tip affects the yield only slightly. If the curve of the clipped kernels is taken as showing that the normal peak due to nutrition occurs at about node 9 or 10, the greater length of awn on node 6 is seen to move the peak of the 470 Journal of Agricultural Research Vol. XIX, No. 9 kernels from awned spikelets nearer to the base of the spike than is the case in the clipped spikelets. DISCUSSION OF RESULTS The results in both Minnesota and Idaho have a direct bearing on the two chief field problems in the production of hooded and awnless barleys. These barleys have not yielded as well as the bearded sorts, and they have shattered. The barleys from which the awns were removed did not give as high yield in these experiments as the awned plants growing beside them. This conforms to the experience of Zoebl and Mikosch, Schmid, Perlitus, /^/77. ^Z C/r?. 2S /^a-^sssye a /o// /^/s/^/s/s/7/3/9JX?^/^^^s£|^^s2lS' A/p o/^ A/aa^ /^/?o/^ ^y^S£- o^ ^s^^/A^r Fig. 13.— Graph showing relation of length of awn to weight of clipped kernels and undipped spikelets on a 2-rowed barley grown at Arlington Farm, Va. and some other investigators. In this study it was evident that the reduction in yield was not due to any injury to the plant, as the differ- ence in growth was not apparent for several days after the awns had been removed. The early growth of the kernels in clipped and normal spikes was equally vigorous. It was only when starch infiltration be- came rapid that the awned spikes showed greater activity. The dififer- ence in ultimate weight was largely due to the difference in the quantity of starch present. There was little difference in the quantities of ash and nitrogen. Zoebl and Mikosch looked upon the awn as an organ of transpiration. Whether the reduction of transpiration alone is suffi- cient to account for the lower rate of starch production is a question. That transpiration has an influence on the behavior of the hooded Aug. 2, 1920 Development of Barley Kernels in Clipped Spikes 471 barleys is indicated in the field experiments. These barieys have proved relatively better in dry years on the northern plains than in wet years. In the "good" years the hooded varieties have been far inferior to the best bearded sorts, but in "bad" years they often have been better. In any case, these two experiments show that the awn has a function, and the loss of the awn has resulted in a reduced yield. The second field problem is that of shattering. The common hooded and awnless varieties have a tendency to shatter at maturity. The clipped spikes of Manchuria and Hannchen barleys showed a ten- dency in this direction; the normal spikes did not. The spikes from which the awns were removed proved to be fragile, and many of them fell to pieces as maturity approached. An explanation of this behavior was found in the determination of ash in the awns, rachises, and paleae. The ash that normally went into the awn was deposited largely in the rachis of the clipped spikes. The additional ash seems to have been sufficient to cause the rachis to be brittle. It would seem that the awn also served as a place in which to store the excess of ash. More mineral matter probably is taken up in growth than is needed by the plant. There is no method of elimination. The extra mineral is deposited in cells which probably serve little purpose other than storage. The removal of tissues and organs containing cells which can be devoted to this end must, in itself, cause some derange- ment of the normal processes of development. From the experiments conducted, it would seem that awnless and hooded barleys are limited by the loss of the awns. It appears that high yields are not to be expected from such varieties. It is to be expected that such sorts will shatter more than awned kinds. This has been the experience in breeding also. For the most part awnless hybrids have been brittle and of low yielding capacity. It is thought that there is little use in attempting to secure valuable awnless or hooded varieties by means of hybrids with most varieties. One possible method of breeding has been indicated by experiments not yet published. Some varieties of awned barley have normally a much lower content of ash in the rachis than others. It is possible that the progeny of crosses with these and the hooded sorts may yield well in semiarid climates and that they will not shatter. One or two such hybrids are now giving promise. When the first elementary experiment conducted in Minnesota indi- cated the physiological difficulties in the way of producing desirable varieties of hooded and awnless barleys, work was amplified in another line. Several hundred hybrids with smooth awns have been produced and tested. Much of this work has been done in the cooperative experiments with the Minnesota Agricultural Experiment Station, but many strains have been tried elsewhere. Several of these give promise of good yielding capacity. 4^2 Journal of Agricultural Research voi. xix,no.9 The awns of these hybrids are smooth. All the large scabrous teeth on the basal two-thirds of the awn have been eliminated. The tips of the awns are slightly rough, but this roughness is not sufficient to be objec- tionable to either growers or feeders of barley. Whether varieties of this type can be made to yield equally as well as the awned sorts remains to be determined. SUMMARY The removal of the awns from a barley spike has a marked effect on the development of the kernels of the spike. Kernels from clipped spikes have smaller volume and a lower weight of dry matter at maturity than do those from normal spikes. The difference is not due to the injury or shock of removing the awns; the kernels in the clipped spikes develop as rapidly as those in the normal spikes for several days after the awns are clipped. About one week after flowering the deposit of dry matter in the ker- nels of the normal spikes begins to exceed that in the kernels of the clipped spikes. This is about the time that rapid starch infiltration begins. The daily deposit of nitrogen and ash is more nearly equal in the two classes of spikes than is the deposit of starch. In normal spikes at Aberdeen, Idaho, the awns contained more than 30 per cent of ash at maturity. When the awns were removed a part of this ash apparently was deposited in the rachis. The rachises of the clipped spikes contained about 25 per cent more ash than the rachises of the normal spikes. The additional ash in the rachises of the clipped spikes probably was responsible for the tendency of these spikes to break. The indications are that the elimination of the awns results not only in lower yields but in shattering as well. Hooded and awnless barleys generally yield less and shatter more than awned varieties, and there seem to be physiological reasons for this fact. It may be possible to produce nonshattering hooded and awnless sorts by using parents which normally have a low percentage of ash in the rachises. It may be possible to obtain strains that will give good yields under arid conditions. Under humid conditions it is likely that the objections to the awns are more easily met by the use of strains with smooth awns, which, so far as known at present, have no physiological limitations. . Vol. XIX AUOUSX 16, 1920 No. lO JOURNAL OP AGRICULTURAlv RESEARCH CONXE^NTS Investigations in the Ripening and Storage of Bartlett Pears -------.»_ 473 J. R. MAGNESS ( Contribution from Bureau of Plant Industry ) Further Data on the Orange-Rusts of Rubus - - - 501 ' L. O. KUNKEL ( Contribution from Bureau of Plant Industry ) Germ-Free Filtrates as Antigens in the Complement- Fixation Test ------»_ 513 WILLIAM S. GOCHENOUR ( Contribution from Bureau of Animal Industry) Mosaic Disease of Corn -----». 517 E. W. BRANDES (C«ntTibutIon from Bureau of Plant Industry) PUBUSHED BY AUTHORITY OF THE SECRETARY OF AGRICULTURE, WITH THE COOPERATION OF THE ASSOCIATION OF LAND-GRANT COLLEGES WASHINOTON, r>. C. WAaHINOTON ; OOVERNUENTPBINTINaOPriOe : UM EDITORIAL COMMITTEE OF THE UNITED STATES DEPARTMENT OF AGRICULTURE AND THE ASSOCIATION OF LAND-GRANT COLLEGES FOR THE DEPARTMENT KARL F. KELLERMANi Chairman Physiologist and Associate Chief, Bureau of Plant Industry EDWIN W. ALLEN Chief, Office of Experiment Stations CHARLES L. MARLATT Entomologist and Assistant Chief, Bureau of EntOTnology FOR THE ASSOCIATIOH J. G. LIPMAN Dean, Stale College of AgriculHir*, * r^ M »o o a; " u '^00 t^- ^ r-» ro Q, loo ti 1 •K lil O i^ n r^ O ~ " •* c<) moo M Q^ »>• t^oo oo 00 6 ^ •* f*l O O O f*5 D, 4 ^%o >o vo 00 vo »o o t^ I fO wi ^ r^ 1\0 o o t~» Q^ o o t^ t^ j>» «-« ^ c4 ♦J P o£ Q,, f^ fo »^ »oo «^ ^ « OOO ^00 Tj- > O 00 o o^o O O OvOO 00 f 00 V) M ro fo 5-9 la's ^ § '>-^ ^ p^ -^ • o ■ •* « s 0 ^ ■ o -O VO a; :" CI) 01 . Q. ^ fci ^ «00 to o 00 ■* •a OS 0 o;^? M M % V '" t/i Q a, M w » « « d • Ji3 Oi ft, M » « » M >. 01 u Ui ^ i8 t m •? m c« o cC :« M M ■'1 0) • k. CS % TiS-^ OlOO 3 °2 41 " % o;'^- W M g- a to ^SS'S^? ■1^ m o a, " " " " " |22 S.ti .^J t~ o o o. •* < o- a;-oro-^«- HI b ■« : g o t^o ct o o ^ q; -i^ t^oo a be I 1" fe V f; in t» O looo ^ o vO o. o o\ D, in tioo Oi d\ ■~ 0 0 t^ 0 IN ft, CO ino o CO !S fe ■Q • O ?15 CIS o a; :o si t^ a afi to ~ « t» lO lo C (4 ^-OOO o o 0 ft, 'h^ .al V "> S to " In 0 0* 'SS 3 w o ft, ^ f>-o6 t-» «>• S.ti ~oo OvinTfoo 5-e ft, « ■* invo v6 Im u:, '■■S^ ■ a o o ft; :•:: • 0 . „ ,, d to ..i w* • 00 r~ g ts ^ q;^c. " " S a to •a (J r>. O »n r^oo R a,--:- n^ .^.XOO O 0 N " O- ^,.... "3 y 0 HI o « o IS -a §« 1 l-lh-1 < o-a 1 ^ 0 V m Aug. i6, 1920 Ripening and Storage of Bartlett Pears 483 l:^ Qi '-^ D ^ if . to ^ ov M a 0 ^ UHO t^ a,-"" a a 4> to ~ : sO PO -a 1 °2 ^ Q, : . ■* 0. 0 >. I- 0 0 R Q,' " " " a . 0 t^oo a^ -• 3 6- b ai " " " >> CS ?g •§£3 9. 3 ^ 0 r<) fc to •~ 0 N 0 S! 0 0) 0 ft, t^oo 6 3 a V tu ^ a to t ;^ 5 ;^ a °2 % ft, ti t^ Ci s V "' to' •Q o> •-■ 0 ,0 W R ft, 0606 d s. J '^00. H 1: 0- 5 ai vo'o u to ^ V . a V ?> ft, \0 t^ r^ a . •- M 1- a to ■{• 00 >A r~ 3 u M 8 .5 3 i ^ q;«« ^ to' °R & 9. M ■Q Ov » 0- a:,^^, 5 to _: 00 >* M ii . r~ 0 >o a 0 ft ': ?? 0 0 to ^ t^oo 0 a "2 " ? !Q I^ a a 0 ft,-7 to -: 00 a M u 0 -^ 'O < w 0 ■^o ' ? ? 9 M 1^ 3 ft;-? **. 000 CO >>b« s^ fi-Q 6 Z :$ H ^ >. •^ ^ "1^ U > 0 to ^ N r^ f. 1 "o '^ ^ i>> A .SSi to ~ 0 0 •s o2 0 ft,' M^ r* 1 0 (U " a to • 0 0 00 en 0 ft;.^:^^ S-M- ~2S J! "3 Oi ft, " - i s W S to »: « 0 0 •iS 0 ft;""" ■§l to ■Q ooo 3 " »2 0 a. M Ht g- a to ^ ■* (?. Ol l-g^ m °? ft; "" " |S3 9d ti5-S' ^•0 ot "2 < q; "« li t> to' ■Q 0 0 >o 0 (J! 0 ft, 0 06 d 3 0, ii tx 'u a to 73 R^ 01 00 a % ft, K.M to' 73 S! 0 i 03 0 ft;^ A §•1 a; i^o to CO a) 0 a, >n« >d be . •nS to ts sa q; 0 0 0 bo S .9 3 0) " to' ft; vd 9 l^ f«S £■3 0* '^ 01 fO ft; -i-o "3 to 0 »I "> N 0 .5 si t^ 0 6 to' >n 0 a < 2I 4) a 0 a; 6 to 0 a; ^?^ 0 \ 3 -«-' 5I ^S2 a, ': ? 0 . :^4 o-a d 2 :i „ 484 Journal of Agricultural Research Vol. XIX, No. 10 For purposes of comparison and discussion, however, the results are also presented as a series of curves, in which it is possible to bring similar substances under the varied treatments into direct comparison. influence; of time of picking upon sugar content of fruit Figures i to 4, inclusive, summarize the data on the development of sugars in the fruit at the various intervals at which pickings were made OAT£ P/Cf<£0 F/iOAf TR££ AUGUST Fig. I.— Sugars in Bartlett pears from Sacramento, Calif,: Curve i, reducing sugars in green fruit when picked from the tree; curve 2, total sugars in green fruit when picked from the tree; curve 3, reducing sugars in collateral lots after ripening at 70° F.; curve 4, total sugars in collateral lots after ripening at 70°; curve 5, reducing sugars in collateral lots after ripening at 40°; curve 6, total sugars in collateral lots after ripening at 40°; curve 7, reducing sugars in collateral lots after ripening at 30°; curve 8, total sugars in collateral lots after ripening at 30°. and the influence of various types of storage upon the sugar content of the fruit picked at these same intervals. Curve i in each figure repre- sents the reducing material present, figured to percentage of green weight in the fruit fresh picked from the tree. According to Thompson Aug. i6, 1920 Ripening and Storage of Bartlett Pears 485 and Whittier (17) this consists mainly of levulose, but it has been figured as dextrose here because comparative results are of primary interest. Curve 2 represents the total sugar or reducing material after inversion, so that the distance between curves i and 2 represents the amount of Ji? S /O JO ^ ju/^e Fig. 2.— Sugars in Bartlett pears from Suisun, Calif.: Curve i, reducing sugars in green fruit when picked from the tree; curve 2, total sugars in green fruit when picked from the tree; curve 3, reducing sugars in collateral lots after ripening at 70° F. ; curve 4, total sugars in collateral lots after ripening at 70°; curve 5, reducing sugars in collateral lots after ripening at 40°; curve 6, totil sugars in collateral lots after ripen- ing at 40°; curve 7, reducing sugars in collateral lots after ripening at 30°; curve 8, total sugars in collateral lots after ripening at 30°. sucrose present. There is in every case a marked increase in the amount of reducing sugar present at successive dates of picking. This increase is somewhat more rapid early in the season, although a distinct increase occurs as long as any pickings are made. It is unfortunate that it was 486 Journal of Agricultural Research Vol. XIX, No. lo impossible to secure even later pickings to see if this increase in reducing sugar continues until the fruit is fully ripe on the tree. The amount of sucrose remains nearly constant at less than i per cent throughout the early season. In the late season there was an increase in the amount of sucrose to i}4 per cent in the late picks of California -^ ^ A^ /S £4^ ^3 Fig. 3 . — Sugars in Bartlett pears from Medf ord , Oreg. : Curve i , reducing sugars in green fruit when picked from the tree; curve i, total sugars in green fruit when picked from the tree; curve 3, reducing sugars in collateral lots after ripening at 70° F. ; curve 4, total sugars in collateral lots after ripeningat 70°; curve s, reducing sugars in collateral lots after ripening at 40°; curve 6, total sugars in collateral lots after ripening at 40°; curve 7, reducing sugars in collateral lots after ripening at 30°; curve 8, total sugars in collateral lots after ripening at 30°. fruit. The increase in sucrose in the late pickings is such that the increase in total sugar shows no falling off in rate up to the date of the last pickings secured. The less rapid increase in reducing sugar is counteracted by the increase in sucrose. Curves 3 and 4 in each figure represent reducing and total sugar, respectively, when the fruit reached prime eating condition in a storage Aug. i6, 1920 Ripening and Storage of Bartlett Pears 487 ^.Si '^S JO ^ ^ /='/^OM TR££ Fig. 4— Sugars in Bartlett pears from Yakima, Wash.: Curve i , reducing sugars in green fruit when picked from the tree; cur\e 2, total sugars in green fruit when picked from the tree; curve 3. reducing sugars in collateral lots after ripening at 70° F. ; curve 4, total sugars in collateral lots after ripening at 70°; curve 5. reducing sugars in collateral lots after ripening at 40°; curve 6, total sugars in collateral lots after ripening at 40°; curve 7, reducing sugars in collateral lots after ripening at 30°; curve 8. total sugars in collateral lots after ripening at 30°. 488 Journal of Agricultural Research voi. xix, No. 10 temperature from 65° to 70° F., curves 5 and 6 in 40° storage, and curves 7 and 8 in 30° storage. The distance between these curv^es and curves i and 2, at the various dates, represents the increase in sugar as the fruit ripened in the dififerent storages. It will be noted that the sugar runs uniformly highest in the fruit ripened at 70°. This is true for both total and reducing sugar in fruit picked at all the different dates. Apparently, either the loss of sugar from respiration is less, or more sub- stances insoluble or nonreducing in the green fruit are changed to soluble reducing material when the fruit is ripened at this optimum ripening temperature than when ripened at lower temperatures. In every lot, regardless of the section from which it came or the date at which it was picked, fruit held at 30° F. was higher in sugar than that stored until ripe at 40°. It must be borne in mind, however, that the 30° fruit was not completely ripened in storage but was held for periods of from a little over three months in the early picked lots to a little over six weeks for the last lots from Oregon and Washington. Then it was removed and held at warm room temperature until ripe, the time required being four to six days. This may have made some difference in the analytical results. Also, at one period, because of a sudden drop in temperature, the fruit picked in the earlier lots was partly frozen. It was thawed very gradually, and no ill effects of the freezing were notice- able afterwards. The fact that the frozen lots and those of the later pickings that did not freeze showed no marked difference in analyses other than that to be expected from the results with the same lots in other storages is also evidence that the carbohydrates of the fruit were not materially affected by the freezing. The general effect of storage upon the sugar content of the fruit was very similar, however, in fruit from the different sections. The curves for total sugar — No. 2, 4, 6, and 7 — cross in only one point in all the figures, showing that the relative amounts of sugar in the different storages run the same in all cases. It seems well estabhshed, there- fore, that the highest amount of sugar will be secured by holding the fruit at optimum temperature for ripening. In case it is necessary to prolong the time of keeping the fruit, holding it at very low temperature until near the time it is needed and then ripening it up at optimum tem- perature gives a higher sugar content than holding it at a temperature just low enough to retard the ripening processes. From the results obtained by Gore (5) on the respiration activity of fruits at different temperatures, it would be expected that respiration would occur at least three times as rapidly in the 70° as in the 40° F. The number of days required to ripen the fruit in the two storages was about in the proportion of three days at 40° to one at 70°, so the total respiration activity would seem to be about equal. If this is true, it would seem that certain factors other than respiration must enter into the relative amounts of sugar present in the different storage lots. Aug. i6. 1920 Ripening and Storage of Bartlett Pears 489 RELATION OF SUCROSE) TO REDUCING SUGAR DURING STORAGE There is a marked increase in sucrose during the time between picking and the full ripening of the same fruit. This is shown by a comparison of the distance between curves i and 2, representing the sucrose in the fruit fresh from the tree, and between curves 3 and 4, showing the sucrose in the same fruit when ripe. There is a very marked increase in sucrose during storage in the earlier pickings, and this increase is even more marked in the late pickings. The late pickings show very little increase in reducing sugar between the time of picking and the time the fruit was ripened, while the increase in sucrose was very marked, being sufficient to make the total sugar increase between the time of picking and full ripeness practically as much in late-picked as in early picked fruit. There seems to be little relation between temperature and kind of sugar in the fruit, the 70°, 40°, and 30° F. storage lots being quite similar in the proportion of sucrose to reducing sugar. A review of all the curves indicates that, whereas in the early picked fruit almost all of the sugar is in the form of reducing substances, the increase in reducing sugars in successive lots, as the season progresses, is much less marked than is the increase of sucrose. In all the lots, reducing sugar in the late picks seemed to run to between 7 and 8 per cent of the green weight of the fruit, after which there was very little increase in reducing substances, while sucrose continued to increase rapidly until after the last pickings were made. RELATION OF ACIDITY TO TIME OF PICKING In the relation of acidity to the time of picking there is not so distinct a correlation in all cases as there is for the sugars. Fruit from different districts seemed to respond somewhat differently in this regard, though certain general tendencies hold for all regions. Figures 5 to 8 summarize the results on acidity, computed as malic acid in terms of percentage of wet weight of the fruit. Curve i in each plot represents acid in the green fruit, curve 2 in fruit ripened at 70° F., curvx 3 in fruit ripened at 40°, and curve 4 in fruit ripened at 30°. In fruit from Suisun, Calif., (fig. 6) there is a constant decrease in acid in the green fruit from the time of the first picking until the last. On the other hand, in fruit from Sacramento (fig. 5) there is a slight rise until July 5, about the opening of the picking season, followed by a drop toward the end of the season. In fruit from the more northern sections, however, there is an increase in acid instead of a decrease. The increase is rather slight in the Medford pears (fig. 7), but very marked in those from Yakima (fig. 8). It is interesting to note that, whereas the acidity of the fruit decreased in the California sections, fruit from the Medford section showed a slight increase, and that from the still more northern Yakima section showed a very marked increase 183719°— 20 2 490 Journal of Agricultural Research Vol. XIX, No. lo as the season advanced. While the data are much too limited to justify the assumption that this relation of acidity to latitude generally holds, the results of the one year's work are of sufficient interest to warrant further study along this line. KFFECT OF STORAGE UPON ACIDITY A somewhat greater uniformity exists in the relation of temperature of storage to the acidity of the ripened product than was found in connection with the time of picking. In the first place, it will be noted so JO s /o 2S JO /iUGUST /^ Fig. s. — Acids in Bartlett pears from Sacramento, Calif.: Curve i, acid in green fruit when picked from the tree; curve 2, acid in collateral lots after ripening at 70° F.; curve 3, acid in collateral lots after ripening at 40°; curve 4, acid in collateral lots after ripening at 30°. that in the early picks there is a wide variation in amount of acid, due to the temperature of the storage used; and in most cases there is a greater amount of acid in the green fruit than in the ripened fruit, regardless of the temperature at which it was held. Fruit picked in a very immature condition has less acid when ripened than at the time of picking. (Curves 1-4, fig. 5-6, early pickings.) Fruit picked at about the time of the opening of the commercial season, however, behaved somewhat differently. In every case the fruit ripened at 70° F. contained a higher percentage of titratable acid than did the same fruit when picked from the tree. This is of interest especially in connection vath the question of whether fruit acids are synthesized in the fruit itself or whether they are carried to the fruit from the leaves. The fact that there is an increase in the acid between the time the fruit is removed from the tree and the time of its becoming ripe is evidence that there is an actual synthesis of acid in the fruit itself. Aug. i6, 1920 Ripening and Storage of Bartlett Pears 491 r4tS?| eo es OATS' i^yOTiirZ? /=>«?C?A^ T-Zf/^S: -4: ^ Fig. 6. — Acids in Bartlett pears from Suisun, Calif.: Curve i, acid in green fruit when picked from the tree; curve 2, acid in collateral lots after ripening at 70° F. ; curve 3, acid in collateral lots after ripening at 40°; curve 4, acid in collateral lots after ripening at 30". '/S ^o '^ 9 /9 24 2^ Fig. 7. — Acids in Bartlett pears from Mcdford, Oreg. : Curve i, acid in green fruit when picked from the tree; curve 2, acid in collateral lots after ripening at 70° F. ; curve 3, acid in collateral lots after ripening at 40°; curve 4, acid in collateral lots after ripening at 30°. 492 Journal of Agricultural Research voi. xix, no. w 0/fSO \ X \Q3S0 ^0.325 \ s a2So I Ka22s /s Fig. 8.-Acids in Bartlett pears from Yakima. Wash.: Curve i, acid in green fruit when picked from the tree; curve 2, acid in collateral lots after ripening at 70° F. ; curve 3, acid in collateral lots after ripening at 40°; curve 4, acid in collateral lots after ripening at 30°. It was almost invariably true that fruit held at 30° F. had a lower acid content than that held in either 40° or 70° storage. This is especially interesting in connection with the fact that the sugars were much higher in the 30° storage lots than in the 40° pears. (Curves 2, 3, 4, fig. 5-8-) Aug. i6, 1920 Ripening and Storage of Bartlett Pears 493 In the latest picked lots from all sections the acid was very nearly the same, both in the green fruit and in the ripened fruit, regardless of the temperature of storage used. The acid content seems to become more nearly stabilized in the late season. It will be noted, however, that whereas the acid content in the California pears was very low at this time, in the fruit from the northern regions it was higher than at any other time during the season. The question as to why the acid content should remain more nearly constant in late-picked than in early picked fruit is naturally suggested by these results, but until something more is known of the synthesis of the acids and the r61e they play in fruit and plant respiration a solution seems improbable. ALCOHOL-INSOIvUBIvE, ACID-HYDROLYZABLE; REDUCING SUBSTANCES Results of analyses made as these were, by hydrolyzing the residue from an alcohol extraction with dilute acid and determining the reducing substances present, have usually been reported as starch. That this may be very misleading is showTi by the fact that ripe pears contain no starch, as proved by iodin tests, yet the residue from the alcohol extraction after being hydrolized contained a considerable amount of reducing material. It is almost certain that an equal or even greater amount of such material found in the green fruit is also made up of substances other than starch. For this reason the percentage weight of this group of substances has been figured as dextrose, and the figure includes starch, together with certain hemicelluloses, galactans, pectin materials, etc., which may be present in varying amounts. A study of the data presented in Tables I to IV shows that these reducing substances run highest in the earliest lots when first picked from the tree. There is a decreasing amount in the green fruit at successive pickings, until the last lots contain only about two-thirds as high a percentage of these substances as do the earliest pickings. There is a very marked drop in the amount of alcohol-insoluble, acid- hydrolyzable reducing substances present in the storage-ripe fruit as compared to the similar lots when picked. Compare column 15, green, with columns 16 and 18, ripe. This, of course, is natural, since all the starch and probably some of the other material have disappeared. It is interesting to note, however, that there is also a decrease in the amount of these reducing substances in the ripe fruit from late pickings as com- pared to ripe fruit from early pickings. This decrease in many cases amounts to 50 per cent of the total and seems to indicate that as the fruit develops on the tree much material other than starch changes over to sugar or is in condition to change over after picking. These results are interesting when considered in connection with the findings of Lewis, Mumeek, and Cate (9) on the decreasing resistance of 494 Journal of Agricultural Research voi.xix, No. w the pear tissue to pressure as the season advances. The pectose material is generally thought to be largely responsible for the thickening and cementing together of the cell walls and hence for the firm texture of fruit. The association of the decrease in amount of this and related material with decreasing resistance of the tissue to pressure is evidence in support of this theory. The temperature at which the fruit was stored has no marked influence on this material. A comparison of the lots picked at the same time and stored under the different temperature conditions (Tables I-IV, columns 16-18) shows little variation. influence; of time of picking and temperature of storage upon percentage of dry weight In this report, sugars and acids have been figured to percentage of the wet weight, as it is considered that wet weight rather than dry weight percentage will give the most accurate index of quality. The data for total dry matter in the fruit are of much importance, however, especially in connection with the pear dehydration industry, and for the purpose of throwing light upon the question of how much shrinkage, due to loss of moisture, occurs in the fruit during storage. From the data on dry weights presented in Tables I to IV it will be noted that while considerable variation seems to occur in various indi- vidual lots, one or two things stand out as of special interest. In the California fruit, in which the first pickings were made much in advance of the commercial season and when the fruit was very immature, the percentage of dry weight was higher in the earliest lots than it was in those lots picked during the main commercial shipping season, a month later. (Tables I and II, columns 19-22.) Toward the end of the season, however, the percentage of dry matter increased until the last pickings gave the highest dry-weight figures of all lots. The first pickings of the Oregon and Washington fruit (Tables III and IV) were made at a some- what later relative date than the first pickings from California, so the fact that the earliest pickings show a low dry weight is in accord with the data for California sections. Thus it is at once apparent that, for purposes of dehydration, pears left on the tree as long as possible will give not only the greatest tonnage, because of the size of the fruit, but will also give the greatest weight of the dried product per pound of green weight. Consequently, it is of special importance that pears intended for dr5dng be left on the trees as long as possible. If the dry weight of the fruit at the time of picking is compared with the dry weight of the same lots when they come from storage fully ripe it is seen that for well-matured fruit there is very little moisture loss dur- ing storage. A comparison of the earliest lots from Suisun and from Sacramento (Tables I and II) is interesting in that the Sacramento fruit was wrapped, whereas that from Suisun was loose in the box and un- wrapped. There is no increase in dry weight in the Sacramento fruit, Aug. i6, 1920 Ripening and Storage of Bartlett Pears 495 while the early lots from Suisun show a marked increase during storage. This indicates the value of wrapping in preventing loss of moisture from fruit. Examination of somewhat later lots picked during the commercial season shows no increase in dry weight while the fruit is in storage, and in many cases it shows an actual decrease. All the storages used were comparatively high in humidity, otherwise there might have been a loss due to more rapid evaporation from the fruit. An examination of the lenticels of the fruit of the dififerent lots was made as the fruit was freshly picked. A number of pears were put in methylene blue solution and after soaking a short time were removed and the lenticels examined under a microscope. It was found that the methylene blue readily penetrated the lenticels of the immature, early picked fruit. Fruit picked at the opening of the commercial season, how- ever, had a layer of brown, suberized tissue formed in the lenticel, which prevented the penetration of the blue solution. Later in the season pears immersed for a considerable length of time and then rinsed in water showed only a very faint blue ring about the outside of the len- ticel. The corky layer had apparently almost completely stopped pene- tration of the solution. Even when an immersed pear was placed under reduced pressure for a time and then under full atmospheric pressure the solution did not penetrate the lenticels. With practice, this condition of the lenticels can be detected by the brown color of the corky growth without the use of a microscope and dye solution. It appears that this change in the lenticels may be a valuable aid to present methods of determining when the fruit is in condition to pick and handle without danger of shriveling or wilting. EFFECT OF TIME OF PICKING UPON LENGTH OF TIME FRUIT MAY BE STORED Table V shows the number of days between the time the fruit was picked from the tree and the time of full yellow ripeness. The Yakima fruit is not included, since the number of days in transit and the fact that one lot was delayed en route makes an accurate comparison impos- sible. Table V. — Number of days required for fruit to become soft, yellow ripe at different temperatures of storage Sacramento, Calif. Date of picking. June 12 18 July 5 12 Aug. 13 Num- ber of days at 70° F. 14 14 14 14 Num- ber of days at 40° F. 45 32 33 24 Suisun, Calif. Date of picking. June 10. July I . 10. 22. Aug. 6 . Num- ber of days at 70° F 15 14 13 12 Num- ber of days at 40° F 49 41 32 28 Medford, Oreg. Date of picking. July 19. Aug. 8. 28. Num- ber of days at 70° F Num- ber of days at 40° F. 31 26 23 496 Journal of Agricultural Research voi. xix, No. 10 The variations in the length of time required for the fruit from the different locaHties to become ripe in 40° F. storage may be due in part to the different lengths of time spent in transit to place of storage. From this it is apparent that the results attained are similar to those found by other investigators — namely, that at the higher temperatures of storage, early picking gave somewhat longer keeping time than later picking. It has been impossible in this work to determine the relative keeping time at temperatures lower than 40° F. because of the necessity of removing the fruit from storage before it reached a full ripe condi- tion. At the 40° F. storage it was found, however, that the early fruits tended to scald and become brown rather than to ripen in good condi- tion, while the later pickings ripened to full yellow and prime condition with practically no scald. Another very important observation was that although late-picked fruit tends to become yellow more quickly than early picked lots, it remains in firm, prime eating condition for a much longer period after becoming yellow than the fruit picked early. GENERAI^ DISCUSSION OF RESULTS AS APPLIED TO COMMERCIAL HANDLING The disposition of the commercial pear crop of the Pacific coast may be grouped under three divisions, which include practically the entire output — namely, (i) fresh shipment, for consumption as fresh fruit or for home canning; (2) commercial canning; and (3) drying or dehydra- tion. The method of handling must, of necessity, be varied consider- ably, depending upon which of these methods of marketing is to be followed. When pears are to be shipped fresh, certain factors other than those which determine the very highest quality of fruit must be considered. Fruit picked comparatively early in the season will remain sound some- what longer, even at the lowest temperatures that it is possible to secure while the fruit is in transit, than will that picked too late; and this must always be an important consideration in determining the time to pick for fresh shipment. It must be remembered, however, that late- picked fruit is richer in sugar and of much higher dessert quality than fruit picked and shipped very early. Furthermore, while late-picked fruit, especially in the relatively high temperatures necessary in cars in transit, comes to prime eating condition in a shorter length of time, it remains in prime condition for a longer period, a consideration of much importance to the retail trade. In the cannery and dehydrated fruit trades, it is possible to sacrifice something in keeping quality for a higher dessert quality product. Most of the fruit is utilized near the point of production. In the can- nery industry the largest problem is to secure a good product and at the Aug. i6. 1920 Ripening and Storage of Bartlett Pears 497 same time to plan so that the tremendous tonnage that comes on within a short period is utilized before the fruit becomes overripe and breaks down. Almost every year canners lose a considerable quantity of pears because the fruit becomes overripe before the cannery can handle it. The first consideration of the canner should be the securing of a high quality product by leaving the fruit on the trees until well devel- oped. Pears picked very early are low in the natural fruit sugars and are of very inferior quality, whether eaten fresh or canned. A high- grade canned product can be secured only by using a high quality of fruit. If this is done, it becomes practically necessary for the cannery man to store part of his season's supply. If certain conditions of storage are carried out, the keeping of Bartlett pears in storage, even up to two months, and still securing a high-quality product is a practical certainty. These conditions may be summarized briefly as follows : (i) Use only well-developed fruit for storage. Early pickings tend to "scald" or turn brown and decay and break down much faster when removed from storage. (2) Put fruit into storage immediately after it is picked. The maxi- mum time that should elapse between picking and storing should not be more than three or four days. The cannery man will know the capacity of his plant; and, if more tons are being picked each day than he can handle, unless some go directly into storage, he can be sure that his cannery will be "flooded" when the fruit ripens. The fruit should go to the storage as soon as picked, rather than when it begins to soften. Much loss in pears in cold storage occurs because the fruit is in an almost soft-ripe condition when put in. (3) Fruit should be cooled as quickly as possible after being placed in storage. It is especially desirable that a room with a large amount of direct expansion or brine piping be used, so that the temperature can be reduced quickly to 30° F. The fruit will cool somewhat more slowly than the air, although, if the fruit is loose in lug boxes, it will follow the air temperature rather closely. (4) An even temperature should be maintained. If the storage rooms are large, it will be well worth while to use certain rooms for cooling down thefruit when it arrives, after which it may be transferred to other rooms for holding. This eliminates putting warm fruit into a room in which other fruit, already cooled, is being held. While this necessitates an extra handling, it is well worth while if it is desired to hold the fruit for some time. Especially is this system desirable if certain rooms having greater cooling capacity can be utilized for this precooling. (5) The temperature should be held down to 28° or 30° F. if a long storage period is desired. Well-developed Bartlett pears will store at that temperature, ripen in excellent condition if removed at any time up to two or three months, and give a high-quaUty product. If it is desired 498 Journal of Agricultural Research voi. xix. no. 10 to hold the fruit for only a few weeks, somewhat higher temperatures are permissible; but even for short storage periods a low temperature, followed by the removal of the fruit and ripening at outside air tempera- ture, gives a better product. (6) The cooling capacity of the storage plant should not be overtaxed. It is possible in the case of Bartlett pears to "store on the tree" to a very marked extent. Two weeks' time on the tree makes only a small differ- ence in the length of time pears will remain sound after removing from the tree, so for cannery trade it is not necessary to pick the entire crop within a very short time. Of course, other factors, such as amount of drop, load on the trees, etc., must be considered. The foregoing suggestions presuppose a very close working agreement between producer, canner, and cold storage; and this is essential for successful handling of Bartlett pears through cold storage. The fruit must be sent to the storage plant quickly if it is to be held in storage, and the cooling capacity must be such that the fruit can be cooled down within a short time. The temperature and storage recommendations apply only to Bartlett pears, since other varieties have been found to give different responses under storage treatment (8, 9). For the dehydration of Bartlett pears, if a drying plant is used, the same principles apply as for canning. On the other hand, if sun drying is used, the problem is much simplified, as the fruit can be handled in almost any quantity within a short time. For drying, however, it is of twofold importance that the fruit remain on the trees as long as possible, for the quality is not only improved but the accumulation of sugars gives an increase in the weight of dried product per pound of green fruit. SUMMARY There is a marked and quite uniform increase in total sugar in Bartlett pears from early summer until after the time of the close of the commer- cial picking season. The increase during the latter part of the season is mainly due to an accumulation of sucrose, while the earlier increase is due mainly to reducing sugar. A distinct relationship was found between the total amount of sugar present in the ripe fruit and the temperature of the storage at which it had been held from the time of removing from the tree until ripe. Pears ripened at 70° F. contained the highest percentage of sugar, those ripened at 40° possessed the lowest total sugar content, and those held at 30° for from 6 to 14 weeks and then ripened at room temperature were inter- mediate in amount of total sugar. There was no marked relation between temperature of storage and relative amount of sucrose and reducing sugar. Percentage of titratabie acid in the fruit tended to decrease in fruit from the California sections as the season advanced, while it tended to Aug. i6, 1920 Ripening and Storage of Bartlett Pears 499 increase in that from Oregon and Washington. There was an increase in acid between the time of picking and the time of full ripening of the fruit when held at 70° F. There was much less acid in fruit ripened at 40° than in that ripened at 70°, and still less in fruit that had been held at 30°. The acid content of the fruit that was allowed to become well matured on the tree remained nearly constant during storage, There was a progressive reduction in the alcohol-insoluble, acid-hydro- lyzable reducing material as the season advanced, not only in the fruit fresh picked from the tree, but also in the same fruit after ripening. There is a marked reduction in these substances between the time when the fruit is first picked and the time when the same fruit becomes ripe. The percentage of total solids is lowest at about the opening of the commercial season. This tends to increase with the accumulation of sugar in the late-picked lots. With proper precautions of picking, handling, and storing, Bartlett pears can be held two to three months in storage and then taken out in good condition. LITERATURE CITED (i) Clark, W. Blair. 1917. A SAMPLING PRESS. In Jour. Indus, and Engin. Chem., v. 9, no. 8, p. 788-790. (2) CruESS, W. v., and StoxE, P. M. I916. PRELIMINARY OBSERVATIONS ON THE RIPENING OF BARTLETT PEARS. In Mo. Bui. State Com. Hort. [Calif.], v. 5, no. 12, p. 425-429. (3) Dunbar, P. B.. and Bigelow, W. D. 1913. THE ACID CONTENT oP FRUITS. (Abstract.) In Science, n. s. v. 38, no. 983, p. 639-640. (4) EWERT. 1910. DIE KORRELATIVEN EINFLtJSSE DES KERNS BEIM REIFEPROZESS DER FRUCHTE. In Landw. Jahrb., Bd. 39, Heft 3, p. 471-486, pi. 12-13. (5) Gore, H. C. 1911. STUDIES ON FRUIT RESPIRATION. U. S. Dept. Agr. Bur. Chem. Bui. 142, 40 p., 17 fig. (6) Kelhofer. 1899. WEITERE UNTERSUCHUNGEN tJBER DIE VERTEILUNG VON ZUCKER, SAURE UND GERBSTOFF IN DEN BIRNENFRUCHTEN. In VI. & VII. Jahresb. Deut. Schweiz. Vers. Sta. u. Scliule Obst-, Wein- u. Gartenbau Wadensweil, 1895/96-1896/97, p. 68-71. (7) KULISCH, P. 1892. BEITRAGE ZUR KENNTNISS DER CHEMISCHEN ZUSAMMENSETZUNG DER AEPFEL UND BIRNEN MIT BESONDERER BERUCKSICHTIGUNG HIRER VERWENDUNG ZUR obstweinberEitung. In Laiidw. Jahrb., Bd. 21, p. 427-444. (8) Lewis, C. I., Macness, J. R., and Cate, C. C. 1918. preliminary report OF PEAR HARVESTING INVESTIGATIONS IN ROGUE RIVER VALLEY. Oreg. Agr. Exp.' Sta. Bui. 154, 24 p. (9) MuRNEEK, A. E., and Gate, C. C. 1919. pear HARVESTING AND STORAGE INVESTIGATIONS IN ROGUE RIVER VALLEY. (Second report.) Oreg. Agr. Exp. Sta. Bui. 162, 39 p., 10 fig. 500 Journal of Agricultural Research voi. xix, no. lo (10) Powell, G. Harold, and Fulton, S. H. 1903. COLD STORAGE, WITH SPECIAL REFERENCE TO THE PEAR AND PEACH. U. S. Dept. Agr. Bur. Plant Indus. Bui. 40, 26 p., 7 pi. (partly col.). (11) RiTTER, George. 1910. UEBER DEN CHEMISCHEN REIPUNGSPROZESS DER FRtJCHTE, MIT BESON- DERER BERiJCKSICHTIGUNG DES OBSTES. In Deut. Obstbauztg., Jahrg. 1910, Heft 31, p. 429-435- (12) RivifiRE, Gustave, and Bailhache, Gabriel. 1908. 6TUDE relative a la PROGRESSION ASCENDANTE DU SUCRE ET A LA PROGRESSION DESCENDANTE DE L'aCIDIT^, DANS LES FRUITS POIRIER, DEPUIS LEUR FORMATION JUSQU'a LEUR MATURITl^. In JoUT. Soc. Nat. Hort. France, s. 4, t. 9, p. 284-289. (13) 1910. DE l'iNFLUENCE DES FEUILLES QUI ACCOMPAGNENT IMM^DIATEMENT LES FRUITS DU POIRIER, SUR LEUR ACCROISSEMENT EN POIDS ET SUR LEUR COMPOSITION CHiMiQUE. In JouT. Soc. Nat. Hort. France, s. 4, t. 11, p. 678-680. (14) Shamel, a. D. 191 7. some observations upon the relation of humidity to the ripening AND STORAGE OF FRUITS. In Mo. Bul. State Com. Hort. [Calif.], v. 6, no. 2, p. 39-41. (15) Stubenrauch, a. v., and Ramsey, H. J. 1913. BARTLETT PEAR PRECOOLING AND STORAGE INVESTIGATIONS IN THE ROGUE RIVER VALLEY. In U. S. Dept. Agr. Bur. Plant Indus. Circ. 114, p. 19-24. (16) Taylor, R. H., and Overholzer, E. L. 1919. SOME EFFECTS OF HIGH TEMPERATURE AND HUMIDITY UPON THE KEEPING QUALITY OF BARTLETT PEARS. In Mo. Bul. State Com. Hort. [Calif.], V. 8, no. 3, p. 118-125. (17) Thompson, Finnan, and Whittier, A. C. I913. FORMS of sugar FOUND IN COMMON FRUITS. In PrOC, SoC. Hoft. Science, 9th Ann. Meeting, 1912, p. 16-22. FURTHER DATA ON THE ORANGE-RUSTS OF RUBUS By L. O. KuNKEL * Pathologist, Cotton, Truck, and Forage Crop Disease Investigations, Bureau of Plant Industry, United States Department of Agriculture In 1 91 6 the writer showed that there exists in the United States two orange-rusts on species of Rubus (i).^ Morphologically these rusts closely resemble each other in their caeoma stages, but in the behavior of the orange spores when germinated and in life cycle they were shown to differ. During the last two seasons further observations were made on the orange-rusts, and it is the object of the present paper to report the results obtained in this study. Atkinson (2) has performed some experiments which to him seemed to indicate that there is only one orange-rust on species of Rubus in the United States. He admits that the orange spores show two distinct methods of germination but attributes this to the influence of tempera- ture. According to his view, promycelia are produced at high tem- peratures and germ tubes at low temperatures. He suggests that this may explain the behavior of the orange-rusts in different parts of the country. In the north and in mountainous regions where the spring temperatures are relatively low the aeciospores produce germ tubes, while in southern sections of the country where temperatures are high they produce promycelia. The writer (7) has previously rifeported the behavior of the aeciospores of the two orange-rusts when germinated side by side at a room tem- perature of about 25° C. This experiment seemed to show conclusively that the aeciospores of the two rusts differ in manner of germination. Nevertheless, in view of Atkinson's results some further germination tests have been made. GYMNOCONIA INTERSTITIALIS ON BLACK RASPBERRY In the fall of 191 6 the writer collected the telia of Gymnoconia inter- stitialis on leaves of wild black raspberry plants growing on the Virginia side of the Potomac River near Washington, D. C. These telia showed that the long-cycled rust is present in the locality just mentioned. Since that time many collections of orange-rust have been made from both wild and cultivated Rubus plants in the vicinity of the city of Washing- ton. A study of these specimens has shown that the rust on the black raspberry is always long-cycled while the rust on the blackberry and • The writer wishes to acknowledge here the help he has received from many colleagues who have offered suggestions or have sent him living specimens of the rusts. ' Reference is made by number (italic) to *' Literature cited," p. 512. Journal of Agricultural Research, Vol. XIX, No. 10 Washington, D. C. Aug. 16, 1920 Qt Key No. G-202 C02 Journal of Agricultural Research voi. xix, no. lo dewberry plants, so far as has been observed, is short-cycled in this region. Wild blackberry and black raspberry plants occur abundantly along the Potomac River in both Maryland and Virginia. They are frequently intermingled with each other, and often both are infected with orange-rust. During the springs of 191 7 and 191 8 the two rusts were many times found growing close to each other, and during both seasons the telia of Gymnoconia were found occurring sparingly on leaves of wild black raspberry plants. The telia were always found on or near those plants that had borne caeomas of Gymnoconia. They were never found on any blackberry host. Many cultivated blackberry and black raspberry fields in the vicinity of Washington are troubled with orange- rust. In every instance the germination tests have shown that the raspberry plants are infected with the long-cycled rust. Rust found in the cultivated blackberry fields is always the short-cycled form. Plate 92 shows the way the aeciospores of the two rusts germinate on Beyerinck agar at room temperature (about 25° C). The spores shown in Plate 92, A, were taken from leaves of wild black raspberry at West Falls Church, Va. They have produced long germ tubes. Those shown in Plate 92, B, were collected at the same place on wild blackberry. They have produced promycelia-bearing sporidia. INFLUENCE OF TEMPERATURE ON GERMINATION In order to study the effect of temperature on germination numerous collections were made from both wild and cultivated blackberry and raspberry plants. The spores were germinated in Petri dishes on water and on Beyerinck agar. The cultures were incubated at temperatures varying by 5° intervals and ranging from 0° to 30° C. None of the spores of either rust germinated at 0°. At 5° excellent germination was obtained, but growth was slow. At all of the higher temperatures — 10°, 15°, 20°, 25°, and 30° — the spores of the two rusts germinated equally well. It was noted that at low temperatures such as 5° and 10° the spores of the long-cycled rust began to germinate somewhat sooner than those of the short-cycled rust. Germination in cultures of both kinds of spores took place more rapidly at 30° than at any of the lower temperatures. Fewer spores germinated, however, at this tem- perature than at the lower temperatures. The spores of both rusts germinated well at all the temperatures tested between 0° and 30°. Spores taken from blackberry leaves always produced promycelia, while those from the black raspberry leaves produced long germ tubes. Mature aeciospores of the two rusts collected at the same time and often within a few feet of each other and incubated at the same tem- peratures and on the same media always showed the same differences in manner of germination. The promycelia produced by the spores from blackberry leaves are typical in every way. They become divided into Aug. i6. 1920 Further Data on the Orange-Rusts of Ruhus 503 four or more cells, and usually four of these contain one nucleus each. Each nucleated cell is capable of producing a sporidium. The germ tubes arising from spores borne on raspberry leaves are long and sinuous. By suitable methods of staining they have been shown to contain two nuclei. At an early stage in germination they may be distinguished from promycelia by their smaller diameter and more rapid longitudinal growth. Temperature, within the range tested, has no ejBfect on the manner in which the aeciospores of these two rusts germinate. In the vicinity of Washington, D. C, at Mountain Lake, Va., and at French Creek, W. Va., both rusts occur side by side under the same conditions of temperature and climate. The writer is, therefore, unable to accept the theory that temperature determines whether spores of a given orange-rust specimen will produce germ tubes or promycelia. COLOR OF SPORES IN MASS The finding of the two orange-rusts growing within a short distance of the laboratories of the Bureau of Plant Industry made it easy for the writer to compare them more carefully than was possible when they had to be brought from dififerent parts of the country. The comparison of the rusts as they occur side by side on their living hosts has brought to light certain dififerences that were not noticed earlier. One of the most important of these is the color of the spores in mass. It soon became evident that the spores of the short-cycled rust are lighter in color than those of Gymnoconia. The spore colors of the two rusts were matched on Ridgeway's color chart. According to this chart the spores of the short-cycled rust are cadmium orange, while those of the long-cycled rust are xanthine yellow. These two colors do not differ greatly from each other and stand side by side in the chart. Nevertheless they can be easily distinguished after one has once noted the difference between them. It is surprising that this difference was not seen earlier, especially since account was taken of the color of the spores in mass. It seems that failure in this regard was due to the fact that the color of the spores of both rusts begins to fade within a few weeks after they are collected, and differences in shade of color were attributed to fading. It was at first thought that the difference in color between the spores on raspberry and on blackberry leaves might be due to the difference in host. In order to test this hypothesis a number of collections were made during the spring of 1917 and 19 18. The long-cycled rust was collected on both wild and cultivated black raspberry at French Creek, W. Va. It was collected on wild blackberry (identified as Ruhus alleghaniensis) at Mountain Lake, Va. Numerous collections were made in the Adirondack Mountains near Old Forge, N. Y., and in the White Mountains near Glen and Jackson, N. H. It was also collected 504 Journal of Agricultural Research voi. xix. No. 10 on black raspberry at Rouses Point, N. Y. Collections of the short- cycled rust were made on wild dewberry and wild blackberry at French Creek, W. Va., on wild dewberry at Mountain Lake, Va., and on wild blackberry and dewberry plants at many other points. The material collected in 191 7 and 191 8 was brought together for comparison at the end of each season and before there was serious fading in the color of the spores. This comparison has shown that for the material at hand the two orange-rusts exhibit the same color differences regardless of the hosts on which they occur or the localities from which they are collected. The color difference makes it possible to identify the two rusts in the field without resort to spore germination. Plate D illustrates the difference in the color of the spores in mass. Figure i shows an infected black raspberry leaf, figure 2 an infected blackberry leaf. MORPHOLOGY OF AECIOSPORES In a former paper the writer (7) has pointed out that no morphological differences could be observed between the aeciospores of certain speci- mens of the two orange-rusts. While this statement was true for the specimens under study, it does not hold when larger numbers of speci- mens of the two rusts are compared. A study of more than 100 different collections has shown that the spores of the two rusts differ considerably from each other both in size and in shape. While a few specimens may not reveal this fact, a more extended study shows that the aeciospores of the two rusts are, on the whole, morphologically different. In order to show this difference more clearly than is possible by description, an outline drawing has been made of a few typical aecio- spores from a number of different specimens of the two rusts. The drawings were made with the aid of a camera lucida. The same magni- fication was used for all spores, so that the different drawings may be readily compared. There is always a certain amount of variation in the size and shape of the spores of a given specimen. This is greater for some specimens than for others, and it was not always easy to select spores that would be typical. Before material for drawing was chosen, spores from several mature caeomas on each specimen were transferred to separate drops of water on glass slides. They were then observed under the microscope, and a group was finally chosen that seemed to be typical for the specimen in question. Table I gives information regarding the place and time of collection, host, manner of germination, and color of spores in mass for most of the specimens collected in 19 17 and 191 8. Numbers given in the last column of the table indicate the drawings in Plates 93 and 94, which show an average sample of spores for each specimen. Aug. 16, 1920 Further Data on the Orange-Rusts of Ruhus 505 Table I. — Place and time of collection, host, manner of germination, and color of the aeciospores in m.assfor most of the specimens collected in igi^ and igi8 Place of collection. Falmouth, Mass Arlington, Va Massachusetts Berlm.Md Hyattsville, Md West Falls Church, Va. Auburn, Ala Fayetteville, Ark Potomac Heights, D. C. Morgantown, W. Va. . Blacksburg, Va Ithaca, N.V Blacksburg, Va West Falls Church, Va. Gainesville, Fla Ithaca, N. Y Chico, Calif Ithaca, N. Y Arlington, Va Berkeley, Calif Hanunonton, N. J Fayetteville, Ark West Falls Church, Va. French Creek, W. Va. Congress Heights, D. C. Connellsville, Pa Chico, Calif Vienna, Va Bryan, Ohio Cameron, N. C Athens, Ohio. . . Blacksburg, Va. Chico, Calif , Arlington, Va , Ithaca, N.Y Mountain Lake, Va. . French Creek, W. Va Blacksburg, Va , Do Vienna, Va Ithaca, N.Y... . Blacksburg, Va. Janassee Junction, Ga. Thunderbolt, Ga Hyattsville, Md Do Blacksburg, Va Vienna, Va Willard, N. C Connellsville, Pa Auburn, Ala Butte Creek Canyon, Calif. Stark, Fla Blacksburg, Va Thunderbolt, Ga Hyattsville, Md Blacksburg, Va Hammond, La Orlando, W. Va Ithaca, N. Y Blacksburg, Va West Falls Church, Va Arlington, Va Do French Creek, W. Va. Madrid, Me French Creek, W. Va. Glen, N. H Time of collection. June 20: June, June 27 June 8 June IS May 21 Apr. 27 Jime 7 June, June 7 June II Jtme 4 May 25 May 21 Mar. 25 June, May I Jime 4 June, Mar. 30 June 20 June 7 May 21 Jime 8 May 20 June 6 May 19 May 21 Jtme 12 June 4 June II do. May I Jime, do. June II June 8 June II do. do. June 4, 19 June 11,19 May 14, 19 Mar. 17,19 June IS, 19 do. June II, 19 May 29, 19 June 26, 19 June 8,19 Mar. 4, 19 May 19, 19 191 7 June 1 1 , 19 Mar. 17, 19 June IS. 19 May 2s, 19 Mar. 29, 19 June «, 19 June 4, 19 June 12, 19 May 21, 19 June, 19 do. June 8,19 July 3,19 June 8,19 June 22, 19 Host. Wild dewberry Wild blackberry ....do Cultivated blackberry. Wild dewberry ....do ....do Wild blackberry ....do Cultivated blackberry, variety Eldorado. Cultivated blackberry, variety Iceberg. Wild blackberry ....do ....do Rubus cuneifolius Wild blackberry R. ursinus Wild dewberry Wild blackberry R. parviflorus Cultivated blackberry. Wild blackberry ....do Wild dewberry Wild blackberry ....do R. ursinus Wild blackberry Cultivated blackberry, Wild blackberry ....do Cultivated blackberry, variety Early King. R. ursinus Wild blackberry Wild dewberry R. americanus Wild dewberry Cultivated blackberry . Cultivated blackberry, variety Mersereau. Cultivated blackberry . Wild blackberry Cultivated blackberry, variety Ancient Britain. Wild blackberry R. procumbens Wild dewberry Wild blackberry Cultivated blackberry Wild blackberry do do do R. ursinus Wild blackberry ....do.. R. hispidus Wild blackberry ....do do do do do Cultivated blackberry Wild blackberry do do do Black raspberry. R. nigrobaccus. . . Manner of germination. Promycelia. ...do ...do ....do ....do ....do ....do ....do ....do .do. .do. .do .do .do .do .do .do .do .do .do .do .do .do .do .do Color of spores in mass. Cadmium orange. do do do do do do do do .do. ....do ....do ....do Germination very poor. Promycelia... do ....do .do .do .do .do .do .do .do .do. .do. .do. .do. .do. .do. .do. .do. .do. .do. .do. .do. .do. .do. .do. .do. .do. .do. .do. .do. .do. .do. .do .do .do do Germ tubes. do do .do. .do. .do. .do. .do. .do. .do. .do. .do. .do. .do. .do. .do. .do. .do. Plate No. do do do Color faded. Cadmium orange. do do ..do. ..do. ..do. ..do. ..do. ..do. ..do. .do. .do. .do. .do. .do. .do. .do. .do. .do. -do. .do. .do. .do. do do do do do do do do do do do do do Xanthine yellow do do 93 -fig- 1 93 > fig- 2 93 r fig- 3 93 > fig- 4 93. fig- S 93. fig- 6 93, fig- 7 93, fig- 8 93, fig- 9 93, fig. 10 93, fig- II 93, fig. 13 93. fig- 13 93, fig. 14 93. fig- IS 93, fig. 16 93. fig- 17 93, fig. 18 93, fig. 19 93 , fig. 20 93, fig- 21 93, fig. 23 93. fig- 23 93, fig. 24 93jfig-2S 93, fig. 36 93. fig- 27 93, fig. 28 93, fig- 29 93, fig- 30 93, fig- 31 93, fig- 32 93, fig- 33 93. fig- 34 93, fig- 35 93, fig- 36 93, fig- 37 93, fig- 38 93, fig- 39 93, fig. 40 93, fig. 41 93. fig- 42 93. fig- 43 93. fig 93, fig 93. fig' 93, fig' 93, fig 93, fig 93. fig 93. fig 93. fig 93. fig 93. fig 93, fig' 93, fig' 93, fig' 93, fig' 93. fig 93, fig' 93. fig 93. fig 93. fig 93. fig 93 , fig 94- fig' 94, fig 94, fig 183719°— 20- 5o6 Journal of Agricultural Research Vol. XIX, No. lo Table I. — Place and time of collection, host, manner of germination, and color of the aeciospores in m^issfor most of the specimens collected in igiy and igi8 — Continued Place of collection. Time of collection. Host. Manner of germination. Color of spores in mass. Plate No. West Falls Church, Va. Vienna, Va OldForge, N. Y Glen,N.H Moimtain Lake, Va. . West Falls Church, Va. Portland, Me Madison, Wis Phillips, Me Portland, Me East Lansing, Mich. . . Sebago Lake, Me Michigan OldForge, N.Y Madison, Wis OldForge, N.Y West Falls Church, Va. Do Do Glen, N. H Mountain Lake, Va.. French Creek, W. Va. Portland, Me Smugglers Notch, Vt, Madrid, Me OldForge, N.Y Smugglers Notch, Vt. Bancroft, Wis Julet, N.Y French Creek, W. Va, OldForge, N.Y Madison, Wis Portland, Me Sebago, Lake, Me Smugglers Notch, Vt. Sebago Lake, Me Rouses Point, N.Y... Sebago Lake, Me Moimtain Lake, Va . . Boxmd Brook, N.J... May 21 do. June 27 June 22 June II May 21 June 24 June 4 July 3 June 24 June 30 June 23 June, June 26 July 17 June 27 May 21 July 4 May 21 June 22 June II Jime 8 Jvme 24 July II July 3 June 27 July II July, June 26 June 6 June 26 June 4 June 24 June 23 July II June 23 June 29 J une 23 Jime II June 17, 1918 Cultivated raspberry. do R. canadensis Wild blackberry R. alleghainensis Cultivated black rasp- berry. R. canadensis Wild blackberry do R. canadensis Wild blackberry R. triflorus Wild blackberry R. canadensis Wild blackberry R. canadensis Wild black raspberry Cultivated black rasp- berry. do R. canadensis R. alleghaniensis Black raspberry R. canadensis R. strigosus Wild blackberry R. canadensis do. Glen, N. H Jime 22, 1916 R. hispidus R. canadensis Black raspberry. . . R. canadensis Wild blackberry. . R. triflorus do R. canadensis R. triflorus Black raspberry. . . Wild blackberry . . R. alleghaniensis . . . Black raspberry. . . R. canadensis Germ tubes. do do do do do .do .do .do .do .do .do .do .do .do .do .do .do .....do do do do do do do do do do do do do do do do do do do do do No germina- tion. Germ tubes . Xanthine yellow do do do do do .do. .do. .do. .do. .do. .do. ."do. .do. .do. .do. .do. .do. ...do ...do ...do ...do ...do ...do ...do ...do ...do ...do ...do ...do ...do ...do ...do ...do ...do ...do ...do ...do ...do . . .do Color faded. 94) fig' 94. fig 94. fig 94) fig 94) fig 94) fig 94) fig 94. fig' 94. fig 94. fig 94) fig 94) fig' 94. fig 94) fig' 94. fig 94) fig 94) fig' 94) fig' 94) fig' 94) fig' 94. fig' 94) fig 94. fig' 94. fig 94. fig 94. fig 94. fig 94. fig 94) fig 94, fig 94) fig 94) fig 94) fig 94) fig 94) fig 94, fig. 94. fig' 94. fig 94. fig 94. fig 94. fig- 44 Plate 93 shows spores from 65 different specimens of the short-cycled orange-rust. Figures i to 44 on Plate 94 show spores from 44 different specimens of the long-cycled rust. The figures on these plates demon- strate that, on the whole, the aeciospores of the two rusts are mor- phologically different. The spores of the short-cycled rust are smaller than those of the long-cycled. They are also more angular and more elongated. Their shape is more irregular. It will be seen that the size and shape of the aeciospores from different specimens of the two rusts vary considerably. On this account it is not always possible by observing spores under the microscope to determine with certainty to which group a given specimen may belong. On the other hand, spore characters do make possible a rather accurate separation of specimens belonging to the two rusts. It is difficult to say just how accurate such determinations will be. Much depends on the specimens at hand and on the judgment of the one who undertakes such a task. The writer's determinations by this means have proved to be correct in about 85 Aug. 16. 1920 Further Data on the Orange-Rusts of Rubus 507 per cent of the specimens studied. By this method it is possible to identify with a fair degree of accuracy orange-rust specimens in herbaria long after the spores are dead and have lost their color. It must be remembered, however, that it is always necessary to have a liberal quantity of mature spores in order to make determinations of value. The figures on Plate 93 show the variation in the size and shape of spores from different collections of the short-cycled rust. The spores shown in most of the figures are relatively small and angular. Those shown in figures 9, 13, 39, 45, and 55 are large and round. They look like the spores of the long-cycled rust, but their color and manner of germination prove that they belong to the short-cycled rust. Figures 3, 10, 28, 29, 33, 35, 36, 44, and 57 show spores that resemble somewhat those of the long-cycled rust. The specimens from which these aecio- spores were taken can not be satisfactorily identified on the basis of spore characters. The spores shown in all of the other figures on Plate 93 are characteristic for the short-cycled rust; but even in such cases one can not be absolutely sure that they belong to this fungus, for occasionally a specimen of the long-cycled rust bears spores like those shown in figure 35 of Plate 94. The aeciospores shown in this figure are small and angular; they do not look like spores of the long-cycled rust. Spores shown in figures 4, 6, 8, 9, 19, 25, 42, and 43 of Plate 94 resemble to a certain degree spores of the short-cycled rust. On the whole, however, aeciospores from different samples of the long-cycled rust are more uniform as regards size and shape than are those of the short-cycled rust. GENETIC RELATIONSHIP BETWEEN THE TWO ORANGE-RUSTS Although the two orange-rusts differ from each other in several char- acters, it must not be denied that they are alike in many respects. They are both systemic on species of Rubus. Their caeomas look much alike, and in many specimens the aeciospores are quite similar. These points of resemblance suggest a genetic relationship. Along with the further evidence that the two rusts are distinct and different from each other have come certain facts that strengthen this suggestion. In an earlier paper the writer mentioned finding a promycelium in a culture of aeciospores of the long-cycled rust. It was thought at the time that the spore producing the promycelium might have entered as a contamination. During the spring of 1917 and 1918 aeciospores of Gymnoconia collected in different parts of the country were germinated in great numbers. Each culture was carefully examined under the microscope. In many cultures only germ tubes could be found. A few promycelial germinations have been obtained, however, from spores of every collection of Gymnoconia which the writer made during the last two seasons. Sometimes such germinations are exceedingly rare, 5o8 Journal of Agricultural Research voi. xix, no. w but if enough spores are germinated the promycelia will be found. The spores of some collections produce them more often than those of other collections. In general it may be said that promycelia are pro- duced more abundantly by aeciospores collected late in the season. They are not entirely absent, however, from cultures made with spores collected early in the season. The possibility of these promycelia being produced by spores of the short-cycled rust that have contaminated the cultures has been excluded in most instances. In order to do this, the aeciospores were taken from caeomas that had not yet opened. Moreover, many of the spores were collected and germinated in parts of the country where the short-cycled rust is not known to occur. If the promycelia appeared only in cultures from aeciospores collected in the South where the short-cycled rust is abundant, mixed infection might offer a possible explanation. But since they also occur in cul- tures of spores collected in the North where the short-cycled rust has never been found, this explanation is unsatisfactory. In the vicinity of Glen, N. H., orange-rust has been collected each spring since 1913. Spores taken from a number of different places in this vicinity have been germinated, but the short-cycled rust has not been found. In both 1 91 7 and 191 8, cultures made at Glen were studied and found to contain a few promycelia. Promycelia have also been observed in cultures of the aeciospores of the long-cycled rust collected at Old Forge, N. Y., where several seasons' search has failed to reveal the presence of the short-cycled rust. They have been found in cultures of aeciospores taken from the black raspberry at Rouses Point, N. Y., French Creek, W. Va., and at points in the vicinity of Washington, D. C. Promycelia were also found in cultures of aeciospores collected at Mountain Lake, Va., on Rubus alleghaniensis. When spores of the short-cycled rust are incubated at room tem- perature (about 25° C.) on a favorable medium, they produce promycelia- bearing sporidia within 24 hours. Spores of the long-cycled rust placed under similar conditions produce long germ tubes within 24 hours. Promycelia have seldom been found in these cultures after so short a time. They occur in cultures of Gymnoconia only after a rather long period of incubation. They can usually be found after 3 or 4 days. In order to study the production of promycelia by the aeciospores of Gymnoconia, it is best to incubate cultures at a fairly low tem- perature. Temperatures varying from 10° to 15° are favorable. Promycelia are always slow to make their appearance in cultures of this rust. If incubation temperatures are high, many germ tubes die before they have time to develop into promycelia. Moreover, cultures kept at high temperatures are usually overgrown by saprophytic mold fungi and bacteria after a few days. Low temperatures check the growth of these organisms. Promycelia can be found most easily in Aug. i6, 1920 Further Data on the Orange-Rusts of Rubus 509 cultures of aeciospores kept at about 10° for a week or longer. It must be understood, however, that they are not produced very abun- dantly even under favorable conditions. Sometimes 1,000 germinated spores may be observed without the finding of a single promycelium, but usually several promycelia will be found for each 1,000 spores observed if the cultures are more than 4 days old. In some cultures they occur more abundantly. It is interesting to note that most of the promycelia developing in cultures of the aeciospores of the long-cycled rust are abnormal, though normal ones are also present. Many of the abnormal promycelia pro- duce one or more normal sporidia, and there can be no doubt regarding their true nature. The abnormal promycelia and their tardy appearance in the cultures seem to the writer to suggest that nuclear fusions and the subsequent reduction divisions are steps accomplished with diffi- culty in these spores. It would be highly interesting to study these phenomena cytologically, but the relatively small number of promycelial germinations makes such a task rather difficult. From a study of many abnormal promycelia the writer has come to recognize certain structures as indicating an attempt at the production of sporidia. Some of these are cross walls, branches, especially those having a diameter less than that of the germ tube, and sporidia-like processes borne on structures that show more or less resemblance to sterigmata. In cultures of the aeciospores of the long-cycled orange- rust it is possible to find all gradations between normal promycelia and germ tubes that can hardly be recognized as promycelia at all. In order to show some of the stages between these two extremes a few drawings have been made of abnormal promycelia. Figure 53 of Plate 94 shows a tube with two rather typical sporidia borne on typical sterigmata. No cross walls occur in this tube. Figure 49 shows a branched germ tube. A cross wall occurs just below the branch. No sporidia are borne on this tube, but there can be little doubt that this is an attempt at promycelium production. Figure 51 shows a germ tube with one cross wall and forked branches. One of these branches has produced a rather long club-shaped tube, while the other has developed into a sterigma-like process bearing a typical spo- ridium. A tube with one cross wall and several branches of small diameter is shown in figure 50. One of these branches is pointed like a sterigma and bears a spore that resembles a sporidium. Figure 45 represents a tube having one cross wall and several short branches. One of these branches is considerably enlarged toward its distal end and presents curves that suggest those of the normal sporidium. A short tube is sho%vTi in figure 47. This tube has one short branch which bears a sporelike body having curves that closely resemble those of a sporidium. The curves of the upper end of this body are especially like those of the 5IO Journal of Agricultural Research voi. xix, no. lo upper part of normal sporidia. Another short tube is shown in figure 46. This bears a branch with an enlarged end resembling a sporidium. A constriction at the point of the first bend in the branch would give rise to a fairly normal sporidium. Figure 48 shows a similar branch, but this time it arises from a very long germ tube. An unbranched tube is shown in figure 54. The diameter of the distal end of this tube is much less than that of the average diameter of germ tubes. Such a decrease in diameter frequently accompanies cross-wall production, branching, and other indications of an attempt at the production of sporidia. Figure 52 shows a tube with a branch of small diameter and one cross wall. The end cell has broken away from the remainder of the tube. It must not be supposed that abnormal promycelia are uncommon in cultures of the short-cycled orange-rust or in cultures of germinating teliospores in general. Abnormal promycelia much like those described above have occasionally been found in cultures of the short-cycled rust. They are not common, however, under ordinary conditions of germina- tion. In cultures of the aeciospores of Gymnoconia, on the other hand, most of the promycelia produced are abnormal. The production of promycelia by the aeciospores of Gymnoconia interstiiialis seems to the writer to be strong evidence that a close genetic relationship exists between the two orange-rusts. One of them is a typical short-cycled rust. It produces three kinds of spores : Spermatia, aeciospores, and sporidia. So far as the writer has observed it possesses one and only one life cycle. There is nothing unusual about this rust. The other orange-rust is long-cycled, but it is not a typical long-cycled rust. It is unusual in that it possesses two life cycles. In addition to the long cycle there is a much repressed short cycle, as shown by the occasional production of promycelia. We know that the germ tubes produced by these spores reinfect Rubus leaves. It is not known whether the sporidia can cause infection. Some of the sporidia have been seen to germinate. They appear normal in every way, and there seems to be no reason why they should not function. So far as the writer knows no one has yet observed the production of promycelia in cultures of the European orange-rust of Rubus. Both Fischer (4) and Lindfors (8) have recently studied the manner of germi- nation of the spores of this rust and have observed only germ tubes. Fischer, however, has shown a branch of small diameter coming from the end of one of his germ tubes. This suggests an attempt at promycelium production and leads the writer to believe that if large numbers of aeciospores of the European orange-rust are germinated and carefully observed promycelia will be found. The findings of an occasional promycelium in cultures of the aeciospores of Gymnoconia interstiiialis at once raised the question as to whether or not such a performance is usual among the rusts. It is not possible to Aug. i6, I920 Further Data on the Orange-Rusts of Rubus 511 get much information on the question from the literature on the germi- nation of rust spores. Most workers have not germinated aeciospores in large numbers, and a few promycelia in their cultures might easily have been overlooked. In order to settle this point it would be necessary to germinate the aeciospores of many different rusts in large numbers, and the writer has not undertaken this task. Nevertheless, it has seemed desirable to make a thorough study of the aeciospore germination of some other rust. For this study the aeciospores of Aecidium fraxini were chosen. These aeciospores are produced in large numbers and germinate readily on both water and Beyerinck agar. A. fraxini was found in abundance on black ash trees growing along the shore of Lake Champlain near Rouses Point, N. Y. Many cultures were made with aeciospores of this rust. The germinations were carefully observed, but not a single promycelium was ever found. Long, wavy germ tubes were produced. No cross walls or branches were observed. These spores produce germ tubes only. Promycelia in cultures of the aeciospores of Gymnoconia inter stitialis indicate that the two nuclei which ordinarily pass out into the germ tube and remain apart through many nuclear and cell divisions occasionally fuse in the spore or perhaps in the young germ tube. If we assume that reduction in chromosome number occurs here as in other promycelia and that the sporidia produced are capable of reinfecting the host, then G. inter stitialis has a double life cycle such as has not been demon- strated for any other rust. It is not believed that promycelia are commonly produced even in small numbers by the aeciospores of most rusts. On the other hand, it seems probable that other rusts will be found that possess double life cycles. Eriksson (j) has reported that the aecia of Aecidium graveolens which occur on species of Berberis are able to reproduce themselves, although they may also infect Avena elatior and give rise to Puccinia arrhenatheri. This strange behavior, which has never been accounted for, may be due to the production of promycelia by a certain number of the aeciospores. Recently Klebahn (5) reports that the aecia of Periderinium pini repro- duce themselves on the pine. He states that the aeciospores give germ tubes, but a further study may show that some of them produce promycelia. In an earlier paper (6) the writer expressed the opinion that the short- cycled orange-rust is more primitive than the long-cycled one. The fact that the long-cycled rust has a double life cycle is further evidence in favor of this view. Arthur (j) considers the diflferences between the two orange-rusts sufficient to place them in separate genera. Moreover, these genera are widely separated in his classification. It would seem that the evidence of a genetic relationship between these rusts should be given considera- tion in any natural system of classification. 512 Journal of Agricultural Research voi. xix, no. lo LITERATURE CITED i) Arthur, J. C. 1917. ORANGE RUSTS OF RUBUS. In Bot. Gaz., V. 63, no. 6, p. 501-515, map. 2) Atkinson, George F. 1918. SELECTED CYCLES IN GYMNOCONIA PECKIANA. In Amef. Jour. Bot., V. 5, no. 2, p. 79-83. 3) Eriksson, Jakob. 1898. sTUDiEN USER den hexenbesenrost der berberitie (puccinia arrhenatheri kleb.) In Beitr. Biol. Pflanz., Bd. 8, p. i-i6, 3 pi. Literatur, p. 14-15. 4) Fischer, Ed. 1916. MYKOLOGISCHE BEiTRAGE. In Mitt. Naturf. Gesell. Bern, 1915, p. 214-234, 2 fig. 5) Klebahn, H. 1905. KULTURVERSUCHE MIT ROSTPILZEN. XH. BERICHT (1903 UND I904). In Ztschr. Pflanzenkrank., Bd. 15, Heft 2, p. 65-ioS, 4 fig., pi. 3. 6) Kunkel, L. O. i914. nuclear behavior in the promycelia of caeoma nitens burrill AND PUCCINIA PECKIANA HOWE. In Amer. Jour. Bot., v. i, no. i, P- 37-47. pl- 3- Bibliography, p. 45-46. 7) 1916. FURTHER STUDIES OF THE ORANGE RUSTS OF RUBUS IN THE UNITED STATES. In Bul. Torrey Bot. Club, v. 43, no. 10, p. 559-569, 5 fig. 8) LiNDFORS, Thore. 1918. MYKOLOGISCHE NOTizEN. In Svensk Bot. Tidskr., bd. 12, no 2, p. 221-227, 4 fig. Further Data on the Orange-Rusts of Rubus Plate D I I ■■!■— W»t— Journal of Agricultural Research Vol. XIX. No. 10 AHoeni CoBattimare PLATE D I. — Infected black raspberry leaf covered with the caeomas of Gymnoconia inter- stitialis. The spores in mass are xanthine yellow, 2. — Blackberry leaf infected with the short-cycled orange-rust. The spores of this rust are cadmium orange in color. PLATE 92 Manner of germination of the spores of the two orange-rusts. The spores were collected at the same time, placed in Beyerinck agar, and incubated at about 25° C. A. — Spores taken from leaves of the black raspberry, showing long, wavy germ tubes. X38. B. — Spores taken from wild blackberry leaves, showing promycelia and numerous sporidia. X85. Further Data on the Orange-Rusts of Rubus Plate 92 //'ja^-#r-^. •^■^. r---'*-- v**^. .rte. \f^ .^^' B% - I"- , Journal of Agricultural Research Vol. XIX, No. 10 Further data on the Orange-Rusts of Rubus PLATE 93 0 o o-V ^-c? O-^'' ""oO -o'^o o 'OO o o on 0 o ^ X3 ^ ^" n" ^-'^ ■ o""^ ^^^o """O o O 32 a o C t5'6 0^ OO 0=' 0 c9 ^ O o 0 o o o ^o^ OJ^ O^ O O ^ =0 0 0 3x0 0 o 0 O Q - O^ o^ O^Q ^3, o c o o o°o 5 0 o"o° 0^ OO 0 oo..,o ooB Oo §p o06 -' 0 o"^ 0 ^ o 0 O 0 0 0 65 "O O '°o O ° O 0,63 0 U Q Journal of Agricultural Research Vol. XIX, No. 10 PLATE 93 Short-cycled orange-rust. X loo. I. — Spores collected at Falmouth, Mass., on wild dewberry. 2. — Spores collected at Arlington, Va., on wild blackberry. 3. — Spores crjllccted in Massachusetts on wild blackberry. 4. — Spores collected at Berlin, Md., on cultivated blackberry. 5. — vSpores collected at Hyattsville, Md., on wild dewberry. 6. — vSpores collected at West Falls Church, Va., rm wild dewberry. 7. — Sp<'jres Cfillected at Auburn, Ala., on wild dewberry. 8. — vSp^jres collected at Fayetteville, Ark., on wild blackberry. 9. — vSpores collected at Potomac Heights, D. C, on wild blackberry. 10. — Spores collected at Morgantown, W. Va., on cultivated blackberry, variety Eldorado. II. — Spores collected at Blacksburg, Va., on cultivated blackberry, variety Iceberg. 12. — Spores crjllected at Ithaca, N. Y., on wild blackberry. 13. — Sfxjres cfjllected at Blacksburg, Va., on wild blackberry. 14. — vSpores collected at West Falls Church, Va., on wild blackberry. 15. — Spores collected at Gainesville, Fla., on Ruhuy cunei/oUui. 16. — vSpores crjllected at Ithaca, N. Y., on wild blackberry. 17. — Spores collected at Chico, Calif., on R. uninuA. 18. — Sp^jres collected at Ithaca, N. Y., on wild dewberry. 19. — Spores collected at Arlington, Va., on wild blackberry. 20. — Spores collected at Berkeley, Calif., on R. parvijlorus. 21. — Spores crjllected at Hammonton, N. J., on cultivated blackV>erry. 22. — SfK^res collected at Fayetteville, Ark., on wild blackberry. 23. — Spores crjllected at West Falls Church, Va., rjn wild blackberry. 24. — Spores collected at French Creek, W. Va., cm wild dewberry. 25. — Spores crjllected at Congress Heights, D. C, fjn wild blackberry. 26. — Spores crjllected at Connellsville, Pa., on wilrl blackberry. 27. — Spores crjllected at Chioj, Calif., on R. uninui. 28. — Spores collected at Vienna, Va., on wild blackberry. 29. — Spores crjllected at Bryan, Ohio, on cultivated blackberry. 30. — Sjxires crjllected at Cameron, N. C, rjn wilrl blackberry. 31. — Spores crjllected at Athens, Ohio, on wild blackberry. 32. — vSpores collecterl at Blacksburg, Va., on cultivated blackljerry, variety Early King. 33. — Sfxjres collected at Chicrj, Calif., on R. uninus. 34. — Sprjres collected at Arlington, Va., on wild blackberry. 35. — Spores collecterl at Ithaca, N. Y., on wild dewberry. 36. — Spores crjllected at Mountain Lake, Va., on wild dewberry. 37. — Spores collected at French Creek, W. Va., on wild dewberry. 38. — Spores crjllected at Blacksburg, Va., on cultivaterl blackberry. 39. — SfKjres collected at BlacksVjurg, Va., rjn cultivated blackberry, variety Mer- screau. 40. — Spores collected at Vienna, Va., rjn cultivated blackberry. 41. — Spores collecterl at Ithaca, N. Y., on wild blackberry. 42. — Sprjres collected at Blacksburg, Va., on cultivaterl Vjlackberry, variety Ancient Britain. 43. — Spores collected at Janassee Junction, Ga., on wild blackberry. 44-— Spores collected at Thimderbolt, Ga., on R. procumbens. 45.— Spores collected at Hyattsville, Md., on wild dewberry. 46.— Spores collected at Hyattsville, Md., on wild blackberry. 4^ —Spores collected at Blacksburg, Va., on cultivated blackberry. 48.— Spores collected at Vienna, Va., on wild blackberry. 4g._Spores collected at Willard, N. C, on wild blackberry. 50.— Spores collected at Connellsville, Pa., on wild blackberry. ^i,_Spores collected at Auburn, Ala., on wild blackberry. 52.— Spores collected at Butte Creek Canyon, Calif., on R. ursinus. 53,_Spores collected at Stark, Fla., on wild blackberry. 54.— Spores collected at Blacksbiu-g, Va., on wild blackberry. 55 —Spores collected at Thunderbolt, Ga., on R. hispidus. j5,_Spores collected at Hyattsville, Md., on wild blackberry. 57.— Spores collected at Blacksburg, Va., on wild blackberry. 58. — Spores collected at Hammond, La., on wild blackberry. jg.—Spores collected at Orlando, W. Va., on wild blackberry. 60.— Spores collected at Ithaca, N. Y., on wild blackberry. 61 .—Spores collected at Blacksburg, Va., on wild blackberry. 62.— Spores collected at West Falls Church, Va., on cultivated blackberry. 63. — Spores collected at Arlington, Va., on wild blackberry. 64.— Spores collected at Arlington, Va., on wild blackberry. 65.— Spores collected at French Creek, W. Va. on wild blackberry. Further data on the Orange-Rusts of Rubus Plate 94 O a o o o o 0 Po o o o o oo o o OO 06 o 00 o o o o o 0,0 ^o o o o"o 00 o» O 0;i Q, o oO o o o 00 18 o o o o ^ 00 o 21 0 0 ^O O Oo o ^ o 6 ^0 00 o 0-0 Q o O 00 O " o O^ O" o o o op n O C)0 0 o O o 0.0 o .0 O r^^D ^^ Journal of Agricultural Research Vol. XIX, No. 10 PLATE 94 Gymnoconia interstitialis . Xioo, except figures 48 and 54, which are XS3^. I. — Spores collected at Madrid, Me., on wild blackberry. 2. — Spores collected at French Creek, W. Va., on wild black raspberry. 3. — Spores collected at Glen, N. H., on Rubus nigrobaccus. 4. — Spores collected at West Falls Church, Va., on cultivated black raspberry. 5. — Spores collected at Vienna, Va., on cultivated black raspberry. 6. — Spores collected at Old Forge, N. Y., on R. canadensis. 7. — Spores collected at Glen, N. H., on wild blackberry. 8. — Spores collected at Mountain Lake, Va., on R. alleghaniensis. 9. — Spores collected at West Falls Church, Va., on cultivated black raspberry. 10. — Spores collected at Portland, Me., on R. canadensis. II. — Spores collected at Madison, Wis., on wild blackberry. 12. — Spores collected at Phillips, Me., on wild blackberry. 13. — Spores collected at Portland, Me., on R. canadensis. 14. — Spores collected at East Lansing, Mich., on wild blackberry. 15. — Spores collected at Sebago Lake, Me., on R. iriflorus. 16. — Spores collected in Michigan on wild blackberry. 17. — Spores collected at Old Forge, N. Y., on R. canadensis. 18. — Spores collected at Madison, Wis., on wild blackberry. 19. — Spores collected at Old Forge, N. Y., on R. canadensis. 20. — Spores collected at West Falls Church, Va., on wild black raspberry. 21. — Spores collected at West Falls Church, Va., on cultivated black raspberry. 22. — Spores collected at West Falls Church, Va., on cultivated black raspberry. 23. — Spores collected at Glen, N. H., on R. canadensis. 24. — Spores collected at Mountain Lake, Va., on R. alleghaniensis. 25. — Spores collected at French Creek, W. Va., on black raspberry. 26. — Spores collected at Portland, Me., on R. canadensis. 27. — Spores collected at Smugglers Notch, Vt., on R. strigosus. 28. — Spores collected at Madrid, Me., on wild blackberry. 29. — Spores collected at Old Forge, N. Y., on R. canadensis. 30. — Spores collected at Smugglers Notch, Vt., on R. canadensis. 31. — Spores collected at Bancroft, Wis., on R. hispidus. 32. — Spores collected at Juliet, N. Y., on R. canadensis. 33. — Spores collected at French Creek, W. Va., on black raspberry. 34. — Spores collected at Old Forge, N. Y., on R. canadensis. 35. — Spores collected at Madison, Wis., on wild blackberry. 36. — Spores collected at Portland, Me., on R. iriflorus. 37. — Spores collected at Sebago Lake. Me., on R. Iriflorus. 38. — Spores collected at Smugglers Notch, Vt., on R. canadensis. 39. — Spores collected at Sebago Lake, Me., on R. Iriflorus. 40. — Spores collected at Rouses Point, N. Y., on black raspberry. 41. — Spores collected at Sebago Lake, Me., on wild blackberry. 42. — Spores collected at Mountain Lake, Va., on R. alleghaniensis. 43. — Spores collected at Bound Brook, N. J., on black raspberry. 44. — Spores collected at Glen, N. H., on R. canadensis. 45. — Spores collected at Glen, N. H., on R. canadensis. 46. — Spores collected at Chain Bridge, near Washington, D. C, on black raspberry. 47. — Spores collected at Chain Bridge, near Washington, D. C, on black raspberry. 48. — Spores collected at Chain Bridge, near Washington, D. C, on black raspberry. 49. — Spores collected at Glen, N. H., on R. canadensis. 50. — Spores collected at Glen, N. H., on R. canadensis. 51. — Spores collected at Glen, N. H., on R. canadensis. 52. — Spores collected at West Falls Church, Va., on black raspberry. 53. — Spores collected at Vienna, Va., on black raspberry. 54. — Spores collected at West Falls Church, Va., on black raspberry. GERM-FREE FILTRATES AS ANTIGENS IN THE COMPLE- MENT-FIXATION TEST By William S. Gochenour Veterinary Inspector, Pathological Division, Bureau of Animal Industry, United States Department of Agriculture In the production of germ-free blackleg filtrates, to insure uniformly good results it is of prime importance to check or control properly each lot of culture flasks, so as to know definitely that the blackleg organism alone has been growing therein. After the culture flasks have been inoculated and have been incubated for from six to nine days it is quite simple to remove one or more cubic centimeters of the culture and test it for the presence or absence of aerobic organisms. The detection of for- eign anaerobes, however, should any be present, is not at all a simple procedure. Moreover, it would be quite impracticable to resort to the complicated process of anaerobic plating or fishing in the search of contaminating anaerobic microorganisms as a routine procedure with each lot of culture flasks. For this reason serological studies were made with pure germ-free blackleg filtrates to ascertain whether they would act as antigens in the complement-fixation test and, if so, what range of specificity could be obtained therewith, using the results as an index to what a satisfactorily produced product should possess. Accordingly, blackleg filtrates were prepared, and a horse was repeat- edly injected with them at intervals extending over a period of approxi- mately three months. Blood serum drawn from this animal constituted the positive or immune serum. Antigenic titrations were then made, using 0.2 cc. of positive and 0.2 cc. of negative (normal horse) serum; and grading amounts of the germ-free filtrate were added as the antigen. The titration given in Table I will exemplify the character of reaction that has been obtained. When the filtrate is concentrated over sulphuric acid in vacuo to one-half or one-third its original volume, the antigenic unit and the anticomplementary dose are reduced in the same ratio. So far as the writer is able to learn by search through the literature, the use of a germ-free filtrate as an antigen in the complement-fixation test is an entirely new phenomenon; and it promises to serve a very important role in the separation and differentiation of the spore-bearing anaerobes. With this purpose in mind it is contemplated to parallel this reaction with the other pathogenic spore-bearing anaerobes as Bacillus edemaiiens, vibrion septique, B. tetanus, B. hotulinus, etc. ; and evidence of the feasibility of doing this is shown in the tests already made with Journal of Agricultural Research, Vol. XIX. No. lo Washington, D. C. Aug. id, 1920 nu Key No. A-51 (SI3) 514 Journal of Agricultural Research Vol. XIX, No. lo B. boiulinus filtrate. Good fixations were obtained by using B. hotulinus (type B) filtrate with B. hotulinus (type B) immune serum, but type B filtrate and type A serum would not produce a fixation, nor would type A filtrate produce a fixation with type B serum. Considering that type B immune serum does not protect guinea pigs against type A filtrate, which contains the type A toxin, and vice versa, the absence of fixation when using one type of serum and the other type of filtrate as antigen is quite important from a differential standpoint and also serves to indicate the specificity of the reaction obtainable by this method. Table I. — Titration of germ-free blackleg filtrate antigen Tube No. I Cc. O. 2 I . 2 2 . 2 A . 2 e . 2 6 . 2 7 . 2 8 . 2 0 . 2 lO . 2 II . 2 12 . 2 I 2 2 e 6 7 8 o lo II 12 I-J 14 Serum. Posi- tive. « Nega- tive. & Cc. Physio- logical salt solution. Cc. Antigen.'" Cc. Comple- ment. Cc. O Hemolytic rabbit se- rum and sheep cor- puscle. <* Cc. Result, e + + + + + + + + + + ± + + + + o Horse serum hyperimmunized to germ-free blackleg filtrate. 6 Normal horse serum. c Germ-free blackleg filtrate. d Hemolytic system employed consisted of a 3 per cent suspension of sheep red cells, 2 J4 units of hemolytic amboceptor, and i.Vi tmits of complement, the latter being titrated against the amboceptor and sheep cells. t + indicates complete inhibition of hemolysis; ±, partial inhibition of hemolysis; and — , no inhibition of hemolysis. Since a blackleg filtrate produced from a pure culture of Bacillus chauveaui and grown under favorable conditions will possess antigenic value in the quantities shown in the preceding table, if a filtrate were encountered that failed to approximate such a titre then only would it Aug. i6. I920 Germ-Free Filtrates as Antigens 515 seem necessary to resort to the anaerobic cultural examination of the culture flasks. Calves inoculated with blackleg filtrate showing a satis- factory antigenic value were rendered sufficiently immune, after a period of three to four weeks, to withstand intramuscular injections of 100 to 200 mgm. of virulent blackleg muscle powder, a quantity sufficient to kill unvaccinated calves in two to three days. Failure of a blackleg filtrate to possess an antigenic titre of from i/io to 1/20 the anticomplementary dose should arouse the suspicion that the blackleg organism did not grow under favorable conditions, that some contamination is present, or that the organism being used was not the blackleg organism at all. CONCLUSION From the data at hand it can be said that — (i) A blackleg filtrate produced under favorable conditions will possess a distinct antigenic value demonstrable by the complement- fixation test. (2) Those blackleg filtrates that conferred a solid immunity on calves were found to possess a high antigenic titre. (3) The complement-fixation reaction should be of much value as a laboratory control test to determine whether the filtrate has been pro- duced under conditions favorable to the blackleg organism or whether the blackleg organism has been supplanted in part or wholly by con- taminating anaerobic microorganisms. (4) Botulinus filtrate also acts as an antigen in the complement- fixation test when type B serum is used with type B filtrate' but fails to cause fixation when one type of serum is used with the other type of filtrate as antigen. (5) The phenomenon of germ -free filtrates acting as antigens in the complement-fixation test is new and promises to play an important part in the differentiation of the spore-bearing anaerobes, more especially those having closely similar cultural characteristics. 183719°— 20 i MOSAIC DISEASE OF CORN ' By E. W. Brandes Pathologist, Office of Sugar-Plant Investigations, Bureau of Plant Industry, United States Department of Agriculture DISTRIBUTION In connection with an investigation of the mosaic disease of sugar cane, a similar disease of corn has been observed by the writer on several occasions in widely separated regions.^ On April i8, 191 9, corn of an unknown variety was seen to be affected with typical mosaic symptoms in a field just west of Penuelas, P. R. The percentage of affected plants was small, however, only 20 individuals being found in the field of some 5 acres. The corn averaged about 24 inches in height at this time and was planted between rows of sugar-cane stubble which had not been completely killed out in preparing the land for the com. All the sugar cane was affected with mosaic. In July, 191 9, corn of the White Creole variety was seen at the Sugar Experiment Station, New Orleans, La., in which the same condition was apparent. This com was more than half grown, and the typical streaking of the leaves was somewhat ob- scured by certain leafspot diseases, among them the leafspot caused by Physoderma zeae-maydis , by which the com was severely attacked. About 10 per cent of the plants in the field were affected with mosaic. In adjoin- ing fields of sugar cane nearly 100 pe'r cent of the plants were affected with the sugar-cane mosaic. In 1920 corn of the same variety was examined early in the season, and a much more serious infestation was found. The com had been planted following sugar cane, and occasional diseased stools of the latter not killed by the plow were found all through the com field. More than 30 per cent of the corn plants were affected. The cases were more abundant in the vicinity of the sugar-cane stools re- ferred to above, but cases could be found many rods from any living cane. Of course, it is possible that a stool of cane had sprouted between the rows in such a situation and later had been killed by the cultivator. In May, 1920, identical cases of mosaic were seen in a field of corn near Cairo, Ga. As in the cases reported previously, a neighboring field of sugar cane was slightly infested with mosaic. • The study of this disease was undertaken on account cf its relation to sugar-cane mosaic. 2 Brandes, E. W. the mosaic disease of sugar cane and other grasses. U. S. Dept. A^. Bui. 829, 26 p., 5 fig., I col. pi. 1919. Journal of Agricultural Research, Vol. XIX, No. lo Washington, D. C. Aug. i6, 1920 va Key No. 0-203 (517) 5i8 Journal of Agricultural Research voi. xix, no. w Diseases of corn bearing a decided resemblance to the one in question have been reported from other countries. Dr. H. L. Lyon states ^ that in the Hawaiian Islands a disease of corn which resembles sugar-cane mosaic is very serious. William H. Weston ^ describes a disease of com in Guam which may be identical with the one under discussion. He mentions yellowing and dwarfing among the symptoms and states that the leaves exhibited mottling and striping. VARIETAL SUSCEPTIBILITY Just enough work has been done on varietal susceptibility to prove that all varieties of corn do not respond in the same way. The writer has never seen such excessive injury as that described for the unknown variety in Guam by Weston. In Louisiana the injury to com of the White Creole variety, while marked in some individuals, was not ex- cessive, excepting when the plants were infected early in the spring. The variety U. S. Select No. 182 is very susceptible to mosaic, but is not especially injured by it. Golden Bantam sweetcorn could not be infected in the greenhouse by methods which were successful with U. S. Select No. 182. Golden Bantam was planted unprotected in a greenhouse with hundreds of infected sugar-cane and sorghum plants. The corn aphis quickly migrated to the young com plants from diseased sorghum in great numbers, but no cases appeared among the Golden Bantam seedlings. It seems probable that this variety is immune. IMPORTANCE No figures are available on the amount of loss sustained on account of injury to corn. The writer is inclined to believe that in this country no great damage has been done thus far. Probably the disease was intro- duced on sugar cane within comparatively recent years, in which case it - may become more important in the future. At present, however, our chief concern is with its relation to the sugar-cane crop. Com is almost invariably used in the rotation on sugar-cane land, so that no planta- tion is ever without com in some of its fields. This means, of course, that the possibility for spread of the disease is greatly increased. Over- wintering by the vims has been demonstrated only in the vegetative portions of the sugar-cane plant, but the existence of other graminaceous hosts certainly complicates the problem of control. SYMPTOMS In com as in sugar cane the most conspicuous symptom of mosaic is the streaked and irregularly mottled appearance of the leaves. In corn, however, the lower, older leaves have a greater tendency to resume their normal color, so that it is sometimes difficult to demonstrate the 1 In verbal communication, January, 1920. ' Weston, W. H. report on the plant disease situation in guam. Guam Agr. Exp. Sta. Rpt. X917, p. 45-62- 1918. Aug. :6, 1920 Mosaic Disease of Corn 519 mosaic patterns in such leaves. In the youngest leaves, either the nor- mal dark green or the pallid, affected tissue may predominate in a given specimen, but the latter condition is most frequently met with. In such cases the areas which remain normal are in the shape of broken or inter- rupted streaks or lines extending in the general direction of the long axis of the leaf (PI. 95), and the contrast in color between these areas and the surroimding pallid areas is very decided. The streaks vary greatly in size, ranging from mere points to elongated "islands" of dark green 2 or 3 cm. or more long and several millimeters wide. The margins of such streaks may be straight or undulating. In most cases the mosaic pattern is more prominent at the base of the leaf, where it diverges from the leaf sheath. Where the normal dark green is predominant, the light green, affected tissue appears usually as a very fine mottling or as irregular elongated streaks on the darker background. From the fore- going description it can be seen that the patterns vary considerably, and yet they have certain general characteristics which make it almost impos- sible to confuse this condition with any other affecting the leaves. Infected plants are always lighter in color than healthy plants. When viewed from a distance such plants can be picked out with a fair degree of accuracy on this account. The top of the plant is especially pale, much more so than normal freshly unrolled young leaves. In some cases the color becomes decidedly yellow. In this connection it must be stated that the pallid color referred to heretofore as characteristic of the dis- eased areas is not a yellowish green but a lighter or more dilute tint of the normal green. In plants which become markedly yellow a decided stunting of the whole plant takes place. At no time has a case been observed to terminate fatally, but certainly considerable injury results from the lack of functioning chloroplastids, and where a large percent- age of the plants are affected the loss due to decreased size of ears is appreciable. When infection takes place early in the growing season, partial or complete sterility of the ears results. This serious feature of the disease was first noticed in Louisiana in 1920. In May, 1920, the writer tagged 20 diseased and 10 healthy plants in a field of White Creole com. The diseased and healthy plants were equally vigorous to all appearances at that time and were in the same rows, alternate diseased and healthy plants in the same row being selected as far as it was practi- cable. When the crop was harvested in August, 17 of the diseased plants were found to be completely sterile, while 3 of them had set a few scattered kernels. The 10 healthy plants were normal, excepting for slight com earworm injury, and produced large well-filled ears (PI. 96). During the course of experiments in the greenhouse several cases of apparent recovery have been observ^ed. Plants which became infected and exhibited the typical symptoms resumed their normal color after several weeks. These plants were held under observation until the ears were mature, but there was no recurrence of the mosaic symptoms. 520 Journal of Agricultural Research voi. xix, no. ro This interesting behavior was also noted in stools of crabgrass (Synr- therisma sanguinalis) and foxtail (Chaetochloa lutescens). There were no changes of growing conditions that could be correlated with these apparent recoveries. In this connection it may not be out of place to record that suckers from diseased stools of sugar cane and sorghum have been observed to come up with no sign of mosaic. These instances are by no means common, but several have been seen in both plants men- tioned. INSECT TRANSMISSION OF CORN MOSAIC The manner in which corn mosaic is transmitted to healthy plants and the relation of this disease to mosaic in other grasses was demon- strated by the following experiments. Experiment i. — On March 12, 1920, 12 com plants of the variety U. S. Select No. 182 were placed in each of two insect-proof cages. All of the plants were from the same lot of seed furnished by the Office of Cereal Investigations. The seed had been planted in one flat, and the seedlings were replanted in 5-inch pots on the date of removal to the cages. They were then 12 inches tall. About 12 individuals of Aphis maydis were carefully removed by means of a small camel's-hair brush from sorghum plants affected with mosaic to each corn seedling in one of the cages. The sorghum plants had been infected by aphids from mosaic sugar cane. Twelve aphids were transferred in the same way from healthy sorghum to each of the com seedlings in the adjoining control cage. On March 28, 6 of the 12 com seedlings in the first cage showed typical signs of mosaic in the two youngest leaves. On April 6, 8 of the plants, or 66^ per cent, were typical cases. The 12 control plants remained healthy up to the time of removal several weeks later. Experiment 2. — On April 6, 1920, 20 com seedlings, variety U. S. Select No. 182, in 5-inch pots were placed in each of two insect-proof cages in the greenhouse. Several specimens of Aphis maydis were transferred from infected corn plants to each corn seedling in the first cage. Aphids from healthy corn in another greenhouse were placed on each corn plant in the second control cage, which was used as a control. On May 4, 7 of the com seedlings in the first cage were found to be infected. On May 28, 15 of the 20 plants were observed to be unmistakable cases. The aphids had increased enormously in both cages. Not a single case could be found in the control cage, nor had any appeared up to June 25, although the plants had been repotted twice and were approaching maturity. These experiments demonstrate conclusively that provision is made for almost unlimited dispersal of the vims through the medium of the com aphis. There is no reason for supposing that transmission in nature is limited to this insect or to this method. It is not yet known whether the virus can survive the winter in seed, but experiments are now under Aug. i6. 1920 Mosaic Disease of Corn 521 way that may throw some light on this phase of the problem. It has been proved that the virus of corn mosaic is identical with that of sugar- cane and sorghum mosaic, so that even if it is found not to be seed- borne, perpetuation of the disease in the perenniai grasses would explain its appearance on com in the spring. Artificial transmission of the disease by means of inoculation with expressed cell sap of afifected plants has not been attempted for com. This method has proved successful in sugar cane, however,^ and there is little doubt that the infectious material is contained in the cell sap of com. Just what this infectious material is can not be stated definitely, but the evidence points strongly toward a living organism. No evi- dence incompatible with this view has been put forward for any mosaic disease, excepting the failure to demonstrate any visible organism. CONTROL Control measures for this disease must be based fundamentally on the removal of sources of the inoculum. So far as is known the only sources of inoculum are the living host plants. Destruction of these plants, then, will effectively eradicate the disease from any region. Practically, the destruction of all affected host plants presents almost unsurmountable obstacles. An immense amount of sugar cane is now infected in the River District of Louisiana and in southern Georgia. Destruction of large numbers of plants by roguing or plowing up is viewed with great concern by the planters, most of whom oppose any plan to control the disease by eradication. The substitution of immune varieties of com as well as cane does not offer any immediate solution, since the most susceptible varieties happen to be the ones most esteemed. Elimi- nation of this disease is dependent upon the education of the planter to an understanding of its seriousness. When this is accomplished public sentiment will permit of the passage of compulsory roguing and quarantine laws, which will be necessary before any hope can be enter- tained of eliminating the disease. 1 Brandes, E.W. ARTIFICIAL AND INSBCT TRANSMISSION OF SUGAH-CANB MOSAIC, /n JouT. Agr. Research, V. 19, no. 3, p. 131-138. 1920. Literature cited, p. 138. PLATE 95 Mosaic disease of corn : The first leaf at the left shows the typical interrupted streaks of normal green in a pallid green background. The next leaf shows a more irregular, mottled pattern. In these specimens the normal green was similar to "nickel green" and the pallid green was similar to "rejame green" in Ridgeway.* The two leaves at the right are' from a healthy plant and are presented for comparison. ' RiDGEWAY, Robert, color standards and cou)R NOMENOUAnjRB. 43 p., S3 col. pi. Washington, D. C, 1912. (522) Mosaic Disease of Corn Plate 95 Journal of Agricultural Research Vol. XIX, No. 10 Mosaic Disease of Corn Plate 96 ^'*H''i ^ ' ■' f ' ' •• •• ' ■* •' • •' ■' •' ■ •' •' •• •• •' •• V ■> • . |!V%l«IMlltUNiiill«UjiiiiiiM>iii»|ir Journal of Agricultural Research Vol. XIX, No. 10 PLATE 96. Mosaic disease of corn; Effect of early infection on the ear. White Creole variety. The three ears at the top were produced by plants naturally infected in the field. In 17 out of 20 marked plants no kernels at all were developed. The two lower ears are typical of all ears produced by healthy plants in the same row with the diseased plants. ADDITIONAL COPIES OF THIS PUBLICATION MAY BE FROCUKED FROM THE SUPERINTENDENT OF DOCUMENTS GOVERNMENT PRINTING OFFICE ■WASHINGTON, D. C. AT 25 CENTS PER COPY Subscription Price, $1.50 Per Year A Vol. XIX SEF>T:e:Jv1BER 1, 1920 No. ll JOURNAL OP AGRICULTURAI/ RESEARCH CONXKNTS Page Genetics of Rust Resistance in Crosses of Varieties of Triticum vulgare with Varieties of T. durum and T. dicoccum _-_--_--_ 523 H. K. HAYES, JOHN H. PARKER, and CARL KURTZWEtt (Contribution from Minnesota Agricultural Experiment Station and Bureau of Plant Industry) Line-Selection Work with Potatoes - - - - 543 O. B. WHIPPLE (Contribution from Montana Agricultural Experiment Station) Occurrence of the Fixed Intermediate, Hordeum interme- dium haxtoni, in Crosses between H. vulgare pallidum and H. distichon palmella ------ 575 HARRY V. HARLAN and H. K. HAYES (Contribution from Bureau of Plant Industry and Minnesota Agricultural Experiment Station) PUBUSHED BY AUTHORITY OF THE SECRETARY OF AGRICDITURB. WITH THE COOPERATION OF THE ASSOCIATION OF LAND-GRANT COLLEGES WASHINOXON, D. C. WMHINQTON : QOVCRNMENT PRINTINa OTFIOB I I»t0 EDITORIAL COMMITTEE OF THE UNITED STATES DEPARTMENT OF AGRICULTURE AND THE ASSOCIATION OF LAND-GRANT COLLEGES FOR THE DEPARTMENT KARL F. KELLERMAN, Chairman Physiologist and Associate Chief, Bureau of Plant Industry EDWIN W. ALLEN Chief, Office of Experiment Stations CHARLES L. MARLATT Entomologist and Assistant Chief, Bureau of Entomology FOR THE ASSOCIATION J. G. LIPMAN Dean, State College oj Agriculturei and Director, New Jersey Agricultural Experi- ment Station, Rutgers College W. A. RILEY Entomologist and Chief, Division of EnUf mology and Ecoruwiic Zoology, Agricul- tural Experiment Station of the University of Minnesota R. L. WATTS. Dean, School of Agriculture, and Diredor; Agricultural Experiment Station; Tk* Pennsylvania State College All correspondence regarding articles from the Department of Agriculture should be addressed to Karl F. KfeUerman, Journal of Agricultural Research, Washington, D. C. All correspondence regarding articles from State Experiment Stations should be addressed to J. G. Lipman, New Jersey Agricultural Experiment Station, New Brunswick, N. J. LIBRARY NEW VORk BOTANIC A I JOMAL OF AGKICHTDRAl ffiSEARCH Vol. XIX Washington, D. C, September i, 1920 No. 11 GENETICS OF RUST RESISTANCE IN CROSSES OF VA- RIETIES OF TRITICUM VULGARE WITH VARIETIES OF T. DURUM AND T. DICOCCUM ^ By H. K. H.AYUS, Head of Section of Plant Breeding , Division of Agronomy and Farm Managem,ent, Department of Agriculture, University of Minnesota, John H. Parker, Scientific Assistant, and Carl KurtzwEil,^ Assistant Pathologist, Office of Cereal Investigations, Bureau of Plant Industry, United States Department of Agriculture INTRODUCTION The black stemrust (Puccinia graminis) of small grains causes enor- mous losses. Reduction in yield of from 10 to 50 per cent of the wheat crop is common. At irregular intervals the black stemrust of wheat causes almost complete failure, especially in the spring-wheat area of the upper Mississippi Valley. Control of this disease, which develops into such terrific epidemics, is impossible by any method now available to the individual grower. For this reason the development of resistant varie- ties assumes great importance. While the barberry eradication cam- paign now being carried on over a wide area Vill certainly reduce the amount of rust, local outbreaks may perhaps be expected even after barberries have apparently been eradicated. The attempt to develop resistant varieties, therefore, should continue. There is every reason to hope that the stemrust problem can be solved by barberry eradication and the development of resistant wheat varieties. When the present study was outlined, the evidence seemed to show that parasitic action of the rust was constant. Recent extensive studies {22, 2sY have confirmed this view and indicate that the bridging hypothe- sis (jj), which was supposed to account for the increase or decrease in ' Published with the approval of the Director as Paper 187, Journal Series, Minnesota Agricultural Experiment Station. Cooperativ(f investigation between the Minnesota Agricultural Experiment Station and the Office of Cereal Investigations, Bureau of Plant Industry, United States Department of Agriculture. ^ The breeding of spring wheat for rust resistance was begun by Dr. E. M. Freeman and E. C. Johnson in 1908 and has been continued without interruption until the present time. The present cooperative arrangement between the Division of Plant Pathology and Botany and the Section of Plant Breeding, Division of Agronomy and Farm Management, of the University of Minnesota, was made in the spring of 1916. The writers wish to acknowledge the helpful cooperation of Dr. E. M. Freeman and Dr. E. C. Stakman in this investigation. t^, ' Reference is made by number (italic) to "Literature cited," pp. 541-542. CvJ O^ Journal of Agricultural Research, Vol. XIX, No. 11 Washington, D. C. Sept. i, 1920 <^'' uv Key No. Minn. -41 J (523) )— o 524 Journal of Agricultural Research voi. xxx. No. n virulence of the rust, is probably incorrect. More recent investigations of the wheatrust fungus have shown that there are numerous biologic forms (25) which can be differentiated only by their action on various pure-line wheat varieties. This is a very serious obstacle to the produc- tion of rust- resistant varieties by breeding, although the fact that Khapli (C I 4013),^ an emmer imported from India, is resistant to all biologic forms so far isolated shows that the problem is not entirely hopeless. Former breeding investigations have determined which varieties of wheat are commonly resistant to stemrust at University Farm, St. Paul, and have also indicated the behavior of rust resistance in crosses. In the light of this information it was decided to make a careful genetic study of one or two crosses, hoping thus to solve the plant-breeding phase of the rust-resistance problem. The second generation of the crosses reported in this paper had already been studied before it was known that there were sometimes numerous biologic forms of the rust in the same locality. It seemed worth while to complete the study of inheritance of rust resistance in these crosses by growing a small F3 family of each Fj plant which produced viable seed. All barberry bushes were removed from the immediate vicinity of the rust plot early in the spring of 1918, and the artificially induced epidemic was produced with a known racial strain of the biologic form Puccinia graminis tritici Erikss. and Henn. Therefore, the greater part, if not all, of the rust infection present during the season of 191 8 was due to this one biologic form. One uredinium found on Kanred (C I 5146), which is known to be resistant to the strain which was used to induce the epidemic, proved to belong to another biologic form. The present paper is a report of the inheritance of rust resistance in its correlation with botanical and morphological characters of crosses between Triticiwi vulgare with varieties of T. durum and T. dicoccum. SOME PREVIOUS CROSSES OF WHEAT SPECIES Because of the many differential characters and the great economic importance of the crop, wheat has been frequently used in studies of the laws of inheritance. Vilmorin {26, 27) reported quite extensive tests of crosses between Triticum sativum T-, T. turgidum, T. durum, Desf., T. polonicum T-, and T. spelta. From the Mendelian standpoint the results obtained are interesting. He observed remarkable uniformity in the F^ generation, wide diversity in the Fj generation, and states that the predominating force after the F3 generation is that of heredity, which compels the plants to reproduce their characters in their immedi- ate descendants. Vilmorin placed spelt and common wheats in one group and poulard and durum in another, because he found that either spelt or common crossed with poulard or durum produced all four types in the 1 Cereal investigations number. Sept. 1. 1920 Genetics of Rust Resistance 525 segregating generations. These types were easily fixed after several years of selection. The four types were believed to have a common ancestry and to belong to one botanical species. Unsuccessful attempts were made to cross these species with T. monococcum ly. This led Vilmorin to conclude that T. monococcutn belonged to a separate species. Tschermak (25) confirmed Vilmorin's conclusions regarding the pro- duction of common, durum, spelt, and poulard forms from crosses when the groups differed in the solid and hollow stem character. He stated that the polonicum type was obtained only when it was used as one of the parents. Relationships were illustrated on a factor hypothesis by assum- ing (i) that Triticum polonicum contained two dominant factors, (2) that common wheat lacked these factors, and (3) that durum contained one of these factors but not the other. Tschermak also thinks that T. mono- coccum is different from other wheat species. He obtained hybrids between T. monococcum and T. vulgare and one Fj plant, which soon died. Blaringhem (6), however, succeeded in crossing Triticum durum and T. polonicum with T. monococcum. Sterility was at first of frequent occurrence in these crosses, but in later generations it was much less common. Explorations by Aaronsohn in 1906 (i) confirm the experimental evidence cited above. A wild emmer, Triticum dicoccum dicoccoides,^ found in 1906, is interesting because it tends to confirm the report that such an emmer was collected as early as 1885. Crosses between Black Winter emmer and Fultzo-Mediterranean (i^) had strongly keeled glumes with hard, adherent chaff in the F^ generation. The F2 plants varied widely in type and exhibited transgressive segrega- tion for some characters. There was considerable sterility in this cross. The hairy chaff and black color of the emmer parent were linked in inheritance. Freeman (12) found evidence of linkage between high ratio of width to thickness of head and hardness of grain in a cross between durum wheat and a variety of bread wheat. The bread wheat parent had a square head with soft, opaque grains, while the durum parent had a much flattened head and produced hard, translucent grains. In a recent article on heredity of quantitative characters in wheat {13) a cross between durum and common wheat is reported. It gave normally vigorous plants in the F^ generation. In the Fg generation, how- ever, many seeds failed to germinate; and among those with a normal vegetative development were found plants exhibiting every degree of sterility, from perfectly sterile to fertile plants. Linkage has been reported by Engledovv and Biffen in crosses between Rivet, Triticum turgidum, and common Fife wheat. Gray glume color 1 From a cross between durum and common varieties. Love and Craig have produced a form which closely resembles the wild emmer. LovE, H. H., and Craig, W. T. the synthetic production of wild wheat FOR.MS. In Jour. Hered., v. lo, no. 2, p. 51-64, illus. 526 Journal of Agricultural Research voi. xix, No. n (4) was always found associated with hairy chaff, and partial linkage (jo) was found between the factors for black color and the factors for glabrous chaflf. Since sterility has been found by many investigators in wheat crosses and has been confirmed by the results presented in this paper, it seems rather difficult to reach any conclusion regarding linkage, because some combinations may be eliminated. The inheritance of the principal botanical characters of our cultivated wheats is well known and, therefore, need not be summarized in this paper. The inheritance of beards will be mentioned in connection with our results on sterility. In crosses between the so-called beardless wheats such as Marquis and Bluestemand a bearded variety, the F^ gen- eration has intermediate awns and in the F2 generation a i to 2 to i ratio is obtained. Fully bearded plants breed true in the F3 generation. How- ard and Howard {14) found that there are two classes of wheats with short awns which, when crossed, give fully bearded plants in the F^ gen- eration and breed true in the F3 generation. Likewise, crosses between bearded and true beardless forms gave i fully bearded plant in the Fg generation out of 16 plants. The fact of interest for our studies is that fully bearded plants breed true for this character. PREVIOUS STUDIES ON INHERITANCE OF RUST RESISTANCE The most successful attempt to breed rust-resistant wheats was made by BifFen (2, j, 5), who found that resistance to striperust {Puccinia glumarum Enkss. and Henn.) was a recessive character. Definite segregation occurred in the F2 generation, and forms bred true in the F3 generation, the ratio of resistant to susceptible in the segregating families being i to 3. From the practical standpoint these experiments have been very valuable. A new variety, Little Joss, was produced, which, because of its rust resistance, yields more on the average than susceptible sorts and has desirable milling characters. Nilsson-Ehle (20) has likewise made studies of the inheritance of resistance to striperust. Distinct dominance of susceptibility was seldom found. Ordinarily the F^ generation was intermediate, and in other cases resemblance to one or the other parent was observed. Segregation was obtained in the Fj generation, but without definite ratios. Transgressive segregation occurred, forms being obtained which were more susceptible than the susceptible parent and others which were more resistant than the resistant parent. The results were explained on the basis of multiple factors. While many observations have been made on resistance of wheat varieties, the two experiments cited above are the only carefully con- trolled studies so far reported which show the mode of inheritance. Sept. 1, 1920 Genetics of Rust Resistance 527 METHODS OF STUDYING INHERITANCI? OF RUST RESISTANCE In the studies here recorde'd crosses were made between Marquis and resistant durum and emmer wheats. Precautions were taken to protect the emasculated heads from foreign pollen. The Fj plants were grown in individually spaced plots, and seed from each Fj plant was grown sepa- rately. The F2 families of crosses between the same parent varieties gave similar results and were considered as a single cross. No effort was made with either the Fj or Fj plants to protect them from natural crossing, and an error was thus introduced which will be discussed later. The correlation between resistance in the Fj and the Fj generations gave unusual results which indicated that some uncontrolled factor was causing complications. For example, the Fj cross between emmer and Marquis which was grown in 19 16 appeared resistant, while in the Fj generation which was grown in 19 17 the number of resistant plants was much smaller than would be expected if resistance were a dominant character. This led to the belief that very likely more than one biologic form of stemrust was present in 191 7. Each F2 plant which produced viable seed was tested in 1918. As previously mentioned, the barberry bushes were removed from the immediate vicinity of the rust plot early in the spring of 1918. An epi- demic was obtained with a known strain of Puccinia graminis triiici which had been cultured in the greenhouse by the Section of Plant Pathology for several generations. This strain had been tested repeatedly on varieties of wheat and proved to be constant. The 19 1 8 results have been used to determine the resistance or sus- ceptibility of F2 plants which have been grown the previous season. The Fg as well as the Fj data have been used as a basis for placing the Fj plants in certain botanical groups, for the problem was chiefly to deter- mine the mode of inheritance and correlation of resistance or suscepti- bility with those botanical characters which are commonly used in differentiating wheat species, some of which are also of economic im- portance. STERILITY IN THE CROSSES OF MARQUIS WITH DURUM AND EMMER VARIETIES In order to determine whether there is any interrelation between botanical characters and rust resistance, the data for each V^ plant were taken in a correlated manner. The practical significance is obvious. For example, if durum head and seed characters were ralher closely linked with rust resistance in inheritance, it would be necessary to grow a larger Fj population to obtain the desired form than if each character were inherited independently. In deciding regarding possible linkage it is important to know whether sterility is involved in the crosses. 528 Journal of Agricultural Research Vol. XIX, No. II Sterility might cause the elimination of certain gametic or zygotic combinations, thus actually eliminating the sort desired. Zygotic com- binations are sometimes eliminated as is the case with the homozygous yellow mouse combination (7, 17) and with lethal factors in Drosophila (jp). In some cases gametic combinations are eliminated. Such elimi- nations occur as a result of pollen or ovule abortion. POLLEN ABORTION The presence of shriveled or abortive pollen grains is one means of recognizing sterility. The method here used was to shake out pollen from the heads upon a clean glass slide and then place a minute drop of fuchsin on the pollen, then a drop of lactic acid, and a cover glass. ^ By this method pollen was preserved for several weeks and could be studied when it was convenient. The parent plants and F^ crosses were grown in 6-inch pots in the green- house. A small percentage of small, globular, clear pollen grains were observed. These, together with the occasional shriveled grains, were counted as sterile. The results of these counts are given in Table I. Table I. — Counts of sterile pollen grains in wheat species and F, crosses between them Variety. Good pollen. Poor pollen. Parental species: Triticum vulgare TriticuTTi durum < Triticum durum Triticum dicoccum Fj crosses: Marquis X emmer (Minn. 1165). Emmer (Minn. 1165) X Marquis. Kubanka (C I 2094) X Marquis. . Marquis X Kubanka (C I 2094). . Marquis X Mindum Marquis Mindum Kubanka (C I 2094) . Emmer (Minn. 1165). 513 135 122 169 475 340 113 151 lOI 3 o 19 37 9 20 12 These results show that there is a larger percentage of shriveled and abortive grains in the crosses than in the parental varieties. There is also an indication of more pollen abortion in the durum-Marquis crosses than in the cross between emmer and Marquis. COMPARISON OF NUMBER OF SEEDS SET IN Fj GENERATION AND PARENTS As a further test of sterility, counts were made of the number of barren florets in several F^ crosses and their parents. There are two or more flowers in each wheat spikelet, and usually two or three kernels mature. The outer florets of each spikelet are usually most vigorous, and when only two florets per spikelet produce kernels these are. usually the outside ones. Therefore, the two outer florets of 1 Outlined to the writers by Dr. C. O. Rosendahl, Professor of Botany, University of Minnesota. Sept. I, 1920 Genetics of Rust Resistance 529 each spikeiet were examined. The result of this examination, together with the percentage of barren florets, is summarized in Table II. Table II. — Number of florets not setting seed in Triticum vulgare, T . durum, T. dicoc- cum, and Fi crosses ofT. vulgare uith T. durum and T. dicoccum Variety or cross. Marquis Preston (Minn. Pioneer Average . r)-4 lumillo (CI 1736). . Kubanka (C I 2094). Acme Average Emmer (Minn. 1165). . . Kubanka (C I 2094) X Marquis. Acme X Preston D-4 X Pioneer Average Marquis X Kubanka (C I 2094). Marquis X lumillo (C I 1736). Pioneer X D-4 Preston X Acme Average Emmer X Marquis . Emmer X Preston . Average Marquis X emmer. Preston X emmer . Average Classification. Vulgare. ....do.. ....do.. Durum . ....do. ...do. ....do. Dicoccum Durum X vulgare .do. .do. Vulgare X durum. . . , ...do .do. .do. Dicoccum, X vulgare ....do Vulgare X dicoccum . ....do Num- ber of heads. 15 17 26 16 16 25 49 4 9 10 Num- ber of florets exam- ined. 309 538 494 898 622 630 426 104 262 I, 692 135 296 348 410 446 270 94 Num- ber of good seed. 341 500 761 499 464 864 604 611 206 62 97 704 128 167 299 293 62 Num- ber of very badly shriv- eled seed. 14 97 9 17 25 Per- cent- barren florets. 3-» 5-1 5-2 4-7 4.8 5-9 3-6 2.9 4-3 49-5 35-6 56.5 47.2 52-7 31- I 49. I 46.5 22. 9 28. 7 25.8 28.2 29.8 29. o The main spike of individual plants of the parent and crosses was used for this study. There was an average of 4.7 per cent of barren florets in three varieties of Tritictim vulgare. In four varieties of T. durum there was an average of 4.3 per cent, while emmer (Minnesota 1165) produced only 2.1 per cent of florets which formed no kernels. Since these are presumably homozygous, it seems fair to conclude that the percentages given show the average number of florets which did not set seed on account of causes other than sterility. Three Fj crosses were studied in which a durum sort was the female parent and four in which durum was the male parent. The average 530 Journal of Agricultural Research voi. xix, No. n percentages of barren florets were 47.2 and 46.5, respectively. There was about the same percentage of barrenness in emmer X Marquis, emmer X Preston, and reciprocals — namely, 25.8 when emmer was the female parent and 29 when emmer was the male parent. Thus, there is apparently more sterility in the durum-Marquis cross than in the cross between emmer and Marquis. Although no cytological examination was made to determine the time of degeneration, it seems very likely that there is ovule as well as pollen abortion. Barren florets, however, might be due to incompatibility of certain genetic combinations or to slow growth of the pollen tube which has been shown to occur in some species crosses ( Density was calculated by dividing the length of the head in millimeters by the number of spikelets less one. Sept. I, 1920 Genetics of Rust Resistance 535 near-common, and common are 3.30, 3.77, 4.08, 3.94, and 4.40, respec- tively. A correlation coefficient of -f-0.244 ±0.028 for density and head characters was obtained. Table VI. — F^ Marquis X lumillo {CI 1^216) and Marquis X Kuhanka {CI 2094)^ [R=resistant to stemrust. S=susceptible to stemrust] Density. Durum. Near- diu-um. Inter- mediate. Near- common. Common. Emmer. R. S. R. s. R. s. R. s. R. s. R. S. Mm. 2-5 3-0 3-5 4..O 3 3 2 4 8 5 4 4 2 5 2 6 3 12 28 29 22 20 9 5 3 3 6 2 2 I 9 20 19 26 21 18 9 8 I I 12 16 14 19 14 7 5 6 4 6 12 16 10 18 6 9 I 3 4-5 e.o s-s 6.0 I 8 25 19 128 14 130 2 93 0 81 4 0 Ratio i:-2 T 1:6 7 I '.if' e 0 «T ' The coefficient of correlation between density of head and classes for durum, near-durum, intermediate, near-common, and common is -l-o.244io.028. F2 generations in which the resistant durum parent was the female are given in Table VII, a total of 632 plants being classified on the basis of the Fg breeding test. Of these Fj plants 100 bred true to durum habit, 47 resembled common, while 3 were emmerlike. Sixteen out of 100 durumlike plants were resistant, while only 2 out of 47 classified as common were resistant. Of the three emmerlike plants, one was susceptible and the other two were resistant. Here, again, as with the reciprocal cross, there is an indication of linkage between common wheat head characters and susceptibility to stemrust. One intermediate susceptible F2 plant was grown in the F3 generation, and several plants were obtained with common head characters. One of these was very resistant and vigorous (Pi. 99). The average densities of the durum, near-durum, intermediate, near- common, and common segregates were 3.76, 3.53, 4.06, 4.38, and 4.60, respectively, although both durum and common wheats were found in the extreme dense and lax groups. The coefficient of correlation be- tween density and head type was H-o.33o ± 0.024. In order to confirm field observations, a study was made in the green- house of the more resistant and susceptible durum, common, and emmer segregates obtained from the durum-common crosses. For head char- acters of types studied see Plate loi. Seedlings of F3 plants were 536 Journal of Agricultural Research Vol. XIX, No. II inoculated with the form of Puccinia graminis tritici which was used in obtaining the field epidemic. Three different tests of resistant and susceptible segregates of emmer, common, and durum F3 lines were m.ade. The results are given in Table VIII. (See PI. 102.) Table VII. — F2 lumillo (C 1 1^36) X Marquis and Kubanka (C / 2094 ) X Marqiiis^ iR= resistant to stemrust. S= susceptible to stemrust] Density. Durum. Near- durum. Intermedi- ate. Near- common. Common. Emmer. R. S. R. S. R. S. R. S. R. s. R. s. Mm. 2-5 ■2.0 I 8 3 2 I I 7 24 18 7 10 13 3 4 II 6 3 I I 19 40 27 22 19 9 2 I 2 2 2 2 II 22 30 47 43 31 2 6 2 I I I 2 15 8 24 21 16 20 6 I I 3 2 6 6 7 6 6 9 I I ^.r 4.0 4-5 c.o I 5-5 6.0 16 84 26 139 8 192 5 112 2 45 2 I Ratio 1:5-3 1:5-3 i: 24.0 1:22.4 1:22.5 1:0.5 1 The coefficient of correlation between density and type as durum, near-durum, intermediate, near- common, and common is -I-0.330 ±0.024. Table VIII. — Comparison of resistant and susceptible F^ lines with Marquis, Ktibanka (C I 2og4), and lumillo (C / 1736Y Source. Plant No. Type. Num- ber inocu- lated. Num- ber in- fected. Remarks. Marquis Minn. 1239. . . . C I 2094 Common . . Durum Emmer. . . ...do 10 12 18 II 14 21 17 20 22 10 10 8 10 12 4 8 19 II Kubanka X Marquis F3 . . . slight chlorosis under in- fected areas. Kubanka X Marquis F3 . . . minute', surrounded by sharp hypersensitive areas; 8 plants were strongly flecked. lumillo X Marquis F3 Common . . ...do normal. I-jmillo X Marquis F3 186-13-S normal. lumillo X Marquis F3 228-37 Durum ...do chlorosis under infected areas; 7 plants were strongly flecked. Moderately resistant with chlo- rosis under infected areas; 2 plants were distinctly flecked. lumillo X Marquis Fs 186-15 lumillo C I 1736 do. . normal. Slightly susceptible; 4 plants were distinctly flecked with slight chlorosis under some infected areas. ' The authors wish to express their thanks to M. N. Levine, Assistant Pathologist, Office of Cereal Inves- tigations. Bureau of Plant Industry, United States Department of Agriculture, for preparing Table VIII and Plate loi and for conducting the greenhouse experiment on which they are based. sept. 1, 1920 Genetics of Rust Resistance 537 These families were selected in the field as examples of resistant and susceptible F3 lines. Since the inoculation results in the greenhouse corroborate the field observations, there is good reason for believing that the field observations are reliable. This greenhouse experiment shows clearly that transgressive segregation occurred. Common, durum, and emmer sorts were obtained with a higher degree of resistance than that of either of the durum parents. These results show that, if the numbers are sufficiently large, it is pos- sible to obtain resistant wheats with common head characters by cross- ing resistant durum and susceptible common varieties. In the F2 generation, out of a total of 128 com^mon segregates only 2 were rust- resistant, and both of these were of little commercial value. Hov\^ever, several resistant plants with the head characters of common wheats were obtained in the F3 generation. Emmer-common crosses. — Only a few plants were available for the study of crosses between emmer and common varieties. Two different white spring emmers (Minnesota 1165 and C I 1524) were used in this study. Minnesota 11 65 is a very vigorous variety which was practically immune from the form of stemrust experimented with, both in the field and in the greenhouse. C I 1524 is quite similar to Minnesota 1165, except that it occasionally produces small uredinia. The difference between emmer and common wheat is greater than that between durum and common wheat. The shape of the head of the emmer parent, which is proportionally very narrow in face view, together with the strongly keeled glumes, dift'erentiates it from the Marquis parent. The kernels of emmer are tightly inclosed by the glumes, and recognition is easy. The Fj generation of the cross between emmer and common is inter- mediate for the differential characters mentioned above as separating the parent sorts (PI. 98). The keels of the F^ generation are inter- mediate, but they more closely resemble those of the emmer parent. The head shape of the F^ generation is likewise more nearly emmerlike. Fifty-six per cent of the kernels were naked and 44 per cent had adher- ent glumes after they were thrashed in an individual plant thrasher. The emmer parent produced 19 per cent naked kernels and 81 per cent hulled kernels. The F^ generation (Table V) is nearly as resistant as the emmer parent. Thus, the Fj generation more nearly resembles emmer. Fifty F2 plants of the cross between Marquis and emmer (Minnesota 1 165) were grown in the Fg generation. Of this number 5 plants were emmers, and i was a very susceptible, lax, common type. The progeny of 5 F2 plants thrashed like common wheats in the F3 generation, but the heads were very compact (PI. 100). Two F, plants bred true to the common keel condition and segregated, producing compact, intermediate, 538 Journal of Agricultural Research voi. xix, no. n and lax heads. Thirty-seven produced progeny consisting of types with inclosed emmerlike kernels, intermediates for thrashing, and common plants. Of these 37 F3 lines, 17 produced progeny of only two sorts, emmerlike for thrashing and compact keelless (PI. 100). It is, of course, impossible to determine the actual factors involved in this cross, because some gametes or zygotes were eliminated by sterility. The cross in which emmer (C I 1524) was the female parent gave results similar to those obtained when emmer was used as the male parent. This is shown in Table IX. One family, 169-5, which was grown from an intermediate F, plant, produced emmers, intermediates, and common- headed sorts and was quite resistant (PI. 100). Table IX. — Classification of crosses between emmer {Minnesota Il6^ and C I 1524) "with Marquis on the basis of rust class in the F^ generation as determined by the Fo '^'^'^ ^a generations , and the m,ain character differences separating emmer from common wheats as determined by the Fj and F3 generations MARQUIS X EMMER I165 Rust class. Emmer. Segre- gating emmer to common. With nonadherent glumes. c— !S. Common. Total. I 4 I 9 21 7 I I IS 22 2 ■2 4 I I 13 Total 5 37 5 2 I 50 EMMER (C I 1524) X MARQUIS Rust class. Emmer. Segre- gating emmer to common. With nonadherent glumes. Compact. Segre- gating. I 3 3 6 2 4 2 2 I 10 2 9 4 Total 3 II 8 I 23 It is interesting that out of a total of 73 plants from these two crosses, 8 bred true to the emmer habit for thrashing and for head shape. Of these 8, 7 plants were put in rust class 1 while i was practically resistant, the progeny in the F3 generation being placed in rust class 2. Only one lax, common plant was obtained in the Fj generation, and this was sus- ceptible. Several plants were obtained in the F3 generation which were not only rust-resistant but also resembled common wheat. Sept. 1. 1920 Genetics of Rust Resistance 539 These facts show that it is possible to transfer the rust resistance of emmer wheats to common wheats by crossing and subsequent selection. There is, however, an apparent partial linkage between rust resistance and the emmer head type which makes it essential to grow large num- bers in the F.^ and F3 generations. SUMMARY (i) Recent studies of the parasitism of the black stemrust of wheat {16, 18, 21, 24) have shown that there are many biologic forms of Puc- cinia graminis which can be differentiated only by their action on pure line wheat hosts. This seriously complicates the breeding of wheat for rust resistance. In the light of this knowledge differences of infection of certain crosses in 1917, as compared with 1916 or 1918, shov/ that the conflicting results may be explained logically by supposing that more than one biologic form was present in the rust nursery in 191 7. (2) Sterility is a factor which must be considered in a study of crosses between common wheats and durum or emmer varieties. Sterility was shown in three ways: (a) pollen abortion; (b) the fact that F^ florets of durum-common crosses set nearly 50 per cent less kernels than the parent sorts, while Fj emmer-common crosses produced about 25 per cent of barren florets; and (c) the large number of natural crosses which occurred in some F2 plants as shown by the F3 results. (3) Crosses between durum and common wheats produced many dif- ferent forms in the F2 generation, such as compact keelless commons resembling club wheats, lax sharply keeled durums, both emmer and spelt, as well as types \vhich resembled the poulard group. Lax and compact durum, common, and emmerlike forms were obtained which bred true in the F3 generation. The segregation in the Fj generation of emmer-common crosses was not so wide as in the durum-common cross, although both lax and compact keelless wheats which bore naked kernels, as well as lax and compact wheats with adherent-glumed kernels, were obtained. (4) The study of inheritance of rust resistance was made in a specially prepared disease plot. Because of the conflicting results of 191 6 and 1 91 7 all barberry bushes were removed early in the spring of 1918 from the immediate vicinity of the rust plot and the epidemic was induced with a known form of rust. The data on rust infection are based on these 1 91 8 results. (5) The following species and varieties were used in the study: Triticum vulgare, varieties Preston, Marquis, and Pioneer; Triticum dummy varieties Acme, D-4, Kubanka (C I 2094), and lumillo (C I 1736); Triticum dicoccum, White Spring emmer (Minnesota 1165 and C 1 1524). The three common wheats were susceptible, the durums were com- merically resistant, Kubanka (C I 2094) being somehwat less resistant 417°— 20 2 540 Journal of Agricultural Research voi. xix. no. n than lumillo (C I 1736), while Acme and D-4 were slightly more resistant than either of the other durum sorts. The emmer varieties were very resistant, Minnesota 1165 being practically immune. (6) The Fj generation of crosses between durum and common varieties was as susceptible as the common parent, while F^ crosses between the practically immune emmer parents and susceptible commons were about as resistant as the durum varieties. Thus, in the cross where emmer is one parent, resistance is partially dominant, while in the durum-common cross susceptibility is completely dominant over resistance. (7) Each F2 plant which produced viable seed was tested in the F3 generation for both rust infection and botanical characters. These F3 notes were used to determine the genotypic nature of individual Fj plants. In the crosses between durum and common in which Marquis was the female parent, 404 Fj plants were tested in the F3 generation and no rust-resistant common wheats were obtained. Likewise, no plants in the F3 generation seemed especially promising for both common wheat characters and rust resistance. In the crosses in which durum was the female, one or two Fj common-headed plants were resistant; but their progeny were worthless from a practical agronomic standpoint. In one F3 family which was grown from a susceptible Fj plant, a number of resistant, vigorous plants were obtained which had common head charac- ters. There is an indication of linkage of durum or emmer characters and rust resistance, since the production of rust-resistant durums or em- mers in the Fj and F3 generations is comparatively easy and the produc- tion of resistant common wheats much more difficult. (8) Resistant and susceptible plants obtained either in the Fj or F3 generation from crosses of durum and common parents were selected. Resistant and susceptible common, emmer, and durum wheats were available for this study. Greenhouse inoculation studies with a known strain of Pticcinia graminis tritici showed that durum, common, and emmer type plants were obtained in the Fj or F3 generation which were more resistant than the resistant durum parents. Thus, we have transgressive segregation for rust resistance. (9) The number of plants available for a study of inheritance between resistant emmer parents and Marquis was not very great. In the F3 generation several lax-headed wheats were obtained which had the head shape and naked kernels of common wheats and which were rust-resistant. This shows that rust-resistant common wheats can be obtained by cross- ing susceptible common varieties with resistant emmers. (10) The mode of inheritance of rust resistance seems entirely com- parable with the general Mendelian manner of inheritance of botanical and morphological characters. The technic of breeding for rust re- sistance is similar to that of breeding for agronomic characters. Sept. 1. 1920 Genetics of Rust Resistance 541 LITERATURE CITED (i) Aaronsohn, Aaron. 19IO. AGRICLXTURAL AND BOTANICAL EXPLORATIONS IN PALESTINE. U. S. Dept. Agr. Bur. Plant Indus. Bui. 180, 64 p., illus., 9 pi, (2) BiFFEN, R. H. 1907. STUDIES IN THE INHERITANCE OF DISEASE-RESISTANCE. In Jour. AgT. Sci., V. 2, pt. 2, p. 109-128. (3) I912. STUDIES IN THE INHERITANCE OF DISEASE RESISTANCE. II. In Jour. Agr. Sci., V. 4, pt. 4, p. 421-429. (4) 1916. THE SLTPRESSiON OP CHARACTERS ON CROSSING. In Jour. Gcnetics, V. 5, no. 4, p. 225-228. (5) 1917. SYSTEMATIZED PLANT BREEDING. In Sewafd, A. C, ed. Science and the Nation. ... p. 146-175. Cambridge, [Eng.] (6) Blaringhem. 1914. sur la production d'hybrides entre l 'engrain (triticum mono- coccuM L.) ET DiFF^RENTS bl6s cultiv^s. In Compt. Rend. Acad. Sci. [Paris], t. 158, no. 5, p. 346-349, i fig. (7) Castle, W. E., and Little, C. C. 1910. ON A MODIFIED MENDELIAN RATIO AMONG YELLOW MICE. In Science, n. s. V. 32, no. 833, p. 868-870. (8) DORSEY, M. J. 1919. A STUDY OF STERILITY IN THE PLUM. In Genetics, V. 4, no. 5, p. 417-488, 5 pi. Literature cited, p. 481-487. (9) East, E. M., and Park, J. B. 1918. STUDIES ON SELF-STERILITY. II. POLLEN-TUBE GROWTH. In Genetics, V. 3, no. 4, p. 353-366. Literature cited, p. 365-366. (10) Engledow, F. L. 1915. repulsion in wheat. In Amer. Nat., v. 49, no. 578, p. 127. (11) Evans, I. B. Pole. 1911. SOUTH AFRICAN CEREAL RUSTS, WITH OBSERVATIONS ON THE PROBLEM OF BREEDING RUST-RESISTANT WHEATS. In JoUT. Agr. Sci., V. 4, pt. I, p. 95-104. Bibliography, p. 104. (12) Freeman, George F. 191 7. LINKED quantitative CHARACTERS IN WHEAT CROSSES. In Amer. Nat., V. 51, no. 611, p. 683-689. (13) 1919. THE HEREDITY OF QUANTITATIVE CHARACTERS IN WHEAT. In GeneticS, V. 4, no. I, p. 1-93. Literature cited, p. 93. (14) Howard, Albert, and Howard, Gabrielle L. C. 191 5. ON THE INHERITANCE OP SOME CHARACTERS IN WHEAT. II. In Mem. Dept. Agr. India Bot. Ser., v. 7, no. 8, p. 273-285, 8 pi. (15) Kezer, Alvin, and Boyack, Breeze. 1918. MENDELIAN INHERITANCE IN WHEAT AND BARLEY CROSSES. Colo. Agr. Expt. Sta. Bui. 249, 139 p., 10. fig., 9 col. pi. (16) Levine, M. N., and Stakman, E. C. 19 18. A THIRD BIOLOGIC FORM OF PUCCINIA GRAMINIS ON WHEAT. In JoUT. Agr. Research, v. 13, no. 12, p. 651-654. (17) Little, C. C. i919. a note on the fate of individuals homozygol^ for certain color FACTORS IN MICE. In Amer. Nat., v. 53, no. 625, p. 185-187. 542 Journal of Agricultural Research voi. xix, no. n (i8) Melchers, Leo E., and Parker, J. H. 1918. ANOTHER STRAIN OP pucciNiA GRAMiNis. Kansas Agr. Exp. Sta. Circ. 68, 4 p. (19) MuLLER, Hermann J. I918. GENETIC VARIABILITY, TWIN HYBRIDS AND CONSTANT HYBRIDS, IN A CASE OB" BALANCED LETHAL FACTORS. In Genetics, V. 3, no. 5, p. 422-499. Literature cited, p. 497-498. (20) Nilsson-EhlE, H. 1911. kreuzungsuntersuchungen AN hafer UND weizen. II. Lunds Univ. Arsskr., n. F. Afd. 2, Bd. 7, No. 6, 82 p. (21) Stakman, E. C, and Piemeisel, F. J. 1917. a new strain OP puccinia GRAMINIS. (Abstract.) In Phytopathology, V. 7, no. I, p. 73. (22) Parker, J. H., and Piemeisel, F. J. 1918. CAN biologic forms OF STEMRUST ON WHEAT CHANGE RAPIDLY ENOUGH TO INTERFERE WITH BREEDING FOR RUST RESISTANCE? In Jour. Agr. Research, v. i, no. 2, p. 111-123, pL (23) Piemeisel, F. J., and Levine, M. N. 1918. plasticity op BIOLOGIC FORMS OP PUCCINIA GRAMINIS. In Jour. Agr. Research, v. 15, no. 4, p. 221-249, pL 17-18. (24) Levine, M. N., and Leach, J. G. 1919. NEW BIOLOGIC FORMS OF PUCCINIA GRAMINIS. In Jour. Agr. Research, V. 16, no. 3, p. 103-105. (25) Tschermak, Erich v. 1913. ubER SELTEnE getreidebastardE. In Beitr. Pflanzenzucht, Heft 3. p. 49-61, fig. 10. (26) ViLMORIN, M. H. 1880. ESSAIS DE CROISEMENT ENTrE BL^S diPP:6rENTS. In Bul. Soc. Bot. France, t. 27 (s. 2, t. 2), p. 356-361, pL 6-7. (27) 1883. EXPERIENCES DE CROISEMENT ENTRE DES bl6s DIFFJ^RENTS. In Btd. Soc. Bot. France, t. 30 (s. 2, t. 5), p. 58-63. PLATE 97- A. — Pollen grains of Marquis wheat. B. — Fi Marquis XKubanka (C I 2094), showing sterile grains. C. — Pollen grains of Kubanka (C I 2094). D. — Stems of Kubanka (C I 2094) grown under rust-epidemic conditions. Note absence of rust infection. E. — Fi Kubanka (C I 2094) X Marquis, showing normal uredinia. F. — Marquis, the susceptible parent. H. — Fj emmer (Minnesota 1 165) X Marquis, showing no normal uredinia. I. — Minnesota 11 65, the resistant emmer parent. Note that in the durum-common cross the Fj generation is susceptible while in the emmer-common cross the Fi gener- ation is resistant. Genetics of Rust Resistance Plate 97 ^^^^■^^^^■^^^■^^^■■■j^H ^^Pr-^^^^g a ■1^^^^^^^^ ^^^^■■H ^^^^— - M f-: mmmmgmm^m^^ ^^^^j^ ^^^^^^H ■Sfl BSSj ^^^^^^^t^^^m- : ^^^^1 2Zi_^^^^B ~ ^^^B 9B ■■ ^^^^^^^l^^^^^^^^^^l E - y.i'. -ipii '' BHBHim ^■IBHHH • •••• • • • o 1 Journal of Agricultural Research Vol. XIX, No. 11 Genetics of Rust Resistance Plate 98 Journal of Agricultural Research Vol. XIX. No. 11 PLATE 98. A, B, C. — Face and side views, respecti vely , of heads of lumillo (C I 1736), Fi lumillo X Marquis, and Marquis. The Fi heads are intermediate in density and have tipped awns. The outer glumes are keeled, although not so strongly as lumillo. D, E, F. — Face and side views, respectively, of heads of emmer, Minnesota 1165, Fi emmer X Marquis, and Marquis. The Fj generation approaches the emmer in some head characters and has intermediate awns. G, H, I.— Kernels of Marquis, F^ emmer X Marquis, and emmer. The Fj ker- nels are longer than Marquis, approaching those of the emmer parent in average length. PLATE 99- Representative heads of F3 families of the cross between durum and Marquis. A, B, C, D. — F3 families which were classified as dimims. Note that these repre- sent all types of head density. E- — Heads of an awnless F3 emmer family. F. — Four heads of an F3 plant which resembles common wheat in head shape and is rust-resistant. Genetics of Rust Resistance Plate 99 Journal of Agricultural Research Vol. XIX, No. 11 Genetics of Rust Resistance Plate 100 Journal of Agricultural Research Vol. XIX, No. 11 PLATE loo. A. — F3 family of a cross between emmer (Minnesota 1165), and Marquis, showing face and side view. This family proved rust-resistant and was also pure for keelless, compact heads and thrashed like common wheat. B. — Heads of an F3 family which resembled common wheat. It was rust-suscep- tible. C. — Heads of different plants of an F3 family of a cross between emmer (C I 1524) and Marquis. The plants varied from emmerlike and intermediate forms to those resembling common wheat as shown at the left. This family bred true for rust resistance. D. — Represents a very frequent sort of segregation obtained in the F3 generation. Many families gave only emmerlike and keelless, compact headed sorts which thrashed like common wheat. PLATE loi. Heads of resistant and susceptible wheat obtained in the F3 generation from the cross between Marquis and durum : A. — Head representing resistant emmer type F3 family 222-32, Kubanka X Marquis. B. — Head representing susceptible emmer F3 family 222-30, Kubanka X Marquis. C. — Susceptible common F3 family 181-24, lumillo X Marquis. D. — Resistant individual plant 186-13-5 which was obtained in the F3 generation of the cross between lumillo X Marquis. E. — Resistant durum F3 family 228-37, lumillo X Marquis. F. — Susceptible durum F3 family 186-15, Marquis X lumillo. Genetics of Rust Resistance Plate 101 Journal of Agricultural Research Vol. XIX, No. 11 Genetics of Rust Resistance Plate 102 ^kl^S.r-t*!.'^^ „ O vi) 00 CM to 1 »o I C: ^ o CD E ~ E CM « 5 o Cvt % CM CM C CD ^ ounces, and the Early Six Weeks to 6}^ ounces. The remaining portion of the tuber was then quartered lengthwise. Since 1 91 6 seed pieces have been cut with a special cutter which cuts a seed piece approximately hemisperical in shape, iX inches across the face and X inch deep. Such seed pieces weigh approximately }i ounce, or about 1 1 gm. Because of variation in depth of eyes, these seed pieces will differ somewhat in weight, the maximum variation within a variety usu- ally not exceeding 2 gm. and in many cases not exceeding i gm. Groups of 20 seed pieces seldom show variation of over 15 gm. While these seed pieces may appear small, they have apparently given just as strong plants as seed pieces cut in the usual manner and weighing iX ounces. Seed pieces have been planted by hand and covered to a uniform depth of 3 inches. Single-plot tests have been used entirely. Such tests carried on over a series of years should give just as reliable results as duplicate or triplicate tests conducted for a shorter period. Duplicating plots com- plicates the work of seed selection. It is not safe to select seed from any one plot to continue the line the following season, and it is difficult to select a small com.posite sample from two or three plots. It is for this reason that the single-plot test was decided upon. The 1 91 6 plots consisted of 4 hills, and the 191 7 and 191 8 plots of 20 hills. The latter is the standard-sized plot adopted here for this work. The selections from each variety are arranged in a group with from 9 to 12 plots in a row and from 9 to 12 rows^ There is no space between plots in the rows. The control plot consists of a single row near the center of the group running the length of the group of plots. Hills are planted 15 inches apart in rows 3 feet 9 inches apart. Seed is chosen from the plots before the tubers are assembled, and in this wa)^ it is secured from different portions of the plot. To illustrate, from a standard- sized plot 5 seed tubers are selected. Four seed pieces from each tuber are planted the following season in adjacent hills, and when this crop is harvested, effort is made to select one seed tuber from each group of four hills, or from each unit, as they are called later. The plots receive rather deep cultivation, are well ridged, and are commonly irri- gated once. At any rate they all receive the same cultivation and irriga- tion. All hills are thinned to single stems as soon as the vines are strong enough to pull, or from five to six weeks after planting. Sept. 1, 1920 Line-Selection Work with Potatoes 549 Since these small seed pieces give an average of only about 1% stems per hill, the thinning is not a very great task. Careful records are kept of the number and weight of marketable tubers and culls. Yields are computed upon the basis of a perfect stand. Naturally this favors the plot \vith an imperfect stand, for a missing hill tends to raise the yield of adjacent hills. Stand records are kept, showdng the position of missing hills. As yet there is no basis upon which to apply such records in correcting yields, but I hope to make use of these later. Careful notes are also kept upon vine characteristics. We are particularly inter- ested in those that indicate degeneration. Three general types of vines — vigorous, semi-curlydwarf or with curlydwarf tendencies, and curly- dwarfs — are recognized. The results of three seasons' work upon this project are presented here, not with the idea of proving the presence or the absence of high-yielding tuber lines within these 316 selections but rather for the purpose of call- ing attention to the difficulty of interpreting these performance records. If such a method of seed improvement is to be of real and practical value, especially in the hands of the average potato grower, it would seem that a 3-year test should give rather definite and dependable results. If the process is to be a longer one than this, then the method should be proposed for the potato specialist rather than for the rank and file of potato growers. GREEN MOUNTAIN TUBER LINES 300 TO 408 In variety tests at this Station, the Green Mountain types have on an average given the highest yields of all varieties grown. Varieties belonging to this group have, however, varied rather widely in yields. They are more susceptible to scab, more inclined to degeneration, and produce tubers less desirable as to commercial form than the Rural New Yorker; but withal the type is a very promising main-crop commer- cial potato. In Table II will be found the 3-year performance records of the 108 Green Mountain lines, expressed in yields of pounds per acre. Of the three groups of tuber lines, these have given the most noticeable varia- tions in yields. I/ine 332 ranks first mth a 3-year average yield of 27,849 pounds per acre, while line 344 has a 3-year average of 3,841 pounds. When ranked on 3-year average tuber production by weight, the 20 highest-yielding lines assembled in the last column of Table III have given an average yield of 25,000 pounds per acre, while the 20 lowest-ranking, appearing in the last column of Table IV, have averaged 12,083 pounds. In Tables III and IV it will be noticed that only a few of the lines appear as consistent high or low yielders. Several numbers will be found in both tables, showing that lines have swung from one extreme to the other. 550 Journal of Agricultural Research Vol. XIX, No. II Table II. — Three-year summary of yields of marketable tubers per acre from io8 Green Mountain tuber lines Tuber line No. 300 301 302 304 305 ■ 306. 307 308. 309 310 311, 312. 314 315- 316. 317 318. 319- 320. 321. 322. 323 324. 325 326. 327. 328. 329' ZZ°- 332. 333- 334. 335' 336. 337. 338. 339' 340. 341. 342. 343' 344 345' 346. 347' 348 349' 350' 351' 352. 353' 354 355 356 357 358 359 1916 Pounds. 453 453 034 776 357 680 61S 969 776 938 61S 19s 872 034 615 608 099 938 938 099 292 615 938 680 615 034 453 453 453 292 549 099 S18 453 615 034 195 164 195 357 807 969 19s 742 680 292 453 518 130 453 938 518 195 422 518 453 776 453 034 Pounds. 20, 442 23 22 26 22 3O1 21 27 26 20 23 19 25 26 31 24 33 30 17 22 22 26 32 37 29 34 37 15 28 27 26 39 37 24 33 36 15 29 19 28 29 23 35 9 14 10 17: 18 23 24 27 28 26 2,3 22 19: 18 23 27 230 300 946 553 199 836 411 017 442 230 513 553 482 128 623 915 199 190 765 300 017 057 168 734 380 632 796 340 876 946 026 632 623 915 238 269 513 340 269 230 309 292 402 685 654 584 230 159 875 495 946 451 765 048 584 694 876 1918 Pounds. 26, 017 16, 261 19, 071 14, 247 15,486 29, 340 14, 402 11,873 11,615 26, 482 21, 603 13,938 13, 705 10, 840 16, 626 ",736 23, 472 10, 269 21,371 30, 663 15, 099 16, 626 13, 938 21,836 19, 164 20, 907 14, 867 10, 453 13,936 30,318 23,230 29,424 25,320 10, 221 16, 028 14, 670 7,665 22, 738 10, 513 24, 262 8,595 11,615 26, 249 489 1,366 8, 001 6,968 30,715 12,776 15, 796 15,099 19, 048 20, 674 31, 128 20, 293 8,362 18,351 ",751 13, 936 3-year average. Pounds. 18, 970 16, 648 17, 468 17, 989 17,132 25, 073 15,951 15,417 16, 802 20, 287 18,816 15,215 16, 376 16,118 19, 789 17,322 24, 162 18, 135 17, 499 22, 842 15, 563 18, 086 19,977 24, 894 20, 171 22, 107 20, 984 12, 234 17,576 22,495 19, 241 27, 849 25,823 15, 099 20, 519 20, 647 11,730 23, 723 14, 073 21,986 14, 557 13, 938 24, 584 3,841 10, 482 9, 326 II, 691 21, 272 14,712 16, 802 18,970 20, 687 19, 938 27, SZ3 19, 19a 12,621 16, 570 15, 299 17,615 Sept. I, 1920 Line-Selection Work with Potatoes 551 Table II. — Three-year summary of yields of m,arketable tubers per acre from, 108 Green Mountain tuber lines — Continued Tuber line No. 360 361 362 3^3 364 3(>S 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 3S1 3S2 383 384 385 386 387 ■ 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 Yearl}^ average 1916 Pounds. 13.936 11,615 11, 615 12, 195 7. 549 10, 453 9,292 8,130 11,034 9, 292 10, 453 11,034 13,357 15, 099 6,388 11,613 9,872 8,516 6,969 10, 453 12, 195 8, 130 13, 938 12, 195 8,711 5,807 12, 776 14, 518 16, 261 9,872 15, 099 16, 259 12,776 13,357 7,549 5,807 7,549 10, 065 8, 130 9,872 8, 130 9, 292 7,549 6,388 15,099 11,615 12, 776 11, 613 9,291 11,335 Pounds. 25,553 22, 300 31,592 32, 766 21,836 23,230 34, 380 27,876 23,230 22, 300 16, 261 31,128 30,321 34, 845 20, 442 29, 343 33,915 33, 849 27,411 22, 765 25, 553 21, 029 27,876 35,309 26, 017 21,371 32,522 23,230 38, 581 24, 941 32, 522 30, 663 24, 623 25, 088 34, 380 34, 722 36, 238 21,836 30, 663 29, 269 28, 590 23,694 20, 907 26, 017 34, 845 30, 663 30, 199 23, 694 19, 298 1918 26, 595 Pounds. 20, 674 26,714 29, 966 30, 807 9, 292 11,150 14, 670 19, 745 9,808 13, 692 7,824 17, 604 31,296 24, 205 5,342 II, 980 30, 073 22, 997 19, 513 15, 099 13, 274 7,433 19, 100 22, 983 16, 261 4, 181 13, 008 12,079 27,873 12, 714 28, 362 32, 986 12,079 13,241 20, 674 12,079 25, 088 19, 513 28, 572 22,997 29, 269 11,873 6, 504 13, 203 22,713 15,159 22, 005 18, 584 11,615 3-year average. 17.493 Pounds. 20, 054 20, 202 24, 399 25,251 12, 896 14, 944 19.447 18, 583 14, 690 15,094 11,512 19,922 24. 991 24, 716 10, 724 17,645 24, 620 21,787 17, 964 16, 105 17,007 12, 197 20, 304 23, 495 16, 996 10, 453 19, 435 16, 609 27,571 15, 842 25,327 26, 636 - 16, 492 17,228 20, 867 17, 536 22,958 17, 138 22,455 20, 712 21, 996 14, 953 11.653 15, 202 24, 219 19, 145 21, 660 17.963 13. 401 18, 474 552 Journal of Agricultural Research Vol. XIX. No. ir Table III. -Tzventy highest annual rankings and 3-year average rankings from 108 Green Mountain tuber lines'^ 1918 Line No. *372- 354 *363 **348 ** 320 . 330. ■ *** *3 **376. . . *362... 332 . • **3o6 . . *400 . . *398 . . 390 . . ** *** *36i ** ** "310. 343- "300. "333- Yield. Pounds. 32, 986 31,296 31,128 30, 807 30,715 30, 663 30,318 30, 073 29, 966 29, 424 29, 340 29, 269 28,572 28, 362 27.873 26, 714 26, 482 26, 249 26,017 25.320 Line No. Yield. Pounds. 332 39. 026 388 38, 582 3Zi 37. 633 *327 37. 633 **324 37, 168 *396 36, 239 *336 36. 239 343 35.310 J383 35,310 **404 34, 845 **373 34, 845 *395 34, 723 *394 34, 380 *366 34, 380 *326 34, 380 376 33,916 *335 33,916 **3i7 33,916 *377 2h 849 354 33,451 I9I6 Line No. *338. 354- 388. 391- 306. *3i6. *345- 324- 317- 373- 404. 332- 390. 320. 33,3- *3&7. *355- *352- 348. 310. *3i8. *3i9- *323- *35i- ♦382. Yield. Pounds. 9,165 7,423 6, 261 6,259 5,680 5,680 5,680 5, 680 5.099 5.099 5. ,099 ,099 .099 ,099 519 14, 519 14, 519 14. 519 14, 519 13,938 13,938 13,938 938 13, 938 13, 938 3-year average. Line No. *** *** *** **390 *363 **3o6 332 354 391 333 1=372 324 ^^373 **376 **343 *362 **404 **3i7 *338 ♦383 **320 *33o Yield. Pounds. 27,849 27,571 27,333 26, 636 25,823 25,327 25,256 25,073 24,991 24, 894 24, 716 24, 620 24, 584 24, 391 24, 219 24, 162 23,723 23, 495 22, 842 22,495 ''The number of stars indicates how many times the line has appeared among the 20 highest-yielding lines. Lines not starred in columns 3 and 5 appear and are starred elsewhere in the table. Four lines appear three times, 12 appear twice, and 29 appear only once. Table IV. -Twenty lowest annual rankings and j-year average rankings from 108 Green Motintain tuber lijiesc- I9I8 1 1917 I9I6 3-year average. Line No. Yield. Line No. Yield. Line No. Yield. Line No. Yield. **344 **345 ***385 ***374 ***402 **347 ***38i **337 *37o **346 *!356 **34i **^64 *368 *334 *3i8 **328 **339 *3i4 *365 Po^mds. 489 1,368 4, iSi 5,342 6,504 6,968 7, 433 7,665 7,824 8, 001 8, 362 8, 595 9,292 9,808 10, 221 10, 269 10, 453 10, 513 10, 840 11, 150 344 346 345 337 328 *330 *3i9 347 :348 *357 356 *4o8 339 *3i2 *3io 374 *30o 402 381 385 Pounds. 9,292 10, 686 14, 403 15,332 15, 796 16, 261 17, 190 17, 65s 18, 584 18, 584 19, 049 19, 299 19,513 19,513 20, 442 20, 442 20, 442 20, 907 21,029 21,372 341 *395 374 *403 *378 *342 *3o8 402 364 *394 *396 *33i ""349 *367 *398 *400 3S1 *377 *384 Pounds. t;,8o8 5, 808 5, 808 6,388 6,388 6, 969 6,969 6, 969 7,550 7,550 7,550 7,550 7,550 8,131 8,131 8,131 8,131 8,131 8,517 8,711 **344. • **346.. ***38S-- **345-- *37o. . ***402 . . **347 • • 337- • ***38i . . **328.. **356.. **364.. *4o8. . *342 . . **339-- **34i . . *368.. *349-- *365-- Pounds. . 3, 841 9,326 • 10, 453 10, 482 • 10, 724 11,512 • 11,653 11,691 11,730 • 12,197 • 12,234 12,621 12, 892 ■ 13,401 • 13, 938 • 14, 073 • 14, 557 14, 690 14, 712 • 14, 944 n The number of stars indicates how many years the line has appeared among the 20 lowest-yielding lines. Lines not starred in columns 3 and $ appear and are starred elsewhere in the table. Four appear three times, 10 appear twice, and 28 appear only once. Sept. 1, 1920 Line-Selection Work with Potatoes 553 If with these records before us we attempt to point out apparently promising hnes, we must base our judgment either upon the frequency with which tiie line appears among the high-yielding lines or upon its performance as indicated by the 3-year average. The following 191 8 field notes are upon the lines that have appeared at least two out of the three seasons among the 20 highest or that rank with the 20 highest according to the 3 -year average. LINE NO. REMARKS. 332. First two units lacking in vigor but others vigorous. Variations may be due to soil. 388. Fourth unit a typical semi-ciu-lydwarf . 354. First unit semi-curlydwarf. 391. A good vigorous type. 2,^2i- A good vigorous type. 390. A good vigorous type. 363. A good vigorous type; third unit inclined to semi-curlydwarf. 306. First unit curlydwarf ; others fairly vigoroxis. 372. A good vigorous type. 324. A fairly good type; a few weak plants, but these may possibly be due to soil. 373. Variable as to vigor; some plants with curlydwarf tendencies, though all are fairly vigorous. 343. A good vigorous type with exception of first unit. 362. A good vigorous type. 404. A good vigorous type. 317. A good vigorous type. 338. A good vigorous type. 383. A very good type; some variation as to vigor, but this may possibly be due to soil. 320. Vigorous with exception of third unit, which is a typical semi-curlydwarf. 330. A good vigorous type. 348. Units I, 4, and 5, very vigorous; other two very good type. 310. First unit with curlydwarf tendencies; others very vigorous. Of these 22 lines, only 9 were noted as having uniformly good vine characteristics in 1918. Eight contained four or more hills with degen- erate tendencies, and 5 showed variations in vigor which possibly might have been attributed to soil but which may, on the other hand, turn out to be the first signs of degeneration. Yield records do not promise to be very effective in dealing with degeneration. It is true that those lines which have appeared frequently among the lowest-yielding are typical degenerates, but degenerate tendencies are also noticeable among the high-yielding ones. On the other hand, 7 of the 19 lines noted with good vine characteristics in 191 8 have each appeared once among the low- yielding 20 in Table IV. This indicates that low yields are not always associated with degeneration. The number of tubers produced in the hill is often suggested as a point worth considering in making seed-potato selections. In Table V will be found a summary of numerical tuber production in the 108 Green Moun- tain tuber lines. There is a maximum variation in the 3-year averages of 3.45 tubers per single-stemmed hill. The 20 ranking highest in 554 Journal of Agricultural Research Vol. XIX, No. II numerical tuber production (Table VI) have as a group a 3-year average record of 4.04 tubers per hill, while the 20 ranking lowest (Table VII) have averaged for the 3-year period 2.40 tubers per hill. The number of tubers produced in the hills varies from year to year, as will be shown in the yearly averages of the different lines. The 108 Green Mountain lines averaged 1.96 tubers per hill in 1916, 3.97 in 1917, and 2.78 in 1918. Variable soil moisture conditions during the setting period are no doubt responsible for a goodly portion of this fluctuation in tuber production, although delayed thinning may have contributed to the low yields of 1916. In a general way, numerical tuber production and production by weight rise and fall together, but this is by no means a fixed rule. Lines of this variety show the greatest variance in numerical tuber production, but this is no doubt accounted for by the more pronounced degenerate tendencies among a larger proportion of these lines. Degenerate lines do not fall off so much in total number of tubers produced, but a larger proportion of the tubers grade below marketable size. Once degen- eration begins in this variety, the line rapidly loses in vitality and yield- ing power. To illustrate this we might call attention to the record of line 345. This yielded 15,680 pounds per acre in 1916, while the average of all Green Mountain lines for that year was 11,335 pounds. In 1917 it yielded 14,402 pounds as compared with an average of 26,595 for ^^^ Green Mountain lines; and in 191 8 it yielded 1,366 pounds, as compared with an average of 17,494 for ^^^ these lines. Expressed in plus and minus variations, this line yielded 4,345 pounds above the average in 1916, 12,193 below the average in 1917, and 16,127 below the average in 1918. Table V. — Tllree-year average numerical production of marketable tubers per hill from 108 Green Mountain tuber lines Line No. 300. 301. 302. 304- 305- 306. 307- 308. 309- 310. in- 314- 315- 316. 317- 318- 319- 320. 321. 322. 323- 324- 325- 326. Yield. 1 3-56 ! 3 22 3 22 ' 3 30 ! 2 88 3 74 2 13 2 92 2 95 3 75 3 50 2 88 3 06 3 16 3 II 3 30 3 48 3 34 ; 3 36 > 3 88 1 2 78 1 3 16 3 50 3 75 i 3 27 3 61 3 43 Line No. 328. 329- 331- 332. 333- 334- 335- 336. 337- 338. 339- 340. 341- 342. 343- 344- 345- 346. 347- 348. 349- 35°- 351- 352- 353- 354- Yield. 3. 12 3-65 4. 18 3-29 4.64 4-36 2-75 3-54 3. 02 2.79 4. 06 2. 65 3. 21 2. 90 3.02 3-84 1. 19 1-95 2-95 2- 59 3-40 2. 56 3-09 2.87 2-93 3-93 Line No. 355- 356. 357- 358. 359- 360. 361. 362. 363- 364- 365- 366. i 367- 368. 369- 370- 371- 372- 373- 374- 375- 376. 377- 378. 379- 380. 381. Yield. 3-69 2.79 3- 40 2. 90 3- 04 3. 06 3-84 37 20 45 13 50 57 2.44 2. 16 3- II 3-42 4- 32 2. 70 3-14 4.41 3-53 3-38 2.68 2. 78 1.86 Line No. 382 383 384. 385. 386. 387. 388. 389 ■ 390 • 391 • 392- 393' 394. 395' 396. 397. 398. 399' 400 . 401 . 402 . 403 • 404, 405 ■ 406 , 407. 408. Sept. I, 1920 Line-Selection Work with Potatoes 555 Table VI. — Twenty highest- yielding of 108 Green Motmiain tuber lines when ranked on 3-year average numerical production of marketable tubers per hillO' Line No. *332 *404 *376 *333 *373 Yield. Line No. Yield. *383 4.27 *330 4.18 *338 4.06 *391 3.97 382 3-95 Line No. 396 *354 *320 406 *343 Yield. 3.93 3-93 3-88 3.86 3-84 Line No. *362 394 *324 310 *3o6 Yield. 3.84 3.81 3-75 3-75 3-74 "The IS starred lines also appear among the 20 highest-yielding when ranked on the 3-year average production of marketable tubers by weight. Ten of these lines, printed in bold-face type, appear among the 19 lines selected in 1918 as having uniformly good vine characteristics. Average of group, 4.40 tubers per hill. Table VII.- — Twenty lowest-yielding of loS Green Mountain tuber lines when ranked on j-year average numerical production of marketable tubers per hill o- Line No. Yield. Line No. Yield. Line No. Yield. Line No. Yield. *344 *38i *34S *385 307 1. 19 1.86 1-95 2.13 2. 13 *370 *4o8 *402 369 *349 1 2. 16 J 2. 30I 2. 40 : 2.44 2. 56 *368 *347 *339 J79 *374 2-57 2.59 2.65 2.68 2.70 389 387 334 321 380 2. 71 2. 72 2.75 2.78 2.78 a The 12 lines starred also appear among the 20 ranking lowest in weight of marketable tubers. None of these 20 low-yielding lines have good vine characteristics. Average of group, 2.40 tubers per hill. The 19 lines selected in 191 8 as having uniformly good vine charac- teristics have been grouped in Table VIII. The 191 8 average of these 19 lines is 26,260 pounds. In Table IX appear an equal number of the heaviest-yielding lines of 191 8, with an average production for the year of 29,267 pounds. The 108 Green ^Mountain lines gave an average yield of 17,493 pounds in 191 8. It will be noted that selection based upon vine characteristics alone came very near isolating the high-yielding lines for that season. The 19 lines appearing in Table VIII have as a group a 3-year average record of 23,270 pounds of marketable tubers per acre. The 19 lines standing highest when ranked on their 3-year average yield of marketable tubers have as a group an average of 25,131 pounds. The variation between the two groups, 1,861 pounds per acre, hardly seems sufficient to repay the labor involved in keeping yield records. Table VIII. — Yields of ig Green Mountain lines chosen upon vine characteristics alone as the best of the 108 lines of this variety in igi8 ^ Line No. Yield. Line No. Yield. Line No. Yield. Line No. Yield. 300 *3i7 *330 *332 *333 Pounds. 26,071 23. 472 30,318 29,424 25,320 *338 340 348 355 *362 Pounds. 22, 738 24, 262 30,715 20, 293 29, 966 *372 377 *383 =^390 *39i Pounds. 31,296 22,997 22, 983 28, 362 32, 989 394 396 400 *404 Pounds, 20, 674 25, 088 29, 269 22,713 •^ The lines starred appear among the 20 highest-yielding lines when ranked on the 3-year average production by weight of marketable tubers. Average of group, 26,260 pounds per acre. 556 Journal of Agricultural Research Vol. XIX, No. II Table IX. — Nineteen higlies t-yielding of loS Green Mountah 1 lines in igi ga Line No. Yield. Line No. Yield. Line No. Yield. Line No. Yield. 391 372 354 363 348 Pounds. 32, 986 31,296 31,128 30,807 30,715 320 330 376 362 332 Pounds. 30, 663 30,318 30, 073 29, 966 29, 424 306 400 398 390 388 Pounds. 29> 340 29, 269 28, 572 28, 362 27,873 361 310 343 300 Pounds. 26, 714 26, 482 26, 249 26, 017 " The lines printed in bold-face type also appear in the group with good vine characteristics. This leaves 10 of the 19 lines with vines suggesting degenerate tendencies. Average of group, 29,267 potinds per acre. RURAL NEW YORKER TUBER LINES 409 TO 516 Rural New Yorker has ranked very high in the variety tests conducted at this Station. Experience would lead to the belief that, with the possible exception of Russet Burbank, no main crop variety tested in these experiments will give more uniformly satisfactory results upon heavy types of irrigated land. Usually little difficulty is experienced in maintaining good, vigorous stock where the seed is carefully selected after the digger. The average yield of all strains of Rural tested in 191 6, 191 7, and 1 91 8 was 13,449 pounds of marketable tubers per acre. It will be noted that the average yield of the 108 tuber lines for the same period was 16,820 pounds. The 3-year performance record of these lines is presented in Table X. Here, as well as in Table II, it is interesting to note the variation in the average annual yield of all lines for the three seasons. Within any one season the variation between tuber lines is not so pronounced as it is in Green Mountain; but as yet none of these lines has reached the final stage of degeneration as several Green Mountain lines have. No effort is made to draw conclusions from Table X as a whole, but in Tables XI and XII an effort has been made to present a portion of the data included in Table X in such form that we may get some idea of the value of a 3-year performance record in isolating high- yielding lines. Table X. — Three-year summary of yields of marketable tubers per acre from 108 Rural New Yorker tuber lines Tuber line No. 1916 409 410 411 412 413 414 415 416 417 418 419 420 421 Pounds. 12,776 9,872 9,292 13,357 9,292 9,872 12, 776 12,776 8,130 4,065 11,034 7,549 8,711 Pounds. 22, 007 20, 907 22, 300 26,327 27, 876 23,230 25,919 21,371 24, 623 17,869 24, 159 24, 159 25,088 1918 Pounds. 17, 422 18,351 22,956 24, 939 15, 648 26, 585 23, 961 28, 108 22, 683 14, 867 20, 442 20, 442 18, 584 3-year average. Pounds. 17, 401 16,376 18, 182 21, 541 17, 605 19, 895 20, 885 20, 751 18,478 12, 267 18, 545 17,383 17,461 Sept. I, 1920 Line-Selection Work with Potatoes 557 Table X. — Three-year summary of yields of marketable tubers per acre from 108 Rxiral New Yorker tuber lines — Continued Tuber line No. 422. 423- 424. 425- 426. 427. 428. 429. 43°- 431- 432- 433- 434- 435- 436. 437- 438- 439. 440. 441. 442. 443- 444. 445' 446. 447- 448, 449. 450. 451' 452. 453 454' 455' 456' 457 • 458 459 460. 461. 462. 463 464, 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 1916 Pounds. 615 776 615 711 711 453 742 7,742 6 6 9 9 Pounds. 28, 340 291 872 969 549 969 130 292 969 388 453 453 615 613 292 938 776 195 034 034 388 807 453 615 195 292 292 034 292 453 034 711 130 130 549 807 872 807 226 549 807 549 969 130 807 130 711 453 230 386 836 159 836 070 765 230 946 430 482 031 230 230 088 836 442 230 300 623 088 048 513 482 002 452 694 230 430 017 017 402 765 938 230 300 765 442 907 159 261 230 048 836 694 676 513 300 977 261 119 977 442 694 513 300 513 230 1918 Pounds. 3-year average. 18 25 190 867 049 362 165 654 351 626 650 551 159 623 714 516 765 119 582 428 371 132 873 876 oiS 977 745 589 229 422 142 428 648 938 093 048 035 758 071 654 049 473 071 519 035 939 093 093 692 907 538 582 351 359 422 035 442 544 067 857 409 Pounds. 19, 048 957 017 636 on 647 054 711 756 961 626 431 654 112 602 419 964 467 628 139 512 035 280 627 539 964 189 630 233 447 767 422 196 394 887 570 175 557 894 604 325 372 019 445 058 764 215 595 053 761 982 815 422 621 165 693 697 558 Journal of Agricultural Research Vol. XIX, No. II Table X. — Three-year summary of yields of viarketable tubers per acre from lo8 Rural New Yorker tuber lines — Continued Tuber line No. 483 484 485 486 487 48S 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 Averas;e 1916 Pounds. io>453 11,615 9,292 8, 711 II, 615 10, 453 9, 292 5, 207 6, 969 10,453 11, 613 8, 711 II, 615 9,872 12, 195 13.' 938 io> 453 8, 130 8, 130 4, 646 4, 646 6, 969 6,388 4, 646 9,872 4, 646 5, 226 7.549 8,711 4, 646 8, 711 6, 969 6, 969 6, 969 9, 292 8,831 Pounds. 16, 725 18, 584 16, 261 12, 195 20, 907 17,654 18, 119 19, 048 18, 584 18, 584 22, 300 23, 694 18, 584 20, 442 19, 048 i9> 513 23, 694 19,513 21, 029 22, 300 22, 765 22, 765 20, 907 23,230 18, 584 24, 623 19, 048 21,371 21, 029 25, 088 21,836 22, 300 22, 300 22,300 22, 300 17, 190 1918 Pou^ids. 17,035 20, 538 11,247 17,551 18, 119 24, 049 21,371 19, 048 22, 494 13, 692 17,887 23,488 18, 826 16, 260 19,315 21, 760 20, 674 22,455 17, 190 14, 670 22, 765 19, 977 19, 513 20, 442 18, 093 20, 049 21, 516 17,115 12,389 25,552 19, 560 19, 560 20, 674 20, 442 17,654 20, 907 21, 864 1 19, 787 3-y«jr average. Poun 14 16 12 12 16 17 16 14 16 14 17 18, 16 15 16 18 18 16 15 13 16 16 15 16 15 16 15 15 14 18 16 16 16 16 16 15 ds. 737 912 266 819 880 385 260 634 015 243 266 631 341 524 852 403 273 699 449 872 725 570 602 106 516 430 263 345 043 428 702 276 647 570 415 409 16, 820 The data included in these tables indicate that with few exceptions high-yielding lines are not consistent high yielders, neither are all low- yielding lines consistent low yielders. The following brief field notes taken in 191 8 do not indicate that yield records will be found very satis- factory in dealing with degeneration. For instance, line 448 has a very good performance record. It ranked first in 1916, appeared among the highest-yielding 20 in 191 8, only missed b}- a narrow margin a place among the best 20 in 1917, and ranks seventh according to the 3-year average. But unless the vine characteristics are very deceiving, its yielding power will drop at least 50 per cent during the next two years. One-third of these lines with promising performance records contained some plants with degenerate tendencies. The field notes of 191 8 indicate to what extent degeneration appears among the lines that have ap- peared for at least two out of three years among the 20 highest-yielding or would be ranked with the 20 highest according to the 3-year average. Sept. 1, 1920 Line-Selection Work with Potatoes 559 Table XI. — Twenty highest annual rankings and 3-year average rankings from 108 Rural New Yorker tuber lines ° 1918 1917 1916 3-year average. Line No. Yield. Line No. Yield. Line No. Yield. Line No. Yield. *425- •• •■ **4i6 **443 **442 **434 430 ''414 **5io **45i *439 450 *465 ***4I2 **433 ^^424 *486 **^4i5 492 **448 *4i I Pounds. 28, 362 28, 108 27,876 27,873 26, 714 26, 6^0 26, 585 25, 552 25, 428 25, 438 25, 142 24, 939 24, 939 24, 623 24, 049 24, 049 23,961 23, 488 23, 229 22,956 *428 **422 *4i3 424 434 *43i 433 **446 412 > 452 415 ^451 *432 Sio /43 *437 421 442 /506 *4i7 Pounds. 34, 070 28, 341 27, 876 27, 387 27,031 26, 947 26, 482 26, 482 26, 327 26, 081 26, 081 25,920 25, 430 25, 430 2:;, 088 2-, c88 25, 088 25, c88 24, 624 24, 624 24, 624 448 *496 412 *449 *423 416 *409 *49S 457 450 424 422 *493 *485 *482 *4S6 *455 446 *49i Pounds. 13, 938 13, 938 13,357 12,777 12,777 12,777 12,777 12,777 12, 196 12, 196 12, 196 II, 615 II, 615 II, 615 II, 615 II, 615 11,615 11,615 II, 614 II, 614 ***4I2 . . . ^**443 • • ■ ^424. . . *^4i5--- •■416. . . **45i... **448... **433--. **434--- *!45o. . . *4i4. . . ..;42s--- '^*442 . . . **446... *452... **422... *430. .. 480... *492 . . . 419... Pounds. ■ 21,541 • 21, 139 21,017 . 20, 885 • 20,751 20, 630 • 20, 539 • 20,325 20, 238 20, 189 ■ 19' 895 19, 636 . 19, 628 19, 280 • 19, 233 19, 048 • 18, 756 . 18, 697 ■ 18, 631 • 18, 545 o The number of stars indicates how many times the line has appeared among the 20 highest rankingrs. Lines not starred in columns 3 and 5 appear and are starred elsewhere in the table. Three lines appear three times, 11 lines appear twice, and 30 Hues appear only once. Table XII. -Twenty lowest annual rankings and 3-year average rankings from 108 Rural New Yorker tuber lines °- 1918 Line No. '^457 **483 *509 **477 *46i **468 **490 **453 **5oo ***4i8 *423 *432 *4i3 *494 **463 *429 **48i *464 *475 **456 **444 Yield. Pounds. 0,758 1,247 2,389 2,544 3,473 3,692 3, 692 3,938 4,670 4,867 4,867 5,159 5,648 6, 260 6,519 6,626 7,035 7,035 7,035 7,035 7,035 1916 Line No. *484 456 *454 ■*447 *472 463 483 481 *5i6 *486 418 **473 *487 *489 490 *505 *482 *493 *495 **488 444 *465 **507 Yield. Line No. Pounds. 12, 196 13,938 14, 403 16, 003 16, 261 16, 261 16, 261 16, 726 17, 190 17,665 17, 869 18, 119 18, 119 18, 584. 18, 584 18, 584 18, 584 18, 584 19, 049 19, 049 19, 049 19, 049 19, 049 *428 418 *5o6 ;5io "^504 *5oi 500 507 *47i 473 *450 468 477 488 *47o *442 453 *43i *503 *439 430 *438 Yield. Pounds. 1,742 4,065 4,646 4,646 4, 646 4, 646 4, 646 5,227 5,227 5,808 5,808 5,808 5,808 5,808 5, 808 6,388 6,388 6,388 6,388 6,388 6,388 6,388 3-year average. Line No. Yield. **483 ***4i8 **477 *454 *484 **468 ;473 500 *509 '^■472 **456 **490 *46i *47i **463 **4 **4 475 474 ^507 Pounds. 12. 266 12. 267 12, 621 12,767 12, 819 13,058 13,761 13,872 14, 043 14,053 14, 196 14, 243 14,557 14, 595 14, 604 14, 634 14,737 14,815 14, 982 15,262 ° The number of stars indicates how many times the line has appeared among the 20 lowest rankinsB. Lines not starred in columns 3 and 5 appear and are starred elsewhere in the table. One appears thiee times, ij lines appear twice, and 37 lines appear only once. 560 Journal of Agricultural Research voi. xix. no. h Of these 21 lines, 7 have curlydwarf tendencies. Twelve appear among the best 33 lines chosen in 191 8 on vine characteristics alone. This leaves two, 443 and 452. in the doubtful list. IINE NO. REMARKS. 416. Third unit a curlydwarf; others vigorous though lacking in uniformity. 44.3. A very good type; first and second unit not especially vigorous. 442. A good vigorous type. 434. A good vigorous type. 510. A good vigorous type. 451. A good vigorous type. 450. Not a very good type. 412. Uniform and fairly vigorous. 433. A good vigorous type. 424. Vines not a good type. 415. A good type and fairly vigorous. 448. All have more or less curlydwarf tendency. 422. Lacking in vigor but type good. 446. First four units have curlydwarf tendencies. 414. First unit with curlydwarf tendencies. 425. Very good and vigorous type. 452. Not very vigorous but type good. 430. First unit very vigorous; others good. 480. Fairly vigorous. 492. A very good type. 419. Some curlydwarf tendencies. In Table XIII data are presented showing the 3-year average numeri- cal production of tubers per single-stemmed hill. In Tables XIV and XV a portion of this data is used to show the lack of correlation between yields in numbers and in weight, as well as the relation between numeri- cal tuber production and good vine characteristics. There is in Table XIII a maximum variation of 1.29 tubers per hill. Between the average of the 20 highest and 20 lowest (Tables XIV and XV) there is a variation of 0.79 tubers per hill. Considering that these are single-stemmed hills, this variation is rather pronounced. But the data are apparently no more reliable than weight records in pointing out promising lines. Low numerical tuber production does not always mean low production by weight, and such records are apparently of little value as a check on degenerate tendencies. Fluctuation^ in tuber production in different seasons has been even greater than the variation between lines in any one season. These 108 lines averaged 2.05 marketable tubers per hill in 1916, 3.78 in 1917, and 2.44 in 1918. In Table XVI the 33 lines with promising vine characteristics have been assembled with their 19 18 yields of marketable tubers. The average yield of these 33 lines is 21,142 pounds per acre. In Table XVII the 33 heaviest-yielding lines of 1918 are assembled for comparison. Their average yield is 24,109 pounds, while the 191 8 average of the 108 Rural New Yorker lines was 19,787 pounds per acre. The lines chosen upon vine Sept. I, 1920 Line- Selection Work with Potatoes 561 characteristics alone yielded well above the average of the 108 lines but did not equal the 33 heaviest-yielding lines of 19 18. Field notes show that 15 of these heavy-yielding lines of 19 18 have poor vine characteristics. This means that if mass selection based on tuber characteristics is practiced within these 15 lines their progeny \\i\\ be almost certain to contain some rather advanced degenerate types that will materially decrease the yield of the lines in 191 9. On the other hand, we can be almost sure that no well-advanced or low^-yielding degenerates will appear among the 19 19 progeny of those with good 1918 vine character- istics. It is interesting to go back beyond the 19 18 records and compare the 3-year performance of these lines assembled in Table XVI with the 33 highest-yielding Rural New Yorker lines when ranked upon their 3-year average production. The 33 lines in Table XVI have produced during the 3-year period an average yield of 17,740 pounds per acre. The 33 of the 108 Rural New Yorker lines ranking highest according to the 3-year average production by weight have yielded on an average 19,181 pounds. Between the two groups there is a variation of 1,441 pounds in favor of the 33 lines ranked upon their performance record. But when vine development so nearly foretells the yielding power of the line it would really seem that this is a more practical basis for selection than yields measured in either weight or number of tubers. Table XIII. — Three-year average numerical prodiiction of marketable tubers per hill frotn 108 rural New Yorker tuber lines Line No. 409. 410. 411. 412. 413- 414. 415- 416. 417. 418. 419. 420. 421. 422. 423- 424. 425- 426. 427. 428. 429. 430- 431- 432- 433- 434- 435- Yield. ■ Line No. 436. 437. 438. 439' 440. 441. 442. 443' 444. 445' 446, 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 Yield. 3-29 3.02 3.18 3. 00 2-95 3- 19 3-48 2.68 3-04 3-09 3-32 3. 10 3-07 2.88 3- 04 3. 21 3-38 3. 20 2-93 2.81 2.78 2.86 3-04 3- 29 3. 16 2. 70 2. 90 Line No. 463- 464. 465- 466. 467. 468. 469. 470. 471. 472. 473- 474- 475- 476. 477- 478. 479- 480. 481. 482. 483. 484. 485- 486. 487. 488. 489. Yield. 3. 00 3-07 3.02 2. 90 3-04 2. 60 2.77 2-95 3. 16 3. 21 3- 04 3.02 2-95 2-55 2. 92 2-95 3- 19 2-93 3.02 Line No. Yield. 490. 491. 492. 493- 494- 495- 496. 497- 498. 499. 500. 501- 502. 503- 504- 505- 506. 507- 508. 509- 510. 5"- 512. 513- 514- 515- 516. 2. 76 2.79 3. 16 2.83 2. 69 2-93 2.83 3-25 2. 78 2.79 2.83 3- 29 3.06 2. 72 2-93 2. 64 3.60 3-27 2. 74 2. 60 3-19 3" 3- 09 3-43 2-75 2.79 2-45 S62 Journal of Agricultural Research voi. xix. no. h Table XIV. — Twenty highest-yielding of lo8 Rural New Yorker tuber lines when ranked on 2-year average numerical production of marketable tubers per hill<^ Line No. Yield. Line No. Yield. Line No. Yield. Line No. Yield. *434 411 *4i9 506 426 3.74 3-65 3-63 3.60 3-59 *422 420 *425 432 *442 3.56 3-54 3.53 3-48 3.48 410 *433 513 *430 *452 3-47 3.43 3.43 3.41 3-38 *412 421 *446 *4i6 409 3.34 3-34 3-32 3-3^ 3-30 2 The II lines starred appear among the 20 highest when ranked on 3-year average production by weight of marketable tubers. Nine of these lines (printed in bold-face type) have uniformly good vine charac- teristics, and the other 11 have vine characteristics which suggest degeneration. Average of group, 3.47 tubers per hill. Table XV. — Twenty lowest-yielding of loS Rural New Yorker tuber lines when ranked on j-year average numerical production of marketable tubers per hill « Line No. Yield. Line No. Yield. Line No. Yield. Line No. Yield. 516 *483 428 *509 *468 2.45 2-55 2.55 2. 60 2. 60 *47i 443 494 *46i 2. 64 2.65 2.68 2. 69 2. 70 503 508 514 *490 476 2.72. 2.74 2.75 2. 76 2.77 *456 498 491 499 515 2.78 2.78 ^' 2. 79 2.79 2.79 a The 7 lines starred appear among the 20 lowest when ranked on 3-year average production by weight of marketable tubers. The 6 lines printed in bold-face type appeared among the 33 chosen in 19 18 as hav- ing the best vine characteristics, and these 6 have very good 3-year averages for tuber production by aweight. Line 428 appeared among the heaviest-yielding 20 in 1918 and only missed by a small margin a place among the 20 heaviest-yielding when ranked on 3-year average production. Average of group, 2.68 tubers per hill. ' Table XVI. — Yields for 1918 of 33 Rural New Yorker lines chosen upon igi8 vine characteristics alone as the best of the 108 lines of this variety « Line No. Yield. Line No. Yield. Line No. Yield. Line No. Yield. *4I2 V3 *4i5 *422 *425 *428 *430 *433 *434 Pounds. 24, 939 15, 648 23,961 17, 190 28, 362 18,351 26, 650 24, 623 26, 714 *442 *45I 465 *48o 486 488 489 491 Pounds. ^1,^13 25,428 24, 939 22, 409 24, 049 19, 048 22, 494 17,887 *492 493 495 *496 *497 498 501 503 Pounds. 23, 488 18, 826 19-315 21, 760 20, 674 22,455 22, 765 19. 513 504 506 507 508 *5io 513 514 516 Pounds. 20, 442 20, 049 21, 516 17. "5 25.552 20, 674 20, 442 20, 907 oThe 16 lines starred appear among the 33 highest-yielding lines when ranked on their 3-year average record. Lines 488 and 507 rank among the 20 lowest-yielding according to the 3-year average. Average of group, 21,142 pounds per acre. Table XVII. — Thirty-three highest-yielding lines of 108 Rural New Yorker tuber lines in igi8 °- Line No. Yield. Line No. Yield. Line No. Yield. Line No. Yield. 425 416 443 442 434 430 414 510 439 Pounds. 28,362 28, 108 27, 876 27,873 26,714 26,650 26, 58s 25,552 25,428 451 450 412 465 433 424 486 415 Pounds. 25, 428 25,U42 24,939 24, 939 24,^623 24, 049 24, 049 23, 961 492 448 411 436 501 417 1 489 498 Pounds. 23,488 23,229 22,956 22, 765 22,765 22, 683 22, 494 22,455 480 496 447 435 507 440 487 426 Pounds. 22,409 21,760 21,589 21,516 21,516 21,371 21,371 21,156 o The 18 lines printed in bold-face type appear in Table XVI. This means that of the 33 highest-yielding Rural New Yorker lines in 1918, 15 have poor vine characteristics. Average of group, 24, 109 pounds per acre. Sept. I, 1920 Line-Selection Work with Potatoes 563 EARLY SIX WEEKS TUBER LINES 517 TO 622 While these lines were selected from stock being grown under the name of Early Six Weeks, they all seem more typical of Red Early Ohio. At first a mixture of Early Ohio and Early Six Weeks was sus- pected, but observations on vine characteristics have led to the belief that the less vigorous lines, more typical of Early Six Weeks, are nothing more than degenerate types of Early Ohio. Such types are constantly appearing within vigorous dark green lines typical of Early Ohio. With this explanation I will still refer to these as Early Six Weeks lines. Varie- ties of the Early Ohio types have proved very satisfactory under soil and climatic conditions at the Station, yielding much better than other standard early varieties like Irish Cobbler and Early Triumph, and with less tendency to degeneration. The 3 -year performance record of these 100 lines is presented in Table XVIII. The average annual yield has followed about the same curve as in the other two varieties. There has apparently been ample variation between lines in any one season and in the 3 -year average to furnish a basis for selection. As in the other varieties, however, these variations in yield have in many cases been rather inconsistent. Table XVIII. — Three-year summary of yields of marketable tubers per acre from loo Early Six Weeks tuber lines Tuber line No. 517 518 519 520 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 545 546 547 548 417°— 20 4 1916 Pounds. I5>099 16, 261 10, 453 9,872 12,195 12,195 i3>357 14,518 12,776 13,357 11,615 13,938 15,099 12,776 16, 261 16, 261 11,615 6, 969 8,711 11,615 9,292 7,549 8,711 11,615 9, 292 12,776 6, 969 9,872 9, 292 11,034 Pounds. 20, 907 21,681 24, 159 24, 004 28, 340 20, 907 21,371 16,958 23, 230 19,977 22, 300 26, 482 22, 765 27,565 22, 997 25,320 14, 402 22, 765 25,088 22, 300 19,977 21,836 24, 159 22, 765 20, 648 18,351 17,654 25, 088 24, 624 26,017 1918 Pounds. 19,513 8,802 17,654 11,247 19. 804 13,692 13, 008 21, 271 20, 293 II, 150 12, 469 18,351 15, 099 22,713 18, 584 21,271 15, 099 16,725 18, 119 15,331 11,615 19, 048 19,977 21, 603 18, 119 22, 300 7,652 22,765 13, 473 25, 320 3-year average. Pounds. 18,506 15,581 17,422 15,041 20, 113 15, 598 15,912 17,582 18, 766 14,828 15,461 19, 590 17,654 21, 018 19, 280 20, 950 13, 705 15,486 17,306 16,415 13,628 16, 144 17,615 18, 661 16, 019 17,809 10, 758 19, 308 15, 796 20, 790 5^4 Journal of Agricultural Research Vol. XIX. No. II Table XVIII. — Three-year summary of yields of marketable tubers per acre from loo Early Six Weeks tuber lines — Continued Tuber line No. 549 550 551 552 553 554 555 556 558 559 560 561 562 563 564 565 566 568 569 570 57I' 572 573 574 575' 576 577 578. 579 580 581. 582. 583 584 585 586, 588. 589' 590' 591' 592. 593 594. 595 596 597 599- 600. 601 602. 603. 604, 605. 606 607. 608. 609. 610. 611. 1916 Pounds. 13*938 12, 19s 19. 745 17,422 10, 453 9,872 9, 292 13,357 12, 195 8,711 12,776 9,292 12,776 iS>099 11,034 15,680 11,034 11,615 15, 680 12,195 10, 453 12,776 9, 292 8,130 10, 453 10, 453 17,422 15,099 14, 518 15,099 13,938 12,776 14,518 14, 518 9,292 12, 195 12, 195 15,099 10, 453 12,776 13, 938 14, 518 15, 099 12, 195 14,518 11,615 11,615 10, 453 7,549 11,615 15,099 12,195 14, 518 13,357 13, 938 11,615 9,292 11, 615 13, 938 Pounds. 25, 553 21,528 24, 623 25,088 14, 228 25, 553 19, 745 20, 540 26, 186 17,654 17,422 20,442 19,978 13,473 23, 694 23,229 19, 745 37, 942 16,725 18, 584 18, 816 22,300 15,331 21,836 21,139 17, 190 26, 017 20, 907 21,371 23, 694 22, 300 24, 452 25, 553 ZZ, 683 32, 986 26, 017 28,908 31,360 23, 230 24, 159 24, 623 28, 968 26, 548 30, 663 30, 199 28, 340 25,088 26, 017 27,876 25, 533 28, 853 30, 810 29, 343 28, 340 29, 966 30, 199 29, 734 27, 876 24, 623 1918 Pounds. 24, 391 19, 745 25, 553 25, 553 14, 181 20, 907 13, 936 15,099 20, 049 17, 887 20, 538 29, 502 27,873 12,958 19, 048 20, 907 13, 203 14, 634 20, 210 15, 648 21, 139 18,351 15, 564 16, 028 19,977 14, 402 19, 513 13,241 14, 181 21, 271 16,397 13,241 10, 918 19, 071 19, 048 20, 210 14,425 20, 132 16, 028 20, 907 9,524 13, 938 17,654 21, 027 17,359 ^2,, 473 10, 685 17, 654 19, 048 10, 513 16, 870 21,516 20, 442 15, 403 18, 584 22, 005 21,516 16, 261 11,615 3-year average. Pounds. 21,294 17, 822 23, 307 22, 687 12,954 14, 324 16,332 19, 476 14, 750 16, 912 19, 745 20, 209 13, 843 17,925 19, 938 14, 660 21,397 17, 538 15,475 16, 802 17,809 13,395 15,331 17, 189 14, ois 20, 984 16,415 16, 690 20, 021 17, 545 16, 823 16, 996 22,424 20,442 19, 474 18,509 22, 197 16, 570 19, 280 16, 028 19, 141 19, 767 21,29s 20, 692 17,809 15, 796 18, 041 18, 157 15, 887 20, 274 21, 507 21,434 19, 033 20, 829 21,273 20, 180 18, 584 16,725 Sept. I, 1930 Line-Selection Work with Potatoes 565 Table XVIII. — Three-year summary of yields of m,arketable tubers per acre from, 100 Early Six Weeks tuber lines — Gintinued Tuber line No. 612 613 614 61S 616 617 618 619 620 62 1 622 Average 1916 Pounds. 12.195 15.099 11.034 13.938 13. 938 15.099 9. 292 11,034 8,711 12, 195 12,334 Pounds. 29, 734 25,088 29, 269 25, 088 25,919 27, 876 30, 663 25, 088 21, 518 16, 725 26, 485 24, 048 I9I8 Pounds. 15.331 16, 137 18, 584 13.473 12, 544 18, 582 19, 048 15. 564 18, 119 4, 646 19.358 17,344 3-year average. Pounds. 20, 054 17, 806 20, 984 16,531 17,467 20, 132 21, 603 16, 648 16, 890 10, 027 19,345 17,909 Table XIX. — Twenty highest annual rankings and j-year average ranking 9 from. loo Early Six Weeks tuber lines ^ 1918 Line No. *S6i. *t;62. 552- **55i- *548. *549- *546. *53i- *543- **6o8. *54i- **6o9. **6o4. **58o. **52 5- **533- *57i- *595- 591- **565- *554. Yield. Pounds. 29, 502 27,873 25.553 25.553 25.320 24,391 22,765 22,713 22, 300 22, 005 21, 603 21,516 21,516 21, 271 21, 271 21, 271 21, 139 21, 027 20, 907 20, 907 20, 907 Line No. *568. **584. *585. **589. 604. *595- **6i8. 608. **t;96. *6o7. 609. **6l2 . **6o5. **6i4. **593 • *588. **6o3 . *6o6. *597- *522. Yield. Pounds. 37,942 33, 684 32,987 31,361 30, 810 30, 663 30, 663 30, "9 30, 119 29, 966 29, 734 29, 734 29, 343 29, 269 28, 969 28, 908 28,854 28,341 28, 340 28, 340 1916 Line No. S5I- *577- *5i8. *532- 533- 565. *569- *578. 580. *594. 589- 618. 612 . 614. 603. *563- *530- *Si7- *583- *579- 584- 593- 596- 605. Yield. Pounds. 9,746 7»423 7,423 6, 261 6, 261 6, 261 5,680 5,680 5,099 5,099 5,099 5,099 5,099 5,099 5,099 5,099 5,099 5,099 5,099 4,518 4,518 4,518 4,518 4,518 4,518 4,518 3-year average. Line No. Yield. Pounds. **55i 23.307 *''552 22,687 **584 22,424 **589 22,197 **6i8 21,603 **6o4 21,507 **6o5 21,434 !568 21,397 *S95 21,29s 549 21,294 **6o8 21,273 *53i 21,018 **6i4 20, 984 *577 20, 984 *^533 20, 950 :6o7 20, 829 *548 20, 790 **696 20, 692 *585 20,442 **6o3 20, 274 "The number of stars indicates how many times the line has appeared among the 20 highest rankings. Lines not starred in columns 3 and s appear and are starred elsewhere in the table. None appear three times, 18 appear twice, and 31 appear only once. In Tables XIX and XX a portion of the data from Table XVIII is assembled for the purpose of a more critical study of the performance of the highest-yielding and lowest-yielding 20 of the 100 lines. 566 Journal of Agricultural Research Vol. XIX, No. II Table XX. — Twenty lowest annual rankings and 3-year average rankings from 108 Early Six Weeks tuber lines <* I9I8 Line No. Yield. Pounds. ^'fi^h\o T 4, 646 ***545 7,652 *5i8 8,802 *592 9>524 *602 10, 513 *599 10, 685 *583 10, 918 *527 II, 150 **520 11,247 *6ii 11,615 ***538 11,615 *S28 12,469 *6i6 12,544 **563 12,958 *524 13, 008 **566 13, 203 *578 13,241 *582 13,241 **547 13,473 *597 13, 473 *6i5 13,473 1917 Line No. Yield. Pounds. 563 13,473 *553 14, 228 *534 14, 402 **S73 15,332 *569 16, 725 62 1 16, 726 *525 16, 958 *576 17, 190 *56o 17,422 **S59 17,654 545 17, 654 ''543 18,351 *57o 18, 584 *S7i 18,816 **555 19, 745 566 19, 746 *527 19,977 538 19,977 *562 19, 978 **56i 20, 442 I9I6 Line No. Yield 535 545 *6oi *539 *574 *536 *540 621 559 538 555 573 *542 *6o9 *585 *6i9 547 *546 *554 *5o5 520 Pounds. 6, 969 6, 969 7,550 7,550 8,131 8,711 8,711 8,711 8,711 9,292 9,292 9,292 9,292 9, 292 9,292 9,292 9, 292 9,292 9,872 9,872 9,872 9,872 3-year average. Line No. ***62 I ***545 J553 573 ***538 *534 **563 *576 **555 **566 **559 *52 7 **520 *574 *528 *57o *535 *5i8 **547 *599 Yield. Pounds. 10, 027 10, 758 12, 954 13,39s 13,628 13, 705 13, 843 14, 015 14, 324 14, 660 14, 750 14, 828 15,041 15,331 15,461 15,475 15,486 15, 581 15, 598 15, 796 15, 796 a The number of stars indicates how many times the line has ranked among the lowest 20. Lines not starred in columns 3 and 5 appear and are starred elsewhere in the table. Three appear three times, 8 appear twice, and 38 appear only once. It will be noted from Tables XIX and XX that in either case the greater number of these lines have appeared only once in either of tliese groups. line 561 appears in the low-yielding groups of 191 6 and 191 7 but is the highest-yielding line of 191 8. It also appears among the 40 lines selected in 191 8 as having the best vines. While the performance record may not so indicate, it is undoubtedly one of the good lines of this variety. Aside from this one line, none of those that have ranked twice among the lowest 20 according to the 3-year average appear among those selected in 191 8 as having good \-ine characteristics. We may safely say that the perfonnance record is more reliable as a means of eliminating the lowest-yielding lines than in pointing out the higher- yielding ones. The following 191 8 field notes show that in this as well as in the other varieties yield records will not eliminate lines with de- generate tendencies. Sept. 1. 1920 Line-Selection Work with Potatoes 567 LINB NO. REMARKS. 552. A good dark green type. 551. A good dark green type. 608. A good dark green type. 609. A good dark green type. 604. A good dark green type. 580. A good dark green type . 525. Third unit lacking in color; others good. 584. A good dark green type. 589. First unit not a good type. 618. A good dark green type. 596. A good dark green type. 612. First unit lacking in color. 605. A good dark green type. 614. Third, fourth, and fifth units not good type. 593 . Color not good although the line is quite vigorous. 603. A good dark green type. 568. Lacking in color but fairly vigorous. 595. A good dark green type. 549. A good dark green type. 531. Last unit showing a tendency to bum. 577. A good dark green type. 607. A good dark green type. 548. A good dark green type. 585. Last unit light-colored and burning; others very good. Of these 26 lines which have appeared at least twice among the highest- ranking 20 or would be ranked among the highest 20 on their 3-year average production by weight, 17 have good records as to vine charac- teristics and 9 contain some vines of a degenerate type. It will be remembered that loss of vitality is indicated in this variety by a loss of color and a tendency for the leaves to bum about the edges in typical degenerates. In Table XXI will be found data upon the 3-year average numerical tuber production of Early Six Weeks. The maximum variation found is 1.78 tubers per hill. The annual average tuber production shows even greater variation. These 100 tuber lines averaged 2.12 tubers per hill in 1916, 4.04 in 1917, and 2.63 in 1918. The 20 highest and 20 lowest yielding lines when ranked upon tuber production in numbers have been grouped in Tables XXII and XXIII. The first 8 lines starred in Table XXIII are typical degenerates with bad yield records by weight. There are, however, 5 lines in Table XXIII which also appear in Table XXIV among the 40 with good vine characteristics in 191 8. Going back to Table XVIII, it will be noticed that these 5 lines have very good records when judged on tuber production by weight. It is true that none of them are high yielders, but they fall very close to the average of the 100 lines. Fourteen of the lines in Table XXII appear in Table XXIV, or among those selected in 191 8 as having good vine characteristics. 568 Journal of Agricultural Research Vol. XIX. No. II TabIvE XXI. — Three-year average numerical production of marketable tubers per hill from loo Early Six Weeks tuber lines Line No. Yidd. 519 520 522 523 524 526 528 529 530 531 532 533 534 535 536 537 538 539 540 541 543 3.02 2. 41 2. 84 2. 69 3.00 3-07 2.86 3- 23 2. 90 2.79 3- 16 3-45 3- 13 3- 27 3-47 3-69 2. 72 3.00 2.79 3-43 2.81 3-^5 3. 12 3-47 3- 02 Line No. 543 545 546 547 548 549 550 551 552 553 554 555 556 558 559 560 561 562 563 564 565 566 568 569 570 Yield. 23 33 93 86 36 i 39 ^ 65 : 40 i 00 74 i 80 , 06 i 79 i Line No. 571 572 573 574 575 576 577 578 579 580 581 582 583 584 58s 586 588 589 590 591 592 593 594 595 596 Yield. 3.18 3. 20 2.70 3- 15 3-27 2.84 3-95 3. 00 2.86 3- 23 3- 40 3-48 3- 20 3- 64 3-5° 3-3° 3.00 3-78 3-34 3- 40 3- 29 3-39 2-93 3-8r 3. 60 Line No. 597- 599- 600. 601 . 602 . 603. 604. 605. 606. ! 607. ! 608. 609. I 610. I 611. I 612. i 613. 614. 615. 616. 617. 618. 619. 620. 621 . 622. Yield. 25 22 15 29 27 47 85 25 32 50 95 90 II 97 43 23 40 77 II 16 52 C50 02 17 28 Table XXII. — Twenty highest of 100 Early Six Weeks tuber lines when ranked on j-year average numerical production of marketable tubers per hill « Line No. Yield. Line No. Yield. Line No. Yield. Line No. Yield. *608 *577 609 *604 *595 3.95 3.95 3.90 3.85 3.81 *568 :589 533 *548 *584 3- 80 1 3-78 1 3- 69 3.65 3.63 *549 :596 *618 *5S1 *585 3.63 3. 60 3.52 3.50 3-50 *607 582 532 541 *603 3.50 3-48 3.47 3-47 3.47 » The lines starred also appear among the 20 highest-yielding when ranked on 3-year average production by V eight. Thirteen of these lines (in bold-face type) have uniformly good vine characteristics; the othei 7 lines contained plants suggesting degeneration. Average of group, 3.65 tubers per hill. Table XXIII. — Twenty lowest of 100 Early Six Weeks tuber lines when ranked on J-year average numerical production of marketable tubers per hill^ Line No. Yield. Line No. Yield. Line No. Yield. Line No. Yield. *62I *Si8 ;545 *520 2. 17 2.41 2.41 2. 65 2. 69 *^73 534 *566 543 615 2. 70 2.72 2.74 2.77 2.77 *527 536 *57o *538 519 2.79 2.79 2.79 2.81 2.84 *576 524 560 578 526 2.84 2.86 2.86 2.86 2.90 " The lines starred also appear among the 20 lowest yielding when ranked on 3-year average production by weight. The lines printed in bold-faced type appear among the 40 chosen in 1918 as having good vine characteristics. Average of group, 2.71 tubers per hill. Sept. I, 1920 Line-Selection Work with Potatoes 569 Table XXIV. — Yields in igi8 of 40 Early Six Weeks tuber lines chosen upon igi8 vine characteristics alone as the best of the loo lines of this variety " Line No. Yield. Line No. Yield. Line No. Yield. Line No. Yield. 517 ^,5^9 *522 *526 *529 *532 536 540 543 *548 Pounds. 19, 513 17,654 19,804 20, 293 18, 351 18, 584 18, 119 19, 977 22, 300 25,320 *549 *55^ :55- 554 *558 560 *56i *562 564 565 Poutids. 24,391 25, 553 25, 553 20, 907 20, 049 20, 538 29, 502 27, 873 19, 048 20, 907 569 571 ^577 *58o *584 *59i *594 *59S 600 Pounds. 20, 210 21, 139 19,977 19, 513 21, 271 19, 071 20, 907 17. 654 21, 027 17, 654 601 *6o4 *6o5 *6o7 *6o8 *6o9 610 *6i8 620 *622 Pounds. 19, 048 21,516 20, 442 18, 584 22,005 21,516 16, 261 19, 048 18, 119 19,358 " The 26 lines starred also appear among the 40 highest-yielding when ranked on the 3-year average tuber production by weight. The others have very good records for the 3-year period. Average of group, 20,71^ pounds per acre. The vine growth of the other 6 hnes is reported in the following 191 8 field notes: ^^ REMARKS. 568. A little lacking in color but fairly vigorous. 589. A very good type with possible exception of first unit. 533. First unit light colored and burning; other plants green and vigorous. 596. A very good green type; first two hills lacking in vigor, but this is possibly due to soil. 585. Type fair with exception of last unit, which shows some biuning. 582. A light-colored, biiming type. 541. A good green type with exception of first unit, which is distinctly lighter and burning. In Table XXIV data are assembled to show the relation between good vine characteristics and tuber production in Early Six Weeks. The 40 lines with best vine characteristics gave an average yield of 20,713 pounds per acre in 1918, while the 41 heaviest-yielding lines of 1918 grouped in Table XXV averaged 21,350 pounds. The 191 8 average of the 100 Early Six Weeks lines was 17,344 pounds of marketable tubers per acre. In this variety vine growth is almost as reliable as yield records in pointing out high-yielding lines. The 3-year average yield of the 40 lines in Table XXIV was 19,515 pounds per acre, while the 40 lines ranking highest on 3-year average tuber production by weight yielded during the same period an average of 20,455 pounds. Will it pay to keep yield records when 40 per cent of the heaviest-yielding lines chosen with the aid of 3 years' data do not produce 1,000 pounds per acre more than do 40 per cent selected in less than 2 hours upon vine charac- teristics alone? 570 Journal of Agricultural Research voi. xix, no. h Table XXV. — Forty-one highest-yielding of too Early Six Weeks lines in 1918 "• Line No. Yield. Line No. Yield. Line No. Yield. Line No. Yield. 561 562 551 552 548 549 546 531 543 608 541 Pounds. 29, 502 27, 873 25, 553 25, 553 25, 320 24, 391 22, 765 22,713 22, 300 22, 005 21,603 604 609 525 533 580 571 595 554 565 591 Pounds. 21,516 21,516 21, 271 21,271 21,271 21, 139 21, 027 20, 907 20, 907 20,907 560 605 526 569 586 589 558 540 575 522 Pounds. 20, 538 20,442 20, 293 20, 210 20, 210 20, 132 20, 049 19, 977 19,977 19, 804 55° 517 577 622 584 IS::::::: 58s 601 618 Pounds. 19. 745 19, 513 19, 513 19, 358 19, 071 19, 048 19,048 19, 048 19, 048 19,048 a The 3 1 lines printed in bold-face type also appear in Table XXIV. This leaves 10 of these high-yielding lines with poor vine characteristics. Average of group, 21,350 pounds per acre. , CONCLUSIONS The data presented, so far as the author is able to judge, do not furnish very strong evidence of the presence of high-yielding lines within the common population of the varieties studied. The real test of the exist- ence of such lines is ability to maintain a high-yielding progeny by indiscriminate mass selection. Short performance records are fairly reliable in eliminating lines with low-yielding tendencies, but they are not so reliable as a basis upon which to select plus variations if such really exist. If degenerate tendencies exist within certain clonal lines and not in others, short performance records are of little value in eliminating the undesirable lines. Degenerate individuals appear with such persistent regularity within line selections as to become a real stumblingblock. If there are no exceptions to this rule, before a hill or tuber line can be increased to a point where it is of real value in commercial potato production it is almost certain to contain degenerate types which soon reduce its yielding power to that of the common population of the variety. The data presented will not justify an indorsement of the plan of clonal line selection as a practical method of potato-seed improvement. This does not mean that the hill-selection method of choosing potato seed is without merit. Generally speaking, high-yielding hills selected upon production by weight will produce the following season a high- yielding progeny. It does mean, however, that to be effective hill selec- tion becomes an annual task. I believe there is a more practical method of potato-seed improvement. Since certain vine characteristics are so closely correlated with yields, selection based on vine development alone promises to be more reliable than selection based on tuber production either by weight or number, and much more practical. Sept. 1.1920 Line-Selection Work with Potatoes 571 At present, selection based chiefly on vine characteristics seems to be the only hope in dealing with degeneration. The success of such selec- tion is measured by one's ability to identify intermediate types as well as typical curlydwarf degenerates. A special seed plot in which the seed pieces from each tuber are planted in consecutive hills promises to afford the best solution of the problems of degeneration and the maintenance of high-yielding variety populations. Seed for such a plot may be selected in one of two ways. Especially vigorous hills may be marked in the field during the growing season, and the crop from these may be dug and used for planting the special seed plot of the following season, or seed may be selected in the field at digging time or from the bin. The first method will insure the greater return from the first seed plot, especially if the field from which the seed is selected con- tains a large percentage of degenerate plants. In selecting seed in the field after the digger or from the bin, one may to a certain extent avoid degenerate types by observing certain tuber characteristics, (i) Choose seed tubers rather above the average size for the variety. (2) Select those that are long rather than short lor the variety and with both ends full and rounded, not pointed. (3) Choose those that are oval rather than round in cross section, or, in other words, flattened tubers. (4) Pick tubers with conspicuous and moderately deep eyes. The seed plot can best be planted by hand. The rows should be spaced the usual distance apart and the hills placed 12 or 15 inches apart in the row. On good, fertile, irrigated land, where the recommendations in regard to thinning are to be carried out, the hills should be spaced at 12 or even 10 inches apart. The tubers should be taken to the field whole and then cut as planted. The seed pieces may be cut rather small if seed is not abimdant. Single-eye seed pieces weighing from X to i ounce will give just as good results as larger pieces. These are conven- iently planted with a hoe, and if small should not be covered over 3 inches deep. Each tuber is cut by itself, and the pieces are dropped in con- secutive hills. Not only is it important to have the hills from each tuber in a group, but it is desirable to have these groups marked off one from the other. This may be accomplished in several ways. A good way is to drop a hill of corn between the groups. The important point, however, is to have the hills from each tuber in a group. This is the only way if intermediate types are to be eliminated from the seed plot by roguing. Mixtures are also easily detected where the hills are in tuber groups. The seed plot should be thinned by removing all but one stalk from each hill. Thinning should be done as soon as the vines are large enough to pull, usually from five to six weeks after planting. Proper compari- sons can best be made between hills with a uniform number of stems, and this is the real object of thinning. But single-stemmed hills also produce tubers of a more uniform size, and this is desirable, especially if the seed stock is to be offered for sale. Regardless of the merits of the case, the public commonly thinks of good potato seed as uniform-sized stock. 572 Journal of Agricultural Research voi. xix. No. n Roguing the seed plot is the most important step in the production of good seed. This consists in going over the field several times during the season and removing undesirable hills. In the first place, there may be variety mixtures in the seed plot. These mixtures are detected chiefly by variations in foliage, color of blossoms, amount of blossoms produced, and dates of ripening. There will also be some weak and degenerate types and diseased hills to be removed. The plot should first be gone over about blooming time, as variety mixtures are then very likely to be most conspicuous. A group of hills may show blossoms that are off type in color, or a certain group of hills may bloom more or less than the other plants in the plot. These hills should be dug and the tubers removed from the field. A little later the plot should be gone over again to catch hills lacking in vigor or infected with disease. In early varieties, this examination should be made just about the time the most vigorous vines show signs of maturing. At this time some vines will be found fully mature. This premature ripening is probably brought on by disease, and the hills should be removed. Such hills most frequently appear at random, but occasionally all the hills from a seed tuber may be diseased. In looking for weak plants, the hills from each tuber should be studied as a group. Weak hills will show a more crinkled foliage and usually less vigorous vines. It is well to remember that the intermediate types are the real trouble makers. They are usually as vigorous as normal hills, but show a slight crinkling of the foliage and are very often of a lighter shade of green. ^ In looking for these weak groups it is best not to work too near the row being examined. First look each row over at close quarters and then from two or three rows distant. In roguing out late varieties, the work must always be done before the vines are touched with frost. Remember that in roguing the tubers as well as the vines must be dug and removed from the seed plot. If the seed plot has been carefully rogued, the entire crop may be saved for seed. Occasionally a group of hills may produce tubers of poor form, and these may be discarded. The small seed from such a seed plot is just as satisfactory as the large, and in the great majority of cases variations in form are nothing more than fluctuations and as such are not transmitted to the following crop. In many instances the improvement secured by growing a variety in a special seed plot of this kind for one season may be apparent for two or three years after returning to the practice of selecting seed at ran- dom from the field at digging time or from the bin; but permanent improvement can be maintained only by continuing the special seed plot year after year. With most varieties it is allowable, if the roguing is carefully done, to increase the seed secured from the seed plot by planting it in a certain part of the commercial field and then saving iWHIPPLB, O. B. DBGBNBRATION IN FOTATOSS. Mont. Agr. Exp. Sta. Bui. Z30, 39 p., 16 fig. 119x9. Sept. 1. 1920 Line-Selection Work with Potatoes 573 from this part of the field seed for the entire acreage the following season. But if one is not absolutely sure he can pick out the interme- diate types in roguing, this practice may prove disappointing. Once the special seed plot is established, it may well be continued from year to year, the seed for the special seed plot of each year being selected from the seed plot of the year before. It is even advisable to go into the special seed plot and mark especially promising groups of hills to be dug to furnish seed for the seed plot of the following year. OCCURRENCE OF THE FIXED INTERMEDIATE, HOR- DEUM INTERMEDIUM HAXTONI, IN CROSSES BE- TWEEN H. VULGARE PALLIDUM AND H. DISTICHON PALMELLA^ By Harry V. Harlan, Agronomist, Office of Cereal Investigations, Bureau of Plant Industry, United States Department of Agriculture, and H. K. Hayes, Head of Section of Plant Breeding, Division of Agronomy and Farm, Managem.ent, Minnesota Agricul- tural Experiment Station INTRODUCTION The cultivated barleys belong to the genus Hordeum and are charac- terized by the presence of three single-flowered spikelets at each node of the rachis. The floret of the central spikelet develops a normal kernel in all forms of barley. The lateral florets, on the other hand, may develop normal or undersized kernels or may be sterile or even abortive. In classifying the varieties, Harlan (2)^ recognized four species, basing them on the degree of fertility in the lateral florets. Varieties of three of these species are concerned in the data presented herein. In all varieties mentioned, the central florets are long-awned. The 6-rowed parents in all hybrids belonged to the botanical variety H. vulgare pallidum, and include Manchuria, S. P. I. No. 20375,Sex-radigt, Odessa, Reid Triumph, Surprise, and a hybrid 6-rowed X 2-rowed barley. The lateral as well as the central florets of these varieties are long-awned and produce well- developed kernels. The 2-rowed parents all belong to the variety H. distichon palmella. The central florets are long-awned, while the lateral ones are awnless with rounded tips and are sterile. Of the varieties mentioned in the text, Svanhals, Garton, and Primus belong to this group. The intermediates are products of hybridization, and the selections are unnamed. All, how- ever, belong to the botanical variety H. intermedium haxtoni. In these the central florets are long-awned and produce normal kernels, while the lateral florets are awnless, as in the 2-rowed barleys, but fertile, pro- ducing undersized but viable kernels. The percentage of fertility in these spikelets is not so high as in the 6-rowed, but neither are they in- fertile, as are the 2-rowed. This form, obviously intermediate in char- acter between the 6-rowed and 2-rowed barleys, has proved to be constant and is as distinct in its genetic behavior as either the 6-rowed or the 2-rowed forms. The history of the intermediate barleys is very interesting, but not always clear. The first recorded observation found is that of John Hax- ton (5) in Scotland. In Morton's Cyclopedia of Agriculture of 1851 he mentions having observed such a plant in a field of " Bere." At the time • This paper is based upon experiments conducted cooperatively by the Ofi&ce of Cereal Investigations, Bureau of Plant Industry, United States Department of Agriculture, and the Minnesota Agricultural Experiment Station. ' Reference is made by number (italic) to " Literature cited," pp. 590-591. Journal of Agricultural Research, Vol. XIX, No. 11 Washington, D. C. Sept. i, 1920 ux I Key No. G-204 (575) 576 Journal of Agricultural Research voi.xix.No.it of this description he obviously had the heterozygous intermediate or a mixture of this and the homozygous intermediate, since he obtained all types from the seeding. Later, the homozygous form seems to have been isolated. In 1883, Drechsler of Gottingen sent Rimpau an intermediate which he had found in a field of H. disiichon palmella zeocriton and which apparently bred true from the time of its discovery. Both Rimpau (6, 7) and Komicke (4) observed the intermediate. According to Komicke's statement of 1885, he found the intermediate as an accidental hybrid in a field of 2 -rowed winter barley. At first the selection was heterozygous, but, according to the statement of his son (5), in 1908, it later became constant. Both Rimpau and Kornicke made mass selections which evidently included heterozygous forms, at least at first. It is probable that homozygous types were eventually obtained. These four cases of probable isolation of the fixed intermediate all occurred before 1900 and, therefore, before the importance of plant selec- tions in hybrids was generally recognized. It is worthy of note that all four were instances where accidental hybrids were observed, rather than the products of any of the numerous crosses made by those investigators. The history since 1900 is more surprising than that previous to 1900. The modern methods of breeding lead to the direct and ready isolation of this type, yet, so far as the authors know, it has not been reported. Since 1900, indeed, two plant breeders who have worked extensively with barley have assumed that a fixed intermediate does not occur in crosses between 6-rowed and 2-rowed barleys. Von Tschermak {8) says as late as 191 4 that the intermediate form with fertile lateral florets is heterozygous and, therefore, is never constant. Biffin (z), while not so sweeping in his statement, must have held a similar view. In 1907, he reported nine crosses in which a homozygous intermediate might occur. From the Fj progeny of one of these he selected a considerable number of intermediates of two types and planted them. All proved to be hetero- zygous. He did not test the progeny of the other crosses but assumed that they would behave in the same way. In the discussion he then states : The heterozygote, therefore, of forms with hermaphrodite staminate lateral florets is potentially hermaphrodite and of the type known to systematists as H. intermedium. In 1907 J. H. Wilson (9) reported an Fj generation of Standwell X Bere, in which he made the following observation: Examples occurred in which the grains of all six rows were fully developed, but the lateral ones were without awns. In this case Mr. Wilson probably noticed the homozygous type. It is interesting that he should have used as the 6-rowed parent the variety of barley in which Haxton found the first recorded intermediate. The observations of theseniorauthorinthesestudies extend overthepast 10 years. In the summer of 1909 he made a large number of hybrids on the grounds of the Minnesota Agricultural Experiment Station. A con- Sept.i.i92o Fixed Intermediate, Hordeum intermedium haxtoni 577 siderable number of these were between 2-rowed and 6-rowed parents. In 1 910 the Fi generation was grown, and in 191 1 the Fj generation. In the Fj generation of 1910 the heterozygous intermediate was very vig- orous, and a considerable number of selections were grown in 191 1 from the small lateral kernels to see if they would produce vigorous plants. In the F2 generation it was apparent that there were intermediates pres- ent which differed from those of 1910, and 56 selections of different types of intermediates were grown in 191 2. These came from 12 different crosses and included the progeny of combinations of 10 different 2-rowed and 7 different 6-rowed parents. A number of these proved to be homo- zygous, and others were isolated from the F3 generation by further selec- tion. Unfortunately, the records of 191 2 are incomplete, and the total number of homozygotes obtained can not be determined. It is also impossible to establish the complete list of crosses from which fixed in- termediates were obtained. Beyond all question, from one to five fixed intermediates were obtained from each of five different crosses. These crosses involve three different 2-rowed and five different 6-rowed parents. In these cases original material or complete records of the progeny are still at hand. The crosses were as follows : S. P. I. No. 20375 X Svanhals, Surprise X Primus, Primus X (2-rowed X 6-rowed), Garton 2-rowed X Sex-radigt, and Manchuria X Svanhals. There is a definite statement in the field records that a selection of zeocriton X Manchuria and another of Chevalier X South African were fixed intermediates, but there are neither specimens nor progeny records to confirm these statements. It will be seen that in the five crosses where the evidence is complete the 2-rowed parents were dense-spiked. Aside from the fact that those which have been preserved came from such crosses, there is no evidence in the data now at hand that indicates inability to secure intermediates from crosses in which both parents are lax-spiked. The common varie- ties of lax 2-rowed barleys have less vigorous lateral florets than the common varieties of dense 2-rowed barleys. It is possible that the prog- eny of crosses where the lax forms were used would be lower in fertility and the intermediates correspondingly less conspicuous. Hence, they would be less desirable and less likely to be retained as specimens. In 1 91 5, after the junior author became connected with the Minnesota Agricultural Experiment Station, it was decided to repeat one of the crosses from which fixed intermediates had been secured, in order to determine the inheritance and, if possible, to discover an explanation of the occurrence of this form. By making crosses and growing an occasional generation in the greenhouse in Washington in the winter and sowing the seed in Minnesota in the spring several generations have been studied since that date. The cross chosen for this purpose was Manchuria X Svanhals. The F2 and F3 generations are reported in de- tail. Other crosses were studied in which no intermediates were pro- duced; but since the data are negative, they have not been included, 578 Journal of Agricultural Research voi. xix. no. u although reference is made in the discussion to them and their signifi- cance. OCCURRENCE OF H. INTERMEDIUM AND OTHER SEGREGATES IN THE PROGENY OF A CROSS OF MANCHURIA AND SVANHALS BARLEYS The Manchuria parent in this cross is a selection of the common 6-rowed barley of the northern Mississippi Valley. The side florets are fertile and long-awned. The Svanhals parent is a well-known 2-rowed variety with long-awned, fertile, central florets and awnless, rounded, sterile lateral ones. The Manchuria is illustrated in Plate 103, C, and the Svanhals in Plate 104, A. This cross was made in the Washington greenhouse in the spring of 191 7, and the Fj generation was grown the same season at the University of Minnesota. In the fall of 1917, 100 of these Fi seeds were sown in pots in the Washington greenhouse. Of these, only 87 matured as Fj plants in the spring of 191 8. As soon as the grain was sufficiently mature to harv^est seed was again sent to Minnesota, where the F3 generation was grown. PHENOTYPIC CLASSES IN THE PROGENY The data on the F3 generation are reported in Table I, which needs con- siderable explanation. In the first place, the material itself is not simple. It is far from easy for a reader unfamiliar wdth barley to visualize six classes of segregates. Added to this difficulty is the fact that the F3 classifications are of varying accuracy and are phenotypic in character. The phenotypic classes of the table give a misleading impression by over- emphasizing the importance of the amount of fertility present. There are other characters than fertility which have much significance in the genetic groups. Fertility, however, is a definite, tangible, measurable condition, and the phenotypic classes founded on it are usable. In some instances the phenotypic classes of the F3 generation coincide with the genetic classes, while in others a phenotypic class contains more than one genetic class. There is a certain amount of unavoidable confusion in describing both phenotypic and genetic classes at the same time. For this reason, until the genetic classes are established the plants exhibiting similar phenotypic classes in their progeny are placed together and re- ferred to as genetic groups. It must be remembered in studying the table that the main object of the F3 classification was to determine the genetic classes of the Fj generation. It is well also to recall at this time that all classifications in this paper are based on the nature of the lateral florets, and to avoid endless repetition the modifying adjective is fre- quently omitted. In such cases the terms sterile, awn-pointed, high fer- tility, etc., refer to the lateral florets only. Table I represents the data as taken. The first two columns of the table contain the identification numbers and the description of the Fj plants. In the remaining colum.ns the F3 progeny of these plants are classified. These progeny fall into two major divisions, in one of which Sept. 1, 1920 Fixed Intermediate, Hoy deum intermedium haxtoni 579 all the kernels are of normal size, while in the other they are subnormal or absent. This second division is separated into plants which have lateral florets with awned or awn-pointed lemmas and plants in which the lemmas of the lateral florets are awnless and more or less rounded. Further subdivisions were based on the percentage of fertility of the lateral florets. The nature of the six resulting classes is illustrated in Plates 103 and 104. It is readily apparent in Plate lofi that the three classes with awned or awn-pointed lemmas on the lateral florets are easily separated. It is just as apparent in Plate 104 that the accident of season or nutri- tion can easily affect the classification of the segregates with rounded lemmas. Errors of classification in this case happen to be unimportant, for, as previously stated, the object of the F3 classification is only to determine the nature of the Fj parent. This is made evident in all cases by the classification used. To make this clear, one of the findings must be anticipated. Of the six classes in which the Fg plants are divided, three are of special importance in the interpretation. These are the fully fertile long-awned, the high-fertility awnless, and the no-fertility awnless classes. These represent the 6-rowed, the intermedium, and the 2 -rowed barleys. In all cases the classification is sufficiently accurate to determine which of these classes are to be found in the progeny of each Fj plant. Table I. — Classification of Sy F^ plants and their F^ progeny in a cross between Man- churia and Svanhals barleys according to the nature of their lateral florets GROTJP I, PLANTS WmCH GAVE ONLY 6-ROWED PROGENY Fj parent type. Number of individuals in Fs progeny, by classes. Kernels normal size, lemmas long- awned, fully fertile ce- re wed). Kernels subnormal in size or absent. F2 parent Lemmas awned or awn-pointed. Lemmas awnless. Total. No. High fertil- ity (80-100 per cent). Low fertil- ity (0-79 per cent). Fertil- ity 5-100 per cent {inter- me- dium). Fertil- ity less than SPer cent. Fertil- ity none (2- rowed). * 2 6-rowed 20 24 20 21 19 24 20 45 23 19 21 20 44 23 22 18 20 24 20 21 19 24 20 45 23 19 21 20 44 23 22 18 7 11 do do 13 23 38 41 do do do do do '\\ do 45 49 '^0 do do do 60 do 70 72 75 do do do 58o Journal of Agricultural Research Vol. XIX, No. II Table I. — Classification of S7 F2 plants vnd their F3 progeny in a cross between Man- churia and Svanhals barleys according to the nature of their lateral florets — Con. GROUP I, PLANTS WHICH GIVE ONLY 6-ROWED PROGENY- -contintied . Fj parent type. Number of individuals in F3 progeny, by classes. Kernels normal size, lemmas long- avraed, fully fertile (6- rowed). Kemels subnormal in size or absent. F, parent Lemmas awned or awn-pointed. Lemmas awnless. Total. No. High fertil- ity (80-100 per cent). Low fertil- ity (0-79 per cent). Fertil- ity 5-100 per cent (iyiter- me- dium). Fertil- ity less than 5 per cent. Fertil- ity none (2- rowed). 79 82 .do 20 45 34 21 19 15 20 do . . . 45 34 21 88 do 94 97 100 . . .do .do 19 15 .do PLANTS WHICH GAVE 6-ROWED AND INTERMEDIUM BUT NO 2-ROWED SEGRE- GATES 29 62 64 85 98 99 37 80 per cent fertile, 25 per cent fertile, 90 per cent fertile, 2^ per cent fertile, 90 per cent fertile, do 5 per cent fertile, awnless. i, awned. ;, awned. s, awned. ;, awned. ;, awTied. 12 18 16 7 17 7 7 23 20 20 15 13 23 30 2 I 12 6 10 10 10 5 6 3 47 44 46 34 40 39 43 GROUP 3, PLANTS WHICH GAVE 6-ROWED, INTERMEDIUM, AND 2-ROWED SEGREGATES 17 19 25 26 28 31 32 39 40 46 48 55 56 58 59 69 71 74 76 77 78 81 84 96 5 per cent fertile, av/ned. .. . 3 per cent fertile, awned. .. . None fertile, awned ....do ....do ....do ....do .. ....do do do do do do 9 per cent fertile, awned. .. . None fertile, awned do do 13 per cent fertile, awned. . . 12 per cent fertile, awned. .. None fertile, a\\Tied do do do do do 15 14 8 0 0 4 II 12 7 0 I 10 9 6 4 I 0 3 15 9 II 2 I II 14 16 10 0 I 9 8 13 12 3 0 8 10 19 5 3 0 8 16 7 13 0 0 10 9 12 13 I 2 II 12 13 8 I I 12 12 16 12 2 0 5 14 5 14 0 I 5 10 14 4 3 I 15 10 16 4 7 I 8 17 15 3 4 0 5 12 17 6 2 II 24 6 13 2 4 13 6 3 0 3 10 22 0 4 8 II 8 15 3 7 9 6 14 I 8 II I 13 0 19 16 7 16 0 7 15 II II 0 6 7 10 IS 0 0 10 41 41 23 49 50 44 45 47 47 39 47 46 44 49 50 26 44 45 39 45 47 44 42 Sept.i.i92o Fixed Intermediate, hordeuni intermedium Haxtoni 581 Table I. — Classification of 8y F^ plants and their F3 progeny in a cross between Man- churia and Svanhals barleys according to the nature of their lateral florets — Con. GROUP 4, PLANTS WHICH GAVE 6-ROWED AND 2-ROWED BUT NO fNTERMEDIUM SEGRE- GATES Fj parent type. Number of individuals in F3 progeny, by classes. Kernels normal size, lemmas long- awned, fuUy fertile (6- rowed). Kernels subnormal in size or absent. F, parent Lemmas a-wned or awn-pointed. Lemmas awnless. Total. No. High fertil- ity (80-100 per cent). Low fertil- ity (0-79 per cent). Fertil- ity 5- J 00 per cent {inter- me- dium). Fertil- ity less than SPer cent. Fertil- ity none (2- rowed). 9 34 35 36 47 54 65 68 None fertile, awned 5 17 6 7 II 12 13 12 7 II 15 19 14 26 25 25 T^ 16 21 19 4 9 4 II II 10 19 15 10 13 24 45 24 do do do do 47 47 47 43 38 43 do . . .do do 83 86 do do . GROUP 5, PLANTS WHICH GAVE ONLY INTERMEDIUM FORMS 10 10 per cent fertile, awnless . . 47 39 44 41 38 45 35 47 16 None fertile, awnless 7 46 18 do 44 30 66 do 5 46 do ». 38 60 93 95 do 15 20 per cent fertile, awnless. . . 35 GROUP 6, PLANTS WHICH GAVE INTERMEDIUM AND 2-ROWED BUT NO 6-ROWED SEGRE- GATES 5 6 None fertile, awnless 7 9 10 8 8 6 8 12 II 4 10 10 9 I 4 5 3 0 0 4 40 38 30 24 II 36 30 33 32 42 29 47 47 50 41 do 14 15 24 27 42 do do do 20 do 46 do 43 48 43 46 51 67 80 do do do 92 do 43 GROLTP 7, PLANTS WHICH GAVE ONLY 2-ROWED FORMS 4 8 None fertile, awnless 46 46 22 45 48 46 do 46 22 .do 2 24 73 89 .do 45 48 do 582 Journal of Agricultural Research voLxix.no. n GENETIC GROUPS INDICATED BY THE PROGENY CLASSES The F2 plants in Table I are arranged in groups according to the segre- gation shown in the F3 generation. Group I consists of plants which gave only 6-rowed progeny in the F3 generation. Since the parents were 6-rowed, the plants in this group are unquestionably homozygous for this character. Group 2 consists of plants which gave 6-rowed and intermedium but no 2 -rowed forms. The Fj plants evidently correspond to those in the high -fertility column of the Fg classification The classification of the high-fertility grOup is readily made and is highly accurate. Since this group gave 6-rowed, high-fertility, and intermedium segregates in ap- proximately a I to 2 to I ratio, it is safe to assume that the high- fertility plants are heterozygous for 6-rowed X intermedium which dififer by a single factor. Group 3 gave all three homozygous classes and all heterozygous classes as well. From this it is obvious that this group is comparable to the F2 generation and is heterozygous for the same factor differences as separate the original 6-rowed X 2-rowed forms. Group 4 gave 6-rowed, low-fertility awned, and 2-rowed forms in approximately a i to 2 to i ratio. Since this low-fertility awned corre- sponded in appearance to the Fj parents, it would seem that this group also was heterozygous for 6-rowed X 2-rowed, which in this instance, however, represents a single factor difference. It gave no intermedium forms, nor did it give the high-fertility awned or the low-fertility awnless forms. In other words, the intermedium character in both its homozygous and its heterozygous aspects was absent. The fact that this group as well as group 3 is heterozygous for 6-rowed X 2-rowed, even though more factors are involved in one case than the other, can be reconciled only on the basis that there are two types either of the 6-rowed or of the 2-rowed forms, only one of which carries the possibilities of pro- ducing intermedium forms. Group 5 consists entirely of Fj plants which gave only intermedium forms in the F3 generation. In other words, out of 87 Fj plants there were 7 homozygous for the the type which has been noticed so rarely in barley studies and whose very occurrence has been questioned so fre- quently. That the Fg plants showed no higher fertility was doubtless due to their being grown in the greenhouse. Those plants of the F, generation falling in the low-fertility class merely show the variation which exists. They do not cast any doubt on the genetic character of the F2 parents, because there were no 2-rowed segregates from any of the seven F2 plants. Group 6 gives homozygous intermediates, low-fertility awnless, and no-fertility awnless. The numbers in the classes are not significant in this case, for some sterile spikes of intermedium and the larger part of Sept. 1. 1920 Fixed Intermediate, Hordeum intermedium haxtoni 583 one heterozygous group are included with the homozygous 2 -rowed as phenotypes. The F2 plants in group 6, however, unquestionably are heterozygous for intermedmm X 2-rowed, a single main factor difference separating the intermedium and 2-rowed forms. Group 7 may be considered as composed of plants homozygous for 2-rowed, since with the exception of the two plants noted there was no fertility exhibited in the progeny. The seven groups just discussed consist of Fj plants assembled in groups because of their evident similarity of genetic constitution. In other words, these groups are genetic classes in contradistinction to the pheno- typic classes of the F3 generation. In Table II the data on these groups are summarized. The second column of this table gives the genetic constitution of each group. Except for the subdivisions ot group i, this classification was verified in the F3 generation. In the third column a hypothetical formula is suggested for each group. These formulas are based on a 2 -factor difference between the 6-rowed and the 2-rowed forms. The Manchuria 6-rowed parent is supposed to possess both fac- tors, while in the 2-rowed Svanhals parent both supposedly are lacking. Table II-^ — Summary of Fn plants of the cross between Manchuria and Svanhals Group No. as in Table I. Genetic constitution as shown by progeny. 1 Hypo- thetical formulas. Ex- pected Men- delian ratio. Ex- pected num- ber on basis of 87 plants. Num- ber ob- tained in 87 plants. Description of F2 plants. I I I 3 3 4 S 6 7 Homozygous for 6-rowed Heterozygous for 6-rowed X regressive 6-rowed. Homozygous for regressive 6-rowed. Heterozygous for 6-rowed X iv termed turn. Heterozygous for 6-rowed X 2-rowed. Heterozygous for regressive 6-rowed X 2-rowed. Homozygous for uiterviedium . Heterozygous for intermedium X 2-rowed. Homozygous for 2-rowed AABB AABb AAbb AaBB AaBb Aabb aaBB aaBb aabb I 2 I 2 4 2 I 2 I II 22 II 5 II 5 22 7 25 10 7 II S [Lateral florets long-awned, < fully fertile, kernels nor- [ mal size. Lateral florets short-awned, highly f ertile,kemels small. Lateral florets awn-pointed to short-awned, no fertility to low fertility. Lateral florets awn-pointed to short-awned, no fertility. Lateral florets enlarged, awn- less, low fertility to no fer- tility, kernels small. Lateralflorets awnless, some- what enlarged, no fertility. Lateral florets small, awn- less, nofertiUty. This hypothesis accounts for the results very well. The 2-rowed segregates of group 7 are homozygous in the absence of both factors. The fixed intermediates or H. intermedium forms of group 5 are homo- zygous for the presence of one factor and for the absence of the other. The heterozygous groups, 2, 3, 4, and 6, are all heterozygous for one or both factors. Groups 2, 4, and 6, which are heterozygous for only one factor, give F3 progeny of limited distribution. Group 3, which is heterozygous for both factors, gives all classes in the F3 generation.- 584 Journal of Agricultural Research voi. xix. No. n Group I is the only group offering any complications, and these are of no particular complexity. If the 2-factor hypothesis is tenable, group i must include three different classes of 6-rowed forms, two of which are homozygous. If the factor A A is considered epistatic to the factor BB, the three classes of group i become phenotypic for 6-rowed and may be considered as a single class. In genetic constitution, however, the homozygous 6-rowed segregate AAbb differs from the parent 6-rowed AABB; and, inasmuch as the BB factor has been lost, the form AAbb is referred to here as a regressive 6-rowed. If this regressive 6-rowed AAbb were crossed on the Svanhals parent aabb, there would be no possibility of securing the homozygous intermediate aaBB. Heterozy- gous types corresponding to such a cross probably are found in group 4, which is supposed to be heterozygous for this regressive 6-rowed AAbb X the 2 -rowed aabb. On the other hand, group 3 is heterozygous for the parent 6-rowed AABB X the 2 -rowed aabb. It will be seen in Table II that the plants obtained in each group came very close to the expectancy. The numbers of 6-rowed and 2 -rowed forms produced coincide exactly with the expected numbers,, while the number of fixed intermediates is only two greater than in the calculated ratio. The nature of the factors is better discussed in connection with the description of the genetic classes which is given in Table II. It will be seen from the table that except in group i these various classes differ in appearance as well as in genetic constitution. In some instances the separations are easily made on appearance alone, while in others only a part of the plants belonging to a group can be easily identified. The first four groups have lateral florets, the lemmas of which bear awn points or awns of varying lengths. Group i, of course, presents no difficulties, for 6-rowed segregates are unmistakable. Group 2 is almost as definite, for there are no other high-fertility segregates with small- kerneled, awn-pointed lateral florets. Group 3 is readily separated from group 2 by the lower fertility of the individuals in group 3. The sterile individuals of group 3, however, are easily confused with the individ- uals of group 4. Probably 80 per cent of such plants in group 3 can be identified by the more obtuse lemma tip on the lateral florets. Groups 5, 6, and 7 differ from the other groups in the absence of awns on the rounded tips of the lateral florets. These conditions are not absolute, in that an occasional floret may possess an awn. In this case, however, other florets on the same spike have the characteristic rounded obtuse tips. Group 5 is characterized by the possession of lateral florets which, even when sterile, are larger than those of the 2-rowed. Under field conditions higher fertility probably would have been present in the members of this group. When the Fg material was being studied, it was possible, however, to isolate the sterile spikes of intermedium forms by inspection. In group 6 the lateral florets of many individuals are sept.i,i92o Fixed Intermediate, Hordeum intermedium haxtoni 585 obviously larger than in the 2-rowed segregates, but as a rule they are not so large as in group 5. Groups 6 and 7 can be accurately sep- arated only by the breeding test. It will be seen, then, that although fertility is the basic distinction, it is expressed in other ways than in the production or nonproduction of kernels. Indeed, the percentage of fertility probably varies more with environment than do the morphological differences. Such varia- tion, however, does not imply any lack of reliabihty of the fertility factors. According to the factor hypothesis, the first four groups have the inherent possibility of producing 6-rowed segregates. Only plants belonging to these four groups have lateral florets the lemmas of which are awned or awn-pointed. The A factor, either as AA or Aa, is found only in these groups. This factor, then, must be associated with the possibility of awns. When A is homozygous, fully fertile, long-awned florets which develop kernels of normal size are produced. When A is heterozygous, all lateral florets are short-awned or awn-pointed. When homozygous, it is epistatic to B, the spike being normal 6-rowed irrespective of the condition B. When heterozygous, A has little effect on fertility, and the amount of fertility present in this case is in direct relation to B. The lateral florets of AaBB are quite fertile, those of AaBb occasionally so, while those of Aabb as a class are sterile. Aa may have a slight effect on fertility, for AaBB has a higher percentage of fertility than aaBB. The factor BB, then, is a fertility factor which at its highest expression produces a lower percentage of fertility than AA. It is not accompanied by the presence of awns. The lateral florets under its stimulation produce undersized kernels, the largest form of floret seeming to be associated with AA. INDICATION OF A THIRD FACTOR The presence of a third factor of fertility was first suggested by a study of the data in Table I. Column 2 of that table shows the per- centage of fertility of the Fg parents. In group 2, three plants exhibited a much lower fertility than the other four. vSince one of these was abnormal, there were, excluding this plant, two low-fertility and four high-fertility plants. It is difficult to measure the inheritance of these variations in this group. The percentage of fertility of the intermedium segregates in the F3 generation of four lines was studied. The F3 inter- medium segregates of the others were missing. Three high-fertility lines gave almost identical fertility of 37+ per cent. One low-fertilfty line gave 29.5 per cent of fertility. In group 3, five out of twenty-five plants showed some fertility. The F3 plants were not studied for differ- ences in percentage of fertility. In group 5, two out of seven plants had fertile lateral florets. In this group the progeny were studied carefully. This is the only group 586 Journal of Agricultural Research Vol. XIX, No. ir where this can be done readily, because the others, where the differences in percentage of fertility occur, are heterozygous for one or both main factors. If there is a third factor, the Fj generation indicates that it is expressed here as a i to 3 ratio with the heterozygous forms indis- tinguishable from those homozygous for the absence of the factor. By the same reasoning, lines 10 and 95 in Table I, which showed fer- tility, should be homozygous for the presence of the factor and should exhibit a higher percentage of fertility than the progeny of the other five lines. The F3 progeny of lines 10 and 95 averaged 54.6 and 48.4 per cent of fertility in the lateral florets, while the progeny of the other lines v/ere, respectively, 35.7, 33.2, 30.3, 25, and 17.8 per cent fertile. The data indicating a third factor in fertility are summarized in Table III. The evidence at hand is not more than an indication of a minor factor, yet the material itself is far more suggestive of such a factor than the evidence. Table III. -Summary of data indicating a third factor of fertility in the cross between Manchuria and Svanhals Group No. Genetic constitution as shown by progeny. Hypothetical formulas. Ex- pected mendel- ian ratio. Num- ber of plants in group. Ex- pected number of plants by sub- groups. Num- ber oi plants ob- tained. 2 3 4 Heterozygous for 6-ro\ved X inter- medium, Heterozygous for 6-rowed X 2- rowed Heterozygous for regressive 6-rowed X 2-rowed fAaBBCC . . AaBBCc... [AaBBcc... . I'AaBbCC... JAaBbCc... 1 AaBbcc . . . AabbCC... AabbCc. . . [Aabbcc. . . . faaBBCC. .. aaBBCc... [aaBBcc . . . 2 4 2 4 8 4 2 4 2 I 2 I 6 ■ 25 1 ^° n f ' I 6 } ^ 2 > 20 2 6 5 Homozygous for iritermedium 2 2 } s DISCUSSION OF RESULTS FIXED INTERMEDIATE, H. INTERMEDIUM Since tne purpose of this study was to gain a better understanding of the occurrence of H. intermedium, this form is of greater interest here than any of the other segregates. From the standpoint of historical and present interest there are three questions which can now be answered. There is no doubt that such forms occur; they have been found to be stable; and in the particular cross reported here in detail the ratio of appearance is approximately i to 16. The stability of this form was established by obser^'^ations on interme- diates obtained from several crosses. The field experiences fit in well Sept. 1, 1920 Fixed Intermediate, Hordeum intermedium haxtoni 587 with the 2 -factor hypothesis suggested. The stability of the homozygous intermediate has been thoroughly tested. Some of these have been grown since 191 2 under widely varying conditions with no indication of reverting to either the 6-rowed or 2 -rowed parental type. Many hybrids have been made with the intermedium form as one parent. There are complete rec- ords of the progeny of crosses on five 6-rowed and on two 2-rowed forms. No 2-rowed segregates appeared among the progeny of the 6-rowed X in- termedium crosses, and no 6-rowed segregates appeared among the progeny of the intermedium X 2-rowed crosses. Every hybrid studied so far indi- cates that H. intermedium has a genetic rank equal to H. vulgare or H. distichon. In field culture there is always variation in the amount of fertility present in the lateral florets of the intermedium form. This is true even on a single plant. The earlier, better-nourished spikes in some strains may have as high as 90 per cent of the lateral florets fertile. The later spikes of the same plant usually have a diminishing percentage, while the last-appearing may have lateral florets which are entirely sterile. Such variations, however, are the result of variation in nutrition, for their progeny are uniformly intermedium in type. There are variations of another sort in the intermedium forms which have more significance. Frequent mention has been made of the fact that the fertility of the different classes may be expected to vary with the parents used. From crosses showing more vigor in the lateral florets of the segregates than those of the Manchuria X Svanhals, very vigorous intermediates may be isolated. From those crosses which show less fer- tility in the segregates, homozygous intermediates probably may be iso- lated in which only occasional kernels may be produced. Instances may be possible where a potential intermediate may exhibit no fertility what- ever. The form atterhergii shown in Plate 106 is probably an infertile inter- medium barley. This form is somewhat anomalous in the taxonomy of barley. It has never been grown by the authors, but from the statements of those who have grown it, it appears to have shown no fertility. The illustration is a photograph of a spike presented to the United States Department of Agriculture several years ago by Mr. E. S. Beaven. It has greatly enlarged lateral florets which closely resemble those of spikes of H . intermedium, in which potentially fertile lateral florets are sterile because of environment. REGRESSIVE 6-ROWED FORM The recognition of the two different forms homozygous for the 6-rowed character offers a field for further investigation. The regressive 6-rowed AAbb is apparently not optically distinguishable from the AABB 6-rowed. There is no method known to the authors of separating these by inspec- tion. In hybrids the regressive 6-rowed form should behave quite 588 Journal of Agricultural Research Voi. xix. No. h differently. If this regressive type were crossed on the intermedium form the results, according to the hypothesis advanced, would not be the same as when the Manchuria is so crossed. The Manchuria X intermedium gives 6-rowed, high-fertility, and intermedium segregates in a i to 2 to i ratio in the Fj generation. The regressive 6-rowed X the intermedium- form should give all classes, including 2-rowed forms, in the Fg generation. On the other hand, if the regressive 6-rowed were crossed on the Svanhals, no intermedium segregates would be expected. Unfortunately, neither of these crosses has been made, but the latter condition has been met in other crosses. Strains of Odessa and Reid Triumph when crossed on Svanhals produced no intermediates, indicating that the regressive 6- rowed AAbb does occur. The progeny of these crosses segregates in the same way as the group which is heterozygous for the regressfive 6-rowed X 2-rowed in Table I. Incidentally, there is evidence that the strains used, of both Reid Triumph and Odessa, originated as selections from a previous hybridization. No evidence is at hand to indicate whether or not the regressive 6-rowed type AAbb occurs elsewhere than in the progeny of hybrids. Because fixed intermediates have been secured from hybrids in which several old agricultural races of 6-rowed barley were used as one parent, it is assumed that forms such as the Manchuria are the normal 6-rowed forms. It is not known how frequently the regressive types occur among our agricultural sorts, nor whether they are associated with other char- acters than the failure to produce intermediates. Heritable variations in the amount of fertility in the lateral florets of 6-rowed barleys have been noticed, however, and since the conception of a regressive 6-rowed type offers a possible explanation of their behavior, a statement con- cerning these unusual varieties may be of interest. In 1909 the senior author noticed a decided tendency to sterility of lateral florets in the agronomic varieties known as Mansfield and Summit. In each case the earliest spikes to appear were normal 6-rowed spikes. The lateral florets were highly fertile, long awned, and produced large kernels. In the later spikes, especially where the plants produced several culms, the lateral florets exhibited increasing sterility. Late spikes, in which none of the lateral florets were fertile, were found in many plants. One of these barleys, Mansfield, was the progeny of a cross between a 6-rowed and a 2-rowed barley, made at Guelph, Ont., in 1891. The peculiar behavior of this variety was so striking that selections of fertile and infertile spikes were made in 1909. These were planted in 1910, but, as would be expected, there was no difiference in their inheritance. Owing to seasonal conditions the percentage of spikes with no fertility in the lateral florets was unusually large in 1910, the sterile spikes being approximately equal to the fertile in number. Spikes from a single plant are shown in Plate 105, D. It will be seen that the sterile florets are sept.i.i92o Fixed Intermediate, Hordeum intermedium haxtoni 589 long awned and that there is no approach toward the homozygous inter- mediate. The second variety, Summit, which for other reasons is thought also to be of hybrid origin, produced spikes with infertile lateral florets much more rarely. It is suggested that these barleys may be forms of the 6- rowed segregate AAbbcc, homozygous for the absence of the two secondary factors of fertility. POSITION OF FERTILE LATERAL SPIKELETS Aside from the question of the amount of fertility in the lateral florets, the location of fertile florets on the spike is of interest. In the common varieties of barley the largest kernels are produced about one-third of the distance from the base of the spike to the tip, in both the central and the lateral spikelets. The longest awns are attached to these florets, and they exhibit a progressive decrease in length toward the tip of the spike. The fertility of the homozygous intermediate runs in the reverse order. When only a few kernels are produced they are found in the upper third and gradually extend downward from the tip in the more fertile intermediates. The most fertile lateral florets in the Mansfield variety are found near the center of the spike. ABERRANT LINE NO. 2)7 In Table I line No. 37 was found to diflfer greatly from the other members of group 2 in both appearance and fertility. It is much lower in fertility than the other individuals and, with the exception of one or two florets, is similar in appearance to the fixed intermediate. This plant is the only one of the 87 whose appearance is difficult to explain. It is not believed that the hypothesis of its being homozygous for the absence of the third factor is adequate explanation for its wide departure from type. For this reason this line was excluded in the later discussion. It is possible that the supposed Fj plant is the result of a natural hybrid. The appearance of the forms in any class varies with the parents used. Some combinations result in greater fertility in the various classes, others result in much less. If the Fi kernel from which this Fj plant developed had been fertilized by some neighboring 6-rowed variety less vigorous than the Manchuria or by a low-fertility intermedium form, such a plant as No. 37 might result. The opportunity for accidental crossing in the F^ generation is unusually good. The lateral florets of the Fj plants are much more likely to open at flowering time than are the normal barley florets, and they flower after the spike is exserted. The flowering glumes are less interlocked and open more readily and more widely. The flowers are more frequently deficient in pollen, and those not self-pollinated remain open for hours or even days, affording unusual opportunity for cross pollination. 5 go Journal of Agricultural Research vo1.xix.no.ii SUMMARY H. intermedium is a form of barley in which the awnless lateral florets exhibit a fertility greater than that found in the 2-ro\ved and less than that occurring in the 6-rowed barleys. The occurrence of this form as a homozygote has been questioned frequently. H. intermedium forms which are stable under all conditions of culture have been isolated from numerous crosses reported in this paper. This form appears to be genetically as distinct as either H. vulgare or H. dis- iichon. A 2-factor hypothesis for fertility in the lateral florets is suggested. On the presence-and-absence hypothesis the 6-rowed barleys are sup- posed to be homozygous for the presence of the epistatic factor, the intermedium to be homozygous for the presence of the hypostatic factor and for absence of the epistatic factor, and the Svanhals to be homozygous for the absence of both factors. According to the hypothesis, there are two types of 6-rowed barleys. The Manchuria parent is supposed to be homozygous for the presence of both factors, while certain regressive 6-rowed segregates are thought to be homozygous for the presence of the epistatic factor and for the absence of the hypostatic factor. There is evidence suggestive of a third factor which affects the vigor of the lateral florets and their percentage of fertility. LITERATURE CITED (i) Biffin, R. H. 1907. THE HYBRIDIZATION OF BARLEYS. In JoUf. Agr. Sci., V. 2, pt. 2, p. 183- 206. (2) Harlan, Harry V. 1918. THE identification OF VARIETIES OF BARLEY. U. S. Dept. AgT. Bul. 622, 32 p., 4 pi. Literature cited, p. 31-32. (3) H[axton], J[ohn]. 1851. BARLEY, /n Morton, John C, ed. Cyclopaedia of Agriculture . . . v. i, p. 1 76-19 1. Glasgow, Edinburgh, London. Article is signed "J. H." and is credited by Komicke to Haxton. See following reference, p. 173. (4) KORNiCKE, Friedr, 1885. DIE ARTEN und varietaten des getreides. 470 p., lo pi. Berlin. (Komicke, Friedr., and Werner, Hugo. Handbuch des Getreidebaues. V.I.) (5) 1908. DIE ENTSTEHUNG UND DAS VERHALTEN NEUER GETREIDEVARIETATEN. Hrsg. von M. Kornicke. In Arch. Biontol., Bd. 2, Heft 2, p. 389-437. (6) RiMPAU, Wilhelm. 1891. kreuzungsprodukte landwirtschaftlicher kltlturpflanzen. In Landw. Jahrb., Bd. 20, p. 335-371- (7) 1892. die genetische entwickelung der verschiedener FORMEN UNSERER Saatgerste. In Landw. Jahrb., Bd. 21, p. 699-702. Sept. 1, 1920 Fixed Intermediate, Hordeum intermedium haxtoni 591 (8) TscHERMAK, Erich von. I914. DIE VERWERTUNG DER BASTARDIERUNG FUR PHYLOGENETISCHE FRAGEN IN DER GETREiDEGRLTPE. In Ztschr. Pflanzenzucht. , Bd. 2, Heft 3, p. 291-312. (9) Wilson, Jolm H. 1907. THE HYBRIDIZATION OF CEREALS. In Jouf. Agr. Sci., V. 2, pt. I, p. 68-88, pi. I. PLATE 103 Individual heads representing the three phenotypic progeny classes in which the lateral florets bear awns: A. — Low-fertility class. B. — High-fertility class. C. — Manchuria parent representing the 6-rowed segregates. ^592; Fixed Intermediate, Hordeum intermedium iiaxtoni Plate 103 Journal of Agricultural Research Vol. XIX, No. 11 Fixed Intermediate, Hordeutn intermedium haxtoni Plate 104 Journal of Agricultural Researcli Vol. XIX, No. 11 PLATE 104 Individual heads representing the phenotypic progeny classes in which the lemmas of the lateral florets are rounded and awnless: A. — The 2-rowed, represented here by the Svanhals parent. B. — Low-fertility class. C. — Hordeum intermedium. PLATE 105 A. — Awn-pointed individual, heterozygous for regressive 6-rowed and 2-rowed characters. B. — Short-awned individual, heterozygous for regressive 6-rowed and 2-rowed characters. C. — Long-awned individual, heterozygous for regressive 6-rowed and 2-rowed characters. D. — Three spikes of Mansfield barley from the same plant, showing t£e variations of fertility in the lateral florets. Fixed Intermediate, Hordeum intermedium liaxtoni Plate 105 Journal of Agricultural Research Vol. XIX, No. 11 Fixed Intermediate, Hordeum intermedium haxtoni Plate 106 o Journal of Agricultural Research Vol. XIX, No. 11 PLATE io6 A.— Infertile spike of potentially fertile Hordeum intermedium. B. — Fertile spike of H. intermedium. C. — Var. atterbergii, probably a sterile interm^diwn. 417o_20 6 ADDITIONAL COPIES OF THIS POTLICATION MAY BE PROCURED FROM THE SUPERINTENDENT OF DOCUMENTS GOVERNMENT PRINTING OFFICE WASHINGTON, D. C. AT 20 CENTS PER COPY Subscription Price, $1.50 Per Year V Vol. XIX QE:]PXENIBKR 15, 1921 No. 12 JOURNAL OP AGRICULTURAL RESEARCH CONTENTS AND INDEX OF VOLUME XIX PUBUSHED BY ADTHOWTT OF THE SECRETARY OF AGRICUITDRB WITH THE COOPERATION OF THE ASSOQATION OF LAND-GRANT COLLEGES \SrASHINGXON, D. C. WAtHIMTON 1 OOVMNIHHT rniNTINQ Off Nl I mi EDITORIAL COMMITTEE OF THE UNITED STATES DEPARTMENT OF AGRICULTURE AND THE ASSOCIATION OF LAND-GRANT COLLEGES FOR THE DEPARTMENT KARL F. KELLERMAN, Chairman Physiologist and Associate Chief, Bureau of Plant Industry EDWIN W. ALLEN Chief, Office of Experiment Stations CHARLES L. MARLATT Entomologist and A ssistant Chief, Bureau of EnUyinology FOR THE ASSOCIATION J. G. LIPMAN Dean, State Collect of Agriculture, and Director, New Jersey Agricultural Experi- ment Station, Rutgers College W. A. RILEY Entomologist and Chief, Division of Ento- mology and EconcMiic Zoology, Agricul- tural Experiment Station of the University of Minnesota R. L. WATTS Dean, School of Agriculture, and Director, Agricultural Experiment Station, The Pennsylvania Slate College All correspondence regarding articles from the Department of Agriculture should be addressed to Karl F. Kellerman, Journal of Agricultural Research, Washington, D. C. All correspondence regarding articles from State Experiment Stations should be addressed to J. G. Lipman, New Jersey Agricultural Experiment Station, New Brunswick, N. J. INDEX Page A-amino nitrogen in frozen wheat 185-186 Acid — citric, effect on germination and growth. . . . 74-96 malic, in Bartlett pears 482-499 phosphoric — effect of sulphur on release in greensand . 241 effect on germination and growth 74-96 Acid-hydrolyzable substances in Bartlett pears 480-499 Acidity — effect on persistence of Pseudovumas cilri in soil 217-218 of Bartlett pears 489-494 of greensand 243-2SS of silage 174-176 of wheat, effect of freezing 187-188 Acids, toleration by Bacterium coronafa- ciens 132-153 Aegle marmelos, susceptibility to Pseudomo- nas citri 34i-342>345 Aeglinae, susceptibility to Pse%idomonas cilri 341-342 Aeglopsis Chevalieri, immunity to Pseuso- monas citri 342i34S Aerobism of Bacterium coronafaciens 157-158 A gropyron — caninum — host of Puccinia glumarum, 258 on northern Great Plains 65-72 tepens, host of Puccinia coronaia agropyri. . . 258 smiihii on northern Great Plains 65-72 fe««rMW on northern Great Plains 65-72 Aguingay. See Rottboellia exaltata. Alang-alang grass. See Imperata sp. Alcohol — effect on milk and fat production 124 in silage 174-177 Alcohol-insoluble substances in Bartlett pears 480-499 Alkalinity, effect on persistence of Pseudo- monas cilri in soil 217-218 Aloes, effect on milk and fat production 124 Aluminum — oxid in greensand 244 sulphate, effect on availability of potassium of greensand 242-256 Amid nitrogen in f rozeti wheat 185-186 Ammonia nitrogen in frozen wheat 185-186 Ammonia production by Bacterium corona- faciens 151 Andropogon — /urcc/iti on northern Great Plains 66-72 scoparius on northern Great Plains 65-72 sorghum — host of Sdcrospora philippinensis 104-1 22 var. halcpcnse resistance to Sclerospora philippinensis 104 Page Anias. See Andropogon sorghum, var. hale- pense. Anthony, Stephen, and Harlan, Harry V. (paper): Development of Barley Kernels in Normal and CUpped Spikes and the Limitations of Awnless and Hooded Varie- ties 431-472 Antigens, germ-free filtrates S13-S1S Antimony, black sulphid, effect on milk and fat production 125-130 Aphis, com. See Aphis maidis. Aphis maidis, carrier of sugar-cane mosaic. . 133, 135 Arislida longiseta on northern Great Plains. . . 65-72 Arsenic, effect on milk and fat production . . 1 25-130 Artemisia — dracunculoides on northern Great Plains. . . 65-72 frigida on northern Great Plains 65-72 gnaphalodes on northern Great Plains 65-72 Artificial and Insect Transmission of Sugar- Cane Mosaic (paper) 131-138 Ash content of barley kernels — change during development 412-428 from normal and clipped spikes 452-469 Ashby, R. C, and Malcomson, A. W. (paper): Variation of Individual Pigs in Economy of Gain 225-234 Aster multiflorus on northern Great Plains 65-73 Atalantia — Ceylonica, susceptibility to Pseudomonas citri 343. 34S-347 cilrioides, susceptibility to Pseudomonas citri 343. 345- 347. 348 disticha, susceptibility to Psexidomonas citri 34S glauca. Syn. Eremocitrus glauca. Hindsii. Syn. Fortunella Hindsii. Atterhergii, variety, infertile intermediate barley 587 Avena saliva, halo-blight i39-i7» Awnless barley, limitations 431-472 Bacillus — avenae, causal organism of blade-blight of oats 167 botulinus, separation by germ-free filtrates 513-515 edematiens, separation by germ-free filtrates. 513 tetanus, separation by germ-free filtrates 513 Bacterium coronafaciens, n. sp.— control 170-171 cultural characters 1 4S-1S8 description of lesions 139-140 geographical distribution 140 isolation No. 36 iS4-iSS overwintering and dissemination 158-162 stock halo organism 148-154 yellow organism iSS-is3 Balsamocitrus Daivci, immunity to Pseudo- monas citri 343i 345 (593) 594 Journal of Agricultural Research Page Banana Root-Borer (paper) 39-46 Barley kernels, development — daily 393-430 in normal and clipped spikes 431-472 Bartlett pears, ripening and storage 473-500 Bacterium iranslucens var. undulosum, effect of- formalin 366-379 presoak method 379-392 Beckmannia erucaeformis, host of Puccir.ia coronata 257 Beetles — Colorado potato. See Leptinotarsa decem- lineata. flea. See Epiiriz cucumeris. Behavior of the Citrus-Canker Organism in the Soil (paper) 189-206 Bicarbonate, sodium, effect on milk and fat production 125-130 Bigaraldin, susceptibility to Pseudomonas ciiri 3S6 Black raspberry, host of Gytnnoconia inter- stitialis 501-507 Blackberry, host of Gymnoconia inier- stitiatis 505-506 Blackleg filtrates, germ-free 513-515 Blish, M. J. (paper): Effect of Premature Freezing on Composition of Wheat 181-188 Blue grama. See Bouteloua gracilis. Bouteloua — curtipendula on northern Great Plains 65-72 gracilis on northern Great Plains 65-72 Brachypodium pinnatuin, host of Puccinia glumaru-m 258 Brandes, E. W. (paper) — Artificial and Insect Transmission of Sugar- cane Mosaic 131-138 Mosaic Disease of Com 517-522 Braim, Harry (paper): Presoak Method of Seed Treatment: A Means of Preventing Seed Injury Due to Chemical Disinfectants and of Increasing Germicidal EfGcieucy . . 363-392 Bromus mollis, host of Puccinia glumarum. . . 258 Buffalo grass. See Bulbilis dactyloides. Bulbilis dactyloides on northern Great Plains . . 66-72 Calamovilfa longifolia on northern Great Plains 65-72 Calandra sordida. Syn. Cosmopolites sordi- dus. Calarin, susceptibility to Pseudomonas citri. . 338 Calashu, susceptibility to Pseudomonas citri. . 358 Calcium — carbonate, effect on availability of potas- sium of greensand 242-256 chlorid, effect on germination and growth. . 74-96 oxid in greensand 244 phosphate, effect on solubility of soils 50-51 sulphate, effect on solubility of soils 47-54 Caraway, effect on milk and fat produc- tion 125-130 Carbohydrates — in silage 174-178 in wheat, effect of freezing 186-187 Carbon dioxid, effect of calcium sulphate on production in soil 51-52 Page Carbonate, calcium, effect on availability of potassium of greensand 242-256 Carex — filifolia on northern Great Plains 65-72 heliophila on northern Great Plains 65-72 Casimiroa edulis, susceptibility to Pseudo- monas citri 341-346 Castor oil, effect on milk and fat production. . 124 Chaetospcrmuni glutinosum, susceptibility to Pseudomonas citri 342>345)346, 348 Chaetochloa lutescens, susceptibility to sugar- cane mosaic 134 Chalcas exotica, susceptibiUty to Pseudo- monas citri 341-346 Chlorid, calcium, effect on germination and growth 74-96 Cicitrange, susceptibility to Pseudomonas citri 355 Citradia, susceptibility to Pseudomonas citri. . 354 Citrandaria, susceptibility to Pseudomonas citri 3S4-3SS Citrangarin, susceptibility to Pseudomonas citri 355 Citrange, susceptibility to Pseudomonas citri. . 354 Citrangedin, susceptibility to Pseudomonas citri 3SS Citrangequat, susceptibility to Pseudomonas citri 356 Citranguma, susceptibility to Pseudomonas citri 356 Citric acid. See Acid, citric. Citropsis Sckweinfurthii, susceptibility to Pseudomonas citri 343)345)347)348 Citrumelo, susceptibility to Pseudomonas citri 354 Citrimshu, susceptibility to Pseudomonas "Iri 355 Citrus-canker. See Pseudomonas citri. Citrus — aurantifolia — infection resembling Pseudomonas citri. 203-204 susceptibility to Pseudomonas citri. . . . 350-361 australasica. Syn. Microcitrus australasica. australis. Syn. Microcitrus australis. decumana. Syn. Citrus m,axima. ezcelsa, susceptibility to Pseudomonas citri 352-361 Garrowayi. Syn. Microcitrus Garrowayi. grandis — susceptibiUty to Pseudomonas citri 350-361 Syn. Citrus maxima. hystrix, susceptibility to Pseudomonas citri 349-361 japonica. Syn. Fortunellajaponica. limetta. See Citrus aurantifolia. margarita. Syn. Fortunella margarita. maxima, effect of inoculation with Pseudo- monas citri 190-199 medica, susceptibility to Pseudomonas citri. . 349 mitis, susceptibility to Pseudomonas citri 351-361 nobilis — var. unshiu, susceptibility to Pseudo- monas citri •^49-361 susceptibility to Pseudomonas citri. . . . 351-361 Apr. i-Sept. 15, 192 1 Index 595 Citrus — Continued. Page sinensis — effect of inoculation with Pseudomonas cilri 203-204 susceptibility to Pseudomonas citri 350-361 trifoliata — effect of inoculation with Pseudomonas citri 200-202 Syn. Poncirus trifoliata. Claucena lansium, susceptibility to Pseudo- monas citri 341-346 Clemelo, susceptibility to P seudoifionas citri. . 357 Clonal varieties of Irish potatoes S43~573 Cogon. See Imperata cylindracea. Coil — lachryma-iobi, resistance to Sclerospora philippinensis 104 ma yuen, resistance to Sclerospora philip- pinensis 104 Collins, G. N., and Kempton, J. H. (paper): A Teosinte-Maize Hybrid 1-38 Colorado potato beetles. See Leptinotarsa decemlineata. Comandra pallida on northern Great Plains. . 65-72 Complement-fixaticn test, germ-free filtrates as antigens 513-515 Composition and Density of the Native Vege- tation in the Vicinity of the Northern Great Plains Field Station (paper) 63-72 Composts, manure-sulphur, effect on availa- bility of potassiuai of greensand 239-256 Conidia, production ixxGibberellasaubinetii. 235-237 Copper sulphate, prevention of injury to seeds by presoak method 363-392 Corn. See Zea mays. Corn aphis. See Aphis maidis. Cosmopolites sordidus — control 45-46 description 41-44 history and distribution 39-40 hosts 40-41 life history ; 44-45 Crabgrass. See Synlherisma sanguinalis. Daily Development of Kernels of Hannchen Barley from Flowering to Maturity at Aber- deen, Idaho (paper) 393-430 Decline of Pseudomonas citri in the Soil (paper) 207-223 Development of Barley Kernels in Normal and Clipped Spikes and the Limitations of Awnless and Hooded Varieties (paper). . 431-472 Dewberry, host of Gymnoconia inter stitialis. 505-506 Dextrin and soluble starch in frozen wheat . . 187 Dickson, James G.. and Johann, Helen (paper): Production of Conidia in Gib- berella saubinetii '. 235-237 Dioxid, carbon, effect of calcium sulphate on production in soil 51-52 Dipotassium phosphate, effect on germina- tion and growth 74-96 Downy mildew of maize. See Sclerospora philippinensis. Dox, Arthur W., and Yoder, Lester (paper): Influence of Fermentation on the Starch Content of Experimental Silage 173-179 Draeculacephala molipes, failure to carry sugar-cane mosaic J34, 13s Page Drugs, effect on milk and fat production. . . 123-130 Dry matter in kernels from normal and clipped barley spikes 452-469 Echinacea angustifolia on northern Great Plains 65-72 Effect of Calcium Sulphate on the Solubility of Soils (paper) 47-54 Effect of Drugs on Milk and Fat Production (paper) 123-130 Effect of Manure-Sulphur Composts upon the Availability of the Potassium of Green- sand (paper) 239-256 Effect of Premature Freezing on Composition of Wheat (paper) 181-188 Effect of Reaction of Solution on Germination of Seeds and on Growth of Seedlings (paper) 73-96 Eleusine coracana, effect of field heterogeneity on yield 291-294 Elliott, Charlotte (paper): Halo-Blight of Oats 139-172 Elym.us canadensis, host of Puccinia mon- tanensis 257 Epitrix cucumeris, failure to transmit potato mosaic 329 Epsom salts, effect on milk and fat produc- tion 124 Eremocitrus glauca, susceptibility to Pseudo- monas citri 344, 345, 347 Euchlaena — luzurians, host of Sclerospora philippinen^ sis 104-123 mexicana, hybrid with maize 1-38 Euphorbia dentcta, host of Uromyces euphor- biae 259 Evodia — latifolia, susceptibility to Pseudomonas citri 345, 346 rtdleyei, susceptibility to Pseudomonas citri 345> 346 "False hybrids," susceptibility to Pseudo- m.07ias citri 359 Fat production in milk, effect of drugs. . . . 123-130 Faustrime, susceptibility to Pseudomonas citri 354 Faustrimon, susceptibility to Pseudomonas citri 354 Faustrimedin, susceptibility to Pseudomonas citri 354 Fennel, effect on milk and fat production. . 125-130 Fermentation, influence on the starch content of experimental silage 173-179 Feronia — elephantum. Syn. Feronia limonia. limonia, susceptibility to Pseudomonas citri 342,34s Feroninae, susceptibility to Pseudomonas citri 342 Feroniella lucida, susceptibility to Pseudo- monas citri .^42, 34Si 347 Ferric oxid in greensand 244 Ferrous sulphate, effect on availability of potassium of greensand 242-^56 Fiber, wool, influence of humidity on strength and elasticity 55-62 Filtrates, germ-free, in complement-fixation test 513-SIS 596 Journal of Agricultural Research Vol. XIX Page Fixed intermediate in barley crosses S7S-S93 Flea beetles. See Epitrix cucumeris. Folsom, Donald, and Schultz, E. S. (paper): Transmission of the Mosaic Disease of Irish Potatoes 315-338 Formalin, prevention of injury to seeds by presoak method 363-392 Fortunella — crassifolia, susceptibility to Pseudomonas ciiri 344, 345 Hindsii, susceptibility to Pseudomonas ciiri 344, 345> 347. 348 japonica, susceptibility to Pseudomonas ciiri 344;34S)348 margarita, susceptibility to Psevdomonas ciiri 344.34S.348 Foxtail. See ChaetoMoa lutescens. Frederich, William J., and Peltier, George L. (paper): Relative Susceptibility to Citrus- Canker of Different Species and Hybrids of the Genus Citrus, Including the Wild Rela- tives 339-362 Freezing, effect on — Bacterium coronafaciens 153 composition of wheat 181-188 Fulton, H. R. (paper): Decline of Pseudo- monas citri in the Soil 207-233 Fiuther Data on the Orange-Rusts of Rubus (paper) S01-513 Fiuther Studies on the Influence of Hiunidity upon the Strength and Elasticity of Wool Fiber (paper) 55-62 Gas formation of Bacterium coronafaciens. . 157-158 Genetics of Rust Resistance in Crosses of Varieties of Triticum vulgare with Varieties of T. durum and T. dicoccum (paper) 523-542 Genetics, teosinte-maize hybrid 1-38 Gentian, effect on milk and fat production. 125-130 Germ- Free Filtrates as Antigens in the Com- plement-Fixation Test (paper) S13-S15 Germination of — Gymnoconia inter stitialis, influence of tem- perature 502-503 seeds, effect of reaction of solution 73-96 wheat, effect of formalin 366-379 Gibberella saubinetii, production of conidia. 235-237 Ginger, effect on milk and fat production. . 125-130 Glucose in Bartlett pears 482-499 Glycosmis pentaphylla, immunity to Pseudo- monas ciiri 34J-346 Gochenour, William S. (paper): Germ-Free Filtrates as Antigens in the Complement- Fixation Test 513-515 Grama, blue. See Bouteloua gracilis. Grapefruit — infection of roots by Pseudomcmas citri . . 231-223 See Citrus maxima. Grass — alang-alang. See Imperata sp. buffalo. See Bulbilis daciyloides. Greensand, effect of manure-sulphur on availability of potassium 239-256 Gymnoconia inter stitialis — color of spores 503-504 germination 502-503 morphology of aeciospores 504-507 Page Halo-Blight of Oats (paper) 139-172 Hannchen barley, daily development of kernels 393-430 Hardy, J.I. (paper): Further Studies on the Influence of Humidity upon the Strength and Elasticity of Wool Fiber 55-62 Harlan, Harry V. (paper): Daily Develop- ment of Kernels of Hannchen Barley from Flowering to Maturity at Aberdeen, Idaho 393-430 Harlan, Harry V., and Anthony, Stephen (paper): Development of Barley Kernels in Normal and Clipped Spikes and the Limitations of Awnless and Hooded Varie- ties 431-472 Harlan, Harry V., and Hayes, H. K. (paper): Occurrence of the Fixed Intermediate, Hordeum intermedium haxtoni, in Crosses between H. vulgare pallidum and H. dis- tichon palmella 575-593 Harris, Arthtu" J. (paper): Practical Univer- sality of Field Heterogeneity as a Factor Influencing Plot Yields 279-314 Hayes, H. K., and Harlan, Harry V. (paper): Occurrence of the Fixed Intermediate, Hordeum intermedium haxtoni, in Crosses between H. vulgare pallidum and H. dis- tichon palmella 575-592 Hayes, H. K., Parker, John H., and Kurtz- weil, Carl (paper): Genetics of Rust Resist- ance in Crosses of Varieties of Triticum vul- gare with Varieties of T. dunun and T. dicoccum 523-542 Hays, Frank A., and Thomas, Merton G. (paper): Effect of Drugs on Milk and Fat Production 123-130 Hesperihusa crenulata, susceptibility to Pseudomonas citri 343. 345) 347. 348 Heterogeneity in fields — influence on yield of— alfalfa hay 286-291 com 296-299 hops 295 kherson oats 294 mangolds 284 nitrogen content of wheat 294-295 orchard crops 299-300 potatoes 284-286 ragi 291-294 rice 295-296 timothy hay 286 wheat 291 physical and chemical basis 300-311 Hooded barley, limitations 431-472 Hordeum. — distickon palmella — cross with H. vulgare pallidum 575-592 zeocrilon, intermediate 576 intermedium haxtoni, in barley crosses 575-592 jubaium, host of Puccinia graminis iritici. . . 270 vulgare — host of Puccinia glumarum 258 pallidum, crossvn\h.H.disiichon palmella 575-592 Humidity, effect on strength and elasticity of wool fiber 55-62 Apr. i-Sept. IS, 1921 Index 597 Page Hungerford, Charles W. (paper): Rust in Seed Wheat and Its Relation to Seedling Infection 257-278 Hybrid, teosinte-maize 1-38 Hybrids, "false," susceptibility to Pseudo- monas citri 359 Hydroxid, sodium, effect on germination and growth 74~96 Imperata — cylindracea, resistance to Sderospora philip- pinensis 104 sp., resistance to Sderospora javanica 105 Indol production by Bacterium coronafadens . 151 Influence of Fermentation on the Starch Con- tent of Experimental Silage (paper) 173-179 Insect transmission of — corn mosaic 520-521 sugar-cane mosaic 131-138 Intermediate, fixed, in barley crosses 575-592 Investigations in the Ripening and Storage of Bartlett Pears (paper) 473-500 Irish potato. See SolanuTn luberosum. Johann, Helen, and Dickson, James G. (paper): Production of Conidia in Gibber- ella saubinetii 235-237 Juniper berries, effect on milk and fat produc- tion 125-130 Kempton, J. H., and Collins, G. N. (paper): ATeosinte-Maize Hybrid 1-38 Kernels, barley — daily development 393-430 development in normal and clipped spikes 431-472 Koeleria cristata on northern Great Plains. . . 65-72 Kinnquat. See Fortunella spp. Kunkel, L. O. (paper): Further Data on the Orange-Rusts of Rubus 501-512 Kurtzweil, Carl, et al. (paper): Genetics of Rust Resistance in Crosses of Varieties of Triticum vulgare with Varieties of T. duriun and T. dicoccum S23-S''2 Ladnaria pundata on northern Great Plains. 65-72 Laduca sparsifolia on northern Great Plains. 65-72 Lansium domesticum, susceptibility to Pseu- domonas diri 34' Laptinotarsa decemlineata, failure to transmit potato mosaic 329 Lavanginae, susceptibility to Psextdomonas dirt 343 Leafhopper, sharp-headed grain. See Drae- culacephala molipes. Lee, H. Atherton (paper): Behavior of the Citrus Canker Organism in the Soil 189-206 Lime— air-slaked, effect on milk and fat produc- tion 125-130 inefiBcacy in preventing seed injury from formalin 363-392 milk of, effect on germiaation of wheat 379 Linie. See Citrus aurantifolia. Limelo, susceptibilicy to Pseudomonas diri. . 356 Limequat, susceptibiUty to Pseudomonas diri 356 Limonia glutinosa. Syn. Chaetospermum glutinosum. Page Line-Selection Work with Potatoes (paper). 543-573 " Loco weed." See Oxyiropis lamberii. Magnesium — oxid in greensand 244 sulphate, effect on germiaation and growth . 74-96 Magness, J. R. (paper): Investigations in the Ripening and Storage of Bartlett Pears. . 473-500 Maize — hybrid with teosinte 1-38 See Zea mays. Malcomson, A. W., and Ashby, R. C. (paper): Variation of Individual Pigs in Economy of Gain 225-234 Malic acid. See Acid, malic. Manchuria barley, effect of removing awns. 432-453 Manure-sulphur composts, effect on availa- biUty of potassium of greensand 239-256 McCall. A. G., and Smith, A. M. (paper): Effect of Manure-Sulphur Composts upon the AvailabiUty of the Potassium of Green- sand 239-256 McCool, M. M., and Millar, C. E. (paper): Effect of Calcium Sulphate on the Solu- bility of Soils 47-54 Mcllvaine, T. C, and Salter, Robert M. (paper): Effect of Reaction of Solution on Germination of Seeds and on Growth of Seedlings 73-96 Mealy bug, sugar-cane. See Pseudococcus boninensis. Mdia azedarach, immunity to Pseudomonas dtri 341 Melicope iriphylla, susceptibility to Pseudo- monas dtri 345) 346 Microdtrus — australasica — susceptibility to Pseudomonas dtri 344, 345. 347. 348 var. sanguinea, susceptibility to Pseudo- monas dtri 344, 345, 347 australis, susceptibility to Pseudomonas dtri 344, 345.347.348 Carrowayi, susceptibiUty to Pseudomonas dtri 344.345-347 Mildew, Philippine downy. See Sderospora philippinensis. Milk production, effect of drugs 123-130 Millar, C. E., and McCool, M. M. (paper): Effect of Calciimi Sulphate on the Solubility of Soils 47-54 Moisture — relations of Bacterium coronafadens 151 soil, influence on persistence of Pseudo- monas dtri 214-217 Mosaic Disease of Com (paper) 517-323 Mosaic — Irish potato — transmission by — grafting 3^9-3«o insects 326-332 plant juice 320-326 soil harboring 335-337 sugar-cane, artificial and insect transmis- sion 131-138 Moznette, G. F. (paper): Banana Root-Borer 39-46 598 Journal of Agricultural Research Vol. XIX Page Muhlenbergia cuspidaia on northern Great Plains 65-72 Murraea exotica. Syn. Chalcas exotica. Musa spp., hosts of Cosmopolites sordidus... . 39-46 Navel orange. See Citrus sinensis. Nitrate — potassium, effect on germination and growth 74-96 sodium, effect on germination and growth. 74-96 Nitrogen content — change during development of barley ker- nels 412-428 of wheat, effect of freezing 183-186 of kernels from normal and clipped barley spikes 452-469 Nonrutaceous plants, susceptibility ot Pseu- domonas citri 34i Nux vomica, effect on milk and fat produc- tion 124 Oats. See Avena sativa. . Occurrence of the Fixed Intermediate, Hor- deum intermedium haxtoni, in Crosses be- tween H. vulgare pallidum and H. disti- chon palmella (paper) S7S-S92 Oil, castor, effect on milk and fat production . . 1 24 Orange, navel. See Ctlrus sinensis. Orangelo, susceptibility to Pseudomonas citri. 356 Orangequat, susceptibiUty to Pseudomonas citri 357 Orange-rusts of Rubus. See Gymnoconia inter stitialis. Oxid— aluminum, in greensand 244 calcium, in greensand 244 ferric, in greensand 244 magnesium, in greensand 244 Oxytropis lamberti, on northern Great Plains . . 65-72 Panicum dichotomiflorum, susceptibihty to sugar-cane mosaic 134 Paramignya longipedunculata, susceptibility to Pieudomanas citri 345 Parker, John H., et al. (paper): Genetics of Rust Resistance in Crosses of Varieties of Triticum vulgare with Varieties of T. du- rum and T dicoccum 523-542 Pears, Bartlett, ripening and storage 473-300 Peltier, George L.,and Frederich, William J. (paper): Relative Susceptibility to Citrus- Canker of Different Species and Hybrids of the Genus Citrus, Including the Wild Rela- tives 339-362 Pentoxid, phosphorus, in greensand 244 Peronospora viaydis. Syn. Sclerospoa javan- ica. Petalostemon — candidum on northern Great Plains 65-72 purpureum on northern Great Plains 65-72 Phihppine Downy Mildew of Maize (paper). 97-122 Phosphate — calcium, effect on solubility of soils 50-51 dipotassium, effect on germination and growth 74-96 Phosphoric acid. See Acid, phosphoric. Phosphorus pentoxid in greensand 244 Physostigmine — Page effect on milk and fat production 124 sulphate, effect on milk and fat produc- tion 1 25-130 Pigs, variation in economy of gain 225-234 Pilocarpine, effect on milk and fat production 1 24 Pitutarian, effect on milk and fat production. 124 Poly gala alba on northern Great Plains 65-72 Poncirus trifoliaia, susceptibility to Pseudo- monas citri 343>34S)347i348 Potassium — efi'ect of manure-sulphur composts on avail- ability in greensand 239-256 nitrate, effect on germination and growth. 74-96 sulphate, effect on germination and growth. 74-96 Potato, Irish. SiQ Solanmn tuberosum. Practical Universality of Field Heterogeneity as a Factor Influencing Plot Yields (paper) 279-314 Presoak Method of Seed Treatment: A Means of Preventing Seed Injury Due to Chemical Disinfectants and of Increasing Germicidal Efficiency (paper) 363-39J Production of Conidia in Gibberella saubinetii (paper) 235-237 Pseudococcus boninensis, failure to carry sugar- cane mosaic 136 Pseudom,onas— avenae, causal organism of bladebhght of oats 167 citri — behavior in soil 189-206 decline in the soil 207-223 relative susceptibiUty of species of Citrus 339-362 Psoralea argophylla on northern Great Plains. 65-72 Puccinia — coronata — agropyri, pathogenic on Agropyron repens. 258 pathogenic on oats 257, 265 glumarum — on caryopses of Gramineae 258, 265 resistance of wheats 526 graminis — in Australia 257 resistance of wheat crosses 523-542 tritici — effect on germination 263 effect on growth of seedlings 263-275 parasitic on Hordeum jubatum 270 resistance of wheat crosses 524-542 malvacearum, pathogenic on hoUyhock seeds 259 montanensis, pathogenic on Elymus cana- densis 257 simplex, pathogenic on barley kernels 258 triticiiia, parasitic on wheat 265 Pummelo. See Citrus maxima. Ragi. See Eleusine coracana. Raspberry, black, host of Gymnoconia intersti- tialis 501-507 Ratibida columnaris on northern Great Plains . 65-73 Relative Susceptibility to Citrus Canker of Different Species and Hybrids of the Genus Citrus, Including the Wild Relatives (paper) 339-362 Apr. i-Sept. IS. 1921 Index 599 Page Rhizobium leguminosarum, critical soil re- action 8s Root-borer, banana. See Cosmopolites sor- didus. Rotlboellia exaltata, resistance to Sderospora philippinensis 104 Rubus — alleghainensis, host of Gymnoconia inter- stitialis S06 americanus, host of Gymnoconia intersti- iialis 505-506 canadensis, host of Gymnoconia instersti- tialis S06 cuneifolius, host of Gymnoconia intersti- tialis S03~5o6 hispidus, host of Gymnoconia inter stiiialis. 505-506 nigrobaccus, host of Gymnoconia intersii- tialis 505-506 parvifloTus, host of Gymnoconia intersti- tialis 505-506 procumbens, host of Gymnoconia intersti- tialis 505-506 strigosus, host of Gymnoconia inter stitialis . . 506 triflorus, host of Gymnoconia interstitialis . . . 506 ursinus, host of Gymnoconia interstitialis . 505-506 spp., hosts of Gymnoconia interstitialis. . . . 501-512 Rust in Seed Wheat and Its Relation to Seedling Infection (paper) 257-278 Rust resistance in wheat crosses S23-S42 Rutaceous plants, susceptibility to Psevr domonas citri 341 Rye, wild. See Elymus canadensis. Saccharum — officinarum, resistance to Sderospora philip- pinensis 104 spontaneum, host of Sderospora sp 104 Salter, Robert M., and Mcllvaine, T. C. (paper): Effect of Reaction of Solution on Germination of Seeds and on Growth of Seedlings 73-96 Salt, common, effect on milk and fat pro- duction 124 Salts, epsom, effect on milk and fat produc- tion 124 Sarvis, J. T. (paper): Composition and Den- sity of the Native Vegetation in the Vicinity of the Northern Great Plains Field Station . 62-72 Satsuma. See Citrus nobilis var. wnshiu. Satsumelo, susceptibility to Pseudomonas "'" 3S7 Schultz, E. S., and Folsom, Donald (paper): Transmission of the Mosaic Disease of Irish Potatoes 31S-338 Sderospora — graminicola — dissimilarity of Sderospora philippinensis. . 105 var. andropogonis-sorghi, classification. . . 119 javanica, hosts 10s maydis, suspected occurrence on teosinte ... 105 philippinensis, n. sp. — causal organism 105-122 hosts 103-10S sacchari, pathogenicity on sugar cane, maise, and teosinte 105 Sderostylis Hindsti. Syn. Fortunella Hindsii. Page Seed treatment, presoak method 363-392 Seeds, germination, effect of reaction of solu- tion 73-96 Sezerinia buxifolia, immunity to Pseudomonas "''■' 343, 345; 347 Siamelo, susceptibility to Pseudomonas citri. . 357 Siamor, susceptibility to Pseudomonas citri. 357 Silage, influence of fermentation on starch content 173-179 Smith, A. M.. and McCall, A. G. (paper): Ef- fect of Manure-Sulphur Composts upon the AvailabiUty of the Potassium of Green- sand 239-256 Sodium. — bicarbonate, effect on milk and fat produc- tion 125-130 hydroxid, effect on germination and growth 74-96 nitrate, effect on germination and growth. . 74-96 Soil- effect of calcium sulphate on solubility 47-54 in vicinity of Northern Great Plains Field Station 63-64 Psciidomonas citri in 189-206, 207-223 Solanum titberosum — host of mosaic disease 315-338 line-selection work S43-573 Solidago pulcherrima on northern Great Plains 65-72 Solubility of soils, effect of calcium sulphate . . 47-54 Sorghum. See Andropogon sorghum. Species, new 97-122, 139-172 Starch — in experimental silage, influence of fermen- tation 173-179 in frozen wheat 187 in silage 175-178 Stemrust. See Puccinia graminis. Sterility in wheat crosses 527-532 Stipa — comata on northern Great Plains 63-72 spartea on northern Great Plains 65-72 viridula on northern Great Plains 65-72 Storage, effect on ripening of Bartlett pears. 473-500 Striperust. See Puccinia glumarum. Sucrose in frozen wheat 1S7 Sugar-cane mosaic, artificial and insect trans- mission 131-138 Sugars — in Bartlett pears 480-499 in frozen wheat 187 in silage 174-177 Sulphate— aluminimi, effect on availability of potas- sium of greensand 242-256 calcium, effect on solubility of soils 47-54 copper, prevention of injury to seeds by presoak method 363-392 ferrous, effect on availability of potassium of greensand 242-256 magnesium, effect on germination and growth 74-96 physotigmine, effect on milk and fat pro- duction 125-130 potassium, effect on germination and growth 74"96 6oo Journal of Agricultural Research Vol. XIX Page Sulphid of antimony, black, effect on milk and fat production 125-130 Sulphur, effect on milk and fat production. 125-130 Sulphur-manure composts, effect on availa- bility of potassitam of greensand 239-256 Sxmlight, effect on Bacterium coronafaciens . . . 153 Syntherisma sanguinalis, susceptibility to sugar-cane mosaic 134 Tangelo, susceptibility to Pseudomonas ciiri. 357 Temperature — influence on — dry weight of Bartlett pears in storage. 494-49S germination of Gymnoconia inter siiiialis, 502-503 relation to — Bacterium coronafaciens 151 rust infection 269-270 soil, effect on persistence of Pseudomonas citri 213-214 Teosinte-Maize Hybrid, A (paper) 1-38 Teosinte. See Euchlaena — luxurians. Tnexicana. Tetranychus binaculatus, transmission of su- gar-cane mosaic 133 Thomas, Merton G., and Hays, Frank A. (paper): Effect of Drugs on Milk and Fat Production 123-130 Toddalia asiatica, susceptibility to Pseudo- monas citri 345; 346 Transmission of the Mosaic Disease of Irish Potatoes (paper) 315-338 Triphasia — glauca. Syn. Eremocitrus glauca. tri folia , immunity to Pseudomonas ci'ri 343 . 345)347 Triticuin — dicoccum — dicoccoides, discovery 525 rust resistance in crosses with T. vuigare 523-542 durum, — crosses 524 rust resistance in crosses with T. -vuigare 523-542 polonicum, crosses 524 Triticum — Continued. Page sativum, crosses 524 spelta, crosses 524 iurgidum, crosses 524 vuigare — host of Puccinia glumarum 258 rust resistance in crosses 523-542 Uromyces euphorbiae, pathogenic on Euphor- bia dentata 259 Variation of Individual Pigs in Economy of Gain (paper) 225-234 Vibrion septique, separation by germ-free fil- trates S13 Vicia sparsifolia on northern Great Plains 65-73 Water- effect on production of carbon dioxid by calcium sulphate in soil 51-Sa in barley kernels — change during development of barley kernels 412-428 from normal and clipped barley spikes. 452-469 Weed, "loco." See Oxyiropis lamierti. Weston, William H., jr. (paper): Philippine Downy of Mildew of Maize 97-122 Wheat, relation of rust to seedling infection. 257-278 Whipple, O. B. (paper): Line-Selection Work with Potatoes 543-573 Wild rye. See Elymus canadensis. Wool fiber, influence of humidity on strength and elasticity 55-62 Xanikoxylum — clava-hercules, susceptibility to Pseudo- monas citri 346 fagara, susceptibility to Pseudotrumas citri. 346 rhetsa, immunity to Pseudomonas citri 345 sp., immimity to Pseudomonas citri 341 Yoder, Lester, and Dox, Arthiu- W. (paper): Influence of Fermentation on the Starch Content of Experimental Silage 173-179 Zea — mays, host of — mosaic disease 517-522 Philippine downy mildew 97-122 ramosa, growth of branches 25 New York Botanical Garden Librar 3 5185 00263 3723