UMASS/AMHERST ^ : KJ nDaDDannnDanonnDnnDnnDDDDnnnaaa D D D a n D D D D D D D D D D D D D D D D D ^tRSt UNIVERSITY OF MASSACHUSETTS LIBRARY S 73 E4 NO. 171 -190 nDDDnnnnnDnnnDDaDDnnDDDnnnnDDDDD lU HI »L HL in DATE DUE m ^^ '* HWl3"68 GAYLORD ! PR'NTEDINU.S A. BULLETIN No. 171 DECEMBER, 1916 MASSACHUSETTS AGRICILTIRAL EXPERIMENT STATION A CHEMICAL STUDY OF THE ASPARAGUS PLANT By F. W. MORSE This bulletin is the report of an investigation of the chemical composition of the asparagus plant and the effect of different fertilizers upon the proportions of the more important plant constituents. Its object is to supply more knowledge for the efficient culture of asparagus. Requests for bulletins should be addressed to the Agricultural Experiment Station, Amherst, Mass. IMassachusetts Agricultural Experiment Station. V3^i- Trustees. OFFICERS AND STAFF. COMMITTEE. Charles H. Preston, Chairman, Wilfrid Wheeler, ... Edmund Mortimer, Arthur G. Pollard, Harold L. Frost, . . . . The President of the College, ex officio. The Director of the Station, ex officio. Hathorne. Concord. Grafton. Lowell. Arlington.' STATION STAFF. Administration. William P. Brooks, Ph.D., Director. Joseph B. Lindsey, Ph.D., Vice-Director. , Fred C. Kenney, Treasurer. Charles R. Green, B.Agr., Librarian. Mrs. Lucia G. Church, Clerk. Miss F. Ethel Felton, A.B., Clerk. Agricultural Economics. Alexander E. Cance, Ph.D., Agricultural Economist. Agriculture. William P. Brooks, Ph.D., Agriculturist. Henry J. Franklin, Ph.D., In Charge Cranberry Sub- station. Edwin F. Gaskill, B.Sc, Assistant Agriculturist. Robert L. Coffin, Assistant. Botany. A. Vincent Osmun, M.Sc, Botanist. George H. Chapman, Ph.D., Research Physiologist. Paul J. Anderson, Ph.D., Associate Plant Pathologist. Orton L. Clark, B.Sc, Assistant Plant Physiologist. Miss Grace B. Nutting, Ph.B., Clerk. Chemistry. Joseph B. Lindsey, Ph.D., Chemist. Edward B. Holland, Ph.D., Associate Chemist in Charge {Research Section). Fred W. Morse, M.Sc, Research Chemist. Henri D. Haskins, B.Sc., Chemist in Charge (Fertilizer Section) . Philip H. Smith, M.Sc, Chemist in Charge (Feed and Dairy Section). Lewell S. Walker, B.Sc, Assistant Chemist. Cableton P. Jones, M.Sc, Assistant Chemist. Carlos L. Beals, M.Sc, Assistant Chemist. James P. Buckley, Jr., Assistant Chemist. W. A. Allen, B.Sc, Assistant Chemist. John B. Smith, B.Sc, Assistant Chemist. James T. Howard, Inspector. Harry L. Allen, Assistant in Laboratory. James R. Alcock, Assistant in Animal Nutrition. Miss Alice M. Howard, Clerk. Miss Rebecca L. Mellor, Clerk. Entomologry. Henry T. Fernald, Ph.D., Entomologist. Burton N. Gates, Ph.D., Apiarist. Arthur I. Bourne, A.B., Assistayit Entomologist. S. C. ViNAL, B.Sc, Graduate Assistant. Miss Bridie E. O'Donnell, Clerk. Horticulture. Frank A. Waugh, M.Sc, Horticulturist. Fred C. Sears, M.Sc, Pomologist. Jacob K. Shaw, Ph.D., Research Pomologist. Harold F. Tompson, B.Sc, Market Gardener. R. P. Armstrong, M.Sc, Graduate Assistant. Miss Eleanor Barker, Clerk. Meteorology. John E. Ostrander, A.M., C.E., Meteorologist. J. S. Sims, Observer. Microbiology. Charles E. Marshall, Ph.D., Microbiologist. F. H. Hesselink v.\n Suchtelen, Ph.D., Research Mi- crobiologist. Poultry Husbandry. John C. Grah.\m, B.Sc, Poultry Husbandman. Hubert D. Goodale, Ph.D., Research Biologist. Lloyd L. Stewart, B.Sc, Graduate Assistant. Miss Marcella C. Curry, B.Sc, Clerk. Veterinary Science. James B. Paige, B.Sc, D.V.S., Veterinarian. G. Edw.\rd Gage, Ph.D., Research Pathologist. J. B. Lentz, V.M.D., Assistant. A. C. Edwards, V.M.D., Assistant. CONTENTS. Introduction, Crowns and roots, ......... Asparagus stalks, ......... Asparagus tops, ......... Progressive changes in composition of the asparagus plant, . The inorganic constituents of the asparagus plant. Effect of fertilizers on the composition of the asparagus plant. Effect of fertilizers on asparagus roots, ..... Effects of fertiUzers on asparagus stalks, ..... Effects of fertilizers on asparagus tops, ..... Effect of fertiHzers on asparagus roots at the end of the cutting season, Reserve material required to produce a crop of young stalks, Amount of vegetable matter contained in ripened asparagus tops. Relation of asparagus roots to weights of stalks, Summary, .......... Practical conclusions from the chemical study of the asparagus plant, P.\GE 265 265 270 272 274 275 276 277 285 286 289 290 292 293 295 296 Publication of this Document approved by the SuPEBVisoR OP Administration. ]^ULLETIX ]^o. 171. DEPARTMENT OF CHEMISTRY. A CHEMICAL STUDY OF THE ASPARAGUS PLANT. BY F. W. MORSE. Introduction. The chemical composition of the asparagus plant {asparagus officinalis) has been under investigation in this laboratory for several years. The studies were begun in connection vnth a series of fertilizer experiments which have been conducted at Concord, Mass., where asparagus culture is an important industry. The chemical composition of the asparagus plant has heretofore re- ceived comparatively httle attention. Rousseaux and Brioux,^ in a study of commercial asparagus culture in France, include numerous determinations of the inorganic constituents. Tanret^ has investigated the soluble carbohydrates, or sugars. Wichers and ToUens"* have re- ported the composition of the roots and crowns at different seasons. A few scattered analyses of the edible stalks have been found in different publications.^ Our studies have included several stages in the development of the asparagus plant, and also the effects produced by different methods of fertilization. Crowns and Roots. The first lot of material collected for the investigation consisted of cro'mis and roots taken from the experiment field at Concord early in November, 1908. One-year-old plants had been set in tliis field in the spring of 1907; therefore the roots when sampled were two and one-half years from the seed. 1 The author's indebtedness to Director Wm. P. Brooks and Dr. J. B. Lindsey for important suggestions regarding the work is gratefully acknowledged. 2 Rousseaux and Brioux: Ann. Sciences Agron., 3d Series, I. (1906), pp. 188-326. » Tanret: Bull. Soc. Chim. (4) 5, p. 889 (1909). * Wichers and Tollens: Journ. fur Landwirthsch., 1910, p. 113. ' N. y. Agr. Expt. Sta. Bull. 265; Office Expt. Sta. Bull. 28, p. 37. 266 MASS. EXPERIMENT STATION BULLETIN 171. The material was collected at this time for the purpose of determining the influences of the different fertilizers on the proportion of the reserve plant foods stored in the roots. The first crop of stalks would be cut from the plots in the following spring, and it was desirable to ascertain if any relationship could be demonstrated between the reserve food stored in the roots and the amount of growth made in the spring. At the time the roots were dug the tops of the plants had been killed by frost and the stems were breaking down. It was consequently assumed that the roots had stored aU the reserves of plant food which the stalks would have for their growth in the following spring. Since these samples were primarily for studying the effects of fertihzers, each plot was represented by four plants which were selected by the size and number of their stalks, on the assumption that a plant with an average amovmt of tops would possess an average lot of roots. The crown and attached roots of each plant were dug with spade and trowel by means of wliich the longest roots were followed to their tips. The word "roots" is used here to designate the rod-like storage roots of the plant, and not the fibrous feeding roots wliich were rubbed off during the waslihig process. The roots in this lot were selected and the digging supervised by Mr. E. F. Gaskill, assistant agriculturist. The subsequent preparation of the samples for chemical analysis was supervised by Mr. P. H. Smith, in charge of the feed and dairy section of this department. The writer was assigned to this investigation in January, 1910, and the work has since been wholly in liis charge. A second lot of roots was collected on Nov. 4, 1910, by the writer and Mr. Gaskill after the plants had been set in the field tliree and one-half years. Two crops of stalks had been cut for market during their life, — a short crop in 1909 and a full crop in 1910. Plants were selected as before by the size and number of the matured stalks, wliich were in the same condition of decay as in 1908. The roots had now ramified to such an extent that those of adjacent plants were more or less intermingled, making it impracticable to follow all roots of selected plants to their tips. Therefore a circle with a radius extending halfway to the adjacent plants in the row was cut with a spade around the chosen plant, after which the crown and attached storage roots were removed from the soil. It was noted that most of the roots ended in the characteristic pointed tips without cut ends, and were there- fore fully representative of the plant. The roots were shaken free of soil, put in sacks and shipped to Amherst. Two days elapsed between the removal of the roots from the soil and their reception at the laboratory. Upon their arrival they were placed in a cool cellar used for vegetable storage. Each crown was next separated into small sections in order to remove adhering soil, and the parts, together with the attached roots, were scrubbed with a stiff brush, after which they were rinsed in clean water. A CHEMICAL STUDY OF THE ASPARAGUS PLANT. 267 The material was next spread on a large sheet of paper in a cool place until the surface was free of adhering moisture. Each individual crown and its accompanying roots were then weighed and the weight noted doTVTi for the subsequent calculations as the fresh or green weight from the field. The first stage of preparation of the material for analysis was to pass a sample, consisting of one crown and its corresponding roots, tlirough a hand-lever feed-cutter, by which they were cut to lengths of about 1 inch (2.5 centimeters). The sample was then placed in a large steam- heated drying oven, where the temperature was about 55° C, and dried until sufficiently brittle to be easily pulverized. In pulverizing asparagus roots for analysis certain properties of their constituents made serious trouble. During the preparation of the first lot of roots in 1908 Mr. Smith found the dried material to be so hygro- scopic that in damp weather it would quickly become sticky and gum the mill. The friction of grinding was also apt to produce sufficient heat to make the material sticky and hopelessly cement the grinding plates together. By using a ball mill in dry weather he finally succeeded in reducing the samples to powder. The WTiter's procedure with the samples of 1910 was as follows: im- mediately after removing the dried sample from the oven the material was allowed to cool a short time in the air and then weighed. Directly after weighing the sample was passed through a large drug mill, by which it was reduced to a mixture of coarse fiber and fine powder, the fiber coming from the outer walls of the roots and the powder from the interior and the crown. The mixture was subsampled by tvv^o successive quar- terings. The subsample was next sifted by means of a millimeter sieve, which separated nearly aU of the fine powder from the fibrous shreds. By tliis step the hygroscopic, gummy constituent was largely eliminated from subsequent milling and the coarse fiber was pulverized about as readily as wheat bran, until it also passed through the millimeter sieve. The entire material of the subsample was thoroughly mixed and preserved in a tightly corked bottle for analysis. Care was taken to conduct all the operations in a dry atmosphere. On June 23, 1911, at the end of the cutting season, a third lot of samples was taken for the purpose of determining the amount of exhaustion wliich the reserve material in the roots had undergone in producing the crop recently harvested. This lot of roots was collected under the supervision of Mr. C. W. Prescott, who was in charge of the Concord experiment field. There was practically no top growth by which to judge the size of a crowai, and the roots were therefore necessarily chosen more at random than in the previous cases. On arrival of the roots at the laboratory they were treated in the manner described for the samples of 1910. The average fresh weight of forty-four roots gathered from eleven different plots was found to be for each of two years, as follows: 1908, 268 MASS. EXPEKIMENT STATION BULLETIN 171. 1,092 grams; 1910, 2,440 grams. In two years the crowns and roots had more than doubled in size and weight. The average weight of sixteen roots from four plots in each of three years is as follows: 1908, 1,120 grams; 1910, 2,393 grams; 1911, 2,401 grams. The roots obtained in 1911 actually averaged slightly heavier than those selected the fall before. This may in part be due to the more random choice of samples in the summer before there was sufficient top growth to guide the selection, but is more probably the result of a higher water content in the growing season, as will be seen in the table of composition. It has already been mentioned that the first object in collecting the different series of roots was to ascertain the effects of different fertilizers on their composition, but it is deemed best to present first the average composition of the roots at different stages of development, and follow with the composition of other parts of the plant before taking up the specific influences of methods of fertilization. In furtherance of the primary object of the investigation, forty-four Cfowns, representing eleven different plots, were collected in the faU of 1908; seventy-six from nineteen plots in the fall of 1910; and sixteen from four plots in the summer of 1911. A complete analysis was not made of every sample. Nitrogen was determined in every individual sample of each year. Total sugar was determined in about two-thirds of the samples obtained in 1908, and in every sample of the lots of 1910 and 1911. Ash and ash constituents were determined in every sample of the lot of 1908, but only in composite samples representing the individual plots in the series of 1910 and 1911. Dry matter was determined in every sample of 1910 and 1911, but was not calculated in the samples of 1908 because the weights after the first drying were omitted. The other constituents — fiber, pentosans and fat — were determined in selected samples in each series, chosen from some with average percentages of nitrogen or sugars, and others with maximum or minimum proportions. In compiling averages for each year from the numerous analyses of individual samples above mentioned there were included only those figures obtained on samples from plots receiving complete fertilizers in some form, and results from plots receiving no nitrogen, no potash or no phosphoric acid were omitted. A CHEMICAL STUDY OF THE ASPARAGUS PLANT. 269 Composition of Asparagus Roots. November, 1908. November, 1910. June, 1911. Dry matter - 21.10 18.62 Ash in dry matter 6.56 6.89 8.93 Protein, 12.25 12.44 12.75 Fiber 15.39 19.77 23.66 Fat, .98 1.77 59.13 1.63 Nitrogen-free extract 64.82 53.03 Sugar in dry matter, 39.98 31.52 23.20 Pentosans 8.91 10.96 11.66 Lignin, etc 15.93 16.65 '8.17 Total nitrogen, 1.96 1.99 2.04 Protein nitrogen, 1.19 1.05 1.30 Amino nitrogen, . ._ .77 .94 .74 Note. —The analytical methods employed throughout this work were those of the Association of OfiBcial Agricultural Chemists in all essentials. The comparison shown by the table is of great interest. As the roots increased in size from 1908 to 1910 there was not a marked change in all constituents. The slight increase in ash may have been due to increased absorption and storage, and in part caused by the impossibility of thor- oughly removing the adhering soil in washing the roots. The nitrogen percentage was practically unchanged, showing that the roots demanded and received that element as fast as new growth developed. There was a change in the relative proportions of the non-nitrogenous materials. In the soluble and active form the sugar was much less in the older roots, while the different inactive forms had aU increased (fiber, pentosans, lignin and fat). There was a small change in the porportion of protein and amino nitrogen, which may have been a seasonal difference. The sixteen random roots selected in 1911 from four plots, as already shown, weighed a trifle more than the roots gathered the fall before from the same plots. The analyses showed, however, a lower percentage of dry matter and actually lower weight on that basis. There was a pro- nounced exhaustion of sugars in the spring growth, but none of the other constituents; instead, the other constituents were increased in pro- portion to the loss of sugars. Nitrogen, which would be also indispen- sable to new growth, was not consumed at the rate of sugar, but was transferred to the growing stalks at a rate which left its proportion in the parent crown almost unchanged. Total ash was not reduced but largely increased as the organic matter was consumed. These points will be considered again in connection with the development of the tops of the plant. 270 MASS. EXPERIMENT STATION BULLETIN 171, Asparagus Stalks. The marketable portion of the asparagus plant consists of the young stalks cut from the crowns during the spring and early summer. Their constituents must be mainly derived from the reserve materials stored the previous summer in the roots, and the total quantity removed in a season represents the drain which the roots must be prepared to meet. Our first samples of 3'oung stalks were obtained from the experiment field at Concord in 1910, but it was clearly evident that during the two or more days which elapsed between cutting in the field and delivery at the laboratory there were destructive changes taking place in the soluble carbohydrates or sugar of the cells. Consequently in the spring of 1911 a series of samples of 3^oung stalks was gathered from the experiment field at Amherst, which had been fertilized in a similar manner to the field at Concord. Samples of stalks were cut from four different plots in the home field on four different dates, beginning May 17 and ending June 14. The stalks were cut as close to the crown as possible, and averaged about 10 inches (25 centimeters) in length. The common practice of asparagus growers in Massachusetts is to grow the crop so that most of the stalk is above ground, and when trimmed to the standard length of 8 inches (20 centi- meters) it is nearly all green. The material used in our investigation rep- resented the crop as cut from the crowns before it is bunched and trimmed. Each plot sample consisted of all the stalks which were tall enough to be marketable on the day of cutting. Immediately after the samples were cut they were taken to the labora- tory, where the stalks were wiped with a dry cloth to free them from adhering soil, after which the samples were weighed. The stalks were then broken into short pieces and spread on a tray which was placed in the steam-heated drjang oven at a temperature of 55° to 60° C. In preparing asparagus stalks for analysis it was found necessary to avoid a large amount of cut or broken surface, since the contents of the ruptured cells changed rapidly during the early drjdng stage by a process of fermentation with a loss of soluble sugar. Too high a temperature would soften the tender tips or buds of the stalks and cause them to stick to the tray. Pieces of stalks about 3 inches (7.5 centimeters) in length withered quickly in a temperature of 55° to 60° C, and at the end of twenty-four hours the largest butts were split in half, longitudinallj^, to promote further rapid drying. Samples dried in this manner were subse- quently found to have retained their sugar unchanged, or at least under such conditions there was obtained the maximum proportion of sugar. Composite samples from all plots represented each date of cutting, in order to determine the rate of change in their composition as the season advanced. The following table shows this composition : — A CHEMICAL STUDY OF THE ASPARAGUS PLANT. 271 Composition of Asparagus Stalks in Spring. [Parts in 100.] May 17. June 1. June 8. June 14. Water Dry matter 92.31 7.69 92.35 7^.5 92.30 7.70 92.24 7.76 Composition of Dry Matter. Ash, . Protein, Fiber, . Fat, . Nitrogen-free extract, Total sugars, Reducing sugars. Pentosans, . Lignin, etc.. Total nitrogen, . Protein nitrogen, Amino nitrogen, 8.77 33.25 18.90 2.84 36.24 9.91 7.75 14.23 12.10 5.32 3.07 2.25 9.07 31.19 17.15 3.03 39.56 15.47 11.66 12.80 11.29 8.47 29.75 18.82 3.20 39.76 15.64 12.04 13.39 10.73 4.76 8.47 28.87 17.92 3.22 41.52 19.87 13.22 13.21 8.44 4.62 3.15 1.47 Two notable sets of changes occurred in the composition of the series of samples. Sugars increased remarkably in the successive periods, while protein and lignin decreased. Dry matter was practically constant. In 1914 two other lots of stalks were analyzed primarily for another purpose, but protein, sugar and dry matter behaved in a manner similar to that of the earlier samples. May 25. June 2. Dry matter, Total sugar in dry matter, . .' Reducing sugar in dry matter, Protein 7.64 20.55 14.25 29.30 7.68 27.39 20.29 28.45 It seems probable that this change in amount of sugar is due to photo- sjTithesis, since so much of the stalk is above ground and supplied with chlorophyl. Growth is somewhat slower as the season advances after 272 MASS. EXPERIMENT STATION BULLETIN 171. the first rapid development in warm days of May, giving more time for the photosjTithesis to go on. It does not seem reasonable that the drain on the roots should be inversely proportional to the reserves in them. The decrease in nitrogenous matter does follow the exhaustion of the roots. The change in protein is a steady decrease in the amino nitrogen, while the true protein remains practically constant. This points also to more self-support and slower growth. Asparagus Tops. The development of reserve food material by the asparagus plant has been studied by the analysis of samples of fully grown tops in midsummer and ripened tops in late fall. Two series of samples were collected from the fertilizer plots at Concord, — one in October, 1911, and the other in August, 1912. These were taken for the purpose of ascertaining whether the reserves were affected in any manner by the different fertihzers em- ployed. Upon analyzing them it was noted that soluble carbohydrates were very low, and the possible destruction by respiration during the time required to transport the samples from Concord to Amherst led to taking parallel samples at Amherst for the study of their composition at the two stages of growth. To avoid serious injury to the crowns, representative samples for each stage of growth were obtained by pulling only one stalk from a crown. Seven average plants yielded in this manner an abundance of material for a sample, and two parallel samples were thus selected on the different dates. To ascertain how fast translocation of reserves was taldng place the tops were divided into two portions. Each top was trimmed to a single stalk and thus was formed two samples, — stalks and branches. The lot of tops was weighed as soon as removed from the field, then divided into stalks and branches, each portion being weighed. Each separate sample was now spread in the sun in the glass house for twenty- four hours, and then run through a fodder cutter. The samples were next dried in the large steam-heated oven until brittle enough to be ground, when they were cooled in the air, weighed and pulverized for subsequent analysis. The summer stage of growth was after blossoming was about over, and the stalks chosen bore no berries. This stage was considered by analogy with other crops to be the stage of maximum growth of tops, and that the reserve material in the tissues would be at the maximum. The ripened stage was when the stalks had turned yellow and the needles were falling from some of the stalks. The tops selected were those wliich shed but few when handled. A CHEMICAL STUDY OF THE ASPARAGUS PLANT. 273 Comfosition of Asparagus Tops. Seven stalks, Aug. 16, 1912, weighed, green, 1,791 grams, were 60 per cent, and stems were 40 per cent, of total weight. Seven stalks, Oct. 23, 1912, weighed, green, 1,859 grams, were 64 per cent, and stems were 36 per cent, of total weight. Branches Branches ~" Summer Tops. Fall Tops. Stems. . Branches. Stems. Branches. Dry matter 23.76 28.43 24.18 32.15 Ash in dry matter, . 7.39 7.31 9.36 8.51 * Protein, . 7.94 17.31 4.47 11.00 Fiber, 44.83 29.76 45.11 32.02 Fat 1.38 4.89 1.35 5.23 Nitrogen-free extra -t. 38.46 40.73 39.71 43.24 Total sugar, 14.28 8.68 9.34 7.09 Pentosans, 15.90 14.15 15.86 14.41 Lignin, 8.28 17.90 14.51 21.74 Reducing sugar, 12.50 2.99 8.76 3.99 Protein nitrogen. 1.03 2.42 .74 1.56 Amino nitrogen, .24 .35 - .20 Protein and sugar both disappear with ripening in about the same pro- portion, and appear to be the only groups of constituents subjected to translocation. The translocation of sugars as they are formed is indicated by the higher percentages in the stalks than in the branches, both in midsummer and in autumn. In November (the 4th), 1914, six tops were gathered which were golden yellow in color but bare of needles. Dry matter, sugar and protein were determined with the following results : — Per Cent. Dry matter 49.45 Sugar 4.08 Protein, . . . . . . . . . . . .4.70 It is probable that neither sugar nor protein is completely transferred to the root, because until killed by frost the U\dng cells must stiU contain active protoplasm and its supply of food. The more extensive series of samples collected at Concord completely corroborate these changes in kind, but respiration undoubtedly affected the sugars. The average composition of the lots is given in the following table: — 274 MASS. EXPERIMENT STATION BULLETIN 171. Composition of Dry Matter. Summer Tops, 11 Samples. Fall Tops, 7 Samples. Ash 9.34 8.65 Protein, ' . . . 17.47 7.94 Fiber 33.04 43.75 Fat 2.71 3.49 Nitrogen-free extract, 37.44 36.17 Sugars, 5.29 - Pentosans, 15.58 20.90 Lignin, etc., 16.57 15.27 Total nitrogen 2.79 1.27 Protein nitrogen, 1.63 1.27 Amino nitrogen, 1.16 - Progressive Changes in Composition of the Asparagus Plant. The following table has been arranged in order to compare the compo- sition of the successive stages of growth which have been studied: — Ash, . Protein, Fiber, . Fat, . Total sugars, Reducing sugars. Pentosans, . Lignin by difference, Total nitrogen, . Protein nitrogen, Amino nitrogen, . Autumn Roots, 1910. Summer Roots, 1911. Young Stalks. Summer Tops. Autumn Tops. Water, Dry matter. 78.90 21.10 81.38 18.62 92.30 7.70 73.44 26.56 70.73 29.27 Composition of Dry Matter. 12.44 19.77 1.77 .52 10.96 16.65 1.99 1.05 8.93 12.75 23.66 1.63 .20 11.66 18.17 2.04 1.30 .74 30.77 18.20 3.07 15.22 11.17 13.41 10.64 4.92 3.07 1.85 7.34 13.56 3.48 10.92 6.79 14.85 14.06 2.17 1.86 .31 .83 7.90 5.70 14.92 19.16 1.26 .12 A CHEMICAL STUDY OF THE ASPARAGUS PLANT. 275 The relation of water to intensity of growth is clearly shown by the changes in the proportion of water at the different stages of development. The summer roots procured in the midst of the growing season contained more water than the dormant roots obtained the fall before. The tops when just at their full height in midsummer were more watery than those that were ripening in the follo\\dng October. But the most striking pro- portion of water was found in the tender, succulent stalks of spring and early summer at the period when growi;h is so rapid that it can be readily measured from hour to hour. The active part performed by sugar is indicated by the difference in the percentages of tliis substance found in the various stages of the devel- opment of the plant. The large proportion of reducing sugar in the stalks and tops at the successive stages sampled, and its absence from the differ- ent series of roots, is in accord with distinction between active and reserve forms of sugars. The sugar in the roots at the seasons chosen for their study was wholly a reserve substance, and being readily soluble in water passed unchanged toward the actively growing stalks. The insoluble non-nitrogenous substances which form the bulk of the plant at each stage of growth undergo the usual inverse changes in pro- portion which accompany the increase and decrease of more active con- stituents. Amino compounds are an important part of the reserve nitrogenous material in the fall roots, as their nitrogen forms almost one-half of the total percentage of the element at that stage. This is a larger proportion than at any other stage, and points to its probable value for rapid transfer to the young stalks in the spring. The Inorganic Constituents of the Asparagus Plant. For comparing the progressive changes in the mineral constituents of the different stages of the asparagus plant we have used the averages of all results from the plots receiving complete fertilizers. At first sight the average composition of the three series of roots appears to be practically alike, but a closer scanning reveals consistent variations in some of the constituents from year to year. Calcium, sulfur and so- dium steadily increased in percentages from stage to stage in the roots, and also between the summer and fall stages of the tops. On the other hand, potassium, magnesium and phosphorus varied between narrow limits in the different stages of root development, and were noticeably diminished in the final ripening stage of the tops. These three elements are evidently translocated from the old tops to other parts of the plant, while the three first mentioned go in only one direction and accumulate as those parts of the plants grow older. Sulfur is considerably in excess of phosphorus, which is unusual in our common garden crops. While no provision was made for this in planning the fertilizer, there was apparently enough of the element present in the stable manure or superphosphate used. 276 MASS. EXPERIMENT STATION BULLETIN 171. The translocation of potash, magnesia and . phosphoric acid back to the roots is indicated but not proven, since there are the blossoms and berries to be considered as a possible destination in their transfer. These two sets of organs were not collected, however, as it was nearly impossible to get anything approaching accurate amounts of them from a series of stalks, because the red asparagus beetle destroys them in preference to other parts of the plant. Inorganic Constituents of the Asparagus Plant at its Different Stages {Per- centages in Dry Matter). Autumn Roots, 1908. Autumn Roots, 1910. Summer Roots, 1911. Young Stalks. Summer Tops. Autumn Tops. Calcium oxide, . Magnesium oxide, Potassium oxide. Sodium oxide, . Phosphoric acid. Sulfuric acid, . .316 .151 2.445 .245 .507 .509 .360 .192 2.465 .368 .464 .627 .436 .184 2.374 .366 .442 .730 .387 .346 5.270 .330 .538 .833 .994 .243 3.436 .203 .472 .472 1.635 .190 2.189 .431 .169 .532 Effect of Fertilizers on the Composition of the Asparagus Plant. The material for the study of the effects of fertilizers on the composi- tion of the different parts of the asparagus plant was chiefly obtained from the experiment field ^ at Concord, but some was taken from the plots at the experiment station in Amherst. The soil of the experiment field is typical of the soils chosen in Massa- chusetts for asparagus culture, i.e., a coarse, sandy loam. Samples of the soil from four sections of the field were analyzed by the conventional method, and the results are given in the following table: — Soil Analyses. Vola- Insol- Cal- Magne- Potas- Phos- Sul- Total tile uble cium sium sium phoric furic Nitro- Humus. Matter. Matter. Oxide. Oxide. Oxide. Acid. Acid. gen. Surface. Southeast, . 4.26 89.43 .20 .07 .09 .25 .04 0.13 1.97 Southwest, . 4.55 89.86 .22 .02 .09 .21 .04 0.14 1.94 Northeast, . 4.14 90.49 .23 .02 .10 .27 .04 0.13 1.85 Northwest, . 4.25 90.27 .22 .01 .07 .20 .05 0.13 1.78 Subsoil. Southeast, . 2.61 91.01 .08 .01 .09 .03 - - - Southwest, . 2,11 93.32 .13 .03 .09 .04 - - - Northeast, . 2.17 93.30 .06 .02 .10 .05 - - - Northwest, . 2.71 92.84 .04 .01 .08 .08 - - - Second Foot. Southeast, . 1.88 92.29 .07 .02 .10 .03 - - - Southwest, . 1.17 94.03 .09 .01 .12 .06 - - - Northeast, . .79 95.62 .06 .04 .09 .05 - - - Northwest, . .80' 96.03 .06 .02 .09 .04 ~ ~ " See annual reports for 1908 and following years for description of fertilizer experiments. A CHEMICAL STUDY OF THE ASPARAGUS PLANT. 277 These analyses were made by Messrs. E. B. Holland and R. D. Mac- laurin before the field was planted in 1907. It will be readily seen that the samples show a striking uniformity in composition. The mamier of fertilizing the experiment plots has been described in a previous paper, ^ but for the sake of clearness the scheme is here briefly outlined. Plot. Application. Pounds per Acre, Nitrate of Soda. Pounds per Acre, Acid Phos- phate. Pounds per Acre, Muriate of Potash. 1 No nitrates _ 200.1 260.0 31 Low nitrate, in spring 311.2 200.1 260.0 32 Low nitrate, in summer 311.2 200.1 260.0 33 Low nitrate, half in spring, half in summer, 311.2 200.1 260.0 34 Medium nitrate, in spring, 466.6 200.1 260.0 35 Medium nitrate, in summer 466.6 200.1 260.0 36 Medium nitrate, half in spring, half in summer, 466.6 200.1 260.0 37 High nitrate, in spring, 622.4 200.1 260.0 38 High nitrate, in summer 622.4 200.1 260.0 39 High nitrate, half in spring, half in summer, . 622.4 200.1 260.0 40 No nitrate, - 200.1 260.0 5 No phosphate 466.6 260.0 6 Low phosphate 466.6 133.4 260.0 7 Medium phosphate, 466.6 200.1 260.0 8 High phosphate 466.6 266.8 260.0 9 No potash, ......... 466.6 200.1 _ 10 Low potash, 466.6 200.1 173.4 11 Medium potash, 466.6 200.1 260.0 12 High potash, 466.6 200.1 346.8 Effect of Feetilizers on Asparagus Roots. The roots of 1908 represented only the plots that had received different applications of nitrate of soda; the samples of 1910 included these plots and the plots to which different quantities of acid phosphate and muriate of potash were applied. The weights of the roots are given by individuals and by plots in the following table : — • 1 Ann. Rept., Mass. Agr. Expt. Sta. 25, p. 156. 278 MASS. EXPERIMENT STATION BULLETIN 171, Weights of Asparagus Roots lohen taken from the Field {Grc Series of 1908. Plot. Root I. Root II. Root III. Root IV. Plot Average. 1 566 1,177 792 974 877 31 1.268 1,177 770 1,024 32, 997 861 952 884 923 33 1,020 635 1,701 1.134 1,020 1.020 975 680 1.179 907 1.043 1.927 1.315 895 34, 1,338 1,264 35 1,360 1,275 36 1,338 1.213 37 1,542 1,224 1.179 1.406 1.338 38 1,837 1,519 1.020 839 1,304 39, 544 1.020 476 884 731 40 907 1.474 1.701 635 1,179 Series of 1910. Plot. Root A. Root B. Root C. Root D. Plot Average. 1 2.262 2.070 1.816 1,951 2.025 5, 1.896 1.633 1.561 2,043 1.783 6, 2,960 3.012 2.573 2,868 2.853 7 2,885 2,791 2.869 2,393 2.734 8 2,182 2.265 1.282 1,833 2.110 2.703 1,792 2,246 9 1,509 1,673 10 2.827 3.015 1.993 1,745 2.395 11 3.410 2,402 2,661 3,097 2.892 12 1.986 2,967 2,691 3,194 2,709 31 . 3.317 1.486 1,985 3,393 2,545 32 1,918 2.570 1,526 2,000 2,003 33 2,655 2.440 1,861 3,195 2,538 34 3,540 2.119 2.677 2.595 2.733 35. 1,957 1.700 2,470 3,029 2.289 36 2,043 4.432 3,227 2,448 3,598 3,062 3,089 37 2.677 2,853 38 2,807 2.313 2,446 1,676 2,310 39 1,989 1.927 2,065 2,927 2,227 40 3.042 1.717 2,197 1,967 2,231 A CHEMICAL STUDY OF THE ASPARAGUS PLANT. 279 There cannot be said to have been any specific effect of the nitrate of soda on the size of roots in 1908. The weights of the four roots from any given plot varied more widely among themselves than the plot averages differed from one another. There were some consistent variations in the weights of the roots dug in 1910. The roots from plots 5 and 9, lacking phosphoric acid and potash, respectively, were consistently lower in weight than the roots from any other plot. Tlie results of the absence of a nitrogen application to plots 1 and 40 were not positive because there were numerous roots from other plots receiving nitrogen that were no heavier individually, and the average weights for plots 32 and 39 were as small. Comparing plot averages in the series 31 to 39, the average weights of roots from plots 32, 35 and 38 were consistently lower than those of the roots from plots 31, 34 and 37, which indicated the probable effect of a spring top-dressing to be an increase in the size of the roots. Neverthe- less, the variations in weights of individual roots from any one of the plots is wide, and renders the conclusion from averages doubtful. . The effect of fertilizers on the inorganic constituents was thoroughly studied by the complete ash analysis of each root dug in 1908, and similar work on composite samples from the different plots in 1910. All the ash analyses were made in the fertiUzer section by Messrs. H. D. Haskins and L. S. Walker, to whom the writer is indebted for the data which appear in th^ tables. Inorganic Composition of Asparagus Roots {Percentages in Dry Matter). Roots of 1908. Averages by Plots. Plot. Total Ash. Calcium Oxide. Mag- nesium Oxide. Potas- sium Oxide. Sodium Oxide. Phos- phoric Acid. Sulfuric Acid. 1, 5.53 .30 .14 2.12 .07 .44 .35 31, 5.96 .26 .14 2.03 .24 .48 .39 32. 6.63 .29 .13 2.62 .18 .56 .38 33, 6.61 .29 .15 2.33 .18 .52 .45 34, 7.12 .35 .16 2.62 .31 .55 .49 35. 6.46 .31 .16 2.51 .22 .48 .51 36. 6.49 .29 .14 2.23 .27 .49 .48 37. 6.41 .32 .14 2.15 .25 .52 .52 38. 7.01 .40 .16 2.44 .21 .56 .56 39. 6.41 .30 .14 2.47 .32 .47 .47 40, 5.89 .29 .12 2.45 .07 .50 .45 280 MASS. EXPERIMENT STATION BULLETIN 171. Inorganic Composition of Asparagus Roots Roots of 1910. Concluded. Averages bt Plots. Plot. Total Ash. Calcium Oxide. Mag- nesium Oxide. Potas- sium Oxide. Sodium Oxide. Phos- phoric Acid. ' Sulfuric Acid. 5 6 7 8 9 10 11 12 6.81 7.09 7.54 7.34 5.94 6.17 6.18 7.10 .41 .32 .37 .38 .33 .34 .40 .18 .16 .21 .19 .19 .18 .19 .20 2.36 2.66 2.73 2.55 1.44 2.10 2.21 2.53 .43 .35 .38 .33 .55 .48 .33 .33 47 46 46 49 44 42 46 48 .63 .69 .63 .66 .57 .62 .62 There was no specific effect of fertilizers observable in the ash con- stituents, except on plots 1 and 40 in the 1908 series, and plot 9 of the 1910 series. Soda was notably lower in the roots from the first-named plots, which had received no nitrate of soda, than in all other roots which had been dressed with that salt. The composite sample representing the last-named plot, which had received no potash salt, showed a much lower percentage of potassium oxide than any other sample of that year, and a small increase in sodium oxide. The most notable fact observable in the ash constituents was the high percentage of sulfuric acid relatively to phosphoric acid. Withholding acid phosphate from plot 5 had no apparent effect in reducing either the phosphoric acid or the sulfuric acid in the sample from that area. Total Nitrogen in the Dry Matter of Asparagus Roots. Roots of 1908. Plot. Root I. Root II. Root III. Root IV. Plot Average. 1 1.21 1.29 1.36 1.30 1.29 31 1.69 1.36 1.30 1.89 1.56 32 1.96 1.93 1.65 , 1.54 1.77 33 1.70 1.36 1.51 2.31 1.72 34 2.43 2.01 2.18 2.12 2.18 35 2.23 2.14 2.51 2.01 2.22 36 1.56 1.92 2.05 2.16 1.92 37 1.92 1.99 2.10 1.87 1.97 38 • 2.20 2.51 1.92 1.22 2.51 2.08 1.21 2.19 2.10 1.20 2.35 39 1.84 1.98 40 1.50- 1.28 A CHEMICAL STUDY OF THE ASPARAGUS PLANT. 281 Total Nitrogen in the Dry Matter of Asparagus Roots — Concluded, Roots of 1910. Plot. Root A. Root B. Root C. Root D. Plot Average. 1 1.67 1.77 2.20 2.25 2.24 2.05 2.46 1.93 2.45 1.69 2.28 1.97 1.92 1 79 5 2.14 2.27 6 2.12 2.07 2.13 2.18 S 1.94 2.08 1.81 1.99 2.33 2.15 2.44 2.04 9 1.82 2.10 10 2.40 1.98 1.61 2.18 2.04 11, . • 2.26 1.90 1.86 2.23 2.06 12 1.91 2.27 1.43 2.25 1.76 1.87 1.77 2.07 31 1.72 1.67 32 1.81 1.96 2.02 2.02 2.02 1.59 2.23 2.01 1.99 2 00 33 1.73 1.84 34 2.02 35 2.01 1.95 1.87 2.23 2.01 36 2.07 1.79 2.24 1.91 1.91 2.00 1.79 1.94 37 1.90 1.96 38 2.32 2.07 2.44 2.60 1.66 1.89 2.02 2.22 39 1.82 1.98 40 1.59 1.30 1.06 1.22 1.29 Total nitrogen was determined in every root sample. The results in- dividually and by plot averages are consistent. The absence of nitrogen in the top-dressing results in a low percentage of nitrogen in the roots from plots 1 and 40. The minimum and medium applications of nitrate show results on the percentages of nitrogen in the roots following the same order in relative quantities. The maximum application of nitrate of soda produced no result in excess of the medium apphcation. The application of the nitrate in midsummer was accompanied by a positively higher percentage of nitrogen in the roots from those plots, viz., plots 32, 35 and 38. There was no apparent effect of fertilizers on the organic constituents of the roots, except that due to the influence on the nitrogenous group. High protein was accompanied by a lessened sugar percentage, but low sugar percentages also frequently occurred with low protein, in which condition there was a high fiber content. Consequently sugar and fiber fluctuated -widely in samples from the same plot on account of some condi- tion that was independent of fertilizers. Tliis wide fluctuation was most extreme in plot 9 of the 1910 series. 282 MASS. EXPERIMENT STATION BULLETIN 171. and if the average for the plot were compared with those of the others in the series it would appear clearly to be an illustration of the effect of potassium on the formation of sugar; but there were two roots with normal percentages of sugar from the plot, while there were roots in plots 5, 7 and 8 which were abnormally low where muriate of potash was regu- larly applied in the normal quantity. It is the writer's opinion that these variations in sugars on tliis group of plots may have been due to an attack of rust in the summer of 1910, although special pains were taken to avoid plants wlxich had thus suffered, when the sample roots were selected. Furthermore, it is believed that there were two positively different types of plants in these series in mode of growth, viz., one type with numer- ous slender, long roots, and the other with fewer but thicker, fleshier roots. This fact was not noted soon enough to correlate the observations with the analytical data, but it is reasonable to assume that the slender roots would have more epidermis in proportion to volume than the fleshy roots, which renders it probable that the former would have more fiber and less sugar than the latter. Orgmiic Composition of Roots. Roots of 1908. Plot and Root. Moisture. Protein. Fiber. Sugars. Pentosans Fat. 1(1.) 2.09 7.37 14.91 47.12 - - 1 (II.). 2.14 7.90 12.70 49.72 7.17 .80 1 (III.), 2.77 8.34 18.20 40.24 8.82 1.20 1 (IV.). 2.33 7.93 16.30 44.44 - - 31 (I.). 2.00 10.36 15.07 42.28 - - 31 (II.), 2.70 8.34 14.78 43.96 8.91 1.04 31 (III.). 2.16 7.93 14.92 44.36 9.00 .98 31 (IV.). 3.08 11.58 14.92 40.00 - - 32 (I.), 2.37 12.07 15.10 - - - 32 (II.). 2.59 11.82 14.86 - - '- 32 (III.), 2.18 10.10 13.98 42.00 8.25 1.05 32 (IV.). 2.59 9.45 18.22 38.04 9.18 1.05 34 (I.). 4.06 14.65 14.15 35.24 8.17 .56 34 (II.), 4.00 12.08 14.66 37.12 8.51 .68 34 (III.), 3.64 13.19 14.19 - - - 34 (IV.). 3.42 12.81 14.62 - - - 35 (I.). 2.06 13.69 17.23 36.00 8.46 1.09 35 (II.). 3.19 13.00 12.98 40.60 7.88 1.07 35 (III.). 4.31 15.12 15.22 - - - 36 (IV.), . 3.34 12.13 15.26 - - - A CHEMICAL STUDY OF THE ASPARAGUS PLANT. 283 Organic Corn-position of Roots ■ — ■ Continued. Roots of 1908 — Concluded. Plot and Root. Moisture. Protein. Fiber. Sugars. Pentosans. Fat. 37(1.) 2.34 11.68 13.50 43.20 7.88 .81 37 (II.), 3.16 12.07 14.86 - - - 37 (III.). 3.01 12.69 13.52 - - - 37 (IV.). 2.91 11.32 18.31 33.16 9.65 1.20 38 (I.). 3.35 13.31 15.54 - - - 38 (II.), 3.02 15.27 17.40 24.28 10.24 1.13 38 (III.), 3.60 15.19 13.18 - - - 38 (IV.), 2.92 13.33 12.31 44.52 7.94 .84 40 (I.), 2.50 9.14 13.69 - - - 40 (II.), 2.71 7.31 13.91 - - - 40 (III.), 2.07 7.37 14.54 44.32 8.27 1.31 40 (IV.), 2.13 7.30 11.97 48.72 8.48 .87 Roots of 1910. 1(A) 3.56 10.07 - 34.80 - - 1 (B), .... 3.07 10.77 19.33 28.10 9.44 1.27 1 (C) 4.49 12.19 - 24.04 - - 1(D) 4.90 9.94 - 27.48 - - 5(A) 5.50 12.63 - 19.16 - 1.57 5(B) 4.70 13.14 ~ 30.24 - 1.47 5(C) 4.94 14.70 16.20 - " 5(D) 4.85 13.56 23.60 15.80 11.72 2.07 6(A) 4.00 12.69 - 23.64 - - 6(B) 5.20 13.31 - 25.76 - - 6(C) 3.21 11.64 - 21.08 - - 6(D) 4.33 11.75 - 23.84 - 7(A) 4.90 12.62 - 27.48 - - 7 (B) 3.70 13.58 - 26.20 - - 7 (C) 3.97 14.75 22.93 11.28 11.10 2.35 7 (D) 4.24 11.45 - 18.76 - - 8(A) 5.10 11.45 - 25.96 - - 8(B) 5.50 12.26 - 22.12 - - 8(C) 4.04 11.94 18.77 29.16 - 1.82 8 (D) 4.78 12.81 - 17.92 - - 9(A) 5.40 10.69 - 30.04 - - 284 MASS. EXPERIMENT STATION BULLETIN 171. Organic Composition of Roots — Continued. Roots of 1910 — Continued. Plot and Root. Moisture. Protein, Sugars. 9(B). 9(C), 9(D), 10 (A). 10 (B), 10 (C), 10 (D), 11 (A). 11 (B), 11 (C), 11 (D), 12 (A), 12(B), 12 (C). 12 (D), 31 (A), 31 (B), 31 (C). 31 (D), 32 (A). 32 (B), 32 (C), 32 (D), 33 (A), 33(B), 33 (C), 33 (D), 34(A), 34 (B), 34 (C), 34 (D), 35 (A), 35 (B), 35 (C), 35 (D), 36 (A), 4.21 4.65 5.40 5.40 3.61 4.56 4.20 5.40 4.03 4.27 5.20 5.30 3.59 4.95 4.48 3.42 3.28 3.47 3.54 4.70 4.93 4.55 4.42 3.58 3.49 4.67 4.90 3.69 3.42 4.88 4.00 25.10 21.76 16.83 15.83 24.08 11.04 9.56 23.64 29.40 30.68 30.44 32.84 30.88 32.64 33.28 27.48 28.12 39.48 32.40 31.12 32.84 27.04 23.20 34.60 29.25 33.08 25.32 29.60 33.70 38.92 11.79 11.63 10.71 9.87 A CHEMICAL STUDY OF THE ASPARAGUS PLANT. 285 Organic Composition of Roots — Concluded. Roots of 1910 — Concluded. Plot and Root. Moisture. Protein. Fiber. Sugars. Pentosans. Fat. 36(B) 5.53 10.51 17.83 28.32 - 1.40 36 (C). 4.19 11.40 - 29.80 - - 36 (D), 4.27 11.94 - 25.76 - - 37 (A). 4.49 11.31 - 35.24 - - 37(B). 4.74 13.44 - 32.00 - - 37 (C), 3.40 11.50 - 33.28 - - 37 (D), 4.18 10.69 _ 35.44 - - 38 (A). 3.95 13.94 - 39.80 - - 38 P). 4.32 12.37 - 36.96 - - 38 (C). 3.96 15.70 - 25.32 - 38 (D), 4.41 11.20 - 36.12 ~ - 39 (A), 4.28 10.87 - 37.30 - 39 (B). 4.64 14.63 - 35.04 - - 39 (C). 3.69 9.94 - 33.28 - - 39 (D), 3.54 12.19 - 26.60 - - 40 (A). 3.42 9.56 - 38.08 - - 40(B), 3.51 7.81 17.13 38.20 9.04 1.22 40 (C), 3.25 6.32 - 33.08 - - 40 (D), 3.70 7.32 - ».« - - Effects of Fertilizers on Asparagus Stalks. An attempt was made to determine the effect of fertilizers on the com- position of the young stalks, and on that of the tops in midsummer and late fall. On May 13, 1910, the day's crop from each of four plots in the Concord field was shipped by Mr. Prescott to the laboratory at Amherst. The four samples represented three plots dressed with the maximum amount of nitrogen and one plot which received no nitrogen. The analyses were limited to determinations of dry matter, ash and total nitrogen, and the results were as follows : — With Nitrogen. No Plot 37. Plot 38. Plot 39. Nitrogen, Plot 40. ■ 7.00 6.50 10.57 6.80 9.81 4.57 Ash in dry matter Nitrogen in dry matter, .... 10.14 4.72 10.76 4.49 286 MASS. EXPERIMENT STATION BULLETIN 171. There was a small variation in favor of the plots dressed with nitrogen in both nitrogen and dry matter. On May 17, 1911, a series of samples was collected in a similar manner from the home field in Amherst, where the material could be prepared for drying as soon as cut. These samples represented one plot without nitrogen, one without phosphoric acid, one without potash and one with a complete fertihzer. Nitrogen and dry matter were determined, and the figures are arranged below. No Nitrogen. No Phosphoric Acid. No Potash'. Complete Fertilizer. 8.04 7.50 5.31 7.61 5.17 7.57 Nitrogen in dry matter 5.33 "' In this series there was again a slight gain in nitrogen in the sample from the plot receiving a complete fertilizer, but there was no effect on the dry matter. On June 1, June 8 and June 14 the entire day's crop from each of four plots was saved and analyzed. These plots represented variations in quantities of nitrogen, phosphoric acid and potash applied as a dressing. The results are shown below for dry matter and nitrogen. Dry Matter. Nitrogen in Dry Matter. Plot. June 1. June 8. June 14. June 1. June 8. June 14. N+P+K 2N+P+K, .... N+2P+K. .... N+P+2K 7.61 7.49 7.62 7.91 7.65 7.70 7.63 7.84 7.73 7.51 7.90 5.00 4.89 5.12 4.98 4.72 4.77 4.87 4.67 4.61 4.37 4.84 There was little effect on the composition of the young stalks to be perceived by comparing the results of the first plot with those of each of the other plots. The dry matter varied within narrow limits, while the nitrogen showed a progressive decrease as the season advanced, which was independent of the fertilizers. There was a slight but consistent advantage shown by the double quantity of potash on dry matter results from the last plot. Effects of Fertilizers on Asparagus Tops. The period immediately following blooming was chosen as one of the stages of growth at which to study the effect of fertilizers on the develop- ment of reserve material in the tops for translocation to the roots. Up to this period the asparagus plant increases steadily in size, and presumably A CHEMICAL STUDY OF THE ASPARAGUS PLANT. 287 draws continuously through its roots on the soil for its required mineral matter while building up its organic matter in its green branches. There has been probably but little transfer of sugars and proteins to the roots during this growing time, and it seemed as if any effect of the fertilizers on the formation of those constituents should be perceptible at this season. The material for this study was gathered from the experiment field at Concord by the selection of samples of tops from eight of the fertilizer plots, which represented wide variations in the method of fertilization. In order to disturb the subsequent growth of these plots as little as possible, not more than one stalk was removed from any plant. Each sample con- sisted of six stalks selected from as many typical plants on a plot. The samples were weighed immediately after being gathered, and were then packed in burlap sacks for shipment to the laboratory at Amherst. The samples were gathered on Aug. 13, 1912, and were delivered at the labora- tory forty-eight hours later. On the arrival of the samples at the laboratory they were again weighed and were found to have lost 13 per cent, of their field weight, of which loss a large part must have been due to respiration and the consequent destruc- tion of sugars. The branches were cut from the main stalks, and the latter broken in short pieces to facilitate drying, which was carried out by spreading the samples on benches in the greenhouse. It was not possible to dry all the samples simultaneously in the oven, so the greenhouse was selected as providing uniform conditions for them. At the end of five days each sample was cut into short lengths by a fodder cutter, after which a small subsaraple was separated by quartering. The small samples were next dried in the oven untU they were in a condition to be easily ground, when they were pulverized and passed through a millimeter sieve. Weights of Samples of Green Tops (Pounds). Plot 1 5.25 Plot 5 6.60 Plot 9 5.25 Plot 11 6.75 Plot 31, 6.25 Plot 32, 6.25 Plot 34, 6.15 Plot 35, 5.65 The absence of nitrogen on plot 1 and of potash on plot 9 was accom- panied by the lightest weights of samples. Plots 11 and 34 received equal amounts of the complete fertilizer in the spring, and their samples ex- ceeded in weight those of 1 and 9. The absence of phosphoric acid from plot 5 did not affect the weight. Just before the needles had dropped in the fall was selected as another stage at which to study the effect of the fertilizers on the composition of the tops and on the development of reserve material. For this purpose Mr. Prescott was asked to procure some samples from the Concord field. 288 MASS. EXPERIMENT STATION BULLETIN 171. In accordance with our instructions Mr. Prescott selected four average plants on each of the plots from which a sample was desired, and removed the entire tops from the crowns. Each plot sample was wrapped in paper and then put in a jute sack for shipment to the laboratory. The samples arrived at the laboratory on October 23 with the outer sacks somewhat wet as though rained upon, which was not unlikely since the period was especially rainy. On opening the sacks the tops were found to be damp, and a slight mold was observed on some of the twigs. The material was cut into short lengths with a fodder cutter and spread above the steam coils in the greenhouse. A few days later the samples were quartered and the subsamples were dried in the steam-heated oven until they could be readily ground and sifted. Partial Composition of Asparagus Tops. Midsum7neT Tops. Plot 1. Plot 5. Plot 9. Plot 11. Plot 34. Ash in dry matter, 10.61 9.00 8.55 9.21 9.69 Protein, 17.87 17.00 17.44 17.56 18.50 Fiber 32.62 33.62 31.58 34.34 - Ether extract 2.46 2.70 3.04 2.66 - Sugars 5.11 5.32 6.33 4.41 4.96 Late Fall Tops. Ash in dry matter, 12.12 7.84 6.68 _ 8.97 Protein 8.44 8.31 7.50 - 8.62 Fiber 41.89 44.93 3.28 46.30 3.56 : 41.23 Ether extract 3.68 3.40 Pentosans, 20.71 21.44 21.60 - 20.14 Partial Composition of Asparagus Tops. Midsummer Tops. Plot 31. Plot 32. Plot 34. Plot 35. Ash in dry matter, 9.58 17.44 10.33 17.12 5.00 9.69 18.50 4.96 8.70 17.94 Sugars, 5.66 3.75 Late Fall Tops. Ash in dry matter, Protein, Pentosans, 8.90 8.12 20.46 7.95 7.50 21.15 8.97 8.62 20.14 8.06 7.06 20.83 A CHEMICAL STUDY OF THE ASPARAGUS PLANT. 289 Plot 1 lacked nitrogen, plot 5 lacked phosphoric acid and plot 9 lacked potash. Plots 11 and 34 received the complete fertilizer in medium amount. Plots 34 and 35 received one and one-half times the amount of nitrogen that was applied to 31 and 32. Plots 31 and 34 received their nitrogen in the early spring, while 32 and 35 had their portions applied in late June. The high ash occurring in both seasons in the tops from plot 1 was apparently due to fine earth wliich adhered to them, as there was much insoluble residue after testing the ash with strong acid. On the other hand, the samples from plot 9 showed a low ash, which was without doubt due to the lack of potash. The development of protein and sugar was not perceptibly affected by the lack of fertiUzers, since there is no consistent relation between the percentages and the amounts. A comparison of the two pairs of plots which received nitrogen at different seasons shows that the tops from the plots dressed with nitrates in summer contained slightly less protein than those from the plots dressed in the spring. This was also the result on the single pair of plots (37 and 38) from wliich the young stalks were sampled in 1910. With the two pairs of plots under comparison there was a slight advantage in the amounts of protein found in the tops from the larger quantities of nitrogen. The effect of fertilizers on the proportions of inorganic constituents in the different stages of tops was not studied because the slight effects pro- duced on the roots did not warrant such a laborious comparison. Effect of J'ertilizers on Asparagus Roots at the End of the CuTTE^fG Season. The summer samples of roots were dug from plots receiving two differ- ent quantities of nitrogen at two different seasons for the purpose of measuring whether the exhaustion of the roots during the growth of the crop was influenced by amount or season of application of nitrate of soda. Plots 34 and 35 received one and one-half times as much nitrogen as 31 and 32, while 31 and 34 received it in the spring and 32 and 35 in the summer. Total nitrogen and sugar showed consistent variations relative to the different treatments, but none of the other constituents could be corre- lated and are not tabulated. The roots dressed with the larger amount of nitrogen contained higher percentages of nitrogen and sugar than those which received the smaller amount. Roots receiving their nitrogen in summer after the cropping season still contained a little more nitrogen than the others. Sugar, however, was more exhausted than in the roots which had received their nitrogen in spring. 290 MASS. EXPERIMENT STATION BULLETIN 171. Comparative Effects of Spring and Summer Top-dressing on Asparagus Roots at End of Cutting Season. Fresh Weight (Grams). Per Cent. . Plot and Root. Dry Matter. Total Nitrogen. Total Sugar. 31 (I.), 600 17.15 1.37 22.40 31 (II.), . 2.744 21.37 1.78 34.54 31 (III.). 1.995 18.52 2.24 19.92 31 (IV.). . 1,970 19.77 1.86 26.17 Average, 1.825 19.20 1.81 25.76 32 (I.), . 1.400 16.04 1.94 19.60 32(11.). . 2.060 15.87 2.06 7.68 32 (III.), 3,830 14.43 2.47 7.40 32 (IV.). . 3,375 19.97 1.79 18.53 Average. 2,666 16.58 2.06 13.30 34 (I.). . 2,750 17.88 1.84 26.43 34 (II.), . 3,150 19.81 1.80 32.67 34 (III.), 3,400 20.58 2.22 36.14 34 (IV.). . 1.805 15.59 2.33 16.13 Average, 2.776 18.46 2.06 27.84 35 a.), . 2.945 18.88 2.12 29.87 35(11.). . 860 23.14 2.22 31.17 35 (III.). 3.180 21.25 2.18 26.70 35 (IV.), . 2.355 17.67 2.10 15.27 Average. 2.335 20.24 2.15 25.90 Sugar fluctuated widely in individual roots, and the value of the aver- ages is somewhat doubtful. The weights of roots from the same plot vary as widely as the weights from different plots, so that no conclusions can be drawn from the size of roots. The general effect of varying the season of top-dressing with nitrate of soda was very small and inconclusive. Reserve Material required to produce a Crop of Young Stalks. An attempt is here made to determine the amount of reserve material drawn from the roots during the spring cutting season. For this purpose use is made of the average composition of fall roots, spring stalks and summer roots, and the average weights obtained from the four plots A CHEMICAL STUDY OF THE ASPARAGUS PLANT. 291 numbered, respectively, 31, 32, 34 and 35 of fall roots, summer roots and the spring crop of stalks. The calculated results are necessarily approximate because identical roots cannot be analyzed at two successive stages of growth, but the com- parison suggests possibilities if not absolute conditions. The average weights of roots were obtained from the samples collected in 1910 and 1911. The average weight of the crop of stalks is calculated from the total weights cut on the four plots in 1911. The number of plants per plot was originally 250, but four roots were removed in 1908 and four more in 1910. Grams of Constituents in Roots and Crop of an Average Playit. Autumn Roots, 1910. Summer Roots, 1911. Spring Crop, 1911. Green weight 2.393.00 2,401.00 447.00 Dry matter 504.90 447.00 34.40 Total sugar, 159.24 103.70 5.23 Fiber, pentosans and lignin 239.22 239.10 22.25 Fat 8.93 7.28 1.05 Protein, 62.81 56.99 10.52 Ash 34.78 39.91 2.97 Total nitrogen, 10.05 9.12 1.68 Protein nitrogen, 5.35 5.81 1.05 Amino nitrogen, 4.70 ?.31 .63 Potassium oxide, 12.44 10.61 1.80 Sodium oxide 1.85 1.63 .11 Calcium oxide, 1.81 1.95 .13 Magnesium oxide, .97 .82 .12 Phosphoric acid, 2.34 1.97 .18 Sulfuric acid 3.12 3.26 .28 The average weight of crop per plot was 238.6 pounds (108.3 kilos) which, divided between 242 plants, gave a Kttle less than a pound, or 447 grams, per plant. When the combined weights of the different constituents of summer roots and spring crop were balanced against the weights of the same con- stituents in the autumn roots there was noted a marked loss in organic matter and a pronounced gain in inorganic matter. The loss of organic matter was confined almost wholly to the sugar, as there was but a small deficit in the quantity of fat. The total carbo- hydrate matter in the spring crop amounted to 27.48 grams, while the difference between the quantities of sugar in the autumn and summer roots was 57.54 grams. There was an increase in protein of 4.7 grams 292 MASS. EXPERIMENT STATION BULLETIN 171. over the amount present in the autumn roots, which might require a little of the sugar in its synthesis; but, on the other hand, the study of the progressive changes in composition of young stalks indicated that they synthesized a part of their sugar before thej'- were of marketable size. Therefore the comparison in this case showed that for every gram of carboh5'-drate developed in the j^oung stalk at least two grams disappeared from the parent root, one of which must have been used in maintaining the energy of the growing plant, just as the young animal uses a large part of its food in maintaining its body energy. The gain in protein during the growth of the crop is of interest in con- nection with the problem of nitrogen fertilization. The transfer of nitro- gen from the autumn root to the growing stalk was apparently accom- plished by using only the amino nitrogen of the reserve in the parent crown, and drawing on the soil nitrogen. The increase in nitrogen of summer roots and crop over the amount in the autumn roots is .75 gram, or 7.5 per cent., and is not of sufficient amount to show the necessity of a spring application of nitrogen. The gain in ash was confined to calcium oxide and sulfuric acid of the determined constituents, while a part of the variation was undoubtedly due to the very fine sand of the soil which had escaped the cleaning process to which roots and stalks were subjected. Calcium oxide and sulfuric acid gained, respectively, .27 gram and .42 gram, or 14 per cent, and 13 per cent. Potassium oxide and magnesium oxide were almost exactly balanced on the two sides, while sodium oxide and phosphoric acid had shght amounts unaccounted for, which may have been due to the difficulties in exact determinations of these constituents in organic substances. These comparisons show but little, if any, immediate effect on the spring crop of a spring application of fertilizers. There was a slight apparent absorption of nitrogen, a more marked intake of lime and sulfuric acid, perhaps in combination, and no apparent use at this period of potash and phosphoric acid. But as already remarked, these comparisons can be regarded as merely suggestive. Amount of Vegetable Matter contained in Ripened Asparagus Tops. The method of asparagus culture now followed by many growers in Massachusetts leaves the tops to die down in the autumn and in the spring works them into the soil by means of a disc harrow. On the experi- ment field a number of the plots have received no annual dressing of manure, and the humus in the soil has been replenished only by the annual growth of tops. In the autumn of 1912 Mr. Prescott was requested to determine the weights of the ripened tops on several plots that had received only chemi- cal fertilizers. Mr. Prescott selected one rod of row on each plot, where A CHEMICAL STUDY OF THE ASPARAGUS PLANT. 293 there were seven consecutive plants to the rod. The stalks were cut level with the ground and weighed. This work was done in the last week in October when the sap had mostly left the stalks. The weights per plot were as follows : — Weights of Tops per Rod of Roiv, Autumn of 1912 (Pounds). Plot 1, without nitrate of soda, . . . . . . . .3.5 Plot 3, complete fertilizer, . . . . . . . . .5.5 Plot 5, without acid phosphate, . . . . . . . .4.0 Plot 7, complete fertilizer, . . . . . . . . .4.0 Plot 9, without muriate of potash, . . . . . . . .4.0 Plot 11, complete fertilizer, ......... 6.5 Plot 34, complete fertilizer, . . . . . . . . .4.0 Plot 40, without nitrogen, . . . . . . . . .3.0 Average, . . . . . . . . . . .4.3 At the rate of 250 plants per plot, or 5,000 plants per acre, these results from 7 plants would give 3,071 pounds of dying tops per acre. Samples of stalks gathered early in November at Amherst contained 49 per cent, of dry matter, by which it is estimated that there were about 1,500 pounds per acre of dry vegetable matter added to the soil of the asparagus field per year. Rousseaux and Brioux ^ report, as the result of five different fields in France, a range of from 891 to 2,128 kilos per hectare for the dry matter in the crops of the tops removed in late autumn from the fields, in accord- ance with French practice, "^heir average dry matter per hectare was 1,579 kilos, or about 1,400 pounds, per acre. In percentage of soil per acre this amount of tops is really small. On such sandy soil as the Concord field the tops would be worked into the surface 4 inches, or mixed with approximately 1,000,000 pounds of soil, which would enrich the soil with not more than .15 per cent, of organic matter. Nevertheless, several of the best plots in the experiment field have received no more organic matter than is contained in the tops, which is a good illustration of the effectiveness of small annual additions of organic matter to our soils. Relation of Asparagus Roots to Weights of Stalks. It was expected that there would be a close relationship found between the size of roots from a plot and the total weight of stalks cut from it, and an attempt was made to correlate the weights of sample roots in 1910 with the weights of crops over a period of five years. In the phosphate group of plots, 5, 6, 7 and 8, the smallest roots were obtained from the plot that received no phosphate in the top-dressing; but the crop jdelds were not invariably the lowest in the series. Plot 8, which received the maximum dressing of acid phosphate, yielded much I Annal. d. Sci. Agron., 1906, pp. 188-326. 294 MASS. EXPERIMENT STATION BULLETIN 171. smaller roots than plots 6 and 7, but its crop yield was the maximum in every year but the fifth, when its yield was exceeded by plot 7 with a fraction of a pound. The weight of roots in the potash group of plots numbered 9, 10, 11 and 12 increased from 9 without potash to 11 with a medium application. The ^-ield of stalks followed the same order each year. The nitrate of soda group included ten plots numbered 31 to 40, inclu- sive. The weights of individual roots from any one plot varied consider- ably from the average for that plot, but the plot averages showed fairly consistent changes in size of roots with amount of nitrogen applied in the top-dressing. The weights of roots from plots 31, 32 and 33 were, respec- tively, smaller plot by plot than the weights of roots from plots 34, 35 and 36. The weights of crops did not follow the same order, but were in several instances reversed. The application of nitrate of soda in the spring on plots 31, 34 and 37 resulted in much larger roots than the summer dressing apparently pro- duced on 32, 35 and 38. On the other hand, the weights of crops from the summer-dressed plots were in nearly all cases the larger. Plot 40 without nitrate yielded roots no lighter in weight than plot 39, which received a maximum dressing of nitrate of soda, divided between spring and summer. The yield of stalks was, however, much smaller on plot 40 than on 39. The small roots with large yields contained higher percent- ages of nitrogen than the roots bearing smaller crops, so there was diffi- culty in correlating roots with crops of stalks, since the variations in pro- portions of root constituents were possible factors in influencing growth of stalks. • Weight of Asparagus Stalks cut in the Spring (Pounds).'^ Plot. Application. 1910. 1911. 1912. 1913. 1914. 5 No phosphate, 232.3 221.1 270.9 388.0 404.2 6 Minimum phosphate 241.4 221.1 2'3.4 385.8 420.4 7 Medium phosphate, 241.6 240.4 281.1 387.7 436.9 8 Maximum phosphate, ...... 252.8 251.6 298.4 403.3 436.4 9 No potash, 208.6 210.6 258.6 324,0 366.7 10 Minimum potash, 237.2 237.3 284.7 373,6 408.4 11 Medium potash 276.5 289 9 342,0 446,8 478.9 12 Maximum potash, 262.7 269.6 302,8 409.8 458.5 31 Minimum nitrate, spring 220.9 223.7 272,4 375.1 395.7 32 Minimum nitrate, summer, 221.3 242,2 284.4 401.6 406.3 33 Minimum nitrate, half in spring, half in summer, . 222.6 239.7 291.2 378.4 389.8 34 Medium nitrate, spring, ...... 214.2 240.6 288.0 381.9 378.6 35 Medium nitrate, summer, 216.0 247.8 288 9 368.3 368.5 36 Medium nitrate, half in spring, half in summer, . 210.2 224.2 268.5 357.4 362.2 37 Maximum nitrate, spring, ..... 193.9 223.2 283.8 345.2 340.9 38 Maximum nitrate, summer, 196.2 234.9 303.0 367,1 347.5 39 Maximum nitrate, half in spring, half in summer, . 214,2 230.7 288,4 358,6 351.6 40 No nitrate 181.2 202.2 263.4 307.5 314.3 For Table of Weights of Roots see p. 278, series of 1910. A CHEMICAL STUDY OF THE ASPARAGUS PLANT. 295 SUMMAEY. During the earlier years of the asparagus field the crowns and roots steadily increased in size, doubhng in weight between the second and fourth years after setting. The proportion of protein remained nearly constant in the dry matter of the roots during the period observed, while the sugar decreased and the cellulose and alUed compounds increased. The composition of the young stalks cut in the spring changed as the cutting season advanced. Dry matter was practically constant, but sugar increased in proportion while protein decreased somewhat. The development of the asparagus tops to maturity was accompanied by a continuous increase in the cellulose and its related groups, — pento- sans and lignin. Protein and sugar decreased in their proportions, but were not wholly translocated to the roots from the ripened tops. Water was the dominant constituent of the asparagus plant in all the stages studied. It was highest in the young stalks. The summer or grow- ing roots were a little more watery than the late fall or storage roots. Calcium oxide and sulfuric acid steadily accumulated in the asparagus tops as they grew old, but potash and phosphoric acid were transferred either to the fruit or back to the roots. Withholding one of the constituents of a complete fertilizer from the annual top-dressing was accompanied by a smaller average weight of roots in the samples taken from the plot thus treated. Withholding ni- trate of soda lessened the percentage of nitrogen and of soda in the roots; withholding muriate of potash lessened the proportion of potash in the roots; withholding acid phosphate produced no apparent change in the constituents of the roots. An increase of nitrate of soda from the minimum to the medium amount in the top-dressing caused an increase in the percentage of nitrogen in the dry matter of the roots. An increase in the amount of muriate of potash produced some increase in the percentage of potash in the roots. Asparagus roots taken from plots receiving the nitrate of soda in the spring were noticeably heavier in weight and a little poorer in nitrogen than roots from plots that were top-dressed with nitrate in the summer. During the cutting season the production of young stalks drew most heavily on the sugar contained in the roots, but there was no approach to exhaustion of that constituent. Fully twice as much sugar was consumed as would have been required to produce the carbonaceous matter in the young stalks. The roots apparently absorbed nitrogen, lime and sulfuric acid during the cutting season. Potash and phosphoric acid were apparently supplied to the young stalks wholly from the reserves in the roots. 296 MASS. EXPERIMENT STATION BULLETIN 171. Pkactical Conclusions from the Chemical Study of the Asparagus Plant. Asparagus roots that had been set in the spring of 1907 were found to have doubled in size and weight between November, 1908, and November, 1910. During this period of rapid growth the percentages of the different fertilizing constituents in the dry matter remained constant or else in- creased slightly. Absence of nitrogen, phosphoric acid or potash from the annual top- dressing was found to limit the growth of the roots. Withholding nitrate of soda from the top-dressing, or appl3dng it in relatively small amounts, resulted in lessening the percentages of nitrogen in all parts of the plant. A complete fertilizer rich in nitrogen is clearly shown to be required in generous amounts in order to produce a continuous strong development of the asparagus plant. Water is of prime importance in all parts of the asparagus plant at all stages of growth. It is especially important in the spring months during the cutting season, since the young stalks contain about 92 per cent, of water, while the roots at this period are more watery than in the fall. The physiological need of water, together with the sandy quaHty of most asparagus soils, indicates that irrigation would be advantageous if not necessary in the production of maximum crops. The reserve material stored in autumn in the roots was found to be principall}^ sugars. Sugars were also prominent in the spring stalks and both summer and fall tops. The synthesis hi sugar in the tops and its translocation to the roots appeared to continue untU the tops were killed by frost. Destruction of the tops by rust, or their premature removal to be rid of the berries, must lessen the amount of sugar which can be stored in the roots. The fertilizing constituents which were stored in the roots over winter appeared to be nearly, if not quite, sufficient for the full development of the succeeding spring crop. There was evidence of a small intake of nitrogen during the cropping season, and a pronounced absorption of lime and sulfuric acid. Sulfuric acid was found to be equally, if not more, important than phosphoric acid among the constituents of the asparagus plant. Never- theless, the sulfate of lime in the acid phosphate appeared to suffice fully for the needs of the crop. BULLETIN No. 172 MARCH, 19 17 MASSACHUSETTS AGRICILTIRAL EXPERIMENT STATION EXPERIMENTS IN KEEPING ASPARAGUS AFTER CUTTING By F. W. MORSE The object of this experiment was to determine some of the changes which take place in asparagus from the time when it is cut in the field until it is ready to be cooked. It is not usually desirable to hold asparagus more than a few days to prevent market gluts. The usual methods of keeping asparagus at sum- mer temperatures cause rapid deterioration in quality, and should be remedied if a discriminating patronage is desired. Requests for bulletins should be addressed to the Agricultural Experiment Station, Amherst, Mass. IMassachusetts Agricultural Experiment Station. OFFICERS AND STAFF. Trustees. COMMITTEE. Charles H. Preston, Chairman, Wilfrid Wheeler, Edmund Mortimer, Arthur G. Pollard, Harold L. Frost, . Hathome. Concord. Grafton. Lowell. Arlington. The President of the College, ex officio. The Director of the Station, ex officio. STATION STAFF. Administration. William P. Brooks. Ph.D., Director. Joseph B. Lindsey, Ph.D., Vice-Director. Fred C. Kennet, Treasurer. Charles R. Green, B.Agr., Librarian. Mrs. Lucia G. Church, Clerk. Miss F. Ethel Felton, A.B., Clerk. A|:ricultural Economics. Alexander E. Canoe, Ph.D., Agricultural Economist. S. H. DeVault, A.m., Graduate Assistant. Agrriculttire. William P. Brooks, Ph.D., Agriculturist. Henry J. Franklin, Ph.D., In Charge Cranberry Substation. Edwin F. Gaskill, B.Sc, Assistant Agriculturist. Robert L. Coffin, Assistant. Botany. A. Vincent Osmun, M.Sc, Botanist. George H. Chapman, Ph.D., Research Physiologist. Paul J. Anderson, Ph.D., Associate Plant Pathologist. Orton L. Clark, B.Sc, Assistant Plant Physiologist. Miss Grace B. Nutting, Ph.B., Curator. Miss Ellen L. Welch, Stenographer. Chemistry. J. B. LiNDSEY, Ph.D., Chemist. Edward B. Holland, Ph.D., Associate Chemist in Charge (Research Section). Fred W. Morse, M.Sc, Research Chemist. Henri D, Haskins, B.Sc, Chemist in Charge (Fertilizer Section). Philip H. Smith, M.Sc, Chemist in Charge (Feed and Dairy Section). Lewell S. Walker, B.Sc, Assistant Chemist. Carleton P. Jones, M.Sc, Assistant Chemist. Carlos L. Beals, M.Sc, Assistant Chemist, James P. Buckley, Jr., Assistant Chemist. W. A. Allen, B.Sc, Assistant Chemist. J. B. Smith. B.Sc, Assistant Chemist. Robert S. Scull, B.Sc, Assistant Chemist. James T. Howard, Inspector. Harry L. Allen, Assistant in Laboratory. James R. Alcock, Assistant in Animal Nutrition. Misa Alice M. Howard, Clerk. Miss Rebecca L. Mellor, Clerk. Entomology. Henry T. Fbrnald, Ph.D., Entomologist. Burton N. Gates, Ph.D., Apiarist. Arthur I. Bourne, A.B., Assistant Entomologist. S. C. Vinal, B.Sc, Graduate Assistant. Miss Bridie E. O'Donnell, Clerk. Horticulture. Frank A. Waugh, M.Sc, Horticulturist. Fred C. Sears, M.Sc, Pomologiat. Jacob K. Shaw, Ph.D., Research Pomologist. Harold F. Tompson, B.Sc, Market Gardener. R. P. Armstrong, M.Sc, Graduate Assistant. Miss Eleanor Barker, Clerk. Meteorology. John E. Ostrander, A.M., C.E., Meteorologist. Microbiology. Charles E. Marshall, Ph.D., Microbiologist. F. H. Hbsselink van Suchtblen, Ph.D., Research Microbiologist. Poultry Husbandry. John C. Graham, B.Sc, Poultry Husbandman. Hubert D. Goodale, Ph.D., Research Biologist. Lloyd L. Stewart, B.Sc, Graduate Assistant. Miss Rachael G. Leslie, Clerk. Veterinary Science. James B. Paige, B.Sc, D.V.S., Veterinarian. G. Edward Gage, Ph.D., Research Pathologist. J. B. Lentz, V.M.D., Assistant. Publication of this Document approved by The Supervisor of Administration. BULLETII^ ]^o. 172. DEPARTMENT OP CHEMISTRY. EXPERIMENTS IN KEEPING ASPARAGUS AFTER CUTTING. BY F. W. MOKSE, The object of this experunent was to determine some of the changes which take place in asparagus from the time when it is cut in the field until it is ready to be cooked. This period varies from a few hours to several days, and during it there is seldom any care taken to preserve the asparagus stalks in a fresh, crisp condition. Sometimes the stalks are kept with their butts in water; but this is not a general practice among the dealers in this vegetable. Fruits and vegetables are living things and life is maintained by respira- tion, which requires a supply of food just as with animals. When animals fast they lose weight because their body material is used in respiration. When vegetables and fruits are removed from the plants on which they grew they steadily lose ia weight because of respiration, and their chemical composition continually changes. Experiments with apples ^ have clearly shown that after the fruit is picked from the tree respiration is maintained by which carbon dioxide and water are continually exhaled, while analysis has proved that sugar steadily diminishes and the fruit loses in weight. It was found, too, that low temperatures slowed down the respiration while high ones speeded it up, and that retarding respiration was an important factor in the preserva- tion of fresh fruits. Besides investigating the nature of the change in asparagus after it has been cut from the plant, the effects of high and low temperatures on the rate of change have been studied as an important part of the experiment. The following table ^ gives the average composition of asparagus stalks when prepared for analysis as soon as practicable after they were cut from the plants: — Table I. Composition of Asparagus Stalks when Fresh {Per Cent.). Water, 92.30 Dry matter 7.70 ' F. W. Morse: The Respiration of Apples and its Relation to their Keeping. Biil. 135, N. H. Agr. Expt. Sta., 1908, 8 pp. Bui. 171, Mass. Agr. Expt. Sta., p. 274. 298 MASS. EXPERIMENT STATION BULLETIN 172. Per Cent, in Dry Matter. Ash 8.69 Protein, 30.77 Fiber 18.20 Fat 3.07 Total sugars 15.22 Reducing sugars, . . . .. . . . . . .11.17 Pentosans, . . . . . . . . . . .13.41 Lignin, etc., . .• . . . . . . . . .10.64 It will be noted that the succulent stalks contained over 92 per cent, of water, and that protein, fiber and sugar were the most abundant con- stitutents of the dry matter. Fiber forms the framework of the stalks, while the protein and sugar are the substances utilized most freely by the cells for food and growth. The two latter substances were studied as the means of determining the kind and rate of change occurring in the asparagus after cutting. Several experiments were conducted, each one varying a little in detail from its predecessor; therefore each experiment v,ill be separately de- scribed. Two were conducted in 1914 and the remainder in 1916. Experiment 1 . — This experiment was begun May 25, 1914. A quantity of stalks was brought to the laboratory immediately after the}^ were cut in the field. Each stalk was rinsed clean from adhering soil and wiped dry with a towel. The lot was then divided into three bunches of uniform size and appearance, and each bunch was weighed and placed under its assigned conditions. One bunch. A, was prepared at once for quick drying. The stalks were broken into pieces 2 to 3 inches long, which were spread in a single layer on a tray and placed in a large drying oven. The oven was heated by a steam coil which maintained a temperature between 50° and 60° C. This heat was sufficient to expel the water from the succulent stalks without softening them, as in cooking. The second bunch, B, was set in a jar with the butts in shallow water and left in the laboratory where the temperature would remain at sum- mer heat, or from 70° to 80° F. day and night. The third bunch, C, was loosely wrapped in paper and laid on the shelf in a refrigerator of the usual family size, kept well suppUed with ice, which held the temperature between 45° and 50° F. At the end of three days Cseventy-two hours), bunches B and C were again wiped dry with towels and weighed, after which they were pre- pared for the drying oven in the same manner as A. B was firm and brittle and had increased in weight over 15 per cent, by imbibing water. C was somewhat limp but not withered, and had lost a little over 3 per cent, of its original weight. When dried to a condition which permitted the asparagus to be easily ground to a powder, the samples were removed from the large oven, KEEPING ASPARAGUS AFTER CUTTING. 299 weighed and pulverized. The samples were then analyzed for absolute dry matter, total sugar, reducing sugar, protein, protein nitrogen and amino nitrogen, and the results are arranged in Table II. Table II. A, B. C. Weight fresh (grams) Weight after keeping (grams) 823 804 927 803 776 Per Cent, calculated on Fresh Weight. Water Dry matter, 92.36 7.64 93.20 6.80 92.75 7.25 Per Cent, in Dry Matter. Total sugars, Reducing sugars, Total protein, . Protein nitrogen, Amino nitrogen. 20.55 14.25 29.33 3.72 .97 14.11 10.10 30.75 4.01 .91 Experiment 2. — This experinient was begun June 2, 1914, and was carried out as nearly as possible in the same manner as Experiment 1, and the data are given in Table III. Table III. A. B. C. Weight fresh (grams), Weight after keeping (grams), 715.5 719.5 836.0 719.0 698.5 Per Cent, calculated on Fresh Weight. Water Dry matter 92.32 7.68 93.19 6.81 92.72 7.28 Per Cent, in Dry Matter. Total sugars, Reducing sugars, Total protein, . Protein nitrogen, Amino nitrogen. 27.39 20.29 28.46 3.49 1.06 12.41 7.46 32.90 3.57 1.69 18.91 12.68 31.39 3.92 1.10 300 MASS. EXPERIMENT STATION BULLETIN 172. Although B imbibed water and increased in weight, there was really greater destructi n of dry matter than in the bunch C, which was kept in the refrigerator. The actual amount of change under each condition is shown on the basis of 100 parts of fresh asparagus in Tables IV. and V. Table IV. — Experiment 1 . A. B. C. Dry^matter (per cent.), 7.64 6.80 7.26 Sugar (per cent.), . 1.57 .71 1.02 Protein (per cent.), . • • • 2.24 2.34 2.23 Protein was little changed, but sugar was partly destroyed, of^sugar was a little in excess of the loss of dry matter. The loss Table V. — Experiment A. B. C. Dry matter fper cent.) Sugar (percent.), Protein (per cent.) 7.68 2.10 2.19 6.81 .84 2.24 7.28 1.37 2.28 There was a marked change in the relative proportions of protein nitrogen and amino nitrogen in B in both experiments, as shown in Tables II. and III. The chemical activity changed the form of nitrogen compounds but not their total amount, as shown in Tables IV. and V. The work was not continued in 1915 on account of other investigations that seemed more important. In the spring of 1916 the investigation was resumed and several different experiments were conducted. Experiment 3. — This experiment was begun May 29, 1916, This lot of stalks was brought to the laboratory from the plots as soon as cut. The plots had not been cut over for two days and the stalks were too tall and the heads had begun to open too much for good marketable asparagus. The stalks were washed and scrubbed with a brush to remove all adhering soil, and wiped dry with towels. The lot was then separated into five bunches as uniform as possible in appearance, after which each bunch was weighed and placed under its assigned conditions. A was broken in short pieces, spread on a tray and placed in the oven at a temperature between 50° and 60° C. B was set upright in a jar with the butts in water and left in the laboratory at the room temperature. C was wrapped loosely in paper and laid on the shelf beside B. D was laid directly upon the cake of ice in the refrigerator. E was stood upright KEEPING ASPARAGUS AFTER CUTTING. 301 in a jar with its butts in water and set in the food compartment of the refrigerator. At the end of forty-eight hours bunches B, D and E were unbound and the stalks were wiped with towels. C, having been kept dry, needed no such dr v'ing. Each bunch was then weighed, after which it was prepared and put in the oven as A had been. The stalks in B were firm and crisp, but the heads were much opened. The stalks in C were limp and slightly withered, and a few would not break, but were cut into the proper lengths for drying. Those in D, Ijdng directly on the ice, were somewhat limp but unwithered, while those in E, standing in the water, were plump and firm, and the heads were unchanged in appearance. Both B and E had imbibed water, but B had gained almost 15 per cent, in weight, while E had gained only 10 per cent. C and D both lost weight; the former shrunk 21.7 per cent., while the latter lost only 3.7 per cent. Dry matter and total sugar were the only determinations made after the dried stalks were pulverized for analysis. Table VI. Weight Fresh (Grama). Weight after Keeping (Grams). Dry Matter from Fresh Weight (Per Cent.). Total Sugars from Dry Matter (Per Cent.). A B C, D. E, ...... . 677 654 590 714 633 751 512 688 697 6.72 6.47 6.30 6.65 6.49 15.96 12.14 12.35 14.63 11.23 Experiment 4. — This experiment was begun June 5, 1916. The ma- terial was much like that of the previous experiment, — a little too much developed for the best marketable condition. The stalks were washed and dried and arranged in five bunches which were subjected to conditions like those of Experiment 3. B and C were held but twenty-four hours, while D and E were continued throughout four days Cninety-six hours). B, in twenty-four hours, imbibed water and increased in weight 16.8 per cent. E, in four days, increased 13.7 per cent. Of the bunches kept dry, C, in the warm room, lost 8.2 per cent, in twenty-four hours, and D, on the ice, lost 5.4 per cent, in four days. The determinations in the dried material were confined to dry matter and sugar. 302 MASS. EXPERIMENT STATION BULLETIN 172. Table VII. Weight Fresh (Grains). Weight after Keeping (Grams). Dry Matter from Fresh Weight (Per Cent.). Total Sugars from Dry Matter (Per Cent.). A B, C, D E .528 534 549 589 544 624 504 557 619 7.50 7.18 7.31 7.34 7.20 20.60 16.31 17.34 17.91 17.99 Experiment 5. — The stalks were brought to the laboratory on the morning of June 15, 1916. The weather for two days had been cooler than usual, so that the asparagus had grown less rapidly than at the time of the two previous trials. The stallcs were about 10 inches long, with close heads. The lot was divided into six bunches, A, B, C, D, E and F. As usual, A was prepared for the drying oven at once. The other five bunches were stood upright in a tin box with a tight cover and with no water in it. The box with its contents was placed in the refrigerator. The two previous experiments had shown that the asparagus stalks would become Ump even when on the ice, unless their butts were in water. The tight box was chosen in order to reduce the evaporation to the lowest point by keeping the stalks in a close atmosphere. Tliis atmosphere was soon saturated with moisture by the exhalations from the stalks, but there was no water for imbibition. The imbibed water promotes chemical activity, and the stalks with butts in water, while remaining firm and crisp, actually lose dry matter more rapidly than those held out of water, which become limp, as shown by B and E when compared with C and D in Tables VI. and VII. One bunch at a time was removed from the box, at intervals of two to four days. June 19, four days after cutting, B was taken out. Stalks were firm and crisp, apparently as fresh as when placed in the box. Drops of moisture appeared on the walls of the box and on the stalks. The stalks were wiped dry with a towel and then weighed. After being weighed the stalks were broken and spread on a tray and dried in the oven as usual. June 21, six days after cutting, C was removed. Stalks were apparently as sound and fresh as B. Subsequent treatment was as usual. June 23, eight days after cutting, D was removed. The stalks in this bunch were slightly limp, but not as limp as bunches kept on ice for a day or two in the circulating atmosphere of the refrigerator. The bunch was treated as usual. June 26, eleven days after cutting, E was removed. The stalks were firm and plump, but this may have been due to imbibition of water KEEPING ASPARAGUS AFTER CUTTING. 303 through the butts, as there was now a positive accumulation of exhaled moisture on the bottom of the box. The refrigerator temperature held at 45= to 50° F. June 29, fourteen days after cutting, F was taken out. The stalks were firm and crisp. The butts looked dry and old on their surfaces; but if freshly trimmed by cutting off one-fourth of an inch of their length, the bunch would have passed for freshly cut asparagus. Much moisture had accumulated on the bottom of the box. The stalks were prepared for drjdng in the usual manner. The usual determinations of dry matter and total sugar were made in the dried material. Table VIII. Weight Fresh (Grams). Weight after Keeping (Grams). Dry Matter from Fresh Weight (Per Cent.). Total Sugars from Dry Matter (Per Cent.). C, D, F, 535 474 512 453 578 531 470 504 444 570 6.76 6.73 6.79 6.64 6.25 6.00 19.39 20.65 13.79 There was one unaccountable discrepancy in this series, — A had a lower sugar content then B, C or D, There may have been some con- dition during the first hours of drying this sample which favored the transformation of sugar into some of the lignified matter, but that is mere conjecture. Ordinarily, a lowering in sugar has been accompanied by a pronounced lessening of dry matter, which did not appear in this instance. Experiment 6. — This experiment was begun June 19, 1916. The stalks were a poor average lot, some having grown too tall and having heads much opened, but a portion of the stalks were in excellent form for market. The lot was divided into four bunches of as uniform quaUty and size as could be estimated. Bunch A was immediately prepared for drying in the accustomed man- ner. The other three bunches were set upright in the tin box with those of Experiment 5, and none of them was removed mtil July 5, sixteen days after cutting. As a whole, these three bunches were in poor condi- tion when taken out. Some of the tips were attacked by a white mold and some of the butts were soft with decay. Some stalks were shriveled throughout their length. The stalks were wiped dry with towels and weighed. Then all stalks showing signs of decay or mold were rejected from further study, and the remainder was sorted into firm and shrunken lots. Of the original lot of stalks, 34 per cent, was rejected, 35 per cent. 304 -MASS. EXPERIMENT STATION BULLETIN 172. was firm and crisp in appearance, and the remaining 31 per cent, was more or less shrunken or withered. These latter two lots of stalks were prepared for analysis in the cus- tomary manner, and dry matter and total sugar were determined. Table IX. Weight Fresh (Grams). Weight after Keeping (Grams). Dry Matter from Fresh Weight (Per Cent.). Total Sugars from Dry Matter (Per Cent.). A B,C, D Firm, Shrunk, 519 1.769 1,714 6.24 5.32 5.38 20.54 2.53 3.83 This lot of stalks proved quite inferior in dry matter to any of the other lots; but in total sugar, A was equal to any of the others of this season. To determine whether the less of sugars was the only destructive change in the dry matter, the losses of both sugars and dry matter were com- pared, as shown in Table X. It was noted that in all but two instances, namely, Experiment 3, C, and Experiment 4, E, the loss of sugar slightly exceeded the shrinkage in dry matter. This excess though small was persistent. Table X. Comparative Losses of Dry Matter and Sugars {Per Cent.). Experiment 1, A, B, C, Experiment 2, A, c'. Experiment 3, A, B, C, D. E. Experiment 4, A, B, C, D, E, Experiment 5, B, C E, F. Experiment 6, A, Firm, Shrunk, Dry Matter. Origi- nally. After Keeping. 6.81 7.28 6.65 6.49 7.18 7.31 7.34 7.20 6.25 6.00 6.32 5.38 Total SuQARa. Origi- nally. After Keeping. .71 1.02 1.17 1.27 1.31 1.29 1.15 l.OS KEEPING ASPARAGUS AFTER CUTTING. 305 The disappearance of the sugar is probably, in part, a transformation into the cellulose group of carbohydrates. This view was suggested by the work of Mrs. K. G. Bitting, who kindly allowed me to read the proof sheets of her bulletin on "Deterioration in Asparagus,"^ in which she has shown that asparagus tissues develop increasing areas of lignin when the stalks are kept for twenty-four hours or more after being cut from the crown. In order to elucidate further the character of the changes in the groups of constitutents in the asparagus, Mr. C. L. Beals determined the crude fiber and fat in the dry matter of the six samples described in experiments 1 and 2. The results are given in per cent, of dry matter and absolute weights calculated in the fresh stalks Table XI. Per Cent, in Dry Matter. A B C Fresh. Kept Warm. Kept CooL Experiment 1 : — Fiber 10.54 15.51 12.71 Fat 2.74 2.04 2.75 Experiment 2: — Fiber 10.99 17. 59 13.01 Fat .S5 2.29 2.94 Grams in Fresh Material. Experiment 1 : — Dry matter Fiber Gain Fat Loss Experiment 2: — Dry matter, . . Fiber Gain Fat Loss 7.64 .80 .21 7.68 .84 6.80 1.05 .25 .14 .07 6.81 1.19 .35 .15 .07 7.25 .92 .12 .20 .01 7.28 .95 .11 .21 .01 'K. G. Bitting: Bulletin 11, National Canners' Association, Washington, D. C, 1917, 18 pp. plates. 306 MASS. EXPERIMENT STATION BULLETIN 172. This series of determinations fully corroborated the increase of Ugnified tissue, as there was a positive gain in the absolute amounts of crude fiber or cellulose in the samples held for three days, which gain was more than twice as great in the warm room. At the same time there was a positive loss of the fatty extract in the warm room, but an almost negligible shrinkage in the refrigerator. The pronounced destruction of sugar by respiration and the increase of lignified tissue must affect the flavor and tender crispness of the young stalks, and these changes were much lessened by the lower temperatures. The development of fiber or cellulose at the expense of sugars and fatty matter is a logical consequence of the continued growth of asparagus stalks after they have been cut from the crown. The comparative amounts of this growiih at summer temperatures and the cooler ones of the refrigerator have been studied with interesting results. Freshly cut stalks of asparagus were divided into two lots, one of which was left in a warm room over night, or about ten hours, while the other was placed in the refrigerator for the same period. Both lots of stalks stood with butts in shallow water. The temperatures of room and refrigerator were noted at the beginning and end of the period, and as neither was opened during the time, it was assumed that the temperatures had remained within the limits noted. The increase in length of each stalk was carefully measured. The total number of stalks used in the different trials was 25. The average results for each trial are tabulated in Table XII. Table XII. Temperature (Degrees F.). Growth (Millimeters). Temperature (Degrees F.). Growth (Millimeters). June 2, 75-76 12.3 52-56 4.3 June 4 70-71 14.3 49-54 2.5 June 7 68-71 11.7 49-54 4.0 June 20 80 18.6 45 " The average rate of growth in the warm room was more than four times as fast as that in the refrigerator. At no time was the refrigerator cold enough to stop entirely the elongation of the tips, but at 45° F. it was nearly negligible. Summarizing the results of these varied experiments, it is clear that in Experiments 1 and 2, the changes in the warm room were fully double those in the refrigerator. In Experiment 3, the bunches in the warm room changed three times as fast as the bunch on ice. In Experiment 4, the bunches in the warm room changed more in one day than those in the refrigerator changed in four days. In Experiment 5, the asparagus changed very little in a week, when kept in a close atmosphere in the KEEPING ASPARAGUS AFTER CUTTING. 307 refrigerator. Experiment 6 showed that two weeks was too long a period to hold asparagus under the conditions. In conclusion, the experiments clearly show the possibility of holding asparagus for a week with very little deterioration in quality, by keeping the stalks at a low temperature and in a close atmosphere with httle air circulation. The temperature should be as low as 45° F. if possible, as this point is about the lowest limit for plant growth to take place, although respiration, or the destruction of sugar, will still persist. Experiments on a commercial scale have not been tried, but the feasible plan appears to be as follows : cool the asparagus as soon as possible after cutting. Lay the stalks loosely in boxes, place on ice in the icehouse and cover with canvas to maintain a low temperature and reduce the circula- tion of air. The common market boxes would probably allow any moisture exlialed and later condensed to drain off and not accumulate in the bottom of the box. Under this treatment the asparagus should not deteriorate appreciably in three or four days, when it may be bunched and trimmed to the proper length. By this treatment the market gluts occurring on account of Sundays and hoUdays, or hot waves, can be tided over with better prices and less waste. Any prolonged holding of asparagus in cold storage is a problem not yet studied. It presents a different set of conditions from those of most other vegetables or fruits. Fruits and most vegetables are matured storage organs of plants, and their structure and composition are adapted to preservation for a longer or shorter time. Asparagus, on the contrary, consists of the youngest stage of the plant at the period of most active growth. Its external and internal structures are adapted to rapid change in composition and devel- opment. The cell protoplasm persists in its activity at a reduced rate, while the delicate cuticle favors evaporation of the cell moisture and the attack of external molds. Hence, it is a difficult matter to arrest the changes and permanently hold the stalks in their pristine tenderness and flavor. It is not usually desirable to hold asparagus more than a few days to prevent market gluts. The usual methods of keeping asparagus at sum- mer temperatures cause rapid deterioration in quahty, and should be remedied if a discriminating patronage is desired. BULLETIN No. 173 MAY. 1917 MASSACHUSETTS AGRICILTIRAL EXPERIMENT STATION The Cost of Distribiting Milk in Six Cities and Towns in Massachusetts By ALEXANDER E. CANCE and RICHARD HAY FERGUSON Department of Agricultural Economics Co-operating with the Office of Markets, United States Depart- ment of Agriculture Worcester * Springfield Location of Cities and Towns Covered in this Investigation. Requests for bulletins should be addressed to the Agricultural Experiment Station, Amherst, Mass. IMassachusetts Agricultural Experiment Station. Trustees. OFFICERS AND STAFF. COMMITTEE. Charles H. Preston, Chairman, Wilfrid Wheeler, . Edmund Mortimer, Arthur G. Pollard, Harold L. Frost, . Hathorne. Concord. Grafton. Lowell. Arlington. The President of the College, ex The Director of the Station, ex o. fficio. Icio. STATION STAFF. Administration. William P. Brooks, Ph.D., Director. Joseph B. Lindsey, Ph.D., Vice-Director. Fred C. Kenney, Treasurer. Charles R. Green, B.Agr., Librarian. Mrs. Lucia G. Church, Clerk. Miss F. Ethel Felton, A.B., Clerk. Agricultural Economics. Alexander E. Cance, Ph.D., Agricultural Economist. S. H. DeVault, A.m., Graduate Assistant. Agriculture. William P. Brooks, Ph.D., Agriculturist. Henry J. Franklin, Ph.D., In Charge Cranberry Substation. Edwin F. Gaskill, B.Sc, Assistant Agriculturist. Robert L. Coffin, Assistant. Botany. A. Vincent Osmun, M.Sc, Botanist. George H. Chapman, Ph.D., Research Physiologist. Paul J. Anderson, Ph.D., Associate Plant Pathologist. Orton L. Clark, B.Sc, Assistant Plant Physiologist. W. S. Krout, M.A., Field Pathologist. Miss Grace B. Nutting, Ph.B., Curator. Miss Ellen L. Welch, A.B., Stenographer. Chemistry. J. B. LiNDSET, Ph.D., Chemist. Edward B. Holland, Ph.D., Associate Chemist in Charge {Research Section). Fred W. Morse, M.Sc, Research Chemist. Henri D. Haskins, B.Sc, Chemist in Charge (Fertilizer Section). Philip H. Smith, M.Sc, Chemist in Charge (Feed and Dairy Section). Lewell S. Walker, B.Sc, Assistant Chemist. Carleton p. Jones, M.Sc, Assistant Chemist. Carlos L. Beals, M.Sc, Assistant Chemist. James P. Buckley, Jr., Assistant Chemist. W. A. Allen, B.Sc, Assistant Chemist. J. B. Smith, B.Sc, Assistant Chemist. Robert S. Scull, B.Sc, Assistant Chemist. James T. Howard, Inspector. Harry L. Allen, Assistant in Laboratory. James R. Alcock, Assistant in Animal Nutrition. Miss Alice M. Howard, Clerk. Miss Rebecca L. Mellor, Clerk. Entomology. Henry T. Fernald, Ph.D., Entomologist. BuKTON N. Gates, Ph.D., Apiarist. Arthur I. Bourne, A.B., Assistant Entomologist. S. C. ViNAL, B.Sc, Graduate Assistant. Miss Bridie E. O'Donnell, Clerk. Horticulture. Frank A. Waugh, ' M.Sc, Horticulturist. Fred C. Sears, M.Sc, Pomologist. Jacob K. Shaw, Ph.D., Research Pomologist. Harold F. Tompson, B.Sc, Market Gardener, R. P. Armstrong, M.Sc, Graduate Assistant. Miss Eleanor Barker, Clerk. Meteorology. John E. Ostrander, A.M., C.E., Meteorologist. Microbiology. Charles E. Marshall, Ph.D., Microbiologist. F. H. Hesselink van Suchtelen, Ph.D., Research Mi- crobiologist. Poultry Husbandry. John C. Graham, B.Sc, Poultry Husbandman. Hubert D. Goodale, Ph.D., Research Biologist. j Lloyd L. Stewart, B.Sc, Graduate Assistant. I Miss Rachael G. Leslie, Clerk. Veterinary Science. James B. Paige, B.Sc, D.V.S., Veterinarian. G. Edward Gage, Ph.D., Research Pathologist. J. B. Lentz, V.M.D., Assistant. I On leave. CONTENTS. PAGE Foreword, ............ 1 Introduction, ........... 2 The problem 2 Co-operative investigation, ......... 3 Scope of the investigation, ......... 3 Processing costs and delivery costs, ........ 8 Difficulties in obtaining data, ......... 9 Analysis of costs, ........... 10 Investment, ............ 10 Depreciation problems, ......... 10 Maintenance, ........... 12 Working capital, .......... 12 Labor . 12 Costs of processing and delivering summarized, . . . . .13 Costs classified by size and kind of business, ...... 14 Investment and size of business, ........ 15 Percentage analysis of costs, ......... 20 Comparative costs by localities, ........ 22 Amherst v. Walpole, ......... 27 Haverhill v. Pittsfield, 29 Springfield v. Worcester, ......... 30 The producer as a distributor, 39 Cost of delivery of special milk, ........ 43 Cost of collection and distribution of wholesale milk in cans, ... 43 Motor truck delivery, ......... 44 Cost of distribution of cream, ......... 45 Significant facts of distribution, showing individual variations, ... 45 Some obvious disadvantages of competitive distribution of milk, . . 50 Suggestions for improving conditions, ....... 53 Publication of this Document approved by the Supervisor of Administration. BULLETIJSr :^o. 173. DEPARTMENT OF AGRICULTURAL ECONOMICS. THE COST OF DISTRIBUTING MILK IN SIX CITIES AND TOWNS OF MASSACHUSETTS/ BY ALEXANDER E. CANCE, PROFESSOR OF AGRICULTURAL ECONOMICS, AND RICHARD HAY FERGUSON, EXTENSION PROFESSOR OF AGRICULTURAL ECONOMICS, MASSACHUSETTS AGRICULTURAL COLLEGE, CO-OPERATING ■W^TH OFFICE OF MARKETS, UNITED STATES DEPARTMENT OF AGRI- CULTURE. Foreword. The facts presented in this bulletin show that the cost of distributing retail milk by more than SO distributor^, some of them producers, some of them dealers, was 2.64 cents a quart in 1914 and 1915. It cost 42 distributors in Worcester and Springfield 2.79 cents a quart on the average. These costs included (1) all labor costs — labor Hired, labor of the members of the family, labor of the operator and proprietor in preparing the milk for dehvery, and delivering it (labor made up more than half of the total cost) ; (2) all depreciation or replacement costs on all buildings, equipment and horses used in preparation or deliverj^; (3) all maintenance charges, or cost of upkeep of plant and equipment — repairs, oil, bottles, etc.; (4) aU overhead or fixed charges and all supplies used but once — rent, interest, taxes, insurance, license, soap, caps, hght, fuel, stationery, bad debts, spoilage, etc. The charges made were adequate and the figures obtained mean that, according to the accounts and statements of 85 dis- tributors, the average milkman in 1914 and 1915 was able to pay himself wages and interest and account for all expenses and losses when he received from his retail customers 2.64 cents more than he paid for a quart of milk delivered at his plant; or 2.79 cents if he lived in Springfield or Worcester. ' Practically all of the data for this bulletin were personally collected by the late Professor Richard Hay Ferguson, who was responsible also for most of the tabulations and for much of the bulletin in its present form. Mr. Ferguson died Dec. 1, 1915. This bulletin was his last work. Z MASS. EXPERIMENT STATION BULLETIN 173. Prices have risen since 1915. Labor and supplies of all kinds are higher. Just how much the increase has been cannot be stated with accuracy. Retail food prices have advanced nearlj^ 30 per cent. Perhaps 25 per cent, will fully cover the advance in milk-distributing costs. Assuming the increase to be 25 per cent, the cost of retaihng milk in the fall of 1916 would probably average 3.30 cents per quart for all distributors here cited and 3.49 cents per quart for the milkmen investigated in Springfield and Worcester. The authors will not, however, vouch for these figures. Actual present costs may be Mgh'fer or lower than 3.30 cents or 3.49 cents. Introduction. It is well known that for a number of years the price of milk to the consumer has been increasing. Not long ago milk was retailed at 6 cents a quart, whereas to-day the price is 9, 10 and, in many instances, 11 cents. Producers complain that notwithstanding the increased price paid by consumers they are, at the prices paid to them, producing milk at a loss and unless some change is made whereby they can get a fair return for Location of Cities and Towns Covered in this Investigation. their product, the whole dairy industry in Massachusetts is doomed. On the other hand the consumers view with alarm the increase in price and cannot understand why they must pay 10 cents a quart for milk when the producer is receiving but 4^ to 5^ cents net. The Problem. The milk question has many phases and many relations. Some of these have been indicated in the very enlightening bulletin on the milk situation in New England, issued in June, 1915, by the Boston Chamber of Commerce. The Massachusetts Agricultural College, in its outline of the problem, has recognized three important lines of study and investigation: COST OF DISTRIBUTING MILK. 6 1. The cost and methods of production. 2. Collection and primary transportation of milk and cream. 3. Methods and costs of distributing; i.e., preparing for delivery and delivering milk and cream. Closely related with all three is the problem of milk inspection. Problems 1 and 2 are quite as important as No. 3, the cost of distri- bution, but this prehminary study deals mainly with distribution and incidentally with transportation. Several studies have been made of the cost of producing milk in the North Atlantic States but, in the authors' opinion, none of these deal -ttith the problem of milk production on the typical dairy farms of New England in a detailed and thoroughgoing way over a sufficiently long period.^ Comparatively little serious work has been done on the methods and cost of transporting milk. Co-operative Investigation. The Department of Agricultural Economics of the Massachusetts Agri- cultural College and the Office of Markets of the United States Depart- ment of Agriculture formulated a plan for making an accurate, first-hand study of milk distribution in a number of Massachusetts cities and towns, perhaps the first study of its kind ever organized. The data used in this study were collected by agents of the Department of Agricultural Economics and the Office of Markets during the fall of 1914 and the winter of 1915. Altogether, rather accurate data were obtained from 85 distributors of milk, each of whom was visited from one to several times in order to obtain as reliable figiu^es as possible. Several of the tabulations were made by the Federal Office of Markets, where all the figures were checked. Scope of the Investigation. Recognizing the fact that the cost of distribution may vary according to the size and location of a towoi or city, as well as with the size and method of doing business, it was decided to investigate three groups of towns. Amherst and Walpole, each having a population approximating 5,000, — the former a college town in the Connecticut valley and the latter an industrial center jn the southeastern part of the State, — were chosen as typifying small town conditions in different parts of Massachusetts. Both Amherst and Walpole draw their supply of milk from the immediate ' Harwood, P. M.: What it costs to produce Milk in New England. Mass. State Bd. of Agr. Cir. No. 9. Boston, Mass., 1914. Hopper, H. A., and Robertson, F. E.: The Cost of Milk Pro- duction. Cornell University in co-operation with Jefferson County Farm Bureau, Bui. No. 357. Ithaca, N. Y., 1915. Lindsey, J. B.: Record of the Station Dairy Herd and the Cost of Milk Production. Mass. Agr. Exp. Sta. Bui. No. 145. Amherst, Mass., 1913. Rasmussen, Fred: Cost of Milk Production. New Hampshire Coll. and Exp. Sta. Exp. Bui. No. 2. Durham, N. H., 1913. Thompson, A. L.: Cost of producing Milk on 174 Farms in Delaware County, N. Y. Cornell Univ. Bui. No. 364. Ithaca, N. Y., 1915. Trueman, J. M.: Records of a Dairy Herd for Five Years. Storrs Agr. Exp. Sta. Bui. No. 73. Storrs, Conn., 1912. 4 MASS. EXPERIMENT STATION BULLETIN 173. neighborhood. The greater portion of Amherst's milk is distributed by dealers, while that of Walpole is marketed by the producers themselves. Haverhill and Pittsfield, industrial centers of approximately 30,000 population each — the former in the northeastern part of the State, in the midst of good dairy farms which supply the requirements of the city, and the latter in the heart of the Berkshires in western Massachusetts surrounded mainly by the homes of summer residents and drawing its milk supply from a greater distance — form the second group. Springfield and Worcester, commercial and manufacturing cities of over 100,000 population, constitute the third group, the one located in the Con- necticut valley, where the land is given over chiefly to the raising of tobacco, onions and other intensive crops, while the other is situated in the center of Massachusetts' best dairying county. Naturally, in Worces- ter and Haverhill a rather large portion of the milk is distributed by the producers themselves. In some cases the producers distribute not only the product of their own dairies but also that of neighboring farmers, thus in a measure becoming middlemen. Table I. — Firms interviewed, classified by Location and Quantity of Retail and Wholesale Milk, Cream and Skim Milk handled daily. Place. 5 1 1 1 -2 J u i 1 1 1 1 < 003 Amherst 5 3 2 - - - Walpole, . 5 3 2 - - - Haverhill, . 22 4 8 1 - 2 Pittsfield, , . 12 3 3 3 - 1 - Worcester, . 31 4 10 10 3 2 1 1 Springfield, 11 3 2 3 1 - - Totals, . 86 20 27 21 10 3 2 3 Per cent, of number, 100 23 31 24 12 3,5 3 3.5 Routes, 170 22 38 42 38 25 2 3 In each locality sufiicient typical distributors were interviewed to insure the reliability of the figures and the representative nature of the facts. The distributors interviewed and the volume of business represented were as follows : — COST OF DISTRIBUTING MILK. Place. Distributors interviewed. Quarts of Milk and Cream distributed daily. Total Number of Distributors in Locality. Total Quarts daily Distribution Estimates. Amherst Walpole Pittsfield Haverhill Worcester Springfield, 5 5 12 22 31 11 1,320 1,409 7,690 10,828 22,809 10,149 5 5 46 40 167 110 20,000 75,000 65,000 In Amherst, Walpole, Haverliill and Pittsfield about 60 per cent, of the total milk distributed is represented. In Springfield there are approxi- mately 65,000 quarts distributed daily, and in Worcester 75,000. The figures presented include approximately 16 per cent, of the Springfield distribution and 30 per cent, of that in Worcester. Some idea of the size of the milk business in Worcester and Springfield and the number and character of distributors may be gained from Tables II. and III. These figures were obtained in April and September, 1916. It is interesting that Springfield is supplied from 694 sources, the milk passing through the hands of 608 distributors handling a daily average of 27 gallons each. The average milkman in Springfield sells 118 gallons of milk and cream daily; in Worcester, 107 gallons. Table II. — Springfield, Sources, Quantities and Methods of securing City Milk and Cream Supply. Sources of SuPx'ly. Number. Approximate Daily Quantities. Number of City Milk (Gallons). Cream (Gallons). Dealers supplied directly. Producers hauling to city, . . Individual producers shipping to city, . Country creameries and milk stations, . Farmers' stations, Totals 15 650 24 5 1,025 14,480 25 500 - 694 15,505 525 560 MASS. EXPERIMENT STATION BULLETIN 173. -sS S 2 § K S g s S o o ^ «i <» «-. eo 0=3 -- H- a s ' s i s s n ^ u d a 1 a •s . 1 ua^ 1 >n III lO ' t- "^ m Is te 1 u 1 1 i-sS O 1 . 1 |s ' i ' 1 i 1 g e s O" 'Z „ 1 < 11 g g ' ' ' ' i g ^ « T3 < ll ' ' 11^ 1 1 i (2 1 &^ si ill g 1 «l to 1^ 2 • 52 Q :S l-l »2 < < 1 * 2 s 1 =• i i "3 1 =» g s ^ to g 1 n 1 IS i o" ' i - s 1 1 i " § o 11 g SS CO " - i § Q z 1 MASS. EXPERIMENT STATION BULLETIN 173. Processing Costs and Delivery Costs. The costs of distributing milk fall naturally into two classes — prepa- ration for deUvery or processing, and delivery to customers. The trans- portation of milk from the producer to the dealer is an additional item of expense, but usually the producer delivers his milk to the dealer. In this study the transportation cost has not been considered. The analysis of costs begins with the preparation of the milk for delivery and ends with the collection of money from customers. Simple as this analysis appears, a number of items cannot well be When the consumer pays 9 cents. charged exclusively either to preparation or to deliver}^ — administration and clerical expenses, hght, telephone, etc.; insurance and taxes, perhaps; slirinkage, spoilage and bad debts. In the summary of costs these have been called "overhead" expenses; usually they might well be distributed betv/een processing and delivery. From the standpoint of health, pure, clean milk is as necessarj^ as a good water supply. Milk just drawn from a healthy cow under sanitary con- ditions is at its best, and could it reach the consumer in this condition it would be ideal. To preserve it and to overcome the bad effects of un- healthy stock, unsanitary methods and conditions in the barn and reduce to a minimum the unavoidable deterioration in handling, transit and storage, milk has to be "prepared" for the customer. This preparation may be called "processing," and, so far as the distributors interviewed COST OF DISTRIBUTING MILK. 9 were concerned, consists cliiefly in cooling the milk and bottling, i.e., washing, filling and capping the bottles. Milk is almost universally deliv- ered to the consumer in bottles; in fact, only one instance of dipped milk was discovered; this was in Worcester. In addition to this, however, some of the larger dealers clarify their milk by running it through a macWne which removes the visible dirt, or pas- teurize it to retard bacterial development. Tliis materially adds to the cost of processing. Tables II and III show that only a minor percentage of the milk distributed in Springfield is pasteurized. In Haverliill, of 20 distributors visited, but 2 had pasteurizers. In Springfield 16 were visited and but 1 had a pasteurizer and clarifier. In Worcester 35 were visited; 2 had pasteurizers and 2 others possessed clarifiers. Some few distributors produced milk under unusually good sani- tary conditions, almost always keeping the bacterial count much lower than in ordinary milk. This they called "special" milk, and maintained that processing other than cooling and bottling was unnecessary and that pasteurizing was more likely to prove harmful than helpful to their trade. Under ordinary conditions the investment in processing macMnery was very small indeed, and the labor involved in caring for the milk was con- fined to the most ordinary precautions to prevent souring. Difficulties in obtaining Data. Many difficulties were met in securing the necessary data to determine the cost of distribution. Very few producers or dealers kept proper books; in fact, any sort of bookkeeping was the exception rather than the rule. Complications also arose when the producer distributed the milk, for it was difficult to separate the items of production and distribution, the stable, shed, horse and harness being used for both. In many cases, there- fore, estimates only could be given, but great care was taken that such estimates should cover the actual cost. The figures quoted are fairly accurate, and those on the cost of distribution of "special" milk can be relied upon in every detail, since most fortunately these distributors have kept accurate records for a period of several years. Mixed Business. — The greatest problem, however, that confronted the investigators arose from the fact that in almost all cases the distributors not only deliver bottled milk directly to the individual consumer, but deliver wholesale milk both in bottles and in cases to other retailers and restaurants and also deliver cream both wholesale and retail. By good fortune figures were obtained from a dealer who kept accurate cost accounts and dealt entirely in wholesale milk. His accounts show that it cost liim three-quarters of a cent (S0.0076) per quart to collect his milk from pro- ducers and distribute it in wholesale quantities. This figure is not appli- cable in most instances, however, for the reason that ordinarily the dis- tributor does not go out of his way to deliver his M^holesale milk; that is to say, his route is no longer and his apparent costs vary but httle, whether he delivers retail milk only or adds a few wholesale deliveries. Careful 10 MASS. EXPERIMENT STATION BULLETIN 173. thought indicates that an allowance of one-half cent per quart for whole- sale milk dehvered by a retail dealer covers the cost of this service in most instances; consequently this figure has been uniformly used. This method of accounting, which very evidently lays the burden of costs on the retailed milk and rather arbitrarily estabUshes the costs of incidental wholesale distribution, is presented with full recognition of its weakness and its limitations. It does not mean that wholesale milk can be delivered at this cost, nor that a mixed business should not be considered on its merits; but it is manifestly unfair to assume that it costs as much to dehver 200 quarts at wholesale to two customers as to deliver 200 quarts at retail to 200 customers; and, since three-fourths of the quantity is retailed and nine-tenths of the equipment is for retailing, the arbitrary figures given are very reasonable interpretations of the facts. The same question arises as to the delivery of retail cream. Based somewhat on the cost of dehvering retail milk and estimating filling, cap- ping, boxing and icing, loss of bottles and other contingent expenses, a charge of 3 cents per quart is deducted for its distribution. These deduc- tions may be open to criticism but they were reached after making full investigations and obtaining the opinions of many distributors. Analysis of Costs. Cost data may be grouped under comparatively few heads: — 1. Investment in land, buildings, horses and all equipment that is more or less permanent in its nature. 2. Depreciation on buildings and equipment. 3. Maintenance of plant and equipment. 4. Circulating capital, i.e., current operating supplies used but once — fuel, soap, ice, etc.; and " overhead," i.e., fixed charges, rent, insurance, taxes, etc. 5. Labor. As previously noted these items may be assigned to processing, delivery and overhead or to processing and delivery. Investment. Investment includes the inventory value of real estate, horses and equipments used in the processing or delivery of mUk and the housing of the horses and equipment. Depreciation was reckoned on all items of investment and was charged for one year. Some specific problems may be mentioned. Depreciation Problems. Horses. — No hard and fast rule was followed in determining the depre- ciation of horses. It was asserted by many that a horse worth $300 after giving ten years' service could be sold for $100; after five years' service, for $200, thus giving an annual depreciation in each case of $20. Some distributors affirmed that no depreciation of horse flesh could honestly be COST OF DISTRIBUTING MILK. 11 charged, since they usually disposed of their horses after three or four years for more than they cost. Other animals eighteen and twenty years of age were giving good service. Rate of Depreciation. — For these reasons each individual case was dealt with on its merits under this general formula: first cost of animal, less the selling price or the present worth, divided by number of years of service equals the annual depreciation. This method of calculation takes no account of losses by death; only horses now in service are considered. Where such losses had occurred in recent years some allowance was made, however. The figures obtained show that the depreciation of horse flesh increased in proportion to the size of the town or city, and also of the load hauled. In Amherst and Walpole annual horse depreciation averaged 7.5 per cent. In Worcester the average was 9.5 per cent. Buildings. — To compute the investment in buildings and the necessary allowance for depreciation was also a source of some difficulty. Ln Walpole and Worcester a number of dairies were housed in basements, some in basements of residences. Moreover, the majority of the country dairies visited are in the barn, stable or shed, a partitioned space in these buildings being all that is considered necessary for the plant. In all these instances an estimate was made of the value of the whole building; this was multi- plied by the fractional space occupied by the milk plant and to this was added the outlay for fitting up the plant itself. When the valuation was arrived at, 3 per cent., as a rule, was charged off for depreciation; 2 per cent, for taxes and insurance; and 5 per cent, for interest. This may be a trifle high, but in some cases the actual charges for taxes and insurance were more than 2 per cent. Equipment. — The equipment varied exceedingly, but without exception fairly reliable data were obtained. No arbitrary rule was followed in com- puting the depreciation, since each individual item has a different period of service and these periods vary with the different plants and users. Many distributors had experience sufficient to enable the investigator to arrive at a fairly exact figure; in other plants estimates were necessary. In a number of cases the equipment was very meager and the methods employed crude; filling bottles by hand, heating water over a smaU gas burner, and washing bottles by hand were not unusual. Except in the case of the large dealers in the cities and a few of the more progressive pro- ducers who distribute, live steam was not used for washing or sterilizing and in several cases the heating apparatus was entirely inadequate. Harness. — The almost unanimous opinion was that the Life of a set of harness costing from $35 to $40 is five years, provided it is kept in good repair; the repairs usually amount to $5 a year. This bears out the statement of harness makers that harness costs $1 a month. Wagons and Sleighs. — There was very little difference of opinion re- garding the upkeep and Life of wagons and pungs. The price of wagons ranged from $175 to $275, with a Ufe of approximately eight years. They are usually varnished every year and painted and overhauled every alter- 12 MASS. EXPERIMENT STATION BULLETIN 173. nate year. Pungs or sleighs cost an average of $50 and last about fifteen years, very little being spent on upkeep. Other Equipment. — Boxes worth 80 cents to $1.25 are good for five years. There is a difference of opinion as to the relative merits of the wooden and the steel boxes. Five complete sets of cans are necessary for the average dealer, one set being replaced each year. This item, how- ever, should be charged to transportation except in the case of the deUvery of wholesale milk. Mamtenance. Maintenance includes the expenditure necessary for the repair and upkeep of the buildings and equipment, including feed of horses and the loss of bottles and cans. In general, the outlay necessary to maintain the plant in working order is maintenance. Such items as grease and oil, veterinary service, shoeing, stable sundries, brushes, brooms, blankets, feed bags, carriers, hose, medicine, paint and other sundries required to keep up the buildings and equipment fall under this head. Working Capital. Working capital (or overhead and current supplies) includes such items as soap, ice, light, fuel, stationery, telephone, rent, insurance, taxes, interest on investment, spoilage, surplus, shrinkage and bad bills. It was difficult in many cases to separate these items, spoilage and surplus being included by some in shrinkage and by others in bad bills; fuel was con- sumed for other purposes than the dairy; the telephone included private use; and insurance, taxes and water rates often covered the residence or buildings used for other purposes in addition to the dairy. Assessed values and tax rates vary greatly, but in general 2 per cent, of the actual value was allocated to taxes and insurance. Insurance averaged about 1^ per cent, for three years. Interest was uniformly computed at 5 per cent, on the entire investment. Labor. Labor is classified as hired, home and personal. Home labor is labor provided by members of the family, such as assistance in the dairy or on the milk wagon, but more often in keeping the books. Usually home labor does not represent an expenditure, but is charged at the prevailing rates. Personal labor is the labor of the proprietor himself and is valued at his own estimate, never less than 25 cents per hour. In no case was the accepted estimate considered excessive or below a reasonable remun- eration. There is much individual variation in each of these items, especially among the producers who board the hired help. The wages paid varied from $25 to $35 per month and board; the estimates for board vary from $15 to $30 per month. The time, too, must often be distributed more or less unequally and arbitrarily between farm work and the preparation and delivery of milk. In all instances these adjustments were made carefully, but except as averages they cannot be considered in all respects infallible. Table IV. — Swnmary of Total Costs, and Cost per 1,000 Quarts, of distributing Milk and Cream {Forty-two Plants). V.,.n.OK- TOTil.. Ho.e=. 1 S. Ho'u^^es. Stables. BoUers. Pumps. Tanis. Washers. Fillers. lee Chests. Wagons. Pungs. Bojes. cans. Office. — ^QL,r Amount. ""izr'-' . . . Per 1,000 quarlB. . . .37,013 00 ,20.00.00 ,7,320 00 ,24,333 34 ,7,743 92 .1.560.00 .220 00 M,7S6 00 K,508 00 M,310 00 ,1,240 00 «10_00 K!,523 50 ,6,moo «..S7100 ,0,78100 «,907 55 ,2,629 70 ,1.340 00 W45 00 ,n"l5 ,170.151 01 DepreciaUon:- 13,971 71 $780 06 .359 60 .730 00 .764 58 .150 00 .16 00 M8100 M09 73 M07 33 ,118 00 Ml 00 .203 73 ,1.372 93 .3,787 85 .397 56 .676 82 ,376 29 03 ,132 00 ,128 4"58 383 12'77 Per 1,000 quarts. . . - Repairs. Sundries. Shoeing. Feed. Carriers. Bottles. Cans. I 1 1 ! "Siir.lL,- : je,896 31 „., H103 80 ,31,773 g 02 •-S ... 54,897 41 Rent. Soap. Caps. Ice. LigMand Fuel. Stationery. '1£- Spcjla^e Shrinkage. Bad Bills. Sundries. Cipculating capital: — «,334 00 .1.1.07 .2.519 19 ,7,376 45 .1,578 30 ,3.872 60 .1,93.45 -■"IS K,328 21 ,2.938 70 .6,560 52 ,3,227 74 "-Sr K!2 44 ^msMso^ Retail: — Daily (quarU) 24.421.70 Wbo*S^e-^-^^'' * 8,913.025.00 (^^^'■JyJquarU) 2.890.595.00 Yearly (quarts) 222,344 00 Total yearly oo8t of retail distribution '. . $248809 39 Cost per quart retail distribution (cents), '. ' ' 2 79 « p. ^ -2 2 g COST OF DISTRIBUTING MILK. 13 Costs of Processing and Delivering summarized. Table IV is an itemized summary of costs tabulated for 42 plants in Springfield and Worcester. Facts obtained in these cities are fairly com- parable and the conclusions are quite as satisfactory as if the data for all six locaHties were included in the tabulations. The summary represents an annual business of approximately 9,000,000 quarts of retail milk, 3,000,000 quarts of wholesale milk and 222,000 quarts of cream out of a total distribution of about 15,000,000 quarts of retail milk, 4,700,000 quarts of wholesale milk and 300,000 quarts of cream — or about 60 per cent, of the total deliveries considered in this investigation. The milk of these 42 distributors was delivered to about 21,000 customers. The total investment in plants and equipments amounts to about 1| cents per quart of milk delivered. The largest investment items are milk sheds, horses and stables; boilers and ice houses come next but are com- paratively insignificant. The chief items of depreciation apply to horses, wagons and harness. These account for three-fifths of the total depreciation; another fifth is assigned to milk shed, stable, boxes, cans and boiler. By ascertaining the first cost, the present value and the time used, most of the items of depre- ciation are easily calculated. Nearly $55,000 is classified under maintenance. More than tliree-fifths of this is for horse feed and just about 80 per cent, is for feed, repairs and horseshoeing. Lost bottles and cans are classified as maintenance and make up most of the remainder. Circulating or working capital is here used to include overhead and fixed charges and supplies which are destroyed in one using. The largest item is interest on the investment, computed at 5 per cent.; the second is ice; and the third is bad bills. These items, with rent, insurance and taxes, fuel and loss by spoilage and shrinkage, account for 75 per cent, of this charge. Other items are soap, caps, stationery, light and oil. Labor of all kinds is by far the largest item, amounting to nearly three-fifths of the entire cost, or one and tlaree-fifths cents per quart of milk retailed. The average cost of processing and retailing milk is 2.79 cents per quart for an average daily delivery of 175 quarts of retailed milk per horse the year round. This cost is arrived at by deducting from the total expenses one-half cent a quart for the wholesale milk distributed and 3 cents a quart for retail cream. 14 MASS. EXPERIMENT STATION BULLETIN 173. Table V. — Cost per^Quart and Percentage of Total Cost for Deprecia- tion, Maintenance, Circulating Capital and Labor. Cost per Quart (Cents). Percentage. Depreciation, .16 5.69 Maintenance .57 20.34 Circulating capital, .48 17.06 Labor L58 2.79 56.91 100.00 Preparation .758 27.19 Delivery, 1.528 55.14 Overhead, .492 2.79 17.67 100.00 Costs classified by Size and Kind of Business. Perhaps a better analysis of 80 plants is presented in Table VI. In this analysis an attempt has been made to classify the distributors by size of business and to set forth the items of cost under processing, delivery and overhead. Only three plants do a business exceeding 2,000 quarts daily, hence the figures for these must be used with caution. Sixty plants do a mixed business, about three-fourths retail and one-fourth wholesale. Twenty plants deUver retail milk only. None of the all-retail plants do a daily business of 500 quarts. They are of one and two wagon capacity and so far as size of business is concerned should be classified with the "under 500" group. The actual per quart costs, wliich include both wholesale and retail milk, run from about 1.6 to 2.9 cents per quart. The discrepancy between per quart costs given in Tables IV, V and VI is accounted for by the fact that in Table IV only 42 firms are considered and the cost of distributing all wholesale milk is computed at one-half cent per quart. Plants of 500 to 1,000 quarts capacity do business most economically — 1.64 cents a quart for all milk delivered and 2.05 cents per quart for milk retailed. These costs are 25 per cent, and 22 per cent., respectively, below the average of all the plants investigated (2.21 cents for all deliveries and 2.64 cents for retailed milk). Plants of 1,000 to 2,000 quarts dis- tribute for 1.82 and 2.23 cents per quart. The 27 plants of less than 500 quarts daily capacity average 2.04 and 2.66 cents a quart. The 3 plants doing a mixed business of more than 2,000 quarts daily and the 20 exclusively retail plants show the highest per quart costs for retailing — 2.92 and 2.93 cents for all expenses. COST OF DISTRIBUTING MILK. 15 The overhead expense is the smallest item and in reality should be dis- tributed between processing and deUvery. It varies from 12.3 to 18.9 per cent, of the total cost in mixed, and 14.7 per cent, in retail business. This item seems to vary directly ■with the size of the business, i.e., with the quantity handled. The processing expense runs from 24.7 to 31.8 per cent, of the total. In general this expense varies inversely with, the quan- tity handled. Delivery costs a little more than one-half of the total, running rather uniformly around 55 per cent. The 1,000 to 2,000 quart group averaged 57.7 per cent, for deliver}^ but the individual variations are wide. On the whole the figures show comparatively little correlation between costs and size of business. Investment and Size of Business. The relation between size of business and average total amount invested in plant and equipment is of interest. The tabulations in Table VII., as might be expected, show a consistent correlation between investment and size of business. But when the investment per 1,000 quarts of milk dis- tributed is considered, tliis consistent correlation is not shown. The strik- ingly high investment ($22.61 per 1,000 quarts) of the retail dealers is, perhaps, rather surprising when compared with an investment of $4.30 per 1,000 quarts in plants during a mixed business of the same size. Enter- prises of the second and fourth classes have also a very high investment ratio. One might suppose that a milk-distributing plant could increase its volume of business by corresponding increase in plant, but an increase from an average of 360 quarts per day to an average of 710 quarts a day seems to multiplj^ the total investment nearly six times, whereas men who do a retail business exclusively have four times the total investment of those who do a mixed business of the same size. 16 MASS. EXPERIMENT STATION BULLETIN 173. a. ►i? I^-H C5 1^ -^ -^ •"S C5 pS 't^ b' !=: ^^ c 1^ ;S t§ 1 -^ '2 ^ g s? ?i. o t« S ^^ ^ < O OJ-OX=0-Ht^0 GCO 1^ T," « O — -.=> W • f- !« "^ s ,. ^SS^SSS §s o ° ° 'sSSg5SS gjO o s ^" ■3 --a • i- • 1 s a ■s ^ > o o o o •B -3 o • § § .s . 1 >> 1 s • 1 ■3 >i e >'a . .1 . . . .IS- j: .2 0. 1 i. 1 ^ "Is VEKHBAD InVESTME irhead expenses: — nistration and clerici , telephone, statione: ance, taxes, license, kagc and spoilage. ? a ■ h 1^ 1 i 11 O Ovc dnii ight isun hrin ad a liiiii 1 1 <:,j;„a2ffi 1- o < H H I 18 MASS. EXPERIMENT STATION BULLETIN 173. Pt)lC£NTAQE5 Of TOTAL C05T5 P£1l QUAU BY ^izE OP Total 5u>3ine55 re lOOI- ^ooo 7.6 Over aooo 18.9 147 Delivery ^ Processing ^^ OverheadCH COST OF DISTRIBUTING MILK. 19 Actual Total Lxpense of Milk DI5TI11BUTI0N PER QUAUT 1001- ^000 .0032 0\/er 1000 00^7 All. Retai .00^3 )I6 \.6'f^ 1.8^f Z.i-Si^ ^.93^ ?roce5Sin3 ^^ Overhead \:z3 20 MASS. EXPERIMENT STATION BULLETIN 173. Table VII. — Percentages of Total Cost per Quart of Wholesale and Retail Milk {80 Plants), by Size or Character of Business. I Under 500 Quarts. II 500- 1,000 Quarts. III 1,001- 2,000 Quarts. IV Over 2,000 Quarts. V All Retail. Average. Number of establishments, 27 20 10 3 20 - Total cost SO 0204 SO 0164 SO 0182 SO 0249 SO 0293 SO 0218 Per cent., 100 100 100 100 100 100 Processing expense, . SO 0065 $0 0046 SO 0045 SO 0067 SO 0090 $0 0064 Per cent.. 31.8 28.1 24.7 26.9 30.7 29.3 Delivery expense. $0 0114 SO 0089 SO 0105 SO 0135 SO 0160 SO 01214 Per cent.. 55.9 54.2 57.7 54.2 54.6 55.7 Overhead expense. SO 0025 SO 0029 SO 0032 SO 0047 $0 0043 SO 00322 Per cent., 12.3 17.6 17.6 18.9 14.7 15.0 Investment: — Per plant $566 $3,325 S5,279 $20,594 S2,277 - Per 1,000 quarts milk sold, . 4 30 12 84 9 51 19 30 22 61 - Percentage Analysis of Costs. The cost analysis presented in Table VIII shows the importance of labor both in processing and delivery, although the percentual importance varies greatly with the size of the business. The labor item differs also in the major processes of distribution. The relative importance of the labor item in the fourth group is the striking feature — 70 per cent, of the processing expense as contrasted with a maximum of 59 per cent, and a minimum of 46J per cent, in the other groups. The labor factor in delivery costs is more uniform but even here the labor item in the fourth group reaches the maximum — 61.9 per cent. It is significant that the labor item in preparation is lowest in the third and the all-retail groups, although the third group shows an actual proc- essing cost of .45 cents, and the all-retail a cost of .90 cents per quart. The principal point of emphasis in the overhead analysis, aside from the notable variation in the importance of the various items, is the high percentage of shrinkage and spoilage in the "over 2,000" group. Bad accounts average more than one-eighth of the overhead and, curiously ertough, are percentually highest in Groups I and II, which show the lowest actual overhead. The interest item, of course, varies with the investment. Its percentual importance averages from about 9 per cent, in Group I, to 26.3 per cent, (three times as much) in the all-retail group. COST OF DISTRIBUTING MILK. 21 Table VIII. — Percentages of Costs in Relation to Size of B^isiness. Amounts handled and I terns of Expenses classified in Groups. Percentages accokdinq to Size or Kind of Business. I II ni IV V Number of quarts sold daily. . Under 500. 500-1.000. 1,001-2,000. Over 2,000. All Retail. Number of establishments. 27 20 10 3 20 Average per cent, quarts sold daily: — Wholesale Retail. . . . 28.4 71.6 26.1 73.9 23.6 76.4 17.6 82.4 100.0 Preparation expenses in per cent, of total. 31.8 28.1 24.7 26.9 30.7 Depreciation and maintenance, SuppUes Labor 8.1 33.0 58.9 8.6 34.7 56.6 14 3 39.1 46.6 15.6 14.2 70.2 18.9 34.6 46.5 Delivery expenses in per cent, of total. 55.9 54.2 57.7 54.2 54.6 Depreciation and maintenance, Supplies Labor 14.8 25.7 59.5 17.8 28.1 54.1 12.5 26.1 61.4 12.8 25.3 61.9 19.3 24.1 56.. Overhead expenses in per cent, of total. 12.3 17.6 17.6 18.9 14.7 Administrative and clerical salaries, Light, telephone, stationery. . Insurance, taxes, license. Shrinkage and spoilage. . Bad accounts Interest 49.8 13.8 4.6 7.0 16.1 8.7 43.0 6.4 5.1 8.1 15.5 21.9 6.0 12.0 5.8 13.0 15.0 28.6 6.5 12.5 19.7 12.3 20.4 37.8 9.2 6.6 8.8 11.3 26.3 Expenses in per cent, of receipts: — Preparation or processing. Delivery Overhead. Total expenses in per cent, of receipts. 7.9 13.9 2.9 24.7 5.8 11.1 3.4 20.3 5.0 11.6 3.3 19.9 8.4 17.0 5.7 31.1 9.4 16.7 4.3 30.4 22 MASS. EXPERIMENT STATION BULLETIN 173. The relation of costs to receipts is the really significant fact to the distributor. Costs run from a minimum of 19.9 per cent, to a maximum of 31.1 per cent, of total receipts. This means that the costs of the all- retail and "over 2,000" groups, for example, absorb 30 to 31 per cent, of the total receipts, a portion more than 50 per cent, greater than the part taken by the second and third groups. This percentage which the expenses bear to receipts may be called the operating ratio. It is lowest in Groups II and III and highest in Group IV. The lower the ratio the more economical the operation of the plant. The operating ratio in any business is very significant. In milk distri- bution 20 per cent, is probably a low ratio and 30 per cent, a high ratio, but much more accounting must be done to determine this. In all in- stances the more expensive distribution is due both to higher processing and higher delivery costs and, in the fourth and all-retail groups also to higher overhead expenses. Comparative Costs by Localities. Table IX presents comparative cost data by towns. In these figures no attempt has been made to separate costs into processing and delivery. All the firms operating in Amherst and Walpole are in the "500 quarts or under" class; all but three of the Haverhill and Pittsfield firms are dis- tributing less than 1,000 quarts per day; hence the firms interviewed doing a daily business of 1,000 quarts and more are almost all in Spring- field and Worcester. The data show plainly the greater cost per quart in the two larger cities, a cost which is seen in practically all items entering into distribution. Few conclusions of significance as regards variations by localities can be drawn from the figures giving total locality costs. COST OF DISTRIBUTING MILK. 23 ^ " 0 0 .-: § ^ cq ri 1 ^ ;o JO rt a> Qo ,0 i 0 CO i «, r, H "' 0 0 1 s S 2 eo s g S 5 M s t^ 0 00 i c3 •V s s '3 »» ^ CO « % " » £ a S5 § s s § g a a -»■ CO 0 0 T)< •* Q 1 '3 n «» e-1 3 g s ^. i 2 s S s ;:: S S s CO § 10 0 « C^J § s s 2 g g "3 S § 0 § S g 0 CO » CO r-. H g s CO g a c s § § « s s i 1 i 1 1 i 1 3 t^ t^ »» C^ •»< H g " H 3) s s § s s ^ H 2 ;3 1 s s ^. 1 Z ■3 U 00 J? CI 2 £5 m I- 0 0 0 00 0 r~ 0 ^ 'J- CO ^ 0 s 3 •z H 0 s § § s g s a i i i 1 1 i 1 > s . - g z ^ 0 H , . a 0 >< H 6 2 u ■3 0 J i 1 1 11 1 < 1 M E 24 MASS. EXPERIMENT STATION BULLETIN 173. Is S o ^ I u ^ ■o c^ O o oo s s. » s — S £ «^ eg —" B» (S K O < h-! S § 22 g i t^ § B ' >o S 1 a § lO lO eo t^ W> 05 1 s i i 5 i iS gl «r lO S S .£-• « M S s g § oo It = s 1 CO g S s N cT to" o" u 1 s S S S ^ g c s 1 g S ". 1 1 « o c «■» -1 CO 'S S z it o H K O ^ £ U ■c ^- s TS' •3 fc 1 < 1 03 2 1 1 1 11 COST OF DISTRIBUTING MILK. 25 5 O O I en *. « ^1 s- !5 5^ r^ d o|2 ' l ^ g s g s « OS s s S§ o 2 a s s? i S >o u 11 1 00 i i s i i 1 o ^ ^2 (N 13 o % s i K i S «5 ^ o o m K Sis5 t~ M ^ J ^g M co O 2; s s o o o t^ t-- o '^r » >>-2 s ^ s 2 § s «* -' lO 00 2 § •i ^ H c >• S O ■c ^ . s ^ ■p a - 1 i 1 -< ■;£ e. t^ ^ 11 26 MASS. EXPERIMENT STATION BULLETIN 173. U m •"= « ^ t>. m » M N •Sfe^a „ S^fS(§ ^ S 03 r- S ^ s? s E-S? Z « S d ■ g § ^ g g g » uj' W| ^ £^ "5 i« in g ^?^ 0 s s s 6 u^ ^ g % § § § § § § £H ^ ^ §s i° 1. s s § s s ^. m |g s 0 CO t- « r- 0 o — •V 10 ^S " " > W to 0) . t-. •* ^ % g § S ^« g O ■3 s ira CO g i ^ lo- s s 5 § 2 t^ gs s " o li i s i i i i H a;*' "5 0 0 0 0 0 *^ s Q s z ^ H « c H u -6 J -i I ,"? 1 1 :S a ft- % ^ COST OF DISTRIBUTING MILK. 27 The comparative analysis of costs, including both processing and delivery, of retailing milk by cities and towns is exhibited in Table X. Before comparing localities it may be well to note that by far the most important item is labor, which varies from one-half to more than two- thirds of the whole distributing cost. This includes onh'^ man labor, horse labor being carried in the other items. This expense is greatest in Springfield, where it amounts to nearly 2 cents a quart, and lowest in Haverhill, where it is scarcely more than 1 cent. Depreciation is the smallest charge, and runs about 6 per cent, of the total; actually it is lowest in Haverhill and highest in Springfield. Maintenance and circulating capital show great relative variation. Both are relatively and actually lowest in Amherst and actually highest in Worcester and Springfield. The two charges amount to .52 cents a quart in Amherst, .85 in Walpole, .88 in Pittsfield, .92 in Haverhill, 1.03 cents in Worcester and 1.04 cents in Springfield. In general these items increase with the size of the town. Amherst v. Walpole. Amherst seems to process and distribute its supply of milk more eco- nomically than Walpole, notwithstanding the labor bill is slightly higher. Omitting cream, our figures show in round numbers 500,000 quarts of wholesale and retail milk delivered yearly in Walpole and 471,000 in Amherst. On this basis, Walpole's labor costs SI 1.65 per 1,000 quarts, and Amherst's SI 1.87; for retailed milk the labor expense is $12.58 per 1,000 quarts in Walpole and S13.69 in Amherst. Hired help is a little cheaper and more plentiful in the eastern part of the State, though the personal labor in both towns was computed at 25 cents per hour. The time occu- pied in delivery is the same, though the average milk route in Walpole is 25 per cent, shorter. Walpole serves more customers per wagon, 180 to 143 for Amherst, but delivers less milk per customer. The dealers in Amherst, however, expend less for maintenance and working capital. The lower maintenance is due in part to the greater load per horse, the average retail load per horse being 175 quarts, in contrast with 143 quarts in Walpole. It must be noted, however, that Walpole hauls more per wagon — including wholesale milk and cream, 234 quarts to 214 for Amherst; the explanation is a two-horse wagon. In working capital there is a margin of .19 cents per quart (43 per cent, less) in favor of Amherst. Table X shows that these two items amount to nearly 40 per cent, of the total in Walpole as compared with less than 26 per cent, in Amherst. With the exception of the items stationery and shrinkage, the Amherst figures for circulating capital show a big saving. The greater stationery charge is accounted for by the use of tickets by several of the Amherst dealers. The wisdom of this expenditure is justified by the small loss in bottles and a minimum loss by bad debts. It cost the five Walpole dealers $340 a year for bottles, or 72 cents per 1,000 quarts of retail milk delivered. 28 MASS. EXPERIMENT STATION BULLETIN 173. 1&^ 1 o M =a Ki tn o o atf 00 o s i o o s g I" atf s i 03 O I i K .S COST OF DISTRIBUTING MILK. 29 Five Amherst dealers expend §140.69 for bottles, or 37 cents per 1,000 quarts of retail milk; this includes one dealer who does not use tickets. Eliminating this dealer for the sake of accurate comparison, the results may be presented in tabular form, as follows: — Number. Dis- tribute 1,000 Quarts. Expend for Bottles. Bad Debts. Dealers in — Total. Per 1,000 Quarts. Total. Per 1,000 Quarts. Walpole, Amherst using tickets, 5 4 470.5 346.7 S340 00 133 40 $0 72 38 S182 31 $0 40 09 It is significant that of $82 reported as lost through bad debts by Amherst distributors, $51 were reported by one dealer who did not use the ticket system. Comparing the figures of Amherst and Walpole dealers who do and who do not use tickets, it appears that where five Walpole dealers using no tickets suffer by bad debts a loss of 40 cents per 1,000 quarts of milk sold at retail, and one Amherst dealer loses similarly 62 cents, the four Amherst distributors using tickets have but 9 cents of bad debts for each 1,000 quarts retailed. Under the ticket system the cost of collection is somewhat less, but since the drivers do the collecting it is difficult to approximate this difTerence. Tickets, of course, mean cash in advance; just how long in advance depends on the price of milk, and the amount used per family, since tickets are usually sold in $1 strips. The price per quart is exactly the same, whether the customer buys tickets in advance or pays in currency when the milk is delivered. Ice cost Walpole dealers $1 per 1,000 quarts ($0,001 per quart) of milk, and the Amherst dealers SO cents per 1,000 quarts ($0.0008 per quart). Haverhill v. Pittsfield. The difference in the figures for these towns is not marked. Pittsfield expends a very little less per quart for maintenance and circulating capital, but this is more than offset by higher labor costs. Labor is comparatively expensive, due to the competition of the summer homes in the vicinity. Although Haverhill distributed milk at a lower cost per quart than any of the four cities, it was not at the expense of service, but rather as the result of the low labor cost coupled with the number of quarts delivered per horse, in other words, by getting the best service out of the horse. Haverhill averages 176.3 retail quarts per day per horse, while Pittsfield averages but 141.2 quarts per horse. Moreover, Pittsfield distributors deliver more cream and wholesale milk per route to a smaller number of customers than do Haverhill milkmen — about 100 quarts as against 75 for Haverhill. 30 MASS. EXPERIMENT STATION BULLETIN 173. It may be said in passing that the milk supplied by Haverhill dealers is exceptionally pure and clean. These qualities are popularly supposed to be expensive. If they are, HaverhUl dealers have met the increased cost by economies elsewhere. The city's entire supply comes from local pro- ducers. Thus any impure milk can be at once traced to the source of supply and the producer of exceptionally clean milk be quickly recognized. Frequent inspections and monthly tests by a competent bacteriologist are made. The methods of inspection and the publication of the results of the monthly bacterial analyses have educated the Haverhill public to appreciate the value of clean milk and have stimulated a healthy rivalry among the producers and distributors. Only one dealer uses a pasteurizer and he is the only distributor who purchases milk outside an 8-mile radius. Springfield v. Worcester. It costs the Springfield dealers studied 16 per cent, more than Worcester dealers to distribute retail milk; and 25 per cent, more than the average of all dealers investigated. Except in the amount spent for maintenance, all the costs of distribution are lower in Worcester than in Springfield. As a matter of fact, differences in depreciation, maintenance and overhead are negligible. The labor item alone requires attention. Worcester has cheaper labor because a large proportion of the distributors are producers, and farm labor at S50 a month (cost of board included) is much lower than labor in the city. In addition to this, a fair proportion of Worcester's milk supply is distributed by foreign-born dealers who value their services cheaply. A short time ago an ordinance was passed doing away with basement dairies in Springfield. This has been productive of much good, although it entails considerable expense. Depreciation has naturally increased in this city but without a corresponding increase in maintenance. COST OF DISTRIBUTING MILK. 31 S § g g i 1 1 2 2 s i S H S " «- g s § g g c ^ s ^ « O § g s g g o a «» 6 § g 8 g 8 ^ o o S S 1 i s s 8 g g t2 g S » CO K s , g g| s s o a© 9% 8 § 8 g g « fa " § § 8 g g 1 § S 12 ^ g & ;s -' " , § 8 1 i 1 e» S s 8 8 8 li i i i 1 § 8 8 8 ^i S R s^ «. . s g 8 8 8 1 "3 i i i g i ffi 1 M CO CO ^ o 1 Fi Z « PQ s D Z -IK K H 3 O ■ H 1 £ ro N 2 ?5 1 ?5 2 s « S S S s s fs a» 8 S S ^ CO s 8 s s 2 3 g i ?5 s g ?5 § fs g s 5? «o g 8 g g g ^ S -«< CO S o o " " g g g 8 t~ CO ■«< •4S g 8 g 8 g U5 o lO t^ « 9» ^M ^ 1 q ' 8 i ' ' 8 i 8 g 8 8 ' o CO 00 99 1 g g § § » 8 CO 8 8 5 i t^ ' g i « « 00 - o ■ H oT 2 2 ?5 32 MASS. EXPERIMENT STATION BULLETIN 173. 0^ . f3 § ? 3 i 6 S S ' ' f2 1 § S g s 1 § s g g 1 § g s s s s s § g « „ ^ g S S ^ 1 g g g g K S S M g •i S 5 i2 S g S ?3 2 1 1 o g g g g 1 1 § S g 1 i U 1 0 PL, S3 1 S S 2 s § o i s i i i H s § 8 S g g 8 o S "" « Cl CO 3 «e cc J2 8 8 g g g a i - § o i to S 9 rt -^■^S. 3 rt-S M ^ M "i S K s ^ O CI 00 o t^ CO (M -rf ^ « a 8 . s s § g 8 t. fi X § § o •* 3 c3,c3 c H «« M . b g 8 g g g is 2 =■ (3, Q to S§ 89 PJ '^ " ^ ^ g 8 g g o' 00 o o ' 00 3 fa ^ •" »o ^ § 8 lO 8 SI .S O CO o eq oo M e© c^ '-' ^ 3 g g g 8 g 1 § s § g i s g gs g g§ s. M S ^^ 2 rt a » s g 8 8 § d o in o § 55 " s m g , "¥" "¥ 8 QO ■«< (§ s t- N i «■ n S s 2: _aj 3 rt" ^ & . o H PL, o rf 2 S 11 COST OF DISTRIBUTING MILK. 33 1 1 S K a S z JO a ^1 3 H $1,790 20 2,536 85 1.578 63 B S W Q o M 2 gs t~ « CO 00 ™ S S t^ ^ to ill laia-:! 34 MASS. EXPERIMENT STATION BULLETIN 173. 8 K 8 sj K f^ 1 J5 i" ta ^ s 1 i o 8 8 O i i m 8 8 g g 8 o o U o >o o o U5 o s a ^ - s cu i i g g g 8 g 8 1 IS ^ S ^ ^ *- a; -- SS S 8 8 8 8 g 8 £ o g o S ^ K «» 8J 8 g 8 o o o 8 8 8 II o 1 ' 5-, ' s ^ i i: 8 8 8 g g 8 i s E m S 2 o ^ fe 8 g 8 8 8 o ' !5 5: s» ° ^■ g 8 8 ?^ g S o g o 8 CO ."S S O s ^g "^ ■* S 6 8 8 8 3 ' 1 § ' ' 1 U ^ d g g 8 8 ^1 1 N 8 1 ^^ 8 8 8 i| 1 1 S O 8 8 g g 8 g H a 2 1 i S. |i " . S gj g ■Z IS ■ 3 H Q CO 2 ?: S iS 1 $263 02 382 02 1,810 09 295 79 403 13 i $5 00 g $20 00 s i $10 00 14 85 131 25 16 00 13 75 g ^ $77 34 124 50 602 17 54 66 140 63 1 $28 00 30 00 100 00 42 50 58 75 en 87 00 10 00 g $4 00 53 33 ?? g $20 00 20 00 10 73 10 00 f2 $0 00 8 50 100 00 2 33 S3 o $15 00 22 50 26 67 17 82 45 00 $60 00 60 00 8 1 $9 00 200 00 1 50 8 o s $21 00 150 00 g $115 68 66 67 360 00 76 92 125 00 S5 "■"""'" S d f S g S 3 COST OF DISTRIBUTING MILK. 35 § ^ « C-i — " »o o t^ >n o CO o g s s s § g g § s s s s ;? g g § s g s C K ^ fS s i H ^ o. 8 o ■a 52 >o S ^ ss c " §^ i g 8 S 8 8 s a § fK s g g o T3 S «=. O t*> 2 li f^ g g R m ' ' S ^ Vl X. CC 1 S 5! 8 P^ § S i i 1 § g ^- i i/j g g - g g - PI >o cc en s ig t-H in 8 ■2| cog i o i »« ^ 1 8 s 8 8 8 S § o g S S g M fe " ^ s « S S R s x -^ ?:? ?i !=i ?2 s >-) * . 8 8 8 g 8 g 1 1 i 8 S S 1 8 00 1 S4 Pr. m :s S s O 6» g 8 8 g g g g S5 M "* CO § g 8 8 -g ' ' o o § ^ g "^ «. ^ d s 3 2: (K « — < H « 2 S s S 36 MASS. EXPERIMENT STATION BULLETIN 173. 1 5 g 22 «■ B S Q S g 1 1 § g ? 1 ! i Q 1 1 W5 00 O 1^ o CO irt ao i ,rS g ft .S 2 ^ Q COST OF DISTRIBUTING MILK. 37 1 i ao>i ' ' ' ' ' s 1 ^ l{ 1 ■«! P 1 i" £ ^" B i >> 1 i 1 i 1 § S § S 1 Q III. i I . . - ^ - 1 s s § s s s i s „ ^ ^ ^ 1 1 ^ g S 2 § g s s s 1 s i i SI is f2 S S S g e § i 1 |§ ^ 2 s g 1 i H i 2 s 1 s s 3 - S S ?5 |, o "is o %\ O CO o 38 MASS. EXPERIMENT STATION BULLETIN 173. "S Q d CQ -2F >, § 1 , , o o §£g c^- ci .« > s? o o o o s tr> c^ cq o o .J g S § s l^ i 1 S i '^ 1 < >? S5 § § g i s Of Q " N ira s « « ^ = o d &: 1 fe § g s s g S s i i i S H S CO s 00 i s s o g g s 1 1 ^A s 3 5 i 1 ^ ^ o « ^ i e Q o o o l^.l o ^' ^ fn s o t§ ■» " '^ g o A § s § s i:: s ■S 2 .5 c c3 ci is — S ^ " ^ 1 c-1 fl CS o e<5 l« 5:t ?^ o !? S :?; ^ «• Q s« n a u Z M « J ' 1 Q • 1 2 3 s" S ?, II > i o 1 300 1.000 1,900 600 650 5 « o 1 ' S § 2 ^ 2 g S g S S g p 1 ■§ i s i 1 3 >1 a s ;^ g :3 g § I i 1 2 § 3 s? s 2 is 23.6 45.8 59.4 50 0 26 0 5J 106.25 306.66 173.33 162.50 100.00 g 1 6 < (£ $0 0295 0145 0225 0182 0272 $4,574 27 5,843 10 17,067 70 4,324 69 7,031 64 i 3 5 2 3 ^' S S 1 1 COST OF DISTRIBUTING MILK. 39 The Producer as a Distributor in Comparison with the Dealer. Any comparison of costs that fails to recognize the difference between the business of the producer who distributes his own milk, or his owti milk plus some purchased from his neighbors, and the dealer who buys all the milk he distributes, is surely inadequate. The data in Tables XI and XII are inserted to exhibit this comparison in some detail. The records of four producers and five distributors whose cost accounts were kept with unusual care are chosen for this comparison. As usual the figures on cost per quart (Table XI) are based on milk sold at retail. From the total cost of doing business 3 cents per quart were deducted for retail cream sold and one-half cent per quart for milk delivered at wholesale. The most striking reflection in the whole comparison is the great differ- ence in costs as between individuals whether producers or dealers. Pro- ducers' retailing costs run from 2.51 to 1.67 cents per quart, and dealers' from 2.95 cents to less than half that much, or 1.45 cents per quart. Such wide variations between individuals indicate the fruitlessness of drawing any but the most general conclusions from the final averages. It is evident that much remains to be done in the study of economical and efficient methods of distribution and in profitable investment in equipment and buildings. 1. According to these figures, the average producer is able to distribute retail milk more cheaply, it costing him 2 cents per quart against 2.16 cents for the dealer. An analysis of the figures, however, shows that the dealer's investment is about 12 per cent, greater than the producer's per 1,000 quarts of milk handled. There is some difference in maintenance, but on the whole this is in favor of the dealer. 2. The labor bill of the average dealer is noticeably greater per quart, notwithstanding he is near his market and saves in time. This is indicated by the fact that the dealer retails 42 quarts per mile to the producer's 20 — more than double. The dealer almost always has the advantage of shorter delivery routes. The producer must often travel several miles from his farm before he reaches his first customer and retrace this distance after his load has been delivered. In this instance the producer averaged 12f miles per wagon; the dealer, only 6 miles per wagon. 3. The producer has the advantage in depreciation and working capital. In other words, the dealer invests more in his equipment and buildings, naturally increasing the depreciation and circulating capital accounts. The items of shrinkage and bad bills are significant. These two items are the most important of the overhead costs of the dealers here noted. As a whole the overhead charges and current supplies, i.e., the circulating capital, of the dealers per 1,000 quarts handled are more than 60 per cent, higher than those of the producers. 4. The dealer gives better service in pasteurizing and clarifying and his labor account is also somewhat reduced by use of better labor-saving devices for washing, filling, etc. 40 MASS. EXPERIMENT STATION BULLETIN 173. 5i . 1 s ■g "^ oi § s g C-. CO H g lO w k: (MOO H "a S K § ^ ^ « 2 "* " m CO C-. o « a o c. « <: "rt g g § s s . B g g § ^ s g'3 c ? -a ■ ■ ■ 1 1 . i: — co" M-" T)." t: w- eo- Q 6 o 6 £ ;?, 12; z z z ?1 g COST OF DISTRIBUTING MILK. 41 One must bear in mind, however, that the expenses of collecting the milk are not charged to the dealer. The above figures are calculated from the time the milk arrives at the dairy or distributing plant until it reaches the consumer, the cost of transportation from the producer to the dealer's plant, including freight and haulage from producer to shipping point and from shipping destination to milk plant, not being included, whereas the producer's costs include haulage to the city. To this degree the figures are not comparable. The dealer sometimes collects from the producer, sometimes pays a higher price for milk delivered at his plant, sometimes paj^s freight charges. Usually the difference between milk col- lected by the dealer and milk delivered to the dealer is about one-half cent per quart. When milk is shipped from a distance it is usually laid dovvTi at the dealer's plant for a price equal to or less than the local producing dis- tributor can produce it. In such case the dealer and the producer who sells his own milk may both start from their doors with loads of milk equal in value. When the dealer procures local milk he usually pays one-half cent per quart more for it if brought to his dairy. Further analysis, both from a collective and an individual standpoint, indicates that the variation in the cost of distribution is related closely to the number of quarts delivered per horse in conjunction with the quarts delivered per mile. One dealer (No. 14) with three horses delivers 1,600 quarts daily (including 500 quarts of wholesale milk in cans). Although his mileage per horse (8 miles) is higher than most of the dealers, his ex- ceptionally hea\'y delivery, 45.8 quarts per mile, helps to bring his retail cost dowTi to 1.45 cents per quart. Of the producers. No. 23 delivers at less cost than others in the group, although his mileage is 15 per horse; this is accounted for by the large load hauled — 230 retail quarts per horse — and his comparatively small overhead charges. Producer No. 9 carries 520 quarts on two wagons. His horse load is good and his delivery per mile (29.3 quarts), retail and wholesale, is larger than any other pro- ducer in the group — ■ in fact, nearly 50 per cent, above the average. Table XII will repay careful study. The analysis of cost per 1,000 quarts of milk delivered daily is excellent for comparative study and reveals very striking individual variations. No. 13, who uses four horses and travels 18 miles, with an average load of 107 quarts per horse to deUver 430 quarts daily, has high cost items in all respects. His labor and working capital accounts are nearly thrice those of No. 14 and his other items twice as great. Dealer No. 24 makes up for his high invest- ment and large depreciation and overhead costs by a low maintenance expense and a small labor bill. His labor charge is only one-half that of No. 13, and S700 less per 1,000 quarts than that of the average producer. The efficiency of No. 14 has been noted above. His economies extend to every division of his business. His labor bill is e.xtremely small and except for horse feed his maintenance costs are very low. 42 MASS. EXPERIMENT STATION BULLETIN 173. z o H p 2 5 1 2 1 i 1 n i 1 1 o « 1 i c 1 g 8 in o S S5 a S S 1 25 2 1 1 .si Ij 1 i i g S 1^ "S 1 1 > 1 c i i is 2 2 ^ ■ ^ ^ feci 00 t^ II S 2 II - ° 1. ^ ° i s S5 g ^ 2 3fi S i 1 If 2? S i S Ill "1 § K S " 1^ i| II 5? ?: 2 " 1 S 2 I 1 1 11 to M a a 2 ,2 H a .s o o — 6 J % I '^- s »» -3 .a ja 3 ^ II COST OF DISTRIBUTING MILK. 43 Cost of Delivery of Special Milk. Fortunately reliable data were secured from four distributors who had kept accurate accounts for a number of years. Two of these produced and distributed what they termed "special" milk — unpasteurized, but held to be equal in purity and cleanliness to certified milk. The term "special" is very unsatisfactory. There is no standard for such milk. Whether the term means anything depends on the producer and seller. Frequently the milk is of excellent quality. In these instances it is sold to the consumer at 12 cents per quart. This "special" milk entails extra care, extra labor and good equipment and requires a special market; moreover, the distributors must of necessity travel far to dispose of their product. Distributor No. 1 traversed 47 miles daily to dispose of 350 quarts — but 7.45 quarts per mile traveled. In case No. 2, 15 miles were traveled daily to dispose of 83 quarts of "special" milk, 19 quarts of skimmed milk, and 4.9 quarts of cream; disregarding the skimmed milk, this is equal to 5.86 quarts of "special" milk and cream per mile traveled. No. 1 has much higher depreciation and maintenance expense than No. 2, due to the use of a Ford car and White motor truck. The extra cost, however, is offset by the reduced cost of labor, which is but a trifle more than a third that of No. 2 (Sll.ll as against S31.42 per 1,000 quarts). At least twelve hours of labor were saved daily at 15 cents per hour. As in the case of distributors of market milk, the same conclusion can be drawn from the above figures, namely, economic distribution depends on the number of quarts per horse, in conjunction with the quarts per mile. Cost of Collection and Distribution of Wholesale Milk in Cans. These figures demonstrate the reasonableness of calculating one-half cent per quart for the cost of delivering wholesale milk, as we have done in the case of mixed delivery in the figures given in the previous pages. In this plant the cost was a little more than three-fourths of a cent per quart including collection from producers. Two hours daily were occu- pied by a man and two horses for collecting and six hours for delivery. It is contended, however, that the motor truck is more economical for wholesale delivery, provided the truck can be kept fully occupied and the location will permit its use during the winter. 44 MASS. EXPERIMENT STATION BULLETIN 173. Table XIV. Investment. Depreciation. Maintenance. Buildings Equipment, Horses 81,280 597 600 S3S 40 62 50 65 00 S70 50 258 37 Totals S2,477 $165 90 $328 87 Circulating capital: — Ice $100 00 Interest 123 85 Shrinkage, 86 68 Other, . 91 20 Total, . 8401 73 Labor $803 00 Total costs ?1,699.50 Milk handled: - Daily (quarts), 6000 Yearly (quarts) 219,000 Cost per quart (cents), ,78 Cost per mile (cents), 24.00 Mileage- - Collection, 4 Delivery, 15 Customers, 12 Quarts per customer, 50 Miles per customer, 1.58 Quarts per mile, 31.60 Quarts per horse 300 Motor Truck Deliver ij. The actual cost figures of motor truck milk delivery are of interest in view of the increasing prevalence of these vehicles. Notice that the per mile cost for horse delivery as given above is 24 cents based on about 7,000 miles traveled yearly. The costs below are based on 10,000 miles annually. Under ordinary conditions the truck equipment would deliver the milk on the above route in four hours, one-half the time taken by horses. The operating cost of a motor truck suitable for distribution of whole- sale milk or of "special milk," where the haul is long or loads are heavy, is given below. These figures apply to a White motor truck, three-quarters to 1 ton, m actual operation (1915) by a producing distributor of milk. Per Mile. Gasoline, SO. 0100 Oil 0016 Grease, waste, etc., ........... 0010 Running expenses, ........... 0050 Tires, total cost per set, $175; guaranteed mileage, 5,000, . . -,' .0350 Overhauling and painting after 20,000 miles, approximately §350, . .0175 Interest 5 per cent, depreciation 20 per cent, on an investment of $2,250 = $562.50 on approximate yearly mileage of 10,000, . . . . .0562 Insurance (fire IJ per cent., collision 2| per cent.) on $2,250 = $96.18 on mileage of 10,000 0096 Driver, $S50 per year, over mileage of 10,000, ...... 0850 Total cost per mile. $0.2209 COST OF DISTRIBUTING MILK. 45 Cost of Distribution op Creaai, The distribution of cream exclusively is analogous to the distribution of "special" or of certified milk, excepting that the cost of delivery is increased because the overhead charges are high in comparison ^vith the quantity delivered. Cream from dealers who delivered a small quantity of cream to their regular milk customers is not subject to this high overhead charge and need not be considered here. Only one plant delivering cream exclu- sively is included in this study. A summarized statement of its expenses is presented below. These figures take no account of bottles which were paid for by the customers. Notwithstanding this fact, the long route and small daily delivery raises the cost to more than 7.5 cents (S0.07o9) per quart, as against 4.5 and G.l cents for retailing "special" milk. Summary of Costs of delivering Cream (Orn Plant). Depreciation, $112 23 Maintenance, ......... 543 25 Circulating capital, 399 10 Labor, 1,155 90 Total yearly cost, $2,210 48 Cost per 1,000 quarts yearly, $75 91 Cream delivered yearly (quarts), .... 29,120 Cream delivered daily (six days a week) (quarts). 93.3 Customers, 95 Quarts per customer, .98 Cost per quart to deliver. $0.0759 Miles traveled, 18 Cost per mile. $0.33 Quarts per mile, . 5.18 Miles per customer. .19 Customers per mile, 5.3 Significant Facts of Distribution showing Individual Variations. Table XV is an attempt to exhibit the saHent facts of milk delivery by I indi\idual milkmen. Amherst and Walpole distributors are not included; wholesale dealers and those using motor trucks, cream and skimmed-milk I handlers and those who furnished imperfect data are also omitted. 46 MASS. EXPERIMENT STATION BULLETIN 173. OQ ^1 il •is C^ t^ :ii ';? "i* C^ •j^ tf -? §. •"N ,^ ^ r-^ .hi c2 -1 1 c 1^ -^ <=^ -^ r* s a- c o cc c ^ 6- 1 .„, ^ h is Is sSsS g M 55^3 i illi oooo oodooo o en >• i| a 1 1^ isii i i isi| s II 1:: — ' u 1 :2 > 1 2«-.^ ' o 1 ' c 1 2 5 ■<* 1 < 1.550 550 720 247 750 930 662 750 312 520 215 500 445 400 i ■>JCecO — MQOOOQOt^OQO s gg^2J^5g?g?;5?:?3SS2 30 0 37.5 70 0 15 3 23 2 50 0 41.5 39.3 56.0 29 3 28.6 16 7 21.4 18.1 CO oooooeoinoomoooo s c»2j^l2S^^^«-'-2g = iliiiSSgiiiiii i 1,200 450 560 230 650 650 622 550 280 440 200 400 428 400 i — --- ' sssssssssggsss ' to ■ 14, 1 < COST OF DISTRIBUTING MILK. 47 J^lOCM ^ (M CO »0 C^ <>) Tf CS CJ ^ ;ic^cooGScoxot-.cciccsieoMc »iCiocn «300t^t^OOOO(MOt^lO § ^^^^^^.^^^^S^ti o j;222g2g?iSgSI52 ooooooooooooo °°2°°!2222'"S2'"°""°° Ol giSgssiiiiigs s 1,100 450 300 464 800 175 2,000 320 380 1,400 320 425 300 1 — — SSSJoKSSSSSSsSS 1 e^O)INC^C»IMCN(N(MC<(NCMM s ' 13, i < 48 MASS. EXPERIMENT STATION BULLETIN 173. 6>Q O '^ S 42 ^ O 675 376 834 360 324 300 1,315 810 451 218 3,302 353 s If 5- oioiooccioooococooor- s 2£;S?§2ggg2"^S *! OOCOOOOOOMCOOt-CO CI 2S§S§g5SS*S2 illl 000000100000.0 r~ O-*- 1 SS§=:22;;^gsS?5S • coeocoracocoeococococoeo s 2 O »3 «' 2 ' c C a % N i 1 1,065 152 374 306 289 107 295 3S.9 8.0 16.7 41.6 18.7 9 7 9 3 o 59 1 10.0 30 8 31.0 22.0 6.8 11.6 o 6 0 15.0 12.0 6 0 4.0 15.0 15.0 o CJ ilis^iS 1 1,065 150 370 186 175 102 175 n — ' 3.66 3.68 3.82 4.47 4.70 5 34 5.57 ' n o ' ?• i < COST OF DISTRIBUTING MILK. 49 The per quart costs of retail delivery of the 66 distributors considered are approximately as follows : — 4, or 6 per cent., less than 1.5 cents. 14, or 21 per cent., between 1.5 and 2 cents. 16, or 24 per cent., between 2 and 2.5 cents. 13, or 20 per cent., between 2.5 and 3 cents. 12, or 18 per cent., between 3 and 3.5 cents. 7, or 11 per cent., over 3.5 cents. The first striking observation is the wide variation in costs, and the comparatively uniform distribution between 1.5 and 3.5 cents. The second is the fact that there is no marked correlation between costs and size of business; dealers distributing 300 quarts or less and dealers distributing more than 1,000 quarts daily are found in every group except the first. The third group contains as many dealers handling less than 500 quarts daily as any group and more dealers handling more than 1,000 quarts daily than any other group. Third, considered by groups, the cost per quart of retailing increases and the size of the retail load decreases from the first to the sixth group. It should be noted that the high average retail load of the first group is due to one dealer whose load was exceptionally heavy. Fourth, some correlation is discernible between the number of quarts retailed per mile of haul and the cost per quart, the more quarts per mile the less the cost; but the correlation is not consistent. The average delivery for Group III is 29.5 quarts per mile; that of Group V is 30.9 quarts per mile, though the average cost per quart of delivery of the latter is about 50 per cent, higher than the former. These two factors, how- ever — the size of the load and the density of delivery (quarts per mile) — are two very important considerations in milk delivery. Fifth, the individual variations in the number of quarts retailed per mile per wagon, within the groups, are very significant. In Group I, for example, one dealer distributes 23 and another 68 quarts per mile. In Group II the variations run from 15 to 70; in Group III, from 10 to 70, and in Group V, from 8 to 56 quarts per mile. Under these condi- tions it is very evident that the costs of milk delivery must vary tre- mendously. Finally, the cost of delivery is closely related to the miles traveled per customer (or, inversely, the number of customers per mile), running from one-thirtieth of a mile between deliveries in the first group to one-nine- teenth of a mile in the sixth group. Nothing more strikingly indicates the individual differences in delivery conditions than the customers served per mile traveled. The first group contains one dealer with a record of 68 customers and another with only 10 customers a mile. The third group shows variations between 9 and nearly 60 customers. Group V has one dealer who serves 62 customers a mile, and another who serves less than 3. The significance of these relationships will be considered under "Disadvantages of Competitive Distribution." 50 MASS. EXPERIMENT STATION BULLETIN 173. Some Obvious Disadvantages in Competitive Distribution of Milk. The investigation clearly indicates the very wide diversity of costs in the retailing of milk. At the same time the milk-retailing serv-ice under competitive conditions is faii-ly satisfactory. The consumer usually gets his milk on time and in such quantities as he requires. If the quality of milk delivered by one dealer is not satisfactory, several others are available. It is questionable, however, whether the consumer does not pay roundly for this competitive service. Several economic disadvantages may be indicated. 1. Overcapitalization. — The great majority of the plants visited are of one or two wagon capacity. Eighty -four per cent, of them deliver 1,000 quarts or less daily; 59 per cent., 500 quarts or less; and 23 per cent., 300 or less. To meet the demands of his customers, comply with the milk regulations and compete wuth other milkmen the progressive dealer in- stalls machinery for washing, filling and capping bottles, clarifying, pas- teurizing and cooling his milk. One recognizes that milk is highly perishable and that the time for the processing is necessarily short. Some dealers, however, have installed pasteurizers capable of disposing of 400 gallons per hour, although their total quantity handled is but 900 quarts per day. Some have bottle-fillers filUng 12 bottles at once when handling only 350 bottles daily. This means running the plant below its capacity. A few dealers have buildings or horses and wagons much more ample and expensive than necessary. In some instances the total investment runs to 1.5 cents ($0,015) a quart sold yearly, whereas the average investment for that size of business is less than one-half cent ($0.0043) a quart; in other instances the invest- ment is •?.4 cents a quart when the average is less than 1 cent ($0.0095) per quart for plants of similar capacity. 2. Small Daily Deliveries per Horse. — A load for a good horse over a good road is 300 quarts of milk in bottles but the investigation disclosed the fact that the usual load is much less. The average load of 10 dis- tributors in Springfield is 216 quarts per horse (307 per wagon), and of 28 Worcester milkmen, 234.4 quarts per horse (346.1 per wagon), including wholesale milk in cans. On the other hand, a rather large percentage of dealers haul 300 quarts or more per wagon. More than 12 per cent, of the milkmen retail 15 quarts or less per mile of travel in contrast to nearly 14 per cent, who average more than 55 quarts a mile. The average delivery is about 32 quarts per mile per wagon. That the size of load bears a direct relation to the cost of deUvery is showm in Table XV. 3. Long Hauls are usually Uneconomical. — Several instances can be cited of distributors who traveled from 10 to 15 miles to retail from 100 to 200 quarts of milk. When the distributor is a long distance from his market or when the distance between stops is great, there is a consider- able waste both in man and horse labor through lost time. This is some- what offset by the drivers making their daily entries during these intervals. COST OF DISTRIBUTING MILK. 51 More than 20 per cent, of the routes average 14 miles long and almost half of them average 13 miles. 4. Loss of Bottles. — In Worcester 30 dealers, delivering 15,809 quarts per day, claim a loss of $4,913.42 yearly in bottles. Most of the loss in bottles is the fault of consumers. Bottles are frequently unfit for service when returned and many dealers state that they destroy such bottles. Milk bottles are handy receptacles during preserving season, and one dealer told of a housewife who proudly exhibited 100 quart bottles filled with preserves and, to add insult to injury, asked liim for a sufficient number of caps to cover them. 5. Bad Debts. — This waste is common to all businesses which extend credit but the competitive milk dealer suffers more than ordinary loss because imscrupulous persons have a variety of methods for evading the pa>Tiient of small bills. To prevent this loss many dealers make special trips for collecting. Bad debts cost Springfield and Worcester about 2| per cent, of all costs of distribution. These losses aggregate SO. 54 per 1,000 quarts in Springfield and $0.82 per 1,000 quarts in Worcester. The loss depends entirely on the class of trade, however, and no comparisons or general conclusions should be drawn from these figures. 6. Shrinkage. — This loss, seemingly insignificant, amounts to a con- siderable sum in the course of a year. It cannot, however, be wholly charged to distribution, as a certain amoimt is lost in transportation through carelessness in transit and leaky and dented cans. A good filling apparatus reduces this loss to a minimum in the dairy and whatever loss may be sustained in transit is probably borne by the producer who ships in cans. In general the shipper receives payment for only 8 quarts per can, though the can usually contains 8j to 8^ quarts. 7. Surplus and Spoilage. — This item is considerable in all towns and cities visited and it is one of the great and ever-present problems which the dealer is trying to overcome. Three factors contribute to the problem of surplus milk : — (a) Restaurants and lunch. counters which close on Sunday. (6) Decreased demand owing to depopulation of cities during summer. (c) Excessive production of milk at certain seasons. The solution of the first factor is the business of the dealer. But to solve the question of decreased consumption, which occurs regularly and covers a long period, and of overproduction during certain months of the year is really the business of the producer. Closely allied to shrinkage and surplus is spoilage. Milk which cannot be delivered at once is very likely to sour and so become a total loss. Naturally this waste is more prevalent during the summer at the time of surplus production. The producer who delivers his own milk can some- times regulate the supply by producing more winter milk, by feeding some milk to calves or pigs, or he may be able to sell it to a creamery. The small dealer can do Kttle but dump the surplus into the sewer. In the aggregate the question of surplus milk is a big one which many 52 MASS. EXPERIMENT STATION BULLETIN 173. dealers, large and small, have WTestled with for years with little success. An emergency butter and cheese factory managed co-operatively, which will utiHze part of the existing equipment and take care of all the extra milk, is, perhaps, the best suggestion. Some relief will come from a form of contract with the producers which provides for definite variations in supply. At best there will always be a loss at this point. The loss sustained by 10 dealers in Springfield, delivering 9,600 quarts wholesale and retail daily, amounted to Sl,661.50 per year, or 52 cents per 1,000 quarts retailed annually. This does not represent the whole value of the milk; it was disposed of at the above loss. 8. Duplication in Routes. — The economic waste through duplication of milk routes was evident in all the towns and cities visited. From per- sonal observation, at an apartment house containing four families, three milkmen called to deliver 4 quarts of milk; at another fourth-floor tene- ment three different milkmen climb four flights everj' day to deliver 6 pints to four families. Between the hours of 3 a.m. and 7 a.m. 42 milk wagons were observed to pass down Bowdoin Street, Worcester; only one failed to deposit milk within a distance of 400 yards from the observer. Similar conditions were found in all the other towns and cities visited. In Worcester 103 one-horse milk wagons and 62 two-horse wagons average approximately 8^ miles per wagon per day; the 64 Worcester retail routes considered in this study aggregate 565 miles, 8.83 miles per route. Eight and one-half miles is probably a conservative estimate for approximately 265 milk wagons distributing milk daily in Worcester. The total public street mileage uithin the city limits is 220, but several miles are practically unoccupied. These milk wagons cover approxi- mately 2,250 miles daily to supply the houses on less than 220 miles of streets. Probably they travel 10 to 14 times the populated street mileage every day. Duplication of deUvery routes is common to all retail business, but in large cities measures have been taken to overcome this waste through central deUvery agencies, where the parcels are assembled, sorted and delivered regularly. The system has proved economical but the objections to this method for the deUvery of milk are too serious to overcome, except by the establishment of a co-operative milk plant. 9. Another economic waste generally overlooked, common to other commodities as well as milk, is shipping to other markets than the local one. Why should Worcester, the center of one of the flnest dairying sections, draw on Maine for its milk supply, when milk produced in the vicinity of Worcester is shipped to Boston? Other things being equal, the local market is the best market. Long-distance shipments are expen- sive to some one, and cause shrinkage and deterioration in quality. The producer in Massachusetts is in the very favorable position of having his market at his very door, yet he frequently seeks one further afield at necessarily increased cost to the consumer or a smaller return to him. COST OF DISTRIBUTING MILK. 53 Suggestions for improving Conditions. 1. Keeping adequate accounts to show cost of operation and calling attention to wasteful methods and inefficiencies. A little study will show many leaks which can often very easily be stopped. 2. Standardizing distribution. The data indicate the need of deter- mining what is adequate and efficient equipment for a 500, 800 or 1,200 quart delivery. Is a two-horse load with one driver and a helper or the one-man, one-horse unit the more economical? None of these things has been worked out. To answer these questions completely means standardizing the milk- distributing business; the answer will indicate means of eliminating waste, lessening costs and increasing service. Many such studies as this must be made but even this first one indicates some points of attack. Not only should the individual distributor study his business, but organizations of distributors should be formed in each town and city for mutual improve- ment and the discussion of points of economy, and for agreement on some dixision of territory to lessen duplication of routes and to protect their mutual interests. 3. The introduction of the ticket system to lessen collection costs and save time in delivery. The investigation indicates that the use of tickets tends to eliminate loss of bottles and bad accounts. 4. Large daily deliveries per horse and per driver. Several progressive firms in cities not here considered give a bonus to the driver for all de- liveries and collections, and a commission on all new business above a certain minimum. This makes it an object for the driver to increase his sales, stop at a few more doors, obtain new customers and climb addi- tional stairs. Long hauls from farm to delivery district are costly and the longer the initial haul the more milk deliveries necessary in order that this high initial cost may be offset. 5. Co-operative delivery. But, after all is said, the final adequate solu- tion of milk distribution will come only through municipal delivery or the organization of producing distributors. In small cities and towns a co-operative milk plant, owned and managed by dairymen, is very feasible. One plant could easily process and deliver the necessary 2,500 to 10,000 quarts per day and solve most if not all of the problems of economical and adequate supply. 6. Central milk plants. The problem of milk distribution in large cities is difficult but the organization of the small milkmen operating in one section of a city into a distributing agency would cure many ills and bring about cheaper delivery. Organization of selling is an old matter to manufacturers and merchandisers but not to dairymen. The difficulties are personal, but sometimes personal jealousies and suspicions are fatal to progress and profits. The solution of the milk problem is in the hands of the milk producers and dealers. If they have sufficient courage, foresight, perseverance and 54 MASS. EXPERIMENT STATION BULLETIN 173. determination to organize for the study of their own business and the efficient disposal of their own product, all concerned will benefit. The dairjonen supplying a large percentage of the milk of Erie, Pa., have owned and operated their own plant for years. They handle milk, cream and ice cream and not only distribute an excellent quaUty of milk at low cost, but turn over to the producer a much larger percentage of the consumer's price than he ordinarily obtains. Their success commends their methods to the attention of progressive distributors. They point to the following achievements: (1) a pure milk supply with an amazingly low bacterial count; (2) a lower price than in many other cities; (3) elimination of duplicate routes, resulting in (4) large deliveries per horse and driver; (5) concentration in large and convenient plants; (6) economical disposal of surplus milk by means of a condensery which the association operates; (7) better wages to employees and (8) satis- factory prices to the producers ; (9) practical elimination of the difficulties which usually arise between producer and dealer; (10) no wasteful com- petition and (11) not a cent paid either in interest or dividends to the original shareholders; (12) every cent of net receipts has gone to the producers, to the plant or to a reserve fund. Not only this, but this method places the distribution on such a basis that the town authorities could supervise the supply at a minimum cost by co-operating with other towns similarly situated. The cost of upkeep of a laboratory for a chemist and inspector in a small town is prohibitive at present, but if borne jointly by several towns the expense would be reduced to a figure well within their means. The advantages obtained by milk inspection are too well known to need consideration here. BULLETIN No. 174 NOVEMBER. 1917 MASSACHISETTS AGRICILTIRAL EXPERIMENT STATION THE COMPOSITION, DIGESTIBILITY AND FEEDING VALIE OF PUMPKINS By J. B. LINDSEY This bulletin contains a detailed report on the composition, digestibility and feeding value of the ordinary field pumpkin. On the first two pages will be found a brief statement of the results secured, together with various suggestions relative to the place of the pumpkin in farm economy. Requests for bulletins should be addressed to the Agricultural Experiment Station, Amherst, Mass. Massachusetts Agricultural Experiment Station. Trustees. OFFICERS AND STAFF. COMMITTEE. Charles H. Preston, Chairman, Wilfrid Wheeler, . Edmund Mortimer, Arthur G. Pollard, Harold L. Frost, . The President of the College, ex officio. The Director of the Station, ex officio. Hathorne. Concord. Grafton. Lowell. Arlington. STATION STAFF. Administration. William P. Brooks, Ph.D., Director. Joseph B. Lindsey, Ph.D., Vice-Director. Fred C. Kenney, Treasurer. Charles R. Green, B.Agr., Librarian. Mrs. Lucia G. Church, Clerk. Miss F. Ethel Felton, A.B., Clerk. Agricultural Economics. Alexander E. Cance, Ph.D., 7?i Charge of Departnunt. .Samuel H. De Vault, A.M., Assistant. Agriculture. William P. Brooks, Ph.D., Agriculturist. Henry J. Franklin, Ph.D., In Charge of Cranberry Invrsti gallons . Edwin F. Gaskill, B.Sc, Assistant Agriculturist. Robert L. Coffin, Assistant. Botany. A. Vincent Osmun, M.Sc, Botanist. George H. Chapman, Ph.D., Research Physiologist. Paul J. Anderson, Ph.D., Associate Plant Pathologist. Orton L. Clark, B.Sc, Assistant Plant Physiologist. W. S. Krout, M.A., Field Pathologist. Miss Mae F. Holden, B.Sc, Curator. Miss Ellen L. Welch, A.B., Stenographer. Entomologry. Henry T. Fernald, Ph.D., Entomologist. BtTRTON N. Gates, Ph.D., Apiarist. Arthur I. Bourne, A.B., Assistant Entomologist. Stuart C. Vinal, M.Sc, Assistant Entomologist. Miss Bridie E. O'Donnell, Clerk. Horticulture. Frank A. Waugh, M.Sc, Horticulturist. Fred C. Sears, M.Sc, Pomologist. Jacob K. Shaw, Ph.D., Research Pomologist. Harold F. Tompson, B.Sc, Market Gardener. Miss Ethelyn Streetbk, Clerk. Meteorology. Microbiology. John E. Ostrander, A.M., C.E., Meteorologist. Charles E. M-uishall, Ph.D., In Charge of Department. Arao Itano, Ph.D., Assistant Professor of Microbiology. George B. Ray, B.Sc, Graduate Assistant. Plant and Animal Chemistry . J. B. Lindsey, Ph.D., Chemist. Edward B. Holland, Ph.D., Associate Chemist in Charge {Research Division). Fred W. Morse, M.Sc, Research Cliemisl. Henri D. Haskins, B.Sc, Chemist in Charge {Fertilizer Division). Philip H. Smith, M.Sc, ClicmiM in Charge {Feed and Dairy Division). Lewell S. Walker, B.Sc, Assistant Chemist. Carleton P. Jones, M.Sc, Assistant Chemist. Carlos L. Beals, M.Sc, Assistant Chemist. James P. Buckley, Jr., Assistant Chemist. WiNDON A. Allen, ' B.Sc, Assistant Chemist. J. B. Smith, ' B.Sc, Assistant Chemist. Robert S. Scull, B.Sc, Assistant Chemist. James T. Howard, Inspector. Harry L. Allen,' Assistant in Laboratory. James R. Alcock, Assistant in Animal Nutrition. Miss Alice M. Howard, Clerk. Miss Rebecca L. Mellor, Clerk. Poultry Husbandry. John C. Graham, B.Sc, In Charge of Department. Hubert D. Goodale, Ph.D., Research Biologist. Lloyd L. Stewart, B.Sc, Graduate Assistant. Miss Rachael G. Leslie, Clerk. Veterinary Science. James B. Paige, B.Sc, D.V.S., Veterinarian. G. Edward Gage, Ph.D., Associate Professor of Animal Pathology. J. B. Lentz, 1 V.M.D., Assistant. 1 On leave on account of military service. CONTENTS. Summary of the results, '..... Composition of the pumpkin, .... Digestibility of pumpkins, ..... Feeding experiments with pumpkins, . Feeding pumpkins to milch cows at this station. PAGE 55, 56 66, 67 67-71 Publication of this Document approved by the Supervisor oi'- Administration. BULLETII^ ^o. 174. DEPARTMENT OF CHEMISTRY. THE COMPOSITION, DIGESTIBILITY AND FEEDING VALUE OF PUMPKINS. BY J. B. LINDSEY. Summary of the Results. 1. The pumpkin contains some 17 per cent, of shell, 73 per cent, of flesh, and 9 to 10 per cent, of seed and connecting tissue. It is a watery- fruit, showing extremes of 84 to 91 per cent., with an average of 88 per cent. 2. The whole pumpkin is relatively rich in ash; the seed shows notice- ably less ash than the remainder of the fruit. On the basis of dry matter, the entire pumpkin contains rather more total protein than is found in grains and roots. It also contains some 18 per cent, of total sugars, of which one-third was found to be present in the form of cane sugar. The fruit minus the seeds contains nearly 43 per cent, of total sugars, which explains in a measure its desirability as a hu- man food. The pumpkin seeds are very rich in fat, and are composed substantially of one-third fat, one-third protein and one-fifth fiber, the balance being carbohydrates and ash. 3. A number of digestion trials were made with sheep, and showed the pumpkin to be about 81 per cent, digestible. On substantially the same water basis, and allowing for the increased food value of the fat, the pump- kin appears to have about 20 per cent, greater feeding value than mangels and turnips. 4. Feeding experiments were made with daily cows, substituting in the ration 30 pounds of cut pumpkins for 5 pounds of hay. The results se- cured indicated that 5 to 6 pounds of pumpkins were equal in food value to 1 pound of hay. The Vermont station concluded that 2| pounds of pumpkins were about equal to 1 pound of silage, and that 6^ pounds were fully equal to 1 pound of hay. On page 66 will be found the conclusions of other investigators. 56 MASS. EXPERIMENT STATION BULLETIN 174. The puinpkiu had a tendency to increase temporarily the fat percentage in the milk, due evidently to the oil contained in the seed. 5. The seeds appeared to be free from any injurious effects upon the animals when fed in the amounts found in the entire fruit, contrary to the notion prevalent among many farmers. In foreign countries they are often dried and ground, and serve as a very nutritious and harmless food, if not fed in too large amounts. 6. It is not considered good economy to grow pumpkins exclusivel}^ as a food for either cows or pigs, because of their high water content and poor keeping quality. For the latter reason it is advisable to feed them in the late fall or earlj'^ winter. In one instance a yield of 9 tons is reported when they were grown exclusivelj^, on which basis they would yield about 2,000 pounds of actual food material (digestible organic matter plus fat multiplied by 2.2) as against 3,000 pounds derived from corn. Their place in the farm economy seems in a way to have been discovered by the farmer, namely, in their limited cultivation together with corn. 7. They may be fed cut reasonably fine at the rate of 30 to possibly 50 pounds per head daily, in place of 6 to 10 pounds of hay, in addition to hay and a reasonable amount of grain. It is not advised to feed them with other watery foods such as roots and silage. The}' also may be fed (cut fine) to pigs, mixed with a combination of equal parts, by weight, of corn meal and fine wheat middlings, or with a mixture, by weight, of 95 parts corn meal and 5 parts of digester tankage. It is doubtful if it pays to cook them. If fed in too large amounts daily they furnish too much bulk but insufficient nutriment, and as a result the animals are likely to lose in flesh. COMPOSITION, ETC., OF PUMPKINS. 57 COMPOSITION OF THE PUMPKIN. The ordinary field pumpkin (Cucurbita pepo) is planted more or less by New England farmers, frequeiitly in the field with corn. It is used as a human food, particularly for pies, and is also fed to pigs and to dairy and beef cattle. Ulbricht and Kosutany ^ have shown that in twelve different varieties of the genus Cucurbita the parts were present in the following propor- tions: — Per cent. Shell 17 Flesh, 73 Seed, 2 Seed and supporting tissue, ......... 7 The pumpkin is a watery fruit. We have found variations of from 84.08 to 91.18 per cent., with an average of 87.53 per cent, in four lots grown on two farms in two different years. In the pumpkin minus the seeds and connecting tissue variations of from 90 to 94 per cent, were noted, with an average of 92.78 per cent., while the seeds contained from 43 to 47 per cent. The seeds, it will be noted, were much less watery than the other portions of the fruit. It was noted that the ripe pumpkins without the seeds contained 4 per cent, less water than the same material less mature. The riper the fruit and the drier the autumn the higher will be the percentage of dry matter. Other investigators, including Dahlin,^ Braconnet,^ Zeunak,^ Gerardin,* Wanderleben,2 found in 10 sorts of the entire fruit extremes of from 85.8 to 94.2 per cent, of water, with an average of 90 per cent. Storer and Lewis,^ with 5 varieties, noted variations of from 84.3 to 94.6 per cent., with an average of 90.41 per cent. Hills ' found 87.9 and 90.1 per cent, in two lots of field pumpkins. On the basis of the natural moisture the four lots of the fruit examined by us tested as follows: — • Landw. Versuchsstationen, 32, p. 231. 2 .\fter Ulbricht, already cited. ' Vermont Experiment Station, fourteenth report, Appendi,x, p. iv., and sixteenth re Appendix, p. iii. / 58 MASS. EXPERIMENT STATION BULLETIN 174. 1 ^ . , . « II ^ ' ' ' " ^ § 3 s; § Cs 1l II s S ^ 5 s CO t- 00 lO to 5 S S S s 1 " . . , s s ■ • • ^ 5| S S ? ^ ^ 1 5 S 2 £: E 1 2 S § S s S § S 88 S5 -r N ci ^- < .32 1.89 .57 " 1.67 2.60 2.14 ■* ^ s q r^ 3.26 6.14 3.43 ^ .98 1.22 .98 ° SI S g ^ ^ ^ ?? ^ .78 1.63 1.09 ^ S f^ i§ ^ S? S 1 S 8 S s g ^ M CO i 1 1 1 ' ' ' 1 q s s CO -ji 8 o " ^ . S - » S r- " s S ^ 1 CO r- § .^- - material, the average results, as shown in table on page 58, have been calculated to a water-free basis, as shown in table on page 59. The whole pumpkin contains rather less ash than carrots or mangels, although it is much richer in mineral matter than the ordinary grains. The seed is much poorer in ash than the other portion of the fruit. The dry matter of the entire pumpkin contains rather more total protein than roots or grain, with a portion of it in the amido form. The seeds were found to be very rich in true protein. The fiber content of the fruit is noticeably higher than in roots. The seeds have more fiber than the other portion, due to the tough seed coat. Nearly all of the fat is con- tained in the seed, the analysis of the two samples showing an average of 37.49 per cent. The pumpkin contains large amounts of sugars; in the entire fruit one notes nearly 18 per cent., of which substantially one-third is in the form of cane sugar, while in the portion free from seeds 42.52 per cent, total sugars are noted. While sugar was not determined in the seeds, it is evident that they contain Uttle, being made up chiefly of protein, fat and fiber. Ulbricht ^ and Hills " made analyses of the ordinary field pumpkins, and Zaitschek,^ of the so-called giant pumpkin {Cucurbita maxima), with the following results: — ' Already cited. 2 Vermont Experiment Station, fourteenth report, Appendix, p. iv., and sixteenth report Appendix, p. iii. » Landw. Jahrbucher 35, p. 245. COMPOSITION, ETC., OF PUMPKINS. 61 2QSi III 9.04 14.19 14.11 55.72 6.95 II 89.40 .96 1.50 1.50 5.90 .74 i < tsi ^">.| m 4.32 30.58 29.08 15.96 12.68 36.46 5.36 624.80 1 SSSSsSc^SSS s ....... g ai d^ ^^^ 10.55 15.44 10.49 13.04 56.79 4.18 5.17 412.40 (2 gSf:S5S?3Sg § ^ g s ^W ■ 5i . ::52 • . Id 26.3 3.4 26.5 1.2 4.9 37.7 sis m eo .115 2^ ■-3 3 O lO CO . o Water, . Ash. Crude protein. True protein, . Fiber, Extract matter. Fat. . . . Pentosans, Calories per 100 grams. 62 MASS. EXPERIMENT STATION BULLETIN 174. These figures agree with those secured in this laboratory. They show a high water content in the natural fruit and a relatively high percentage of crude protein. The seed is shown to be particularly rich in protein and oil, and quite low in carbohydrate matter. DIGESTIBILITY OF PUMPKINS. A number of digestion trials were made in two successive years, using two sheep in each case. The pumpkins were fed together with hay and also with hay and gluten feed as basal rations. The entire details of the experiment will be published elsewhere. The coefficients of digestibility only are given in the table on page 63. COMPOSITION, ETC., OF PUMPKINS. 63 ^ Ip s E5 S5 55 2 S S5 S s § S S 28 S S s =3(2 ^ |s|| £:S mj3 |l|g s g § § ^ SS S S8 • ^ S § S § c3 S § -s^S • fill g S S g S S S

i? !"£ fi ^ 1 1 1 1 § II ' c-g-S -2 "a . ' ' ' g 2 S ' o|o i S i i S § e 1 wS a II - (N ^ (M -H CQ -H 1 >H • ^ ^ •* fe 3 S 2 s s 2" % « 2 S 3 2: 4. 4. s 0 S S 0 S 2 i I 64 MASS. EXPERIMENT STATION BULLETIN 174. One notes wider variations in the digestibility of the different ingredients by the two sheep than are desirable. Thus, there are extremes of from 75.41 to 89.32 per cent, in case of the dry matter; 67.20 to 83.63 per cent. in case of the protein; and still wider variations in the fiber. The coefficients for the pumpkins minus the seeds and connecting tissue are much higher, and indicate that if the seeds had been removed the animals would have digested practically the entire fruit. Careful observations failed to note any whole seeds or large portions of seeds in the faeces. It seems evident that in case of sheep No. 1 the pumpkins must have exerted a favorable influence on the digestibiUty of the hay. Zaitschek carried out digestion experiments on the Giant pumpkin with two steers, feeding a combination of hay and pumpkins. His results are tabulated below in addition to our own for comparison. — Source. Is i Q .sl 1 j |l 1 ll 1 1^ Massachusetts Station (2 sheep), . Zaitschek (2 steers), . 8 2 80.7 81.4 82.3 65.4 72.6 76.6 70.3 _ 63.7 61.0 67.6 88.7 89.4 91.6 90.1 - 68.7 80.1 In spite of the variations in results secured at this station with sheep, our average results agree surprisingly well with those secured by Zaitschek. Applying the digestion coefficients to the composition of the pumpkin in its natural state, we have the following digestible organic nutrients in 2,000 pounds: — COMPOSITION, ETC., OF PUMPKINS. 65 li O t- 00 1 § g 3^-: H^ H § § § -3^ -2 0 0 >o o o 1 § § K 4 g S ' s jlj OJ o o o 2 3 S 1 it So ■* t- o o d d "^ g III ■* OJ o o § 2 g g pi s § § g ^ s ^ s a ' a a .2 .2 .2 J • B 3 CO m cQ J • S :§ 111! ^ s ^ ^ f ^i 66 MASS. EXPERIMENT STATION BULLETIN 174. The above data indicate that on the basis of substantially the same water content, 2,000 pounds of pumpkins contain some 9 pounds more of digestible crude protein, 16 pounds more of digestible fiber, 43 pounds less digestible extract matter, and some 27 pounds more digestible fat than are contained in a hke amount of mangels. Mangels, then, are richer in car- bohydrate matter, but less rich in protein and particularly in fat than is the pumpkin. The pumpkin contains more digestible protein than the ruta baga, about the same amount of fiber, rather less carbohydrate matter and noticeably more fat. On the basis of total digestible nutrients, allowing for the increased energy value of the fat, the two roots appear to have about 20 per cent, less feeding value than the same weight of pumpkin. These figures, of course, cannot be taken too Uterally. It is doubtful if the computation of net energy values — because of the scan- tiness of the data — would tl^row any additional Ught on the relative values of the several feeds. FEEDING EXPERIMENTS WITH PUMPKINS. A number of experiments are recorded relative to the value of pump- kins as a feed for cows and pigs. Hills ^ fed three cows in three periods of fifty-four days each on hay, silage, a grain mixture and pumpkins. Dur- ing the first and third periods the cows received the hay, silage and grain, and in the second period, hay, silage, grain and pumpkins. Two and one- half pounds of pumpkins with 90.1 per cent, of water were substituted for 1 pound of silage, with apparently like results. In a second experiment with four cows, feeding pumpkins in the second of three periods at the rate of 40 pounds per cow daily, he concluded that 6| pounds of pumpkins with 87.9 per cent, water were equal to 1 pound of hay. French ^ fed six Berkshire pigs that were eight months of age on a ration of wheat shorts and field pumpkins (cooked) with the seeds removed. The experiment covered five periods of eighty-four days each, and in the last two periods the pigs consumed an average each of 26 pounds of pump- kins per day. The average daily gain in live weight was 1.5 pounds, and the results were considered quite satisfactory. Burkett ' fed several lots of three pigs on combinations of skim milk, corn meal and pumpkins cooked and uncooked; also on milk and raw pumpkins versus milk and corn meal; and on milk, pumpkins and apples, half and half, cooked, versus milk, corn meal and bran, half and half. The general conclusion was that cooking did not increase the feeding value of pumpkins, and that a combination of skim milk, corn meal and pumpkins gave the most satisfactory results. Pott * reports that in England pumpkins are quite generally fed to fat- > Already cited. ' Oregon Experiment Station, Bui. No. 53, p. 22. • New Hampshire Experiment Station, Bui. No. 66. * Handbuch der tierischen Ernahrung, etc., 11. Band, pp. 424, 425. COMPOSITION, ETC., OF PUMPKINS. 67 tening pigs, together with ground barley and beans; also to milch cows at the rate of 25 to over 100 pounds daily, cut fine and mixed with cut straw; and to fatl^ening cattle as high as 100 pounds daily, preferably cooked. Pumpkins are also fed in Austria to cows, fattening cattle, pigs and horses. Pott states that the claim made that the seeds are injurious is without foundation. Feeding Pumpkins to Milch Cows at this Station. In order to observe the effect of pumpkins upon the quantity and quality of milk and on the general condition of the animals, two grade Jersey cows were selected and fed with 30 pounds each of pumpkins daily, ill addition to hay and grain. The data and plan are as follows: — History of Cows. Name. Breed. Age (Years). Last Calf dropped. Approx- imate Milk Yield (Pounds). Fat (Per Cent.). Weight of Cows (Pounds). Samantha, . Redm.. Grade Jersey. Grade Jersey. 11 9 August 25 August 11 36.7 23.5 4.1 3.9 950 910 Plan and Duration op Experiment. The two cows were fed in three distinct periods of twenty-one days each, exclusive of the preliminary periods. In the first period they each received a ration of hay, bran and cottonseed meal and hominy meal; in the second period the same ration, excepting that 5 pounds of the hay were replaced by 30 pounds of the pumpkins; in the third period the ration fed was the same as in the first period. The results secured in the first and third periods were averaged and compared with those secured in the second. Five pounds of hay were therefore compared with 30 pounds of pumpkins. Care of Animals. — The animals were well cared for and turned into a barnyard about eight to nine hours each day. They were fed twice daily; the hay was given sometime before milking and the grain just before milking, while in the morning the grain was given just before, and the hay just after, milking. Water was supplied constantly by aid of a self-watering device. Character of Feeds. — The hay and grains were of the usual good qual- ity. The pumpkins were grown by one farmer and were the ordinary yellow field variety of different sizes.' Most of them were ripe. Sampling Feeds and Milk. — The hay was sampled at the beginning and end of each period by taking forkfuls of the daily weighing, running the 68 MASS. EXPERIMENT STATION BULLETIN 174. same through a power cutter, subsampUng and placing the laboratory samples in large glass-stoppered bottles with proper markings. The grains were sampled daily by placing definite amounts in glass-stoppered bottles, and these bottles properly labeled were brought to the laboratory at the end of each period. The pumpkins were cut into small pieces before being fed. The analytical data serving for the digestion experiment also served for this experiment. Analysis of the Milk. — The milk of each cow was sampled daily for five consecutive days of the last two weeks of each period, the samples preserved with formalin, and the five-day composite sample tested for solids and fat. Weighing the Animals. — The animals were weighed for two conseci^- tive days at the beginning and end of each half of the period before the afternoon feeding. Analysis of Feedstuffs. Water. Ash. Protein. Fiber. Extract Matter. Fat. Hay. 11.34 0.16 5.14 31.03 45.57 1.76 Bran, 12.45 6.47 15.73 10. ?7 50.68 4.40 Cottonseed meal, 8.81 6.37 41.63 10.19 25.91 7.09 Hominy meal 11.24 2.05 10.41 4.48 64.67 7.15 Pumpkins 84.77 1.14 2.50 2.10 7.77 1.72 Total Feed consumed {Pounds). Average, Periods I. atid III. Name. Hay. Bran. Cotton- seed Meal. Hominy Meal. Pump- kins. Red III., Samantha, 378 504 63 84 42 63 42 84 ■ Period II. Red III Samantha 273 399 63 84 42 42 63 84 630 630 COMPOSITION, ETC., OF PUMPKINS. 69 Daily Feeds consumed (Pounds). Hay -\-Grain (Periods I. and III.). Name. Hay. Bran. Cotton- seed Meal. Hominy Meal. Pump- kins. Red III., Samantha 18 24 3 4 ' ' 2 4 - Hay +Grain+Pumpkins (Period II.). Red III., Samantha, ..... 13 19 3 4 2 3 2 4 30 30 Estimated Digestible Nutrients in Daily Rations. Hay -\-Grain (Periods I. and III.). Name. Protein. Fiber. Extract Matter. Fat. Total. Nutritive Ratio. Red HI Samantha, 1.73 2.51 3.60 4.86 7.63 11.03 .53 .83 13.49 19.23 1:7.2 1:7.1 Average, .... 2.12 4.23 9.33 .68 16.36 - Hay -\-Grain -{■Pumpkins (Period II.). Red III., . Sfimantha, . . 2.16 2.95 3.05 4.31 8.31 11.71 .97 1.27 14.49 20.24 ■ :0 1:6.1 Average, 2.55 3.68 10.01 1.12 17.36 The above nutrients were estimated on the basis of actual analysis and the application of average digestion coefficients. The 30 pounds of pump- kins fed contained 1 pound more digestible nutrients than 5 pounds of hay. This was due to the fact that the pumpkins had rather less water than was expected, and that they contained such a high percentage of digestible matter. On the basis of digestible matter, 1 pound of haj' is equivalent to some 4| pounds of pumpkins. 70 MASS. EXPERIMENT STATION BULLETIN 174. Weights of the Animals (Pounds). Red III. Samantha. Period, I. III. II. I. III. U. Be^nning, .... End 915 948 930 930 905 928 1,095 1,140 1,153 1,148 1.095 1,118 Gain or loss, + 33 ± +23 + 23 +45 1 -5 +23 Average, . . + 17 + 20 +23 Gain or Loss for Both Cows. Periods I. and III. (hay+grain) = 37 pounds+. Period II. (liay+grain+pumpkins) = 46 pounds+. There seems to have been very Uttle difference in the changes in weight as a result of feeding the two rations. Total Yield of Milk Products. Hay -[-Grain {Period I.). Name of Cow. Total Milk (Pounds. Daily Milk (Aver- age). Total Solids (Pounds). Total Fat (Pounds). Average Per Cent. Total SoUda. Average Per Cent. Fat. Red III Samantha, 364.4 532.1 17.4 25.3 47.88 76.73 17.49 29.11 13.14 14.42 4.80 5.47 Hay+Grain (Period III.). Red III Samantha 301.6 460.0 14.4 21.9 42.07 69.18 16.47 26.40 13.95 15.04 5.46 5.74 Hay -{-Grain + Pumpkins (Period II.). Red III.. Samantha, 341.7 495.3 16.3 23.6 48.15 76.08 19.24 29.87 14.09 16.36 6.63 6.03 COMPOSITION, ETC., OF PUMPKINS. 71 Total Yield of Milk Products — Concluded. Hay -'rGrain {Average, Periods I. and III.). Xame of Cow. Total Milk (Pounds). Daily Milk (Aver- age). Total Solid3 (Pounds). Total Fat (Pounds). Average Per Cent. Total Solids. Average Per Cent. Fat. Average Per Cent. Solids not Fat. Red III Samantha, 333.0 496.1 15.9 23.6 45.09 73.08 17.08 27.83 13.54 14.73 5.13 5.61 8.41 9.12 Average, . 414.6 19.7 59.09 22.46 14.25 5.42 8.83 Hay +Grain+ Pumpkins {Period II.). Red III Samantha, 341.7 495.3 16.3 23.6 48.15 76.08 19.24 29.87 14.09 15.36 5.63 6.03 8.46 9.33 Average, . 418.5 19.9 62.12 24.56 14.84 5.87 8.97 The yield of milk was substantially the same on each ration. The total solids showed an increase as a result of feeding the pumpkins, and this was due evidently to an increase in the percentage of fat in the milk. Attention has been called to the fact that the pumpkin seeds are rich in fat. By referring to the average daily rations consumed (page 69) it may be seen that the ration without pumpkins contained .68 pound daily of digestible crude fat, and with the pumpkins 1.12 pounds, the excess of .44 pound of pure fat being derived from the pumpkin seeds. This additional food fat evidently temporarily increased the fat in the milk. In so far as the results of a single experiment with two cows are concerned it appears that 6 pounds of pumpkins fully replaced 1 pound of hay. On the basis of digestible nutrients our calculations show that 4^ pounds of pumpkins with 84.8 per cent, of water replaced 1 pound of hay with 11.34 per cent, of water. It is quite possible that 25 pounds of pump- kins would have replaced 5 pounds of hay with equal results. Because of the rather wide variations in the moisture content of the fruit, one could say only on the basis of results secured, that from 5 to 6 pounds of pump- kins were equivalent to 1 pound of first-class cow hay. BULLETIN No. 175 MAY, (917 MASSACHUSETTS AGRICILTLRAL EXPERIMENT STATION Mosaic Disease of Tobacco By G. H. CHAPMAN This bulletin deals with the mosaic disease of tobacco, includ- ing a brief review of the work of previous investigators. It also gives results obtained at this station relating to the cause, reac- tion and control of the disease. It is shown that more than 80 per cent, of field infection may be traced originally to the seed bed. Specific methods for control are recommended. Requests for bulletins should be addressed to the Agricultural Experiment Station, Amherst, Mass. IVIassachusetts Agricultural Experiment Station, Trustees. OFFICERS AND STAFF. COMMITTEE. Charles H. Prkston, Chairman, . Hathorne. Wilfrid Wheeler, . . Concord. Edmund Mortimer, . Grafton. Arthur G. Pollard, . Lowell. Harold L. Frost, . . Arlington The President of the College, ex officio. The Director of the Station, ex officio. STATION STAFF. Administration. William P. Brooks, Ph.D., Director. Joseph B. Lindsey, Ph.D., Vice-Director. Fred C. Kenney, Treasurer. Charles R. Green, B.Agr., Librarian. Mrs. Lucia G. Church, Clerk. Miss F. Ethel Felton, A.B., Clerk. Agricultural Economics. Alexander E. Cance, Th.D., Agricultural Economist. S. H. DeVault, A.M., Graduate Assistant. Agriculture. William P. Brooks, Ph.D., Agriculturist. Henry J. Franklin, Ph.D., In Charge Cranberry Sub- station. Edwin F. Gaskill, B.Sc, Assistant Agriculturist. Robert L. Coffin, Assistant. Botany. A. Vincent Osmun, M.Sc, Botanist. George H. Chapman, Ph.D., Research Physiologist. Paul J. Anderson, Ph.D., Associate Plant Pathologist. Orton L. Clark, B.Sc, Assistant Plant Physiologist. W. S. Krout, M.A., Field Pathologist. Miss Grace B. Nutting, Ph.B., Curator. Miss Ellen L. Welch, A.B., Stenographer. Chemistry. J. B. LtNDSET, Ph.D., Chemist. Edward B. Holland, Ph.D., Associate Chemist in Charge {Research Section). Fred W. Morse, M.Sc, Research Chemist. Henri D. Haskins, B.Sc, Chemist in Charge {Fertilizer Section) . Philip H. Smith, M.Sc, Chemist in Charge {Feed and Dairy Section) . Lewell S. Walker, B.Sc, Assistant Chemist. Carleton p. Jones, M.Sc, Assistant Chemist. Carlos L. Beals, M.Sc, Assistant Chemist. James P. Buckley, Jr., Assistant Chemist. W. A. Allen, B.Sc, Assistant Chemist. J. B. Smith, B.Sc, Assistant Chemist. Robert S. Scull, B.Sc, Assistant Chemist. James T. Howard, Inspector. Harry L. Allen, Assistant in Laboratory. James R. Alcock, Assistant in Animal Nutrition. Miss Alice M. Howard, Clerk. Miss Rebecca L. Mellor, Clerk. Entomology. Henry T. Fernald, Ph.D., Entomologist. Burton N. Gates, Ph.D., Apiarist. Arthur I. Bourne, A.B., Assistant Entomologist. S. C. Vinal, B.Sc, Graduate Assistant. Miss Bridie E. O'Donnell, Clerk. Horticulture. Frank A. Waugh, i M.Sc, Horticulturist. Fred C. Sears, M.Sc, Pomologist. Jacob K. Shaw, Ph.D., Research Pomologist. Harold F. Tompson, B.Sc, Market Gardener. R. P. Armstrong, M.Sc, Graduate Assistant. Miss Eleanor Barker, Clerk. Meteorology. John E. Ostrander, A.M., C.E., Meteorologist. Microbiology. Charles E. Marshall, Ph.D., Microbiologist. F. H. Hesselink van Suchtelen, Ph.D., Research Mi- crobiologist. Poultry Husbandry. John C. Graham, B.Sc, Poultry Husbandman. Hubert D. Goodale, Ph.D., Research Biologist. Lloyd L. Stewart, B.Sc, Graduate Assistant. Miss Rachael G. Leslie, Clerk. Veterinary Science. James B. Paige, B.Sc, D.V.S., Veterinarian. G Edward Gage, Ph.D., Research Pathologist. J. B. Lentz, V.M.D., Assistant. On leave. CONTENTS Introduction, Historical summary, Names, .... Description of the mosaic disease of tobacco, Occurrence, .... Economic importance, Infectious nature of the disease. Contagious nature of the disease. Pathological anatomy, Leaves, .... Stems, .... Roots, .... Fungi and the mosaic disease, . Bacteria and the mosaic disease, Dissemination agents. Insects, .... Workmen, Seed Fertilization in relation to mosaic disease Effect of colored light on mosaic disease. Experimental data, . Biochemical studies. Enzyme activities in healthy and diseased plants. Reaction of mosaic sap with various substances. Probable character of the causal agent, Prevention and control, ..... Summary, ....... PAGE 73 74 78 78 80 81 81 82 83 83 84 85 85 86 87 87 88 89 90 91 93 105 110 113 117 Publication of this Document approved by the Supervisor of Administration. BULLETII^ JSTo. 175. DEPARTMENT OF BOTANY. MOSAIC DISEASE OF TOBACCO. BY G. H. CHAPMAN. Introduction. The observations and conclusions reported in the following pages are the results of several years of more or less continuous investigation on the part of the writer, and deal with the probable causes, occurrence, appearance and methods of control of this well-known disease of tobacco and related plants. Enough has been accomplished so that it is believed wise to add still another paper to the already long list of Hterature which has been published on this disease. During the time in which these experiments have been in progress much new literature has appeared dealing with this subject, some of which has helped the writer by verifying his results and by bringing out new facts concerning the disease; but, on the other hand, some of the work appears to have been done in a hasty manner, and possibly erroneous conclusions drawn in some cases, thus adding to the large amount of confusing subject-matter which has to do with this disease. The experiments carried on by the writer were begun in a general way in 1907, and have been repeated several times during the years subsequent to that date, new Hues of investigation both in the field and laboratory having been added as occasion demanded. Some considerable time has been spent in verifying the results obtained by other recent investigators, and an attempt has been made to gather together in a broad, general way, as well as in detail, all the reliable information possible about this interesting disease, as well as to bring out new facts in regard to it. More attention has been given to the biochemical aspects of the problem than has heretofore been done by investigators. " Also presented in part to the facility of the graduate school of the Massachusetts Agricultural College, June, 1916, as a major thesis in partial fulfillment of the requirements for the degree of doctor of philosophy. 74 MASS. EXPERIMENT STATION BULLETIN 175. Historical Summary. In the following paragraphs is given a brief r^sum^ of the more important work done on the mosaic disease of tobacco up to the present time, and as an excellent critical review of the Hterature, etc., up to 1902 is given by A. F. Woods 1 in his work on the subject, the same is quoted in full below. He states : — Adolph Mayer ^ was the first to make a careful study of the trouble. He demon- strated that it could not be caused by an insufficient supply of mineral nutrients. He found as much nitrogen, potassium salts, phosphates, calcium and magnesium present in the soils and plants v/here the disease occurred as in the soils where the disease did not occur. He also found that the trouble was apparently distributed over the field without regard to the soil conditions. Since tobacco requires much lime, liming the soil was tried, but the disease was not prevented thereby. Mayer further kept hotbeds in some cases rather moist, in others dry, and then again, richly or poorly manured with nitrogen; but in no case could he determine that the conditions in question caused the disease. He also found that variations in the temperature of the hotbeds apparently had no effect; neither did crowding, which produced partial etiolation, appear to have any effect on the disease. Seeds from flowers in which self-fertilization was pre- vented he found to be just as susceptible to the disease as seeds produced without such precautions, but on the soil on which the disease had once appeared it was again produced. According to his observation, also, the trouble was not often found on soil used for the first time for tobacco. He further proved that the juice of the diseased leaves injected with the juice of healthy plants did not develop the disease. He was not able to produce it by injecting diseased juice into other solanaceous plants. Where the diseased juice was injected into tobacco the same trouble developed in from ten to eleven days. Heating to 60° C. did not destroy the infectious substance; at 65° to 75° it was attenuated, and at 80° it was killed. After Mayer had shown the absence of animal and fungous parasites he. sup- posed bacteria to be the cause of the disease, but all his efforts with bacteria cul- tivated from the surface of diseased leaves, and also with different mixtures of bacteria, failed to produce it. Nevertheless, he thought that there must be certain pathogenic bacteria present in those soils in which the disease appeared, and therefore proposed to change the soil in the hotbeds and to devote the fields where tobacco had been cultivated to other crops. He also recommended the use of mineral rather than organic manures. These general results were confirmed by several subsequent investigators. Not, however, till Beijerinck ' took hold of the question was much of importance added to our knowledge of the malady. He proved the absence of bacteria in the devel- opment of the disease. He showed that the juice of the plant filtered through Chamberland filters, while remaining perfectly clear and free from bacteria, still retained the power of infection. A small drop of it injected hypodermically into the growing bud was sufficient to give the plant the disease. He found that only dividing (meristematic) cells can become diseased. Diseased tissue kept its in- fectious qualities even after drying, and retained its injurious properties in the > Woods, A. F.: Observations on the Mosaic Disease of Tobacco. U. S. D. A., Bur. Plant Ind. , Bui. No. 18 (1902). 2 Mayer, Adolph: U'ber die Moaaikkrankheit des Tabaks. Landw. Versuchsstation, 32: 4 51^67 (1886). Review of the same article in Journ. of Mycology, 7: 382-385 (1894). ' Beijerinck, M. W.: Verhandelingen der Koninklijke Akademie van Wetenachappen te Amsterdam. Deel 6: No. 5. See also Centb. f. Bakt., Par., etc., II: 5: 27-33 (1899). MOSAIC DISEASE OF TOBACCO. 75 soil during the winter. Weak solutions of formalin did not kill the virus, but heating to boiling point did. Fresh, unfiltered juice was more effective than an equal amount of filtered juice. He found that soil around diseased plants may infect the roots of healthy plants, but he did not determine whether direct trans- ference is possible through healthy root surfaces, or whether insects, by injuring the roots, favored infection. He defines the milder form of the disease as a suffer- ing of the chlorophyll bodies. Later a general disease of the plasmatic contents of the cells sets in. In field conditions as a final stage the swollen green areas become marked with small dead spots, but these did not appear on plants grown under glass. Under certain conditions he observed that plants apparently recover from the disease; i.e., the new growth appeared to recover. He found that the infective material, whatever it might be, could be transported through considerable distances in the plant, but could cause the disease only in the dividing cells. He assumed the virus to be a non-corpuscular, fluid-like material, which had the power of growth when in contact, in a sort of symbiotic way, with the growing cells, — "a living fluid contagium." Shortly after Beijerinck's paper, Sturgis ' published a critical review of the work done on the disease up to that time, with numerous valuable results and observa- tions made in Connecticut, where the trouble is known as "calico" or "mottled top." The results obtained by Sturgis and observations made by him on tobacco in Connecticut bore out the statements of other careful and critical workers, and greatly cleared up the field for further investigation. He came to the conclusion that on close, clayey soils the disease may be more abundant than on an open, porous soil. The disease is not conta- gious, but he could not state definitely as to its infectiousness; it is not caused by fungi, nematodes or parasitic insects, and the facts observed by him were not favorable to the theory of bacterial origin. He also came to the conclusion that the disease is not inherent in the seed, and looked upon it as a purely phj^siological trouble brought about by sudden interruptions of the normal plant metabohsm. Koning, ^ in his work, verified much of the work of Beijerinck and Mayer, and Woods ^ later verified the work of these investigators and pointed out that in the diseased leaves there was an excess or excessive activity on the part of an enzyme belonging to the oxidases, and that the power of oxidation in the cells was inversely proportional to the amount of chlorophyll present, using the color as a basis of comparison. He also pointed out that there was a marked structural difference between the cells of the dark green and Ught green areas, and proved to his own satisfaction that the Hght green areas are the truly diseased portions, a fact that will be referred to later in this paper. In a later careful investigation of the disease Woods * arrived at the following conclusions, which were a great stride forward in our understanding of some phases of this baffling disease. He states: — > Sturgis, W. A.: Mosaic Disease of Tobacco. Conn. Agr. Exp. Sta. Rept., 250-254 (1898). ' Koning, C. J.: Die Flecken oder Mosaikkr&nkheit des hollandischen Tabaks. Zeitschrift fur PBanzenkr., 9: 65-80. * Woods, A. F.: Inhibiting Action of Oxidase on Diastase. Science, n. s.. No. 262, 17-19. « Woods, A. F., loc. cit. 76 MASS. EXPERIMENT STATION BULLETIN 175. The disease is not due to parasites of any kind, but is the result of defective nutrition of the young dividing and rapidly growing cells, due to a lack of elabo- rated nitrogenous reserve food accompanied by an abnormal increase in the activity of oxidizing enzymes in the diseased cells. The unusual activity of the enzyme prevents the proper elaboration of the reserve food, so that a plant once diseased seldom recovers. On the decay of the roots, leaves and stems of both healthy and diseased plants, the enzyme in question is liberated and remains active in the soil. The enzyme is very soluble in water and appears to pass readily through plant membranes. If the young plants take it up in sufficient quantity to reach the terminal bud, they become diseased in the characteristic way. Under field conditions there is little danger from infection in this manner, but in the seed bed the danger is much greater on account of the greater susceptibility of the young plants to the disease, and the greater amount of free oxidizing enzymes likely to be in the soil due to the decay of the roots and plants. New or steam sterilized soil should therefore be used for the seed bed. I have shown that transplanting, especially when the roots are injured, may produce the disease. Great care must, therefore, be taken not to injure the roots in this process or in the subsequent cultivation, or to check the growth of the plants. There is evidence that rapid growth, caused by too much nitrogenous manure or too high a temperature, is favorable to the disease. Why this should be the case has not been determined. It is probably connected with the manufacture of reserve nitrogen by the cells and its distribution to the rapidly growing parts. Plants grown under such conditions are less able to stand successfully marked variations in temperature and moister conditions of soil and atmosphere. Varia- tions of this kind favor the development of the disease in the less resistant plants. Close, clayey soils, packing hard after rains and requiring constant tillage, are not favorable to the even growth of either the tops or roots of tobacco plants. In moist, cloudy weather the plants will grow too fast, and in hot, dry weather the soil is likely to bake, checking growth and making probable injury to the roots in cultivation. Such soils are very favorable to the development of the mosaic disease, as pointed out by Thaxter. i He found that loosening the soil by liming and giving partial shade, thus causing a more even condition of growth, very greatly reduced the disease. Crops grown under cheesecloth covers protected at the side are said to be re- markably free from the disease. The plants make a steady rapid growth, much greater than in ordinary field culture. . . . The disease is not, so far as observed, produced by a lack of soil nutrients, though from its nature we would expect that a deficiency of nitrogen, phosphoric acid, lime and magnesia might favor its development. Koning^ says that manuring with kainit and Thomas slag diminishes the extent of the disease. Mayer, Beijer- inck and other investigators, however, agree that the trouble is not caused by the lack of any soil nutrients. It appears, so far as my own investigations go, that the trouble cannot be cured by giving the plants additional food of any kind. Over- feeding with nitrogen favors the development of the disease, and there is some evidence that excess of nitrates in the cells may cause an excessive development of the ferments that cause the disease. Very slight attacks of the disease known as "mottled top" are said not to injure the quality of the leaf to a sufficient extent to be noticeable commercially, though they may be less elastic and have a poorer burn and aroma than healthy leaves. Hunger, ^ in his work on the mosaic of Deli tobacco, verified much of the work of previous investigators, and later, in carefully planned and « Thaxter: Conn. Agr. Exp. Sta. Rept., Ill, 253 (1899). ' Koning, C. J., loc. cit. > Hunger, F. W. T.: De Mozaiek-ziekte bij deli Tabak. Med. s'Lands Plantentium, Batavaia. Deel 1: 63 (1903). MOSAIC DISEASE OF TOBACCO. 77 executed experiments,^ proved that the disease was not contagious but was highly infectious. He beUeved that it could be carried from diseased to healthy leaves simply by touching, especially in the case of the young leaves, a fact that makes it necessary for the workman to use great care when looking for the tobacco bud worms, etc., in the buds. He was of the opinion that a rupture of the leaf was not necessary to induce the mosaic disease in plants. Selby 2 a year later showed this to be apparently true for tobacco grown in Ohio, and Hunger's statements were in his opinion in all respects con- firmed. He also reported that "Blossoms of various plants were inocu- lated through the nectar by transmission of nectar from diseased plants, as by insect visitation. A slender brush of horse hair was used for this purpose. No evidences of the disease were observed as a result of this method." CUnton* was able to produce the trouble on tomatoes by inoculating with juice from a diseased tobacco plant and from the tomato so infected was able to reproduce the disease on the tobacco again by inoculation from the tomato, again showing the infectious nature of the disease, and that the troubles on the tomato and tobacco were practically identical. This has been repeatedly verified by the writer and many other investi- gators. Jensen,* in his work on the disease, came to the conclusion that the right way to get at the methods of control of the disease was by experi- mentation to obtain a resistant strain of tobacco, no matter what the cause of the disease might be, and he carried on some experiments along these lines. As yet no definite results have been reported by the in- vestigators, but the time has probably been too short to obtain results. Lodewijks* stated that by subjecting diseased plants to diiferent col- ored fights he was able to bring about a cure in some cases. He states: — The mosaic disease cannot be diminished or prevented by lessened light intensity. Neither diffused nor colored light has any effect on the disease if the healthy leaves are not able to function in normal daylight. Under the latter condition, however, diffused light exerts a retardation, red light diminishes the trouble, and blue light effects a cure. All the results may then be explained by the hypothesis that the virus formation diminishes with the intensity of the light, while in the healthy leaves, through the action of the virus, an anti-virus is formed, the action of which destroys the virus (immunity and antitoxin formation in the case of animals). . . . Normally in the metabolism of the tobacco plant a substance is formed, the action of which is opposed to that of the equally normally occurring virus of mosaic disease, perhaps because it binds itself chemically to the latter. ' Hunger, F. W. T.: Die Verbreitung der Mosaikkrankheit infolge der Behandlung dea Tabaka. Centralbl. f. Bakt. Par., etc., II: 11: 405-408 (1908). 2 Selby, A. D.: Tobacco Disease. Ohio Agr. Exp. Sta. Bui. No. 15, 88-95 (1904). ' Clinton, G. P.: Notes on Fungous Diseases, etc. Conn. Agr. Exp. Sta. Rept., 1907-08, 857- 858. * Jensen, H.: tJber die Bekampfung der Mosaikkrankheit der Tabakpflanze. Centralbl. f. Bakt. Par., etc., II: 15: 440-445 (1906). ' Lodewijks, T. A., Jr.: Zur Mosaikkrankheit des Tabaks. Rec. Trav. bot. Neerlandais, VII. (1910). 78 MASS. EXPERIMENT STATION BULLETIN 175. Both substances, virus and anti-virus, may be increased by external factors or conditions. In the first instance the plants become diseased with the mosaic disease; in the latter an immunity against the disease is brought about. Decrease in intensity and cure occur if the virus formation ceases or stops, while at the same time the formation of an anti-virus is taking place normally or is increased. * A discussion of Lodewijks' work is to be found later in this paper. Allard^ in a recent work on the disease states that from the results of his experiments he is of the opinion that the trouble is not primarily physiological but is parasitic in nature, but he is unable to throw any light on the nature of the parasite, and in spite of the conclusions drawn by him, none of his results, at least in so far as the writer is able to judge, has in any way weakened the theory that the trouble may be physiological in nature; and some of his results, from the writer's point of view, seem to substantiate this idea of a physiological agency. Two points of great interest are brought out by him, viz., the mosaic as affecting the color of the corolla by blotching, etc., and the carrying of the disease by certain aphids. These points have not been noted before. In the follo-\\ing pages some of his work wiU be taken up in detail in so far as it seems to bear out or refute work done by the writer. It may be seen from the foregoing r6sum4 that the theory that the disease is physiological in character has been in the past pretty generally accepted, but the identification of the ultimate causes producing the symptoms varies widely with the different investigators. The writer's conclusions with regard to this point are taken up later in this paper. Names. By right of priority the term "mosaic" is the one which should be applied to this disease. It has, however, many local names, and these sometimes are appUed differently to the different manifestations of the symptoms; among them may be mentioned the following: "calico," "brindle," "mongrel," "mottle-top," "string leaf," "frenching," etc. Other terms have also been used, but they do not in all cases apply to the "mosaic" alone, hence they are here omitted. The term "infectious chlorosis" as suggested by Clinton is perhaps best descriptive of diseases of this general character, with "mosaic" as a specific type under this division, there being many other infectious, chlorotic diseases of plants quite distinct from the mosaic type. Description of the Mosaic Disease of Tobacco. Descriptions of the mosaic disease of tobacco have been repeatedly presented, and the disease itself is so well known that there is little need of repetition at this point, but a brief r6sum6 of the salient characteristics 1 Translation from abstract of Lodewijks' paper in Bot. Centralbl., 114-518 (1910). « Allard, H. A.: Mosaic Disease of Tobacco. U. S. D. A., n. s.. Bur. Plant Ind., Bui. No. 40 (1914). MOSAIC DISEASE OF TOBACCO. 79 of the disease will be given so that no misunderstanding may arise, as several other leaf troubles more or less chlorotic in character have often been confounded with the true "mosaic." The disease may show on the leaves at all stages of the growth, from the seedling to the mature plant. It is often difficult in seedlings to diagnose the trouble definitely, as the slight mottling and curl of the leaves may be due to other factors. As a rule, in young plants the leaf is rougher and a permanent mottling is observed, very slight in character, however, and not to be confounded with the mottling due to normal metabolic processes which occurs under certain conditions of growth. As the disease progresses, however, the leaf is found to be divided into light and dark green areas; in mild cases there does not appear to be any marked leaf distortion, and the light green areas sometimes verge on the yellow in color. The dark green areas apparently deepen in color with the intensity of the disease, and in extreme cases the leaf is much distorted and the dark portions appear blister-like, due to their more rapid growth. The leaves, as a rule, are much stiffer and thicker to the touch than are the normal healthy leaves. Sometimes in the later stages of the disease there are found dry, dead, brown patches or spots on the leaves, sometimes where the dark green areas were originally, but more often the light green portions show this extreme condition. Both the light and dark areas show abnormalities in structure; nevertheless, the light green areas are the more truly diseased ones, the dark green areas presenting different characteristics, and although showing changes in cell arrangement, etc., function more normally in many respects. Most investigators have held that the light green areas are the diseased portions of a leaf, but some have been of the opinion that the dark green areas are the diseased portions. As will be seen from the writer's experiments the former is the more correct view, as the increase in color intensity and the blistering of the dark green areas is due to the necessarily increased func- tioning thrown on these portions of the leaf. Occasionally a leaf may be distorted in such a manner as to present the appearance of being little more than a long filament consisting principally of midrib, with but very little leaf surface. This condition has been observed by the writer in some instances, but should not be confounded with a similar trouble occurring on tobacco in certain regions, which is of an unknown character but which is not the true mosaic as it is not infec- tious. This latter trouble has been noted particularly in Java, etc., as is reported by Peters ^ in his work on the diseases of tobacco. It has not been observed in tobacco fields in this region by the writer. It is thought that soil and moisture conditions are responsible at least partially for this disease. > Peters, L.: Krankheiten und Beschadigung des Tabaks. Mitteil. aus der Kaiser. Anstalt F. Land- u. Forstwirtschaft. Heft, 13: 64 (1912). 80 MASS. EXPERIMENT STATION BULLETIN 175. OCCUKRENCE. The mosaic disease has been known for years both in Europe and America, and may be said to be present everywhere that tobacco is grown. It apparently is a more serious disease in the tropics and in certain parts of Europe than it is in this country. In New England it has been known for some time, and, although present to a certain extent each year, is not of such great economic importance as in some other locahties. In Massa- chusetts it is found practically everywhere, and some years appears to be much more prevalent over widespread areas than in others. As a rule, however, the disease is not epidemic in character, and often only a com- paratively few plants in a field will be found affected. On certain fields, however, — and these most often are such as have been cropped to tobacco for many years without the practice of cover- cropping or rotation, — mosaic disease is present year after year, and a large percentage of the crop is always badly affected, the plants beginning to show the trouble in from three to four weeks after planting in the field. The prevalence of the disease in the field, aside from the special cases above noted, is apparently related in some way to conditions in the field during the growing season, or during the time the plants are in the seed bed. There is no question that a large percentage of the infection found in the field, exclusive of that appearing on the sucker growth after topping, or due to infection at the time of transplanting, is due to a primary infec- tion from the seed bed. While the disease as a rule is first noticed in the field some time after transplanting, very often the seedlings in the beds are affected. This is particularly true in the case of old or carelessly treated beds. It is often very difficult for the casual observer to identify the disease on the seedlings, as the macroscopic or visible symptoms are either very slight or lacking. In this way many plants are transplanted to the field by workmen without their being aware that they are diseased, and the disease becoming more pronounced in the later stages of growth, the infection is laid to the soil in the field, when in reality the infected soil of the seed bed is responsible and not the field soil. As has been stated, the closest examination of the seedlings is necessary to identify the trouble in the seed bed, particularly in mild cases of infection. From observations made repeatedly, not only on seed beds but also experimentally under controlled conditions in the greenhouse with soils from old beds, afterwards transplanting the seedlings to soil previously not used for tobacco, and using as checks healthy plants from new soil, the writer has come to the conclusion that at least 80 per cent, of our field infections come from the seed bed and do not originate in the field as is commonly supposed. Mosaic disease on tobacco leaves. (1) Older leaves showing mottling. (2) Leaves showing marked distortion and tendency to string leaf (on right). More or less indistinct types of mosaic disease, except (a), which is a young leaf with pronounced blisters. MOSAIC DISEASE OF TOBACCO. 81 Economic Importance. • It is very difficult to estimate the loss to growers due to mosaic disease, as the prevalence in different localities varies greatly, as also does the intensity of the attack in different seasons. The damage resulting from mosaic disease is twofold: first the plants when severely attacked are smaller and the leaves poorer in quality; secondly, the buyer, if he sees much mosaic in a field, will invariably cut the price a few cents a pound, as the leaves affected do not in many cases make a valuable wrapper and are much poorer in quality. The writer has observed certain fields where the loss would run into hundreds of dollars from this cause alone. The amount of damage done by late mild attacks when the plants are maturing, or appearing on the sucker growth after topping, is practically negligible, and, so far as can be learned, does not in any way injure the commercial leaf. It is always well to clean off the diseased suckers, however, as they present a very ragged appearance, and might injure the sale of the crop to a certain extent. There is no question but that during certain seasons the loss due to mosaic is quite large, but an exact estimate of this loss is difficult to obtain, owing to the many other factors involved. Infectious Nature of the Disease. That the mosaic disease is very infectious is well known, and a discus- sion of the detailed experiments on this point is not necessary. Experi- mentally it has been repeatedly shown that the juice from all parts of a diseased plant is capable of transmitting the disease, although it should be stated that the percentage of infection obtained from the root extract is considerably lower than that obtained from the leaves. A few of the results obtained are given in the foUomng table, however: — Table I. — Infectivity of the Juice from Different Parts of Diseased Plants, August, 1909. Part of Diseased Plant used (Plants from Field). Number of Healthy Seedlings inoculated. Number of Plants Dis- eased Three Weeks after Inoculation. Leaves showing disease, Control, Leaves showing disease, Control, Basal leaves (not showing disease) Control, Roots Control, Roots 10 (juice; needle pricks), 10 (distilled water; needle pricks), 10 (insertion of tissue into veins), 6 (insertion of healthy tissue into veins) 12 (juice; needle pricks), 5 (distilled water; needle pricks), 21 (juice; needle pricks), 7 (distilled water; needle pricks), 16 (insertion of tissue into veins), 82 MASS. EXPERIMENT STATION BULLETIN 175. Later experiments with the roots of other diseased plants gave similar low results. It is a very easy matter to infect seedUngs at the time of transplanting, and the writer has repeatedly seen many cases in the field which could only have been brought about by such infection. It is only necessary to get some of the juice from the diseased plant on to the hands to transmit the disease by handling healthy plants, the causal agent gaining entrance through the broken ends of roots, leaf hairs or broken and abraded leaf areas. In some of the experiments conducted relative to this point, a very high percentage of infection has been obtained. In one case where the juice from a diseased plant was very thoroughly rubbed on the hands, and 40 healthy seedlings immediately set, no care being used to guard against bruising the leaves, etc., 31 plants developed the disease in two weeks' time. In another experiment where 62 seedhngs were subjected to the same treatment, 30 plants developed the disease; in still another, series of 28 seedhngs, 21 developed the disease. Controls planted at the same time, handled vdth a hand rubbed with the juice of a healthy leaf developed the mosaic in only a few isolated cases. From the above it can easily be seen that great care should be exercised in the matter of handUng the seedlings, especially diseased seedlings. Contagious Nature of the Disease. In spite of the fact that it is held by some investigators that the mosaic disease is contagious, the -writer has never been able to satisfactorily dem- onstrate that it is. Under carefully controlled conditions in the green- house, guarding against accidental infection, it has been impossible to demonstrate the contagious nature of the disease. In isolated instances, indeed, apparent contagion has occurred, but it is beheved that these cases were due to accidental infection, as the percentage was so low, — less than 2 per cent., — and under the conditions the plants were subjected to, such as contact, spraying of the juice on leaves, etc., the percentage should have been much higher if contagion was to be held responsible. It is a fact that it is only necessary to break or rupture the trichomes or hairs on the leaf, subsequently spraying with diseased juice, to obtain infection, although this method does not give a very high percentage. It can easily be seen that such a rupture may be very easily brought about, and hence apparent contagion occur. As is stated elsewhere in this paper, insect and other carriers may also play a part in these so- called cases of contagion. MOSAIC DISEASE OF TOBACCO. 83 Pathological Anatomy. I Leaves. As might be supposed, there are great differences in structure between normal, healthy leaves and leaves affected with the mosaic disease. These differences are greatest, naturally, in badly diseased leaves. Woods ^ was one of the first to point out this fact, and his statements have been re- peatedly verified by the WTiter. He stated that the light colored areas were not normal, and that "this difference consists in the fact that in badly diseased plants the palisade parenchyma of the light colored areas is not developed at all. All the tissue between the upper and lower epider- mis consists of a spongy or respiratory parenchyma rather more closely packed than normal. In moderately diseased plants the palisade paren- chyma of the light area is greatly modified. Normally the palisade parenchyma cells of a healthy plant are from four to six times as long as broad. In a moderately diseased plant, however, the cells are nearly as broad as they are long, or at most, not more than twice as long as broad. As a rule, the modified cells of the leaf pass abruptly into the normal cells of the green area." From the above it can be seen that Woods was of the opinion that the light green areas were abnormal or diseased, and that the dark green areas were normal and healthy. The writer in his observations found this to be true in general, but occasionally the dark green areas showed a more closely packed parenchjTna than in normal leaves, and always the 'palisade layer was well developed and approached the normal in character. The development or non-development of the palisade layer, as Woods hinted, is dependent on the degree of severity of the disease. The lighter the attack the less are the palisade cells and parenchyma tissue altered, and vice-versa. This the writer found to be true in so far as anatomical differences were concerned, but as will be noted later, the dark green, apparently normal, healthy tissue contained some of the infective agent of the disease. The structure of the dark green areas varies only slightly from that of the normal leaf, with the few exceptions above noted, and may be con- sidered normal in character. The writer has sectioned many leaves in aU stages of disease, and these structural differences have always been found to occur in the manner above indicated. These differences in structure have been taken up more or less in detail, as some investigators have held, and still hold, that the dark green areas are the part diseased, and that the light green areas are normal, inasmuch as they approach the normal leaf in color in many cases, most probably basing their assumption on the fact that the dark areas form blister-Uke growths and are sometimes darker in color than normal leaves. No one recently appears to have > Woods, A. F.: Inhibiting Action of Oxidase on Diastase. Science, n. s., XL, No. 262, 17-19 (1900). 84 MASS. EXPERIMENT STATION BULLETIN 175. investigated the structure of the dark and light areas carefully in the case of the tobacco, except Woods. It was to verify Woods' statements that the writer took up this phase of the matter, and mention will again be made of it in connection with the biochemistry of the leaf. There can be no doubt as to the correctness of Woods' contention that the light green areas are abnormal and diseased; but that the dark green areas are not diseased, at least in certain cases, cannot be so definitely stated. Their structure may be somewhat modified by the increased functioning thrown on the healthy cells. On the other hand, it is fallacious to state that the Ught green are the healthy, and the dark green are the diseased, portions of a leaf. Plates III. and IV. show three cross sections from leaves. III. showing the cross section of a healthy leaf; IV., that of the light green area of a diseased leaf and of a dark green area of the same leaf. It wiU be noted that the palisade layer is practically suppressed in IV. (1), or the Ught green portion, while in IV. (2) the palisade layer approaches the normal in character except for a closer packing of cells in general. Milder cases of diseased leaves vary between these limits. These figures are from camera Incida drawings of material killed and fixed in medium chrom- acetic acid. In the material used the normal leaf section is somewhat thicker than those of the diseased leaf, but for comparative purposes is perfectly satisfactory. Stems. The anatomical differences in the leaves of healthy and diseased to- bacco plants have been given in the preceding paragraphs, and as it was desired to carry the investigations further to cover the entire plant, re- peated examinations were made of both cross sections and longi-sections of stems of plants in various stages of disease, and also of healthy, normal plants grown both in the field and greenhouse. It should be stated at this point that occasionally the writer has observed on the stems of some badly mosaicked plants a mottling, or, rather, a streaking of the stem, a portion of which would be darker green than the remainder, and this is without question a manifestation of the mosaic disease. Sections of such stems, however, showed absolutely no variation in structure from those of normal plants, and in no case, although the examinations covered an extended period of time, was it possible to show any structural difference between the stems of badly diseased mosaic plants and those of healthy plants of the same age. Examinations of the stem close to the terminal apex of the plant revealed the same conditions as those of other parts of the stem. No differences were observable except in the matter of size and arrangement of cells, such as would naturally be expected when we take into consideration the differences in size and development of the stem near the terminal apex and progressively towards the base. Section through normal tobacco leaf : (a) epidermis; (6) palisade cells; (c) parenchyma tissue. oodUBD^PrPv^ Sections tlirough mosaic-diseased leaves. (1) Light green area: (a) epidermis; (6) palisade cells; (c) parenchyma tissue. (2) Dark green area: (a) epidermis; (6) palisade cells; (c) parenchyma tissue. MOSAIC DISEASE OF TOBACCO. 85 Roots. In the same manner roots of mosaicked and healthy plants were ex- amined at various times under all conditions of growth and severity of disease, and in every case the root structure was found to be normal. Root tips from healthy and diseased plants showed absolutely no differ- ences in structure. It might be anticipated that, as the disease first mani- fests itself in the di\dding cells of the leaves, there might be a supple- mentary differentiation, so to speak, of the meristematic tissue at the growing point of the root, functioning co-ordinately with that of the aerial part of the plant. No such condition was observable, however, and, so far as the writer has been able to find, there is no manifestation of local cell disturbances in the root such as are found in the leaf tissue. The causal agent of the disease, however, as has previously been noted, is without question present in all parts of the plant, and it should not be stated that it is confined to those parts which show structural variation. Fungi and the Mosaic Disease. Almost from the first it has been established that no fungi are asso- ciated with the cause and development of the mosaic disease of tobacco. In no case where careful work has been conducted under conditions elimina- ting the possibility of accidental infection has any fungus been found associated with the trouble. Cultures of fungi obtained occasionally from leaves have always been traceable to careless manipulation or ex- ternal infection, and the fungus obtained failed to infect healthy plants, no matter what methods of inoculation were used. The writer has occasionally obtained cultures on various media such as oat agar, tobacco leaf agar and prune agar, from the tissue of the so- called "rusted" spots which are sometimes a late manifestation of the last stages of the mosaic; but, as with previous investigators, it was found impossible to infect healthy plants from these cultures, either by needle pricks, spraying, or inserting the fungus into incisions in the leaf or stem. These experiments with fungi were made merely to demonstrate to the writer's own satisfaction that they could not be the causative agents of the disease, as there might be a possibility that they were latent in the plant during the earher stages of the disease and only developed super- ficially during the later stages. According to Jenkins^ and others these rusted spots which are some- times observed are primarily caused by a drying out and disintegration of the cell tissue, which has been weakened by the disease and which thus forms a suitable medium, under favorable conditions, for the develop- ment of secondary fungi and micro-organisms. This view is also held by the writer as a result of observations extending over a series of years. » Jenkins, E. H.: Studies on the Tobacco Crop of Connecticut. Conn. Agr. Exp. Sta. Bui. No. 180, p. 56 (1914). 86 MASS. EXPERIMENT STATION BULLETIN 175. Bacteria and the Mosaic Disease. Among the many theories advanced regarding the cause of the mosaic the chief one for some time, particularly among the earlier investigators, was that of bacterial infection either through the agency of infected soil or otherwise. Mayer, ^ in his rather extended study of the disease, came to the conclusion that it was caused by bacteria, but was unable to isolate the organism. Prilleux and Delacroix ^ claimed to have found an organism associated with the mosaicked leaves, but their descriptions leave one in doubt as to whether they were working with the true mosaic disease or not. It is very probable that they were dealing with another disease which occurs in France, but which is somewhat different from the mosaic disease. The next important work on the bacteria supposedly connected with this disease was done by Iwanowski. ^ He isolated several organisms from the juice of diseased leaves, and by reinoculation was able to cause infection, but only in a very small number of instances. This he explains by a probable attenuation of the organism when grown on artificial media. Hunger,* in a very critical review of the bacterial theory, stated that he was unable in any way to substantiate the findings of Iwanowski, and that although he observed certain bodies in the cells, he was not able to classify them as either bacteria or plasmodia, as they disappeared after heating with phenol chloral hydrate, while the rest of the cell contents were unaffected. More recently Allard^ has advanced the opinion as a result of his investigations that the disease is parasitic in nature but does not attempt to discuss the character of the parasite, and apparently has made little attempt to demonstrate anatomically the presence or absence of bacteria. Hunger's work is probably the most satisfactory of its kind along this line. The writer has made examinations of diseased plants, sectioning leaves, stems and even the roots, but has never been able satisfactorily to demonstrate the presence of bacteria in the tissues. In this work a variety of stains were used, chief of which, however, were Ziehl's carbol fuchsin and Heidenhain's iron hsemotoxylin. It is to be noted in this connection that all investigators have apparently confined their studies to the leaves or part of the plant in which the disease showed itself, and very few attempts, if any, have been made to study the question of the possible presence of bacteria in tissue far removed from the diseased portions. In view of the fact that the juice from all ' Mayer, A.: Over de in Nederland dikwijk voorkomende Mozaikziekte der Tabak. Land. Tijdschr. (1885). » Prilleux, E. E. and Delacroix, G.: Maladies bacillaires de divers v6g6taux. Ck)mpt. Rend. Acad. Sci. Paris, 118: 668-671 (1894). » Iwanowski, D.: Uber die Mosaikkrankheit der Tabakspflanze, Zeit. f. Pflanzenkrank, 13: 1-41. pi. 1-3 (1903). * Hunger, F. W. T.: Untersuchungen und Betrachtungen uber die Mosaikkrankheit der Tabakspflanze. Zeit. f. Pflanzenkrank, 15: 257-311 (1905). ' AUard, H. A.: Mosaic Disease of Tobacco. U. S. D. A., Bur. Plant Ind. Bui. No. 40 (1914). MOSAIC DISEASE OF TOBACCO. 87 parts of a diseased plant will cause infection, it would be natural to sup- pose that if bacteria were the causal agent, it should be possible to demon- strate their presence in the different parts of a diseased plant. This has never been done, and in the writer's study of the anatomy of diseased plants it has never been possible to demonstrate the presence of bacteria in the different tissues. The writer has many times attempted to obtain cultures of bacteria from diseased tissue, and in some cases cultures of organisms were obtained on various media, but they proved in every case to be secondary in character, and were not capable of reproducing the disease. In the hght of all later investigations the evidence points over- whehningly to the absence of bacteria, in the present-day sense of the term, as the causal agent of the disease. Dissemination Agents. Insects. The fact that many fungous and bacterial diseases are often transmitted by insects, as well as other agents, has been long known and thoroughly established, but until Allard (loc. cit.) called attention to the fact that the mosaic disease could be carried by aphids, and one in particular (Macrosiphum tobaci Perg.), nothing had been published on this phase of the matter. Allard in well-controlled experiments demonstrated beyond a reasonable doubt that the disease was so communicated. CUnton (loc. cit.) made a few observations on the infection of healthy plants by the tobacco horn worms which had been feeding on diseased leaves, but was unable to demonstrate that the disease could be so transmitted either by the excreta ejected by the worm or by its biting and feeding on the healthy plants. His results were negative in the few experiments made. Observations made in the field during the progress of the writer's work have not shown conclusively that the disease is communicated by biting insects, such as the tobacco horn worm, grasshoppers and a small black flea beetle of more or less common occurrence in our fields. Occasionally aphids have been found infesting the leaves of tobacco in our fields, but so far as could be judged were present in too small numbers to be active agents in transmitting the trouble. As a rule, comparatively few aphid infestations are found in our tobacco fields. In the greenhouse during several winters tobacco plants grown in benches were infested with white fly, and it was at first feared that they might carry the infection from diseased to healthy plants in the same benches. This, however, was not the case, and it has never been possible to demon- strate positively that the white fly is an active agent in the spread of the disease. This insect is, of course, of rare occurrence in our fields, but may possibly do damage in the south. It apparently feeds and breeds freely under greenhouse conditions on the underside of the leaves. In order to ascertain more definitely the possibility of infection by these insects, adult whit^ flies from badly mosaicked leaves were carefuUy re- 88 MASS. EXPERIMENT STATION BULLETIN 175. moved and placed on the underside of the leaves of tobacco plants, en- closed in a small cloth-covered cage, and were allowed to remain on the tobacco leaves of the plants in these cages for four days. After this length of time the plants were removed from the cages and placed on a bench at some distance from those containing mosaicked plants badly- infested with white fly. On none of the plants did mosaic develop. The plants were later placed in close juxtaposition to those in the original benches, which, as indicated, were at this time heavily infested with the white fly and badly mosaicked, but although the plants remained until maturity, no cases of mosaic developed on them in spite of a heavy infestation of white fly. The writer's observations on the activities of aphids as carriers of in- fection have not been so extensive as in the case of the white fly, as only minor infestations of the former occurred in the greenhouses; and the indications pointed to the fact that, although there were a certain number of aphids present on the leaves of both healthy and diseased plants, so far as was observable no cases of infection from this source arose, as the mosaic developed only on an average of 1 case out of 30, except on the plants which were artificially inoculated with the juice from diseased leaves. It should be stated, however, that aphids present in the green- house were not of the same species as that tmder consideration by Allard, and there is no reason to doubt the accuracy of his observations on the species tabaci Perg, The question of insects as carriers of the mosaic disease as well as of many other diseases is still open to discussion; and it may be that in the case of the mosaic a very heavy infestation of aphids is necessary to bring about a successful infection of healthy plants, as the amount of active infective material carried by such insects would in any case be very small, and accumulative effects of the activities of several insects might be necessary to introduce a sufficient amount of the active principle to trans- mit the disease. Workmen. It has been shown that the disease is highly infectious and it has also been proved repeatedly by many investigators that it is very easy to transmit the disease to healthy plants at the time of transplanting. A workman handling diseased seedlings, and subsequently healthy ones, will very often infect them. Several instances of this have come to the writer's attention, every other plant for some distance in a row developing mosaic within a month after transplanting. The same condition has also been observed by Clinton {loc. cit.) in Connecticut, and can only be ex- plained by the fact that the workman's hands were infected through handling a diseased plant, and the infection then transmitted to healthy ones, the causal agent being introduced through broken tissue of the leaves or roots of the seedlings. This method of transmission is particu- larly striking in the above case, as the same individual plants every other plant in a row when working the ordinary planter. Of course, there MOSAIC DISEASE OF TOBACCO. 89 have been many cases where eveiy plant for some distance in a row has developed mosaic, but this might be explained if it is assumed that both workmen had handled diseased seedlings, or if a number of plants in the lot were diseased. In time, the causal agent becomes so attenuated that infection ceases, and the remainder of the row remains healthy. Experimentally, this method of transmission has also been shown to be possible, and a high percentage of infection has been obtained. In one experiment, after thoroughly rubbing the hands with the tissue of a dis- eased plant, and then pulling and transplanting healthy seedlings, over 80 per cent, of the transplants became mosaicked within a month. Only a relatively small number of seedlings in this instance were treated in this way, however, the total being 28, of which 24 developed mosaic symptoms within three weeks. Another manner of transmission is by cultivation. If some of the sap from a diseased plant comes in contact with the tools, etc., employed, there is a possibility that the infection might be carried to healthy plants by this means, but the percentage of infection of this character is probably very low in actual field practice. The workmen when budding and topping are very often carriers of infection, as they are not as a rule careful to leave untouched the plants showing mosaic symptoms but take the plants as they come, and thus spread the disease to many healthy plants. This method of dissemina- tion has been very often observed, and perhaps is the most fruitful source of infection in the field. The subsequent new growth will almost in- variably be mosaic in character, as will also the suckers developing later. The amount of damage to the marketable leaves, however, providing the suckers are removed, is very sHght, if any, and cannot be said to injure the leaf in any way, at least in so far as our observations bear on this point. If the suckers are left, however, the plants present a ragged ap- pearance, and the mosaic on the suckers is quite noticeable, and might injure the sale of the crop at the price it ought to command. Seed. The causal agent is not carried by the seed, and seed from mosaic plants has never produced a larger percentage of mosaicked seedlings than seed collected from healthy plants, when germinated and grown under the same conditions. It is difficult to conceive of this, as it has been shown by Allard (loc. cit.) that the tissues closely enveloping the seed in the pod are capable of causing infection; but the writer has saved seed from badly mosaicked plants for three successive years, and the seedlings from such seed showed no signs of the disease, unless infection was pro- duced artificially through some external agency. It should be pointed out, however, that there is the possibility that the vigor of the seed from mosaicked plants may be less than that from healthy ones, and consequently the plants developed from such seeds, being weaker, might be more susceptible to the factors active in the product on of 90 MASS. EXPERIMENT STATION BULLETIN 175. mosaic symptoms. It is impossible to make a definite statement on tbds point, however, as the writer has not been able to gather sufficient data over a series of years to prove or disprove it. Fertilization in Relation to Mosaic Disease. It has been repeatedly shown by many investigators (see historical summary) that a lack of plant food alone will not suffice to produce the mosaic disease, and the writer has also, in connection with the tomato, shown that an excess of nitrogen, potash, phosphoric acid and lime will not produce nor intensify the disease. ^ The same has been found to be true for tobacco. In our experiments on tobacco, the method made use of was to add to each pot the proper amount of a complete tobacco fertiUzer (in this case applied at the rate of 3,000 pounds per acre), and then to add an additional amount of nitrogen, potash and phosphoric acid in quickly available forms, equal to that already present. No mosaic was produced in any case, although where the amount of nitrogen was trebled a rather peculiar malformation of the leaves was observed which at first sight might have been mistaken for mosaic symptoms. All inoculations failed to take, however, and the trouble therefore could not have been the true mosaic. It has been held that liming would lessen the prevalence of the disease, but the writer's observations and experiments do not bear out this state- ment. Under field conditions this may be the case in certain seasons, but continued observations from year to year on heavily limed areas show no appreciable lessening of the number of mosaicked plants. SeedUngs and plants grown in the greenhouse in soil kno'mi to be heavily infected in- dicated the same results, as also did the work on new soil with mosaicked seedlings. Here lime was applied in varying amounts at the rate of from 500 to 6,000 pounds per acre. No appreciable effect on the mosaic disease was observable. The results obtained are given in the following tables : — [New soil, lime, mosaicked seedlings.] Lime (Pounds per Ache). New Soil in Pots (Number planted with Mosaicked SeedUngs). Number of Plants show- ing Recovery One Month after Planting. 500 1.000 2,000 4,000 6,000 No lime (check) 40 28 34 12 10 5 - 1 Twentieth Annual Report, Mass. Agr. Exp. Sta. (1908), p. 140. MOSAIC DISEASE OF TOBACCO. 91 The lime was applied to this new soil, in the different amounts indicated, one week previous to the setting of the plants. No appreciable differences were observable in the subsequent growth as regards intensity of mosaic symptoms, all the plants being comparatively evenly mosaicked. There was not a single case of recovery. Table III. — Effect of Liming on Mosaic. [Infected seed bed soil, lime, seed.] Lime (Pounds pee Acre). Per Cent Infection (Seedlings Twelve Weeks Old). 600 12.0 1,000, 18.4 2,000, 9.8 4,000 21 0 6,000 No lime (check), . 13.7 The lime was here applied to a soil which was heavily infected, and the seed sowed very thinly in the flats containing the various amounts of lime and soil. The seedlings were allowed to grow in the flats until they were counted. They were naturally crowded somewhat, but were free from insects during the period of growth. It is possible that some infection may have occurred, however, but there are very strong indications that liming had no beneficial action in lessening the disease. As the results are so variable the matter cannot be considered as absolutely settled, but certainly no consistently favorable results were obtained in this experi- ment from the use of lime. Effect of Colored Light on Mosaic Disease. In connection with work on the mosaic disease of tobacco it had long been noted, in that section of the Connecticut Valley where the crop was grown under shade, that the plants appeared in general to be much less affected with the mosaic disease than were those grown in the open. This fact has already been noted by Sturgis ^ in Connecticut. Investi- gations were outUned, in conjunction with other work on this disease already under way, relative to a study of the effects of various light conditions on the intensification or reduction of the disease. While the writer's preliminary work was in progress, Lodewijks ^ published a paper » Sturgis, W. C: On the Effects, on Tobacco, of Shading and the Application of Lime. Conn. Agr. Exp. Sta. Ann. Rept., 23, 252-261 (1899). * Lodewijks, J. A., Jr.: Zur Mosaikkrankheit des Tabaks. Rec. Trav. Neerlandais, Vol. 7, 107-129 (1910). 92 MASS. EXPERIMENT STATION BULLETIN 175. on the effects of colored light on mosaic-diseased plants, and as a result of his experiments stated that a cure was effected by blue light, red light diminished the disease, and suffused light checked it somewhat. In brief, his methods of experimentation and conclusion were as follows : — The diseased leaves of a plant were covered with a cloth hood of the desired color, of a sufficient size to allow ample room for growth. The apparently healthy basal leaves were left uncovered and exposed to the normal daylight. After a time the hoods were removed, and it was found that in the case of the plants exposed under the blue hood a cure was ef- fected; those exposed under a red hood showed a duninution in the severity of the disease; and in the case of plants exposed to the suffused light alone the disease was somewhat checked. The cloth used for the red and blue hoods was a rather coarse cotton material similar to that used for making flags. Several investigators had noted the apparent beneficial effects resulting from growing diseased plants in suffused light, but Lodewijks was the first to really study the effects produced by colored light, although Bauer appears to have made some observations on this point. As in no case could the writer find that Lodewijks in his work had reinoculated from the apparently cured plants to healthy ones, to prove the presence or absence of the causal agent, and as it is often present and active in apparently healthy leaves of diseased plants, as has been shown many times, it was thought necessary to settle the point as to the presence or absence of the causal agent in plants treated as in Lodewijks' work. Method. — The method of treatment of diseased plants was in every way similar to that employed by Lodewijks as to texture of cloth, methods of covering the plants, etc. The cloth covers were held away from the plant by means of wire hoops, and the cloth was tied around the stem of the plant below the diseased leaves. Plate V. shows a hood in place over a field-grown plant, and gives a clear idea of the arrangement of the hoops, etc. The cloth used was a coarse grade of cotton, and the colors were cad- mium orange, ox-blood red and induHn blue. ^ Plants showing well-developed symptoms of the mosaic disease w^ere selected for the experiment, none of which had less than four character- istically diseased leaves, the lower remaining leaves apparently healthy. The hoods were placed over the diseased leaves as above noted, and left on for the required time, in most of the experiments twenty to thirty days. At the end of this period the hoods were removed and the plants carefully examined for visible symptoms of the disease. Two leaves from the upper (i.e., the part under the hood) portion of the plant were removed under absolutely aseptic conditions, the juice expressed and healthy plants inoculated therewith by means of glass capillaries inserted just below the terminal leaflets. Control inoculations with distilled water and boiled juice were also made at the same time. The plants, after the > Ridgway, Robert: Color Standards and Color Nomenclature. Washington, D. C. (1912). Effect of colored light on mosaic disease; showing method of attaching lioods over leaves. MOSAIC DISEASE OF TOBACCO. 93 removal of the leaves above mentioned, were allowed to grow to matm-ity under normal light conditions. Most of the experiments were carried on in the greenhouse, where tem- perature and other conditions were under more direct control than in the field, although field experiments later repeated gave the same results, but, of course, in this case there was a greater chance of subsequent infection through careless handling, insect attacks, etc. In the following paragraphs are tabulated the results of a typical series of experiments relative to the effects of light on mosaicked plants. Experimental Data. Red Cloth. — Three plants were covered with the red cloth hoods for twenty days. The covers were then removed, and in all cases visible symptoms of the disease were still present, although the color variation between light and dark green areas was not so marked as at the beginning of the experiment. All the new growth, in addition to the leaves diseased at the time the hoods were put on, also showed the mottling distinctly. A week after the hoods were removed all the plants still showed the disease in undiminished severity. Healthy plants inoculated with the juice from the leaves confined under the hood became diseased in from nine to eighteen days' time. Controls inoculated in the same manner with boiled juice from the same leaves, and with distilled sterile water, remained with very few exceptions healthy. Table IV. gives the results of the inoculation experiments in one series. Table IV. Result of Inoculation with Juice from Plants grown under Red Hoods. Plant No. Number of Healthy Plants Inocu- lated with Juice from Leaves of Treated Plant. Number of Inoculated Plants show- ing Mosaic at the End of Eighteen Days. A-1 B-1 6 7 4 6 6 C-1 4 Controls inoculated with boiled juice, 10; diseased in eighteen days, 1. Controls inoculated with distilled, sterile water, 10; diseased in eighteen days, 0. From the above results it may be seen that there was a diminution in the color variation in diseased leaves;, it was not of a permanent character, the plants all showing the disease in undiminished severity again after a short exposure to normal dayhght. The causal agent of the disease was still highly infectious. 94 MASS. EXPERIMENT STATION BULLETIN 175. In a second series the hoods were allowed to remain over the plants for tliirty days, as it was thought that a twenty-day exposure might have been too short, but no appreciable variation in the results was obtained as a result of the longer treatment. Orange Cloth. — In this series two plants were covered with orange hoods for a period of thirty days. On removing the covers it was found that the visible symptoms of the disease were, if anything, intensified. The gro"\\i:,h was somewhat more spindling, the leaves narrower, and the light and dark green areas very clearly defined. Infection was produced from both plants by inoculation into healthy plants. The causal agent was very active and highly infectious. Blue Cloth. — The diseased parts of three plants were covered with blue cloth hoods, as in the preceding experiments, for a period of twenty-five days. The covers were then removed and a careful examination of the leaves made. On plants A-2 and B-2 no visible symptoms of the mosaic disease could be observed, although a slight tendency towards curling was noticeable on a few of the leaves. The leaves were all uniformly light green in color, and aside from this, appeared normal. Plant C-2, however, showed on two leaves a slight mottling. Two weeks after the hoods were removed, plants A-2 and B-2 did not show any marked symptoms of the mosaic disease other than a faint mottling of a few leaves, not sufficient, however, to seriously injure the leaf. Plant C-2 developed mosaic again in the same length of time, but not as seriously as before the treatment. It may be that the mottling on A-2 and B-2 was due to the maturing of the plant, although this mottling is usually distinctive enough to be readily differentiated from that caused by the mosaic disease. Healthy plants inoculated with the juice of leaves from plants A-2, B-2 and C-2 contracted the disease almost without exception. Controls inoculated with boiled juice failed to develop the disease. Table V gives the results of the inoculations. Table V. Results of Inoculations with Juice from Plants grown under Blue Hoods. Number of Number of Healthy Inoculated Plants show- Plant No. lated with ing Mosaic at End of Leaves of Eighteen Treated Plant. Days. A-2. . . 8 5 B-2, . . 4 4 C-2. . . 10 9 Controls inoculated with boiled juice, 6; diseased in eighteen days, 0. Controls inoculated with sterile distilled water, 6; diseased in eighteen days. MOSAIC DISEASE OF TOBACCO. 95 The above results show that when blue light is used there is a suppres- sion of leaf color variation more or less permanent in character, the treated plants, with one exception, showing no typical symptoms of the disease for at least two weeks subsequent to the removal of the hoods. It cannot be said, however, that the disease was controlled, as inoculation of healthy plants with the juice from these leaves produced the disease in nearly every case. The causal agent of the disease was still very active in the apparently normal fully recovered leaves, and was highly infectious. Discussion of Results. — The results of these experiments do not agree entirely with those obtained by Lodewijks, particularly in the case of action of the blue light, inasmuch as the plants covered with the blue hoods, although showing an apparent recovery from the mosaic, still contained the causal agent of the disease, and by inoculation with the juice expressed from these plants into healthy plants the disease was again produced in practically all cases. It should be noted that the visible symptoms of the disease were suppressed, the reason for which may be as Allard {loc. cit.) suggests in his work on the mosaic disease of tobacco. He states, with respect to Lodewijks' observations, "If the malady in question was true infectious mosaic disease, one is inclined to believe that covering the young plants temporarily reduced the color contrasts of the mottled areas. These changes may have led Lodewijks to conclude that a partial or a complete cure had been effected in his experiments." It might be inferred from the above that on the removal of the hoods exposing the plants to normal daylight, they would soon regain the color contrast, but this is not entirely so in the case of the blue light, as has been shown. The apparent recovery, therefore, is- not entirely the result of a suppression of color contrast due to the action of blue light on the leaves as suggested by Allard, but is undoubtedly so in part. It is evident that the treatment of plants as above recorded does not destroy the causal agent of the mosaic disease, whatever may be its char- acter, the treated leaves apparently still containing the causal agent, very probably in the same manner as do the parts of a plant which do not show visible symptoms of the disease, as the stem, lower leaves, roots, etc., the juice of which is often highly infectious. It would appear from the re- sults that the new terminal growth subsequent to the removal of the hoods would develop the trouble, and this was the case in plant C-2, but not apparently so with plants A-2 and B-2. Lodewijks' opinion, therefore, that in the plant a "virus" and "anti- virus" are present, and that certain abnormal conditions cause the "virus" to be produced in excess, bringing about a mosaicked appearance, while if the "anti- virus" is produced in excess, immunity is secured, will hardly hold, as it is clearly shown that even after apparent cure the causal agent is present and active. It is significant to note that under the influence of blue light both assimilation and starch formation are decreased, thus bringing about a 96 MASS. EXPERIMENT STATION BULLETIN 175. partial starvation, as it were, not, however, serious enough to reduce greatly the total starch formation and assimilation of the whole plant; while at the same time the clilorophyll production is very little changed if a comparison of the color of the normal and treated leaves can be taken as a basis of such a comparison. This latter fact has already been noted by Lodewijks in his work on the disease. It is, therefore, indicated by the results obtained in the preceding ex- periments that the different colors have little or no effect on the causal agent of the disease, but in the case of the blue there is a strong depres- sion of the macroscopic symptoms of the disease. Biochemical Studies. Enzyme Activities in Healthy and Diseased Plants. The study of enzymes in relation to diseases, particularly those of a so-called physiological nature has not been extensively gone into as yet by investigators, but it is believed that a study of their activities and reactions should be made, not only in the case of physiological troubles, but also those caused by fungi and bacteria, as it is the writer's firm belief that the activities of a large number of the fungi, and their effects on the respective hosts, are in a great measure due to the action of either exo- enzjines or endoenzymes produced by the fungi concerned. There is a possibility that the future may show a great advance in the study of host resistance, etc., when the conditions under which enzyme activity in fungi and bacteria takes place are better known, and plants may pos- sibly be bred to a condition of producing either a sap in which these activities cannot take place, or wiU produce anti-enzymes which will inliibit the activities of the enzymes contained in the respective fungi. Although many have made a study of this disease, very few have con- cerned themselves with the question of the enzyme activities; among the first to make mention of this phase of the question was Woods {loc. cit.), who found that the enzymes designated as peroxidases were at least dif- fusable, and occurred apparently in larger amount in diseased leaves than in healthy ones; also that their action was twice as strong in the Ught green areas as in the darker portions of the leaf. Koning (loc. cit.), as a result of his investigations, came to the conclusion that the disease was caused by a certain enzyme, which he stated to be oxidase, and the action of wliich he described. !Ee beUeved that it was formed in the plant under certain conditions. HeintzeH also found oxidizing enzj^mes present wliich were more active, if not present in greater amounts, in diseased plants than in the normal plants. Woods later (1902), in his work on the mosaic disease, verified liis former observation, and stated further that the diastase activity was much inhibited in the case of diseased plants. He attributed the lessened diastase activity to the presence of excessive ' Ileintzcl, K.: Contagiose Pflanzcnkrankheiten ohne Mieroben mit besonderer Beriicksichti- gung der Mosaikkrankheit der Tabaksbliitter. Erlangen, 46 p., 1 pi. (Inaugural Dissertation) (1900). MOSAIC DISEASE OF TOBACCO. 97 amounts of oxidizing enzymes, and showed experimentally that diastatic action is inliibited by the presence of oxidizing enzj-mes. This is the only work that has been accomplished up to the present time, so far as relates to a study of the enzyme activities involved in tliis disease. Only two enzymes have been considered, namely, the oxidase and diastase, and it should be stated that in the light of later developments in the de- termination and estimation of enzyme preparations and activities the results obtained in some cases might well be open to some criticism. Loew,^ wliile worldng with tobacco, discovered the presence of an enzyme wliich he called catalase, but he made no observations relative to its activities in the case of mosaic-diseased plants. The results of the writer's studies on enzyme activities of healthy and mosaic plants are detailed below. Method. — In the experiments here detailed the enzymes under dis- cussion were studied, in so far as was possible, (1) with regard to their presence or absence in (a) leaves, (b) stems and (c) roots of healthy and diseased plants (tliis was considered necessary, as it has been found that, irrespective of the parts shoTsdng visible symptoms of the disease, the sap from all other parts also is capable of transmitting the trouble) ; (2) with regard to the age of the plant; (3) with regard to the growth of the plant under different conditions. These will be discussed in detail under their respective sections. The methods employed for the estimation were for the most part those which by experience have been found satisfactory, and in the main give quantitative results; in some cases the results are more or less qualitative in nature, owing to our present insufficient knowledge of the methods of isolation and action of the enzyme involved. It should be stated that plants used in the experiments were both field and greenhouse grown, but no essential differences in results were obtained from the two series. The individual experiments will not be given in detail, but as the determinations of any given series were made in every case in the same manner, only average results with the maximum and minimum readings will be given. The experiments are, however, described in sufficient detail to enable those interested to follow the methods em- ployed closely enough to check up the work of the writer. Catalase (leaves). — A comparison was made of the catalase activity of healthy and diseased leaves, as it had been noted as early as 1908 by the writer that there was apparently a great difference between the cat- alase activity of healthy and mosaic-diseased tomato leaves, and later the same was found to be true in the case of tobacco. At that time only rough determinations were made, but since then the ^vriter has made hundreds of determinations, the results of which have borne out the ob- servations made then, and indisputably established the fact that there is a wide difference in the catalase activity of healthy and diseased leaves. » Loew, O.: Catalase: A New Enzyme of General Occurrence, with Special Reference to the Tobacco Plant. U. S. D. A., Bur. Plant Ind., Bui. No. 68 (1901). 98 MASS. EXPERIMENT STATION BULLETIN 175. In all the experiments fresUy collected material was used, and the determinations made almost immediately after collection. The usual pro- cedure was as follows : — A weighed amount of leaf was ground thoroughly with a weighed amount of acid-washed sand and a certain volume of double distilled water, and the whole washed into the apparatus with sufficient double distilled water to bring the volume up to the standard volume used in the particu- lar series in question. This, of course, gave to each flask a standard con- stant dilution value. To this mixture was then added a like volume of 1 per cent, solution of Merck's perhydrol, thus making the H2O2 concen- tration of the total mixture .5 per cent. The amount of oxygen liberated in ten minutes was arbitrarily taken as the measure of enzyme activity. Several different forms of apparatus were used, but for large amounts of leaf any ordinary water displacement method was found to be very sat- isfactory. (Care should be exercised where this mode of analysis is used, to take into accoxint the absorption of oxygen by the water.) In making determinations where the amount of material was very small, the ap- paratus designed by Lohnis for use in milk examinations was found to be more convenient. Practically all determinations were made at tempera- tures ranging from 17° to 23° C. The action of the catalase is much accelerated by shaking, as pointed out by Loew, and each test was shaken imder exactly similar conditions in all the determinations made. It was found necessary to use this method for the determination of the catalase activity, as any method involving titration, such as the permanganate method, was unsatisfactory, due to the reaction of certain constituents in the tissue with the reagents. Table VI. shows the relative amounts of oxygen developed in normal tobacco leaves, and it is to be noted that the catalase of the dark green leaves was much more active than that of the light green leaves. This was found to hold true, to a certain extent, for light and dark green leaves even on the same plant. The basal leaves of older plants, which in some cases were almost mature, and of a Ughter color than the middle and upper leaves, developed in every case relatively less oxygen. This was partic- ularly true in the case of Havana tobacco. Broadleaf did not show such a wide divergence, but it should also be stated that in the Broadleaf plants employed in the determinations the basal leaves did not show any great color difference. As will be noted, some of these experiments were made with plants grown under field conditions, but a greater number were made with plants grown in the greenhouse, under control conditions. MOSAIC DISEASE OF TOBACCO. 99 ^ o ^ 5 . . (U i "o 5 g d 1 1 >> bO i a 1 1 'c3 0 8 2! ^ i-2 •^ ^Q J o a - a >• 6 1 i ^" 5 5 ,M ^■■ 5 .is •a 5 " hJ hJ Q " tj p p a a i 00 U5 t- 00 (M O en CO t^ CO CO CO o -< S^ <) a B ^^ Q o o >o o U3 CM U3 1-1 o 8 s s " " I^o i g S s s s s s g !3; S . bO M M '"' Si c c W i > i 8 1 1 1 i 1 1 <: ffl o P w fe c W « CO 100 MASS. EXPERIMENT STATION BULLETIN 175. These results show that the catalase activity varies somewhat even in healthy plants, dependent upon age and also, apparently, on the general condition of the plant. It shows clearly, also, that in plants of approxi- mately the same age the catalase activity varies somewhat between plants with dark green leaves and those with light green leaves. Even on the same plant tliis holds true, as can be seen from the results tabulated below. Table VII. — Catalase Activity of Light and Dark Leaves from Same Plant. [Plants nearly mature; procedure as in Table VI.] Plant No. Number of Determina- tions. Light Leaves, Cubic Centi- meters of Oxygen developed (Average). Dark Leaves, Cubic Centi- meters of Oxygen developed (Average). Bi Xii 104 Al7 4 3 3 6 51.8 62.0 71.4 58.1 119.8 125.5 93.7 79.3 An, examination and determination of the catalase activity in diseased leaves shows that the amount of oxygen developed is relatively much less than in the case of healthy leaves. In the table below are given some of the results obtained from diseased leaves. In these experiments the leaf tissue was used without reference to the light and dark areas of the in- dividual leaf. It is significant that the activity is very much less than in healthy leaves. All the plants used in this series were badly diseased. It should be stated that in apparently mild cases of the disease the variation from the normal catalase content was not so great. The results shown here can hardly be compared with those given in Table VII., as the plants were not in some cases of the same age, nor were they grown at the same time. Table VIII. — Catalase Activity in Diseased Leaves. [Plants badly diseased; procedure as in Table VI.] Plant No. Number of Determina- tions. Cubic Centi- meters of Oxygen developed (Average). Pe 8 9 11 47.2 n, 32.8 3, 54.5 a 69.6 Total, . 34 51.0 MOSAIC DISEASE OF TOBACCO. 101 In the next table will be found a comparison of the results of catalase activity in healthy and diseased leaves from plants grown at the same time and under identical conditions. The plants were inoculated artificially in as uniform a manner as possible. Table IX. — Catalase Adivihj in Leaves of Healthy and Diseased Plants of Same Age. [Procedure as in Table VI.] Leaves. Number of Determina- tions. Cubic Centi- meters of Oxygen developed (Average). Diseased, 10 10 52.3 Healthy, 119.0 The values here obtained simply substantiate those given in preceding tables, but in addition allow of a direct comparison. The leaf tissue was used in the preceding experiments wdthout regard to the light and dark green patches on the individual leaf. It was thought that an examination of the light and dark green areas of indi^ddual leaves of mosaicked plants might give a clue as to whether the activities of the catalase were inhibited in one or both of these areas in comparison with a leaf from a healthy plant of approximately the same age and color. It was found that the catalase acti\ity of the dark green areas ap- proached that of the normal leaf of the same color, while the catalase activity of the light green areas was much below normal, even in the case of a Hght green normal leaf being used for comparison. The values obtained are given in Table X. Table X. — Catalase Activittj in Diseased Leaves. [Comparison of light and dark green areas; procedure as in Table VI.] Series. Number of Determina- tions. Cubic Centimeters of Oxy- gen DEVELOPED (AVERAGE). Light. Dark. X 0. . 21, 4 3 8 42.1 37.0 54.3 73.6 95.4 103.0 Diastase. — It is a well-kno^\Ti fact that diastase is intimatelj^ con- nected with metabolism in the leaf in practically all chlorophyll-bearing plants, as well as in many of the fmigi, and the relations of the activities 102 MASS. EXPERIMENT STATION BULLETIN 175. of diastase in the mosaic disease are of rather significant import, as can be easily shown. It was pointed out several years ago by Woods {loc. cit.) that the action of oxidizing enzymes when present in solutions containing diastase tended greatly, under ordinary conditions, to inhibit the activities of the diastase. Turning more particularly to the mosaic disease, he made the observation that in the cells of the light green areas, although they formed starch practically in a normal manner, so far as could be observed the starch was not translocated, and that in the morning there was prac- tically as much starch present as at night, which is not the case in a normally functioning leaf. In this case it was found that practically all the starch disappeared in the night and was translocated. Recently there has been more or less contention as to the exact method of action of diastase on starch, and within the last two or three years important investigations have resulted in the opinion, substantiated more or less in detail, that the diastase of the older writers is not one enzyme alone, but is made up of at least two components. The first of these breaks down the starches into, or as far as, the erythro-dextrine and achro-dextrine stage, the second component taking up the action from this point and completely hydrohzing the starch to the sugar compoimds which are found to be present, as the next step in the process of metabolism. It was in the light of these investigations that the writer took up the question of the diastase activity in the mosaic disease, and it was foimd to be less active in the leaves which showed severe symptoms of the disease than in those which showed only a sHght trace. There was, how- ever, apparently a greater or less breaking down of the starch in all the leaves examined, so far as could be determined by the colorimetric methods, which, although not altogether satisfactory, may be rehed upon as much as any of the present-known methods of determination. At the morning examinations the starch did not in some cases take on the color of the normal starch in the healthy leaves, but was accompanied by a yellow brown to a reddish or violet coloration, dependent somewhat on the strength of the indicator used. The strength of the iodine solution used in this case was a fiftieth normal iodine-potassium iodide solution. This reaction would indicate that the starch to a certain extent had been acted upon at least partially by the diastatic enzymes, and would indicate also that it was possibly the first of the components above mentioned which was more active, and that the second was more or less inhibited in its action. In the normal leaf, of course, there was a certain amount of starch present indicated by the blue coloration of the granules. The amount was slight, however, compared to that in the diseased leaves, and in no case was there any of the brown or violet color, almost complete hydrolysis having apparently taken place very rapidly. This would indicate, as pointed out by Woods, that the oxidizing enzymes, of which we will make mention, and which are present in excessively large amounts in the diseased areas of the leaf, do play an important role in the controlling or inhibiting of the activities of the diastatic enzymes, but not on the MOSAIC DISEASE OF TOBACCO. 103 diastase in the old conception of the term. Rather it might be said the action is on the primary enzyme concerned in diasta.ic activity, if the newer concept of diastatic activity above advanced is true, as it would seem to be from the impubHshed investigations of Roessler of the Uni- versity of Prague, who was able to separate by salting out from a very carefully prepared solution of the ordinary diastase at least two compo- nents having the respective actions above mentioned. In no case, as indicated by the color reaction obtained, did we get a complete hydrol- ysis of a large amount of starch, the process only being carried on, ap- parently, as has been indicated, — as far as the erythro-dextrine and achro-dextrine stage. It was attempted in our experiments to isolate or rather separate out diastase in a more or less pure form from the leaves of healthy and diseased plants, and although certain results were obtained, it was rather a difficult matter, as in the writer's experience it has been found that diastase is one of the most difficult of the enzymes to purify to any extent. The protective colloids, etc., during the purification are separated away from the enzyme aggregate, and the purer ferment be- comes less active. The reason for tliis cannot be very well explained at the present, but it is the experience of all investigators with diastase that this is a fact. However, results were obtained which seemed to indicate that the diseased leaves contain relatively less "diastase" than do the normal healthy leaves. Chlorophj/llase. — This enzyme has been found to be always present with chlorophyll in amounts directly proportional to the amount of chlorophyll present, and according to Willstatter and Stoll^ does not bring about an hydrolysis but an "alcoholysis," RCOOC20H39 — C2H5OH . RCOOCaHs — C20H39OH in the presence of ethyl alcohol. It forms the alcohol phytol, C90H39OH, from the radical in the presence of ethjd alcohol and not water only. Very Httle is known about its action in the plant cell, and although the writer was able to demonstrate its presence in both healthy and diseased leaves, no quantitative data were secured as to its relative activity in healthy and diseased tissue. Until better methods are worked out for its purification and rapid determination it "would be futile to hazard an opinion in regard to its specific action in the cells of healthy and diseased leaves. Oxidases and Peroxidases. — Woods {loc. cit.) was one of the first to observe that in mosaic-diseased leaves the oxidase activity was greatly increased. Since then it has been found that in the curly dwarf disease of the potato and sugar beet the oxidase activity is greatly increased in the diseased leaves as compared with that of the normal. These two diseases have been for the most part regarded as physiological, and it is 1 Willstatter and Stoll: Unt. uber Chlorophyll XI und XIII. Uber Chlorophyllasa. Liebig's Ann. der Chemie., 378, 18 (1910); 380, 148 (1911). 104 MASS. EXPERIMENT STATION BULLETIN 175. a significant fact that this excessive activity of oxidizing^ enzymes has been more frequently noted in diseases of this character than in those which are caused by bacteria or fungi. The reaction of the host is ap- parently different in the latter case. BunzeP has noted that the oxidase activity varies with the age of the plant in the curly dwarf disease of potato, reaching its greatest activity when the plant growth ceases. The writer has also found this to be true for tobacco to a certain extent, and always met with greater activities of the oxidases as the leaves were approaching maturity. This was marked in the case of normal plants, but not so much in the case of diseased leaves. In the writer's examinations of healthy and diseased tissue not only qualitative colorimetric methods were employed, but also a simpHfied Bunzel's oxidase apparatus was made use of. This has been found to be the most satisfactory method for the quantitative estimation of oxidase activity.^ A few of the quantitative results obtained are given in Table XI. Table XI. — Oxidase Activity in Normal and Mosaic Sap. [Manometer readings in centimeters of mercury. Bunzel apparatus mod.] Experiment. Time in Minutes. Normal. Diseased. 0 0 0 30 -0.60 -0.80 A 60 —1.09 -1.23 75 —1.12 —1.29 120 —1.22 -1.43 0 0 0 30 -0.32 -0.50 B, 60 —0.80 -0.70 75 -1.02 -0.96 120 —0.92 -1.21 0 0 0 c 30 -0.51 -0.46 75 -0.63 -0.88 100 -0.70 -0.91 It will be noticed that the mosaic sap is higher in total and average in every instance. For the qualitative determinations the usual guaiac test was employed. The guaiac test for oxidases and peroxidases is too well known to require 1 Bunzel. H. H.: Oxidases in Healthy and Curly Dwarf Potatoes. Jour. Agr. Research, Vol. II.. 5. 373-404 (1914). 2 Bunzel, H. H.: The Measurement of the Oxidase Content of Plant Juices. U. S. D. A., Bur. Plant Ind.. Bui. No. 238 (1912). MOSAIC DISEASE OF TOBACCO. 105 an extended explanation. The results obtained by this method in every case showed the diseased leaves to contain much more oxidases than the healthy ones of the same age; this was also true for peroxidases, but here, of course, the reaction with guaiac was somewhat masked owing to the presence of the oxidases and their reaction. In examinations of the roots of healthy and diseased plants the same condition was observable; there was always an excessive activity of the oxidizing enz3''me to be noted. In going over the results of the experiments with the enzymes in ques- tion, the main point brought to the attention is that there is in all diseased plants an excessive activity of the oxidizing enzymes, and a corresponding decrease in the activity of the diastatie enz3rmes and catalase. This at least indicates a very much disturbed equiUbrium and a consequent derangement of normal function on the part of the cells. Naturally the ones most affected by this disturbance are the dividing or meristematic cells, as these are the cells upon which the plant is dependent for its sub- sequent growth, and any deviation from the normal is more Ukely to be indicated in the development of these ceUs than in those of the other parts of the plant. Any change in function induced here will leave its imprint to a greater or less extent on the cell during its subsequent exist- ence, hence the peculiar manifestations of the disease in the leaves. It is true that plants attacked by parasites sometimes show an exces- sive activity on the part of certain enzymes, but, as a rule, the disturb- ance is more local in its nature. It is also a fact that malnutrition, such as partial starvation, drought, etc., will bring about an excessive produc- tion or activity of the oxidizing enzymes in particular, as has been pointed out by Bunzel, of general distribution throughout the plant; but tliis, except in cases of maturing plants, changes upon restoration of normal conditions, and tends to become normal. Reaction of Mosaic Sap with Various Substances. We have seen that the enzymatic activities of the plant are very much disturbed in disease; also that it has been impossible to demonstrate the presence of any forms of bacteria or fungi either in the tissues themselves or in the expressed juice. It is a fact, as shown by practically aU investigations, that the disease is very infectious. This fact alone in the minds of many is sufficient to place the causative agent among the parasitic organisms. The field, however, is hmited to that class of organisms designated as "ultramicro- scopic" organisms, about which very little is known, and in the case of plant diseases not even a semblance of the demonstration of the activities of such organisms has been made. Owing to the fact that the enzyme activities are much changed, as has been demonstrated in the preceding pages, and also to the fact that not only the activities of the oxidizing enzymes are changed, but also the 106 MASS. EXPERIMENT STATION BULLETIN 175. activities of others; it was believed by the writer, with Woods and others, that the disease might be physiological in nature, particularly in so far as the causal agent, not being a Uving organism in the ordinary conception of the word, was concerned. So Uttle is known about the action of the so-called ultramicroscopic organisms that it is an open question in the writer's mind whether this division should be the dumping ground for all infectious diseases about the etiology of which little or nothing is known. It is conceivable that other causes, not organic in nature, may be able to produce the manifestations of parasitism. Under this type of infection would be included infectious diseases caused by enzjones or the resultant product of the activities of a group of enzymes. Certain reactions of the juice from diseased plants tend to confirm this view, and in the following pages are given the results obtained by the writer and other investigators relating to the reactions of these juices with various reagents. Drying. — It has been shown by various investigators that the dried leaves of the mosaic-diseased plants retain their infectious quahties for a long time. Beijerinck and Allard found that diseased leaves were capa- ble of causing infection after being dried for periods of two years and eighteen months, respectively. The writer has used material three years old, and obtained infection in a great majority of cases. The results obtained are given below. Table XII. — Air-dried Mosaic Leaves, finely ground and macerated with Cold, Distilled Water. [Leaves (herbarium specimens) three years old.] Number of Plants inoculated. Point of Inoculation. Number of Plants infected. Per Cent. Infection. 10 12 7 13 Below terminal leaflets Main stem near base Midribs of a basal leaf, Midribs of a basal leaf 10 11 6 12 100 91 86 90 Filtration. — The use of various filters such as the Chamberland, Berkefeld and Kitasato types, as a means for the separation of bacteria and other organisms in a fluid, has been widely adopted in recent years, and more recently filters possessing different sized pores have been used for differential diagnostic purposes in work on the so-called "ultramicro- scopic" organisms, enzymes and toxins. While these methods are without doubt of importance, it should always be borne in mind that to obtain true filtration effects comparatively large volumes of the fluid should MOSAIC DISEASE OF TOBACCO. 107 be used, otherwise there is a strong possibility, particularly in the case of enzymes, that instead of a filtration occurring at once, a large amount of certain constituents may be adsorbed (dependent on the nature of the filter), and that true filtration may not take place until compara- tively large amounts have been drawn through the filter. The writer has noted this particularly in work with enzymes, many of which are strongly adsorbed by various substances. Aside from the "ultramicro- scopic" organisms, however, the bacteria cannot pass through many of these filters. With reference to the causal agent in mosaic sap it has been found that it passes through both the Chamberland and Berkefeld filters, and even the finer grade of Berkefeld filter allows the passage of the causal agent. Beijerinck {loc. cit.) showed that the juice was still infectious after being passed through the Chamberland filter, and Allard {loc. cit.) and CUnton {loc. cit.) have both shown that the juice was infectious after passage through the Berkefeld (normal) filter. The results obtained by the writer agree with these observations, and also the juice was found to be infectious after passing it through the fine Berkefeld candle. The Kitasato filter was also used, and here positive infection was also obtained, although the percentage was small. The writer attempted to repeat these experi- ments with the Kitasato filter during the past year, but was unable to obtain the filter. In all cases relatively large amounts of the sap were used after filtration through paper. The average percentage of infection obtained with each filter in the writer's experiments was as follows: — Per Cent. Chamberland (average of 3 examinations, 1911), . . . . .91.0 Berkefeld (normal; average of 5 examinations, 1911), .... 03. 0 Berkefeld (fine; one test only, 1914) 47.0 Kitasato (average of 2 examinations, not dated) , . . . . . 40 . 5 The work with the fine grade of Berkefeld and Kitasato filters should be repeated, but there are sufficient indications to warrant the insertion of these results at this time. Resistance to Antiseptics. — The writer has at various times treated filtered and unfiltered juice with many of the antiseptics such as are commonly used to prevent bacterial action. The following table contains the data and results obtained in one typical series of experiments of this character: — 108 MASS. EXPERIMENT STATION BULLETIN 175. Table XIII. Antiseptic. Amount of Sap used (Cubic Centimeters). Period of Treatment. Infection. Toluol (2 c. c.) 10 2 months. ++ Toluol (2 c. c). 10 4 months. ++ Chloroform (saturated at beginning), . 10 2 months. ++ Chloroform (in excess) 10 2 months. — Chloroform (saturated at beginning), . 10 4 months. + Chloroform (in excess), 10 3 days. -+ Thymol (2 per cent.) 10 2 months. + Thymol (2 per cent.) 10 4 months. + Ether (saturated) 10 2 months. + Ether (saturated) 10 4 months. + Formaldehyde (1-4 H2O, 1 c. c. added). 10 2 months. - Formaldehyde (1-4 H2O, 1 c. c. added), . 10 10 days. - Carbolic acid (5 per cent., 10 c. c. added), . 10 2 days. - Chloralhydrate (M moL), .... 10 2 days. - Chloralhydrate (H mol.) 10 20 hours. - ++=very infectious. — l-=one or two cases of infection, possibly accidental. +=infectious (over 40 per cent.). — =no infection. From the preceding table it may be seen that the sap containing the causal agent of the disease varies greatly in its reaction to so-called anti- septics and other compounds. The writer^ has already pointed out in a previous publication that the influence of certain capillary active sub- stances on enzymes is very variable, aside from the specific toxic qualities of certain of these substances. In comparing the reaction of the sap con- taining the causal agent to certain of these compounds we find that there is a similarity of reaction to that shown by the enzjTnes. In the paper above cited it was shown that those compounds which changed the sur- face tension had, as a rule, dependent on their physical properties (hydro- colloidal or hpocolloidal), a certain definite effect on enzjmie activities. Taking up the discussion of the results in detail we find in toluol a compound which is not soluble in water to any great extent, and hence, behaving hke a Upocolloid, having no effect on the action of the causal agent contained in the sap. Toluol, as a rule, has a more or less definite inhibitory action on living organisms. Chloroform, when present in the sap not to exceed saturation, behaves also lUce a lipocolloid, as it is only very slightly soluble in the water, and J Chapman, George H.: The influence of Certain Capillary-Active Substances on Enzyme Activity. Internat. Zeitschrift fiir Physik.-chem. Biologie., I Band, 5 u. 6 Heft (1914). MOSAIC DISEASE OF TOBACCO. 109 we find in this case that the acti^aty of the agent is not destroyed. Chloro- form in excess, however, does destroy apparently the causal agent of the disease. It is noteworthy that tliis action of chloroform exactly parallels that found to be the case with enzjines. ThjTnol, when used in 2 per cent, concentration is very often used as a preventive to bacterial action, and also prevents the growth of fungi. We find, however, that when it is present in concentration not exceeding 2 per cent, in the sap the causal agent still possesses its infectious qualities for some time. Ether is a substance which, like chloroform, has lipoid-Uke properties, but which has a definite action on the surface tension, lowering it con- siderabl3^ Sap containing ether to the saturation point, which lowers the surface tension from 1 to about .619, was still infectious four months after treatment, although the percentage of infection was much decreased. A solution of the sap containing approximately .8 per cent, of actual formaldehyde was very injurious, and at the end of two months no infec- tion was obtained. At the end of ten days in one experiment, however, plants were inoculated and two cases of mosaic disease developed from a series of eight plants, but it is believed that this may possibly have been an accidental infection, as in no other instance was infection obtained. In formaldehyde, however, we have a compound which has a specific narcotic action on certain enzymes aside from its surface activities. Where carbolic acid was added to a solution of the sap the active prin- ciple was apparently destroyed. In cliloralhydrate we have a substance very soluble in water, but not possessing any relatively great surface activity. It has, however, a specific toxic action on the causal agent of the disease, and even after twenty hours no infection was obtained. These results with chloralhydrate are in complete accord with those obtained in the enzyme work previously mentioned. Most of the substances used in the above experiments possess a very definite toxic action to all organisms, particularly bacteria and fungi. As to their effect on the so-caJed ultramicroscopic organisms the writer is unable to state, not having had the opportunity of working with so-called cultures of these organisms. The parallelisms between the surface-ten- sion effects of these substances on enzymes and on the sap containing the active principle of the mosaic disease are very striking. Having shown that the causal agent is not bacterial or fungous in character, we must ehminate for the present the supposition of the presence of a toxin or virus in the pathologist's conception of these terms, as it is usual to conceive of these substances as being either the produi^^t of an organism or the activity manifested by the organism itself. As to the production of toxins and viruses by the so-called ultramicroscopic organ- isms httle is known. Noguchi was the first to apparently demonstrate that such organisms do e.xist, and was able to cultivate an organism obtained from the brain of patients suffering from infantile paralysis. 110 MASS. EXPERIMENT STATION BULLETIN 175. However, these organisms were always mixed with certain bodies probably of a protein nature, and Noguchi, himself, so far has been unable to state absolutely which may be the active agent, although he naturally infers from his inoculation experiments that the organisms found must be the causative agent owing to the extreme infectious character of the disease. He, however, states that it is not absolutely clear to him whether the organism alone or a combination of this organism with the bodies found in culture associated with it are capable of producing infection. He does state, however, that in the case of animal pathology no such symbiotic relationship has so far been observed. From the character of his statement, however, it is clearly indicated that he does not preclude the possibUity of such a condition arising. Probable Character of the Causal Agent. The question as to the exact character of the causal agent of mosaic disease has been an extremely interesting one to investigators, and studies on this phase of the problem have narrowed the field by the elimination from consideration of fungi and bacteria, as has previously been shown not only in this work, but also by many other investigators. This also precludes the presence of a virus or a toxin resultant from the activities of such organisms. This leaves, then, for consideration as the causal agent an "ultrami- croscopic" or "invisible" organism and the enzymic activities in their fullest conception. The reactions of the so-called "ultramicroscopic" organisms are little known at present, and about the only grounds for admitting of such a class of organisms is the infection factor, and possibly reproduction to a certain extent. We do know, however, many reactions of the class of substances called enzjones and toxins, but fundamentally the differentiation of the three above mentioned is difficult, and is per- haps in many cases impossible. Working with filtered sap from mosaic- diseased plants, we get the following results in comparison with reactions of some of the so-caUed "ultramicroscopic" organisms and toxins. Temperature. — The sap containing the causal agent of mosaic disease becomes non-infectious; in other words, becomes inactive when heated to about 80° C. for a short time. It is reported that ultramicroscopic organisms and toxins are killed or rendered inactive, respectively, by exposure to heat for any length of time at temperatures somewhat below 100° C. Enzymes are also rendered inactive at temperatures somewhat below 100° C. All three react practically alike as regards temperature. The causal agent in mosaic sap, as may be seen, is also rendered inactive at temperatures below 100° C. Size. — As to size, nothing can be definitely stated, but it is a fact that the ultramicroscopic organisms, enzymes and toxins must have a diameter of less than .1 //, otherwise they would become visible under the higher powers of the microscope. In no case has it been possible to demonstrate the presence of organisms under even the highest powers available. MOSAIC DISEASE OF TOBACCO. Ill Reaction to Antiseptics. — It is stated that the ultramicroscopic organ- isms are not, to any extent, affected by the ordinary antiseptics, and the same is true for toxins in general. On the other hand, enzymes and their activities are very strongly affected by the substances usually made use of as antiseptics, and this is found to be true, with one or two possible exceptions, in the case of mosaic sap. It has been shown that formalin, carbolic acid, chloralhj^drate, and even chloroform in excess, will inhibit the activities of the causal agent in mosaic sap, while, on the other hand, such substances as ether, toluol, thjnnol and chloroform in dilution have little or no effect. While all three classes are to a certain extent affected by antiseptics in general, the enzjTne group is most strongly affected, and in the case of the mosaic we find this reaction; also, as has been pointed out, the effect of substances possessing marked surface-active properties is, in the case of mosaic sap, quite analogous to that of these substances on enz}Tnes. It had been hoped to carry on more detailed work on this point, but as yet no opportunity has offered to take up this phase of the matter. Allard^ has studied the effects of alcohol, ether and other sub- stances on mosaic sap, and an interpretation of his results, with particu- lar reference to the surface-active properties of the substances under con- sideration by him, parallel the author's findings in the case of enzymes to a marked degree. It is believed that more work of this character might throw considerable light on this matter. Activity. — So far as can be judged from laboratory results the activity of the causal agent in mosaic sap is continuous, and as this holds true not only for organisms but, within limits, for enzymes and toxins as well this property cannot be made use of for differential purposes. "Koch's Laws" or "Postulates," so called, are followed by all three of the classes under consideration, and the same is true in the case of mosaic disease; the causal agent obeys these laws, and might well be placed in any one of the classes so far as this property is concerned. The Kitasato filter has been used by some as a means of separation of "ultramicroscopic" organisms from enzjones and toxins, and although the arbitrary use of any one filter as a standard, unless the size of pores, adsorption properties, thickness of walls, etc., are carefully taken into consideration, may be open to question, this procedure has been followed in some instances in animal pathology, and it has been found that the Kitasato filter held back the organisms and that no infection could be obtained from the filtrate. In the case of the mosaic disease, however, we find that apparently, as has been previously indicated in this paper, where large volumes are used, the causal agent passes through the Kitasato filter, and we do get infection from the filtrate. The disease is infectious, but whether the infection may be indefinitely transferred through several plants with undiminished virulence is open to question. On some varieties of tobacco this does not apparently take 1 Allard, H. A.: Some properties of the virus of the mosaic disease of tobacco. Journal Agr. Research, Vol. VI., No. 17 (July, 1916). 112 MASS. EXPEEIMENT STATION BULLETIN 175. place, but so far as the writer's observations go the virulence of the causal agent is not lessened appreciably. This property is one of the strongest points advanced by those favoring the theory of the presence of a definitely organized parasite as the causal agent of the disease. It is, however, entirely possible that enzj^ines or similar substances intro- duced into a plant even in extremely small quantities, are capable of regeneration of a certain kind, andindeed it is held by some that enzymes do grow and even reproduce themselves under certain conditions. The difficulties encountered in the study of this phase of enzyme work are very great, however, and it is questionable if such statements can be as yet definitely accepted. Organisms, even of the ultramicroscopic class, in their reactions would not follow the law of proportion aUty, but in the case of mosaic sap and its reactions we find, by measuring the relative activities and reactions of the enzjines present that apparently a proportionaUty of reaction for any one lot of sap does hold. The writer has very often found in the measurement of the activities of the catalase and oxidase particularly that not only a fairly definite relation exists between the various enzymes, but that reaction of any one is dependent on the amount of sap used. Of course, here we are dealing with a mixture, and it may be open to question if the measurement of the enzyme activities is a true measure of the activities of the causal agent. The whole subject of the differential diagnosis of enzjones, toxins and ultramicroscopic organisms is an extremely difficult one, and no sharply dividing Unes can properly be drawn between them. It would appear to the writer that in some cases, at least, it is entirely dependent on the view- point and interpretation of the investigator as to the class to which certain diseases should properly be ascribed. The factors of reproduction and infection, as ordinarily understood, have proved a stumbhng block to the acceptance of the idea that there may be other forms of matter aside from organisms capable of reproducing a disease, but there is in reaUty very httle real ground for taking this attitude. In the case of the mosaic disease there are certainly many reac- tions which wiU not allow of placing the causal agent in the class of ultra- microscopic organisms. The general distribution of the causal agent in a diseased plant, its exceedingly locahzed action on the meristematic tissues, this action being apparently confined to the nascent chlorophyll, the non-uniformity of response to apparently favorable conditions during any one season even on one field, and also its individuaUsra as shown by plants growing together (one often diseased and the other not) are to the writer indicative of something of a different character. It is also possible that in the search after the infinitesimal the fact that a highly organized plant as a whole may react in the same manner as some of the simpler organisms has been overlooked. It is as a rule not the presence of an organism alone which is responsible for the manifesta- tions of disease, but the products of the metaboUsm of the organism. MOSAIC DISEASE OF TOBACCO. 113 If the metabolic processes are changed ever so slightlj^, due to any stimu- lus, far-reaching effects may be induced throughout the organism, and this we find to be the case in the mosaic disease, and the writer believes that it is justifiable to look upon the matter in this hght, as it is no more hypothetical than the concept of an "ultramicroscopic" parasite, which, if demonstrated (and no amount of concentration or methods of culture have indicated in any way the presence of aggregates or colonies), certainly would become visible if multipHcation occurred. Theoretically is it possible to conceive of an organism, functioning as such, to be made up of so few molecules of protein, fat and carbohydrate that it would be impossible to demonstrate its presence? If so, our ideas of relative size of molecules of protein, etc., must be changed. Prevention and Control. The question of the prevention and control of mosaic disease is of prime importance to the grower, entirely aside from more technical considera- tions as to the exact cause or causes of the disease, and it is believed that with reasonable care it is possible for the grower to lessen materially the amount of mosaic in the field. Many recommendations have been made regarding treatment of dis- eased plants after they have once contracted the disease, but so far the writer has never observed a plant which, once attacked by the disease, recovered at any subsequent period of its growth. On the other hand, it has never been observed that the disease killed a plant, at least in this region. It is doubtful, owing to the character of the disease, if it can ever be entirely ehminated on some soils and under certain unfavorable conditions occurring during some seasons. As has been indicated previously there is apparently little or no relation to be found between excess or lack of food materials and the prevalence of the mosaic. It has been in some instances stated that favorable results have been obtained from the use of lime in different forms, but this treatment cannot be recommended for various reasons. Experimentally it has been shown that heavy liming has little or no effect on the disease once a plant has contracted the disease, and even when appUed to soils from old beds no consistently favorable results have been obtained (see page 91). Used in the larger quantities it might be inferred from the results that the lime apparently did exert a beneficial action, but to apply hme generally in such amounts would be folly, as it would in many cases bring the soil to a comparatively neutral or alka- line condition, which reaction would favor the development of root rot, caused by the fungus, Thielavia, and this, once thoroughly established, in a field or seed bed, is much more injurious to tobacco than is the mosaic As has been pointed out, the writer, from his observations, is strongly of the opinion that much of the field infection may be traced to the seed 114 MASS. EXPERIMENT STATION BULLETIN 175. bed, and as a rule those beds which have long been used or carelessly- handled are found to be producers of mosaicked seedlings in far larger numbers than are found on new beds or on beds which have been carefully sterihzed either by steam or formalin. It has been found that the soils of old beds do tend to produce more mosaicked plants than do those of new beds, although it may be possible that under field conditions the differences in amount during different seasons may vary. Soils brought into the greenhouse gave the following results: — Table XIV. — Experiments with Soils from Old and New Beds. [Seedlings transplanted in sterilized soil.] Soil. Number of Seedlings transplanted. Number Diseased Four Weeks after Trans- planting. Diseased (Per Cent.). Son A (old bed) SoU 21 (old bed) SoUIo Soil B (new bed), Soil C (new bed) 60 43 50 30 49 45 17 21 3 2 75.0 40.0 40.0 10.0 4.0 The soil from the old beds was in very bad condition and had been very carelessly handled, apparently. A count of mosaicked seedUngs left in these old-bed soils six weeks after the transplants was taken, showing, respectively, an infection of A, 43 per cent.; 21, 32 per cent.; la, 17 per cent.; B, 6 per cent.; and C, 7+ per cent. It is evident that some of the seedlings were infected during trans- planting, probably by handling diseased seedUngs and then healthy ones, thus transmitting the disease. This method of transmission at the time of transplanting is very common, as has been pointed out repeatedly. It has been shown that much of our infection may originally come from the seed bed as a result of the soil becoming infected for any reason. The use of tobacco stems and tobacco water has also been found by many investigators to cause infection. The amount of infection resulting from watering beds with water extract of diseased stems is, however, prob- lematical, and it is not believed by the writer that this is an important factor in mosaic transmission, especially if the stems are steeped in hot water. The broken, decaying roots of diseased plants left in the beds also carry the causal agent of the disease as do the stems of diseased plants, and freezing has apparently little or no effect on it, so the use of stems on the seed bed should be carefully attended to in order not to apply any from diseased plants. Where stems and tobacco water are applied year MOSAIC DISEASE OF TOBACCO. 115 after j^ear without attention to this point the bed usually becomes more seriously infected. One of the cheapest methods for the control of this disease in the seed bed, where it can be advantageously carried out, is to change the location of the beds to soil on which no tobacco has been grown, and to avoid the use of stems and tobacco water. Occasionally, however, some sHght in- fection will occur even here, but as a rule not to any great extent. If proper attention is paid to watering, ventilation, etc., Httle trouble of this character is to be expected in new seed beds. It has been shown in Connecticut and elsewhere that a thorough ster- ilization of the seed bed by steam at a boiler pressure of from 70 to 90 pounds is also a satisfactory method for the control not only of fungous diseases but weeds also, and the same holds true for the mosaic disease. The writer has seen this tried a number of times with excellent results where the above-mentioned pressures have been used. Some growers, however, seem to be of the opinion that the prime value of steaming is to kill weed seeds, and so use low pressures. While low pressures will kill weed seeds, it is questionable if they will sterilize the soU sufficiently to kill the spores of fungi or render inactive the causal agent of the mosaic, although under laboratory conditions it is rendered inactive at tempera- tures of about 80°C, equivalent to 176°F. In some of our experiments conducted some years ago it was strongly indicated that improper partial sterihzation would not entirely rid the soil of the causal agent of mosaic. It might be stated here that, in many cases where the growers have reported failure in the control of diseases after steam sterilization, inquiry has usually brought out the fact that too low pressure was used, and as a result thorough sterilization was not obtained. Another source of fail- ure of beds after sterihzation with steam, under high pressure, has been that the grower has not paid sufficient attention to watering. This mat- ter should be closely attended to, as a sterilized bed, particularly on light soils, dries out very quickly, and needs much more attention than is usually given a bed under ordinary conditions. If the watering is neglected there is very often a severe checking of the germination of the seed, and in some cases a partial loss of the bed. Formahn sterilization may also be used, and is quite as satisfactory, especially when used on light soils. On heavy soils it is not quite so con- venient to apply, however. Where formahn is used the beds cannot be sown until all the formalin is out of the soil, which usually takes from ten days to two weeks. This very often is too long a delay, particularly where spring sterihzation is practiced. It has been pointed out that the workmen may be a rather important factor in transmitting the disease (page 88), and in cases where at trans- planting time diseased seedhngs are handled it has been recommended by Clinton^ that the hands be thorouglily washed in soap and ■water ' G. P. Clinton: Chlorosis of Plants with special reference to Calico of Tobacco. Conn. Agr. Exp. Sta. Rept., 1914, p. 417. 116 MASS. EXPERIMENT STATION BULLETIN 175. before again handling healthy seedhngs. If these precautions are taken, according to CUnton, a considerable amount of mosaic infection will be avoided at the time of planting. It has been repeatedly shown that care should be exercised during early cultivation not to cut the roots or touch broken or abraded leaves of plants and then subsequently touch other plants, for the disease is very easily transmitted in this way, as the fine hairs or epidermis may be broken and infection occur. The amount of infection due to cultivation is, however, in the writer's opinion, slight, but as much care as is com- mensurate with efficiency should be exercised by the workmen during cultivation. The advisabiUty of the removal of diseased plants is open to question, and on the whole it cannot be economically recommended unless the plants can be replaced early in the season. As has been previously pointed out, the disease may be carried from plant to plant when topping, etc., and the subsequent sucker growth will become mosaic. At this time, how- ever, the commercial leaves are of such size that their value will not be materiall}^ impaired, but if possible, to prevent a certain amount of infec- tion, only healthy or diseased plants should be topped at any one time. Of course, all suckers developing later, diseased or otherwise, should sub- sequently be removed from all plants, not only for the sake of the com- mercial leaves, but to prevent a ragged looking field, giving the appear- ance of a large amount of mosaic. It has been very difficult to associate any particular type of soil with general occurrence of mosaic disease, but on the whole, from data gathered at different times, the heavier types of soil in the valley appear to be more generally favorable for the production of mosaic-diseased plants. This cannot be definitely stated, however, as the data are complicated by the fact that in some cases, on both heavy and light soils, the condition of the soil as regards organic matter present enters into the question. The writer has observed that on many heavy soils where comparatively large amounts of organic matter are present during certain seasons, in com- parison with similar soils deficient in organic matter, the mosaic is much less. To a certain extent this holds true also for the lighter soils. The exact relation existing between the mosaic disease and these factors is at present not enough studied to warrant definite conclusions, but Sturgis (loc. cit.) was of the opinion that clayey soils were favorable to its pro- duction. It is a significant fact that many of our tobacco soils are some- what deficient in organic matter, however. Well-cultivated and conse- quently well-aerated soils do not apparently produce as many mosaicked plants as those which are not well cultivated. Another factor which should be carefuUy attended to is that of the moisture conditions in the bed at the time the plants are pulled. It should not be too moist nor too dry, as in either case the roots are apt to be broken and infection from handling result more certainly than when the plants are removed with a minimum of root injury. MOSAIC DISEASE OF TOBACCO. 117 Summary. 1. The mosaic disease is not caused by fungi or bacteria. It has never been possible to demonstrate the presence of these organisms in the tis- sue of any part of the plant. 2. The disease is highly infectious, particularly when inoculated into young plants, all subsequent growth exhibiting marked symptoms. 3. The disease is not contagious. 4. Until more is known about the action of the so-called "ultramicro- scopic" organisms, the disease cannot be ascribed to an organism of that class, as the character and reactions of the causal agent do not in many respects coincide with reactions of that class of organisms. 5. Many of the reactions of the causal agent are of such a nature as to indicate that it is either an enzyme, an aggregate of enzjTnes, or the prod- uct of enzyme activities. 6. The enzyme activities of diseased plants are greatly altered, far more than is usually the case in plants which are attacked by pathogenic fungi or bacteria. 7. As a result of the writer's experiments, it is believed that the disease is primarily induced by a disturbance in the enzj^ine activities and their relation to each other, due to abnormal metabolism, and not by any parasite. 8. The pathogenicity of a disease is not necessarily a proof that it is of parasitic origin, as it is conceivable that similar conditions may exist relative to enzjTne activities, although the extent of such action is not known at present. 9. On fields where the mosaic disease is prevalent, the primary infec- tion can usually be traced to the seed bed, and many healthy seedUngs are infected by the workmen when setting the plants. It is estimated that about 80 per cent, of the infection occurs in this manner. 10. Owing to the nature of the disease the matter of absolute preven- tion and control is difficult, but with careful attention to details of ster- ilization of the seed bed, and handhng of the plants at time of trans- planting, a large percentage of infection may be avoided. BULLETIN No. 176 OCTOBER, 19(7 MASSACHISETTS AGRICILTIRAL EXPERIMENT STATION The Cause of the Injurious Effect of Sulfate of Am- monia when used as a Fertilizer By R. W. RIPRECHT and F. W. MORSE This Bulletin is a continuation of Bulletin No. 165, " The Effect of Sulfate of Ammonia on Soil." It shows that soluble salts of iron, manganese and aluminium, severally or collectively, were always found in soils which had been dressed with sulfate of ammonia without an addition of lime, and that these several compounds were positively injurious to clover seedlings in cul- tural experiments. Requests for bulletins should be addressed to the Agricultural Experiment Station, Amherst, Mass. Massachusetts Agricultural Experiment Station. OFFICERS AND STAFF. COMMITTEE. Charles H. Preston, Chairman, . Hathorna Wilfrid Wheeler, . Concord. Trustees. Edmund Mortimer, . . Grafton. Arthur G. Pollard, . Lowell. . Harold L. Frost, . Arlington. The President of the College, ex officio. The Director of the Station, ex officio. STATION STAFF. Administration. William P. Brooks, Ph.D., Director. Joseph B. Lindsey, Ph.D., Vice-Director. Fred C. Kenney, Treasurer. Charles R. Green, B.Agr., Librarian. Mrs. Lucia G. Church, Clerk. Miss F. Ethel Felton, A.B., Clerk. Agricultural Economics. Alexander E. Cancb, Ph.D., In Charge of Department. Samuel H. DeVault, A.M., Assistant. Agriculture. William P. Brooks, Ph.D., Agriculturist. Henry J. Franklin, Ph.D., In Charge of Cranberry Inves- tigations. Edwin F. Gaskill, B.Sc, Assistant Agriculturist. Robert L. Coffin, Assistant. Botany. A. Vincent Osmun, M.Sc, Botanist. George H. Chapman, Ph.D., Research Physiologist. Paul J. Anderson, Ph.D., Associate Plant Pathologist. Orton L. Clark, B.Sc, Assistant Plant Physiologist. W. S. Krout, M.A., Field Pathologist. Miss Mae F. Holden, B.Sc, Curator. Miss Ellen L. Welch, A.B., Stenographer. Entomology. Henry T. Fernald, Ph.D., Entomologist. Burton N. Gates, Ph.D., Apiarist. Arthur I. Bourne, A.B., Assistant Entomologist. Stuart C. Vinal, M.Sc, Assistant Entomologist. Miss Bridie E. O'Donnell, Clerk. Horticulture. Frank A. Waugh, M.Sc, Horticulturist. Fred C. Sears, M.Sc, Pomologist. Jacob K. Shaw, Ph.D., Research Pomologist. Harold F. Tompson, B.Sc, Market Gardener. Miss Ethelyn Streeter, Clerk. Meteorologry. John E. Ostrander, A.M., C.E., Meteorologist. Microbiology. Charles E. Marshall, Ph.D., In Charge of Department. Arao Itano, Ph.D., Assistant Professor of Microbiology. George B. Ray, B.Sc, Graduate Assistant. Plant and Animal Chemistry. Joseph B. Lindsey, Ph.D., Chemist. Edward B. Holland, Ph.D., Associate Chemist in Charge (Research Division). Fred W. Morse, M.Sc, Research Chemist. Henri D. Haskins, B.Sc, Chemist in Charge {Fertilizer Division). Philip H. Smith, M.Sc, Chemist in Charge {Feed and Dairy Division). Lewell S. Walker, B.Sc, Assistant Chemist. Carlbton p. Jones, M.Sc, Assistant Chemist. Carlos L. Beals, M.Sc, Assistant Chemist. James P. Buckley, Jr., Assistant Chemist. Wxndom A. Allen, i B.Sc, Assistant Chemist. John B. Smith, i B.Sc, Assistant Chemist. Robert S. Scull, B.Sc, Assistant Chemist. James T. Howard, Inspector. Harry L. Allen, Assistant in Laboratory. James R. Alcock, Assistant in Animal Nutrition. Miss Alice M. Howard, Clerk. Miss Rebecca L. Mellor, Clerk. Poultry Husbandry. John C. Graham, B.Sc, In Charge of Department. Hubert D. Goodale, Ph.D., Research Biologist. Miss Rachael G. Leslie, Clerk. Veterinary Science. James B. Paige, B.Sc, D.V.S., Veterinarian. G. Edward Gage, Ph.D., Associate Professor of Animal Pathology. John B. Lentz, i V.M.D., Assistant. ' On leave on account of military service. CONTENTS PAGE Part I., Chemical Investigations 1^^ Part II., Water Cultures, ^^^ Conclusions, ...••••••*' Publication of this Document approved by the SuPEnviBOu OF Administration. 134 BULLETIN No. 176. DEPARTMENT OF CHEMISTRY. THE CAUSE OF THE INJURIOUS EFFECT OF SULFATE OF AMMONIA WHEN USED AS A FERTILIZER.^ BY R. W. RUPRECHT AND F. W. MORSE. Part I. — CnEJvncAL Inves^tigations. In a previous report- there has been described how the continued use of sulfate of ammonia on the experiment plots called "Field A" caused the removal of lime in the drainage waters in the form of calcium sulfate, and when lime was not present in sufficient quantity there were formed noticeable amounts of aluminium sulfate and iron sulfate, but that no accumulation of free acid could be found. Since comparatively little material had been pubUshed on the forma- tion of salts of aluminium ard iron in soils, it was considered advisable to continue the investigations, and as the work progressed it was found that soluble manganese salts were also present in some of the soils to which sulfate of ammonia had been applied. The present bulletin is a report of our investigations into the relations between sulfate of ammonia and salts of aluminium, iron and manganese, and the quantities of these salts which will injure clover seedUngs. Soils from plots 1, 6, 7 and 8 of Field A were used to determine how freely ammonium sulfate solutions would extract manganese from them. The soils have been fully described in Bulletin No. 165, but for con- venience the fertilizers used on these four plots will be described here. Each plot received dissolved boneblack at the rate of 500 pounds per acre, and muriate of potash 250 pounds per acre. Plot 1 received 300 pounds of nitrate of soda per acre; plots 6 and 8 received 225 pounds of ■1 The work reported in this bulletin, together jvith the material published in Buls. Nos. 161 and 165, was submitted by Mr. Ruprecht to the faculty of the graduate school of the Massa- chusetts Agricultural College in part fulfillment of the requirements for the degree of doctor of philosophy. 2 Bui. No. 165, "The Effect of Sulfate of Ammonia on Soil." 120 MASS. EXPERIMENT STATION BULLETIN 176. sulfate of ammonia per acre; and plot 7 received no nitrogenous fertilizer. In 1909, and again in 1913, hydrated lime was applied to one-half of Field A, crosswise of the plots. The total amount in the two dressings was 9,000 pounds per acre. The ammonium sulfate solutions were used in the manner described in Bulletin No. 165, viz.: 150 grams of air-dry soil were placed in a large flask with 750 cubic centimeters solution and shaken frequently for two hours. The solution was then filtered through paper, which gave a clear filtrate with a yellowish tint. Manganese was determined by the colorimetric method described by Schreiner and FaUyer,^ in which the manganese salts are oxidized to permanganate by nitric acid and lead peroxide. The strengths of the solutions were tenth-normal (N/10) and normal (N). The results obtained by the extracts from unlimed soils are tabu- lated in Table I., together with the amounts of iron obtained from the same soils in our previous work, and reported on page 81 of Bulletin No. 165. Table I. — Milligrams Manganese Oxide (MnoOi) arid Iron Oxide (FcoO^) obtained from 100 Grams Air-dry Soil by Ammonium Sulfate Solution. Manganese Oxide. Iron Oxide. Plot. N/10 Solution. N Solution. N/10 Solution. N Solution. 1 Trace. .58 .40 .79 6 .88 1.52 .46 .51 7, Trace. 1.18 .43 .50 «• .63 1.45 .89 1.21 The stronger solution removed much more manganese than the weaker, but not in proportion to its strength. The fertilization of plots 6 and 8 with ammonium sulfate evidently produced some manganese compounds that were readily soluble in the solutions, since there was more manganese obtained from those plots than from the other two. From the limed soils of these four plots there was removed no man- ganese by N/10 or N solutions, but when stronger solutions of am- monium sulfate were used (2§ N and 5 N), traces of manganese were found in the soil extracts. Tliis would appear to be due to the presence of enough ammonium sulfate in the concentrated solutions to overcome the lime and act upon the manganese in the soil. Since iron had been found by color tests to be generally present in water extracts from the unlimed soils of Field A, while aluminium could rarely 1 Bui. No. 31, Bureau of Soils, U. S. Dept. Agr., 1906. INJURIOUS EFFECT OF SULFATE OF AMMONIA. 121 be detected by the precipitation test with ammonium hydroxide, it was decided to try larger quantities of soil and larger volumes of water, which would permit subsequent concentration and perhaps yield measurable quantities of these elements by the usual analytical methods. Eight kilograms of air-dry soil were put in a percolation jar, the tubulure of wliich was covered with a piece of linen and plugged loosely with glass wool. Enough water was added to saturate the soil, which was then left in the wet condition for two days. Water was then added in portions of 1 liter at a time, each of which ceased dropping from the bottom of the jar before another was added. Eight liters were thus used, and the percolated water was evaporated in a porcelain dish on the water bath until the volume was reduced to 1 liter, which was next filtered tlirough paper and finally through a porcelain filter under pressure, as there was a turbidity which paper would not remove. The clear soil extract was next heated and made slightly alkaUne with ammonium hydroxide. A copious flocculent precipitate formed, which was collected on a filter, washed and then analyzed. When the filtrate was further heated and a few drops of ammonia added, a second precipi- tate, similar to the first, formed and was also analyzed. The two pre- cipitates differed but little in composition, and the results obtained were combined in Table II. Table II. — Constituents of Precipitate obtained in Concentrated Soil Ex- tract, expressed as Milligrams in 100 Grams of Soil. Plot 1. Plot 6. Plot 8. Aluminium oxide (AI2O3I .074 .152 .105 Silica (Si O2), .381 .538 .835 Manganese oxide (Mn304), None. 1.596 .362 Calcium oxide (Ca O) 1.955 None. .225 The precipitate was found to contain but a trace of iron, which is not tabulated as such, but is really included in the aluminium oxide. The calcium which separated in the ammonium hydroxide precipitate was apparently in the form of carbonate, as the precipitate from the extract of plot 1 effervesced vigorously when dissolved in hydrochloric acid, as the first step in analysis. There is a striking difference between the precipitate obtained in the soil extract from plot 1 and those from plots 6 and 8. The protective effect of nitrate of soda on the calcium in the soil is shown in contrast to the depleting influence of ammonium sulfate, with the consequent forma- tion of salts of manganese and aluminium. No effort was made to esti- mate possible calcium or manganese not precipitated by the successive additions of ammonium hydroxide. 122 MASS. EXPERIMENT STATION BULLETIN 176. A second series of percolation experiments was tried in wliich but 1 kilogram of soil was used, and proportionately smaller amounts of water were percolated through it, until the total percolate amounted to 1 liter. The percolate was filtered through porcelain and subsequenth'' 3'ielded no precipitate with ammonium hydroxide. Iron and manganese were both found and determined by the colori- metric methods. Both Umed and unlimed soils from plots 1, 6, 7 and 8 were used in this series. All the extracts yielded colorimetric tests for iron, but only those from the unlimed soils showed any manganese. The results on the unUmed soils are given in Table III. Table III. — Milligrams Manganese Oxide (Mn^Oi) and Iron Oxide {Fe^O'i) removed in Water from 100 Grams of Unlimed Soil. Manganese oxide, Iron oxide, . Trace. .04 1.49 .07 The amounts of manganese from the soils of plots 1, 6 and 8 are closely like those obtained in the previous series with 8 kilograms of soil. The iron obtained is about one-haK the amount of aluminium oxide tabulated in the previous series. There were in the laboratory samples of soil from plots 5 and 6 which were collected four years before, in 1912. Plot 5 had received the same amount of sulfate of ammonia that had been applied to plot 6. Both samples were from the unlimed halves of the plots. One kilogram of each was treated as in the previous experiment. The extracts showed the presence of aluminium and iron, but were most striking in the tests for manganese. Plot 5 yielded 2.36 mg. MugO^, and plot 6 yielded 3.18 mg. Mn304, from 100 grams of soil. This shows that the formation of salts of aluminium, iron and manganese by ammonium sulfate was as marked four years ago as in 1916. All these experiments showed that ammonium sulfate persistently formed soluble salts of aluminium, iron and manganese in the soil of Field A. It was next decided to secure samples of soils from other fields that had received ammonium sulfate as a fertilizer over a considerable period of time. The desired soils were obtained from the agricultural experiment stations of Ohio and Rhode Island by the kindly co-operation of Director Thorne and Director Hartwell. The soil of the Ohio experiment field is a rather heavy clay loam. The samples were taken from Section C of the continuous five-year rotation experiment described in Circular No. 144 of the Oliio Agricultural Experi- ment Station. The plots selected for our purpose were Nos. 8 and 24. INJURIOUS EFFECT OF SULFATE OF AMMONIA. 123 Since 1893 each plot had received acid phosphate and muriate of potash, but plot 8 had not received any nitrogenous fertiUzer, while plot 24 had been dressed with sulfate of ammonia at the rate of 220 pounds per acre during each five-year period. One-half of each plot had received ground Umestone annually at the rate of 2 tons per acre since 1908, wliile the other half had received none during that period. The plots were seeded with clover at the time the soil samples were taken in the faU of 1915. In a letter regarding the samples, Director Thorne said : — For several years there has been practically no clover on the unlimed ammonium sulfate plots in our work. There are occasionally a few scattering plants, but probably not 20 plants on the twentieth-acre plot. . . . When ammonium sulfate is neutralized with lime we get a luxuriant growth. . . . There are usually at the beginning of the season as many clover plants on the unlimed as on the limed land, but they do not get much beyond the nutriment furnished by the seed, and by harvest have disappeared. The soil of the Rhode Island experiment field is a sandy loam. The samples for our use were taken from the permanent plots numbered 23, 25 and 29, wliicb have been repeatedly described in the annual reports of the Rhode Island Agricultural Experiment Station. All tlii-ee plots have received acid phosphate and muriate of potash smce 1893. Plots 23 and 25 have been supphed with nitrogen in sulfate of ammonia, while plot 29 has had nitrate of soda. Plots 25 and 29 have at irregular intervals received apphcations of lime, and in 1915 all three plots received a dressing of it, but in different amounts. Plot 23 received the equivalent of 500 pounds calcium oxide per acre, plot 25 received 1,500 pounds, and plot 29 received 1,000 pounds. This apphcation of 500 pounds per acre on plot 23 was the first in its history, and was made, as Director Hartwell stated, ". . . because it was becoming so very unsuit- able for crop growth." The soils were prepared for investigation by drying them at a moderate temperature, and then sifting them through a coarse screen with seven meshes to the linear inch, which is the same treatment that was used with the soils from Field A. The samples from Rhode Island were used in percolation experiments with quantities of 1 kilogram of soil and 1 liter of percolated water. The clay of the Ohio soils rendered this method impracticable because the water percolated very slowly. The Ohio samples were accordingly put in stoppered bottles, with twice as much water as there was soil by weight, and shaken continuously for two hours in a machine. The solu- tions were first filtered tlii'ough paper and finally through porcelain filters. Aluminium, iron and manganese were tested for, and when present in measurable quantities their amounts Were determined. Aluminium could not be obtained in appreciable quantity from any but the soil from plot 23 of the Rhode Island field. No manganese was 124 MASS. EXPERIMENT STATION BULLETIN 176. found in the extracts from any Rhode Island sample, but was obtained from all the Ohio samples. Iron was extracted from aU but the more heavily limed soils. Table IV. — Milligrams of Aluminium Oxide {AUO3), Iron Oxide {Fe^O^, and Manganese Oxide {Mn^O^^ removed in Water from 100 Grams of Soil. [Soils representing Ohio and Rhode Island experiments with ammonium sulfate.] ^'-^r I™- Oxide. Ma^nganese Ohio plot 8, limed, Ohio plot 8, unhmed, Ohio plot 24, limed, Ohio plot 24, unlimed, Rhode Island plot 23, Rhode Island plot 25, Rhode Island plot 29, None. Trace. Trace. None. .05 .16 None. None. .03 None. .03 .64 3 .27 None. None. Trace. None. None. None. None. The Ohio soil which had received sulfate of ammonia (plot 24) without lime gave a striking reaction for soluble manganese salts similar to our own soils; but in the soils from Rhode Island the sulfate of ammonia seemed to exert its influence on aluminium and iron compounds (plot 23). At a later period samples of soil were received from Prof. F. D. Gardner of Pennsylvania State College, which were taken from different plots on the permanent experiment field at that institution. The soil of the field is a clay loam. The samples were taken from plots 31, 32 and 36. Plots 31 and 32 had received equal amounts of dissolved boneblack and muriate of potash. Plot 31 had sulfate of ammonia appUed at the rate of 240 pounds per acre every two years, while plot 32 received 360 pounds per acre in the same period. Plot 36 received no fertilizer. This treat- ment had been in vogue since 1885. One kilogram of air-dry soil was treated with water by the percolation method. Plot 32 with the heavier application of ammonium sulfate jdelded strik- ingly more iron and a little more manganese than plot 31. The unfertilized soil, plot 36, yielded the most iron, but a negligible amount of manganese. INJURIOUS EFFECT OF SULFATE OF AMMONIA. 125 Table V. — Milligrams of Iron Oxide (Fe-yOs) and Manganese Oxide {Mn-iOi) removed in Water from 100 Grams of Soil. [Soils representing Pennsylvania experiments with sulfate of ammonia.] Plot 31. Plot 32. Plot 33. Iron oxide, Manganese oxide, .28 .13 .58 .15 .01 The results of the chemical investigation of the effect of sulfate of am- monia as a fertilizer in constant use on soils of four different experiment fields show the accompaniment of soluble salts of either aluminium, iron or manganese, or all three together, in the absence of a base like lime. In the presence of calcium carbonate, water has removed no observable amounts of aluminium or manganese salts, and bare traces of iron salts, indicating that Ume either reacts with the ammonium salt promptly, or subsequently breaks up the salts of aluminium and manganese, and also iron salts, almost completely. Paet II. — Water Cultures. Our investigation of the effects of sulfate of ammonia on the soils of Field A included in its progress several series of water cultures in wliich seedlings of rye, barley and clover were used to study the possibilities of poisonous effects from the presence of soluble substances in the soils. In the earliest series there were used water extracts made from soils of plots 1, 6, 7 and 8 for the purpose of learning whether the injurious effect of am- monium sulfate apphed to the soil would appear in the solution obtained from the soil. The soil extracts were prepared in sufficient quantity by mixing soil and water in the proportion of 1 part by weight of soil to 2 parts of water, shaking frequently during a period of two hours, and then allowing the liquid to clear by settling. The water extract was then carefully decanted from the soil. A part of this extract was filtered through porcelain, under pressure, to see whether the poisonous substances, if present in the extract, were colloidal in their nature. Discs of paraffine, reinforced by wire gauze and punctured with numer- ous holes, were arranged by means of suitable corks to float on a basin of water flush with the surface. On these discs the seeds were moistened sufficiently to germinate, and their radicles then penetrated through the holes into the water below. The plan was essentially that described in Bulletin No. 70, Bureau of Soils. As soon as the seedlings were large enough for the purpose, selected ones were transferred to wide-mouthed bottles, which contained the soil extracts. Each bottle contained 250 cubic centimeters, and 4 seedlings 126 MASS. EXPERIMENT STATION BULLETIN 176. were supported in each one through notches cut in the cork stopper. The different series were grouped as follows: — Plot 1. Rye Seedlings. Unlimed soil, iinfiltered extract. Unlimed soil, filtered extract. Limed soil, unfiltered extract. Limed soil, filtered extract. Clover Seedlings. Unlimed soil, unfiltered extract. Unlimed soil, filtered extract. Limed soil, unfiltered extract. Limed soil, filtered extract. The same arrangement was maintained for the soils of plots 6, 7 and 8, and each extract was tested in three different bottles with a total of 12 seedlings. The cultures were maintained for four weeks, at the end of which the seedhngs had begun to wilt. Differences in the seedlings were noted by the end of the first week. Those growing in the extracts from the limed soils were noticeably better as a whole than those in extracts from unlimed soils. Rye seedlings in the unlimed extracts had reddish stems and grew less rapidly. Roots of the clover seedhngs in unlimed extracts began to appear stunted; es- pecially so in the unhmed extracts from plots 6 and 8. When the experi- ment was discontinued the best seedhngs had developed in the extracts from the limed soils of plots 6 and 8, while the poorest plants were in the extracts from the unhmed soils of the same two plots. The roots of the clover in these two extracts were short and thick and lacked branches. Filtered extracts produced the same results as unfiltered ones. A lot of barley seedhngs was next used in the unfiltered soil exi;racts. At the end of the first week the roots in the unhmed extract from plot 6 began to look stunted. By the end of two weeks the seedlings in aU the unlimed extracts showed a tendency to wilt and the tips of the leaves turned white. At the end of the fourth week, when the experiment was stopped, the seedhngs in the extracts from the Hmed soils were uniformly superior to those in the extracts from the unhmed. The poorest seedhngs were in the extract from the unhmed soil of plot 6. The strikingly inferior growth of the different kinds of seedhngs in the extracts from the unlimed soils of plots 6 and 8, which had been dressed with ammonium sulfate, suggested that the poisonous effect might be due to sulfates of aluminium, iron or manganese, wliich were known to occur in extracts from those soils. More culture experiments were accordingly tried from time to time, in which standard nutrient solutions were used instead of soil extracts. Vari- II aniiaiiB o d No. 1, Nutrient sol.; No. 2, Nutrient sol.+CaCOs; No. 3, Nutrient sol.+CaS04; No. 4, Nutrient sol. +2 c.e. Al sol.; No. 5, same as No. 4+CaC03; No. 6, same as No. 4+CaS04; No. 7, Nutrient sol.+l c.c. Al sol.; No. 8, same as No. 7+CaC03; No. 9. same as No. T+CaSOi. ; y ^f*^^^^-^ No. 4, No. 1, Nutrient sol.; No. 2, Nutrient sol.+CaC03; No. 3, Nutrient sol.+CaSOj Nutrient sol. +5 c.c. Fe sol.; No. 5, same as No. 4+CaC03; No. 6, same as No. 4+CaS04; No. 7, Nutrient sol. +2 c.c. Fe sol.; No. 8, same as No. 7+CaCo,i; No. 9, same as No. 74-CaSOi; No. 10, Nutrient sol.+ l c.c. Fe sol.; No. 11, same as No. lO+CaCOa; No. 12, same as No. 10+CaSO4. INJURIOUS EFFECT OF SULFATE OF AMMONIA. 127 ous proportions of ferrous sulfate were added in one series, aluminium sulfate was used in a second series and manganous sulfate in a third. The standard nutrient solution was prepared in two parts: (a) 20.5 gi-ams manganesium sulfate in 350 cubic centimeters of water; and (b) 40 grams calcium nitrate, 10 grams potassium nitrate, 20.56 grams disodium phosphate in 350 cubic centimeters of w^ater. From each of the solutions (a) and (b) were taken 100 cubic centimeters and added to 9,800 cubic centimeters of water, together with a few drops of ferric chloride solution. Tliis diluted nutrient solution was used in the culture bottles. SeedUngs of red clover were used in all these experiments with nutrient solutions, because clover had sho^vTi the greatest susceptibility to the soil influences on Field A. The experiments with sulfates of aluminium and iron have been fully- described in Bulletin No. 161 of this station, and only a summary of the results is given here. Effects of the aluminium and iron salts began to show by the end of the first week, in stunted, tliickened roots, followed in a few days by a smaller growth of leaves, when compared with seedhngs in the check nutrient solutions. Cultures with 43 parts of aluminium in a million, or with only 44 parts of iron, produced these effects, wliile in the higher concen- trations employed the roots were killed.^ Calcium hydrate and calcium carbonate added to the bottles contain- ing aluminium or iron neutraUzed their injurious effects in the lower con- centrations, but were ineffective with high concentrations. Calcium sul- fate w^as entirely ineffective as an antidote. The poisonous effects of the salts appeared to be exerted upon the tips or growing parts of the roots. The rootlets died leaving a tliick, stubby taproot. Microscopic examinations of the roots by Dr. G. H. Chapman showed the cells in the growing parts to be either killed or arrested in their development. Photographs of the clover seedlings which were published in Bulletin No. 161 are reproduced here to show the characteristic effects of the poisonous sulfates of aluminium and iron. Culture experiments in which manganous sulfate was added to the nutrient solutions in graduated quantities were begun after it had been demonstrated that ammonium sulfate fertilization was accompanied by soluble manganese salts in the soils to which no lime had been added. A solution of manganous sulfate, MnS04.4 H2O, was prepared of Vio molecular concentration, and measured amounts were made up to 250 cubic centimeters with the nutrient solution. Certain bottles received fine calcium carbonate and others calcium suKate, so that the solutions in those bottles were approximately saturated with the calcium salt. The scheme of the series is outlined below. 1 In preparing this bulletin it has been noted that in Bui. No. 161, by an unfortunate error in the decimal point, all figures relating to parts per million of iron in the nutrient solutions are only one-tenth as large as they should be. This error caused iron to appear much more toxic than aluminium, as compared in the tables of that bulletin. 128 MASS. EXPERIMENT STATION BULLETIN 176. No. 1. No. 2. No. 3. No. 4. No. 5. No. 6. No. 7. No. 8. No. 9. No. 10. No. 11. No. 12. No. 13. No. 14. No. 15. Standard nutrient solution. With calcium carbonate. With calcium sulfate. With 40 parts manganese per million of solution. With 40 parts manganese and calcium carbonate. With 40 parts manganese and calcium sulfate. With 100 parts manganese per million of solution. With 100 parts manganese and calcium carbonate. With 100 parts manganese and calcium sulfate. With 200 parts manganese per million of solution. With 200 parts manganese and calcium carbonate. With 200 parts manganese and calcium sulfate. With 300 parts manganese per million of solution. With 300 parts manganese and calcium carbonate. With 300 parts manganese and calcium sulfate. The experiment was conducted outdoors in the pot yard instead of in the greenhouse, the seedlings being put under cover at night and during inclement weather. The experiment was continued four weeks. The effect of the manganese was noticed after the first week. The seedlings with manganese did not grow as fast as the checks, and also began to show chlorosis of the leaves. The roots did not have a stunted appearance as was noticed when iron and alummium salts were used, but seemed to be simply underdeveloped. Neither the presence of calcium carbonate nor calcium sulfate had any beneficial effect. In some cases the calcium carbonate seemed to aggravate the to.xicity rather than alle- viate it. When the experiment was discontinued the tops in the most concentrated manganese solutions had died and those in the most dilute had apparently lost all their chlorophyl. The tops and roots of the plants were dried and manganese determina- tions were made on them. The table shows the amounts of manganese found in 1 gram of oven-dried samples. Table VI. — Milligrams of Manganese Oxide {MrizO^ in 1 Gram of Clover Plants. Standard, 40 ppm Mn, 100 ppm Mn, 200 ppm Mn, 300 ppm Mn, The results show that manganese is taken up by the plants in consider- able amounts and is carried into the tops. Concentrations above 100 parts of manganese per million of solution have little effect in increasing INJURIOUS EFFECT OF SULFATE OF AMMONIA. 129 the amount taken up by the plant. While some manganese is carried into the tops, most of it remains in the roots. In order to determine whether calcium carbonate or sulfate had any beneficial action in more dilute solutions of manganese a second experi- ment was undertaken. In this series 10 parts and 20 parts of manganese in a million parts of nutrient solution were, respectively, compared with the standard and with equal amounts of manganese supplemented by calcium carbonate and by calcium sulfate. At the end of three weeks all the seedlings except those in the standard solution showed chlorosis by the hght green or yellowish color of the leaves. The more dilute manganese still had a detrimental effect on the clover plants, but not so marked as in the previous experiments with higher concentrations. Neither of the calcium compounds exerted any beneficial effects, but as in the first experiment seemed, if anytliing, to increase the injury. A third series of cultures was conducted during the winter in the green- house, and concentrations of from 10 parts to 40 parts of manganese per milHon of nutrient solution were again tried with and without calcium carbonate added to the solution. Much cloudy weather caused an in- ferior growth of the clover plants, but the experiment was continued four weeks, and at the end there was the same chlorosis of the leaves when manganese was present. Again, calcium carbonate failed to prevent the chlorosis in the presence of manganese, and instead apparently increased it. Masoni,^ PugUese^ and Aso^ have found that iron salts seem to counter- act the toxicity of manganese. In order to confirm their conclusions one series of experiments was undertaken using a combination of these two salts, another series using manganese plus aluminium salt, and still another series using ii-on and aluminium together. To the standard nutrient solution were added 20 parts of manganese and 2 different quantities of aluminium, 21.6 parts and 43 parts, respec- tively, per milhon of solution, with and without calcium carbonate. A similar series was prepai'ed containing 22 and 44 parts of iron per million, respectively. All the solutions containing iron produced seedlings with darker color than the rest. The roots in the solutions containing aluminium or iron became stunted in appearance whether calcium carbonate was present or not. Manganese and aluminium or iron had no apparent antagonistic effects when present together in a nutrient solution. This toxicity with calcium carbonate is unlike the results reported by McCool,^ who found that calcium chloride would counteract the toxicity of manganese to a marked extent. This may be due to the difference in the solutions and seedlings used, as he used manganese chloride, calcium cliloride and Canada field peas. 1 Staz. Sper. Agr. Ital. 44 (1911), p. 85; Abs. E. S. R. 26. ! Atti R 1st Incoragg. Napoli 6 ser. 65 (1913), p. 289; Abs. Chem. Abs. 9, p. 641. 3 Bui. Agr. College, Tokyo, V. p. 177. « Cornell Agr. Exp. Sta. Memoir No. 2 (1913). 130 MASS. EXPERIMENT STATION BULLETIN 176. Having found that manganese is carried up into the tops of the plants the following experiments were tried to determine if there was an increase in the amount of manganese in the tops of clover grown on plots where the poor vegetation was thought to be due to manganese. The first crop of clover analyzed was the same as that reported in Bulletin No. 161. The tops only were analyzed, and the results were based on dry matter. Table VII. — Milligram of Manganese Oxide {MnJD^ in 1 Gram of Clover. Plot. Fertilizer. Limed Soil. Unlimed Soil. 1 Nitrate of soda Trace. .076 5 Sulfate of ammonia, .054 .193 6 Sulfate of ammonia, .054 .193 7 None .031 .114 8 Sulfate of ammonia, . Trace. .171 The clover from the limed portions of the plots shows very little differ- ence between the different plots. The plants from the unlimed portions show a marked increase of manganese in those plots receiving sulfate of ammonia. In the spring of 1915 samples of clover, grass, clover roots, and grass roots were taken from the limed and unlimed portions of plot 5.^ From the unUmed end two samples were taken, one of normal looking plants and another of poor plants. The plants were brought into the laboratory and the roots carefully washed free of soil, especial care being taken not to break many of the finer roots. The tops were then cut from the roots, and the clover separated from the grass, the same being done with the roots. They were then dried at 75 degrees and ground. The tops were then analyzed for iron, manganese and silica. The roots were only an- alyzed for manganese as it is almost impossible to wash them entirely free from soil wliich would invalidate the results for iron and silica. Plot 5 is fertilized as follows: (NH4)2S04, dissolved boneblack, low-grade sulfate of potash. INJURIOUS EFFECT OF SULFATE OF AMMONIA. 131 Table VIII. — Composition of Clover and Grass Tops and Roots, in Milli- grams per 1 Gram of Dry Sample. Iron Oxide Fe^Os. Manganese Oxide Mm04. Silica Si02. Plot 5, limed clover tops. Plot 5,'limed grass. Plot 5, unlimed good clover, Plot 5, unlimed good grass, . Plot 5, unlimed poor clover. Plot 5, unlimed poor grass, . Plot 5, limed clover roots, . Plot 5, limed grass roots, . Plot 5, unlimed good clover roots Plot 5, unlimed good grass roots, Plot 5, unlimed poor clover and g rass roots, 1.14 1.91 1.34 2.97 Faint trace. .053 Trace. .158 .096 .272 Trace. .138 .091 .218 .245 1.72 19.25 4.82 26.64 5.36 57.35 A study of the table shows that the manganese is taken up to a greater extent by the poor plants, both clover and grass, than by the good plants. The grass seems to be more tolerant than the clover, much more being taken up than by the clover. The results would also seem to indicate that the manganese was not evenly distiibuted throughout the plot, but was more concentrated in spots. As it was rather difficult to find normal clover on the plot it might be said that the spots of better plants were the places of smaller amounts of manganese. A somewhat similar condi- tion has been found by Guthrie and Cohens on a golf green. The variations in the iron content of the good and poor plants are so small as to come within the limit of experimental error. The increased amount of silica in the poor plants is probably due to their more mature state. As the foregoing experiments with manganese salts in nutrient solutions had shown that calcium carbonate did not counteract the toxicity of the manganese, while in the field an application of lime to soil supposedly infertile because of the presence of manganese salts corrected the toxicity, pot cultures were started to determine whether calcium carbonate in the soil could counteract the toxicity of manganese. The soil used was from the unlimed end of plot 7 and the unlimed end of plot 6. As the soil from the unlimed end of plot 6 already contained a large amount of soluble manganese it was first extracted by shaking it for two hours on a mechanical shaker with a volume of water twice that of the soil. The soil was then air-dried and passed through the large sieve (7 holes to the hnear inch). » Agr. Gaz. New South Wales, 21 (1910). 132 MASS. EXPERIMENT STATION BULLETIN 176. Earthenware pots 6 inches in diameter and 5 inches deep were used. Each pot was filled with 2 kilos of the air-dried soil. The lime was applied to the surface and thoroughly worked in. The manganese sulfate was applied in solution. The soil was kept at a 25 per cent, moisture content. The clover seed was first soaked for eight hours in a solution of calcium hypochloride, and then seeded on the surface of the soil and pressed into contact with it. The soil was then covered with a half-inch layer of washed quartz and sand to act as a mulch. The treatment employed is shown in the table, there being two pots in each treatment. The Series of Pot Cultures. Pot. Plots. Soil Treatment. 1 2 3 4 5 G 7 8 9 10 11 12 13 14 Plot 6, Plot 6, Plot 6, Plot 6, Plot 6, Plot 6, Plot 7, Plot 7, Plot 7, Plot 7. Plot 7, Plot 7, Plot 7, Plot 7, None. 2 tons calcium carbonate per acre. Extracted with water. Extracted, and 2 tons calcium carbonate per acre. Extracted, and 80 pounds manganese sulfate per acre. Extracted, and 2 tons calcium carbonate and 80 pounds manganese sulfate per acre. None. 2 tons calcium carbonate per acre. 80 pounds manganese sulfate per acre. 2 tons calcium carbonate and 80 pounds manganese sulfate per acre. 100 pounds manganese sulfate per acre. 2 tons calcium carbonate and 100 pounds manganese sulfate per acre. 150 pounds manganese sulfate per acre. 2 tons calcium carbonate and 150 pounds manganese sulfate per acre. The seeds were planted on March 7 and 8, and began to show above the sand on the 9th, and most of them had sprouted by the 15th, when all the pots were watered for the first time. The plants came up rather unevenly, and some replanting was necessary. The replanting was done with seedlings sprouted on paraffine plates. On April 3 all the pots were thinned to 25 plants. The poorest pots at this time were Nos. 3 and 5, the extracted soil with and without the addition of manganese. All of the pots treated with manganese sulfate without lime were poorer than those receiving lime. On April 24 the above differences were even more striking. The plants on No. 5 had practically all died, while on No. 6, where calcium carbonate had been added, they made a small growth. All of the plants on the extracted soil were poorer than those on the other pots. The extraction had probably removed most of the soluble nutri- ents. The clover was weighed in both the green and dry states, with the INJURIOUS EFFECT OF SULFATE OF AMMONIA. 133 results given in Table IX. The crops were subsequently analyzed for total nitrogen, iron oxide, silica and manganese, the results of which are shown in Table X. Table IX. — Grams of Clover obtained from Pot Cultures. Pot. ' Treatment. Green Weight. Dry Weight. 1 2 3 4 5 6 8 10 11 12 13 14 N'one, Calcium carbonate, Extracted with water, Extracted, and calcium carbonate Extracted, and manganese sulfate, .... Extracted, and calcium and manganese, None, Calcium carbonate, Manganese sulfate (80 pounds) Calcium carbonate and manganese sulfate, . Manganese sulfate (100 pounds), Calcium carbonate and manganese sulfate, . Manganese sulfate (150 pounds), Calcium carbonate and manganese sulfate, . 8.15 22.55 7.00 11.03 5.88 17.83 30.00 32.98 25.78 35.23 25.58 34.78 19.80 34.00 1.20 3.55 1.05 1.60 .70 2.50 4.10 4.65 3.30 4.60 3.00 4.80 2.40 4.95 The soil from plot 6 was noticeably inferior in productivity to that from plot 7, when used in the pots as well as in the field. This is shown by comparing pot 1 vdth. pot 7 and pot 2 with pot 8. Extracting the soil with water diminished the crop, as shown in pots 3 and 4, indicating that soluble plant food was removed by the water, whether toxins were removed or not. The addition of manganese sulfate to the soil produced a marked de- pression in yield on both soils when unaccompanied by calcium carbonate, while the employment of the calcium with the manganese resulted in each instance in an increase of crop beyond that produced by the calcium carbonate alone. These results are in accord with field experiments lately reported by Skinner and Reid.^ Chemical analysis of the clover was confined to the crops from the soil of plot 7. Manganese was found to increase in the clover tops nearly in proportion to the quantities added to the soil. The presence of calcium carbonate in the soil did not prevent the absorption of the manganese to a marked extent; therefore it would seem to have been an antidote for the poisonous effect of -the manganese witlain the plant. The consistent increase of the percentage of nitrogen in the crops Action of Manganese under Acid and Neutral Soil Conditions," Bui. No. 441, U. S. Dept. Agr., 134 MASS. EXPERIMENT STATION BULLETIN 176. treated with carbonate of lime is striking, and has been noted before in our field work, and reported in Bulletin No. 161. There is a singular discordance between the ill results obtained with manganese sulfate and calcium carbonate used together in water cultures and the good effects produced by their joint action in experiments with soil cultures. It is possible that in solutions the greater solubility of manganese sulfate permitted its rapid absorption by the roots in compari- son with the intake of the less soluble calcium carbonate, and injurious results w^ere produced in advance of any possible antidotal effect of the calcium. Table X. — Percentage Corn-position of Dry Clover from Pot Cultures.- Pot. Treatment. Nitrogen. Silica. Iron Oxide. Manganese, Parts in 1,000,000. 7 None 3 04 1.03 .14 Trace. 8 Calcium carbonate, .... 3.25 1.24 .16 Trace. 9 Manganese sulfate (80 pounds), 2.88 .71 .17 .345 10 Calcium and manganese, . 3.73 1.74 .24 .345 11 Manganese sulfate (100 pounds). 3.28 1.00 .20 .640 12 Calcium and manganese, . 3.71 2.39 .29 .599 13 14 Manganese sulfate (150 pounds), Calciumfand manganese, . 3.-15 3.54 2.38 .19 .26 1.157 .893 The roots were carefully washed free of soil, dried and analyzed, but the quantities were very small and determinations could not be made in dupUcate in most instances; therefore the figures have not been included here. Conclusions. The positive presence of soluble salts of iron, aluminium and manganese in soils which have been repeatedly dressed with ammonium sulfate with- out adding lime; the formation of one or more of these salts in soils that were extracted with solutions of ammonium sulfate; and the positively injurious action of manganese sulfate, iron sulfate and aluminium sulfate on seedling plants in w^ater cultures and pot cultures when taken together form a chain of facts which clearly indicates that the injurious effects of sulfate of ammonia when used freely without the accompaniment of lime are due to the formation of these soluble salts in the soils of the fields so dressed. BULLETIN No. 177 OCTOBER, 1917 MASSACHUSETTS AGRICULTURAL EXPERIMENT STATION Potato Plant Lice and their Control By W. S. REGAN This Bulletin is a report on a serious outbreak of potato plant lice in Massachusetts during the summer of 1917, together with details of injury, identification, life cycle and natural factors influencing the destructiveness of the pest, experiments with various insecticidal materials and apparatus, and recom- mendations for control. Requests for bulletins should be addressed to the Agricultural Experiment Station, Amherst, Mass. Massachusetts Agricultural Experiment Station. Trustees. OFFICERS AND STAFF. COMMITTEE. Charles H. Preston, Chairman, Wilfrid Wheeler, Edmund Mortimer, Arthur G. Pollard, Harold L. Frost, The President of the College, ex officio. The Director of the Station, ex officio. Hathorue. Concord. Grafton. Lowell. Arlington. STATION STAFF. Administration. William P. Brooks, Ph.D., Director. Joseph B. Lindsey, Ph.D., Vice-Director. Fred C. Kenney, Treasurer. Charles R. Green, B.Agr., Librarian. Mrs. Lucia G. Church, Clerk. Miss F. Ethel Felton, A.B., Clerk. Agricultural Economics. Alexander E. Cance, Ph.D., In Charge of Department. Samuel H. DeVault, A.M., Assistant. Agricultiire. William P. Brooks, Ph.D., Agriculturist. Henry J. Franklin, Ph.D., In Charge of Cranberry Investi- gations. Edwin F. Gaskill, B.Sc, Assistant Agriculturist. Robert L. Coffin, Assistant. Botany. A. Vincent Osmun, M.Sc, Botanist. George H. Chapman, Ph.D., Research Physiologist. Paul J. Anderson, Ph.D., Associate Plant Pathologist. Orton L. Clark, B.Sc, Assistant Plant Physiologist. W. S. Krout, M.A., Field Pathologist. Miss Mae F. Holden, B.Sc, Curator. Miss Ellen L. Welch, A.B., Stenographer. Entomology. Henry T. Fehnald, Ph.D., Entomologist. Burton N. Gates, Ph.D., Apiarist. Arthur I. Bourne, A.B., Assistant Entomologist. Stuart C. Vinal, M.Sc, Assistant Entomologist. Miss Bridie E. O'Donnell, Clerk. Horticulture. Frank A. Waugh, M.Sc, Horticulturist. Fred C. Sears, M.Sc, Pomologist. Jacob K. Shaw, Ph.D., Research Pomologist. Harold F. Tompson, B.Sc, Market Gardener. Miss Etheltn Streeter, Clerk. Meteorology. John E. Ostrander, A.M., C.E., Meteorologist. Microbiology. Charles E. Marshall, Ph.D., In Charge of Department. Arao Itano, Ph.D., Assistant Professor of Microbiology. George B. Rat, B.Sc, Graduate Assistant. Plant and Animal Chemistry. Joseph B. LiNDSEY, Ph.D., Chemist. Edward B. Holland, Ph.D., Associate Chemist in Charge {Research Division). Fred W. Morse, M.Sc, Research Chemist. Henri D. Haskins, B.Sc, Chemist in Charge (Fertilizer Division). Philip H. Smith, M.Sc, Chemist in Charge (Feed and Dairy Division) . Lewell S. Walker, B.Sc, Assistant Chemist. Carleton p. Jones, M.Sc, Assistant Chemist. Carlos L. Beals, M.Sc, Assistant Chemist. James P. Buckley, Jr., Assistant Chemist. WiNDOM A. Allen,* B.Sc, Assistant Chemist. John B. Smith,* B.Sc, Assistant Chemist. Robert S. Scull, B.Sc, Assistant Chemist. James T. Howard, Inspector. Harry L. Allen, Assistant in Laboratory. James R. Alcock, Assistant in Animal Nutrition. Miss Alice M. Howard, Clerk. Miss Rebecca L. Mellor, Clerk. Poultry Husbandry. John C. Graham, B.Sc, In Charge of Department. Hubert D. Goodale, Ph.D., Research Biologist. Miss Rachael G. Leslie, Clerk. Veterinary Science. James B. Paige, B.Sc, D.V.S., Veterinarian. G. Edward Gage, Ph.D., Associate Professor of Animal Pathology. John B. Lentz,* V.M.D., Assistant. On leave on account of military service. CONTENTS. Economic importance of the pest. Description of potato plant lice, Manner of feeding and nature of injury, Life cycle of the potato plant louse, . Control measures, .... Practical considerations and fundamentals of control, Efficiency of various contact insecticides for the control'of potato Discussion of results, "Black Leaf 40," "Black Leaf 40" and Pyrox, etc, "Nico-Fume" liquid. Fish-oil or whale-oil soaps Kerosene emulsion, Miscible or soluble oils, Lime-sulfur, Spraying apparatus, Summary of control measures. Natural agents in the control of potato plant lice, Acknowledgments, ...... PAGE 135 136 136 137 138 138 139 140 140 141 141 142 142 143 143 143 144 145 146 Publication of this Document appboved bt the Supervisor of Administration. BULLETIN No. 177. DEPARTMENT OF ENTOMOLOGY. POTATO PLANT LICE AND THEIR CONTROL. BY W. S. REGAN. Economic Importance of the Pest. Potato plants among other crops have suffered severely from the attacks of plant lice during the present summer. The extent of injury- has varied considerably according to the infestation. Some potato patches with a mild infestation have shown little injury, and the loss in yield from this source will be negligible. In other fields, judging from the extent to which the tops have been killed, the crop will suffer a loss of from 10 to 50 per cent., and in some instances the destruction has been so complete that it will hardly pay to harvest the crop. The potato plant louse (Macrosiphtim solanifolii Ashm.) is not a new insect to this section, but conditions appear to have been ideal during the spring and early summer for its multiplication to such numbers as to cause devastation in many places where no measures were taken to check it. Nor has injury by this pest been exceptional in Massachusetts this season. Reports from Connecticut, New York, New Jersey, Maryland, Virgmia and Ohio indicate that the potato crop of these States has suf- fered equally as much; and in some of these States, in addition to killing the potato plants in many localities, these lice were becoming dangerously abundant on tomatoes. The potato crop of Maine and Canada has also been severely curtailed during some years in the past due to these pests. In Massachusetts injury to potato plants by plant lice became evident during the second week of July, and rapidly increased in severity until the latter part of the month and early August, when no progressive injury could be noticed, and an examination of previously badly infested fields showed these insects present only in very small numbers, and cer- tainly not numerous enough to cause further material injury this season. This indicates a period of about three to four weeks' time when the plant lice are dangerously prevalent upon potato plants, and reports from other sections, as well as the past history of outbreaks of this pest, indicate that this is about the length of time, dating from their first ap- 136 MASS. EXPERIMENT STATION BULLETIN 177. pearance in injurious numbers, when damage by these insects need be feared. During this brief period potato fields showed injury varying from little to the complete destruction of the plants. Some patches were completely free from infestation, while others near by, apparently no more attractive, were badly injured or destroyed before insecticidal treat- ment could be applied. The gradual disappearance of the plant lice from the potato plants, usually about a month after infestation becomes evident, has, in many cases, been the salvation of the crop. This disappearance is due mainly to natural controlling factors, such as parasitic and predatory enemies, weather conditions and disease, all of which contribute to the destruction of myriads of these insects, and to a natural migration of the plant lice from potato plants to other host plants during the latter part of July and August. These factors will be discussed at greater length later. Description of Potato Plant Lice. Potato lice are yellowish or greenish in color, with an occasional pink form. Some are furnished with wings and can fly readily, while others are wingless and have to depend upon crawling for getting about. When full grown these insects are no larger than a pin head, and because of their color and small size, and the fact that they occur for the most part upon the underside of the leaves, plants may be badly infested and considerable injury result before their presence is noticed. Man^ter of Feeding and Nature op Injury. Plant lice are sucking insects and obtain their food by inserting a bristle-like beak into the host plant, from which the juices are extracted. Thus all feeding is done beneath the surface and within the tissues of the plant. On plants badly attacked the leaves begin to turn yellow, curl up, gradually turn brown and die. Under conditions favoring their growth, an attack by plant lice of a week or two will suffice to kill a large portion of the top of a potato plant, and the development of the tubers must necessarily be affected on plants thus injured. When a leaf or stem becomes too dry to afford suitable feeding ground the plant lice crawl to a fresh leaf, or migrate to other plants and continue their injury. Where plant lice are abundant enough to cause apprehension, the underside of the leaves, stems and blossom stalks will be covered with these tiny creatures, and the plants become covered with honey dew, a sticky substance excreted by these insects. This honey dew is often attacked by a black fungus, which, together with the molted skins ad- hermg to the sticky surface, gives the plants an unhealthy appearance and midoubtedly interferes with proper functioning. In spite of its mumtcness, the beak of the plant louse makes a wound which furnishes a suitable entrance for disease, and even if the infesta- tion with plant lice is insufficient to injure the plants, the infection with disease thus caused may entirely ruin the crop. POTATO PLANT LICE AND THEIR CONTROL. 137 Life Cycle of the Potato Louse. Numerous observations have been made on the life cycle and habits of the potato louse (Bulletin No. 147, Maine Agricultural Experiment Station), but many important details are yet to be learned. Infestation of potato plants during the late spring and early summer is accomplished by a migration of the plant lice, either by flight or by crawling from neigh- boring vegetation. These new arrivals are all females, and begin at once to feed upon the sap of the plants. These females lay no eggs, but in a short time produce living offspring, which are the first of a long series of females, and these likewise in the course of eight to ten days produce living young. Plant lice are prolific breeders, a single female often pro- ducmg as many as 20 young per day. It is, therefore, not astonishing that they should multiply so rapidly and cause such devastation in a comparatively short time. No males or egg-laying females ever occur upon potato plants. The first few generations may be wingless, or at any time winged individuals may appear and fly away to seek fresh plants for their own feeding and for their progeny, thus causmg a more or less even infestation of potato fields. After spending a few weelcs or months upon potato plants, winged individuals called ''fall migrants" appear and leave the potato plants for winter hosts, — plants of the same kind as those from which the spring migration took place to the potatoes. As previously stated, the migration to the winter hosts here in Massachusetts takes place probably to some extent during the latter part of July, but mainly during August, the exact time, however, varying according to seasonal fluctuations of temperature and moisture, and the condition of the potato plants. The early drying out or dying of the potato tops will, no doubt, hasten the appearance of "fall migrants," regardless of whether the drying out is due to injury by the plant lice or to other factors. Observations by Miss Edith Patch, State Entomologist of Maine, seem to indicate that buckwheat and shepherd's purse are among the winter hosts sought by these insects. The migration to the winter host plants is followed by the production of winged males and wingless, egg-laying females. These females lay glistening brownish black eggs upon the leaves and stalks, and in this stage the winter is passed. Control Measures. Practical Considerations and Fundamentals of Control. Under the topic "Manner of Feeding and Nature of Injury," discussed on an earlier page, it was pointed out that plant lice obtain their food by piercing the host plant and sucking the juices from within, no feeding being done on the outer surface. Therefore any poison, such as arsenate of lead or Paris green, which is sprayed over the foliage and must be eaten in order to be effective, would be absolutely useless against plant 138 MASS. EXPERIMENT STATION BULLETIN 177. lice, since these insects pierce beneath the poison before feeding is begun. Accordingly, a contact insecticide, a material which kills by contact with the body, is required to deal effectively with these sucking insects, and satisfactory results with an insecticide of this nature can be expected only when apphcation is absolutely thorough. Each insect must be hit by the spray in order to be killed. Careless work will merely lead to a waste of material, time and energy and to a continuation of the infestation. Such carelessness, frequently due to ignorance of the essentials of apphcation rather than intent, is often the source of complaint that material recom- mended for the control of plant hce is ineffective. Almost invariably unsatisfactory results with standard contact insecticides are attributable to improper apphcation. Since potato Uce confine their feeding almost wholly to the underside of the leaves, care must be taken to direct the spray upward so that the underside of each leaf will be well covered. To apply such a spray before the infestation reaches the distinctly dangerous stage, while it might kill manj^ of the scattered plant Uce, might, on the other hand, be merely a waste of energy, for the amount of injury which the plant hce are going to inflict is purely problematical, so many elements of uncertainty enter in. For instance, weather conditions play an important part in the welfare of the plant hce. Heavy rains wash these deUcate insects from the plants, and cold weather retards their increase. Warm, damp weather is favorable to a parasitic fungous disease which may destroy the plant Uce over large areas. Parasitic and predatory enemies, when conditions are favorable, often destroy such numbers of the plant lice, even after considerable injury to the plants is evident, that control measures are superfluous. Then, too, the natural migration of the plant Uce from potato plants to the winter hosts is an element of some uncertainty. The greater amount of injury may be com- pleted and the plant lice soon be ready to leave the potato plants for the winter hosts before injury to the vines is extensive enough to become particularly noticeable. At tliis time, if the fact were known, it would hardly appeal to the average grower as an economical proposition to insti- tute control measures. AU of these factors combine to make the matter of the dcsirabiUty or necessity of artificial control measures for potato plant lice often a diffi- cult one to determine. Furthermore, it has been the obsei-vation of the writer that in many cases where control measures have been carried out, particularly where improper apphcation made several sprayings necessary, more actual injury was done the plants by the handling and trampUng incidental to such work with a contact insecticide than, it is probable in most cases, the plant lice would have inflicted had the infestation been aUowed to run its course. One application with the proper material, properly appUed to the under- side of the foliage, when the infestation is severe enough to cause evident wilting of the leaves, can in most cases be made economically and to advantage, especially if injury is noticeable before the early part of POTATO PLANT LICE AND THEIR CONTROL. 139 August, when the infestation is more hkcly to be progressive than other- wise. This is especially the case with the average garden potato patch. Over larger areas the practicability of appljdng treatment must be deter- mined by the severity of the infestation, its seasonal importance, — that is, whether it is Uable to be progressive or is past the dangerous stage, — accessibility, available apparatus, etc. Reference has already been made to the fact that the winter is passed in the egg stage of the plant louse upon such plants as buckwheat, shep- herd's purse and possibly various other weeds. On this account "clean culture;" the destruction by burning of potato vines, weeds and other refuse about gardens and potato fields after harvest, unless such material is composted; the burning over of grassy and weedy fields in the vicinity of potato patches in the late fall or early spring; and late fall plowing of gardens are worthy of more general practice. The increased danger to the potato crop from "blight" after infesta- tion with potato lice has already been pointed out. This should em- phasize the need of frequent spraying with Bordeaux mixture or similar fungicide for the remainder of the growing season. Ejjicienqj of Various Contact Insecticides for the Control of Potato Lice. During the early part of July, when injury by potato lice began to cause considerable apprehension, many conflicting reports were received concerning the efficiency of different contact insecticides recommended for the control of these insects. On this account, as well as from the fact that the demand at this time for nicotine sprays so exceeded the supply in many localities that it was impossible to obtain this material, it was thought desirable to have at hand some more definite knowledge con- cerning the effectiveness of the various common contact sprays, in order to be better able to recommend a substitute where any material desired was unobtainable. With this end in view a badly infested potato field, already showing severe injury to the tops, due to the sucking of the plant lice, was selected to carry out these trials, which were conducted by Mr. A. I. Bourne of the Massachusetts Agricultural Experiment Station staff and the writer. All plants treated were thoroughly drenched, the under and upper sides of the foUage ahke, and carefully tagged, check plants being left for com- parisons. It must be kept in mind that a satisfactory contact insecticide combines safety and efficiency with reasonable cost. It must be strong enough to kill the insects and yet not injure the foliage of the plant to which it is applied, and the cost of application must not be excessive. It will be seen from the following report on these experiments that only a comparatively few dilutions of the materials tried met this test. It was impossible, in most cases, to make a very accurate estimate of the per- centage of plant Hce killed, so that where a percentage estimate is given it is intended to show the comparative efficiency of the various insecti- cides tried, and is at best only roughly approximate. It is hardly to be 140 MASS. EXPERIMENT STATION BULLETIN 177. expected that the spraying operations of the average grower will result as successfully as those reported here, where all possible care was taken to thoroughly drench the plants. It should be kept in mind, however, that it is only necessary to reduce the numbers of the plant lice 75 per cent, or more, when they can no longer continue an aggressive attack that will result in serious injury, but must take, figuratively speaking, a defensive position against their ene- mies. The parasitic and predatory enemies of the plant lice are much more resistant to contact sprays than the plant lice themselves, and in no case with the insecticides used where the plants were not injured were these beneficial insects destroyed, although they were present in numbers when appUcation was made. The few plant lice which escape an eflacient spray appUcation fall ready prey to these enemies. A report of the results of these tests follows: — Material and Dilution. Plant Lice killed. Injury to Plants. "Black Leaf 40" (1-400) with soap, "Black Leaf 40" (1-800) with soap, "Black Leaf 40" (1-800) with Pyrox, i soap. "Black Leaf 40" (1-1,000) with soap, "Black Leaf 40" (1-1,600) with soap, "Nico-Fume" liquid (1-750) with soa Fish-oil soap (1-5), Fish-oil soap (1-6), Fish-oil soap (1-8), Kerosene emulsion (1-9), Miscible or soluble oil (1-25), Miscible or soluble oil (1-40), Miscible or soluble oil (1-50), Miscible or soluble oil (1-64), Lime-sulfur, 34° Beaumd (1-22), Lime-sulfur, 34° Beaumfi (1-43), 99-100 per cent., 98-99 per cent., 98 per cent., Not over 75 per cent., Ineffective, few killed, 98 per cent., 98-99 per cent., 98 per cent.. Not over 50 per cent., 90 per cent., Perfect kill, Perfect kill, 98-99 per cent., 98 per cent., Ineffective, not over 20 per cent., Ineffective, .... No injury. No injury. No injury. No injury. No injury. No injury. No injury ? No injury. No injury. No injury. Plants killed. Considerable injury. Some injury. Some injury. Some injury. No injury. Discussion of Results. 1. "Black Leaf 40." — This material is perhaps the insecticide most commonly used for the control of plant hce, but any of the other nicotine preparations of a similar nature now on the market should give satisfac- tory results. It is a concentrated solution of nicotine sulfate, containing 40 per cent, of nicotine by weight. It was tried with four dilutions — 1-400, 1-800, 1-1,000, and 1-1,600 — in each case, with the addition of soap at the rate of 2 pounds to 50 gallons of the diluted "Black Leaf 40." Both ordinary hard laundry soap and liquid soap were used with similar POTATO PLANT LICE AND THEIR CONTROL. 141 results, the hard soap being cut into small pieces and dissolved in boiling water before adding to the solution. If liquid soap is used, 1 quart should be added to every 50 gallons of the diluted "Black Leaf 40." In addition to increasing the effectiveness of this nicotine preparation the soap aids materially as a spreader, thus insuring a more uniform coating of the foUage and a more perfect "hit" of the plant lice. All of the four dilutions tried showed no foliage injury, but only the 1-800 strength met the test of reasonable economy and efficiency. This strength showed nearly a perfect kill. The dilution 1-800 reduced to practical terms is as follows: — "Black Leaf 40," § pint. Hard soap, dissolved in boiling water, . . 2 pounds (liquid soap, 1 quart). Water, ........ 50 gallons. Reduction to a small amount would be as follows: — "Black Leaf 40," IJ teaspoonfuls. Hard soap, dissolved in boiling water, . . . . . f ounce. Water, .......... 1 gallon. The cost of this spray material will depend mainly upon the quantity of the "Black Leaf 40," or similar nicotine preparation, purchased. In an amount of 10 pounds, which diluted as recommended (1-800) would give 1,000 gallons of spray mixture, the cost amounts to but little over 1 cent per gallon. If purchased in an amount as small as an ounce the cost is increased to something over 4 cents a gallon. 2. "Black Leaf 40" and Pyrox, etc. — The question has frequently been asked as to whether or not "Black Leaf 40" can be safely combined with Pyrox, Bordo-lead and other materials, such as arsenate of lead and Bordeaux mixture, thus reducing the labor involved in making separate applications. Pyrox and Bordo-lead are a combination of an arsenical and a fungicide, and are used for the control of leaf-eating insects, such as the potato beetle, and fungous diseases. "Black Leaf 40" and Pyrox or Bordo-lead can be safely combined with equally as good results as when these materials are used separately. However, soap should not be used with such a combination, and should never be used in any combination containing Pyrox, Bordo-lead or Bordeaux mixture, as an "incompatible mixture" results. "Black Leaf 40," or any similar nicotine preparation, may also be safely combined with arsenate of lead or Bordeaux mixture . — but without the addition of soap. 3. "Nico-Fume" Liquid.- — This material is somewhat similar to "Black Leaf 40," being a nicotine preparation containing 40 per cent, free nicotine. There appears to be little or no difference in the effective- ness of these two materials, and since the "Nico-Fume" liquid is the more expensive, it is suggested merely as a possible substitute in case the "Black Leaf 40" is not obtainable. ■ It was used at approximately the same strength as the "Black Leaf 40," and with the addition of a like 142 MASS. EXPERIMENT STATION BULLETIN 177. amount of soap. Combinations of "Nico-Fume" liquid with other in- secticides and fungicides can be made with the same restrictions as for "Black Leaf 40." 4. Fish-oil or Whale-oil Soaps. — These soaps have long been used for the control of plant lice. Three dilutions were tried, — 1 pound to 5 gallons of water, 1 pound to 6 gallons of water, and 1 pound to 8 gallons, the soap being cut up into small pieces, dissolved in boiling water, and diluted with cold water to the required strength. The 1-5 and 1-6 strengths showed high efficiency. The 1-8 strength was unsatisfactory, not more than half of the plant lice being killed. There was some sus- picion of foliage injury at the 1-5 strength, but this was not extensive, and, since some of the tops had been killed by the plant lice, this point could not be definitely determined. The 1-6 strength proved efficient and showed no injury. Used at this strength the cost of fish-oil or whale-oil soap spray is approximately that of the "Black Leaf 40" solution, 1-800; that is, less than 2 cents per gallon where a quantity of the soap to the amount of 5 pounds or more is purchased. Since the amount of soap to be dissolved in case the fish-oil or whale-oil soap is used is greater than the quantity used with the "Black Leaf 40" solution, the latter is perhaps somewhat preferable because of the smaller outlay of time and bother thus involved. These soaps, however, furnish an excellent substitute in case of difficulty in obtaining the nicotine preparation. Pyrox, Bordo- lead, Bordeaux mixture or similar materials should never be used with soap of any kind. 5. Kerosene Emulsion. — This material was made according to the usual stock formula, as follows: — Hard soap, ....... | pound (liquid soap, ^ pint). Water, 1 gallon. Kerosene, 2 gallons. The soap is cut into small pieces and dissolved in the water, wliich should be boiling. The soap solution is then poured into the kerosene while hot, and churned back and forth with a spray pump until a creamy mass is formed and no free oil is present. This can usually be done satis- factorily in from ten to fifteen minutes. The emulsion formed is a stock solution, which should be diluted at the rate of 1 part to 9 parts of water for plant lice. It was supposed that kerosene emulsion, a standard remedy for plant lice and other soft-bodied insects, would prove highly effective against potato lice, but the trials with this material proved disappointing, as not more than 90 per cent, of the insects were killed. This indicates an effi- ciency for kerosene emulsion considerably less than that of the "Black Leaf 40," 1-800, and the fish-oil soap, 1-6. Furthermore, the trouble and time involved in making the emulsion, as well as the danger of foliage injury when this material is improperly made, militate against its use where the other materials referred to above are obtainable. The cost of POTATO PLANT LICE AND THEIR CONTROL. 143 the kerosene emulsion per gallon of the diluted spray is something over 1 cent, or approximately the same as for the "Black Leaf 40" and the fish- oil soap solutions. G. Miscible or Soluble Oils. —4bne of the standard commercial brands of miscible oils was used in these tests, this being tried with four dilutions, — 1-25, 1-40, 1-50 and 1-64. This material in all four dilutions showed a very high killing efficiency, but even at the greatest dilution, 1-64, showed distinct oil injury to the potato foliage. In justice to this material, how- ever, it must be said that the sample experimented with was not perfect, as there was some free oi|, evident; an ever-present danger, nevertheless, with this material. Time did not permit obtaining a fresh sample of mis- cible oil, so that this material must be placed in the questionably danger- ous class until further experiments prove to the contrary. The cost of this material is less than that of any of the other insecticides referred to, and obtained in any quantity would amount to less than 1 cent per gallon of diluted spray material. 7. Lime-sulfur, — A standard commercial brand of this material, hav- ing a density of 34 Beaum6, was used in these tests. Two dilutions were tried, — 1-22, which is about twice the normal strength for application to foliage, and 1-43, which is about the usual dilution for foliage spray- ing. Even at the 1-22 strength this material killed only a comparatively small number of plant lice, and could in no way be considered an effective aphidicide. Furthermore, at this strength there was evident foliage in- jury shortly after application, which took the form of a wilting or droop- ing of the plants. The next day, however, the plants thus injured seemed to have entirely recovered. Spraying Apparatus. Satisfactory spraying outfits for applying insecticides are equally as important as efficient spray materials. Ordinary hand atomizers are use- less, since it would be necessary to turn over every plant so that the under- side of the leaves could be reached. Such handling would probably result in as much injury to the plants as the plant lice would be likely to inflict. For small garden potato patches, perhaps up to a quarter of an acre, a knapsack or compressed-air spray pump will prove satisfactory. These pumps hold from 3 to 5 gallons of spray, but the frequent need of refilling makes them less desirable for use where larger areas are to be treated. In spraying operations involving fairly large potato fields a barrel pump, traction outfit, power sprayer or similar apparatus will be found the only practicable thing. Regardless of the type of pump used, an extension rod and an under- spray nozzle at a right angle to the rod are essential in order that the underside of the leaves may be easily reached. For a knapsack or com- pressed-air pump a 3 or 4 foot extension rod of iron or brass is perhaps most convenient. A 4 or 5 foot length of iron pipe is, perhaps, most satis- factory when directing the spray by hand from a barrel piunp, power 144 MASS. EXPERIMENT STATION BULLETIN 177. sprayer or similar apparatus, but numerous combinations of rods and nozzles may be made to increase the spraying area or the number of rows treated at one time. In the case of traction sprayers or other direct row- spraying apparatus the common inverted T method is ordinarily used with two nozzles attached to throw spray in opposite directions, so that two rows may be treated from each T. By attaching several T's to the main cross rod, so that the T's come between the rows, a number of rows may be sprayed simultaneously. It is essential with such apparatus that the T's be made sufficiently long and the nozzles attached at the proper angle to thoroughly drench the underside of the foliage. Work with such apparatus must be done slowly if satisfactory results are to be expected. Some growers have adopted an arrangement with traction sprayers whereby a cross piece, located a short distance in front of the nozzles, tips over the plants. The nozzles are directed forward and downward so that, theoretically, while the plants are thus tipped over, the underside of the leaves are covered with the spray. Not only is the efficiency of this method open to doubt, but the effect upon the plants of such treatment is worthy of consideration. A nozzle giving a fine mist spray is essential. The disk and Vermorel are two tyi^es of nozzles well adapted for the work. The disk nozzle must be of the angle form, which gives a suitable underspray at a right angle to the rod, and covers a fairly large area, being on this account pref- erable to the Vermorel nozzle. The Vermorel nozzle cannot be purchased in the angle form, but a 45° elbow can be obtained or a bend made in the extension rod to overcome this difficulty. It is fairly well adapted for use with a knapsack or compressed-air pump. Where a considerable length of hose is needed it is desirable to have this as light as possible in order to faciUtate handling among the rows with the least possible injury to the plants. One-fourth inch Meruco tubing has been found highly satisfactory for this purpose, especially for the leading hose. Attachments for this tubing to rubber or cotton hose of larger size can be readily obtained. Long-tail hose couplings will also be found advantageous in preventing a "blow-out" where pressure of any amount is used. Summary of Control Measures. 1. Potato plant hce can be readily controlled by the use of a contact insecticide of "Black Leaf 40" or similar nicotine preparation at the rate of 1 part of this material to 800 parts water, with the addition of com- mon laundry soap, dissolved in boiling water, at the rate of 2 pounds (liquid or soft soap, 1 quart) to 50 gallons of the diluted "Black Leaf 40" solution. The formula in practical terms is given on an earlier page. Fish-oil or whale-oil soap at the rate of 1 pound to 6 gallons of water is about equally as effective, but is less desirable on account of the extra time and bother involved in dissolving larger quantities of soap. "Black Leaf 40" can be combined safely with Pyrox, Bordo-lead, Bor- POTATO PLANT LICE AND THEIR CONTROL. 145 deaux mixture or arsenate of lead, but soap should be omitted when such combinations are made. These combinations are equally as effective as when the materials are used separately. Kerosene emulsion is not liiglily effective against potato plant lice, and the labor involved in preparing this material is also against its use. Tests with miscible or soluble oils seem to indicate that these materials are dangerous to use upon potato foUage. Lime-sulfur is ineffective for the control of potato plant lice even at double the ordinary strength used upon foliage. 2. Satisfactory results with an efficient contact spray can be expected only when thorough work is done. Each insect viust he hit with the spray. Since plant lice confine their work almost wholly to the underside of the leaves, the spray must be directed upward from underneath the plants. An angle disk nozzle or similar underspray nozzle is necessary for such work. One thorough application with an efficient spray should control potato plant lice so that a second treatment will be unnecessary. Too much handling or trampling about the plants will often result in more injury than the plant lice are likely to cause. 3. The practicability of applying treatment for the control of potato lice, especially over large areas, must be determined by the severity of infestation, its seasonal importance, — that is, whether it is likely to be progressive or is diminishing in severity, — accessibility, available appa- ratus, etc. If injury to the plants has not been severe enough to kill por- tions of the tops of the plants to an evident extent before the 1st of August, it is probable that the injury likely to be done will not exceed the cost of applying treatment. When severe injury is noticeable before the 1st of August, a thorough treatment should be made at once. Application before the insects are present in numbers will be merely a waste of time and energy. 4. The destruction by burning of potato vines after harvest, together with all weeds and other refuse about gardens and potato fields, unless such material is composted; the burning over of grassy and weedy fields in the vicinity of potato patches in the late fall or early spring; and late fall plowing of gardens are methods of clean culture which may materially reduce future infestation. 5. Injury by potato lice renders the plants more susceptible to "blight," and should emphasize the need for frequent sprays with Bordeaux mixture. Natural Agents in the Control of Potato Plant Lice. Many factors contribute to a natural control of potato lice; in fact, to such an extent that during most seasons in the past their injury has been unimportant in Massachusetts. Weather conditions rank very high among controlling influences. Cool or wet weather offers quite a decided. check to aphid development, and heavy or continuous rains undoubtedly destroy many of these delicate insects. 146 MASS. EXPERIMENT STATION BULLETIN 177. Among the predatory enemies of plant lice, lady beetles and their young, and the larvse of syrphus flieS; are most important. Both as adults and during the immature stages, lady beetles are voracious feeders upon plant lice as well as upon other tiny insects. The average person readily recognizes a lady beetle and knows its beneficial habits, but the lady beetle young, being of an entirely different appearance, are often mistaken for injurious forms and unfortunately are destroyed. These young vary in length all the way. up to about a half inch, are bluish or blackish in color, often with orange spots on the back, and resemble very much a miniature alligator in general appearance. They crawl about freely, destroying large numbers of the plant lice. The syrphus fly young are maggot-like forms, being pointed at the head end and somewhat broader behind, and are of variable length but average about one-fourth of an inch. These are ordinarily orange, greenish or whitish in color, are very sluggish, but destroy, nevertheless, numbers of the plant lice. Tiny, almost microscopic, wasp-like insects also aid in the destruction of plant lice, their young living parasitically in the bodies of these pests. During certain seasons^ especially when there is an abundance of warmth and moisture, a fungous parasite attacks these plant lice and destroys large numbers. In some localities this disease has been credited with having practically exterminated the plant lice after they had become numerous enough to menace seriously the potato crop. Acknowledgments. The foregoing is not presented as a "distinct contribution to scientific knowledge," but is merely an attempt to present in available form facts already determined by others, together with results of personal observa- tions and experience. The writer wishes to acknowledge credit to Bulletin No. 147, Maine Agricultural Experiment Station, for certain facts and suggestions made use of in this paper; and is indebted to Mr. A. I. Bourne of the Massa- chusetts Agricultural Experiment Station staff for assistance in carrying out the insecticide tests. The work has been carried out under the direct supervision of Dr. H. T, Fernald, whose kind co-operation has been of much help. BULLETIN No. 178 DECEMBER, 1917 MASSACHUSETTS AGRICIILTIRAL EXPERIMENT STATION THE EUROPEAN CORN BORER PYRAUSTA NUBJLALIS HUBNER A RECENTLY ESTABLISHED PEST IN MASSACHUSETTS By S. C. VINAL This Bulletin reports the discovery of the fact that the European com borer, Pyrausta nubilalis Hiibner, has gained a foothold in Massachusetts; gives a brief account of its life his- tory and habits; and suggests methods of control and the prob- able necessity of action by the State or Federal governments. Requests for bulletins should be addressed to the Agricultural Experiment Station ^ Amherst, Mass. Massachusetts Agricultural Experiment Station. Trustees. OFFICERS AND STAFF. COMMITTEE. ' Charles H. Preston, Chairman, . Hathorne. Wilfrid Wheeler, . Concord. Edmund Mortimer, . Grafton. Arthur G. Pollard, . Lowell. Harold L. Frost, . Arlington The President of the College, ex officio. The Director of the Station, ex officio. STATION STAFF. Administration. William P. Brooks, Ph.D., Director. Joseph B. Lindsey, Ph.D., Vice-Director. Fred C. Kennet, Treasurer. Charles R. Green, B.Agr., Librarian. Mrs. Lucia G. Church, Clerk. Miss F. Ethel Felton, A.B., Clerk. Agricultural Economics. Alexander E. Cance, Ph.D., In Charge of Department, Samuel H. DeVault, A.M., Assistant. Agriculture. William P. Brooks, Ph.D., Agriculturist. Henry J. Franklin, Ph.D., In Charge of Cranberry Investi- gations. Edwin F. Gaskill, B.Sc, Assistant Agriculturist. Robert L. Coffin, Assistant. Botany. A. Vincent Osmun, M.Sc, Botanist. George H. Chapman, Ph.D., Research Physiologist. Paul J. Anderson, Ph.D., Associate Plant Pathologist. Orton L. Clark, B.Sc, Assistant Plant Physiologist. W. S. Krout, M.A., Field Pathologist. Mi.ss Mae F. Holden, B.Sc, Curator. Miss Ellen L. Welch, A.B., Stenographer. Entomology. Henry T. Fernald, > Ph.D., Entomologist. Burton N. Gates, Ph.D., Apiarist. Arthur I. Bourne, A.B., Assistant Entomologist. Stuart C. Vinal, M.Sc, Assistant Entomologist. Miss Bridie E. O'Donnell, Clerk. Horticulture. Frank A. Waugh, M.Sc, Horticulturist. Fred C. Sears, M.Sc, Pomologist. Jacob K. Shaw, Ph.D., Research Pomologist. Harold F. Tompson, B.Sc, Market Gardener. Miss Etheltn Streeter, Clerk. Meteorology. John E. Oste-a^nder, A.M., C.E., Meteorologist. Microbiology. Charles E. Marshall, Ph.D., In Charge of Department. Arao Itano, Ph.D., Assistant Professor of Microbiology. George B. Ray, B.Sc, Graduate Assistant. Plant and Animal Joseph B. Lindsey, Ph.D., Chemist. Chemistry. Edward B. Holland, Ph.D., Associate Chemist in Charge (Research Division). Fred W. Morse, M.Sc, Research Chemist. Henri D. Haskins, B.Sc, Chemist in Charge (Fertilizer Division). Philip H. Smith, M.Sc, Chemist in Charge (Feed and Dairy Division) . Lewell S. Walker, B.Sc, Assistant Chemist. Carleton p. Jones, M.Sc, Assistant Chemist. Carlos L. Beals, M.Sc, Assistant Chemist. James P. Buckley, Jr., Assistant Chemist. WiNDOM A. Allen, * B.Sc, Assistant Chemist. John B. Smith, 2 B.Sc, Assistant Chemist. Robert S. Scull, 2 B.Sc, Assistant Chemist. Bernard L. Peables, B.Sc, Assistant Chemist. James T. Howard, Inspector. Harry L. Allen, Assistant in Laboratory. James R. Alcock, Assistant in Animal Nutrition. Miss Alice M. Howard, Clerk. Miss Rebecca L. Mellor, Clerk. Poultry Husbandry. John C. Graham, B.Sc, In Charge of Department. Hubert D. Goodale, Ph.D., Research Biologist. Miss Rachael G. Leslie, Clerk. Miss Grace MacMullen, B.A., Clerk. Veterinary Science. James B. Paige, B.Sc, D.V.S., Veterinarian. G. Edward Gage, Ph.D., Associate Professor of Animal Pathology. John B. Lentz, * V.M.D., Assistant. On leave. » On leave on account of military service. CONTENTS. PAGE and identification, ......... 147 Description of the insect, ......... 148 European history, . . . . . . . . . . . 148 Status of the pest in eastern Massachusetts, . . . . . . 148 Importation, ........... 148 Present distribution, ......... 149 Food plants 149 Importance 149 Character of injiiry, .......... 150 Life history and habits, .......... 150 Control, 151 Co-operation, . . . . . . . . . . . 152 Publication of this Document approved by the supebtisob of administration. BULLETIN ^o. 178. DEPARTMENT OF ENTOMOLOGY. THE EUROPEAN CORN BORER, Pyrausta nuhilalis Hiibner, A RECENTLY ESTABLISHED PEST IN MASSACHUSETTS. BY S. C. VINAL. Nearly every year we find a new insect pest of foreign origin has become established in some section of the United States. To the long list of Euro- pean pests now found in Massachusetts this article adds one more, — the European corn borer or corn pyralid, Pyrausta nubilalis Hubner, recently established in the vicinity of Boston, Mass. This species has long been recorded as one of the most serious enemies to maize culture in Europe, and if not checked may in time become a very serious pest to America's great corn crop. Discovery and Identification. During the past summer the writer found many corn plants in the vicinity of Boston, Mass., being tunneled by light colored caterpillars, the identity of which was unknown . During July nearly every infested plant could be readily detected, having its tassel broken over and hanging pen- dent just above the first two or three spikes. This was due to the larval tumiels in the pith of the main tassel stalk so weakening it that the wind readily blew it over. Early in August moths emerged from pupas collected in the field, and having Dr. C. H. Fernald's collection of both native and exotic moths available, a successful attempt was made to determine the species. Speci- mens of both male and female pyxaUd moths which corresponded identi- cally to those obtained from infested corn stalks in eastern Massachusetts were found in his European collection. These were determined by M. Ragonet, a French lepidopterist, and were labeled Pyrausta (Botys) nubilalis Hubner. Further proof of the identity of this moth was obtained by submitting specimens to Mr. H. G.'Dyar of the United States National Museum, Washington, D. C, who determined them to be Pyrausta nubilalis Hubner, a native of Europe. 148 MASS. EXPERIMENT STATION BULLETIN 178. Description of the Insect. When full grown the larva is 1 inch in length; the body is flesh-colored, often somewhat smoky or reddish above, while the head is flat and dark brown in color. On close observation a transverse row of four light colored spots, with two smaller ones immediately behind them, can be seen on each abdominal segment. From each of these light colored areas a short, stout spine arises, and this character distinguishes the European com borer from the mature caterpillar of the potato and corn stalk borer {Pa-pai-pema nitella Gn.). The female moth has a robust body, is pale yellow in color and has a wing expanse of a little over 1 inch. The outer third of the fore wing is traversed by two serrated lines darker than the rest of the wing, while the hind wings are light yellow in color. The male moth has a long, slender body, is slightly smaller in wing ex- panse, and in color is reddish brown, being much darker than the female. Between the two serrated lines mentioned above is a pale yellow streak, and near the middle of the fore wdng are two small yello-\vish spots. The hind wings are grayish and crossed by a broad band of pale yellow. European History. Pyrausta nubilalis is widely distributed in Europe and Asia, having been reported in literature as occurring in Central and Southern Europe, West Central and Northern Asia and Japan. Its food plants in these widely separated localities consist of corn (except fodder corn), hemp, hops, miUet and several wild grasses. Corn and hop plants are severely damaged by this pest, 50 per cent, of these crops being destroyed in some sections of Central Europe. Foreign literature contains a large number of references to the serious damage caused by the larvse of P. nubilalis, but there is a decided lack of literature dealing with its biology and control. . Status of the Pest in Eastern Massachusetts. Importation. The questions naturally arise as to how, when and where the European corn borer was introduced. At the present time these cannot be definitely answered, but a few deductive conjectures may be given. The important European food plants of P. nvbilalis consist of corn, hemp, hops and millet. Of these the only food plant offering ideal condi- tions for its importation is hemp. This crop is grown to some extent in Southern Europe, and probably some plants infested by larvai of P. nvbilalis were cut and shipped during the fall and winter months to a cordage company in the vicinity of Boston, Mass. These plants were not used inmiediately, and the larvse transformed to pupse in early spring, and soon emerged as moths. On finding corn plants growing in the THE EUROPEAN CORN BORER. 149 vicinity, oviposition took place and the European corn borer became established. '*^ Early sweet corn grown in market gardens 10 to 12 miles inland has been seriously attacked by this pest for the past three or four years, and from this we might infer that it was imported about 1910. A survey of eastern Massachusetts showed that some towns located at the mouth of the Mystic River were more generally infested than others. At the mouth of this river is located the Charlestown Navy Yard, which probably has one of the largest "rope walks" in Eastern United States. Whether the European corn borer was first introduced at the Navy Yard, or at some cordage company located on the opposite bank of the river, it has been impossible to ascertain, but enough has been written to show that it probably was first established in this vicinity. Present Distribvtion. The area infested by the European corn borer in Massachusetts is approximately 100 square miles in extent, and is located immediately north and northwest of the city of Boston. The places most severely infested during the past season were Somerville, Medford, Maiden, Everett, Chelsea, Revere, Lynn, Saugus, Melrose, Stoneham, Winchester, Arling- ton, Belmont, Cambridge, Brookline and the follomng parts of Boston: South Boston, Brighton, Roxbury and Dorchester. Food Plants. At the present time sweet corn is the only valuable commercial crop seriously attacked by this pest, for the other food plants — hops, hemp and millet — are not grown within the infested region of Massachusetts, The most commonly infested weeds and grasses are barnyard grass (EcJiinochloa crus-galli Beauv.), pigweed {Amaranthus retroflexus L.) and foxtail grass {Setaria glauca Beauv.). Dahlia stems are also injured by the European corn borer. The moths apparently prefer to oviposit on corn, and will not infest weeds and grasses unless com plants are not available in sufficient numbers. Importance. Sweet corn is practically the only corn grown within the infested area, and the amount of damage caused by the European corn borer depends upon whether it is an early or late variety. The early crop of sweet com is picked during late July and early August, and by reference to the life history it will be seen that these plants are subjected to the attack of the first brood of larvse only. The late corn, however, suffers from the attack of both the first and second broods of larvai. While the early crop may be damaged to the extent of 10 to 20 per cent., the loss to late corn plant- ings may be as high as 75 to 80 per cent. This higher percentage of damage to late corn is caused by the habit of the small second brood larvse of boring through the husk and tunneling in the developing ear, making it worthless for market. 150 MASS. EXPERIMENT STATION BULLETIN 178. Chaeactee of Injuey. With the exception of the leaf blades the whole corn plant above ground is subject to the attacks of these voracious caterpillars. The larvse after emerging from the egg either commence feeding on the unopened staminate flowers borne by the tassel, or immediately pierce the sheath near its junction with a node. Those which feed on the tassel bore a hole in the side of the buds and- feed on the internal succulent parts. Soon these small caterpillars leave the tassel buds and enter the tassel stalks, or terminal internode, where they tunnel through the pith and finally complete their larval life in this inteniode. These tunnels so weaken the terminal internode that it soon becomes broken over, a type of injury which is especially noticeable on the early com crop. It is quite evident that this injury indirectly affects the formation of corn on the cob by destrojdng the pollen necessary for fertilizing the com silk. Those larvae which do not feed on the tassel inmiediately pierce the sheath surrounding an internode, 'usually where the edges overlap at its junction with a node. Here they feed on the internal surface of the sheath, excavating a groove halfway around the stalk, and then bore directly into the pith where they form long winding tunnels. Whenever the larvae during their tunneling operations reach a node, a rather large cavity is usually formed. From this cavity the larvse sometimes bore through the node, "but more often they turn and tunnel in the opposite direction in the originally infested internode. At the termination of the feeding period nearly all of the central portion of the stalk has been eaten, and this so weakens the plant that a strong wind is likely to break over the stalk, thus completing the destruction commenced by the caterpillars. A number of these stalk-boring larvse very often attack the small stalk or pedicel bearing the ear, and in some cases may bore directly through this into the developing ear. This injury to the pedicel causes the ear to wither and die. The most serious damage to the crop is caused by the large percentage of the second brood lai-vae which immediately enter the ear after hatching. The injury by this brood to the corn ear is very similar to that caused by the well-known corn ear worm {Chloridea obsoleta Fab.). Besides feeding ■on the kernels in a similar manner to the corn ear worm, the European com borer exhibits characteristic tuimeling habits and bores through the cob. Life Histoey and Habits. As the life history has not been thoroughly worked out, it is only pos- sible to give a brief r6sum6 of it at the present time. There are two broods a year of the European corn borer. Hibernation takes place as full grown or nearly full grown larvae, within their tunnels in the corn stalks, and in some cases in the cob. These larvae pupate in the spring and emerge as moths, probably the latter part of May. Soon after emergence the females begin laying eggs on the corn stalks, and in a THE EUROPEAN CORN BORER. 151 few days these hatch. The young larvse begin feeding at once, and quickly eat their way through the sheath before they tunnel in the main stalk. On reaching maturity, which occurs the latter part of July, the larvjB clear out a portion of the burrow, prepare an opening through which the adults can escape, and after spinning a thin silken partition across the top and bottom of this cleared space, transform to pupse. The moths emerge for the second brood in about two weeks. This brood of larvoe becomes full grown by late fall, but does not transform to pupae at once as in the first brood. Instead, the winter is passed as larv'se within the stalks, pupation taking place the following spring. Control. From the brief sketch of the life history it is apparent that there is no hope of destroying this pest during the summer by the use of insecticides, since all of its transformations take place within the plant. Our main hope lies in the possibility of establishing a system of cultural methods which wall enable us to prevent injury. The fact that the winter stage is passed in the food plant suggests control measures which should result in killing the great majority of the hibernating insects. These measures, if carefully followed, should reduce the injury of the following season ma- terially. 1 . Burning the Stalks during the Fall or Winter. — While this is un- doubtedly one of the most effective measures for the destruction of the hibernating insects which can be adopted, it is somewhat wasteful, for the stallcs are valuable either for feed or as a source of humus so necessary for maintenance of fertility and texture in the garden soil. Burning, there- fore, is inadvisable when other effective methods can be used. 2. Burying the Stalks. — In home gardens the stalks may be put in trenches and covered by at least 1 foot of soil. In larger market gardens the stalks may be placed in the center of manure piles until decomposed. In some cases plowing under might be resorted to, but the work must be thorough or it will be ineffective. Any stalks left on the surface are likely to harbor a crop of borers for the next season. If corn stalks are distributed over the land and then cut up by running a disk harrow over the field in both directions it should be possible to turn them practically all under. It should be clearly understood that half-hearted work is of little value. Occasional stalks which it may seem hardly worth the trouble to clean up are likely to harbor enough borers to severely infest the spring crop. 3. Feeding the Stalks. — From the economic point of view this is the best possible means of destroying the hibernating insects, since the value of the stalks for fodder is not materially affected by the presence of the insects, and if properly carried out this method must result in the destruc- tion of practically all of them. Feeding the stalks whole will be relatively ineffective, since parts not eaten by the animals are likely to harbor in- sects. Shredding the stalks, whether to be fed green or dry, must greatly 152 MASS. EXPERIMENT STATION BULLETIN 178. reduce the chances that any of the insects will survive. Ensilage by- ordinary methods must prove a highly effective method of destroying the insects present in the stems or other parts of the affected plants, for it would seem to be ia the last degree improbable that they could survive under the conditions existing in the silo. Co-operation. It has been pointed out that the caterpillars which survive the winter emerge as moths which fly freely the following spring. Consideration of this fact makes it apparent that no method of control can be even fairly satisfactory unless all those cultivating corn in an infested district co- operate to insure as far as may be possible the destruction of all hiber- nating insects. A few neglected gardens in any vicinity may harbor enough borers to infest a wide area. Measures for insuring or compelling satisfactory handling of all infested material are, therefore, very necessary, and, while the desired end might possibly be obtained by local organizations of farmers and gardeners and vigorous action, it seems probable that the matter must be taken in hand by the State or Federal government if the insect is to be brought under control. BULLETIN No. 179 NOVEMBER. 1917 MASSACHISETTS AGRICULTIRAL EXPERIMENT STATION The Greenhouse Red Spider attacking Cucumbers and Methods for its Control By STLART C. VINAL This Bulletin deals primarily with the development and dis- covery of an efficient control for the common red spider attack- ing cucumbers grown under glass; and secondarily with the more biological phases, including distribution, importance, life history and habits of this pest under greenhouse conditions. Requests for bulletins should be addressed to the Agricultural Experiment Station, Amherst, Mass. Massachusetts Agricultural Experiment Station. ■Trustees. OFFICERS AND STAFF. COMMITTEE. Charles H. Preston, Chairman, . Hathorne Wilfrid Whekler, . . Concord. Edmund Mortimer, . . Grafton. Arthur G. Pollard, . Lowell. Harold L. Frost, . Arlington The President of the College, ex officio. The Director of the Station, ex officio. STATION STAFF. Administration. William P. Brooks, Ph.D., Director. Joseph B. Lindsey, Ph.D., Vice-Director. • Fred C. Kenney, Treasurer. Charles R. Green, B.Agr., Librarian. Mrs. Lucia G. Church, Clerk. Miss F. Ethel Felton, A.B., Clerk. Agricultural Economics. Alexander E. Cance, Ph.D., In Charge of Department. Samuel H. DeVault, A.M., Assistant. Agriculture. William P. Brooks, Ph.D., Agriculturist. Henry J. Franklin, Ph.D., In Charge of Cranberry Investi- gations. Edwin F. Gaskill, B.Sc, Assistant Agriculturist. Robert L. Coffin, Assistant. Botany. A. Vincent Osmun, M.Sc, Botanist. George H. Chapman, Ph.D., Research Physiologist. Paul J. Anderson, Ph.D., Associate Plant Pathologist. Orton L. Clark, B.Sc, Assistant Plant Physiologist. W. S. Krout, M.A., Field Pathologist. Miss Mae F. Holden, B.Sc, Curator. Miss Ellen L. Welch, A.B., Stenographer. Entomology. Henry T. Fernald, Ph.D., Entomologist. Burton N. Gates, Ph.D., Apiarist. Arthur I. Bourne, A.B., Assistaiit Entomologist. Stuart C. Vinal, M.Sc, Assistant Entomologist. Miss Bridie E. O'Donnell, Clerk. Horticulture. Frank A. Waugh, M.Sc, Horticulturist. Fred C. Se.ars, M.Sc, Pomologist. Jacob K. Shaw, Ph.D., Research Pomologist. Harold F. Tompson, B.Sc, Market Gardener. Miss Etheltn Streeter, Clerk. Meteorology. Microbiology. John E. Ostrander, A.M., C.E., Meteorologist. Charles E. Marshall, Ph.D., In Charge of Department. Abac Itano, Ph.D., Assistant Professor of Microbiology. George B. Ray, B.Sc, Graduate Assistant. Plant and Animal Chemistry. Joseph B. Lindsey, Ph.D., Chemist. Edward B. Holland, Ph.D., Associate Chemist in Charge (Research Division) . Fred W. Morse, M.Sc, Research Chemist. Henri D. Haskins, B.Sc, Chemist in Charge (Fertilizer Division) . Philip H. Smith, M.Sc, Chemist in Charge (Feed and Dairy Division). Lewell S. Walker, B.Sc, Assistant Chemist. Carleton p. Jones, M.Sc, Assistant Chemist. Carlos L. Beals, M.Sc, Assistant Chemist. James P. Buckley, Jr., Assistant Chemist. WiNDOM A. Allen, i B.Sc, Assistant Chemist. John B. Smith, » B.Sc, Assista7it Chemist. Robert S. Scull, i B.Sc, Assistant Chemist. James T. Howard^ /nspedor. Harry L. Allen, Assistant in Laboratory. James R. Alcock, Assistant in Animal Nutrition. Miss Alice M. Howard, Clerk. Miss Rebecca L. Mellor, Clerk. Poultry Husbandry. John C. Graham, B.Sc, In Charge of Department. Hubert D. Goodale, Ph.D., Research Biologist. Miss Rachael G. Leslie, Clerk. Veterinary Science. James B. Paige, B.Sc, D.V.S., Veterinarian. G. Edward Gage, Ph.D., Associate Professor of Animal Pathology. John B. Lentz, i V.M.D., Assistant. » On leave on account of military service. CONTENTS, PAGE Introduction, ........... 153 History and distribution, ......... 154 Food plants, ............ 165 Nature of injury to cucumbers, ......;. 156 Economic importance of the pest on cucumbers, ..... 156 Life history, ............ 157 Feeding habits and dispersion, ........ 159 Natural enemies, ........... 160 Introduction to experiments, ......... 160 Experiments conducted in the laboratory, ...... 161 Fumigation experiments, ......... 161 Spraying experiments, ......... 163 Summary of materials found to be efficient experimentally, . . . 169 Experiments conducted in commercial greenhouses, ..... 169 Lemon oil, . . . . . . . . . . 170 Linseed oil emulsion, ......... 170 Conclusions drawn from commercial spraying experiments, . . . 172 Prevention, . . . . ... . . . . . . 172 Control measures, ........... 172 Preventive measures, . . . . . . . . .172 Repressive measures, ......... 174 Control of red spiders attacking other crops, . . ... . . 179 Summary, . ........... 180 Bibliography, 181 Publication of this Document approved by the Supervisor of Administration. BULLETIN No. 179. DEPARTMENT OF ENTOMOLOGY. THE GREENHOUSE RED SPIDER ATTACKING CUCUMBERS AND METHODS FOR ITS CONTROL. {Tetramjchus bimaculatus Harvey.) (Class, Arachnida; Order, Acarina; Family, Tetranychida.) BY STUAET C. VINAL. INTRODUCTION. The minute spimiing mites, commonly called red spiders, have long been known as among the most troublesome of greenhouse pests, although they also cause a great deal of damage to flowers, vegetables and trees growing out of doors. A greenhouse affords an almost ideal environment for the development and rapid multiplication of red spiders, and as a consequence we find this pest taking advantage of the opportunity offered and doing great damage to many of the principal crops groMoi in green- houses. The production of vegetables under glass is an expensive process, in- volving a large investment of capital and a continual expense to maintain such an establishment. To counterbalance this expense the value of the crop must be proportionally high, and anything which interferes with the fullest development of the plants reduces the profits materially. Without doubt the common red spider {Tetranychus bimaculatus Harvey) is the most widely distributed and destructive pest of green- house cucumbers. Nowhere in America is the cucumber forcing industry more highly developed than in the market-garden district of Boston, Mass., and therefore the injury caused by this pest assumes its greatest economic importance in this section. During the last few years numerous inquiries have been received by the Massachusetts Experiment Station from market gardeners in regard to the control of red spiders attacking greenhouse cucumbers. Because of the lack of an efficient method of control very few recommendations 154 MASS. EXPERIMENT STATION BULLETIN 179. could be given, and in many cases the injury by these mites resulted in serious losses. Thus it soon became evident that some line of investiga- tion should be conducted on the control of this mite attacking greenhouse crops, and in October, 1915, this problem was assigned to me. The investigations upon which this paper is based were carried on under the direct supervision of Dr. H. T. Fernald. The thanks of the wTiter are due Dr. H. T. Fernald, Dr. G. C. Crampton and Dr. W. S. Regan for their interest throughout the progress of the work. Acknowledgments are also due the chemistry department of the station for its co-operation, expecially to Dr. E. B. Holland for his interest and careful manufacture of many complicated spray materials which led to the discovery of an efficient control for the greenhouse red spider. The writer is also under obligations to Mr. H. F. Tompson, professor of market gardening, for suggesting this research and for much valuable information concerning the efficiency of control measures when used in commercial houses. To Mr. M. E. Moore of Arlington and Mr. J. Winthrop Stone of Watertown the writer gratefully acknowledges his indebtedness for their kind co-operation in allowing promising materials to be thoroughly tested on a commercial scale in their greenhouses. As this paper has to deal primarily with the control of the greenhouse red spider, other more biological phases will be discussed only briefly, unless they have a direct bearing upon control measures. HISTORY AND DISTRIBUTION. The greenhouse red spider of New England was first described by Harvey in 1893 as Tetranychus bimaculatus. He considered it distinct from the European species Tetramjchus telarius Linn., and later workers have failed to prove conclusively the identity of these species. The first account of serious injury caused by this mite in the United States came from the New England States, where it caused much damage to greenhouse plants. In 1855 a mite, since described by Banks as T. gloveri, but now known as T. bimaculatus Harvey, was reported bv Glover as doing injury to the cotton plants of the south. This injury increased in importance, and in 1900 the Bureau of Entomology, United States Department of Agriculture, established a southern laboratory to work on the control of this pest. With the development of greenhouses in the west the ravages of the red spider scon appeared and caused serious darnage to greenhouse plants as well as to many cultivated garden plants and fruit trees. A closely related mite has long been a serious pest of hop plants in Europe; therefore it is not surprising that our species of red spider assumes a great importance in seriously damaging hop fields both in the east and far west. The red spider, therefore, is very generally distributed throughout the United States, extending from Maine to Florida and westward to Texas and California, only a few States in the western arid region being exempt from the ravages of this pest. GREENHOUSE RED SPIDER. 155 FOOD PLANTS. Tetranychus bimaculatus is very cosmopolitan in its feeding habits, having been listed by McGregor as feeding on 183 species of plants, 55 per cent, of which were cultivated, in the southeastern part of the United States. Much confusion has arisen because of the large number of host plants and the variability in color of mites feeding on these different plants. New species have been described based upon these color varia- tions, but they have been discarded by later workers as synonymous. Under New England conditions of climate the red spider as a rule does not seriously damage plants except those which are usually grown in greenhouses. A few exceptions to this statement may occur near badly infested greenhouses or during very dry seasons. As this paper has to deal with greenhouse control, only those plants found most often infested in, and in the vicinity of, greenhouses will be enumerated. The greenhouse vegetables most subject to attack are (1) cucumbers, (2) egg plants and (3) tomatoes. Cucumbers grown under glass in the market-garden district of Boston are rarely exempt from the attacks of red spiders. These plants are first attacked when only two leaves have unfolded, and injury continues until the death of the plant, which in the majority of cases is due primarily to the removal of chlorophyll from its leaves by the mites. Egg plants, although very susceptible to attack, are not generally grown in the vicinity of Boston. Greenhouse tomatoes appear to be practically immune from red spider injury except when very young. Several times the writer has seen a greenhouse containing approximately 1,500 full-grown cucumber plants, with a row of tomatoes planted at each end of the house. The cucumber plants were rapidly dying from the injuries caused by millions of .red spiders, while the tomatoes remained unaffected. This was an extremely severe infestation, and shows to what extent greenhouse to- matoes are immune. Almost all weeds found in infested greenhouses harbor mites, and if not destroyed are liable to infect a following crop. The greenhouse flowers subject to attack are (1) roses, (2) violets, (3) sweet peas, (4) carnations, (5) chrysanthemums and (6) many others of minor importance. In floriculture perhaps the most important infestations occur on roses and violets, with sweet peas, carnations and chrysanthemums next in order. Usually a very large number of widely differing plants are grown in a florist's greenhouse, and many of these will become more or less seriously infested by the migration of mites from one or more of the above- mentioned plants. However, these infestations are usually not of great importance. The plants in the vicinity of greenhouses subject to attack are (1) beans, (2) egg plants, (3) celery, (4) tomatoes, (5) strawberries, (6) clover, (7) grasses and (8) weeds. Plants subject to attack which are found near greenhouses may serve 156 MASS. EXPERIMENT STATION BULLETIN 179. as sources of inside infestation, or may in turn become infested from plants or parts of plants thrown out of the greenhouse during or after an infestation. The most important garden crops attacked are the bean, egg plant and celery. Tomatoes grown out of doors are more susceptible to red spider injury than when grown in greenhouses. Strawberry plants are also subject to attack, but usually this does not assume great im- portance undgr New England climatic conditions. The most important plants, as far as the greenhouse man is concerned, are those found around most greenhouses, consisting of clover, grasses and weeds, as these are undoubtedly important factors in causing inside infestation. NATURE OF INJURY TO CUCUMBERS. The first signs of injury appear soon after the plants have been trans- planted in the greenhouse, and in the majority of cases on the oldest, basal leaves. The pests usually attack the leaves of a cucumber plant progressively; that is, the older, basal leaves first show injury, then those just above are attacked, and thus the ravages of the pest progress upward as the plant grows. As a general rule very young, hairy leaves around the terminal shoot are exempt from attack until the plant becomes very heavily infested. The injury is caused by the puncturing of the under surface of the leaf and the extraction of the liquid contents of the leaf cells immediately sur- rounding the puncture, which results in a very characteristic and notice- able injury. In the process of feeding, the green chlorophyll is withdrawn, leaving a small dead area which soon appears on the upper surface of the leaf as a small whitish speck. As the mites continue feeding, the removal of chlorophyll and specking increases until ultimately the leaf becomes yellowish, lifeless and useless for food assimilation. The characteristic red spider injury is quite easily recognized, even-in its early stages of development. The normal leaf is opaque, allowing no light to pass through it, while around injured areas considerable light passes through the leaf tissue, due to the lack of chlorophyll in this vicinity. The contrast between the opaque normal leaf tissue and the lightness seen around affected areas is especially noticeable when the cucumber plants have become full-grown and have leaves and terminal shoots running over the top wires, for at this time the leaves are between the source of light and the observer walking beneath them. The appearance on the upper surface of the minute, pitted dead specks or spots, usually arranged in clusters, will also point to infested areas. ECONOMIC IMPORTANCE OF THE PEST ON CUCUMBERS. The damage caused by red spiders in cucumber houses varies in severity. The factors influencing this have not been determined, but at least they are very complicated. The severest injury seems to occur in houses containing a light sandy soil, while houses having heavy soils are better GREENHOUSE RED SPIDER. 157 able to withstand the attacks of this pest. Nearly every cucumber grower in the Boston district, so far as the writer has been able to determine, is forced to fight red spiders in order to bring his crop to maturity. In many cases vvhole houses of j'oung cucumber plants have been destroyed with sulfur fumes because the mites were so numerous and the injuries so severe that it was deemed wise by the grower to destroy the plants and reset the house. The usual methods used by greenhouse men to combat this pest consist of severe pruning of infested plants and spraying with as strong a stream of water as these delicate plants ^vill stand, repeating this as often as possible without allowing mildew to seriously injure the leaves. In nearly all cases the mites whi out m the struggle for existence, and shorten the life of a cucumber plant over one month. Under normal conditions the plant should betir a large amount of fruit during this time. The loss, therefore, to cucumber men by red spider infestation is due to shortening the life of the plant during its productive period. A conservative estimate of the value of the cucumber crop grown within the market-garden district of Boston is $1,500,000 per season. The cucumber growers suffer a loss of approximately $150,000, or 10 per cent, of the whole crop, from the ravages of the red spider alone. Many individual growers have estimated their loss between $2,000 and $5,000 annually. LIFE HISTORY. An examination of infested cucumbers will reveal the presence of tiny transparent eggs, resembling minute dewdrops, attached to the under surface of a leaf or interwoven among the silvery threads which the mites are capable of spinning. In developing from the egg to the adult stage the red spider follows one of two distinct courses, depending on the sex. With the female the egg hatches in about four or five days to a tiny colorless, six-legged form known as the larva, which feeds actively for a little over one day. At the end of this time the larva becomes firmly attached to the leaf and enters a quiescent premolting period which lasts for one day. At the termination of this time the skin is shed and there appears an eight-legged form called the primary nymph or protonjrmph, which feeds for approximately one day and then enters a quiescent pre- molting period. The duration of this period is approximately the same as that of the larval quiescent stage. From this premolting period there emerges the secondary nymph or deutonymph, which is probably the most voracious of the immature mites. The deutonyinphal stage is divided into an active feeding period and a quiescent period, each of which requires one day for its completion, after which the adult female emerges from the deutonymphal molt. For the development from egg to adult it takes seven to eight days under favorable conditions of tem- perature. (See table on page 159.) The stages of the female red spider and their duration may be represented as follows : — v„,r To,-rro Quiescent Proto- Quiescent Deuto- Quiescent aj u o i.gg. l.arva. j nymph. II. nymph. III. ^d"'* ?" I 1 1 1 1 1 1 H— 1 4-5 days. 1}^ days. 1 day. IH days. 1J4 days. IJ^ days. 1 day. 15-20 days. 158 MASS. EXPERIMENT STATION BULLETIN 179. Immediately following the deutonympharmolt the full-grown female establishes herself upon a cucumber leaf and feeds for about two or three days before oviposition^takes place. | [During this short period it mates *'^eA?r Q and shows a tendency to migrate. Following this period for about eight to ten days it deposits about six eggs per day, thus making a total of fifty to sixty eggs laid by a single female. The average duration of life of the adult female in summer is about two weeks, but this period in- creases as the weather becomes colder. The development of the male is very similar to that of the female, with the exception that the second nymphal stage is lacking. The other stages, however, require a little longer period for development, so that the time from the egg to the adult is onlj^ one day shorter than the development of the female. The different stages of development and the length of each stage of the male red spider may be represented as follows: — Nymph. Quiescent TI. Adult cf . -I 1 1 2 days. l>-4 days. 2 days. 1^ days. 5-7 days. GREENHOUSE RED SPIDER. 159 Development of Female Mite from Egg to Adult. Date. 1. 2. 3. 4. 1916. May 21, a.m., P.M., Hatched" Hatched? Hatched. : May 22, a.m., P.M., Larva. Larva. Larva. Larva. Larva. Larva. Hatched. Larva. May 23, a.m., P.M., Quiescent I. Quiescent L Larva. Quiescent I. Quiescent I. Quiescent I. Larva. Quiescent I. May 24, a.m., P.M., Molted. Protonymph. Quiescent I. Molted. Molted. Protonymph. Quiescent I. Molted. May 25, a.m., P.M., Protonymph. Quiescent H. Protonymph. Protonymph. Protonymph. Quiescent II. Protonymph. Protonymph. May 26, a.m., P.M., Quiescent II. Molted. Quiescent II. Molted. Molted. Deutonymph. Quiescent II. Quiescent II. May 27, a.m., P.M., Deutonymph. Quiescent III. Deutonymph. Deutonymph. Deutonymph. Quiescent III. Molted. Deutonymph. May 28, a.m., P.M., Quiescent III. Molted (adult 9)- Quiescent III. Quiescent III. Quiescent III. Molted (adult ?). Quiescent III. Quiescent III. May 29, a.m., P.M., - Molted (adult $). - Molted (adult ?). FEEDING HABITS AND DISPERSION. A mite which has become full-grown, on finding a suitable spot on the under surface of the leaf, settles down to feed, and the results soon become apparent on the upper surface. At first this injury shows as a few small dead or corky specks, but as feeding continues these few are added to until we find a small area literally made up of them. The mite also im- mediately begins to lay eggs, which soon hatch into young mites. These, however, usually remain feeding in the immediate vicinity of their birth, thus causing more or less concentrated injury at different points on the leaf where older mites have established themselves, forming what might be termed different colonies. As these colonies increase in number the feeding areas also increase, until finally they coalesce and cover prac- tically the whole leaf. This is now absolutely useless to the plant and worthless as a food supply for the large number of mites which inhabit it, and they therefore migrate to other leaves. This migration may be up the plant or may extend to the next plant, provided their leaves are in contact. This new plant may have hitherto escaped injury so that the basal leaves remain uninjured, while an infestation occurs part way up the plant. In natural dispersion the migration is nearlj^ always by full- grown females previous to the egg-laying period. In the majority of cases dispersion within a greenhouse is accomplished wholly by natural agencies. In artificial dispersion the most important factors are the men engaged in pruning, picking or "rolling up" cucumber plants. They pass from an infested to a non-infested plant, but cari'y over infestation on their clothing, hands or tools. This means of dispersion becomes exceedingly 160 MASS. EXPERIMENT STATION BULLETIN 179. important when the plants have become so badly hifested that webs have been spun over the leaves, as the pickers passing from one house to another carry infestation with them. NATURAL ENEMIES. Red spiders out of doors have a very large number of enemies be- longing to widely different groups, nine groups of predacious forms em- bracing thirty-one species having been listed (McGregor, 1917) as attack- ing the red spider. Under greenhouse conditions, however, red spiders are exceptionally free from enemies. It appears that the red spider enemies are unable to develop in the high temperatures which are necessary for most greenhouse crops. In cucumber houses the writer has repeatedly examined infested leaves in the hope that some enemy would be found able to withstand greenhouse conditions and prove useful in the control of this mite, but these examinations have proved fruitless. On violets which are grown in a humid atmosphere and at a low temperature, a few predaceous mites belonging to the order Acarina, family Gamasido', are very beneficial. INTRODUCTION TO EXPERIMENTS. Before taking up the experiments conducted on the artificial control of red spiders a few facts will be summarized in order that the failure of some fumigants and sprays may be better understood. Cucumber plants grown out of doors are very delicate and susceptible to injury of many kinds, while those grown in forcing houses are much more sxj. Therefore the sprays and fumigants which can be used with safety to the foliage are very few, while the red spiders are exceptionally hard pests to combat. These two opposing factors have been found extremely hard to satisfy. Many greenhouse men ask the following question, "Why is fumigation not effective in controlling red spiders?" It has been known for many years that these mites are very resistant to fumigation with our ordinary poisonous gases, such as tobacco and hydrocyanic acid gas. To explain this peculiarity we must contrast the respiratory systems, through which all poisonous gases act, of mites and insects. The latter are efficiently controlled, while only a very few of the former succumb to such treatment. In insects the respiratory system is composed of several large main air tubes which repeatedly divide, forming very small tubes which ramify into all parts of the body. This system of tracheal tubes opens to the exterior by several small segmentally arranged openings called spiracles, and through these the poisonous gas enters the air tubes, which conduct it to every tissue in the body, and i)roduces sudden death. Although the tracheal system of the red spider is better developed than in most mites, it is far simpler than in the majority of insects, containing a much smaller number of tubes. GREENHOUSE RED SPIDER. 161 The number and location of the spiracles in red spiders have not been determined because of their minuteness, but they are probably two in number and are situated in the vicinity of the head region. Therefore, although the red spider can be killed by fumigation with hydrocyanic acid gas, it is impossible to do so without severely damaging plant life, due to the concentration of the poisonous gas required. An infested plant has at all times every developmental stage of the red spider on its leaves, but in artificial control methods we need to con- sider only three general stages. , 1. Egg Stage. — At the present time no spray is known which will affect this stage without severely injuring the plant. 2. Quiescent Stage. — As explained under the life history, the young larvae on hatching feed for a day, and then settle down on the leaf in a premolting or quiescent state during which time no nourishment is taken. These quiescent mites form a new chitinous layer beneath the old external skin covering of the preceding stage. Thus during this period a red spider has two chitinous layers covering the body instead of the normal one, and because of this it has been found very difficult to kill by contact sprays. By reference to the life history it will be seen that each female mite passes through three of these quiescent periods before reaching the adult state. If red spiders in this stage of development are not killed by the spray material recommended for control, it will be almost impossible to eradicate this pest unless sprayings are conducted daily. As soon as the spray applied to an infested plant has evaporated, the mites will be found inactive, and many workers have concluded that all mites above the egg stage have been killed. However, if the leaves were kept under careful observation it would be seen that many of the mites quiescent at the time of application later molt and establish themselves. This point has been overlooked by former workers on the control of red spiders, but is a very important one. 3. Feeding Stages. — ■ A large number of spray materials efficiently control mites in the active feeding stages, but because of their inefficient control of the quiescent stages have been discarded. EXPERIMENTS CONDUCTED IN THE LABORATORY. Fumigation Experiments. Several fumigation experiments were conducted in the hope that some gas might be found effective for red spiders without being injurious to cucumber plants. (a) Sulfur Dioxide (SO2). In many commercial forcing houses sulfur is burned between crops, in order to rid the house of all insects, fungous diseases and mites. To prove whether this was an efficient method, the following experiments were performed. 162 MASS. EXPERIMENT STATION BULLETIN 179. Powdered suKur was burned at the rate of one-quarter of a pound per 1,000 cubic feet of space in a tight fumigating box containing a badly- infested plant. After twelve hours' fumigation the plant was removed. Results. — The cucumber plant was severely injured and died. All mites were killed, those quiescent failed to molt and the eggs did not hatch. This experiment was repeated several times and the results checked with those above. Fumigation with sulfur dioxide is an inexpensive and efficient method of ridding an infested house of mites between crops. Painting Sulfur on Steam Pipes. — This is an old practice of florists in combating the red spider, but has been proved beyond a doubt to be absolutely worthless. (b) Hydrogen Sulfid (H.2S). Potassium sulfid (liver of sulfur) dissolved in water has been widely recommended as an efficient spray for controlling red spiders, and it is claimed that its efficiency depends upon the fact that it combines with the carbon dioxide of the air, forming potassium carbonate and hydrogen sul6d according to the following formula?: — Monosulfid: K2S + H2O + CO2 = K2CO3 + HzS. Polysulfid: KjSs + H2O + CO2 = K2CO3 + H2S + 4S. As an insecticide it is claimed that this sulfid acts by virtue of its caustic properties and the hydrogen sulfid given ofT by its decomposition, this gas being for insects almost as poisonous as hydrocyanic acid gas. To determine whether hydrogen sulfid could be used with safety to plants and still be effective in killing red spiders the following experiment was performed: a plant infested with mites was placed for twelve hours in a fumigating box containing a 1 per cent, atmosphere of hydrogen sulfid. Results. ■ — The plant was severely injured and died, while the mites and eggs were unaffected. (c) Carbon Bisulfid (CS2). Experiments using carbon bisulfid at the rate of 2 pounds per 1,000 cubic feet proved to be inefficient in controlling the mites even after a twelve-hour fumigation. The plants in this case were not injured. Carbon bisulfid at a higher rate would be too expensive to use in commercial houses, and therefore further experiments were discontinued. (d) Benzene or Benzol (CgHg). Early in the experiments on the control of red spiders it was found that benzene vapor had a very active effect upon the mites. However, this proved to be only a temporary stupefication, and mites which had GREENHOUSE RED SPIDER. 163 been removed from the fumigating box containing benzene vapor soon recovered in fresh air. The expense and danger accompanying the use of benzene precludes its use on a commercial scale. Nitrobenzene and para-dichlorobenzene were experimentally used as fumigants, but proved to be as unsatisfactory as benzene, while nitro- benzene severely injured foliage. Spraying Experiments. At present the only kno\vn method of controlling red spiders is by the use of sprays. The majority of these act as adhesive sprays, while only a few are truly contact poisons. (a) Water. Water alone has been found very useful in the control of this pest on certain plants, such as the carnation, violet and rose. The usefulness of a water spray lies in the fact that frequent syringing dislodges many mites from the leaves. The majority of these fall to the moist ground and become permanently pasted into the mud. Frequent use of water also prevents the formation of webs, which are quite necessary as a means of travel and dispersal when a leaf becomes thickly populated. Although > water is very useful in controlling these mites on certain plants, others cannot be grown in a humid atmosphere without being seriously attacked by fungous diseases, and this is especiallj^ true of cucumber plants. The tenderness of the forcing house cucumber also limits the usefulness of a strong stream of water. (b) Adhesive Sprmjs. 1. Flour Paste. — Perhaps the most widely known and thoroughly tried adhesive spray is flour paste, recommended by W. B. Parker (1913) in controlling mites attacking hops in the Sacramento Valley, Cal. He found that flour paste made according to the following formula proved to effectively control 99 to 100 per cent, of the mites: 8 pounds of flour boiled in 8 gallons of water to form a paste, and diluted to make 100 gallons of spra3^^ In order to obtain an accurate estimate of the effectiveness of this spray when used on cucumbers the following experiment was performed: a stock solution of flour paste was made and diluted according to Parker's formula. This spray was applied thoroughly to an infested plant. Results. — The spraj'' has excellent spreading qualities, and as an adhesive is Cjuite efficient in controlling all mites which at the time of spraying are actively feeding. However, this spray does not affect either the hatching of the eggs or the emergence of the mites from the quiescent stages. 1 In a recent government bulletin McGregor and McDouough recommend the use of laundry starch, thus simplifying the process of cooking in forming the stock paste solution. 164 MASS. EXPERIMENT STATION BULLETIN 179. 2. Soaj). — The addition of soap to a spray material increases its spreading qualities and at the same time adds to its adhesive properties. For red spider control soap is inefficient as a contact poison, but if used in faii'ly concentrated solutions it proves to be an excellent adhesive spray. Ivory soap used at the rate of 1-1 pounds in 25 gallons of water was tried as a spray and found to be as effective as flour paste (8-8-100), with the advantage of being much easier to make and not requiring constant agitation. Results. — After this spray has been applied the water evaporates, leav- ing a brittle film of soap over the mites, which is fairlv efficient in sticking these pests to the leaves. However, nearly all mites which are in the quiescent stage molt and establish themselves, and quite a few of the actively feeding mites are able to break the brittle film of soap covering their bodies and thus become liberated to feed on the leaf as before. The eggs are not affected. A common brand of fish oil soap, at the rate of 1 pound in 10 gallons of water, was applied to mites on cucumbers. The efficiency of this over ordinary soap proved to be very little, if anj\ (c) Sulfur and Compounds of Sulfur. Sulfur and many of its compounds have been recommended for the control of red spiders attacking various plants. The following have been tried thoroughly, but have proved, for the most part, inefficient. 1. Dry Sulfur. — In southern California, where the temperature is high, dusting plants early in the morning so that the dew on foliage will cause the particles of suKur to adhere has been found very suc- cessful, especially upon low-growing plants. The use of resublimed or flowers of sulfur on plaats which are not prostrate has proved very unsatisfactory as a control for red spiders. Many of the market gardeners of Boston have thoroughly tried out this method without any material success. Several experiments were conducted, but dusting did not seem to affect the red spiders in the least, even though the temperature was high. 2. Sulfur as a Liquid Spray. — This spray has been recommended for controlling red spiders, but experimentally proves to be of very little value. 3. Sulfur Compounds, (a) Potassium Sulfid {Liver of Sulfur) KoS. — This spray has been recommended by McGregor as being very effective in controlling red spiders attacking cotton. Using 3 pounds of potassium sulfid to 100 gallons of water, McGregor found that 100 per cent, of the mites on cotton were killed by this spray. This is an easily prepared material which may be a])plied ^vith safctj'' to foliage, but at the present time, on account of the increasing demand for potassium salts for use in the manufacture of munitions and fertilizers, this is very difficult to GREENHOUSE RED SPIDER. 165 obtain, while the price is rather high. In using this material on cucumbers it is necessary to add soap to the solution in order to increase its spreading qualities. Results. — This spray proved to be efficient in controlling actively feeding mites, but only a few of those quiescent failed to molt. The eggs were not affected. (b) Calcium Sulfid (CaS2). — This spray proved to be of little value as it killed but few mites. Soap cannot be added to this solution as it forms an insoluble calcium soap which is precipitated. Had this material proved of value it could be obtained more cheaply in lime-sulfur, of which it is a constituent, than in the form of the pure white calcium sulfid. (c) Sodiuvi Sulfid (Na2S). — To determine whether a substitute for potassium sulfid could be obtained by the use of sodium sulfid, a spray was made by the following formula: — Pounds. Commercial NaOH» .......... 2^ Flowers of sulfur, . . . . . . ' . . . . .5 After solution is complete add water to make 100 gallons of spray. Results. — Although this spray proved to be as effective in killing all actively feeding mites as did the potassium sulfid solution, its effect on the quiescent stages was materially less. The eggs were not injured. (d) Soluble Sulfur. — This is a commercial compound made up prin- cipally of sodium sulfid, and as a spray the results check with those given above, with the exception that this spray is very apt to injure the foliage. (e) Barium Sulfur (B. T. S.). — This material, used at the rate of 3 pounds to 50 gallons of water, is not injurious to folfege, but is inefficient in controlling mites. Soap cannot be added, as it forms an insoluble barium soap. (/) Lime-sulfur and Nico-fume Liquid. — This has been recommended as a spray for spider mites as well as the clover mite (Bryobia), and has the following composition: — Lime-sulfur, commercial (quarts), ........ 2 Nico-fume (pint), ........... 2 Water (gallons) 25 Residts. — The application of this material caused considerable injury to the cucumber foliage, while it was only fairly efficient in controlling the mites. Several greenhouse men have sprayed with dilute lime-sulfur solution, but have found it both inefficient in controlling these pests and injurious to the foliage. Nicotine sprays are also inefficient when used alone. (d) Oil Sprays. 1. Spraijs containing Petroleum Oils, (a) Arlington Oil. — This is a chemically miscible oil containing approximately 90 per cent, petroleum oil. Used at the rate of 1 part oil in 50 parts of water it was found effective 166 MASS. EXPERIMENT STATION BULLETIN 179. in controlling aphids and thrips, but killed only 50 per cent, of the actively feeding mites. At the above strength this spray severely injured cucumber foliage, and even when diluted to 1 part oil in 100 parts of water, injury still occurred. (h) Arlington Oil and Black-LeaJ-Jfi . — Formula: oil, 1 part to 125 parts of water; Black-Leaf-40, 1 part to 2,000 parts of water. This com- bination spray is much more active than the ingredients used separately, but is injurious to the cucumber foliage. (c) Kerosene Emulsion. — This is recommended as being efficient in controlling red spiders, but it severely injures tender foliage. 2. Sprays containing Vegetable Oils, (a) Lemon Oil. — This is manu- factured by the Lemon Oil Company, Baltimore, Md., and is at present sold at 11.75 per gallon in 5-gallon lots. It is a completely saponified oil soap, and is guaranteed to contain the following ingredients: — Per Cent. Soap, . • 6 Vegetable oil, . . . " 3^ Potassium carbonate, .......... J Terabenthine (Tiirpentine?), ......... 5 Water (not over), . . . . . . . . . .85 Of the many commercial insecticides used experimentally in the control of red spiders this proved the most satisfactory. Results. — Used at the strength of 1 part lemon oil in 20 parts of water, or 1 pint in 2^ gallons of water, it killed all actively feeding mites, as well as those in the quiescent stage, without injuring the foliage. The eggs are not materially affected by this spray. If young potted cucumber plants are dipped in the above mixture some injury will result to the terminal growing point, but if the plants are sprayed this injury does not occur. During the spring and summer months of 1916 this spray was thoroughly tried out on a commercial scale, and proved to be very satisfactory, but its expensiveness precludes its free use as a general spray for red spiders. (b) Ex-periments on the Duplication of Lemon Oil. — With the co- operation of Dr. E. B. Holland of the Massachusetts Agricultural Experi- ment Station a number of spray materials were made in order to deter- mine the killing agent in lemon oil, and for the purpose of duplicating the efficiency of this oil by a substitute which would be less expensive. The following table will briefly show the composition of these mixtures and their relative effectiveness in controlling red spiders: — GREENHOUSE RED SPIDER. 167 o o o o o S o o o o o "P ™ CO « 2 to s " o o o o o i o o o o o m o § '" 1 CO o 1 CO o 1 f" s I § ° § § «o 1 n g ui 1 ' ' ' ^ ' ' < ° o ' g lO s o o o o o o o m o o 2 „ CD " CO •o s -^ 1 o o o o o o o m ' >o 1 ' o o ' >o ' o ' o o o ■o o o . >o t- •a •o ? « N '^ ^ M g o o o o o o o o VH o "* •o JO '" g- o s o s § ° J'" o U5 n ^ _ o 1 § a 1 g 1 a i 1 c 1 1 1 1 • s i 1 1 1 o i i 1 1 1 1 a i 3 -*J 0. 0. UD PQ fe CO 03 S 1 g s 1 i2 J c "g 1 £ >1 1 H 1 1 i ^ 1 s 2 i B o J i 0. 3 a 0 1 0 e 1 1 m o o & p « 168 MASS. EXPERIMENT STATION BULLETIN 179, (c) Linseed Oil Emulsion. — Thus, out of nine mixtures, only those containing linseed oil proved at all promising. Mixtures 7 and 8 were rather poorly saponified (chemically), while 9a and 96 were completely saponified; but 7 and 8 proved efficient, while 9a and 9& were not. This could only be explained by the fact that the free linseed oil was really the toxic 'agent, and when it was only partly saponified there remained some free linseed oil which established the efficiency of the spray. Upon this supposition were based other preparations containing linseed oil mechanically emulsified in a solution of soap in water. These emulsions proved to be efficient when a 1 per cent, oil spray was used. Two types of linseed oil emulsion may be made, depending upon the length of time these emulsions are to be retained before use. Experimentally it was found that the most stable stock emulsion could be made as follows : one-eighth of a pound of Ivory soap (one-half a 5-cent cake) dissolved in a pint of very hot water. After the soap is completely in solution add 1 pint of cold water followed by the addition of 1 pint of raw Unseed oil. The oil should be completely emulsified by the use of a bucket pump. This solution is stable, provided the water contained in it is not allowed to evaporate. In using this stock emulsion, especially after it has been kept for some time, it is best to mix one part of stock with an equal volume of water before diluting to desired strength. One part of stock emulsion in 20 parts of water proved to be efficient in killing mites, both in the quiescent and feeding stages. If spraying is to be done soon after mixing the emulsion it is best to increase the amount of water and soap, and make the emulsion as follows: shave 6 ounces of Ivory soap (1| 5-cent bars) into 1 gallon of hot water. Add 2 quarts of cold water to cool the solution, then add 1 quart of raw linseed oil and emulsify with a bucket pump. This emulsion, used at the rate of 1 part in 9 parts of water, is very efficient, kiUing quiescent and feeding mites without injury to leaf tissue. Soy bean oil substituted for linseed oil proves to be efficient, and in some locaHties could be used to advantage. Action of Linseed Oil Emulsion upon Mites. — The majority of oils used as insecticides are regarded as contact poisons. These poisonous oils are supposed to enter the body of the insect, either directly through the thin membraneous chitin of the body segments or by entering the spiracles, where they immediately pass through the tracheal lining and produce an active effect upon the internal structures essential to the life of the insect, ij In a previous part of this paper it has been shown that the spiracles are very few, — probably two in number, ■ — and that the body of a red spider is covered by a rather thick and continuous coating of- chitin. For these reasons sprays which prove effective in killing aphids are of little value when applied to mite-infested plants. Many of the spray materials which have given partial success in con- trolling mites have a marked adhesive action, and from this property GREENHOUSE RED SPIDER. 169 linseed oil emulsion derives its efficiency. The spray as made (see "Re- pressive Measures") contains the amount of soap necessary to hold the oil in suspension and give the spray material excellent spreading quahties. Raw linseed oil contains two types of oils, — (1) drying oil and (2) resinous oil. Upon this fact is based its usefulness in paints, as well as its efficiency as a red spider spray. A leaf thoroughly covered by the spray soon becomes dry, the water evaporating, while the oil and soap become more and more concentrated as this evaporation continues. Finally there is formed a very thin layer of oil and soap which gradually settles dovm on to the leaf surface, cover- ing all mites ^vhich were feeding on the leaf at the time of application. This film gradually envelops the mite, and the volatile parts of the linseed oil are given off, leaving behind a resinous or waxy oil which securely cements the legs of the mite to itself and to the leaf. Thus the mite is helpless, and the waxy residue of the linseed oil remains, sticking the mite until it dies of starvation. Without doubt some of its effectiveness may be due to its being a contact poison, but its most important quality is its adhesiveness. Summary of Materials found to be Efficient Experimentally. No fumigant was efficient in killing red spiders without severely damag- ing cucumber plants. Sulfur burned to form sulfur dioxid proved to be very effective in killing all stages of mites. Although this gas is deadly to plant life, its application as a fumigant to rid -empty houses of all mites is extremely useful. Many spray mixtures proved to be efficient in controlling actively feeding mites, but did not affect those in the quiescent stages of develop- ment. For the control of all stages above the egg stage lemon oil, a com- mercial product, and linseed oil emulsion proved to be the most satis- factory. Soapy solutions should also receive some attention as among the most readily prepared spray materials, although their efficiency is only temporary and treatment must be repeated often in order to control these mites. EXPERIMENTS CONDUCTED IN COMMERCIAL GREEN- HOUSES. The materials found to be most efficient in the laboratory experiments were applied to cucumber plants in commercial establishments in order to determine the practicability of sprajdng for the control of these mites before any recommendations were made. It was found impossible for the writer to be stationed at these green- houses during the whole spraying period. Therefore the efficiency of these sprays under commercial conditions has been determined largely by the statements of the growers, checked by more or less frequent per- sonal observations. 170 MASS. EXPERIMENT STATION BULLETIN 179. Lemon Oil. The first of these commercial experiments commenced during May, 1916, and continued until the middle of June. Lemon oil, 1 part in 20 parts of water, was thoroughly tested in several greenhouse^, and in all cases the spray proved very efficient, provided it was thoroughly applied to the infested plants. At the time the first commercial applications were made the plants were nearly full-grown, and the mites were at that time rapidly spreading through the houses. All that could be expected of this spray was to hold the red spiders in check, so that they would not materially damage the whole house before a good crop of cucumbers had been picked. Owing to the scarcity of labor it was found impossible to apply sprays at weekly intervals, and therefore the results were not as satisfactory as they would have been under other conditions. However, these sprayings held the red spiders in check and prolonged the life of the cucumber plants, which would have died early in the season had no treatment been applied. In several instances young potted cucumber plants were dipped in a 1 to 20 dilution of lemon oil as they were being set in the greenhouse. This proved to be injurious to the succulent leader, although the leaves gave no indication of injury. Linseed Oil Emulsion. During the summer of 1916 experimental work on the determination of the killing property of lemon oil led to the discovery of linseed oil emulsion and its efficiency in controlling mites. This emulsion has re- ceived a very thorough trial in commercial greenhouses this season (1917), and proves to be satisfactory in many respects. The ingredients are always at hand, the initial cost is low, being one-fourth that of lemon oil, and the method of preparation is simple. Experiment No. 1. Early in the spring of 1917 this spray mixture was thoroughly tested on a commercial scale in greenhouses located in Watertown, Mass. This range is naturally divided into two groups. Group L contained the oldest cucumber plants and Group II. the youngest. It was decided that appli- cations should be made to the youngest plants, although they were really too old for effective spraying. The cucumber plants became badly in- fested in the seed-plant house before being set out. Therefore this in- festation became serious soon after the plants were transplanted to the greenhouses. Severe pruning was resorted to, but this did not hold the mites in check. For efficient control, these plants should have been thoroughly sprayed at the time they were transplanted. Group II. consisted of three greenhouses. In greenhouse No. 1 the plants were very heavily infested, and were 5 feet tall at the time of the GREENHOUSE RED SPIDER. 171 fii-st application. In No. 2 the plants were 2| feet tall and generally infested, although not shomng any noticeable injury to the plants from the red spider attack. In No. 3 the plants were 4 feet high and rather severely infested. In each of these houses three applications were made at weekly intervals. The finnl results of these experiments are as follows: the greenhouses of Group I. were not sprayed, and though the plants were very little older than those in Group II. they died from the red spider injury after being in the range approximately three months. In Group II. the plants were sprayed and produced fruit for over a month longer than the un- sprayed plants of Group I. Houses No. 1 and No. 3 contained such large cucumber plants that a thorough application of a spray was found im- possible, but the ravages of these mites were checked during the spraying period. Although a complete control was impossible, the productive life of the crop was lengthened approximately one month. In house No. 2, containing the youngest cucumber plants in Group II., the control was much more efficient, primarily because the plants were smaller and a thorough spraying could be given them. However, even these plants were too large to insure a thorough application after the first spraying. Experiment No. 2. Further tests of the efficiency of linseed oil emulsion were made in commercial greenhouses at Arlington, Mass. In this establishment all plants were infested in the seed-plant house while still in pots. Soon after they were set in the greenhouses the first spray was applied, and one week later the second application was made. These two applications were made at the proper time, and controlled the mites so effectually that during midsummer some of these houses were yielding good crops, while only a few scattered plants were beginning to show marked red spider injury. At approximately the same time in former years the plants in these houses have been severely infested and dying from the ravages of the red spider. This range of greenhouses consists of twelve large houses, and therefore it is not surprising that the whole establishment could not be thoroughly covered each week. An excellent demonstration of the efficiency of linseed oil emulsion was made in the seed-plant house. As stated above, when the cucumber plants were still in pots in this house they were noticeably infested by red spiders. The grower, knowing that this house contained many mites, determined that sprayings should be given wath special care, in order to eradicate these pests. Soon after the potted plants were set out in the seed-plant house the first application was given, care being taken to cover thoroughly all the leaf surface. One week after this the second thorough spraying was applied. These applications were made so thor- oughly that very few if any mites which originally infested the cucumber plants survived, and the plants attained full growth without showing any red spider injury. 172 MASS. EXPERIMENT STATION BULLETIN 179. Conclusions drawn from Commercial Spraying Experiments. Sprayings conducted on bright, sunny days with a rather high tem- perature in the greenhouse resulted in slight injury to the edges of the leaves, but if applications were made on cool, cloudy days this injury did not occur. For a thoroughly efficient control at least three applications should be given the cucumber plants at weekly intervals, as soon after they have been set out in the greenhouses as possible. PREVENTION. The writer has been unable to conduct a thorough test in eliminating red spiders from the whole range by cultural methods, because it was found impossible to procure an establishment which would serve for this purpose. In commercial greenhouses many factors enter into the red spider problem which cannot be solved unless a suitable range is found which will eliminate these confusing factors in order that some definite knowledge may be gained by using preventive measures. However, under greenhouse conditions, it is the writer's firm conviction that the red spiders can be totally exterminated from commercial ranges by clean culture, both within and outside the greenhouse. It is hoped that some experimental work may be conducted on this important control measure in the near future. CONTROL MEASURES. The general biology and development of experimental and commercial control measures have already been discussed, but only in a general way. Under this heading the methods used for the prevention and repression of red spiders will be taken up more in detail. Having established the efficiency of the repressive measures, only the preparation and application of spray materials will be considered. Preventive Measures. The solution of the red spider control problem in cucumber greenhouses should be accompUshed through preventive efforts rather than by re- pression, if it is to be done most economically. The commercial grower should do everything possible to eliminate these pests, both within and outside his greenhouses. In the majority of cases cucumber plants are infested either in the plant house or soon after they have been set out in the greenhouse. The origin of this infestation may be weeds which have harbored mites through- out the winter inside the greenhouse, or weeds and grasses immediately surrounding the house at the base of which the mites winter over and migrate into the greenhouse early in the spring. The first is very im- GREENHOUSE RED SPIDER. 173 portant when plants are started very early in the season, while the second is of importance only after the warm days of spring have started these outside weeds. Fumigation of Greenhouses and Equipment loith Sulfur Fumes. Immediately before setting the cucumber plants in a house, and before fumigation is begun, all boards which are to be used either between the cucumber rows or to make "A" trellises should be taken inside the green- house. Do not lay the boards on the ground, but stand them against the steam pipes or in some similar manner to allow the poisonous gas free access to all parts. Other equipment which has been in any way con- nected with a previous infestation and is to be used during the cucumber season should also be placed in the house for fumigation. Do not intro- duce living plants until after a thorough fumigation and a subsequent airing of the houses, as sulfur fumes are deadly to plant life. In fumigating, each house should be tightl}^ closed and sulfur used at the rate of one-third of a pound to every 1,000 cubic feet of space. (In- crease to one-half pound in case of houses that are not fairly tight.) Directions for Fumigation. — Weigh the required amount of sulfur and divide it into four equal parts upon pieces of paper. This is about the right number for a 150-foot house. Metal pans with plenty of breadth are perhaps the best containers for the fumigating operation. First cover the bottom of each pan with chips that have been soaked in kerosene, and distribute these containers at various points tlirough the house, placing beside each the sulfur to be used. When all is in readiness set fire to the chips, and when these are burning well drop in the sulfur. Be certain that the sulfur has ignited and then withdraw from the house. Allow the sulfur fumes to act for at least twelve hours before opening the house. This fumigation may be done during the day or at night, accord- ing to the convenience of the grower, and if the method is followed out carefully the red spiders will be completely exterminated within the house. Special attention should be paid to the house in which potted cucumbers are to be grown, and fumigation should be very thorough, for in many cases the seat of infestation occurs here. At the conclusion of the cu- cumber crop in the late summer the whole house should be fumigated with sulfur before the plants have died, thus preventing the borders from becoming infested from thrown-out cucumber plants, and reducing the number of red spiders which would otherwise winter over and attack the next cucumber crop. Destroi/ing Outside Sources of Infestation. The next problem which confronts the grower is to eliminate the possi- bility of infesting the houses from outside sources. Investigation has shown that many weeds and grasses, often found around greenhouses, serve as breeding places for these pests, and undoubtedly are the source 174 MASS. EXPERIMENT STATION BULLETIN 179. of inside infestation. In the fall red spiders are found in large numbers on these grassy borders, and being capable of wintering over out of doors, it follows that a large perceDtage of those found in the fall will also be present in the spring, and are quite certain to migrate to the more at- tractive cucumber plants within the greenhouse. Methods of Exterminating Grassy Borders. 1. The border for at least 10 feet away from the house should be thor- oughly cultivated, preventing the gro^vth of weeds throughout the season. 2. Where cultivation is not practicable, burning the border may be resorted to. ' 3. If neither of the above methods can be employed, kill all vegetation around the greenhouse by sprajdng with sodium arsenite used at the rate of 1 pound to 20 gallons of water. It must be remembered, however, that sodium arsenite is a poison, and care should be taken to prevent animals from grazing on treated borders. Repeat as often as necessary. Elimination of Artificial Dispersion. As described under "Feeding Habits and Dispersion," the most im- portant factors in artificial dispersion are the men working in the green- houses. The grower should systematize, as far as it is practicable, all work which must be done in his houses according to the infestation; for example, in two greenhouses, one showing red spider injury, the other apparently free, pruning or "rolling up" of plants should first be done in the house apparently free from infestation, and later in the infested house. Also in picking cucumbers, the young houses — which usually are not as badly infested as older ones — should be picked first, and older, badly infested houses last. Special care should be exercised not to allow the men who have finished picking in a badly infested house to start pruning or " rolling up " a very young house. Baskets used in picking cucumbers should never be used in a younger house as a receptacle for pruned parts of young plants. The writer realizes that these recommendations are not all applicable under commercial conditions, but every precaution which is practicable should be taken if artificial dispersion and infestation are to be reduced. Repressive Measures. During the early stages of infestation it is frequently found advisable to destroy plants which are found to be badly infested. These badly infested plants should be pulled out before the leaves begin to die, so as to prevent dispersion due to lack of food. If a few leaves, usually near the ground, are badly infested the pruning of these will lessen the numbers of mites materially. In all cases, whether a plant has been pulled or pruned, the red spiders on these leaves should be destroyed by burning. Do not throw them outside of the house, but GREENHOUSE RED SPIDER. 175 destroy them immediately, thus eliminating the chance of infesting plants surrounding the greenhouse. Pruning is especially useful when judi- ciously applied to the young plants in a greenhouse. Such prmnng should be supplemented by spraying for a thoroughly efficient control. Spraying. If there is any possibility of infestation, spraying should commence soon after the cucumber plants have been set out in the greenhouse. If spraying is done at this time less material will be used, and a very thorough application can be given in a minimum amount of time. In experiments conducted in commercial greenhouses it was found that red spider sprays applied to young cucumber plants gave very satisfactory results, while on older plants these sprays did not prove as efficient. This can be ex- plained by the fact that a good-sized cucumber plant has a large amount of leaf surface which must be thoroughly covered by the contact spray if efficiency is to be expected. This is economically impossible after the plants have become nearly full-grown, because of the length of time and amount of material necessary to accomplish it. Early spraying will con- trol red spiders at a minimum expense of time, labor and materials. Linseed oil emulsion is especially adapted for use in conamercial green- houses on a rather large scale. If only a few plants need to be treated, lemon oil, manufactured by the Lemon Oil Company, Baltimore, Md., may be purchased at nearly all stores carrying insecticides. This, diluted at the rate of 1 part in 20 parts of water, gives a very efficient spray, but for commercial spraying this material is too expensive. Soapy solutions sprayed upon delicate plants on several successive days prove to be useful. In making this solution a high-grade soap (Ivory soap) should be dissolved at the rate of 4 ounces in 3 or 4 gallons of water. Preparation of Linseed Oil Emulsion. {a) The necessary articles for preparation are as follows: — 1. Bucket pump. 2. Container or mixing tank. This should hold at least 8 or 9 gallons. For this purpose a small washtub is perhaps the most available. Pails may be used, provided the materials are mixed proportionally. 3. Ivory soap. 4. Raw linseed oil. 5. Hot water. (6) The following proportions of materials for 100 gallons of spray are used: — 1. Five gallons of hot water. 2. One and one-half pounds of Ivory soap. (Six 5-cent cakes or three 10-cent cakes.) 3. One gallon of raw linseed oil. 176 MASS. EXPERIMENT STATION BULLETIN 179. (c) Steps in the preparation of stock solution follow: — 1. Put the required amount of hot water in the container. 2. Shave the Ivory soap into this and stir until completely dissolved. 3. If at this time the temperature of the soap solution is too hot for the hand to bear, dilute with 1 gallon of cold water and let it stand until about body temperature or lukewarm. The cooling of this solution is necessary in order to prepare a permanent emulsion; otherwise the oil will come to the surface on standing (see No. 6). It also prevents the chemical and physical killing properties of the linseed oil from being changed by heat. 4. Add slowly, whUe stirring vigorously, 1 gallon of linseed oil. 5. Completely emulsify by using the bucket pump. Pump the emulsion from the container through the pump and back into the container again, keeping the nozzle below the surface of liquid. Five minutes' vigorous pumping should completely emulsify this solution. 6. Set aside for a few minutes while preparing spray tank in order to see that oil does not come to the surface. [d) The following are directions for the preparation of spray tank and spray: — 1. Fill the 100-gallon spray tank about one-half full of water. If the water used is too cold, upon the addition of the stock solution the soap will solidify into small lumps, thus spoiling the emulsion. This may occur early in the spring, when the water is very cold, but later in the season ordinary tap water may be used without danger of the soap solidifying on the addition of the stock solution. 2. Add stock solution made above. (See (c) 1, 2, 3, 4, 5, 6.) 3. Agitate. (If lumping occurs, the addition of a few pails of hot water will remedy this.) 4. Fill the 100-gallon spray tank. Application of the Spray. Outfits and Methods of Spraying. — In commercial greenhouse spraying either a barrel pump or power sprayer should be employed, the latter being the more economical, provided it is available and the size of the establishment warrants its use. For spraying a few plants, or in a very small greenhouse, perhaps the most satisfactory outfit consists of a com- pressed air sprayer. The length of hose necessary in spraying cucumber houses depends upon the size of the house and the method of growing cucumbers. If the vertical trellis system is used, in most cases it is best to have the hose of sufficient length to reach from the sprayer down the middle aisle and across the opposite end of the house, thus eliminating the necessity of changing the sprayer during the spraying operations. By passing in a zigzag manner across the house and gradually working backward the house may be thoroughly covered in the least amount of time. If cucum- bers are grown on the "A" trellis sj'^stem the man spraying should travel GREENHOUSE RED SPIDER. 177 up on one side of the row and back on the other. In either case a boy should be employed to guide the hose, so that it will not injure the plants as it is pulled from one row to the other. These are the most common methods of spraying, but there are many modifications which the grower can make according to the conditions surrounding his houses and the manner of growing his plants. An extension rod made from small piping with an elbowed tip or angle nozzle is absolutely necessary for thoroughness in spraying. If cucumber plants are grown on the vertical trellis s,ystem the extension rod should be about 2§ feet in length, while if grown on the "A" treUis system the rod should be 4 feet in length, as this will allow the man spraying to reach the basal leaves of the plants readily. It is perhaps more satisfactory to use a 45° angle nozzle, several of which may be purchased {e.g., Friend and Simplex angle nozzles^ thus eliminating the necessity of a separate elbow. Methods of Application. — From the fact that the red spider as a rule passes its entire existence upon the under surface of a single leaf, early in the season, when the plant is only slightly infested, it is plainly necessary in spraying to cover the entire under side of every leaf. Special attention should be paid leaves showing typical red spider injury, especially the lower leaves of the plant, near the ground, as these are usually most severely infested. To facihtate this under-surface spray an extension rod with an elbow tip or angle nozzle is essential. The pressure necessary in power spraying varies from 50 to 125 pounds, depending upon the type of nozzle. Do not allow the spray to bombard the under surface of the leaf if a coarse nozzle is used. As this Unseed oil emulsion is a contact spray, it is necessary that the whole under surface of a leaf should be covered by a film of this material. If the spray is deposited on the leaf in fine droplets which do not run together, this can be remedied by the adjustment of the pressure until they unite to form a film. If a coarse nozzle is used, as the Simplex, a low pressure will be required for film formation, while with a fine nozzle, as the Friend, a higher pressure will be necessary. A preference should be given the fine nozzle and high pressure, as this is less apt to injure the leaves, while it proves very satisfactory in forming the film. The success or failure of the spraying depends upon this film formation and thorough application of the material. When Applications should be made. — In general greenhouse practice spraying on bright days is and should be the rule, as with sunshine there is less danger that conditions favorable for disease will result. In the application of the linseed oil emulsion, however, spraying conducted on sunny days with a rather high temperature in the greenhouse may result in a slight injury to the edges of the leaf, while if spraying is done on cool, cloudy days no injury is caused by the applications. Therefore, as far as possible, spraying for the red spider should be done on cloudy days when the temperature in the house is not over 80°. The injury on bright 178 MASS. EXPERIMENT STATION BULLETIN 179. days has never been serious, but should be eUminated as far as possible by proper management of greenhouse temperature and the selection of suitable days for spraying. In order to effectively control red spider infestations, at least three sprayings given at weekly intervals are necessary. The first spraying should usually be applied one week after the plants have been set in the greenhouse. If the young plants show mite injury before this time the application should be made as soon as possible. Usually young cucumber plants do not appear to be affected early in the season. However, on closer examination it will be found that the majority of these plants harbor a few mites which, if allowed to develop unhin- dered, will later become so numerous, and the plant so large by the time injury is noticeable, that an efficient control will be found extremely difficult and expensive. Since this spray does not destroy red spider eggs it is clear that a second appUcation is necessary to kill the individuals which were eggs at the time of the first spraying. This should be applied seven to eight days after the first. If the second spray is not appUed at the proper time it will be almost impossible to control these pests, for many mites will have become adult and laid eggs unless the application is made as rec- ommended. Some mites are sure to escape the first and second sprayings, and there- fore a third application must be given in order to kill these mites, which if not controlled will rapidly multiply and severely injure the plants. As previously mentioned in the discussion of the "Economic Im- portance of the Pest," the loss to cucumber growers due to red spider infestation consists in shortening the life of the plant during its pro- ductive period. It is absolutely essential that these three sprayings be made as directed, otherwise the producing period of the plants will be reduced at least one month. Under normal conditions the few mites found early in the season re- produce rapidly until finally the plant becomes seriously affected by the injuries caused by their progeny, and usually dies before producing a full crop. If the mites are held in check by weekly applications early in the season the length of the period during which these regular applica- tions are made will later be added to the adult life of the plant. The longer the spraying period the longer the productive life of the cucumber plant. It is therefore of great financial importance to the grower to see that these sprayings are thoroughly applied at weekly intervals during the early life of the crop. Cost of Spraying. — The comparative cost of 100 gallons of spray containing lemon oil and linseed oil is as follows: lemon oil, $8.75; Hnseed oil emulsion, $1.50. If sprayings are made with a power sprayer it will take a man, with the help of a boy, approximately three hours to spray thoroughly a green- GREENHOUSE RED SPIDER. 179 house containing 1,600 cucumber plants about 4 feet high. The material used wdll amount to 100 gallons. Thus the cost of one spraying when the plants are nearly half grown is approximately $3. Spray materials, . . . . . . . . . $1 50 Man, three hours, . . . . . 1 00 Boy, tliree hours, .......... 50 This is a fair estimate of the cost of the third spraying. The first and second sprayings taken together should cost approximately $3. Thus, for three applications of hnseed oil emulsion to 1,600 plants, the invest- ment for labor and materials will be approximately $6. This should be considered insurance on the crop. At the above rate the cost for three applications is less than one-half cent per plant. The original investment for spray materials and labor will be repaid many times over by prolonging the fruit-bearing period of the plants. CONTROL OF RED SPIDERS ATTACKING OTHER CROPS. Perhaps a few words relative to the control of these mites attacking some of the other crops will prove useful, especially to florists. Although the writer has confined most of his attention to the control of this pest on cucumbers, it is reasonable to suppose the same control measures will give as satisfactory results in eliminating this pest on other plants. While this is true, a few factors must be thoroughly understood in order to procure these results. On small or rather smooth-leaved plants, such as the violet, rose, carnation, sweet pea and bean, the linseed oil emulsion spray as used on cucumbers does not prove as satisfactory. The reason for this is that the greater part of the spray applied to these plants runs off the leaf, and not enough linseed oil is deposited on the mites to render them helpless. To remedy this difficulty the stock linseed oil emulsion should not be diluted as much as recommended for cucumber spraying. In some cases where very delicate plants are to be sprayed the same dilution may be made, but the solution of soap should be stronger. In spraying cucumbers a 1 per cent, linseed oil mixture is used. On plants such as the violet it is best that the original linseed oil stock solution be diluted only one-half as much, making a 2 per cent, linseed oil mixture and a more concentrated soap solution. In the majority of cases proper experimentation by the grower will furnish satisfactory evidence for the required dilution for efficiency on his special crop. During July and August, 1917, the writer had the opportunity of thor- oughly testing the efficiency of this. 2 per cent, linseed oil emulsion for the control of red spiders attacking violets in the field at Mr. William Sims' greenhouses, Cliftondale, Mass. This field of violets, containing about 180 MASS. EXPERIMENT STATION BULLETIN 179. 100,000 plants, was sprayed, using a power sprayer, three times between July 15 and September 1. The object of this spraying was not to rid the plants of red spiders, although this undoubtedly could have been accom- plished, but to keep their numbers so reduced during the dry summer months that they could not seriously injure the new and tender foliage or kill the plants as they had done in previous years. The results were entirel}^ satisfactory, and the violet plants were kept practically free from these pests. Those plants rather seriously damaged before spraying began regained their dark green foliage, and during the middle of August only a few leaves could be found in the field showing typical red spider injury. Thus the damage caused by red spiders was reduced to a minimum by spraying, while in previous years and under similar conditions they had practically stripped the plants of their foliage. The difficulty of thoroughly applying a spray to the lower surface of the leaves of a low-growing plant is well recognized, for our modern nozzles are not adapted to this type of spraying. This difficulty, however, may be overcome in violet spraying by the use of a simple spray nozzle consisting of a "Skinner System" plug. This plug is often used in green- houses, where it is inserted at intervals in the side of a water pipe. Water passes from the pipe through a small hole in the center of the plug, and then strikes a curved lip which transforms the solid stream to a fine, fan- like spray. This plug is placed in the end of an extension rod 5 feet in length, made from one-eighth-inch piping. The rod is then bent until the fan-like spray travels parallel to the surface of the ground. This type of nozzle proved very satisfactory, and could be held close to the plant without injuring the leaves. SUMMARY. The common greenhouse red spider (Tetranychus bimaculatus Harvey) is very generally distributed throughout the United States, extending from Maine to Florida, and westward to Texas and California, only a few States in the western arid region being exempt from the ravages of this pest. The red spider is very cosmopolitan in its feeding habits. In market- garden greenhouses the most important vegetable attacked is the cu- cumber. In floriculture greenhouses the rose, violet, sweet pea, carnation and chrysanthemum are seriously injured. The most important outside plants, as far as the greenhouse man is concerned, are those found around most greenhouses, consisting of clover, grasses and weeds, as these are undoubtedly important factors in causing inside infestation. It is estimated that the annual loss to cucumber men in the Boston market-garden district, due to red spider injury, amounts approximately to $150,000, or 10 per cent, of the whole crop. Experimentation on the control of this mite attacking cucumbers gave no f umigant which could be used with safety to the foliage. Sulfur burned to form sulfur dioxide proved to be very effective in killing all stages of GREENHOUSE RED SPIDER. 181 mites. Although this gas is deadly to plant life, its application as a fumigant to rid empty greenhouses of red spiders is extremely useful. Many spraj^ mixtures proved to be efficient in controlling actively feeding mites, but did not affect those in quiescent stages of development. For the control of all stages above the egg stage linseed oil emulsion proved to be the most satisfactory. The control of the red spider may be accomplished by combining pre- ventive and repressive measures. Clean culture, or the eradication of weeds and plants which harbor mites during the winter period, either within or outside the greenhouse, is by far the most vital means of prevention in cucumber greenhouses. Dispersion within the greenhouse may be hindered by destroying plants or parts of plants which harbor the initial infestation. Applications of linseed oil emulsion at weekly intervals during the early life of the plant prove very effective if made with extreme care. At least three applications must be made for an efficient control. By checking red spider infestation early in the season the producing period of the plants is lengthened approximately one month. BIBLIOGRAPHY. The following bibliography includes only the more important economic works on the red spider: — Britton, "W. E., 1901. "Common Soap as an Insecticide." First Rept. State Ent., Conn., pp. 227, 278. (Red Spider Remedy, pp. 271-273.) Chittenden, F. H., 1901. "Some Insects Injurious to Violet, Rose, and Other Ornamental Plants." Bull. 27, n. s., Bur. Ent., U. S. Dept. Agri. (The Two- spotted Red Spider and Control, pp. 35-42, Figs. 9-14.) Chittenden, F. H., 1909. "The Common Red Spider." Circ. 104, Bur. Ent., U. S. Dept. Agri. Ewing, H. E., 1914. "The Common Red Spider or Spider Mite." BuU. 121, Oregon Agri. Exp. Sta., 95 pp., 30 figs. Fleet, W. J., 1900. "Some Comparative Trials of Insecticide Pumps in Relation to the Treatment of Tea Blights and Experiments in the Treatment of Red Spider." Indian Mus. Notes, Vol. IV., No. 3, pp. 113-117. Gillette, C. P., 1889. "The Red Spider." Bull. 4, Iowa Agri. Exp. Sta., pp. 183, 184. (Greenhouse Control.) Glover, T., 1855. "Insects Frequenting the Cotton Plant." Rept. U. S. Comm. Patents, Agri., pp. 64-119, Pis. VI-X. (Reference to red spiders, p. 79.) Harvey, F. L., 1892. "The Two-Spotted Mite." Annual Rept. Me. Agri. Exp. Sta., pp. 133-146, PI. III. (Original description of Tetranychus bimacu^atus Harvey.) Maynard, S. T., 1889. "Experiments in Heating Greenhouses." Bull. 4, Hatch Exp. Sta., Mass. Agri. College. (Reference on control of red spiders, pp. 14, 15.) McGregor, E. A., 1912. "The Red Spider on Cotton." Circ. 150, Bur. Ent., U. S. Dept. Agri., pp. 1-13, 5 figs. McGregor, E. A., 1913. "The Red Spider on Cotton." Circ. 172, Bur. Ent., U. S. Dept. Agri., pp. 1-22, 12 figs. McGregor, E. A., 1914. "Red Spider Control." In Journ. Econ. Ent., Vol. VII., No. 4, pp. 324-326. 182 MASS. EXPERIMENT STATION BULLETIN 179. McGregor, E. A., 1916. "The Red Spider on Cotton and How to Control It." Farmers' Bull. 735, Bur. Ent., U. S. Dept. Agri., 12 pp., 10 figs. McGregor, E. A., and McDonough, F. L., 1917. "The Red Spider on Cotton." Bull. 416, Bur. Ent., U. S. Dept. Agri., prof, paper, 72 pp., numerous figs. Morgan, H. A., 1897. "Observations on the Cotton Mite." Bull. 48, La. Agri. Exp. Sta., pp. 130-135. Parker, W. B., 1913. "The Red Spider on Hops in the Sacramento Valley of California." Bull. 117, Bur. Ent., U. S. Dept. Agri., pp. 1-41, 6 pis., 9 figs. Parker, W. B., 1913. " Flour Paste as a Control for Red Spiders and as a Spreader for Contact Insecticides." Giro. 166, Bur. Ent., U. S. Dept. Agri., 5 pp., 2 figs. Perkins, G. H., 1897. "The Red Spider." Rept. of Ent., 10th Ann. Rept. Vt. Agri. Exp. Sta., pp. 75-86, Figs. 1-4. Quayle, H. J., 1912. "Red Spiders and Mites of Citrus Trees." Bull. 234, Cal. Agri. Exp. Sta., pp. 483-530, Figs. 1-35. Quayle, H. J., 1913. "Some Natural Enemies of Spiders and Mites." Journ. Econ. Ent., Vol. VI., pp. 85-88. Russell, H. M., 1908. "Experiments for the Control of the Red Spider in Florida." Journ. Econ. Ent., Vol. I., pp. 377-380. Sirrine, F. A., 1900. "Insects Affecting Carnations." Amer. Florist, Vol. XV., No. 613, pp. 909-913, 6 pis. (Red Spiders on Carnations and Control, p. 910.) Surface, H. A., 1906. "Mites or Red Spiders on Leaves." Pa. Dept. Agri. Mo. Bull., Div. ZooL, Vol. IV., No. 3, pp. 95, 96. (Recommends spraying with potassium sulfid.) Taylor, W., 1896. "Notes on Destroying Red Spider." Journ. Hort., Ser. 3, Vol. XXXIIL, No. 854, pp. 440, 441. Titus, E. S. G., 1905. "Red Spider on Cotton." Bull. 54, Bur. Ent., U. S. Dept. Agri., pp. 87, 88. Titus, E. S. G., 1905. "The Cotton Red Spider." Giro. 65, Bur. Ent., U. S. Dept. Agri., 5 pp., 2 figs. Volck, W. H., 1903. "Sulfur Sprays for Red Spiders." Bull. 154, Cal. Agri. Exp. Sta., 11 pp., 3 figs. Volck, W. H., 1913. "The Control of Red Spiders." Monthly Bull. State Com. Hort., Cal., Vol. 2, pp. 356-363. Webster, F. M., 1899. "The Chinch Bug. Experiments with Insecticides." Bull. 106, Ohio Agri. Exp. Sta., pp. 235-256, 5 figs. (Carbon Bisulfid against Red Spider, pp. 254, 255.) Weldon, G. P., 1909. "Two Common Orchard Mites." Bull. 152, Colo. Agri. Exp. Sta., 12 pp., 7 figs. Weldon, G. P., 1910. "Life History Notes and Control of the Common Orchard Mites." Journ. Econ. Ent., Vol. III., No. 5, pp. 430-434. Wilson, H. F., 1911. "Notes on the Red Spider Attacking Cotton in South Caro- lina." Journ. Econ. Ent., Vol. IV., pp. 337-339. Woglum, R. S., 1909. "Fumigation Investigations in California." Bull. 79, Bur. Ent., U. S. Dept. Agri., 73 pp., 28 figs. (Citrus Red Spider, p. 11.) Woodworth, C. W., 1902. "The Red Spiders of Citrus Trees." Bull. 145, Cal. Agri. Exp. Sta., 19 pp., 5 figs. Woodworth, C. W., 1903. "Entomology." Univ. Cal. Agri. Exp. Sta. Rept., 1901-03, pp. 104-110. (Red Spider Remedies, p. 105.) Worsham, E. L., 1910. "The Cotton Red Spider." Bull. 92, Ga. Agri. Exp. Sta., pp. 135-141, 5 colored plates. BULLETIN No. (80 NOVEMBER, 19 17 MASSACHUSETTS AGRICILTIRAL EXPERIMENT STATION Report of the Cranberry Sub- station FOR 1916 By H. J. FRANKLIN AND OBSERVATIONS ON THE SPOILAGE OF CRANBERRIES DUE TO LACK OF PROPER VENTILATION By C. L. SHEAR and NEIL E. STEVENS, Pathologists, and B. A. RIDOLPH, Scientific Assistant, Eruit- Disease Investigations, Bureau of Plant Industry, Inited States Depart- • ment of Agriculture Requests for bulletins should be addressed to the Agricultural Experiment Station, Amherst, Mass. fVlassachusetts Agricultural Experiment Station. OFFICERS AND STAFF. Trustees. COMMITTEE. Charles H. Preston, Chairman, Wilfrid Wheeler, Edmund Mortimer, Arthur G. Pollard, Harold L. Frost, . The President of the College, ex officio. The Director of the Station, ex officio. Hathorne. Concord. Grafton. Lowell. Arlington. STATION STAFF. Administration. William P. Brooks, Ph.D., Director. Joseph B. Lindset, Ph.D., Vice-Director. Fred C. Kenney, Treasurer. Charles R. Green, B.Agr., Librarian. Mrs. Lucia G. Church, Clerk. Miss F. Ethel Felton, A.B., Clerk. Agricultural Economics. Alexander E. Cancb, Ph.D., In Charge of Department. Samuel H. DeVault, A.M., Graduate Assistant. Agriculture. William P. Brooks, Ph.D., Agriculturist. Henry J. Franklin, Ph.D., In Charge Cranberry I nvestiga' tions. Edwin F. Gaskill, B.Sc, Assistant Agriculturist. Robert L. Coffin, Assistant. Botany. A. Vincent Osmun, M.Sc, Botanist. Georob H. Chapman, Ph.D., Research Physiologist. Paul J. Anderson, Ph.D., Associate Plant Pathologist. Orton L. Clark, B.Sc, Assistant Plant Physiologist. W. S. Krout, M.A., Field Pathologist. Miss Mae F. Holden, B.Sc, Curator. Miss Ellen L. Welch, A.B., Stenographer. Entomology. Henry T. Fbrnald, Ph.D., Entomologist. Burton N. Gates, Ph.D., Apiarist. Arthur I. Bourne, A.B., Assistant Entomoloaiat. Stuart C. Vinal, M.Sc, Assistant Entmriologist. Miss Bridie E. O'Donnell, Clerk. Horticulture. Frank A. Wauqh, M.Sc, Horticulturist. Fred C. Sears, M.Sc, Pomologist. Jacob K. Shaw, Ph.D., Research Pomologist. Harold F. Tompson, B.Sc, Market Gaidener. . Miss Ethelyn Streeter, Clerk. Meteorologry. John E. Ostrander, A.M., C.E., Meteorologist. Microbiology. Charles E. Marshall, Ph.D., In Charge of Department. Arao Itano, Ph.D.. Assistant Professor of Microbiology. George B. Ray, B.Sc, Graduate Assistant. Plant and Animal Chemistry. Joseph B. Lindsey, Ph.D., Chemist. Edward B. Holland, Ph.D., Associate Chemist in Charge (Research Division) . Fred W. Morse, M.Sc, Research Chemist. Henri D. Haskins, B.Sc, Chemist in Charge (Fertilizer Division). Philip H. Smith, M.Sc, Chemist in Charge (Feed and Dairy Division) . Lewell S. Walker, B.Sc, Assistant Chemist. Carlbton p. Jones, M.Sc, Assistant Chemist. Carlos L. Beals, M.Sc, Assistant Chemist. James P. Buckley, Jr., Assistant Chemist. Windom a. Allen, i B.Sc, Assistant Chemist. John B. Smith, ' B.Sc, Assistant Chemist. Robert S. Scull, i B.Sc, Assistant Chemist. James T. Howard, Inspector. Harry L. Allen, Assistant in Laboratory. James R. Alcock, Assistant in Animal Nutrition. Miss Alice M. Howard, Clerk. Miss Rebecca L. Mellor, Clerk. Poultry Husbandry. John C. Graham, B.Sc, In Charge of Department. Hubert D. Goodale, Ph.D., Research Biologist. Miss Rachael G. Leslie, Clerk. Veterinary Science. James B. Paige, B.Sc, D.V.S., Veterinarian. G. Edward Gage, Ph.D., Associate Professor of Animal Pathology. John B. Lentz, ' V.M.D., Assistant. > On leave on account of military service. CONTENTS. PAGE Report of the cranberry substation for 1916: — Blueberry culture, . . . . . . . . . .183 Weather observations, ......... 183 Frost protection, .......... 184 Fungous diseases, .......... 186 Storage tests, ........... 193 Tentative practical conclusions based on the results of the storage tests 216 Resanding, ........... 218 Fertilizers, 222 Insects 223 The cranberry rootworm, ........ 223 The gypsy moth 224 The cranberry tip worm, ........ 226 The black-head fireworm 226 The cranberry fruit worm, . . . . . . . . 227 Bog management, .......... 232 Observations on the spoilage of cranberries due to lack of proper ventilation: — Introduction, ........... 235 Temperature tests in open and closed cans, ..... 236 Effect of carbon dioxide on cranberries, ...... 237 Effect of different relative humidities on spoilage due to carbon dioxide, 238 Relation of fungi to spoilage due to carbon dioxide, .... 238 Effect of carbon dioxide on fungi in the berries, ..... 239 Publication of this Document approved bt the Supervisor op Administration. BULLETIlSr JSTo. 180. DEPARTMENT OF AGRICULTURE. REPORT OF THE CRANBERRY SUBSTATION FOR 1916. BY H. J. FRANKLIN. The investigations were mainly along the lines pursued in 1915. Many- storage tests were conducted with the fruit, the description and results of which will be found particularly interesting. Blueberry Culture. A quarter of an acre was planted with six distinct strains of specially selected and bred swamp blueberry stock provided by the Bureau of Plant Industry of the United States Department of Agriculture. This was done under the direction of Prof. Frederick V, Coville, for the most part on August 3 1, about 375 plants being set out. The rows were 8 feet apart, and the plants were set at intervals of 4 feet in the row. Most of these plants made some growi^h during the fall, and seemed in good condition when winter began. A check row of unselected stock, taken from a neighboring swamp and planted on May 18, grew well during the summer. Many supe- rior wild plants were selected when in fruit and marked for planting in 19 17 as an additional check. It is hoped that the selected blueberry may prove a satisfactory substitute for cranberries on bogs where conditions make the growing of the latter fruit unprofitable. The commercial growing of the blueberry may also develop enough to compete with that of the cranberry in the cultivation of swamp soils, and thus provide a new industry for Massachusetts. Weather Observations. Weather observations were made as in previous years, thermometer readings and amounts of precipitation being telegraphed daily to the Bos- ton office of the Weather Bureau during the periods of frost danger, and frost conditions being telephoned to growers on cold nights when asked for. The frost damage on the Cape this season was negligible. 184 MASS. EXPERIMENT STATION BULLETIN 180. Beginning with the second decade in May, wet weather prevailed more or less until about the 1st of August, culminating on July 24 in an all-day rain in which 4.20 inches fell at the station bog in twenty-four hours, this, because of the previous saturation of the ground, causing the streams to rise so much that the bogs located in considerable watersheds were generally flooded in spite of all efforts to keep the water down, 'It was estimated that over 1,000 acres of bearing bog on the Cape, either in or a little past the blooming period, were entirely submerged in this way. The wet season provided unusual chances to study the effects of water on the blossoms and small berries. As a rule, the bogs bloomed heavily, and for a time a record-breaking crop was expected, but an unusually large proportion of the blossoms failed to set fruit. This failure took place especially among the under berries, for the crop turned out to be more "on top" than usual. Almost no berries were commonly found in thick clumps of vines where the blossoms had been very abundant, while in thin vines near by there was a fair amount of fruit. These conditions were gen- eral, though less so on bogs that either had no winter-flowage or had it taken off early. The wet weather evidently caused this failure of the set, though it is hard to say definitely how it did so. The rain may have pre- vented a proper fertilization of the flowers either by washing off the pollen or by preventing bees from working actively. Perhaps an unusual preva- lence of fungous diseases induced by the excessive moisture blasted the blossoms. It is the wi'iter's present opinion, based on general observation and ex- perience, that late holding of the winter-flowage so throws the blossoming period out of its normal season that the danger of its meeting unfavorable conditions for the setting of the fruit is usually considerably increased thereby. That flooding when the berries are small is dangerous was shown by the effects observed on some bogs submerged for not over fifteen hours with the blooming period past and crop fully set. These bogs lost half their berries in spite of the cloudy weather that prevailed when the water was let off and for three days afterward. The largest of the berries injured under these circumstances were somewhat over a quarter of an inch in diameter. Many of the larger berries on some bogs, however, endured submergence two or three days without apparent injury. Frost Protection. In the fall of 1915 tests with new tobacco cloth, used in various ways on a bog with much moss under the vines, showed no considerable temperature advantage. In the spring of 1916 this cloth was tried on a bog that was fairly well sanded and with only a little moss. Green registering thermometers were used in all the tests. Under one thickness of cloth spread on the vines they showed a higher minimum temperature than thermometers not cov- ered, — by 3 degrees in some cases, though the usual difference was less REPORT OF CRANBERRY SUBSTATION FOR 1916. 185 than 2 degrees. Two thicknesses spread on the vines raised the minimum temperature from 31 to 5 degrees, according to wind conditions, above that over the unprotected bog. One thickness supported on wires about hip high gave a medium advantage as compared with the single and double thicknesses spread on the vines. In the fall these tests were continued on patches of unpicked vines on the station bog, and a maximum advantage of about 3 degrees with a single thickness and of 6 degrees with a double one was obtained. More- over, this advantage continued after the vines had been covered with the cloth continuously day and night for nineteen days in a test begun Sep- tember 25 and ended October 14. The experience with this cloth justifies the following conclusions: — (a) This protection is not satisfactory on bogs with much moss under the vines because of the reduced radiation on such bogs. (b) Good secondhand cloth is so hard to get that its use is not practi- cable. (c) One thickness of new cloth is not enough when spread on the vines. (d) The difficulties and expense of wire supports prohibit their use. (e) With two thicknesses spread on the vines, the protection is proba- bly sufficient for most of the Cape bogs, and this seems the best way to use it. It is too bulky to handle easily on large areas, but it may be left on a bog continuously during quite a long cold period without reducing the protection afforded. (/) It is better to protect with water if it can be done at reasonable expense. Howes ^ berries that had undergone various low temperatures were picked and examined on November 15, as follows: — 1. Of 433 berries that had endured a temperature of 15^* F., 375 were entirely sound and 58 were soft. Eighteen of the latter showed unmis- takably that they had decayed from fungous disease, leaving only 40, or 9.64 per cent., that could have been softened by frost; and perhaps even this figure should be reduced on account of fungous rot that could not be distinguished. 2. Of 442 berries that had undergone a temperature of 13|° F., 340 were sound and 102 soft. Of the latter, 26 showed that they had rotted because of fungous diseases, this leaving 76, or 18.27 per cent., that might have been frosted. 3. Of 444 berries exposed to a temperature of 9° F., 200 seemed entirely sound, 244 being soft. Twenty of the latter evidently had been softened by diseases, leaving only 224, or 52.83 per cent., that could have been hurt by frost. • This variety has been called "Late Howe" in previous reports of the cranberry substation. The writer is informed that it was first taken from the wild, and cultivated by the late James P. Howes of East Dennis, Howes being a common family name in that part of Cape Cod. As "Howe" is evidently a corruption, and as "late" is superfluous, all the varieties that have been called " Howe" being late, the name Howes is considered more appropriate and is therefore used in this report. 186 MASS. EXPERIMENT STATION BULLETIN 180. The temperatures here recorded were taken with Green minimum regis- tering thermometers hung just over the vines bearing the berries. The fruit was well colored when it underwent these temperatures. Several tests in both 1915 and 1916 showed that the temperature at which freezing begins among ripened Early Black or Howes cranberries is at or slightly above 22° F., no softening resulting from exposure to 23°. The records of minimum temperatures at the station bog from 1911 to 1916, inclusive, show that no temperature low enough to harm well-colored berries appreciably occurred in any picking season of those six years. The results of these investigations show that, for bogs in warm or aver- age locations that are flooded by pumping, it is unprofitable in the long run to try to protect well-colored berries from frost, especially if the crop is light. Fungous Diseases. These investigations were conducted, as in previous years, in co-opera- tion with the Bureau of Plant Industry of the United States Department of Agriculture, Dr. C. L. Shear and his assistant, Dr. Neil E. Stevens, visiting the Cape several times during the season, the latter spending sev- eral weeks at the station, and both giving sustained and aggressive atten- tion to the more technical side of the work during a considerable period in the growing season and throughout the fall and early winter. Table 1 is the season's record of the principal Bordeaux mixture spraying plots, experiments with which have been described in previous reports. None of these plots were treated this year, but the record is included here to show the effects on the 1916 crop of the spra^-ing done in former years. Plots A, B, C, D and E were all sprayed in 1911, 1912 and 1913. The treatment was continued on plots A, B and D in 1914, but was stopped on C and E. It was further continued on A (entire plot) and on one-half of B and one-half of D in 1915. Plots 15 and " 1913" were sprayed in 1913, 1914 and 1915. The whole of plot 15 has been treated with a complete mixture of commercial fertilizers for several years, as was also the middle part of A in 1913 and 1914. All the plots were picked with scoops as hereto- fore. Where two checks were taken they were laid out on opposite sides of the plot. The entire sections on which D and E are located, being small, were used as checks. The fruit used in the storage tests was stored, without separating, in quart cans with the covers on tight, but not sealed, the berries being taken by hand from different parts of the picking crates, all the crates picked being thus represected in the cans in most cases. KEPORT OF CRANBERRY SUBSTATION FOR 1916. 187 Percent- age of Eotten and Partly Rotten Berries at End of Storage Test. S § in g 2 ?5 5 S5 ° 2 " s 5 3! S § S 1 s s 1 3 02 hJ Si S s S! S5 S s; t^ E5 « « ^ ^ § s 00 00 ^ ^ ^ % > ^ ^ o > 8 i S? S^ > o >• o § § :? Z 'A iz; ^ Is i;; 15 ^ 0 P P P 12; ^ p p = « s s S 5 S S Z 5 5 s 5 5 3 2 s s s -^ "* •V ^ T,. '^ *^ ^ r.. " s M S S •* " S g5 ^ ^- ^ ^ *-• *.• CU o o o o o o O 5 o eg eg eg eg 6 o eg tg Quan- tity of Fruit placed in Storage Test (Quarts). a> - « 00 CO « - «, « «, « S S CO ^ *< §3 S5 «s- o J a I J 1 1 1 a 1* M *" ^ a 1 „• ^ .2 .^ .M ^ r!<" > 1^- s ^ ^ § S s P3 W pq n s i s s S? S S Si >> >, ^ ^ s s >> >1 ^ Pi. fe fe ^ -j; ^ ^ w w w w s S s w w w w W M w H W §«:t -s "S "S t~ «> s ^^1 ' ' ' ' ■* ' ' a ll 1 OS &s 'm ■y •^ •% ■^ ■M 'O y. ■^ "O • "S Is •a "2 n p. -s D< o< =a D. 2 2 0. •a 2 -g 1 00 .4^ ^ (& OQ ^ ^ ^ ^ ^ ^ ^ o ^ ^ Z ^ 1 ^ 1 s ^ «■ ^ ^ . as u ■■3 s < 1 1 ,g 1 1 ^ - ? i ^ o Pli :? -s ■"^ §• a -p: ■g 2 ^ ^ ■a McFarlin, . McFarlin. . Early Black, . Early Black, i S' « 00 N 1 Sprayed, . Not sprayed, . , Sprayed, . Not sprayed. o B (part not sprayed in 1915), . B (check) D (part not sprayed in 1915), . D (check), REPORT OF CRANBERRY SUBSTATION FOR 1916. 189 The table shows that as a rule the areas sprayed in 1915 were less pro- ductive in 1916 than their untreated checks, and that the fruit from these sprayed areas was inferior in keeping quality in all cases in 1916. In this connection the figures given for plots B and D in Table 2, taken from the last report of the substation (Bulletin No. 168, page 3), are of interest. Judging by the results of the 1915 and 1916 storage tests given in Tables 1 and 2, the resistance of the plants to the attack of fungous dis- eases had been weakened by the injury caused by Bordeaux mixture described in previous reports. Three plots, numbered, respectively, B. L. 1, B. L. 2 and B. L. 3, were sprayed with "Black-Leaf 40" used at the rate of 1 part to 400 parts of water, 2 pounds of resin fish-oil soap to 50 gallons being added to spread and stick the spray, on the dates and with the results shown in Table 3. These plots and their checks were all picked with scoops. The storage- test fruit was stored, without separating, in quart cans with covers on tight but not sealed, the berries being taken by hand from different parts of the picking crates, all the crates being thus represented. The spray evidently did not much affect the quantity of fruit, and the storage tests showed no fungicidal value for it. This was not entirely ai fair test, as all the sprayed areas had been treated with complete commer- cial fertilizer mixtures in 1915, but the impairment in keeping quality shown by the sprayed berries as compared with the check fruit was in all cases greater than that heretofore found by the writer to have resulted' from the use of fertilizers. Did this spray have a harmful effect in this regard in some way? Two plots, numbered A. L. 1 and A. L. 2, were sprayed with "Corona" arsenate of lead, used at the rate of 3 pounds to 50 gallons of water, on the dates and with the results shown in Table 4. These plots and their checks were picked with scoops, and the storage-test fruit was selected and stored! in the same way as that of the "Black-Leaf 40" plots. 190 MASS. EXPERIMENT STATION BULLETIN 180. Percent- age of Rotten and Partly Rotten Berries found at End of Storage Test. ^ S S § g 5 5 § g s s ?g 1 CO M « O ui 3 2 2 ?! S "> t i. ^ s & S g S S ^ 1 1 z ^ ^ 2 2 2 "> S S III 00 00 00 00 to «5 2" lll^Pp g s s § sg ?; 3 § g ^ ■^ t-^ 00 S SI - •I -J4 §5 §5 "^ III "= s s 1 11 1 $5 III i ' i I til ■ < ^ lib r i t llii 1 g 1 Q 00 00 00 to ■* •>»< .s > ts 1 pq M - H W W '. 1 1" . pa m 1 >> >. Www E a < 1 CO h4 h4 hJ n m PQ 1 ■ 1 CJ CO M H^ h4 >4 m PQ pa REPORT OF CRANBERRY SUBSTATION FOR 1916. 191 Percent- age of Rotten and Partly Rotten Berries found at End of Storage Test. si SI g 2 § 5 S s S s g s . > > t> > t» •s| Z 12; 'A ^ ;? 15 B 5 s 5 5 S S S ^ S S §3 m 1=;, ^, eg eg eg eg eg eg ^ a „> '- 00 00 00 2: 2 S fl*©.- m feT3'S_; i^ ;:3 § s S § -^ m^Pt- •^ :S S'S'-I-S^ * S .^ § o 2 § mit' o 03 OJ «o to 55 m lO ^ ^ -H s oSo-o M i Hh 0. eg eg eg 1 ii ^.^ Ol i ^ 1 1 bi . . « -» o !=» tt^ OS o o e» «5 <0 ?a . >. >> >. ^ ^ ^ fS ^ [^ ^ ;! s o ;:j fj « N a ^ '-' 2 M J3 g ^j vii -S Hi s < '^ - : B.' B.' S.° p^ b B<' h B.' B: B,° S^ §5gS§§ 2^§§§g tn M 0l«0>0 g^SStgg W 0. ^cico^_;- |6 ««eo-*u5ffl -Hoo«o o d ^(§ . O M -§l ^ 1 11 S 11^ 1 § 1 § ^s w 2 1 2 1 .4-1 1 Qua tity Fru (Bus els) to O «N0S«OO 2:2:s2;s2: ^332:2:2: . 3 to Dec . 3 to Dec . 3 to Dec . 3 to Dec 3 to Dec . 3 to Dec . S to Dec . 3 to Dec . 3 to Dec . 3 to Dec . 3 to Deo . 3 to Dec 686668 666686 81° F. 81° F. 74° F. 76° F. 76° F. 75° F. 81° F. 81° F. 73° F. 76° F. 76° F. 75° F. 11.30 A.M. 12.00 M. 3.00 P.M. 4.00 P.M. 4.00 P.M. 4.00 P.M. 11.30 A.M. 12.00 M. 3.00 P.M. 4.00 P.M. 4.00 P.M. 4.00 P.M. 1 1 81° F. 81° F. 74° F. 79° F. 79° F. 75° F. 81° F. 81° F. 73° F. 79° F. 79° F. 76° F. s § a a a a a a a aa a a a as 11.15 a 11.45 A 2.35 p 10.30 A 10.45 A 2.00 P 11.25 A 11.55 A 2.45 p 10.35 A 10.55 A 2.10 P 1 -HC,«^U,« i 1 1 1 s 1 s 1 1 1 ! 1 $ to to 1 1 1" 1 o Q REPORT OF CRANBERRY SUBSTATION FOR 1916. 199 In partial confirmation of the evidence presented above, that scoop- picking is not especially harmful to the keeping quality of cranberries, a recital of the experience with 14 bushel crates of Early Black berries picked with scoops in two different ways from narrow alternating par- allel and adjacent strips of vines is here included. In picking seven of these crates the scoops were allowed to fill to a considerable extent as usual before emptying, the berries churning back and forth as they accumulated. With the other boxes the berries were not allowed to collect as they were picked, but were poured out of the scoops after each pull through the vines. The results of the storage of this fruit are shown in Table 8. The churned berries kept as well as the unchurned. The crates were examined by the "nine-sample" method. Table 8. — Picking Test. — The Scoop-churning of Berries during the Process of Picking does not materially affect Keeping Quality. How Berries WERE SCOOPED. Date picked and stored. Quan- tity stored (Bush- els). How stored. Date ex- amined to de- termine Rot Per- centage. Percent- age of Rotten and Partly Rotten Berries found at End of Test. With churning, Without churning, . Oct. 8 Oct. 8 7 7 Unseparated, in picking crates, . Unseparated, in picking crates, . Dec. 19 Dec. 19 29.51 29.64 4. Relative Keeping Quality of the Upper and Under Berries of the Vines. — The three tests to determine this were carried out as indicated by Table 9, the results showing rather conclusively that the berries most exposed to sun and wind during their growth are considerably better keepers than those produced under the protection of the vines. Moreover, the top berries were much more highly colored and averaged considerably larger in size than the others when picked. These berries were aU picked by hand under the supervision of the writer, who did much of the work himself. They were stored in quart cans. 200 MASS. EXPERIMENT STATION BULLETIN 180. Table 9. — C/pper and Under Berries compared as to Keeping Quality. Test No. Variety. Berries. Date picked. Quan- tity placed in Storage Test (Quarts). Period of Storage Test. Percent- age of Rotten and Partly Rotten Berries found at End of Storage Test. 1, . 2, . 8, . Early Black, . Early Black, . Howes, . . Only sound upper berries. Only sound under berries. Only sound upper berries. Only sound under berries. Only sound upper berries. Only sound under berries. Sept. 30 Sept. 30 Oct. 6 Oct. 6 Oct. 13 Oct. 13 6 6 14 12 6 6 Sept. 30 to Dec. 2 Sept. 30 to Dec. 2 Oct. 6 to Nov. 20 Oct. 6 to Nov. 21 Oct. 13 to Dec. 9 Oct. 13 to Dec. 9 31.55 38.83 28.74 37.93 15.49 18.44 It seems to be the general experience with Cape Cod bogs that late holding of the winter-flowage improves the keeping quality of the berries. As the writer has observed that late holding of the water frequently re- duces the quantity of under berries as compared with the amount of fruit produced in the tops of the vines, the results of these tests may partly explain this improvement. They also suggest that the generally recog- nized good comparative keeping quality of the 1916 crop may have been due largely to the very general failure of the under berries to set in their usual abundance. The deeper the scoops are run through the vines in picking, the greater the proportion of under berries that are gathered and the greater, also, the quantity of unattached cranberry leaves and sand that gets mixed with the fruit. On account of the inferior keeping quality of the under berries here shown, and because of the harm done by admixtures of loose leaves proved by tests described below (No. 7, page 206), the desirability of closely scooping berries that are to be stored long is rendered doubtful. 5. Housing promptly v. Leaving Crates of Berries in the Sun on the Bog, as affecting Cranberry Keeping. — Eight series of tests were carried out in this connection, four with Early Black and four with Howes fruit. Four of these were conducted in connection with the picking experiments described above (No. 3, page 197), Table 7 showing their arrangement and results. Dr. Stevens took all the temperatures given in this table with chemical thermometers, their bulbs being plunged to the centers of the crates. At 8 a.m., September 19, the temperatures of the twelve boxes of Early Black berries ranged from 68° to 70° F., and at 8 a.m., September 20, they ranged from 61° to 62°, from which there was little change for several days after. REPORT OF CRANBERRY SUBSTATION FOR 1916. 201 The records in Table 7 show that as a rule the temperature of berries left in crates on the bog exposed to the sun for several hours did not change more than 3 degrees. The temperatures of some of these crates were taken every thirty minutes from the time they were picked until they were housed, almost no variation being discovered until very near the latter time. The averages of percentages given in the table indicate that the Early Black berries housed at once kept somewhat better than those left on the bog, whereas these results with the Howes fruit were reversed. This difference in the storage of the two varieties corresponded with the differ- ence in the average temperatures of the different lots when housed, the Early Black berries housed at once averaging to have lower temperatures when placed in storage than did those left on the bog, whereas the Howes fruit housed at once had a somewhat higher average temperature when stored than did that left on the bog. The four other experiments under this head were carried out in connec- tion with some of the tests of the effect of wetness on cranberry keeping described below (No. 6 (a), page 201), Table 10 exhibiting their arrange- ment and results. As in the first four series of tests, Dr. Stevens took all the temperatures with chemical thermometers at the centers of the crates. It was partly cloudy all day the day that the Early Black berries used in these tests were picked. The averages of percentages in the table show that with both varieties the wet berries kept better after having been left on the bog, whereas the dry ones kept better when housed at once. On the whole, the results of these tests were inconclusive, though they failed to show much harm to the keeping quahty resulting from leaving the crated fruit on the bog for several hours under ordinary harvesting and storage conditions. 6. Wet and Dry Cranberries compared as to Keeping. — Three series of tests come under this head, as follows: — (a) An area 60 feet square laid out on Early Black vines on the station bog was divided into equal parts by lines running diagonally between the corners. Two of the opposite triangles thus formed were scooped while the berries were wet with dew, the other two being left until they were dry. The ways in which these berries were tested and the results obtained are shown in test No. 1 of Table 10. (b) An area 100 by 30 feet laid out on Howes vines on the station bog was divided into triangles by diagonal lines between the corners. Two opposite triangles were picked with scoops while the vines were more or less wet with dew, and the other two when they were dry. The manner of testing this fruit and the results obtained with it are shown in test No. 2 of Table 10. 202 MASS. EXPERIMENT STATION BULLETIN 180. 1^1 36.19 23.05 Averages of Per- centages of Rotten and Partly Rotten Berries. 5 S S § Percent- age of Rotten and Partly Rotten Berries found at End of Storage Test. SOCOIOCOCOIO totoiocototo PQQQQQ QQQQQQ S33$2S 22S32S iiii ^:5Sggg 55:s»"" 222""" d::::^^;^ Tem- perature of Berrries when picked. hi s s s s s s s s s s s s i 1 ■ ■ 5 S - - w fa t§ ;§ M ^- ^' « i u ~ e^ REPORT OF CRANBERRY SUBSTATION FOR 1916. 205 Percentage of Rotten and Partly Rotten Berries found at End of Storage Test. 33.74 30.62 19.70 16.84 18.02 25.17 11.21 11.01 9.96 14.83 14.04 12.56 20.95 a .2 1 With neither vines nor leaves, ' . . . With vines and leaves attached, 2 . With neither vines nor leaves. With vines and leaves attached, . With vines stripped of leaves, ' . . . With leaves only,* With neither vines nor leaves, . ... With vines and leaves attached, . With vines stripped of leaves. With leaves only With neither vines nor leaves. With vines stripped of leaves. With leaves only Quan- tity of Fruit placed in Storage (Bush- els). -- «»«„ c«„.. «co« d 3 i 1 Sept. 20 to Dec. 18, . . \ Sept. 29 to Dec. 18, . . j Oct. 5 to Dec. 19, Oct. 7 to Dec. 18, llll Sept. 20 Sept. 29 Oct. 5 Oct. 7 I o > Early Black, . . Early Black Howes, Howes II -■ El .i'l a o a o — > 111 «:-^£ -gab" c o fe-S iS iz; H <; 206 MASS. EXPERIMENT STATION BULLETIN 180. The table shows that the results of this test strongly confirmed those of the first two, giving striking evidence of the harmful effect of excessive moisture among cranberries in storage. 7. Effects of Admixtures of Vines and Leaves on Cranberry Keeping. — The four series of tests in this connection were carried out as shown in Table 12. The fruit was picked with scoops and was stored in bushel picking crates. The crates were examined by the "nine-sample" method. The table shows that these tests gave convincing evidence of the harm- ful effect of an admixture of unattached cranberry leaves in the storage of the fruit. They also indicated that the berries keep as well with the ad- mixture of vines and leaves attached, commonly obtained in scooping, as any way. The entire removal of the vines and leaves, aside from the injury done in the process, however, seems to do no harm. 8. Berries separated with Hayden and with White Machines and Berries screened intho^d separating compared as to Keeping Quality. — The berries used in these two series of tests were handled throughout in the same way. The three lots of fruit in each series came from the same source, individual crates of berries as they came from the bog being divided as evenly as pos- sible into three separate parts by successive pourings into barrels to pro- duce them, care being taken to handle the berries of the different lots as nearly alike as possible. As there was no White separator in w^orking order in East Wareham at the time, all this fruit was carted in open barrels in a farm wagon (without springs) to the Makepeace screenhouse at Wareham, two of the lots of each series being there run through Hayden and White separators, respectively. The berries were received into barrels from both the Hayden and the White machines, those of the fb'st box (the "good" box) also being used in the test in the case of the former. The berries of all the lots were carted back in the open barrels to the station screenhouse, where they were hand-screened, the fruit in all cases being received into picking crates placed close to the mouths of the screens and being stored in those crates. The arrangement and results of these tests are shown in Table 13. The " nine-sample " method was used in examining the crates. REPORT OF CRANBERRY SUBSTATION FOR 1916. 207 |g^c^.-a i ^ 2 3 ° ^ oi ■»^ -l ^ s~ a (^ S f^ la •^ • -s^ r^ •si ^ SPh - - - . » ■ s M ?5 c3 c^ ^ g g ■g 1 1 "S ■§ t o „- s § 1 1 1 P 1 11 -sl o «5 CO to to to ci CJ CT ^ c <3 i >, i| 1 o s 1 1 i 1 "O 2 5 •X3 (5 i 1 g i 1 1 535 I g S 1 a S S- 5 1 -•§ ■§ i 1 ^ S s ^ w w ^ w w 1 ?5i i#l? ^ -T- ^ ■* ^ ^ ■j? 1 1 c6 > w J >-■ < s H s > J 208 MASS. EXPERIMENT STATION BULLETIN 180. The figures of the table indicate that, in both tests, the White machine apparently affected the keeping qualities of the fruit about the same as did the Hayden. This result is surprising, and must be verified by future experiments. The difference in the tendency to rot between the separated and unseparated berries was not as great as in last year's tests. This may have been partly due to the injury that all the lots of fruit probably re- ceived in the carting, this perhaps partly hiding the real difference in the damage done by the various methods of cleaning. 9. The Injury to the Keeping Quality of Cranberries caused by Separators employing the Bouncing Principle and by the Drop in the Barrel. — That this varies greatly -with different lots of berries was indicated by the results of half a dozen minor experiments conducted by Dr. Stevens. The range in the increase of decay caused by these factors in these tests was from about 14 to about 127 per cent. A new arrangement devised by the writer for preventing the barrel in- jury, for use both in screening and in connection with separators, works Avell mechanically and promises to be generally satisfactory, though no storage tests have been conducted to determine the degree of its effective- ness. This device is on exhibition at the offices of the New England Cran- berry Sales Company, Middleborough, Mass., and the J. J. Beaton Growers' Agency, Wareham, Mass., and it also may be seen at tk station screenhouse at East Wareham at any time during the cranberry season. 10. The Effect of Grading on the Keeping of Cranberries. — The two fol- lowing series of tests come under this head : — (a) Two lots of Early Black berries picked in the same location on the station bog were treated as shown in Table 14. To make sure of their being well cleaned thej^ were run through a Hayden separator twice imme- diately before they were stored. Only the berries going into the separator barrels were used in the test. Neither lot was hand-screened. They were stored in bushel picking crates of the same dimensions and construction. The Hayden grader was used. A board was in the grader frame in place of the grader while the second lot was run through. The spacing of the grader, fourteen thirty-seconds of an inch, was wider than that commonly used, and it took out from a fifth to a quarter of the entire quantity of berries put through the separator while it was in use. REPORT OF CRANBERRY SUBSTATION FOR 1916. 209 Percentage of Rotten and Partly Rotten Berries found at End of Storage Test. !5 m a a & S c a ^ Average Cup-count of Berries at End of Storage Test. 1 1 ii s s Quantity of Berries placed in Storage Test (Bushels). eo eo Ill - , i 4 IS «• eJ lU i si St 1 210 MASS. EXPERIMENT STATION BULLETIN 180. The figures of the table show that the closely graded berries kept con- siderably better than the ungraded ones, there being nearly 22 per cent, more rot among the latter at the close of the test. The cup-counts were taken with the inspectors' cup of the New England Cranberry Sales Com- pany. (6) Two lots of Howes berries were obtained for this series of tests by dividing boxes of fruit, just as they had been stored when they came from the bog on October 7, into equal parts by alternate dippings with a quart measure. They were put through a Hay den separator, with the upper set of bounce-boards set at the middle notch, on December 26. A board five- eighths of an inch thick was kept in the grader frame in place of the grader while the second lot was run through. The grader took out about a quarter of the quantity of berries separated while it was in use. Only the berries that went into the barrels from the separator were used. They were poured from the barrels into boxes and were taken into the warm screening room a box at a time, so that thej^ might undergo a high temperature no longer than necessary during the screening. Both lots were carefully screened at the same time on December 29, the berries being run into picking crates placed close to the mouths of the screens. They were carefully shaken down and stored in these crates at once. The arrangement and results of these tests are shown by Table 15. It will be seen that after a winter storage of nearly ten weeks almost 32 per cent, more berries showed rot among the ungraded fruit than among that which had been closely graded. At no time during the test did the temperature of the storage room range more than 8° above the freezing point of water, and for considerable periods it ran more or less below it. The cup-counts given in the table were taken, as in the first series of tests, with the Sales Company's cup. While it cannot safely be said that the results of these tests prove that grading improves the keeping of cranberries, they bring out a point of much importance. Closely graded berries, being larger and more uniform in size, are much more desirable in appearance than ungraded ones. If they also keep better, the advisability of preparing them for market in this way as a means of inducing greater consumption is much confirmed. If close grading were generally practiced it could be made a powerful factor in properly controlling the cranberry market, for, while it tended strongly to increase consumption on one hand, it would in a sense cut down production on the other. In the writer's opinion it would be the best possi- ble means for dealing with overproduction, for if any part of a crop had to be thrown away it would be only the berries of inferior size or quality. The results of these grading tests are entirely in line with last year's findings of the writer, in the study of ventilation as affecting cranberry keeping, and with those brought out by Dr. Shear and his collaborators in their paper published as a part of this bulletin. The small berries as well as the leaves, conclusive experiments with which are described above (No. 7, page 206), might be expected to check ventilation, not only by REPORT OF CRANBERRY SUBSTATION FOR 1916. 211 Percentage of Rotten and Partly Rotten Berries found at End of Storage Test. O to (M O ^ ^ §3 ^ S ^ g sj s? s s: S5 s Average Cup-count of Berries at End of Storage Test. eo t~ o •-< - o> lO to >o t^ O 0> O Oi g S § 2 § § g i . . „ . ■sgs5 llli Seven-sample, Seven-sample, 1 Dec. 29 to Mar. 7, Dec. 29 to Mar. 7, Quantity of Berries placed in Storage Test (Bushels). •4< U3 1 o . •r 1 Whether graded OR NOT. Graded, Not graded, . . . 1 M " c^" 1 212 MASS. EXPERIMENT STATION BULLETIN 180. mechanically reducing the spaces for the passage of air and gases among the fruit, but also by themselves using up oxygen and giving off additional carbon dioxide, in this way being especially harmful. 11. The Relative Effect of Barrel and Crate Containers on Cranberry Keeping in Shipments. — Three lots of Early Black and two lots of Howes berries, each lot consisting of a barrel and two half-barrel crates, made up an experimental shipment to determine this. All the berries of each lot came from the same place on the station bog, the different lots being picked in various locations, the Early Black on October 2 and the Howes on October 5. All five lots were run through a Hayden separator and screened on November 7. On account of difficulties encountered in ar- ranging for shipping this fruit with other berries in a carload, it was then kept in open barrels, all of which were nearly full, until November 17, when it was packed for shipment. The berries shipped in barrels were packed in the usual way, while the crated fruit was placed in 4-quart baskets like those used as containers for strawberries.^ All the lots were left in the packed condition in a cold room until November 20, when they were carted in a farm wagon (without springs) from East Wareham to Tremont Station. They were kept in the railroad freight-house over night and placed in different parts of a car on top of a carload of other berries the next morning. The car left Tremont November 21 and arrived in Washington, D. C, on Saturday, November 25. They were there left in the freight-house until the following Monday morning. They were then taken to Arlington Farm and stored at a temperature of about 50° F. until December 9. The barrels and crates were opened and stored in a laboratory, the temperature of which varied from 60° to 85° F., from December 9 until December 14 and 15, when they were sampled and examined, as follows: — (a) The eight following samples were taken from each barrel : — Nos. 1 and 2, two quarts near the top, just below the layer crushed in heading, — distinguished in Table 16 by the word "top." No. 3, one quart taken a quarter of the distance down from the top, — indicated by "J". Nos. 4 and 5, two quarts taken near the middle, — marked "|". No. 6, one quart taken from three-quarters of the distance from the top toward the bottom, — designated as "f ". Nos. 7 and 8, two quarts from near the bottom, — distinguished as "bottom." The berries were dipped out of the barrels down to the parts sampled, the samples being taken from all parts of the surface of the fruit exposed by the dipping, except within 2 inches of the staves. (6) Four 1-quart samples were taken from each crate of each lot at various places in the crate, so as to make up as fair an average as possible, each sample representing different baskets. ' The crates and baskets were furnished through the courtesy of Mr. J. J. Beaton of Wareham, Mass. REPORT OF CRANBERRY SUBSTATION FOR 1916. 213 PLiSSloSoiofe ►io--coaiN o t~ OJ (M O lO ■* 00 N KMC^MeOCMC ;3SSsS53ggSg O 00 CO t^ »-^ QO «* -^ CO «0 <«=) t^ 1-t t^ CD CD »0 Tf rt lO to 00 O -H — 1 M CO -< >o o ci ■<»<■ od t-^ N eo e^ M f5 CM ifj N 3 o c.a^ a >3!°- 0>CO*-HOiCOCOCOC 1 U-, C >>C S o » Sj gsssssss -■^cor«-coCOt^03T-IOi»i t-JfcOOO-* <=P5?!=» :§g5 S; S !2 J3 S !5 5S S! |l ■a MI'S e-r" I I g| "S ® E g S ,1 I J« cc I & ° »£S 22^?:s:si-'Ji ^ 2 fS goo 214 MASS. EXPERIMENT STATION BULLETIN 180. The sampling was done by Dr. Stevens. The results of his examina- tions are given in Table 16. They show that the crated fruit was in much better condition than that in barrels in all the lots, especially those of the Howes variety. The results of these tests accord with the conclusions given in last year's report (pages 23 and 24) regarding the use of crates instead of barrels as shipping containers for cranberries. These results Avere confirmed by those obtained with shipments of berries from another bog to Portland, Me., made by Dr. Stevens, but not described here. 12. The Relative Development of Decay in Different Periods of the Storage Season. — The four series of tests to determine this were conducted as follows : — (a) On September 22, 20 quart cans were filled with entirely sound ber- ries from each of 7 half-filled crates of Early Black fruit picked at the same time in the same general location on the station bog three days before. This fruit was stored at once, and the different 20-can lots were examined one after another at intervals of two weeks. (b) On October 4, 10 quart cans were filled with sound berries from each of 12 half-filled crates of Howes fruit picked at the same time and in the same place on the station bog the day before. These cans were stored at once, and the different 10-can lots were examined one after another at weekly intervals. (c) Quart cans were filled with sound Early Black fruit in lots of 10, from each of 13 half-filled crates successively, at weekly intervals from September 20 to December 13, inclusive, the berries all having been picked at the same time and in the same general location on the station bog on September 19. The cans of each lot were stored as soon as filled and were examined at the end of a two-week storage. (d) Quart cans were filled with sound Howes fruit in lots of 10, from each of 1 1 half -filled crates successively^ at weekty intervals from October 4 to December 13, inclusive, the berries all having been picked at the same time and in the same location on the station bog on October 3. The cans of each lot were stored as soon as filled and were examined at the end of a two-week storage. The arrangement and results of all these series of tests are given in order in Table 17. They failed to show any distinct difference in the rate of rot development in the various periods of the storage season, this general result differing from that of last j^ear's experiment ^ in this connection. The writer now thinks that the handling of the berries in selecting them for these tests, and their lack of ventilation in the tightly covered cans, may have so affected their keeping as to hide different results that perhaps would have been obtained under more normal storage conditions. The description of the tests is included here for its possible value in making future comparisons, and as a record of work done. Further experiments along this line should be tried. 1 Bui. No. 168, Mass. Agr. Expt. Sta., 1916, p. 18. REPORT OF CRANBERRY SUBSTATION FOR 1916. 215 Table V, Rot Development amonx] Cranberries stored in Tin Cans in Different Periods of the Storage Seaso7i. Test and V.iriety. Quan- tity of Berries used (Quarts). Date stored. Date ex- amined to deter- mine Rot Per- centage. Total Number of Berries. Number of Rot- ten and Partly Rotten Berries when ex- amined after Storage. Percent- age of Rotten and Partly Rotten Berries found at End of Storage. (o), Early Black, 20 Sept. 22 Oct. 6 11,415 450 3.94 20 Sept. 22 Oct. 20 11,641 1,516 13.02 20 Sept. 22 Nov. 3 11,506 3,069 26.67 20 Sept. 22 Nov. 17 11,630 4,167 35.83 20 Sept. 22 Dec. 1 11,781 5,118 43.44 20 Sept. 22 Dec. 15 11,599 6,316 54.45 20 Sept. 22 Dec. 29 11,412 6,532 57.24 (6), Howes, . 10 Oct. 4 Oct. 11 4,903 71 1.45 Oct. 4 Oct. 18 4,905 100 2.04 Oct. 4 Oct. 25 4,960 228 4.60 Oct. 4 Nov. 1 4,961 418 8.43 Oct. 4 Nov. 8 4,888 503 10.29 Oct. 4 Nov. 15 4,981 776 15.58 Oct. 4 Nov. 22 4,948 860 17.38 Oct. 4 Nov. 29 4,877 939 19.25 10 Oct. 4 Dec. 6 4,894 1,147 23.44 Oct. 4 Dec. 13 5,029 1,494 29.71 Oct. 4 Dec. 20 4,821 1,353 28.06 Oct. 4 Dec. 27 4,845 1,553 32.05 (c), Early Black, Sept. 20 Oct. 4 5,779 301 5.21 Sept. 27 Oct. 11 5,530 308 5.57 Oct. 4 Oct. 18 5,602 137 2.45 Oct. 11 Oct. 25 5,782 222 3.24 Oct. 18 Nov. 1 5,441 240 4.41 Oct. 25 Nov. 8 5,363 140 2.61 Nov. 1 Nov. 15 5,379 201 3.74 Nov. 8 Nov. 22 5,487 220 4.01 Nov. 16 Nov. 30 5,693 295 5.18 Nov. 22 Dec. 6 5,684 315 5.54 Nov. 29 Dec. 13 5,510 307 5.57 Dec. 6 Dec. 20 5,763 304 5.28 Dec. 13 Dec. 27 5,513 476 8.63 216 MASS. EXPERIMENT STATION BULLETIN 180. Table 17. — Rot Development among Cranberries stored in Tin Cans in Different Periods of the Storage Season — Concluded. Number Percent- of Rot- age of Quan- tity of Berries used (Quarts). Date ex- ten and Rotten amined Total Partly and Test and Variety. Date stored. to deter- mine Number of Rotten Berries Partly Rotten Rot Per- Berries. when ex- Berries centage. amined found at after End of Storage. Storage. (i). Howes 10 Oct. 4 Oct. 18 4,643 118 2.54 10 Oct. 11 Oct. 25 4,730 104 2.20 10 Oct. 18 Nov. 1 4,908 191 3.89 10 Oct. 25 Nov. 8 4,570 117 2.56 10 Nov. 1 Nov. 15 4,546 103 2.27 10 Nov. 8 Nov. 22 4.633 129 2.78 10 Nov. 15 Nov. 29 4,808 116 2.41 10 Nov. 23 Dec. 7 4,747 112 2.36 10 Nov. 29 Dec. 13 4,915 145 2.95 10 Dec. 6 Dec. 20 4,943 155 3.14 10 Dec. 13 Dec. 27 4.849 142 2.93 T 13. Incubator Test of Keeping Quality of Cranberries. — A few lots of Early Black berries were moistened and tested as to their keeping quality in quart cans, with the covers on tight but not sealed, in a chicken incubator run at a temperature of 80° F. The results seemed to show that the rela- tive keeping quality of cranberries can be determined in this way in a period of about forty-eight hours. Tentative Practical Conclusions based on the Results of the Storage Tests. 1. Cranberries should not be picked wet. 2. Scoop-picking is not particularly harmful to keeping quality. 3. Deep scooping is likely to affect cranberrry keeping adversely be- cause it gathers maximum amounts of under berries, loose leaves and sand, these materials being harmful in storage. 4. Cranberries left in the sun on the bog for a good part of the day during picking seem to keep about as well as those housed at once, under average storage-house conditions. There might be a great difference in this regard, however, if cooler storage were practiced, for the relatively high temperature usually had by the berries when they are picked proba- bly has a hurtful effect, hence the sooner they are cooled the better. 5. Lack of sufficient ventilation affects cranberry keeping adversely, apparently by interfering with the process of respiration, not by prevent- KEPORT OF CRANBERRY SUBSTATION FOR 1916. 217 ing the evaporation of moisture, as suggested in last year's report (pages 6 to 17). Cranberries, like other fruits, are living, breathing organisms when picked, and must take in oxygen and give off carbon dioxide freely to continue their life processes. They may do this for several months after they are taken from the vines. Lack of ventilation probably affects them in much the same way that smothering does an animal, — by per- mitting the accumulation of the carbon dioxide gas given off by their tissues and thus reducing their supply of oxygen. The harmful effect of the carbon dioxide appears to be pretty well demonstrated by the experiments described by Dr. Shear and his associates in another part of this bulletin (page 237). This gas appears to collect in injurious quan- tities among cranberries, both in storage and shipment, because of the closeness -nith which the fruit packs together and of the size of the con- tainers used. As has been so splendidly demonstrated with apples,^ the rapidity of the life processes in fruits varies directly vnth temperature, much more carbon dioxide being given off at high than at low temperatures. While cranberries may not behave exactly as apples do, it seems to follow that low temperatures are important to cranberry keeping both in storage and shipment, for "^vith such temperatures the need of ventilation is probably less. The general problem divides itself naturally into two parts, as follows: — (a) Storage previous to Shipment — Low temperatures, because of their retarding effect on the process of respiration and on the growth of rot- producing fungi, seem most important. The storage house, therefore, probably should be constructed and managed to maintain such tempera- tures, without resorting to artificial cold storage, at as little expense as possible. This in turn, however, is likely in practice to depend largely on arrangements for free but controllable ventilation. If, as the results of the experiments described by Dr. Shear and his collaborators on page 238 seem to tend to show, a damp atmosphere does not injure the keeping of this fruit, the thorough ventilating of the storage room during the night and on cold days would be the cheapest means of obtaining low tempera- tures, and they probably should be maintained as far as possible by the use of dead-air spaces in the walls. To combine satisfactory arrangements for free but controllable ventilation and for effective heat insulation at a reasonable expense is probably, therefore, the main problem to be solved by future builders of cranberry storage houses. Artificial cold storage for cranberries has not been investigated much yet, and therefore is not con- sidered here. (6) Preparation for Shipment. — Wliile a low temperature is still prob- ably desirable for cranberries after they leave the producer, this factor, except as it may be utilized by cooling previous to shipment or by shipping in refrigerator cars, is largely out of his control. He should, therefore, « F. W. Morse, Bui. No. 135, New Hampshire Agr. Expt. Sta., 1908, and Journal of the Ameri- can Chemical Society, Vol. 30, No. 5, 1908. 218 MASS. EXPERIMENT STATION BULLETIN 180. make the most of careful handling of the fruit in packing and of proper ventilation for it while in transit and in the market. The latter seems to call especially for close grading and for the use of as small and open con- tainers as practicable. 6. The separator problem is still unsolved. Resanding. The year's experience with the plots, results with which have been dis- cussed in previous reports, is shown in Table 18. The check areas were in each case laid oat adjacent to and on opposite sides of the plot. AH the plots and checks were picked with scoops. The storage-test berries were selected by handfuls from different parts of the crates as they came from the bog and put in quart cans, each can representing one crate. The cans were stored with covers on tight but not sealed. This, the seventh year since resanding was discontinued on plots 0 and V, is the first one except 1913 in which their yield has been noticeably reduced as compared with that of the checks. Throughout the season these unsanded plots presented a marked contrast to the surrounding bog which was resanded in 1912 and 1914, their vines being comparatively very thin and sickly in appearance. REPORT OF CRANBERRY SUBSTATION FOR 1916. 219 Percent- age of Rotten and Partly Rotten Berries found at End of Storage Test. o— icocoo-*a>oocoooiocjMio<^>oaot^ SSSwocoStj.cocomcotonco'N-"-' r^t^t— 00D0CO00CSCOCOCOCO»O»'5»O'^^^'— ' 1 CN>>>>>>>>>>■>>■>>■> > >■ oooooooooooooooooo *? . Z2;^^S5 2:2;^^^^.5iZS^^^Z^?^^ ~t 5SS2S22SS222555P22 i" ^sjsssssssssssss-*^^ s ft a ftft ft ft ft ft ft ftftft ft ft g.-^ -g -g ft, ntity ced arage !St rts). g^r^H § &^_a a ^ s .> -s ■| S.|^^ «00O00Cv3-S<00C3iC0 0>00O-HO-*-*02'^ 03iraoco«=^'-.«« i °-3 &« 3 a £=« m itity ■uit ned ). ????g§?S§2£SSSg2gS§^5S §fe-3 3^ odc2c5ioc»Modooj'c22"«»o«D^OM §z^s« " Co-g- t^t^I:^t^500 0iX>00 0000 00>0>0>OTl<-*Tfl 5'S C^(MN(N > > 1 • • -1 • -1 • • "o "3 "3 .2 .S .S lOiOio' 1 So'o'o§2S2^222^S2_22 £ tr"3 "S "a c'o 'o'oo^'o'o'o^'o'oS ° ° |^^^||||3-|3|s-|l=fll 1 i i § § i^^^.i^^^^^'d^-^-c-a >(Ncrt«(ueciCa)C!Ccuc:c3 |sss|^^^|^^^|^^|^^ S'S'B'o 8222-S22S-222-S22 •S ^ §> ^-S -3 "S "o ^-3 o -o ^ o -o -^'^ o ^•r-ri-r *j^ — — ' -^ — — — ' c3ss S::3S H° ftaanP'SIa cu »13 03 ca^^a) ca cs^a* c« es S^MMM2Cfofr,fe>Htnfefi<>^fofc>, ^".ir.iir.ij'^".:ir.^"/!r.ii".;ir.a-.ir^''.ii'.ir JJJJJJJJJJJJJJJ ■ ■ U >>>,>,>.>.>i>> >>■>>>>>> >> >> >- >> 1 1 1 > WHWWHWWWWWHWWWHKMW H^ 00000. o>^o.o>o.oa>o«=oo.o>o. <|l 5 ' _ a O z '^^^ '^^^ '^.-c-c '^.-:: ,-c^ ■* ^.^.ii ^^^ .-^^T^-^ .-^.-^ .-^-^ 1 •gss %%% sss SS gg ^.S.S A^X. AX,A X.A x.Ji PL, >■> » do 0 0 :zr^ ^ z rt rt « ^"h Eh 220 MASS. EXPERIMENT STATION BULLETIN 180. Summary of Table 18. Plots and Checks. Total Area (Square Rods). When resanded. Total Quantity of Fruit picked (Bushels) Average Quantity of Fruit per Square Rod ) Average Percentage of Rotten and Partly Rotten Berries at End of Storage Test. Plots O and V, Checks O and V, . Plots N, R and T, . Checks N, R and T, 49H 27 56 Not since November, 1909, Twice since 1909, Yearly in the fall, 1911 to 1915, inclusive. Twice since 1909, 64.73 32.55 77.87 1.31 1.21 58.61 49.38 32.10 29.70 The keeping qualities of the fruit of the sanding plots and their checks were determined by storage tests each year from 1912 to 1916, inclusive. The results of these tests and their averages are given in the following table : — KEPORT OF CRANBERRY SUBSTATION FOR 1916. 221 « PQ 1 1 iJ u 1 i PL, lis III g S5!K§§SS§g s ??SSg§SS?32 1 t^ S^gSSSPilSS s §S5JS;^J^S?32 s s? SSSSSS^SS g2g52^§^S22 i s? SSSSSSSSSK g ?3?5S^2SS22 i •o g§g^2:§gj2§ si g 5gSS§gS?SS§§g d s 18.75 15.30 21.90 Not started Not started Not started S I II 912 and fall of 1914, 1909, I and fall of 1914. . the fall, 1911 to 1915, inclusive I and fall of 1914, , the fall, 1911 to 1915, inclusive the fall. 1911 to 1915, inclusive 1 and fall of 1914, . i 1 "o §2 92 92. 2 2 .s 1 i i i i i 1 i ;- 1 g 1 1 1 s 1 s < I 's o 13J^ to 24 9 18 to 25M 9 12 to 27 9 6 to 12 9 15 to 24 >> -s 1 § M >> 1 M J4 M .^ M M ^ M M M ^ M M M g o > > d o 15 ;? rt rt H H 222 MASS. EXPERIMENT STATION BULLETIN 180. Fertilfzers. The season's results with the station bog fertiUzer plots are given in Table 20. The area of each plot, as stated in the report for 1912, is 8 square rods, and the variety of berries tested is the Early Black. The plots are on a peat bog with a covering of sand ranging from 6 to 8 inches in thickness. Table 20. — Fertilizer Plots in 1916. Yield and Relative Keejnng Quality of Berries. Plot. Fertilizer used. Date treated in 1916. Date picked. Quan- tity of Berries pro- duced (Bush- els). Quan- tity of Berries in Storage Test (Quarts). Date Stored Berries were ex- amined to de- termine Rot Percent- age. Percent- age of Rotten and Partly Rotten Berries found at End of Storage Test. 1 0 - Sept. 22 10.67 8> Dec. 4 46.86 2 N, . June 24 Sept. 22 9.33 8 Dec. 4 50.21 3 P, . June 24 Sept. 22 9.00 8 Dec. 4 45.90 4 K, . June 24 Sept. 22 9.60 8 Dec. 4 53.78 5 0, . - Sept. 22 9.20 8 Dec. 4 49.61 6 NP. June 24 Sept. 22 6.33 8 Dec. 5 57.64 7 NK. June 24 Sept. 22 6.60 8 Dec. 5 56.33 8 PK, June 26 Sept. 22 8.00 8 Dec. 7 49.00 9 0, . - Sept. 22 9.00 8 Dec. 7 45.14 10 NPK. June 27 Sept. 22 6.88 8 Dec. 7 43.80 23 Peat 2, - Sept. 22 8.00 8 Dec. 9 39.39 11 NPKL, June 27 Sept. 23 2.86 8 Dec. 7 59.07 12 NPKcl. June 27 Sept. 23 6.00 8 Dec. 7 50.98 13 0, . - Sept. 23 7.67 8 Dec. 8 41.12 14 Ni.PK, June 26 Sept. 23 5.50 8 Dec. 8 55.84 153 NaPK, June 26 Sept. 23 4.52 12 Dec. 8 63.10 16 NKPn, June 26 Sept. 23 7.20 8 Dec. 8 55.81 17 0, . - Sept. 23 9.33 8 Dec. 8 39.87 18 NKPo, June 26 Sept. 23 8.33 8 Dec. 8 47.36 19 NPKi,. June 26 Sept. 23 7.75 8 Dec. 8 53.08 20 NPKo," June 26 Sept. 23 9.00 8 Dec. 8 59.94 21 0, . - Sept. 23 10.33 8 Dec. 8 49.63 1 The storage-test berries from each plot were stored, without being run through a separator or otherwise cleaned, in quart cans on the day they were picked, each can being filled with handfuls of fruit taken from different parts of a separate picking crate, its contents thus rep- resenting as fairly as possible the contents of the crate as it came from the bog. The covers of the cans fitted tightly during the storage, but were not sealed. 2 Leaf mold worked into a condition in which it could be spread easily with a shovel. 3 The figures for plot 15 arc probably misleading, as half of that plot was used in spraying tests with Bordeaux mixture in 1913, 1914 and 1915, and certain effects of that treatment may have remained in 1916; though, if the whole plot had yielded at the same rate as did the portion that never had been sprayed, it would have produced only 5.33 bushels. The rot percentage given for this plot is an average of the percentages obtained in the tests of the fruit of the sprayed and the unsprayed parts. REPORT OF CRANBERRY SUBSTATION FOR 1916. 223 Plots 1, 5, 9, 13, 17 and 21 are all untreated checks. The meanings of the symbols used in the table are as follows: — 0 = Nothing. N = 100 pounds nitrate of soda per acre. P =400 pounds acid phosphate per acre. K =200 pounds high-grade sulfate of potash per acre. L = 1 ton of (slaked) lime per acre. Kcl =200 pounds muriate of potash per acre. Nij =150 pounds nitrate of soda per acre. Nj =200 pounds nitrate of soda per acre. Pi;^ =600 pounds acid phosphate per acre. P2=800 pounds acid phosphate per acre. In combination they mean, for example, as follows: N2PK = 200 pounds of nitrate of soda + 400 pounds of acid phosphate + 200 pounds of high-grade sulfate of potash per acre. As the table shows, the fruit of the fertilized areas this season was, as a rule, much inferior in both quantity and keeping quality to that of the checks, this being especially marked with the plots treated with lime and with the maximum amount of nitrate of soda. Considering all the expe- rience with these plots since they were started in 1911, it is the writer's judgment that, in general, whatever slight advantage in yield has been gained by the use of the fertilizers has been balanced by the cost of the treatment, the deterioration in the quality of the fruit and the greater cost of picking due to the increased vine growth. Insects. The Cranberry Rootworm (Rhabdopterus picipes (Oliv.)). The rearing of the beetles definitely identified the mfestation by the cranberry rootworm {Rhabdopterus picipes (Oliv.)) tentatively recorded in last year's report (pages 32 and 33). By the beginning of winter the grubs of this insect nearly complete their growth. They are then, except the head, for the most part nearly white in color and somewhat over a quarter of an inch long. They hibernate without growing larger. They do some feeding in the spring and change into pupa? in June. No beetles of the infestation under observation had yet emerged on June 30, this season, a collection of the insects taken that day consisting of 4 grubs and 32 pupse. One beetle was found on July 1, and during the following two weeks they practically all came out, the period of most rapid emergence extending from the 3d to the 11th of the month. It was anticipated that the adults might feed freely on the cranberry foliage, and at the WTiter's suggestion an arsenical spray was applied to the infested area on July 3 and repeated on the 11th and 18th. In the first two applications, 2j pounds of "Corona" arsenate of lead and 1 heaping teaspoonful of white arsenic to 40 gallons of water were used. For the last treatment the mixture was the same, except that the arsenic was increased 224 MASS. EXPERIMENT STATION BULLETIN 180. to IJ teaspoonfuls to 40 gallons. The writer suggested only the arsenate of lead, fearing arsenic would do harm. The latter was added by the fore- man of the bog to do a thorough job, and fortunately no injury resulted. The writer visited the bog on July 20 and found dead rootworm beetles in large numbers under the vines, most of them being in a dry and brittle condition. Only a very few were crawling about. The cranberry foliage on the infested area showed that the beetles had fed freely upon it. As 6 of 15 beetles, collected July 11 and kept at the station screenhouse, were still active on the 26th, the condition of those found on the bog on the 20th seemed to indicate that the spraying had been effective. This bog was kept under observation until the end of the season, and no evidence of the continued presence of the pest was discovered, it having been practically exterminated by the treatment. Prof. H. B. Scanamell has published a valuable bulletin on this insect.^ The Gypsy Moth (Porthetria dispar L.). Several quarts of egg masses were collected from trees late in December, 1915, and early in January, 1916, and divided into lots of about a half quart each, two of these being put in cans with moist sand in the bottom and placed in the basement of the station screenhouse for checks, the others being enclosed in cloth netting sacks and submerged for the winter in 3 feet of water in a pond. The eggs of the check lots hatched almost perfectly. The dates on which the various submerged lots were taken from the water, and the WTiter's estimates of the percentages of eggs that hatched, were as follows: lot 1, April 2, 25 per cent.; lot 2, April 18, 20 per cent.; lot 3, April 23, IS per cent.; lot 4, May 1, 25 per cent.; lot 5, May 5, 20 per cent.; lot 6, May 13, 20 per cent.; lot 7, May 24, 5 per cent. The submergence did not seem to kill the eggs as readily in these tests as in those reported last year. This may have been due to the unseasonable coldness of the spring this season, which probably caused the water in the pond to warm up more slowly than usual. On May 29, 59 gypsy-moth caterpillars from one-eighth to five-six- teenths of an inch long were submerged on the leaves of an oak branch just as they were taken from the woods, in 8 inches of water in a washtub. All but 3 of the worms clung to the branch and went down into the water with it. At the end of a forty-three-hour submergence, 8 floated on the surface, 4 had sunk to the bottom of the tub, and 47 still clung to the leaves. These worms were watched for two days after the close of the test, but only 1 of the 59 showed any sign of life. On May 31, 50 caterpillars from one-quarter to five-sixteenths of an inch long were submerged, as before, on the leaves of an oak branch in 9 inches of water. All these worms clung to the leaves tenaciously when submerged. After twenty-two hours in the water, 2 floated on the surface, > The Cranberry Rootworm, Bui. No. 263, U. S. Dept. Agr., 1915. REPORT OF CRANBERRY SUBSTATION FOR 1916. 225 3 had sunk to the bottom, and 45 still clung to the leaves. They were then taken from the water, and within seven hours 26 had nearly or en- tirely recovered. On June 1, 152 worms from one-quarter to three-eighths of an inch long were submerged on the leaves of an oak branch, as before, in 9 inches of water. After thirty-eight and one-half hours of submergence, 46 floated on the water, most of them being alive and active, 40 had sunk to the bottom, and 66 still clung to the leaves. Those clinging to the branch were then taken from the water and watched, and only a few ever showed any sign of recovery. As a rule, the worms that came to the surface of the water were among the largest of those submerged, as was also the case in later tests, descriptions of which are not included here. The results of these experiments and of observations of bog flooding operations, in which the small gypsy caterpillars behaved similarly, have led the writer to the following conclusions : — 1. That reflowing for this insect will be most satisfactory if done while the worms are small and probably before the largest are more than five- sixteenths of an inch long. The sooner it is done after the eggs are all hatched the less will be the damage fr.om the feeding of the worms and the less the trouble from their floating ashore alive, as it is evidently the habit of the very young caterpillars to cling to their support when sub- merged. 2. To be entirely effective, even when the worms are small, a flowage must probably be held nearly forty hours. Mr. C. W. Minott of the Bureau of Entomology of the United States Department of Agriculture conducted some interesting investigations during May and June, 1916, concerning the wind-spread of gypsy-moth caterpillars on cranberry bogs. With his permission the following con- densed account of these studies is given here : — Two bogs in Carver, Mass., were selected for experiments on wind dispersion, namely, Muddy Pond bog, containing about 100 acres, and John's Pond bog, con- taining about 44 acres (including pond). Six screens made of cotton cloth tacked to a frame in two sections, each being 3 by 10 feet, were set up horizontally just above the tops of the vines at various distances from the neighboring woodlands. Each screen contained 60 square feet of cloth upon which " tanglefoot " was applied. Daily examinations of each screen were made and data were taken concerning the temperature and the direction and velocity of the wind during the dispersion period. The screens were located on the bogs at various distances, ranging from 400 to 1,200 feet, from woodland infestations. From one screen, located 600 feet from infested woodland on the northwest and 900 feet on the west, 62 small caterpillars were removed during the season, or shghtly more than 1 to the square foot. A total of 143 small worms was wind-borne on to the six screens, which indicated that an average of about 17,000 per acre blew on to the bogs. The infestations around these bogs are as yet only medium in extent, this showing what may be expected when the surroundings of bogs become thickly infested. ' 1 Collins, C. W.: Methods used in determinihg Wind Dispersion of the Gipsy Moth and Some Other Insects, Journal of Economic Entomology, Vol. 10, p. 174, 1917. 226 MASS. EXPERIMENT STATION BULLETIN ISO. The Cranbemj Tip Worm {Dasyneura vaccinii Smiths). The season's observations of the effect of resanding on the abundaDce of this pest sustained the conchisions heretofore reported. One species of Chalcidicl {Tetrastichus sp. ^) and two of Proctotrypid {A'phanogmtis sp. ^ and Ceraphron sp. ') parasites were reared from the larva; of the last brood after they had encased themselves in their cocoons this season. Two of these {Tetrastichus sp. and Aphanogmus sp.) emerged in only small numbers, but the Ceraphron species had infested a large, though undetermined, majority of the maggots collected by the writer, and its adults kept coming out from August 9 to September 14, inclusive, their period of most rapid emergence being from August 12 to August 22. The eggs of the tip worm are not "white" as they have been described.* They are watery translucent in appearance, with scattered pinkish pig- ment, and are about one-third of a millimeter long. They are elongate, usually slightly curved from end to end, with rounded and slightly nar- rowed ends and without noticeable surface markings. The Black-Head Fireworm {Rhopobota vacciniana (Pack.)). Prof. H. B. Scammell, in cranberry insect investigations in New Jersey for the Bureau of Entomology, had much success last year in treating both broods of this insect in the worm stage with a form of nicotine sulfate known as "Black-Leaf 40." He used 1 part of this insecticide to 400 parts water, and added resin fish-oil soap at the rate of 2 pounds to 50 gallons to make the spray spread and stick. When the writer saw the plots Professor Scammell had treated in this way, they were green and had a fair amount of fruit, whereas the surrounding bog, and even plots sprayed with arsenate of lead, had been turned brown by the insect and bore prac- tically no crop. The writer tried this treatment against the first brood on two large plots this season, and while it failed to control the insect entirely, it checked it so much that the plots remained green while the surrounding bog was turned rather brown, the contrast being striking. This insecticide must be tested further before it can be said at what strength it should be used or how many times it should be applied to either brood. At the strength in which it has so far been tested it is a rather expensive treatment, costing about $7 per acre per application. It may be found, however, that weaker mixtures suffice. At any rate, this treatment stands at present as the only really effective method of controlling the first brood of this insect, burning and flooding excepted, and in spite of its expense it will, therefore, find favor in the management of many bogs. Two, and perhaps three, applications for the first brood are advisable. > Bui. No. 175 of the New York State Museum, p. 151. 2 Determined by Mr. A. A. Girault of the Bureau of Entomology. 3 Determined by Mr. J. C. Crawford of the Bureau of Entomology. * Smith, J. B.: Insects Injurious in Cranberry Culture, Farmers' Bulletin No. 178, U. S. Dept. Agr., 1903, p. 19. REPORT OF CRANBERRY SUBSTATION FOR 1916. 227 As a treatment for the second brood, it may have to compete with arsenate of lead, for there is danger of injuring tender foliage, and especially blos- soms, in spraying with any contact insecticide, and arsenate of lead is far more effective with the second brood than with the fii'st. Proper treatment of the first brood with "Black-Leaf 40" may check the pest so well that a thorough treatment of the second brood will not be so necessary as it is at present. In anj^ case, not more than one application of "Black- Leaf 40" for the second brood is likely to be desirable. The writer gave some cranberry uprights sprayed with "Black-Leaf 40" to some gypsy-moth caterpillars, providing another lot with unspraj^ed vines as a check. The latter were eaten much more freely than the former. This suggests that the effectiveness of this insecticide may be partly due to a deterrent property. The second brood of the fireworm did less damage than usual this sea- son, and less than might have been expected from the abundance of the first brood. The wet season seemed to check it strongly somehow. The Cranberry Fruit Worm {Mineola vaccinii (Riley)). This insect did the least injury this season of any year in the wTiter's experience. ' It has not been less prevalent since 1903. We have no relia- ble information concerning its abundance in years previous to 1904. The writer has tried to determine, as far as possible, the relative abun- dance of this pest in the various cranberry-growing regions. It is most harmful on Cape Cod and in Wisconsin, being far less troublesome in New Jersey, the amount of injury on dry bogs (without -winter-flowage) in the latter section, when the writer was there in 1915, being about the same as that on the flowed bogs of the Cape in the same season. It does about the same damage on Long Island and Nantucket as in New Jersey, being far less prevalent there than on Cape Cod. It appears to be almost if not entirely, unknown on the Pacific coast of Oregon and Washington. It will be seen that this insect is not usually very troublesome except in the regions with comparatively cold and dry climates, a heavier total precipi- tation as well as a higher average temperature being characteristic of the warmer sections. One might expect from this that any variation in the Cape Cod climate toward that of the warmer regions would be likely to tend to reduce the pest, whereas any variation in the opposite direction woidd be likely to tend to make it more abundant. Cape Cod Data appear to strongly substantiate this Conclusion. — The season of 1905 was the worst on record for fruit-worm injury. The Cape had a lower mean temperature in 1904 than in any subsequent year up to the present time, and in 1905 had a smaller total precipitation than in any year since, in spite of the fact that the rainfall in all the last five months of the j^ear except October was heavy. Of the severity of the winters 1903-€4 and 1904-05, the Annual Summary of the New England Section of the Climate and Crop Service of the Weather Bureau for 1905 (page 3) remarks as follows: — 228 MASS. EXPERIMENT STATION BULLETIN 180. February — the last of the winter months, with its remarkably low temperature record — completes one of the coldest winters of official record. At Boston the mean temperature for the three months, December, January and February, 1904-05, 24.8 degrees, is the lowest for the winter months since 1871, excepting 24.4 degrees in 1903-04, and 24.5 degrees in 1873-74. The winter for New England, as a whole, was the coldest since the estabUshment of the weather service of this section in 1884. The mean temperature was 17.9 degrees, and the next lowest is 18 degrees for the winter, 1903-04. As far as the writer can determine, the greatest reductions in fruit-worm activity in recent years, aside from that of this season, occurred in 1906 and 1913. The records of the Weather Bureau show that the total pre- cipitation of 1906 on the Cape was the greatest of any year since 1904, May, June and July being especially wet months. The winter of 1905-06 was mostly an open one. Both temperature and precipitation ran abnor- mally high throughout the greater part of the period beginning with October, 1912, and ending May 1, 1913, the winter being very open. As affecting the abundance of the pest in 1916, it should be noted that September, 1915, was a month of record high temperatures for its season, that the winter 1915-16 was mostly very open, and that the first half of this gro\ving season was very wet throughout. In the latter part of May the writer covered large numbers of fruit worms in their cocoons, in quart cans partly filled with moist sand, with different measured and uniform depths of sand ranging from three-six- teenths of an inch to a full inch, and made records of the subsequent emergence of the adult insects. Unfortunately, no check of worms not covered with any sand was kept for comparison, but, judging from the freedom with which the parasites and moths emerged through three- six- teenths, one-fourth, three-eighths, one-half, five-eighths, two-thirds and even three-fourths inch depths, it appears that resanding as commonly done does not much affect the abundance of either the fruit worm or its worm parasites. The full inch covering of sand seemed to smother most of the moths and parasites, though a few of both came out even from that depth. The WT-iter liberated a number of apparently female moths from a boat on a pond on July 25, and three of them were seen to fly to the shore, a measured distance of about 272 feet, in a single flight, a toy balloon being anchored in the pond at their point of departure to measure from, and the measuring being done with twine. This demonstration of this insect's powers of flight is of interest in connection with the speculation concerning the annual infestation of bogs from surrounding uplands and from neigh- boring bogs. Fruit-worm eggs showed a range in Chalcidid {Trichogramma minuta) parasitism of from about 25 to 75 per cent, on dry bogs and from none to about 75 per cent, on those with winter- flowage this year. This parasite was not found at all on half the flowed bogs examined, more than a quarter of the eggs showing its presence on only 3 out of 30 such bogs. It appeared REPORT OF CRANBERRY SUBSTATION FOR 1916. 229 to be entirely absent on some flowed bogs on which it infested from 76 to 89 per cent, of the eggs in 1915. Its great reduction on the flowed bogs may have been due to the long period of wet weather in the first half of the growing season. The Braconid (Phanerotoma franklini ^ Gahan) parasitism was found to range from 24 to about 55 per cent, on dry bogs (without winter-flowage) and from none to about 33 per cent, on flowed ones. On one bog which had the \nnter-flowage held until May 25, 24 per cent, of the fruit worms were infested with this parasite, and on another, bared of the winter water on May 14, 21 per cent, were infested, these figures indicating that moder- ately late holding of the flowage perhaps does not reduce this parasite in proportion to its host as seriously as was suggested by the writer in last year's report (page 40.) It should be stated in this connection that the percentages of Phanerotoma and Pristomeridia parasitism given in this and previous reports only show the amounts of these parasitisms among the worms at work in the berries when the examinations were made, and indicate the parasitism of the entire season only in a very rough way. It was discovered this year that the parasitized worms leave the berries somewhat sooner than the unparasitized ones, examinations made toward the end of the pest's period of activity showing greatly reduced percentages for the worm parasitism as compared with those made earlier. Worms from the same location on one bog showed percentages of Phanerotoma parasitism on different dates, as follows: September 3, 33.3 per cent.; September 6, 40 per cent.; September 13, 2.3 per cent. The percentages of Pristomeridia parasitism found in this same location were as follows: September 3, 5.5 per cent.; September 6, 6.6 per cent.; September 13, 0. Pristomeridia agilis - was very scarce this year, the percentage of its parasitism being found to range from none to 5^ on flowed bogs and from 4J to about 10 on strictly dry ones. The examinations by which the percentages of Phanerotoma and Pristo- meridia parasitism given in this and previous reports w^ere determined were made by crushing fruit worms between glass slides in such a way as to expel their viscera through the anal opening, the parasite larva, when present, apparently always being ejected with them and being found easily with a good hand lens. A number of eggs deposited at the same time by Phanerotoma females under observation in eggs laid by fruit-worm moths in confinement where they were secluded from parasites, and subsequently kept in closed bottles, were e.xamined with a microscope successively at various times after depo- sition. None of these parasite eggs examined after either thirty-six or forty-two hours showed any sign of hatching. Two of three examined at the end of forty-six hours had hatched, but the larvee showed no sign of life. After forty-nine hours all the eggs had hatched, and some of the 1 This parasite, called Phanerotoma tibialis in the writer's previous reports, has recently been described as new to science, and given the name here used, by Mr. A. B. Gahan of the Bureau of Entomology. Cf. Proc. U. S. Nat. Mus., Vol. 53, 1917, p. 200. ' The exact identity of the species is still in doubt. 230 MASS. EXPERIMENT STATION BULLETIN 180. larvae moved their mouth parts considerabl3^ The weather was cool during the entire period (Julj^ 29 and 30) in which this investigation was in progress, the maximum temperature in the sun at the station bog being 80° F. and the minimum bog temperature being 40°. Cocoons of parasitized fruit worms are usually much smaller and more delicate than those of unparasitized ones. Submergence tests were conducted with fruit worms in their cocoons, as follows : — 1. Six small cheesecloth sacks, each containing 20 cocoons, were sub- merged to a depth of 2 feet in a pond at 10.30 a.m., September 14. They were all taken from the water and examined in the afternoon of September 26, and all the worms were found dead, a majority of them being partly decomposed. Most of them had left their cocoons and were on the inside of the sacks. 2. Three lots of cocoons of 20 each were submerged in cheesecloth sacks to a depth of 2 feet in a pond at 9 a.m., September 30. These were all taken from the water and examined between 11 a.m. and 1 p.m., October 12. All the worms were found dead, most of them being more or less decomposed. About half had left their cocoons and were clinging to the inside of the sacks. 3. Two cheesecloth sacks, each containing 20 cocoons, were submerged in 2 feet of water in a pond at 3 p.m., October 12. These sacks were taken out and examined at 5 p.m., October 24. Most of the worms were found dead and more or less decomposed, as in the previous tests, but 7 were alive in one sack and 2 in the other. 4. Two cheesecloth sacks, each containing 20 cocoons, were submerged to a depth of 2 feet in a pond at 8 a.m., October 25. They were taken out and examined on November 6,^17 being found alive in one sack and 8 in the other. In all these tests the sacks were of the same material, were tied up and submerged in the same way, to the same depth in the same place and for practically the same length of time. It will be seen that as the season advanced the submergence had much less effect on the worms. As the pond grew colder fast while these tests were in progress their results sug- gested that the temperature of the water largely determined its effect. At 1 P.M., Jan. 3, 1917, a weighted cheesecloth sack, containhig 15 fruit worms in their cocoons, was placed in the bottom of each of two 1-quart cans full of water, the water being at a temperature of 59^° F., and the cans, with their covers on tight, were placed in a chicken incubator to- gether with Green maximum and minimum registering thermometers, the incubator being set to run at a temperature of 60° F. As a check on these cans, two similar cans containing similar lots of fruit worms were placed in a pail of water at the same time, the temperature of the water in the cans and in the pail around them being about 35° F. The pail, together with maximum and minimum registering thermometers, was placed in a barrel the temperature of the air in which was about 37° F. The barrel was headed up and buried in hay to keep its contents at an even tempera- REPORT OF CRANBERRY SUBSTATION FOR 1916. 231 ture. The cocoons in both the incubator and the barrel were taken from the water at 9 p.m., January 15, and were examined the next day in a warm room. All but 9 of the 30 worms that had been in the incubator were dead, whereas all but 3 of the 30 from the pail were alive. Those taken from the pail were as a rule very lively after they got warmed up, most of them crawling actively. On the other hand, none of those from the incubator became active, the live ones showing they were so only when prodded con- siderably, their movements even then being very sluggish. None of the dead worms had begun to decompose. The temperature of the incubator was shown by the thermometers to have ranged from 52° to 66° F. during the test. The temperature of the water in the cans kept in it was 57° F. at the end of the test, and had probably averaged a little under 60°. The temperature in the barrel had ranged from 31° to 39^° F., that of the water in the pail being 35° at the end of the test. This incubator and pail experiment was duplicated by a test carried out similarly in all details, except that vaseline bottles of 3^-ounce capacity, with tightly inserted cork stoppers, were used instead of the cans, the cocoons being submerged at noon, Jan. 29, 1917, and being taken from the water at 3 p.m., February 13. Of the 30 worms kept in the incubator 16 were dead and 14 alive at the end of the test, while of the 30 tested in* the pail 27 were alive and only 3 dead. Moreover, the live worms frorrk the bottles in the pail were much more active after they got warmed up- than were those from the incubator. None of the dead worms had begun to decompose noticeably. In this test the temperature in the barrel ranged from 32° to 36° F. The incubator got out of order twice, — on the seventh and tenth days of the test, — its temperature the first time falling to 40°' and the second to 33° F. With these exceptions it ran between 52° andl 62°, and probably averaged about 56°. Many of the cocoons used in these tests were carefully opened under water at the end of the submergence, and, while they were all found to be largely filled with water, none were without a little air or gas, this indi- cating that the findings in this regard previously reported by the writer " were not quite accurate, the former examinations apparently not having been sufficiently careful. The results of these experiments seem to prove that the effect of sub- mergence of the worms in their cocoons depends largely, if not principally, upon the temperature of the water, and they suggest that a flowage after picking, if it is begun before October 1 and continued for twelve or possibly even ten days, may control this insect as well as late holding of the winter- fiowage usually does. It may be said that such a flooding would interfere with harvesting, but as late picking is usually a result of late holding of the previous winter-flowage, and as late holding is most commonly practiced as a treatment for the fruit worm, this objection does not seem valid. Flooding practiced annually after picking would probably have a much less harmful effect on a bog than late holding of the winter-flowage every year has. » Bui. No. 160, Mass. Agr. Expt. Sta., 1915, p. 113. 232 MASS. EXPERIMENT STATION BULLETIN 180. Bog Management. Prof. H. B. Scammell has recently reported ^ a destructive visitation of 1;he fall army worm (Laphjgma fricgiperda S. & A.) this year on widely :separated cranberry bogs in New Jersey following closely, and evidently somehow caused by, the removal of the winter-flowage in mid-July. This insect feeds on a variety of plants, but has not heretofore been known as a ■cranberry pest. As its frequent outbreaks, which start in the southern ^States, sometimes reach as far north as Canada, by the spreading of the successive broods of strong-fljdng moths, in a single season, though it is unable to endure the winter in the north, there is ground for fearing that midsummer removal of the winter-flowage may more or less regularly invite serious trouble from this insect on Cape Cod as well as in New Jer- sey. This unexpected development must be regarded as a possible com- plication in connection with certain phases of the biennial cropping system suggested by the writer in last year's report (page 46). Late holding of a deep winter-flowage is sometimes dangerous. This flowage was started off from a bog in Assonet, Mass., on June 10, its with- drawal being completed on the 11th. When the writer visited this bog on June 30 the vines seemed completely dead where the flowage had been deepest (5 feet deep), whereas they showed no injury, aside from the re- tarded seasonal development of growth, where the water had been shal- lowest (2 feet deep), their leaves having been well retained and appearing green and healthy. "Where the water had been deepest the leaves were all off, the buds at the tips of the uprights were gone, and the vines were brittle and showed no green in the break when broken off. There was a complete gradation from this condition to that where the flowage had been shallowest, corresponding with the variation in elevation. Part of the vines on this bog were set out in the spring of 1914, and part in the spring of 1915, strips of both plantings running from the lowest to the highest parts of the bog. The WTiter is informed by the manager that the one-year sets where the flowage was deep finally recovered somewhat, but that the two-year plantings were killed entirely. A large bog in Rochester, Mass., the winter-flowage of which ranged in depth from 4 feet to nothing, had this flowage held until May 31 this sea- son. This is an old bog, with vines well established. Where the water was deepest the leaves all came off, leaving the uprights alive but bearing only the terminal bud. On the other hand, there was no abnormal falling of the leaves where the water was shallow. As on the Assonet bog, there was a complete gradation in the injury corresponding with the variation in the depth of the flowage. A new 60-acre bog at Assonet, Mass., was flowed on the night of May 31, the vines being completely submerged for forty-eight hours, the water ranging from 3 feet to a few inches in depth, and averaging about 2^ feet. » Proc. 47th Ann. Meet, of the Amer. Cranb. Grow. Assoc, p. 11, January, 1917. REPORT OF CRANBERRY SUBSTATION FOR 1916. 233 The flooding and draining were done entirely at night. A few days later the writer's attention was called to an injury that had resulted. He visited the bog and found the buds and even the tops of the new growth of the uprights on parts of it seriously hurt. The injury was mainly on the central portion of the bog, and centered around a large pile of ashes left from the burning of stumps and brush when it was built. Vines at con- siderable distances from this pile showed at most but slight injury, except in a streak parallel to the end of the dike toward which the wind had blown during the flooding. Leaves of bushes which had hung down into or stood in the water of the reflow, around the margin of the bog, showed a marked and unusual burning injury, and they bore traces of a white powder which appeared to be ash that had floated in the water from the pile at the cen- ter of the bog. The situation as a whole led all those who observed it to conclude that the ash pile had caused the trouble. The pile was estimated to be 2^ feet deep over an area 25 feet square and about 6 inches deep over another area 75 feet square. Piles of ashes on bogs are probably danger- ous because of the lye leached from them. Many unaccountable spots where vines refuse to grow thriftily on bogs may be the result of effects remaining from ashes left from the burning of brush piles . It is well known that alkalies in the soil are inimical to cranberry growth. A portable sectional bridge devised by the writer for use in carting ber- ries across bog ditches proved valuable at the station bog this year. With its help it was easy to cart berries without killing the vines in tracks by repeated passages of the wheels over the same ground. A light truck probably could be used to great advantage with this bridge, though the writer has tried only a horse and wagon with it so far. At any rate, it will make it possible to much reduce the present expense of removing berries from bogs. It may be seen at the station bog at any time during the cran- berry season. With many Cape Cod bogs a desirable reduction in the cost of resanding could probably be effected by the development of a sanding rim around the margin, t With such a rim the sand for any part of the bog could always be brought from the nearest point. The rim should be wide enough for a good roadway, and it should be built level with the bog surface, so that it may serve as a sanitary catch-basin for floating berries and leaves. If, as the results of some of the writer's storage experiments seem to indicate, the berries from the marginal portion of a bog, other conditions being the same, are usually of poorer keeping quality than those from the center, the condition may naturally be laid to the continual deposition of diseased cranberry material floating on the surface of repeated flowages and wafted to the margin by the wind. Thus the possible value of a marginal catch- basin as suggested becomes evident. The sanding rim would also have some value as fire protection for a bog. As the sanding rim becomes suflBciently widened by the removal of sand in repeated resandings, the bog can be gradually enlarged by planting 234 MASS. EXPEEIMENT STATION BULLETIN 180. on the inner side of the rim, this increase in property being mostly clear gain. The sanding rim can be constructed most advantageously when a bog is built. Its development after the bog is planted is attended with some difficulties. Among these the extra cost of turfing the upland adja- cent to the bog, and the liability in resanding of seeding the bog more or less with certain troublesome weeds, should be especially considered. OBSERVATIONS ON THE SPOILAGE OF CRAN- BERRIES DUE TO LACK OF PROPER VENTILATION. BY C. L. SHEAR AND NEIL E. STEVENS, PATHOLOGISTS, AND B. A. RUDOLPH, SCIENTIFIC ASSISTANT, FRUIT-DISEASE INVESTIGATIONS, BUREAU OF PLANT INDUSTRY. Introduction. The injury to cranberries due to keeping them in tightly closed packages was brought strikingly to the writers' attention during temperature tests conducted in the fall of 1916. Uniform samples of Early Black cranberries from bogs near Wareham, Mass., were put up in pound coffee cans and sent to Washington by mail. There they were placed in the constant tempera- ture apparatus used by Drs. Brooks and Cooley of this office, and described by them in their recent paper. ^ One can from each lot was placed at each of the following temperatures: Centigrade, 0, 5, 10, 15 and 20 degrees (equal to 32, 41, 50, 59 and 68 de- grees Fahrenheit). They were kept at these temperatures from early in September until about the middle of November. When the berries were removed from the cans and sorted, it was found that spoilage at the lower temperatures had been much greater than the previous experience of the writers had led them to beUeve could be due to fungi alone. Many of the spoiled berries had a peculiar lusterless appearance, and were of a uniform dull red color differing both from normal and from typical rotten berries. Among various factors considered as possible causes of this condition the excessive accumulation of carbon dioxide seemed the most probable. The work of F. W. Morse,^ Gore ' and others has proven that large amounts of this gas are given off in the respiration of various fruits, while the studies of Fulton * indicate that the spoiling of strawberries and raspberries which he noted in tight packages is due to the accumulation of carbon dioxide. Fulton found that if strawberries were kept in tightly closed bottles for » Brooks, Charles, and Cooley, J. S.: Temperature Relations of Apple-rot Fungi. Journal of Agricultural Research, 8, 13a«-163, 1917. ' Morse, Fred W.: Effect of Temperature on the Respiration of Apples. Jour. Amer. Chem. Soc, 30, 876-881, 1908. » Gore, H. C: Studies on Fruit Respiration, . U. S. Dept. Agr., Bur. of Chem., Bui. No. 142, 1911. * Fulton, S. H.: The Cold Storage of Small Fruits, U. S. Dept. Agr., Bur. of Plant Indus., Bui. No. 108, 1907. 236 MASS. EXPERIMENT STATION BULLETIN 180. three days the oxygen of the air was practically exhausted, and more than 35 per cent, by volume of carbon dioxide had accumulated. Under these conditions, as well as in cartons tightly wrapped, "The fruit softened and had the characteristic bad flavor of fruit confined in an atmosphere of carbon dioxide" (3, p. 22), Dr. Charles Brooks and Dr. E. M. Harvey of this office, who have sepa- rately studied storage conditions in apples and other fruits, examined the cranberries referred to and were of the opinion that the condition might very likely be due to the accumulation of an excessive amount of carbon dioxide. Although it was then too late in the season (November 20) to undertake a thorough investigation of the subject, preliminary tests were made which gave results of considerable interest. Temperature Tests in Open and Closed Cans. In order to compare directly the keeping of cranberries in open and closed cans, uniform lots of sound berries were divided, one portion being placed in tightly closed cans, and the other portion in similar cans with the covers removed. The result of one of these tests, which is typical of sev- eral, is given in the following tables: — Temperature Tests on Howes from State Bog, Massachusetts, beginning November 21, ending December 16. Closed Cans. Tempebatubb in Degrees C. Sound. SpoDed. Spoiled (Per Cent.). 20 328 172 34.5 16 357 147 29.5 10 444 67 13.0 6 472 29 5.5 0 483 20 4.0 Open Cans. 20 291 69 19.0 15 333 61 15.5 10 333 29 8.0 6 341 18 5.0 0 340 8 2.5 It will be noted that in all cases the amount of spoilage is greater in the closed cans than in the open cans. REPORT OF CRANBERRY SUBSTATION FOR 1916. 237 Effect of Carbon Dioxide on Cranberries. Several series of tests were made in which cranberries from various sources (Early Blacks and Howes from Massachusetts, and Howes from New Jersey) were kept for short periods in an atmosphere of nearly pure carbon dioxide. It was noticed in each case that at the end of three days practically all the berries in the carbon dioxide were spoiled, whereas ber- ries from the same lots kept in similar containers with air showed very little rot even at the end of two weeks. The berries which had been kept in an atmosphere of carbon dioxide had the peculiar uniform dull, lusterless, red color which had been noticed in many of the berries which had spoiled in closed cans. On sectioning these berries it was found that the tissue of the berry, which is white in a normal berry, had taken on the same uniform red color. Berries which have been treated in this manner have a peculiar, bitter taste, which is very characteristic. They are no longer firm, as in the sound fruit, nor elastic to the touch as in rotten fruit, but have become flaccid. The same effect on the berries was readily produced by sealing up a quantity in an air- tight container, and allowing them to remain at room temperature for a week. That this injurious effect is produced by the accumulation of carbon dioxide is indicated by preUminary tests made in December, 1916. Equal quantities of sound Early Blacks or Howes were put in similar containers (Hempel desiccators). One of these desiccators was filled with carbon dioxide, the other two contained air, but the upper portion of one of them was filled with a saturated solution of potassium hydroxide, which would absorb the carbon dioxide almost as fast as given off by the berries. The berries in the first lot were thus exposed to an atmosphere of carbon dioxide throughout the test; those in the second lot were exposed to air containing practically no carbon dioxide; and those in the third to an at- mosphere in which the carbon dioxide given off in respiration was allowed to accumulate. The results of one of these tests which was typical of all are given in the following table : — Condition op Berries at End op Test. •WEBB KEPT. Sound. Spoiled. Spoiled (Per Cent.). CO2 Air exposed to water Air exposed to KOH solution, 35 56 45 34 39 19 50 40 29 It will be noted that the amount of spoilage, including rot due to fungi, is greatest in the berries exposed to carbon dioxide and least in the con- tainer from which this gas was removed, which apparently indicates that a large portion of the spoilage was due to the carbon dioxide. 238 MASS. EXPERIMENT STATION BULLETIN 180. Effect of Different relative Humidities on Spoilage due to Carbon Dioxide. Most of the tests described above had been made in atmospheres having relatively high moisture content. In order to determine whether the hu- midity of the air in any way influenced the spoUage, a series of tests was run in which sound cranberries of the Howes variety weie kept in tightly sealed Hempel desiccators which were maintained at constant humidity by sulfuric acid solutions of different densities. This method has been described by one of the writers in an earlier paper. "• All these tests were made at a temperature of about 24° C. Chambers having relative humidities of 100 per cent, (saturated atmos- phere), 75 per cent., 50 per cent., 25 per cent, and approximately 0 per cent, were used, and so far as could be detected by careful observation there was no difference in the rate of spoilage at the different humidities. Relation of Fungi to Spoilage dub to Carbon Dioxide. It is of course possible that one effect of accumulation of carbon diox ) at least in small amounts, may be to make the berries more susceptible to the attacks of fungi. It seems certain, however, that the injury to the fruit is in many cases wholly independent of the action of fungi. On March 13, 1917, we received from Dr. FrankUn a box of Pride cran- berries taken from a crate of fruit which had been kept in storage in the basement of the screenhouse at the State experimental bog at East Ware- ham. These were taken to represent the average condition of the spoiled fruit at the time. This lot contained 271 berries. They were carefully sorted, and 195 were somewhat softened and flaccid, having much less resil- iency than the rotten fruit, in which the tissues are more or less destroyed by the growth of fungi. They had the same general appearance as berries treated with carbon dioxide, and their condition was believed to be due to the time and manner in which they had been kept rather than to fungous disease. Fifty of these berries were taken at random and cultures made by transplanting the bulk of the pulp from the cranberries, the skin being re- moved. Of these cultures, but 2, or 4 per cent., produced fungi. Assum- ing that this represents the average number afl'ected with fungous disease, deducting 4 per cent, from the total, 195, would leave 187 presumably free from fungous disease. Cultures were also made from the tissue of the remaining 76, which had more the appearance and character of fruit at- tacked by fungi. The results of these cultures showed, however, that 49 of these berries were apparently destroyed by some other cause than fun- gous disease, thus maldng a total of 236 out of 271, or 87 per cent., not destroyed by fungi but presumably by the period and conditions of storage since picking. » Stevens, Neil E.: A Method for studying the Humidity Relations of Fungi in Culture. Phytopathology, 6, 428-432, 1916. REPORT OF CRANBERRY SUBSTATION FOR 1916. 239 From a sample of cranberries of the cherry variety taken July 2, 1917, at Madrid, Me., which had been kept in the cellar of a house all winter, 50 softened berries were chosen at random and cultures were made from their pulp, as described above. Twenty of these berries, or 40 per cent., yielded the end-rot fungus, while 22 berries, or 44 per cent., showed no fungi, and were presumably destroyed by the other causes discussed in this paper. Effect of Carbon Dioxide on Fungi in the Berries. That carbon dioxide in high concentrations injures fungi in the cran- berries as well as the berries themselves is indicated by a test in which equal numbers of rotten cranberries from a single lot were placed in similar ves- sels, one of which was filled with carbon dioxide and the other left open. At the end of one week transfers of tissue were made from each berry. Of the berries which had been kept in an atmosphere of carbon dioxide 70 per cent, contained no viable fungi and the others yielded Penicillium, or the end-rot fungus. Of the berries kept in the open vessel only 15 per cent, contained no living fungi, and the others jdelded fungi of six different species. The rate at which carbon dioxide is given off by cranberries in storage and the variation of this rate with temperature, the concentration of the gas necessary to cause injury, and the concentration which occurs urder storage conditions, have not been determined, and further investigations on this line are planned. It seems very probable from the facts now in hand, however, that this spoilage is a considerable factor in the loss during storage, and throws new light on the results of Dr. Franklin, ^ which indi- cate the importance of ventilation, as well as on this year's results in shipping cranberries in tight as compared with ventilated packages. 1 Franklin, H. J.: Report of Cranberry Substation for 1915, Mass. Agr. Expt. Sta., Bui. No. 168, 1916. BULLETIN No. 181 NOVEMBER, 19(7 MASSACHUSETTS AGRICILTIRAL EXPERIMENT STATION Digestion Experiments with Sheep By J. B. LINDSEY, C. L. BEALS and P. H. SMITH The complete data of one hundred and fifty-three digestion experiments with sheep, together with short discussions of each, are contained in this bulletin. Complete summaries and aver- ages are given in the latter part of the publication for the con- venience of the reader. The bulletin is not intended for general distribution, as it is of a technical character; it will prove of most use to experiment station workers and others interested in determining the relative values of feeding stuffs. Requests for bulletins should be addressed to the Agricultural Experiment Station, Amherst, Mass. Massachusetts Agricultural Experiment Station. Trustees. OFFICERS AND STAFF. COMMITTEE. Charles H. Preston, Chairman, . Wilfrid Wheeler, Edmund Mortimer, Arthur G. Pollard, Harold L. Frost, The President of the College, ex officio. The Director of the Station, ex officio. Hathorne. Concord. Grafton. Lowell. Arlington. STATION STAFF. Administration. William P. Brooks, Ph.D., Director. Joseph B. Lindsey, Ph.D., Vice-Director. Fred C. Kenney, Treasurer. Charles R. Green, B.Agr., Librarian. Mrs. Lucia G. Church, Clerk. Miss F. Ethel Felton, A.B., Clerk. Agricultural Economics. Alexander E. Cance, Ph.D., In Charge of Department. S. H. DeVault, A.m., Assistant. Agriculture. William P. Brooks, Ph.D., Agriculturist. Henry J. Franki in, Ph.D., In Charge Cranberry Investiga- tions. Edwin F. Gaskill, B.Sc, Assistant Agriculturist. Robert L. Coffin, Assistant. Botany. A. Vincent Osmun, M.Sc, Botanist. George H. Chapman, Ph.D., Research Physiologist. Paul J. Anderson, Ph.D., Associate Plant Pathologist. Orton L. Clark, B.Sc, Assistant Plant Physiologist. W. S. Krout, M.A., Field Pathologist. Miss Mae F. Holden, B.Sc. Curator. Miss Ellen L. Welch, A.B., Utenographer. Plant and Animal Chemistry. J. B. LiNDSEY, Ph.D., Chemist. Edward B. Holland, Ph.D., Associate Chemist in Charge {Research Division). Fred W. Morse, M.Sc, Research Chemist. Henri D. Haskins, B.Sc, Chemist in Charge (Fertilizer Division). Philip H. Smith, M.Sc, Chemist in Charge (Feed and Dairy Division) . Lewell, S. Walker, B.Sc, Assistant Chemist. Carleton p. Jones, M.Sc, Assistant Chemist. Carlos L. Beals, M.Sc, Assistant Chemist. James P. Buckley, Jr., Assistant Chemist. W. A. Allen, ' B.Sc, Assistant Chemist. J. B. Smith, > B.Sc, Assistant Chemist. Robert S. Scull, i B.Sc, Assistant Chemist. Bernard L. Peables, B.Sc, Assistant Chemist. James T. Howard, Inspector. Harry L. Allen, Assistant in Laboratory. James R. Alcock, Assistant in Animal Nutrition. Miss Alice M. Howard, Clerk. Miss Rebecca L. Mellor, Clerk. Entomology. Henry T. Fernald, 2 Ph.D., Entomologist. Burton N. Gates, Ph.D., Apiarist. Arthur I. Bourne, A.B., Assistant Entomologist. S. C. ViNAL, M.Sc, Assistant Entomologist. Miss Bridie E. O'Donnell, Clerk. Horticulture. Frank A. Waugh, M.Sc, Horticulturist. Fred C. Sears, M.Sc, Potnologist. Jacob K. Shaw, Ph.D., Research Pomologist. Harold F. Tompson, B.Sc, Market Gardener. Miss Ethelyn Streeter, Clerk. Meteorology. John E. Ostrander, A.M., C.E., Meteorologist. Microbiology. Charles E. Marshall, Ph.D., In Charge of Department. Arao Itano, Ph.D., Assistant Professor of Microbiology. George B. Ray, B.Sc, Graduate Assistant. Poultry Husbandry. John C. Gr.^jiam, B.Sc, Poultry Husbandman. Hubert D. Goodale, Ph.D., Research Biologist. Miss Rachael G. Leslie, Clerk. Miss Grace MacMullen, B.A., Clerk. Veterinary Science. James B. Paige, B.Sc, D.V.S., Veterinarian. G. Edward Gaqe, Ph.D., Associate Professor of Animal Pathology. J. B. Lentz, 1 V.M.D., Assistant. ' On leave on account of military service. ' On leave. Publication of this Document approved by the Supervisor of Administration. CONTENTS. PAGE Introduction, . . . . . . . . . . ' . 241 Composition of feedstuffs (per cent.). ....... 242 Composition of feces (per cent.), ........ 249 Weight of animals at beginning and end of each period, and average daily water consumed, ......... 256 Digestion coefficients of basal ration used in the computation of digestion coefficients, .......... 263 Computation of digestion coefficients, ....... 265 Discussion of results, .......... 306 English hay — basal, . . . . . . . . . 307 English hay and gluten feed — basal, ...... 308 English hay, potato starch and Diamond gluten meal — basal, . . 309 Gluten feed — present experiments, ....... 310 Gluten feed — earlier experiments, . . . . . . .311 Diamond gluten meal, . . . . . . . . .312 English hay fed with wheat gluten flour (to note effect of the wheat gluten flour), ......... 314 Corn bran, . . . . . . . . . . - . 316 Distillers' grains, . . . . . . . . . .317 Feterita, 318 Alfalfa 319 Roots and vegetables, . . . . . . . . .319 Cabbage . 319 Carrots 321 Mangels 322 Pumpkins, ......... . 323 Turnips, ........... 324 Comparative summary, ........ 325 Vegetable ivory meal, ......... 325 Vinegar grains, .......... 326 New Bedford garbage tankage, ........ 327 New Bedford pig meal, ......... 328 Rowen 328 Soy bean hay, .......... 329 Stevens' "44" Dairy Ration, ........ 329 Sudan grass, . . . . . . . . . . . 330 Sweet clover, ........... 333 Complete summary of the averages of all coefficients, .... 334 BULLETIN No. 181. DEPARTMENT OF CHEMISTRY. DIGESTION EXPERIMENTS WITH SHEEP. J. B. LINDSEY, C. L. BEALS AND P. H, SMITH. Introduction. The digestion experiments reported in this bulletin were made during a number of years, beginning with the autumn of 1912. They include portions of Series XVIII. and XIX. and all of Series XX., XXI. and XXII., with the exception of one experiment in Series XXII. Each series in- cludes a period of time between the early autumn and the following spring. A few of the results have been given in other publications. The basal ration in the majority of cases was English hay, or English hay and gluten feed. The usual method of conducting the tests was employed, and has been fully described elsewhere. ^ The composition of the feeds tested in the several series is presented in the tabulation known as Table I., which is arranged alphabetically. Table II. is arranged by series, beginning with Series XVIII. It con- tains the average amount of feces excreted daily by each sheep, the weight of one- tenth of the feces in air-drj- condition, the percentage of dry mat- ter in the air-drj^ feces, and the composition of the dry matter. Table III. contains the weight of the animals at the beginning and end of each digestion period, and the average amount of water consumed daily. In Table IV. will be found the digestion coefficients of basal rations used in the computations which follow in Table V. This table, headed "Com- putation of Digestion Coefficients," presents the detailed data of each trial, together with the resulting coefficients. Following the complete data will be found a summary of the coefficients secured for each material, together with a discussion of the results. Table VI. gives an average of the coefficients secured for each feed tested. It may be stated that the period in nearly all cases extended over four- teen days, the first seven of which were preliminary, the collecting of the feces being made on the last seven. Ten grams of salt were fed each sheep daily, and water ad libitum. The sheep were grade Shropshires, as nearly as possible of the same age and weight. 1 Mr. Smith and Mr. Beals did the larger part of the analytical work and the tabulations; the work at the feeding barn was carried out by Mr. J. R. Alcock. 2 Eleventh report of the Mass. State Agri. Exp. Sta., pp. 146-149 (1893). 242 MASS. EXPERIMENT STATION BULLETIN 181. 1 ^ g 2 ^ § 2 § § s :s g § s s s s s . . . 05 o> CO •«»< CO ^ . Sgl ?;? H t~ ?! s s s s s s< R 2 f5 s s § § .-2"~ X t^ s g B < m ^ I-. '-2.,^^- s ^ § g f3 S 2 § ^ S3 s ss § s g S £ 6|«|S £ S . 2 ^ s s ;: 8 § s g § s § § 8 ■ § § P=H 1 a a 1 a' 8 1 3 73 1 ? > 1 s s ^ bO •? •3 1 1 £> "i^ - - - s "i^ '?; i ^ ^ Sf rt' ^ 5 ^ i2 i ii 3i ji a d s s •3 ■3 •S •^ •^ 2 g 2 s g S !^ "^ ■^ ^ ^ d 6 6 c3 a d a d [S O O o o 1 Ph 1 1 X U ><■ x' x' x' ^ t^. X ^ >< >< X X d x H X X X X R DIGESTION EXPERIMENTS WITH SHEEP. 243 s S a> OS S? CO o CO oo 00 CO OS •* <-, 05 OS OS OS CJS o g OS o o o OS s s S? JO cc _ s s o g s OS 2 s s § §5 2 s u oo S? 00 . . t~ . cc CO ^ ^ t^ r- '^ . . . . (M " m CO " s s s ^ 3 OS OS s CO lO CO ire to OS CO CO CO s CO CO . . . . . M « CO " CO « ire ire CO <-l ire ire o § ire o o o § ire ire o CO o s s s s o g s § s s s s s s § S 5 § s §8 88 -s r3 -? e g e Q fi g g OJ ■ Q e s 0 Q P ;- s 3 ;^- 1 1 1 1 1 1 1 1 1 1 1 1 i a a a a i ;3 ;s c ^ e C e s c a S n (H 0 g g C; a Q (j a a l^ P^ [^ B ^ i^ ^ ii i! s ^ ^ .1^. 4^ *^ •2 ."ii .iH !^ -2 ?^ .2 3 ;) 3 3 3 3 3 3 3 3 3 3 W o o O o O o O O O o o O O o o O o o W W ffi c ^ « o = 2 2 - n 00 - 2 o ^ 2 lO CO - CO Tt> ire CO x' X XI X ^ X! X X X >< X >< d >< X X X X X X > X i « 1 244 MASS. EXPERIMENT STATION BULLETIN 181. 5 s -* o o> a> o o 00 00 00 t^ t- t>. r- 00 t^ t^ to Oh s to U3 o eo to § •fl eo CM -« CM ■s ■* 02 00 t^ -< o u? "5 ui to bsJ- p; s s ^ s S; m t^ Ifl s K? K fe g g J= § °l"f OS 00 §8 S5 §8 S5 s 18 §8 s § s s § g 8 § s t^ ^ la ,. _. ^ 2 ^ _. ^- A ^ s 2 3 .a "^ 2 .a .Hi ,"3 .a .li .2 .2 .2 .2 .2 Wl To Nl bO "Sb bD bO Tn N, "ib "bb "m "m a s S H. a B (i fei y fel a fel a li S S 8 >> J^ >> >, >> >i >1 >i >1 >, >> >, >. >, >>>>>. ^ W ^ ^ ^ f§ £ 5 s^ ^ ^ ^ a iS ^ ^ ^ 1 ^ M CO •* US o ^ CO 05 o ^ CM CO ■* W - CM (2 i ^ «■ i*^' >< M x" ><■ x" >< M «■ M «' M >< >< y. ^ X X 1 > t^ « ;«! >< ><1 « « >^ « « ^<) DIGESTION EXPERIMENTS WITH SHEEP. 245 s » S " " « £ ° ^ •* s s " ^ « " s s S IS s o to S g - s s s g S s s {2 5 g ?5 CO s § « s s s 5 S ?5 is ^ ?? CO s sj 5 s s s ?2 ?2 ?3 f3 ?5 s « oo s S s >rt oo ;. " to s s S t^ s S t^ 2 § g to to to to to to °. s to ^ to 5 to to ^ to to to ° SI to 8 s g § s 8 s s 53 ^ § 1 38 §8 S5 J2 s § S s § •1 c e 1 J 1 1 1 1 .s 1 3 c! s 1 1 1 1 1 1 1 1 1 1 1 1 .2 1 i I 1 U1 1 a i 1 1 c 8 1 1 .2 1 CO - lO to - - o 2 S ^ 2 - - . CO M. ^ o - 00 o 2 X X 1^ ^ X >< >< >< X y. >< X X >< >< X k1 X X 1.x X 246 MASS. EXPERIMENT STATION BULLETIN 181. lO o to "S t^ •* [^ N M i . t^i § s S § S S 5: S s S n s § S g E5 S^^ ^ 5 g s S § g S g § 00 g w « OS a> t^ o § ^ _§ s? g f=:. CO H Q d § 2 s ^ S ^ 5 ^ O o Ol 00 00 o 00 tn to PLI s t^ O — 1 OS to « •o 00 U9 o> •s ■* « , >> >)>>>.>>>>>> >) >, >> >> >, >. !» C C M W K W W W W W w K K W W S S i Ol (S 8 y. y. ij ti P5 >< M >< •^ '^ ^^ X ^< x „• ^J ^• a ^• M h-i h-i 1-i >^ X! XI X. XI x: XI X5 X! X 248 MASS. EXPERIMENT STATION BULLETIN 181. ^ 2 % < 1 SSSS^ggf^gSK „ . o ill Sgj;?SS2§§^?355 g„-o«oo^u,g,oo J S §g§;:£§SSK22 S?S2 '-■^^gg d 1 SISS5S22SglS?3 ^;3°'Sg§'=''"'Sg < §gSSSE:S8E52 2SE; ... . . Dry Matter as weighed out. 16.55 14.50 13.79 91.66 92.96 94.96 87.36 91.25 89.10 92.22 92.53 » fe Sweet clover Sweet clover Turnips (Swedish) Wheat gluten (high grade) flour Wheat gluten (high grade) flour Vegetable ivory meal Vegetable ivory meal, Vinegar grains (Fleischmann's), Vinegar grains (Fleischmann's), 1 1 •*coi^ci-*o>OMc<3roo 1 XXI XXII XVIII XX XXI XXI XIX XX XXI XXII XXII DIGESTION EXPERIMENTS WITH SHEEP. 249 ' o •c lO C-l S o t^ •o t^ fe CO ^ ■o rr ^ M- "^ u? w CO CO CO CO CO CO CO CO l»l g - o fe S S3 s g g ?3 K X s J* s s? 2l-i3 m CO < C3 z w J § § g g g U5 s § g t^ o o s s ^ s S3 s 13 s S « S3 S s ?3 S; s s § s s S 1 g 2 S s s s £ s § S s s s s s s (N M CO M — • 1 i 1 1 1 1 c § I J2 1 1 1 1 1 1 1 a a c 1 1 E S C3 2 s 1 1 1 2 s s 3 3 "bi 3 "bb "d 03 03 >> ^ >) >1 >, >> >1 >, >, ;>, ;., >> ;;■ >> >> >> o3 03 ^ ^ ^ J5 .a .4 r- ^ j- J- 2 bC ■^ bD bO bO s* bO c C C c C e c c W w W w W W W W W w W w w W w w g- > r: « K - d > ^ >■ > >' > -■ K h-; j-j M •g « « ^ 'J- o o o o t- t- M CM CO CO Tf 'i- £ .1 1:3 a ti «• H a H g <;^ H >^" H ;^' >< d d fe > > > > > > > > > > >< t^ >< X X ^ y. X >< X X X! X X 250 MASS. EXPERIMENT STATION BULLETIN 181. =5 CO >0 O M 00 O OO t^ CkO^«coo — 10IO*0^*^'^CO»0'^ t^ — . .O CJ -« ^ ^ e ^- _- a a HWWWWHWWW 2 g w w w w '>< X X X X X X >< >< >< '^ >< >< ^ >< >< >< ?, X X X X X X X X X X X X X >< X X DIGESTION EXPERIMENTS WITH SHEEP. 251 s ^ eo § ° •* " o 5! S5 CO "5 S3 S o - M •* oo S ^ l^ ° " i^ 5 5 CO s 5 s 5 ". g 5 » " ° g S to 5 tn ^ ^ §5 s S5 s s s s s " IS S5 « S S3 s 5 g s si U S s s g o a. °. 5 05 g d § « •* « •* « ^ ^ s o s oo s " g s s 2 o s ^ 2 S s 2 S " o 2 ?5 2 s s ^ " ?5 U5 s S § s g s s •>»< ^ s S S S g s o g s £; S SI ^ i s* i §5 s 5 i i s i ci S3 i i si i i i S3 1 t, o ■* .2 2 J 2 .S S S S S S -s -^ a « S -2 2 2 ^ o o a WHWWWWWWWHHWWHW W W W W H H > >5-S-S-S-H-S>>^ > ^ > &; > > s 2 S ------------- CO - " - - - 252 MASS. EXPERIMENT STATION BULLETIN 181. 2 "S fe « M o SJ g s s f^ s f2 § S5 S s §3 o CS 00 ■* ■* < 1 w « fe S s g s s B s 5 5 u s g Si ;s § S5 'i E S g s s s g S s 13 S5 s s s § s S S s >< g 5 Q £ OJ o= 1- «n eo o o CO CO M o — o o ^ — C-- to to Ol o g O: C-. -* OO — OO OO — ' —1 !2 o c b-a-oS s? 2 s 2 s « CO s g 2 s gg 2 g g S 5 °|-^|^ § S g g s g § s s S s s § g g g s 1^0 i 1 g B i i i i ^ i 1 i g 2 s g = -mi s §3 s s s §5 S5 s s ?? ^ s s S5 ?:j ^ g M w s s S? 5? to s s s? o i ^ S H? >p^6 -rt' ■s e s s « g g g S g 0. 0. o. >> >> >> o => s >• .t 1 « 1 u 1 1 cS i g Z 2; :S S 1 1 > fe c -? 3 1 •s 13 s 1 t 1 1 1 1 y. -? 1 M 1 m 1 3 3 S^ S •2 •2 4^ 5 ii s i: 3 3 ?^ ^ 3 3 ■a T3 "0 bC bl a g •a CJl •a g ^ 8 ea i ea >> >i >> >. >, >i J>, >> >, >> >> >. >i >. >i >. ;> ^ i ^ ja .a .a J3 4:: _g j- j- j3 j3 jS .2 OJ en «i «i .!!! .2 .2 .2 .2 ": ^ Tn M N, b£ "3) "m u> •a Tn -a M •iii ■si bf) a 3 W W w H W W M w W w W W w W W W W i > a 1-H > ^ H S > t^ > > > > > > ^ K :§ c. ., eo CO CO ^ ^ •«• ^ „ „ IM 1 « X >< >< « x' X X X X X X d X X X X X X X X X 1 >^ t«1 X (>< X X X X X X X X X DIGESTION EXPERIMENTS WITH SHEEP. 253 § s » 2 § s f2 S s § s S S S o g § " o S 2 2 ■«< eo ^ M< eo "* "* "* "*■ -* "" M< re CO CO CO Co s s CO s s 2 g § s s S s S § s ?3 s; s; ?3 §5 S5 2 s s; § S5 s g g S s s 05 00 o> <3> 05 § o t~ — CO •o< —1 t^ ° " " " OS " " " o 2 2 ^ 2 2 l« 2 2 2 o » ° ° s ^ g ^ 2 ' =^ OJ 2 =o. CO 05 o o CO Ol ^ 2 2 3 o o> CT> Oi o g e > > „ -O -O T) 1 -g ■« I § g g I I I 5 1 1 1 £ ^ <£ *«■«"£ a a c "S 5 5 5 S S S * -s ■§ ■§ ^ ^ ^ -g o o o tD bt bC g 0< O. S 2 2 w w w w g s 1 s 1 s s ■g S 6 1 3 1 3 c s 3 -2 3 3 -M M M H ii 1 2 1 1 1 1 R R R s 5 o ^ f^ -2 5 S H. a H. a K K >, >; j;; >. >. >, J3 ^ ^ D c W w f^ w w W ^ ^ HWWHWWWW X y. X y, X X X X X X >i X X X y, >i X X X y. y, X X X X X X X X X X X X X X X X X X X X X X X 254 MASS. EXPERIMENT STATION BULLETIN 181. 2£| r-. op o H S -*oe«5-.-* O CJ »0 CO .-I l£3 r~r^-HO-*oo(MOt~oocooo-*ooo>o 1^-<:jn^h.-hcC00O05'**j<«eo(M'cococc-*M-*cocoodcio to OS fe 3 § g o >« -^ .a 1 ^ piJHHUHP'^P^H W W H W W H H W H W 1 ^ " X ^ >^ " ^ t ^ ^ ^ ^' ^ d ><■ d 1 ^^MMCOMM-a" -* rH -H M 00 CO 2 CO ^ ° '^ ^ e<> 0 00 03 i X X X '^. X X X X X X X X 256 MASS. EXPERIMENT STATION BULLETIN 181. II s g «2 g g g S5 S K g g f2 g g "5 8 s ^ 1 s 2 s 2 a S § S g 1 S s g 5 5 a fe % «f, s g J2 g § s JO s s § K S s s g S J5 H s| s - « 2 2 2 s s - g s s S s 53 S 2 ^ " — g O M K g K § g S g g s K g g g g S g 8 o g s S i S M K 2 s i S s s s 2 5 2 ^ !Z -fn s g K g 8 g s S s g R s s s 8 8 g E| g lO s. i 2 2 g g 5 i 2 s 5 s s g S •rt fl . .._ — — — — " II m i s i i 1 i s i i ■Of 1 g » s ?? « i i* ill t s >, ?, -^ ^ P x! ^ ^ p P H fe 1 ■i 1 C9 1 ^2 1 .2 1 1 T3 s 1 "i -a 1 1 1 1 1 1 a 3 a c 0. D. 1 s s 1 § 1 § a g 3 3 3 "bb §. S g -3 -0 3 tS >> >> >, >i >, >> >1 >, K^ >i j>, >) ^ >, >, >> >. C3 JS J3 J3 X ,j3 >4 - J- ^ r- J3 ^ j3 j- J- ^ ^ ^ _a J= J3 ^ J 2 .•ii .a .a .a .ffi 2 .la 2 .2 .2 .2 .ii tc Tn Tn "3) Nl "m "bb bC "m "bb -a Tf ■bil ■bh c c c e. W w W W W w w W w W w W W w W W w d >' H KH- ^• « ^ >' «• >" > >' > «• a- -• tij > CO 1 M CO ■«• -9< o lO to « f. t^ „ M CO CO -X T^ Ifl 1 s a ►-; «• ;-; « a B h; hJ a X ^ ki X X X X > > > > > > > > > >< >^ ;^ X X X X M X >^ >< y. x }-! >^ X y. X DIGESTION EXPERIMENTS WITH SHEEP. 257 s s 5 ^ !^ ° S § g 5 3 e g g 8 s 2 g 8 1 8 -. S ^ g i g i 8 B 2 s 2 1 s s § § i 8 i 8 S g g 8 i >t5 i g i 8 g K S 1 s B i s e § K •* U5 ^ 8 2 8 1 8 g s ^ 8 8 ° K ° 8 2 s ^ g i 1 § 52 g 3 K i 8 8 i i 8 g 2 s s s 8 S s g 2 g i g S ^ 8 12 S 1 8 s s :s 8 a s 2 £ .2 -a 1 1 5 S I I a I S -2 I I a a a a HWWWWWKWKWWWWHBWWW y, y, X 'y, X y, '/i X y, X y, y, X y, y, y, y, y, y, y, y. X X X X X X X X X X X X X X X X X X X X X X X X X 258 MASS. EXPERIMENT STATION BULLETIN 181. Eh S i^ T3« O M go S § S lOOiOOOiCWSOOiOOOiO o »o o s s s s ^ ^ § s § s o o o lO O to »C lO s ^ s s s g s g g S B s .2 .2 .2 .2 .2 W W H W W W K DIGESTION EXPERIMENTS WITH SHEEP. 259 g K s « S g .o s g § g •o § s s s g s s g t2 s s § s s SS § 1 M g g 2 s 2 g 2 ^ s 2 s g S3 2 S s § s g K § g s s s S t^ 8 s s s g § o g g g g § t> s s ss § g 5 g g 2 s 2 B 1 s s 2 S 2 S 2 S § s g o § S g K § § s g g g s K § g K g S S s g i s 2 § § t^ X5 s t^ s g 2 to § S s § g 2 S 2 00 S s t- § s S g S § § g s S g s JO g g § s S g g s 2 S § -^ '^ -^ - i 1 s 5 i 1 1 s 1 1 s 1 1 i ? 1 1 S 1 g 1 ' ^ ?: ?: f c £ £ 1 £ ci £ c E a >, >) >, E -a 1 1 S D 1 1 i £ pa 2 S g Z 1 1 > 1 1 1 -2 3 -3) "bb ■a £ •3) ■3 "2 S 1 1 1 3 •a -0 -a 1 1 £ 1 1 1 1 03 T3 1 1 1 1 1 S J i ^ J ^ y oi -S y ^ s ^ ^ S "S 5 m 1 c c d a jj g ^ g p fl jS o o O S ii -2 J^ bU bll bl) i^ _g ^ "S s C3 3 s 3 M ■bi •3) i 'bi bD § e § § i Ti -3) ■a § n ft ft >> >> >, >) >> >. ^ ^ J3 ^ ^ ^ J3 .a ^ J= j= J3 j; x JS _jg ^ _l- 43 pl3 j:j s: ^ j3 _J-. J3 J _i; X ^ ^ _j- .2 .2 .2 .2 .2 .^4 .2 .2 .2 .2 .2 .2 tc bO bC bl Ul bl bC bl bC U) bC c c c S a K Ui W W w W W w W W w W W W W w w W W W W W W > > > td > > > > > > > >' > > ?* d > > > > > > t: :: = = ?^ - ^ CO 2 3 S 2 - - - c, - « eo eo - o w lO X ^ i X X X X X X X X X >< X X X X X X X X X X X X X X X X X X X X X X X X X X X 260 MASS. EXPERIMENT STATION BULLETIN 181. H ^ 9 5^ ^.r: s S s !^ S S ^^ s s w K? ^ ^ ^ g « s s 2; t»^ -< ^ p ^:s s S § § s s g g g § J5 s t2 S § S 8 W E| 2 g - ;2 2 1 S 2 2 f^ ^ ti to 2 s 2 ^ ^ a ■?2S s S >c? f^ s ^ g in g K ira ^ !5 ^ s g s g-Sf o o OS to a> 05 o fe g K^ ■^•1 s s § 8 g K s g K § g § g g § g g 3 S| Is g s S s § s g 2 K g to 2 § g B § m » "S*-^ ^felog s § !S o S s; go ^ 1^ s s S Sfl 5^ ^ s Aver Wat consu Cubic timet* S- co- -" - M --' N- « > >> jT >, >> >> >, >, sT >, >, >, >i >. >> >> >> ,4 A A J3 A .s — j3 j3 ^ j3 ^ .2 "1 ,"^ "'. "1 "< ,"i "i "? «i .a ■'i "hfi "3) "hb Tn -n •hn -n To ■bii M Hi 3 W w W w W w w w w W W K W W w W g" >■ > „. y, X >< > „. >* ^ S "-1 >^ l-l ■^ ■^ ■^ ■^ m t CO CO o r. i. I^ 00 «> 00 a> OJ » o o ^ -;; c. ^ i 'A y. M Pi >i P5 >< >< ^ P, « >< P3 >< >^ X X •g 'A y. 1^ ;^ y. >< X X X >,>>>,>> OD cm ex I ^ >>>.>.>.>> H W W W s 3 3 e c CO 02 CO X X X X X >^ X X X X X X! XXXXXXXXXXXXXXXXX X X X X X X X X X X X X >^ X X X X 262 MASS. EXPERIMENT STATION BULLETIN 181. 1 -§5 s s >o g § s g g g g S t2 t~ s s S ^ s 3 § o 5 § S S: S o o § g S 1 t3 o Q iz; IS s CO K g s s g K § § K K g s s s o p.^ m ^ i2 ^ s 2 00 s S § s § O 00 a H ■ < i 11 K s S § K t2 K § s s S K s g g S-s ^ esi t-i CO lO CO 1^ fe cg^ 2 CO 2 o oc O) 00 131 CO o sn O 00 < o P z 1 U5 lO o U5 S S o o •o o lO o o o o a B N t^ •n t- « lO t^ o I^ o >o o o S S "i « ?j i §g § S si S S; o g S5 <; H ^ K PQ El <1 ISUI g 1 § s T 00 § s S 5! K s § 1 Q W t^ CO «-)_ f-^ o > >. S: Q b< -2 i a 1 1 % ■3 "O tt ■> m m s -C !B CC ^ T3 ■o T) "3 3 "3 "S ■« -o ■B '0 S rt o s c c < S cj ■C T3 •o "H C8 cS c« ci M >, >> >. .S ^ ^ © s >> >) 2 03 2 A C3 cS 3 > > s s X. ji ^ o _o s 1 1 •s S ^ ■s 1 03 ^ g " i :! ^ M "3) 1 rt s "m M 3 3 ^ ^ s § ■hi M < w ^ ^ w ^ ^ f§ (§ 1 02 a 1 3 a i > " > - > t^ >< >< 1 ^ s >1 ^ ^ S w ^ :§ o « _. c^ 2 „ CO 2 3 2 s s CO t~ r~ s fS a .J <5 1 Pi Pi X 1^ X Pi Pi >< Pi ^ Pi Pi Pi Pi Pi Eh ^ ?; y. >^ >< Kl y. >^ K X ^< X k1 'A y 'A A DIGESTION EXPERIMENTS WITH SHEEP. 263 ^ -_ ..^ ^ _ . . t;". . > §^ ' - P, "• X "^ J>S ^ j^" J 1 «• «• r^ X X! y, X X X X X X |> G X X X ^ K^ o > > ^ >< P P p X X % S o -^ CO 0 ira 0 r^ CO QO "o ■» CO "* "5 >o Tt< 0 •»< 0 -"S* -* •* 1 ^ a ^ >> m 5 £ s 1 s S g s § f2 S S S 8 ?; is— >4 . c g « a ® S 0 S g 5 s 1 s S § 3 s s 5 S S S S 'N S n d 1 i i- o S s ' s t- g s s us S S 5 5 S 0 i p-( s=i I.S to 00 c« LO 0 =0 t^ 00 OS 0 -^ ■< CO ^ CO CO CO CO CO •* e-j ■* CO (M °l s s s 8 s s § s 3 5 IS 2 g 5 5 \l s C P ^ Tc -3" ~ •€ § 8 1 1 • J ■ • s :s 0 1 c c s c H 3 -2 3 . -2 . . -2 J 3 3 3 3 2i rf "m "tb Ti "m "m J n T! t) "2 . -3 . . T5 8 C c C fl '5 5 C3 _ C3 a >; s^ t^ _£ >^ >^ >> >> >i >v ;^ >.>;>; >. C3 03 o3 o3 c« o3 03 c3 03 ea =3 c3 c3 03 J3 ja t. JS J= J3 X 4: ^ J= J= _J^ ^ j3 ■S X J- •X j- J3 ^ "" " -d X X .2 .2 .2 tn .2' 6 §= .2 2 [2 3 .2 .2 .2 M M M g "Si •g, "bb To 'm M M M M M e e O c a a c c c c §g H 1 W O w W W W W W H H W W H „" ss > 1 >! > a > > a§ a 1 . ^ C3 . t-< J . . 1 -§ > -g -B ss J;;^ > > y' >, > y y' CS 03 -a C3 =3 M 1 j ' > -t3 03 > 1 > e3 a > > ^ > > a a ■43 -^ "a "3 2 pa P3 •3 -- . CO . II O t~^ "2 ^ eS t2 ^ « e< ^ - i s •a tC -5 ■* ^ - ' i i 3 --0 o i o o 2 ■^ g S "■_ j^- ci c3 ^ M C-j" ■* *" '"' C-l 00 .^ 0 -H -. CO j^ «• «■ x 'd 1^' d y. X X X X X X d ri 1 > > k1 X P, p. P X X X X X X r:! j:^ 264 MASS. EXPERIMENT STATION BULLETIN 181. < - =. "5 < .2 *"■ K X '"' ^ ^ Q 1 y. > 'A o ^ ^ „• "^^ "~ "" h-; I-; i 1 ' >< x ^ o. "5 CO fe ■o "* TJI i ^ III S § S S R s s § § 1 w t.' o oc cc 0-- CO CI ^ CO » CO o CO CO CO E d CO CO 00 ^ £ -j< t^ lO «^ >« in .* ira 00 40 CO < CO CO s ^ ^ ^ si 00 I fe g r; s s g g = i • J 1 :z e 0 C3 c » +J f m pi 1 "bk 5 -is 1 pd .' - >, . t^ j^ t^ kT ^ >> C3 ci— c4 cj cS S -g-C j3 J= J= .J3 j3 -ss-s J= J3 J3 J3 J= .22 .2 00 .2 .2 c ^g^ "si "m "a •a M a o c->^ a fi c a d Ci W O H W W W H W W 1 r 1 > > _^ - fUj (£. -S -B i-i i_i 1 C3 c3 > > > T5 a. >■ >■ -§ c C a S C3 tr* t t > ^ l-H d X .9 21 -a 1 1 "^ CO C3 _ <1 DIGESTION EXPERIMENTS WITH SHEEP. 265 Table V. — Computation of Digestion Coefficients. Series XVIII., Mangels, Period 3. Sheep V. Item. 1 i .S 1 1 s 1 ii pi, 400 grams English hay fed, . 1,400 grams mangels fed, 347.00 228.76 575.76 151.42 19.67 13.57 34.25 12.03 109.51 13.63 174.24 188.89 9.33 .64 Amount consumed, Minus 159.94 grams feces excreted. 33.24 23.98 46.28 19.06 123.14 35.67 363.13 65.43 9.97 7.28 Amount digested Minus hay digested, 424.34 225.55 9.26 9.05 27.22 22.26 87.47 73.37 297.70 116.76 4 29 Mangels digested, .... Per cent, digested, 198.79 86.90 .21 1.55 4.96 41.21 14.10 103 .45 180.94 95.79 " Sheep VI. Amount consumed as above. Minus 155.50 grams feces excreted, 575.76 147.21 33.24 21.73 46.26 17.86 123.14 35.62 363.13 64.49 9.97 7.51 Amount digested, Minus hay digested, 428.55 225.55 11.51 9.05 28.42 22.26 87.52 73.37 298.64 116.76 2.46 4.29 Mangels digested Per cent, digested, 203.00 88.74 2.46 18.12 6.16 51.18 14.15 103.81 181.88 96.29 : Average percent, digested, . 87.82 9.84 46.20 103.73 96.04 - Series XVIII., Cabbage (Heads), Period 4. Sheep I. 400 grams English hay fed, . 1,600 grams cabbage (heads) fed, . 357.40 154.56 18.80 12.70 33.06 27.79 111 97 15.21 184.06 97.02 9.51 1.84 Amount consumed, Minus 131.90 grams feces excreted, 511.96 125.34 31.50 15.44 60.85 17.08 127.18 29.82 281 .08 57.67 11.35 5.33 Amount digested, .... Minus hay digested. 386.62 232.31 16.06 5.83 43.77 20.17 97.36 78.38 223.41 123.32 6.02 5.04 Cabbage digested. Per cent, digested. 154.31 99.84 10.23 80.55 23.60 84.92 18.98 124.79 100.09 103.16 53.26 266 MASS. EXPERIMENT STATION BULLETIN 181. Table V. — Computation of Digestion Coefficieijts — Continued. Series XVIII., Cabbage (Heads), Period 4 — Concluded. Sheep II. Item. 1 1 i a 1 ll p2 Amount consumed as above. Minus 138.19 grams feces excreted, 511.96 131.56 31.50 16.27 60.85 16.64 127.18 33.63 281 .08 59.30 11.35 5.72 Amount digested, .... Minus hay digested. 380.40 232.31 15.23 5.83 44.21 20.17 93.55 78.38 221.78 123.32 5.63 5.04 Cabbage digested, Per cent, digested. 148.09 95.81 97.83 9.40 74.02 24.04 86.51 15.17 99.74 98.46 101 .48 .59 32.07 Average per cent, digested, . 77.29 85.72 112.32 102 .32 42.67 Series XVIII., Cabbage (Leaves), Period 5. Sheep I. 400 grams English hay fed, . 1,200 grams cabbage (leaves) fed, 355.20 228.60 20.32 33.12 33.64 27.29 111.82 29.99 179.08 132 .69 10.34 5.51 Amount consumed. Minus 186.23 grams feces excreted. 583.80 177.27 53.44 32.00 60.93 22.21 141.81 39.35 311.77 75.84 15.85 7.87 Amount digested, .... Minus hay digested, 406.53 230.88 21.44 6.30 38.72 20.52 102.46 78.27 235.93 119.98 7.98 5.48 Cabbage digested. Per cent, digested. 175.65 76.84 15.14 45.71 18.20 66.69 24.19 80.66 115.95 87.38 2.50 45.37 Amount consumed as above. Minus 199.50 grams feces excreted, .\mount digested, . Minus hay digested , Cabbage digested. Per cent, digested. Average per cent, digested 583.80 189.72 163.20 71.39 53.44 32.46 14.68 44.23 37.14 20.52 16.62 60.90 141.81 40.81 101.00 78.27 22.73 75.79 311.77 83.91 227.86 119.98 107.88 81.30 DIGESTION EXPERIMENTS WITH SHEEP. 267 Table V. — Computation of Digestion Coefficients — Continued. Series XVIII., Mangels, Period 6. Sheep V. Item. 1 1 1 .■sw 1 400 grams English hay fed, . 1,800 grams mangels fed. 355.68 312.84 23.23 19.65 33.40 20.08 115.06 21.30 174.67 251.00 9,32 .81 Amount consumed. Minus 179.45 grams feces excreted. 668.52 170.08 42.88 24.07 53.48 22.06 136.36 40.19 425.67 74.20 10.13 9.56 Amount digested, .... Minus hay digested, 498.44 231.19 18.81 10.69 31.42 21.71 96.17 77.09 351.47 117.03 .57 4.29 Mangels digested Per cent, digested, 267.25 85.43 8.12 41.31 9.71 48.36 19.08 89.58 234.44 93.40 - Sheep VI. Amount consumed as above. Minus 173.44 grams feces excreted. 668.52 164.54 42:88 21.90 53.48 19.12 136.36 41.10 425.67 73.52 10.13 8.90 Amount consumed Minus hay digested, 503.98 231.19 20.98 10.69 34.36 21.71 95.26 77.09 352.15 117.03 1.23 4.29 Mangels digested, Per cent, digested, 272.79 87.20 10.29 52.36 12.65 63.00 18.17 85.31 235.12 93.67 : Average per cent, digested, . 81.32 46.84 55.68 87.45 93.58 - Series XVIII., Turnips (Swedish), Period 7. Sheep V. 400 grams English hay fed, . 1,600 grams turnips fed, Amount consumed , . . . Minus 156.48 grams feces excreted. Amount digested, .... Minus hay digested, Turnips digested, ... Per cent, digested, 356.28 23.12 33.21 111.41 179.03 9.51 220.64 16.17 21.14 24.25 157.34 1.74 576.92 39.29 54.35 135.66 336.37 11.25 149.45 19.70 17.93 39.74 66.19 5.89 427.47 19.59 36.42 95.92 270.18 5.36 231.58 10.64 21.59 74.64 119.95 4.37 195.89 8.95 14.83 21.28 150.23 .99 88.78 55.34 70.15 87.75 95.48 56.90 268 MASS. EXPERIMENT STATION BULLETIN 181. Table V. — Computation of Digestion Coefficients — Continued. Series XVIII., Turnips (Swedish), Period 7 — Concluded. Sheep VI. Item. < fe 2 1 f2 Amount consuraed as above, Minus 155.75 grams feces excreted. 576.92 148.59 39.29 20.34 54.35 15.62 135.66 42.70 336.37 64.37 11.25 5.56 Amount digested, Minus hay digested, 428.33 231.58 18.95 10.64 38.73 21.59 92.96 74.64 272 .00 119.95 5.69 4 37 Turnips digested, Percent, digested, 196.75 89.17 8.31 51.38 17.14 81.08 18.32 75.55 152.05 96.64 1.32 75.86 Average per cent, digested, . 88.98 53.36 75.62 81.65 96.06 66.38 Series XIX., English Hay and Gluten Feed, Sheep V. Gluten Feed, Period 2. 550 grams English hay fed, . 150 grams gluten feed fed, . 484.11 134.33 28.22 1.41 46.09 37.40 152.01 11.75 246.80 77.62 10.99 6.15 Amount consumed, Minus 226.53 grams feces excreted. 618.44 209.47 29.63 19.38 83.49 26.79 163.76 57.86 324.42 97.92 17.14 7.52 Amount digested, .... Minus hay digested, 408.97 285.62 10.25 7.90 56.70 23.97 105.90 94.25 226.50 153.02 9.62 5.16 Gluten feed digested, . Per cent, ration digested, 123.35 66.13 2.35 34.59 32.73 67.91 11.65 64.67 73.48 69.82 4.46 56.13 Per cent, gluten feed digested. 91.80 167.00 87.50 99.00 94.70 72.50 Sheep VI. Amount consumed as above. Minus 223 grams feces excreted, . 618.44 206.48 29.63 21.62 83.49 26.06 163.76 54.70 324.42 96.65 17.14 7.45 Amount digested Minus hay digested, 411.96 285.62 8.01 7.90 57.43 23.97 109 06 94.25 227.77 153.02 9.69 5.16 Gluten feed digested, Per cent, ration digested, .... 126.34 66.61 .11 27.03 33.46 68.79 14.81 66.60 74.75 70.21 4.53 56.53 Per cent, gluten feed digested, 94.70 .78 89.60 126.00 96.30 73.60 Average per cent, gluten feed digested, . 93.25 83.89 88.50 112.50 95.50 73.05 Average per cent, ration digested, . 66.37 30.81 68.35 65.64 70.02 56 33 DIGESTION EXPERIMENTS WITH SHEEP. 269 Table V. — Computation of Digestion Coefficients — Continued. Series XIX., English Hay, Period 3. Sheep I. Item. S Q < a 1 1 • II u £ 800 grams English hay fed, . Minus 296.19 grams feces excreted, 696.00 279.99 40.79 32.39 65.77 33.49 215.34 75.71 358.51 128.80 15.59 9.60 English hay digested, Per cent, digested, 416.01 59.77 8.40 20.59 32.28 49.08 139.63 64 84 229.71 64.07 5.99 38.42 SCO grams English hay fed Minus 315.10 grams feces excreted. 696.00 294.78 40.79 31.04 65.77 30.83 215.34 88.43 358.51 134.84 15.59 9.64 English hay digested, Per cent, digested, 401.22 57,65 9.75 23.90 34.94 53.12 126.91 58.93 223.67 62.39 5.95 38.17 Average per cent, digested, . 58.71 22.25 51.10 61.89 63.23 38.30 Series XIX., Pumpkins (Seeds removed), Period 4. Sheep I. 500 grams English hay fed, . 2,000 grams pumpkins fed, . 437.35 108.40 25.85 9.55 41.20 14.89 137.68 18.79 222.04 62.39 10.58 2.78 Amount consumed, Minus 180.47 grams feces excreted 545.75 169.30 35.40 19.67 56.09 21.30 156.47 45.27 284.43 76.54 13.36 6.52 Amount digested, . Minus hay digested, 376.45 258.04 15.73 5.69 34.79 21.01 111.20 85.36 207.89 139.89 6.84 4.02 Pumpkins digested, Per cent, digested. 118.41 109.23 10.04 105.13 13.78 92.55 25.84 137.52 68.00 108.99 2.82 101.44 Sheep II. Amount consumed as above, Minus 198.05 grams feces excreted. 545.75 185.99 35.40 24.03 56.09 21.09 156.47 53.23 284.43 80.63 13.36 7.01 Amount digested Minus hay digested, 359.76 258.04 11.37 5.69 35.00 21.01 103.24 85.36 203.80 139.89 6.35 4.02 Pumpkins digested Per cent, digested, 101.72 93.84 5.68 59.48 13.99 93.96 17.88 95.16 63.91 102.44 2.33 83.81 Average per cent, digested, . 101.54 82.31 93.26 116.34 105.72 92.63 270 MASS. EXPERIMENT STATION BULLETIN 181. Table V. — Computation of Digestiox Coefficients — Continued. , Series XIX., Vegetable Ivory Meal, Period 5. Sheep V. Item. 2 < a S J s a o .11 ^ 550 grams English hay fed 150 grams gluten feed fed 200 grams vegetable ivory meal fed, 484.28 134.90 174.72 28.14 1.50 2.39 45.67 36.37 10.52 153.52 11.74 12.27 245.91 79.97 148.33 11.04 5.32 1.21 Amount consumed, Minus 252.10 grams feces excreted, 793.90 237.73 32.03 21.78 92.56 37.09 177.53 61.71 474.21 107.74 17.57 9.41 Amount digested, Minus English hay and gluten feed digested. 556.17 408.66 10.25 9.19 55.47 55.79 115.82 109.07 366.47 228.12 8.16 9.16 Vegetable ivory meal digested, . Per cent, digested, 147.51 84.43 1.06 44.35 - 6.75 55.01 138.35 93.27 - Sheep VI. Amount consumed as above. Minus 241.94 grams feces e.xcreted. 793.90 228.63 32.03 22.41 92.56 33.61 177.53 57.93 474.21 106.82 17.57 7.86 Amount digested, Minus English hay and gluten feed digested. 565.27 408.66 9.62 9.19 58.95 55.79 119.60 109.07 367.39 9.71 9.16 Vegetable ivory meal digested, . Per cent, digested, 156.61 89.63 .43 17.99 3.16 30.04 10.53 85.82 139.27 93.89 .55 45.45 Average per cent, digested, . 87.03 31.17 30.04 70.42 93.58 45.451 Series XIX., Pumpkins (Entire), Period 6. Sheep I. 550 grams English hay fed, . 2,000 grams pumpkins fed, . 483.29 176.20 26.82 13.39 43.50 31.22 151.46 29.71 248.51 76.01 13.00 25.87 Amount consumed, . . . • Minus 253.33 grams feces excreted, 659.49 240.66 40.21 25.63 74.72 30.52 181.17 69.51 324.52 105.98 38.87 9.02 Amount digested Minus hay digested. 418.83 285.14 14.58 5.90 44.20 22.19 111.66 93.91 218.54 156.56 29.85 4.94 Pumpkins digested. Per cent, digested. 133.69 75.87 8.68 64.82 22.01 70.50 17.75 59.74 61.98 81.54 24.91 96.29 One sheep only. DIGESTION EXPERIMENTS WITH SHEEP. 271 Table V. — Computation of Digestion Coefficients — Continned. Series XIX., Pumpkins (Entire), Period 6 — Concluded. Sheep II. as Amount consumed as above, Minus 228.69 grams feces excreted Amount digested, . Minus hay digested, Pumpkins digested, Fer cent, digested, . Average per cent, digested. 659.49 216.96 40.21 25.75 74.72 27.34 442.53 285.14 14.46 5.90 47.38 22.19 181.17 61.62 119.55 231.14 156.56 157.39 89.32 25.64 86.30 74.58 98.12 38.87 8.87 30.00 4.94 Series XIX., Cabbage (Whole), Period 7. Sheep I. 450 grams English hay fed, . 1,600 grams cabbage fed, 398.57 187.68 22.16 22.90 32.48 40.95 126.86 19.33 207.42 100.90 9.65 3.60 Amount consumed. Minus 192.77 grams feces excreted 586.25 183.40 45.06 26.50 73.43 22.23 146.19 46.36 308.32 81.29 13.25 7.02 Amount digested, . Minus hay digested. 402.85 235.16 18.56 51.20 16.56 99.83 78.65 227.03 130.67 6.23 3.67 Cabbage digested, Per cent, digested, 167.69 89.35 13.68 59.74 34.64 84.59 21.18 109.57 96.36 95.50 2.56 71.11 Sheep II. Amount consumed as above. Minus 200.09 grams feces excreted, 586.25 189.95 45.06 27.77 73.43 20.97 146.19 53.53 308.32 80.56 13.25 7.12 Amount digested Minus hay digested 396.30 233.98 17.29 52.46 16.56 92.66 78.65 227.76 130.67 6.13 3.67 Cabbage digested Per cent, digested, 162.32 86.49 12.41 54.19 35.90 87.67 14.01 72.48 97.09 96.22 2.46 68.33 Average per cent, digested, . 87 .,92 56.97 86.13 91.03 95.86 69.72 272 MASS. EXPERIMENT STATION BULLETIN 181. Table V. — Computation of Digestion Coefficients — Continued. Series XIX., Carrots, Period 8. Sheep I. Item. 1 >- Q < £ s ti .11 ^s 500 grams English hay fed, . 1,500 grams carrots fed. 443.75 188.10 25.43 16.10 41.53 30.55 36.08 15.05 153.80 15.52 218.41 139.27 10.03 2.16 Amount consumed, Minus 211.46 grams feces excreted. 631.85 202.45 51.13 24.90 169.32 53.49 357.68 86.85 12.19 6.66 Amount digested Minus hay digested, 449.40 261.81 10.98 5.59 26.23 18.40 115.83 95.36 270.83 137.60 5.53 3.81 Carrots digested Per cent, digested. 167.59 89.10 5.39 33.48 7.83 52.03 20.47 131.89 133.23 95.66 1.72 79.63 Sheep II. Amount consumed as above, Minus 194.84 grams feces excreted, 631.85 186.79 41.53 28.47 51.13 21.01 169.32 49.97 357.68 80.93 12.19 6.41 Amount digested Minus hay digested 445.06 261.81 13.06 5.59 30.12 18.40 119.35 95.36 276.75 137.60 5.78 3.81 Carrots digested Per cent, digested 183.25 94.42 7.47 46.40 11.72 77.87 23.99 154.57 139.15 99.91 1.97 91.20 Average per cent, digested, . 93.26 39.94 64.95 143.23 97.79 85.42 Series XIX., English H.\y, Period Sheep V. 800 grams English hay fed, .... Minus 294.67 grams feces excreted, 710.00 279.97 40.75 27.10 60.49 28.11 221.66 81.30 363.53 133.44 23.57 10.02 Hay digested, Per cent, digested, 4,30.03 60.57 13.65 33.50 32.38 53.53 140.36 63.32 230.09 63.29 13.55 57.49 Sheep VI. 800 grams English hay fed Minus 317.10 grams feces excreted. 710.00 301.56 40.75 27.38 60.49 29.58 221.66 91.64 363.53 142.35 23.57 10.61 Hay digested Per cent, digested, 408.44 57.53 13.37 32.81 30.91 51.10 130.02 58.66 221.18 60.84 12.96 54.99 Average per cent, digested, . 59.05 33.16 52.32 60.99 62.07 56.24 DIGESTION EXPERIMENTS WITH SHEEP. 273 Table V. — Computation of Digestion Coefficients — Continued. Series XIX., Exglish Hay, Potato Starch and Gluten Meal (Diamond), — Gluten Meal (Diamond), Period 10. Sheep III. Item. 2 P < 1 1 300 grams English hay fed 125 grams potato starch fed, 100 grams gluten meal (Diamond) fed. 270.81 113.23 94.40 15.60 .81 20.5. 42.47 86.85 2.04 141.55 113.23 47.37 6.31 1.71 Amount consumed Minus 132.41 grams feces excreted. 478.44 127.01 16.41 10.88 62.97 16.70 88.89 35.89 302.15 58.27 8.02 5.27 Amount digested Minus hay and starch (100 per cent.) digested. 351.43 273.01 5.53 4.37 46.27 10.66 53.00 53.85 243.88 200.99 2.75 2.96 Gluten meal (Diamond) digested. Per cent, ration digested, .... 78.42 73.45 1.16 33.70 35.61 73.48 59.62 80.71 34.29 Per cent, gluten meal (Diamond) digested, . 83.07 143.20 83.85 - 90.54 - Sheep IV. Amount consumed as above, Minus 147.44 grams feces excreted, 478.44 141.26 16.41 14.24 62.97 18.25 88.89 39.64 302.15 63.73 8.02 5.40 Amount digested Minus hay and starch (100 per cent.) digested. 337.18 273.01 2.17 4.37 44.72 10.66 49.25 53.85 238.42 200.99 2.62 2.75 Gluten meal (Diamond) digested. Per cent, ration digested 64.17 70.47 13.22 34.06 71.02 55.41 37.43 78.91 32.67 Per cent, gluten meal (Diamond) digested, . 68.00 - 80.00 - 79.00 - Average per cent, ration digested, . 71.96 23.46 72.25 57.52 79.81 33.48 Average per cent, gluten meal (Diamond) digested. 75.54 71.60 81.93 - 84.77 - 274 MASS. EXPERIMENT STATION BULLETIN 181, Table V. — Computation of Digestion Coefficients — Continued. Sebies XIX., English Hay, Potato Starch and Gluten Meal (Diamond), — Gluten Meal (Diamond), Period 11. Sheep IV. Item. 1 Q < a i II ^ fe 400 grams English hay fed 125 grams potato starch fed, 125 grams gluten meal (Diamond) fed, 369.40 112.80 117.25 20.91 1.34 25.93 52.52 118.47 2.16 195.37 112.80 58.86 8.72 2.37 Amount consumed Minus 174.32 grams feces excreted. 599.45 166.84 22.25 14.70 78.45 19.60 120.63 47.37 367.03 79.30 11.09 5.87 Amount digested, Minus hay and starch (100 per cent.) digested. 432.61 330.75 7.55 5.85 58.85 13.48 73.26 73.45 287.73 233.93 5.22 4.10 Gluten meal (Diamond) digested, Per cent, ration digested 101.86 72.17 1.70 33.93 45.37 75.02 60.73 53.80 78.39 91.40 1.12 47.07 Per cent, gluten meal (Diamond) digested, . 86.90 127.00 86.40 - 47.30 Series XIX., Distillers' Grains (Corn), Period 12. IV. 400'grams English hay fed, .... 359.00 21.72 26.42 115.10 184.99 10.77 125'grams gluten meal fed, .... 117.51 1.57 52.87 2.36 58.44 2.27 125.''grams potato starch fed, 113.49 - - - 113.49 - 200rgrams distillers' grains fed, . 187.72 3.44 55.28 23.69 86.73 18.58 Amount consumed, 777.72 26.73 134.57 141.15 443.65 31.62 Minus 240.66 grams feces excreted, 229.42 20.35 31.20 65.64 104.06 8.17 Amount digested, 548.30 6.38 103.37 75.51 339.59 23.45 Minus hay, potato starch and gluten meal digested. 424.80 7.92 59.44 71.65 278.40 6.13 Distillers' grains digested 123.50 - 43.93 3.86 61.19 17.32 Per. cent, digested 65.79 - 79.47 16.29 70.55 93.22 DIGESTION EXPERIMENTS WITH SHEEP. 275 Table V. — Computation of Digestion Coefficients — Continued. Series XIX., Corn Bran, Period 13. Item. 1 Q 1 a 1 2 400 grams English hay fed, . 350 grams corn bran fed, 363.48 316.47 22.28 2.72 25.52 16.52 119.33 45.89 188.21 247.32 8.14 4.02 Amount consumed. Minus 191.02 grams feces excreted. 679.95 180.42 25.00 14.38 42.04 20.82 165.22 45.39 435.53 93.77 12.16 6.06 Amount digested, .... Minus hay digested. 499.53 214.45 10.62 4.90 21.22 13.02 119.83 73.98 341.76 118.57 6.10 3.09 Corn bran digested. Per cent, digested. 285.08 90.08 5.72 210.29 8.20 49.64 45.85 99.91 223.19 90.24 3.01 74.88 Sheep II. Amount consumed as above. Minus 234.41 grams feces excreted, 679.95 221.52 25.00 16.59 42.04 24.72 165.22 60.54 435.53 112.43 12.16 7.24 Amount digested, Minus hay digested, 458.43 214.45 8.41 4.90 17.32 13.02 104.68 73.98 323.10 118.57 4.92 3.09 Corn bran digested Per cent, digested 243.98 77.09 3.51 129.04 4.30 26.03 30.70 66.90 204.53 82.70 1.83 45.52 Average per cent, digested, . 83.59 169.67 37.84 83.41 86.47 60.20 Series XIX., Gluten Feed, Period 14. Sheep V. 650 grams English hay fed, . 150 grams gluten feed fed, . 584.68 135.78 35.14 1.29 48.47 37.92 184.47 12.65 300.93 77.21 15.67 6.71 Amount consumed. Minus 237.30 grams feces e.xcreted, 720.46 223.20 36.43 20.47 86.39 29.04 197.12 58.17 378.14 105.94 22.38 9.58 Amount digested, .... Minus hay digested. 497.26 344.96 15.96 11.60 57.35 25.20 138.95 112.53 272.20 186.58 12.80 8.78 Gluten feed digested, . Per cent, digested. 152.30 112.09 4.36 337.98 32.15 84.78 26.42 208.85 85.62 110.89 4.02 59.91 276 MASS. EXPERIMENT STATION BULLETIN 181. Table V. — Computation of Digestion Coefficients — Continued. Series XIX., Gluten Feed, Period 14 — Concluded. Sheep VI. Amount consumed as above. Minus 262.21 grams feces excreted. Amount digested, . Minus hay digested, Gluten feed digested, Per cent, digested, Average per cent, digested. 720.46 247.05 36.43 21.49 197.12 69.87 378.14 118.50 473.41 344.96 14.94 11.60 59.21 25.20 127.25 112.53 128.45 94.54 14.72 116.36 73.06 94.63 22.38 10.01 12.37 8.78 Series XIX., English Hay and Gluten Feed, — Gluten Feed, Period 15. Sheep V. 650 grams English hay fed, . 125 grams gluten feed fed, . 587.73 113.59 33.62 2.61 44.26 29.35 185.02 9.28 311.90 66.97 12.93 Amount consumed. Minus 246.94 grams feces excreted. 701.32 233.51 36.23 20.88 73.61 26.34 194.30 65.36 378.87 112.55 18.31 8.38 Amount digested Minus hay digested. 467.81 346.76 15.35 11.09 47.27 23.02 128.94 112.86 266.32 193.38 9.93 7.24 Gluten feed digested, . Per cent, ration digested. 121.05 66.70 4.26 42.45 24.25 64.22 16.08 66.36 72.94 70.29 2.69 54.23 Per cent, gluten feed digested. 106.57 163.22 82.62 173.28 108.91 50.00 Sheep VI. Amount consumed as above. Minus 280.06 grams feces excreted 701.32 264.94 36.23 21.75 73.61 29.67 194.30 77.04 378.87 127.00 18.31 9.48 Amount digested, . Minus hay digested. 436.38 346.76 14.48 11.09 43.94 23.02 117.26 112.86 251.87 193.38 8.83 7.24 Gluten feed digested, . Per cent, ration digested. 89.62 62.22 3.39 39.97 20.92 59.69 4.40 60.35 58.49 66.48 1.59 48.23 Per cent, gluten feed digested. 78.90 129.89 71.28 47.41 87.34 29.55 Average per cent, ration digested, 64.46 41.21 61.96 63.36 68.39 51.23 Average per cent, gluten feed diges ted, . 92.74 146.56 76.95 110.35 98.13 39.78 DIGESTION EXPERIMENTS WITH SHEEP. 277 Table V. — Computation of Digestion Coefficients — Continued. Series XX., English Hay, Period 1. Sheep I. Item. Q < 1 1 s 2 800 grams English hay fed Minus 281.27 grams feces excreted, 737.36 260.32 44.32 26.73 53.24 26.16 239.13 70.75 381.94 126.89 18.73 9.79 English hay digested Per cent, digested, 477.04 64.70 17.59 39.69 27.08 50.86 168.38 70.41 255.05 66.78 8.94 48.67 Sheep II. 800 grams English hay fed Minus 291.37 grams feces excreted, 737.36 269.63 44.32 27.93 53.24 , 26.42 239.13 75.74 381.94 129.62 18.73 9.92 English hay digested Per cent, digested 467.73 63.43 16.39 36.98 26.82 50.38 163.39 68.33 252.32 66.06 8.81 47.96 Average per cent, digested, . 64.07 38.34 50.62 69.62 66.42 48.32 Series XX., Pumpkins (Entire), Period 2. Sheep I. 500 grams English hay fed, . 2,000 grams pumpkins fed, . Amount consumed. Minus 223.14 grams feces excreted Amount digested, . Minus hay digested. Pumpkins digested. Per cent, digested. 449.50 268.40 717.90 211.14 27.96 20.94 25.48 10.62 14.86 70.96 145.10 34.92 180.02 57.13 122.89 100.12 22.77 65.20 377.25 93.09 284.16 152.87 43.04 10.47 32.57 6.79 25. Sheep II. Amount consumed as above, Minus 203.70 grams feces excreted. 717.90 193.41 48.90 25.09 68.69 23.87 180.02 50.71 377.25 83.99 43.04 9.75 Amount digested, Minus hay digested 524.49 287.68 23.81 10.62 44.82 15.84 129.31 100.12 293.26 152.87 33.29 6.79 Pumpkins digested Per cent, digested, 236.81 88.23 84.93 13.19 62.99 28.98 76.20 29.19 83.59 140.39 96.40 26.50 91.76 Average per cent, digested, . 66.98 72.05 74.39 93.62 90.52 278 MASS. EXPERIMENT STATION BULLETIN 181. Table V. — Computation of Digestion Coefficients — Continued. Series XX., Pumpkins (Entire), Period 3. Sheep I. Item. - 1 < S ll .11 ^ 550 grams English hay fed, . 150 grams gluten feed fed, . 1,200 grams piimpkins fed, . 495.00 136.65 191.04 30.15 2.51 13.87 36.23 34.91 29.94 159.09 11.30 27.49 256.96 81.58 98.90 12.57 6.35 20.84 Amount consumed. Minus 248.90 grams feces excreted, 822.69 236.31 46.53 25.62 101.08 30.51 197.88 62.06 437.44 108.34 39.76 9.78 Amount digested Minus hay and gluten feed digested. 586.38 435.84 20.91 11.43 70.57 45.53 135.82 122.68 329.10 243.75 29.98 11.62 Pumpkins digested. Per cent, digested. 150.54 78.80 9.48 68.35 25.04 83.63 13.14 47.80 85.35 86.30 18.36 88.10 Sheep II. Amount consumed as above, Minus 255.64 grams feces excreted, 822.69 242.78 46.53 28.19 101.08 30.83 197.88 62.49 437.44 110.78 39.76 10.49 Amount digested Minus hay and gluten feed digested, 579.91 435.84 18.34 11.43 70.25 45.53 135.39 122.68 326.66 243.75 29.27 11.62 Pumpkins digested, Per cent, digested, 144.07 75.41 6.91 49.82 24.72 82.57 12.71 46.23 82.91 83.83 17.65 84.69 Average per cent, digested, . 77.11 59.09 83.10 47.02 85.07 86.40 Series XX., Pumpkins (Entire), Period 4. Sheep I. 412 grams English hay fed, . 112 grams gluten feed fed, . 2,000 grams pumpkins fed, . 373.07 102.73 234.40 22.72 2.09 19.48 27.42 26.56 34.81 120.99 8.49 36.87 190.86 113.24 11.08 4.77 30.00 Amount consiuned. Minus 216.87 grams feces excreted. 710.20 204.77 44.29 20.62 88.79 28.20 166.35 58.93 364.92 88.79 45.85 8.23 Amount digested, .... Minus hay and gluten feed digested. 505.43 328.30 23.67 60.59 34.55 107.42 93.23 276.13 181.21 37.62 9.35 Pumpkins digested. Per cent, digested, 177.13 75.57 14.99 76,95 26.04 74.81 14.19 38.49 94.92 83.8. 28.27 94.23 DIGESTION EXPERIMENTS WITH SHEEP. 279 Table V. — Computation of Digestion Coefficients — Continued. Series XX., English Hay and Gluten Feed, — Gluten Feed, Period 5. Sheep I. Item. 1 Q 1 = o s 1 It as .11 1 550 grams English hay fed, . 150 grams gluten feed fed, . 497.86 136.71 29.92 2.67 43.21 36.54 159.22 11.39 251.62 79.82 13.89 6.29 Amount consumed. Minus 197.54 grams feces excreted 634.57 188.22 32.59 19.20 79.75 26.76 170.61 44.16 331.44 90.14 20.18 7.96 Amount digested, . Minus hay digested. 446.35 318.63 13.39 11.87 52.99 22.04 126.45 109.86 241.30 166.07 12.22 6.71 Gluten feed digested, . Per cent, ration digested, 127.72 70.34 2.02 41.09 30.95 66.45 16.59 74.12 75.23 72.80 5.51 60.56 Per cent, gluten feed digested, 93.42 75.66 84.70 145.65 94.25 87.60 Sheep II. Amount consumed as above. Minus 221.18 grams feces excreted ed, . 634.57 211.36 32.59 23.19 79.75 30.77 170.61 51.87 331.44 97.12 20.18 8.41 Amount digested, . Minus hay digested. 423.21 318.63 9.40 11.37 48.98 22.04 118.74 109.86 234.32 166.07 11.77 6.71 Gluten feed digested, . Per cent, ration digested. 104.58 66.69 28.84 26.94 61.42 8.88 69.60 68.25 70.70 5.06 58.33 Per cent, gluten feed digested. 76.50 - 73.73 77.96 85.50 80.45 Average per cent, ration digested. 68.52 34.97 63.94 71.86 71.75 59.45 Average per cent, gluten feed diges 84.96 75.66' 79.22 111.81 89.88 84.03 Series XX. , English H Sheep IV. AY, Pe RIOD 6. 800 grams English hay fed, . Minus 300.50 grams feces excreted. 717.84 44.22 25.63 53.41 27.59 230.28 84.97 272.99 139.85 16.94 10.29 Amount digested Per cent, digested. 429.51 60.08 18.59 42.04 25.82 48.34 145.31 63.10 233.14 62.51 6.65 39.26 1 Cm sheep or Jy. 280 MASS. EXPERIMENT STATION BULLETIN 181. Table V. — Computation of Digestion Coefficients — Continued. Series XX., Soy Bean Hay, Period 7. Sheep V. Item. >> a < 1 1 it II 400 grams English hay fed, . Minus 29.03 grams waste. ■ 367.68 27.24 22.39 1.64 26.95 1.74 120.19 9.57 189.51 13.86 8.64 .43 Amount consumed, 340.44 20.75 25.21 110.62 175.65 8.21 400 grams soy bean hay fed, Minus 55.1 grams waste. 353.08 51.00 23.41 2.77 56.00 3.76 123.15 22.93 143.21 21.13 7.31 .41 Amount soy bean hay fed, . 302.08 20.64 52.24 100.22 122.08 6.90 Amount consumed, ^ . Minus 284.68 grams feces excreted, 642.52 270.13 41.39 30.69 77.45 27.80 210.84 87.25 297.73 116.69 15.11 7.70 Amount digested, .... Minus hay digested. 372.39 214.48 10.70 8.30 49.65 12.61 123.59 74.12 181.04 114.17 7.41 3.69 Amount soy bean hay digested, . Per cent, digested. 157.91 52.27 2.40 11.63 37.04 70.90 49.47 49.36 66.87 54.78 3.72 53.91 Sheep VI. 400 grams English hay fed, . 400 grams soy bean hay fed, 367.68 353.08 22.39 23.41 26.95 56.00 120.19 123.15 189.51 143.21 8.64 7.31 Amount consumed, Minus 287.90 grams feces excreted, 720.76 273.65 45.80 29.99 82.95 25.31 243.34 94.16 332.72 116.86 15.95 7.33 Amoxint digested, .... Minus hay digested. 447.11 231.64 15.81 8.96 57.64 13.48 149.18 80.53 215.86 123.18 8.62 Soy bean hay digested. Per cent, digested, 215.47 61.03 6.85 29.26 44.16 78.86 68.65 55.75 92.68 64.72 4.73 64.71 56.65 20.45 74.88 52.56 59.75 59.31 Series XX., Carrots, Period Sheep IV. 500 grams English hay fed, .... 1,500 grams carrots fed, .... 449.85 196.95 30.23 20.31 35.72 22.12 146.34 17.39 225.95 135.12 11.61 2.01 Amount consumed Minus 228.61 grams feces excreted. 646.80 216.81 50.54 28.25 57.84 22.79 163.73 58.69 361.07 98.39 13.62 8.69 Amount digested Minus hay digestetl 429.99 283.41 22.29 12.09 35.05 17.86 105.04 98.05 262.68 146.87 4.93 5.22 Carrots digested Per cent, digested, 146.58 74.42 10.20 50.22 17.19 77.71 6.99 40.19 115.81 85.71 - DIGESTION EXPERIMENTS WITH SHEEP. 281 Table V. — Computation of Digestion Coefficients — Continued. Series XX., Carrots, Period S — Concluded. V. Item. 1 < c s t .11 fc Amount consumed as above, Minus 200.63 grams feces excreted, 646.80 190.44 50.54 25.27 57.84 21.10 163.73 47.95 361.07 88.24 13.62 Amount digested, Minus hay digested, 456.36 283.41 25.27 12.09 36.74 17.86 115.78 98.05 272.83 146.87 5.74 5.22 Carrots digested Per cent, digested 172.95 87.81 13.18 64.89 18.88 85.35 17.73 101.96 125.96 93.22 .52 25.87 Sheep VI. Amount consumed as above. Minus 208.06 grams feces excreted. 646.80 198.49 50.54 29.56 57.84 20.84 163.73 51.41 361.07 88.48 13.62 8.20 Amount digested, Minus hay digested, 448.31 283.41 20.98 12.09 37.00 17.86 112.32 98.0 ; 272.59 146.87 5.42 5.22 Carrots digested Per cent, digested 164.90 83.73 8.89 43.77 19.14 86.53 14.27 82.06 125.72 93.04 .20 9.95 Average per cent, digested, . 81.99 52.96 83.20 74.74 90.66 17. 9U Series XX., Carrots, Period Sheep IV. 550 grams English hay fed, . 150 grams gluten feed fed, . 1,000 grams carrots fed. 497.04 137.03 114.00 30.97 3.04 11.18 37.78 36.41 12.67 151.30 10.63 263.72 80.36 79.22 13.27 6.59 1.24 Amount consumed. Minus 250.91 grams feces excreted, 748.07 240.07 45.19 27.03 86.86 28.28 171.62 59.30 423.30 116.31 21.10 9.15 Amount digested, .... Minus hay and gluten feed digested. 508.00 393.12 18.16 9.86 58.58 47.48 112.32 103.64 306.99 223.65 11.95 9.93 Carrots digested Per cent, digested, 100.70 8.30 74.24 11.10 87.61 8.68 89.58 83.34 105.20 2.02 162.90 Two sheep only. 282 MASS. EXPERIMENT STATION BULLETIN 181. Table V. — Computation of Digestion Coefficients • — Continued. Series XX., Carrots, Period 9 — Concluded. Sheep V. Item. - S 1 1 £ Amount consumed as above, Minus 233.76 grams feces excreted, 748.07 222.94 45.19 25.06 86.86 25.95 171.62 53.57 423.30 109.73 21.10 8.63 Amount digested, Minus hay and gluten feed digested, . 525.13 393.12 20.13 9.86 60.91 47.48 118.05 103.64 313.57 223.65 12.47 9.93 Carrots digested Per cent, digested 132.01 115.80 10.27 91.86 13.43 106.00 14.41 148.71 89.92 113.51 2.54 204.84 Sheep VI. Amount consumed as above. Minus 210.84 grams feces excreted, Amount digested Minus hay and gluten feed digested, . 748.07 200.99 547.08 393.12 45.19 24.58 20.61 9.86 86.86 23.17 63.69 47.48 171.62 48.84 122.78 103.64 423.30 96.06 327.24 223.65 21.10 8.34 12.76 9.93 Carrots digested Per cent, digested, 153.96 135.05 10.75 96.15 16.21 127.94 19.14 197.52 103.59 130.76 2.83 228.23 Average per cent, digested, . 113.85 87.42 107.18 145.27 116.49 198.66 Series XX., English Hay, Period 10. Sheep VII. 600 grams English hay fed UinuB 244.84 grams feces excreted, 546.00 235.85 35.65 20.42 44.50 21.25 174.83 70.68 277.37 115.74 13.65 7.76 Amount digested Per cent, digested, 310.15 56.80 15.23 42.72 23.25 52.25 104.15 59.57 161.63 58.27 5.89 43.15 Sheep VIII. 600 grams Enghsh hay fed Minus 228.90 grams feces excreted. 546.00 220.06 35.65 19.89 44.50 21.04 174.83 65.67 277.37 105.52 13.65 7.94 Amount digested Per cent, digested 325.94 59.70 15.76 44.21 23.46 52.72 109.16 62.44 171.85 61.96 5.71 41.83 Average per cent, digested, . 58.25 43.47 52.49 61.01 60.12 42.49 DIGESTION EXPERIMENTS WITH SHEEP. 283 Table V. — Computation of Digestion Coefficients — Continued. Series XX., New Bedford Pig Meal, Period 11. Sheep IV. Item. 1 1 C J3 < .11 pS 550 grams English hay fed, . 150 grams gluten feed fed, . 200 grams New Bedford pig meal fed. 455.35 137.18 182.40 30.96 2.96 35.84 39.02 34.94 43.03 146.35 11.48 16.69 227.64 81.46 80.80 11.38 6.34 6.04 -Amount consumed, Minus 295.87 grams feces excreted. 774.93 281.73 69.76 42.71 116.99 40.68 174.52 70.43 389.90 121.09 23.76 6.82 Amount digested Minus hay and gluten feed digested. 493.20 367.37 27.05 9.84 76.31 47.33 104.09 101.01 268.81 200.92 16.94 8.86 New Bedford pig meal digested, . Per cent, digested. 125.83 68.99 17.21 48.02 28.98 67.35 3.08 18.45 67.89 84.02 8.08 133.77 Sheep VI. Amount consumed as above, Minus 295.20 grams feces excreted. 774.93 281.18 69.76 45.24 116.99 38.94 174.52 69.14 389.90 121.59 23.76 6.27 Amount digested Minus hay and gluten feed digested. 493.75 367.37 24.52 9.84 78.05 47.33 105.38 101.01 268.31 200.92 17.49 8.86 New Bedford pig meal digested, . Per cent, digested, 126.38 69.29 14.68 40.96 30.72 ' 71.39 4.37 26.18 67.39 83.40 8.63 142.88 Average per cent, digested, . 69.14 44.49 69.37 22.32 83.71 138.33 Series XX., English Hay and Wheat Gluten Flour, Period i; Sheep VII. 600 grams English hay fed, .... 554.70 37.00 39.77 173.73 289.00 15.20 40 grams wheat gluten fed 36.66 .32 33.88 .04 2.28 .14 .\mount consumed 591.36 37.32 73.65 173.77 291.28 15.34 Minus 238.80 grams feces excreted. 227.70 19.17 22.72 66.08 111.53 8.20 Amount digested 363.66 18.15 50.93 107.69 179.75 7.14 Minus wheat gluten (assumed to be all di- gested). 36.66 .32 33.88 .04 2.28 .14 Hay digested 327.00 17.83 17.05 107.65 177.47 7.00 Per cent, digested, . . . . 58.95 48.19 42.87 61.96 61.44 46.05 To note effect of wheat gluten flour. 284 MASS. EXPERIMENT STATION BULLETIN 181. Table V. — Computation of Digestion Coefficients — Continued. Series XX., English Hay and Wheat Gluten Flour, Period 12 — Concluded. Sheep VIII. Item. 1 < 1 1 E II z 1 Amount consumed as above, Minus 235.59 grams feces excreted, 591.36 224.59 37.32 19.81 73.65 23.96 173.77 62.80 291.38 110.11 15.34 7.91 Amount digested Minus wheat gluten (assumed to be all di- gested). 366.77 36.66 17.51 .32 49.69 33.88 110.97 .04 181.17 2.28 7.43 .14 Hay digested Per cent, digested, 330.11 59.51 17.19 46.46 15.81 39.75 110.93 63.85 178.89 61.90 7.29 47.96 Average per cent, digested, . 59.23 47.33 41.31 62.91 61.99 47.01 Series XX., Vegetable Ivory Meal, Period 13. Sheep IV. 500 grams English hay fed, . 150 grams gluten feed fed, . 200 grams vegetable ivory meal fed. 460.10 137.70 182.50 29.45 2.96 2.17 33.59 35.58 8.61 148.24 11.31 15.09 237.46 81.72 155.22 11.36 6.13 1.41 Amount consumed. Minus 263.03 grams feces excreted. 780.30 248.69 34.58 23.80 77.78 32.60 174.64 61.05 474.40 121.64 18.90 9.60 Amount digested Minus hay and gluten feed digested, 531.61 370.64 10.78 9.40 45.18 44.27 113.59 102.11 352.76 207.47 9.30 8.75 Vegetable ivory meal digested. Per cent, digested. 160.97 88.20 1.38 63.59 .91 10.57 76.08 145.28 93.60 .55 39.01 Sheep V. Amount consumed as above. 780.30 34.58 77.78 174.64 474.40 18.90 Minus 242.74 grams feces excreted. 229.05 23.14 30.53 54.35 111.73 9.30 Amount digested, 551.25 11.44 47.25 120.29 362.67 9.60 Minus hay and gluten feed digested, . 370.64 9.40 44.27 102.11 207.47 8.75 Vegetable ivory meal digested. 180.61 2.04 2.98 18.18 155.20 .85 Per cent, digested, 98.96 94.01 34.61 120.48 99.99 60.28 DIGESTION EXPERIMENTS WITH SHEEP. 285 Table V. — Computation of Digestion Coefficients — Continued. Series XX., Vegetable Ivory Meal, Period 13 — Concluded. Sheep VI. Item. 1 Q < 1 11 i Amount consumed as above, Minus 237.89 grams feces excreted. 780.30 224.04 34.58 24.22 77.78 28.92 174.64 53.03 474.40 108.17 18.90 9.70 Amount digested, Minus hay and gluten feed digested, . 556.26 370.64 10.36 9.40 48.86 44.27 121.61 102.11 366.23 207.47 9.20 8.75 Vegetable ivory meal digested, Per cent, digested 185.62 101.71 .96 44.24 3.59 41.70 19.50 129.22 158.76 102.28 .45 31.91 Average per cent, digested, . 96.29 67.28 28.96 108.59 98.62 43.73 Series XX., English Hay and Gluten Feed, Sheep IV. •Gluten Feed, Period 14. 550 grams English hay fed, . 150 grams gluten feed fed, . 509.41 138.14 31.94 2.94 35.40 37.45 165.25 11.81 263.63 79.74 13.19 6.20 .\mount consumed. Minus 271 grams feces excreted. 647.55 257.02 34.88 25.42 72.85 26.63 177.06 70.22 343.37 125.24 19.39 9.51 Amount digested, . Minus hay digested. 390.53 320.93 9.46 12.78 46.22 17.70 106.84 110.72 218.13 171.36 5.94 Gluten feed digested. . Per cent, ration digested. 69.60 60.31 27.12 28.52 63.45 60.34 46.77 63.53 3.94 50.95 Per cent, gluten feed digested. 50.38 76.15 - 58.65 63.55 Sheep V. Amount consumed as above, Minus 250.96 grams feces excreted. 647.55 238.11 34.88 24.76 72.85 26.22 177.06 60.22 343.37 116.58 19.39 10.33 .■Vmount digested, Minus hay digested, 409.44 320.93 10.12 12.78 46.63 17.70 116.84 110.72 226.79 171.36 9.06 5.94 Gluten feed digested, Per cent, ration digested 88.51 63.23 29.01 28.93 64.01 6.12 65.99 55.43 66.05 3.12 46.73 Per cent, gluten feed digested, 64.07 - 77.25 51.82 69.51 50.32 286 MASS. EXPERIMENT STATION BULLETIN 181. Table V. — Computation of Digestion Coefficients — Co?i^i«ue(/. Series XX., English Hay and Gluten Feed, — ■ Gluten Feed, — Period 14 — • Concluded. Sheep VI. Item. i Q i c: 1 E 1. II Amount consumed as above. Minus 246.79 grams feces excreted, 647.55 234.80 34.88 23.74 72.85 25.71 177.06 60.34 343.37 115.69 19.39 9.32 Amount digested, .... Minus hay digested. 412.75 320.93 11.14 12.78 47.14 17.70 116.72 110.72 227.68 171.36 10.07 5.94 Gluten feed digested, . Per cent, ration digested. 91.82 63.74 31.94 29.44 64.71 6.00 65.92 56.32 66.31 4.13 51.93 Per cent, gluten feed digested, . 66.47 - 78.61 50.80 70.63 66.01 62.43 29.36 64.06 64.08 65.30 49.87 Average per cent, gluten feed diges ed. 60.31 - 77.34 51.311 66.26 60.16 Series XXI., English Hay and Gluten Feed, Sheep IV. •Gluten Feed, Period 1. 550 grams English hay fed, . 150 grams gluten feed fed, . 491.43 135.48 33.96 4.55 37.10 38.33 157.55 10.07 248.91 80.08 13.91 2.45 Amount consumed. Minus 236.19 grams feces excreted, 626.91 219.61 38.51 26.27 75.43 23.85 167.62 328.99 57.80 j 102.86 16.36 8.83 Amount digested Minus hay digested. 407.30 280.12 12.24 12.90 51.58 15.95 109.82 96.11 226.13 149.35 7.53 5.98 Gluten feed digested, . Per cent, ration digested. 127.18 64.97 31.78 35.63 68.38 13.71 65.52 76.78 68.73 1.55 46.03 Per cent, gluten feed digested, 93.87 - 92.96 136.00 95.88 63.27 Sheep V. 550 grams minus 1.86 grams waste 548.14 grams English hay fed. 150 grams gluten feed fed. . equal 3 489.76 135.48 33.84 4.55 36.98 38.33 157.02 10.07 248.06 80.08 13.86 2.45 Amount consumed, Minus 222.51 grams feces excreted, 625.24 206.82 418.42 279.16 38.39 29.24 75.31 23.70 167.09 52.64 328.14 92.72 16.31 8.52 Amount digested Minus hay digested. 9.15 12.86 51.61 15.90 114.45 95.78 235.42 148.84 7.79 5.96 Gluten feed digested, . Per cent, ration digested. 139.26 66.92 23.83 35.71 68.53 18.67 68.50 86.58 71.74 1.83 47.76 Per cent, gluten feed digested, 102.79 - 93.16 185.00 108.12 74.69 Two sheep only. DIGESTION EXPERIMENTS WITH SHEEP. 287 Table V, — Computation' of Digestion Coefficients - Series XXI., Exglish Hay and Gluten- Feed, — Gluten Feed, Concluded. Sheep VI. Contimied. - Period 1 — Item. 1 1 Q J3 < a 1 &4 1 1 il ^ Amount consumed as for Sheep IV., Minus 211.77 grams feces excreted. 626.91 195.55 38.51 25.99 75.43 22.37 167.62 48.50 328.99 90.38 16.36 8.31 Amount digested, .... Minus hay digested. 4.31.36 280.12 12.52 12.90 53.06 15.95 119.12 96.11 238.61 1 8.05 149.35 5.98 Gluten feed digested, . Per cent, ration digested, 151.24 68.81 32.51 37.11 70.34 23.01 71.07 89.26 2.07 72.53 49.21 Per cent, gluten feed digested. 111.63 - 96.82 228.00 111.46 , 84.49 Average per cent, ration digested. 66.90 29.37 69.08 68.36 71.00 j 47.67 Average per cent, gluten feed diges ed. 102.76 - 94.31 183.00 105.15 74.15 Series XXI., English Hay, Period 2. Sheep VII. 700 grams English hay fed, .... Minus 301.12 grams feces excreted, 623.84 281.40 43.11 30.93 47.10 27.32 200.01 82.25 315.97 130.74 17.65 10.16 Amount digested, Per cent, digested 342.44 54.89 12.18 28.25 19.78 42.00 117.76 58.88 185.23 58.62 7.49 42.44 Sheep VIII. 700 grams English hay fed, .... Minus 256.07 grams feces excreted, 623.84 238.91 43.11 26.88 47.10 25.87 200.01: 63.88 315.97 112.51 17.65 9.77 Amount digested Per cent, digested, ..... 384.93 61.70 16.23 37.65 21.33 45.08 136.13 68.06 203.46 64.39 7.88 44.65 Average per cent, digested, . 58.30 32.95 43.54 1 63.47 61.51 43.55 288 MASS. EXPERIMENT STATION BULLETIN 181, Table V. — Computation of Digestion Coefficients — Co7itin Series XXI., Vegetable Ivory Meal, Period 3. Sheep IV. lied. 550 grams English hay fed 150 grams gluten feed fed 200 grams minus 1 .86 grams waste equals 198.14 grams vegetable ivory meal fed. Amount consumed, Minus 258.16 grams feces excreted. Amount digested, .... Minus hay and gluten feed digested, Vegetable ivory meal digested. Per cent, digested, 486.20 135.66 176.54 798.40 243.01 555.39 416.65 138.74 78.59 32.48 4.73 2.10 39.31 24.45 14.86 10.79 4.07 193.81 36.51 38.05 9.41 83.97 32.37 51.60 51.45 .15 1.59 183.56 61.36 122.20 114.31 51.07 £ 245.05 80.55 146.83 472.43 115.26 357.17 231.18 125.99 85.81 Sheep V. 550 grams minus .7 gram waste equals 549.3 grams English hay fed. 150 grams gluten feed fed, .... 200 grams minus 1.43 grams waste equals 198.57 grams vegetable ivory meal fed. Amount consumed, Minus 253.24 grams feces excreted. Amount digested Minus hay and gluten feed digested. Vegetable ivory meal digested. Per cent, digested, 485.58 135.66 176.93 798.17 238.55 559.62 416.23 143.39 81.04 32.44 4.73 2.11 39. 5.64 10.78 36.47 38.05 9.43 .95 51.87 51.42 .45 4.77 183.39 53.72 129.67 114.18 15.49 100.06 244.72 80.55 147.15 472.42 109.14 363.28 230.94 132.34 89.94 Sheep VI. 550 grams English hay fed 486.20 32.48 36.51 158.21 245.05 13.95 150 grams gluten feed fed 135.66 4.73 38.05 9.90 80.55 2.43 200 grams minus 1.57 grams equals 198.43 grams vegetable ivory meal fed. 176.81 2.10 9.42 15.47 147.06 2.76 Amount consumed 798.67 39.31 83.98 183.58 472.66 19.14 Minus 248.26 grams feces excreted. 2.33.44 26.61 31.96 58.99 106.22 9.66 Amount digested, 565.23 12.70 52.02 124.59 366.44 9.48 Minus hay and gluten feed digested, . 416.65 10.79 51.45 114.31 231.18 7.86 Vegetable ivory meal digested, 148.58 1.91 .57 10.28 135.26 1.62 Per cent, digested 84.03 90.95 6.04 66.45 91.98 58.70 Average per cent, digested, . 81.22 142.38' 4.13 72.53 89.24 55.87 Two sheep only. DIGESTION EXPERIMENTS WITH SHEEP. 289 Table V. — Computation of Digestion Coefficients — Continued. Series XXI., Exglish Hay and Wheat Gluten- Flour, Period 4. i Sheep VII. Item. J -5) c s PL, 1 1 1^ 1 700 grams English hay fed 40 grams wheat gluten fed, .... 621.95 37.18 40.99 .29 47.21 34.51 203.19 .03 312.77 2.20 17.79 .15 Amount consumed, . . . . . Minus 293.60 grams feces excreted. 659.13 278.48 41.28 26.20 81.72 25.84 203.22 84.71 314.97 131.70 17.94 10.03 Amount digested Minus 40 grams wheat gluten (assumed to be all digested). 380.65 37.18 15.08 .29 55.88 34.51 118.51 .03 183.27 2.20 7.91 .15 English hay digested Per cent, digested, 343.47 55.22 14.79 36.08 21.37 45.27 118.48 58.31 181.07 57.89 7.76 43.62 Series XXI., English H.^y, Potato Starch and Gluten Meal (Diamond), — Gluten Meal (Diamond), Period 5. Sheep IV. 300 grams English hay fed 269.61 18.52 20.03 91.05 132.95 7.06 125 grams potato starch fed. 111.95 - - - 111.95 - 100 grams gluten meal (Diamond) fed. 90.97 1.03 41.06 1.87 45.46 1.55 Amount consumed, 472.53 19.55 61.09 92.92 290.36 8.61 Minus 135.77 grams feces excreted. 126.67 15.76 • 17.64 33.21 56.23 5.83 Amount digested, 343.86 3.79 43.45 59.71 234.13 2.78 Minus hay and starch (100 per cent.) digested, 265.62 6.85 8.61 55.54 191.72 3.03 Gluten meal (Diamond) digested. 78.24 - 34.84 4.17 42.41 - Per cent, ration digested, .... 72.77 19.38 71.12 64.26 80.63 32.29 Per cent, gluten meal (Diamond) digested, . 86.00 - 84.80 - 93.30 - Sheep V. Amount consumed as above. 472.53 19.55 61.09 92.92 290.36 8.61 Minus 113.25 grams feces excreted. 107.36 17.71 15.07 23.48 46.02 5.08 Amount digested 365.17 1.84 46.02 69.44 244.34 3.53 Minus hay and starch (100 per cent.) digested. 265.62 6.85 8.61 S5.54 191.72 3.03 Gluten meal (Diamond) digested. 99.55 - 37.41 13.90 52.62 .50 Per cent, ration digested 77.28 9.41 75.33 74.73 84.15 41.00 Per cent, gluten meal (Diamond) digested, . 109.40 - 91.20 - 120.10 32.25 To note effect of wheat gluten flour. 290 MASS. EXPERIMENT STATION BULLETIN 181. Table V. — Computation of Digestion Coefficients — Continued. Series XXI., English Hay, Potato Starch and Gluten Meal (Diamond), - Gluten Meal (Diamond), Period 5 — Concluded. VI. Item. M « S 1 < I 1 II z 1 Amount consumed as above, Minus 132.14 grams feces excreted, 472.53 125.27 19.55 16.95 61.09 17.15 92.92 31.17 290.36 54.46 8.61 5.54 Amount digested Minus hay and starch (100 per cent.) digested. 347.26 265.62 2.60 6.85 43.94 8.61 61.75 55.54 235.90 191.72 3.07 3.03 Gluten meal (Diamond) digested, Per cent, ration digested 81.64 73.49 13.30 35.33 71.93 6.21 66.46 44.18 81.24 35.66 Per cent, gluten meal (Diamond) digested, . 89.70 - 86.00 - 97.20 - Average per cent, ration digested, . 74.51 14.03 72.79 68.48 82.01 36.32 Average per cent, gluten meal (Diamond) digested. 95.03 - 87.33 - 70.20 32.25' Series XXI., Distillers' Grains, Period Sheep IV. 300 grams English hay fed, . 125 grams potato starch fed, 100 grams gluten meal (Diamond) 200 grams distillers' grains fed. Amount consumed. Minus 196.28 grams feces excreted Amount digested, . Minus basal ration digested, Distillers' grains digested. Per cent, digested, 112.84 91.06 193.74 480.60 354.86 125.74 64.88 1.06 4.48 23.45 17.90 41.09 51.15 112.35 28.15 84.20 44.68 39.52 77.26 2.03 28.50 119.95 45.25 12.51 43.89 Sheep V. Amount consumed as above. 666.89 23.45 112.35 119.95 384.73 26.41 Minus 190.84 grams feces excreted. 181.18 19.84 29.70 39.90 83.30 8.44 Amount digested 485.71 3.61 82.65 80.05 301.43 17.97 Minus basal ration digested. 354.86 2.66 44.68 62.19 240.53 2.95 Distillers' grains digested 130.85 .95 37.97 17.86 60.90 15.02 Per cent, digested, 67.54 21.21 74.23 62.67 66.63 82.48 One sheep only. DIGESTION EXPERIMENTS WITH SHEEP. 291 Table Y. — Computation of Digestion Coefficients — Continued. Series XXI., Distillers' Grains, Period 6 — Concluded. Sheep VI. Item. 1 1 <' 8 1 E 11 .11 ' 1 Amount consumed as above, Minus 189.96 grams feces excreted, 666.89 180.41 23.45 19.75 112.35 28.16 119.95 42.09 384.73 80.99 26.41 9.42 Amount digested, Minus basal ration digested. 486.48 354.86 3.70 2.66 84.19 44.68 77.86 62.19 303.74 240.53 16.99 2.95 Distillers' grains digested Per cent, digested, 131.62 67.94 1.04 23.21 39.51 77.24 15.67 54.98 63.21 69.16 14.04 77.10 Average per cent, digested, . 66.79 36.31 76.24 53.85 66.32 80.52 Series XXI., Corn Bran, Period 7. Sheep IV. 300 grams English hay fed, .... 271.41 18.40 19.24 87.20 140.06 6.51 125 grams potato starch fed, 109.78 - - - 109.78 - 100 grams gluten meal fed, . 91.55 .88 40.81 1.85 46.50 1.51 200 grams corn bran fed. 180.48 2.35 15.38 23.84 134.49 4.42 Amount consumed. 653.22 21.63 75.43 112.89 430.83 12.44 Minus 164.70 grams feces excreted 157.47 14.74 22.55 36.25 77.19 6.74 Amount digested, . 495.75 6.89 52.88 76.64 353.64 5.70 Minus basal ration digested, 354.56 2.70 43.84 60.55 243.00 2.89 Corn bran digested. 141.19 4.19 9.04 16.09 110.64 2.81 Per cent, digested. 78.23 178.29 58.78 67.49 82.27 63.57 Sheep V. Amount consumed as above. Minus 157.82 grams feces excreted. 653.22 150.86 21.63 14.92 75.43 28.12 112.89 29.43 430.83 71.72 12.44 6.67 Amount digested Minus basal ration digested, 502.36 354.56 6.71 2.70 47.31 43.89 83.46 60.55 359.11 243.00 5.77 2.89 Corn bran digested Per cent, digested, 147.80 81.89 4.01 170.63 3.47 22.56 22.91 96.10 116.11 86.33 65.16 292 MASS. EXPEKIMENT STATION BULLETIN 181, Table V. ■ — Computation of Digestion Coefficients — Continued. Series XXI., Corn Bran, Period 7 — Concluded. VI. Item. i 1 Q < 1 a, jl 2 1 Amount consumed as above, Minus 169.38 grams feces excreted, 653.22 162.10 21.63 18.75 75.43 22.08 112.89 40.61 430.83 74.58 12.44 6.08 Amount digested, Minus basal ration digested. 491.12 354.56 2.70 53.35 43.84 72.28 60.55 356.25 243.00 6.36 2.89 Corn bran digested Per cent, digested, 136.56 75.66 .18 7.66 9.51 61.83 11.73 49.20 113.25 84.21 3.47 78.51 Average per cent, digested, . 78.59 152.19 47.72 70.93 84.27 69.08 Series XXI., New Bedford Garbage Tankage, Period Sheep IV. i 550 grams English hay fed 150 grams gluten feed fed, .... 150grams New Bedford garbage tankage fed. 497.86 136.58 137.21 34.65 4.67 21.57 36.19 38.05 30.21 167.93 9.97 13.27 246.00 81.53 69.87 13.09 2.36 2.29 Amount consumed Minus 286.48 grams feces excreted. 771.65 272.18 60.89 35.87 104.45 42.98 191.17 67.58 397.40 117.31 17.74 8.44 Amount digested, Minus basal ration digested. 499.47 425.07 25.02 11.40 61.47 51.23 123.59 120.97 280.09 232.55 9.30 7.42 New Bedford garbage tankage digested. Per cent, digested, 74.40 54.22 13.62 63.14 10.24 33.90 2.62 19.74 47.54 68.04 1.88 82.09 Sheep V. 550 grams English hay fed Minus 29.28 grams waste hay, 497.86 26.36 34.65 1.83 36.19 1.88 167.93 8.76 246.00 13.27 13.09 .62 English hay consumed, .... 150 grams gluten feed fed 150 grams New Bedford garbage tankage fed. 471 50 136.58 137.21 32.82 4.67 21.57 34.31 38.05 30.21 159.17 9.97 13.27 232.73 81.53 69.87 ...47 2.36 2.29 Amount consumed Minus 244.60 grama feces excreted, 745.29 231.78 59.06 32.43 102.57 43.57 182.41 48.05 384.13 100.10 17.12 7.63 Amount digested, Minus basal ration digested. 513.51 407.41 26.63 10.87 59.00 49.93 134.36 115.02 284.03 223.12 9.49 7.12 New Bedford garbage tankage digested. Per cent, digested r ■■ ■ 106.10 77.33 15.76 73.06 9.07 30.02 19.34 145.80 60.91 87.18 2.37 103.48 Excluded from average. DIGESTION EXPERIMENTS WITH SHEEP. 293 Table V. — Computation of Digestion Coefficients — Continued. Series XXI., New Bedford Garbage Tankage, Period 8 — Concluded. Sheep VI. Item. 1 Sheep IV. 800 grams English hay fed, . Minus 60.57 grams waste, English hay consumed, 50 grams wheat gluten fed, . Minus 34.71 grams waste. Wheat gluten consumed, Amount consumed. Minus 264.12 grams feces excreted Amount digested, . Minus wheat gluten digested, English hay digested, . Per cent, digested. 714.96 60.57 654.39 49.55 4.20 45.35 47.48 34.71 12.77 .42 667.16 252.60 414.56 12.77 21.16 .11 401.79 61.40 21.05 46.42 52.26 4.43 47.83 359.20 30.43 328.77 42.35 59.23 25.51 215.37 74.01 329.81 118.27 33.72 11.40 211.54 1.04 22.32 46.67 141.35 65.63 210.50 64.03 1.58 17.08 17.29 10.51 6.78 .21 6.57 38.47 Sheep VI. 800 grams English hay fed, . 50 grams wheat gluten fed, . 714.96 47.48 49.55 .42 52.26 42.35 235.29 .04 359.20 18.66 ■ .79 Amount consumed. Minus 288.47 gram.o feces excreted 762.44 275.89 49.97 27.56 94.61 28.14 235.33 77.17 363.08 131.65 19.45 11.37 Amount digested, . Minus wheat gluten digested. 486.55 47.48 22.41 .42 66.47 42.35 158.16 .04 231.43 3.88 8.08 .79 English hay digested, . Per cent, digested. 439.07 61.41 21.99 44.38 24.12 46.15 158.12 67.20 227.55 63.35 7.29 39.07 Average per cent, digested, 61.41 45.40 46.41 69.42 63.69 38.77 ' -To note effect of wheat gluten floiu-. DIGESTION EXPERIMENTS AVITH SHEEP. 295 Table V. — Computation of Digestion Coefficients — Continued. Series XXI., English Hay and Gluten Feed, — Gluten Feed, Period 11. Sheep V. Item. 1 >> Q <; S 1 s .11 r? 550 grams English hay fed, . 150 grams gluten feed fed, . 499.40 137.31 42.65 4.63 50.94 38.57 152.27 10.50 240.01 81.10 13.53 2.51 Amount consumed. Minus 198.58 grams feces excreted 636.71 187.88 47.28 28.13 89.51 26.55 162.77 42.63 321.11 82.38 16.04 8.19 Amount digested, . Minus hay digested. 448.83 294.65 19.15 15.78 62.96 27.00 120.14 95.93 238.73 153.60 7.85 6.90 Gluten feed digested, . Per cent, ration digested. 154.18 70.49 3.37 40.50 35.96 70.34 24.21 73.81 85.13 74.35 .95 48.94 Per cent, gluten feed digested. 112.30 73.00 93.20 231.00 105.00 38.00 Sheep VI. Amount consumed as above, Minus 193.37 grams feces excreted, 636.71 183.65 47.28 30.78 89.51 25.79 162.77 39.61 321.11 79.50 16.04 7.87 Amount digested Minus hay digested, 453.16 294.65 16.50 15.78 63.72 27.00 123.16 95.93 241.61 153.60 8.17 6.90 Gluten feed digested, . Per cent, ration digested, 158.51 71.17 .72 34.90 36.72 71.19 27.23 75.67 88.01 75.24 1.27 50.94 Per cent, gluten feed digested, 115.40 15.00 95.10 259.00 108.50 51.00 Average per cent, ration digested, 70.83 37.70 70.77 74.74 74.80 49.94 Average per cent, gluten feed diges ed, . 113.85 44.00 94.15 245.00 106.75 44.50 Series XXI., Feterit.a, Period 12. Sheep V. 550 grams English hay fed, . 150 grams gluten feed fed, . 200 grams feterita fed, . 495.94 137.25 179.18 41.96 4.68 3.23 47.86 38.35 23.71 151.61 10.14 2.51 240.97 81.94 143.75 13.54 2.14 5.98 Amount consiimed. Minus 244.74 grams feces excreted. 812.37 229.57 49.87 35.51 109.92 35.72 164.26 49.56 466.66 98.59 21.66 10.19 Amount digested, .... Minus hay and gluten feed digested. 582.80 449.56 14.36 17.72 74.20 61.21 114.70 121.31 368.07 242.18 11.47 7.84 Feterita digested, .... Per cent, digested, 133.24 74.36 ~ 12.99 54.79 - 125.89 87.58 3.63 60.70 296 MASS. EXPERIMENT STATION BULLETIN 181. Table V. — Computation of Digestion Coefficients — Continued. Series XXI., Feterita, Period 12 — Concluded. Sheep VI. Item. 1 Q < g 1 .ll 1 Amount consumed as above, Minus 243.61 grams feces excreted, 812.37. 228.55 49.87 32.25 109.92 39.63 164.26 48.43 466.66 97.57 21.66 10.67 Amount digested Minus hay and gluten feed digested, 583.82 449.56 17.62 17.72 70.29 61.21 115.83 121.31 396.09 242.18 10.99 7.84 Feterita digested Per cent, digested, 134.26 74.93 " 9.08 38.30 ~ 126.91 88.29 3.15 52.68 Average per cent, digested, . 74.65 - 46.55 - 87.94 56.69 Series XXI., English Hay, Period 13. Sheep XII. 700 grams English hay fed, . Minus 259.31 grams feces excreted, English hay digested, . Per cent, digested, 390.57 60.69 52.48 31.96 20.52 39.10 57.87 26.15 31.72 54.81 128.70 65.30 200.47 64.70 XIII. 700 grams English hay fed, . Minus 6.43 grams waste. Amount consumed, Minus 266.30 grams feces excreted English hay digested, . Per cent, digested. 52.10 33.40 18.70 35.89 57.87 .27 57.60 26.49 31.11 54.01 197.09 2.54 194.55 72.71 121.84 62.63 309.85 2.98 306.87 108.97 197.90 64.49 Sheep XIV. 700 grams English hay fed Minus 280.22 grams feces excreted, 634.55 263.60 52.48 33.95 57.87 28.42 197.09 76.73 309.85 115.91 17.26 8.59 English hay digested Per cent, digested, 370.95 " 57.64 18.53 35.31 29.45 50.89 120.36 61.07 193.94 62.59 8.67 50.23 Average per cent, digested, . 59.49 36.77 53.24 63.00 63.93 50.69 DIGESTION EXPERIMENTS WITH SHEEP. 297 Table V. — Computation of Digestion Coefficients — Continued. Series XXI., Sweet Clover (Green), Period 14. Sheep IV. Item. 1 i C5 1 1. 1 500 grams English hay fed, . 1,600 grams sweet clover, 442.15 264.80 35.81 25.31 39.93 45.89 138.13 89.50 216.17 96.53 12.11 7.57 Amount consumed, Minus 302.46 grams feces excreted, 706.95 274.48 61.12 35.74 85.82 30.11 227.63 86.41 312.70 112.50 19.68 9.72 Amount digested, .... Minus hay digested. 432.47 260.87 25.38 13.25 55.71 21.16 141.22 87.02 200.20 138.35 9.96 6.18 Sweet clover digested, . Per cent, digested, 171.60 64.80 12.13 47.93 34.55 75.29 54.20 60.56 61.85 64.07 3.78 49.91 Sheep VI. 1,600 grama sweet clover fed. Minus 26.14 grams waste, 264.80 24.94 25.31 2.27 45.89 2.47 89.50 11.78 96.53 8.11 7.57 .31 Sweet clover consumed, 500 grams English hay consumed. 239.86 442.15 23.04 35.81 43.42 39.93 77.72 138.13 88.42 216.17 7.28 12.11 Amount consumed. Minus 270.07 grams feces excreted 682.01 245.33 436.68 260.87 58.85 34.32 83.35 28.07 215.85 72.59 304.59 100.81 19.37 9.54 Amount digested, . Minus hay digested. 24.53 13.25 55.28 21.16 143.26 87.02 203.78 138.35 9.83 6.18 Sweet clover digested, . Per cent, digested. 175.81 73.30 11.28 48.96 34.12 78.58 56.24 72.36 65.43 74.00 3.65 60.28 Average per cent, digested. 69.05 48.45 76.94 66.46 69.04 50.10 Series XXII., Sud.'Ln Grass (Green, Second Crop), Period 1. Sheep IV. 500 grams English hay fed 433.50 33.42 39.75 131.31 218.05 10.97 1,600 grams Sudan grass (green, fourth cut- ting; fed. 367.00 24.44 44.29 104.79 190.67 11.81 Amount consumed, 809.50 57.86 84.04 236.10 408.72 22.78 Minus 316.87 grams feces excreted. 294.78 39.71 35.93 72.78 135.72 10.64 Amount digested 514.72 18.15 48.11 163.32 273.00 12.14 Minus hay digested 268.77 9.36 20.27 90.60 143.91 4.52 Sudan grass digested 245.95 8.79 27.84 72.72 129.09 7.64 Per cent, digested 65.41 37.97 62.86 69.40 67.70 64.69 298 MASS. EXPERIMENT STATION BULLETIN 181. Table V. — Computation of Digestion Coefficients — Continued. Series XXII., Sudan Grass (Green, Second Crop), Period 1 — Concluded. Sheep VI. Item. ■ 1 0 < 1 P-( 1 1. .11 1 Amount consumed as above. Minus 318.78 grams feces excreted, 809.50 295.99 57.86 42.56 84.04 33.62 236.10 72.75 408.72 135.68 22.78 11.37 Amount digested, Minus hay digested, 513.51 268.77 15.30 9.36 50.42 20.27 163.35 90.60 273.04 143.91 11.41 4.50 Sudan grass digested Per cent, digested 244.74 65.09 5.94 24.30 30.15 68.07 72.75 69.42 129.13 67.69 6.91 58.51 Average per cent, digested, . 65.25 30.14 65.47 69.41 67.70 61.60 Series XXII., English Hat, Period 2. Sheep IV. 800 grams English hay fed, . Minus 285.59 grams feces excreted, English hay digested, . Per cent, digested. 419.48 61.21 52.02 36.77 15.25 29.31 59.28 27.49 31.79 53.63 218.84 70.96 147.88 67.57 337.75 120.34 217.41 64.37 Sheep VI. Amount consumed as above. Minus 276.71 grams feces excreted. 685.36 256.90 52.02 38.18 59.28 30.85 218.84 65.18 337.75 112.32 17.48 10.38 English hay digested Per cent, digested, 428.46 62.52 • 13.84 26.61 28.43 47.96 153.66 70.21 225.43 66.75 7.10 40.62 Average per cent, digested, . 61.87 27.96 50.80 68.89 65.56 40.79 Series XXII., Sudan Grass (Dry, Second Crop), Period 3. Sheep IV. 400 grams English hay fed, .... 353.00 28.45 33.43 107.49 174.63 9.00 500 grams Sudan gra.ss (dry, fourth cutting) fed. 390.60 33.55 53.00 129.99 167.68 6.37 Amount consumed 743.60 62.00 86.43 237.48 342.31 15.37 Minus 309.29 grams feces excreted. 290.42 38.92 38.57 67.61 135.92 9.41 Amount digested 453.18 23.08 47.86 169.87 206.39 5.96 Minus hay digested, 218.86 7.96 17.05 74.17 115.26 3.69 Sudan grass digested 234.32 15.12 30.81 95.70 91.13 2.27 Per cent, digested, 59.99 45.07 58.13 73.62 54.35 35.63 DIGESTION EXPERIMENTS WITH SHEEP. 299 Table V. — Computation of Digestion Coefficients — Co7\linupd. Series XXII., Sudan Grass (Dry, Second Crop), Period 3 — Concluded. Sheep VI. Item. 1 P 1 = 1 1 2 1 fc Amount consumed as above, Minus 313.86 grams feces excreted, 743.60 292.83 62.00 40.53 86.43 36.84 237.48 68.73 342.31 137.31 15.37 9.43 Amount digested, Minus hay digested 450.77 218.86 21.47 7.96 49.59 17.05 168.75 74.17 205.00 115.26 5.94 3.69 Sudan grass digested Per cent, digested, 321.91 59.37 13.51 40.27 32.54 61.40 94.58 72.76 89.74 53.52 2.25 35.32 Average per cent, digested, . 59.68 42.67 59.77 73.19 53.94 35.48 Series XXII., Sudan Grass (First Crop, Third Cutting), Period 4. Sheep IX. 700 grams Sudan grass (third cutting, dry) fed. Minus 270.21 grams feces excreted, Sudan grass digested, Per cent, digested, 616.00 258.02 357.98 58.11 45.40 21.11 24.29 53.50 73.24 220.47 27.56 74.23 45.68 146.24 62.37 66.33 268.02 8.87 130.04 5.08 137.98 3.79 51.48 42.73 Sheep XI. 700 grams Sudan grass (third cutting, dry) Minus 292.87 grams feces excreted. 616.00 278.93 45.40 26.22 73.24 30.96 220.47 77.60 268.02 138.32 8.87 5.80 Sudan grass digested, . . . . . Per cent, digested 337.07 54.72 19.18 42.25 42.28 47.73 142.87 64.80 129.70 48.39 3.07 34.61 Average per cent, digested, . 56.42 47.88 60.05 65.57 49.94 38.67 Series XXII., Sl^dan Grass (First Crop, Second Cutting), Period 6. Sheep IX. 700 grams Sudan grass (second cutting, dry) fed. Minus 266.71 grams feces excreted, Sudan grass digested Per cent, digested, 616.49 251.56 59.61 26.54 95.49 34.03 205.41 64.90 246.53 119.29 9.43 6.79 300 MASS. EXPERIMENT STATION BULLETIN 181. Table V. — Computation op Digestion" Coefficients — Continued. Series XXII., Sudan Grass (First Crop, First Cutting), Period 7. Sheep XI. 2t5 700 grams Sudan grass (first cutting, dry) fed Minus 5.57 grams of waste, . Amount consumed. Minus 283.26 grams feces excreted, Sudan grass (first cutting) digested, Per cent, digested. 615.44 5.23 610.21 266.94 .49 88.44 38.36 2.15 202.18 67.96 250.73 1.84 343.27 56.25 34.27 55.92 50.08 56.63 134.22 122.55 49.24 Sheep XII 700 grams Sudan grass (first cutting, dry) fed. 615.44 61.97 88.93 204.33 250.73 9.48 Minus 94.43 grams of waste, . 81.92 10.78 9.72 29.42 31.09 .91 Amount consumed. 533.52 51.19 79.21 174.91 219.64 8.57 Minus 253.21 grams feces excreted. 239.33 24.87 35.47 58.73 112.58 7.68 Sudan grass (first cutting) digested, 294.19 26.32 43.74 116.18 107.06 .89 Per cent, digested, 55.14 51.42 55.22 66.42 48.74 10.38 Sheep XIII. 700 grams Sudan grass (first cutting, dry) fed, Minus 278.76 grams feces excreted. 615.44 263.73 61.97 26.69 37.50 204.33 57.80 250.73 124.11 9.48 7.62 Sudan grass (first cutting) digested, . Per cent, digested, ..... 351.71 57.15 . 35.28 56.93 51.43 57.83 133.53 66.82 126.62 50.51 1.86 19.62 Average per cent, digested, . 56.18 54.76 56.56 66.54 49.50 17.67 Series XXII., English Hay, Period 8. Sheep IV. 800 grams English hay fed, .... Minus 295.57 grams feces excreted. 719.36 279.11 49.28 29.20 59.35 29.50 236.60 77.65 357.23 133.42 16.90 9.35 English hay digested Per cent, digested 440.25 61.20 20.08 40.75 29.85 50.30 158.95 67.18 223.81 62.65 7.55 44.67 DIGESTION EXPERIMENTS WITH SHEEP. 301 Table V. — Computation of Digestion Coefficients — Continued. Series XXII., English Hay, Period 8 — Concluded. Sheep VI. Item. 1 1 <; 1 1 s II 800 grams English hay fed, .... Minus 282.37 grams feces excreted. 719.36 266.70 49.28 28.51 59.35 28.86 236.60 72.62 357.23 127.51 16.90 9.20 English hay digested, Per cent, digested, . . . . . 442.66 61.54 20.77 42.15 30.49 51.37 163.98 69.81 229.72 64.31 7.70 45.56 Average per cent, digested, . 61.37 41.45 50.84 68.25 63.48 45.12 Series XXII., Vinegar Grains, Period Sheep IX. 250 grams vinegar grains fed, 550 grams English hay fed, . 230.55 495.72 5.79 35.59 47.40 41.49 46.41 167.55 116.01 239.18 14.94 11.90 Amount consumed. Minus 309.94 grams feces excreted. 726.27 297.60 41.38 27.14 88.89 37.91 213.96 78.21 355.19 145.44 26.84 8.90 Amount digested, .... Minus English hay digested. 428.67 302.39 14.24 14.59 50.98 21.16 135.75 113.93 209.75 150.68 17.94 5.38 Vinegar grains digested, Per cent, digested. 126.28 54.77 - 29.82 62.91 21.82 47.02 59.07 50.92 12.58 84.20 Sheep XI. Amount consumed as above. Minus 318.83 grams feces excreted. 726.27 297.05 41.38 29.11 88.89 39.63 213.96 76.55 355.19 143.45 26.84 8.32 Amoimt digested, Minus English hay digested. 429.22 302.39 12.27 14.59 49.26 21.16 137.41 113.93 211.74 150.68 5 36 Vinegar grains digested Per cent, digested, 126.83 55.01 ~ 28.10 59.28 23.48 50.59 61.06 52.63 13.16 88.08 Average per cent, digested, . 54.89 - 61.10 48.80 51.77 86.14 302 MASS. EXPERIMENT STATION BULLETIN 181. Table V. — Computation of Digestion Coefficients — Continued. Series XXII., Vinegar Grains, Period 10. Sheep IV. Item. 1 1 1 1 250 grams vinegar grains fed, 550 grams English hay fed, . 231.33 495.00 5.95 36.23 46.77 43.07 46.54 159.10 116.45 244.33 15.61 12.18 Amount consumed, Minus 285.43 grams feces excreted, 726.33 272.59 42.18 25.60 89.84 35.38 205.64 69.10 360.78 143.14 27.79 8.37 Amount digested, .... Minus English hay digested. 453.74 301.95 16.58 14.85 54.46 21.97 136.54 108.19 217.64 153.93 19.42 5.48 Vinegar grains digested. Per cent, digested, 151.79 65.60 1.73 29.08 32.49 69.47 28.35 60.92 63.71 54.71 13.94 89.30 Sheep VI. Amount consumed as above. Minus 281.57 grams feces excreted, 726.33 268.28 42.18 28.97 89.84 37.00 205.64 63.07 360.78 130.46 27.79 8.77 Amount digested Minus English hay digested. 458.05 301.95 13.21 14.85 52.84 21.97 142.57 108.19 230.32 153.93 19.02 5.48 Vinegar grains digested Per cent, digested 156.10 : 30.87 66.00 34.38 73.87 76.39 65.60 13.54 86.70 Average per cent, digested, . 66.54 29.081 67.74 67.40 59.66 88.00 Series XXII., Stevens' "4-1" Dairy R.\tion, Period 11. Sheep IV. 250 grams of Stevens' "44" Dairy Ration fed, 550 grams English liay fed 227.65 499.13 9.49 33.69 61.35 41.08 164.56 112.82 247.47 14.66 12.33 Amount consumed Minus 269.57 grams feces excreted. 726.78 257.14 43.18 26.90 102.43 31.09 193.88 68.68 360.29 122.50 26.99 7.97 Amount digested Minus English hay digested, 469.64 304.47 16.28 13.81 71.34 20.95 125.20 111.90 237.79 155.90 19.02 5.55 Stevens' "44" Dairy Ration digested. Per cent, digested, 165.17 72.55 2.47 26.03 50.39 82,14 13.30 45.36 81.89 72,58 13.47 91.88 1 One sheep only. DIGESTION EXPERIMENTS WITH SHEEP. 303 Table V, — Computation of Digestion Coefficients — Continued. Series XXII., Stkvens' "44" Daihy Ration, Period 11 — Concluded. Sheep VI. Item. 1 i 4 1 1 1 . .-Sw 1 Amount consumed as above, Minus 277.79 grams feces excroted, 726.78 43.18 30.03 102.43 34.10 193.88 65.85 360.29 126.28 26.99 9.02 Amount digested, Minus English hay digested, 460.60 304 47 12.25 13.81 68.33 20,95 128.03 111.90 234.01 155.90 15.97 6 55 Stevens' "44" Dairy Ration digested. Per cent, digested, 156.13 68.58 : 47.38 77.23 16.13 55.01 78.11 69.23 10.42 71.08 Average per cent, digested, . 70.57 26.031 79.69 50.19 70.91 81.48 Series XXII., New York Alf.\.lfa (Third Cutting), Period 2. Sheep IV. 800 grams New York alfalfa (third cutting) fed. Minus 306.21 grams feces excreted, Alfalfa digested, Per cent, digested, 700.96 42.27 105.92 248.28 291.11 293.69 24.73 28.43 131.78 97.59 407.27 17.54 77.49 116 50 193 52 58.10 41.50 73.16 46.92 06.48 Sheep VI. 800 grams New York alfalfa (third cutting) fed. Minus 4.57 grams waste 700,96 4.14 42.27 .17 105.92 .27 248.28 2,19 291.11 1.47 13.39 .03 Amount consumed, Minus 330. .53 grams feces excreted. 696.82 315.33 42.10 30.27 105.65 33.61 246.09 134.99 289.64 105.13 13 36 11.32 Alfalfa digested Per cent, digested, 381.49 54.75 11.83 28.10 20.04 68.19 111.10 45.15 184.51 63.70 2.04 15.27 Average per cent, digested, 56.43 34.80 70.68 46.04 65.09 15.97 Series XXII., English Hay, Period 13. Sheep XII. 700 grams English hay fed, . Minus 208.70 grams feces excreted, English hay digested, . Per cent, digested. 636.09 45.73 51.78 211.56 311.81 256,15 23.54 24.23 78.46 121.20 379,94 22.19 27.55 133.10 190.61 59.73 48.52 53.20 62.91 61.13 One sheep only. 304 MASS. EXPEEIMENT STATION BULLETIN 181. Table V. — Computation of Digestion Coefficients — Continued. Series XXII., English Hat, Period 13 — Concluded. Sheep XIII. Item. 15 1 1 .11 1 700 (minus 1.67 grams waste) equals 698.33 grams English hay fed. Minus 281.36 grams feces excreted, 634.57 266.79 45.63 25.43 51.65 26.68 211.06 80.12 311.07 125.87 15.17 8.70 English hay digested Per cent, digested, 387.78 57.96 20.20 44.27 24.97 48.37 130.94 72.04 185.20 57.54 6.47 42.65 Average per cent, digested, . 58.85 46.40 50.77 62.48 60.34 42.77 Series XXII., New York Alfalfa (Third Cutting), Period 14. Sheep XII. 700 grams Now York alfalfa (third cutting) fed. Minus 269.07 grams feces excreted. 638.75 257.96 44.37 22.13 99.45 27.42 221.65 114.64 260.10 84.59 13.16 9.18 Alfalfa digested, Per cent, digested, 380.79 59.61 22.26 50.15 72.03 72.43 107.01 48.28 175.51 67.48 3 98 30.24 Sheep XIII. 700 grams New York alfalfa (third cutting) Minus 277.50 grams feces excreted. 638.75 265.04 44.37 21.89 99.45 26.53 221.65 121.39 260.10 86.32 13.16 8.91 Alfalfa digested Per cent, digested, 373.71 58.50 22.50 50.68 72.92 73.32 100.26 45.23 173.78 66.81 4.25 32.29 Average per cent, digested, . 59.06 50.42 72.88 46.76 67.15 31.26 Series XXII., Rowen, Period 15. Sheep XII. 700 grams rowen fed, 636.09 51.65 82.88 179.06 300.80 21.69 Minus 259.10 grams feces excreted, 249.25 33.97 32.83 57.08 110.19 15.18 Rowen digested 386.84 17.68 50.05 121.98 190.61 6.51 Per cent, digested 60.81 34.23 60.39 68.12 63.37 30.01 Sheep XIII. 700 grams rowen fed, Minus 257.74 grams feces excreted. 636.09 247.04 51.65 33.18 82.88 32.93 179.06 57.17 300.80 109.56 21.69 14.20 Rowen digested Per cent, digested, 389.05 61.16 18.47 35.76 49.95 60.27 121.89 68.08 191.24 63.57 7.49 34.53 Average per cent, digested, . 60.99 35.00 00.33 68.10 63.47 32.27 DIGESTION EXPERIMENTS WITH SHEEP. 305 Table V. — Computation of Digestion Coefficients — Continued. Series XXII., Sweet Clover (Green), Period 16. Sheep IX. Item. 1 >> P 1 .11 400 grams English hay fed, . 1,600 grams sweet clover fed. 359.48 232.00 25.92 10.67 32.57 49.79 109.03 62.13 182.04 101.52 9.92 7.89 Amount consumed. Minus 246.07 grams feces excreted. 591.48 228.30 36.59 27.36 82.36 28.67 171.16 72.71 283.56 107.38 17.81 9.94 Amount digested, .... Minus hay digested. 363.18 208.50 9.23 11.66 53.69 15.63 98.45 68.89 176.18 109.22 7.87 4.46 Sweet clover digested, . Per cent, sweet clover digested, . 154.68 66.67 _ 38.06 76.44 29.56 47.60 66.96 65.96 3.41 43.22 Sheep XI. Amount consumed as above, 591.48 36.59 82.36 171.16 283.56 17.81 Minus 231.36 grams feces excreted. 214.52 24.89 25.91 69.78 102.26 8.51 Amount digested, 376.96 11.70 56.45 101.38 181.30 9.30 Minus hay digested 208.50 11.66 15.63 68.89 109.22 4.46 Sweet clover digested 168.46 .04 40.82 32.49 72.08 4.84 Per cent, sweet clover digested, . 72.61 .03 81.98 52.29 71.00 61.34 Average per cent, sweet clover digested. 69.64 .03 79.21 49.95 68.48 52.28 Series XXII., Sudan Grass (Green), Period 17. Sheep XII. 400 grams English hay fed 1,600 grams Sudan grass (green, first cutting) fed. Amount consumed, Minus 235.01 grams feces excreted. 352.28 313.28 24.69 22.49 30.79 44.58 117.63 95.02 169.27 136.43 9.90 14.75 665.56 215.74 47.18 16.27 75.37 24.18 212.65 58.53 305.70 109.14 24.65 7.62 Amount digested Minus hay digested, 449.82 207.85 30.91 11.36 51.19 15.70 154.12 72.93 196.56 101.56 17.03 4.26 Sudan grass (green, first cutting) digested, . Per cent. Sudan grass (green, first cutting) digested. 241.97 77.23 19.55 86.93 35.49 79.61 81.19 85.45 95.00 69.63 12.77 86.57 306 MASS. EXPERIMENT STATION BULLETIN 181. Table V. — Computation of Digestion Coefficients — Concluded. Series XXII., Sudan Grass (Green), Period 17 — Concluded. Sheep XIII. Item. 1 1 >> P i 1 s .11 1 Amount consumed as above, Minus 253.04 grams feces excreted. 665.56 235.78 47.18 26.83 75.37 24.38 212.65 65.90 305.70 110.13 24.65 8.54 Amount digested Minus hay digested, 429.78 207.85 20.35 11.36 50.99 15.70 146.75 72.93 195.57 101.56 16.11 4.26 Sudan grass (green, first cutting) digested, . Per cent. Sudan grass (green, first cutting) digested. 221.93 70.84 8.99 39.96 35.29 79.16 73.82 77.69 94.01 68.97 11.85 80.34 Average per cent. Sudan grass (green, first cutting) digested. 74.04 63.45 79.38 81.57 69.30 83.46 Discussion of the Results. Having presented in the foregoing pages a statement of the general purpose of these experiments, an explanation of the tables, and the data of the composition of the feeds and feces, as well as the detailed data of the e.xperiments, including the computation of the digestion coefficients, it is intended in the pages which follow to state briefly the general character of each feed, summarize the coefficients secured, and draw such conclu- sions as the results indicate. In noting the variations which occur when the same feed is fed to different sheep, the fact must not be lost sight of that digestibility is made up of a number of processes. Armsby states the matter clearly when he says "digestibility in ruminants is a very complex affair, depending on many factors; ... it may be characterized as a series of fermentations effected in part by a variety of organized ferments, and in part by enz5mies secreted by the digestive organs or contained in the feed itself. Changes in the composition of the contents of the digestive tract, or in the rapidity with which they move forward through it, can hardly fail to influence in a variety of ways the course of these fermenta- tions, and it seems, on the whole, rather surprising that they go forward as rapidly as they do." DIGESTION EXPERIMENTS WITH SHEEP. 307 Summary of Coefficients of English Hay — Basal. Lot. Series. Period. Sheep. Dry Matter. Ash. Protein. Fiber. Nitrogen- free Extract. Fat. 1 XIX. 3 I. 59.77 20.59 49.08 64.84 64.07 38.42 1 XIX. 3 II. 57.65 23.90 53.12 58.93 62.39 38.17 1 XIX. 9 V. 60.57 33.50 53.53 63.32 63.29 57.47 1 XIX. 9 VI. 57.53 32.81 51.10 58.66 60.84 54.99 Average, 58.88 27.70 51.71 61.45 62.72 47.26 2 XX. 1 I. 64.70 39.69 50.86 70.41 66.78 84.67 2 XX. 1 II. 63.43 36.98 50.38 68.33 66.06 47.96 2 XX. 6 IV. 60.08 42.04 48.34 63.10 62.51 39.26 2 XX. 10 VII. 56.80 42.70 52.25 59.57 58.27 43.15 2 XX. 10 VIII. 59.70 44.21 52.72 62.44 61.96 41.83 Average 60.94 41.12 50.91 64.77 63.12 51.39 3 XXI. 2 VII. 54.89 28.25 42.00 58.88 58.62 42.44 3 XXI. 2 VIII. 61.70 37.65 45.08 68.06 64.39 44.65 3 XXI. 9 IX. 58.21 40.25 42.18 62.65 60.86 46.52 3 XXI. 9 X. 56.37 43.47 46.81 59.36 58.50 40.87 3 XXI. 9 XI. 53.82 38.14 40.07 56.46 57.13 39.27 Average 57.00 37.55 43.23 61.08 59.90 42.71 4 XXI. 13 XII. 60.69 39.10 54.81 65.30 64.70 53.07 4 XXI. 13 XIII. 60.15 35.89 54.01 62.63 64.49 48.78 4 XXI. 13 XIV. 57.64 35.31 50.89 61.07 62.59 50.23 4 XXII. 2 IV. 61.21 29.31 53.63 67.57 64.37 40.96 4 XXII. 2 VI. 62.52 26.61 «.« 70.21 66.75 40.62 Average, 60.44 .31.24 52.26 65.36 64.58 46.73 5 XXII. 8 IV. 61.20 40.75 50.30 67.18 62.65 44.67 5 XXII. 8 VI. 61.54 42.15 51.37 69.31 64.31 45.56 5 XXII. 13 XII. 59.73 48.52 53.20 62.91 61.13 41.89 5 XXII. 13 XIII. 57.96 44.27 48.37 72.04 57.54 42.65 Average, 60.11 43.92 50.81 67.86 61.41 43.69 Grand average • • 59.47 36.31 49.78 64.10 62.35 46.34 308 MASS. EXPERIMENT STATION BULLETIN 181. Five distinct lots of hay were used in these experiments. The hay was cut when in bloom from an old mowing, and was composed largely of Kentucky blue grass (Poa pratensis) and sweet vernal grass {Anthoxan- thum odoratum) "with an admixture of more or less clover. The results, on the whole, are reasonably uniform, although one notes occasional variations, particularly in the fiber and also in the protein, due evidently to the individuality and perhaps to particular condition of the sheep. The last two lots were evidently of somewhat better quality, or per- haps cut a little earlier than the first two, for they showed a somewhat superior digestibility. All five lots were more fully digested than is timothy hay. Note that the fiber in the hay has a digestibility slightly above the extract matter. This is characteristic of many coarse feeds. Summary of Coefficients of English Hay and Gluten Feed — Basal Lot. Series. Period. Sheep. Dry Matter. Ash. Protein. Fiber. Nitrogen- free Extract. Hay. Gluten Feed. Fat. 1 XIX. 2 V. 66.13 34.59 67.91 64.67 69.82 56.13 1 XIX. 2 VI. 66.61 27.03 68.79 66.60 70.21 56.53 2 XIX. 15 V. 66.70 42.45 64.22 66.36 70.29 54.23 2 XIX. 15 VI. 66.22 39.97 59.69 60.35 66.48 48.23 2 XX. 5 I. 70.34 41.09 66.45 74.12 72.80 60.56 2 XX. 5 II. 66.69 28.84 61.42 69.60 70.70 58.33 2 XX. 14 IV. 60.31 27.12 63.45 60.34 63.53 50.95 2 XX. 14 V. 63.23 29.01 64.01 65.99 66.05 46.73 2 XX. 14 VI. 63.74 31.94 64.71 65.92 66.31 51.93 3 XXI. 1 IV. 64.97 31.78 68.38 65.52 68.73 46.03 3 XXI. 1 V. 66.92 23.83 68.53 68.50 71.74 47.76 3 XXI. 1 VI. 68.81 32.51 70.34 71.07 72.53 49.21 3 XXI. 11 V. 70.49 40.50 70.34 73.81 74.35 48.94 3 XXI. 11 VI. 71.17 34.90 71.19 75.67 75.24 50.94 Average 66.59 33.25 66.39 67.75 69.91 51.89 In many cases it was thought wise to use a basal ration composed of English hay and gluten feed in order to secure a combination better balanced as regards protein and carbohydrates than is hay. Gluten feed was selected to be used with the hay because it contained a moderate amount of protein and is usually quite fully digested. In Series XIX. a combination of 650 grams of hay and 125 grams of gluten feed was used, and in the other cases 550 grams of hay and 150 grams of gluten feed. DIGESTION EXPERIMENTS WITH SHEEP. 309 The results of Period 14, Series XX., are rather surprising, and in a way hardly to be explained, being noticeably below Series XXI., Periods 1 and 11, which are reasonably uciform. They will be discussed further in considermg the digestibility of gluten feed. Series XIX., Period 15, has more hay in proportion to gluten feed, and the coefficients are some- what below the other series, with the exceptions mentioned. Summary of Coefficients of English Hay, Potato Starch and Diaynond Gluten Meal — Basal. Lot. Series. Period. Sheep. Dry Matter. .\sh. Protein. Fiber. Nitrogen-" free Extract. Hay. Starch and Gluten. Fat. 1 XIX. 10 III. 73.45 33.70 73.48 59.62 80.71 34.29 1 XIX. 10 IV. 70.47 13.22 71.02 55.41 78.91 32.67 1 XIX. 11 IV. 72.17 33.93 75.02 60.73 78.39 47.07 2 XXI. 5 IV. 72.77 19.38 71.12 64.26 80.63 32.29 2 XXI. 5 V. 77.28 9.41 75.33 74.73 84.15 41.00 2 1 XXI. 5 VI. 73.49 13.30 71.93 66.46 81.24 35.66 Average, 73.27 20.16 72.98 63.54 80.67 37.16 In order to study the digestibility of fiber in distillers' grains and corn bran, a basal ration composed of a limited amount of hay plus potato starch and Diamond gluten meal was used. This ration naturally con- tained but little fiber, and would permit the intestinal juices and ferments to e.xert their maximum effect upon the fiber of the two by-products. Sheep IV. in Series XIX. received 100 grams more hay and 25 grams more gluten meal daily in the combination than did the other-three sheep. The coefficients of this basal ration are fairly uniform, excepting that Sheep V. appeared to have digested noticeably more of the ration than did the other sheep. 310 MASS. EXPERIMENT STATION BULLETIN 181. Summary of Coefficients of Gluten Feed (Present Experiments). Series. Period. Sheep. Brand. Dry Matter. Ash. Protein. Fiber. Nitrogen- free Extract. Fat. XIX. 2 V. 91.80 167.00 87.50 99.00 94.70 72.50 XIX. 2 VI. - 94.70 - 89.50 126.00 96.30 73.60 XIX. 14 V. - 112.09 337.98 84.78 208.85 110.89 59.91 XIX. 14 VI. - 94.54 258.91 89.69 116.36 94.63 53.50 XIX. 15 V. Clinton 106.57 163.22 82.62 173.28 108.91 50.00 XIX. 15 VI. Clinton 78.90 129.89 71.28 47.41 87.34 29.55 XX. 5 I. Clinton 93.42 75.66 84.70 145.65 94.25 87.60 XX. 5 II. Clinton 76.50 - 73.73 77.96 85.50 80.45 XX. 14 IV. Clinton 50.38 - 76.15 - 58.65 63.55 XX. 14 V. Clinton 64.07 - 77.25 51.82 69.51 50.32 XX. 14 VI. Clinton 66.47 - 78.61 50.80 70.63 66.01 XXI. 1 IV. BufiFalo 93.87 - 92.96 136.00 95.88 63.27 XXI. 1 V. Buffalo 102.79 - 93.16 185.00 108.12 74.69 XXI. 1 VI. Buffalo 111.63 - 96.82 228.00 111.46 84.49 XXI. 11 V. Buffalo 112.30 73.00 93.20 231.00 105.00 38.00 XXI. 11 VI. Buffalo 115.40 15.00 95.10 259.00 108.50 51.00 Aver ige, 91.59 - 85.44 142.41 93.77 64.41 The gluten feed represented in these trials comprised three different lots of the same general type of chemical composition. It contained approximately 9 per cent, of water; and in dry matter the ash varied from .95 to 3.49 per cent., the protein from 25.47 to 28.29 per cent., the fiber from 7.30 to 8.70 per cent., the extract matter from 56.86 to 59.70 per cent., and the fat from 1.56 to 4.94 per cent. In general appearance the three samples resembled each other closely. The variations in percentage of ash and fat indicated some little difference in the manufacturing process, but not sufficient to warrant any noticeable variations in the digestibility of the several lots. In fact, the gluten feed used in Series XIX., Periods 2 and 14, and the same series, Period 15, were two different lots, and yet they resemble each other closely in digestibility. Here follow the results of a number of early experiments. The process of manufacture was somewhat different, more of the germ being retained resulting in a higher fat percentage. The ash also was not much over 1 per cent, because the evaporated steep water was not added. Rather wide variations are noted as in the later experiments. DIGESTION EXPERIMENTS WITH SHEEP. 311 Summary of Earlier Work with Gluten Feed. Digestion Coefficients. Year. Series. Period. Sheep. Dry Matter. Ash. Protein. Fiber. Nitrogen- free Extract. Fat. 1893 - 2 II. 75.53 - 85.97 39.92 78.44 82.25 1893 - 2 IV. 80.44 - 83.94 46.28 84.37 80.58 1894 - 5 III. 89.35 - 88.69 94.69 88.93 92.74 1894 - 5 IV. 91.11 - 88.88 104.56 89.76 95.61 1896 - - - 87.00 - 86.00 77.00 90.00 81.00 1906 - 8 IV. 93.78 78.07 89.26 123.46 93.42 75.70 1906 - 8 V. 97.83 98.67 92.91 128.92 97.03 79.68 1909 XI. IV. 92.25 85.05 90.02 107.23 92.30 76.29 1909 XI. V. 99.24 93.69 92.22 153.69 97.98 77.29 1909 XII. IV. 94.58 - 91.22 127.83 96.09 77.09 1909 XII. * V. 95.18 - 89.65 146.29 97.60 57.82 1909 XII. 14 II. 75.30 - 68.83 63.62 83.32 67.06 1909 XIV. I. 99.18 77.54 87.31 134.91 103.30 97.79 1909 XIV. II. 90.98 34.40 83.32 119.00 99.04 81.26 1909 XIV. III. 85.81 51.15 81.84 81.74 94.04 79.46 1909 XIV. IV. 101.55 64.83 88.58 147.10 108.24 84.21 Average Results. Present € xperimen ts, . . 91.59 - 85.44 142.41 93.77 64.41 Earlier e tperiment s. . 90.57 - 86.79 106.01 93.36 80.36 Averages are rot particularly satisfactory, especially when the figures from which they are made up vary widely among themselves. The fore- going averages show, however, the gluten feed to have a high digestibihty. A study of the numerous results brings out at least two striking facts. In the first place, in some experunents the coefficients are very much higher than in others. Thus, Series XX., Period 14, gave results very noticeably below the others. It is the beUef of the writer, however, that at least a part of the varia- tion is due to the lessened activity of the digestive processes, even though such a condition may not be indicated by any outward signs. The changing from one ration to another may also change the intestinal flora. In the second place, it is observed that in a number of instances the gluten feed appears to be over 100 per cent, digestible. It seems reason- able to assume that this is due to its favorable effect in increasing the 312 MASS. EXPERIMENT STATION BULLETIN 181. digestibility of the hay; this condition was particularly pronounced in case of the fiber and to a lesser extent in the extract matter, and is in accord with the accepted teaching of the favorable influence of a protein concentrate on the fiber and extract matter of a basal ration having a wide nutritive ratio. The digestibility of the protein varied in proportion to the digestibility of the extract matter, and is shown to be quite well utiUzed. The fat show^ed wide variations, due in part to the small amount present, and in part to other causes. The ash content of gluten feed is not large, and ia most cases more ash was excreted from the total ration than was con- tained in the gluten feed fed, so that coefficients for this ingredient cannot be deduced. Average Coefficients for All Results. Different lots, ........... 7 Number of single trials, ......... 32 Dry matter 91.08 Ash, Protein 86. 12 Fiber 124.21 Nitrogen-free extract, . . . . . . . .93.57 Fat, 72.39 The average results for all samples indicate very clearly that gluten feed is a highly digestible nitrogenous concentrate, and that in all prob- abihty it exerts a favorable influence upon the digestibility of a basal ration having a wide nutritive ratio. Summary of Coefficients for Diamotid Gluten Meal. Series. Period. Sheep. Dry Matter. Ash. Protein. Fiber. Nitrogen- free Extract. Fat. XIX XIX XIX XXI XXI XXI 10 10 5 5 5 IV. IIH IV. IV. V.i VI. 83 68 87 86 109 90 143 127 83.8 80.0 86.4 84.8 91.2 86.0 m 100 100 90.5 79.0 91.4 93.3 120.1 97.2 47.3 Average,! Average of previous results (8), 86 87 - 85.0 100 88.0 93.0 I Results from Sheep III. and V. omitted from average. A combination of 300 to 400 grams of hay, 125 grams of potato starch, and 100 to 125 grtims of Diamond gluten meal were fed as a basal ration in order to study the digestibility of distillers' dried grains and com DIGESTION EXPERIMENTS WITH SHEEP. 313 bran. It seemed worth while in this connection to get at the digestibility of the Diamond gluten meal. In order to accomplish this the digestion coefficients found for the hay were applied to the hay consumed, and to the resulting product was added the amount of starch consumed, which was assumed to be entirely digested. The sum of the hay and starch digested was taken from the total amount digested, and the remainder represented the gluten meal digested. The coefficients used for the hay in case of Series XIX. represented an average of those secured by using the results from Sheep I., II., V. and VI., all of which agreed closely. Those used in Series XXI. were the average of those for Sheep VII., VIIL, IX., X. and XT., as IV., V. and VI. had not been used in getting the digestibility of this lot of hay. The coefficients for the hay were as follows: — Sebies. Dry Matter. Ash. Protein. Fiber. Nitrogen- free Extract. Fat. XIX XXI., 59 57 28 37 52 43 62 61 62 60 47 43 The nutritive ratio of the basal ration in Series XIX. averaged 1:6.5 and in Series XXL, 1:6.8. In passing, attention is called to the fact that the ash, fiber and fat content of gluten meal are quite low, showing less than 2 per cent, of each on a dry-matter basis, and the coefficients secured were, as might be expected, of uncertain value, although it is reasonable to assume that these several constituents were quite fully digested. The content of protein and extract matter, on the other hand, on the basis of dry matter, was 45 and 50 per cent., respectively, showing this feedstuff to be made of these two food groups in nearly equal proportions. A study of the coefficients secured shows some "wdde variations. Sheep III., Series XIX., for some reason gave quite low results, and in Series XXI., Sheep V gave results considerably above the others. In making the average, therefore, it seemed wise to omit the coefficients obtained with these two sheep. The results show the gluten meal to have a high digestibility; in fact, it is believed that if a method sufficiently accurate were available it could be shown that the meal was practically all utilized. The coefficients given for previous results represent eight single trials with four different lots, and were secured a number of years ago with gluten meal made by a little different process and averaging 40 per cent. protein and 54 per cent, extract matter in dry matter. The latter co- efficients are in substantial accord with those recently secured. 314 MASS. EXPERIMENT STATION BULLETIN 181. Summanj of Coefficients showing Effect of High-grade Wheat Gluten Flour upon Digestibility of Hay. j 1 Dry Matter. Ash. Protein. Fiber. Extract Matter. 1 Fat. i 1 i i i 1 % ^ 1 i 5 S .J 1 XX. XX. 10 and 12, 10 and 12, VII. VIII. 59 60 1 57 60 48 46 43 44 43 40 52 53 62 64 6C 62 61 62 58 62 46 48 43 42 Average, . 59 58 47 43 41 52 63 61 62 60 47 42 The object of this trial was to observe the effect of a high-grade wheat gluten flour, composed largely of protein, upon the digestibility of the hay. In the hay experiment 600 grams were fed to each of two sheep, and in the experiment immediately following 40 grams of the gluten were added to the hay. The hay contained in dry matter 6.66 per cent, ash, 8.36 protein, 32.08 fiber, 50.40 extract matter and 2.50 fat, and had a nutritive ratio of 1:12. The wheat gluten contained in dry matter .86 per cent, ash, 92.41 protein, .11 fiber, 6.23 extract matter and .39 fat, being nearly pure gluten meal, with traces of ash, fiber and fat, and a small amount of ex- tract matter. The nutritive ratio of the hay-gluten mixture was 1:6. A study of the comparative coefficients of the hay when fed with and with- out the gluten — assuming the gluten to have been entirely digested — indicates that the latter improved the digestibility of the hay slightly, particularly the fiber, extract matter and fat. The protein, on the other hand, showed an apparent lessened digestibility, due perhaps to the fact that the protein of the gluten was not completely assimilated. Appljdng the coeflScients secured for the hay when fed by itself to the same hay fed in combination with wheat gluten, and subtracting the result from the total amount of hay plus gluten digested, we find that in case of one sheep 47.48 grams, and in case of the other, 33.95 grams, were digested against 36.36 grams fed. This indicates that in one case at least the gluten was not only fully digested but improved somewhat the digestibility of the hay. DIGESTION EXPERIMENTS WITH SHEEP. 315 Summary of Coefficients showing Effect of High-grade Wheat Gluten Flour upon Digestibility of Hay — Continued. i 1 Dry Matter. Ash. Protein. Fiber. Extract Matter. Fat. 1 1 3 M ^ 3 1 J3 i i ^ i 1 XXI. 4> VII. 55 55 36 28 45 42 58 59 58 59 44 42 ' In case of hay alone, period 2. This experiment was with a new lot of hay, testing in dry matter 6.59 per cent, ash, 7.59 per cent, protein, 32.67 per cent, fiber, 50.29 per cent, extract matter and 2.86 per cent, fat, and having a nutritive ratio of about 1:17, being very wide. The wheat gluten was the same as the lot previously fed, and the combination of 700 grams hay and 40 grams wheat gluten had a nutritive ratio of 1 :5.7. In other words, the addition of 40 grams of gluten to 700 grams of hay produced a much narrower ration than if the hay had been fed by itself. A study of the coefficients shows no particular improvement in the digestibility of the hay as a result of adding the gluten, although such an improvement was antici- pated. Applying the coefficients secured for the hay when fed by itself to the same hay fed in combination with wheat gluten, and subtracting the result from the total amount of hay plus gluten digested, we have 38.58 grams of gluten digested as against 37.18 grams fed, showing the gluten to have been completely digested. Summary of Coefficients showing Effect of High-grade Wheat Gluten Flour upon Digestibility of Hay — Concluded. Dry Matter. Ash. Protein. Fiber. Extract Matter. Fat. ^• -ti -^i *: g 8 :fl eg (S .13 03 ^ ^ g g g g g g S % ^ ^ XXI. 101 IV. 61 57 46 37 47 43 66 61 64 60 38 45 XXI. 101 VI. 61 57 44 37 46 43 67 61 63 60 39 45 Average, 61 57 45 37 46 43 66 61 63 60 38 45 Average of all trials (5), 58 57 43 3. 44 46 62 61 61 60 43 43 1 In case of hay alone, periods 2 and 9. 316 MASS. EXPERIMENT STATION BULLETIN 181. The hay was the same as fed in the former trial; the gluten was a new lot, but did not vary in composition much from the previous sample used. Unfortunately, Sheep IV. and VI. were not used in testing the digesti- bility of the hay, and the coefficients represent the average obtained by using Sheep VII., VIII., IX., X. and XI. It is evident in this trial that the gluten did improve the digestibility of the hay somewhat, particularly the fiber and extract matter. Experiments by numerous investigators ^ have shown that when a ration containing considerable starch, and having a nutritive ratio of 1 :12 or more, is fed to ruminants more or less of the starch is found in the feces, and if to this ration a protein concentrate is added the starch dis- appears, and the digestion coefficients, not only of the extract matter but also of the fiber, are improved. In our own case the addition of a small amount of a very rich protein food to hay improved the digestibility of the latter, but not in as marked a way as was expected. Summary of Coefficients of Corn Bran. Seeies. Period. Sheep. Dry Matter. Ash. Protein. Fiber. Nitrogen- free Extract. Fat. XIX 13 I. 90.08 210.29 49.64 99.91 90.24 74.88 XIX.. . . . 13 II. 77.09 129.04 26.03 66.90 82.70 45.52 XXI IV. 78.23 - 58.78 67.49 82.27 63.57 XXI., . V. 81.89 - 22.56 96.10 86.33 65.16 XXI VI. 75.66 - 61.83 49.20 84.21 78.51 Average, 80.59 - 43.77 75 92 85.15 65.53 Average of prev ious trials (2), . 71 - 55 65 75 . Average of all i )revious trials (6), 71 - 60 71 80 80 The corn bran represents the hull or skin of the kernel, together with pieces of broken germ and more or less of the starchy portion which it is not possible to separate by mechanical means. It is often found in the markets of Massachusetts, and has been offered at a very reasonable price. In dry matter it contained 1 .08 per cent, ash, 6.87 per cent, protein, 13.86 per cent, fiber, 76.33 per cent, extract matter and 1.86 per cent. fat. While low in ash and protein, its fiber content is not excessive, and it is quite rich in extract matter. The hay-gluten meal-starch combination served as the basal ration. For some reason Sheep I., as indicated by the digestion coefficients, ap- peared to have utilized the bran quite fully. The results secured with the other sheep were as uniform as was to be expected, although Sheeji II. and V. apparently made less use of the protein, while the latter sheej) gave a high coefficient for the fiber. 1 See brief r6sum6 in Die Erniihrung d. landw. Nutzthiere, by Kellner, sixth ed., pp. 53, 54. DIGESTION EXPERIMENTS WITH SHEEP. 317 The results are higher than those formerly secured by us, where the corn bran was fed together with hay, exceptiog those for protein and fat. It is evident that the fiber is quite well digested, much more so than that contained in wheat and oats. Comparing the corn bran with corn meal on the basis of net energy values it is found that if corn meal is placed at 100 com bran equals 82. Summary of Coefficients of Distillers' Grains. Series. Period. Sheep. Dry Matter. Ash. Protein. Fiber. Nitrogen- free Extract. Fat. XIX 12 IV. 65.. 79 - 79.47 16.21 70 55 93.22 XXI 6 IV. 64.88 64.51 77.20 43.89 63.17 81.99 XXI 6 V. 67.54 21.21 74.23 62.67 66.63 82.48 XXL, . 6 VI. 67.94 23.21 77.24 54.98 69.16 77.10 Average 66.54 36.31 77.05 44.44 67.38 83.70 Average of all prcN-ious trials for corn grains (17). Average of all previous trials for rye grains (2). 79 58 - 73 59 95 81 67 95 84 The object of this experiment was to study particularly the digestibility of the fiber. For this purpose the grains were added to the hay-Diamond- gluten-meal-starch basal ration, which was quite low in that ingredient. Distillers' grains represent the residues from the manufacture of dis- tilled spirits. Those containing a high protein percentage are derived largely from corn. On the basis of 10 per cent, water the two samples contained 26.51 and 23.76 per cent, of protein, and may be considered of fair quality. The best grades usually contain 30 or more per cent, of protein. On the dry matter basis the average of the two samples con- tained 2.07 per cent, ash, 27.92 per cent, protein, 13.67 per cent, fiber, 46.69 per cent, extract matter and 9.65 per cent, fat. In the present experiments variations are observed in the percentages of the several ingredients digested. It is rather surprising that such differences occur in the percentages of fiber digested. It is evident, in spite of the low fiber content of the basal ration, that the sheep did not utilize the fiber from the distillers' grains very well, which indicates that other grains than corn were used in the mash. Previous trials with corn grains showed higher coefficients for the total dry matter and for the extract matter and fat (see above), while the coefficients for the fiber were believed to have been "too high. It seems probable that in the former trials, where the distillers' grains were fed with hay, the addition of the former increased the digestibility of the hay fiber. It is believed that the extent of the digestibility of distillers' grains will depend upon 318 MASS. EXPERIMENT STATION BULLETIN 181. the kind of grains composing the mash. If much rye, barley and wheat are used the coefficients, especially those for fiber, will be lower than when corn is the predominating grain. Summanj of Coefficieiits of Feterita. Series. Period. Sheep. Dry Matter. Ash. Protein. Fiber. ' Nitrogen- free Extract. Fat. XXI XXI 12 12 V. VI. 74.36 74.65 - 54.79 46.55 - 87.58 87.94 60.70 66.69 Average, Texas Station, i Corn for comparison (12), . 74.51 88.99 90 - 50.67 90.03 74 50.00 57 87.58 96.60 94 60.70 74.52 93 Feterita, or Sudan durra, is one of the grain sorghums, which include also Kafir, milo, durra and kaoliang. According to Morrison "it has slender stems carrying more leaves than milo but less than kafir, and erect heads bearing flattened seeds. Over much of the drier western portion of the grain sorghum belt these crops are more sure, and even on good soil return larger yields than corn." It has been stated that the average crop is 25 bushels per acre, with a maximum of 80 bushels (56 pounds) for feterita. The sample tested by us came from a carload received by an eastern grain dealer, and contained 10.41 per cent, water. Its dry matter consisted of 1.80 per cent, ash, 13.23 per cent, protein, 1.40 per cent, fiber, 80.23 per cent, extract matter and 3.34 per cent, fat. In chemical composition it resembles corn, being a little higher in protein and lower in fat. Hay and gluten feed served as a basal ration, and the feterita constituted 30 per cent, of the total ration. The results of the trial agree closely. It is surprising, however, that in total dry matter the coefficients fall so much below corn. Neither the protein nor the fat appear to be as well digested; the extract matter, however, ap- proaches in digestibility that contained in corn. Com contains sub- stantially 85.7 pounds of digestible organic nutrients in 100, and on the basis of our results feterita contains 71.06 pounds, thus indicating that the latter has only 83 per cent, of the nutritive value of corn. There are no data from which to compute its net energy value. It is doubtful, however, if such data would show any wide variations from that secured as a result of digestion data. Further experiments with the feterita should be made, however, before drawing positive conclusions.' 1 See note 2. 2 Since the above was written, Fraps of the Texas Station, Bui. No. 203, reports results with this grain showing higher digestion coefficients than those secured by ourselves. These co- efficients are inserted above, together with our own. DIGESTION EXPERIMENTS WITH SHEEP. 319 Summary of Coefficients of Alfalfa. Series. Period. Sheep. Dry Matter. Ash. Protein. Fiber. Nitrogen- free Extract. Fat. XXII XXII XXII XXII 12 12 14 14 IV. VI. XII. XIII. S8.10 54.75 59.61 58.50 41.50 28.10 50.15 50.68 73.16 68.19 72.43 73.32 46.92 45.15 48.28 45.23 66.48 63.70 67.48 66.81 16.66 15.27 30.24 32.29 Average Average all previous trials third cutting (6). Average all previous trials (109), . 57.74 58 60 42.61 44 50 71.78 70 71 46.40 40 43 66.12 70 72 23.62 42 38 The alfalfa was quite free from foreign material. It represented the third cutting, and was grown in the State of New York. It averaged in dry matter 6.49 per cent, ash, 15.34 per cent, crude protein, 35.06 per cent, fiber, 41.13 per cent, extract matter and 1.98 per cent, crude fat. The results are satisfactory and are quite uniform with those previously secured. The fiber in alfalfa hay has relatively a low, and the protein a high, digestibility. Roots and Vegetables. It is generally assumed that roots and vegetables are quite fully di- gested by animals. Relatively few digestion trials have been made to determine the rate of digestibility and to note the effect, if any, of such materials upon the digestibility of feeds with which they are fed. (a) Cabbages. The whole cabbage, the head minus the outside leaves, and the leaves themselves were analyzed and digestion experiments carried out. The whole cabbage contained 88.27 per cent, water, and its dry matter con- sisted of 12.20 per cent, ash, 21.82 per cent, protein, 10.30 per cent, fiber, 53.76 per cent, extract matter and 1.92 per cent. fat. The heads minus leaves contained 90.34 per cent, water, and the dry matter consisted of 8.22 per cent, ash, 17.98 per cent, protein, 9.84 per cent, fiber, 62.77 per cent, extract matter and 1.19 per cent. fat. The outside leaves contained 80.95 per cent, water, and the dry matter consisted of 14.49 per cent, ash, 11.94 per cent, protein, 13.12 per cent, fiber, 58.04 per cent, extract matter and 2.41 per cent. fat. The exterior leaves contained about twice as much dry matter as the heads. Cabbage is rich in protein, — in fact, considerably richer than the legumes, — on an equal moisture basis. It is rich also in ash, particularly the leaves, which may have been due in part to the adherence of soil particles. The percentages of fiber and fat are relatively low. 320 MASS. EXPERIMENT STATION BULLETIN 181. The cabbage was fed in combination with hay, and constituted 25 to 34 per cent, of the dry matter of the total rations, the latter having nutritive ratios of from 1 :6.6 to 1 :9. Summary of Coefficients for Cabbage. Whole Cabbage. Series. Period. Sheep. Dry Matter. Ash. Protein. Fiber. Nitrogen- free Extract. Fat. XIX XIX 7 7 I. II. 89.35 86.49 59.74 54.19 84.59 87.67 109.57 72.48 95.50 96.22 71.11 68.33 Average, 87.92 56.97 86.13 91.03 95.86 69.72 Heads Minus Leaves. XVIII.. XVIII., . . 4 4 I. II. 99.84 95.81 80.55 74.02 84.92 124.79 08.15 99.74 103.16 101.48 53.28 32.07 Average, 97.83 77.29 76.54 112.27 102.32 42.67 Leaves. XVIII., XVIII., . . 5 5 I. II. 76.84 71.39 45.71 66.69 44.23 60.90 80.66 75.79 87.38 81.30 45.37 29.40 Average 74.12 44.97 63.80 78.23 84.34 37.39 The whole cabbage was quite well digested, with an average dry matter percentage in case of the two sheep of 88 per cent. The fiber averaged 91 per cent, digestible, showing in case at least of one of the sheep that it had improved the digestibility of the fiber in the hay. The extract matter also had a high digestibility (96 per cent.). The heads proved rather more digestible than the whole cabbage, namely, 98 per cent., the protein 77 per cent., and both the fiber and extract matter over 100 per cent. It seems evident that the cabbage exercised a beneficial effect upon the hay with which it was fed. The leaves did not prove as digestible as the center, although one notes that the dry matter averaged 74 per cent, digestible, the protein 64 per cent., the fiber 78 per cent, and the extract matter 84 per cent. The whole cabbage, head minus leaves, and leaves would contain of digestible organic matter, on the basis of our data, in 2,000 pounds, the following: — DIGESTION EXPERIMENTS WITH SHEEP. 321 Water (Per Cent.). Protein (Pounds) Fiber (Pounds), Extract Matter (Pounds). Fat (Pounds). Total Fat X 2.2 (Pounds) Nutritive Ratio. Whole cabbage, Head, Leaves, 88.3 90.3 81.0 43.88 26.73 29.02 21.92 19.32 38.80 120.74 123.20 185.24 3.12 1.17 3.38 193.40 171.82 260.50 1:3.4 1:5.4 1:8.0 Because of the less moisture content the leaves show a larger amount of total organic nutrients than either the total cabbage or the interior. On the basis of 88.3 per cent, water, — that found in the Avhole cabbage, — the interior shows 207.2 and the leaves 160.4 pounds of digestible organic nutrients per ton. The whole cabbage, head and leaves, have the following relative values based upon digestible organic nutrients and natural moisture, or an equivalent moisture content of 88.3 per cent.: — Natural Moisture Basis. Equal Moisture Basis. Whole cabbage Head Leaves 100 89 135 100 106 83 (6) Carrots. Summary of Coefficients of Carrots. Series. Period. Sheep. Dry Matter. Ash. Protein. Fiber. Nitrogen- free Extract. Fat. XIX.. . 8 I. 89.10 33.48 52.03 131.89 95.66 79.63 XIX., . 8 II. 94.42 46.40 77.87 154.59 99.91 91.20 XX., . 8 IV. 74.42 50.22 77.71 40.19 85.71 - XX.. . 8 v. 87.81 64.89 85.35 101.96 93.22 25.87 XX., . 8 VL 83.73 43.77 86.53 82.06 93.04 9.95 XX., . 9 IV. 100.70 74.24 87.61 89.58 105.20 162.90 XX., . 9 V. 115.80 91.86 106.00 148.71 113.51 204.84 XX.. . 9 VL 135.05 96.15 127.94 197.52 130.76 228.23 Average, 100.95 64.40 89.05 129.47 104.75 114.66 Two different lots of carrots were fed. They averaged 87.64 per cent, water, and in dry matter contained 9.56 per cent, ash, 10.11 per cent, protein, 8.53 per cent, fiber, 70.71 per cent, extract matter and 1.09 per 322 MASS. EXPERIMENT STATION BULLETIN 181, cent. fat. They are low in protein, fiber and fat, and quite high in ash and in extract matter. In the first and second experiments they were fed in combination with hay, and constituted about 30 per cent, of the total dry matter which had a nutritive ratio of 1 : 10 to 1 :13.6. In the third experiment they were fed together with hay and gluten feed, and composed about 15 per cent, of the dry matter of the ration, which had a nutritive ratio of 1 :7.6. Sheep rV. in Series XX., Period 8, showed such a low rate of digestibility that the results were not included in the average. With this exception the coefficients resulting from the hay and carrot combination agree reasonably well, and show 88.76 per cent, of the dry matter to have been digested. The protein and exi;ract matter are also shown to have been quite well assimilated. The fat is so small in amount that the results have no particular meaning. In most cases a high fiber digestibility is observed; in fact, more was apparently digested than was consumed. Where the carrots were fed with hay and gluten feed more of the dry matter was apparently digested than was fed. Thus one observes co- efficients of 117 for the dry matter, 107 protein, 145 fiber and 116 extract matter. This, it is believed, was due to the coefficients used for the digestibility of the basal ration, composed of hay and gluten feed. These coefficients for some reason averaged only 62.43 for the dry matter, as against 68.4, the average for all of the other experiments. If, however, one uses the average figure of 68.4, the coefficients for the dry matter of the carrots varj' from 67.4 to 101.66. The coefficients as a whole indicate that carrots were quite fully utilized, and that they seemed to improve the digestibility of the basal ration with which they were fed. It is proposed to study this matter more fully. (c) Mangels. Summary of Coefficients of Mangels. Series. Period. Sheep. Dry Matter. Ash. Protein. Fiber. Nitrogen- free Extract. Fat. XVIII., XVIII., . . XVIII., XVIII., 3 3 6 6 V. VI. V. VI. 86.90 88.74 85.43 87.20 1.55 18.12 41.31 52.36 41.21 51.18 48.36 63.00 103.45 103.81 89.58 85.31 95.79 96.29 93.40 93.67 - Average Average of all previous trials (6), 87.07 84 30.58 50.94 59 95.54 78 94.76 94 - Four single trials were carried out with one lot of mangels which con- tained 83.10 per cent, of water, — less than is found usually in this root. In the dry matter there was 6.10 per cent, ash, 5.84 per cent, protein, DIGESTION EXPERIMENTS WITH SHEEP. 323 6.38 per cent, fiber, 81.40 per cent, extract matter and .28 per cent. fat. The mangels were very low in protein, fiber and fat, and high in extract matter. They were fed in combination with hay only, and constituted from 40 to about 47 per cent, of the total dry matter of the combined ration, which had a nutritive ratio of 1:11 to 1:13. The coefficients are quite satisfactory, showing the dry matter to be 87, the protein 51 and the fiber and extract matter 95 per cent, digested. It is possible that the mangels improved the digestibility of the hay somewhat, but it is regretted that they were not fed also with a combination of hay and a protein concentrate in order to note if they would not have had a more pronounced effect. (d) Pumpkins. Summary of Coefficients of Entire Pumpkins. Series. Period. Sheep. Dry Matter. A.h. Protein. Fiber. Nitrogen- Ex^rl^t. Fat. XIX 6 I. 75.87 64.82 70.50 59.74 81.54 96.29 XIX., 6 II. 89.32 63.93 80.69 86.30 98.12 96.87 XX., 2 I. 81.62 70.96 67.89 65.20 90.84 89.27 XX.. 2 II. 88.23 62.99 76.20 83.59 96.40 91.76 XX., 3 I. 78.80 68.35 83.63 47.80 86.30 88.10 XX., 3 II. 75.41 49.82 82.57 46.23 83.83 84.69 XX.. 4 I. 75.57 76.91 74.81 38.49 83.82 94.23 Average 80.69 65.40 76.61 61.05 88.69 91.60 Pumpkins minus Seeds and Connecting Tissue. XIX XIX 4 4 I. II. 109.23 105.13 93.84 59.48 92.55 93.96 137.52 95.16 108.99 102.44 101.44 83.81 Average, 101.54 82.31 93.26 116.34 105.72 92.63 Two lots of pumpkins, grown on two different farms in successive years, were used. One lot was tested whole, and also without the seeds and connecting tissue. The whole pumpkins averaged 87.53 per cent, water, and the dry matter contained 7.74 per cent, ash, 15.60 per cent, protein, 15 per cent, fiber, 49.37 per cent, extract matter and 12.29 per cent. fat. The edible portion contained 94.58 per cent, water, and its dry matter consisted of 8.81 per cent, ash, 13.74 per cent, protein, 17.33 per cent, fiber, 57.56 per cent, extract matter and 2.56 per cent. fat. Wider variations occur in the digestibility of the different ingredients by the two sheep than are desirable. In case of Series XX., Periods 3 324 MASS. EXPERIMENT STATION BULLETIN 181. and 4, where the pumpkins were fed with a basal ration of hay and gluten feed, the coefficients for the fiber, extract matter and fat appear to be lower than when the basal ration consisted of hay only. One would expect contrary results, for the combination of hay and pumpkins had a nutritive ratio of 1:9 to 1:11, and the hay, gluten feed and pumpkins a ratio of approximately 1:7.5. The lower digestibility of the pumpkins in the hay-gluten-feed-pumpkin ration may have been caused by the extra amount of total dry matter fed (approximately 100 grams daily). The coefficients for the pumpkins minus the seeds are considerably higher, and, so far as one is able to judge from the results, indicate that the pumpkins had a favorable effect upon the digestibility of the hay. When the entire fruit was fed no seeds or parts of seeds were found in the feces. In general, it may be said that the entire pumpkins appear to be fairly well digested, but not quite as fully as are mangels, turnips and carrots. Their relative feeding values will depend considerably upon their content of dry matter. The large percentage of fat in the pumpkin tends to increase slightly its feeding value pound for pound over most of the root crops. (e) Turnips. Summary of Coefficients of Turnips. Series. Period. Sheep. Dry Matter. Ash. Protein. Fiber. Nitrogen- free Extract. Fat. XVIII., XVIII.. . . 7 7 V. VI. 88.78 89.17 55.34 51.38 70.15 81.08 87.75 75.55 95.48 96.64 56.90 75.86 Average, 88.98 53.36 75.62 81.65 96.06 66.38 One lot only of Swedish turnips was tested, which contained 86.21 per cent, water; the dry matter tested 7.33 per cent, ash, 9.58 per cent, protein, 10.99 per cent, fiber, 71.31 per cent, extract matter and .79 per cent. fat. They were rather richer in protein and fiber than mangels, and somewhat lower in carbohydrate matter. At the same time they may be regarded as carbohydrate in character. They were fed together wdth hay, and constituted 38 per cent, of the total ration, which had a nutritive ratio of 1 :10.4. The results with the two sheep agree very closely, the sheep digesting 89 per cent, of the dry matter, 76 per cent, of the pro- tein, 82 per cent, of the fiber and 96 per cent, of the starchy matter. DIGESTION EXPERIMENTS WITH SHEEP. 325 Com-parative Sumvic ry of Coefficients for Roots and Vegetables. Digestion Coefficients. Digestible Or- Water (Per Cent.). Dry Matter. Ash. Protein. Fiber. Nitrogen- free Extract. Fat. trients in 2,000 Pounds (Basis, 88 Per Cent. Water). Whole cabbage, 88 88 57 88 91 96 70 193 Carrots, . 88 101 64 89 129 105 115 233 Mangels, . 83 87 31 51 96 95 196 Turnips, . 86 89 53 76 82 96 66 201 Pumpkins, 88 81 65 77 61 89 92 212 The total dry matter of the carrots appears to be more fully digestible and the dry matter of the pumpkin less digestible than that of the mangels, turnips and cabbage, the coefficients of which are quite uniform. The protein shows a high and uniform digestibility excepting that contained in the mangels. The fiber — excepting in the pumpkins, with its hard shell and seed covering — is shown to be quite well digested, as is also the extract matter. The fat is not of much consequence excepting in the pumpkin, which contains over 12 per cent, with a high digestion coefficient. On a uniform moisture basis of 88 per cent., the total digestible organic nutrients (including the fat multiplied by 2.2) do not vary widely from each other, with the exception of the carrots, which merit further study. Summary of Coefficients of Vegetable Ivory Meal. Series. Period. Sheep. Dry Matter. Ash. Protein. Fiber. Nitrogen - free Extract. Fat. XIX., . V. 84.43 44.35 - 55.01 93.27 - XIX., VI. 89.63 17.99 30.04 85.82 93.89 45.45 XX., IV. 88.20 63.59 10.57 76.08 93.60 39.01 XX., V. 98.96 94.01 34.61 120.48 99.99 60.28 XX., VI. 101.71 44.24 41.70 129.22 102.28 31.91 XXI., 3 IV. 78.59 193.81 1.59 51.07 85.81 61.82 XXI., 3 V. 81.04 - 4.77 100.06 89.94 47.10 XXI., 3 VI. 84.03 90.95 6.04 66.45 91.98 58.70 Average, . 88.33 • 78.42 18.47 85.52 93.84 49.18 Corn meal for com parison, 88 - 67 92 90 326 MASS. EXPERIMENT STATION BULLETIN 181. This material represents the sawdust or shavings from the vegetable ivory, or the coro^o nut {Phytelephas macrocarpa). A complete report on its composition, digestibility and feeding value has been published elsewhere.^ The details of the several digestion tests, however, were not given. The nut is used in the manufacture of buttons and similar ma- terials; the residue is practically tasteless and of a tough, horny nature. Animals will not eat it when fed by itself, but usually consume it readily if mixed with one or more grains. It averaged in composition 10.76 per cent, water, and in dry matter 1.25 per cent, ash, 5.36 per cent, crude protein, 8.01 per cent, fiber, 84.37 per cent, extract matter and 1.01 per cent. fat. Its extract or carbohydrate matter is nearly all in the form of mannan, yielding mannose on hydrolysis. The material in all cases was fed with 550 grams of hay and 150 grams of gluten feed as a basal ration, and constituted some 30 per cent, of the total ration, A glance at the results show that the coefficients secured in Period 13 (hitherto unpublished) are noticeably above the others. This is believed to have been caused by the use of the coefficients secured for a basal ration of hay and gluten feed, which gave 62 as the digestibility of the dry matter as against 66 for the basal ration of hay and gluten feed employed in the other experiments. The average of the coefficients secured in Periods 5 and 3 (as published) gave 84 for the dry matter and 92 for the extract matter, and are believed to be more nearly correct. The coefficients secured for the protein, fiber and fat are not surprising, in view of the smallness of the amounts present in the ivory meal in com- parison with the total amounts of these ingredients consumed. The larger part of the ivory meal consists of carbohydrate matter, and it was quite well digested. How the mannan was decomposed in the digestive tract is not clear; it was found, however, to have largely disappeared in the feces. The ivory meal evidently is as fully digested as corn meal, and our published results of experiments with dairy animals demonstrate it to have considerable nutritive value. Summanj of Coefficients of Vinegar Grains. Series. Period. Sheep. Dry Matter. Ash. Protein. Fiber. Nitrogen - free Extract. Fat. XXII XXII XXII XXII 9 9 10 10 IX. XI. IV. vi. 54.77 65.01 65.60 67.48 29.08 62.91 59.28 69.47 66.00 47.02 50.59 60.92 73.87 50.92 52.63 54.71 65.60 84.20 88.08 89.30 68.70 Average Dried brewers' grains for comparison (5). 60.70 61 64.42 81 58.10 49 55.97 57 82.57 89 » Beals and Lindsey: Journal of Agricultural Research, Vol. VII., No. 7. DIGESTION EXPERIMENTS WITH SHEEP. ;27 Vinegar grains were put out by the Fleischmann Company, Chicago, and represent the residue in the manufacture of yeast, or possibly of yeast and distilled liquors. They tested 7.63 per cent, water, and the dry matter contained 2.54 per cent, ash, 20.39 per cent, protein, 20.12 per cent, fiber, 50.33 per cent, extract matter and 6.62 per cent. fat. They were fed together with hay to four sheep. For some reason Sheep IX. and XI. did not digest them as well as did Sheep IV. and VI. The average results from the four sheep show that in total digestible matter, fiber and extract they compare well with dried brewers' grains, although the pro- tein of the latter is more completely utilized. They are certainly an addition to our supply of protein concentrates, and can be used in the grain ration in a similar way to dried brewers' grains. Summary of Coefficients of New Bedford Garbage Tankage. Series. Period. Sheep. Dry Matter. Ash. Protein. Fiber. Nitrogen- free Extract. Fat. XXI XXI.. . XXI., . . . 8 S 8 IV. 1 V. VI. 54.22 77.33 81.12 63.14 73.06 74.22 33.90 30.02 45.18 145.8 116.8 68.04 87.18 92.01 100.0 147.0 Average 79.22 73.64 37.6 131.3 89.6 123.5 Excluded from average. This tankage represents the garbage collected in the city of New Bed- ford which was treated by the so-called Cobwell process. Briefly stated, the method of treatment consists in removing, so far as possible, from the material as received, all glass, tin cans, banana and orange peel, after which the residue is placed in large iron tanks and treated with benzine to re- move the fat, which process also takes out the larger part of the water. It is then run over conveyors, and any other objectionable material is removed, after which it is ground. The tankage contained 8.53 per cent, water, and in dry matter 15.72 per cent, ash, 22.02 per cent, protein, 9.67 per cent, fiber, 50.92 per cent, extract matter and 1,67 per cent. fat. It. was in good mechanical con- dition, was fed with hay and gluten feed, and constituted about 18 per cent, of the ration, which had a nutritive ratio of 1:7. Sheep IV. digested the tankage poorly, and it has seemed wise to exclude the coefficients from the average of those secured with the other two sheep. The protein was not well digested^ which indicated its inferiority as compared with material derived from slaughterhouses. This was con- firmed by subjecting the tankage to the action of the alkaline permanganate method for determining nitrogen availability, and the securing of an 328 MASS. EXPERIMENT STATION BULLETIN 181. availability coefficient of 44.66. Any nitrogenous matter testing below 50 by this method is considered of poor quality. The extract matter was quite well utilized, and likewise the small amount of fat. The fiber for some reason appeared to be completely digested, which is not probable. The non-nitrogenous matter of the tankage was quite well utilized, but the protein is likely to prove inferior to the better grades of animal or vegetable nitrogenous concentrates. Summary of Coefficients of New Bedford Pig Meal. Series. Period. Sheep. Dry Matter. Ash. Protein. Fiber. Nitrogen- free Extract. Fat. XX.. . . . XX 11 IV. VI. 68.99 69.29 48.02 40 96 67.35 71.39 18.45 26.18 84.02 83.40 133.77 142.88 Average 69.14 «,« 69.37 22.32 83.71 138.33 I This material according to the manufacturers was composed of 73 per cent, garbage tankage, 18 per cent, standard middlings, 7 per cent, pre- pared molasses feed and 2 per cent, linseed meal. It tested 8.80 per cent, water, and the dry material consisted of 19.65 per cent, ash, 23.59 per cent, protein, 9.15 per cent, fiber, 44.30 per cent, extract matter and 3.31 per cent. fat. The sheep digested the entire mixture fairly well. Evidently the addition of the vegetable concentrates improved the digestibility of the total protein in the mixture. The fiber was poorly digested, but the extract matter and particularly the fat showed high coefiicients. It is quite reasonable to assume that garbage tankage is likely to vary considerably in quality. S'ummmy of Coefficients of Rowen. Series. Period. Sheep. Dry Matter. Ash. Protein. Fiber. Nitrogen- free Extract. Fat. XXII 15 XXII 15 XII. XIII. 60.81 61.16 34.23 35.76 60.39 60.27 68.12 68.08 63.37 63.57 30.01 34.53 Average, Average previous trials (12), 60.99 65 34.40 60.33 70 68.10 66 63.47 65 32.27 47 Rowen represents the second growth of meadows, and contains in addition to the grasses a considerable admixture of clover. The samples tested contained 9.13 per cent, of water, and in dry matter showed 7.19 DIGESTION EXPERIMENTS WITH SHEEP. 329 per cent, ash, 8.14 per cent, protein, 49.02 per cent, extract matter, 2.39 per cent, fat and 33.26 per cent, fiber. While of satisfactory appearance it was inferior in composition to the average, which has been shown to test 11.4 per cent, protein and 24.1 per cent, fiber on a 14 per cent, water basis. The digestion tests agree exceedingly well, but confirm the analysis, showing it to be rather less digestible than the average of previous trials. Summary of Coefficients of Soy Bean Hay. Series. Period. Sheep. Dry Matter. Ash. Protein. Fiber. Nitrogen- free Extract. Fat. XX XX 7 7 V. VI. 52.27 61.03 11.63 29.26 70.90 78.86 49.36 55.75 54.78 64.72 53.91 64 71 Average Average previous trials (4), 56.65 60 20.44 74.88 73 52.56 57 59.75 64 59.31 44 The medium green soy beans were grown upon the station grounds, and were cut to put in the silo about the middle of September. They had not sufficiently matured to warrant their use as a seed crop. At the time of making the test the hay contained 11.73 per cent, of water, and, on a dry matter basis, 6.63 per cent, ash, 15.86 per cent, protein, 34.88 per cent, fiber, 40.56 per cent, extract matter and 2.07 per cent. fat. The tough, fibrous nature of the straw is in evidence in the high fiber content of the hay. Sheep V. was not able to digest the hay as well as Sheep VI. The results for the latter sheep agree fairly well with the average of the four other trials reported. With the exception of the protein the ingredients in soy bean hay appear to be about equal in digestibility to those contained in average English hay. The higher digestibility of the protein is due to the presence of the beans. It is believed soy beans should be ensiled with com rather than made into hay. Summary of Coefficients of Stevens' "44-" Dairy Ration. Series. Period. Sheep. Dry Matter. Ash. Protein. Fiber. Nitrogen- free Extract. Fat. XXII XXII 11 11 IV. VI. 72.55 68.58 26,03 82.14 77.23 45.36 55.01 72.58 69.23 91.88 71.08 Average 70.57 26.03 79.69 50.19 70.91 81.48 330 MASS. EXPERIMENT STATION BULLETIN 181. The Stevens' "44" Dairy Ration is one of the numerous proprietary dairy rations offered in Massachusetts markets. It is claimed to be a mixture of a great variety of the most desirable grains and by-products. It had 8.94 per cent, water, and in dry matter 4.17 per cent, ash, 26.95 per cent, protein, 12.88 per cent, fiber, 49.56 per cent, extract matter and 6.44 per cent. fat. Its high fiber content indicated the presence of some unsatisfactory material, and this was confirmed by the digestion test. The mixture proved to be fairly well digested, but not equal in total digestibility to mixtures of bran, cottonseed meal, gluten feed and corn or hominy meal. The fiber digestibility was considerably below that secured for hay, while the extract matter was below what one would expect in high-grade material. The protein, on the other hand, was quite well digested. Dige&tibility of Sudan Grass. This grass (Andropogon sorghum var.) was introduced into the United States in 1909, and has been tried at this station for a number of years, A full report on its merits will be given elsewhere. The green material contained from 76.5 to 80.42 per cent, of water when cut, and the hay averaged 14.47 per cent, of water. On the basis of dry matter the two samples of green material averaged 6.84 per cent, ash, 13 per cent, crude protein, 29.10 per cent, fiber, 47.13 per cent, extract matter and 3.93 per cent. fat. The hay averaged in drj- matter 8.93 per cent, ash, 13.85 per cent, crude protein, 33.85 per cent, fiber, 41.80 per cent, extract matter and 1.53 per cent. fat. The green material was fed with English hay, and the ration had a nutritive ratio of 1:8.3. The Sudan hay in three out of four experiments was fed exclusively, and had a nutritive ratio of 1 :5.7. Summary of Coefficients of Sudan Grass. Series. Period. Sheep. Condition of Grass. Dry Matter. Ash. Pro- tein. Fiber. Extract Matter. Fat. XXII. XXII. 17 17 XII. XIII Green, first crop (heading). Green, first crop (heading). 77.23 70.84 86.93 39.96 79.61 79.16 85.45 77.97 69.63 68.97 86.59 80.34 Average, 74.04 63.45 79.39 81.71 69.30 83.47 XXII. XXII. 1 1 rv. Green, second crop, Green, second crop, 65.41 65.09 37.97 24.30 62.86 68.07 69.40 69.42 67.70 67.69 64.69 58.51 Average 65.25 31.14 65.47 69.41 67.70 61.60 xxn. XXII. 3 IV 3 VI. Dry, second crop, . Dry, second crop. 59.99 59.37 45.07 40.27 58.13 61.40 73.62 72.76 54.35 53.52 35.63 35.32 Average 59.68 42.67 59.77 73.19 63.94 35.48 DIGESTION EXPERIMENTS WITH SHEEP. 331 In Period 17, first crop, Sheep XII. digested the material rather better than Sheep XIII. In Period 1 the green material, second crop, scarcely in head, was cut and fed in September. At the same time, some of it was made into hay and fed later. The total dry matter of the hay was over 4 per cent, less digestible than the same material fed green. Strange to say, the fiber showed a somewhat higher digestibility in the hay, while the extract matter was noticeably less digestible. As might have been expected, the fat (ether extract) showed a lower digestibility in the hay, due probably to the fact that the sheep were able more thoroughly to extract such sub- stances out of the green plant. For some reason the sheep digested the second crop (green) less fully than they did the first. The latter was cut in 1917, and the former in September, 1916. Whether the lessened digestibility was due to the climatic variations prevailing in two different years, or because a second growiih was actually not as digestible as the first, it is not possible to say. The average of the coefficients of the two lots of green Sudan grass follows, together with green barnyard millet, sorghum and com for comparison. A verage Coefficients for Comparison. Number of Different Lots. Single Trials. Dry Matter. Ash. Pro- tein. Fiber. Extract Matter. Fat. Sudan grass, Barnyard millet (blossom), Sorghum (past blossom), . Corn fodder (dent) milk, . 2 3 2 7 4 6 4 17 69.64 70.00 65.00 70.00 47.30 56.00 42.00 39.00 72.42 65.00 44.00 62.00 75.56 73.00 55.00 64.00 68.50 71.00 73.00 77.00 72.54 58.00 64.00 76.00 The above comparison indicates that Sudan grass in digestibility is fully equal to other important green feeds. Summary of Coefficients of Sudan Hay. Series. Period. Sheep. Character of Hay. Matter. Ash. Pro- tein. Fiber. Nitro- gen-free Ex- tract. Fat. XXII. XXII. XXII. 7 7 7 IX. XII. XIII. Before heading, first crop. Before heading, first crop. Before heading, first crop. 56.25 55.14 ,57.15 55.92 51.42 56.93 56.63 55.22 57.83 66.38 66.42 66.82 49.24 48.74 50.51 23.01 10.38 19.62 Average, 56.18 54.76 56.56 66.54 49.50 17.67 332 MASS. EXPERIMENT STATION BULLETIN 181. Summary of Coefficients of Sudan Hay — Concluded. Series. Period. Sheep. Character of Hay. Dry Matter. Ash. Pro- tein. Fiber. Nitro- gen-free Ex- tract. Fat. XXII. XXII. XXII. 6 4 4 IX. IX. XI. Heading, first crop, Full blossom, first crop. Full blossom, first crop. 59.19 68.11 54.72 55.48 53.50 42.25 64.36 62.37 47.73 68.40 66.33 64.80 51.57 51.48 48.39 28.00 42.73 34.61 Average, 56.42 47.88 55.05 65.57 49.94 38.67 XXII. xxn. 3 3 IV. VI. Heading, second crop. Heading, second crop. 59.99 59.37 45.07 40.27 58.13 61.40 73.62 72.76 54.35 53.52 35.63 35.32 Average, Average of all of above 59,68 57.49 42.67 50.11 59.77 57.96 73.19 68.19 53.94 50.98 35.48 28.66 Results at Texas Experiment Station. Series. Period. Sheep. Character of Hay. Dry Matter. Ash. Pro- tein. Fiber. Nitro- gen-free Ex- tract. Fat. - 39 Iand2 Headed, . . . 30.00 17.70 63.10 57.60 48.70 - 60 land2 Full tassel, . - 23.50 58.30 58.60 41.80 45.20 - 62 land2 Headed, blooming, . - 15.00 64.20 60.20 52.60 61.10 - 73 land2 Late, mixed with crab grass. - 32.20 57.30 62.80 59.60 61.10 24.80 49.40 61.20 52.90 54.00 Timothy hay, for comparison 55 39.00 48.00 50.00 62.00 50.00 Barnyard millet, well headed 57 63.00 64.00 62.00 52.00 46.00 In the above trials an effort was made to note the digestibility of Sudan grass cut at successive stages of growth. The results do not indicate any particular difference. The second cutting of hay appeared to be more digestible than the first. Whether this would hold true in all cases is of course not established. It is just the opposite from the results secured with the green Sudan grass. The probability is that much will depend upon the climatic conditions prevailing during growth. If the weather should be warm, with plenty of sunlight and moisture, it is possible that the second growth would fully equal and perhaps e.xceed the first growth in digestibility. DIGESTION EXPERIMENTS WITH SHEEP. 333 Results recently reported^ from the Texas Experiment Station are somewhat below those secured by us, at least in case of the fiber. If one should eliminate the protein coefficient of Period 39 the remaining protein coefficients would be some two points above the Massachusetts figure. In all of the trials one notes particularly the high digestibility of the fiber and the low coefficients secured for the extract matter and fat. This holds true also for the millet. The digestibility of Sudan grass is shown to be above that for timothy, and equal to barnyard millet. The difficulty in curing satisfactorily the coarse grasses, of which Sudan and millet are examples, render them less satisfactory for hay than that ob- tained from the finer grasses. Digestibility of Sweet Clover. Sweet clover {Melilotus Alba) is a biennial legume found quite widely distributed in southern Canada and the United States. The two samples used were grown on the experiment station grounds. The clover was fed green to the sheep, beginning about June 12 and ending June 26. At the close of the trials the clover was budding to early blossom, and the lower portion of the stalks was woody. The two samples averaged 84.50 per cent, of water, and in dry matter contained 7.08 per cent, ash, 19.40 per cent, protein, 30.29 per cent, fiber, 40.10 per cent, extract matter and 3.13 per cent. ash. The green clover was fed with hay, and the rations had an average nutritive ratio of 1 :6.4. Summary of Coefficients of Sweet Clover Series. Period. Sheep. Condition of Clover. Dry Matter. Ash. Pro- tein. Fiber. Nitro- gen-free Ex- tract. Fat. XXI. XXI. 14 14 IV. VI. Early blossom, Early blossom. 64.80 73.30 47.93 48.96 75.29 78.58 60.56 78.58 64.07 74.00 49.91 50.28 Average, 69.05 48.44 76.93 69.57 69.03 50.10 XXII. XXII. 16 16 IX. XI. Budding, Budding, 66.67 72.61 ' 76.44 81.98 47.60 52.29 65.96 71.00 43.22 61.34 Average Average of both samples, .... Alfalfa' for comparison, . . . . . Clover^ for comparison 69.64 69.45 61.00 66.00 48.45 79.21 78.07 74.00 67.00 49.95 59.76 42.00 53.00 68.48 68.76 72.00 78.00 52.28 51.19 38.00 65.00 Bulletin No. 203, 1916. Henry and Morrison. 334 MASS. EXPERIMENT STATION BULLETIN 181, Siveet Clover Hay, Wyoming Station, Bulletin No. 78. Series. Period. Sheep. Condition of Clover. Dry Matter. Ash. Pro- tein. Fiber. Nitro- gen-free Ex- tract. Fat. XIV. 1.2.3 Rank, late cut, 60.88 65.79 75.46 33.63 72.04 30.94 Alfalfa I hay for comparison Clover 1 hay for comparison, .... 60.00 62.00 45.00 58.00 74.00 61.00 46.00 53.00 70.00 68.00 28.00 54.00 » Massachusetts Station'. Sheep IV. in Scries XXL, and Sheep IX. in Series XXII. did not seem able to digest the clover as well as the other two sheep. The slight variation in the stage of growth of the clover appeared to be without influence on its digestibility. The young sheep IX. and XL did not digest the fiber as well as did the old sheep IV. and VI. Sweet clover cut previous to blooming appeared to be quite well utilized, and showed rather higher coefficients than those for alfalfa or clover cut in bloom. The results of the Wyoming Station with sweet clover hay cut at an advanced stage of growth indicate that with the exception of the fiber it is as fully digestible as either alfalfa or clover hay. Table VI. Complete Summary of the Averages of All Coeffi- cients, ARRANGED ALPHABETICALLY. Number of Single Trials. Dry Matter Pro- tein. Nitro- gen-free Ex- tract. Alfalfa, . Cabbage (entire), . Cabbage (heads), . Cabbage (leaves), . Carrots, . Corn bran. Distillers' grains, . English hay — basal English hay and gluten feed — basal, English hay, potato starch and gluten meal (Diamond) — basal. English hay and wheat gluten flour {to note effect of the flour). Feterita, Gluten feed, 42.61 56.97 77.29 44.97 64.40 36.31 36.31 33.25 20.16 43.00 71.78 86.13 76.54 63.80 89.05 43.77 77.05 49.78 66.39 72.98 44.00 50.67 85.44 46.40 91.03 112.27 78.23 129.47 75.92 44.44 64.10 67.75 63.54 62.00 66.12 95.86 102.32 84.34 104.75 85.15 67.38 80.67 61.00 87.76 93.77 23.62 69.72 42.67 37.39 114.66 65.53 83.70 46.34 51.89 37.16 43.00 58.70 64.41 DIGESTION EXPERIMENTS WITH SHEEP. 335 Table VI. — Complete Summary of the Averages of All Coeffi- cients, ARRANGED ALPHABETICALLY — Concluded. Ration. Number of IS Dry Matter. Ash. Pro- tein. Fiber. Nitro- gen-free Ex- tract. Fat. Gluten meal (Diamond),! 86.00 - 85.00 100.00 93.00 - Mangels 87.07 30.58 50.94 95.54 94.76 - New Bedford garbage tankage, 79.22 73.64 37.60 131.30 89.60 123.50 New Bedford pig meal, . 69.14 44.49 69.37 22.32 83.71 138.33 Pumpkins (entire). 80.69 65.40 76.61 61.05 88.69 91.60 Pumpkins (seeds removed), . 101.54 82.31 93.26 116.34 105.72 92.63 Rowen, 60.99 34.40 60.33 68.10 63.47 32.27 Soy bean hay, .... 56.65 20.44 74.88 52.56 59.75 59.31 Stevens' "44" Dairy Ration, 70.57 26.03 79.69 50.19 70.91 81.48 Sudan grass (green). 4 69.64 47.30 72.42 75.56 68.50 72.54 Sudan hay, ^ 57.49 50.11 57.96 68.19 50.98 28.66 Sweet clover (green), 69.45 48.45 78.07 57.76 68.76 51.19 Turnips, 88.98 53.36 75.62 81.65 96.06 66.38 Vegetable ivory meal, . 88.33 78.42 18.47 85.52 93.84 49.18 Vinegar grains 60.70 " 64.42 58.10 55.97 82.57 See page 312. BULLETIN No. 182 MARCH, 1918 MASSACHISETTS AGRICILTIJRAL EXPERIMENT STATION SOY BEANS AS HIMAN FOOD By ARAO ITANO This publication has been prepared to furnish information concerning the many forms in which soy beans may be utilized Requests for bulletins should be addressed to the Agricultural Experiment Station Amherst, Mass. Massachusetts Agricultural Experiment Station. OFFICERS AND STAFF. COMMITTEE. Charles H. Preston, Chairman, . Wilfrid Wheeler, Edmund Mortimer, Arthur G. Pollard, Harold L. Frost, The President of the College, ex officio. The Director of the Station, ex officio. Hathorne. Concord. Grafton. Lowell. Arlington. STATION STAFF. Administration. William P. Brooks, i Ph.D., Director. Fred W. Morse, M.Sc, Acting Director. Joseph B. Lindsey, Ph.D., Vice-Director. Fred C. Kenney, Treasurer. Charles R. Green, B.Agr., Librarian. Mrs. Lucia G. Church, Clerk. Miss F. Ethel Felton, A.B., Clerk. Agricultural Economics. Alexander E. Cance, Ph.D., In Charge of Department. Samuel H. DeVault, A.M., Assistant. Agriculture. William P. Brooks, ^ Ph.D., Agriculturist. Henry J. Franklin, Ph.D., In Charge of Cranberry Investi- gations. Edwin F. Gaskill, B.Sc, Assistant Agriculturist. Robert L. Coffin, Assistant. Botany. A. Vincent Osmun, M.Sc, Botanist. George H. Chapman, Ph.D., Research Physiologist. Paul J. Anderson, Ph.D., Associate Plant Pathologist. Orton L. Clark, B.Sc, Assistant Plant Physiologist. W. S. Krout, M.A., Field Pathologist. Mrs. F. W. Wheeler, B.Sc, Curator. Miss Ellen L. Welch, A.B., Clerk. On leave. Entomology. Henrt T. Fernald, 1 Ph.D., Entomologist. Burton N. Gates, Ph.D., Apiarist. Arthur I. Bourne, A.B., Assistant Entomologist. Stuart C. Vinal, M.Sc, Assistant Entomologist. Miss Bridie E. O'Donnell, Clerk. Horticulture. Frank A. Waugh, M.Sc, Horticulturist. Fred C. Sears, M.Sc, Pomologist. Jacob K. Shaw, Ph.D., Research Pomologist. Harold F. Tompson, B.Sc, Market Gardener. Miss Ethelyn Streeter, Clerk. Meteorology. John E. Ostrander, A.M., C.E., Meteorologist. Microbiology. Charles E. Marshall, Ph.D., In Charge of Department. Arao Itano, Ph.D., Assistant Professor of Microbiology. George B. Rat, B.Sc, Graduate Assistant. Miss Louise Hompe, A.B., Graduate Assistant. PlaHt and Animal Chemistry, Joseph B. Lindsey, Ph.D., Chemist. Edward B. Holland, Ph.D., Associate Chemist in Charge {Research Division). Fred W. Morse, M.Sc, Research Chemist. Henri D. Haskins, B.Sc, Chemist in Charge {Fertilizer Division). Philip H. Smith, M.Sc, Chemist in Charge {Feed and Dairy Division). Lewell S. Walker, B.Sc, Assistant Chemist. Carleton p. Jones, M.Sc, Assistant Chemist. Carlos L. Beals, M.Sc, Assistant Chemist. WiNDOM A. Allen, - B.Sc, Assistant Chemist. John B. Smith, - B.Sc, Assistant Chemist. Robert S. Scull, ^ B.Sc, Assistant Chemist. Bernard L. Peables, B.Sc, Assistant Chemist. Harold B. Pierce, fe.Sc, Assistant Chemist. James T. Howard, Inspector. Harry L. Allen, Assistant in Laboratory. James R. Alcock, Assistant in Animal Nutrition. Miss Alice M. Howard, Clerk. Miss Rebecca L. Mellor, Clerk. Poultry Husbandry. John C. Graham, B.Sc, In Charge of Department. Hubert D. Goodale, Ph.D., Research Biologist. Miss Grace MacMullen, B.A., Clerk. Miss Elizabeth E. Mooney, Clerk. Veterinary Science. James B. Paige, B.Sc, D.V.S., Veterinarian. G. Edward Gage,' Ph.D., Associate Profesi Pathology. John B. Lentz, « V.M.D., Assistant. ' On leave. 2 On leave on account of military service. of Animal CONTENTS Introduction, .... Chemical composition and digestibility Human food prepared from soy beans, Soy bean milk (toniu), The ordinary method employed in Japa Toniu from the soy bean meal Author's method, Synthetic toniu, Condensed, Evaporated (yuba), Soy bean curd (tofu). Fresh tofu, Frozen tofu (kori tofu), Fried tofu (abura-age), Baked beans, . Boiled beans, . Roasted beans, Powdered beans. Roasted, . Raw, Green beans, . Soy bean pulp (kara) , Fermented boiled beans (natto). Ripened vegetable cheese (miso). Soy bean sauce (shoyu), . Vegetable butter, ice cream, oil and lard, PAGE 1 2 3 3 3 4 4 5 5 5 10 pcblication of this document approved bt the Supervisor of Administration. BULLETIN No. 182. DEPARTMENT OF MICROBIOLOGY. SOY BEANS {GLYCINE HISPIDA) AS HUMAN FOOD. BY ARAO ITANO. INTRODUCTION. For centuries the importance of soy beans as human food has been well known in oriental countries. Kellner,i Atwater ^ and others ^ bear testimony to this importance by their studies of the chemical composi- tion, digestion and assimilation. Soy beans have furnished the chief source of protein to the people of Japan and China; they are in universal use, and have played the role of meat and milk for these nations. A lack of animals, the economic conditions and rehgious rites have all had their influence in making soy beans the leading protein food crop in this, one of the most densely populated sections of the globe. Although a great favorite and very important, the position of the white bean of the United States is scarcely comparable with the conspicuous place occupied by the soy bean in these eastern countries. It is the richest, cheapest and most productive of aU legumes, and is prepared by nearly as many methods for human consumption as cow's milk. At this particular time, when this country as well as others is searching out economical food and food production, it may be well to inquire into this article of food and its methods of preparation for humans, for it is doubtless one of the most promising in sight. This being a popular presentation, the technical and theoretical dis- cussions of the subject will be held for future treatment. Not only from the standpoint of food supply, but also from the standpoint of nitrogen supply to the soil and industrial uses, the soy bean occupies a very im- portant place. 1 O. Kellner: U. S. Dept. Com., Bur. For. and Dom. Com., Special Agents Series, No. 84, Pt. I., 35. 2 W. O. Atwater: Farmers' Bull. No. 142, 1902, U. S. Dept. of Agr. 3 The Japanese investigations. Bulletins from College of Agriculture, Tokyo and Sapporo, Japan. MASS. EXPERIMENT STATION BULLETIN 182. CHEMICAL COMPOSITION AND DIGESTIBILITY. Table I. — Chemical Composition of Dry, Ripened Soy Beans. ^ Sot Beans from — China. Hungary. France. United States of America (Goess- mann). Japan. Crude protein, . 38.69 31.21 34.92 33.36 42.05 Fat, 17.87 18.29 12.78 15.53 12.81 21.89 20.46 Crude fiber, 12.69 4.53 Starch, 3.49 3.51 3.53 - - Ash, 5.39 5.63 5.97 5.35 4.19 Other organic matter. 21.01 28.09 26.53 34.18 28.82 Table I. plainly indicates the very high percentage of protein, 31.21 to 42.05 per cent., and of fat, 15.53 to 21.89 per cent., which compares with beef (round steak), containing an average of 19 per cent, proteins and 12.8 per cent. fats. While it is a w^ell-estabhshed fact that these substances, namely, pro- teins and fats, are essential materials in animal nutrition, the results of recent investigations indicate that indi^Tidual proteins difi'er in their digestibility and nutritive value, and that this difference is due to the particular amino acids which they yield upon hydrolysis. The interpre- tation, however, of such experimental results as have been thus far secured is somewhat confused. In case of the soy beans, the digestibility of the crude protein and fat is estimated at somewhere between 65 and 92 per cent., and 70 and 80 per cent., respectively, by the different investigators, such as Oshima,2 KeUner ^ and others. Although these figures may not necessarily be indicative of actual food value, the relative merit of the soy bean as human food is very significant. The author feels that there is still much to determine in the case of vegetable and animal proteins, and that we have not yet reached the stage in our knowledge where definite recommendations can be made. Prauswitz'* conception, one of many, may have some bearing in the case of this particular food, for the preparation of soy beans does seem to have a distinctive effect upon their digestive and assimilative values. It is possible that the fundamental differences in the nature of the nutrients, or proteins, may be disregarded. The long-continued, successful use of soy beans in oriental countries, over two thousand years, cannot be con- sidered lightly in scientific interpretation. ' M. Inouye: Bull. 2, 209, 1894-97, College of Agriculture, Tokyo, Japan. * K. Oshima: Bull. 159, p. 191, 1905, U. S. Dept. of Agr., Office of Exp. Sta. ^ O. Kellner: U. S. Dept. Com., Bur. For. and Dom. Com., Special Agents Series, No. 84, Pt. I., p. 35. « Prauswitz: Ztschr. Biol., 35 (1897), p. 335. SOY BEANS AS HUMAN FOOD. 6 HUMAN FOOD PREPARED FROM SOY BEANS. The various food articles prepared from soy beans which are known to the author are named below (names in parentheses indicate the Japanese name) : — 1. Soy bean milk (toniu). Ordinary method employed in Japan Toniu from the soy bean meal. Author's method. Synthetic toniu. Condensed. Evaporated (yuba). 2. Soy bean curd (tofu). Fresh tofu. Frozen tofu (kori tofu). Fried tofu (abura-age). 3. Baked beans. 4. Boiled beans. 5. Roasted beans. 6. Powdered beans. Roasted. Raw. •7. Green beans. S. Soy bean pulp (kara). 9. Fermented boiled beans (natto). 10. Ripened vegetable cheese (miso). 11. Soy bean sauce (shoyu). 12. Vegetable butter and ice cream. 13. OU (table use). 14. Lard (cooking). Soy Bean Milk (Toniu). The author suggests a Japanese term, toniu, meaning milk from beans, to designate the liquid preparation from soy beans, the so-called "milk" from soy beans, to avoid confusion of terms. The toniu may be prepared by any one of the following processes, varying somewhat in quality and, accordingly, adaptation to use. The Ordinary Method employed in Japan. 1. Soak the beans in water for twelve hours at room temperature, changing the water frequently. 2. Grind the beans to a fine smooth paste by means of a grinder, prefer- ably a millstone, adding water to the ground mass from time to time, to the amount of three times the bulk of beans. 3. Boil the mass to foaming for one hour. 4. Strain through fine cheesecloth. The strained fluid should be white and opaque. Note. — The toniu thus prepared resembles cow's milk. This is indi- cated in Table II. Upon standing, fat globules separate out on the 4 MASS. EXPERIMENT STATION BULLETIN 182. surface. After standing several days souring takes place as in cow's milk. It can be used very satisfactorily for various family foods, as in the preparing of bread, cake, vegetable stews, soups, chocolate, candies, etc. It has a sUght vegetable flavor wliich may be objectionable to some people for drinking purposes, although it is used to a considerable extent in oriental countries. Table II. — Composition of Soy Bean Milk compared with Coiv's Milk {Per Cent.)} Soy Bean Milk. Cow's Milk. Water, Albuminoids, Fat Fiber, Ash, Non-nitrogenous extract including carbohydrates, . Milk sugar, 92.53 3.02 2.13 .03 .41 86.08 4.00 3.05 .70 5.00 Table II. indicates the similarity in composition between toniu and cow's milk. Toniu from the Soy Bean Meal} 1. Add water to the amount of five times the bulk of the bean meal. 2. Let it stand for twelve hours at room temperature. 3. BoU it to foaming for one hour. 4. Strain tlirough fine cheesecloth. The strained fluid should be white and opaque. Author's Method. 1. Add water to the amount of five times the bulk of the bean meal. 2. Inoculate the content with B. coli and with B. lactis ocrogenes as used in salt rising bread. 3. Let it stand for sixteen hours at room temperature. 4. Boil to foaming for one hour. 5. Filter through fine cheesecloth. 6. Add table salt to the amount of one-half teaspoonful per quart. The addition of 5 per cent, milk sugar (lactose) improves the taste, and may be desirable unless the milk is intended for diabetic patients. > M. Inouye: Bull. 2, 212, 1894-97, College of Agriculture, Tokyo, Japan. "- The soy bean meal may be obtained by grinding the beans in a wheat flour mill; a fine coffee mill works satisfactorily also. This preparation may be used in the same manner as the previous product. SOY BEANS AS HUMAN FOOD. 5 Note. — The advantage of this method over the others may be sum- marized as follows: — 1. Ehmination of disagreeable flavor. 2. Adjustment of taste. 3. Reducing the probability of flatulence in the alimentary canal. 4. Adaptability as a Uquid food for diabetic patients. The results of further investigation of the method and also of its nutri- tive value are withheld for the present. Synthetic Toniu. Toniu of very high quahty, which resembles cow's milk very closely in composition, can be produced through both chemical and biological means; in fact, the author has been informed that tliis end has been accomphshed in one of the London chemical laboratories. The author, however, doubts its practicabihty for domestic use. Condensed Soy Bean Milk {Condensed Toniu). '^ 1. Add 4 grams of dipotassium phosphate and 600 grams of cane sugar to 4 liters of soy bean milk. 2. Concentrate the solution in vacuo to a very thick Uquid. Note. — It can be used like condensed cow's milk for the preparation of chocolate, etc. It gives an agreeable taste, but has a very feeble odor of raw beans. Evaporated Soy Bean Milk (Yuba). 1. Boil the soy bean milk until a film is formed on the surface. 2. Collect the film and cut it in any shape desired. Note. — The film consists of coagulated albuminoids and fat. It may be used as an article of food, cooked in soup, etc. Soy Bean Curd (Tofu). Table III. — Chemical Composition of Some Preparations (Per Cent.).- Water. Protein. Fat. Carbo- hydrates. Ash. Fresh tofu, Frozen tofu, Fried tofu Tofu cake (kara). .... Yuba, 88.11 18.72 57.40 84.49 18.31 6.29 48.65 21.96 5.23 49.65 3.38 28.65 18.72 1.58 18.00 1.64 2.33 .57 8.04 11.82 .58 1.65 1.35 .66 2.22 Table III. indicates the chemical composition of various preparations from soy bean milk. The digestibiUty of the nutrients in tofu has been 1 T. Katayama: Bull. 7, 113, 1906-08, College of Agriculture, Tokyo. Japan. 2 K. Oshima: Bull. 159, 28, 1905, U. S. Dept. of Agr., Office of Exp. Sta b MASS. EXPERIMENT STATION BULLETIN 182. found to be as high as 95 per cent, for protein, 95 per cent, for fat, and 99 per cent, for carbohydrates.^ Thus the composition and the digesti- biUty of tofu estabhsh it as a very nutritive food substance. The methods of preparation of these articles will be given in the follow- ing pages. Fresh Curd (Tofu). 1. Prepare the soy bean milk either from whole beans or from bean meal as described previously. 2. Add 2 per cent, of any one of the following substances while it is hot, stirring constantly: — (a) Mother liquid of sea salt.^ (b) Magnesium and calcium chloride solution.^ (c) Saturated solution of alum.* (d) Vinegar.^ 3. Filter off the liquid. 4. Press the precipitate in a wooden frame, 5. Let the pressed curd float in a large quantity of fresh cold water in order to free the coagulum from chemicals added. Note. — In Japan tofu is prepared and sold in the market as baked goods are in this country. Its preparation may be too involved for the domestic kitchen. Among the coagulants the mother liquid of sea salt and the magnesium mixture are preferred to the others because the excess of these substances is almost completely removed by immersing in cold water. Frozen Tofu {Kori Tofu). 1. Cut the fresh tofu into small pieces. 2. Subject the pieces to freezing. 3. Dry in vacuo after freezing. Note. — The product thus prepared can be preserved for years and transported very easily. Freezing hastens the removal of water. The final product is porous and can be eaten in soups. Fried Tofu (Abura-age). 1. Cut the frozen tofu into the desired size. 2. Fry it in rape-seed oil, sesame-seed oil, or in a large quantity of lard until the surface becomes brown. Note. — It makes a very palatable, rich food, and may be eaten like fried egg or meat, or in soup. 1 When eaten with rice. 2 This is commonly used. ^ Mix the saturated solution of magnesium and calcium chloride in proportion of 4 : 1. (The author's recommendation.) * Recommendation of the author. 5 Recommendation of the author; ordinary table vinegar. SOY BEANS AS HUMAN FOOD. Baked Beans. 1. Soak the beans, suspended in a cloth bag, in a large quantity of hot water over night. (Soaking for twenty-four hours in cold water which is changed occasionally will give the same result.) 2. Change the water, when hot water is applied, in the morning and an hour or two before cooking. 3. Add 1 teaspoonful of soda per quart of beans and boil until the beans become soft. 4. Bake like other beans. Note. — The characteristic strong flavor of the beans is removed by soaking before cooking; the addition of soda makes the beans soft. Cook- ing -with salt pork, potatoes, onions, molasses and other substances makes the beans more palatable to some tastes. Boiled Beans. Treat the beans as in the case of the baked beans, and boil them in a double boiler four to five hours until they become soft. Note. — The addition of any one of the articles recommended for use with the baked beans may make the beans more agreeable to some people. Roasted Beans. 1. Roasting can be done either in an oven or in an ordinary corn popper. 2. Roast until the skin of the bean is burst by popping. Note. — The beans can be kept soft by immersing them in a syrup while they are hot. Thus very wholesome candy is prepared. Powdered Beans. Roasted. 1. Roast as in the roasted beans. 2. Let them stand until cool to harden them. 3. Grind them in a coffee mill or any other suitable grinder. Note. — The powder can be used as salad dressing or cooked with cookies Uke peanuts and other nuts, or employed as a substitute for cofTee. Raw {Soy Bean Meal). Grind the raw beans to a fine powder. Note. — One part of bean meal mixed with 4 parts of wheat flour in bread makes a quite palatable bread, which is very nutritious; it is also used for biscuit, muffins, etc. Bread made of soy bean meal alone is recommended for diabetic patients, as it contains only very small amounts of starch, sugar and dextrin. ^ 1 a. L. Winton: Conn. State Exp. Sta. Rept., 30, 153-165, 1906. MASS. EXPERIMENT STATION BULLETIN 182. Green Beans. 1. Pick them when the beans are three-fourths to full grown. 2. Boil them in salt water. 3. Discard the pods. 4. Serve the beans with butter or milk. Note. ■ — The pods are tough and they can be removed easily on boihng. Soy Bean Pulp (Kara). 1. This is the residue after the milk is extracted in the process of prep- aration of soy bean milk. 2, Cooked like any other vegetable with proper seasoning. Note. — Makes a very rich dish; an addition of green onions, cabbage or parsnip may improve it. Fermented Boiled Beans (Natto). 1. Boil beans for five hours. 2. Wrap inside of a straw bundle. 3. Smoke them in a closed cellar by building a wood fire and closing the door. 4. Let them ferment in a warm, moist atmosphere at 40° C. for twenty- four hours. Note. — In making the bundle rice straw is preferred. This may not be suited to American palates on account of its peculiar flavor, which is due to the ripening protein. This recipe may also be undesirable on account of the difficulties involved in the process. Table IV. — Chemical Composition of Natto {Per Cent.).^ Nitrogen proteids, .......... 4.033 Nitrogen of amides, .......... 1.892 Nitrogen of peptone, 1.617 Total nitrogen 7.542 The relatively high percentage of total nitrogen may be due to the loss of carbon as carbon dioxide during the fermentation. Ripened Vegetable Cheese ^ (Miso). L Preparation of "mother miso," or koji.^ 2. Steam soy beans for twenty-four hours. 3. Rub into a thick, uniform paste. 1 K. Yabe: Bull. Vol. 2, 72, 1894-97, College of Agriculture, Tokyo, Japan. 2 Koji used for manufacturing miso ia similar to that used in making sak§, — Japanese rice wine. It consists of barley or rice with a culture of certain forms of fungi, chiefly Aspergillus oryzae. It contains diastatic, inverting and proteolytic ferments. SOY BEANS AS HUMAN FOOD. 9 4. Add proper amount ^ of koji, salt and water. 5. Mix well and store in a vat at 15° to 20° C. 6. Let it ferment for a certain period of time according to the variety of miso. Note. — Preparation of miso at home is not easily done because of the complexity of the technic, although it is very often practiced in Japan. Koji is sold in Japan on the market from special factories. It can be used very extensively for preparing soups, cooking vegetables, making sandwiches, etc. Different kinds of miso are produced through the use of different manipulations and components. Table V.— Composition of Red and White Miso {Per Cent.) } 1 1 . •5 1 Q i 11 3: 1 fo 1 ¥ 5 1 < S 6 h White miso, . 59.27 39.78 22.13 10.18 5.10 1.09 6.31 8.32 .95 5.99 7.70 Red miso, 50.16 48.66 32.28 12.48 6.46 2.31 2.72 10.40 1.18 10.84 12.40 Table V. indicates a high percentage of substance soluble in cold water. This fact makes it very convenient material to be used in soups. A trace of alcohol is present also. Soy Bean Sauce (Shotu). 1. One part each of beans, wheat and common salt and 2 parts of water are used. 2. Roast and pulverize wheat. 3. Steam and mash the beans as in case of miso. Cool to 40° C. 4. Add powdered wheat in the proportion of 70 parts of the caked beans to 30 parts of the wheat by weight. Mash and mix thoroughly. 5. Add spores of Aspergillus oryzoo, then mix. Spread upon wooden vessels or traj^s, about 3 liters per tray. The trays are stacked away in a cellar where the temperature is kept somewhat above 40° C. (After twenty to twenty-five hours, the mycelium of the fungus will be found; evolution of CO2 and heat is observed as the fermentation proceeds; after about six days the growth of the fungus is completed, and an abund- ance of yellowish spores, " perithecia," is present. The temperature is kept approximately at 27° to 28° C.) Dry the material and grind. This is the shoyu-koji. 6. Heat the required amount of water and salt to 115° to 118° C. Cool to room temperature. 1 The amount to be added varies according to the kind of 2 K. Oshima: Loo. Cit. p. 30. iso desired. 10 MASS. EXPERIMENT STATION BULLETIN 182. 7. Mix sho>ai-koji with the salt solution. 8. Allow the mixture to ferment in casks for one to two years \nth frequent stirring. 9. On the completion of fermentation, filter and press. 10. Allow filtrate to settle for two or three days. 11. Remove the clear supernatant liquid and heat it at 70° to 100° C. in a double boiler from two to three hours. 12. To improve the taste it is common to add a certain quantity of sugar or sweet sake during the heating process. Note. — This sauce is mainly manufactured in zymo factories in Japan, for its preparation at home is too difficult. It is a thick, dark brown liquid and used extensively in Japan and China. It may be used in American kitchens for soups, gra\des and vegetable stews, and makes a good substitute for Worcestershire sauce or any other table sauce. It has very slight food value, but its merit lies in its flavor, which seems to sharpen the appetite and accelerate the digestive functions. ^ Table VI. — Chemical Composition of Shoyu {Per Cent.).^ Number Specific Grav-ity. Water. Protein.3 Carbohydrate. Free Acid, mostly Lactic. Ash. Com- mon Salt. Phos- phoric Acid. Sample. Glucose. Dextrin. 1, 2, . . 3, . . 1.185 1.190 1.208 62.39 62.82 60.58 9.28 9.53 9.15 2.70 3.33 5.85 .69 1.43 1.18 1.33 .92 18.48 18.70 20.14 16.03 15.67 17.47 .53 .51 .46 Vegetable Butter, Ice Cream, Oil (Table Use) and Lard (Cooking). The manufacture of these articles from soy beans needs further in- vestigation. To say anything further concerning their economical and industrial importance at the present time would be premature. 1 Pawlow: The Work of the Digestive Glands, London, 1902. 2 K. Oshima: Bull. 159, 32, 1905, U. S. Dept. of Agr., Office of Exp. Sta. 3 Consists of soluble albumin, peptone and further cleavage products. Eisei Shiken Jho: Bull. Imp. Sanit. Lab., Tokyo, No. 8, 35, 1897. BULLETIN No. 183 MAY, 1918 MASSACHISETTS AGRICILTIRAL EXPERIMENT STATION Rose Canker and Its Control By P. J. ANDERSON This bulletin records results of investigations on a new and serious fungous disease of roses, and describes successful control methods Requests for bulletins should be addressed to the Agricultural Experiment Station, Amherst, Mass. Massachusetts Agricultural Experiment Station. OFFICERS AND STAFF. COMMITTEE. Charles H. Preston, Chairman, Wilfrid Wheeler, Edmund Mortimer, Arthur G. Pollard, Harold L. Frost, The President of the College, ex officio The Director of the Station, ex officio. Hathorne. Concord. Grafton. Lowell. Arlington. Administration. STATION STAFF. William P. Brooks, i Ph.D., Director. Fred W. Morse, M.Sc, Acting Director. Joseph B. Lindsey, Ph.D., Vice-Director. Fred C. Kenney, Treasurer. Charles R. Green, B.Agr., Librarian. Mrs. Lucia G. Church, Clerk. Miss F. Ethel Felton, A.B., Clerk. Agricultural Economics. Alexander E. Cance, Ph.D., In Charge of Department. Samuel H. DeVault, A.M., Assistant. Agriculture. William P. Brooks, ^ Ph.D., Agriculturist. Henry J. Franklin, Ph.D., In Charge of Cranberry Investi- gations. Edwin F. Gaskill, B.Sc, Assistant Agriculturist. Robert L. Coffin, Assistant. Botany. A. Vincent Osmun, M.Sc, Botanist. George H. Chapman, Ph.D., Research Physiologist. Paul J. Anderson, Ph.D., Associate Plant Pathologist. Orton L. Clark, B.Sc, Assistant Plant Physiologist. W. S. Krout, M.A., Field Pathologist. Mrs. S. W. Wheeler, B.Sc, Curator. Miss Ellen L. Welch, A.B., Clerk. Entomology. Henry T. Fernald, Ph.D., Entomologist. Burton N. Gates, Ph.D., Apiarist. Arthur I. Bourne, A.B., Assistant Entomologist. Stuart C. Vinal, M.Sc, Assistant Entomologist. Miss Bridie E. O'Donnell, Clerk. Horticulture. Frank A. Waugh, M.Sc, Horticulturist. Fred C. Sears, M.Sc, Pomologist. Jacob K. Shaw, Ph.D., Research Pomologist. Harold F. Tompson, B.Sc, Market Gardener. Miss Ethelyn Streeter, Clerk. Meteorology. John E. Ostrander, A.M., C.E., Meteorologist. On leave. Microbiology. Charles E. Marshall, Ph.D., In Charge of Department. Arao Itano, Ph.D., Assistant Professor of Microbiology. George B. Ray, B.Sc, Graduate Assistant. Miss Louise Hompe, A.B., Graduate Assistant. Harold L. Sullivan, B.Sc, Graduate Assistant. Plant and Animal Chemistry. Joseph B. Lindset, Ph.D., Chemist. Edward B. Holland, Ph.D., Associate Chemist in Charge {Research Division). Fred W. Morse, M.Sc, Research Chemist. Henri D. Haskins, B.Sc, Chemist in Charge (Fertilizer Division). Philip H. Smith, M.Sc, Chemist in Charge (Feed and Dairy Division) . Lewell S. Walker, B.Sc, Assistant Chemist. Carleton P. Jones, M.Sc, Assistant Chemist. Carlos L. Beals, M.Sc, Assistant Chemist. WiNDOM A. Allen, i B.Sc, Assistant Chemist. John B. Smith, i B.Sc, Assistant Chemist. Robert S. Scull, ' B.Sc, Assistant Chemist. Harold B. Pierce, B.Sc, Assistant Chemist. James T. Howard, Inspector. Harry L. Allen, Assistant in Laboratory. James R. Alcock, Assistant in Animal Nutrition. Miss Alice M. Howard, Clerk. Miss Rebecca L. Mellor, Clerk. Poultry Husbandry. John C. Graham, B.Sc, In Charge of Department. Hubert D. Goodale, Ph.D., Research Biologist. Miss Grace MacMullen, B.A., Clerk. Mrs. Nettie A. Gilmore, Clerk. Veterinary Science. James B. Paige, B.Sc, D.V.S., Veterinarian. G. Edward Gage, ^ Ph.D., Associate Professor of Animal Pathology. John B. Lentz, i V.M.D., Assistant. 1 On leave on account of military service. CONTENTS Introduction, Symptoms, .... Description of the causal fungus. Mycelium, Sclerotia, Conidiophores, Conidia, .... Life history of the fungus. Germination of the spores, Temperature relations, Effect of freezing the spores. Thermal death point of spores, . Effect of desiccation on the spores, Parasitic life of the fungus, Pathogenicity, .... Infection court. The mycelium in the host tissues. Normal structure of the stem, Path of the mycelium. Effect on the host cells. Saprophytic life of the fungus, . Longevity of mycelium in the soil. Growth on other substrata. Depth of penetration of the soil, . Rate of growth of the mycelium. Effect of freezing the mycelium, . Thermal death point of mycelium. Dissemination, ..... Original source of the pathogene. Spread from one grower to another, . Local dissemination. Occurrence of two species of Cylindrocladium on roses. Morphological characters. Cultural characters, Latin description of Cylindrocladium panum. Control, .... Exclusion of the pathogene, Eradication of the pathogene. Disinfection by chemicals. Laboratory tests. Formaldehyde, . Sulfuric acid. Copper sulfate, . Lime-sulfur, Dry sulfur. Soot, .... Greenhouse tests with formaldehyde. Disinfection by heat. Laboratory tests. Time required to disinfect soil by steaming. Green house tests of disinfection by heat. Disinfection of pots, tools, etc., . Protection of the host, . . . . Comparative value of different fungicidal Lime-sulfur, .... Dry sulfur flour, Ammoniacal copper carbonate. Lime, ..... Bordeaux mixture, . Treatment of the walks in the house, . Immunization of the host, Summary of control measures, . Literature cited, ..... Publication of this Document approved by the Supervisor of Administration. BULLETIN No. 183. DEPARTMENT OF BOTANY. ROSE CANKER AND ITS CONTROL. BY P. J. ANDERSON. INTRODUCTION. Rose canker is a serious disease of greenhouse roses which was first described in 1917. It has probably been long prevalent in America, but has escaped notice largely on account of its obscure symptoms and con- sequent difficulty of diagnosis. Its ravages were formerly assigned to other causes or left unexplained. Rose growers who first brought it to the attention of this station in November, 1916, stated that they had been suffering severe losses for at least four years. After conditions in the rose houses had been investigated, the situation was considered so serious that work was immediately begun to determine more of the nature of the disease, and especially to find a remedy for it. The investigation was started in co-operation with L. M. Massey, pathologist of the Ameri- can Rose Society, who first observed the disease two months before this, and had already decided that its seriousness warranted a thorough inves- tigation. Research at the Massachusetts station has been largely confined to determination of the best methods of controlling the disease and inves- tigation of such facts in the life history of the causal fungus as have a direct bearing on control measures. Massey undertook investigation of other phases of the disease, and has recently published his results (1917). A successful method of control has been evolved and is presented in this bulletin, but it is hoped that, as a result of long-term experiments now in progress in commercial houses, this method will be improved and, pos- sibly, other easier methods found. However, since this will require a number of years, the present method is published in order that rose growers who are troubled with the disease may have the benefit of all that we alread}'^ know about canker and its control. ' The writer is greatly indebted to Prof. A. Vincent Osmun, head of the department of botany at this station, for much valuable assistance, suggestions and criticism of the manuscript of this bulletin. 12 MASS. EXPERIMENT STATION BULLETIN 183. Only roses under glass are known to be affected. Some varieties, e.g., Hoosier Beauty, are more susceptible than others, but there is yet no evidence that any are immune. Massey (1917) observed the disease on Hoosier Beauty, Ophelia, Hadley, Russell, Sunburst, American Beauty and many seedlings. It has been reported only from the northern and eastern United States, but closer observation will probably show that it has a much wider range. SYMPTOMS. The disease is most easily recognized by brown dead areas (cankers) in the bark of the stems. These are more frequent and larger at the crown than higher up, but any part of the stem or branches may be at- tacked. Crown cankers may be below the surface, just at the surface, or, more often, extending up the stem, sometimes several inches (Plate I., Fig. 1). They may be confined to one side or may girdle the stem. The young canker is blue-black or purplish in color and smooth, but as it becomes older the part above ground becomes reddish brown, dry, hard and cracked longitudinally. The margin is definite, and the dead area becomes sunken. Frequently the part of the stem immediately above the canker is swollen (Plate II.). When the subterranean part of the canker becomes old it is soaked and "punky," and the bark may be rubbed off between the thumb and forefinger, or it may rot away entirely "(Plate I., Fig. 1). Sometimes a callus is formed around the edge of the canker. Two types of cankers occur on the stem and branches higher up. The larger ones start from wounds, especially the stubs which are left after the blossoms are cut (Plate I., Fig. 2). Cankers from these stubs run back down the stems. The canker may stop at the first live branch below, but very commonly it continues to progress downward, and each successive branch dies as it is encircled by the descending canker. Can- kers may also start from other wounds besides cut stubs. They are usually oval in outline and may be several inches long. The second type of aerial canker does not originate with wounds, but starts directly in the healthy green bark. First, small round purple areas appear, some- times singly but more often in groups. As these increase in size the cen- ters become light brown and the margins remain dark, giving a "bird's- eye" effect. When they occur in groups they coalesce and form large irregular dead areas in which, however, the individual cankers may still be distinguished for some time (Plate III., Fig. 2). The depth of the canker varies, depending on such factors as the age of the part attacked, size of the infection court, environmental condi- tions and probably others. This is particularly a disease of the bark, and commonly the discolored area will be located outside the cambium entirely. But in more severe cankers it may extend to, or entirely through, the pith. If the shoot is young and has not yet hardened, the canker goes deeper and the entire shoot dies. This is frequently evidenced in the Fig. 1. — Old canker running up from the crown. Fig. 2. — Canker running down from a cut stub. Canker on a lateral branch showing hypertrophy. PLATE III. Fig. 1 . — Canker resulting from coalescence of a number of small ones from stomatal infections. Fig. 2. — Five cankers on a single stem. ROSE CANKER AND ITS CONTROL. 13 sudden wilting and dying of shoots which have gi'own up rapidly from below the surface of the ground. Older shoots are rarely killed outright. Only occasionally have we seen entire plants killed by this disease. One, several or all of the shoots of a plant may be attacked. Dead " brush " and dead small shoots are usually much in evidence in affected houses. The seriousness of the disease, however, lies not in the number of plants killed but in the fact that affected plants are small and weaker, resulting in diminished yields of inferior roses. The diseased plants cannot be forced, no matter how much fertilizer is applied and how well they are cultivated. New shoots do not grow from beneath the surface of the soil, but all come from the tops. These latter symptoms are the ones which the florist usually notices first, and, in fact, may be the only ones he notices. Diagnosis of this disease is rendered difficult by two natural develop- ments in the life of the rose plant which may easily be confused with disease: (1) Many varieties of roses naturally turn black at the crown very early; this, however, is a superficial blackening, and rarely runs up much above the surface of the ground. (2) The bark of all rose stems ' cracks with age, especially at the base, just as the bark of trees does. These two developments often resemble canker so closely that even one experienced in diagnosis may be misled. DESCRIPTION OF THE CAUSAL FUNGUS. Rose canker is produced by the parasitic growth of a fungus, Cylin- drocladium scoparium Morg., within the tissues of the host (rose plant). Previous to 1917 this fungus had not been reported as a parasite. It was first found in Ohio by Morgan (1892) growing on an old pod of the honey locust {Gledilsia triacanthus L.). Seven years later it was reported again by EUis and Everhart (1900) as growing on dead leaves of the papaw tree {Asimina triloba Dunal), and described as a new species, Diplocladium cyUndrosporum E. and E.; but a study of the type materials of the two species by Massey showed them to be the same. As far as the literature shows, these are the only times that the organism had been observed up to 1916, and both times as a saprophyte. The body of the fungus is composed of (1) mycelium, (2) sclerotia, (3) sporophores (conidiophores), and (4) spores (conidia). These four parts, or organs, of the fungus are here described separately. Mycelium. The mycelium is the part of the parasite which lives inside the tissues of the rose stem. It is composed of many microscopically slender, branch- ing, tubular threads (hyphse) which grow in every direction through the host cells for the purpose of securing nourishment from them for the fungus. Incidentally, in this process, the cells are killed and turn brown, thus producing the canker. The hyphje are 4 to 6 « in diameter, and are divided by cross-walls (septa) into cells 5 to 20 times as long as their 14 MASS. EXPERIMENT STATION BULLETIN 183. Fig. 1. — Young mycelium from culture. -Old mycelium, showing chlamydospores. diameter. The manner of branching and septation is shown in Fig. 1. When the myceUum is young the walls are thin and not constricted, or, at most, only slightly constricted, at the septa. The contents consist of homogeneous protoplasm. Both the walls and contents are colorless, and when seen in mass, in pure cul- ture, look like white cotton. But when the mj'celium be- comes older it be- comes brown, the hypha^ are gnarled and twisted, deeply constricted at the septa, the cells short and oval or globose, giving one the im- pression of strings of beads (Fig. 2). The cells now con- tain large drops of reserve food, and the walls are thick. These cells are probably more resistant to adverse conditions, and serve to carrj^ the fungus through unfavorable periods. They may be called chlamydospores. Their diam- eter is much greater than that of the ordinary hyphre, as indicated by the figures. SCLEROTIA. Sometimes the surface of old cankers is dotted over with minute shining black pimples (Plate II.). They are usually not much larger than a pin point and never as large as a pin head. To the naked eye they look like pycnidia, but microscopic examination always proves them to be sterile balls of thick- walled pseudoparenchjanatous fungous cells (typical sclerotia). They are directly under the epidermis, but this does not obscure their shining black prominence. In certain culture media they are pro- duced in great abundance. The cells are much like the chlamydospores; in fact, the sclerotia seem to be only a further development of the chlamydospore-forming hypha", and all gradations between the two may be found. Their function is probably the same as that of the chlamydospores. A thin cross-section of one is shown in Fig. 3. Fig. 3. — Thin section through sclerotium. ROSE CANKER AND ITS CONTROL. 15 Fig. 4. — Tuft of conidiophor on a dead rose stem. CONIDIOPHORES. The conidia, or ordinary spores, — as distinguished from the chlam- ydospores, — are borne on special upright branches, — conidiophores. These are produced in great abundance in artificial culture, but are rarely seen on the cankers. The writer has found them occasionally just at the surface of the ground on young shoots recently killed by the pathogene. But in badly infested rose beds which are kept wet they are produced in great abundance on dead shoots and parts of the rose plants which are cut off and left to decay on the ground under the bushes. To the naked eye the dead shoots seem to be dusted over in patches with a white powder. Under a strong hand lens — or better, a binocular microscope — each particle of this white powder is seen to be composed of a tuft of slender-stalked "brooms" with glistening white heads. One of these tufts is shown in Fig. 4. Each little broom is a conidiophore with its mass of conidia on the apex. The number of conidiophores in a tuft varies from 5 to 40, or more. No de- tails, further than shown by Fig. 4, can be made out under the bin- oculars. Under the compound microscope, however, it is possible to de- termine accu- rately the struc- ture of these little brooms. Examined in the dry condi- tion they ap- pear as in Fig. 5, where the conidia are cemented to- gether into a solid head. But when mounted in water the cement which holds them together dissolves, many of them float away, and the head becomes loose as represented in Fig. 6. The main stem of the conidiophore may be unbranched up to just below the conidia, as repre- sented by Fig. 5, or it may show one or more monopodial branches at Fig. 5. — Conidiophores and conidia as seen in a dry condition. Fig. 6. — Conidiophores as seen when mounted in water, many of the conidia washed away. 16 MASS. EXPERIMENT STATION BULLETIN 183. various heights. The spores are frequently borne on lateral branches of this stem (Fig. 6), while the main stem is continued upward and terminates in an enlarged club. The ultimate branchlets, and one or two series below them, are usually in threes, as shown in Fig. 5, but twos are not uncommon. In regard to the dimensions of the co- nidiophore, Morgan (1892) -uTites: "the fertile hyphie have a simple septate stem 5 to 7 /^ in thickness, and are dissolved above into a level- topped cyme of branches; their height, exclusive of the spores which easily fall off, is 125 to 150 //." Ellis and Everhart (1900) give the di- mensions as 50-110 X 5-6 z^. In pure culture the writer, has found them taller than the above measurements; an average of 50 conidiophores grown on potato agar gave 291 fi, and the diameter of the stalk, 6.6. ;^. CONIDIA. The conidia are long, cylindrical, obtuse at each end, hyaline, divided into 2 cells by a septum at the center (Fig. 7). The A^ contents are at first homogeneous, but later show vacu- oles or oil drops (Fig. 8). Morgan (1892) gives the dimensions as 40-50 x 4 /^ at the apex, and 3 /^ at the base; Ellis and Everhart (1900), 40-50x4-5/^; Massey (1917), 36-55 x 3.3-4.51 fj-, with an average of 48.3 x 4.13 ^u. The writer found the aver- age of 50 on a young potato agar culture to be 48.8 x 5.1 //; 50 on a two-months' culture, 39.2 x 4.03 /*; 50 produced on a pod of Gleditsia, 41x4.1/^. LIFE HISTORY OF THE FUNGUS. Before am^ measure of control could be intelligently^ attempted it was first necessary to become intimately acquainted with the life history of the causal organism (the pathogene). In the studies which are recorded below most attention was directed to those points which appeared to have a direct connection with control. Nevertheless, in order to become familiar with the entire life cycle, certain phases of development which have no obvious connection had to be investigated. For convenience in discussion, the life history is treated under three heads: — 1. Germination of the spores. 2. Parasitic life of the fungus (pathogenesis). 3. Saprophytic life of the fungus. Fig. 7. — Germinating conidia. ROSE CANKER AND ITS CONTROL. 17 Germination of the Spores. The life cycle begins with germination of the spores. The first essential condition for germination is the presence of water. Spores never ger- minate except when they are directly in water. A moist atmosphere is not sufficient. Germination takes place through the production of one or more tubes from each of the two cells of the spore. Usually the tubes do not start at the same time; one in each cell begins to grow, and this is later followed by another. Four germ tubes to each spore is the most frequent condition, but there may be more or fewer. The tubes may come out from any place on the surface of the spores, as illustrated in Fig. 7. They elongate very rapidly at laboratory temperatures, quickly develop septa, branch repeatedly and soon a mycelium is produced. The brown thick- walled cells of the mycelium, which we have called chlamydospores, germinate by the production of slender hyaline germ tubes similar to those of the conidia and under the same conditions. Other detached cells of the mycelium also possess the power of germina- tion. Especially is it common to see germ tubes arising from the cells of the main stem of the conidiophore when detached and kept in water. Such germ tubes usually arise from the end walls of the cells, and may grow directly through one or more old cells before emerging. Temperature Relations. The relation of temperature to germination of spores was studied carefully in the hope of evolving some method of control bj'- keeping the rose houses at temperatures which are unfavorable for germination and thus retarding progress of the disease. The general effect of variation of temperature and the maximum, minimum and optimum temperature for germination were determined by the following method : — Method. — Viable spores from a young, pure culture were transferred to a drop of water in the center of a glass slide. The slide was supported on two short glass rods in a Petri dish, used as a moist chamber. A few drops of water placed in the bottom of the dish kept the air humid and prevented drying out of the drop containing the spores. The Petri dish was then kept at the desired constant tem- perature in incubator, refrigerator or constant temperature room. Observations were taken and percentages of germination counted at regular intervals. No figures are based on the results from a single slide. Each result tabulated represents the average of several sHdes. Tests at high or low temperatures were controlled by duplicates at ordinary room temperatures. The results of the tests are summarized in Table I. 18 MASS. EXPERIMENT STATION BULLETIN 183. Table ] . — Effect of Teniperature Variation on ^-pore Germination. Temperature (Deg , Centigrade rees). Period before starting to germinate (Hours). Percentage of Germina- tion in 24 Hours. 5, . - 0 8-9, . • 24 1 (2 per cent, in 48 hours). 12, . 5 95 15, . 17, . Not observed before 7 hours, when about 20 per cent, had started. 4-5 95 95 20, . 4i 95 22-23, 3-4 95 25-26, 2-31 95 28, . 30, . Not observed before 6J hours, when 95 per cent, had germinated. 6J 95 95 31, . 6i 70 33.5, 6i 21 (Erratic and abnormal). 36. . 4 70 37.5. - 0 40, . - 0 It is apparent from these tests that spores germinate at any tempera- ture between 8° and 36° C. Between 12° and 30° the percentage of germination was almost total, ranging from 95 to 100 per cent, (all marked 95 per cent, in the table). Within these Umits there was prac- tically no variation of percentage due to temperature. In other words, if the optimum temperature is to be determined by percentage germina- tion alone, it is very wide. Below 12° the percentage drops off rapidly until at 8° to 9° we get but 1 per cent, in twentj'-four hours. Germina- tion ceases altogether below this. Between the temperatures of 31° and 36° it is difficult to express the effects of temperature in percentages. Not only is germination erratic, varying greatly in slides apparently treated alike, but it may also be so abnormal that it is difficult to de- termine just what constitutes germination. The spores assume peculiar shapes by the development of knobs or, more commonly, globose swellings twice the diameter of the spores. These vary in number and location, but most frequently they are on the ends of the spores. Very slender unbranched germ tubes may grow for a time from these. The percentage of spores affected does not gradually diminish to form a regular curve. Thus, in one test at 36°, 70 per cent, were affected in this way. But at 37.5° there was no germination or change in the spores which could be detected with the microscope. The effect of temperature variation is more apparent in the time required for germination to begin than in ROSE CANKER AND ITS CONTROL. 19 the final percentage of germination. In this respect there is a rather regular curve. The optimum is at about 25°, where germination begins in two to three and one-half hours. At 12° it required five hours, and at 8° no germination was apparent until after twenty-four hours. The fact that spores do not germinate at a certain temperature does not mean that they are dead. Spores kept for two days at 5° showed not the least indi- cation of germination, but when brought back to ordinary room tem- peratures they quickly germinated to over 95 per cent. Experiments to be described later show that spores may be kept for long periods at temperatures both lower and higher than indicated in this table and still retain their viability. Apparently there is little opportunity for retarding the progress of the disease by maintaining temperatures in the house unfavorable to the fungus, because the optimum temperature for spore germination is ap- proximately the same as the optimum for growing roses. The latitude of the germination optimum is also unfavorable to such a method of control. Effect oj freezing the Spores. It is a well-known fact that the spores — especially the conidia — of many fungi are quickly killed by freezing, and this weakness may be utilized in checking disease. The purpose of the present investigation was to determine whether the spores of Cylindrocladium can be killed by freezing, and if so, how much exposure is required. Two methods were used. First Method. — Petri dishes containing young cultures with abundance of spores were exposed to out-of-door temperatures of — 3° to — 10°C. Checks were first made at room temperatures to test the viability of the spores. Spores were re- moved from the frozen plates at regular intervals and put to germinate in moist chambers at ordinary room temperatures, as described above in spore germination tests. Bj^ this method the spores were dry when frozen. After about two hours the percentage of germination began to decline; in eight hours it had fallen to 10 per cent.; in twelve hours, to less than 1 per cent.; and at the end of fourteen hours there was no germination whatever. All checks germinated 95 per cent. Second Method. — Spores were transferred from plates along with a portion of the agar to drops of water on slides. All was macerated until the spores were well distributed through the water. They were immediately put outside to freeze and one slide brought into the laboratory at the end of each hour and tested for ger- mination. The results were very similar to those obtained by the first method. Freezing for one hour seemed not to affect them at all; in two hours the percentage dropped to from 75 to 80 per cent.; in three hours, to 30 per cent.; in six and one half hours, to 25 per cent.; in ten hours, to 1 per cent. From 1 to 2 per cent, germinated even after exposures of twenty- 20 MASS. EXPERIMENT STATION BULLETIN 183. four hours, but these were spores in the center of the drop of water, or directl}^ in the agar, which seemed to give them some protection. There was no germination whatever after thirty-six hours. The first method more nearly approximates natural conditions, but under any conditions we may safely draw the conclusion from these experiments that all spores are killed by freezing during thirty-six hours. Thermal Death Point of Spores. Investigation of this point was undertaken with a view to the possi- bility of sterilization by heat. Thermal death point is defined as the lowest temperature at which an organism is killed by an exposure for ten minutes. Since this point might be different for spores than for mycelium, each was tried separately. Method. — Spores from a young culture immersed in a drop of water were placed in a thin pipette tube, sealed at one end and covered with a rubber cap at the other. The tubes were then dropped into vessels of water kept at the desired temperature. Each vessel was supplied with a thermometer, and could be heated by a Bunsen burner when necessary. After ten minutes the tubes were removed, the sealed end filed off, and the spores forced out through it on to a glass slide by pressing the rubber cap at the other end. The slides were then put in moist chambers as previously described in germination tests. These were kept at ordi- nary laboratory temperatures. Temperatures at intervals of 1°, from 40° to 55°, were tried. All tests were made in duplicate several times. Up to and including 46° the spores did not seem to be affected by ten- minute exposures. Above this the percentage remaining alive declined very rapidly to the absolute thermal death point of 49°. At this tem- perature none ever germinated. It was also found that spores can be killed at lower temperatures than 49° by exposing them for longer periods. In some previous experiments it had been determined that they are killed by an exposure to 37.5° for twenty-four hours. At 42° they are killed in two hours. To determine the effect of varying the period of exposure at a given temperature, 40° was selected as a standard, and spores exposed (in drops of water on slides in Petri dishes) during periods differing by intervals of one hour. They were then brought back to room temperature and tested as above. The results of this series are given in Table II. ROSE CANKER AND ITS CONTROL. 21 Table II. — Germination of Spores after Exposure to a Temperature of 40° C. Period of Exposure (Hours). Time required after Removal to Room Temperature before beginning to germinate. Percentage of Germination after 24 Hours at Room Temperature (20-24" C). 3, 4, 5. 5J, . 61 . 71 . 9, 12, 14, 18 20, 95 per cent, in 3§ hours. Not observed sooner. 2J hours. Just starting. 3 hours. Just starting. 63 per cent, after 5 hours. At least longer than 4 hours. 1 per cent, in 7 hours. At least longer than 6 hours. Over 95 Over 95 Over 95 Over 95 50 3 0.5 0 0 It will be noticed that the longer the period of exposure, the longer the time required for germination after being removed to room temperature. There was n« decrease in the percentage of germination until after four hours. From this point it dropped rapidly to less than 1 per cent, in six and one-half hours, and no germination whatever after seven and one- half hours. Effect of Desiccation on the Spores. The length of time during which spores are able to live in a dry condi- tion may have an important bearing on dissemination of a fungus and spread of a disease. Neither the thinness of the walls nor character of the spore contents of Cylindrocladium would lead one to expect great lon- gevity. The following method was used to determine longevity at ordinary room humidity : — Method. — The lids of Petri dishes, containing pure cultures of Cylindrocladium with abundance of conidia, were lifted enough to allow the thin film of agar to become hard and dry within a day or two. At intervals of one day .spores were transferred from these dishes to drops of water on slides in Petri dishes, as pre- viously described for other germination tests. The percentages of germination were determined after the spores were kept in moist chambers for twenty-four hours. All checks — made from the cultures before tilting the lids — germinated to over 95 per cent. Several hundred spores were transferred for each test. Three different Petri dish cultures were used at different times. In every trial the percentage of germination began to* decUne after twenty-four hours. In two days it had dropped to 25 per cent.; in five days, to 10 per cent. After ten days not more than 1 per cent, germinated, and in no case was any germination observed after drying for fifteen days. 22 MASS. EXPERIMENT STATION BULLETIN 183. The longevity of conidia, then, appears to be very Hmited when kept in a dry condition. When the atmosphere is kept very humid the}" live longer, at least several weeks, but no careful investigation has been undertaken to determine just how long with each degree of humidity. If water stands on them, even in the culture dish, they germinate and then quickly die if dried out at once. Parasitic Life of the Fungus. Pathogenicity. In order to prove that an organism is the causal factor of a certain disease there are four requirements — called the four rules of proof — which pathologists all agree must be fulfilled. These are: (1) find the organism constantly associated with the disease; (2) isolate the organism, grow and study it in pure cultures; (3) produce the disease again by inoculation from these pure cultures; (4) reisolate the organism and prove by culture its identity with the organism which was first found. These four rules were comphed with by Massey (1917), and the pathogenicity of Cylindrocladium scopariwm established. The present writer has also given the four rules repeated test, and obtained results similar to those of Massey. These experiments are not described in detail here, but only certain notes on each of the four steps recorded. 1. Constant association of --the pathogene with the canker is not so easy to establish as in most fungous diseases because the fungus can rarely be seen with the naked eye on cankers in rose houses. Nevertheless, the writer has occasionally been able to find a white band of conidia around cankers on yovmg shoots just at the surface of the grovmd. Almost always when a canker is kept in a moist chaml^er for twenty-four hours or longer the mycelium grows out as long, straight, white hypha?, which can readily be recognized as peculiar to Cyhndrocladium by one who has become acquainted with the appearance of this fungus. Also, after a few days in the moist chamber, conidia usually begin to develop on the surface. The presence of the pathogene in old cankers is also often betrayed by sclerotia, ■ — small, flat, shining black specks just under the epidermis. Yet the writer has often found cankers in which the organism could not be determined in any of the above ways. There seems to be only one absolutely sure way of determining association of the pathogene in all cases, and that is by making isolations, which is really a part of the sec- ond rule of proof. 2. The following has been found the most satisfactory method of isolation: — Method. — The surface of the canker is first sponged with mercuric chloride 1-1,000. Scalpels and steel needles are kept in a jar of 95 per cent, alcohol. The epidermis, or at least a thin outer layer of the canker, is then peeled off with a scalpel from which the alcohol has been burned over a Bunsen. Another scalpel sterilized in the same way is used to cut out a portion of the peeled canker. It is ROSE CANKER AND ITS CONTROL. 23 then removed with a flamed needle to a flask of sterile water, washed, and trans- ferred to a potato agar slant — or sometimes poured plates are used. One or two drops of lactic acid are added to the tube of agar when slanted. The acid not only prevents growth of bacteria, but also seems to make the medium more favorable for the growth of Cylindrocladium. Occasionally other agars, such as corn meal, oat, lima bean, Czapek's and Cook's No. 2, have been successfully used, and there is no objection to them. The almost constant use of potato agar in the present investigation is due more to habit and convenience than to any advantage over other media. In the ease of small initial cankers the epidermis was not peeled off. The mycelium grows up into the air and into the agar very quickly, and after some experience one is able with the naked eye to distinguish within twenty-four hours the growth of Cylindrocladium from that of other fungi he is apt to meet with on roses. But if there is any doubt, he has but to wait another day or two, and spores are produced by which this fungus can be absolutely identified. Other methods of isolation besides tissue transfers have been successfully used. Where spores are present, or where they have been developed in moist chambers, cultures are very easily made by touching them with the tip of a sterile platinum needle, — first thrusting the needle into the agar so that more spores will adhere, — and then transferring to agar slants. When the sclerotia were first discovered on the cankers there was some question as to their connection with Cylindrocladium. Some of them were picked out under the binoculars with a sterile needle, freed from all clinging rose tissue, washed in sterile water, and transferred to agar plates. In this way, also, pure cultures were obtained. By the first method described, the organism has been isolated in pure culture from hundreds of typical cankers. In order to determine the very youngest stages, a number of stems showing the little round lesions (de- scribed under "Symptoms"), from the size of a pin point to several millimeters in diameter, were brought into the laboratory, washed merely with sterile water, and transfers made as above. Pure cultures were obtained from even the smallest of them. The relation of the pathogene to dead stubs was also determined in this way. After the flower is cut, one or more shoots quickly grow out from below the cut end of the stem. The topmost one, however, is usually some distance below the cut surface, and a useless stub is left from 1 inch to 3 or 4 inches long. This stub usually dies slowly from the apex back to the first branch, where it is apt to stop. When the canker disease is prevalent in the house, however, the dying frequently does not stop at the first shoot but continues down the stem, and the shoots die as they are encircled by the descending dead area. Frequently the fruiting bodies of various species of fungi, such as Pestalozzia, Phoma, etc., can be found on these stubs, but in other cases no spores could be found. A large number of them were collected from a house known to be infested, and transfers made. Cylindrocladium was obtained from over half of them. After they were found to be infested in some cases, more attention was directed to them and the sclerotia frequently observed. It was from these sclerotia that the pure cultures mentioned above were obtained. Study of the fungus in pure culture will be described later. 3. Plants were inoculated in four different wavs: — 24 MASS. EXPERIMENT STATION BULLETIN 183. Methods of Inoculation. — (a) Stems wounded, inoculated with agar in which the fungus was growing, kept moist several days with moist cotton, {b) Same as (a), but the plants not wounded, (c) Wounded, spores sprayed over the plants with an atomizer, and kept for several days under a bell jar. (d) Same as (c), but plants not wounded. All these methods were controlled by checks treated in the same way except for applying the fungus. Typical cankers were produced by all four methods of inoculation. The shortest incubation period — time between inoculation and first appearance of symptoms — was four days on the wounded plants and five days on the unwounded ones. The rate of development of the canker after it first appears varies greatly. On some plants which were first wounded and kept under bell jars the cankers were over a centimeter across in two weeks, but if the bell jars were removed and the humidity of the air diminished, the cankers grew very slowly. Small aerial cankers usually soon stop growing altogether unless several of them occur close to- gether, or unless they are kept very moist. Crown cankers grow more rapidly than cankers higher up, but their rate of growth becomes de- cidedly slower as they advance above the surface of the soil. 4. Reisolations were very readily made from a number of the cankers produced by artificial inoculation. The fungus was obtained in pure culture, and easily identified by its cultural and morphological characters as Cylindrocladium scoparium. Injection Court. The artificial inoculations described above indicate that a wound is not necessary for infection. All observations indicate, however, that a wound is a very favorable infection court. A great many of the basal cankers start from the union of stock and scion; aerial cankers from the cut surfaces of stubs and from various bruises made by tools, etc. Even where no wound appeared, it seemed possible that there might be small wounds not readily visible to the naked eye. In order to determine whether such was the case, and if not, to determine whether any natural openings in the epidermis serve as infection courts, artificial inoculations were made by spraying spores with an atomizer on what, as far as could be seen with the naked eye, seemed to be perfectly healthy stems. As soon as canlcers began to appear they were cut out, fixed, imbedded in paraffin, cut into serial sections and stained. Twenty-four cankers varying from the size of a pin point to 2 millimeters in diameter were used and cut serially to a thickness of 8 /i. In no case was any wound through the epidermis discovered. But in every case a stomate was located directly at or very near the center of the canker. In the larger cankers there were several stomates, and it was not always possible to determine the point of entry. In the smaller ones, however, only one was present, and it was always approximately at the center. A number of infections were also discovered which were so small that they had not ROSE CANKER AND ITS CONTROL. 25 been seen when the material was fixed. In some cases the affected cells extended no farther than 5 or 6 rows below the stomate. There does not seem to be any reasonable doubt that the stomates serve as infection courts, and that the little round lesions on the smooth stems are largely the result of these stomatal infections. The Mycelium in the Host Tissues. In order to follow the course of the mycelium after it has entered the rose stem, and to determine its effect on the host tissues, cankers in every stage of development, from that where they are not yet visible to the naked eye up to the old, fully developed lesion, were sectioned, stained and studied. Method. ■ — The mycelium is very difficult to follow in unstained sections, but after some experimenting a simple method of treatment was found by which the mycelium could be very distinctly differentiated in the host cells. Cankers were fixed in Gilson's fluid, dehydrated gradually, and cut with a slide microtome from 95 per cent, alcohol, i The sections were then stained one minute in a saturated solution of safranin in 95 per cent, alcohol, excess safranin removed by trans- ferring to 95 per cent, alcohol for one minute, stained one minute in 1 per cent, gentian violet in clove oil, and cleared in clove oil, the oil washed out with xylol and the sections mounted in balsam. This method is very rapid and any number of sections can be stained at one time. Before describing the behavior of the mycelium in the tissues it will first be necessary to review briefly the structure of a normal rose stem. Fig. 9 represents a cross-section of a stem of about the age when cankers are most frequent. Normal Structure of the Stem. — On cutting through a rose stem with a knife, one very readily notices that it is composed of three distinct parts, (1) a rather succulent outer cylinder of bark, (2) a central soft white pith, and (3) a hard cylinder of wood between the two. The cell elements which occur in each of these will be enumerated in order, be- ginning with the outside. First, the stem is covered with a smooth, thin, waterproof coat, — the cuticle. Just beneath this is the one layer of rather flat cells composing the epidermis. Next in order are three or four layers of cells with heavy walls and no intercellular spaces. This is the collenchyma. The cuticle, epidermis and collenchj-ma form an air-tight, water-tight covering of the stem, uninterrupted except by the stomates. These microscopic breathing pores, which are not so numerous on the stem as in the leaves, are guarded and strengthened on either side by crescent-shaped pro- jecting cells. The structure of the stomate can best be understood by reference to the figure. It will be noticed that there is a free passage between the guard cells into the stomatal cavity beneath, and from here to the loose, thin-walled cells of the next underlying tissue, the chloren- ' Very small cankers were imbedded in paraflSn, sectioned and stained in the usual way; but for larger cankers this was found to be unnecessary, and a long and tedious process. 26 MASS. EXPERIMENT STATION BULLETIN 183. chyma. Except under the stomates, where it is thicker, the chlorenchyma is composed of three or four layers of cells containing around the inside of the walls the green chloroplasts which give the color to the bark. Next in order are the large thin-walled cells of the inner cortex, the lowermost healthy rose stem . of which contain abundant starch grains in storage. Next there are areas of angular, very thick-walled cells, the bast fibers. The walls are so thick that there is hardly any opening (lumen) through the center. In longisection these are seen to be shaped like long, sharp-pointed pencils, with the sharp ends overlapping. Their function is to give rigidity and ROSE CANKER AND ITS CONTROL. 27 strength. The areas of bast fibers do not form a complete cylinder, but the inner cortex tissue runs down between them. Just under each bast area there is a region of tissue called phloem. It contains long tubes (sieve tubes) through which the elaborated plant food passes down through the stem from the leaves. Each sieve tube is accompanied by a line of small slender cells (companion cells), which appear in transection as though they were cut out of the corners of the sieve tubes. The remain- ing cells of the phloem are box-like cells called phloem parenchyma. The phloem is bounded below by the cylinder of thin flat cells, the cambium, which marks the line of cleavage between the bark and wood. The wood, or xylem, is composed mostly of four kinds of cells: (1) Box-like parenchyma cells which compose the broad medullary rays as well as the narrow rays one cell in width. (2) Long tubes of large diam- eter (tracheae) through which the water mainly passes from the roots to the parts above. The walls are strengthened bj^ spiral or annular thickenings. (3) Vertically elongated cells (tracheids) of smaller diameter and thicker w^alls, also water carriers. These make up the greater portion of the wood. (4) Wood fibers, somewhat smaller in diameter, with thick walls and long tapering points. They cannot be distinguished from the tracheids in transection. Although the walls of all the xjdem elements are heavj^ they are all marked with pits so that liquids have only a thin membrane through which they must pass to go from one cell to the next. The pith (not shown in the figure) is composed of cells of only one kind, large or small, somewhat isodiametric (parenchyma). The walls are very thin. Path of the Mycelium. — The germ tube, when it attacks the host, is very slender and easily passes between the guard cells down into the stomatal cavity. It could then readily pass between the loose cells of the chlorenchyma and inner cortex, but it does not choose to progress this way. Only rarely has the mycelium been seen progressing for any considerable distance between the cells, but it immediately passes into the cells by means of holes which it is able to dissolve through the walls. From this time on the mycelium is entirely intracellular except for the short distances through which it sometimes passes from one cell to an- other. It branches profusely, but the host cells do not become filled with mycelium. Rarely are more than one or two strands seen in a single cell, except in very old cankers. It is very slender and delicate at first, but in age becomes brown and takes on the various cell forms previously described for the mycelium. It seems to prefer the starch storage cells of the inner ^"'- lO-- Young mycelium in , . , . ,. ■ . the cells of the inner cortex. cortex, and m cankers ot medium age is always found most abundantly in these cells (Fig. 10). However, the other cells are not immune. Mycelium may be found quite abundantly 28 MASS. EXPERIMENT STATION BULLETIN 183. in the collenchyma, the heavy walls of which seem to offer no resistance whatever to the progress of the invader. Occasionally it has been found even in the epidermal cells. The first bar to its inward progress is the area of bast fibers. It does not pass through these at once, but in very old cankers it has been observed even in the bast fibers. There is, how- ever, an easy path between the bast areas through the flaring outer ends of the medullary rays, which do not stop at the cambium but extend up between the phloem areas. From here the hyphce can easily pass lateralh^ into the phloem. Pass- ing down into the xylem elements the invader finds its progress made much easier by the jpr^^^a==f=^rjr-ji-jrJ^^^ prcseuce of pits in all the walls. It does not ^'^^~^~^^>tf^^ confine itself to the medullary rays, but passes Fig. 11. — Mycelium in the laterally into the other elements. The my- ceiis of the medullary rays. celium has been fouiid in every element of the X3dem, least of all, however, in the wood fibers. Often in old cankers the tracheae may be found almost clogged with mycelium, frequently in the form of chlamj^dospores. The method by which it passes through the walls is shown in Fig. 11. From the xylem it passes down into the pith, where it finds progress easj^ through the thin walls. Effect on the Host Cells. — All of the cankers do not extend to the pith. A great many of them, for some unexplained reason, never go deeper than the bark. The fact that the affected plants stop growing, and do not send up any more shoots from below the cankers, is probably due to destruction of the phloem, which prevents any food passing down to the lower stem or roots. The cells somewhat in advance of the invading hyphoe first become filled with a brown, finely granular substance which gradually becomes coarser and later mostly disappears, possibly being used by the parasite, and the cells are left almost empty. The starch, nuclei and chloroplasts also disappear. The walls are not affected except for the holes through which the hyphce pass. The whole effect on the host seems to be entire disorganization of the cell contents. There is no hyperplasia, hypertrophy or other abnormal cell change in the canker. To be sure, there is often a swelling just above the canker, which is pro- duced by an increase both in the size and number of cells of the inner cortex. This is, however, probably due to the amount of elaborated food which is stopped here because it cannot now continue downward on its normal course. As the canker becomes older, the cells of the bark col- lapse, being now empty. The cracks which then appear in the bark may be due to the contraction of the dying tissue, or to the expansion of the growing stem, or both. The cells of the xylem and pith do not collapse, but the affected tissues turn brown. ROSE CANKER AND ITS CONTROL. 29 Saprophytic Life of the Fungus. Early in this investigation it was discovered that the canker pathogens does not necessarily live all the time on the rose plant, but that it is also a natural inhabitant of the soil. This was first proved by isolating it under sterile conditions from soil 4 and 5 inches below the surface in the rose beds. Then it was found that when sterilized soil is inoculated the mycelium spreads rapidly through it and lives and grows normally there for a long time. Since these pure cultures in soil have been used rather extensively in this investigation, the method of making them is described here and omitted in all future references. Method. — Milk bottles of 1 quart capacity were used. Thirty-three cubic inches of rose soil, moistened until muddy, was put in each bottle. The mouth of the bottle was then plugged with cotton and the whole sterilized in an autoclave. After it was cool it was inoculated by transferring a small bit of agar containing mj-celium to the surface of the soil. Soil so treated becomes entirely infested in twelve to twenty-one days at ordinary room temperature. Longevity of Mycelium in the Soil. Before undertaking control measures it was very essential to know whether the fungus lives indefinitely in the soil, or whether it starves out and dies when the rose plant is not present to furnish nourishment. On March 27, 1917, eight milk bottles of soil were inoculated. At the end of every month clods of soil were transferred from these bottles to acidified agar plates. It has been found that when soil particles containing living mycelium are transferred to agar plates the mycelium begins to grow out on to the agar within twenty-four hours, and in a few days produces spores by which it can be definitely identified. The soil bottles were kept in a dry culture room. No water was added to them, but the soil is still somewhat moist at this writing. One year from the date of inocu- lation every plate isolation gave pure cultures of Cylindrocladium. There seems to be no doubt, then, that it will live for a year at least, and prob- ably indefinitely, in the soil without the rose plant being present. Growth on Other Substrata. The longevity of the mycelium may possibly be increased by passing a part of its existence on substrata other than the living rose plant and the soil. The abundant growth and production of spores on dead and decaying rose twigs on the soil has previously been referred to. Dead rose leaves were sterilized and inoculated with spores in moist chambers, and it was found that the mycelium grows luxuriantly and. produces some spores on them. Pods of the honey locust and leaves of the papaw tree — substrata on which the fungus was previously reported — were inoculated in the same way. The fungus grew normally on both, produc- ing spores in great abundance on the pods, and less abundantly on the leaves. The great variety of artificial media on which it can be made to 30 MASS. EXPERIMENT STATION BULLETIN 183. grow in the laboratory also indicates a wide range in feeding habits. Other kinds of decaying vegetable matter in the soil were not tried, but it would not be surprising if it were found capable of living on a great number of them. Depth of Penetration of the aSoU. In the soil isolation tests the fungus was not found below 5 inches, but this was not conclusive, since the method of isolation proved not to be entirely satisfactory, and only a few isolations were made. The soil in the milk bottles was never more than 4 inches deep, but the fungus grew as luxuriantly at that depth as at the surface of the ground. In order to test its ability to penetrate to greater depth, glazed drain tiles 2 feet long were closed at the bottom with an inch of cement, filled with soil, plugged with cotton at the top and sterilized. The soil was then inocu- lated on the surface. Holes had been drilled at regular intervals through the side of the tiles. These were corked, and after the whole was steri- lized the corks were made air-tight and water-tight by covering them with melted paraffin. In order to determine whether the fungus had penetrated to a certain depth a cork at that depth was removed, a portion of the soil next to it transferred to an agar plate, and the hole immedi- ately made tight again, all operations being carried out under aseptic conditions. Unfortunately the soil became dry too quickly, due to the large opening at the top, and it was found necessary to pour more water on to the top of the soil. At this writing the fungus is growing through- out the entire depth of soil in the tiles, and has been isolated from the lowest holes, almost 2 feet below the surface. Whether it was washed down by the water or grew down naturally is not certain, but at present the fungus is growing normally in every particle of soil 2 feet below the surface. If it could be washed down by the water in the tiles, there is no reason why it should not be washed down by water in the rose houses. Judging from these results, and what is known about the penetration of other soil fungi, there seems to be no reason for doubting that the myce- lium may exist several feet below the surface, depending to some extent on the character of the soil. Rate of Growth of the Mycelium. The rapidity with which mycelium grows through soil is dependent on the temperature. The optimum, maximum and minimum tempera- tures for growth were determined for the purpose of finding which tem- peratures in the greenhouse are favorable and which unfavorable to the spread of the fungus. Method. — When the milk bottles of infested soil are kept in a dark place the progress of the white mycelium downward can be readily observed through the sides of the bottles. A number of bottles were inoculated, and when the mycelium was well started downward the limit was marked accurately by blue pencil lines around the bottles. The bottles were then placed simultaneously in incubators, ROSE CANKER AND ITS CONTROL. 31 ice boxes and constant temperature rooms, wherever a constant temperature could be maintained for a week at a time. A new line was drawn at the end of every forty-eight hours. The results of this test are tabulated in Table III. An examination of this table shows that the optimum temperature for growth is 26 to 27° C, the minimum is just above 8.5°, and the maximum between 30° and 32°. At the optimum, the mycelium grows at a rate of approxi- mately three-fourths of a centimeter per day; in other words, it requires about forty days for the mycelium to grow through 1 foot of soil. The results offer little hope of maintaining in the greenhouse a temperature very unfavorable to the growth of the fungus. Table III. — Effect of Temperature Variation on Rate of Mijcelial Growth in Soil. Temperature, Centigrade (Degrees). Number of Measurements. Daily Growth in Centimeters. 5, 10 10 20 150 170 170 130 90 30 25 40 30 10 0 8.5 0 14 16 .37 21-22 .50 23-25 61 25 .63 25-26, .68 25.5-26.5 26-27, 74 30 25 32-36 0 37.5 0 Effect of freezing the Mycelium. It is very important to know whether soil can safely be used in the benches after being frozen out of doors. The following tests were made to determine this point : — Method. — Eight bottles, each containing 33 cubic inches of soil, were plugged, sterilized and inoculated with Cylindrocladium. After seven months the soil was thoroughly infested with the fungus, and probably contained all modifications of the mycelium which ever occur in the soil. Transfers were made and the fungus in all found to be alive. Then, before the ground froze in November, four of the bottles were exposed outside, one on top of the ground, one just under the surface, one 6 inches down, and one a foot below the surface. The other four were kept in the laboratory for controls. Some of these bottles were brought in each month of the winter to see whether the fungus was still alive. 32 MASS. EXPERIMENT STATION BULLETIN 183. The last test was made May 10, after the bottles had experienced the coldest winter on record in Massachusetts. The fungus was still living in the soil. Apparently', then, soil cannot be made safe by exposing it during the winter out of doors. Thermal Death Point of Mycelimn. Anticipating soil sterilization by heat, the thermal death point for the mycelium was determined. Method. — The same method was used as for determination of the thermal death point of spores, except that bits of agar containing mycehum were inserted into the sealed tubes, and after exposure for ten minutes to the desired temperature were transferred to sterile agar plates. If the mycelium was still alive it quickly began to spread to the agar. Temperatures between 42° and 55° C at intervals of 1° were tested. Up to and including 48° the treatment seemed to have no effect on the mycelium. At 49° it was sometimes killed and sometimes not. It never grew after ten minutes' exposure to 50°. We may therefore consider 50° the thermal death point. It will be noticed that the thermal death points of nwcelium and spores differ bj^ only 1 degree. The mycelium tested contained, besides the ordinary white mycelium, also the dark bodies with thick walls which we have called chlamydospores and sclerotia. As was the case with spores, so also the mycelium may be killed by a longer exposure to a lower temperature. Based on an exposure during one hour, the thermal death point was found to be 48°. DISSEMINATION. In deciding on a method of controlling a disease it is of prime impor- tance to find out how the pathogene is spread about, where it comes from, how it reaches the host. In the present case a threefold question is involved: (1) How did the fungus get into rose houses in the first place? (2) How is it spread from the houses of one rose grower to those of another? (3) On the premises of a single grower, how does it pass from house to house, bench to bench, or plant to plant? In the light of what has been learned concerning the life history and habits of the pathogene, we may undertake to answer these three questions. 1. Original Source of the Pathogene. The fungus, from all that is known of its past history, is a native of America. Since it has been reported but a few times, it probably is not very common out of doors. As greenhouse roses are grown in the section of the country where it has been reported, it would not be far-fetched to imagine the fungus being carried into rose houses with rotted leaves, where it was able to adapt itself to parasitic life on the rose. It is not necessary to assume, tlien, that this is an imported pathogene. Early ROSE CANKER AND ITS CONTROL. 33 in the course of the investigation it was suspected that it might have been brought over from Europe on Manetti stocks, which are used ahiiost exclusively by rose growers for grafting. The IManetti is moderately susceptible to the disease, as maj^ be readily determined by examination of IManetti shoots coming from below the graft in a badly diseased house. Pure cultures have frequently been made from these shoots. Massey (1917) also made infection experiments and found Manetti roses suscepti- ble. In the course of these investigations hundreds of Manetti stocks from Scotland were examined for lesions, numerous tissue plants were made, hundreds more were kept in moist chambers to bring out the fungus, and thousands of them watched carefully for a year after being planted in sterilized soil in order to see whether the disease developed. All results were negative, and up to the present we have no reason to suspect that the fungus is being imported on Manetti stock. It would be very helpful if we knew how widely the fungus is distributed over this country in its natural state, and whether it is being carried into the houses again and again. Various investigators have worked on the fungous flora of the soil and published lists of species isolated, but none of them mentions Cylindrocladium. This may indicate that it is only local in its distribution, or may be due merely to difficulties of isolating it. There seems to be little doubt that it infests the soil about rose houses where the disease occurs and where infested soil has been dumped out. 2. Spread from Onk Growt:r to Another. Plants are continually being sent from one grower to another. Small cankers on these would be overlooked even if the sender was famiHar with the disease. Not only could the mycelium be sent in the plant itself, but particles of soil adhering to the plants could easily carry it. It has been proved by laboratory tests that infested particles of soil may be kept dry for at least three months, and probably longer, without killing the mycelium. The disease may be spread in other ways, but this one w^ould be sufficient to account for the present known distribution. 3. Local Dissemination. There are a number of ways in which the fungus spreads from one part of a house to another, or from one plant to another, (a) It may grow for long distances through the soil and enter the plant below the surface of the soil. That infection can take place in this way has been repeatedly proved by setting clean plants in infested soil and thus pro- ducing the disease on them, (b) If the fungus is ia the potting soil it would be effectually distributed in the beds when the plants were transplanted to them, (c) Where "own-root" plants are grown the soil in the cutting bench may be infested, and the disease is then carried with the cuttings when they are planted in the benches, (d) It is easily carried from one part of the house to another on tools, clothes and shoes of workmen. 34 MASS. EXPERIMENT STATION BULLETIN l'83. (e) Insects, centipedes and worms carry the spores, as has been proved in the laboratory by permitting them to pass over sterile plates after being on dead twigs bearing spores. (/) The water used in watering the plants is usually driven from the nozzle with enough force to splash spores and bits of mycelium from the soil or debris on the ground up to the stems. Probably most of the stomatal infections above ground are started in this way. The spores of manj' fungi are so light that they float around in the air and are wafted about by very light air currents. It does not seem likely that the spores of Cylindrocladium are carried about to any great extent in this way. They are bound together in solid heads of spores, which are probably too heavy for currents of air such as usually occur in rose houses. That they can be dislodged and blown some distance by strong air currents was proved in the laboratory by passing a strong current of air from a fan over spores growing on a dead rose stem, and exposing agar plates 1, 2 and 3 feet away. Colonies of the fungus devel- oped on all of them, but it is hardly probable that so strong an air current would normally occur in rose houses. They could also be blown about on dust particles, but the soil in rose houses is rarely permitted to become dry enough to form dust. OCCURRENCE OF TWO SPECIES OF CYLINDROCLADIUM ON ROSES. During these investigations a second species of Cylindrocladium has frequently been isolated. It was first taken from the roots of a plant which had typical cankers on the crown. Later it was secured a number of times from crowns and from dead areas of the plant above the ground. It was commonly isolated directly from the soil in the rose beds, from the surface to 8 inches down. Except for its size, it resembles C. scoparium so closely that the writer was at first inclined to consider it but a dwarf variety of that species. The spores are only about one-third as large as those of C. scoparium. Although numerous isolations have been made, no transition forms between the two have been found. The small form has been grown through many generations in culture, and has remained constant on all media. Infection experiments were carried out, but all attempts to produce the disease by the same inoculation methods as were used for the larger form gave only negative results. The fungus grows and produces spores on the dead tissue about wounds and on cut stubs, but seems to lack ability to spread to healthy tissue. The small form then appears to be a saprophyte, while the larger one is a parasite. In order to determine whether there are cultural differences by which they could easily be distinguished, the two forms were grown simul- taneously on five standard culture media. They show very marked diagnostic differences. Such differences in morphology, pathogenicity ROSE CANKER AND ITS CONTROL. 35 and cultural characters are certainly marked enough to be considered specific rather than varietal. Since no species of Cylindrocladium other than C. scoparium has been described, a new name, Cylindrocladium jyarvuni, is proposed for this small form. The morphological differences and the cultural characters and differ- ences of the two species are given in parallel columns below. Morphological Characters. Since some morphological characters vary somewhat with the condi- tions under which they are grown, all measurements given below were taken from potato agar plates grown simultaneouslj^ under the same conditions, and each is the average of fifty measurements. C. scoparium. Size of spores, 48.8 x 5.1 /i. Height of conidiophore, 291 fj.. Diameter of conidiophore stalk, 6.6 /v. C. parvum. Size of spores, 16.8 x 2.5 fi. Height of conidiophore, 130 //. Diameter of conidiophore stalk, 4.25 fi. Cultural Characters. Most soil fungi can easily be grown on a great variety of artificial media. The characters of the colony differ markedly with the medium used, and very frequently species of fungi, like bacteria, can be distin- guished more easily by macroscopic cultural characters than by micro- scopic morphological characters. Obviously, to grow each fungus on all the possible media, or even a great number of them, would be almost an endless task. Five common media, all easy of preparation, have there- fore been adopted by the writer as standard for all diagnostic work. These five are (1) potato agar (ace. Thom. Bui. 82 U. S. D. A., Bureau of An. Industry); (2) sugar potato agar (the same as the potato agar except for addition of 3 per cent, of cane sugar); (3) gelatin (150 grams gold label to a liter of water); (4) sugar gelatin (same as above with addition of 3 per cent, of cane sugar) ; (5) Czapek's synthetic agar (ace. Waksman in Soil Sc. 2: 113). Petri dishes, each with a single colony started at the center, were used. They were kept in the diffused light of the laboratory at the ordinary laboratory temperature. Every reference to a color in the description below refers to the color given under that name in Ridgway's "Color Standards and Nomencla- ture," 1912. Color "in reverse" in these descriptions refers to the color of the colony when examined from the bottom of the dish. This color may be due to (1) a pigment in the medium itself (extra-cellular), (2) intracellular pigments (i.e., the natural color of the mycelium), or (3) very frequently it is due to a combination of the two. Sometimes a dis- tinction is made between them, but for diagnostic work such a distinction usually adds difficulty instead of simplifying determination. Most emphasis is placed on those characters which appear within the first 36 MASS. EXPERIMENT STATION BULLETIN 183. week after the colony is made. If one has to wait two or three weeks or longer for a character to appear, the long waiting makes diagnosis tedious, and one of the principal purposes of this method of diagnosis is defeated. The more important characters for distinguishing these two species are italicized. Many minor distinguishing characters are not mentioned. Potato Agar. C. scoparium. Growth only moderately good. Starts with abundant, perfectly white, raised, aerial mycelium, but soon falls flat at the center, which becomes cov- ered with spores after two or three days. Always more or less aerial mycelium out toward the margin, which is rather coarse and tow-like. Not a decided color in reverse during the first week, but a dilute cream color to buff. At the end of the second week it turns to avellaneous or wood brown, and after three weeks still darker. Rood's brown. Margin of colony crenulate or wavy. C. parvum. Only moderately good growth. My- celium finer and denser than C sco- parium, perfectly white. Spores pro- duced in great abundance. The edge entirely throughout its growth remains very even and forms a perfectly round colony. Practically no color — possibly a very faint buff — develops in reverse even after three weeks' growth. Sugar Potato Agar. C. scoparium. Very rank growth, abundance of spores, entire plate covered in two weeks. Dense opaque color appears in reverse after three days; vinaceous purple to hwmatite red at the edge, dark- ening to russet or chocolate at the center. At the end of a week a large central area appears almost black, but examined more closely shows various shades of reddish brown, chestnut and bay. Entire reverse opaque after two weeks. The brown color is due to the extremely abundant production of sclerotia and chlamy- dospores on this agar. C. parvum. Rank, white growth of a very much finer texture than C scoparium,. Abun- dant production of spores. Color in reverse, white, or at most, only cream color at end of one week. This is one of the best diagnostic characters. At the end of two weeks it has passed through gray and drab gray to a clear wood brown, with minute patches of army brown here and there which show chlamydospores under microscope. The red-brown colors of C. scoparium never appear. Gelatin. C. scoparium. Growth very poor, consisting of a thin covering of coarse radiating hyphae. Very few spores. Stops growing after about ten days. Gelatin turned to a watery liquid which at the end of a week is orange rufous, but gradually turns darker to Sanford's brown. Lique- faction extends some distance beyond the margin of the colony. C. parvum. Growth very scanty, so much so that it is necessary to look at the plate against a black background to see it at all during first week. Gelatin lique- fied. No color at first, but becomes dilute old gold by end of second week. This medium is hardly suitable for dis- tinguishing the two. ROSE CANKER AND ITS CONTROL. 37 Sugar Gelatin. C. scoparium. Rank growth of coarse radiating aerial mycelium, but few spores. Gela- tin liquefied. After about four days a striking brilliant carmine color begins to appear in reverse, due to a pigment in the gelatin. This gradually spreads to the whole plate and becomes darker, an ox-blood red. This is probably the best diagnostic cultural character for this species. The mycelium covers the plate in ten days. C. parvum. Fine tangled aerial mycelium and more abundant spore production than for C. scoparium. Gelatin liquefied. Covers entire plate in two weeks. At the end of a week the colonies vary from Mars yellow to raw sienna in reverse, and at the end of two weeks have darkened to amber brown and Mars yellow. The color during the entire development of the colony is in strong contrast to the carmine and ox-blood of C. scoparium. Czapek's Agar. C. scoparium. Growth moderately good, aerial mycelium thin. Spores abundant. At the end of a week the colors in reverse are 7nuch the same as for potato agar, — claret brown, russet or amber, with a brick-red color suffused through it. At the end of two weeks the center is prac- tically black, fading through brown and red tints toward the margin. The red color is due to a pigment in the me- dium; the brown, to the chlamydo- spores and sclerotia. Irregular edge. C. parvum.. Finer and denser aerial growth of mycelium. During the first week the reverse remains pearly white: later it changes to dilute wood brown, then Rood's brown and at the end of two weeks ap- proaches Natal brown. None of the red tints of C. scoparium ever appear. Margin much more even than that of C. scoparium. Abundant production of spores in distinct concentric zones.- Latin Description of Cylindrocladium par\tjm. Cylindrocladium parvum n. sp. Album effusum; conidiophoris erectis, base sunjdicibws, apice ternate vel dichotomice ramosis, 130 x 4-2ofi; conidiis cijlindraciis, medio obscure 1-septatis, hyalinis, 16.8 x 2.5fJ: Hab. in caulibus emortuis et radicibus rosarum et in humo, Massachu- setts in Amer. bor. — Simile C. scopario. CONTROL. Ever}' method used in the control of any fungous disease is an appU- cation of one of four principles: (1) exclusion of the fungus, (2) eradica- tion of the fungus, (3) protection of the host, or (4) immunization of the host. Although practically all the work of the present investigation has been on the second of these principles, there are possibilities of using all four of them in the control of rose canker. These four are first con- sidered separatel}' below in the order named, and finally a general scheme of treatment is recommended. 38 MASS. EXPERIMENT STATION BULLETIN 183. Exclusion of the Pathogexe. By exclusion we mean preventing a fungus from entering a given territory in the first place, whether this territory be a country, a State, a region or only one rose house. Since this disease seems to be pretty generally distributed over the country already it is obviously impossible to exclude it from the United States, and probably from any particular State or section. But it is entirely possible to exclude it from the house of a rose grower who finds that none of his plants are already affected, or where new houses are being erected at some distance from old ones. The whole practice, then, consists of taking every possible precaution against carrying any diseased stocks, cuttings or infested soil into the house. Every plant brought in should be carefully examined, and, if there are any suspicious cankers in the bark, it should be discarded. All new plants and cuttings should be taken whenever possible only from houses known to be free from the disease. Eradication of the Pathogene. By eradication we mean the absolute destruction or removal of the fungus from the rose beds or from the whole house, so that it is no longer present in the plants or in the soil, pots, debris, manure or anywhere else from which it can return to the plants. The practice of this method is .of course necessary only when it has been impossible to exclude the pathogene and it has become established in the house. Up to the present this has proved to be the most successful principle applied to controlling canker. The ultimate aim is to eradicate the fungus from the plant itself, but the application of direct methods, such as excision of cankers, pruning off of dead parts, or even absolute destruction of entire plants when cankers are found on them, is altogether useless because the soil all about the plants is infested. From the soil the fungus can grow back into the roses as fast as it can be cut out. Spraying or dusting is of course useless, also, because no fungicide can reach the mycelium in the inner tissues of the plant; and also it is not possible to cover the parts of the plant below the surface of the ground where infection commonly occurs. Obviously, then, eradication resolves itself into destruction of the path- ogene in the soil; in otlier words, soil disinfection. Of the various methods of disinfecting soil only two have appeared to be at all practicable: (1) by heat, and (2) application of chemicals. Freezing, as previously men- tioned, is not effective. Desiccation would take entirely too long. Other methods are either too expensive or too difficult of application. In the course of the present investigation both heat and chemicals have been successfully used. ROSE CANKER AND ITS CONTROL. 39 Disinfection by Chemicals. Laboratory Tests. Some of the chemicals which have been used in the past for disin- fecting soil for the control of other fungous diseases are formaldehyde, sulfuric acid, copper sulfate, sulfur, lime-sulfur. The results obtained by the use of these same chemicals for other fungi could not be used directly in the present investigation because every fungus differs in its resistance to a given chemical. It was first necessary to determine what concentration and what quantity of solution per cubic foot was needed to kill the fungus. These facts could be determined more accurately and conveniently in the laboratory than in the greenhouse. The method used in all these tests was as follows: — Method. — Milk bottles, each containing 33 cubic inches of soil, were steam sterilized and inoculated from pure cultures of the fungus. When the soil was entirely infested (requiring from twelve days to three weeks) it was stirred into a loose condition with a sterile glass rod, and the proper amount of chemical in solution, at the strength to be tested, poured in under aseptic conditions. Since the soil did not dry out as rapidly in these bottles as it would under natural con- ditions in the greenhouse, it was emptied into sterilized porous flowerpots after a few hours. It was found after several trials that the pots dried out too rapidly if left in the open laboratory. Thereafter they were covered with bell jars which were tilted enough to allow free circulation of air beneath them, and the length of the drying process could then be regulated. After eight to ten days in the pots, clods of the soil were transferred from various portions of the pots to sterile agar plates. If the fungus was still alive it spread to the agar; otherwise there was no growth whatever from the clods. At first, the solutions were applied at the rate of 1 gallon to the cubic foot of earth. Afterwards, 2 gallons per cubic foot were used. When dry chemicals, such as sulfur, were tested the required amount was thor- oughly stirred into the infested soil of the bottles with a sterile rod and no water added. Formaldehyde. — First tests were at the rate of 1 gallon per cubic foot at the following concentrations: 1-500 (1 part of commercial formalde- hyde to 500 parts of water), 1-400, 1-300, 1-200 and 1-100. None of these concentrations gave complete success. On the transfers from the last two, however, only a few of the clods contained living mycelium. This indicated a lack of complete penetration by the solution. In the next series of tests the same concentrations at the rate of 2 gallons per cubic foot were used. The 1-100 and 1-200 then gave absolute control, while the 1-300 usually did; but occasionally a single clod developed a mycelium on the agar. The death point concentration lies somewhere between 1-200 and 1-300. But to be well within the margin of safety, 1-200 (1 pint of commercial formaldehyde solution to 25 gallons of water) was decided upon as the best strength to use in the greenhouse. Sxdjuric Acid. — This chemical has been successfully used in the past in the control, particularly, of certain root diseases of nursery trees. At the rate of 2 gallons per cubic foot, concentrations of 1, 2, 3, 4, 5 and 8 per cent, were used. The 5 per cent, solution killed most of the myceUum, 40 MASS. EXPERIMENT STATION BULLETIN 183. but not all of it. The 8 per cent, killed all of it. The death point con- centration lies between 5 and 8 per cent., but such a high concentration is hardly practicable in the rose house, and the exact point was not de- termined. Copper Sulfate. — Concentrations of 1, 2, 3, 4, 5 and 10 per cent, were used at the rate of 2 gallons per cubic foot. The 5 per cent, seemed hardly to check the fungus, but 10 per cent, proved entirely effective. Such a high concentration seemed prohibitive for application to soil, and no more accurate determination was made. Lime-sulfur. — This mixture proved to be worthless, even when applied at a concentration of 1 part of commercial product (32° Baume) to 10 gallons of water, and at the rate of 2 gallons per cubic foot. Dry Sulfur. — Finely ground sulfur flour was added to the soil and thoroughly stirred in. First, 10 grams per bottle were used, and when that proved to be ineffective 10 grams more were added, etc. All results were negative, even up to the rate of 7 pounds of sulfur to a cubic foot of soil. This test was performed at a laboratory temperature of 19° to 24° C. Perhaps if higher temperatures had been used the sulfur would have b6en more effective. Dry sulfur seems to be worthless at the tem- peratures tested. Soot. — There is an idea prevalent among florists that soot has fungi- cidal value, but plant pathologists seem never to have made any extensive experiments with it. The same method and rates as for dry sulfur were tried. At the rate of 4 pounds per cubic foot soot did not kill the fungus, but at the rate of 7 pounds no growth of the pathogene occurred. Of all the chemicals tried, formaldehyde seemed to be the only one which would give control at concentrations which could safely be used on the soil. Greenhouse Tests with Formaldehyde. The greenhouse tests on the use of formaldehyde were begun before the laboratory tests were completed, and at a time when it appeared that a concentration weaker than 1 pint to 25 gallons would be sufficient. As a result, the tests on a large scale were made wdth a concentration of about 1 pint to 40 gallons, but, on the other hand, more solution was applied per unit of soil. Two houses, each capable of growing more than 1,000 rose plants, were thoroughly soaked with the solution. One of the houses contained raised benches; the other, ground beds. Both had previously grown diseased roses. The soil was replaced by soil from outside the houses before sterilization. In the light of what we now know of the habits of Cylindrocladium, it is safe to assume that this soil was infested, because soil from the benches in previous years had been thrown out near it. After soaking the soil thoroughly the houses were closed. Fumes of formaldehyde were so strong in the closed houses that it was not possible to remain in them. After the soil had dried sufficiently both houses were planted with roses which had been potted in soil sterilized ROSE CANKER AND ITS CONTROL. 41 with steam, and which had been kept under conditions as sterile as pos- sible. Three months after planting, no disease had appeared in either house. Soon afterward it began to appear in the house with the ground beds, and gradually increased until, almost a year after planting, it was generally prevalent throughout the house. In the bench house, however, no disease has as yet been found, although plant-to-plant inspections have been made frequently throughout the year. The fact that a con- centration of formaldehyde weaker than 1 pint to 25 gallons controlled the disease in the bench house is probably due to the longer action of the more concentrated fumes, and probably, also, partly to the greater amount of the solution applied. The lack of control in the ground bed house can be easily explained in the light of our studies on the depth of penetration of the mycelium in the soil. The surface soil was disin- fected, but it was not possible to disinfect it down as far as the mycelium grows. After the formaldehyde had evaporated the deep mycelium began to grow upward, and during that period the plants remained healthy; but, after the mycelium had grown up to the surface again, the cankers began to appear and the roses became as badly affected as before the house was treated. Two conclusions may be drawn from this experiment: (1) the soil can be disinfected effectively by the use of formaldehyde, and (2) ground beds cannot be sterilized by this method. Disinfection by Heat. Laboratory Tests. The feasibility of destroying any fungus by application of heat to the soil manifestly depends, first of all, on the thermal death point of all stages of that fungus. As has previously been described, this point for Cylindrocladium was found to be 50° C. This comparatively low death point indicated that the soil could be readily disinfected by steaming, because a temperature much higher than 50° C. can be easily obtained by the use of steam. Time required to disinfect Soil by steaming. — This was further confirmed by the following tests : • — ■ Method. — Sterile Petri dishes were filled with soil which was thoroughly in- fested with mycelium. After removing the lids they were subjected to steam at a temperature of 90° to 95° in an Arnold sterilizer for the desired length of time. The lids were then replaced and the soil allowed to cool, when clods of it were • transferred to agar plates as described above. Exposures of five, ten, fifteen, twenty and thirty minutes were tried. No mycelium appeared on any of the transfers, even after five min- utes' exposure. Shorter periods of exposure were not tried because of the uncertainty of securing penetration by steam in less than five minutes. But, to determine what effect shorter exposures would have on mycelium, tests were made bj^ the sealed tube method described for thermal death point tests. In these tests the mycelium was killed in less than one minute when exposed to a temperature of 95° C. 42 MASS. EXPERIMENT STATION BULLETIN 183. From these tests we may conclude that soil can be disinfected by steam in less than a minute if penetration is obtained. Apparently effective- ness is limited only by the time required for the steam to penetrate every particle of the soil. Greenhouse Tests of Disinfection by Heat. Heat may be applied to the soil by steam or by hot water. The first method has been in vise in the greenhouses for the disinfection of the soil used in potting since the beginning of this investigation. Perforated steam pipes were laid a foot apart in a large pit. Soil a foot deep or more was piled over them and the steam turned into the pipes. Burlap or other coverings may be used to cover the soil and make it retain more of the steam. Soil thermometers were used to determine the temperature. It is only necessary to keep the temperature above 50° C. for ten min- utes. A higher temperature, of course, makes for additional safety. The one or two hours of heating frequently recommended for other diseases is only wasted time and expense, being entirely unnecessary for this fungus. Thousands of plants have been potted in soil disinfected in this way during the last year, and canker has never appeared on any of them. No doubt other methods of steam disinfection, such as the inverted pan method, would be equally effective. Either method could probably be used just as effectively on the benches, but the formalde- hj'de treatment is efficient, and quicker and easier of application. If there is any reason to suspect the presence of the fungus in the manure which is used to mulch the beds it may be disinfected in the same way as the potting soil. Soil for the cutting bench may also be treated in the same wa3^ The second method of applying heat — by the use of boiling water — is now being tested. It should be just as effective as steam, and at the same time much more rapid. The boiling water is forced through the water pipes ordinarily used in the house, and is applied to the soil through a hose with a long nozzle and a handle which will not become heated. The water should be applied until a thermometer inserted into the soil at any point and at any depth registers above 50° C. Higher temperatures make for additional safety. This method has the disadvantage of leaving the soil in poorer condition for working. The hot-water method is still in the experimental stage, and is not far enough along to warrant any recommendations. Disinfection of Pots, Tools, etc. In starting new houses with clean plants and clean soil, it is very es- sential that everything which is used should be free from any form of inoculum. The first danger is from pots which h^ve been previously used, and which are apt to contain mycelium or spores in the particles of earth which still cling to them. They can be sterilized by immersing ROSE CANKER AND ITS CONTROL. 43 in boiling water for ten minutes. Steaming is just as effective. The method used is simply a matter of convenience. Usually a grower, when he finds disease in his houses, finds it imprac- ticable to destroy all his roses and start all over again. Therefore he retains some of his old houses and starts disinfection operations on one or more, from which he has removed all the plants. This inevitably results in the constant danger of carrying some infested soil or parts of plants from the infested to the clean houses. Every possible precaution should be taken to guard against this, because a failure here means that the work must all be done again. All sorts of tools offer an easy means of conveying the inoculum. Whenever possible an entirely different set of tools should be used in the clean houses, and no tools from the other houses brought in under any conditions. But, if this is not possible, the next best alternative is to sterilize the tools before bringing them in. The method of sterilizing them is not so important as thoroughness. They may be dipped in boiling water, steamed, or a barrel of Bordeaux mixture or formaldehyde — preferably stronger than 1 pint to 25 gallons in this case — may be used for soaking the tools. It may be necessary to sterilize other things besides pots and tools, e.g., boots and clothes of workmen. Every grower, after learning the habits of the pathogene, must decide for himself on the best way, under his own conditions, of keeping his houses clean. Protection of the Host. By protection we mean the placing of a barrier between a plant and a pathogene which would otherwise attack it and cause disease. This is well exemplified in the extensively used practice of spraying plants, the fungicide forming a poison barrier through which the fungus cannot penetrate. The humicolous habit and underground method of attack of the canker fungus seem to preclude anj'^ hope of important benefit from spraying. There is one place in the propagation of roses, however, where a fungicidal covering might be beneficial. Scions and cuttings should, whenever possible, be taken from houses known to be clean. If they are taken from houses in which the disease occurs there is always a possibiUty of spores being lodged on them, even where lesions have not as yet appeared. To either wash off and kill these spores or, at least, to prevent germination where they are, it has been the practice during this investigation to dip all such cuttings in a fungicide before grafting or planting. Comiparative Valiie of Different Fungicidal Coverings. In order to find the best fungicide to use for dipping, and also to secure data for use in case spraying should be found advisable at any time, the comparative value of a number of fungicides was tested in the laboratory. 44 MASS. EXPERIMENT STATION BULLETIN 183. Method. — Glass slides were sprayed with the fungicide to be tested and per- mitted to dry for varying periods of time. Then spores of the fungus in a drop of water were transferred to the center of the sprayed slide, which was then kept in a moist chamber for twenty-four hours. Checks on unsprayed slides were always made at the same time. Percentages of germination were counted at the end of twenty-four hours, and observations were taken for several days to see if there was any further development; but none of the results in these tests were modified by later observations. When a dry fungicide was used it was dusted on to the slide without water. All checks in these tests germinated over 95 per cent. Lime-sulfur. — Concentrations of 1-10, 1-30 and 1-50 commercial lime-sulfur solution were used. The 1-50 concentration proved to be useless from the start. The 1-30 seemed to check germination at first, but after it had been on the slide four or five days over 50 per cent, of the spores germinated. The 1-10 concentration entirely prevented germination when fresh, but after a week the control was erratic, with over 50 per cent, germination on some of the slides. Commercial lime- sulfur seems to be useless for control of this fungus. Dry Sulfur Flour. — Slides were very heavily dusted and the germina- tion tests made at about 25° C. The presence of the sulfur had no effect whatever on the spores. They germinated just as well as the checks. Dry sulfur appears to be even less effective than the lime-sulfur. Ammoniacal Copper Carbonate. — This fungicide prevented germina- tion twenty-four hours after being dried, but when tried a week later was only 25 per cent, efficient. This would hardly be a safe fungicide. Lime. — Milk of lime sprayed on the slides from an atomizer pre- vented germination from the first, and was just as effective as Bordeaux. Milk of lime is not suitable for dipping cuttings. The lime test was made with a different end in view. Bordeaux Mixture. — This fungicide was made up at a strength of 4-4-50. Germination tests were made every day for twenty-one days after the slides were sprayed. No germination occurred in any of these tests. These fungicidal tests clearly indicate Bordeaux mixture as the most suitable solution for dipping cuttings. Treatment of the Walks in the House. Undoubtedly the walks between the benches of a house which has previously grown diseased roses are infested with the pathogene. One could easily think of a great many ways in which small particles of soil from the walks could be carried into the benches. It is therefore necessary either to keep the fungus killed out of the surface of the walks by repeated applications of some fungicide or to cover the walks with some sub- stance which will be a barrier through which it cannot pass up to the benches. In the beginning of this investigation the walks were kept sterile by frequent applications of formaldehyde. This proved unsatis- factory because the fumes of formaldehyde often injure the roses, pro- ducing dead spots on the leaves. This was abandoned and a search ROSE CANKER AND ITS CONTROL. 45 begun for something more suitable. Up to the present, lime gives the best promise of making a satisfactorj^ barrier. Sterile bottle tests show that the mycelium will not grow in soil containing air-slaked lime at the rate of 1^ pounds per cubic foot. Neither will spores germinate in the presence of lime. Until something more satisfactory is found it is recommended that all walks in the houses be kept covered with lime. Not only will this furnish an effective barrier to the fungus coming up from below, but it will also prevent growth of spores and other inocula brought in from other houses on the shoes of workmen and visitors. Immunization of the Host. By immunization we mean either the development of varieties of roses which are immune, — at least highly resistant, — or rendering them immune by injection or feeding through the roots with some chemical. No work has been done along either of these lines in regard to rose canker. From the first it has been noticed that some varieties of roses are more susceptible than others. No doubt in the course of time desirable varieties will be found or developed which will not suffer from canker. How soon that will be no one can predict. A rose breeder of wide national reputa- tion told the writer that he had spent most of his life producing four or five varieties of roses. It is a long process, and until such varieties are developed it will be necessary to resort to such emergency measures as have been described in this bulletin. Summary of Control Measures. In the light of all that we know about rose canker and its causal path- ogene the following measures are recommended for its control : — 1. Carefully inspect the rose house to see if canker is present. If not, employ every means to prevent its entering, — import as few roses as possible from other houses; examine carefully every plant brought in; reject any with suspicious dead areas in the bark. 2. If it is present on the roses it cannot be eradicated from the infected plants. The only hope lies in starting new plants from clean cuttings in clean soil, and guarding against infection at every step in the plant's development. 3. Dip the cuttings in Bordeaux mixture. 4. Sterilize the pots by dipping for ten minutes in boiling water. 5. Sterilize the potting soil and cutting bench soil by steaming to a temperature of over 50° C. for ten minutes or more. Suspected manure should be treated in the same way. 6. Use raised benches, not ground beds. 7. Remove old soil if diseased roses have been grown in it, and soak the benches thoroughly with (1) formaldehyde at the rate of 1 pint to 25 gallons, or (2) boiUng water. 46 MASS. EXPERIMENT STATION BULLETIN 183. 8. Sterilize the bench soil by one of these two methods. If formalde- hj'de is used, apply at the rate of 2 gallons per cubic foot. If boiling water is used, apply until every part of the soil is heated above 50° C. 9. Use a different set of tools in the clean house, or sterilize all tools before bringing them in. 10. Keep the walks in all houses covered with lime. LITERATURE CITED. Morgan, A. P., 1892. "Two New Genera of Hj-phomycetes." Bot. Gaz. 17: 190-192. Ellis, J. B., and Everhart, B. M., 1900. "New Species of Fungi from Various Localities, with Notes on Some Published Species." Bui. Tor. Bot. Club 27: 49-64. Massey, L. M., 1917. "The Crown Canker Disease of the Rose." Phytopathol- ogy 7: 408-417. BULLETIN No. 184 JULY, 1918 MASSACHISETTS AGRICILTIRAL EXPERIMENT STATION Late Dormant versus Delayed Dormant or Green Tip Treatment for the Control of Apple Aphids By W. S. REGAN This bulletin reports the results of comparative tests in the laboratory with commercial lime-sulfur and miscible oils for the destruction of apple aphid eggs at the late dormant period, and experiments with these materials and lime-sulfur-nicotine-sulfate combination for destroying the living aphids at the delayed dor- mant or green tip period. Observations on the extent of foliage injury under both laboratory and field conditions, and the manner in which these insecticides kill, are also made. Requests for bulletins should be addressed to the Agricultural Experiment Station Amherst, Mass. Massachusetts Agricultural Experiment Station. OFFICERS AND STAFF. Trustees. COMMITTEE. Charles H. Pkeston, C Wilfrid Wheeler, Edmund Mortimer, Arthur G. Pollard, Harold L. Frost, hairman, Hathorne. Concord. Grafton. Lowell. Arlington. The President of the College, ex officii The Director of the Station, ex officio. STATION STAFF. Administration. William P. Brooks, ' Ph.D., Director. Fred W. Morse, M.Sc, Acting Director. Joseph B. Lindsey, Ph.D., Vice-Director. Fred C. Kekney, Treasurer. Charles R. Green, B.Agr., Librarian. Mrs. Lucia G. Church, Clerk. Miss F. Ethel Felton, A.B., Clerk. Agricultural Economics. Alexander E. Cance, Ph.D., In Charge of Department. Agriculture. William P. Brooks, ' Ph.D., Agriculturist. Henry J. Franklin, Ph.D., In Charge of Cranberry Investi- gations. Edwin F. Gaskill, B.Sc, Assistant Agriculturist. Robert L. Coffin, Assistant. Botany. A. Vincent Osmun, M.Sc, Botanist. George H. Chapman, Ph.D., Research Physiologist. Paul J. Anderson, Ph.D., Associate Plant Pathologist. Orton L. Clark, B.Sc, Assistarit Plant Physiologist. W. S. Krout, M.A., Field Pathologist. Mrs. S. W. Wheeler, B.Sc, Curator. Miss Ellen L. Welch, A.B., Clerk. On leave. Entomology. Henry T. Fernald, Ph.D., Entomologist. Burton N. Gates, Ph.D., Apiarist. Arthur I. Bourne, A.B., Assistant Entomologist. Stuart C. Vinal, M.Sc, Assistant Entornologist. Miss Bridie E. O'Donnell, Clerk. Horticultlire. Frank A. W.-^-ugh, M.Sc, Horticulturist. Fred C. Sears, M.Sc, Pomologist. Jacob K. Shaw, Ph.D., Research Pomologist. Harold F. Tompson, B.Sc, Market Gardener. Miss Etheltn Streeter, Clerk. Meteorology. Microbiology. John E. Ostrander, A.M., C.E., Meteorologist. Charles E. Marshall, Ph.D., In Charge of Department. Arao Itano, Ph.D., Assistant Professor of Microbiology. George B. Ray, B.Sc, Graduate Assistant. Miss Louise Hompe, A.B., Graduate Assistant. Harold L. Sullivan, B.Sc, Graduate Assistant. Plant and Animal Chemistry. Joseph B. Lindsey, Ph.D., Chemist. Edward B. Holland, Ph.D., Associate Chemist in Charge (Research Division). Fred W. Morse, M.Sc, Research Chemist. Henri D. Haskins, B.Sc, Chemist in Charge (Fertilizer Division). Philip H. Smith, M.Sc, Chemist in Charge (Feed and Dairy Division). Lewell S. Walker, B.Sc, Assistant Chemist. Carleton p. Jones, M.Sc, Assistant Chemist. Carlos L. Beals, M.Sc, Assistant Chemist. WiNDOM A. Allen, ' B.Sc, Assistant Chemist. John B. Smith, * B.Sc, Assistant Chemist. Robert S. Scull, i B.Sc, Assistant Chemist. Harold B. Pierce, B.Sc, Assistant Chemist. James T. Howard, Inspector. Harry L. Allen, Assistant in Laboratory. James R. Alcock, Assistant in Animal Nutrition. Miss Alice M. Howard, Clerk. Miss Rebecca L. Mellor, Clerk. Poultry Husbandry. John C. Graham, B.Sc, In Charge of Department. Hubert D. Goodale, Ph.D., Research Biologist. Miss Grace MacMullen, B.A., Clerk. Miss Elizabeth E. Mooney, Clerk. Veterinary Science. James B. Paige, B.Sc, D.V.S., Veterinarian. G. Edward Gage, i Ph.D., Associate Professor of Animal Pathology. John B. Lentz, » V.M.D., Assistant. 1 On leave on account of military CONTENTS Object of comparative tests, ........ Delayed dormant period indicative of complete hatching of aphid eggs, Object of delayed dormant spraying, ...... Comparative tests for the destruction of aphid eggs under laboratory con ditions, ......... Discussion of results, ........ Action of lime-sulfur and miscible oils upon the aphid eggs, . Comparative tests for the destruction of the living apple aphids. Discussion of results, ........ Efficiency of lime-sulfur against the aphids, .... Action of lime-sulfur upon the aphids, .... Foliage injury by lime-sulfur, ...... Efficiency of lime-sulfur and nicotine sulfate combined against the aphids, .......... Action of the lime-sulfur-nicotine-sulfate combination upon the aphids, .......... Foliage injury by the lime-sulfur-nicotine-sulfate c Efficiency of miscible oils against the aphids. Action of miscible oils upon the aphids. Foliage injury by miscible oils, . Conclusions, ....... Acknowledgments, ...... Bibliography, ...... ombination. PAGE 47 47 48 49 50 50 51 53 53 53 53 Publication of this Document approved by the Supervisor of Administration. BULLETIN No. 184. DEPARTMENT OF ENTOMOLOGY. LATE DORMANT VERSUS DELAYED DOR- MANT OR GREEN TIP TREATMENT FOR THE CONTROL OF APPLE APHIDS. BY W. S. REGAN. In carrying on field experiments during the summer of 1917 for the control of potato plant Uce, commercial lime-sulfur solution, among other materials, was tested as to its effectiveness. Although this was used at the rate of 1 gallon to 22 gallons of water, about twice the ordinary summer strength, and in spite of the fact that every precaution was taken to drench thoroughly all parts of the plants, the percentage of plant lice killed was so small, under 10 per cent., that it could in no way be considered of value as an aphidicide at a strength safe to use upon potato foliage. Object of Comparative Tests. The results of these tests led the writer to question just how effective the usual dormant strength, 1 to 8, of lime-sulfur would prove against apple aphids when appUed at the delayed donnant period, just after the eggs have hatched. With a view to determining this point, a number of tests have been carried out during the past several weeks. In these experiments commercial lime-sulfur solution was used alone and in com- bination with nicotine sulfate, and several brands of proprietary mis- cible oils were also tried out in comparison. Tests were also made to determine the effect of lime-sulfur and miscible oils upon the unhatched eggs. Delayed Dormant Period indicative of Complete Hatching of Aphid Eggs. Remarks might be prefaced here by the statement that the terni dormant is taken to mean the condition of the buds in the winter or early spring before they begin to swell. By late dormant is meant the swollen condition of the buds at the time just before they split open, or 48 MASS. EXPERIMENT STATION BULLETIN 184. in other words just before the buds show the least bit of green. This condition would normally be reached during the early part of April in Massachusetts. The term delaj^ed dormant is applied to that period in the development of the cluster buds and foliage when they have ex- panded from a quarter to a haK inch. It is more or less axiomatic that the hatching of the aphid eggs is about coincident with the first splitting of the apple buds, and that by the time the buds have expanded from a quarter to a half inch, the delayed dormant period, practically all of the eggs have hatched and the young plant lice have migrated to the new growth for food. Observations have confirmed this. Twigs brought in from the field and examined on April 17 had numerous plant lice eggs upon them, but none of these had hatched. The buds were in the late dormant condition. Twigs brought in on April 19 were found to have a few newly hatched individ- uals, which had migrated to those buds just beginning to expand and show the least bit of green available for feeding purposes. From the 19th to the 24th of April, newly hatched aphids appeared in increasing numbers. After the latter date only a few new individuals appeared, which could be readily determined by their size. It is evident from this that under favorable weather conditions such as existed during the period mentioned the time of maximum emergence is rather brief. The presence of a few newly hatched individuals on some of the twigs on May 1 indicated that a small number of belated aphids were still hatching from the eggs, but in no case observed had the foliage expanded beyond about half an inch before hatching was completed. No viviparously produced aphids were in evidence at this time. Object of Delayed Dormant Spraying. In the past the practice of spraying with lime-sulfur for the control of San Jos6 scale has been confined for the most part to the dormant or late dormant season. Comparatively recently, however, the practice of de- layed dormant sprajdng with lime-sulfur has been quite generally advo- cated, based on the assumption that such treatment is fully as effective as dormant or late dormant season applications against the San Jos6 scale, and that apple plant lice in their active stages would offer less resistance to this insecticide than the unhatched eggs. In other words, it is believed by some that a delayed application of lime-sulfur at full dormant-season strength, just after the buds have split open and have expanded perhaps not over half an inch, will control the San Jose scale, and to quite an extent the apple plant lice as well. AppUcations at this time, practice has shown, can be made with little or no eventual injury to the foUage. Our tests, so far as the efficiency of the delayed applications of lime- sulfur in controlling plant lice is concerned, have by no means borne out this conclusion. From the standpoint of the fungicidal value of lime- sulfur, delayed dormant applications appear to have some advantage over those of the dormant season. TREATMENT FOR CONTROL OF APPLE APHIDS. 49 On the other hand it has been recognized by some that only by the addition of nicotine sulfate to the lime-sulfur solution, when this is ap- pUed as a delayed dormant spray, can the aphids be satisfactorily con- trolled. This would indicate that the nicotine sulfate is mainly respon- sible for the control of the plant lice, and that the only reason for delaying the lime-sulfur treatment and combining it with nicotine sulfate is to make necessary only one application instead of two. Then, too, some advocate the addition of an arsenical to the above combination, at the delayed dormant period, for the control of bud moth, case bearers, etc., making possible, theoretically at least, by this insecticide combination the control of San Jose scale, apple aphids and certain foliage feeders by one application. Comparative Tests for the Destruction of Aphid Eggs under Laboratory Conditions. The first tests were made for the purpose of determining the com- parative efficiency of lime-sulfur solution and miscible oils against the unhatched aphid eggs. The lime-sulfur was a fresh sample of a com- mercial concentrate, having a density of 34° Beaume. This was used at the strength recommended upon the container for dormant appUcations, 1 to 8. Two proprietary miscible oils were tested, these being diluted 1 to 15, the usual dormant-season strength. Although both samples were fresh from the manufacturers, one was evidently imperfect as there was some free oil present. In the tests, however, this imperfect sample showed to less advantage in destroying the eggs than the well-prepared sample, a rather unexpected outcome, perhaps, in view of the presence of free oil. These tests, as in the case of those following in which the aim was to determine the comparative killing efficiency, were carried out in the laboratory, where careful counts could be made and results checked. Dipping the infested apple twigs was resorted to rather than spraying, in order to insure uniformity of treatment, as by the latter method any variabiUty of apphcation might lead to an improper interpretation. On examination, shortly after the infested twigs were brought in from the field, it was impossible to make any estimate of the probable number of eggs that would hatch, since a large percentage of the eggs were apparently dead from some cause, as indicated by their shriveled condition. Twigs of as nearly the same size and degree of infestation as possible were se- lected for insecticide treatment and check, the average length of the twigs being about 8 inches. No definite percentage of efficiency can be given for the tests against the eggs. The results should be taken as merely comparative and in the way of a generalization, and are perhaps in need of further verification both in the laboratory and under field con- ditions. The tests against the unhatched eggs were begun when the buds were in the late dormant condition and at such a short time before hatching occurred that it was impossible to carry out verification checks. The results are given in the following table : — 50 MASS. EXPERIMENT STATION BULLETIN 184. Co7nparative Efficiency of Lime-Sulfur and Miscible Oils against Apple Aphid Eggs in the Late Dormant Period under Laboratory Condi- tions. Material and Dilution. Hatch on Treated Twigs. Hatch on Check. Injury to Twigs. Lime-sulfur, 1 to 8. . Miscible oil A. 1 to 15, . Miscible oil B, 1 to 15, . No hatching on three twigs. Thirty-six eggs hatched on three twigs. Seven eggs hatched on three twigs. Twenty-nine eggs hatched. Twenty-four eggs hatched, Fifty-four eggs hatched, . No injury. No injury. No injury. Discussion of Results. While these results can hardly be accepted as conclusive, for the reasons given above, it seems evident that lime-sulfur thoroughly appUed at the late dormant period is highly effective under favorable conditions in destrojdng the aphid eggs, and is certainly more efficient against this stage of the insect than miscible oils. Of course, in dipping the twigs it is to be expected that better results would be obtained than in the ordinary practice of orchard spraying, and it is also true that under field conditions, as will be pointed out under the topic "Action of Lime-sulfur and Miscible Oils upon the Aphid Eggs," discussed later, the intervention of rain between the time of appHcation and the normal hatching period might alter results to a marked degree. This may account to some extent for the frequent ineffective control of apple aphids by the dormant or late dormant season lime-sulfur treatment, with which absolute thoroughness is practically impossible under field conditions, and which has also the added element of uncertainty of results due to the meteorological factor just mentioned. The hatching of a comparatively small number of eggs that have survived treatment might result in quite a severe infestation before the season is far advanced. There is also to be considered the possibility of reinfestation from other sources by migrants in the case of the green apple aphis. The destruction of the eggs or suppression of the stem mothers in the spring does not always guarantee freedom from these insects during midsummer, when supplementary treatments are sometimes desirable or necessary. The miscible oils do not appear to be very effective against the aphid eggs, even with absolute thoroughness of appUcation; and it is probable that a sufficient number of eggs would withstand the treatment, to produce a severe infestation later in the season, unless other measures were taken for control. Action of Lime-sulfur and Miscible Oils upon the Aphid Eggs. — Obser- vations as to the killing power of the lime-sulfur against the aphid eggs indicate that the effectiveness of this material is due mainly to a me- chanical action. On twigs examined after dipping, it was noticed that as the lime-sulfur dried it tended to stick down the eggs and mat the twig TREATMENT FOR CONTROL OF APPLE APHIDS. 51 pubescence over them in such a manner that the delicate insects were apparently unable to force their way from the eggs. This fact — that the action of lime-sulfur against the unhatched eggs appears to be mainly mechanical — presents an element of great uncertainty concerning results that would obtain under field conditions. For instance, the occurrence of a rain between the time of application and the normal hatching time for the eggs might alter results to a great extent, as many of the eggs which are stuck down and potentially unable to hatch would probably thus be liberated, so that hatching might result. This contingency emphasizes the desirability of making the application of the lime-sulfur at the late dormant period if success against the aphid eggs is aimed at, in order to make the space of time between treatment and the normal hatching period as brief as possible, and to eliminate anj^ unfavorable meteorological factors that might lessen the efficiency. As will be shown later the various elements that combine to make aphid control by lime- sulfur treatment against the eggs during the dormant or late dormant periods a matter of much uncertainty, as compared with other practices discussed later, miUtate against its use at either of these periods, unless no other treatment against the aphids is intended, in which case the late dormant treatment should give the most satisfactory results. No such mechanical action was evident in the case of the miscible oils, so that whatever kilUng of the eggs may have resulted from the use of these insecticides was undoubtedlj^ of a chemical nature. Comparative Tests for the Destruction of the Living Apple Aphids. These tests were made against living apple aphids on twigs whose foliage showed varying degrees of expansion from just after the splitting open of the buds, the real delaj^ed dormant period, up to a development of three-fourths of an inch or more, the latter being tested mainly to determine the extent of foliage injury likely to result from the treatment. Full dormant-season strength of lime-sulfur and miscible oils was used and this same strength of lime-sulfur in combination with nicotine sul- fate, observations being made both as to their killing power and their effect upon the foliage. Careful counts were made of the number of living plant lice present upon the twigs before and after the dipping treatment, and from this the killing efficiency of each material could be readily estimated. The results follow: — 52 MASS. EXPERIMENT STATION BULLETIN 184. 05 ^ • >>-J egg _ M m M o o o o o o o o o S-2U o CO o> OJ § g s § g g g g g wse s WOj o M :^ >J .2 I? t? _3. _3 "o f=H .2 'g s >^ >; m K^ >^ :^ » s^ ^ >, ^ ^ 6 _^ s S ■s 5 5 s ■i s _3 _3 3 _3 _3 Is :2 is Is :s 2 :s !s !s ;c 'a 'b W £ 5 5 £ S Jc _M M Ul ,.M .£? .Sf .M M a 'a W cc O in CO M Q cc o5 M ^ E M , S c a" c c cT c © « a) o a p. 0. 0. 0. 13 a .S o o .s 5 _e 1 1 0 i o o a o PQ a 2 5 s 2 o _d .5 .s 2 2 .S 2 5 ^ :>t i:^ :i^ :5t 5 5 :i: ^ :^ ^ s ^■ """^^ ^^ g li - g 2 s o o o o o o o o o i3 • ^1 r L. 6,0 H °£ „ c £ 2 -o| s W5 CB oo •* o D « CO o t- >o -* U5 cs s N M CJ 'Z P3 a 's < P9 O Q << m o Q w < < P9 O H i i i i ° _o 0 o Q z _ -H — -H »-l 2 oj -S i »- 5 D 2 ^ .S J 23 23 2s 2: " S 3 3 3 Q 03 m o OS a 0 a ?; •*3 •J2 •J •rj •J3 •< 8 g g 8 8 ^ 1 •g •g 'S '3 K •a •T3 ■C •a T3 a g g g w s «- « g co oo 00 oo 00- 00- 00- oo" 00- 2 2 2 2 3 £ _o s _o o 2 _o _o "_ "_ ^_ -«_ - '- -_ < PQ 5 «■ s i g ^ s rs 3 .2 •2, ^2 .H '3 o S o 1 "3 3 3 1 •3 3 "3 3 .2 JJ ^ Ji i S 1 t i t 3 3 3 3 g g G g g B B Q S 1 1 'i 'i 3 3 3 3 ;3 ;3 3 a 3 1^ S s § TREATMENT FOR CONTROL OF APPLE APHIDS. 53 DiscussioJi of Results. Efficiency of Lime-sulfnr against the Aphids. — It is evident from the foregoing that Hme-sulfur alone appHed at the delayed dormant period even at full dormant-season strength is practically worthless in con- trolling apple aphids. Actual count shows this material to be under 10 per cent, efficient, and in every case the deUcate, recently hatched aphids were the only ones affected. In addition to those killed, a few were more or less permanently incapacitated, judging from their feeble condition, but even if these were included in the "kill," it would alter the results given only slightl5\ The count to determine the number of plant lice killed was made at later periods of the day on which treatment was applied and on subsequent days until all deaths due to the treatment could be checked up. It should be kept in mind that all the twigs were thoroughly dipped and that the ordinary orchard spraying would prob- ably be even less effective, unless perhaps the apphcation of the spray under pressure might possibly dislodge some of the plant Hce and thus counterbalance the less thorough application. Observations made after treatment showed that the older plant lice were apparently unaiTected and were quietly feeding, except where the coating or dr>dng out of the buds by the lime-sulfur made it necessary for them to seek suitable feed- ing places elsewhere. Action of Lime-sulfur upon the Aphids. — The action of the lime-sulfur upon the young plant lice, the only stage of the active insects against which it appears to have any particular effect, seems to be mainly me- chanical, in that it sticks these delicate young to the twigs in such a manner that death is probably the result of starvation. Death occurred very slowly in some cases, since the insects were often found feebly struggling to liberate themselves several hours after the treatment. Foliage Injury by Lime-sulfur. — The effect of the lime-sulfur upon the opening foliage was noted both in the laboratory and upon field-sprayed trees, where more reliable data of this nature could be obtained. While a number of elements may enter in to affect results, such as the variety of apple, weather conditions, pressure under which the application is made, etc., our tests showed that little or no eventual injury results from the use of dormant-season strength lime-sulfur where the buds have not expanded beyond a half inch. Upon spraj^ed trees, where expansion beyond this point had occurred, injury was more evident, but even on treated trees, with the foliage out three-fourths of an inch to an inch or more, an examination several weeks after application showed Httle other than tip injury in most cases. It seems advisable, however, from the standpoint of thoroughness if for no other reason, to confine such spraj'- ing within the delayed dormant period. It was noted that the long pubescence on foliage that had expanded to about half an inch, but had not unfolded to any extent, appeared to shed the lime-sulfur readily or absorb it only in occasional spots, which resulted in injury at these 54 MASS. EXPERIMENT STATION BULLETIN 184. points; whereas the shorter, matted pubescence of the bark and bud scales absorbed it readily, and on this account more injury was often caused to those buds just splitting open than to those sUghtly more advanced. Efficiency of Lime-sulfur and Nicotine sulfate combined against the Aphids. — Previous tests have shown that nicotine suKate at the dilu- tion 1 to 800 is practically a perfect aphidicide. The addition of Ume- sulfur probably increases its efficiency very Httle, so that the only logical reason for the use of this combination at the delayed dormant period is for the purpose of saving labor by combining two operations — the San Jos6 scale treatment and aphid treatment — in one. Laboratory tests where absolute thoroughness of application by dipping was possible showed this combination to be 100 per cent, effective. The effectiveness of this combination under field conditions would depend mainly on thoroughness of appUcation. Action of the Lime-sulfur-nicotine sulfate Combination upon the Aphids. — As already indicated the action of lime-sulfur in killing the aphids appears to be mainly mechanical, — by sticking them to the plant so that in most cases death is probably the result of starvation. The action of the nicotine sulfate in killing the aphids is rather slow, requiring from about half an hour to twenty-four hours or more for different individuals. Immediately after the dipping there was no evidence that the treatment had any effect upon the aphids. In about fifteen minutes, however, considerable restlessness was apparent and inside of half an hour a num- ber of the plant lice had begun to drop from the twigs, some being pre- cipitated rather forcefully as if from strong muscular contraction. These lay strugghng feebly but unable to crawl, gradually becoming dark colored and motionless. Those plant hce that survived the treatment for a number of hours appeared after a few hours to be paralyzed and in- capable of either locomotion or feeding, but were feebly moving their legs and antennae and excreting honey dew in abnonnally large amounts. An examination of the twigs forty-eight hours after treatment showed all the plant lice to be dead. The fact that nicotine sulfate kills rather slowly may account for the occasional reports that this material is in- effective against plant lice. Examination of treated plants shortly after application might readily lead to this conclusion, but if sufficient time is allowed before examination there will be no question as to its effective- ness. Foliage Injury by the Lime-sulfur-nicotine-sulfate Combination. — A comparison of the effects from the use of full dormant-season strength lime-sulfur alone and in combination with nicotine sulfate on apple foli- age in various stages of development from the first sphtting of the buds to a development of an inch or more showed no noticeable difference. Even at the latter period of development the amount of foliage injury was not serious. TREATMENT FOR CONTROL OF APPLE APHIDS. 55 Efficiency of Miscible Oils against the Aphids. — Tests against the living aphids with two brands of proprietarj^ miscible oils showed a killing efficiency of 100 per cent, for each of these. Action of Miscible Oils upon the Aphids. ■ — The killing action of miscible oils upon the aphids seems to be almost instantaneous. In fact on twigs examined shortly after dipping no movement of the aphids could be noticed. The action is evidently of a strictly chemical nature. Foliage Injury by Miscible Oils. ■ — While spraying with miscible oils for the control of San Jos^ scale is usually confined to the dormant or late dormant season, our tests would indicate that this material, if perfect, can be used at full dormant-season strength during the delayed dormant period with no more injury to the foliage than results from the use of lime-sulfur. At this period in tests conducted both in the laboratory and in the field only slight tip injury resulted; but where the foliage had developed to three-fourths of an inch or more, the injury from the use of the miscible oils seemed to be slightly greater than that resulting from the lime-sulfur treatment. Even this was not serious and was readily overcome as the season advanced. From the foregoing it is evident that where the use of miscible oils for orchard sprajdng is practiced the most economical time for application is during the delayed dormant period, when one application will serve for both the San Jos6 scale treatment and aphid control. Conclusions. 1. The delayed dormant period is usually indicative of the complete hatching of apple aphid eggs. At this time the buds have expanded from a quarter to a half inch. 2. Lime-sulfur solution at full dormant-season strength is less than 10 per cent, effective against the living aphids when applied at the de- layed dormant period. 3. Lime-sulfur applied at the late dormant period, before the buds spHt open and just before the hatching of the aphid eggs, appears to be highly effective, under favorable conditions, in destrojdng the eggs, but the elements of thoroughness of appUcation and unfavorable meteoro- logical conditions present such uncertainty as to results that this treat- ment can hardly be recommended as an effective control. 4. If lime-sulfur is to be used as a control for San Jose scale and no special treatment for apple aphids is to be made later, best results against aphids, if present, are likely to be obtained by a late dormant-season application just before the eggs hatch. Treatment at this time should also be thoroughly effective against the scale. 5. The application of the lime-sulfur (1 to 8) and nicotine sulfate (1 to 800) combination applied at the delayed dormant period gives practically a perfect control for apple aphids and makes unnecessary a separate earlier apphcation of lime-sulfur for San Jose scale. The per- 56 MASS. EXPERIMENT STATION BULLETIN 184. centage of efficiency will depend mainly upon thoroughness of applica- tion. 6. The ordinary dormant-season treatment of apple orchards with miscible oil against San Jose scale, if applied thoroughly at the delayed dormant period, should result in practically a perfect control of apple aphids also. 7. Delayed dormant apphcations of full dormant-season strength lime- sulfur, lime-sulfur and nicotine sulfate combined, and miscible oils, if perfect, can be made without material injury to apple foliage. Even when the foliage is considerably more advanced, little severe injury usually results. This fact, if taken into account, might make unnecessary separate applications for early and late budding varieties. As the foliage becomes more advanced, however, the success of the treatment involves greater difficulty, since the aphids are very difficult to reach when they have the spreading leaves for protection. 8. The action of lime-sulfur in destroying both the aphid eggs and living insects appears to be mainly mechanical, by sticking them to the twigs. 9. The action of nicotine sulfate in killing the hving aphids is slow, requiring from about half an hour to twenty-four hours or more for different individuals. Death appears to be due to paralysis. 10. Miscible oils are practically instantaneous in their killing action against the living aphids. The action is probably of a chemical nature. Acknowledgments . The writer is greatly indebted to Mr. A. I. Bourne of the Massachu- setts Agricultural Experiment Station staff for assistance in carrying out the insecticide tests, and to Dr. H. T. Fernald for his kind suggestions and assistance. Bibliography. 1908. Gillette, C. P., and Taylor, E. P. "Orchard Plant Lice and Their Reme- dies." Bulletin 134, Colorado Agricultural Experiment Station. 1910. Wallace, E. "Spray Injury Induced by Lime-sulfur Preparations." Bul- letin 288, Cornell University Agricultural Experiment Station. 1911. Shafer, G. D. "How Contact Insecticides Kill." Technical Bulletin 11, Michigan Agricultural Experiment Station. 1914. Tartar, H. V. "On the Valuation of Lime-sulfur as an Insecticide." Journal of Economic Entomology, Vol. VII., p. 463. 1915. Parrott, P. J., and Hodgkiss, H. E. "The Status of Spraying Practices for the Control of Plant Lice in Apple Orchards." Bulletin 402, New York Agricultural Experiment Station, Geneva, N. Y. 1915. Shafer, G. D. "How Contact Insecticides Kill." Technical Bulletin 21, Michigan Agricultural Experiment Station. 1916. Mclndoo, N. E. "Effects of Nicotine as an Insecticide." Journal of Agri- cultural Research, Vol. VII., No. 3, p. 89, United States Department of Agriculture. TREATMENT FOR CONTROL OF APPLE APHIDS. 57 1916. Parrott, P. J., Hodgkiss, H. E., and Lathrop, F. H. "Apple Aphids and Their Control." Bulletin 415, New York Agricultural Experiment Sta- tion, Geneva, N. Y. 1917. Quaintance, A. L., and Baker, A. C. "Aphids Injurious to Orchard Fruits, Currant, Gooseberry and Grape." Farmers' Bulletin 804, United States Department of Agriculture. 1917. Parrott, P. J., Hodgkiss, H. E., and Lathrop, F. H. "Plant Lice Injurious to Apple Orchards" (II.). Bulletin 431, New York Agricultural Experi- ment Station, Geneva, N. Y. 1918. Thayer, P. "Delayed Applications of Lime-sulphur." Monthly Bulletin, Ohio Agricultural Experiment Station, Vol. III., No. 3, p. 82. BULLETIN No. 185 JULY, 1918 MASSACHUSETTS AGRICILTIRAL EXPERIMENT STATION The Inheritance of Seed Coat Color in Garden Beans By J. K. SHAW and JOHN B. NORTON This bulletin is a record of the inheritance of seed coat color among certain varieties of garden beans as shown by intercrossing these varieties. There are presented also certain hypotheses to account for the facts observed. It should be of interest to students of genetics, more espe- cially those engaged in or contemplating investigations with this particular group of plants. Requests for bulletins should be addressed to the Agricultural Experiment Station Amherst, Mass. IMassachusetts Agricultural Experiment Station. Trustees. OFFICERS AND STAFF. COMMITTEE. Chables H. Preston, Chairman, . Hathorne. Wilfrid Wheeler, . Concord. Edmund Mortimer, . Grafton. Arthur G. Pollard, . LoweU. Harold L. Frost, . Arlington The President of the College, ex officio. The Director of the Station, ex officio. STATION STAFF. Administration. William P. Brooks, i Ph.D., Director. Fred W. Morse, M.Sc, Acting Director. Joseph B. Lindsey, Ph.D., Vice-Director. Fred C. Kenney, Treasurer. Charles R. Green, B.Agr., Librarian. Mrs. Lucia G. Church, Clerk. Miss F. Ethel Felton, A.B., Clerk. Asrricultural Economics. Alexander E. Cance, Ph.D., In Charge of Department. Agriculture. William P. Brooks, ' Ph.D., Agriculturist. Henry J. Franklin, Ph.D., In Charge of Cranberry Investi- gations. Edwin F. Gaskill, B.Sc, Assistant Agriculturist. Robert L. Coffin, Assistant. Botany. A. Vincent Osmun, M.Sc, Botanist. George H. Chapman, Ph.D., Research Physiologist. Paul J. Anderson, Ph.D., Associate Plant Pathologist. Orton L. Clark, B.Sc, Assistant Plant Physiologist. W. S. Krout, M.A., Field Pathologist. Mrs. S. W. Wheeler, B.Sc, Curator. Miss Ellen L. Welch, A.B., Clerk. On leave. Entomology. Henry T. Fernald, Ph.D., Entomologist. Burton N. Gates, Ph.D., Apiarist. Arthur I. Bourne, A.B., Assistant Entomologist. Stuart C. Vinal, M.Sc, Assistant Entomologist. Miss Bridie E. O'Donnell, Clerk. Horticulture. Frank A. Waugh, M.Sc, Horticulturist. Fred C. Sears, M.Sc, Pomologist. Jacob K. Shaw, Ph.D., Research Pomologist. Harold F. Tompson, B.Sc, Market Gardener. Miss Etheltn Streeter, Clerk. Meteorology. John E. Ostrander, A.M., C.E., Meteorologist. Microbiology. Charles E. Marshall, Ph.D., In Charge of Department. Arao Itano, Ph.D., Assistant Professor of Microbiology. George B. Rat, B.Sc, Graduate Assistant. Miss Louise Hompe, A.B., Graduate Assistant. Harold L. Sullivan, B.Sc, Graduate Assistant. Plant and Animal Joseph B. Lindset, Ph.D., Chemist. Chemistry. Edward B. Holland, Ph.D., Associate Chemist in Charge {Research Division). Fred W. Morse, M.Sc, Research Chemist. Henri D. Haskins, B.Sc, Chemist in Charge (Fertilizer Division). Philip H. Smith, M.Sc, Chemist in Charge (Feed and Dairy Division). Lewell S. Walker, B.Sc, Assistant Chemist. Carleton p. Jones, M.Sc, Assistant Chemist. Carlos L. Beals, M.Sc, Assistant Chemist. Windom a. Allen, i B.Sc, Assistant Chemist. John B. Smith, i B.Sc, Assistant Chemist. Robert S. Scull, i B.Sc, Assistant Chemist. Harold B. Pierce, B.Sc, Assistant Chemist. James T. Howard, Inspector. Harry L. Allen, Assistant in Laboratory. James R. Alcock, Assistant in Animal Nutrition. Miss Alice M. Howard, Clerk. Miss Rebecca L. Mellor, Clerk. Poultry Husbandry. John C. Graham, B.Sc, In Charge of Department. Hubert D. Goodale, Ph.D., Research Biologist. Miss Grace MacMullen, B.A., Clerk. Mrs. Nettie A. Gilmore, Clerk. Veterinary Science. James B. Paige, B.Sc, D.V.S., Veterinarian. G. Edward Gage, » Ph.D., Associate Professi Pathology. John B. Lentz, i V.M.D., Assistant. 1 On leave on account of military service. of Animal CONTENTS. PAGE Introduction, ........... 59 Review of literature, . . . . . . . . .59 Methods, 60 Recording data, .......... 61 Varieties used, .......... 63 Crosses of pigmented with non-pigmented beans, ..... 65 The inheritance of pigment patterns, ....... 67 Mottling 67 Mottling patterns, .......... 76 The behavior of eyedness, ........ 79 The inheritance of pigments, ......... 82 The behavior of the yellow-black determiners, ..... 84 The behavior of the determiners of the red series, . . . .92 The interrelations of the yellow-black and red series, .... 93 Crosses involving Creaseback, . . . . . . . . 96 Crosses involving Davis Wax, ........ 98 Crosses involving White Marrow, ....... 99 The genetic constitution of the varieties used, ...... 101 Svunmary, ............ 102 Bibliography, 103 Publication of this Document approved by the Supervisor of Administration. BULLETIN :N'o. 185. DEPARTMENT OF HORTICULTURE. THE INHERITANCE OF SEED COAT COLOR IN GARDEN BEANS. BY J. K, SHAW AND JOHN B. NORTON. Introduction. Investigation of inheritance in garden beans at this station was begun in 1907 by Mr. C. S. Pomeroy, then assistant horticulturist, who made several crosses during that summer and grew the Fj generation in 1908. Additional crosses were made during the same summer. In the fall of 1908 this crossed seed and that of the Fi generation above referred to fell into the hands of the senior writer, who has been responsible for the con- duct of the investigations since. In the summer of 1913 the junior author came into the work and has since borne a large share. During all this time Prof. F. A. Waugh has had general supervision, and his helpful criticisms and suggestions made from time to time are gratefully ac- knowledged. Review of Literahire. A number of investigators have given time to the study of the in- heritance of seed coat color in beans. Mendel (1) after his classical experi- ments with peas gave some attention to beans, but he discovered little beyond the fact that he had here a more complex problem than that pre- sented by peas, and he was not able to apply the simple 3:1 formula to explain his results. Emerson (2) made many crosses of different horticultural varieties, and observed among other things the behavior of seed coat color. He con- sidered that the mottled offspring exhibited characters not visible in either parent. In a later paper (3) the same author gives the numbers of seeds resulting from a cross of a dark brown and a yellow brown, and from a cross of a black and a white variety. The results of these crosses were similar to those of Burpee Stringless, Giant Stringless, Challenge Black Wax and White Marrow. 60 MASS. EXPERIMENT STATION BULLETIN 185. Further investigation showed Emerson that the theory of mosaics could not explain the behavior of mottled beans resuhing from crosses of non-mottled parents, and he advanced (4) a theory suggested by ShuU, which supposed one factor was responsible for mottling in fixed races and a different factor responsible for mottling in heterozygous forms mentioned above, which is visible in heterozygous individuals only. In another paper (5) Emerson discusses this theory, and develops another theory suggested by Spillman, which supposes mottling to be due to two factors which may exist separately in the heterozj^gous mottled forms or coupled in those forms which breed true to the mottled characters. By this theory the facts reported in the present paper may be explained. Tschermak carried on numerous investigations of the inheritance of seed coat color in beans along with others with stocks and peas. In his most recent paper (22) he analyzes his results, and is able to account for most of them in a satisfactory fashion by means of simple Mendelian factors. ShuU (17) advanced the hypothesis of the appearance of the mottling factor only in heterozygous indi\aduals referred to above. Kajanus (13) reports investigations of the inheritance of colors and color patterns in garden beans, especially of the behavior of a violet marbled tjT^e of mottling due apparently to distinct factors. He reports also on the chemical nature of the pigments involved. Jarvis (11) and Tracy (19) have given excellent descriptions and a quite stable nomenclature of coimnon bean varieties, and Freeman (7) describes several tyj^es of the Mexican frijoles and teparies, P. acutifolivs var. latifolius. Methods. At first, commercial seed procured from the trade was used, but begin- ning in 1909 steps were taken to breed pure races, and earlier crosses made with plants grown from commercial seed were, so far as possible, repeated. Evidence indicated that a few of these earlier parents were probably hybrids, and such crosses have been ignored in the consideration of results. In all cases the plants used for crossing have been externally true to type, but as will appear later, it is probable that in some varieties two or more races have been encountered. In such cases no external differences have been observed in the parent plants, though their behavior on crossing revealed different genetic composition. In making the crosses the procedure of Emerson has been generally followed; that is, emasculation and pollination have been performed in one operation. This method has given sometimes 30 per cent, or more of successful attempts, and at other times a very low percentage of successes. This is probably due in part to unfavorable environmental conditions, and in part to variations in the procedure, — usually the selection of a female blossom that was not sufficiently mature. In a few cases the resulting plants have been like the female parent, indicating that self-fertilization SEED COAT COLOR IN GARDEN BEANS. 61 had taken place before the foreign pollen was introduced, or at least before it could take effect. Some of the crossing was done in the field and some in the greenhouse during the winter. In the latter case the blossoms were not covered, while in the former a one-fourth pound manilla bag was tied tightly over the flower stalk for five or six days, after which it was torn open, and, if the attempt seemed successful, left to indicate the seed pod at the time of harvest. In all cases four generations from the cross have been grown. In each generation except the fourth a certain number of plants chosen more or less at random have been self-fertilized by enclosing them during the blossoming period in cheesecloth or, in a few cases, waxed paper sacks. Neither of these is satisfactory. Both weaken the plant, the waxed paper sacks more than the cheesecloth ones. It has been the invariable observa- tion that there is a progressive weakening of the plants through the four generations. First generation crosses are invariably strong, vigorous plants, often seemingly more vigorous than the parent varieties, while the fourth generation plants are decidedly weak and unproductive. Whether this is due to the repeated self-fertilization or to the weakening effect of covering the plants does not appear. Possibly both have contributed to the result observed. The blossoms have been more or less infested with thrips. None have been observed on covered plants, but it is entirely possible that there may have been cases of such infestation, and that in rare cases a grain of foreign pollen was introduced in the blossom of a plant supposed to be self-fertilized. A few irregularities that may have been due to such a cause have been observed. Nevertheless, the probability that such cases are extremely rare is indicated by a number of considerations. Bean blossoms are commonly self-fertilized before they open, and, according to our observation, thrips does not infest unopened buds. The pollen- carrying ability of thrips cannot be large, and it appears that it does not commonly enter beneath the bags, or it would have been observed in the many examinations of the covered plants. And finally, cases arousing suspicion of the entrance of foreign pollen are extremely rare. Recording Data. The method of securing records of the plants has been previously described (16). It consists, essentially, in assigning to the exjDression of each supposed Mendelian character a special letter designation. The plants have been examined for blossom color, pod color, and for seed coat color. This involves going over the plants not less than three times, and in most cases two examinations have been made for each character, involving examination of the plants six times. In order to identify the individual plants, each is assigned a number in order. The seeds are planted about 6 inches apart, and a small tag bearing its number is early attached to every fifth plant. Thus, in order to ascertain the number of 62 MASS. EXPERIMENT STATION BULLETIN 185. any plant, one has to examine only a very few plants along the row before finding one bearing a tag with its number. When any plant is self-fertilized the fact is noted on the card along with the rest of the record for that plant. A record of the original crosses is kept so that one may trace readily the ancestry of any individual plant back to the original parents. As soon as the seed was well matured a single bean from each plant representative of those on that plant has been selected and preserved. Unfortunately mice gained access to a portion of these seed samples and destroyed many of them. Still, samples representing a majority of the plants grown escaped destruction and are available for examination. Many of the pigments found in the seed coats are subject to change with age, and due allowance must be made in the study of old seed. It has been said that an attempt was made in recording observations to designate the expression of each independent character by a separate letter. The letter used, the color which each stands for, and the name of a variety bearing each color character are as follows : — Seed Coat Color. Found in — A. White, ........ Davis Wax. B. Bufif Blue Pod Butter. C. Yellow, ........ Giant Stringless. D. Medium or bright red, ..... Red Valentine. E. Dark or purplish red, ..... Mohawk. F. Coffee brown, ....... Burpee Stringless. G. Black Challenge Black Wax. H. Olive, ........ Certain crosses. L. Eyedness, ....... All eyed beans. O. Dark mottled, ....... Red Valentine. P. Light mottled, ....... Golden Carmine. Flower. A. White, ........ All white and eyed sorts. B. Light pink, ....... Burpee Stringless. C. Pink, ........ All black seed sorts. D. Crimson, ........ Blue Pod Butter. E. Waxy pink, ....... Certain crosses. The first eight letters stand for separate and quite distinct colors, most of which may be found in one or more of the varieties used. The color H does not appear in any of the varieties used, but does appear in several of the crosses. Attempts have been made to distinguish different eye sizes in eyed beans. There is no doubt that eye size is inherited, but the data secured do not appear clear and definite enough to warrant any positive conclusions; therefore only a brief general report on different eye sizes is made. Mottled beans are of two distinct kinds, — one, designated as "dark mottled," includes those sorts where the darker color or colors predomi- nate, of which there are many varieties other than Red Valentine, the SEED COAT COLOR IN GARDEN BEANS. 63 one cited; the other, called "light mottled," includes those varieties of the Horticultural type. The different blossom colors have been more fully explained in a previous publication (15). Varieties used. During the eight years that the investigations have been in progress twenty-one varieties have been used in the crosses yielding results deemed worthy of consideration. A few others have been used in a very limited way. Including reciprocals, more than 120 different crosses have been made, some of which have been repeated two or three times. The principal varieties used, their blossom and seed coat color, and the letters used to designate them, are as follows: — Blossom. Seed Coat. Vabiety. Color. Letter. Color. Letter. Black Valentine, . Pink C Black G Blue Pod Butter, . Crimson, . D Buff, . . . . B Bountiful, Pink, . C Greenish buff. B Burpee Kidney, White, . A Red mottled eye. EOL Burpee Stringless, . Light pink, . B Coffee brown, . F Challenge Black Wax, . Pink, . C Black G Creaseback, . White, . A White A Currie, .... Pink, . C Black, .... G Da%as Wax, . White, . A White A German Black Wax, Pink, . . C Black, .... G Giant Stringless, . Light pink, . B Yellow, C Golden Carmine, . Light pink. B Light mottled, . EP Golden Eyed Wax, White, . A Yellow eyed, . . CL Keeney Rustless, . White, . A Dark red eyed mottled, EOL Longfellow, . Light pink. B Red mottled. DO Low Champion, Light pink, . D Red D Mohawk, Light pink. B Dark red mottled. EO Prolific Black Wax, Pink, . C Black G Red Valentine, . White, . A Red mottled. DO Wardwell, White, . A Dark red mottled eye. EOL Warren, .... Light pink, . B Dark red, . E Warwick, Light pink, . B Red mottled. DO White Marrow, White, . A White, . . . . A The nomenclature is according to Jarvis (11), and for a full description of the several varieties the reader is referred to his paper or that of Tracy (19). 64 MASS. EXPERIMENT STATION BULLETIN 185. An examination of the above table reveals several more or less constant correlations between blossom color and seed coat color. All white or ej^ed beans are accompanied by white blossoms. So far as the knowledge of the writers goes this is always true, unless it may be in some cases of eyed beans, when the eye is unusually large. With this reservation no certain exceptions have been observed among either commercial varieties or the crosses made. With the exception of Red Valentine, all totally pigmented or mottled beans show more or less color in the blossom. A few plants in certain lots of Red Valentine have shown shght color in the blossom, while in other lots a careful examination showed no colored flowers. As is shown later, more than one strain of Red Valentine has been encountered, and this may account for the occasional appearance of slightly tinged flowers. There are a number of commercial varieties having pigmented seeds and white flowers. In these varieties black beans and pink flowers always go together, and this seems to be generally the case among commercial varieties whether the bean is solid black or black mottled, unless the mottling is confined to a distinct eye. Our records show a number of instances where a black or black mottled bean is said to have been accompanied by a white flower, but such cases are very few among many where the flower is pink, and we are inclined to ascribe them to erroneous observations, usually of blos- som color. Certain pigmentation of the plant as a whole seems to accom- pany certain blossom colors. The crimson flower of Blue Pod Butter is always accompanied by a deep purplish coloration of the entire plant. It is probable that the factor producing the pink flower and black coloration in the seed coat always causes also fine purplish lines on the stems and possibly a darker foliage than is found in non-pigmented plants. Pod color is undoubtedly independent of other coloration of the plant, except that green podded plants have slightly darker green foliage than wax podded varieties. The purplish coloration characteristic of the foliage of Blue Pod Butter, found also in crosses when it is one of the parents, extends to the seed pods whether they are green podded or wax podded. In many cases a more or less obscure reddish or crimson splashing appears on the outside of the seed pod. This is frequently, but apparently not always, associated with mottled seeds. It is clearly seen in varieties of the Horticultural class. Often it does not show until the pod is about to ripen, and disappears with complete maturity. On account of these facts it has been found dif- ficult to secure accurate data bearing on the genetic behavior of this character. Moreover, our attention has been directed more especially to other characters. Our observations indicate that it is a character worthy of more careful study directed especially upon this point. As has been previously intimated, the inheritance of pigmentation in beans is exceedingly complicated. Many independent factors are involved, and through various interrelations of these, varied colors and color pat- terns are produced. These colors and color patterns are not limited in number to the letter designations given. To put it in another way, many SEED COAT COLOR IN GARDEN BEANS. 65 of the letters have been used to designate more than one color, or colors, of different genetic origin, but always similar colors, and usually those that on first encountering we could not certainly differentiate. For ex- ample, the B seed colors of Blue Pod Butter and Bountiful are similar in appearance, but of entirely different genetic constitution, and can be with some difficulty distinguished from each other in the field. It has been the aim to use a given letter within a given cross always for the same color character, and it is thought that this has been usually successful. The appearance of pigment in the seed coat of beans is usually the ex- pression of a complex factor or the concurrence of several factors. In the absence of any one of the elements of this factor complex the beans are unpigmented. If this be the case, crosses of non-pigmented beans may give rise to pigmented offspring. One such cross has been encountered in this work, that of Davis Wax X Michigan White Wax. This does not signify that such crosses are rare, for only three have been made in the course of this work, the other two, Creaseback X Burpee's Fordhook Favorite, and White Marrow X Burpee White Wax, yielding only non- pigmented offspring. As previously reported, numerous crosses of plants bearing white flowers have given rise to plants with pigmented flowers, but all these have been accompanied by pigmented seeds. Had the pos- sible results from intercrossing non-pigmented beans been realized from the first a much larger number of such crosses would have been attempted. Crosses of Pigmented with Non-pigmented Beans. We have a white-coated bean whenever one or more elements of the factor complex for pigmentation are absent, and crosses of such plants with pigmented plants have shown dominance of pigmentation. The pro- portions of pigmented and non-pigmented beans in the F2 generation have been approximately 3:1, yet most crosses show departures from this ratio that, in view of the large numbers involved, may be significant. These results are shown in Table I. In some crosses there is an excess of pigmented beans and in others a deficiency. We have been unable to settle upon any theory that will explain in detail these seeming irregu- larities. If the non-pigmented parent lacks more than one element of the pigment complex an excess of non-pigmented beans in F2 would result, — an explanation of the observed excess of white beans that may or may not be correct. It is possible that the excess of pigmented beans might be explained on the basis of a complex pigmentation factor were it thoroughly understood, but we are unable at present to offer adequate explanation of all the departures from a 3:1 ratio that have been observed. Some of the crosses involving Creaseback show very great departures from a 3:1 ratio. In 97a and 331a the number of white beans is very few. Both these must be crosses, for Creaseback is a pole bean, and pole beans have appeared in considerable numbers in 97a, and most of the beans in 331a were entirely unlike Warwick, the female parent. This behavior of Creaseback will be more fully discussed later. 66 MASS. EXPERIMENT STATION BULLETIN 185. Table I. — Crosses of Pigmented with Non-pigmented Beans. iNT Varieties. Fs and F4 (Pig- mented Parents only). Blue Pod Butter (P) X White Marrow (W), . White Marrow (W) X Blue Pod Butter (P), . Totals Ratios Burpee Stringless (P) X White Marrow (W), . White Marrow (W) X Burpee Stringless (P), . Totals Ratios Currie (P) X White Marrow (W). . Ratios, White Marrow (W) X German Black Wax (P), Ratios, White Marrow (W) X Golden Carmine (P), . Ratios, Golden Eyed Wax (P) X White Marrow (W), White Marrow (W) X Golden Eyed Wax (P). Totals Ratios, White Marrow (W) X Keeney Rustless (P), . Ratios White Marrow (W) X ProUfic Black Wax (P), . Ratios Red Valentine (P) X White Marrow (W), . White Marrow (W) X Red Valentine (P), Totals, Ratios Blue Pod Butter (P) X Creaseback (W), Creaseback (W) X Blue Pod Butter (P), Creaseback (W) X Blue Pod Butter (P), Ratios, Challenge Black Wax (P) X Creaseback (W), Ratios, Challenge Black Wax (P) X Creaseback (W), Golden Eyed Wax (P) X Creaseback (W), . Ratios Creaseback (W) X Prolific Black Wax (P), . Ratios Pigmented. 77 57 134 2.21 54 46 100 3.22 122 2.49 2.63 12 1.71 14 2.00 33 3.00 71 33 104 3.15 144 5 73 2.43 145 3.92 101 3.33 26 1.S3 White, 35 Pigmented. White. 87 196 497 2.55 302 68 175 205 477 1.59 57 2.70 22 3 3.67 13 110 4.33 185 70 129 63 314 29 60 65 97 3.42 SEED COAT COLOR IN GARDEN BEANS. 67 Table I. — Crosses of Pigmented with Non-pigmented Beans — Concluded. Cross No. Parent Varieties. Fa. F3 and F4 (Pig- mented Parents only). 331 331o 332 73 Warwick (P) X Creaseback (W), . Warwick (P) X Creaseback (W). . Creaseback (W) X Warwick (P), . Totals (omitting 331a) Ratios Challenge Black Wax (P) X Davis Wax (W), . Ratios Pigmented. White. 48 14 78 23 S.S9 : 1 243 84 2.89 : 1 Pigmented. White. 131 5 104 103 40 110 42 2.62 : 1 242 73 329 3.32 : 1 The Inheritance of Pigment Patterns. The disposition of pigments over the surface of the bean may be even, in which case we call it self-colored; or the pigments may be irregularly disposed, revealing the separate colors in short stripes or splashes, when we have a mottled bean. The mottling or the self-color may be limited to a more or less well-defined area around the hilum, giving us an eyed bean. These two pigment patterns, mottling and eyedness, will be separately considered. Mottling. There are many varieties of beans with mottled seeds. The colors involved are various, and the type of mottling differs in different varieties. The inheritance of the various colors is dealt with in a later section. The various types of mottling are without difficulty separated into two classes, — a light mottling shown in various varieties of the Horticultural class, and a dark mottling shown by Red Valentine, Refugee and many others. Many crosses involving both types of mottling have been made, and the mottling always breeds true. There are also many crosses where only non-mottled parents have yielded mottled beans, both of the light and dark mottled types. But in no case have these mottled beans bred true. This is in accord with other investigations, and a theory to account for the facts has been set forth by Emerson (4) on the suggestion of Spillman. This theory supposes that mottling is brought about by two factors, Y and Z, which are coupled in the case of true-breeding mottled varieties, but may be separately borne by distinct varieties, and in such cases are inherited independently. Individuals from such crosses bearing both Y and Z are mottled and always heterozygous, while those bearing either one are not mottled. Whether or not this is the final and complete ex- planation of mottling in beans, it serves to explain the results thus far obtained. The following crosses of mottled beans have bred true, yielding only mottled progeny: — 68 MASS. EXPERIMENT STATION BULLETIN 185. Cross No. Parent Varieties. Total Number of Progeny. 215 Golden Carmine X Mohawk, 77 258 Red Valentine X Keeney Rustless, 109 262 Wardwell X Keeney Rustless 168 273 Mohawk X Red Valentine, . 78 274 Red Valentine X Mohawk 281 Golden Carmine is of the light mottled tjqDe, and Keeney Rustless and Wardwell are mottled-eyed beans; all the others are of the common dark mottled type. Table II. shows the results of crossing mottled and self-colored varieties. In all such crosses the Fi generation has yielded only mottled beans. The F2 generation has been composed of mottled and self-colored beans in proportions approximating 3:1, though rather wide departures will be noted. These departures are subject to the same comments as those in crosses of pigmented and non-pigmented beans shown in Table I. All extracted self-colored beans have bred true and mottled beans have proved homozygous in mottling in some cases and heterozygous in others, as shown in the table. None of the mottled varieties in this table are of the light or Horticultural type. Wardwell and Keeney Rustless have mottled eyes. Golden Eyed Wax has a self-colored eye, while the other self-colored varieties are totally pigmented and of various colors. Table II. — Crosses of Mottled with Self-colored Beans. Cross No. Parent Varieties. F2. Fa and F4. 19 Blue Pod Butter (S) X Mohawk (M), . Mottled. 8 Self. Mottled. 55 43 1 Self. 21 20 Mohawk (M) X Blue Pod Butter (S), . 7 4 - Totals 15 4 55 21 S.75 : 1 7 2.62 26 15 90 : 1 23 Blue Pod Butter (S) X Red Valentine (M), 23 9 Ratios 3.29 : / 10 : ; 29 Blue Pod Butter (S) X Warwick (M), . 38 37 30 Warwick (M) X Blue Pod Butter (S), . 106 51 16 109 3 Totals 144 61 106 40 Ratios ..... 2.36 4 2.65 54 : / 54 Mohawk (M) X Burpee Stringless (S), 24 57 Burpee Stringless (S) X Red Valentine (M), 32 13 82 24 58 Red Valentine (M) X Burpee Stringless (S), 63 30 180 31 Totals 95 43 262 55 2.21 : 1 4.76 : 1 SEED COAT COLOR IN GARDEN BEANS. 69 Table II. — Crosses of Mottled with Self-colored Beans — Concluded. Cross No. Parent Varieties. F2. Fa and F^. 95 Challenge Black Wax (S) X Warwick (M), . . Ratios, Mottled. 34 S.62 Self. 13 : 1 41 Mottled. 136 179 2.61 68 Self. 52 : 1 115 Currie (S) X Mohawk (M) 158 34 116 Mohawk (M) X Currie (S) 20 6 26 7 Totals 178 47 9 94 41 Ratios, S.79 : 1 2.29 : 1 119 120 Currie (S) X Red Valentine (M), Red Valentine (M) X Currie (S), ... Totals 116 287 403 50 151 201 117 201 77 176 194 46 19 65 Ratios 2.00 : 1 2.98 : 1 193 Giant Stringless (S) X Mohawk (M), . 12 2 19 16 67 99 86 7 194 Mohawk (M) X Giant Stringless (S), . 13 2 23 Totals 25 4 30 6.25 : ; 9 54 : / 197 Giant Stringless (S) X Red Valentine (M), 25 36 198 Red Valentine (M) X Giant Stringless (S). 30 5 53 22 Totals 55 14 107 58 Ratios 3.9S : 1 14 75 1.84 95 122 136 72 231 : 1 287 288 Prolific Black Wax (S) X Red Valentine (M), . Red Valentine (M) X Prolific Black Wax (S), . 27 180 28 47 Totals 207 89 75 Ratios 2.3$ : / 7 3.08 23 12 1.64 : 1 348 Blue Pod Butter (S) X Refugee (M), . 5 14 Ratios .71 : 1 : / 27 Blue Pod Butter (S) X Wardwell (M), 10 1 94 56 78 111 172 28 28 Wardwell (M) X Blue Pod Butter (S), 35 11 30 Totals 45 12 58 S.75 : 1 7 2.97 11 9 5.50 43 41 3.31 81 55 2.38 : 1 61 Burpee Stringless (S) X Wardwell (M), 15 2 Ratios 2 14 : 1 : / 191 Giant Stringless (S) X Keeney Rustless (M), 4 13 Ratios 4.00 : 1 21 : 1 201 Giant Stringless (S) X Wardwell (M), . Ratios 42 2.00 34 : / 244 Wardwell (M) X Golden Eyed Wax (S), 21 12 44 18 2.31 3 19 1.00 19 Ratios 1 75 = '^ ; 357 Lotigfellow (M) X Golden Eyed Wax (S), . 4 3 : 1 : 1 70 MASS, EXPERIMENT STATION BULLETIN 185. In cross 54, Mohawk X Burpee Stringless, in the F3 and F4 genera- tions, 54 plants yielded only mottled beans. This is explained by the fact that only two parent plants were involved, and both happened to be homozygous for mottling. In Table III. are shown the results obtained from crosses of mottled and white varieties. In all such crosses the Fi beans have been mottled,, and all extracted whites have bred true. Extracted self-colored beans have sometimes bred true and sometimes yielded self-colored and white in approximately a 3:1 ratio, never mottled beans. As shown in the table, the usual result in F2 seems to be a 9:3:4 proportion. Table III. — Crosses of Mottled with White Beans. Cross No. Parent Varieties. F2. F3 AND F4 (Mottled Par- ents onlt). M. S. W. M. S. W. 141 230 309 309o 310 327 366 331 331a 332 Davis Wax (W) X Keeney Rustless (M). White Marrow (W) X Golden Carmine (M), . Red Valentine (M) X White Marrow (W), Red Valentine (M) X White Marrow (W). . White Marrow (W) X Red Valentine (M), Wardwell (M) X White Marrow (M), White Marrow (W) X Burpee Kidney (M), . Warwick (M) X Creaseback (W), . . . Warwick (M) X Creaseback (W), . Creaseback (W) X Warwick (M), . 17 11 82 38 16 32 6 5 8 8 5 22 7 1 25 24 12 7 34 13 9 5 3 8 1 3 11 18 7 149 45 81 17 5 28 14 15 8 10 18 9 121 16 124 5 19 101 ~ 5 21 36 4 3 11 10 29 3 2 3 26 6 10 6 13 3 3 8 3 6 37 14 Cross 141, Davis Wax X Keeney Rustless, yielded no self-colored beans. It will be shown later that Davis carries the coupled factors YZ, and as soon as pigment is introduced yields mottled beans. This being true, and Keeney Rustless also bearing YZ, no self-colored beans can appear. In cross 309 it is evident that two strains of White Marrow are involved, the one in 309a being like Davis Wax in bearing the coupled YZ, and the other strain only one of these factors, thus permitting the appearance of self-colored beans. In crosses 331 and 332 there are certain irregularities due to Creaseback that will be discussed later. Most of our crosses among self-colored beans have yielded only self- colored progeny, no mottled or white beans appearing. A list of such crosses follows : — SEED COAT COLOR IN GARDEN BEANS. 71 Cross No. Total Number of Progeny. 43 Burpee Stringless (S) X German Black Wax (S) 419 44 German Black Wax (S) X Burpee Stringless (S), 437 50 Golden Eyed Wax (E) X Burpee Stringless (S), 410 55 Burpee Stringless (S) X Prolific Black Wax (S), 75 81 Challenge Black Wax (S) X Golden Eyed Wax (E), 459 87 Challenge Black Wax (S) X Prolific Black Wax (S), 112 Golden Eyed Wax (E) X Currie (S). . 879 189 Giant Stringless (S) X Golden Eyed Wax (E), . 266 190 237 Golden Eyed Wax (E) X Giant Stringless (S), . Golden Eyed Wax (E) X Prolific Black Wax (S), 213 419 346 Black Valentine (S) X Prolific Black Wax (S), . 108 349 Blue Pod Butter (S) X Warren (S), . 18 350 Bountiful (S) X German Black Wax (S), . 11 351 Bountiful (S) X Prolific Black Wax (S), . 75 354 German Black Wax (S) X Bountiful (S), . 81 362 Prolific Black Wax (S) X Bountiful (S) 56 Crosses of a number of self-colored varieties have yielded only mottled individuals in Fi, and mottled and self-colored individuals in F2, in what seems to be roughly a 1:1 proportion. These are shown in Table IV. Table IV. — Crosses of Self-colorsd Varieties yielding Mottled Progeny. Cross No. Parent Varieties. F2. F3 AND Fi (Mottled Par- ents only). M. S. M. S. 1 Blue Pod Butter X Burpee Stringless, . 159 146 165 170 2 Burpee Stringless X Blue Pod Butter, 78 88 28 22 Totals 237 234 193 192 3 Blue Pod Butter X Challenge Black Wax , 39 25 30 4 Challenge Black Wax X Blue Pod Butter , 92 125 38 28 Totals 128 7 164 16 63 8 58 5 Blue Pod Butter X German Black Wax, 5 6 German Black Wax X Blue Pod Butter, 48 27 26 35 Totals 55 2 43 1 34 11 40 11 Blue Pod Butter X Giant Stringless, 12 12 Giant Stringless X Blue Pod Butter, 26 51 20 32 28 20 52 22 31 10 54 15 Blue Pod Butter X Golden Eyed Wax, 11 16 Golden Eyed Wax X Blue Pod Butter, 25 32 43 42 Totals, 45 69 54 59 53 27 53 21 Blue Pod Butter X Prolific Black Wax, 42 22 Prolific Black Wax X Blue Pod Butter. 84 91 39 51 Totals, 153 36 150 26 66 38 93 343 Low Champion X Blue Pod Butter, 48 72 MASS. EXPERIMENT STATION BULLETIN 185. Extracted self-colored individuals have bred true, and no unpigmented beans have appeared. We may note at this point that Blue Pod Butter is one of the parents of all these crosses. The explanation of this is that Blue Pod Butter is the only self-colored bean bearing the factor Y, all other self-colored varieties carrying the other factor for mottling, designated as Z; and as, according to Emerson's theory, mottling can result only when Y and Z are both present, the variety named is the only self-colored variety that can produce mottling when crossed with the other self-colored variety used. While the proportion 1:1 is held quite closely when the total numbers of reciprocal crosses are con- sidered, it may be noted that in all cases except the crosses involving Blue Pod Butter with Golden Eyed Wax and Challenge Black Wax there is an alternate preponderance of mottled and self-colored beans in the two members of the reciprocal crosses, a fact that may have a significance, or be onlv a chance occurrence. Table V. — Crosses of Self-colored with White Beans yielding Mottled Progeny. Cross No. Parent Varieties. F2. F, AND F4 (Mottled Par- ents only). M. S. W. M. S. W. 7 8 33a 33 34 67 68 73 129 184 249 250 298 Blue Pod Butter (S) X Davis Wax (W), . Davis Wax (W) X Blue Pod Butter (S), . Blue Pod Butter (S) X White Marrow (W), . Blue Pod Butter (S) X White Marrow (W), . White Marrow (W) X Blue Pod Butter (S), . Burpee Stringless (S) X White Marrow (W), . White Marrow (W) X Burpee Stringless (S), . Challenge Black Wax (S) X Davis Wax (W), . Currie (S) X White Marrow (W), . White Marrow (W) X German Black Wax (S), Golden Eyed Wax (S) X White Marrow (W), . White Marrow (W) X Golden Eyed Wax (S), . White Marrow (W) X Prolific Black Wax (S). . 14 38 74 40 38 35 182 63 59 46 12 19 3 13 41 17 17 16 11 68 58 19 22 6 14 6 16 24 23 17 14 84 49 32 23 4 11 36 19 18 40 8 93 135 145 21 47 31 85 124 102 23 100 59 72 32 40 209 35 9 3 17 3 49 52 28 21 53 14 10 20 4 9 3 12 60 9 7 48 5 19 29 31 12 1 14 8 2 21 12 13 23 8 22 38 34 23 28 6 30 4 18 15 9 12 6 4 I SEED COAT COLOR IN GARDEN BEANS. 73 At least two of the white varieties used in this work, Davis Wax and White Marrow, seem to carry the factors for mottling, and in most cases they have yielded in F2 mottled, self-colored and white beans in what is probably a 9:3:4 ratio. Crosses with these varieties are shown in Table V. All extracted whites have bred true, and extracted self-colored beans have either bred true or yielded self-colored and mottled beans in approxi- mately a 3:1 ratio. In several cases mottled beans have been extracted which bred true, thus indicating that in some cases at least both Davis and "Wlaite Marrow carry both Y and Z; that is, they are really mottled beans lacking pigment. In cross 33a no mottled beans appear, probably because Blue Pod Butter and the particular strain of White Marrow involved carry the same mottling factor, and both likewise lack the other one. It is certain that a different plant of Wliite Marrow was used and one from a commercial stock, while in 33 and 34, individuals of a selfed strain were used, and this strain was not derived from the plant used in 33a. In the cross of Golden Eyed Wax X WTiite Marrow (Table V.) the beha%aor as regards mottling is as expected from the above considera- tions. In another cross of what were supposed to be the same varieties no white beans appeared. The behavior of the progeny was exactly what would be expected of a cross of Golden Eyed Wax X Warwick. Warwick and White Marrow were grown next to each other in the row, thus making it easy to make an error in obtaining blossoms. We are therefore inchned to believe that the irregularity was due to such an error in pollination. According to Emerson's theory of mottling all mottled varieties have the constitution PYZ in which formula P indicates the factor for pig- mentation and YZ the coupled factors for mottling. Non-mottled pig- mented beans can have only one of these factors bearing either PYz or PyZ. White beans may be either pYZ, pYz or pyZ. The possible re- sults of intercrossing these types of beans are as follows: — Case No. Cross Constitution. Color of Beans. Proportion of MoTTLFD, Self and White in F2. M. S. W. 1 2 3 4 5 6 7 8 PYZ X PYz, . PYZ X PyZ. . PYZ X Pyz. PYZ X pYZ, . PYZ X pYz. . . PYZ X pyZ, PYZ X pyz, . . PYz X PyZ, m X s, . m Xs, . m X s, . m X w. m X .w, m X w, m X w, sXs, . 3 3 3 3 9 9 9 2 1 1 1 3 3 3 2 1 4 4 4 74 MASS. EXPERIMENT STATION BULLETIN 185. Case No Cross Constitution. Color of Beans. Proportion op Mottled, Self, and White in Fz. M. S. W. 9 PYz X Pyz, sXs, . . . - 4 _ 10, . PYz X pYZ, sX w, 9 3 11, . PYz X pYz, sX w, - 3 12, . PYzSXfcyZ. sXw, 6 6 13. . PYz Xfpyz. s X w. - 3 14, . PyZ X Pyz. sXs. - 4 15, . PyZ X:pYZ, sXw, 9 3 ^ 16. . PyZ X pYz, sXw, 6 6 17, . PyZ X pyZ, sX w, - 3 18, . PyZ X pyz. sXw. - 3 19, . Pyz X pYZ, sXw, 9 3 20, . Pyz X pYz. sXw, - 3 21, . Pyz X pyZ, sXw, - 3 22, . Pyz X pyz. sXw, 1 - 3 The results secured in the work here reported can be quite satisfactorily explained on the above theory. All crosses of mottled beans have yielded only mottled beans, as shown on pages 67 and 68. Some crosses of self-colored beans have jdelded mottled progeny. (See Table IV.) In most such crosses Blue Pod Butter is one of the parents. If it has the constitution PYz then the other members of the crosses must be PyZ. Self-colored beans of either of the above types, when crossed with mottled beans, have yielded mottled and self-colored beans in the proportion of approximately 3:1, as shown in Table II. The mottUng factors of white beans are not so readily determined, and there seems to have been more than one strain of some of the white varieties used. Davis Wax seems alwa5'-s to carry the coupled factors YZ. (See Tables III. and V.) It is probable that there are three strains of White Marrow, as follows : • — Constitution. Found in Crosses — pYZ, 33. 34 (case 10), 67, 68. 129. 184, 249, 250. 298 (case 15), pyZ pYz 309o (case 4). 230, 309, 310, 366 (case 6). 33o (case 11). Crosses involving Creaseback. — In crosses involving Creaseback the beans in Fi have always been black or nearly so. In the cross with Chal- lenge Black Wax the beans were nearly black, but with faint signs of SEED COAT COLOR IN GARDEN BEANS. 75 mottling. In later generations black beans predominate, with some signs of indistinct mottling in some cases. The occasional appearance of motthng suggests that one or both mottUng factors are carried by Crease- back. The fact that mottling appears with Blue Pod Butter which in all other crosses seems to carry the Y only, and with Challenge Black Wax which carries the Z, indicates that Creaseback must carry both Y and Z, or that more than one strain has been used. If coupled factors are present there should appear beans breeding true to the mottled character. No such cases have been clearly shown. If we assume that the appearance of sohd or nearly solid black beans is due to the presence of an additional factor X, which renders the black color epistatic to mottUng, we have a hj^pothesis that is fairly well supported by the limited data available. These data are shown in Table VI. Crosses 31 and 32 Table VI. — Crosses involving Creaseback. Parent Varieties. Blue Pod Butter X Creaseback, Creaseback X Blue Pod Butter, Blue Pod Butter X Creaseback, Creaseback X Blue Pod Butter, Challenge Black Wax X Creaseback Challenge Black Wax X Creaseback Challenge Black Wax X Creaseback Challenge Black Wax X Creaseback Golden Eyed Wax X Creaseback, . Creaseback X Prolific Black Wax, . Fs AND F4. were among the earlier crosses made, and while no indi\ddual records of mottled beans in F2 were kept, it is evident that motthng did occur, but it was very faint and nearly obscured by black in most cases. There were a few dark mottled beans, however, and one of these being selfed gave the proportions of mottled, self-colored and white beans shown in the table. In crosses 31 and 32, Table VI., Creaseback may have the formula yZ, for in this case, assuming the presence of X in Creaseback 76 MASS. EXPERIMENT STATION BULLETIN 185. and a formula of Yz for Blue Pod Butter, we should get a proportion of 6 mottled, 42 self-colored, and 16 white, which proportion is rather closely approximated in both crosses 31 and 32. The crosses with Challenge Black Wax seem to present different combinations of characters. Number 97 was one of the early crosses, and the obscure mottling earlier referred to appeared, but no record was preserved. Cross 97c was made later when the appearance of mottling was more clearly appreciated, and these two may be of the same nature. Crosses 97a and 976 are probably ahke, and the failure of any white seeded beans to appear in 97a due to chance. We are unable to explain the small proportion of white beans, unless it may be on the basis of difference in the pigment complex earlier referred to. In cross 247, Golden Eyed Wax X Creaseback, no mottled beans are recorded in F2, but in later generations obscurely mottled beans do appear, and it is not impossible that a closer study of the F2 generation would have revealed their presence. Unfortunately these samples are among those destroyed. This variety is worth further study and a full comprehension of its behavior, and the reasons therefor would probably throw much hght on the inheritance of pigmentation, not only in beans but in a general way. Another variety that apparently behaves in a similar way is Crystal Wax. Oweni reports that crossed with Round Pod Kidney (Brittle Wax) there appeared in Fi colored and dark mottled, nearly black beans, and the F2 plants were 10 mottled, 24 self-colored and 10 white, nearly all of the self-colored seeds being black. Mottling Patterns. Among the commercial varieties of mottled beans two prevailing types of mottling are evident. Both show as a ground color a sort of buif or ecru. In the darker mottling, represented by Red Valentine and Refugee, this color prevails over only a small part of the seed, while in the lighter, represented by varieties of the Horticultural class, it covers three-fourths or more of the surface. Some evidence indicating that this buff color is the same thing in both light and dark mottled beans will be presented later. When crossed, the darker tj^pe of mottling seems to behave as a simple dominant in the single cross that has been made. Table VII. — Light and Dark Mottling. Parent Varieties. F2. F3 AND F4. Cross No. 0 Parents. 0 Parents. 0. 0. 0. 0. 0. 215 Golden Carmine (0) X Mohawk (0), . 1 1 33 6 10 21 Report N. J. Experiment Station, 1908, p. 456. SEED COAT COLOR IN GARDEN BEANS. 77 In the above table and the one following, O represents the dark or Red Valentine tjq^e of mottling, and o the light or Horticultural type. The beha\aor of W^iite Marrow and Davis Wax in crosses with colored beans indicates that both these varieties possess one or both of the factors for mottling, as has already been shown (page 73). There is no evidence that the factor, O, for dark mottling is present in either variety. Crosses of these two varieties with Blue Pod Butter (Table V.) yield no dark mot- tled beans, indicating that Blue Pod Butter does not possess the O factor. Therefore Blue Pod Butter may be described as PYzo, and the two white varieties as pyZo or pYZo. All dark mottled varieties may be described as PYZO. All other pigmented self-colored sorts used in these experiments may be described as PyZO, except Warren, which is probably like Blue Pod Butter so far as mottling factors are concerned. The results of crossing Wliite Marrow and Davis Wax with a number of pigmented varieties are shown in Table VIII. A study of the results Table VIII. — Mottli ng Factors in White Beans. Fj. Fa AND F4. Cross Parent Varieties. 0 Parents. ©Par- S Par- No. ents. ents. 0. 0. S. W. 0. o. S. W. o. W. S. W. 230 White Marrow (W) X Golden Carmine (M). - " 5 7 - - - - 7 107 22 4 3 1 309 Red Valentine (M) X White Marrow (W). 33 9 ~ 13 22 25 4 21 4 ~ 16 19 33 11 ~ 366 White Marrow (W) X Burpee fi 2 3 7 9 _ _ _ 5 2 Kidney (M). 3 30 2 2 14 21 327 Wardwell (M) X White Mar- 10 _ 2 4 _ 23 _ _ 11 4 row (W). 327a Wardwell (M) X White Mar- row (W). 17 4 - 2 12 3 2 2 9 2 - - 141 Davis Wax (W) X Keeney Rustless (M). 14 4 - 2 5 16 6 - 3 - - ~ - 67 Burpee Stringless (S) X White Marrow (W). 39 9 16 17 51 17 18 25 5 30 9 6 29 9 126 39 52 28 21 68 White Marrow (W) X Burpee Stringless (S). 23 12 12 14 25 17 18 - 94 18 30 78 27 184 White Marrow (W) X German Black Wax (S). 42 17 19 32 1 I 0 0 ?. 4 4 5 6 0 0 15 5 13 5 5 73 Challenge Black Wax (S) X Davis Wax (W). 141 51 68 84 46 27 19 4 9 7 2 6 13 18 3 4 23 9 11 3 28 6 1 3 21 209 3 84 55 34 249 Golden Eyed Wax (S) X White Marrow (W). 20 26 22 33 29 14 10 16 61 28 16 45 26 250 White Marrow (W) X Golden Eyed Wax (S). 8 5 6 4 17 5 4 2 2 7 1 38 30 11 38 52 7 here shown indicates that the factor O just described is associated with the Z mottling factor. If this be the case, on crossing a colored bean PyZO with a white bean pYZo we should get in F2 a proportion of six 78 MASS. EXPERIMENT STATION BULLETIN 185. dark mottled, three light mottled, three self-colored, and four white, which is in harmony with the results shown in the table. No dark mottled beans could breed true, and no extracted hght mottled beans could yield self- colored offspring. In cross 230 Golden Carmine, which must be, according to the fore- going hypothesis, of the constitution PYZo, when crossed with White Marrow yields no dark mottled l)eans, but does jdeld self-colored beans. White Marrow must therefore be pyZo, and the proportion in F2 one of 9:3:4. The self-colored beans in F3 and F4 are from the heterozygote parents, and are not, like the other light mottled beans, extracted from the heterozygote. In cross 309a no self-colored beans are produced. Red Valentine must, from its appearance, be PYZO, and White Marrow must be pYZo. The theoretical F2 proportion — 9 dark mottled, 3 light mot- tled and 4 white — is closely approximated. In cross 366 Burpee Kidney is like Red Valentine and White Marrow pyZo as in cross 230, the non- appearance of light mottled beans in F2 being due to small numbers. In cross 327 Wardwell, a bean with a dark mottled eye, when crossed with White Marrow yields no light mottled beans, while in 327a light mottled beans appear, but no self-colored ones. This can be explained on the assumption that in cross 327 the Wliite Marrow plant used was of the pyZo strain, while in 327a a plant of the constitution pYZo was used. In cross 141, Davis Wax X Keeney Rustless, no self-colored beans are produced, and as in all other crosses of Davis Wax it has the formula pYZo, while Keeney is PYZO. In crosses 67 and 68 Burpee Stringless must be PyZO and White Mar- row pYZo. On the assumption that the O and Z factors are associated or coupled, the failure of hght mottled progeny to appear in the proportion 18:0:6:9 must be due to the small numbers involved, and this lot belong properly on the second line above, it being of the same constitution as the F2 heterozygote. Similar cases are found in crosses 181 and 73. The appearance of a single self-colored plant from a light mottled parent in cross 68 is unexplained unless it be a stray plant. Such a plant undoubtedly did appear in a lot all of which were supposed to be from a light mottled parent plant. It is not thought that these seeming irregularities are suf- ficient to throw serious doubt upon the general theory of the inheritance of types of mottling, but they are recorded in order to fully present the facts as they have appeared. Besides the types of mottling here discussed a wholly different type has been encountered in certain crosses involving White Marrow. This is a fine marbling or cloudy motthng, bluish, brownish or bluish black in color. It is similar to that shown by the variety Cut Short. Data bearing on this are limited. In a cross of Prolific Black Wax X White Marrow this type of mottling appeared, sometimes covering the whole bean and sometimes confined to a limited area, giving an eyed bean. Three plants with this type of mottling yield the parent type and white in the numbers of 6:9, 20:4 and 5:1, respectively. They have been extracted from both self- colored and dark mottled parents. SEED COAT COLOR IN GARDEN BEANS. 79 The Behavior of Eyedness. In many varieties of pigmented beans the pigment is centered around the hilum, producing the eyed bean. The eye may be restricted to a very small area near the hilum, or it may extend over nearlj^ the entire bean, and in some varieties there are found detached circular spots on the dorsal or lateral portion of the bean. In most if not all such cases the pigmented area around the hilum is large. Leopard Wax is a variety of tliis sort. The pigments and different types of mottling found in totally pigmented beans may occur in any size or type of eye. In most cases the edge of the pig- mented area is not sharply defined, but in others it is clear-cut and definite. No varieties with this sharply defined edge have been used in the crosses here reported, but they have been extracted from certain of the crosses. The behavior of crosses of totally pigmented and ej^ed beans made in the course of this work is shown in Table IX. It closely resembles that of a monohybrid, but the proportions in the r2 generation are somewhat at variance with the expectation. The total number of plants in r2 is 1705, and the ratio 3.9:1. Nearly all crosses show an excess of totally pigmented beans. The progeny of heterozygous parent plants in Fs and F4, totaling 2,069, show a ratio of 3.02:1. Why this difference in the behavior in heterozygous plants occurs, it is impossible to explain at present. We can only repeat the suggestion made with reference to results shown in previous tables (page 65). All extracted eyed beans have bred true, and in all cases the beans of the Fi generation have been totally pigmented. In Table X. are shown the results of crosses of eyed and white beans. In all these crosses totally pigmented beans are produced in Fi. In the r2 generation totally pigmented, eyed and white beans are produced in the proportions shown. It is probable that these plants are of four classes and may yield all three types, totally pigmented and eyed, totally pig- mented and white, or they may be homozygous for total pigmentation. Eyed beans may be pure or may yield eyed and white. These results are in harmony with the conclusions of Emerson (5) and Tschermak (22), and indicate that total pigmentation is dependent upon two characters, — P for pigmentation and T, which spreads the pigment over the entire bean, and the absence of which, Pt, causes an eyed bean. As has been the experience of previous experimenters we have found no beans with the formula pt. However, we have used only five white seeded sorts, and only three of these at all extensively. The white beans extracted from an eyed parent in crosses 249, 268 and 327 should be of this constitution, and should yield no totally pigmented beans on crossing with an ej^ed form. Unfortunately, none of these few white seeded plants were self-fertilized or retained for seed, making it impossible to test this theory. The fact that eye sizes differ has been mentioned. While too few accurate data have been collected in the course of these experiments to make any definite report, it is evident that these eye sizes are inherited 80 MASS. EXPERIMENT STATION BULLETIN 185. Table IX. — Crosses of Eyed urith Self-colored Beans. Cross No. Parent Varieties. Fs and F* (Totally Pigmented Parents). Blue Pod Butter X Golden Eyed Wax, . Ratios Golden Eyed Wax X Blue Pod Butter, . Ratios, Golden Eyed Wax X Burpee Stringless, Ratios, Giant Stringless X Golden Eyed W'ax, . Ratios, Golden Eyed Wax X Giant Stringless. . Ratios, Golden Eyed Wax X Prolific Black Wax, Ratios, Challenge Black Wax X Golden Eyed Wax, Ratios, Currie X Golden Eyed Wax, Ratios Red Valentine X Golden Eyed Wax, Ratios, Golden Eyed Wax X Red Valentine, Ratios Keeney Rustless X Burpee Stringless, . Ratios Giant Stringless X Keeney Rustless, Red Valentine X Keeney Rustle Ratios Blue Pod Butter X Wardwell, Ratios Wardwell X Blue Pod Butter, Ratios Burpee Stringless X Wardwell, Ratios, .... Giant Stringless X Wardwell, Ratios Totally Pigmented. Eyed. 41 5.9 117 S.3 7.7 110 S.9 2.8 87 3.3 157 3.7 186 3.5 191 i.S 256 5.3 14 5 15 7.5 4 4.0 39 3.3 25 Z.8 43 «.0 Totally Pigmented. Eyed 79 200 80 2.6 125 154 3.3 22 3.1 46 81 3.5 79 80 3.0 225 130 3.0 70 192 2.6 114 36 3.0 60 i.6 59 55 6.6 54 26 2.2 13 131 2.6 123 104 2.7 53 8 6.6 120 42 S.4 SEED COAT COLOR IN GARDEN BEANS. 81 in definite proportions. Larger eye sizes show more tendency to break up than smaller ones. It is probable that the formula Pt above referred to should be taken to indicate the smallest eye size observed, and that Table X. — Crosses of Eyed with White Beans. Parent Varieties. Fa AND F4. Cross No. F2. TOTALLY PIGMENTED PARENTS. EYED PARENTS. It P 1 i 11 1' t3 1 IB i 1 141 247 249 250 268 327 Davis Wax X Keeney Rustless, . Golden Eyed Wax X Creaseback, . Golden Eyed Wax X White Marrow, White Marrow X Golden Eyed Wax, White Marrow X Keeney Rustless, . Wardwell X White Marrow, . 17 9 51 15 4 25 1 1 17 3 8 9 2 4 23 4 7 5 45 14 45 68 3 4 19 133 11 62 4 25 26 7 18 22 23 4 5 4 20 5 4 31 6 6 4 18 18 1 12 44 21 8 9 5 7 7 12 12 10 56 39 20 21 3 9 9 6 the larger eye sizes are due to the presence of other factors. If there are two additional factors for eye size they could yield four homozygous eye sizes, and there are without doubt at least that number known. There could be also four heterozj^gous forms which might exhibit other sizes. Thus the following f ormulse may express various eye sizes : — Formula. Eye Size. Found in — Ptrs PtRs PtrS PtRS Very small eye, .... Small eye, Medium eye, Large eye Maule Butter. Golden Eyed Wax. Keeney Rustless. Leopard. Of course the characters R and S could be carried by any totally pig- mented bean, but could not appear until a cross with some eyed form was made. 82 MASS. EXPERIMENT STATION BULLETIN 185. The Inheritance of Pigments. Thus far we have dealt with the inheritance of pigment patterns without reference to the particular colors involved. All the pigment patterns studied carry many different colors. So far as we have been able to see, there is no relation between the behavior of pigment patterns and the pigments themselves. We will now consider the manner in which the several pigments behave in inheritance. It is evident that there are two classes of pigments found in the varieties of colored beans used in these experiments. One class appears as some shade of red or purplish red, and is found in Red Valentine, Golden Carmine, Mohawk and similar colored varieties. This pigment is readilj^ soluble in water, as shown by laboratory tests and indicated by the readi- ness with which such seeds fade when exposed to the action of dew and rain in the field. The hght reds, such as Red Valentine, take on the purplish color when treated with alkah, and the purphsh reds of Mohawk change to a bright red in acid solutions. The former are unchanged in acid solutions and the latter in alkaline solutions. These reactions indi- Table XI. — Crosses of Blue Pod Butter with other Self-colored Varieties. Cross Parent Varieties. F2. Fs AND F4 (Va- rious Colored Parents only). No. Various Other Colors. B. Various Other Colors. B. 1 2 3 4 5 6 9 10 11 12 21 22 15 16 352 3431 347/ 349 Blue Pod Butter X Burpee Stringless, Burpee Stringless X Blue Pod Butter, Blue Pod Butter X Challeoge Black Wax, . Challenge Black Wax X Blue Pod Butter, . Blue Pod Butter X Currie Currie X Blue Pod Butter Blue Pod Butter X German Black Wax, . German Black Wax X Blue Pod Butter, . Blue Pod Butter X Giant Stringless, . Giant Stringless X Blue Pod Butter, . Blue Pod Butter X Prolific Black Wax. . Prolific Black Wax X Blue Pod Butter, Blue Pod Butter X Golden Eyed Wax, Golden Eyed Wax X Blue Pod Butter, . Brittle Wax X Blue Pod Butter, Blue Pod Butter X Low Champion, . Blue Pod Butter X Warren, 176 116 57 174 25 71 10 63 8 45 123 101 35 37 5 44 1 56 40 18 53 7 11 6 12 1 30 48 34 14 20 1 12 2 231 156 37 95 33 43 51 98 3 38 87 134 31 45 23 30 38 236 26 51 18 23 66 90 5 50 106 68 22 22 20 2 26 15 15 26 10 7 22 3 26 SEED COAT COLOR IN GARDEN BEANS. 83 cate that this pigment is anthocyan. In order to distinguish this from the other series it is called the red series. The other class of pigments encountered in this work shows itself in the various shades of yellow, coffee brown and black seen in Giant String- less, Burpee Stringless and all the Black Wax varieties. This pigment does not fade in the field, and seems only slightly soluble, or possibly in- soluble, in water, but dissolves in alcohol and alkalies. Not enough work has been done with it to determine its identity, and this series of colors is referred to in this paper as the yellow-black series. The variety Blue Pod Butter is, as previously explained, different from most other varieties in seed coat color and in other characters as well. The flower is deeper colored than any other variety and the whole plant deeply tinged wdth purple. The seed is of ecru or buff color, not seen in other self-colored varieties except Bountiful, which is similar. This buff color is of the same appearance as the ground color in all mottled beans. In Table XI. are shown the results of crosses of Blue Pod Butter with other varieties of various solid colors. In all these crosses the Fi genera- tion shows no self-colored buff beans, but all are mottled. In F2 we get a proportion of 1 buff or B bean to 3 of various other colors. In all cases the extracted buff beans have bred true to seed color, and also they carry the deeply colored flowers and purplish foliage of Blue Pod Butter. Of the beans shown in the column headed "various other colors" in Fg, one-fourth are of solid color and jdeld only solid colored beans in F3 and F4, while three-fourths are mottled and break up in F3 in the same manner as do the Fi plants. In no case has a solid colored bean yielded a buff bean like those borne by Blue Pod Butter. In Table XII. are shown crosses Table XII. — Crosses of Blue Pod Butter with Mottled Varieties. Cross F2. F3 AND F4 (Va- Hious Colored Parents only). No. Various Other Colors. B. Various Other Colors. B. 23 Blue Pod Butter X Red Valentine, . 23 7 26 15 106 7* 93 14 52 130 17 29 30 Blue Pod Butter X Warwick. . Warwick X Blue Pod Butter, 39 105 10 51 45 3 19 Blue Pod Butter X Mohawk, . 9 1 6 20 Mohawk X Blue Pod Butter. . 7 4 - 27 Blue Pod Butter X Wardwell. . 5 4 27 28 Wardwell X Blue Pod Butter, . 33 10 87 103 32 of Blue Pod Butter with mottled beans. Their behavior is similar to the crosses shown in Table XI., except that homozygous mottled beans 84 MASS. EXPERIMENT STATION BULLETIN 185. appear. These facts suggest that Blue Pod Butter lacks some factor possessed by the other varieties, and, furthermore, that it is associated with a mottling factor. We have called this factor M. We have already- adopted the explanation of the phenomenon of mottling by assuming a formula for Blue Pod Butter of PTYz, — that is, Blue Pod Butter lacks one of the mottling factors, Z, while the other varieties shown in Table XI. have this factor Z. Blue Pod Butter, then, lacks both Z and M, while all the other varieties carry these factors. We can then express the constitution of Blue Pod Butter by the formula PTYzmo, and Burpee Stringless, for example, by PTyZMO, and the evidence is that Z and M are always associated, or that we have another case of apparently perfect gametic coupling. The varieties other than Blue Pod Butter must possess additional determining factors for the various colors exhibited. These will be dealt with later. It has been said that we have two series of pigments in beans, — one bearing the red series, evidently anthocyan, and the other what we have called the yellow-black series. The crosses given in Table XL, excepting 343, 347 and 349, are of the latter nature, while these two crosses and three in Table XII. are crosses with varieties exhibiting colors of the red series. These behave like those given in the previous table so far as the relation of their colors to the B of Blue Pod Butter is concerned. If we assume that it is the factor just discussed that is the determining element for the class of pigment borne, and assume, further, that there are two of these pigment modifiers, one of which, M, brings about the formation of the yellow-black pigments, and the other, which we may call M', the formation of those of the red or anthocj^an series, we have a theory that seems to explain the facts already presented and others shown later as well. The production of a totally pigmented bean, then, rests on the presence of several factors. First, we must have P, in the absence of which we have a white bean; second, T, in the absence of which the bean has an eye; third, the presence of M or M', the former causing beans of the yellow- black series, and the latter, pigment of the red series. If neither or only one of the mottling factors Y and Z are present the bean is self-colored, while if both are present a mottled bean results. If P and T are present and M and M' absent, the bean is buff-colored, shown in Blue Pod Butter and the lighter shades in mottled beans. All colored varieties used in these experiments carry Y or Z or both; and the factor M or M' or both are, when present, always associated with the factor Z. The Behavior of the Yellow-Black Determiners. When the factors P, T and M are present, a buff or ecru colored bean is produced. The presence of certain additional factors modifies this to the various colors of the yellow-black series. These colors are black, designated by G; coffee brown, designated by F; yellow, designated by C; and a possible light brown or olive brown, designated by H. The first- SEED COAT COLOR IN GARDEN BEANS. 85 named color, G, is found in all black wax beans; the second, F, in Burpee Stringless; and the third, C, in Giant Stringless and Golden Ej^ed Wax. The color H is of a somewhat uncertain nature and our records are doubt- less somewhat confused. It is probable that more than one character has been recorded as H. There is reason to believe that additional determiners of this series may exist, but our data are too fragmentary to afford a basis for any positive assertions. In Table XIII. are shown the results of cross- Table XIII. • — Crosses of Varieties carrying Yellow-hroxon Determiners. Parent Varieties. Fi. F3 AND F4. Cross No. F2. G Parents. F Par- ents. C Par- ents. G. F. C. G. F. C. F. C. C. 190 50 81 43 44 Golden Eyed Wax (C) X Giant Stringless (C). Golden Eyed Wax (C) X Burpee Stringless (F). Challenge Black Wax (G) X Golden Eyed Wax (C). Burpee Stringless (F) X Challenge Black Wax (G). Challenge Black Wax (G) X Burpee Stringless (F). C F G G G 34 84 55 _ 24 2 14 17 all 9 16 5 22 14 51 124 180 101 3 6 17 63 3 8 44 156 21 86 57 23 7 71 ing several varieties carrj^ing yellow-brown determiners. Golden Eyed Wax X Giant Stringless jdelds only yellow beans like the parental vari- eties. In cross 50, a yellow (C) by coffee brown (F), we get apparently a simple monohybrid, the two varieties differing in that only Burpee String- less possesses the determiner F. In all crosses involving Challenge Black Wax the Fi seeds were black. In cross 81 Challenge Black Wax must carry G and F, for coffee brown beans Hke those of Burpee Stringless were extracted in F.2 and later generations. It probably carries also the yellow determiner C, for no beans lacking all three determiners appeared. In the F2 generation the proportions should be 12:3:1, assuming that F is epistatic to C and G epistatic to F. The proportions on record are 34 :2 :16. There is reason to believe that some of the plants recorded as C were really F. The progenj;- of one C plant were mostly F. Usually it is not difficult to distinguish the two colors, but in this case it is probable that some errors were made. In crosses 43 and 44 we probably have a monohj^brid, the Challenge Black Wax carrying the determiner G which is lacking in Burpee Stringless. Both carry the F and C determiners. Following the notation used, the formulae for these varieties seem to be as follows : — Golden Eyed Wax ' . . . . PtyZMm'OgfC Giant Stringless, PTyZMm'OgfC Burpee Stringless PTyZMm'OgFC Challenge Black Wax PTyZMm'OGFC 86 MASS. EXPERIMENT STATION BULLETIN 185. In Table XIV. are shown the results of crossing Burpee Stringless and Golden Eyed Wax with two other black wax varieties, — Prolific Black Wax and Currie. These crosses differ from those shown in the preceding table in that two new colors designated as H and B make their appearance in relatively small numbers. Burpee Stringless carries the yeUow-black modifier M and the deter- miners F for coffee brown, and C for yeUow. Prolific Black Wax probably carries the F and possibly C, though other crosses of this variety seem to show that it lacks C, in which case its non-appearance here may be explained by the small numbers involved. It also carries the black de- terminer G and possibly another one, H, for olive brown, though the be- havior of this color is not at all well understood. In other crosses of this table buff-colored beans (B) appear. According to our hjT^othesis this can occur only when the modifier M is absent, or, if present, only when aU determiners are absent. In these varieties M is present, therefore they must carry no determiner in common. Golden Eyed Wax carries the determiner C, and this must be absent in the vari- eties Currie and Prolific Black Wax. The absence of B beans from the F2 generation may easily be due to the small number involved. In one cross of Golden Eyed Wax with Currie, H beans appear, while in the other none are recorded. This may be due to the absence of a de- terminer for H in the strain of Currie involved. As elsewhere stated the behavior of the type recorded as H is uncertain and not well understood. The data presented in Table XIV. indicate the formula for Currie of PTyZMm'OGFc, with the possible additional determiner H, and for Prolific Black Wax, of PTyZMm'GFc and possibly the H in addition. The latter may carry also the determiner C, preventing the appearance of buff beans, but as other crosses indicate that it does not carry C, it is regarded as more probable that the absence of B beans is due to the small numbers involved. In Table XV. are shown the results of the crosses of Blue Pod Butter with Burpee Stringless (coffee brown), and with two yellow seeded sorts. All these crosses but one give black mottled beans in Fi. While none of the mottled beans breed true in later generations, as has been already explained, there have been many cases where solid black beans have bred true. The appearance of these black beans is explained on the hypothesis that Blue Pod Butter carries the black determiner G, but does not have the yellow-black modifier M, and the lack of this prevents the G determiner from acting. On crossing with a variety carrying M, the G takes effect, producing a black or black mottled bean. In cross 16a no black beans appear. It is probable that another strain of Blue Pod Butter which lacked the G determiner was used in this cross. It must have carried the determiner F, for F is always epistatic to C, and could not be carried by Golden Eyed Wax. No B beans appear in F2, owing, doubtless, to the small numbers, for they do come out in later generations as extractives from F parents, and some of them breed true. SEED COAT COLOR IN GARDEN BEANS. 87 g n , c^ «c , ^ o 1 tooo >o 1 w' , . g « 1 -H 1 1 1 H ft, o 0 2; Ji< p=; a ^ t- 1 [i< « 1 , «^ « cc o ►H '"' o 1 too ^„t^rt 1 rt t- o fe - ■« « « O «-. 00 d ^^'S^^^^ S2"';:"'§S3;:;":SSS5 aj - ' 6 , « c fe ^ "= ?5 g d fc CO o o o • . S 1 . .1 n Black \ (G). (G), c Black > olific urrie urrie rolifi (5 O O Ph ^ M • ><5 f^ £"9. 9 $i Ph less Wax Wax Wax 3" 73 13 -^ a >> >, >> w tq H a no o o 1^ ^ ^ 8 " - - ^ MASS. EXPERIMENT STATION BULLETIN 185. pS m § S ' 1 • 2 ' S ^. 1 - . o , . « d 5:S ' SS' 2^2 ffl 1 - _ « oo^o. ^ 1 1 ' 1 1 1 d . 1 < , . p.- 1 ' -22—' . . 1 , , PS Ah m \ ^ ^ ^- 1 00^ CJ , , , , , 6 ^ S-*3Sg2 'I'll « S2 K PL, O W " H ' 1 1 . «« 1 ^ , d P.- 1 2 - - , . . ^co , , CO . o ;; """S "S -s"-*"^"'^- «'2;:''2' ■^ P3 g S - 3 g ;: 2 w " ' ' ' ' d S OC.C.COOW fe' S 2 ' ' '^ ^ " "" d g S '° -^ ^ -^ 2 ' f^ o ooooo O f^ i s > Blue Pod Butter X Burpee String- less. Burpee Stringless X Blue Pod But- ter. Blue Pod Butter X Giant Stringless, Giant Stringless X Blue Pod But- ter. Giant Stringless X Blue Pod Butter, Blue Pod Butter X Golden Eyed Wax. Golden Eyed Wax X Blue Pod But- ter. Golden Eyed Wax X Blue Pod But- ter. |l c.^2J22 2 1 SEED COAT COLOR IN GARDEN BEANS. b\) Beans classified as H appear in F2 in the crosses with Burpee Stringless only, but they do appear scatteringly in later generations of most of the other crosses. Too small numbers are involved to determine its nature and relations. It is not always easy to separate the several colors F, C and H in making field observations. These colors seem to develop in the ripen- ing beans somewhat in order of their epistasis, the olive H first, and so on up to the coffee brown, and even black, provided determiners for these higher colors are present. The fact that several selfed plants recorded as H gave rise to offspring made up partially or wholly of F beans in crosses 1 and 2 raises the suspicion that these parent plants really carried the determiner F, but for some reason failed to develop their true color. Pos- sibly the weakening effect of covering the plant, which has been already discussed, may have had this effect. The yellow color G is more positively determined in the field, and the records seem clear. Extracted C beans either breed true or yield B beans in the proportions 3C:1B. According to our hypothesis there might be a 9:7 proportion in cases like this when the heterozygote is a hybrid, as Mc mC. Such a heterozygote would be yellow, and would yield 9 yellow to 7 buff. No such proportion is approached among the offspring of G parents, but in the other columns are shown a few cases that approach such a proportion. Their number is too few to be sure whether they are 9:7 or 3:1 proportions. The total numbers of such offspring in the table are 172 G, F, H and C beans to 73 buff. This is a considerable excess of buff beans, and supports the idea that some of these proportions are really 9:7. If such cases do occur the buff beans would be of three kinds, some lacking the modifier M, some the determiner and some lacking both. This raises the question whether these can be distinguished from each other. While this cannot be answered positively, we are quite sure that more than one kind of buff beans does appear. Some further evidence will be presented on this point in connection with a discussion of the relations between seed coat and flower colors. In Table IV. are shown the results of crossing self-colored varieties where mottled progeny resulted. This showed equal numbers of self- colored and mottled beans, in harmony with the hypothesis of Emerson. In Table XVI. are shown those crosses which involve Blue Pod Butter and black wax varieties, separating the self-colored beans into black and buff. These appear in approximately equal numbers and both breed true. It was early observed that buff beans generally bred true in all crosses, and comparatively few were planted. This accounts for the small numbers given in the right-hand column of the table. Our records show some half dozen plants scattered through the several crosses that were called smoky black or brown. None of them were self-fertilized, and it is impossible to say whether they represented types that appear in very small proportion, whether they were mutations, or whether they were the result of environmental conditions. We are inclined to attribute them to the last-named influence. If the constitution of Blue Pod Butter is 90 MASS. EXPERIMENT STATION BULLETIN 185. represented by the formula PTYzmG, and that of the black wax varieties by PTyZMG, either or both having possible additional hypostatic de- terminers, we have in effect a simple monohybrid based on the presence or absence of the modifier M with its accompanying mottling factor Z. This gives a proportion 3M:lm. Two of the plants carrying the modifier are heterozygous and mottled, while one is homozygous and is solid black. Inasmuch as Y and Z are confined to different gametes, according to Emerson's hypothesis, no zygote PTyzm is possible. Thus we have the theoretical proportion 1 black, 2 mottled, 1 buff, which is borne out by the facts presented in the table. Table XVI. — Crosses of Blue Pod Butter with Black Wax Varieties. Parent Vaeieties. Fi. Fa AND F4. Cross No. F2. GBO Parents. G Par- ents. B Par. ENTS. G. GBO. B. G. GBO. B. G. B. 3 4 5 6 9 10 21 22 Blue Pod Butter X Chal- lenge Black Wax. Challenge Black Wax X Blue Pod Butter. Blue Pod Butter X Currie, . Currie X Blue Pod Butter, . Blue Pod Butter X German Black Wax. German Black Wax X Blue Pod Butter. Blue Pod Butter X Prolific Black Wax. Prolific Black Wax X Blue Pod Butter. GBO GBO GBO GBO GBO GBO GBO GBO 21 64 5 23 6 15 48 31 36 110 20 33 4 47 73 70 18 53 7 13 6 12 48 34 6 11 28 11 27 45 25 29 2 64 25 52 81 29 14 1 26 15 26 46 53 71 37 134 21 23 253 68 8 43 70 The variety Bountiful has seeds that bear some resemblance to those of Blue Pod Butter. They have been recorded by the same symbol, B. The flowers are pink instead of crimson, and the plants do not show the marked purplish tinge. It has been used in crossing to a limited extent only. In Table XVII. are tabulated the results of crosses with two black wax varieties. From the results of other crosses we have assigned to the black wax varieties the black, brown and, in some cases at least, the yellow determiner. In these crosses with Bountiful all these colors appear as well as the H color, the behavior of which we do not clearly under- stand. This indicates that Bountiful does not possess any of these de- terminers. Buff-colored beans appear only in small numbers, indicating that it does not lack the modifier M. If we assign to Bountiful the formula PTyZMgfc, and to the black wax varieties the formula PTyZMGFC, the results of crossing would be in harmony with the limited data shown in Table XVII. SEED COAT COLOR IN GARDEN BEANS. 91 PQ II =D 1 i K Ph W P3 1- - ■ K , , d 1 « 1 1 1 Ph" II 1 CO i K (In «■ II .1 K 1 . 1 1 •1 d 1 o, ^ 1 1 t^' o ^-,o , i n , ^ - 1 ^ K 1 1 MOO 1 i d , t,« , fe , ,«-. 1 d J- t~t.c,«c.20g Eh m « e. K " ' - - s d -H -, 1 , 1 fe >0 M II 1 d 1 S - - - l-H H fe a o o o 1 S S 5 ffl 11 1 1 s 1 « 1 ll 1 1 1^ § i ^ i 92 MASS. EXPERIMENT STATION BULLETIN 185. The Behavior of the Determiners of the Red Series. According to the hypothesis already presented (see page 82), some varieties carry a modifier which gives rise to a series of colors different from the yellow-black series just considered. Only two members of this series have been clearly recognized in this work, — one a dark or purplish red designated by E, seen in Mohawk, and a lighter red seen in Red Valentine which we have called D. Beans of the darker shade are changed to the lighter on immersing in acid solutions, and a reversal of this is seen on treatment with a solution of potassium hydrate. The darker alkaline color seems to be dominant, and the limited data presented in Table XVIII. indicate that crosses of these determiners behave as a simple Table XVIII. — Crosses of Light Red with Dark Red Varieties. Parent Varieties. Fi. Fa AND F4. Cross No. Fo. E Parents. D Par- ents. E. D. E. D. D. 215 258 Golden Carmine X Mohawk, Red Valentine X Keeney Rustless, E 2 26 9 81 16 26 7 62 monohybrid. As no light red beans appear in cross 215, both Golden Carmine and Mohawk must carry the factor E. No signs of a buff- colored bean have appeared in cross 258, therefore it is assumed that both Red Valentine and Keeney Rustless carry the factor D, while the latter variety carries the factor for the purplish red determiner E, which is lacking in Red Valentine. The relations of Blue Pod Butter and the several varieties of the yellow- black series have already been discussed. Table XIX. shows in a similar way the relations of Blue Pod Butter and varieties of the red series. The hypothesis of the "red" modifier M' as necessary for the expression of these colors has already been advanced. Upon this hj'pothesis and that of the two determiners E and D the facts shown in the table can be fairly well explained, though a few cases are rather difficult of exj^lanation. Blue Pod Butter carries the determiner E but lacks the modifier M'. When this is supplied by crossing with Red Valentine, Low Champion or Warwick, dark red E beans appear in dominant proportions. For some reason the Fi beans in the Warwick crosses appear to have been lighter in color, and were recorded as light red, or D. In later generations undoubted dark red beans appear. Whether this is due to some environmental influence or to an unknown genetic influence cannot be stated. This has been recorded in two different years, and can hardly be an error of observation. SEED COAT COLOR IN GARDEN BEANS. 93 Table XIX. — Crosses of Blue Pod Butter with Varieties of the Red Series. F3 AND F4. ■ F2. Cross Fi. E Parents. D Par- No. ents. E. D. B. E. D. B. D. B. 23 Blue Pod Butter (B) X Red Valen- tine (D). Dark red 16 7 7 26 15 - 17 - - 343 \ 347) Blue Pod Butter (B) X Low Cham- Dark red 30 14 12 7 4 7 30 16 pion (D). 9 17 57 7 3 U 29 Blue Pod Butter (B) X Warwick (D), Light red 26 13 10 17 9 98 39 5 29 7 3 63 64 29 30 Warwick (D) X Blue Pod Butter (B), Light red 75 30 51 44 16 - 28 - 19 Blue Pod Butter (B) X Mohawk (E). Dark red 8 1 1 10 16 4 6 36 - 20 Mohawk (E) X Blue Pod Butter (B), Dark red 6 1 4 - - - 27 Blue Pod Butter (B) X Wardwell (E), Dark red 5 ~ 4 34 78 4 7 3 8 18 11 1 28 Wardwell (E) X Blue Pod Butter (B), Dark red 25 8 10 28 46 49 43 13 16 21 11 39 The Interrelations of the Yellow-black and Red Series. All the varieties showing pigments of the red series are mottled beans with the exception of Warren, and Warren has not been crossed with varieties of the yellow-black series. Therefore all crosses between red and yellow-black varieties shown in Table XX. are mottled in the first generation. Owing to this fact the colors of both series may usually be seen on examination of the Fi beans. It is possible to separate the beans of the F2 generation into three classes, as shown in the table. The yellow- brown beans are partly self-colored and partly mottled, showing only yellow-brown or black, as the case may be. A larger number are mottled, showing these colors and also light or dark red, or both. A third class shows only red, and these are always mottled. No solid red bean of any shade of color has ever appeared from the crosses shown in Table XX. All plants listed in the yellow-black column breed true to these colors, and the same is true of those belonging to the class of red beans. Those in the middle column break up exactly like the Fi generation. These facts are shown in the columns under F3 and F4. In crosses 198, 119, 191, 194, 115 and 52, buff beans appear in small numbers in F3 and F4, but none have been observed in the F2 generation. In the other crosses more have been observed. If the parent varieties possess a determiner in common the chances of a buff bean appearing would be small, and this may explain their absence. Probably if the 94 MASS. EXPERIMENT STATION BULLETIN 185. numbers involved were larger they would appear in many crosses where they are not shown. According to the hypotheses already advanced, these crosses involve varieties whose constitution may be expressed by PYZmM' X PyZMm', each variety possessing one or more determiners in addition. The mottled beans of the yellow-black series, appearing from these crosses, are the heterozygotes lacking the determiners E and D. No such beans have bred true. Table XX. — Crosses of Varieties of the Yellow-black with the Red Series. F2. Fs AND r4. Cross PARENliVARIETIES. y-b-i-r Parents. y-b Par- rPAR- No. ents. ENTS. y-b. y-b-fr. r. y-b. y-b-l-r. r. y-b. r. . 240 Golden Eyed Wax (y-b) X Red Valentine (r). - - - 12 16 80 - 239 Red Valentine (r) X Golden Eyed Wax (y-b). 12 36 15 8 6 - - 198 Red Valentine (r) X Giant Stringless (y-b). 5 20 10 18 4 41 - 57 Burpee Stringless (y-b) X Red Valentine (r). - - ~ 20 7 25 " 58 Red Valentine (r) X Burpee Stringless (y-b). 20 55 8 25 12 13 16 29 102 72 288 Red Valentine (r) X Prolific Black Wax (y-b). 10 15 5 10 10 202 15 119 Currie (y-b) X Red Valen- tine (r). 21 36 17 11 3 11 - 76 95 Challenge Black Wax (y-b) X Warwick (r). Giant Stringless (y-b) X Ward- 13 14 2 3 2 - - 201 20 33 11 23 32 « 117 26 well (r). 11 7 14 13 6 191 Giant Stringless (y-b) X Keeney Rustless (r). 1 3 1 9 18 9 36 5 193 Giant Stringless (y-b) X Mo- hawk (r). Mohawk (r) X Giant String- 2 6 6 17 1 - - 194 2 5 8 27 49 13 _ 37 less (y-b). 115 Currie (y-b) X Mohawk (r). . 49 90 32 11 16 15 10 6 6 85 15 10 116 Mohawk (r) X Currie (y-b), . 8 19 9 2 11 2 24 8 52 Keeney Rustless (r) X Burpee Stringless (y-b). 3 6 2 7 11 3 30 25 A detailed study of the records of the progeny of crosses like those shown in Table XX., giving consideration to the manifestation of the various pigments, leads to conclusions already advanced in the discussion of the crosses belonging within each series (page 84). Some five or six varieties of red mottled beans have been crossed with a similar number belonging to the yellow-black series. The results do not lend themselves readily to tabular presentation, therefore they are dealt with in a text discussion. These facts are in addition to those shown in Table XX. Red Valentine crossed with Golden Eyed Wax yields buff beans in SEED COAT COLOR IN GARDEN BEANS. 95 small numbers, indicating that these parents possess no determiners in common. One plant \vith red mottled beans yielded in the next genera- tion red mottled and buff beans in the proportion of 3:1, indicating that the parent plant was heterozygous for the factors M and D. Red Valen- tine X Giant Stringless gives results of the same nature, and they indi- cate the same constitution as that of Golden Eyed Wax. In one cross of these two varieties, dark red and even black beans appeared. This is so contrary to the usual experience that it is thought they are due to acci- dental crossing in the field, or some other accident of similar nature. In crosses of Red Valentine with Burpee Stringless we have coffee brown, yellow and light red mottled beans, as would be expected from the formulae already advanced. Buff beans also appear in small numbers, indicating that these two varieties have no determiner in common. Dark red mottled beans appear in numbers greater than those of light red mottled beans, and so distributed as to make it doubtful if they are the result of accident. Their presence can be explained on the supposition that Burpee Stringless carries the determiner E. Small numbers of olive- brown, or H, beans appear as in other similar crosses. The constitution indicated for Burpee Stringless is PTyZMm'FCEd, which is in harmony with the one previously advanced. Dark red mottled beans have been extracted from crosses of Red Valentine with Prolific Black Wax, indicating that Prolific Black Wax carries the alkaline determiner E. This type, self-fertilized, yields dark red mottled and light red mottled beans in the proportion 25:12, probably a simple 3:1 ratio. Buff beans also appear in small numbers, indicating that these two sorts have no determiner in common. Coffee brown, or F, beans appear in considerable numbers, and when selfed sometimes breed true, or may yield yellow (C), buff (B) and olive-brown (H) beans in proportions subordinate to the coffee brown. In this as in other crosses involving Red Valentine, the parent type, light red mottled, always breeds true when extracted. Warwick has a coat color apparently very similar to or identical with Red Valentine. The blossom color is light pink, while the usual strains of Red Valentine are white. This indicates a different pigmentation for the two varieties, which may or may not affect the color of the seed coat. When crossed with Challenge Black Wax, Warwick gives in the Fi genera- tion a mottled bean showing black and red similar to those where Red Valentine is involved. In later generations there is a greater complexity among the mottled beans. Coffee-brown and yellow beans are extracted, also the buff, or B beans, all in rather small nimabers. These solid-colored beans all breed true or yield other hypostatic or recessive colors in com- paratively simple proportions. Among the mottled beans various shades of black, violet, brown, red and yellow may be seen, and in addition the buff color always showing in mottled beans. Beans of these complex colors segregate into self-colored beans or mottled beans of less complex natures. We have observed no case where a mottled bean showing colors 96 MASS. EXPERIMENT STATION BULLETIN 185. of both the red and yellow-black series has bred true. From crosses similar to the one just discussed we have extracted black mottled beans similar to Refugee that have bred true, though not in large numbers. Mohawk has a seed coat color somewhat similar to Red Valentine and Warwick, but the red color is darker and is changed to a bright red by- acid solutions. It is assumed to carry the alkaline modifier E. When crossed with Giant Stringless it yields in F2 numerous plants with coffee- brown beans, indicating that Mohawk carries the determiner F. When crossed with Burpee Stringless no yellow beans appear, for both these varieties carry F, and the hypostatic yellow color cannot appear. Keeney Rustless crossed with Burpee Stringless yields many black beans. This may be explained by assuming that Keeney Rustless carries the black determiner G but not the modifier M, which prevents the appearance of the black color. It does carry M' and E, and is therefore a dark red bean. Burpee Stringless supplies the modifier M which with the determiner G brings forth the black color. The cross Keeney Rustless X Burpee String- less may be expressed by PmM'GfcED X PMm'gFC. It is probable that Burpee Stringless carries an E also. Buff-colored beans appear in this cross, indicating a lack of common determiners. Wardwell crossed with Giant Stringless and Burpee Stringless yields progenies similar to those resulting from a cross of the latter two varieties with Mohawk so far as pigments are concerned. Both Mohawk and Wardwell carry the determiner F, but it is not expressed owing to the lack of the modifier M. When this is supplied by Giant Stringless or Burpee Stringless coffee-brown flecks appear in the mottled beans, and various types of mottled beans and both mottled and self-colored beans of the yellow-black series may be isolated. Crosses involving Creasebach. In Table VI. were presented the manifestation of color patterns in crosses of Creaseback with Blue Pod Butter and Challenge Black Wax. In Table XXI. are shown the same crosses, giving the proportion of plants exhibiting the various seed coat pigments involved. In the discussion of Table VI. (page 74) it was brought out that Creaseback must carry the determiner G, and its formula according to the hypotheses followed is pyZMG. As soon as the factor for pigment is introduced by Blue Pod Butter, which may be assumed to have here the formula PYzmG, black beans appear making up all the Fi generation, and in F2 there follows what is probably a 9:3:4 proportion with the buff of Blue Pod Butter and white. The exact proportion is 9.21:2.55:4.31 when all lots showing the three colors are combined. Wliere black seed parent plants show only buff or white progeny besides black, and where buff seed parent plants yield white seeded progeny, there is evidently a simple 3:1 proportion. In cross 97, Challenge Black Wax X Creaseback, there is evidently a simple 3 : 1 proportion based on the presence or absence of the factor for pigmentation. Cross 97 as tabulated is derived in part from a cross made SEED COAT COLOR IN GARDEN BEANS. 97 in 1909 and in part from a cross made in 1911, which exhibited similar behavior. In the 1911 cross there were four Fi plants, two of which gave the progeny just referred to, and the other two gave the progeny shown in cross 97a. "^Vhy these show such a different proportion we do not know, for 97a must have been a successful cross, as proved by the appearance of pole beans in normal proportions. It may be that the pollen grains were not of the same constitution, or possibly stray pollen grains carry- ing only black were involved in the Fi generation. The facts are here presented in the hope that they may be suggestive to some other investigator. Table XXI. — Crosses involving Creaseback. F,.. F3 AND F4. Cross Parent Varieties. J^'i. G Parents. B Par- No. ents. G. B. A. G. B. A. B. 61 A. 31 Blue Pod Butter X Creaseback, G Ill 33 65 155 46 67 21 113 33 131 34 119 32 Creaseback X Blue Pod Butter, G 55 13 33 66 75 103 65 14 31 16 22 97 Challenge Black Wax X Creaseback, G 295 - 79 136 - 75 - - 97a Challenge Black Wax X Creaseback. G 101 1 29 60 2 " Crosses of other varieties with Creaseback are not shown in the table because the results were complicated and somewhat uncertain. With Golden Eyed Wax the Fi generation gave only black beans, and in F2, 10 black to 3 white. In F3 and F4 there appeared also coffee-brown (F) and yellow (C) seeded plants in moderate numbers. One coffee-brown plant bred true in the 9 progeny grown. When crossed with Warwick the results were complicated beyond hope of comprehension. In the cross Creaseback X Warwick the Fi generation is recorded as black with faint signs of mottling, while in the reciprocal, which may have involved a different strain of the parent varieties, the Fi beans were distinctly mottled, showing many distinct shades of pig- ments. Apparently about all the pigments of both the red and yellow- black series were involved. The behavior of the pigments in these reciprocal crosses does afford some further evidence bearing on the hypothesis of a factor discussed on page 75 and there called X. One strain of Warwick X Creaseback gives the expected number of white beans, the ratio being 24 self-colored, 5 mottled and 9 white. Another strain yields no white beans but gives 30 self-colored and 7 mottled in both cases, approximately four times as 98 MASS. EXPERIMENT STATION BULLETIN 185. many self-colored as mottled. If there is a factor X in Creaseback which inhibits the expression of mottling as previously suggested, the following gametes should be formed: PYZX, pYZX, PyZX, pyZX, PYZx, pYZx, PyZx, pyZx. The zygotes formed would yield 9 mottled without X, 27 with X; 12 self-colored and 16 white. The 27 "mottled" beans with X do not show mottling, making a to^^al of 39 self-colored, 9 mottled and 16 white, or nearly four times as many self-colored as mottled. Of the mottled beans 6 should show colors of both series, and 3 those of the red series only, which are the actual numbers shown in the F2 generation of this cross. Crosses involving Davis Wax, As has already been shown, Davis Wax, a non-pigmented bean, carries factors for light mottling which appear as soon as pigment is supplied. When crossed with Blue Pod Butter the Fi generation is light mottled, like beans of the Horticultural group. In F2 there are produced light mot- tled, buff and white beans in the proportion, presumably, of 9:3:4. In later generations these behave as shown in Table XXII. It is possible Table XXII. — Crosses involving Davis Wax. Cross Parent Varieties. Fi. F2. F3 AND F4 (BEP Parents). BEP. B. A. BEP. B. A. 7 8 Blue Pod Butter X Davis Wax, . Davis Wax X Blue Pod Butter, . BEP BEP 14 38 3 16 22 50 17 49 8 53 0 10 3 18 12 16 to derive from this cross light mottled races that breed true as well as the parent types, as is shown in the table. No black beans appear, as the modifier M is not present. Among these light mottled progeny there appear some plants that produce what seem to be bud sports, in which the darker reddish color predominates over the surface of the bean. These may appear as single pods or as branches bearing several pods, and rarely a portion only of the beans in a single pod is affected. If these dark mottled beans are planted they breed true to seed coat color, while the plants with light mottled seed may breed true in this character, or may give rise to plants bearing bud sports as before. Limited observations suggest that these sporting plants exist in definite proportions. The fact that such plants have appeared so often in the breeding work here reported, and that dark mottled beans are frequently seen in seed of varieties of the Horticultural type offered for sale, suggests that this peculiarity of bud sporting is a SEED COAT COLOR IN GARDEN BEANS. 99 frequent and possibly a constant character of beans of this class. At any rate, we have here a peculiarity which would doubtless yield interest- ing results on further and more specific investigation. Reciprocal crosses of Challenge Black Wax and Davis Wax yielded complicated progenies. Dark mottled beans appear because the former variety carries the factor O for dark mottling, which acts with YZ from Da\ns Wax to bring about this result. Challenge Black Wax carries the modifier M, and Davis Wax brings in M', so that we get beans of both the yellow-black and red series. Owing to the complicated nature of the progeny of this cross it is not shown in tabular form. Crosses involving White Marrow. The only other white variety that has been used at all extensively is White Marrow. The color pattern factors are rather complex, and the pigment factors much more so. Owing to this the crosses of White Marrow with the several varieties used will be taken up one by one. Apparently White Marrow carries several pigment modifiers and determiners in a latent condition, owing to the absence of the pigmentation factor P. When it is crossed with another variety carrying P, and perhaps several additional modifiers and determiners, we have very many classes of beans which are extremely difficult to segregate. Crosses of White Marrow and Blue Pod Butter. — Three crosses of these varieties have been made, including reciprocals. As previously indicated (page 73), the Fi beans have light red (D) stripes and splashes on the usual buff (B) ground color. In the next generation these split up, show- ing, in addition to the two colors mentioned and the parent forms, con- siderable numbers of coffee-brown (F) and yellow (C) beans. No black beans have appeared in this cross, a fact that may be explained on the hypothesis that the particular strain of Blue Pod Butter used lacked the factor G. We have been led to conclude that Blue Pod Butter lacked both modifiers M and M'. The appearance of both series of colors in the progeny of this cross leads to the conclusion that White Marrow carries both modifiers in an inactive state, owing to the lack of the factor P, When both are present the M' is epistatic to M, and the beans are classified as of the red series. Beans showing the dark red color have jaelded in some cases only the parent color (E), and in other cases various combinations of dark red (E), light red (D), yellow (C), buff (B) and white, but we have no record of coffee-brown beans (F) from this parentage, though they do appear in small numbers from light red (D) parents. Yellow (C) parent plants yield progeny of similar color, and, in addition, buff (B) or white or both in subordinate numbers. In a few cases our records show fight red (D) beans in small numbers, which occurrences are difficult to explain. They are rather too frequent to be mere accidents. Further investigation should lead to interesting results. 100 MASS. EXPERIMENT STATION BULLETIN 185. Crosses of White Marrow with Golden Eyed Wax. — The progeny of this cross are less complicated than others having White Marrow in the parentage. The fii'st generation beans, being mottled, show both yellow and red splashes. Those of the F2 generation, showing only yellow either in solid color or motthng, either breed true or yield white beans in the expected ratio. Among some three hundred plants the records show two buff (B) seeded plants. These are probably accidental strangers, yet they may be a definite class occurring in small numbers; if so, no explanation of their occurrence can be presented. Crosses of White Marrow with Burpee Stringless. — Other crosses have shown that Burpee Stringless has a constitution similar to Grolden Eyed Wax, with the addition of the determiner F, making the bean coffee brown. The beans of the Fj generation were of a yellow-olive mottled color. In the next generation a variety of colors appeared among the mottled beans, — coffee brown, yellow, olive, chocolate brown and red. In later generations these differentiated clearly into the coffee brown of Burpee Stringless, yellow (C), light red (D), buff (B) and white. Self- colored coffee-brown seeds have given all brown, brown and yeUow, brown and white, and mixed progeny including all three tj'pes. Light red, light mottled seeds have bred true, and have yielded white seeded plants in the usual proportion of 3:L Crosses of White Marrow with German Black Wax. — The results of this cross are similar to the previous one with the addition of the epistatic black (G). There is the same confusion of colors in the Fj generation, but on further segregation they separate into black, coffee brown, yellow and white. The light red also appears and apparently dark red (E) also, though in small numbers. We have no case where a parent plant of this color has been bred. One yellow seeded plant, being selfed, yielded yellow and white in a 3:1 pro- portion, and one solid black of the F2 generation yielded a mixture of black and coffee brown. Crosses of White Marrow with Red Valentine. — This cross differs from those just considered in that Red Valentine belongs to the red series. There are red, black or brown beans appearing, but yellow does appear in many of the mottled beans. One plant of mostly sohd yellow beans pro- duced a progeny of yellow and light red mottled beans, the former in larger numbers. There is a tendency to produce the dark mottled bud sports referred to on page 98. There are other complications in this cross, some of which can be explained only on the supposition that the White Marrow plant used as a parent was heterozygous in its nature. This might well be, for so long as the factor P is absent the pigment modifiers and determiners might be interchanged without the external appearance being changed. SEED COAT COLOR IN GARDEN BEANS. 101 The Genetic Constitution of the Varieties used. In the following table is given the genetic constitution as indicated by the investigations here reported. It is not asserted that these are correct in ail cases, even should the general hypotheses here presented prove sound. Moreover, there are doubtless in a given variety different strains of indistinguishable external appearances, especially among the non- pigmented varieties. Blue Pod Butter, Bountiful, . Burpee Stringless, Challenge Black Wax, Creaseback, Currie, Davis Wax, German Black Wax, Giant Stringless, . Golden Carmine, . Golden Eyed Wax, Keeney Rustless, . Longfellow, Low Champion, . Mohawk, Prolific Black Wax. Red Valentine, Ward well, Warren, Warwick, White Marrow, PTYzmm'oGfcHEd PTYzmm'oGFcHEd PTYzmm'ogFcHED PtYZmM'OgFCHED PTyZMm'OgFChEd PTyZMm'OGFChED pTYZMm'OXG pTyZMm'oXGFCHED PTyZMm'OGFcH PTyZMm'OGfCHED PTyZMm'OGfCHeD pTYZmM'oged PTyZMm'OGFCHED PTyZMm'OgfChEd PTYZmM'ogfchEd PtyZMm'OgfChed PtYZmM'OGfcHED PTYZmM'OgfcheD PTyZmM'OeD PTYZmM'OgFcHED PTyZMm'OGFcHE PTYZmM'OgfcheD PTYZmM'OgfcHeD PTYZmM'OgfCheD PtYZmM'OgFcHED PTYzmM'ED PTYZmM'OgfcHeD pTYZMM'ogfCheD pTyZMM'ogfChedpTYz pTyZMM'ogfCheD The significance of the letters is as follows : — P is the factor for pigmentation, without which the bean is white. Pre- sumably this factor is the one causing the production of the basic chro- mogen. T is the factor for totality of pigmentation, without which the bean is an eyed bean if P is present. Y and Z are the factors for mottling, which are coupled in mottled vari- eties but may exist separately in non-mottled varieties, and if brought together in crossing give mottled beans which break up in later generations. M and M' are the two modifiers, M giving rise to the beans of the yeUow- black series and M' to those of the red series. They doubtless represent 102 MASS. EXPERIMENT STATION BULLETIN 185. one of the enzjTues that are beheved to be necessary for the production of sap colors in plants. O is the factor for dark mottling in mottled beans, in the absence of which we have the light mottled type of the Horticultural class, provided P, Y and Z are all present. X represents a blackening factor found only in Creaseback. The remaining letters of the formulae are the determiners which in the presence of other necessary factors determine the color of the seed coat. The significance of the colors is as follows: G, black; F, coffee brown; C, yellow; E, dark red; D, light red (see page 84). Summary. It is evident from these and other investigations that the inheritance of seed coat color in beans is very comphcated, and difficult to explain fully and satisfactorily. The problems involved are interesting, and the plants convenient to handle for purposes of investigation. They provide excel- lent material for the fruitful investigation of Mendelian inheritance. In this work 21 varieties have been used in making over 120 different crosses, involving more than 40,000 plants. The work continued over a period of eight years. There are certain correlations in the pigmentation of the plant. All white or eyed beans are accompanied by white flowers; all black or black mottled beans by dark pink flowers. Mottled beans, other than black mottled beans and those of various yellow and brown colors, are usually accompanied by light pink flowers. In a general way the crosses of pigmented and white beans show a 3:1 ratio, but there are some rather wide departures which may or may not be of genetic significance. The inheritance of mottling may be explained by the double factor hypothesis of Emerson and Spillman. Crosses of two mottled varieties have in all cases given only mottled progeny. Crosses of mottled and self-colored varieties have yielded mottled beans in Fi, and the parent types in a 3:1 ratio in Fj. Crosses of mottled and white varieties have given mottled beans in Fi, and usually mottled, self-colored and white in a 9 :3 :4 proportion in F2. In most cases crosses of two self-colored varieties have given only self- colored progeny. The principal exceptional variety is Blue Pod Butter, which, when crossed with most self-colored varieties, yields mottled progepy none of which breed true to the mottled character. White vari- eties may carry the character for mottUng, which can show itself only after crossing with a pigmented sort. Creaseback is peculiar in that it seems to carry factors for mottling and an additional factor causing a blackening which nearly or quite obscures the mottled pattern. There are two types of mottling, — the dark, seen in Red Valentine and Refugee and many others, and the light, seen in varieties of the Horticul- tural class. The former behaves towards the latter as a simple dominant. SEED COAT COLOR IN GARDEN BEANS. 103 Apparently the factor for the dark motthng is associated with one of the motthng factors. Wliite beans may jdeld hght mottled beans, but none have yielded dark mottled beans. There is evidently needed to produce a totally pigmented bean a factor for total pigmentation. If it is absent when the factor for pigmentation is present we have an eyed bean. Eye size is evidently governed by one or more factors, but these investigations do not afford definite data regard- ing their relations. Pigment patterns and pigment colors are controlled by distinct factors. According to the hypothesis presented in this paper, any color shown in a bean seed is, in most cases, dependent on tlu'ee or more factors. The basic factor for pigmentation may be modified into either one of two series, — one including the various yellows, browns and black; and the other, different shades of red. The third factor, called a determiner, finally determines what the color is to be. In some cases the determiners bring about the color through causing an alkaline or acid condition. Possibly in some cases the color is determined by the degree of acidity or alkalinity. The two modifiers discovered are apparently associated with one of the mottling factors, but the determiners are free and independent, though standing often in an epistatic or hypostatic relation to one another. Bibliography. 1. Bateson, W. 1902. Mendel's Principles of Heredity, p. 78. Cambridge. 2. Emerson, R. A. 1902. Preliminary Account of Variation in Bean Hybrids. In Nebr. Agr. Exp. Sta. 15th Ann. Rpt., pp. 30-43. 3. . 1904. Heredity in Bean Hybrids. In Nebr. Agr. Exp. Sta. 17th Ann. Rpt., pp. 33-68. 4. . 1909. Factors for Motthng in Beans. In Ann. Rpt. Amer. Breeders' Assoc, Vol. 5, pp. 368-376. 5. . 1909. Inheritance of Color in the Seeds of the Common Bean. In Nebr. Agr. Exp. Sta. 22d Ann. Rpt., pp. 67-101. 6. . 1914. The Inheritance of a Recurring Somatic Variation in Variegated Ears of Maize. Nebr. Agr. Exp. Sta. Research Bui. 4. 7. Freeman, G. F. 1912. Southwestern Beans and Teparies. Ariz. Agr. Exp. Sta. Bui. 68. 8. Halsted, B. D. 1905. Notes upon Bean Crosses. In N. J. Agr. Exp. Sta. 26th Ann. Rpt., pp. 478-480. 9. . 1906. Experiments with Bush Beans. In N. J. Agr. Exp. Sta. 27th Ann. Rpt., pp. 454-466. 10. . 1907. Experiments with Bush Beans. In N. J. Agr. Exp. Sta. 28th Ann. Rpt., pp. 340-343. 11. Jarvis, C. D. 1908. American Varieties of Beans. N. Y. Cornell Agr. Exp. Sta. Bui. 260. 12. JoHANNSEN, W. 1908. Uber Knospenmutation bei Phaseolus. In Zeitschrift fiir Induktive Abstammung und Vererbungslehre, Band 1, p. 1. 13. Kajanus, Birger. 1914. Zur Genetik der Samen von Phaseolus vulgaris. In Zeitschrift fiir Pflanzenzuchtung, Band 2, p. 378. 14. Mann, Albert. 1914. Coloration of the Seed Coat of Cow Peas. In Jour. Agr. Research, Vol. 2, pp. 33-56. 15. Shaw, J. K. 1913. The Inheritance of Blossom Color in Beans. In Mass. Agr. Exp. Sta. 25th Ann. Rpt., pp. 182-203. 104 MASS. EXPERIMENT STATION BULLETIN 185. 16. Shaw, J. K. 1911. A System of recording Mendelian Observations. In Amer. Nat. Vol. 45, p. 701. 17. Shull, G. H. 1907. The Significance of Latent Characters. In Science, Vol. 25, pp. 792-794. 18. . 1908. A New Mendelian Ratio and Several Types of Latency. In Amer. Nat. Vol. 42, p. 433. 19. Tracy, W. W., Jr. 1907. American Varieties of Garden Beans. U. S. Dept. Agr. Bur. Plant Indus. Bui. 109. 20. TscHERMAK, E. VON. 1901. Weitsre Beitrage iiber Verschiedenwerthigkeit der Merkmale bei Kreuzung von Erbsen und Bohnen. In Zeitschrift fiir das land- wirthschaftliche Versuchswesen in Oesterreich. Band 4, pp. 641-731. 21. . 1904. Weitere Kreuzungsstudien an Erbsen, Levkojen und Bohnen. In Zeitschrift fiir das landwirthschaftliche Versuchswesen in Oesterreich. Band 7, pp. 533-638. 22. . 1912. Bastardierungsversuche an Levkojen, Erbsen und Bohnen mit Riicksicht auf die Faktorenlehre. In Zeitschrift fiir Induktive Abstammung und Vererbungslehre, Band 7, pp. 81-234. BULLETIN No. 186 NOVEMBER, 1918 MASSACHISETTS AGRICILTIRAL EXPERIMENT STATION The Composition, Digestibility and Feeding Value of Alfalfa II The Value of Corn Bran for Milk Production By J. B. LINDSEY and C. L. BEALS Part I of this bulletin contains a summary of all analyses and digestion experiments made with alfalfa and red clover, including those made at this station. It contains also the results of three feeding experiments with milch cows, intending to throw light upon the value of alfalfa as an efficient source of milk protein, its effect upon the action of the kidneys in causing a milk shrinkage, and its value for milk production when fed as the entire source of roughage together with corn meal. Experiments IV and V compare alfalfa with rowen, and VI and VII were conducted to ascertain how best to combine alfalfa with other feedstufis in making up the dairy ration. A summary of all results will be found on pages 105-107. Part II of the bulletin describes two feeding experiments with corn bran as a component of the dairy ration. Conclusions and suggestions will be found on page 142. Requests for bulletins should be addressed to the Agricultural Experiment Station Amherst, Mass. CONTENTS Paet I. PAGB Summary and suggestions, ......... 105 Introduction, ........... 107 The chemical composition of alfalfa and red clover, ..... 107 The digestibility of alfalfa hay, 109 Feeding experiments with alfalfa, . . .111 Experiments I, II and III: Alfalfa, beet pulp and corn meal v. hay, beet pulp and corn gluten products, . . . .111 Experiments IV and V: Alfalfa v. rowen, . .125 Experiment VI : Alfalfa, English hay and grain v. English hay and grain, . 132 Experiment VII: Alfalfa, corn stover, corn-and-cob meal and bran v. English hay, corn-and-cob meal, gluten feed and bran, . . . 137 Part II. Summary and suggestions, . . . . . . . . . 142 The experiments in detail, ......... 142 Publication of this Document approved by the Supervisor of Administration. BULLETI]^ ]^o. 186. DEPARTMENT OF CHEMISTRY. Part I [ON, DIGJ FEEDING VALUE OF ALFALFA. THE COMPOSITION, DIGESTIBILITY AND BY J. B. LINDSEY AND C. L. BEALS. SUMMARY AND SUGGESTIONS. 1. Green alfalfa contains from 70 to 80 per cent, of water, 2 to 2.5 per cent, of ash, 2.9 to 4.7 per cent, of protein, 4.2 to 12.8 per cent, of fiber, 7.98 to 11.3 per cent, of extract or starchy matter, and not over 1 per cent, of fatty matter. 2. Alfalfa hay of good quality should average about 14 per cent, of water, and on this basis will contain some 7 to 9 per cent, of ash, 13 to 14.5 per cent, of protein,^ 27 to 33 per cent, of fiber, 33 to 36 per cent, of starchy matter and 1.5 to 2 per cent, of fat. The earlier it is cut the less fiber and the more ash and protein it will contain, 3. Alfalfa resembles red clover quite closely in chemical composition, although it is likely to be slightly lower in protein and starchy matter. Both alfalfa and clover contain considerably more protein and less fiber and extract matter than do the cereals and grasses. 4. A complete chemical study of the different food groups composing the alfalfa has not been made. In early blossom an average of 71.1 per cent, of itF total nitrogen has been found to exist as true protem, and 28.9 per cent, as non-albuminoid nitrogen. One sample has shown 10.17 per cent, in the form of amino acids, and fully 88 per cent, as true protein. In the carbohydrate group from 3.9 to 16.8 per cent, of pentosans, and as high as 4.71 per cent, of galactan, have been found. 5. Alfalfa, red clover and timothy hay contain about the same amount of digestible organic nutrients in 1 ton (950 to 970 pounds); while rowen averages 1,028 pounds, or 8 per cent, more; and gluten feed, 1,556 pounds, or 64 per cent. more. 1 Cut before bloom, alfalfa may contain 20 per cent, protein. 106 MASS. EXPERIMENT STATION BULLETIN 186. 6. Comparing these several feeds, however, on the basis of net energy- values, as suggested by Armsby, one finds red clover to have 13 per cent, more energy value, timothy hay and rowen 20 per cent, more, and gluten feed 160 per cent. more. This lessened energy value of the alfalfa has been shown to be due to its causing an increased metabolism in the animal organism. 7. In case of an average of three experiments (I, II and III) with cows, the dry matter in a ration composed of alfalfa, beet pulp and corn meal produced substantially as large a yield of milk and milk ingredients as did a like amount of dry matter in one composed of first-cut mixed hay, beet pulp and corn gluten products. The alfalfa seemed to act as a slight stimulus to production. In these experiments alfalfa and hay each fur- nished about 71 per cent, of the total dry food of the rations. 8. The animals showed a total gain in live weight of 13 pounds on the alfalfa ration, and 481 pounds on the hay ration, indicating that the less energy value of the alfalfa might have been responsible for this difference. 9. The protein contained in the alfalfa, beet pulp and corn meal ration, of which 78.2 per cent, was from alfalfa, seemed to be fully as effective in the formation of normal milk as did the protein contained in the hay, beet pulp and corn gluten ration. 10. The diuretic effect of the alfalfa appeared to be without influence in lessening the yield of milk and milk ingredients. 11. In case of the average of two experiments (IV and V), alfalfa proved slightly superior to rowen in the volume of milk produced. The difference, however (4.2 per cent, on the basis of equal amounts of dry matter in the two rations), was not sufficient to warrant any marked claim of superior- ity. This slight stimulating effect may be due to the superiority of the protein contained in the alfalfa. 12. The fat percentage in the milk produced on the alfalfa ration did not keep pace with the increased milk yield, for a like amount of dry matter in the alfalfa and rowen rations produced a like amount of milk fat. 13. The herd made a total gain in live weight of 16 pounds on the alfalfa ration, and lost a total of 24 pounds on the rowen ration, differ- ences not sufficient to warrant any particular conclusion, 14. A good quality of rowen appears to be nearly as satisfactory a source of roughage for milk production as a like amount -^f a similar quality of alfalfa. 15. One experiment (VI) showed that a ration composed of one-half first-cut hay and one-half alfalfa, together with a little wheat bran and corn-and-cob meal, gave as satisfactory results as one consisting of first- cut hay, wheat bran, corn-and-cob meal and gluten feed. The former ration contained substantially home-grown products, and would render it unnecessary to purchase grain, the alfalfa furnishing the necessary extra protein required, and the corn-and-cob meal the necessary extra digesti- ble matter. 16. One experiment (VII) indicated that reasonably good results can FEEDING VALUE OF ALFALFA. 107 be secured from a roughage ration composed of two-thirds alfalfa and one- third corn stover, together with a grain ration of corn-and-cob meal. If the stover is well cured and kept under cover it wiU give more satisfactory results than if left in the open during the winter. The yield of milk, how- ever, on such a ration would not be quite equal to the yield on one com- posed of first-cut hay and a grain mixture of equal parts of wheat bran, corn-and-cob meal and gluten feed. 17. Too high an estimate should not be put upon the alfalfa, for while studies at this station and elsewhere have shown it to contain more pro- tein than most other sources of roughage, and to equal wheat bran in feeding value, it is quite inferior as a source of energy or fat production to most of the concentrates. 18. In the light of our present knowledge it is preferable, particularly in the eastern states, not to use alfalfa as the entire source of roughage for milk production, but to feed one-half alfalfa and one-half hay, or two- thirds alfalfa and one-third corn stover, or 10 to 15 pounds of alfalfa and 1 bushel of silage daily. Such combinations, together with a grain ration of 70 to 80 per cent, corn-and-cob meal, and 20 to 30 per cent, wheat bran or oats or barley, ought to give quite satisfactory results. INTRODUCTION. In the year 1914 this station published Bulletin No. 154, entitled "Al- falfa," which related primarily to the growing of the crop in Massachu- setts, based upon the results of home and co-operative experiments. It included specific directions for the general management of the crop. The present bulletin summarizes the analyses and digestion trials made with alfalfa, both at this station and elsewhere, and presents the results of seven feeding experiments relative to its effect on milk production and its place in the dairy ration. Alfalfa belongs to the same family of plants as the clover, pea and bean. The family name is Leguminosce, and these plants are usually spoken of as legumes. It has been cultivated both in Asia and Europe for a long time, being known in Germany and France under the name of Luzerne. It has been grown with great success in California and in the hot semi- arid regions of the southwestern portions of our country. Of late years it has been c .iltivated with success in the northwestern States, and more recently it has been grown with considerable success in different portions of the Middle Atlantic and New England States. It is an especially deep- rooted perermial, and needs, among other things, a well-drained soil hav- ing a water table several feet below the surface, and an abundance of lime. THE CHEMICAL COMPOSITION OF ALFALFA AND RED CLOVER. The composition of these plants will vary more or less, depending upon the stage of growth at which they are cut, and whether the material is derived from the first, second or third cutting. The analysis of medium 108 MASS. EXPERIMENT STATION BULLETIN 186. red clover is used for comparison. In order to make the analyses com- parable, they have been brought (in case of the green samples) to sub- stantially a like water basis. In case of the hays, a uniform moisture content of 14 per cent, has been employed. Table I. — Chemical Composition of Green Alfalfa and Red Clover. Num- ber of Analy- Water (Per Cent.). Crude Ash (Per Cent.). Crude Protein (Per Cent.). Crude Fiber (Per Cent.). Extract or Starchy Matter (Per Cent.). Crude Fat (Per Cent.). Alfalfa, average, ' . Alfalfa, average, 2 . Clover, average, ^ . Clover, average,' . Alfalfa, before bloom, ^ Clover, before bloom,' Alfalfa, in bloom, ' . . Clover, in bloom, i ' . Clover, in bloom,* Alfalfa, in seed, ' . Clover, in seed,' . 143 6 85 13 11 2 27 36 3 6 2 74.7 74.7 73.8 73.8 80.1 80.0 74.1 72.5 72.5 70.2 70.2 2.4 2.0 2.1 2.4 2.3 2.1 2.5 2.0 2.5 2.2 2.7 4.5 3.4 4.1 4.1 4.7 3.6 4.4 4.1 4.6 2.9 4.6 7.0 7.8 7.3 7.5 4.2 4.7 7.8 8.2 7.9 12.8 8.6 10.4 11.5 11.7 11.5 7.9 9.0 10.4 12.1 11.8 11.3 13.1 1.0 .5 1.0 .8 .8 .6 .8 1.1 .8 .6 .8 Table II. — Chemical Composition of Alfalfa Hay (Red Clover Hay for Comparison). Eitract Num- Water (Per Cent.). Crude Crude Crude or Crude ber of Ash Protein Fiber Starchy Fat Analy- (Per (Per (Per Matter (Per Cent.). Cent.). Cent.). (Per Cent.). Cent.). Alfalfa, average, ' . 250 14 8.1 14.0 26.6 35.1 2.2 Clover, average, ' . 76 14 7.0 12.6 25.2 38.1 3.1 Clover, average,' . 15 14 7.8 13.5 24.6 37.6 2.5 Alfalfa, first cutting, ' . 46 14 8.3 13.1 29.0 34.0 1.6 Alfalfa, first cutting,' . 3 14 6.7 14.5 27.5 35.8 1.5 33 14 8.3 13.6 29.6 32.9 1.6 Alfalfa, second cutting,' 1 14 5.8 13.2 32.7 33. 2 1.1 Alfalfa, third cutting, ' 17 14 9.0 13.8 26.8 34.7 1.7 Alfalfa, before bloom, 1 11 14 9.2 20.2 18.8 ;^3.9 3.9 Clover, before bloom, ' 2 14 6.9 17.9 17.6 4U.1 3.5 Clover, before bloom,* 1 14 9.6 15.3 24.4 35.0 1.7 Alfalfa, in bloom, > 31 14 9.3 13.9 28.1 33.0 1.7 Clover, in bloom,' 1 14 7.7 13.2 25.7 37.8 1.6 Alfalfa, in seed, > . ,0 14 6.7 11.7 26.6 38.7 2.4 A study of the analyses of both the alfalfa and clover shows that these plants resemble each other closely in general chemical composition. They 1 Feeds and Feeding, 15th edition, 1915, Henry & Morrison. * Analyses made at the Massachusetts Agricultural Experiment Station. FEEDING VALUE OF ALFALFA. 109 contain considerably more protein than do the cereals and grasses, and less fiber and extract matter. If anything, the alfalfa is likely to be slightly richer in protein than the clover, and to contain a little more extract matter. Much, however, depends upon the exact stage of growth, the season and the soil on which the crops are grown. ^ THE DIGESTIBILITY OF ALFALFA HAY. The general statement may be made that a food is valuable at least in so far as the animal can digest and assimilate it. A large number of digestion trials, principally with sheep, are on record, of which the fol- lowing is a summary: — Table III. — Coefficients of Digestibility of Alfalfa Hay (Other Feeds for Comparison). Num- ber of Dry Matter (Per Cent.). Crude Ash (Per Cent.). Crude Protein (Per Cent.). Crude Fiber (Per Cent.). Extract or Starchy Matter (Per Cent.). Crude Fat (Per Cent.). Alfalfa, average,' . Clover, red, average,* 109 25 60 59 50' 86' 71 59 43 54 72 66 38 57 Alfalfa, first cutting,' . Alfalfa, second cutting,' Alfalfa, third cutting,' 53 21 6 59 62 58 64' 52' 44' 67 76 70 42 44 40 72 74 70 38 40 42 Alfalfa, bud to bloom,' Clover, in bloom,' 74 4 60 62 58 70 62 43 53 72 39 54 Corn fodder, dent, mature for comparison,' 30 66 23 45 63 73 70 Timothy, average for com- parison,' .... 58 55 39 48 50 62 50 Rowen (largely of grasses),' . 12 65 - 70 66 65 47 In making a study of the above summary one notes, in case of the average results, that the digestibility of the dry matter of the alfalfa is about the ti-me as of the clover. The crude protein of the alfalfa is no- ticeably more digestible than that of the clover (12 per cent, more), while 1 As alfalfa begins to blossom, its nitrogen content has been found to consist of 71.1 per cent, of true protein and 28.9 per cent, of so-called amids, although variations from these averages are pro- nounced (Mentzel u. Lengerke's Kalendar). Hart et als.. Research Bulletin No. 33, Wisconsin Experiment Station, found in a sample .31 per cent, of its nitrogen in the form of ammonia, 1.03 per cent, as an acid amid, and 10.17 per cent, as amino acids; the remainder, 88.49 per cent., existed as true protein. Headden, in Bulletin No. 124, Colorado Experiment Station, gives a considerable amount of data on the chemistry of alfalfa, recognizing sucrose, glucose and starch, 2.89 per cent, of galactanand from 11.44 to 13.38 per cent, of pentosans. Pott (Handbuch d. thier. Ernahrung II Band p. 55) reports from 13.9 to 16.8 per cent.' of pentosans. Lindsey and Holland found 4.71 per cent, of galactan in the alfalfa seed. ' Feeds and Feeding, 15th edition, 1915, Henry & Morrison. • Lindsey's compilation, twenty-third report of the Massachusetts Agricultural Experiment Station, 1911. 110 MASS. EXPERIMENT STATION BULLETIN 186. the crude fiber shows a lower digestibility (11 per cent. less). The extract matter of the alfalfa is more digestible than that of the clover. The second cutting of alfalfa hay appears to be more digestible than the first and third cuttings, which are nearly equal in digestibility. Comparing alfaKa in bloom with clover in bloom, one notes the same differences as in the average analyses of all samples : namely, that in case of the alfalfa the crude protein and extract matter are more digestible, and the crude fiber less digestible, than in the clover hay. A comparison of our own results tells substantially the same story, as the following data show: — Table IV. Coefficients of Digestibility of Alfalfa and Clover Hays {Our Results). Exrtact Dry Crude Crude Crude Crude berof Matter Ash Protein Fiber Starchy Fat Single (Per (Per (Per (Per Matter (Per Trials. Cent.). Cent.). Cent.). Cent.). (Per Cent.). Cent.). Alfalfa hay. . . . 6 60 45 74 46 70 28 Clover hay, .... 4 62 58 61 53 68 54 In comparing the total digestibility of alfaKa hay with that of other feeds we have the following figures: alfalfa and clover, about 60 per cent.; timothy, 55 per cent.; rowen (largely of grasses), 65 per cent.; dent corn fodder, 66 per cent. It is evident, therefore, that in point of digestibihty alfalfa and clover are rather more digestible than timothy hay, but less digestible than mature corn fodder or well-cured rowen. Appl3dng the average digestion coefficients to the average analyses of the several feeds, we have the following digestible nutrients for 1 ton: — Table V. - - Digestible Nutrients in One Ton. Crude Protein (Pounds). Crude Fiber (Pounds). Extract Matter (Pounds). Crude Fat (Pounds). Total Nutri- ents (Pounds). Relative Diges- tion Values; Alfalfa = 100. Relative; Net Energy Values; Alfalfa = 100. Alfalfa. . 199 229 505 17 950 100 100 Red clover. 149 272 503 35 959 101 113 Timothy hay. 60 330 550 30 970 102 126 Rowen, 158 318 524 28 1,028 108 1201 Gluten feed,* 446 110 948 52 1,556 164 260 »LiB dsey'B calc ulations. « For compariso n. FEEDING VALUE OF ALFALFA. Ill One notes that of the several coarse fodders, alfalfa furnishes by far the most digestible protein. Thus, timothy hay yields only 60 pounds, clover and rowen 149 and 158 pounds, and alfalfa substantially 200 pounds in a ton. Alfalfa furnishes the largest amount of protein of any of the more conunon and useful coarse fodders. In case, however, of the total digestible nutrients, one notes but little difference between the timothy, clover and alfalfa. Rowen yields 8 per cent, more, while such a concentrate as gluten feed contains 64 per cent, more, than alfalfa. Total digestible matter, however, is not the most satisfactory unit of measure of the energy value of feedstuffs. The unit known as net energy, obtained by deducting from the total energy in the feed the energy losses in feces, urine and heat radiated, is the best known method of comparison. On this basis Armsby's method of calculation, as indicated in the last column of the table, shows red clover to have 13 per cent, more net energy value than aKalfa, timothy hay 26 per cent., rowen 20 per cent., and gluten feed 160 per cent. While experi- ments conducted with the aid of the respiration calorimeter demonstrate these differences, it may be difficult to show such noticeable variations with the aid of ordinary feeding experiments. FEEDING EXPERIMENTS WITH ALFALFA. Experiments I, II and III. Alfalfa, Beet Pulp and Corn Meal v. Hay, Beet Pulp and Corn Gluten Products for Milk Production. The three experiments immediately following were made by the re- versal method with two groups of six and one group of eight cows. The objects of the several experiments were : — 1. To compare the effect of the dry matter and the protein in the two rations on the yield of milk and milk ingredients, and on the gain or loss in weight. 2. To see if the protein derived largely from alfalfa was as satisfactory for milk production as that secured largely from corn by-products. 3. To note if the diuretic effect of the alfalfa caused any noticeable milk shrinkage.^ 4. To observe the possible adverse effect on milk production of the increased metabolism, caused by the alfalfa. The rations were designated as the alfalfa and hay rations. The former consisted of alfalfa as the total roughage, plus beet pulp and corn meal; the latter, of hay as the roughage, plus beet pulp, gluten feed and gluten meal. The alfalfa ration naturally derived its protein largely from al- falfa, while in the hay ration a large part of the protein came from the gluten products. The digestible nutrients in each ration should be about the same. 1 Research Bulletin No. 33, Wisconsin Experiment Station. / 112 MASS. EXPERIMENT STATION BULLETIN 186. Table VI. — History of Cows. Experiment I. Cows. Breed. Age (Years). La-st Calf dropped. Served. Milk Yield, Begin- ning of Trial (Pounds). Samantha II, . CecUe II, . Betty III, . . Fancy III, Betty II, . . . Idalll, . . . Grade Holstein, . Pure Jersey, . Grade Ayrshire, Grade Jersey, Grade Ayrshire, . Pure Jersey, . 7 3 3 8 10 3 Oct. 31, 1915 Nov. 11, 1915 Sept. 14. 1915 Feb. 24, 1916 Aug. 31, 1915 Jan. 29,1916 Dec. 27, 1915 Jan. 13,1916 Apr. 3, 1916 Jan. 18,1916 Mar. 16,1916 40 17 22 29 21 25 Experiment II. Colantha, Grade Holstein, . 3 June 19, 1916 Sept. 15, 1916 23 Mary, . . . Grade Holstein, . 6 Sept. 1,1916 _ 31 Samantha II, . Grade Holstein, . 7 Aug. 14, 1916 Dec. 27, 1916 32 Samantha III, Grade Holstein, . 3 Aug. 6, 1916 Oct. 30,1916 23 Red III, . . . Grade Jersey, 11 Aug. 18, 1916 Nov. 12, 1916 31 White, Grade Holstein, . 7 Aug. 27, 1916 Nov. 17, 1916 41 Experiment III. CecUe II, . Betty II, . Samantha II, Colantha, Red IV, . Ida II, . White, Samantha III, Grade Jersey, Grade Ayrshire, Grade Holstein, Grade Holstein, Grade Jersey, Pure Jersey, . Grade Holstein, Grade Holstein, Oct. 14, 1916 Oct. 25, 1916 Aug. 14, 1916 June 19, 1916 Sept. 26, 1916 Dec. 27, 1916 Aug. 27, 1916 Aug. 6, 1916 Jan. 16, 1917 Apr. 16, 1917 Nov 7, 1916 Oct. 20, 1916 Feb. 28, 1917 Feb. 15, 1917 Oct. 27, 1916 FEEDING VALUE OF ALFALFA. 113 Table VII. — Duration of Experiments. Experiment I. Dates. Hay-ration Cows. Alfalfa-ration Cows. Length of Period (Weeks). Apr. 10 through May 14, 1916, May 26 through June 29,' 1916, Samantha II, Cecile II, Betty III. Fanny III, Betty 11, Ida II. Fancy III, Betty II, Ida II. Samantha II, Cecile II, Betty III. 5 5 Experiment II. Oct. 20 through Nov. 23, 1916, Dec. 4, 1916, through Jan. 7, 1917. Colantha, Mary, Saman- tha II. Samantha III, Red III, White. Samantha III. Red III, White. Colantha, Mary, Saman- tha II. 5 5 Experiment III. Jan. 29 through Mar. 4, 1917, Mar. 15 through Apr. 18, 1917, Cecile II, Betty II, Sa- mantha II, Colantha. Red IV, Ida II, White, Samantha III. Red IV, Ida II, White, Samantha III. Cecile II, Betty II, Sa- mantha II, Colantha. 5 5 Care of Animals. — The animals were well cared for in all cases, and turned into a barnyard from four to nine hours daily, depending upon the weather conditions. They were fed twice daily; the hay was given some time before milking, and the grain just before milking in the afternoon, while in the morning the grain was given just before and the hay just after milking. Water was supplied constantly by the aid of a self -watering device. During the winter the barn wings were kept at a temperature of about 50° F. with the aid of steam heat. Character of Feeds. — The hay was of mixed grasses with some clover, cut upon the station farm. An effort was made to have it of as uniform quality as possible in each experiment. The alfalfa in the first experiment was said to be second cutting, grown in Michigan. It was bright, leafy and sweet, but rather coarse. In the second experiment about one-third of the alfalfa was from the same source, and two-thirds were second and third cutting grown upon the station farm. In the third experiment it was third cutting grown upon the college farm. The beet pulp in the first and second experiments was molasses beet pulp, and in the third experiment, plain dried pulp, — all of good quality. The gluten feed and Diamond gluten meal were of the usual satisfactory quality. The same may be said of the corn meal, except that it was rather moist, and it was necessary to purchase it in small amounts to prevent heating. 114 MASS. EXPERIMENT STATION BULLETIN 186. Sampling Feeds and Milk. — The hays were sampled at the beginning, middle and end of each half of the trial by taking forkfuls of the daily weighings, running same through a power cutter, sub-sampling and placing the laboratory samples in large glass-stoppered bottles; these bottles prop- erly labeled were brought to the laboratory immediately. The grains were sampled daily by placing definite amounts in glass-stoppered bot- tles, properly labeled, and brought to the laboratory at the end of each half of the trial. Dry matter determinations were made and samples pre- pared for complete analysis. The milk was sampled for five consecutive days in each week, preserved with formalin, and the composite analyzed for total solids and for fat by the Babcock method, and for nitrogen. The method of sampling consisted in mixing the milk as soon as drawn with the aid of a perforated tin disk attached to the end of a stout tin handle, by moving the same up and down gently for a number of times, and then taking out a definite amount with a small long-handled tin dipper. Table VIII. - - Analyses of Feeds (P er Cent.). Feed. Water. Dbt Matteb. Ex- peri- ment. Ash. Crude Pro- tein. True Pro- tein. Fiber. Ex- tract Mat- ter. Fat. I Hay 11.62-13.15 8.02 9.13 7.37 35.07 45.27 2.51 II Hay, . . 11.22-11.67 6.34 8.10 7.20 35.19 48.12 2.25 III Hay. . 10.25-11.04 6.46 8.42 7.35 34.06 48.66 2.40 I Alfalfa, 12.86-15.16 7.24 14.93 11.40 41.22 35.14 1.47 II 1 AlfaKa (old), 12.65-14.18 7.22 15.31 11.12 40.56 35.52 1.39 1 Alfalfa (new), 12.43-12.81 7.92 17.41 14.04 31.42 41.42 1.83 III Alfalfa. 11.21-11.79 7.16 14.89 11.83 35.75 40.05 2.15 I Beet pulp, . 11.40-11.98 4.41 10.40 7.65 17.72 66.86 .61 II Beet pulp. . 12.23-12.76 4.09 10.31 _ 18.24 66.73 .63 III Beet pulp. . 10.21-13.53 2.79 11.02 - 21.22 64.28 .69 I Gluten feed, 8.41-10.85 5.11 30.08 21.39 7.13 54.95 2.73 II Gluten feed. 10.25-11.63 4.82 30.99 - 7.13 54.64 2.42 III Gluten feed. 9.47- 9.72 5.00 31.54 8.03 53.36 2.07 I Gluten meal, 8.61- 8.96 .87 48.78 46.23 1.46 47.95 .94 II Gluten meal, 9.26- 9.88 1.20 49.38 - 1.63 46.88 .91 III Gluten meal, 8.32- 8.57 1.04 50.50 - 1.76 45.74 .96 I Corn meal, . 14.53-16.40 1.70 10.34 9.52 1.72 82.61 3.63 II Corn meal. . 13.27-13.64 1.51 10.37 - 2.62 81.42 4.08 III Corn meal, . 11.53-11.65 1.32 10.49 - 2.51 81.75 3.93 FEEDING VALUE OF ALFALFA. 115 The analjiiical data are expressed in dry matter because of variations in moisture. From an analytical standpoint the hays resemble each other closely; the same may be said of the alfalfa, except that the sample in the third experiment contained somewhat less fiber. The albuminoid matter was determined by the Stutzer method, which includes both the amino acids and the acid amids. In view of the fact that the amino acids are supposed to be valuable in protein synthesis, the Stutzer method of separation is not held to be of as much importance as formerly. The hay contained 14.5 per cent, and the alfalfa 22.64 per cent, of its nitrogen in the non-albuminoid form. The beet pulp used in the third experiment showed rather more fiber and a little less extract matter, because of the lack of the molasses. The several lots of the different grains were quite uniform in character. The one sample of gluten feed on which a non-albuminoid nitrogen test was made showed some 29 per cent, of this ingredient, indicating the ad- dition of considerable "steep water" in its manufacture. Table IX. — Average Daily Ration consumed per Cow (Pounds). Experiment I. Num- ber of Cows. Character OP Ration. Alfalfa. Hay. Beet Pulp. Gluten Feed. Gluten Meal. Corn Meal. 6 6 Alfalfa. . Hay. . . 16.47-24.50 18.96 16.00-24.00 18.77 4.00-5.00 4.17 4.00-5.00 4.17 .00-6.00 1.83 1.00-3 00 2.33 3.31-7.75 4.55 Experiment II. Alfalfa, . Hay, . 19.14-22.8 20.65 .00-5.00 3.67 .00-5.00 3.67 .75-3.88 1.73 0.00-4.00 3.17 4.00-6.00 5.13 Experiment III. Alfalfa, Hay, 16.0Ch22.00 20.63 15.69-22.00 20.36 .00-4.00 3.13 .00-4.00 3.13 .00-2.00 .75 2.67-4.50 3.39 1.09-5.16 4.32 The reason for presenting the above concise tables is to give the inter- ested student an idea of the amounts fed daily in the two different rations, 116 MASS. EXPERIMENT STATION BULLETIN 186. and to emphasize their uniformity. In the first experiment more corn meal was fed than gluten products, because of its larger moisture content. Table X. — Estimated Dry and Digestible Nutrients in Average Daily Rations (Pounds). Experiment I. Dry Matter. Digestible Nutrients. Character op Ration. Protein. Fiber. Extract Matter. Fat. Total. , Nutri- tive. Ratio. Alfalfa. . . . Hay. . . . 23.82 23.93 2.23 2.44 3.37 4.09 9.28 8.48 .22 .27 15.10 15.28 1:5.9 1 :5.4 Experiment II. Alfalfa. . . . 25.64 2.66 3.28 10.08 .25 16.27 1 :5.2 Hay, . . . 25.31 2.67 4.29 9.09 .25 16.30 1:5.2 Experiment III. Alfalfa. Hay, 24.84 24.72 15.71 15.74 1 :5.7 1 :5.1 The above data were secured by applying average digestion coefficients to the analyses of the several feeds, and multiplying by the amounts of dry matter consumed daily. It is at best but an estimate. It serves, however, to give the reader an idea of the uniformity of the two rations, in so far as digestible nutrients and nutritive ratio are concerned. FEEDING VALUE OF ALFALFA. 117 1. The effect of the total dry matter contained in the two separate rations, and also the effect of the dry matter in the hay and in the alfalfa upon the yield of milk and milk ingredients. Table XI. — Total Dry Matter consumed in Each Feed and in the Com- plete Ration {Pounds). Experiment I. i Alfalfa. Hay. Beet Pulp. I Glttten Feed. Gluten Meal. Corn Meal. Total. •o 1 1 1 1 ii 1 ft 1 1 1 >>> r 1 ^1 < 6 3,421 16.29 3,455 16.45 773 3.68 349 1.67 447 2.13 809 3.85 5,003 5,024 Experiment II. 3,776 17.99 3,711 17.69 675 3.21 325 1.54 602 .87 4.44 5,380 5,313 Experiment III. 8 5,116 18.26 5,1 20 777 2.76 866 3.09 1,071 3.82 6,964 6, Totals. 12,259 2,225 - 864 1,915 - 2.810 - 17.347 17, Beet pulp was fed in each half of each trial in substantially like amounts. A study of Table XI indicates that the total dry matter consumed in each ration was substantially the same, the most noticeable variation being in Experiment II. The total dry matter consumed in the three experiments was nearly the same, differing by only about one-half per cent. The dry matter consumed in the form of alfalfa and in hay in the three experiments (12,312 pounds and 12,259 pounds) likewise shows a varia- tion of substantially only one-half per cent. 118 MASS. EXPERIMENT STATION BULLETIN 186. Table XII. — Total Milk and Milk Ingredients produced. Experiment I. Num- ber of Cows. Character>f Ration., Milk duced (Pounds). Solids (Per Cent.). Solids (Pounds). Fat (Per Cent.). Fat (Pounds). Nitro- (Per Nitro- een (Pounds). 6 6 Alfalfa, . Hay. . . . 4,916 4,870 12.78 13.47 628.1 655.9 4.21 4.73 206.6 230.5 .51 .54 25.10 26.38 Experiment II. 6 6 Alfalfa, . Hay. . . . 4,856 4,776 13.21 13.20 641.5 630.5 4.57 4.53 222.1 216.3 .54 .54 26.42 25.61 Experiment III. 8 Alfalfa, . . . 6,094 14.04 855.6 5.02 306.2 .59 35.73 8 Hay. . . . 6,087 14.18 862.9 5.10 310.4 .60 36.46 Totals. Alfalfa, Hay. 15,866 15,733 13.34 13.62 2.125.2 2,149.3 4.60 4.79 734.9 757.2 87.25 ■ 88.45 Table XII shows that in Experiment I the milk yield favored the al- falfa ration by about 1 per cent., while in the second experiment the dif- ference was about 2 per cent. In the third experiment the difference was very slight, — only 7 pounds. The total of the three experiments gives a yield of 15,866 pounds for the alfalfa ration, and 15,733 pounds for the hay ration, a difference of about nine-tenths of 1 per cent, in favor of the alfalfa. In Experiment I, for some reason, the yields of total solids and total fat were noticeably greater (4.4 and 11.1 per cent.) on the hay ration. These results, however, were not made emphatic by the two other experi- ments; hence one is in no way justified in assuming that the alfalfa in- fluenced the milk composition. A definite amount of dry matter, there- fore, in each of the rations produced substantially the same results in the yield of milk and milk ingredients. One sees that the alfalfa stimulated slightly the yield of milk without correspondingly increasing the solids. FEEDING VALUE OF ALFALFA. 119 Table XIII. — Gain or Loss in Live Weight (Pounds). Gain. Loss. Net. Experiment. Alfalfa Ration. Hay Ration. Alfalfa Ration. Hay Ration. Alfalfa Ration. Hay Ration. I II Ill 105 63 26 292 60 136 73 108 _ 7 +105 —10 —82 +292 +53 +136 Totals. - - - +13 +481 In Experiment I, when several of the animals were somewhat advanced in the milking period, each herd showed an increase in live weight. In Experiment II the cows were comparatively fresh and not as much gain was noted; in this experiment the alfalfa ration produced a slight de- crease in the weight of the herd. In Experiment III, conducted during the winter and early spring, a decrease was also noted when the alfalfa ration was fed. It may be remarked that in each experiment it was our object to feed slightly less nutrients than calculations showed to be neces- sary to maintain weight and to meet the demands for milk, so that the full effect of each ration would be felt. It seems evident that while the alfalfa and corn meal ration fully maintained the milk yield, it was not as effective in increasing the live weight as was the hay and gluten ration. 2. The effect of different forms of protein on the yield and character of the milk. Table XIV. — Protein consumed in the Feeds and Rations (Pounds). Alfalfa. Hay. Beet Pulp. 1 Gluten Products. Corn Meal. Totals. Experiment. Alfalfa Ration. Hay Ration. I II Ill 510.8 631.2 761.8 315.4 309.0 428.8 80.4 69.6 85.6 323.0 398.0 497.2 83.7 96.4 112.3 674.9 797.2 959.7 718.8 773.6 1,011.6 Totals, 1,903.8 1,050.2 235.6 1,218.2 292.4 2,431.8 2,504.0 Beet pulp was fed in each half of each experiment in substantially like amounts. 120 MASS. EXPERIMENT STATION BULLETIN 186. Table XV. — Protein Found in the Milk (Pounds), Experiment. Alfalfa Ration. Hay Ration. I 156.9 165.1 223.3 164 9 n 166.0 III. . 227 9 Totals. . 545.3 558.8 In case of the alfalfa ration, the total amount of protein consumed in the three experiments was 2,431.8 pounds, of which 1,903.8 pounds, or 78.2 per cent., came from the alfalfa, and 528 pounds, or 21.7 per cent., came from the beet pulp and corn meal. In the hay ration, of the total of 2,504 pounds consumed, 1,050.2 pounds, or 41.9 per cent., came from the hay, and 1,453.6 pounds, or 58.1 per cent., came from the beet pulp and corn gluten products. The total protein in the milk (N X 6.25) produced by the alfalfa ration was 545.3 pounds, and by the hay ration 558.8 pounds, showing that the alfalfa ration, in which 78.2 per cent, of the protein was derived from alfalfa, produced as much milk protein and substantially as much milk solids as did the hay ration; or, in other words, that the protein of the alfalfa was fully as satisfactory a source of protein for milk formation as was that in the hay and corn gluten. An objection might be raised to this conclusion because 528 pounds of protein (21.7 per cent, of the total amount fed) was derived from beet pulp and corn meal, and this amount of protein was nearly equal to the amount produced in the milk. It must be remembered, however, that of the 528 pounds, scarcely two-thirds would be digestible and hence available for milk production. Although it is quite possible that the protein from the beet pulp and corn meal was also utilized for the formation of the nitrogenous matter in the milk, it is fairly safe to conclude that the alfalfa protein proved fully as satisfactory a source for milk formation as did that contained in the hay and corn gluten products. Hart and Humphrey^ have more completely demon- strated this by feeding to two cows a ration composed of alfalfa and starch, and they found that the protein in the alfalfa was equal to that contained in a ration composed entirely of corn products. S. The diuretic effect of the alfalfa. The same authors have shown in two experiments with two cows that the substitution of alfalfa in place of corn products caused a marked in- crease in the excretion of urine and a shrinkage in the milk yield, in some cases amounting to substantially 25 per cent. FEEDING VALUE OF ALFALFA. 121 Because of the number of cows involved, it was not practicable to determine the urine output nor the water drunk. On the basis, however, of the volume of milk as well as the total solids yielded, as stated in Table XII, it did not appear in the five weeks' period that the alfalfa exerted any adverse effect. A study of the daily records of individual cows, espe- cially during the transition period from the hay to the alfalfa ration, con- firms this conclusion. In fact, the alfalfa seemed to act as a slight stimulus to production. Whether this was due to the favorable character of the proteins or to other causes is not clear. 4. The infltience of the increased metabolism caused by the alfalfa on the yield of milk and on live weight. Armsby^ has shown that by increasing the metabolism alfalfa is de- cidedly inferior as a source of energy to timothy hay, in the proportion of 34.1 to 48.63 therms of net energy per 100 pounds of dry mattter; i.e., a decrease of some 30 per cent.^ Inasmuch as the dry matter in alfalfa and in hay comprised some 71 per cent, of the total dry matter contained in each of the two rations, it would seem as though the influence of the increased metabolism caused by the alfalfa would be noticeable, even though the hay was not what might be classed as timothy. The yields of milk and milk solids fail to show any unfavorable effect of this factor. Only in the case of the live weight (Table XIII) produced does one notice a possible adverse effect of the alfalfa, which might be attributed to its inferior energy value. Table XVI. Additional Experimental Data. Total Rations Consumed by Each Cow (Pounds). Experiment I. Cows. Alfalfa. Hay. Beet Pulp.' Corn Meal. Gluten Feed. Gluten Meal. Fancy III, 735.0 714.5 140 151.0 35.00 105.0 Betty 11 609.5 604.5 140 151.0 52.50 87.5 Ida II 611.0 602.5 140 151.0 52.50 87.5 Samantha II 857.5 840.0 175 271.0 210.00 35.0 Cecilell 576.5 560.0 140 116.0 35.00 70.0 Betty III 589.5 620,5 140 116.0 - 105.0 1 The Nutrition of Farm Animals, pp. 660, 663. ' Most other hays (mixtures of grasses) are also shown to be quite superior to alfalfa as a source of energy. 3 The same amount fed in each half. 122 MASS. EXPERIMENT STATION BULLETIN 186. Table XVI. — Total Rations Consumed by Each Cow Experiment II. Concluded. Cows. Alfalfa. Hay. Beet Pulp. Corn Meal. Gluten Feed. Gluten Meal. Samantha III 735.0 700.0 105 175.0 26.25 140.0 Red III 700.0 665.0 140 140.0 135.64 - White, 799.0 770.0 140 210.0 96.25 105.0 Colantha, . . . . 697.0 700.0 105 183.8 35.00 140.0 Mary 670.0 622.0 105 183.8 35.00 140.0 Samantha II 735.0 735.0 175 183.8 35.00 140.0 Experiment III. Red IV 700.0 688.0 105 144.5 35.00 93.5 Ida II. 700.0 700.0 105 144.5 35.00 105.0 White, 770.0 770.0 105 162.1 ~ 157.5 Samantha III, 770.0 770.0 105 144.5 - 137.0 Cecile II, . 560.0 549.0 140 108.2 - 105.0 Betty II, . 770.0 727.0 105 180.6 70.00 105.0 Samantha II, 735.0 731.0 105 180.6 70.00 105.0 Colantha, . 770.0 765.0 105 144.5 - 140.0 Table XVII. Changes in Live Weight {Pounds). Experiment I. Cows. Alfalfa. Hay. Fancy III ' . . +29 +10 +8 +13 +23 +22 +20 Betty II +93 Ida II +28 Samantha II, Cecile II +52 +51 Betty III +48 Totals, +105 +292 FEEDING VALUE OF ALFALFA. 123 Table XVII. — Changes in Live Weight (Pounds) — Concluded. Experiment II. Cows. Alfalfa. Hay. Samanthalll Red III +17 +38 +8 —32 -41 +5 —5 White, ........ +45 +7 Mary, . ..... +3 Samantha II, —2 Totals. . —10 +57 Experiment III. Red IV, —7 —35 +9 —16 —24 +17 -22 +7 Ida II +19 White, +1 Samantha III, Cecile II, .... . . . +27 +1 Betty II +10 Samantha II Colantha, +16 +55 Totals, —102 + 136 Table XVIII. — Yield of Milk and Milk Ingredients. Experiment I. Alfalfa Ration. Cows. Milk (Pounds). Solids (Per Cent.). Solids (Pounds). Fat (Per Cent.). Fat (Pounds). Nitro- (fe^r Cent.). Nitro- gen (Pounds). Fancy III, Betty II. . Ida II, Samantha II, Cecile II, . Betty III, . 1,044.4 687.2 791.0 1,186.0 549.7 656.2 11.97 12.96 13.45 12.33 14.42 12.52 125.01 89.06 106.39 146.23 79.27 82.16 3.82 4.32 4.68 3.85 5.06 4.04 39.89 29.69 37.02 45.66 27.81 26.51 .46 .53 .51 .50 .61 .51 4.80 3.64 4.03 5.93 3.35 3.35 Totals, 4,914.5 - 628.12 - 206.58 - 25.10 124 MASS. EXPERIMENT STATION BULLETIN 186. Table XVIII. — Yield of Milk and Milk Ingredients — Continued. Experiment I. — Concluded. Hay Ration. Cows. Milk (Pounds). Solids (Per Cent.). Solids (Pounds). Fat (Per Cent.). Fat (Pounds). Nitro- gen (Per Cent.). Nitro- (PoulTds). Fancy III, Betty II, . . . Ida II, Samantha II, . Cecilell, . Betty III, . 992.7 631.3 670,5 1,280.3 581.0 714.2 12.77 13.82 14.38 12.79 14.95 13.28 126.77 87.25 96.42 163.75 86.86 94.85 4.36 4.95 5.26 4.36 5.56 4.56 43.28 31.25 35.27 55.82 32.30 32.57 .50 .56 .58 .51 .62 .54 4.96 3.54 3,89 6.53 3.60 3.86 Totals, . . 4,870.0 - 655.90 ~ 230.49 - 26.38 Experiment II. Alfalfa Ration. Samantha III, . 630.1 14.00 88.21 4.75 29.92 .59 3.72 Eed III, . 847.2 13.12 111.15 4.82 40.84 .52 4.41 White, 953.1 12.54 119.52 4.48 42.70 .52 4.96 Colantha, . 689.8 13.10 90.36 4.19 28.90 .55 3.79 Mary, 874.8 12.71 111.19 4.10 35.86 .50 4.37 Samantha II, . 861.0 14.06 121.06 5.10 43.91 .60 5.17 Totals, 4,856.0 - 641.49 - 222.14 - 26.42 Hay Ration. Samantha III, . 621.7 14.09 87.60 4.72 29.34 .60 3.73 Red III, . . . 631.5 13.51 85.32 5.00 31.58 .55 3.47 White, 954.7 12.75 121.72 4.45 42.48 .53 5.06 Colantha, . 709.1 13.03 92.40 4.25 30.14 .54 3.83 Mary, . . . 955.2 12.58 120.16 4.21 40.21 .46 4.39 Samantha II, . 903.9 13.64 123.29 4.71 42.57 .59 5.33 Totals. 4,776.1 - 630.49 - 216.32 - 25.61 FEEDING VALUE OF ALFALFA. 125 Table XVIII. — Yield of Milk mid Milk Ingredients Experiment III. Alfalfa Ration. Concluded, Cows. Milk (Pounds). Solids (Per Cent.). Solids (Pounds). Fat (Per Cent.). Fat (Pounds). Nitro- (fe^r Cent.). Nitro- gen- (Pounds). Red IV, . Ida II, White, Samantha III, Cecilell, . Betty II, . Samantha II, Colantha, . 775.0 877.0 865.4 648.1 597.6 950.3 711.2 669.3 14.54 14.41 13.13 13.94 15.50 13.87 13.76 13.48 112.69 129.38 113.63 90.35 92.63 131.81 97.86 90.22 5.44 5.40 . 4.73 4.72 6.03 4.89 4.65 4.42 42.16 47.36 40.93 30.59 36.04 46.47 33.07 29.58 60 57 54 61 64 57 61 58 4.65 5.00 4.67 3.95 5.42 4.34 3.88 Totals, 6,093.9 855.57 - 306.20 - 35.73 Hay Ration. Red IV, . . . 738.2 14.72 108.66 5.43 40.08 .62 4.58 Ida II, 756.2 15.05 113.81 5.75 43.48 .61 4.61 White, 883.9 13.45 118.88 4.83 42.69 .58 5.13 Samantha III, . 662.8 14.31 94.85 4.88 32.35 .61 4.04 Cecilell, . 683.1 15.15 88.34 5.72 33.35 .63 3.67 Betty II, . . 899.2 14.25 128.13 5.22 46.94 .59 5.31 Samantha II, 893.8 13.55 121.11 4.74 42.40 .59 5.30 Colantha, . 670.0 13.30 89.11 4.34 20.08 .57 3.82 Totals, 6,087.2 862.89 310.37 - 36.46 Experiments IV and V. Alfalfa V. Rowen for Milk Production. The claims made for alfalfa as a coarse fodder par excellence for milk production led us to compare the same with the second cutting of grass known as rowen. Two experiments were conducted with four cows each by the reversal method. The methods followed in the experiments, such as care of cows, sampling of feeds and milk, are the same as those described in previous experiments of a similar nature. 126 MASS. EXPERIMENT STATION BULLETIN 186. s 4.73 6.43 5.25 3.90 Milk Yield. Beginning of Trial (Pounds). g ;3 S S i i n^ i 1 Oct. 27, 1916 Jan. 10, 1917 Sept. 19. 1916 Aug. 18. 1916 Sept. 1, 1916 4 oo ■* ^ «> ' 1 1 1 c '1 1 c i o 1 i 1 ■* «5 O 1! tj & I O O (» Q g « FEEDING VALUE OF ALFALFA. 127 Table XX. — Duration of Experiments. Experiment IV. Dates. Rowen-ration Cows. Alfalfa-ration Cows. Period (Weeks). Feb. 26 through April 1, 1917, . April 12 through May 16, 1917, . Mary, Red III, . . Fancy III, Peggy, Fancy III, Peggy, Mary, Red III, . 5 5 Experiment V. April 26 through May 23. 1917, . June 3 through June 30, 1917, . Red IV, Wall, . . Cecils II, Betty II, . Cecile II, Betty II, . Red IV, Ida 11, . 4 i Character of Feeds. — The rowen represented the second cutting of grass. It was well cured and in good condition, but it did not show a digestibility equal to the average, as the results stated below will show. The alfalfa was of good quality; it was grown in New York State, and while rather coarse was said to be third cutting. The corn meal and bran were of the usual good quality. Table XXI. — Coefficients of Digestibility secured for Rowen and Alfalfa. Trials. Dry Matter. Ash. Protein. Fiber. Extract Matter. Fat. Rowen Average (previous trials), Alfalfa, Average (previous trials, third cutting). 2 12 4 6 61 65 58 53 34 43 44 60 70 72 70 68 66 46 40 63 65 66 70 32 47 24 42 It will be noted that the digestibility of the protein in the rowen was noticeably below the average. The alfalfa coeflEicients agreed well with the average results of other trials. The protein in the rowen showed a digestibility inferior to that of the protein in the alfalfa, while the fiber in the alfalfa was noticeably less digestible than the fiber in the rowen. The low digestibility of the fiber in the alfalfa is characteristic of the plant. 128 MASS. EXPERIMENT STATION BULLETIN 186. Table XXII. — Analyses of Feeds {Per Cent.). Feed. Water. Dry Matter. Ex- peri- ment. Ash. Crude Pro- tein. True Pro- tein. Fiber. Ex- tract Mat- ter. Fat. IV Rowen, . 10.29-11.13 8.87 12.59 10.05 28.57 45.93 4.04 V Rowen, . 9.08-10.96 7.66 11.99 10.48 26.33 50.36 3.66 Average, . 8.27 12.29 10.27 27.45 48.15 3.85 IV Alfalfa, . 11.84-12.17 7.21 15 58 12.59 33.13 41.89 2.19 V Alfalfa, . 11.58-11.87 7.05 16.29 13.16 29.65 44.68 2.33 Average, . 7.13 15.94 12.88 31.39 43.29 2.26 IV Grain mixture, ' . 12.23-12.65 3.32 12.60 5.10 74.50 4.48 V Grain mixture, ' . 13.07-13.18 3.43 13.02 - 5.24 74.80 3.51 1 The grain mixture consisted of 30 per cent, bran and 70 per cent, corn meal. Applying the digestion coefficients secured by our experiments to the analyses of rowen and alfalfa, tjae following amounts of organic nutrients are found to be digestible in 2,000 pounds of dry matter. Table XXIII. — Digestible Organic Nutrients in 2,000 Pounds Dry Matter {Poxmds). Feed. Protein. Fiber. Extract Matter. Fat. Totals. Rowen Alfalfa 147.48 229.54 373.32 288.72 606.69 571.43 24.64 10.85 1,152.13 1,100.61 The alfalfa furnished 82.06 pounds more of digestible crude protein than did the rowen, but less digestible fiber and extract matter, and rather less total digestible organic nutrients. While the rowen contains noticeably less digestible protein, the above computation indicates that it should prove approximately as valuable for milk production as alfalfa. FEEDING VALUE OF ALFALFA. 129 Table XXIV. — Total Rations consumed by Each Cow (Pounds). Experiment IV. Rowen. Alfalfa. Grain Mixture. Cows. Rowen Ration. Alfalfa Ration. Mary 626 610 210 210 Red III 663 685 195 210 Fancy III 768 770 350 350 Peggy 630 630 210 210 Totals 2,687 2,675 965 980 Experiment V. Red IV 504 504 168 168 Idall 504 504 168 168 Ceoile II 476 476 196 196 Betty II 688 578 224 224 Totals 2,072 2,062 756 756 Totals (both experiments), 4,759 4,737 1,721 1,736 The totals show that in the two experiments substantially like amounts of rowen or alfalfa and grain were fed. Table XXV. — Total Dry Matter consumed in Each Feed (Pounds). MENT. Rowen. Alfalfa. Grain Mixture. EXPER Rowen Ration. Alfalfa Ration. IV 2,401 1,864 2,354 1,828 845 657 V, . 657 Totals. 1 . : 4.265 4,182 1,502 1,514 About 2 per cent, more dry matter in the form of rowen was fed than in alfalfa, while the dry matter in the form of grain was about the same. If the rowen was equal to the alfalfa, one would expect fully as good results in milk yield and live weight. 130 MASS. EXPERIMENT STATION BULLETIN 186. Table XXVI. — Gain or Loss in Live Weight (Pounds). Gain. Loss. Net. Experiment. Rowen Ration. Alfalfa Ration. Rowen Ration. Alfalfa Ration. Rowen Ration. Alfalfa Ration. IV V 26 5 10 59 18 37 30 0 +8 -32 —20 +36 Totals, - - " -24 +16 There appeared to be a slight gain on the alfalfa and a slight loss on the rowen ration. Table XXVII. — Yield of Milk and Milk Ingredients. Experiment IV. Rowen Ration. Cows. Milk (Pounds). Solids (Per Cent.). Solids (Pounds). Fat (Per Cent.). Fat (Pounds). Nitro- (?e"r Cent.). Nitro- gen (Pounds). Mary, . . . Red III, . Fancy III, Peggy. . . . 770.2 596.1 1,132.9 609.0 12.62 14.26 13.34 15.51 97.20 85.00 151.13 94.46 4.17 5.52 4.89 6.23 32.12 32.90 55.40 38.25 .52 .61 .49 .62 4.01 3.64 5.55 3.78 Totals, 3,108.2 13.76' 427.79 5.10> 158.67 .55> 16.98 Alfalfa Ration. Mary, . . . 737.4 12.51 92.25 3.94 29.05 .51 3 76 Red III, . . . 608.3 13.42 81.63 5.01 30.48 .59 3.59 Fancy III, 1,272.1 13.03 165.75 4.47 56.86 .53 6.74 Peggy. . . . 650.7 15.43 100.40 6.36 41.38 .63 4.10 Totals, . . 3,268.5 13.461 440.03 4.82' 157.77 .56' 18.19 Experiment V. Rowen Ration. Red IV, . . . 638.4 15.13 81.46 5.91 31.82 .59 3.18 Ida II, . 555.6 14.75 81.95 5.74 31.89 .57 3.17 CecUe II, . . . 478.3 15.14 72.41 6.69 27.22 .62 2.97 Betty II, . . . 727.1 13.39 97.36 4.42 32.14 .50 3.64 Totals, 2,299.4 14.49' 333.18 6.35' 123.07 .66' 12.96 > Average percentages obtained by dividing total pounds miUc. solids, etc., by total pounds of FEEDING VALUE OF ALFALFA. 131 Table XXVII. — Yield of Milk and Milk Ingredients Alfalfa Ration. Concluded. Cows. Milk (Pounds). Solids (Per Cent.). Solids (Pounds). Fat (Per Cent.). Fat (Pounds). Nitro- (fe^r Cent.). Nitro- gen (Pounds). 554.4 540.2 490.7 714.6 14.26 14.04 15.26 13.69 79.06 75.84 74.88 97.83 27.89 27.17 29.10 34.80 .59 .56 .59 .56 Red IV. . . Ida II, . Cecile II, . . Betty II, . . 6.03 5.03 5.93 4.87 3.27 3.03 2.89 4.00 Totals, Totals rowan. Totals alfalfa. 2,299 9 5,407.6 5,568.4 14.24' 327.61 760.97 767.64 5.17' 118.96 281.74 278.73 .57' 13.19 29.94 31.38 Average percentages obtained by dividing total pounds of solids, etc., by total pounds of In Experiment IV the alfalfa ration apparently increased the yield of milk 5.2 per cent., while in Experiment V the yield was the same on each ration. The total yield for both experiments was 5,407.6 pounds on the rowen, and 5,568.4 pounds on the alfalfa, or an increase of 3 per cent, in favor of the alfalfa. The rowen ration produced a total yield of 760.97 pounds of solids as against 767.64 pounds for the alfalfa; the total yield of fat was 281.74 pounds on the rowen ration, and 278.73 pounds on the alfalfa; the yield of nitrogen was 29.94 pounds on the rowen ration, and 31.38 pounds on the alfalfa. The following table shows the amount of milk and milk ingredients pro- duced by 100 pounds of dry matter derived from each of the two rations: — Table XXVIII. — Milk and Milk Ingredients produced by 100 Pounds of Dry Matter {Pounds). Ration. Milk. Solids. Fat. Rowen, Alfalfa 93.77 97.76 13.19 13.48 4.89 4.89 In case of the volume of milk, and to a less degree in case of the total solids, the jdelds were rather in favor of the alfalfa ration. The fat per- centage, on the other hand, did not keep pace with the increase in the milk yield. Note (Table XXVII) that in Experiment IV, with the rowen ration, the percentage was 5.1 as against 4.82 for the alfalfa ration; and in Experiment V, 5.35 for the rowen ration as against 5.17 for the alfalfa ration. The per cent, of solids not fat was substantially the same in each 132 MASS. EXPERIMENT STATION BULLETIN 186. experiment, namely, 8.66 against 8.64 in the fourth, and 9.14 against 9.07 in the fifth. On the basis of dry matter, the fat yield was the same with each ration. Experiment VI . Alfalfa, English Hay and Grain v. English Hay and Grain for Milk Pro- duction. The object of this particular experiment with milch cows was to compare the feeding value of a ration composed of equal parts of alfalfa and Eng- lish hay, corn-and-cob meal and a Uttle bran (mostly home-grown products) with that of one consisting of Enghsh hay, bran, corn-and-cob meal and gluten feed, in order to see whether reasonably satisfactory results could not be secured from the use of alfalfa as a considerable source of protein, in place of purchased protein in the form of bran and gluten feed. Plan of the Experiment. — Eight cows which had calved during the late summer and autumn were divided into two groups of four each and fed by the reversal method. One group of four received the so-called alfalfa ration at the same time the other four were receiving the English hay and purchased grain ration. In the second half of the trial the feeding was reversed. Table XXIX . — History of Cows Cows. Breed. Age (Years). Last Calf dropped. Served. Milk Yield, Begin- ning of Trial (Pounds). Samantha Fancy II, Samantha Cecile. Red III, Daisy II, Ida, . Betty II, II. Grade Holstein, . Grade Jersey, Grade Holstein, . Pure Jersey, . Grade Jersey, Grade Jersey, Pure Jersey, . Grade Ayrshire, . 8 4 2 6 6 2 4 4 Sept. 23, 1911 Oct. 28, 1911 Nov. 1, 1911 Nov. 21, 1911 Sept. 23, 1911 Nov. 17, 1911 Nov. 16, 1911 Nov. 9, 1911 Nov.- 7. 1911 Dec. 10,1911 Dec. 10,1911 Mar. 12,1912 Nov. 6,1911 Mar. 25, 1912 Feb. 24, 1912 Jan. 8, 1912 26.1 30.6 28.9 23.7 20.2 28.0 31.1 Table XXX. — Duration of Experiment. Dates. Alfalfa-ration Cows. English Hay-ration Cows. Length Period (Weeks). Dec. 28, 1911, through Jan. 24, 1912. Feb. 9 through Mar. 7, 1912, Samantha. Fancy II, Sa- mantha II, Cecile. Red III, Daisy II, Ida, Betty II. Red III, Daisy II. Ida, Betty II. Samantha, Fancy II, Sa- mantha II, Cecile. 4 4 FEEDING VALUE OF ALFALFA. 133 An interval of fifteen days was allowed between the two periods of the experiment. Character of Feeds. — The hay was fine and of fair quality, coming from a meadow that had been in grass for a number of years. The alfalfa was grown upon the college grounds, and was of excellent quality. The corn- and-cob meal was excellent, and the bran and gluten feed of average quahty. The method of care and feeding, weighing of the animals and sampling of the feeds and milk were the same as previously described. Table XXXI. — Analyses of Feeds (Per Cent.). Feed. Water. Ash. Crude Pro- tein. True Pro- tein. Fiber. Ex- tract Matter. Fat. Totals. Alfalfa (farm), . . . 11.20 7.25 16.66 11.79 28.62 35.54 1.73 100 Alfalfa (experiment station), . 10.14 7.88 16.85 13.79 24.86 38.89 1.38 100 English hay, .... 9.49 5.58 8.63 7.74 28.45 45.64 2.21 100 Wheat bran 12.43 6.02 15.46 9.68 52.04 4.37 100 Corn-and-cob meal, 16.04 1.28 8.27 4.38 66.92 3.11 100 Gluten feed, .... 9.97 .82 25.74 6.64 53.37 3.46 100 Table XXXII. — Total Rations consumed by Each Cow (Pounds). Alfalfa Ration. Cows. Hay. Alfalfa. Bran. Corn-and- cob Meal. Gluten Feed. Samantha Fancy II Samantha II Cecile Red III Daisy II Ida Betty II 336 280 336 308 336 224 280 308 326 278 336 304 329 224 280 298 56 56 56 56 56 56 56 56 196 140 168 168 140 140 168 168 - Totals 2,408 2,375 448 1,288 - 134 MASS. EXPERIMENT STATION BULLETIN 186. Table XXXII. — Total Rations consumed by Each Cow (Pounds) — Con- cluded. English Hay Ration. Cows. Hay. Alfalfa. Bran. Corn-and- cob Meal. Gluten Samantha 663 - 66 84 112 Fancy II, . . . . 553 - 56 56 84 Samantha II 663 - 56 56 112 CecUe 586 - 56 84 84 Red III 762 - 56 84 56 Daisy II 442 - 56 84 56 Ida 594 - 56 84 84 Betty II, 604 - 56 84 84 Totals 4,767 - 448 616 672 Table XXXIII. — Average Daily Ration consumed per Coxo {Pounds). Character of Ration. Hay. Alfalfa. Bran. Corn-and- cob Meal. Gluten Feed. Alfalfa English hay 10.7 21.3 10.6 2 2 5.8 2.8 _ 3 The above tables show that the average cow on the alfalfa ration con- sumed 10.6 pounds of alfalfa and 10.7 pounds of hay, or 21.3 pounds of roughage, and in addition, 2 pounds of bran and 5.8 pounds of corn-and- cob meal; while on the hay ration the average cow ate 21.3 pounds of hay, 2 pounds of bran, 2.8 pounds of corn-and-cob meal and 3 pounds of gluten feed. Different cows naturally varied from this average, depending upon their individual requirements. It was a comparison of ration against ration, and not one single feedstuff against another. If similar rations were used by a dairyman, in case of the hay ration he would be obhged to purchase 2 pounds of bran and 3 pounds of gluten feed for each animal; he could produce the hay and the corn-and-cob meal upon the farm. In case of the alfalfa ration he would find it necessary to purchase only the 2 pounds of bran daily, and he could grow the remainder of the ration. In fact, the animals probably would do about as well if the bran were omitted and the corn-and-cob meal correspondingly increased. FEEDING VALUE OF ALFALFA. 135 Table XXXIV, — Estimated Dry and Digestible Nutrients in Average Daily Ratiojis (Pounds) . Character of Ration. Dry Matter. Digestible Organic Nutrients. Nutri- Protein. Fiber. Extract Matter. Fat. Total. tive Ratio. Alfalfa, . English hay, . 25.78 26.09 2.30 2.07 3.32 3.95 9.98 9.76 .41 .46 16.01 16.24 1 :6.17 1 :7.11 The above results were calculated from actual analyses and average digestion coeflBcients. The two rations do not vary greatly from each other; the total digestible nutrients are about the same and hkewise the extract matter. The amount of digestible fiber in the hay ration is a httle higher and the protein a little lower. The daily protein consumption is somewhat higher in the alfalfa ration. One would expect substantially similar results from the two rations. Of the 2.3 pounds of digestible pro- tein in the alfalfa ration, 1.28 pounds, or 55.8 per cent., was from the alfalfa hay, and the balance of L02 pounds from the hay and grain. In the hay ration 1.05 pounds, or nearly 50 per cent., of the protein was from the hay, and the balance of 1.02 pounds from the grain. Table XXXV. Gain or Loss in Live Weight (Pounds). Alfalfa Ration. . a « & a 4 s ,2 1 5 02 1 1 Beginning, .... 827 677 777 873 805 726 1,062 930 _ End 815 675 772 880 783 708 1,030 900 - Gain or loss. —12 —2 —5 +7 -22 -18 -32 -30 —114 English Hat Ration, Beginning, , . , . 860 675 827 895 777 725 1,068 942 - End 832 657 761 890 793 720 1,040 925 - Gain or loss. -28 —18 —66 —5 +16 -5 —28 —17 —161 The cows lost somewhat in weight on both rations. 136 MASS. EXPERIMENT STATION BULLETIN 186. Table XXXVI. — Yield of Milk and Milk Ingredients. Alfalfa Ration. Cows. Total Milk (Pounds). Daily Milk (Pounds). Solids (Per Cent.). Solids (Pounds). Fat (Per Cent.). Fat (Pounds). Samantha, 702.2 25.07 15.96 112.07 6.25 43.89 Fancy II, . 667.0 23.82 13.23 88.24 4.73 31.55 Samantha II, 829.9 29.64 13.28 110.21 4.48 37.18 CecUe. . 762.8 27.24 13.64 104.05 4.85 37.00 Red III, . 664.6 23.73 13.60 90.39 5.24 34.83 Daisy II. . 516.6 18.45 14.15 73.10 4.93 25.47 Ida, . 655.6 23.41 14.59 95.65 5.67 37.17 Betty II, . 741.4 26.48 13.41 99.42 4.52 33.51 Totals, 5,540.1 24.73 13.96' 773.13 5.061 280.60 English Hay Ration. Samantha, 766.9 27.39 14.79 113.42 5.69 43.64 Fancy II, . 641.5 22.91 13.49 86.54 4.81 30.86 Samantha II, 841.1 30.04 13.28 111.70 4.41 37.09 CecUo, . 663.4 23.69 13.81 91.62 5.08 33.70 Red III, . 632.7 22.60 13.68 86.55 5.39 34.10 Daisy II, . 547.3 19.55 13.56 74.21 4.65 25.45 Ida, . . 720.0 25.71 14.43 103 90 5.78 41.62 Betty II, . 788.6 28.16 13.74 108.35 4.91 38.72 Totals, 5,601.5 25.00 13.861 776.29 5.09' 285.18 1 Average percentages obtained by dividing total pounds of solids and of fat by total pounds of milk. Table XXXVII. — Average Composition of the Milk (Per Cent.) Chakacter of Ration. Total Solids. Fat. Alfalfa 13.96 13.86 6.06 English hay, 5.09 FEEDING VALUE OF ALFALFA. 137 Table XXXVIII. — Dnj and Digestible Matter required for Maintenance and to produce Milk and Milk Ingredients (Pounds). Character of Ration. Dry Matter. Digestible NuTRiE>fT.s. 100 Pounds Milk. 1 Pound Solids. 1 Pound Fat. 100 Pounds Milk. 1 Pound Solids. 1 Pound Fat. Alfalfa English hay, . 104.24 104.33 7.47 7.56 20.57 20.49 64.73 64.94 4.64 4.69 12.78 12.76 The tables showing the yield of milk and milk ingredients, the com- position of the milk and the dry and digestible matter required to pro- duce milk all point to the fact that the two rations were equally effective. Only in case of live weight were the results rather against the hay ration. Experiment VII. Alfalfa, Corn Stover, Corn-and-coh Meal and Bran v. English Hay, Corn- and-cob Meal, Gluten Feed and Bran for Milk Production. In Experiment VI the feeding effect of a ration composed of one-half English hay, one-half alfalfa, together with a large amount of corn-and- cob meal and a Uttle bran, was compared with a ration of Enghsh hay, corn-and-cob meal, gluten feed and bran. In the present experiment (VII) a ration composed of alfalfa, cut corn stover and a large amount of corn-and-cob meal with a small amount of bran was compared with a ration of Enghsh hay and substantially Uke amounts of corn-and-cob meal, gluten feed and bran. The question to be answered is, "Can the farmer by growing alfalfa and corn get along without purchasing grain?" Plan. — Eight cows were used and fed by the usual reversal method. Because the cows calved at different times the eight animals were not all fed between the same dates, but in groups of two. Table XXXIX. — Histonj of Cows. Cows. Breed. Age (Years). Last Calf dropped. Served. Milk Yield, Begin- ning of Trial (Pounds). Samantha Red III, Betty, Betty II, Amy, Amy II, Samantha Red III, Grade Holstein, . Grade Jersey, Grade Jersey, Grade Ayrshire, Pure Jersey, . Pure Jersey, . Grade Holstein, . Grade Jersey, 10 8 9 6 6 4 10 8 Aug. 26, 1913 Aug. 23, 1913 Nov. 23, 1913 Oct. 18, 1913 Dec. 9,1913 Dec. 17,1913 Aug. 26, 1913 Aug. 23, 1913 Nov. 19, 1913 Nov. 2,1913 Apr. 13. 1914 Jan. 9, 1914 Mar. 14, 1914 Jan. 30,1914 Nov. 19, 1914 Nov. 2, 1913 19.4 24.5 29.3 26.4 30.1 24.1 21.0 20.5 138 MASS. EXPERIMENT STATION BULLETIN 186. Table XL. — Duration o f Experiment. Dates. Alfalfa, Corn Stover, Corn-and-cob Meal and Bran Ration Cows. English Hay, Corn- and-cob Meal, Glu- ten Feed and Bran Ration Cows. Length Pe?Ld (Weeks). Nov. 19 through Dec. 23, 1913, Jan. 3 through Feb. 6. 1914, . Dec. 24, 1913, through Jan. 27, 1914, Feb. 6 through Mar. 12, 1914, . Jan. 21 through Fob. 24, 1914, . Mar. 4 through Apr. 7, 1914, . . Feb. 28 through Apr. 3, 1914, . Apr. 11 through May 15, 1914, . Samantha, Red III, . Betty II, . Betty, Amy II, . Amy, Samantha, Red III, . Red III, . Samantha, Betty, Betty II, . . . Amy, Amy II, . Red III, . . . Samantha, 5 5 5 5 5 5 5 5 The care and feeding of the animals, time of weighing and method of sampling feeds and milk were the same as in the previous trial. Character of Feeds. — The hay was of quite satisfactory quality, tim- othy predominating; the alfalfa hay was also of average quality. The corn stover was stocked out of doors, and was subject to weather condi- tions. The corn-and-cob meal was made from corn grown upon the station grounds, while the bran and gluten feed were purchased. Table XLL — Analyses of Feeds (Per Cent.). < Feed. Water. Ash. Protein. Fiber. Extract Matter. Fat. English hay, . 11.30 6.04 9.32 29.09 42.50 2.06 Alfalfa hay, . . . 11.90 6.56 14.45 27.99 36.95 2.04 Corn stover, 33.15 4.44 5.96 23.24 32.47 .88 Grain mixture, . 11.16 2.92 17.09 7.76 57.32 3.91 Bran 11.54 5.89 15.58 10.47 51.27 4.80 Corn-and-cob meal, 12.74 1.33 7.94 5.30 69.16 3.27 Table XLII. — Total Rations consumed (Pounds). English Hay. Alfalfa Hay. Corn Stover. Grain Mixture. Bran. Corn-and- oob Meal. English hay ration totals, . Alfalfa ration totals, . 5,873 4,143 1,966 2.240 684 1,559 FEEDING VALUE OF ALFALFA. 139 Table XLIII. — Average Daily Ration consumed per Cow (Pounds). Characteb of Ration. English Hay. Alfalfa Hay. Corn Stover. Grain Mixture. Bran. Corn-and- cobMeal. English hay, Alfalfa 21 14.8 7 8 2.44 5.57 The "grain mixture" was composed of a mixture, by weight, of 30 parts wheat bran, 35 parts gluten feed and 35 parts corn-and-cob meal. The above tabulations show that a ration composed of hay and a grain mixture was compared with a ration of alfalfa, some corn stover, a large amount of corn-and-cob meal and a rather limited amount of bran. On the basis of dry matter, the alfalfa ration contained 80 per cent, alfalfa and 20 per cent, corn stover. In case of the grains, 65 per cent, of the amount fed with the English hay would have to be purchased, and only 30 per cent, of that fed with the alfalfa. Table XLIV. Estimated Digestible Nutrients in Average Daily Rations (Pounds). Character of Ration. Protein. Fiber. Extract Matter. Fat. Total. English hay Alfalfa 2.25 4.32 3.18 9.08 9.54 .44 .38 15.73 15.85 The alfalfa ration furnished rather more digestible protein than the English hay ration, although it is believed the latter ration contained all that was needed by the animals. The English hay ration, as nearly as can be estimated, contained rather more total digestible nutrients than the alfalfa ration. This was due to the rather high moisture content of the corn stover fed as a portion of the alfalfa ration. On the basis of digestible nutrients, one would expect slightly better returns from the English hay ration. Table XLV. Gain or Loss in Lave Weight (Pounds). English Hay Ration. s ,!^ 1 1 ^ 1 1 >> B ■< >, 1 ^ 1 Beginning, .... 1,137, 945 870 907 727 830 1,122 955 _ End 1,147 960 892 903 727 785 1,170 955 - Gain or loss, -MO -1-15 +22 -4 > ,1 1 a < 1 ■3 Beginning 1,090 905 867 870 713 770 1,075 970 - End 1,167 915 850 860 700 734 1,075 975 _ Gain or loss, ' ' ■ —23 + 10 —17 —10 —13 —36 ± +5 —84 It is evident that the cows gained slightly on the hay and grain ration and lost somewhat on the aKalfa, corn stover and grain ration. Cow Amy was not in very good condition and lost noticeably in weight during both feeding periods. Table XLVI. — Yield of Milk and Milk Ingredients. English Hat Ration. Cows. Total Milk (Pounds). Daily Milk (Pounds). Solids (Per Cent.). Solids (Pounds). Fat (Per Cent.). Fat (Pounds). Red III, . Samantha, Betty, Betty II. . Amy, Amy II, . Red III, . Samantha, 765.8 768.0 1,000.2 972.9 727.4 713.2 762.3 21.9 21.9 28.5 28.6 27.8 20.8 20.4 21.8 13.61 15.73 13.84 13.90 13.81 15.15 14.28 14.82 104.23 120.81 138.16 139.03 134.36 110.20 101.84 112.97 4.85 '" 4.71 4.73 5.09 5.77 5.30 5.33 37.14 44.16 47.02 47.31 49.52 41.97 37.80 40.63 Totals, 6.708.1 24.0' 14.33' 961.60 5.15' 345.55 Alfalfa Ration. Red III 740.5 21.2 14.40 106.63 5.52 40.88 Samantha, 658.1 18.8 15.54 102.27 5.96 39.22 Betty, 837.4 23.9 13.51 113.13 4.69 39.27 Betty II, . 970.1 27.7 14.17 137.46 4.92 47.73 Amy, 835.4 23.9 13.87 115.87 5.10 42.61 Amy II, . 789.2 22.6 14.89 117.51 5.59 44.12 Red III, . 637.4 18.2 13.96 88.98 5.19 .33.08 Samantha. 691.5 19.8 14.98 103.60 5.37 37.13 Totals. 6,159.6 22.0' 14.37' 885.45 5.26' 324.04 Average. FEEDING VALUE OF ALFALFA. 141 It is very evident that the hay and grain ration gave noticeably larger returns of mUk and milk ingredients than did the alfalfa ration. The alfalfa ration produced 8.2 per cent, less milk and 8.6 per cent, less milk solids than did the English hay ration. The writer is convinced that the milk shrinkage on the alfalfa ration was due largely to the corn stover. While of good quality it was stooked out of doors and brought to the barn every few days and cut fine before being fed. It varied considerably in moisture content, depending upon the weather. If the stover had been brought from the field in November and stored under cover, in all probability more satisfactory results would have been secured. 142 MASS. EXPERIMENT STATION BULLETIN 186. Part II. THE VALUE OF CORN BRAN FOR MILK PRODUCTION. Summary and Suggestions. 1. Corn bran contains noticeably less ash, protein and fat, and some- what more extract or starchy matter, than does wheat bran. 2. Digestion experiments with sheep recently made at this station showed that 80 per cent, of its dry matter was digestible as against 66 per cent, for wheat bran. 3. A definite amount of dry matter contained in a ration composed of hay, gluten feed, ground oats, cottonseed meal and corn bran produced, in an average of two experiments, substantially as much milk and milk ingredients as a like amount of dry matter in a ration composed of hay, gluten feed, ground oats, cottonseed meal and wheat bran. 4. The gains in live weight were about the same on each ration. 5. Corn bran, if properly combined'in a grain ration, is likely to give as satisfactory returns as wheat bran. It may constitute 30 per cent, of the ration, together with 30 per cent, cottonseed or linseed meal, 20 per cent, corn or hominy meal, and 20 per cent, ground oats 5 or a ration may be combined consisting of 40 per cent, corn bran, 20 per cent, gluten feed, 20 per cent, cottonseed or linseed meal, and 20 per cent, ground oats or barley. A combination of corn bran, gluten feed and corn meal would not be satisfactory because of a deficiency in ash, and because all three constituents would be derived from corn. The Experiment in Detail. What Corn Bran is. — Corn bran is the hull or skin of the corn kernel, together with a small amount of the germ, and the starchy portion which it is impossible to separate out in the process of manufacture of various corn products, such as starch and glucose. The bran thus obtained was formerly dried and sold by itself, but at present it is more often sold as a constituent of hominy feed or proprietary mixed feeds, or is mixed with corn gluten as a component of gluten feed. It is still sometimes found in the markets of Massachusetts, and has been offered at a reasonable price. It has been shown, by means of experiments^ conducted at various times, 1 Massacbusetta Experiment Station Bulletin No. 181, p. 316. VALUE OF CORN BRAN FOR MILK PRODUCTION. 143 to be well digested by ruminants; its energy value as compared with corn meal at 100 is equal to 82. In the minds of many feeders corn bran is considered a quite inferior product, and at best of doubtful feeding value. Method of conducting the Experiment. — In order to demonstrate its value two feeding experiments with cows were carried out at this station during 1917 and 1918. In one case six and in the other eight cows were fed by the reversal method, for two periods of five weeks each, on a basal ration of hay, gluten feed, ground oats and cottonseed meal.^ Half of the cows in each case received in addition 4 pounds of corn bran during the first periods of the experiments, while the other half received a like amount of wheat bran. In the second periods the corn and wheat brans were interchanged. At the outset the cows used in each experiment were as carefully mated in regard to yield of milk and period of lactation as possible, so that the two herds receiving the different rations would vary in general performance but very little. Their names and arrangement may be found in Tables I and II. > A little cottonseed meal was added to each ration in the second experiment in order to insure against the possible ill effect of having too great a proportion of the grains derived from corn in the corn bran half of the trial. 144 MASS. EXPERIMENT STATION BULLETIN 186. 4 o o g s g 3 Si 1^ 5.20 4.20 4.00 3.40 4.40 3.90 ^2| ?3 S ?5 ?3 ?5 ?5 bo d 11 755 1,090 975 1,190 1,140 1,030 1 1 1 1 1 1 1 1 1 1 1 t 1 11 ji Aug. 9. 1917 Aug. 16, 1917 July 3, 1917 July 24, 1917 Aug. 1, 1917 July 5, 1917 1 t^ 00 M CO >« TO ' B* h' q a" a" Grade Jersey, Grade Holste Grade Holste Grade Holste Grade Holste Grade Holste 1 Peggy Samantha II, Colantha Samantha III, Samantha IV, '^ "^ '^ o> ° o «• a ^ tj % ^ :2; ►? ^ ;z: o o 0» 03 O Od 0> OS •* « >0 TO oT CO d >>bb»*»c5 bbti>» TO OS t» B a a a a K (^ h^ I 1 2 2 a o fc 1 P^ Ai VALUE OF CORN BRAN FOR MILK PRODUCTION. 145 As in all feeding experiments, a sufficient preliminary period was al- lowed at the beginning of each trial for the cows to become accustomed to the rations, and for their alimentary tracts to become emptied of what- ever food they may previously have been receiving. For the same reason a transitional period was allowed between the two halves of each experi- ment. These periods were of at least ten days' duration. The exact dates are given in Table II. The amounts of hay and grains fed the various cows daily were care- fully calculated for each animal, on the basis of its milk and maintenance requirements,^ and from personal knowledge of the particular animal's appetite. The general care and management of the animals, as well as the methods of sampling milk, hay and grain, were similar to those already described in the foregoing experiments. The hay which was used in the rations was raised on the experiment station farm, and was of average uniformity and good quality. All the grains were of standard quality. The daily and total amount of each feed per cow may be found in the following table, as well as the average and total amounts per herd: — 1 T. L. Haeker, Minnesota Bulletin No. 140, p. 56. 146 MASS. EXPERIMENT STATION BULLETIN 186. -i 1 2 ^ '3 1 1 " ES ^ 1 , 1 1 1 1 1 1 ^ ^ s i s g § g § g § ;< ^ o O) W g 3 § g § g § § g , K o "^ i >. s g s g § § S ^ ^ ^ ^ CO CO ' CO Ph ^ 1 3 o § g § g s g g , ?? E; s s s s g o H >; ■«1 fQ « ^ 3 ' ' 1 ' ' ' 1 1 & ^ 4 s § § g § s . ^ ^ eo '«< ■.»< ■^»" "* ^ 3 ss S § o 5 s ^ ' o ^ >, g § § g § g . s H p s S s s s c3 § w 1 8 § § § § g 8 i i i § g 1 ' • w a > l-l . ■g ■a ■g 1 1 d 1 1 . . nn ?5 1 a > ^ d •a 1 2 00 If O u 1 r ' ■ ' . . . . g 8 g g M c» ■ ja ^ 9* 1 g r- ^ od ^ !^ O Q VALUE OF COEN BRAN FOR MILK PRODUCTION. 147 ' ' 1 • ' ■ ' g § . s g § S5 g J J g s § , ■>»< ■* 1 ■* o o § • • 1 1 § g § § 1 i § I 1 ii 1 1 a i I ' S c^ 3 05 § g o o o § § g g g s § g g g g g g g (N ca •-< ci .-< g g o o o o o o US IS la lo o t-^ lO t- ^ § >o o g g g ui ci ui pooooooo S S 5 3 g g S S g g g 2 S §3 g g 2 S at § g g S S g g g . c3 j3 ja M -'S S I i * ca tJ -2 2|^ •a i g • o I I » li^l >> *i ^ Q s Ml-" a c -is J3 J: ^ ^ la Z^ ° S J3 S _ p ^ a^ g 9 i"! 8 o 1 1 i ^* -a § -s § (jj -CI ,^ a 148 MASS. EXPERIMENT STATION BULLETIN 186. Q h 6 3.00 1.50 1.00 1.00 3.00 1.50 1.50 1.50 "5 3 §g§8§ggS i SSSSSS^SSS i o 0 z SSSSSSSS . 3 1 g§§gg§8S i ooeoooiOQOt^mt^ u- i , § g g g g • 5 "3 1 , § s § , , g g ^ g 1 s s !■ 1 ^ % H 1 1 ' - 1 ssssggss § ' i5 « M iz; K o o 1 ' • ' • 3 1 1 1 1 1 1 r r 1 1 >. 2 1 gggggsgg ^ s ' ^ 6 s' ^^ i J ^- i i i 1 ^ g ^ ^ 1 g g \ j 5 i \ i 1 1 1 1 t < VALUE OF CORN BRAN FOR MILK PRODUCTION. 149 Table III. — Analyses of Feeds {Per Cent.). Experiment I. Average Mois- ture. > Dry Matter. > Dry Matter. Feed. Ash. Protein. Fiber. Extract Matter. Fat. Hay Corn bran, Wheat bran, Gluten feed, Ground oats, f 11.63 \ 10.252 / 12.90 \ 11.81 f 11.33 \ 10.97 1 10.44 \ 10.30 / 11.16 I 10.23 88.37 89.752 87.10 88.19 88.67 89.03 89.56 89.70 88.84 89.77 ) 0. 8.28 7.76 17.23 30.52 11.52 32.39 11.79 10.23 7.13 11.67 50.32 78.19 60.90 54.78 67.71 2.66 1.21 4.51 2.85 5.22 Experiment 11. Hay / 11.34 \ 10.58 88.66 89.42 5.88 7.83 33.68 50.35 ..26 Corn bran, f 12.62 \ 11.59 87.38 1.31 7.36 12.62 77.36 1 35 Wheat bran, . 1 11.55 \ 11.99 88.45 88.01 7.05 16.31 11.25 59.90 5.49 Gluten feed, . ( 9.42 1 9.29 90.58 90.71 3.87 29.66 8.17 55.49 2.81 Ground oats, . J 10.66 1 10.39 89.34 89.61 3.72 11.99 11.88 66.91 5.60 Cottonseed meal. f 9.60 \ 9.11 90.40 90.89 6.23 38.64 13.10 34.77 7.26 1 The two figures in each case represent the average of three samples taken in each half of the trials. ■ In case of cows Samantha IV and Colantha II special samples of hay had to be taken during the second half of the experiment for moisture determinations, and the figures derived are as follows: Samantha IV, moistiire 10.38, dry matter 89.62; Colantha II, moisture 9.86, dry mat- ter 90.14. The variations in composition of the hay and grain used were com- paratively slight. The average analyses of the corn and wheat brans used in the two experiments compare as follows on the dry-matter basis: — Table IV. — Average Analyses of the Corn and Wheat Brans {Per Cent.). Extract Matter. Corn bran, . Wheat bran. 1.18 7.09 7.56 16.77 12.20 10.74 77.78 60.40 1.28 6.00 150 MASS. EXPERIMENT STATION BULLETIN 186. Wheat bran contains more ash, protein and fat, and noticeably less extract or starchy matter, than does the corn bran. In using corn bran as a component of a dairy ration these differences, particularly the ash and protein, would have to be given consideration. By applying the percentages of dry matter of the various feeds as given in Table III to the amounts fed (Table II), the amounts of dry matter fed can easily be obtained. Only the totals for the herds and the aver- age per animal for each herd are given in Table V. Table V. — Total Amount and Average Daily Amount of Dry Matter consumed (Pounds). Experiment I. Corn Bran Ration. Hay. Corn Bran. Wheat Bran. Gluten Feed. Ground Oats. Ck)tton- eeed Meal. Total Daily average, . 3,834 18.26 733 3.50 - 593 2.83 374 1.78 - Wheat Bran Ration. Total Daily average, . 3,840 18.26 j 747 3.55 696 2.84 375 1.79 - Experiment II. Corn Bran Ration. Total Daily average, . 5,155 18.42 984 3.52 _ 333 1.19 444 1.59 470 1.68 Wheat Bran Ration. Total Daily average, . 5,153 18.41 988 3.53 333 1.19 444 1.59 469 1.68 During the two experiments the total amount of dry matter consumed by the cows receiving the corn bran ration was 12,920 pounds, while the cows receiving the wheat bran ration consumed substantially the same, or 12,945 pounds. Of these totals, 1,717 pounds represented corn bran and 1,735 pounds wheat bran. For convenience the average daily amounts of dry matter consumed per cow in the two rations of both experiments are here tabulated. VALUE OF CORN BRAN FOR MILK PRODUCTION. 151 Table VI. — Average Daily Amount of Dry Matter consumed 'per Cow (Pounds) . Ration. Hay. Corn Bran. Wheat Bran. Gluten Feed. Ground Oats. Cotton- seed Meal.i Corn bran, Wheat bran, 18.34 18 34 3.51 3.54 2.01 2.02 1.73 1.74 1.69 1.59 An application of the percentage composition of each feed as given in Table III to the above figures would give the amounts of protein, fat, fiber, etc., each ration contained, and this in turn multiplied by average digestion coefiicients^ would give the approximate digestible nutrients in each ration. Table VII. — Estimated Dry and Digestible Nutrients in Average Daily Rations (Pounds). Dry Matter. Digestible Nutrients. Nutri- Character of Ration. Protein. Fiber. Extract Matter. Fat. Total. tive Ratio.* Corn bran, Wheat bran, . 26.39 26.42 1.92 2.23 4.12 4.01 9.75 9.16 .44 .62 16.23 15.92 1 : 7.72 It will be seen that the dry and digestible matter consumed in each ration was almost identical. The digestible protein contained in the corn bran ration was some 14 per cent, less than that in the wheat bran ration. It is believed, however, that a surplus remained after making the usual allowance for maintenance and milk requirements. ' Used in Experiment II only. ' Coefficients used were the results of determinations made with sheep. Lack of space pro- hibited printing them here. ' Fat taken to equal 2.2 times carbohydrates. 152 MASS. EXPERIMENT STATION BULLETIN 186. Table VIII. — Yield of Milk and Milk Ingredients. Experiment I. Corn Bran Ration. Dates. Cows. Milk (Pounds). Solids (Per Cent.). Solids (Pounds). Fat (Per Cent.). Fat (Pounds). Oct. 24 through Nov. 27, 1917. Dec. 8. 1917, through Jan. 11, 1918. Peggy, Samantha II, . Colantha II, . Colantha, . Samantha III, . Samantha IV, ' . 659.7 981.7 797.6 633.0 691.4 896.9 16.20 12.99 13.12 12.63 13.82 12.93 90.67 127.52 104.65 79.95 96.55 116.00 6.95 4.61 4.43 4.11 4.90 4.18 38.90 45.26 35.33 26.02 33.88 37.60 Totals, . . . Average, 2 . 4,560.3 13.47 614.34 4.76 216.89 Wheat Bran Ration. Oct. 24 through Nov. 27, 1917. Colantha, . Samantha III, . 601.2 699.4 12.45 13.56 74.85 94.84 4.19 4.89 25.19 34.20 Samantha IV, . 837.4 12.59 105.43 4.15 34.75 Dec. 8, 1917, through Jan. 11, 1918. Peggy, . . Samantha II, 605.6 1,058.1 16.17 13.01 97.93 137.66 6.80 4.53 41.18 47.93 Colantha II. i . 871.6 13.12 114.35 4.33 37.74 Totals, 4,673.3 _ 625.06 - 220.99 Average, 2 . - 13.38 - 4.73 - Experiment II. Corn Bran Ration. Feb. 20 through Mar. 26, 1918. Red IV, . 945.1 13.65 129.01 5.07 47.92 Colantha, . 584.8 12.63 73.86 4.03 23.57 Samantha III, . 672.4 13.68 91.98 4.66 31 33 Samantha IV. . 851.3 12.82 109.14 4.10 34.90 Apr. 6 through May 10, 1918. Fancy III, 889.0 12.42 110.41 4.35 38.67 Peggy, 496.9 15.28 75.93 6.21 30.86 Samantha II, . 886.6 12.91 114.46 4.43 39.28 Colantha II. . 676.9 13.67 92.40 4.59 31.02 Totals, 6,002.0 - 797.19 - 277.55 Average,' . - 13.28 - 4.62 - » See footnote, Table II. 2 Average obtained by dividing total pounds of solids and fat by total poundi of milk. VALUE OF CORN BRAN FOR MILK PRODUCTION. 153 Experiment II — Concluded. Wheat Bran Ration. Dates. Cows. Milk (Pounds). Solids (Per Cent.). Solids (Pounds). Fat (Per Cent.). Fat (Pounds). Feb. 20 through Mar. 26, 1918. Apr. 6 through May 10, 1918. Fancy III, Peggy, Samantha II, Colantha II, . Red IV, . . Colantha, . Samantha III, . Samantha IV, . 552.0 1,004.8 765.8 832.8 463.9 633.9 12.61 15.20 12.84 13 33 13.68 12.99 13.63 12.78 117.69 83.90 129.02 102.08 113.93 60.26 86.40 105.93 4.42 6.24 4.36 4.31 5.02 4.36 4.78 4.18 41.25 34.44 43.81 33.01 41.81 20.23 30.30 34.65 Totals, . . . Average, I . 6,015.4 13.29 799.21 4.65 279.50 - Average obtained by dividing total pound,s of solids and fat by total pounds of milk. The total milk produced in the two experiments on the corn bran ration was 10,562.3 pounds, and 10,688.7 pounds on the wheat bran ra- tion, an increase of 1.19 per cent, in favor of the latter. The total solids produced on the corn bran ration amounted to 1,411.5 pounds as against 1,424.3 pounds for the wheat bran. The corn bran ration produced 494.4 pounds of fat as against 500.5 pounds on the wheat bran ration, an in- crease of 1.3 per cent. Table IX. — Gain or Loss in Live Weight (Pounds). Ration. Experiment I. Experiment II. Totals. Gain. w Gain. Loss. Corn bran, .... Wheat bran 93 91 32 8 86 106 18 50 + 129 + 139 A slight gain was made on each ration. BULLETIN No. 187 NOVEMBER, 1918 MASSACHISETTS AGRICILTIRAL EXPERIMENT STATION CLARIFICATION OF MILK Part I By CHARLES E. MARSHALL and E. G. HOOD Together with Lieutenant R. C. Avery, S. G. Mutkekar, Lieutenant William L. Payne, Mary L. Chase, Harry A. Cheplin, Louise Hompe, John E. Martin, Conrad H. Lieber, James Neil), Louis P. Hastings, John Yesair. The independent work of Lieutenant E. L. Davies has been included Our purpose has been to ascertain, if possible, the real signifi- cance of clarification of milk Requests for bulletins should be addressed to the Agricultural Experiment Station Amherst, Mass. CONTENTS. Part I. I. Introduction: page The significance of the clarifier, ....... 155 What is clarification? ........ 155 Slime produced by clarification, . . ' . . . .158 Milk modified by clarification, ...... 158 II. Slime: Amount of slime removed, ....... 158 Determination of the weight of slime, ..... 159 Effect of temperature upon the amount of slime, . . . 177 Influence of time and acidity upon the amount of slime, . . 178 Food value of slime, . . . . . . . . .180 Leucocytes (so-called) in slime, . . . . . . . 183 The fibrin (so-called) in slime, . . . . . . .184 The dirt in slime 185 Micro-organisms in slime, ........ 190 III. Milk: Corpuscular elements of milk, ....... 196 The fibrin (so-called) in milk, 202 Micro-organisms in milk, ........ 203 Effect of clarification upon the germ-content, . . . 203-214 Effect of clarification upon certified and commercial milks, 215, 216 The development of bacteria in clarified and unclarified milks (plates), 217-227 Influence of temperature upon clariflcation of milk as revealed by germ-content, ......... 228 Repeated clarification of milk, ...... 229 Effect of clarification upon pure cultures of bacteria, molds and yeasts 229-236 Effect of three clarifications upon pure cultures, . . . 237 Colonization of bacteria in milk, ...... 238 Efficiency of the individual organism free and in colony, . . 238 Other considerations, . « . . . . . . . 239 IV. Simimary, 241 Part II. This portion of the work is well under way, and will treat comparatively the changes in unclarified and clarified milk, ...... 242 Publication of this Document approved by the Supervisor of Administration. BULLETI]^ I^o. 187. DEPARTMENT OF MICROBIOLOGY. CLARIFICATION OF MILK. Part I . I. INTRODUCTION. The Significance of the Clakifier. The use of the clarifier has been an outgrowth of the employment of the "separator" in an attempt to clarify or purify milk. Since the func- tion of the "separator" is to remove the fat from milk, the addition of a new function to this machine presented compUcations not easily overcome in a single machine, for as improvement takes place in the primary purpose of the separator, retrogression may be instituted in the secondary, as in this case, — the clarification or purification of milk. The separation of fat from milk is not desired in clarification, yet it is desirable to accomplish what the separator also succeeds in doing in part, — the removal of for- eign and unwholesome elements so far as this is possible. A single- purposed machine is susceptible of higher development simply because it does not have to compromise with other and foreign purposes. Ac- cordingly, there is good reason, as a basis, for endeavoring to perfect a machine wliich will perform the single function of clarification in its highest degree. What is Clarification? It is the work of this comparatively new machine, known as a clarifier, which has been subjected to careful study in this laboratory. Its func- tion, not its mechanism, has been studied. Milk is poured into the machine from which it emerges as milk. In its passage through it has lost that substance which adheres to the bowl of the machine, — the slime. The problem before us, therefore, takes this form : What is the slime, and in its removal from milk has it improved or injured the milk? The fullest answer which can be given at this time is the substance of this continued thesis. The categorical reply to this question*cannot be given till the close of this laboratory's studies, which, 156 MASS. EXPERIMENT STATION BULLETIN 187. it is to be hoped, may eventually have fairly covered the field compre- hensively as well as quite intensively. The present attitude toward the clarifier is reflected by the Commission on Milk Standards^ which offers a status on clarification. Summing up the points bearing upon milk purification by the clarifier are found these Favorable : — (a) It removes visible dirt. (b) It removes inflammatory products, including many of the causative germs. (c) It performs the work of the strainer, but in a much more efficient manner. Against : — (a) It removes visible dirt, but not all disease-producing germs, and hence mis- leads the consumer as to the real purity of the milk. (b) It does not remove urine or the soluble portion of feces; nevertheless the milk appears clean. (c) It adds another process requiring the handling of the milk, complicating the situation. (d) It largely destroys the value of the dirt test, though no more so than good straining. (e) It breaks up clumps of bacteria and distributes them through the milk. (/) The exact nature of the material removed is not yet fully understood. The essence of the above assertions is found in the bewildering effect it produces on the mind of a critical reader, for it both asserts and does not assert. When the summary concludes thus: "The exact nature of the material removed is not yet fidly understood," it neutrahzes the first effect produced and causes a fog to settle on the rather precocious opinions preceding. It is unfortunate that the reader is left to speculate concerning the reaUties wliich actually he submerged beneath tliis opales- cent atmosphere. It is fitting, therefore, to analj^ze these statements, not exhaustively, but a little more closely, just for the purpose of indicating their looseness. Putting several of these statements together, the thought is thrown into one or two channels: — (a) It removes visible dirt. (b) It performs the work of a strainer, but in a much more efficient manner. (c) It removes ^'isible dirt, but not all disease-producing germs, and hence mis- leads the consumer as to the real purity of the milk. (d) It largely destroys the value of the dirt test, though no more so than good straim'ng. In other words, it removes visible dirt more effectively than any strainer. "Confusing the consumer," "the total elimination of organisms," and "the effect on a test" have no relation to its claim. It may be said, too, that straining of milk must be as reprehensible in misleading the consumer as clarifying, for does it not prepare the consumer for a more sightly product? Yet straining is upheld. The authors feel confident that such assertions as the above will mislead the reader. 1 U. S. Public Health Service, Public Health Reports, Vol. 2, No. 7, p. 17. CLARIFICATION OF MILK. 157 Again, "it does not remove all the disease-producing organisms." It would be a rare centrifuging machine which would claim such a func- tion as eliminating all pathogenic micro-organisms, in the light of what is known about centrifuging out such forms. Selective elimination of this nature savors of the superhuman at present, and implies more than is possible. The clarifier is the product of human effort. " It destroys the value of the dirt test." This is rated as an unfavorable quaUty, yet is considered favorable in the case of straining. One might ask whether it is desirable to remove as much dirt as possible, or allow it to remain simply to make the dirt test, occasionally appUed, effective? If the authors were to sum up these statements as they stand, they must conclude that the clarifier is a far more efficient strainer, which is allowed, apparently, than any now in use. A criticism of the clarifier, very peculiar because of its subtle nature, is introduced: — " (b) It does not remove urine or the soluble portion of feces; neverthe- less the milk appears clean." The implication here is far-reaching, for the reader might think that there is such a machine or device, on the one hand or on the other, and that such a claim is made for the clarifier or a centrifuge. Why such an assertion is left in its baldness for lay readers to digest the writers caimot understand. Does any device accomplish it, does even pasteurization of milk, which is a sort of panacea advocated by this commission for aU mUk trouble, overcome what is intimated? That such products exist even in the best of milk, in an infinitesimal degree, cannot be denied, but it seems a strange assertion in connection with a review of clarification. Why not explain? Here is another very interesting assertion (this is properly made): "(e) It breaks up clumps of bacteria and distributes them through the milk." This is well-founded, but what is the result? The need of an answer to this is apparent and it should accompany the statement. Does the commission know? In a general waj^, how often is such a reason given? "(c) It adds another process requiring the handling of the milk, com- pHcating the situation." Here, too, is one of those statements which are so commonly brought forth to "clinch" an argument. Has man ever hesitated to utihze a new device, when such a device, so far as he can de- termine, improves the product, even if it does entaU a new movement? It corresponds very closely with the exclamation of a certain writer who had done no particular work with the clarifier, and who closed his review with, "What next?" The authors have perhaps colored this very brief analysis too highly by specific selections, but not without a purpose. They have not even done it to criticize, although criticism may be merited in a way. The object has been to bring conspicuously before the reader the confused condition of minds and the lack of knowledge as well as the existence of certain substrata of prejudice relevant to a new device (the clarifier) 158 MASS. EXPERIMENT STATION BULLETIN 187. designed to meet a specific demand which had been fostered by the fre- quent use of another device (the separator) for clarification. On the other hand, criticism could be easily framed from a re\aew of literature of manufacturing firms which has for its purpose the setting forth of the merits of the clarifier. Wliile the specific statements have a modicum of truth and a basis in fact, the reader is left to deduce a quantitative estimate which is very misleading. There exists a sinister purpose beneath the surface which is not commendable. How, for in- stance, is the reader to gather the significance of a photograph of slime deposit when he knows nothing about its relation to the milk? Is he to infer that milk which may be highly infectious to man is rendered safe when passed through a clarifier? Such a statement and many others, by inference, are highly reprehensible, and should not be tolerated by in- telhgent men. If the clarifier cannot prove its value per se, then it is rightly questioned and should be weighed in the balance of exacting scrutiny. Let this new contribution be judged by its work stated in concrete and sane speech. It is only fair to the public to have sanitarians and manu- facturers alike deal frankly and honestly with such matters as clarification. Such statements need study, and some of them should not have been written before a careful investigation had been made. Clarification aims to assist in the purification of milk. Does it do it or does it not, and to what degree? This is the definite goal toward which the work of this laboratory has been directed. At the start it is frankly allowed that the best way to secure pure milk is to have a sound cow and obtain the milk free from dirt and disease contamination. This is a recommendation difficult to execute. Human knowledge and performance are weak. It seems impracticable to many minds. The clarifier is offered as a means to assist in accomplishing what man as a machine fails to do. The performance of the clarifier is bound up in what is re- moved in the slime, and in how the removal of this slime affects the milk from which it has been eliminated. II. SLIME. Slime is that material which is removed from the milk during the process of clarification, and which adheres to the bowl of the clarifier. It consists, speaking in a general manner, of the so-called leucocytes or epithelial cells of milk, or corpuscular elements of milk, so-called fibrin which exists in milk in the form of microscopic shreds, traces of casein, traces of fat, traces of milk sugar, inflammatory products such as garget at times, bacteria, yeasts, molds which succeed in entering the milk, and the in- soluble dirt which may be present in the milk, or other foreign insoluble particles which may find their way into the milk, — in short, anything which may be suspended and not in solution in milk and which will respond to centrifugaHzation.' I A clarifier is a centrifuge, accordingly these terms are employed interchangeably as well as centrifugalization and clarification. CLARIFICATION OF MILK. 159 These substances which make up the slime will be subjected to indi\ddual scrutiny as progress is made. Amount of Slime eemoved. The amount of slime removed by the clarifier depends upon many factors, as may be guessed from its component parts. Besides the in- fluence of the constituents of milk, temperature, acidity or age of milk, individuality of the cow, the condition of the machine, the number of revolutions of the bowl, and probably many other factors determine the amount of slime within its total limitations or the amount which is possible within a given amount of milk. Then, too, as clarification proceeds, the character — perhaps more specifically and accurately the consistency • — of the slime changes, which is doubtless attributable to the mechanical action of the clarifier. Determination of the Weight of Slime. As a rule, in literature moist weight is employed to report the amount of slime. If conditions were identical when clarifying, the clarifiers the same, the amount of milk passed of the same measurement, then possibly a fairly representative lot of determinations could be established. This seems very difficult, however, as will be gathered later. Owing to this fluctuation in the moisture content, it is essential that the moisture be eliminated to constant weight before comparisons can be satisfactorily made and a true interpretation of the amount estabUshed. For many purposes this additional labor may be avoided and the moist weight will serve. Accordingly, it was found desirable early in the work to establish the variatiouo in the determinations of the amount of slime from different sources, since difficulty was met in the interpretation of results when based upon moist weight alone. The determinations furnished are based on clarification of milk at the same temperature, the same milk and the same age of milk, the use of the same machine, the same number of revolu- tions per minute, — in short, the same methods and procedures throughout. It is therefore a test of methods and procedures, and has its very im- portant bearing upon slime determination. The weights are always re- corded as moist or dry weight. 160 MASS. EXPERIMENT STATION BULLETIN 187. Table I, — A Determination of the Weight of Slime under Moist and Dry Conditions. [Thirty pounds of milk used for each sample; milk was held at 70° F.J Number of Test. Slime. Sample. MOIST WEIGHT IN — DRY WEIGHT IN — Grams. Per Cent, of Milk. Grams. Per Cent, of Milk. I 6.7100 6.5905 .049 .048 1.6217 1.6017 .011 .011 II 5.6425 5.7036 .041 .041 1.3136 1.3836 .009 .010 Ill, 6.7544 6.4483 .049 .047 1.6317 1.5180 .011 .011 IV 4.4049 4.0133 .032 .029 1.1150 .9540 .008 .007 V 4.8775 4.6215 .035 .033 1.1551 1.2510 .008 .009 VI 5.1382 5.0012 .037 .036 1.359S 1.2746 .009 .009 VII 5.8314 6.4810 4.8073 4.3109 .042 .047 .035 .031 1.3770 1.6286 1.0482 .9965 .010 .011 .007 .007 VIII, .>.... 6.6093 6.7741 .048 .049 1.5088 1.6839 .011 .012 IX 6.8910 6.0158 .043 .044 1.4629 1.4715 .010 .010 X 4.5792 4.2663 4.8683 4.5678 .033 .031 .035 .033 1.1793 1.0558 1.2783 1.1552 .008 .007 .009 .008 XI 6.2538 6.0309 6.1542 6.0529 .045 .044 .045 .044 1.5084 1.4379 1.4725 1.4218 .008 .010 .010 .010 XII 5.3230 5.3092 .039 .038 1.2436 1.2552 .009 .009 XIII 4.6834 4.6892 4.7127 .034 .034 .034 1.2756 1.2417 1.2526 .009 .009 .009 XIV 4.6806 4.7212 4.3300 .034 .034 .034 1.2756 1.2526 1.3900 .009 .009 .010 XV 7.1353 7.0018 .052 .051 1.9734 1.9546 .014 .014 XVI 7.0210 7.1330 .051 .052 1.9232 2.0017 .014 .014 XVII, 5.8702 5.8702 .043 .043 1.3654 1.3821 .010 .010 Literature is quite consistent in the amount of slime given off in clarifica- tion. The necessity for constant weight is evident from the preceding table if exact determinations for comparison are to be made. CLARIFICATION OF MILK. 161 The Determinations of Others. — Bahlman^ says the weight of material deposited in the clarifier from 725 gallons of milk was 2\ pounds. As an average, then, 1 gallon of milk yielded 1.6 grams of moist sludge (.044 per cent.) equivalent to .6 gram (.01 per cent.) of dried material. In his "Studies on the Clarification of Milk," Hammer^ gives the follow- ing amounts of slime secured from different lots of milk: ■ — Table II. — Amounts of Slime obtained from Different Lois of Milk {Hammer). Pounds of Milk Clarified. Amount of Slime Deposited in Cubic Centimeters. ' Per Cent, of Slime Removed. Pounds of Milk Clarified. Amount of Slime Deposited in Cubic Centimeters. ' Per Cent, of Slime Removed. 635 70 .024 953 65 .015 837 - 125 .032 1,249 125 .022 725 90 .027 1,147 250 .048 1,150 70 .013 1,356 125 .020 918 70 .016 1,241 100 .017 1,169 45 .008 There has been contributed to this theme the experience of Mcln- erney:^ — Table III. — Amount of Slime obtained from Different Quantities of Milk {Mclnerney). Experiment. Milk used (Ounces). Slime obtained (Ounces). Per Cent, of SUme. 1, 89,088 5.64 .0063 2 82,964 7.65 .0092 3 87,680 6.98 .0080 4 88,960 6.49 .0073 5 .' . . 89.088 6.80 .0076 6 84,480 12.62 .0149 7 84,480 8.25 .0091 8 84,480 6.45 .0076 » Bahlman, Clarence: Milk Clarifiers. Am. Jour, of PubUc Health, 1916, Vol. VI, No. 8, p. 856. > Hammer, B. W.: Agricultural Experiment Station, Iowa State College of Agriculture and Mechanic Arts. Research Bulletin No. 28, January, 1916. » This appears to be moist sUme measured in cubic centimeters. * Mclnerney, T. J.: Clarification of Milk. Cornell University Agricultural Experiment Sta- tion. Bulletin No. 389, April, 1917, p. 496. 162 MASS. EXPERIMENT STATION BULLETIN 187. The author states: "After all the milk had been passed through, the machine was taken apart and the amount of slime deposited on the walls was carefully removed, placed in a bottle, and weighed." He does not say whether it is moist weight or dry weight. It is apropos that the extensive work of Lieutenant Davies^ be in- serted here, inasmuch as it contributes very suggestive data. The authors present it exactly as it was found. The results secured furnish informa- tion upon slime-yield nowhere else to be found, and it has these advantages: The amount of slime is measured from milk of individual cows, and where it has been possible to point out abnormaUties this has been done. In the interpretation of Lieutenant Davies' results it will be well to keep in mind that very small amounts of milk were used, which usually leads to a high percentage of moisture in the slime; that the weight is moist weight which is subject to great fluctuation; and that the diagnosis of abnormalities appears crude because no intimate study has been made. Yet these data are far more suggestive of what is involved in the process of clarification, so far as sUme production is concerned, than can be gleaned from almost any other source. Clarification op Certified Milk (Davies). Methods. De Laval Clarifier No. 95 was used in this work, its capacity being well suited for the work, the quantities of milk from each cow being very variable and usually small. In place of the tank supplied with the clarifier a funnel was fitted so that given quantities could be easily measured. At the same time there was the advantage that every bit of milk could be passed through the bowl without rinsing with water; also no particles of dirt could remain on the side. While the latter was of no consequence with the certified milk, it does make a difference with the ordinary market mUk. Three bowls were used; this allowed plenty of time for washing and steriliz- ing them. The bowl shell was weighed while quite dry before the test. The milk was clarified immediately after being drawn, 4 quarts being used where possible; if less than 4 quarts, then all the mUk was clarified. The bowl was allowed to run down itself, any attempt to stop it quickly seemed to shake the slime film off on to the discs, and weighing was impossible. The bowl was wiped dry and weighed; the amount of slime was calculated in per cent, of milk clarified. The cows were tested as often as circumstances would aUow. No attempt was made to keep any definite order, it being found best to test whenever the 1 Lieut. E. L. Davies was connected with this department as a graduate assistant at the time this work was done. It was, however, executed independently of this bulletin and as a minor thesis. He was majoring in microbiology and pursuing dairying as one of his minors in his graduate work. He became restless when the war opened and tried many times to enter the Canadian service, but was refused on account of physical disability. He was invited by Prof. Dan H. Jones of the Ontario Agricultural College to become a member of the bacteriological staff. Remaining there for a period, and removing his physical disability at the same time, he again became restless for active service. He was accepted into the officers' school. After several months of training on this side, together with local service, he was sent to France. He experienced active service in the trenches at once. Within six weeks he was shot down by Germans whom he was making prisoners. CLARIFICATION OF MILK. 163 milk and clarifier were ready together; in tliis way no inconvenience due to waiting was caused the milker. In the table on pages 163-175 the breed of the cow is designated by the initial letter of the breed, a prefix "R" designating "registered," prefix "G" desig- nating "grade." Example, G. G., — Grade Guernsey. Ages are given in years and months approximately. Weeks in lactation calculated from the first week of lactation. Numiber of tests made, 440, with 74 different cows. 1 cow tested 11 times. . 1 cow tested 10 times. 5 cows tested 9 times. 14 cows tested 8 times. 13 cows tested 7 times. 12 cows tested 8 times. 9 cows tested 5 times. 6 cows tested 4 times. 6 cows tested 3 times. 5 cows tested twice. 2 cows tested once. Sixty-five, or 14.7 per cent., showed bloody sUme. Seventy-four, or 16.8 per cent., gave .1 per cent, slime or over. Average sUme for 440 tests, .067 per cent. Cow. Date. Weeks in Lac- tation. Milk (Pounds). Slime Cent.). Remarks. No. 1. R. J., age, 12 years, 6 months. June 16 July 1 12 14 .3.9 12.5 .090 .195 Bloody. July 8 15 12.2 .070 July 9 15 13.3 .060 July 18 16 12.0 .065 July 21 16 11.0 .145 Bloody. July 28 18 12.5 .100 Bloody. Aug. 3 19 12.0 .055 No. 19. G. H., age, 11 years. June 20 35 12.5 .115 June 24 June 25 July 10 35 35 37 9.5 10.7 9.5 .340 .187 .030 Very swollen udder, slime bloody. Swelling nearly gone, July 22 39 8.7 .060 July 29 July 30 40 40 6.3 5.3 .215 .105 Swollen udder, slime pussy and bloody. Swollen udder. Aug. 10 41 7.0 .065 Bloody. No. 21. G. H., age, 2 years, 8 months. June 9 June 16 20 21 8.0 7.4 .095 .040 June 23 22 7.4 .020 164 MASS. EXPERIMENT STATION BULLETIN 187. Cow. Date. Weeks in Lac- tation. Milk (Pounds). Slime (Per Cent.). Remarks. No. 21— Continued. July 6 24 7.4 .925 July 20 26 6.8 .015 July 23 26 6.8 .060 July 31 ■ 27 7.3 .045 No. 22. G. H., age, 14 years, June 18 - 11.0 .295 First milking. June 19 - 14.8 .292 June 20 - 14.0 .180 July 3 - 16.0 .110 July 6 - 17.9 .130 July 24 - 17.1 .015 No. 23. G. J., age, 2 years, 7 months. June 9 June 16 18 19 11.3 10.5 .052 .030 » June 26 20 10.7 .080 July 8 22 9.9 .050 July 29 24 9.6 .075 No. 24. G. J., age, 9 years. June 11 2 16.5 .115 No trouble. June 17 3 15.0 .035 June 30 5 14.8 .065 July 1 5 13.3 .045 July 31 9 ■ 13.8 .095 Aug. 10 10 13.2 .025 No. 26. G. A., age, 3 years, 4 months. June 9 June 26 55 57 9.0 8.2 .065 .050 July 6 58 7.9 .035 July 8 58 7.2 .045 July 13 60 6.4 .095 Sore teat. July 27 61 7.8 .060 July 29 61 8.1 .055 Aug. 6 62 6.8 .010 Bloody. No. 48. R. A., age, 8 years, 4 months. June 16 June 24 14 15 15.8 16.0 .090 .062 No trouble. June 30 16 15.7 .075 July 3 16 14.0 .090 July 10 17 14.9 .150 Swollen udder. July 14 17 12.6 .095 Swollen udder. July 20 18 14.2 .235 Slime bloody, udder swollen badly. CLARIFICATION OF MILK. 165 Cow. Date. Weeks in Lac- tation. Milk (Pounds). Slime (Per Cent.). Remarks. ^0.48 — Continued. July 23 18 13.7 .080 July 31 19 13.2 .020 Aug. 8 20 13.6 .020 No. 52. R. J., age, 6 years, 3 months. June 16 June 23 13 14 11.1 10.2 .075 .035 June 26 14 11.4 .025 July 1 15 10.5 .025 July 23 18 10.4 .075 Aug. 1 19 10.6 .065 Aug. 3 19 11.1 .070 No. 56. G. J., age, 6 years, 6 months. June 9 June 19 6 7 15.4 11.8 .060 .295 June 20 7 12.5 .070 June 25 8 12.5 .047 July 1 9 12.2 .055 July 20 12 11.6 .055 July 28 13 10.2 .075 Aug. 6 14 11.0 .020 No. 54. G.J.,age. June 9 30 7.0 .165 Bloody. June 18 31 6.0 .095 June 26 32 7.0 .055 July 1 33 7.1 .050 July 20 36 8.5 .025 July 27 37 7.9 .045 Aug. 1 37 6.2 .030 Aug. 10 38 3.2 .110 No. 59. G. H., age, 13 years, June 20 30 6.5 .055 June 25 30 10.5 .075 July 6 32 11.1 .045 July 18 33 10.0 .050 July 24 34 10.0 .065 Bloody. July 30 35 10.3 .035 Aug. 8 36 10.7 .085 No. 60. G. H., age, 12 years. June 15 32 7.8 .160 Hind quarter sore. June 24 33 7.3 .085 Bloody. July 1 34 5.8 .095 Pus like. 166 MASS. EXPERIMENT STATION BULLETIN 187. Cow. Date. Weeks in Lac- tation. Milk (Pounds). Slime (Per Cent.). Remarks. No. 60— Continued. July 13 36 5.2 .070 July 22 37 3.8 .045 July 31 38 5.0 .100 Bloody. No. 82. G. H., age. 11 years, July 3 - 11.0 .385 First milking. July 4 - 12.2 .260 July 21 3 15.8 .030 July 28 4 17.1 .060 Aug. 8 5 15.1 .085 No. 63. G. S., age, 10 years. June 8 23 13.5 .100 June 18 24 12.4 .080 June 24 25 12.8 .040 Bloody. July 10 27 11.7 .065 No. 64. G. S., age, 10 years. June 8 40 3.8 .138 June 15 41 3.6 .140 No. 66. G. H., age, 11 years, June 11 48 9.5 .085 June 17 49 8.8 .060 ' July 1 51 11.4 .045 July 18 53 8.8 .080 July 19 53 9.2 .095 July 28 54 8.5 .085 Bloody. No. 68. G. G., age, 5 years, 1 month. June 11 32 8.0 .085 Bloody. June 23 33 8.7 .070 July 7 36 7.2 .060 July 22 38 7.0 .040 Aug. 1 39 6.5 .055 No. 69. G. H., age, 4 years, 8 months. June 17 July 1 10 12 19.4 18.2 .140 .032 July 7 13 19.2 .105 Bloody. July 14 14 17.2 .060 July 20 15 18.0 .115 Bloody, swollen ter. Bloody. quar- July 30 16 16.8 .105 No. 71. G. H.,age. . June 8 47 20.8 .050 June 12 47 17.4 .070 June 18 48 16.0 .065 June 23 49 15.3 .070 CLARIFICATION OF MILK. 167 Cow. Date. Weeks in Lac- tation. Milk (Pounds) Slime (Per Cent.). Remarks. No. 71 — Continued. July 9 51 15.0 .015 July 13 51 18.8 .060 July 21 52 14.8 .080 July 29 54 11.9 .065 Aug. 6 55 15.1 .020 No. 72. R. G., age, 6 years, 9 months. June 11 June 17 45 46 6.0 6.0 .070 .140 June 26 47 5.0 .026 July 7 49 7.7 .075 No. 75. G. G., age, . June 9 19 13.5 .110 Bloody, sore teat. June 10 19 13.5 .100 June 15 20 11.0 .060 June 25 21 10.0 .060 July 3 22 10.0 .065 July 22 25 8.4 .055 July 28 26 9.4 .070 Aug. 10 28 5.6 .025 No. 76. G. G., age, . June 10 18 13.5 .085 Sore teat. June 19 19 13.6 .100 Sore teat. June 26 20 10.5 .040 July 1 21 10.1 .105 Sore quarter and teat. July 7 22 9.5 .050 July 20 24 9.0 .090 July 28 25 8.8 .085 Aug. 3 25 8.5 .090 No. 77. R. J., age, 4 years, 9 months. June 12 June 16 50 50 7.8 9.3 .035 .065 June 25 52 8.0 .035 July 3 53 9.5 .050 July 21 55 14.8 .055 July 28 56 8.1 .045 No. 78. G. G., age, . June 23 _ 14.8 ^ .065 July 7 - 12.2 .040 July 21 - 12.0 .065 July 29 - 11.1 .015 168 MASS. EXPERIMENT STATION BULLETIN 187. Cow. Date. Weeks in Lac- tation. Milk (Pounds). Slime (Per Cent.). Remarks. No. 80. R. A., age, 5 years. 2 months. June 24 July 3 - 9.5 8.9 .025 .040 July 8 - 8.5 .055 July 13 - 6.5 .020 July 20 - 7.5 .040 July 30 - ■8.0 .050 No. 82. G. Cage, . June 10 - 15.1 .105 Bloody. June 17 - 12.2 .065 June 23 - 11.0 .115 Sore teat. July 3 - 11.0 .075 July 9 - 10.0 .060 July 21 - 8.5 .085 July 29 - 10.1 .070 Aug. 3 - 10.0 .080 Bloody. No. 84. G. A., age, 6 years, 4 months. June 10 June 19 21 22 8.7 6.7 .075 .040 June 25 23 5.0 .071 July 3 24 6.1 .080 July 10 25 6.0 .060 No. 88. G. H., age, 5 years, 6 months. June 11 June 20 31 32 8.0 7.0 .090 .080 June 30 33 6.0 .030 July 3 34 5.2 .056 July 8 35 4.5 .046 No. 93. G. H., age, . June 19 5 10.2 .085 June 30 6 9.3 .027 July 13 8 8.5 .065 July 21 9 11.2 .055 July 31 11 8.4 .055 Aug. 6 12 7.5 .035 No. 94. G. H., age, 10 years, June 9 14 19.3 .145 June 15 15 16.2 .085 June 23 16 16.0 .085 July 6 18 13.5 .045 July 9 18 13.5 .050 July 14 19 15.1 .050 CLARIFICATION OF MILK. 169 Cow. Date. Weeks in Lac- tation. Milk (Pounds). Slime (Per Cent.). Remarks. No.9i — Continued. July 24 20 14.4 .060 July 29 21 11.6 .050 Aug. 6 22 13.0 .040 Bloody. No. 97. G. H., age, 10 years, June 30 - 12.0 .055 July 3 - 10.8 .060 July 8 - 10.4 .055 July 20 - 11.0 .035 No. 101. R. A., age, 5 years, 6 months. June 10 June 18 31 32 10.2 16.3 .020 .035 July 1 34 13.8 .065 July 7 35 10.5 .085 July 10 36 14.0 .085 July 21 37 14.0 .055 Aug. 3 38 10.0 .065 Aug. 8 39 10.0 .060 No. 102. G. A., age, 10 years, 2 months. June 12 June 26 2 4 20.8 19.2 .190 .280 Bloody. Bloody. July 10 6 18.8 .105 Bloody. July 23 8 17.5 .095 Bloody. July 31 9 17.5 .100 Bloody. No. 103. R. A., age, 11 years, 4 months. July 18 July 30 - 14.4 14.8 .045 .060 Aug. 3 - 14.3 .090 Bloody. No. 104, R. A., age 11 years, . June 8 11 8.5 .070 No. 105. G. J., age, 3 years, 9 months. June 12 June 25 18 20 6.5 7.5 .045 .030 July 1 21 7.8 .010 July 9 22 7.4 .035 July 21 24 6.8 .020 July 30 25 7.5 .010 No. 106. G. H., age, 4 years. June 19 12 14.8 .035 June 30 13 14.4 .080 July 21 16 14.2 .020 July 29 18 12.6 .020 Aug. 3 18 13.4 .055 170 MASS. EXPERIMENT STATION BULLETIN 187. Cow. Date. Weeks in Lac- tation. Milk (Pounds) Slime (Per Cent.). Remarks. No. 107. G. H., age, 3 years, 8 months. June 15 June 30 18 20 9.2 10.7 .095 .040 July 3 20 10.5 .060 July 9 21 10.0 .015 Bloody. July 27 24 10.3 .095 Bloody. July 28 24 9.4 .010 Aug. 8 25 9.0 .045 No. 108. R. H., age, 3 years, 10 months. June 17 June 19 27 27 14.5 14.5 .070 .035 June 24 28 13.5 .060 Bloody. July 1 29 13.5 .055 July 7 July 14 30 31 12.0 12.3 .105 .015 Bloody, quarter July 18 31 11.0 .055 July 27 32 11.0 .055 Aug. 1 32 11.5 .045 Aug. 6 33 11.2 .055 No. 110. R. H.. age 5 years, 4 months. June 17 June 23 33 34 12.3 10.0 .055 .070 July 1 35 10.2 .045 July 7 36 10.4 .100 July 27 38 11.8 .055 July 28 39 8.8 .080 Aug. 6 40 8.5 .095 Bloody. No. 111. G. H., age, . July 23 - 12.0 .100 Bloody. Aug. 3 - 9.0 .085 Bloody. No. 112 . G. H., age, 3 years. July 13 - 8.8 .055 July 23 - 10.7 .055 July 29 - 10.0 .080 Aug. 8 - 10.5 .050 No. 113. R. J., age, 4 years, 6 months. June 25 June 26 - 5.0 6.0 .165 .095 First milking. June 30 1 10.3 .125 Bloody. July 7 2 10.8 .060 Bloody. July 14 3 12.2 .030 July 27 5 12.4 .040 CLARIFICATION OF MILK. 171 Cow. Date. Weeks in Lac- tation. Milk (Pounds). Slime (Per Cent.). Remarks. No. 112 — Continued. July 29 5 11.3 .065 July 30 6 10.8 .125 Bloody. No. 115. G. H., age, . July 13 - 10.3 .020 July 31 - 11.5 .005 Aug. 3 - 13.8 .120 Bloody. No. 116. R. H.,age,5years. June 11 100 7.0 .150 June 15 100 7.6 .052 July 11 103 7.0 .075 July 19 104 7.8 .075 July 24 106 6.5 .045 July 29 107 6.0 .090 July 31 107 6.3 .050 Aug. 10 108 5.8 .045 No. 117. G. H., age. 3 years, 11 months. June 12 June 30 33 35 11.3 12.0 .077 .075 Bloody. July 20 38 12.8 .055 July 31 39 13.6 .040 Aug. 6 40 11.7 .060 No. 118. G. J., age, 5 years, . June 16 30 10.8 .045 June 26 31 10.3 .030 July 7 33 9.7 .020 July 23 35 7.2 .035 July 28 36 8.5 .035 No. 119. G. H., age, 5 years, 2 months. June 15 June 26 35 36 9.8 9.8 .107 .075 Sore teat. July 18 39 7.8 .040 July 19 39 7.5 .060 Aug. 1 41 6.8 .080 Aug. 10 42 6.2 .055 No. 120. G. G., age, . July 24 - 10.6 .055 Bloody. Aug. 1 - 10.1 .090 Aug. 8 - 13.1 .070 No. 125. R. H., age, 7 years, 6 months. June 11 June 16 33 . 33 18.5 19.0 .127 .235 June 17 33 18.5 .205 172 MASS. EXPERIMENT STATION BULLETIN 187. Cow. Date. Weeks in Lac- tation. Milk (Pounds). Slime (Per Cent.). Remarks. No. 125 — Continued. June 18 34 19.0 .155 No trouble foimd at any time with this June 19 34 18.0 .147 cow. June 20 34 20.0 .155 June 24 35 17.9 .185 July 1 36 16.4 .185 July 8 37 14.7 .235 July 13 37 12.9 .100 July 22 39 15.3 .120 No. 127. G. S., age, . June 12 43 13.5 .110 Bloody. June 19 44 13.1 .065 June 30 45 12.5 .065 July 3 46 12.5 .090 July 13 47 12.4 .115 July 18 48 12.0 .105 July 21 48 12.0 .070 July 29 49 12.5 .080 Aug. 10 50 12.0 .055 No. 130. G. H., age, 4 years, 2 months. June 18 June 25 25 26 14.8 13.2 .077 .055 July 1 27 12.2 .065 July 6 27 10.8 .100 July 13 28 11.0 .125 Bloody, udder swollen. July 23 29 11.3 .085 July 28 30 11.5 .090 No. 131. G. H.,age, . June 26 14.5 .115 Bloody, sore teat. July 7 16.5 .070 July 22 15.6 .060 July 30 - 12.0 .060 No. 133. G. A., age, 4 years, 3 months. June 11 Juno 16 15 15 15.0 14.5 .027 .060 June 23 17 11.3 .055 July 7 18 12.1 .060 July 21 20 11.0 .095 Bloody. July 29 21 11.5 .046 Aug. 6 22 10.8 .030 Bloody. CLARIFICATION OF MILK. 173 Cow. Date. Weeks in Lac- tation. Milk- (Pounds). Slime (Per Cent.). Remarks. No. 134. R. A., age, 3 years, 11 months. June 15 June 20 55 55 6.5 6.5 .080 .035 June 24 56 5.0 .060 No. 135. G. H., age, . July 24 - 7.5 .090 July 28 - 8.4 .130 Cow sick. Aug. 8 - 11.0 .095 Bloody. No. 136. G. H., age, . July 23 - 9.2 .140 Aug. 6 - 14.2 .070 No. 141. G. A., age, 2 years, 10 months. June 12 June 18 20 21 9.8 9.8 .095 .075 June 24 22 9.6 .015 July 8 23 8.5 .055 July 9 23 8.6 .045 July 13 24 7.0 .025 July 27 26 7.6 .035 July 29 26 8.6 .045 Aug. 10 27 8.8 .055 No. 143. G. H., age, 3 years, 10 months. June 12 June 15 8 8 9.4 8.7 .075 .075 Bloody. June 26 10 7.5 .105 July 9 12 6.5 .010 July 11 12 5.8 .070 Aug. 10 12 6.5 .050 No. 144. R. A., age, 3 years, 6 months. June 10 June 17 35 36 9.6 9.7 .105 .020 Bloody. July 6 38 8.8 .020 July 8 38 7.6 .045 July 13 39 8.5 .080 July 29 41 7.6 .035 Aug. 10 42 8.1 .055 No. 147. G. A., age, 3 years, 9 months. June 20 June 23 40 40 5.8 .040 .050 July 3 42 5.0 .076 Bloody. July 8 42 5.0 .085 July 22 44 3.8 .105 174 MASS. EXPERIMENT STATION BULLETIN 187. Cow. Date. Weeks in Lac- tation. Milk (Pounds) . Slime (Per Cent.). Remarks. No. Ii7 — Continued. July 27 45 2.8 .030 July 31 45 4.3 .090 Two miUdngs. No. 148. R. A., age, 2 years, 10 months. June 8 June 9 28 28 10.0 10.0 .072 .080 Very bloody, udder bruised. June 10 28 10.0 .085 Bloody. June 20 29 9.5 .045 June 24 30 9.4 .095 Bloody. July 3 31 7.5 .070 July 23 34 5.9 .045 July 31 35 4.9 .075 Aug. 3 35 5.3 .035 No. 149. R. A., age, 2 years, 11 months. June 19 June 23 25 25 8.2 7.6 .052 .050 July 6 27 6.9 .055 July 14 28 5.6 .055 July 20 29 7.2 .025 Aug. 1 30 6.1 .050 Aug. 6 31 6.5 .015 Aug. 8 31 6.5 .030 Blood No. 150. R. G., age, 3 years, 4 months. June 18 June 25 38 39 8.0 7.3 .080 .070 July 6 40 7.0 .055 July 18 42 6.0 .030 July 22 42 6.2 .060 July 28 43 6.0 .060 No. 152. R. H., age, 2 years, 8 months. June 9 20 11.5 .080 Bloody. June 24 22 9.8 .027 July 6 24 8.8 .010 July 9 24 9.0 .055 July 20 26 9.7 .050 July 24 26 9.3 .080 Bloody. July 30 27 9.2 .050 Aug. 3 27 10.4 .020 No. 153. G. A., age, 3 years, 4 months. June 6 June 18 28 29 14.1 11.4 .057 .070 June 26 30 11.8 .085 I CLARIFICATION OF MILK. 175 Cow. Date. Weeks in Lac- tation. Milk (Pounds). Slime (Per Cent.) Remarks. No. 153 — Continued. July 3 31 11.3 .075 July 8 32 10.4 .020 July 21 34 11.0 .010 July 30 35 9.8 .025 No. 81. R. A., age, 5 years, 1 month. June 12 58 2.5 .106 No. 154. R. G., age, 2 years, 9 months. June 11 July 6 15 18 8.0 7.8 .085 .040 Bloody. July 9 19 7.6 .040 July 23 21 7.0 .045 July 30 22 7.0 .055 Aug. 3 22 6.0 .070 No. "B." G. G., age, . June 10 - 8.4 .095 June 19 - 10.6 .012 July 3 - 9.0 .045 July 14 - 9.0 .045 July 22 - 9.5 .105 Bloody, swollen der. ud- July 31 - 8.6 .015 Aug. 10 - 8.0 .080 Bloody. No. "B." G. G., age, . June 10 - 9.3 .042 June 18 - 11.5 .040 June 25 - 11.1 .015 Bloody. July 7 - 10.9 .050 July 21 - 10.5 .060 Aug. 1 - 10.0 .015 No. 28 Aug. 6 - 19.1 .065 Aug. 8 - 17.9 .055 Fresh milking. From the figures in the preceding table conclusions may be drawn which will more or less summarize the results. It was found difficult to take figures for illustrations which were not influenced by some factor other than that under discussion. 1. Different individuals vary greatly in the amount of shme given, even when apparently perfectly normal conditions exist. The following averages of individuals illustrate this : — Per Cent. No. 125, ■ 168 No. 107 051 No. 115, ^ . . . .048 No. 64 139 176 MASS. EXPERIMENT STATION BULLETIN 187. 2. The individuals vary greatly in the amount of slime given at different milkings; in successive tests No. 107 gave .095, .04, .015 and .095 per cent. No. 26 varied even more, from .095 to .01 per cent. 3. A few cows seem to be fairly constant in the amount of slime. Nos. 125 and 118 illustrate this very clearly. 4. The amount of sUme is affected by sore teats and diseased or bniised udder. No. "B" averages .056 per cent, for two successive tests, the follow- ing test she gave .105 per cent. On inquiry of the milker it was found that the cow's udder was bruised. Nos. 48, 75, 76, 108, also illustrate this. 5. It cannot be said that large amounts of sUme indicate sore or diseased udder. No. 125 in eleven tests never gave less than .1 per cent., and no trouble could be found. Nos. 16 and 94 both gave very high tests, but without apparent cause. 6. The presence of blood in the slime cannot be said to indicate a diseased udder in so far as close examination would reveal. Bloody slime is not con- fined to cows giving high amounts of sUme. 7. The period of lactation does have an influence. Cows just freshened give a high per cent, of slime; it is often continued for several weelvs. In late lactation the tendency seems to be to give a high per cent., yet this does not always hold good. Many of the tests given in the table show that cows which have been milking for a long period give very small amounts of slime. 8. The relation between amount of milk secreted and slime is in no way clear; it is doubtful if there is any such relation. The Determinations of this Laboratory. — To Lieutenant Davies' data may be advantageously added further determinations of slime from different breeds and individual cows, together with a few determinations made upon commercial milk from different sources. One of the significant things which comes to light in these determinations, which were made incidental to other work, is the tendency to remain more or less constant over succes- sive days. This does not appear in Lieutenant Davies' work. Table IV. — Amount of Slime from Different Breeds. Certified Milk. [Five pounds of milk used.] Cow. Breed. Condition. Slime (Dry Weight in Grams). 53 Jersey, Normal, . .2041 .1732 .26931 .25901 .3387' - 77 Jersey, Normal, . .2588 .2646 .3650 .24351 .28001 .2140' 78 Guernsey, . - .2882 .3710 .2928 .3266 - - 72 Guernsey, . Abnormal, 3.8232 1.2086' .5963' .4917' .5055' - 85 Ayrshire, . Normal, . .5984 .43421 .45671 .5058' .4974' - 100 Ayrshire, . Normal, . .6171 .74941 .72071 .97931 .6715' - 30 Holstein, . Normal, . .2492 .2506 .3390 .3111 .3462 - 127 Shorthorn, Abnormal, .2020 1.7008 1.1392 1.0180 1.1713' 1.30651 Weights made on successive days. CLARIFICATION OF MILK. 177 Table IV. — Amount of Slime from Different Breeds — Concluded. Commercial Milk} [Ten pounds of milk used.] Slime (Dry Weight in Grams). Cole 1.14080 1.1881 1.2210 1.0141 1.1385 1.1423 Adams, .80275 .7946 .8231 _ - - Farm .84500 .8305 .9834 - - - From the above study it will be gathered that the amount of slime from cows of the same breed and different breeds is subject to great variation, but the daily production from a cow or from a herd, when determined on successive daj^s, appears to be quite uniform. Effect of Temperature upon the Amount of Slime. The temperature of the milk at the time of clarifying exerts some in- fluence upon the amount, as is illustrated in the accompanying tables. The cause of this is not patent unless it may be due to the coalescence of colloidal particles, thus diminishing the extent of surface of the combined particles and increasing the effect of the centrifugaUzing forces. Table V. — Effect of Temperature on Amount of Slime Removed. [Twenty pounds of commercial milk used in each test.] Sample. Temperature (Degrees F.). Slime (Grams, Dry Weight, in Duplicate). I II Ill IV, v, . . 55 75 100 55 75 100 55 75 95 55 75 95 55 75 95 1.9812 1.9474 1.9664 1.9353 1.9888 1.9800 2.0181 2.1736 2.3984 2.4226 2.6228 2.5358 1.1897 1.3322 1.5948 1.2342 1.2679 1.3168 1.3786 1.1244 1.2300 .9631 1.0524 1.4778 1 "Commercial" and "market" as applied to milk ordinary milk that is sold. used interchangeably, meaning the 178 MASS. EXPERIMENT STATION BULLETIN 187. Table V. Effect of Temperature on Amount of Slime Removed ■ eluded. Con- Sample. Temperature (Degrees F.). Slime (Grams, Dry Weight, in Duplicate). VI VII VIII IX X 55 75 95 45 75 90 45 75 90 45 75 90 45 75 90 1.4493 1.7300 1.6632 1.6493 .4210 .3735 .4485 .5093 .6140 .5840 1.0643 1.0433 1.0366 1.1601 1.3069 1.3357 .9360 .9282 1.0009 .9667 1.0092 1.0050 .9849 1.0345 1.0545 1.1404 1.1468 1.2180 Table VI. — Effect of Higher Temperatures on Amount of Slime Removed (Commercial Milk). Sample. 90° F. 110° F. 125° F. 140° F. Held 90° F. for Three Hours. I .9097 1.1783 1.2691 1.3367 1.3358 1.6268 1.6804 1.1385 II 1.0545 1.1423 Influence of Time and Acidity upon the Amount of Slime. That the elements of time and acidity operate with temperature became evident as the work proceeded. It is illustrated in the table below. CLARIFICATION OF MILK. 179 Table VII. — Effect of Time and Temperature on Amount of Slime Removed. [A single sample of commercial milk was used in this test.] Time. Temperature (Degrees F.). Grams (Moist Weight). At once, 24 hours 48 hours At once, 24 hours, 48 hours, 72 hours 90 hours, At once, 24 hours, ........... 42 42 42 68 68 68 68 68 90 90 90 90 1.1015 1.1219 1.2715 1.3034 1.2732 1.0384 1.3680 1.9330> 1.2085 1.4677 48 hours, 1.6412 72 hours 1.9322 1 High acidity. Discussion. — It is readily deducible from the above evidence that the amount of slime differs widely when secured from the milk of the same cow, from milk of different individual cows, and from mixed milks, whether the mixed milks have the same origin or not. It is also manifest from the work of this laboratory that samples from the same milk when clarified under the same conditions yield practically the same amount of slime. It follows, therefore, that the causes for these variations must be found in the condition of the animal, the condition^, which surround the manipula- tion of the milk, and the conditions which are involved in the clarification. From Lieutenant Davies' investigations it seems clear that with the beginning of the period of lactation there is a great increase of slime. This may be attributable to the colostral milk in which colostral cells are numerous. Evidence also seems to point directly to inflammatory con- ditions of the udder as a cause of increase; garget and other products of inflammation and germ action within the udder are common, probably much more so than is usuaUy recognized. As high as 20 per cent.- has been given as the average appearance of garget in milch cows. This does not seem unreasonable when one reflects on the sensitive nature of the mammary gland, and the injuries to udders so frequently encountered by milkers, giving rise to restricted or general mastitis. Doubtless the variability in cell-content must influence the amount of slime to a con- siderable extent. This may or may not be associated wdth inflammatory processes. The so-caUed fibrin may be a variable quantity. These are matters which we shall treat in greater detail later. Whether milk is dirty or clean, whether many micro-organisms are present or not, whether it is fresh from the cow or has stood for some time, whether it has been held at a low or high temperature, are all in some way related to the variation in the amount of slime obtained. Again, the clarifier itself and the manner of manipulation have a de- ' Ernst, W.: Milk Hygiene, translated by Mohler and Eichorn, p. 85. 180 MASS. EXPERIMENT STATION BULLETIN 187. cided influence upon the slime produced. Whether the machine is run at high speed or low speed, whether the temperature of the milk is high or low, whether the machine has passed quantities of milk or only a small amount, whether it is one size or another and whether it is one make or another, — all exert a modifying influence on the amount of slime thrown out. If, for instance, the amount removed when it is greater in one case than in another is to the credit and efficiency of the machine, will depend on whether the material so removed is dirt or some normal content, as leucocytes. However, it would seem that in the light of the primary purpose of a clarifier the greater the amount of sUme removed the better. This will have to be passed over, however, for it has not been the object of the writers to test the efficiency of clarifiers of different manufacturers, or even the different makes of a single manufacturer. This has been studiously avoided. Food Value of Slime. The average amount of slime estimated in terms of the entire milk is less than five one-hundredths of 1 per cent. This weight includes foreign elements, as dirt, hairs and such other materials as are likely to find their way into the milk. Only the normal elements, as the so-called leucocytes, the so-called fibrin, fat and casein, can in any sense be regarded as possess- ing food value. Inasmuch as the 3^ per cent, of fat and the 3 per cent, of casein existing in slime (see analyses below) represent only 3^ and 3 per cent, of five one-hundredths per cent, of the milk, in other words, .00175 and .0015 per cent, of the milk, the conclusion of analysts, that the food value of slime is negligible, is warranted. There is interest attached, however, to the seeming fact that the protein not only comes from the casein that is thrown out, as suggested by Mclnerney, but that it takes the form of purin bodies, too, as suggested by North. The fat also appears not only to be the fat of milk but, as Bahlman states, the fat of epithelial cells and other detritus. Evidently the cellular elements fur- nish a recognizable source of some of the material or substances found in the slime; hence, when taken together with the large number of corpus- cular elements eliminated in the slime which will be shown later, they cannot be overlooked in the interpretation of milk clarification. This raises a question at once, which, so far as the authors are aware, has not been answered : Do these cellular elements in any manner contain a con- stituent or constituents which contribute to nutrition? The work of McCoUum and Davis, ^ McCoUum, Simmonds and Pitz,- Osborne and Mendel,^ Hopkins and Neville,* and others suggests the possibility that ' McCoUum, E. V., and Davis, M.: The Nature of Dietary Deficiencies of Rice. Journal of BioL Chem., 1915, Vol. XXIII., p. 181. s McCoUum, E. V., Simmonds, E. V., and Pitz, W.: The Relation of the Unidentified Dietary Factors, the Fat-soluble A and Water-soluble B, of the Diet to the Growth-promoting Properties of Milk. Jour, of Biol. Chem., 1916, Vol. XXVII.. No. 1, p. 33. 5 Osborne, T. B., and Mendel, L. B.: Milk as a Source of Water-soluble Vitamine. Jour, of Biol. Chem., 1918, Vol. XXXIV., No. 3, p. 537. * Hopkins, F. G., and NevUle, A.: A Note concerning the Influence of Diets upon Growth. Biochem. Jour., 1913, Vol. VII., p. 97. CLARIFICATION OF MILK. 181 in these corpuscular elements there may exist what may be called nutri- tional activators, or bodies which in very small quantities are essential to body maintenance. Cheviical Analyses of Clarifier Slime. Analysis by Bahlman. * Per Cent. Protein (nitrogen X 6.38), 67.9 Fat, 3.4 Milk sugar, . . . . . . . . . . . . .7.8 Crude fiber 2.2 Silica 3.8 Oxide of iron, ............. .5 Oxide of alumina, ............ .6 Calcium phosphate, . . . . . . . . . . .3.6 Potassium phosphate, . . . . . . . . . .6.2 Sodium and potassium chloride, .......... .1 96.1 Undetermined, . . . . . . . . . . . . .3.9 100.0 Anahjsis by Mclncrney. ' Experiment. Fat (Per Cent.). Water (Per Cent.). Total Solids (Per Cent.). Ash (Per Cent.). Nitro- gen (Per Cent.). Casein (Per Cent.). 1, 3! 4, 5, 6, 7, 8, 4.0 5.0 3.4 3.2 4.0 5.0 3.7 4.0 70.13 71.86 70.04 69.92 75.50 71.01 71.35 70.87 29.87 28.14 29.96 30.08 24.50 28.99 28.65 29.13 4.17 2.73 3.81 3.00 2.74 3.36 2.59 2.83 .43 .23 .71 .14 .31 .10 .49 .27 2.74 1.46 4.52 .89 1.97 .63 3.12 1.72 A^ ^erag 3, 4.0 71.33 28.67 3.15 .33 2.13 Analysis by North. Per Cent. Total solids, 30 Fat ,3 Ash 3 Nitrogenous organic compounds, .......... 24 I Bahlman, Clarence: Milk Clarifiers. Am. Jour. Pub. Health, 1916, Vol. VI, No. 8, pp. 8.55, 856. « Mclnerney, T. J.: Clarification of Milk, Cornell University Agricultural Experiment Station. Bulletin No. 389, April, 1917, p. 499. ' North, Charles E.: The Creamery and Milk Plant Monthly, Vol. II, No. 1, p. 19. 182 MASS. EXPERIMENT STATION BULLETIN 187. We may conclude for the present, at least, that the slime cast out by the clarifier has no nutritional significance, for in amount it is negUgible and in quality value there exist no definite data. This laboratory has concerned itself with some determinations of fat in slime to ascertain whether breed or amount of slime affected the per cent, of fat present. No relation can be seen by the authors. The following tables will contribute information which makes this conclusion reason- able:— Table VIII. - — Determination of Fat in Slime from Different Breeds. Cow. Breed. Weight Slime (Dry). Per Cent. of Fat. Weight of Shme (Dry). Per Cent. of Fat. Weight of Slime (Dry). Per Cent, of Fat. Weight SlLme (Dry). Per Cent. of Fat. 53 Jersey, . .2041 4.3 - - - - - - 77 Jersey, . .2588 6.5 - - - - - - 78 Guernsey, .2928 5.0 .3939 4.9 - - - - 72 Guernsey, - - - - - - - 85 Ayrshire, .5984 3.9 .4342 3.8 .4567 3.9 .5058 3.8 100 Ayrshire, .7494 3.5 .7207 3.6 .9793 3.6 - - 30 Holstein, .3390 3.5 - - - - - - 127 Shorthorn, 1.7008 3.5 1.1713 3.6 1.3065 3.5 - - Likewise no relation can be estabHshed between total soHds of the cow's milk and the slime produced. Table IX. — Determination of Total Solids in Slime from Different Cows. Cow 127. Cow 72. Cow 100. Cow 85. Cow 53. Slime (Dry). Per Cent. of Solids. SUme (Dry). Per Cent. of Solids. Slime (Dry). Per Cent. of Solids. Wei* Slime (Dry). Per Cent. of Solids. Weight SHme (Dry). • Per Cent. of Solids. 1.1713 12.22 .2472 12.76 .7207 12.20 .4567 12.31 .2693 12.55 1.3065 11.97 .2740 12.50 .9793 11.78 .5058 12.42 .2590 12.90 1.1940 11.80 .2245 12.08 .6715 11.60 .4974 12.45 - - 1.0472 11.51 .2317 12.45 - - - - - - .8930 11.85 .2962 12.49 - - - - - - 1.6843 11.86 - - - _ - - - - 1.3548 11.90 - - - - - - - - CLARIFICATION OF MILK. 183 The same holds true when these determinations are followed over several successive days. Possibly the differences are so small that they do not become sufficiently evident against the fluctuations in the amount of slime eliminated. Table X. — Determination of Total Solids in Slime over Successive Days Thurs- day. Fri- day. Satur- day. Sun- day. Mon- day. Tues- day. Betty III: — Forenoon, . Afternoon, . Red IV: - Forenoon, . Afternoon, . Weight of slime (dry). Per cent, of solids, . Weight of slime (dry), Per cent, of solids, . Weight of slime (dry), Per cent, of solids, . Weight of slime (dry), Per cent, of solids, . .1440 14.6400 .4385 .1289 12.1290 .1043 13.7300 .4607 14.0900 .4970 14.4900 .2064 13.2800 .1551 13.0800 .4452 13.9500 .4205 .0941 12.9900 .1127 13.6600 .4418 13.5700 .4113 13.5000 .1127 12.9800 13.7300 .1082 12.7800 .4455 14.0200 Leucocytes (so-called) in Slime. That the clarifier throws out of the milk a large proportion of the so- called leucocytes is the testimony from various sources. The number eliminated, moreover, is usually determined by the examination of milk before and after clarification. It is desirable, therefore, to treat this particular subject more fuUy in connection with other corpuscular ele- ments under the discussion of milk. Determinations, however, which have been made from slime directly are quite limited because of the great possibility of error and the difficulties involved, but are helpful in arriving at a knowledge of the clarifier situation. Hammer ^ has estimated as many as 830,000,000 to 1,120,000,000 per cubic centimeter of moist slime. The estimates of this laboratory are ba^ed on certified and market milk and upon individual cow's nulk. The authors do not deem this method as accurate as the determination of leucocytes in milk before and after clarification. This attempt at determination does indicate forcibly that the cellular elements of milk make up a no mean portion of the total slime eliminated. 1 Hammer, B, W.: Agricultural Experiment Station, Iowa State College of Agriculture and Mechanic Arts. Research Bulletin No. 28, January, 1916. 184 MASS. EXPERIMENT STATION BULLETIN 18/ Table XI. — Leucocytes 'per Gram in Slime from Certified Milk. Sample. Cow. Number per Gram. Sample. Cow. Number per Gram. 1 33 104,000,000 14, . 56 90,000,000 2, 77 19,500,000 15, 77 - 3, 33 72,800,000 16, 24 20,000,000 4. 77 62,400,000 17, .33 420,000,000 5, 77 20,500,000 18, 62 670,000,000 6. 146 30,900,000 19, 146 200,000,000 7. 33 40,000,000 20, 77 330,000,000 8, 77 32,000,000 21, 56 442,000,000 9. 33 28,000,000 22, 24 390,000,000 10, 77 24,500,000 23, 62 80,000,000 11, 33 70,000,000 24, . 62 and 33 300,000,000 12, 62 - 25. 62 and 33 600,000,000 13, 146 3,000,000 Note. — The slime was macerated in a definite quantity of physiological solution and the cells determined in the suspension. All cells, however, are not released from the slime by this method. Table XII. — Leucocytes per Gram in Slime from Commercial Milk. Sample. Number per Gram. Sample. Number per Gram. 1, 2, 3 300,000,000 400,000,000 200,000,000 4 5 6 350,000,000 270,000,000 420,000.000 Note. — This slime was treated in the ! manner as in the case of certified milk. Further discussion of this subject will be deferred to the discussion of corpuscular elements of milk, on page 196. The Fibrin (so-called) in Slime. The constituent of milk which has been designated as fibrin because it responds to the methods of staining fibrin is approximately completely removed, as will be gathered from the tables given later (see page 202). CLARIFICATION OF MILK. 185 Table XIII. — Presence of Fib 'in in Slimes from Certified Milk. Sample. Cow. Fibrin. Sample. Cow. Fibrin. 1, . . . . 33 + 14, ... . 56 + 2, 77 + 15, 77 + 33 + 16, 24 + 77 + 17, 33 + 77 + 18, 62 + 146 + 19, 146 + 33 + 20, 77 + 77 + 21, 56 + 33 + 22, 24 + 10, 77 + 23, 62 + 11, 33 + 24, 33 and 62 + 12, 62 + 25, 33 and 62 + 13, 146 + The Dirt in Slime. By dirt is meant those extraneous substances which find their way into milk from without, or after the milk has left the udder. All milks, whether certified or ordinary market milk, contain some dirt. It appears, however, in different quantities in different milks, and the amount present in a gen- eral way corresponds closely to the grade of the milk. An analysis of the dirt found in or gaining entrance to milk has resulted in the recognition of definite substances associated with the cow, stable, milker or utensils. Some of the materials are feces, dust, hairs, straw, hay, epithelial cells, — in short any loose material on the cow or easily detached from the cow, the milker, the stall; substances floating in the air as the result of stirring hay or bedding or any dusty articles in the stable; material adherent to the pail; and other foreign matter reaching the milk through flies, straining, etc. In this particular connection our interests center in what the clarifier may do toward undoing what has been done in milking and handling milk. During the process of milking, as a rule, the dirt is added; then an effort is made to remove it by straining and render it harmless by pasteuriza- tion. The clarifier is now added as a means to assist in the removal of dirt. It is evident that the clarifier as a centrifuge cannot remove that portion of the dirt which goes into solution. No centrifuge can do this as long as the solution diffuses throughout the whole mass; accordingly, this should not be charged against the machine, because it is beyond the reach of any present practical device, mechanical or otherwise. 186 MASS. EXPERIMENT STATION BULLETIN 187. Table XIV. — Does an Increase in Dirt Mean an Increase in Bacteria in Clarified Milk and Water? 1. Determine by adding definite quantities of dirt to water, and esti- mate number of bacteria per cubic centimeter before and after clarifica- tion. Bacteria per Cubic Centimeter. Before. After. Sample I, .5000 gram in 1 liter, Sample II, .5000 gram in 1 Uter Sample III, .2000 gram in 1 liter Sample IV, .1000 gram*n 1 Uter 30,000 40,000 30,000 10,000 40,000 50,000 20,000 10,000 2. Determine by adding similar quantities of dirt to milk, estimating the number of bacteria per cubic centimeter before and after clarification. Adding .5000 Gram of Dirt to Milk Containing 100,000,000 Bacteria per Cubic Centimeter. Bacteria per Cubic Centimeter. Before. After. Sample I, . . . Sample II, 160,000,000 225,000,000 75,000,000 175,000,000 Adding .2000 Gram of Dirt to Milk Containing 15,000,000 Bacteria per Cubic Centimeter. Sample III, 1,000,000 Adding .1000 Gram of Dirt to Milk Containing 22,000,000 Bacteria per Cubic Centimeter. Sample IV, 40,000,000 30,000,000 A determination of the solubility of dirt was undertaken to set before the reader just the nature of the dirt problem. The first series of deter- minations was made by placing a combination of dry manure, curryings and dust of definite weight, which might get into milk easily, into water as a menstruum, then the suspension and solution were filtered or clarified. Later, milk was employed as a menstruum in place of water. CLABIFICATION OF MILK. 187 Table XV. — Determinations of Solvability of Dirt. Insoluble Dirt Re- moved by Filtration. No. 1. Grams. Weight of dirt added to 500 cubic centimeters of water, . . . . . . . 1049 Weight of dirt recovered, 0889 Weight of dirt entering solution, .......... .0160 Per cent, of soluble dirt, 16. No. 2. Weight of dirt added to 500 cubic centimeters of water, . . . . . . . 1000 Weight of dirt recovered 0798 Weight of dirt entering solution, .......... .0202 Per cent, of soluble dirt, 20. No. 3. Weight of dirt added to 500 cubic centimeters of water, ...... .2031 Weight of dirt recovered, 1700 Weight of dirt entering solution, .......... .3310 Per cent, of soluble dirt, 12. Table XVI. — Determinations of Solubility of Dirt. Insoluble Dirt Re- moved by Clarification. No. 1. Grams. Dirt added in 1,000 cubic centimeters of water, ....... .5000 Dirt recovered from clarifier, . . . . . . . . . . .4210 Dirt lost as soluble, 0786 Per cent, entering solution, 15. No. 2. Dirt added in 1,000 cubic centimeters of water, ....... .5000 Dirt recovered from clarifier, . . . . . . . , . . .4210 Dirt lost as soluble 0786 Per cent, entering solution, 16. Dry manure is evidently more soluble than the dirt used in the preced- ing tests. Table XVII. — Determinations of the Solubility of Dry Manure in Water. No. 1. Grams. Manure (dry) added to 1,000 cubic centimeters of water, ..... .2000 Manure recovered, . .1535 Manure entering solution, ........... .0465 Per cent, of solubility, 23.3. No. 2. Manure (dry) added to 1,000 cubic centimeters of water, ..... .2000 Manure recovered, ............ .1520 Manure entering solution, ........... .0480 Per cent, of solubility, 24. No. 3. Manure (dry) added to 1,000 cubic centimeters of water, 2000 Manure recovered, ............ .1501 Manure entering solution, ........... .0499 Per cent, of solubility, 24.5. 188 MASS. EXPERIMENT STATION BULLETIN 187. An attempt to add dirt to certified milk and recover or determine it after passing the clarifier was undertaken by the method of differences. This, however, is subject to the error in clarifying the same sample of milk in two lots; the possibility of such error can be ascertained by consulting page 160. Even though the same conditions are observed throughout as considered previously, except the addition of dirt, the error resulting in clarification is real, and the method of differences here used cannot be accepted as absolute. So difficult is it to extract dirt from slime and weigh it that the results must be considered as indicative only. If, for instance, an addition of a solvent to the slime for releasing the dirt is made, the solution of the dirt is increased. When 1 per cent, of KOH is added to dry manure the per cent, of solution goes to 28.5, 32.5 and 32, instead of 24 and 24.5, as in the case of water. To illustrate the results obtained by the addition of about .2000 to .5000 gram of dirt to one liter of milk, the following determinations are given: — Table XVIII. — Solvbilihj of Dirt in Milk. No. 1. 1 Grams. Slime from 1 liter of normal milk, . . . . . . . . . 2 . 2504 Slime from 1 liter of normal milk + .5000 gram dirt 2.9123 Difference representing dirt recovered, ......... .6619 No. 2. Slime from 1 liter of normal milk, . . . . . . . . . 1 . 1276 Slime from 1 liter of normal milk + .5044 gram dirt, . . . . . . 1.5519 Difference representing dirt recovered, ......... .4243 No. 3. Slime from 1 liter of normal milk, ......... 1.7432 Slime from 1 liter of normal milk + .2000 gram dirt, 1.9340 Difference representing dirt recovered, ......... .1908 1 In this case the difference represents more dirt than was added. In the above samples the certified milk or normal milk represented the minimum amount of dirt present in milk; accordingly, it doubtless had little effect on the results obtained. While it is unjustifiable to say that the amounts recovered from the slime, after the milk has had added a definite amount of dirt and has been through a clarifier, indicate the effi- ciency of the clarifier in removal of dirt, it is justifiable to infer that a portion of the insoluble part is removed. A lack of exact methods, as heretofore hinted, by which dirt is separated from the remainder of the slime precludes drawing more definite conclusions or giving more satis- factory data. The removal of dirt has been approached from another angle, which will help in understanding the nature of dirt in milk in its relation to clarifica- tion. In one instance 5 pints of commercial milk were passed through the Wisconsin Sediment Tester, using individual discs of cotton for each pint. The milk was then allowed to pass directly into the clarifier receiv- ing can and clarified immediately. The slime eliminated by the clarifica- tion was tested by macerating the slime and centrifuging. Visible amounts CLARIFICATION OF MILK. 189 of dirt were present in the bottom of the tubes. From this one gathers that the clarifier still removes dirt after the milk has been passed through the cotton disc of the Wisconsin Cotton Disc or Sediment Tester. In another instance this trial was made with 2 pints of commercial milk. Dirt was recognized after submitting the milk to the same pro- cedures as above. Evidences of dirt appeared on the clarifier bowl also. A little different form of experimentation was then adopted to demon- strate the efficiency of the clarifier in removing insoluble dirt. Definite quantities of milk were run through the clarifier; a sample of clarified milk was taken from time to time, centrifuged and examined for dirt. Table XIX gives the results of this experiment. Table XIX. — Efficiency of Clarifier in Eliminating Dirt. [All samples of milk showed presence of dirt before clarification. Claimed maximum efficiency of clarifier, 45 pounds.] Lot. Pounds of Milk. Centrifuge Test. I 10 No dirt observable. 20 No dirt observable. 30 No dirt observable. 40 No dirt observable. 50 Slight trace observable. 60 Slight trace observable. 70 Slight trace observable. 80 More dirt observable. II, 20 40 No dirt observable. No dirt observable. 60 Slight trace observable. 80 Slight trace observable. III, 20 No dirt observable. 40 No dirt observable. 60 Slight trace observable. 80 Slight trace observable. IV, 1 . 10 No dirt observable. 20 Slight trace observable. 30 Slight trace observable. 40 More dirt observable. 50 More dirt observable. 60 More dirt.observable. 70 Original dirt observable. 80 Original dirt observable. Sawdust was present. 190 MASS. EXPERIMENT STATION BULLETIN 187. It is legitimate to claim that the cotton disc in the Wisconsin Sediment Tester is as good a strainer as is employed, but it is not wholly efficient. The clarifier removes insoluble dirt which has not been removed by the tester. Again, the clarifier removes insoluble dirt to such an extent when running within its prescribed limitations that it is impossible to detect it by any methods used by the investigators. Of course, dirt which has gone into solution is beyond reclamation. It is doubtless true that the clarifier is the most efficient strainer known when the specific gravity of the dirt is not lighter than the milk. It practically removes all- insoluble dirt. Micro-organisms in Slime. It is possible to study the number of micro-organisms in the slime eliminated from milk as well as the number of micro-organisms before and after clarification. It would be better to use the slime in this determina- tion were it feasible to release the micro-organisms from the slime, since in the determination before and after clarification colonization with its difficulties interferes to such an extent as to vitiate the results. To demonstrate this difficulty in the release of micro-organisms from slime, and at the same time to indicate the micro-organisms elimmated from milk which do not reveal themselves in the counts before and after clarification, the following tables are introduced. In these efforts it is doubtful whether 50 per cent, have been made available for counting. CLARIFICATION OF MILK. 191 m g 0 0 1 1 0 ^ 1 '-<■■ § fin ^ a < § § , 1 ' ■^ ° OJ ^ - «' a § § g ' s .9 0 ta t^ H 1 r s < .8 . i i s 1 ' m 1^ g § -^ § (2 " s « «■ s p g s s§ 1 o .2 . . O O -H > -sm s § -^ § P-H fS -H CO » s o o o S § CQ r 1 1 ■ s , ~i"" i i 1 ' i M 1" 1 8 - s fS n- g § § S g ' .9 . o o « M T § 2 -^ I cu eq M '^ g 1 1 s ' i « r 1 1 s" ^ - a fl ntent before ^ ntent after cl if slime (gran acerated in p nal amount o content, whe content, whe 8 8 Z ime m origi Germ Germ % if o o ^ m 11 CLARIFICATION OF MILK. 193 Table XXI. — Releasing of Micro-organisms from Slime. Certified Milk. [One liter employed for each sample.] Bacteria per Cubic Centimeter in Milk. Bacteria per Cubic Centimeter — Sample. In First Suspension. In Second Suspension. In Third Suspension. Before clarification 10,000 5,000 500 200 After clarification, 10,000 3,000 200 100 Before clarification, 15,000 4,000 1,000 100 After clarification, 10,000 1,000 500 100 Before clarification, 2,500 2,000 1,500 150 After clarification, 2,300 1,700 500 200 Before clarification. 14,000 4,200 2,500 -I After clarification. 12,000 3,000 1,000 -' Before clarification, 4,000 2,000 500 200 After clarification, 6,000 2,000 300 100 Before clarification, 15,000 1,500 1,100 300 After clarification. 18,000 1,000 500 200 Before clarification. 500 800 400 40 After clarification. 600 2,000 300 10 Less than 100. Commercial Milk. Bacteria Weight Slime from Milk. First Suspension. Second Suspension. Sample. per Cubic timeter in Milk. Bacteria per Cubic Cen- timeter. Weight Slime. Bacteria per Cubic Cen- timeter. Weight of Slime. Before clarification, . After clarification, . Before clarification, . After clarification, . Before clarification, . After clarification, . 400,000 350,000 75,000 50,000 320,000 280,000 .9150 .9910 .8940 40,000 17,000 20,000 25,000 75,000 40,000 .0300 .0340 .0450 16,000 1,000 1,000 6,000 .0200 .0130 194 MASS. EXPERIMENT STATION BULLETIN 187. Table XX points out that, when the slime is built up to the same amount as the original milk from which it has been obtained by means of sterile physiological salt solution, the number of organisms recovered when agitated may be even more than in the original determination in the milk before clarification. It further shows that agitation has a decided effect in releasing the micro-organisms probably from both the slime and colonies, but, on the other hand, it doubtless falls very much short in its purpose. Table XXI reveals the effect of repeated maceration and agitation upon the releasing of micro-organisms from slime. Both tables seem to reveal the fact that estimates made from milk before and after clarification have little value. To bring out the results obtained by other laboratories and by this laboratory in efforts to count organisms in slime, it is pertinent to insert the following tables, but these should be interpreted in the Ught of the preceding attempts to release the micro-organisms. No other conclusion can be drawn from these figures than the most conspicuous failure to determine the number of micro-organisms in slime, and yet this is the most reUable approach available at the present time. The values secured by repeated macerations and suspensions are far in advance of any other determinations of micro-organisms. Some of Hammer's findings are as follows: — Table XXII. — Micro-organisms Found in Slime (Hammer). Pounds of Milk Clarified. Slime (Cubic Cen- timeter). Bacteria per Cubic Centimeter of Slime. Pounds of Milk Clarified. Slime (Cubic Cen- timeter). Bacteria per Cubic Centimeter of SUme. 635 837 725 1,150 918 1,169 70 125 90 70 70 45 38,000,000 830,000,000 31,000,000 1,445,000,000 710,000,000 790,000,000 953 1,249 1,147 1,356 1,241 65 125 250 125 100 675,000,000 860,000,000 435,000,000 278,000,000 680,000,000 I CLARIFICATION OF MILK. 195 Table XXIII. — An Attempt to Estimate the Number of Bacteria in the Slime Removed from Certified Milk as Produced by Individual Cows. Sample. Cow. Number of Bacteria per Gram of Moist Slime. Sample. Cow. Number of Bacteria per Gram of Moist Slime. 33 570,000 20, . 77 650,000 77 300,000 21, 56 290,000 33 30,000 22, 24 110,000 77 20,000 23, 62 145,000 77 430,000 24, 62 and 33 17,000 146 68,000 25, 62 and 33 550,000 33 50,000 26, 33 360,000 77 33,000 27, 77 200,000 33 90,000 28, 33 152,000 77 52,000 29, 77 300,000 33 50,000 30, 33 100,000 62 - 31, 77 150,000 146 45,000 32, 33 100,000 56 540,000 33, 77 500,000 77 - 34, 33 200,000 24 220,000 35, 77 100,000 33 60,000 36, 33 570,000 62 110,000 37, 77 300,000 146 580,000 Table XXIV. — An Attempt to Estimate the Number of Bacteria in the Slime Removed in Market Milk. Sample. Number of Bacteria per Gram of ! Moist Slime. 1 Sample. Number of Bacteria per Gram of Moist Slime. 3 4 5 6 8 750,000,000 ! 15,000,000 26,000,000 25,000,000 900,000 60,000,000 50,000,000 6,000,000 9, 10, 11, 12, 13, .14, 15, 35,000,000 1,500,000 4,200,000 4,500,000 3,200,000 2,800,000 4,000,000 196 MASS. EXPERIMENT STATION BULLETIN 187 The results of counting the micro-organisms in slime are therefore un- satisfactory, yet it is evident that very large numbers are imbedded in it, sufficient at times, so far as the tables are concerned, to overthrow the counts obtained in milk before and after clarification. It is only through the study of the micro-organisms in slime, and the suspension- of specific organisms which will be given later, that any adequate notion of what occurs in this respect is obtained. For purposes of illustrating the operation of the clarifier in the action on micro-organisms, the following table is furnished. Other than this little significance is to be given to results shown. Table XXV. — Bacteria per Gram of Moist Slime in the Three Seeming Layers. Sample VI. Sample IX. Sample XII. Direct, Bottom Plate, . 30,000,000 350,000,000 50,000,000 1,500,000 200,000,000 24,000,000 Direct Middle Plate 30,000,000 1,100,000 450,000,000 200,000,000 45,000,000 12,000,000 [Direct, Top [Plate 30,000,000 600,000,000 42.000,000 7,000,000 160,000,000 118,000,000 III. MILK. When milk is subjected to clarification slime is removed. What com- poses slime and what its significance is has been considered in the forego- ing discussion. Apparently the nutritional value of milk has not been materially altered so far as can be determined at present; corpuscular elements have been removed, suspended dirt has been eliminated, micro- organisms have been thrown out in large numbers. These, however, have been determined through the slime. It now remains to study the modifications of milk itself, including, as it does under natural circum- stances, all of these elements. Corpuscular Elements of Milk. The so-called leucocytes are very greatly reduced in numbers by clari- fication. This will be established by attached data. Whether this re- moval has any particular meaning -per se other than demonstrating the efficiency of the clarifier under normal or abnormal conditions cannot be stated positively in the light of our present knowledge. However, the large numbers present in inflammatory processes of the udder have a significance from the standpoint of toxic products and pathogenic micro-organisms, and accordingly may be considered objectionable. The thought, too, of enormous numbers existing in milk due to inflammation, whether local or general, is reprehensible in the same way that visible dirt affects the value. Nevertheless, in normal milk large numbers are found, but CLARIFICATION OF MILK. 197 whether they possess any inherent qualities as food value or other significance cannot at the present time be satisfactorily interpreted. The removal of leucocytes or other corpuscular elements, as colostral cells, from milk bears directly upon the interpretation of the efficiency of clarification, in that such products as garget, etc., are removed, and, further, a measure is established. The determinations made by the Biochemical Laboratory of Boston, quoted by Parker,^ by Hammer, ^ and by this laboratory, are therefore appended to illustrate the above views. Table XXVI. — Effect of Clarifying Milk on Cell Counts (Boston Biochem- ical Laboratory). Machine A working at 6,000 Revolutions per Minute. Date. Minutes Elapsed after Starting the Run. Tempera- ture of Milk at Sampling grees F.). Average Number of Cells per Field in Unclarified Milk. Average Number of Cells per Field in Clarified Milk. May 14, 1915, . . . . 5 80 17.0 9.0 25 80 12.0 8.0 35 85 17.0 4.0 45 72 17.0 4 0 47 74 - 13.0 May 18, 1915. " 20 80 4.0 2.2 50 82 4.3 2.3 65 78 13.0 3.4 75 78 8.0 2.4 85 83 6.3 1.2 120 75 3.2 2.3 May 19, 1915 20 76 13.6 6.0 50 74 6.8 5.4 60 106 8.0 4.0 115 96 7.0 5.0 May 20, 1915 20 98 7.0 4.0 50 74 6.7 4.3 80 80 27.6 10.5 90 88 . 20.2 12.6 100 78 18.2 12.0 110 72 19.0 5.0 115 . 78 17.0 1.0 ' Parker, H. N.: The City Milk Supply, 1917, pp. 257, 258. ' Hammer, B. W.: Agricultural Experiment Station, Iowa State College of Agriculture and Mechanical Arts. Research Bulletin No. 28. 198 MASS. EXPEEIMENT STATION BULLETIN 187. Table XXVI. — Effect of Clarifying Milk on Cell Counts — Concluded. Machine B working at 5,400 Revolutions per Minute. Date. Minutes Elapsed after Starting the Run. Tempera- ture of Milk at SampUng grees F.). Average Number of Cells per Field in Unclarified Milk. Average Number of Cells per Field in Clarified Milk. May 14, 1915 5 78 11.0 3.0 25 79 82.0 2.0 35 85 9.0 5.0 45 84 9.0 1.0 May 17. 1915 20 94 8.0 6.0 60 88 11.0 9.0 75 92 17.0 5.0 80 88 4.0 4.0 85 88 4.0 2.0 90 90 24.0 4.0 May 19. 1915 20 92 5.7 1.1 50 88 7.2 3.0 60 90 6.8 3.0 65 94 5.6 2.2 May 21, 1915 20 86 14.5 14.0 50 74 14.0 13.0 70 78 13.0 11.0 85 80 14.7 11.8 95 80 19.0 17.0 105 72 22.0 19.0 CLARIFICATION OF MILK. 199 Table XX\'II. — Cells per Cubic Centimeter before and after Clarification {Hammer) . Temperatcre of Milk. Number of Cells per Cubic Centimeter before Clarification. Number of Cells per Cubic Centimeter after CVification. Per Cent. of Cells thrown out. 58 266,000 206,000 23 -. 120,000 52,000 57 -. 441,000 290,000 34 -. 572,000 259.000 55 56, 407.000 227,000 44 68, 390,000 247,000 37 55, 171,000 93,000 46 46, 258,000 116,000 55 43, 276,000 220,000 20 41, 376,000 193,000 49 51, 177,000 95.000 46 44, 293,000 265,000 10 54, 448,000 140,000 69 54. 303,000 197,000 35 50, 426,000 274,000 36 61, 276,000 202,000 <>7 43, 156,000 93,000 40 60, 208,000 159,000 24 46, 832,000 226,000 73 48, 198.000 90,000 55 48. 484.000 378,000 22 48, 610.000 489,000 20 68, 282,000 152,000 46 67. 405,000 145,000 64 64. 216.000 186.000 14 60, 442,000 244,000 45 54, 209,000 158,000 24 60. 301,000 203,000 33 66, 281,000 216,000 23 59, 367,000 302,000 IS 52, 182,000 169,000 7 59, 209,000 110,000 47 73, 184,000 102,000 45 "0. 230,000 135,000 41 200 MASS. EXPERIMENT STATION BULLETIN 187. Table XXVII. — Cells per Cubic Centimeter before mid after Clarification (Hammer) — Concluded. Temperature of Milk. Number of Cells per Cubic Centimeter before Clarification. Number of Cells per Cubic Centimeter after Clarification. Per Cent. of Cells throw^i out. 70 69, . . . 62 62 68 67, 65 64 68 81 81, 77 71, 72 61, 61, 64, ... . 64, 159,000 324,000 205,000 308,000 258,000 218,000 287,000 267,000 146,000 196.000 216,000 288,000 253,000 220,000 194,000 120,000 393,000 421,000 73,000 173,000 95,000 157,000 129,000 112,000 206,000 184,000 61,000 131,000 89,000 149,000 132,000 140,000 140,000 95,000 212,000 316,000 54 47 54 49 50 49 28 31 58 33 59 48 48 36 28 21 46 25 Average, 297,481 177,442 39 Table XXVIII. — Leucocytes per Cubic Centimeter in Certified Milk be- fore and after Clarification. Sample No. Cow. Before. After. Per Cent. Reduction. 1, 33 455,000 65,000 85 2 77 26,000 11,000 58 3 33 494,000 56,000 88 4 77 440,000 234,000 46 5 77 208,000 30,000 85 6 146 117,000 13,000 88 7 33 182,000 19,000 89 8 77 141,000 11.000 92 9, 33 174.000 23.000 86 CLARIFICATION OF MILK. 201 Table XXVIII. — Lexicocytes 'per Cubic Centimeter in Certified Milk before and after Clarification — Concluded. Sample No. Cow. Before. After. Per Cent. Reduction. 10 11 12, 13, 14, 15, 16, 18 19 20 21 22 23, 24 25 77 33 62 146 56 77 24 33 62 146 77 56 24 62 62 and 33 62 and 33 163.000 260,000 150,000 81.000 340,000 31.000 97,000 520,000 80,000 55.000 21,000 364,000 260,000 200,000 370,000 200,000 21,000 21,000 13,000 17,000 35,000 13,000 17,000 190,000 26,000 13.000 7,000 39,000 26.000 25.000 52,000 20,000 87 92 91 79 89 58 82 63 68 76 67 89 90 87 90 Table XXIX. Leucocytes per Cubic Centimeter in Commercial Milk before and after Clarification. Sample No. Before. After. Per Cent. Reduction. 1 3, 4 5 6 250,000 230,000 130,000 200,000 290,000 400,000 65,000 30,000 12,000 20,000 50.000 30,000 74 87 90 90 82 92 The tables furnish an understanding of the leucocytic situation in clarification. If nothing else is to be attributed to the ejection of cellular elements, it can be safely said that the clarifier does perform its function very satisfactorily in removing normal corpuscular elements, and, further, should there be accumulations or aggregations due to inflammatory conditions, it doubtless eliminates every particle of this heavier suspended mass, inasmuch as the surface is reduced and its power to remain sus- pended long in the milk destroyed. What is gained by this act is to be 202 MASS. EXPERIMENT STATION BULLETIN 187, estimated by the general understanding that, so far as possible, all traces of inflammatory products should be removed from milk. This is to be done whether any tangible reason can be given or not at present; it is the consensus of opinion that at times, at least, these products are dan- gerous, especially the micro-organisms giving rise to them. The Fibrin (so-called) in Milk. A substance which has been designated as fibrin is visible in milk when treated with a fibrin staining process. This is almost invariably removed by clarification. It cannot be our purpose to assign to this particular substance any role other than existence, in accordance with results of staining. That such results are obtainable can be best verified by actual trial. Table XXX. — Presence of Fibrin in Certified Milk before and after Clarification. Sample No. Cow. Before. After. 1, 33 + -. 2 77 - - 3, 33 + - 4 77 + - 5 77 - - 6 146 + - 7 33 + - 8 77 + - 9 33 + - 10 77 + - 11. 33 + - 12 62 + - 13 146 + - 14 56 + - 15 ... 77 + - 16 24 + - 17 , 33 + + 18 62 + - 19, 146 + - 20. 77 - - 21 56 + - 22, 24 + - 23 62 + - 24 33 and 62 - - 2.5, 33 and 62 - - CLARIFICATION OF MILK. 203 Table XXXI. — Presence of Fibrin in Commercial Milk before and after Clarification. Sample No. Before. After. Sample No. Before. After. 1 3 + + 4, 5 6 + Micro-organisms in Milk. This particular aspect of the work seems~^to be the most popular for testing the efficiency of the clarifier, and yet it has a faulty basis which is not always considered in conclusions. Microbial counts may tell a verj^ misleading falsehood unless the full story is told and the conditions are fully understood. Several contributions have been made upon the removal or non-removal of bacteria by the clarifier. Dr. J. Arthur McClintock' divided clarifiers into three types, — A, B and C. Out of 26 tests made with type A, he obtained a reduction of 29.7 to 55.1 per cent. Out of 22 tests made with type B, he obtained a reduction of — 3.5 to 29.8 per cent. Only two instances of increase occurred among the 22 tests. These account for the — 3.5 per cent. Out of 12 tests made with type C, he obtained a reduction of — 631 to 35.9 per cent. Only in one instance among these 12 tests did he have an increase, which alone accounts for the — 631 per cent. These results are so different from those which follow that the reviewer hesitates to accept them without further data, and does not feel at liberty to accord with the deductions from his study of the different types of clarifiers. There must be influences at work which the writer failed to record. There may be gleaned an astounding statement from A. J. Hinkelmann,- in which he says: "I have found that the pathogenic bacteria commonly met with are precipitated much more readily than are the non-pathogenic." kSuch selective power on the part of the clarifier almost bespeaks super- human capacity. It also indicates that if an organism is pathogenic (which, of course, has only restricted application, depending upon species of animal affected and other conditions) it possesses a distinctive specific gravity. T)iis scarcely seems credible, although it can be understood that some organisms are heavier than others. The division, however, » McClintock, J. Arthur: An Investigation of Clarification of Milk. The Milk Trade Journal, 1916, Vol. IV, No. 6. p. 10. ■ Hinkelmann, A. J.: Micro-organic Weight. Reprint from the Illinois Medical Journal, issue of March, 1916. 204 MASS. EXPERIMENT STATION BULLETIN 18/ can scarcely be made from pathogenesis alone, if present knowledge has any weight. More may be said concerning this later, in connection with some evidence which the authors may wish to furnish. A table furnished by W. A. Stocking^ illustrates results commonly obtained with commercial milk. Table XXXII. — Effect of a Centrifugal Clarijjer upon the Germ-content of Milk (Stocking). Sample No. Bacteria before Clarifying. Bacteria after Clarifying. Numerical Increase. Per Cent. Increase. 1 2 3 4 5 ■ . . 6,000 15,000 60,000 133,000 370,000 9,000 22,000 156,000 197,000 643,000 3,000 7,000 96,000 64,000 273,000 50 46 160 48 73 The seemingly universal increase given by Stocking is not "borne out by other workers who furnish extended studies. The explanation for this may be found in the character of the milk used. Parker quotes the findings of the Biochemical Laboratory of Boston.- 1 Marshall, C. E.: Microbiology, 1917, p. 390. 2 Parker, H. N.: The City Milk Supply, pp. 257, 258. CLARIFICATION OF MILK. 205 Table XXXIII. — Effect of Clarifying Milk on the Bacterial Count (Biochemical Laboratory) . Machine A, working at 6,000 Revolutions jier Minute. Date. Bacteria per Cubic Centimeter in Un- clarified Milk. Bacteria per Cubic Centimeter in Clarified Milk. • Numerical Increase. ' Per Cent. Increase, i May 14,1915 1,700,000 1,900,000 200,000 12 1,250,000 920,000 -330,000 -26 950,000 1,500,000 550,000 58 780,000 1,200,000 420,000 54 - 1,330,000 - - Average 1,170,000 1,370,000 200,000 17 May 18, 1915 360,000 360,000 0 0 710,000 880,000 170,000 24 950,000 960,000 10,000 1 800,000 980,000 180,000 23 750,000 850,000 100,000 13 900,000 1,080,000 180,000 20 Average 745,0002 851,6662 76,666 10 May 19, 1915 1,350,000 1,220,000 -130,000 —9 1,600,000 1,300,000 -300,000 -19 850,000 420,000 -430,000 —50 950,000 500,000 —450,000 -47 Average 1,187,5002 860,000 —327,500 -27 May 20, 1915 410,000 270,000 —140,000 -34 230,000 190,000 -^0,000 -17 600,000 580,000 -20,000 -3 860,000 1,000,000 140,000 16 660,000 500,000 -160,000 -24 650,000 700,000 50,000 7 750,000 610,000 —140,000 -18 Average 594,285 550,000 -44,285 -7 This column added by the authors. ' Corrected from table. 206 MASS. EXPERIMENT STATION BULLETIN 187. Table XXXIV. — Effect of Clarifying Milk on the Bacterial Count (Bio- ^ chemical Laboratory). Machine B, working at 5,400 Revolutions per Minute. Date. Bacteria per Cubic Centimeter in Un- clarified Milk. Bacteria per Cubic Centimeter Clarified Milk. Numerical Increase. ' Per Cent. Increase. ' May 14, 1913 1,100,000 650,000 -450,000 -40 1,030,000 820,000 -210.000 —20 600,000 1,010,000 410.000 68 450,000 900,000 450.000 100 Average 795,0002 845,000 50.000 6 May 17, 1915 1,070,000 580,000 —490,000 —45 780,000 980,000 200,000 25 800.000 950,000 150,000 19 1,150,000 780,000 -370,000 -32 850,000 750,003 -100,000 -12 900,000 1,400,000 500,000 55 Average 925,000 906,666 -18,334 —2 May 19, 1915 900,000 800,000 -100,000 -11 1,110,000 910,000 -200,000 -18 780,000 660,000 —120,000 —15 870,000 930,000 60,000 7 Average 915,000 825,000 -90,000 -10 May 21, 1915 200,000 180,000 -20,000 -10 90,000 130,000 40,000 44 280.000 240,000 -40,000 -14 130.000 170,000 40,000 30 550,000 750,000 200,000 36 760,000 820,000 60,000 8 Average, 335,000 381,666 46,666 14 This column added by the authors. Corrected from table. Ill this table it will be noted that there are cases of increase and cases of decrease in the number of bacteria. In this particular this work is at variance with the conclusions drawn from Stocking's table. CLARIFICATION OF MILK. 207 Clarence Bahlman^ made eight tests of market milk in which he finds an average increase of 27 per cent. Table XXXV. — Effect of Clarifying Milk on the Microbial Count (Bahhnan). Test No. Bacteria per Cubic Centimeter. Per Cent. Increase in Raw. Clarified. Bacteria. 1 2 3 4 5 6 7 8 630.000 900,000 1,400,000 455,000 418,000 3,150,000 2,160,000 1,380.000 750,000 980,000 1,800,000 730.000 580,000 4,005,000 2,800,000 ;. 720.000 19 9 28 60 30 27 30 25 Average, 1,312,000 1.670.000 27 These results correspond closely with those contributed by Stocking. All tests have shown an increase in numbers. From Hammer 2 are gathered some modifications which give the nu- merical increase and decrease of micro-organisms in milks containing germ-contents within certain limitations. ' Bahbnan, Clarence: Milk Clarifiers. Amer. Jour, of Pub. Health, 1916, Vol. VI, No. 8. « Hammer, B. W.: Studies on the Clarification of Milk. Iowa Agr. Exp. Sta., 1916. Bulletin No. 28. 208 MASS. EXPERIMENT STATION BULLETIN 187. Table XXXVI. — Bacteria per Cubic Centimeter before and after Clari- fication {Hammer). [Original count under 100,000 per cubic centimeter.] Bacteria per Cubic Centimeter before Clarification. Bacteria per Cubic Centimeter after Clarification. Per Cent. Change in Number. Bacteria per Cubic Centimeter before Clarification. Bacteria per Cubic Centimeter after Clarification. Per Cent. Change in Number. 61,500 58,500 —5 12,700 13,450 6 70,000 61,000 —13 25,800 26,300 2 48,000 71,500 49 22,200 23,800 7 19,550 20,400 4 24,850 20,250 —19 41,000 41,000 0 8,200 7,950 -3 11,650 15,850 36 6,700 6,700 0 83.000 98,500 19 63,000 78,000 24 20,250 15,400 -24 30,500 46,500 52 35,500 31,500 —11 97,000 78,000 ■ —20 91,500 95,500 4 45,500 51,000 12 67,500 70,500 4 22,500 26,700 19 38,000 35,000 —8 16,250 17,300 6 61,000 62,000 2 19,150 20,000 4 56,500 46,500 -18 7,150 9,250 29 35,500 126,500 256 8,300 7,000 —16 24,500 24,500 0' 75,500 111,000 47 24,500 24,500 0 73,500 149,500 103 18,500 53,000 186 15,000 28,500 90 48,500 43,000 -11 37,500 35,000 -7 32,500 36,000 11 48,500 63,500 31 19,050 20,150 6 71,500 147,500 106 42,000 41,000 -2 36,500 50,000 37 7,900 6,500 -18 26,000 53,500 106 5,700 6,150 8 97,500 132,000 35 18,450 24,400 32 59,000 63,000 ■7 9,900 11,100 12 Corrected from table. CLAEIFICATION OF MILK. 209 Table XXX^'II. — Bacteria per Cubic Centimeter before and after Clari- fication {Hammer). [Original count from 100,000 to 500,000 per cubic centimeter.] Bacteria per Cubic Centimeter before Clarification. Bacteria per Cubic Centimeter after Clarification. Per Cent. Change in Number. Bacteria per Cubic Centimeter before Clarification. Bacteria per Cubic Centimeter after Clarification. Per Cent. Change in Number. 257,000 247,000 -4 450,000 345,000 -23 227,000 219,000 -4 460,000 435,000 —5 179,500 150,500 -16 190,000 392,000 106 226,000 233,500 3 365,000 450,000 23 ' 142,500 139,000 -2 105,000 141,000 34 107,000 117,000 9 141,500 177,000 25 128,000 121,000 -5 142,500 194,000 36 111,000 101,000 —9 460,000 605,000 32 101,000 64,500 —36 430,000 1,235,000 187 131,000 149,500 14 340,000 495,000 46 400,000 450,000 12 390,000 540,000 38 480,000 560,000 17 260,000 400,000 54 233.000 320.000 37 179,000 238,000 33 260,000 435,000 67 Table XXXVIII, — Bacteria per Cubic Centimeter before and after Clari- fication (Hammer). [Original count over 500,000 per cubic centimeter.] Bacteria per Cubic Centimeter before Clarification. Bacteria per Cubic Centimeter after Clarification. Per Cent. Change in Number. Bacteria per Cubic Centimeter before Clarification. Bacteria per Cubic Centimeter after Clarification. Per Cent. Change in Number. 1,185,000 1.470.000 24 970,000 705,000 —27 5,450,000 5.700,000 5 580,000 655,000 13 1,885,000 1,800,000 —5 645,000 385,000 —40 1,050,000 1,095,000 4 2,385,000 2,985,000 25 2,110,000 2,265,000 7 765,000 1,275,000 67 960,000 1,080,000 12 1,590,000 1,870,000 18 550.000 1,110,000 102 545,000 785,000 44 Fifty-one comparisons were made on samples showing less than 100,000 organisms per cubic centimeter. In 3 cases (6 per cent.) the bacterial content before and after clarification was the same; in 14 cases (27 per cent.) there was a decrease during clarification varying from 2 to 24 per cent., and averaging 210 MASS. EXPERIMENT STATION BULLETIN 187. 12 per cent.; while in the remaining 34 cases (67 per cent.) there was an in- crease during clarification varying from 2 to 256 per cent, and averaging 41 per cent. If the total 51 samples are considered there was an average in- crease of 24 per cent. Twenty-seven comparisons were made on samples containing from 100,000 to 500,000 bacteria per cubic centimeter in the unclarified milk ; 9 comparisons (33 per cent.) showed a decrease during clarification varying from 2 to 36 per cent, and averaging 12 per cent., while 18 comparisons (67 per cent.) showed increases varying from 3 to 187 per cent, and averaging 43 per cent. Con- sidering all of the samples there was an average increase of 25 per cent. Fourteen comparisons were made on samples containing more than 500,000 bacteria per cubic centimeter in the unclarified milk; only 3 comparisons (21 per cent.) showed a decrease during clarification, 1 of 5, 1 of 27, and 1 of 40 per cent, (averaging 24 per cent.), while 11 comparisons (79 per cent.) showed increases varying from 4 to 102 per cent, and averaging 29 per cent. There was an average increase of IS per cent, when the total 14 samples are considered. The number of samples of milk under 100,000 bacteria per cubic centi- meter does not show a larger percentage of decreased counts than the samples between 100,000 and 500,000 bacteria per cubic centimeter; in fact, the milk samples of over 500,000 bacteria showed a less increase than the samples with a lower number of organisms. All the samples were market milk samples; accordingly, the histories of the samples are un- known. This makes it difficult to draw any specific conclusions. Hammer's work is, however, very interesting in connection with the results of this laboratory, which will be furnished later. A general critical review of the clarifier tests has been written by Prof. E, G. Hastings for the Journal of the American Medical Association for March 24, 1917. His conclusion intimates that the clarifier may not be a progressive step in the purification of milk. This is a somewhat hasty conclusion without his having investigated the results of its action a little more closely. Too much is superficially apparent in its action to turn it aside with the wave of the hand and the cynical remark, "What next?" An extended acquaintance with the machine and its operations will at least suggest very subtle problems, perhaps much more illuminat- ing if solved than any which have been attacked thus far, and causes one to speculate about milk questions which have been heretofore untouched or remotely surveyed. From time to time these suggestions will be hinted at in the text. CLARIFICATION OF MILK. 211 T. J. Mclnerney^ has contributed the following table, which indicates the effect of clarification upon the bacterial count in fresh and old milk: — Table XXXIX. — Effect of Clarification on the Bacterial Content of Fresh Milk {Mclnerney). Bacteria per Cubic Centimeter — 1 Increase — Experiment. In Unclarified Milk. In Clarified Milk. Per Cubic Centimeter. Per Cent. 1 700 1,600 900 128.57 2 2.300 2,400 100 43.48 3 641 1,825 1,184 184.71 4 1,250 2,483 1,233 98.64 5 563 2,900 2,337 415.10 6 1,400 1,475 75 5,36 7 525 1,100 575 109.52 8 6,000 9,000 3,000 50.00 9 10,000 30.000 20,000 200.00 10, 1,100 1,400 300 27.27 11 5,000 10,000 5,000 100.00 12 4,000 4,000 0 - 13, 4,500 18,000 13,500 300.00 14 3,600 5,000 1,400 38.39 15 2,100 2,600 500 23.81 16, 3,650 5,550 1,900 52.05 17 7,000 20,000 13,000 185.71 18 5,480 12,125 6,645 121.26 19 10,000 13,000 3,000 30.00 20 11,320 13,600 2,280 20.14 21 4,280 8,000 3,720 86.91 22 4,600 4,250 —350 - 23, 1,600 4,100 2,500 156.25 24 15,000 22,000 7.000 46.67 25 53,000 71,500 18.500 34.90 26 60,000 156,000 96.000 160.00 27 5,675 5,775 100 1.76 28 10.200 11,000 800 7.84 Average 8,410 15,739 7,329 1 87.15 J Mclnerney. T. J.: Clarification of Milk. Cornell University Agr. Exp. Sta., 1917. Bulletin No. 389. 212 MASS. EXPERIMENT STATION BULLETIN 187. Table XL. Effect of Clarification on the Bacterial Content of Old and Dirty Milk {Mclnerneij). Bacteria per Cubic Centimeter — INCREA8E — Experiment. In Unclarified Milk. In Clarified Milk. Per Cubic Centimeter. Per Cent. 1, 830,000 13,900,000 13.070,000 1,574.70 2 40,000 110,000 70,000 175.00 3 494,000 6,400,000 5,906,000 1.195.55 4 133,500 197,500 64,000 47.94 5 15,000,000 30,000,000 15,000,000 100.00 6 37,800,000 40,000,000 2,200,000 5.82 7 1,500,000 3,200,000 643,000 1,700,000 273,000 113.33 8. ...... . 370,000 73.78 9 600,000 1,300,000 700,000 116.67 10, 55,000 175,000 120.000 218.18 11 19,000.000 160,000,000 141,000,000 742.10 12 248,000 425,000 177,000 71.37 13 558,750 1,863,300 1.304,550 233.48 14 190,000 237,000 47.000 24.74 15 83,400,000 91,030,000 7,630.000 9.15 16 1,590,000 1,831,000 241.000 15.16 17 4,420,000 5.700,000 1,280,000 28.96 Average 9,778,191 21,000,694 11,222,503 114.77 CLARIFICATION OF MILK. 213 James M. Sherman^ has also furnished his results of the bacterial counts before and after clarification. Table XLI. — Effect of Clarification on the Bacterial Count of Milk (Sherman). Bacteria peb Cubic Centimeteb — Test No. Machine. Before Clarification. After Clarification. 1 A 3,700 6,100 2 A 3,800 6,300 3 A 5,500 8,500 4 A 2,900 6.300 , 5 A 4,200 6,200 6. . . A 4,100 6,200 7 A 3,400 7,400 8 A 3,900 6,100 9 A 3,400 4,900 10 A 3,000 4.900 11 A 3,200 6,800 12 . . . A 4,300 9,600 13 B 3,300 5,600 14 B 6,900 7,300 15 B 9,300 13,800 , 16 B 4,800 7,600 17 B 1,800 3,100 18 B 2,500 3,300 19 B 2,900 3,700 20 B 11,400 13,400 21 B 4,300 6.400 22 B 3,600 4,500 23 B 10,300 13,400 24, B 7,800 9,300 Average, . . . . - 4,720 7,120 Again there is the decided increase of micro-organisms following clari- fication. Although realizing that the usual interpretation of microbial counts in this connection has no basis in actual truth, and there can be no increase because the milk passes through the clarifier so quickly that there is no ' Sherman, James M.: Bacteriological Tests of Milk Clarifier. Jour, of Dairy Science, 1917, Vol. I. No. 3, p. 272. 214 MASS. EXPERIMENT STATION BULLETIN 187. time for multiplication, and, further, in the slime large masses of organ- isms are found, this laboratory has felt it desirable, nevertheless, to undertake the determination of the number of organisms in milk before and after clarification, not so much for the purpose of contributing to what has already been given, but rather for the purpose of knowing what is really involved in the determination and what interpretation of the results obtained may be given. Since the operation of clarifying is so short, it is difficult to believe that any multiplication takes place, as has already been stated above. If none takes place then it must be a dis- ruption in colonies, which leads the student to wonder whether there is greater efficiency in micro-organisms liberated from a disrupted colony as compared with the same organisms imbedded in the colony. This will appear later. The authors' studies were carried out under the following conditions: — The clarifier used was No. 98 De Laval. It was run by a ^-horsepower motor at uniform speed of 7,200 to 7,300 revolutions per minute. The temperature was maintained at 60° C. when clarifjdng. As soon as the machine reached full speed the milk was passed through. The bowls, discs, etc., were sterilized in an autoclave at 15 pounds pressure for thirty minutes. The milk both before and after clarification was thoroughly mixed prior to taking the samples, which were placed in sterile flasks. In the case of certified milk, the milk was obtained from the milker in the "certified" stable; in the case of the commercial milk, from the receiving room of the college dairy. The commercial milk came from the farmers in the vicinity of the college, and was not above the average commercial milk. It doubtless reached the clarifier sooner than it would had it been sent to a city from Amherst, then clarified after reaching the city. For estimating the number of bacteria in milk, the Standard Methods of the American Public Health Association were employed. An effort was made to adhere to these methods in all of our work so far as feasible. A determination of the number of bacteria cast out by the clarifier into the slime has been undertaken both by a direct count, mathematical cal- culation, and by repeated maceration and clarification. Methods and discussion will be reserved until after some facts have been placed before the reader. CLARIFICATION OF MILK. 215 Table XLII. — Bacteria in Certified Milk from Individual Cows before and after Clarification. Sample No. Cow. Number of Organisms in 1 Cubic Centimeter of Un- clarified Milk. Number of Organisms in 1 Cubic Centimeter of Clarified Milk. Per Cent. Increase. 77 33 77 33 62 146 56 77 24 33 62 146 77 56 24 62 62 and 33 62 and 33 33 77 33 77 33 77 33 77 5,000 1,100 4,000 2,000 1,100 1,700 4,000 1,600 12,000 9,000 4,000 4,000 1,500 3,800 11,000 3,600 100 500 1,000 2,000 4,000 1,900 2,000 100 5,000 1,700 1,500 1,300 1,000 3,000 1,500 3,000 2,000 1,000 3,000 2,200 5,000 6,000 8,000 3,000 1,100 1,200 5,000 9,000 1,200 500 400 1,200 800 1,000 1,100 600 200 6,000 1,000 500 2,000 1,000 1,500 1,300 600 1,000 —60 —27 76 —45 212 —50 —11 -25 —72 —20 31 -75 -42 —70 100 20 —41 216 MASS. EXPERIMENT STATION BULLETIN 187. Table XLII. — Bacteria in Certified Milk from Individual Cows before and after Clarification — Concluded. Sample No. Cow. Number of Organisms in 1 Cubic Centimeter of Un- clarified Milk. Number of Organisms in 1 Cubic Centimeter of Clarified Milk. Per Cent. Increase. 34 33 700 1,900 171 35 . . 77 700 5,000 614 36, 33 5,000 2,000 -60 37 77 1,100 800 —27 Table XLIII. — Bacteria in Commercial Milk before and after Clarification. Sample No. Number of Bacteria in 1 Cubic Centimeter of Unclarified Milk. Number of Bacteria in 1 Cubic Centimeter of Clarified Milk. Per Cent. Increase. 1 250,000 900,000 260 2 100,000 200,000 100 3. 75,000 65,000 -13 4, 20,000 50,000 150 5 5,000 12,000 14 6 125,000 - 70,000 —44 7 130,000 400,000 207 .8. 25,000 48,000 92 9 20,000 35,000 75 10 350,000 250,000 —28 11 30,000 40,000 33 12 40,000 50,000 25 13 30,000 20,000 —33 14 10,000 10,000 - 15 16,000 33,000 106 The following superficial and provisional conclusions may be drawn from these tables: — 1. In the case of fresh certified milk about 70 per cent, of the tests give an increase in bacterial content in unclarified milk over the same milk clarified. This leaves 30 per cent, showing an increase after clarification. 2. In the case of commercial milk about 85 to 90 per cent, show an CLARIFICATION OF MILK. 217 increase in the bacterial content after clarification over the same milk unclarified. 3. The slime sediment reveals a deposit of bacteria which of course must come out of the milk undergoing clarification (see page 190). There seems to be a tendency, which is not universal because the milk from different cows varies so, for milk at the time of milking (70 per cent, of the cases) to undergo a reduction in the number of bacteria after clari- fication as revealed by plating, while milk which stands increases in its number of bacteria after clarification in direct proportion to the time that it is permitted to stand before clarification. This would indicate that fresh certified milk is freer from colonies and has a greater number of single organisms, and these single bacteria are thrown out with the slime (see "Slime," page 195), in some cases to a considerable extent. In certain instances, however, colonies have formed and are disrupted, thus increasing the bacterial content of certified clari- fied milk (30 per cent, of the cases). The commercial milk appears to admit of so much colonizing with the subsequent disruption by the clarifier that a high percentage (85 to 90 per cent.) of samples will give an increased number of bacteria after clari- fication. Since a large number of bacteria is found in the slime, and there is little opportunity for multiplication during the process of clari- fication, the increase in the number of bacteria is only apparent and not real. Thus far we are substantially in accord with the report of the Biochem- ical Laboratory of Boston, Hammer and Balilman. Assuming that micro- organisms have no time to multiply, it follows that although a count- increase is evidenced by the plating method, the number is actually reduced by those appearing in the slime. Serial Counts of Micro-organisms in Clarified and Unclarified Milk over a Period of Time. Together with the single bacterial counts of milk before and after clarifi- cation should be considered two-hour counts of milk, certified and market, unclarified and clarified, extending over seventy-two hours. This study will give a more precise knowledge of the effect of clarification upon the germ-content of milk in spite of the errors creeping in from colonization and plating. It will be seen at once that the graphs depict a situation not revealed by the single count before and after clarification, and they corre- spond more closely with actual experience. This taken together with other factors, as the character of fermentation resulting from clarification (see page 240), has great significance. 218 MASS. EXPERIMENT STATION BULLETIN 187. 29 w w ) """ ~~ ■^ — — r~i — — — _io m JfiO >_ _c. BU Eil lU ill „ . 5 )oo )09 ) 2 >00 )00 ) ' ~ 2 )00 ) I m )po ) 1 m )W 50(> )00 ) // IW 1P9 ) / P^ &Q m 1 / ^ /• IQ\ m 1 / '/" / 9 poo b / $ ?w [) / 1 e > ] poc <>0f 0 \\ f s'KM 0 1 1 f (J0( o 1 • fKK 0 1 oof 0 / ■KM 0 1 1 , f 00( 0 / 1 « SQi 5 1 / i 00( 0 ,^ i rf>0( 0 1 J 09( 0 / 1 i .oc 0 / / * 0O( f) ' f^fX 0 f Of'f 0 / 1 / ,W 0 / / / n ; 0 / J 0 / { 1 VfK 0 / 1 1 finr 0 1 50C 0 / V)C 0 1 :inr n V V -' ^^ y 1 rmr lfUl f1fl4 Wil ?m 0 -- V \ / \ / 1 fifl Cilil m M^ IK ^_ _ lor 0 ix- -^ ^ _ I cip llf 0 aid ilin -- -- 1 ) 1 1 1 I 1 > 1 J 2 » 2 ! 2 \ 2 ) 3 ! 3 ? * ) 4t 4) 5 ^ ® ) 7 ? HO irs of PI ^ l« L 220 MASS. EXPERIMENT STATION BULLETIN 187. £( 09( 00< 0 r— ~ K 00( 00< 0 .. C! £1. FU D 1 ID .00< 00< 0 sex 09( 0 00( 00( 0 / 50(lOCK 0 // 00(iOO( 0 '/• 50to0( 0 A' 100 '/ i50» ;; c £0|)0 }i 15000 M i tOfiJO ,'i 5 55o|)0 socio £5060 ' [ 20000 / LSOpO LOOpO 90 po 1 80 I TO » i 1 ! m lO ' f>0 10 ' / 40 )<} ^ -1 .-^0 )0 /■ 1 . [Jdar ifl BCi Mil t ?/i )n 7 X \ ^ ''^ 1 / nrit Ind Ml 11: 10 (O ;> 411 4- 4 i C .6 1 7 J : ! Ho ir3 of PI iti ^ 1 L_ CLAEIFICATION OF MILK. 221 ?Q K10 >oo , " 10 )00 )00 - ( ERl IF^ ED mi K - _ ,' 5 )00 )00 y P. ino )no / / "T 2 )00 )00 ) 1 ion 100 '' y m )W ^ ^ / / iOO )00 / r f/ 100 )00 1 / ^ y ^ )0O ' / 7 10 )00 J / , 9 )00 I / / ■ 8 500 ) 1 ^ / 6 )00 b !— - 7 iOO . I )00 ) 1 6 JOO 1 -- 6 )00 3 9 ;oo ) 5 )00 ) i iOO ) 4 )00 ) ^' 3 500 0 ^ 3 )00 3 ^ ^__ / \ / y 2 500 3 ^z' ■^ ^ / ^ £ 000 0 1 / \ jl 1 500 ) / ■7^ \ ^ 1 : 000 0 // / ^ '\ r- 900 0 '1 800 3 if TO^J 0 rf 50(^ ) 7 sod tod 0 ;inf D In? _ 2or ) ni' 1fir= ^Y IOC 9 nif rif iec' IK 0 an( 31 im( -- ( 0 : 2 1 4 ] 6 1 s : 0 2 2 ; 1 J 0 ; 2 ; ; 4 ) 4 i 4 3 5 1 6 } 7 2 . H( an of PI atl >lg 222 MASS. EXPERIMENT STATION BULLETIN >87. £( 00( 00( 0 K 00( 00( 0 _. c: ■xr: FIl 3D1 III 00( OCX 0 EOo( )00( 0 00( 100( 0 50( i00( 0 101 loa 0 5( I00( 0 11 innc 0 / iOO( 0 ' A IfiCX 0 ^ 7 ino( 0 j / 50( 0 r / '00( 10 / J I ,riO( 10 J t ,00( 10 '/ ' ;50( lO // i iOO( 10 / 1 .501 10 ^. OOi 10 / 'I o if.n 10 1 inn in / / 1 M) lO / / 1 ■•on 10 A .•in )0 / / / on in / / in m // 1 80 10 / J vn lO 1 / 1 60 10 \ J \ / / 50 10 \ i \ / ^ ^■^ 40 10 \ / \ ; [^ A A , 7 7.n lO I -/ -- 7 -- \" / IJn Tfi If- nrt Mi It ?,n »0 \ / * - Olf r! ie M Ik _ 10 10 -^ -. ^ ClJ rl M IK 0 d llr le -- ■- ) ) 1 0 1 2 1 4 1 6 1 8 2 ) 2 2 2 I 2 8 ; Z 3 5 4 3 4 i 4 8 5 1 6 } 7 . HC ora ol PI iti n« 1 _ CLARIFICATION OF MILK. 223 pi m 10 pi t) r a lOOi 100( 0 _ Ci MMi sc AL MI. K • _ 100 00( 0 lOO iOOi 0 j^ 11 /00( 10 0( 0 -/ -"X f -^ ;i ?*' iOOi )00: 0 h / ""■ .00( I00( 0 ill >00 )0C( 0 / \' ■"- If lOOl )00( )0 //" \ /. K)0 lOO: )0 / ■^/ / \; /I i50 100( )0 / s V i \ 1 ;on )OUI 10 ^^ A r .50 rjo: lO 00 00 )0 50 )00 10 / 40 )00 )0 iff 30 )00 >0 ll 20 )00( 10 ^^^ Z— 7 10 )00. >0 5 )00 )0 1 )00 10 iOO )0 150 )0 iOO< )0 ?50 )0 ?00 )0 ir>o )0 ;oo )0 >r>o )0 iOO )0 rSO )0 too )0 ?50 )0 KIO 10 ^fir K) -A ino lar ifi id Ill J :oo )0 \^H •If \fA Ml IK _ ISO 10 'rift •{f .ed Mi L)S LOO )0 an 1 S Lirt — ■- • 1 ) 1 '. 1 t 1 ) 1 i a ) 2 \ 2 2 1 3 ; 3 , 4 1 4- 4 i 5 t 6 ) 7 u _ _ — n Ho irs of PI Iti IL _ _ _ _ 224 MASS. EXPERIMENT STATION BULLETIN 187. 50 300 )00 ) 40 TOO )no 1 _ - C im 2ffi lAL MI JC _ ?o ■)on )0f) \ "" 20 wo )00 ^ 10 100 )00 s DOO )00 6 DOO )00 100 jOO 4 100 )on ) 3 )M 3 )00 2 poo )00 1 fiOO ino , / 1 poo )0C ) / / pii')r 100 ^ \ / / WO )00| \\ ^ wo J / ^ Ron )00 ) ' / V <■ LOO )00 ) /i / / I 50 )09 ) / ,^ \~ ° TO 10 r > r-V > ^ 9 100 ) \ i 8 300 ) / \ ft 100 ) \ /• \ 7 iOO ) \ /| 7 )00 ) \ /^ 6 500 ) \ ? f. wo r) I \ \ 'F 5 500 ) 1 \ V )/ M" ) ^ /\ 4 I5Q0 ) \ / ■''' 4 000 ) - '' '~ '\ / 500 :> / 000 D / s • / ? 500 T \ \ /; / IJh :1a -if Led Mi lie ? 000 :> T' y \ CI iri fie L M .lie 1 500 T . fll irl rip i W IlK 1 000 0 a Id Ui le -- -- c f 1 ) 1 ', 1 V 1 7 1 1 ? > ? '. ? » 2 i r '. 3 i 4 ) 4- I 4 i 5 \ 6 ) 7 Ho irs Of PI iti 16 CLARIFICATION OF MILK. 225 5C 00 00( 0 r ' 4<; ooc ooc 0 cJ wa Rc; AL Mil K . _ 5C 00'- ooc 0 2C OOf ooc 0 1 ooc ooc 0 ooc ooc 0 00 i()0( n A ooc: ooc 0 / \j / — ^ \ 7 ^ // 00^ ooc 0 - sr A; 7 ^. i J' ^-- 1 oo(i ooc 0 / \^ /,' — 4\ •, 1" ::rt ■^ ^ •— -^ 50(! ooc 0 / r \ 7 ■" ' ^ \ ooc ooc 0 \ / Bo IDOC 0 Ji oo; lOOC 0 .' 5o; OOi 10 // 401 iOO( 0 ^ -- 7 30C 00( n ■^ / ' 20( «0( 0 X ^ 10( ooc 0 5( iCOl 0 ■I IC ooc n - lonc 0 " ino( in ,oa 10 , ■50( lO '00( .0 ,50C 0 001 lO , ifiOC 10 , iOO( 10 r>o( 10 oo; '0 ir.o( in 100( 10 :50( 10 Unc la tf ed MiJ y- :oo( 10 01 ri ip M llr ^ poi it> nif r11 i« M in- 00( 10 ai d lli -■ — 1 1 ) 1 ?. 1 V 1 i 1 i 2 1 2 > 2 ^ 2 H 7. : 3 i 4 >. V 4 ,r> t 6 ) 7 ; HO ITS of PI itl ig - _^ 226 MASS. EXPERIMENT STATION BULLETIN 187. 5< 09( epJo ^ —r- ■n ^ — 1 1 n n n ~ 4( 00( oodo ■ c 101 asc M. lE « S< 00( ooio ?( 58! oodo ;< oo; 00(| 0 QtX M<5 0 - 00( ooc 0 ,0(X ocx; 0 ■ oo; Of) iO ifini QO '0 . nrx nfn iO 1 Of^ IMl( 1(1 1 50( og '0 y'' / 00( 00( in - 1 ■in' no in ^ w 00 tJ > / 30( 00 10 / ?;o( |00 lO / / / 101 )0n( 10 y / ," z 'On >0 A / K — 1 ^ — — ,/i ii m )0 / / ; k ("iO in / / / OO »o / / ^ .— 1— / / fon m / ; \ ."in m f; 1 inn m / / r >^ )0 ; f lon m / [ t'in m / V V" - / wo< fO 1 ^ / / \ / (■"in '0 1 / ,X mn ^0 / .... / ^ / Tin aa. •1f ed IJl,- k _ _^ '00 in / "" r m \r\ iFI M IV- h- _ / '0 / a Id li » L, — 1 1 ■^ 1 > 1 I 1 1 1 ? 1 ? ' ? 1 ? 1 ;i ' 7\ \ 4' '> 4 ' ^ 1 f^ . R( 7 \ Ho ^3 of PI iti « ... _J CLARIFICATION OF MILK. 227 50 m )0C> } ~" "" *9 m )0(;^ ) _ C m^ EC AL MXJ ,K ■ . f!0 100 inn f) 20 30d )00 E 10 300 )00 D ;^ a )00 )00 0 7 iw JW ? / ^bm 1 4)00 hOQ ? j ; 3)00 .00 3 / 4.W )O0 ? , y / £| )09 P^O / "^x ,-' / 1 500 pno // / 1 lov pw 3 i^ r^ JOO poo 3 !*r ~ i:' / \ _, IftO POO T j "^ }00 300 3 / e. ;m 100 P / f «. 100 POO 3 ^ '^' s 5 POOj ) « ft 00 3 5 500 fi \}m 1 M 500 _^QQ , Roni ? 100 2 ■>rirt , ITn In 1f fHl MV , 2t)00 , 01 irl •\r» M lY 1 ion< ) r,l T )PQ K d ! 11. IS — .- - r j ' f 1 1 ? 1 1 1 ^ 1 f1 ? 1 ? ' ? t ? 1 ;^ ■ ii i 4 1 4 I 4 n r> b <> > 7 '. HO irs of PI iti 3« The conclusion may be drawn from these graphs that there is no great distinction to be made between clarified and unclarified milk so far as bacterial counts are concerned. Yet when the character of change is contrasted, microbial influences are patent as between the unclarified and clarified samples. Incidental questions having more or less relation to the previous dis- cussion may arise. Some of these questions have been anticipated in our work, and have been added as illuminative material. 228 MASS. EXPERIMENT STATION BULLETIN 187. Table XLFV. — A Determination of the Number of Bacteria per Cubic Centimeter in Clarified and Unclarified Commercial Milk Held at lJj° C. and Plated at Intervals of Twenty-four Hours. Sample. 24 Hours. 48 Hours. I. . II. Ill, IV, V, . VT, VII, VIII. IX. X, XI. Before clarificatior After clarification, Before clariication. After clarification, Before clarification, After clarification, Before clarification. After clarification. Before ckrification, After clarification. Before clarification, After clarification. Before clarification, After clarification, Before clarification, After clarification. Before clarification. After clarification. Before clarification. After clarificafton, Before clarification. After clarification. 36,000 34,000 3,600 2,100 28,000 39,000 2,500 2,350 7,750,000 6,340,000 4,000,000 2,740,000 500,000 450,000 330,000 240,000 4,500,000 4,000,000 1,500,000 1,200,000 1,750,000 3,100,000 9,958,000 9,950,000 2,300 3,200 40,000,000 27,400,000 14,600,000 19,600,000 20,400,000 13,500,000 14,500,000 10,000,000 21,200,000 19,200,000 20,000,000 25,000,000 5,600,000 5,200,000 67,100,000 50,600,000 210,000,000 220,000,000 230,000 190,000 539,000,000 465,000,000 201,000,000 209,000,000 120,000,000 160,000,000 41,200,000 40,000,000 340,000,000 210,000,000 500,000,000 420,000,000 250,000,000 250,000,000 80,000,000 40,000,000 7,500,000 15,000,000 350,000,000 400,000,000 25,000,000 39,000,000 752,000,000 441,000,000 400,000,000 187,000,000 237,000,000 135,000,000 166,000,000 100,000,000 750,000,000 450,000,000 650,000,000 560,000,000 Table XLV. — A Determination of the Number of Bacteria per Cubic Centi- meter in Clarified and Unclarified Certified Milk Held at 10° C. and Plated at Intervals of Twenty-four Hours. Test. Sample. . At Once. 24 Hours. 48 Hours. I Before clarification. After clarification, . 940 580 1,000 1,050 900 600 II Before clarification, After clarification, . 1,450 4,200 2,200 3,700 2,400 4,600 III. Before clarification, After clarification, . 1,800 2,600 1,740 2,500 2,000 3,200 IV. Before clarification, After clarification, . 980 810 1,150 860 1,000 650 V Before clarification, After clarification, . 1,400 1,200 1,100 1,750 1,000 1,200 VI Before clarification. After clarification, . 4,000 3,000 4,000 3,100 4,300 2,500 VII Before clarification, After clarification. . 5,000 4.000 4,700 5,000 2,300 1.400 VIII Before clarification, After clarification. . 3,000 4,100 2,300 2,000 3,500 2,500 CLARIFICATION OF MILK. 229 Incidentally only, it is interesting to note the effect of repeated clari- fication upon the same sample. From this it may be seen that neither the slime nor bacteria are removed to such an extent that repeated clari- fication will not eliminate more bacteria and more slime. Table XLVI. Effect of Repeated Clarification on Bacterial Count of Same Sample of Market Milk. Bac- teria Cubic Centi- meter in MUk. ""if" Slime in Grama Second * Clarification. Third Clarification. FOTTRTH Clarification. Bac- teria Cubic Centi- meter. Slime in Grams. Bac- teria cr4 Centi- meter. Weight SltLe in Grams. Bac- teria Cubic Centi- meter. Weight SUme in Grams. Before clarification, . After clarification, . Before clarification, . After clarification, . Before clarification, . After clarification, . 50,000 70,000 7,000 17,000 25,000 18,000 3.122 2.091 3.265 74,000 48,000 25,000 11,000 9,000 22,000 1.379 1.002 1.315 48,000 40,000 11,000 22,000 1.236 .927 .865 40,000 .925 In connection with the single and serial bacterial counts it will be pertinent to study also the effect of clarification upon specific organisms in different substances, for in this manner a possibility is furnished of gaining some adequate notion of how the clarifier acts in centrifuging out certain types of organisms. Table XLVII. — Effect of Clarification on Pure Cultures of Bacteria. B. subtili.s. Suspended in — Before Clarification. After Clarification. Result or Per Cent. Removed. First Test. Water Broth (A. P. H. A.) Second Test. Water Broth (A. P. H. A.) Skimmed milk Whole milk, ....... 105,000 95,000 107,000 75,000 90,000 95,000 92,000 7,000 15,000 48,000 2,000 18.000 25,000 56,000 93.3 84.3 55.2 97.0 80.0 74.0 40.0 230 MASS. EXPERIMENT STATION BULLETIN 187. Table XLVII. — Effect of Clarification on Pure Cultures of Bacteria — Continued. B. coli. Suspended in — Before Clarification. After Clarification. Result or Per Cent. Removed. First Test. Water 480,000 118,000 76 Broth (A. P. H. A.) 465,000 115,000 75 Skimmed milk, 495,000 375,000 28 Whole milk 530,000 320,000 40 Second Test. Water 370,000 90,000 76 Broth (A. P. H. A.) 395,000 135,000 66 Skimmed milk, 315,000 215,000 31 Whole milk 400,000 280,000 30 B. cyanogenes. B. megatherium. Water 10,000 3,000 70 B. subtilis. Water, 70,000 10,000 55,000 85 Gum tragic-water, . . . . 70.000 22 B. subtih s. Specific Gravity. Suspended in — Before Clarification. After Clarification. Per Cent. Removed. 1.000 1.003 1.005 1.009 Water Water+1 per cent, gelatin, . Water+2 per cent, gelatin, Water+4 per cent, gelatin, . 100,000 138,000 110,000 120,000 5,000 25,000 40.000 48,000 95 82 64 60 B. subtilis. 1.000 Water 65,000 3,000 95 1.003 Water+l per cent, sucrose, . 115,000 5,000 94 1.011 Water+3 per cent, sucrose, . 126,000 13,000 89 1.023 Water+6 per cent, sucrose, . 95,000 12,000 87 1.026 Water+8 per cent, sucrose, . 103,000 18,000 83 CLARIFICATION OF MILK. 231 Table XL VII. — Effect of Clarification on Pure Cultures of Bacteria — Continued. Streptococcus pyogenes. Suspended in — Before Clarification. After Clarification. Per Cent. Removed. First Test. Salt solution, Whey solution, Certified milk Second Test. Salt solution, Whey solution Certified milk Third Teat. Salt solution Whey solution, Certified milk 2,120,000 1,750,000 2,600,000 2,370,000 2,000,000 1,900,000 2,000,000 1.750.000 1,400,000 370,000 550.000 2.100,000 345,000 850.000 1,870,000 450,000 700.000 1,000,000 83 19 85 59 16 77 60 27 Staphylococcus albus. First Teat. Salt solution. 130.000 10,000 91 Whey solution. 175.000 44,000 75 Certified milk. Second Teat. 705.000 285.000 59 Salt solution. 800,000 44.000 94 Whey solution, 420.000 143,000 66 Certified milk, 870,000 460,000 47 Third Teat. 1,200,000 400,000 950,000 180,000 82,000 540.000 93 Whey solution, ....... 79 Certified milk 43 232 MASS. EXPERIMENT STATION BULLETIN 187. Table XL VII. — Effect of Clarification on Pure Cultures of Bacteria — Continued. B. prodigiosus. Suspended in — Before Clarification. After Clarification. Per Cent. Removed. Salt solution, First Test. 700,000 400,000 1,350,000 3,100,000 2,290,000 3,400,000 830,000 2,000,000 1,970,000 230,000 170,000 1,100,000 840,000 1,250,000 2,600,000 220,000 1,400,000 1,370,000 67 Whey solution, Certified milk, Second Test. 57 18 72 Whey solution. Certified milk, Third Test. 45 22 73 Whey solution, 30 30 B. iumescens. First Test. Salt solution, 22,500 3,000 71 Whey solution, 13,000 6,000 53 22,000 12,500 43 Second Test. Salt solution. . 31,000 21,500 32,000 6.000 10,000 10,000 80 Whey solution, ....... 53 68 Third Test. Salt solution. 40,000 3.000 92 Whey solution, 20,000 5,000 75 Certified milk 77,000 16,000 79 CLARIFICATION OF MILK. 233 Table XLVII. — Effect of Clarification on Pure Cultures of Bacteria — Concluded. Suspended in — Before Clarification. After Clarification. Per Cent. Removed. First Test. Salt solution Whey solution Certified milk Second Test. Salt solution Whey solution Certified milk Third Test. Salt solution Whey solution Certified milk 6,200,000 4,590,000 5,510,000 2,800,000 1,800,000 2,800,000 1,300,000 1,650,000 2,895,000 1,230,000 4,300,000 4,355,000 440,000 1,600,000 2.590,000 440,000 1,270,000 2,750,000 80 6 21 84 11 7 66 23 5 Streptococcus lacticus. First Test. Salt solution, 4,500,000 1,500,000 66 Whey solution. 3,500,000 1,300,000 63 Certified milk. Second Test. 1,000,000 800,000 20 Salt solution. 700,000 120,000 82 Whey solution. 720.000 60,000 91 Certified milk. 600.000 540,000 10 Third Test. Salt solution, 400,000 35,000 91 Whey solution, 430,000 70,000 82 Certified milk, 1,060,000 600,000 43 Note. — 1. Salt Solution. — Prepared by adding 8.5 grams of sodium chloride to 1,000 cubic centimeters of distilled water. Sterilized by autoclaving at 15 pounds for thirty minutes. 2. Whey Solution. — Prepared from whey secured from the college dairy. Egg albumin was added to the whey, and heated for two hours in the flowing steam. It was then filtered clear through filter paper. To this was added 1 per cent, of bacto-gelatin and sterilized intermittently. 3. Certified Milk. — Fresh certified milk secured from the college herd. 4. In each of the experiments 1,000 cubic centimeters of the material was employed. The pure culture under test was added directly from a twenty-four-hour milk or broth culture, after the quantity of culture to be used had been deterrriined. 5. The specific gravity and viscosity of the whey menstruum were approximately that of cer- tified milk, as determined by preliminary experiments with pyknometer and viscosimeter. 6. Room temperature in which experiments were conducted varied from 19° to 23° C. so that clarification was conducted within this range of temperature. 234 MASS. EXPERIMENT STATION BULLETIN 187. Table XL VIII. — Effect of Clarification on Pure Cultures of Molds and Yeasts. Rhizopus nigricans spores. Before Clarification. After Clarification. Test No. 100 Dilution. 100 Dilution. 1,000 Dilution. 1 2 3 10, or 1,000 per cubic centimeter, 30, or 3,000 per cubic centimeter, 11, or 1,100 per cubic centimeter, 1, or 100 per cubic centimeter. Sterile,! Sterile SterUe. SterUe. Sterile. Penicillium glaucum spor( 10, or 1,000 per cubic centimeter, 40, or 4,000 per cubic centimeter, 20, or 2,000 per cubic centimeter. Sterile, 1, or 100 per cubic centimeter, 2, or 200 per cubic centimeter. Sterile. Sterile. SterUe. Oidium lactis spores. Before Clarification. After Clarification. Test No. 1,000 Dilution. 100 Dilution. 1,000 DUution. 1 2 3 14, or 14,000 per cubic centimeter, . 24, or 24,000 per cubic centimeter, . 11, or 11,000 per cubic centimeter, . SterUe 4, or 400 per cubic centimeter, 1, or 100 per cubic centimeter, SterUe. SterUe. SterUe. Saccharomyces cerevisice. 1 200, or 200,000 per cubic centimeter. 1, or 100 per cubic centimeter. SterUe. 2 370, or 370,000 per cubic centimeter. 1, or 100 per cubic centimeter. SterUe. 3 120, or 120,000 per cubic centimeter. 3, or 300 per cubic centimeter. Sterile. Aspergillus nige.r spores. 16, or 16,000 per cubic centimeter, 7, or 7,000 per cubic centimeter, 5, or 5,000 per cubic centimeter. 6, or 600 per cubic centimeter, 4, or 400 per cubic centimeter, 1, or 100 per cubic centimeter. Sterile. SterUe. SterUe. Note. — Molds were grown in pure culture; spores were swept up with sterile filter paper and introduced into 1,000 cubic centimeters of sterile milk. After thorough agitation milk was clari- fied under sterile conditions. Counts were made immediately before and after. Cultures of Saccharomyces cerevisice were grown on wort medium at room temperature for three days; 5 cubic centimeters of the culture were inoculated directly into 1,000 cubic centimeters of sterile milk. After thorough agitation, milk was clarified under sterile conditions. Counts were made immediately before and after. ' "Sterile" means that no colonies appeared when plates were made of the dilutions indicated. ' Oidium was grown directly in sterile milk at room temperature for three days, untU small colonies appeared on surface. CLARIFICATION OF MILK. 235 Table XLIX. — Effect of Three Clarifications on Pure Cultures. Streptococcus pyogenes. First Clarification. Second Clarification. Third Clarification. Per Cent. Re- moved. Before. After. 1 Before. After. Before. After. First Test. Salt solution, 2,120,000 370,000 370,000 80,000 80,000 14,000 99 Whey solution, . 1,750,000 550,000 550,000 250,000 250,000 75,000 95 Certified milk, . 2,600,000 2,100,000 2,100,000 - 800,000 69 Second Test. Salt solution, 2,370,000 345,000 345,000 78,000 78,000 15,500 99 Whey solution, . 2,000,000 850,000 850,000 325,000 325,000 100,000 95 Certified miUc, . 1,900,000 1,870,000 1,870,000 1,200,000 1,200,000 900,000 52 Third Test. Salt solution, 2,000,000 450,000 450,000 95,000 95,000 22.500 98 Whey solution, . 1,750,000 700,000 700,000 370,000 370,000 110,000 93 Certified milk. , 1,400,000 1,000,000 1,000,000 600,000 600,000 400,000 71 Staphylococcus albus. First Test. Salt solution, 130,000 10.000 10,000 1,200 1,200 100 99 Whey solution, . 175,000 44.000 44,000 7,500 7,500 1,600 99 Certified milk, . 705.000 285.000 285,000 147,000 147,000 56,000 92 Second Test. Salt solution. 800.000 44.000 44,000 3,300 3,300 200 99 Whey solution, . 420,000 143.000 143,000 31,000 31,000 3,600 99 Certified milk, . 870,000 460.000 460,000 260,000 260,000 126,000 85 Third Test. Salt solution. 350,000 31,000 31,000 1.500 1,500 100 99 Whey solution, . 400,000 82.000 82,000 21,000 21,000 4,000 99 Certified milk, . 950.000 540.000 540,000 350,000 350,000 75,000 92 236 MASS. EXPERIMENT STATION BULLETIN 187. Table XLIX. — Effect of Three Clarifications on Pure Cultures Continued. B. tumescens. Suspended in — First Clabification. Second Clarification. Third Clarification. Per Cent. Before. After. Before. After. Before. After. Re- moved. First Test. Salt solution, . 31,000 20,000 20,000 1,000 1,000 70 99 Whey solution. . 20,000 2,000 2,000 6,000 6,000 60 99 Certified milk, . 33,000 70,000 70,000 8,400 8,400 4,100 87 Second Test. Salt solution. 22,500 3,000 3,000 1,000 1,000 550 93 Whey solution, . 13,000 6,000 6,000 500 500 150 98 Certified mUk, . 32,000 10,000 10,000 2,500 2,500 1,300 96 Third Test. Salt solution, 40,000 3,000 3,000 750 750 200 99 Whey solution, . 20,000 5,000 5,000 600 600 500 97 Certified milk, . 77,000 16,000 16,000 4,800 4,800 2,400 97 First Test. Salt solution. 6,200,000 1,230,000 1.230,000 350.000 350,000 90,000 98 Whey solution, . 4,590,000 4,300.000 4,300,000 1.850,000 1,850,000 660.000 83 Certified milk, . 5,510,000 4.355,000 4,355,000 4,000.000 4.000.000 3.625,000 32 Second Test. Salt solution. 2,800,000 440,000 440.000 210.000 210,000 50,000 98 Whey solution, . 2,775,000 2,375.000 2,375,000 1.130,000 1,130,000 430,000 84 Certified milk, . 2,800,000 2.590.000 2,590.000 2.400.000 2,400,000 1,900,000 32 Third Test. Salt solution, 1,300,000 440.000 440.000 62,000 62,000 26.000 98 Whey solution, . 1,650,000 1.270.000 1,270.000 620.000 620,000 320,000 80 Certified milk, . 870,000 1,700.000 1,700,000 950,000 950,000 750,000 14 CLARIFICATION OF MILK. 237 Table XLIX. — Effect of Three Clarifications on Pure Cultures ■ Concluded. B. prodigiosns. Suspended in — First Clarification. Second Clarification. Third Clarification. Per Cent. Re- moved. Before. After. Before. After. Before. After. First Test. Salt solution, 144,000 14,000 14,000 13,000 13,000 1,600 91 Whey solution, . 153,000 99,000 99,000 24,100 24,100 15,000 90 Certified milk, . 1,100,000 1,100,000 1,100,000 600,000 600,000 420,000 61 Secmid Test. Salt solution. 700,000 230,000 230,000 82,000 82,000 17,000 97 Whey solution, . 400,000 170,000 170,000 90,000 90,000 47,000 88 Certified milk, . 1,350,000 1,100,000 1,100,000 800,000 800,000 550,000 59 Third Test. Salt solution. 830,000 220,000 220,000 75,000 75,000 30,000 96 Whey solution, . 2,000,000 1,400,000 1,400,000 630,000 630,000 340,000 83 Certified milk, . 1,970,000 1,370,000 1,370,000 1,312,000 1.312,000 , 796,000 59 Streptococcus lacticus. First Test. Salt solution. 4,500,000 1,500,000 1,500,000 375,000 375,000 125,000 97 Whey solution, . 3,500,000 1,300,000 1,300,000 1,000,000 1,000,000 400,000 88 Certified milk, . 1,000,000 800,000 800,000 360,000 360.000 300,000 70 Second Test. Salt solution. 700,000 120,000 120,000 36,000 36.000 8,400 98 Whey solution, . 720,000 60,000 60,000 20,000 20,000 1,500 99 Certified milk, . 600,000 540,000 540,000 300,000 300,000 80,000 86 Third Test. Salt solution, 400,000 35,000 35,000 12,000 12,000 1,200 99 Whey solution, . 430,000 70,000 70,000 40,000 40,000 10,000 97 Certified milk, . 1,060,000 600,000 600,000 60.000 60,000 50,000 95 238 MASS. EXPERIMENT STATION BULLETIN 187. Table L. — Streptococci Suspended in Milk Subjected to Clarification. I. Bacterial count of whole milk J Before clarification, 33.,000\., , . 1- <■ J J- . i • ^ AT^ .ex- io^r,^ >ol per cent, decrease, before adding streptococci. ^After clarification, 16,000 J Bacterial count of same milk fBefore clarification, 29,000,000 \ . after adding streptococci. \After clarification, 36,000,000 J II. Bacterial count of whole milk f Before clarification, 75,000 1 . before adding streptococci. \ After clarification, 120,000 / Bacterial count of same milk fBefore clarification, 2,000,000 \ . after adding streptococci. \After clarification, 3,700,000/ Colonization of Bacteria in Milk. Little can be stated with any degree of assurance concerning coloniza- tion of bacteria in milk. That colonization occurs, and that the degree of colonization is irregular in different milks, can be attested in several ways. One of these methods is set forth in what might be wisely designated as the provisional conclusions offered by many of the workers who have determined the number of bacteria before and after clarifying, assuming that the increased count is due to the breaking up of the colonies formed. This is, of course, indirect evidence, and must be regarded as tentative until something more direct can be provided. Little is known of a definite character concerning what bacteria wiU do in this respect, so that any conclusions based upon this precarious factor may go far astray. Knowl- edge of exact value upon this subject is almost entirely lacking. Again, the tendency of bacteria to grow into colonies is daily recognized, and yet there are conditions of cultures which do not favor such developments. What can be said about milk, and to what extent does the colony vitiate our crude plating methods and our comfortable conclusions based on them? This is important and is made conspicuous by a shroud of ignorance. Efficiency of the Individual Organism Free and in Colony. This leads to the next step, which is also of significance. Does the individual organism in a colony exercise the same degree of physiological efficiency as when the organism is alone and acting in an individual role? We are told by Mclnerney^ that bacteria increased more rapidly in unclarified than in clarified milk, yet a greater degree of change, as the production of acid, is recorded in the milk influenced by clarification than in the check culture unclarified and uninfluenced. This also occurs in a pure culture of lactic bacteria when shaken. This suggests, possibly, that per individual the clarified culture is doing greater work. What values shall be attached to the individual germ free as against the same germ in a colony? This we must know if we are going to interpret 1 Mclnerney, T. J.: Clarification of Milk. Cornell Univ. Agr. Exp. Sta. Bulletin No. 389, April, 1917. CLARIFICATION OF MILK. 239 milk clarification, provided the present explanation which accounts for the increased number of bacteria after clarification is tenable. At present our knowledge is too restricted to draw stable conclusions. Other Considerations. Centrifugal force has been repeatedly and commonly employed to eject micro-organisms when in suspension, which is the case in hand. Its values for this purpose are in a very general way understood. From the largest micro-organism with limited surface as compared with its content, to the minutest with its extensive surface as compared with its content, there seems to exist a gradation in effectiveness. In other words, the large organisms are easily ejected, while the minutest are with difficulty cast out. In the case of some of the invisible viruses the capacity to produce disease is not reduced materially by centrifugalization. In the foregoing tables it is apparent that the larger micro-organisms, as the spores of Oidiuvi lactis and the cells of Sacch. cerevisice, respond readily to centrifugal force, while such organisms as B. prodigiosus respond poorly. Likewise, colonies seemingly act as large and small cells. Again, it is well known that micro-organisms contain a variable amount of fat, as B. tuberculosis. Fat is easily determined, too, in varying amounts in mold and yeast cells when subjected to certain conditions of growth. The presence of fat must influence the specific gravity of cells, which in turn is closely related to results from centrifugalization. The age of a microbial cell, or the stage of development, is also bound up with its specific gravity, due probably to the degradative changes taking place. This is easily seen in the development of a culture when the old cells settle to the bottom. It is verv' evident from physical laws that the material in which micro- organisms are suspended has a very important and peculiar influence in their sedimentation by mechanical force. Milk, with its higher specific gravity and viscosity, acts as a deterrent in the removal of micro- organisms by centrifugalization, as is clearly evidenced by the preceding tables for specific organisms. In spite of deterrent influences referred to, micro-organisms are removed from milk in as large quantities as 75 per cent, and over. Inasmuch as the plate colony-counts probably represent colonies removed from milk, the percentage may rise much higher. The results presented in the preceding tables, in which the work of the clarifier upon specific organisms is shown, have an illuminating bearing on the action of the clarifier in its practical application to market milk. In considering micro-organisms in milk it is necessary to remember the "ebb and flow" of species. All who are students of milk have learned that in the course of fermentation-development certain types of micro-organisms in milk gradually reach the crest of their growth then gradually decline in numbers, as the rise and fall in numbers of the many species which are present in fresh milk, and which practically dis- 240 MASS. EXPERIMENT STATION BULLETIN 187. appear as conditions change. This is also discernible in the ascendency and decline of the lactic group followed by other types which appear and disappear, leading finally to complete decomposition of the milk. This "special growth-curve" which appears when conditions are favorable is a factor in clarification, for by this mechanical act the conditions for microbial development are apparently somewhat altered, and accordingly there is resulting a more or less 'kaleidoscopic change. It follows, there- fore, that an additional factor to those already controlling the stages of alteration or fermentation in milk has been introduced, naturally jdelding somewhat different changes in the course of milk fermentation. The removal of large numbers of bacteria by clarification, as has been established, must exert some influence upon the changes which take place in the clarified milk. Especially will this be true if the types which yield more readily to centrifugalization are cast out in large numbers and the types which seem to respond but poorly remain behind. The balance of growth equilibrium is disturbed. When conditions of growth are so complex as in mUk, it can at once be surmised that owing to the great variation in the germ-content of milk, both in numbers and kinds, the results must be widely different. It seems that there ought to be evidence which will correlate this great change in germ-content with alterations in clarified milk as different from unclarified. It will not be possible to furnish all of our data at the present writing. Only such evidence as has led us into a more intimate study of these changes will be given. When unclarified and clarified milk of the same original sample is permitted to stand for some time at low temperature (15° C), so that the fermentation changes appearing do not rush by unnoticed, visible altera- tions are evident. The precipitated casein resulting from such a fer- mentation may be collected then on a sterilized filter paper, and, after covering carefully, allowed to stand at ordinary temperatures for some time. The difference in the fermentation changes of the unclarified and clarified milk casein is usually strikingly manifest. This demonstrates that in the unclarified milk and casein there exist organisms which pre- ponderate over those in the clarified milk and casein. Hence the clarifier has ejected certain types of organisms in sufficient numbers to control the character of the fermentation in the clarified milk and casein. Whether these changes can be explained by the elimination of Oidium lactis and other molds and yeasts (see page 234) cannot be definitely stated at present. These observations have induced us into undertaking to demonstrate the factors involved in these differences. To this task our energies have been directed, and some of the data are at present available, but it is felt that the answer should be given as a single answer and as cornpletely as possible. CLARIFICATION OF MILK. 241 IV. SUMMARY. 1. It is evident that our present knowledge of clarification does not enable us to reach a scientific interpretation. 2. An intimate study of clarification not only reveals facts which assist in its understanding, but also leads us into depths beyond our reach. It is constantly presenting suggestions concerning milk investigations which have not been considered heretofore through established channels. A fertile field for research is opened. 3. The slime eliminated and the comparison of the clarified milk with the unclarified seem to offer, at the present time, the best approach to the study of clarification. 4. The amount of slime eliminated from milk is variable, and dependent upon — The condition of the cow, whether normal or abnormal. The period of lactation. The age or freshness of the milk. The acidity of the milk. The temperature at the time of clarifjung. The amount of corpuscular elements. The amount of insoluble dirt in the milk. 5. The food value removed from milk through the elimination of slime may be disregarded, unless there are contained within some of the ele- ments of the slime nutritional activators, as the so-called vitamines, which seems improbable. 6. Masses of cells are thrown out in the slime. This is especially em- phasized when any inflammation exists in the udder. Garget, existing as it does in ropy, tenacious form, is completely ejected. What significance is to be attached to normal cells, so far as the authors are concerned, cannot be stated from our present knowledge. 7. A fibrinous material responding to fibrin stains is practically wholly eliminated from milk in clarification. 8. Practically all insoluble dirt is removed in clarification. The clari- fier is the most effective strainer employed in the diary. Its efficiency in this respect is greater than that of the cotton filters of the Wisconsin Sedimentation Tester. Dirt in solution, of course, is not subject to the action of a centrifuge or clarifier, inasmuch as it diffuses throughout the whole mass. 9. Micro-organisms are found in large numbers — yes, in masses — in the slime. These come from the milk, since there is no other source, and there is not sufficient time to multiply while passing through the clarifier. In certified milk there is also a reduction shown after clarification, as revealed by the plating method. In market milk the number is usually Increased after clarification, as revealed by the plating method. This is doubtless due to the disruption of colonies. Besides the above evidence there are the results of repeated clarification of milk and pure cultures, 242 MASS. EXPERIMENT STATION BULLETIN 187. the action of clarification upon pure cultures, and the results secured by- direct counts, — all of which testify to the elimination of micro-organisms by the clarifier in no small degree. No differentiation between pathogens and non-pathogens can be made. The larger the micro-organisms, speak- ing generally, the greater the proportion cast out. 10. Frequently, yes, commonly, the action of the clarifier upon the micro- organisms is so significant as to alter their respective power or capacity for change in the milk. This is easily detectable by the appearance of clarified and unclarified samples when observed from day to day over a period of time. It is also readily determined by filtering out the curd, when formed, upon filter paper, and allowing it to undergo fermentation for a few days under proper conditions. In Part II we shall consider this alteration in clarified milk as com- pared with unclarified milk. The work has progressed to a point that it is safe and only fair to say that an intimate study is confirming the general statements above. BULLETIN No. 188 DECEMBER, 1918 MASSACHISETTS AGRICILTIRAL EXPERIMENT STATION The NiTRiTioN of the Horse By J. B. LINDSEY Part I of this bulletin contains very brief conclusions of many important investigations conducted chiefly by French and Ger- man scientists. The results of the earlier work (previous to 1890) were based largely upon the relation of digestible nutrients to maintenance and work performed, while that of more recent times has been based to a greater extent upon the application of calorimetry to the intake and outgo of energy. Part II of the bulletin contains the practical conclusions of feeding experiments with alfalfa, brewers' dried grains, velvet bean feed and linseed meal, together with suitable combinations for work horses when these feeds are used as components of the ra- tion. A full description of the feeding trials follows the general conclusions. Requests for bulletins should be addressed to the Agricultural Experiment Station, Amherst, Mass. CONTENTS. Part I. PAGE Some results of important investigations: — A. Early investigations, ......... 243 B. Recent investigations and the application of calorimetry, . . 246 C. Svmamary of investigations, ........ 247 D. Books on horse nutrition, ........ 249 Part II. Feeding trials with horses : — Results and suggestions, ......... 250 A. Alfalfa for horses 252 B. Brewers' dried grains for horses, . . . . . . . 257 C. Velvet bean feed for horses, ........ 259 D. Linseed meal as a grain supplement for horses, .... 262 Publication of this Document approved by the Supervisor of Administration. BULLETII^ ]Sro. 188. DEPARTMENT OF CHEMISTRY. THE NUTRITION OF THE HORSE. BY J. B. LINDSET. Part I. SOME RESULTS OF IMPORTANT INVESTIGATIONS. A. Early Investigations. Much work has been done, especially in Europe, concerning the prin- ciples which underlie the nutrition of the horse, and many experiments made to test the practical application of the knowledge secured. Among the Europeans who have studied these matters most thoroughly may be mentioned Boussingault ; Baudement; Sanson; Grandeau, LeClerc, Bal- lancey and Alikan; Lavalard; Miintz and Gerard; Wolff and Kellner; Ziintz, Hagermann and Lehmann. In the United States many experiments have been made concerning the most suitable feeds and feed combinations for horses. Worthy of especial mention is the one conducted by McCam- bell of the Kansas Experiment Station ^ with the government horses at Fort Riley. The early investigations were based largely on the anal3^sis and digesti- bility of the feeds fed and the relation of digestible nutrients to mainte- nance and work performed. Some of the more important conclusions, in- cluding particularly the modifications of rations and methods of feeding are mentioned below. 1. Of the total food consumed, -fo is needed for maintenance in a state of repose, ^2 for bodily repair, and ^2 for work performed; or 1^ for maintenance in repose and yj for bodily repair and work. (Grandeau-Lavalard.) 2. Work of Grandeau and his associates, 1882-94. (a) Maize was utilized in varying proportions with oats, depending upon the time of year and relative cost. (6) Straw was gradually substituted for hay, followed finally by the complete removal of the hay. (c) Beans were fed in place of brewery by-products. » Bui. No. 186, Kans. Exp. Sta. 244 MASS. EXPERIMENT STATION BULLETIN 188. (d) Limited amounts of oil cakes were used in the ration. (e) The nutritive ratio was widened from 1: 4.5 to 1: 7.1. (/) Glucose was found to be completely digested; starch, from 76 to 98 per cent. ; cellulose that could be hydrolyzed, from 40 to 68 per cent. ; and crude cellulose, from 32 to 58 per cent. (g) The average horse of 1,000 pounds needed for — Maintenance, at rest, .76 pound of digestible protein plus 8.8 pounds of carbohydrates (including fat multiplied by 2.4) which contains 15,000 to 16,000 calories and has a nutritive ratio of 1: 10 to 1: 11. Maintenance and repair, 1 pound digestible protein plus 9.9 pounds di- gestible carbohydrates (including fat multiplied by 2.4) which con- tains 20,000 calories and has a nutritive ratio of 1:10. Light Work, 1.3 pounds digestible protein and 11.6 pounds digestible carbohydrates (including fat multiplied by 2.4) which contains some 25,000 calories ha\'ing a nutritive ratio of 1:7. This amount is suf- ficient for horses doing 500,000 kilogrammeters of work daily. 3. Experiments in the French army. The following nutrients were found to be needed per 1,000 pounds of live weight as a result of experiments made by military officers on French army horses in 1887-89, the ration being composed largely of oats and hay: — Time of peace, 1.1 pounds digestible protein plus 10.8 pounds digestible car- bohydrates, having a nutritive ratio of 1:9. Time of war, 1.35 pounds digestible protein plus 10.8 pounds digestible car- bohydrates ; nutritive ratio of 1:8. 4. Lavalard found that omnibus and hack horses needed 1.45 pounds digestible protein plus 10.4 pounds digestible carbohydrates; nutritive ratio of 1 : 7 per 1,000 pounds live weight. 6. A few of Wolff's conclusions may be mentioned (1876-85). * (a) For maintenance of a 1,100-pound horse on hay alone, 23.1 pounds were required containing 1 .26 pounds of digestible protein and 9.25 pounds of total digestible organic nutrients, with a nutritive ratio of 1: 6.3. (b) An average day's work for a farm or draft horse of 1,100 pounds, in good condition, is 2,000,000 kilogrammeters, which requires 5.09 pounds of digestible nutrients plus 9.25 pounds for maintenance, or a total of 14.34 pounds containing 1.90 pounds protein and having a ratio of 1:6.6. (c) When fed an average quantity of hay exclusively, a 1,100-pound horse cannot take over 26.4 pounds, and can do but little work on such a diet. The addition of some clover hay enables the horse to do about one-fourth of a day's work, while if given a full diet of alfalfa, 26 pounds, the horse is able to do fully one-half an average day's work. (d) The ordinary food for the work horse is like amounts of hay and oats (13 pounds of hay and 13 pounds of oats for a 1,100-pound horse). The proportions of each can be varied, depending upon the amount of the work required. (e) The carbohydrates furnish the chief source of heat and energy for the horse. (/) One kilo of oats (2.2 pounds) added to a work ration enabled the horse to do substantially 530,400 kilogrammeters more of work, and 1 kilo of maize, 700,000 kilogrammeters. Maize proved a very satisfactory food to improve the weight and appearance of the horse. (g) The horse bean when fed in an amount not exceeding 2 pounds daily proved quite satisfactory as a source of increased protein in the ration, but as a source of energy it hardly equaled oats. 1 Grundlagen f. d. rationelle FUtterung des Pferdes, 1886. THE NUTRITION OF THE HORSE. 245 The above results and others that could be cited were based largely upon digestible nutrients in the foods fed and their relation to work per- formed, and did not take into consideration the energy expended in digest- ing the different kinds of feeds resulting in the loss of varying amounts of heat, nor the heat radiation resulting from the increased metabolism caused by certain feedstuffs. B. Recent Investigations and the Application of Calorimetky. The development and application of calorimetry, and its use in studying the intake and outgo of energy, has proved of great help in increasing our knowledge of the principles of nutrition and the nutritive value of animal feeds. The following calorimetric units and methods are employed in measuring the utilization of energy: — (a) The Calorie. — The heat which is given off by a food when combined with or burned in oxygen is the measure of its total energy. The unit of energy is termed the calorie, and represents the amount of heat required to raise 1 kilogram of water 1° C. (Armsby has recently introduced the term therm, or larger unit, meaning the amount of heat necessary to raise 1,000 kilograms of water 1° C.) According to Stohman, Berthelot and Riibner the heat units, or number of calories, in 1 gram of protein or car- bohydrates are 4.1, and in fat, 9.3, and the total energy of a food is the amounts of protein and carbohydrates multiplied by 4.1, and of fat mul- tiplied by 9.3. (6) The Kilogrammeter. — This represents the mechanical equivalent of a definite amount of heat, and is equal to the energy required to raise 1 kilogram of water 1 meter high. A calorie of heat is equivalent in mechanical energy to that requked to raise 427 kilograms 1 meter high (or 427 kilogrammeters), and this unit is called kilogram-calorie. To convert digestible protein, carbohydrates and fat into kilogram- meters, multiply the grams of protein or carbohydrates by 4.1, and the fat by 9.3, and these products by 427. (c) The Respiratory Quotient. — The relation of the oxygen consumed to the carbon dioxide given off has been termed by Pfliiger the respiratory quotient, and is determined by dividing the volume of the carbon dioxide by the volume of oxygen. In case of carbohydrates, glycogen, starch and sugar, the coefficient is equal to 1; in case of albuminoids, .729; ^ of fat, .700; and of alcohol, .666. An animal in a state of repose consumes a definite amount of oxygen in the breaking up or burning of the food, and gives off a definite amount of carbon dioxide, the jneasurement of which forms a basis for the food required for maintenance. The consumption of oxygen and the exhala- tion of carbon dioxide are rapidly increased the moment any work is per- formed. This method has been used with the horse by introducing tubes » After Lavalard, already cited, p. 123; according to Kellner, p. 75, .765. 246 MASS. EXPERIMENT STATION BULLETIN 188. into the trachea and measuring at intervals the intake and outgo of the respiratory gases. {d) The Respiration Calorimeter. — The apparatus consists of an air- tight room in which the animal is placed for different periods of time, and, in addition to collecting the feces and urine, the carbon dioxide exhaled and the heat radiated are accurately measured. It has been employed particularly in nutrition experiments with man, neat cattle, dogs and even smaller animals. An illustration of the value of the calorimetric method over chemical an- alysis and digestibility may be cited in the experiment conducted by Wolff, who found that a horse weighing 500 kilograms (1,100 pounds) required 6 kilos of oats and 6 kilos of hay, equivalent to 5,547 grams of digestible organic nutrients (minus fiber), to keep him in a state of maintenance and to enable him to perform 1,450,000 kilogrammeters of work. Of these nutrients 3,551 grams were necessary for maintenance, leaving 1,996 grams available for work. This amount — 1,996 grams — is equivalent to 3,478,030 kilogrammeters of work (1,996 multiplied by 4.1 calories equals 81,836 calories, which, multiplied by 425, equals 3,478,030), whereas the work actually performed was 1,450,000 kilogrammeters, or 41.7 per cent. Even this percentage was found by other experimenters to be too high, and is explained on the ground that the horse was particularly accustomed to such work. Ziintz and Lehmann, by the use of the respiratory quotient, found that the percentage of similar work in relation to digestible nutrients was reduced to 26 per cent., and Laulonie, by the same method, secured 22 per cent. In other words, after the maintenance requirement is satis- fied, the horse seems to be able to make use of about 25 per cent, of the remaining energy in the form of a definite kind of work (net efficiency of the animal, Armsby). It has been found further by Ziintz and Hagermann, in an extended series of experiments, that the net efficienc}'- of food in case of the horse varies widely, depending upon the character of the work performed. Thus, in case of walking without a load, the average efficiency was 35 per cent.; in different grades of ascent, at a walk without a load, from 33.7 to 36.2 per cent.; and with a load, 22.7 per cent. In case of work at a slow trot with- out a load the net efficiency was 31.96 per cent., and with a load, from 23.4 to 31.7 per cent. On the basis of these studies formulas have been worked out for the amount of food required for definite kinds of work, but it is hardly practicable to employ them under conditions ordinarily prevailing. By this method of procedure Ziintz has determined the net energy value of^a number of foods for the horse, and the results have led to a reduction in the amount of coarse food supplied, and an increase in the amount of concentrates, thus requiring the animal to expend less energy in mastica- tion and digestion, and to care for less inert matter in the intestinal tract. A former ration for the bus horses of Paris, composed of oats, corn, beans, bran, hay and straw, contained 18.5 kilos of dry matter, while a ration based on the results of recent investigations, composed of oats, corn, beans. THE NUTRITION OF THE HORSE. 247 molasses and chopped straw, contained only 12.5 kilos of dry matter, and proved to be less cumbersome, fm-nished a like amount of energy, caused less digestion disturbances and was more economical. C. Summary of Investigations. The many investigations made, some of which have been mentioned, have led to a number of important practical deductions concerning the nutrition of the horse which are stated below. 1. Horses need a definite amount of nutrients per 1,000 pounds of live weight for maintenance, and an extra quantity for work. This amount depends upon the size and temperament of the horse and the character and extent of the work performed. 2. In addition to the data akeady presented, the following recent state- ments by KeUner and Armsby concerning the nutrients and energy re- quirements of the horse are worthy of especial mention: — For Horses of 1,000 Pounds' Live Weight (Kellner). Light Work. Medium Work. Hard Work. Dry matter (pounds), Protein Fat •Carbohydrates, Total (fat X 2.2) Starch equivalent 18-23 1.0 .4 9.8 11.7 9.2 21-26 1.4 .6 11.3 14.0 11.6 23-28 2.0 .8 13.7 17.5 15.0 For Horses of 1,000 Pounds' Live Weight (Armsby). Light Work (2 Hours). Medium Work (4 Hours). Hard Work (8 Hours). Digestible protein, . "Net energy (therms). 1.4 11.1 2.0 18.2 Armsby adopts Kellner's protein standards and substitutes therms of ■energy for the customary fat and carbohydrates, or starch equivalent. He bases his knowledge of therms of net energy in feeding stuffs ^ utilized by horses largely on the work done by Ziintz and Hagermann. The feeding stuffs used by these experimenters were comparatively few in number. 3. Fat should not be suppHed to horses to a greater extent than is recommended for dairy animals, and 1 pound per 1,000 pounds of live weight should be regarded as the extreme amount. 1 The Nutrition of Farm Animals, by H. P. Armsby, p. 721. 248 MASS. EXPERIMENT STATION BULLETIN 188. 4. The proportion which the protein of the food should bear to the car- bohydrates and fat (nutritive ratio) has been a matter of considerable study and dispute. The International Congress of Nutrition ^ in 1900 dis- cussed the matter and concluded that a relation of 1 : 6 to 1 : 7 was the most suitable. Lavalard ^ states, as a result of his experiments, that 1 : 6 to 1:9 are permissible and satisfactory. KeUner'^ states that for horses doing work at a walk a ratio of 1: 10 is allowable, but that for hard work, and especially work done at a trot, a ratio of 1: 7 is preferable, because in such cases extra protein is needed to furnish maximum amounts of blood in order to carry the oxygen required for the rapid breaking down of the food material. 5. Experience has taught feeders, especially in European countries, that it is advisable to crush the coarse grains before feeding, and to cut the roughage and make a mixture of the two. The cut roughage aids in absorbing any moist feeds, particularly molasses, and also serves as a distributor of the heavy concentrates. 6. French investigators have recommended the substituting of corn, barley, rye, oil cakes, sugar and molasses for oats, and the reducing of the coarse fodders to a minimum, particularly for hard-worked horses, — as low in some cases as 6 pounds daily per 1,000 pounds live weight. 7. Cut straw has been highly recommended in place of hay because it is cheaper, is less likely to cause colic, contains less foreign material than hay, and serves as an excellent medium for the distribution of the grain. 8. A mixture which the French authority, Lavalard, recommends con- sists of 8 pounds of oats, 9 pounds of corn, 1 pound of beans, 5 pounds of molassine meal, and 7 pounds of chopped straw. This mixtiu-e contains, of digestible nutrients, 1.7 pounds protein, .47 pound fat, 11.52 pounds carbohydrates, 27.5 pounds total dry matter, and 27,712 calories of energy and is sufficient for hard-worked horses of 1,100 pounds weight. 9. For roughage the coarser hays, including alfalfa and clover, are recom- mended, also oat, wheat and barley straws. 10. Kellner recommends also as satisfactory concentrates, in addition to the cereals (excepting wheat), linseed, cocoanut and palm nut meal in amounts not exceeding 1 to 2 pounds daily. He states that corn, small amounts of brewers' grains, rice and linseed meals can be used in order to reduce the amount of oats to a minimum. 11. In the United States relatively large amounts of corn are fed, while on the Pacific coast barley of good quality predominates. In the semi- arid regions Kaffir corn and alfalfa have been used satisfactorily, particu- larly the latter. 12. The amount of water required daily depends upon the size of the animal, the work performed, and the time of year. The time of watering — whether before or after feeding — is a matter of minor importance. Horses become accustomed to both methods, and care should be taken to avoid sudden changes from the accustomed method. 1 L'Alimentation du Cheval, pp. 100, 101 * Die Ernahrimg d. landw, Natzthiere, Sechste Auflage, p. 455. THE NUTRITION OF THE HORSE. 249 13. Horses are, as a rule, of a nervous temperament, and it is advisable to avoid anything that will prove a source of irritation to the intestines, and that will induce extra water consumption. Inferior fodder, especially moldy stufif, should never be fed. D. Books on Horse Nutrition. The Nutrition of Farm Animals, Armsby. Chapter XIV. Published by the Macmillan Company, New York, 1917. The Productive Feeding of Farm Animals, Well. Chapter XXIV. Published by J. B. Lippineott Company, Philadelphia, 1915. Productive Horse Husbandry, Gay. Published by J. B. Lippineott Company, Philadelphia, 1914. Feeds and Feeding, Henry & Morrison. Chapters XVIII, XIX, XX. Published by the Henry & Morrison Company, Madison, Wis., 1915. A Digest of Recent Experiments on Horse Feeding, Langworthy. United States Department of Agriculture, Office of Experiment Stations, Bulletin No. 125, 1903. Die Ernahrung d. Landw. Niitzthiere, Kellner, Sixth Edition, Part III, Chapter V. Published by Paul Paray, Berlin, 1912. Grundlagen f . d. rationelle Fiitterung des Pferdes, Wolff. Published by Paul Paray, Berlin, 1886. L'Alimentation du Cheval, Lavalard. Published at Librarie Agricole de la Maison Rustique, Paris, 1912. Le Cheval, Lavalard. Published by Librarie De Firmin Didot et Cie, Paris, 1888. Les Aliments du Cheval, Duchambre et Curot. Published by Asselin et Houzeau, Paris, 1903. 250 MASS. EXPERIMENT STATION BULLETIN 188. Part II. FEEDING TRIALS WITH HORSES. Results and Suggestions. (a) Alfalfa for Horses. 1. On the basis of 1,000 pounds' live weight, a ration composed of 1.7 pounds of oats, 6.8 pounds of corn and 8.5 pounds of alfalfa hay did not prove sufficient for horses doing reasonably hard farm work (Kansas ration). 2. Fed such a ration the horses appeared quite restless and nervous, and lost in live weight, indicating insufficient food and possibly an unfavor- able action of the alfalfa upon the nervous system. 3. An increase of 10 per cent, in the above ration checked the loss of live weight, but not the restless, hungry condition. 4. The substitution of a timothy hay mixture for a portion of the al- falfa seemed to check in a measure the restless condition of the horses. 5. During the fall months the same grain ration was maintained, but timothy hay was substituted for all of the alfalfa. The horses fully main- tained their weights and appeared quieter than when the alfalfa ration was fed. This may have been due in part, at least, to the fact that less work was required daily than in the early part of the season. 6. A combination of one-fifth oats and four-fifths corn, together with a mixture of one-half alfalfa and one-half timothy, is likely to prove more satisfactory than a ration in which alfalfa constitutes the entire roughage. 7. A combination of one-third oats and two-thirds corn and timothy hay appears to be quite satisfactory, and furnishes sufficient protein for horses doing ordinary work. Only when quite hard work is required is it necessary to increase the protein by feeding alfalfa or a small amount of a protein concentrate. In such cases the roughage should be reduced and the amount of grain increased. (6) Brewers' Dried Grains for Horses. Brewers' grains, when prepared from perfectly fresh material, may constitute from 15 to 25 per cent, of the daily grain ration for horses, and may replace a like amount of oats. (c) Velvet Bean Feed for Horses. 1. Velvet bean feed represents the ground bean and pods of a tropical legume. THE NUTRITION OF THE HORSE. 251 2. At this station a ration composed of oats, corn, wheat bran and 20 per cent, velvet bean feed was fed to two farm horses for a period of three months, and gave quite satisfactory results. 3. While it would be possible to increase the amount of this feed in the mixture, it would hardly be advisable because the pods render the feed less digestible than corn. 4. Some lots have been found upon the market more or less moldy, due to imperfect drying. Such material is quite unfit for horses. Care should be taken to feed only well-dried, sweet material. (d) Linseed Meal for Horses. 1. During a period of two months the horses received a ration of oats, corn and 7 per cent, linseed meal. They ate the mixture readily and ap- peared in excellent condition during the entire time. 2. It is preferable in feeding this material to have the other grains with which it is mixed at least coarsely ground, otherwise the linseed meal separates out and is not likely to be eaten as readily. The addition of 5 to 7 per cent, of linseed meal to the grain ration for hard-worked horses should prove very helpful. (e) Rations for Work Horses. The amount of roughage fed may vary between 1 and Ij pounds daily per 100 pounds' live weight. Alfalfa may constitute one-half of the roughage. The amount of grain to be fed wiU depend, naturally, upon the character and amount of the work performed. From 1 to 1.4 pounds daily per 100 pounds of live weight should prove sufficient under most conditions. I. IV. 100 pounds of oats. 125 pounds of brewers' dried grains. 400 pounds of corn. 100 pounds of oats. i hay and i alfalfa. 225 pounds of corn. 50 pounds of wheat bran. Timothy or mixed hay. V. II. 100 pounds of oats. 200 pounds of corn. 100 pounds of velvet bean feed. Timothj' or mixed hay. 150 pounds of oats. 200 pounds of corn. 50 pounds of wheat bran. III. Timothy or mixed hay. 100 pounds of brewers' dried grains. 150 pounds of oats. VI. 200 pounds of corn. 100 pounds of oats. 50 pounds of wheat bran. 180 pounds of corn. TimSthy or mixed hay. 20 pounds of linseed meal. Timothy or mixed hay. 252 MASS. EXPERIMENT STATION BULLETIN 188. Hominy meal or crushed barley may be fed in place of one-half of the cracked or whole corn if desired. Molasses may constitute 10 per cent, of the grain mixture. It may be diluted somewhat with water and mixed with the grain. It aids in preventing colic. Inferior hay — weedy or moldy — 'and musty grain are to be avoided as causes of digestion dis- turbances. A. Alfalfa for Horses. The Kansas Experiment Station,^ co-operating with the United States Department of Agriculture, conducted a series of experiments in the feeding of work horses, using the artillery horses at Fort Riley (937 in all), with an average weight of 1,165 pounds. The work performed was called rapid light draft, and consisted of marching and drilling, drawing heavy wagons and guns often at a trot or gallop. Among the many rations tried was one composed, on the basis of 1,000 pounds of live weight, of 6.8 pounds of corn, 1.7 pounds of oats and 8.5 pounds of alfalfa hay, which contained, according to calculations made by the experimenters, the following di- gestible nutrients: — 1 Protein, . Kansas Ration. . 1.655 Carbohydrates, 8.720 Fat, .408 Total (fat X 2.2), . 11.270 Nutritive ratio, . 1:5.800 The alfalfa experiment was conducted with 17 horses for one hundred and forty days, and during the test the horses showed an average gain of 25.6 pounds per head. It was stated that they showed no signs of short- ness of wind, softness, lack of endurance, laxative effect or excessive urination. The amount of grain was reduced 19 per cent, and the amount of hay 30 per cent, from that consumed in a check ration of prairie hay and oats. The observers explain the satisfactory results on the ground that a small amount of alfalfa hay was fed with a relatively large amount of corn, a combination requiring a minimum amount of energy for its digestion. The 1,000-pound horse, working eight hours daily, requires, according to Armsby et als.,"^ 2 pounds of digestible crude protein and 18.2 therms of net energy. The horses in the Kansas alfalfa ration received 1.67 pounds of digestible crude protein and 13.41 therms of net energy. On the basis of digestible matter the following comparison can be made of nutrients required per 1,000 pounds' live weight for medium to hard work: — 1 Bui. No. 186. » The Nutrition of Farm Animals, p. 7H. THE NUTRITION OF THE HORSE. 253 Authority. Protein. Carbo- hydrates. Fat. Total (Fat X 2.2). Nutritive Ratio. Alfalfa ration Lavalard's standard for comparison, Grandeau's standard for comparison, Kellner's standard for comparison, i . Kellner's standard for comparison, ' . 1.655 1.330 1.920 1.600 2.170 8.721 11.170 10.920 12.500 13.700 .408 .400 ;600 .800 11.26 12.50 12.83 14.20 15.87 1 :5.8 1 :8.3 1 :5.7 1 :7.9 1 :6.3 1 Medium work. 2 Hard work. It appears that while the Kansas ration contained ample protein on the basis of accepted standards, it was deficient in total digestible nutri- ents and in therms of net energy. It seems to have been successful for the army horses doing the regular work required of them, but it is doubt- ful to the writer if it would prove sufficient in amount for horses doing medium to hard farm work. Experimental. In order to test the efficiency of this ration, two young western horses designated as Tom and Joe, which were purchased the winter previous, and which had been doing farm work during the spring and summer, were placed, Sept. 11, 1916, on the Kansas ration. Tom received 2| pounds of oats, 9| pounds of cracked corn and 12 pounds of alfalfa hay, and Joe received 2j pounds of oats, 9 pounds of cracked corn and 11 pounds of alfalfa. The hay fed for the first three weeks was grown upon the station grounds, was fine, but mixed with more or less foreign grasses. On Octo- ber 6 it was replaced with a coarser but better grade, this second cutting said to have been grown in Michigan. The ration was fed in three por- tions daily, and the horses weighed on each Monday morning before feed- ing and watering. Weights. Tom. Joe. September 17, . 1,415 1,305 September 24, . 1,415 1.295 October 1 1,415 1,290 Octobers 1,425 1,285 October 15, 1,405 1,285 October 22 1,410 1,285 October 29 1,400 1,280 November 6, . . . 1,415 1,310 November 13, . 1,425 1.335 254 MASS. EXPERIMENT STATION BULLETIN 188. Although, as has been previously shown, this ration was deficient in both total digestible nutrients and therms of net energy, the horses held their weights, due in all probability to the light work performed during the autumn months. They appeared hungry and very restless, the latter condition, in the opinion of the writer, being in part at least, a result of the influence of the alfalfa upon the nervous system. Beginning in the spring of 1917 the two horses which had been used on digestion experiments the preceding winter were worked on the farm and fed the Kansas alfalfa ration. On the basis of live weight Tom received daily 2f pounds of oats, 10| pounds of cracked corn and 12 pounds of al- falfa, and Joe received 2| pounds of oats, 9 pounds, 14 ounces of corn and 11 pounds of alfalfa. Weights. Tom. Joe. April 23 1,390 1.310 April 30 1.390 1,280 May 7. . . . 1,390 1,290 May 14, . . . 1,400 1,295 May 21. . . . 1,380 1,285 May 28. . . . 1,370 1,275 June 4, 1,370 1,260 It was necessary to work them lightly during the first month. As the work was increased in amoimt they began to show a gradual loss in weight and to appear very nervous and hungry. Because of such conditions, and of the additional spring work required of them, the ration was increased 10 per cent. June 4, Tom receiving 13.2 pounds of alfalfa, 3 pounds of oats and 11.5 pounds of corn, and Joe receiving 12.1 pounds of alfalfa, 2.7 pound of oats and 10.9 pounds of corn. Weights. Tom. Joe. June 11, . . . 1,390 1,400 1,400 1,410 1,275 June 18 June 25. 1,280 1,275 July 2 1.270 July 9. July 16 1,420 1,430 1.270 1.300 THE NUTRITION OF THE HORSE. 255 These rations contained the following pounds of digestible nutrients and therms of net energy: — Protein. Carbo- hydrates. Fat. Total (Fat X 2.2). Therms. Fed. Required. Tom Joe Grandeau standard, 2.56 2.37 14.02 13.05 .59 .54 17.80 16.60 17.96 22.21 20.72 25.48 23.66 In so far as weights and digestible nutrients were concerned, the horses appeared 'to have received sufficient food for the work they were doing. The therms fed fell below the standard theoretically required, which leads one to question whether this standard is not too high. The horses still appeared rather restless and hungry, although they performed their daily task in a more satisfactory way. Beginning July 16 the ration was modi- fied by reducing the amount of alfalfa fed daily to each horse to 10 pounds, .and adding 6 pounds of timothy mixture to Tom's ration and 5 pounds to Joe's ration, the grain remaining as in the ration preceding. The object of the change was to attempt to reduce the restless action manifested by the horses, which in a measure was successful, and their weights were maintained. Weights. Tom. Joe. July 16 1,430 1.300 July23 1.430 1.300 July 30, 1,415 1.270 Augusta, ' 1,410 1.270 August 13, 1.410 1,280 August 20. 1,420 1.300 August 27, 1,410 1,270 Septembers 1.410 1.300 Beginning September 4, hay was substituted for the entire amount of alfalfa, the grain ration remaining constant. The calculated digestible nutrients and weights of the horses follow: — 256 MASS. EXPERIMENT STATION BULLETIN 188. SJ .2 Therms. S X S a "3 1 1 1 1 s' £ 1 1 1 1 1 1 Tom, 1.87 2.85 13.36 .52 19.22 1:9.3 20.00 25.48 Joe 1.70 2.66 12.55 .47 17.94 1 :9.5 18.80 23.66 Grandeau's standard (1,400- pound horse). Lavalard's standard (1,400- pound horse). 2.69 1.86 - - - 17.96 17.20 - - - Weights. September 10, September 17, September 24, October 1, . October 8, . October 15, . October 22, . October 29, . The weights were well maintained, indicating that for the work per- formed sufficient nutriment was being supplied. The work was rather irregular during this period, and may be considered as light. The combination of hay, corn and oats evidently was sufficient in total digestible nutrients, but rather deficient in protein, according to Grandeau, for horses doing moderate work. The therms of energy were noticeably below the standard. The ration conformed more closely to that set by Lavalard, who accepts one with less protein and a wider nutritive ratio than other investigators. It is well known that horses keep in good con- dition and do satisfactory work on rations composed of hay, corn and oats. It seems probable, therefore, that only in case of quite hard work is it desirable to increase the protein requirement above the amount furnished by such a combination. Less corn and more oats, i.e., rather more pro- tein and less starch, or a somewhat narrower ration, is desirable in the warm summer months. While recognizing the large number of horses in the Kansas experiment and the satisfactory results secured, on the basis of our own observations and the accepted feeding standards it seems to the writer that the amounts of the several feeds are not likely to be sufficient, nor the combination THE NUTRITION OF THE HORSE. 257 particularly satisfactory, for most work horses. It is believed that for each 100 pounds of live weight a pound of roughage is a reasonable allow- ance, and that one-half of this roughage may consist to good advantage of alfalfa, anii the balance of a timothy mixture. B. Brewers' Dried Grains for Horses. Brewers' dried grains, the residue of the beer breweries, contain from 20 to 28 per cent, of protein, 13 to 17 per cent, of fiber, 5 to 7 per cent, of fat, and from 40 to 46 per cent, of extract matter. They contain more protein, fat and fibre than oats, some 14 to 20 per cent, less extract mat- ter, and possess about 15 per cent, less net energy value. Voorhees^ of the New Jersey station, as a result of feeding trials, stated, " That on the whole a pound of dried brewers' grains was quite as useful as a pound of oats in a ration for work horses." Foreign investigators have stated that they can replace one-half of the oat ration. In New England, while they have been used more or less, one fails to learn of their general employ- ment as a part of the daily ration. If used especially for horses, it is quite important that they be dried before being allowed to sour or decompose. This station has fed them as a component of horse rations with satis- factory results. The same two horses that were used in the alfalfa experi- ment were employed. They did moderate farm work which consisted principally of plowing, harrowing and teaming. Ration I. 5 pounds of ground oats. 3 pounds of brewers' grains. 8 pounds of cracked corn. 2 pounds of wheat bran. 15 pounds of timothy mixture. The ration contained the following digestible nutrients in pounds and net energy value in therms on the basis of 1,000 pounds of live weight: — AUTHORITT. Protein. Total (Fat X 2.2). Nutritive Ratio. Therms. Brewers' dried grain ration, .... KeUner's standard (moderate work), . Lavalard's standard (moderate work), . Grandeau's standard (moderate work), 1.76 1.40 1.33 1.92 12.00 12.62 12.50 12.83 1 :5.9 1 :8.0 1 :8.3 1 :7.9 15.1 The above comparisons indicate that the ration fed contained sub- stantially sufficient digestible protein and total nutrients. The horses were weighed weekly in the morning, before feeding and watering, » Bui. No. 92, N. J. Agr. Exp. Sta. 258 MASS. EXPERIMENT STATION BULLETIN 188. Weights. Tom. Joe. May 22. 1,400 1,240 May 29. 1,400 1.280 Junes, 1,400 1,275 June 12, 1.425 1,285 June 19, . . . 1,425 1 290 ^ It seemed evident that for the work performed the horses were receiving sufficient nutrients to keep them in normal condition, although they did not materially add to their weight. Ration II. On June 19 the ration was modified slightly by replacing 2 pounds of the oats with 2 pounds of the brewers' grains, thus increasing the protein slightly, while the total nutrients received were nearly the same. Weights. Tom. Joe. June 26, Julys, July 10, July 17. 1,420 1.415 1.420 1,400 1.260 1,250 1,240 1,240 During this period there seemed to be a slight loss in weight. Whether this was due to the warm weather or to the modification of the ration is not clear. Ration III. On July 17 the horses were put back on to Ration I and continued until August 14> Weights. Tom. Joe. July 24 1,420 1,300 July 31 1,415 1,270 August 7, 1,410 1,285 August 14 1,405 1,270 Slight shrinkages in weight were noted. THE NUTRITION OF THE HORSE. 259 Ration IV. On August 14, because the horses were doing somewhat less work, Ra- tion I was reduced 1 pound each of oats and cracked corn. Weights. August 21, . August 28, . September 4, September 11, 1,440 1,305 1,435 1,310 1,425 1,265 1,435 1,295 It will be seen that the rations fed the two horses from about the middle of May until September 11 contained from 3 to 5 pounds of the brewers' grains out of a total of 18 pounds of grain (or from 17 to 28 per cent.)- At the beginning the horses weighed 1,400 and 1,240 pounds, respectively, and at the close, 1,435 and 1,295 pounds. During this time variations in weight were noted, due perhaps partly to increase or decrease in work, and partly to weather conditions. The horses kept in uniformly good condition throughout the season, indicating that the brewers' grains in the amounts fed exerted no adverse effect upon them. The writer is inclined to favor Rations I and II as satisfactory combina- tions, especially if the brewers' grains can be purchased for less than the oats. It is not advisable under most conditions to include too large an amount of brewers' grains in the ration, for the reason that they will fur- nish too much protein and not sufficient digestible matter. C. Velvet Bean Feed for Horses. The velvet bean, of which there are many varieties, is a tropical legume and is grown largely in Florida, Alabama and Mississippi. It needs a long season for its maturity and is rarely grown north of Savannah. It is a rank grower, the vines trailing on the ground to a length of from 15 to 75 feet; they are difficult to secure for hay, and have been used largely for grazing. It is now more common to pick the best of the beans and use them without hulling for cattle, or hulled as a food for pigs. Machinery has been devised for drying and gi-inding the unhulled beans, thus pro- ducing the velvet bean feed, and it is said that the industry is increasing rapidly. 260 MASS. EXPERIMENT STATION BULLETIN 188. Analysis and Digestibility of Velvet Bean Feed (Bean and Hulls). Composition. Percentage Digestible. Pounds Digestible in 2,000. Water Ash, . Protein, Fiber, Extract matter, . Fat 12.00 5.11 16.80 12.85 49.00 4.24 32 75 63 "85 81 32.7 252.0 161.9 833.0 68.7 Total 100.00 - 1,348.3 In chemical composition the feed does not vary greatly from wheat bran, except that it has rather more fiber derived from the bean pods. It contains about 175 pounds more digestible organic nutrients per ton than bran, and should have a somewhat greater feeding value. The present spring the experiment station fed it as a component of a ration to the two station horses which were being used on general farm work and which had been employed in digestion experiments the previous winter. Ration I. Ration I, which we began feeding in May, was composed of a mixture of — Oats, . Corn, Velvet bean feed. Wheat bran, . Pounds. . 100 . 160 40 40 The velvet bean feed constituted 11.7 per cent, of the ration. The horses ate the ration freely, Tom receiving 18 pounds and Joe 17 pounds daily, in addition to 15 pounds of hay. Ration II. On June 8 the ration was modified by increasing the velvet bean feed to 60 pounds and decreasing the corn to 140 pounds in the mixture. The velvet bean constituted nearly 18 per cent, of the mixture, and each horse received a little over 3 pounds a day. The weights of the horses follow : — Tom. Joe. June 3, 1,395 1,280 June 10, 1,345 1,245 June 17 1,370 1,265 June 24 1,400 1,285 THE NUTRITION OF THE HORSE. 261 During this period these horses were working eight to nine hours daily for 5^ days each week, doing plowing, harrowing and similar farm work. They maintained their live weight, but were^ not in as good flesh as was desired. Ration III. On June 24 the hay was increased to 18 pounds daily, and so continued until July 15, for the reason that they acted rather hungry, and it was thought a little more bulk would render them more contented. Weights. Tom. Joe. Julyl, 1,370 1,270 Julys, 1,395 1,300 July 15, . . . 1,400 1.300 The work during the above time was of about the same character, but on the whole not as difficult as during June. The live weight appeared to be maintained, but apparently did not increase. Ration IV. On July 15 the grain mixture was increased to 20 pounds for Tom and 19 pounds for Joe, in addition to the 18 pounds of hay, and so maintained until September 1. Weights. I "°"- Joe. July 22 1,400 1,300 July 29, 1,390 1,290 August 5 - - August 12, _ . 1,410 1,320 August 19, 1,395 1,320 August 26, 1,400 1,320 September 2 1,405 1,325 Dm-ing the above period Tom appeared stationary and Joe increased about 25 pounds in weight. Tom is a long-bodied, long-legged horse and not as compact of build as is Joe. In spite of the fact that the live weight was not substantially increased, the horses appeared in better condition than in the early summer. The horses were quite fully employed during August in harrowing, plowing and drawing manure. The estimated pounds of nutrients and therms of energy contained in the daily ration on the basis of 1,400 pounds live weight follow: — 262 MASS. EXPERIMENT STATION BULLETIN 188. Total (Fat X 2.2), Nutri- tive Ratio. Therms fed. Therms needed (Armsby). 15 pounds hay + 18 pounds grain equals IS pounds hay + 20 pounds grain equals Authority: for for comparison comparison Kellner's standard (moderate work). Kellner's standard (hard work). Lavalard's standard for comparison (moderate work). Grandeau's standard for comparison (moderate work). 2.43 2.76 1.86 2.20 20.37 23.37 17.70 24.50 18.10 17.96 1 :7.4 1 :7.4 20.40 23.00 25.5 25.5 It is believed that 15 pounds of hay and 18 pounds of grain, of which velvet bean feed constituted some 3 pounds, were sufficient for the work the horses did from week to week. It is possible that during a few days, or for a week at a time, the nutrients were not sufficient. The other ration, consisting of 18 pounds of hay and 20 pounds of grain, probably was more than was needed. The horses ate the ration, of which velvet bean feed comprised some 18 per cent., continuously for over three months, and the results were in every way satisfactory. D. Linseed Meal as a Grain Supplement for Horses. Beginning September 1 the two horses Tom and Joe were fed a grain ration composed by weight of 100 pounds of whole oats, 160 pounds of whole corn, and 30 pounds of old process linseed meal. Tom received daily 20 pounds of the mixture and Joe 19 pounds, in addition to 18 pounds of hay. This ration was continued until September 28, when it was slightly modified by decreasing the linseed meal to 20 pounds in the mixture, or about 7 per cent. The reason for the reduction was that the linseed did not mix evenly with the corn and oats, owing to the fact that they were not ground or crushed; hence considerable would separate out and the horses were inclined to leave a little. Horses do not seem to care particularly for the linseed if fed unmixed, but will eat a reasonable amount readily if constituting a part of a mixture. This ration was continued until November 11. The horses did regular farm' work during this i^eriod, but did not average as manj^ hours daily as earlier in the season, and the work would be considered only moderate. THE NUTRITION OF THE HORSE. 263 Weights Joe. September 2, September 9, September 16, September 23, September 30, October 7, . October 14, . October 21, . October 28, . November 4, November 11, 1,405 1,325 1,395 1,315 1,405 1,330 1,435 1,345 1,445 1,350 1,450 1,370 1,440 1,340 1,425 1,340 1,415 1,340 1,410 l',350 1,425 1,360 Digestible Nutrients in Ration (Pounds). Protein. Total (Fat X 2.2). Nutritive Ratio. IS pounds hay + 20 pounds grain equals Kellner's standard (hard work), . 3.11 2.80 24.04 24.50 1:6.7 1 :'7.7 On the basis of the calculated digestible nutrients it is evident that the horses were receiving all the food necessary for eight hours of hard work daily. The work actually performed could only be called moderate, which explains to an extent the gain in live weight. It is believed that the addi- tion of 5 to 10 per cent, of linseed meal to a grain ration composed of one or more cereals will prove helpful, especially to hard-worked horses, and will be eaten without trouble. BULLETIN No. 189 MARCH, 1919 MASSACHISETTS AGRICILTIRAL EXPERIMENT STATION The European Corn Borer and its Control By STIART C. VINAL and D. J. CAFFREY Requests for bulletins should be addressed to the Agricultural Experiment Station, Amherst, Mass. Publication of this Document approved by the Supervisor of Administration. CONTENTS. Foreword, .... Introduction, Synonomy, .... Common names applied to species. Foreign history, . History in United States, Discovery of the insect, . Identifying the species, . A previous record in Massachusetts, Preliminary investigations, Plans made for further investigations. Control measures during spring of 1918, Control measures during autumn of 1918, Quarantine measures enacted and their origin. National quarantine measures. State quarantine measures, . Geographical distribution, In the Old World, . In the United States, Territory examined in Massachusetts, Territory examined in New Hampshire, Territory examined in Maine, . Territory examined in Rhode Island and Connecticut, Food plants. In the Old World, In the United States (Massachusetts), List of food plants, . Character and extent of injury, Corn, .... Other food plants, . Descriptions of the different stages, The egg. The larva, The pupa. The adult, Life history, First generation, Second generation. Seasonal history and development, Nimaber of generations, . Seasonal history. Seasonal abundance, Habits of larvjB, . Hatching, Habits when attacking corn. Habits when attacking dock. Habits when attacking Lady's Thumb, Habits when attacking barnyard grass. Molting, ..... Length of larval life without food, . Unusual habits, .... IV CONTENTS. Pupation " 49 Location of pupa, ....... 49 Cocoon formation 49 Changes undergone by the larva previous to pupation, 50 Process of pupation, 50 Changes undergone by the pupa, .... 50 Habits of adults, 51 Emergence of the moths, . . 51 Copulation, . 51 Proportion of sexes, ...... 52 Flight, 54 Oviposition, ........ 55 Details of oviposition, 55 Distribution of egg masses, . . ' . 56 Total niunber of eggs deposited by each female, 56 Duration of fertility, ...... 56 Parasites, 57 European records of parasites, .... 57 Records of parasites in Massachusetts, 57 Predators, 59 Birds, 59 Insects, ........ 59 Control, 59 Destroying plants containing over-wintering larvae, 59 Application of arsenicals to plants, .... 62 Cultural practices to avoid damage, 63 Other insects fr.equently mistaken for the European Corn Borer 64 The stalk borer, 64 The corn ear worm, ...... 64 Cutworms, 64 Svunmary, . . ; • 65 Literature 68 Explanation of plates, 69 BULLETI]^ ^o. 189. DEPARTMENT OF ENTOMOLOGY. THE EUROPEAN CORN BORER AND ITS CONTROL. BY STUART C. VINAL AND D. FOREWORD. During 1918 the Massachusetts Agricultural Experiment Station and the Bureau of Entomology of the United States Department of Agri- culture worked on the European corn borer under a co-operative agreement by which the station was to make a study of the life history, food plants, methods of distribution and methods of control of the insect, while the Bureau was to determine its distribution, develop control measures and prevent its further spread. Mr. Stuart C. Vinal, assistant entomologist of the experiment station, was assigned to this work on the station side, and located in Arlington, He worked day and night on the subject and accomplished an enormous amount, but with such disregard for his health that when attacked by influenza he was unable to resist it and died Sept. 27, 1918. The person best fitted to take up and bring together for publication the information gathered by Mr. Vinal was Mr. D. J. Caffrey, who had been in charge of the Bureau side of the work, and who had been in close touch with Mr. Vinal's investigations throughout the year, and he there- fore took the material left by Mr. Vinal and has brought it together and put it in shape for publication. Fortunately, most of it was already well worked out, but providing the data obtained by the United States gov- ernment as its share of the work, and the form and arrangement of the whole bulletin have been Mr. Caffrey's contribution. The line drawings have been prepared by the WTiter of this foreword, from sketches made by Mr. R. E. Snodgrass of the United States Bureau of Entomology. H. T. Fernald. MASS. EXPERIMENT STATION BULLETIN 189. INTRODUCTION. Practically all insect pests of foreign origin found in the United States have reached our seaports through the agency of commerce. The great variety of living plants, as well as raw materials for use in manufacturing enterprises and the miscellaneous freight and personal effects that are daily received on our shores from all parts of the world, provide an ample opportunity for the entrance of almost any destructive pest. Many of these insect immigrants, on finding favorable climatic and food plant conditions, become permanently established, and in the course of time spread from their point of origin and become of more economic impor- tance each year, unless checked by artificial agencies. The danger existing from these involuntary importations of destructive insect pests is still further increased by the fact that in most instances their natural enemies are not imported with them. Under these cir- cumstances the pest is enabled to extend its activities without being subject to the natural handicaps imposed by nature. This results in a more rapid multiplication and a greater degree of destructiveness than exists in the original habitat of the insect. Such, in brief, is the history of many of our most important and gen- erally distributed insect pests of to-day. To the long list of foreign pests now found in the United States must be added the European corn borer, or corn pyralid, Pyrausta nubilalis Hiibner, which has recently become established in the eastern part of Massachusetts. The caterpillar of this insect has long been recorded in Europe and Asia as one of the most serious insect enemies of corn, hemp, millet, hops and other crops. Corn and hop plants are very severely damaged by this pest, 50 per cent of these crops often being destroyed in some sections of Central Europe. As a result of studies made on the habits and destructive powers of the European corn borer throughout the infested portion of Massachusetts during the seasons of 1917 and 1918, it is evident that this species is without doubt the most dangerous and destructive insect enemy of the corn crop that has yet been introduced into the United States. As corn is one of the bulwarks of American agriculture, and has within the past few years become our most valuable crop from a monetary standpoint, it ^^ill be recognized that the problem of controlling this insect which threatens to destroy a large per cent of the crop each year is not con- fined to Massachusetts, but is a problem of national importance, which must be acted upon promptly and thoroughly to the end that the insect may be at least confined to its present area of distribution, if ultimate extermination is found to be impossible. If this insect is allowed to extend its area of distribution and reach the corn belt of the middle western States, it will be a national calamity. Although Massachusetts is universally considered to be a manufacturing THE EUROPEAN CORN BORER AND ITS CONTROL. 6 State, it should be stated that during 1917 a total of 2,806,000 bushels of field corn were grown in the State which were worth $6,033,000 ac- cording to the prices prevailing the 1st of the following December. This is in addition to the value of the sweet corn, fodder corn and popcorn produced in the State. Aside from the national importance of restrict- ing the spread of this dangerous insect, the State of Massachusetts should take all measures to protect the revenue obtained from its corn crop. There are several other sjjecies of destructive corn borers known to attack corn in the United States, the most important of which are the larger cornstalk borer, Diatraa zeacolella Dyar, and the lesser cornstalk borer, Elasmopalpus lignosellus Zeller. These two species occur in the South, and even to some extent in the northern States, but have never become permanently established in Massachusetts or any other State with a similar climate. They are doubtless unable to withstand the severe winter conditions, and this characteristic has the effect of greatly limiting their range of^istribution. The European corn borer, however, is not limited in its range by ordinary climatic conditions, judging from its range of distribution in the Old World, and from its behavior to date in the infested area of Massachusetts. The species would thus be able to adapt itself to all parts of Massachusetts and ultimately to the entire country. In Massachusetts the only native stalk borer attacking corn is Papai- pema nitela Gn., which more frequently infests the stalks of potatoes, tomatoes and numerous common weeds. This insect, however, does not normally occur in sufficient numbers to cause serious loss. During the past two seasons, however, it has been rather more abundant than usual and because of the fact that its injuries to corn superficially resemble those caused by the European corn borer, much of the damage really caused bv the latter has been attributed to the native stalk borer. SYNONOMY. The species was first described and figured by Jacob Hiibner (2) in 1796. He described the male and female as separate species, — the male as Pyralis riubilalis, and the female as Pyralis silacealis. Owing to this fact the synonomy of the species in Europe is somewhat confusing. Haworth (3) in 1811 refers to the species as Pyralis glabralis. Treitschke (4) in 1829, and Duponchel (5) in 1831, adopted the name Pyralis silacealis Hubn., although recognizing that the Pyralis nuhilalis of Hiibner was the male of Pyralis silacealis Hiibn. Guenee (7) in 1854 accepts the species as being identical with the Botys eupulinalis illustrated in the Icones Insectorum of Clerck (1) in 1759. A study of the figure referred to in Clerck's work, however, con- vinced later workers that it could not be the same insect. Nevertheless, this error by Guenee led to the acceptance of Botys as the generic name by several succeeding workers. 4 MASS. EXPERIMENT STATION BULLETIN 189. During the same year Guenee (8) gave the name Botys zealis to a species from the East Indies very close to Botys eupulinalis. After the descrip- tion he adds this note: "It may be simply a variation of our eupulinalis, or, rather, this latter may have become acclimated among us with the cultivation of maize, and may be of exotic origin." In the present state of our knowledge the first theory seems to be the most probable. Lederer (9) in 1863 retains the species in the genus Botys, where it had been placed through the faulty conception of Clerck's figure, by Guenee, as previously mentioned. Lederer, however, accepts the figure of Hiib- ner's nvbilalis as truly representing the species, and refers to it as Botys nubilolis. This name is accepted by Staudinger and Wocke (10) in 1871. Moore (12) in 1888 refers to the species as Hapalia kasmirica. He is followed by Butler (13) as late as 1889, who designates the species as Hapalia eupulina {non Clerck). Meyrick (14) in 1895 removed the species to the genus Pyrousta, and retained the nuhilalis of Hiibner, in which he has since been followed by Hampson (15), and by Staudinger and Rebel (17) in 1901. We may therefore accept the species as Pyrausta nuhilalis Hiibn. COMMON NAMES APPLIED TO SPECIES. In Europe several different common names are applied to the species under consideration. The names most frequently used are the "corn pyralid;" "maize pyralid;" "pyralid of the maize;" "maize botys;" "botys;" "millet botys;" and "der Maiszunsler." In the literature concerning the insect which has been published in the United States since its discovery, the species has been referred to as the European corn borer and the European cornstalk borer. The former name undoubtedly is more appropriate for the insect, as the larva? attack all parts of the corn plant except the fibrous roots, and do not confine their operations to the stalk as the name cornstalk borer would imply. Although many plants are attacked by the insect, corn is its favorite host, and is injured to a greater extent than any other com- mercial crop attacked by it. The name European is adopted to indicate its foreign origin, although the species is indigenous to other parts of the world. Taking all facts into consideration, it is believed that the name European corn borer is the most appropriate common name for the insect, and as such it will be considered in this bulletin. FOREIGN HISTORY. Foreign literature contains a large number of references to the serious damage caused by P. nubilalis, a loss of 50 per cent of the crops attacked being reported by some writers. There is, however, a decided lack of literature dealing with its biology and control. The only exceptions are the brief and incomplete articles by Robin and Laboulbene (11) in 1884, THE EUROPEAN CORN BORER AND ITS CONTROL. 5 and of Jablonowski (16) in 1899. Robin and Laboulbene detail the habits of the larva? and the character of their damage to corn, hemp, hops and other food plants. The authors give an account of the severe damage which resulted from the attacks of this insect on corn, hemp and hops in the Department of the Aisne (France) during 1878 and 1879, as well as short extracts from the writings of other European authors mentioning the activities of this insect in various food plants. The absence of parasites is noted, and brief descriptions are given of the larva, pupa and adult. The authors recommend the burning of plants contain- ing the overwintering larva^, during the fall or early winter, as the most effective means of control. Jablonowski records a very severe outbreak of P. nuhilalis which de- stroyed a fourth part of the corn crop in Hungary during 1898. This damage was especially pronounced in the large plains of Hungary, which are very fertile. The author describes the character of the damage caused by the larva to corn, millet, hemp, hops and various minor food plants. The adult is described and figured very accurately; its habits of flight are detailed, and also the ovi position habits of the female. Mention is made of a single parasitic fly (Ceromasia interrupta Rdi.) which the author bred from the larva. Reference is also made to KoUar (6), who in 1837 recorded that some Iclineumonidae had been bred from the species. For control measures Jablonowski recommends that early in the season, when most of the larva are confined to the terminal nodes of the plant, these upper portions be cut off and thrown into a water barrel, to be sub- sequently treated with hot water or fluid manure. This procedure can be repeated at short intervals because the treatment will not curtail the harvest. After harvest the infested plants should be pulled up by the roots and burned. In cases where the upper parts of infested plants are harvested the remaining stubble should be lightly plowed up, collected with a rake and burned. The author mentions the fact that the plow- ing under of infested material does not injiu-e the contained larvse. He also states that after shelling the corn the cobs should be used as fuel during the winter. The burning of all wild grasses that may serve as host plants for the overwintering larvse is another general recommenda- tion. These methods were found to be attended with considerable labor and expense, but were very effective in controlling the pest in Hungary during the outbreak of 1898. HISTORY IN UNITED STATES (MASSACHUSETTS). Discovery of the Insect. During the summer of 1917 the senior author found many sweet corn fields in the vicinity of Boston, Mass., which were being very severely injured by light-colored larvae which tunneled in the stalk and later attacked the ears. Further investigation disclosed the fact that the identity of these dep- 6 MASS. EXPERIMENT STATION BULLETIN 189. reflating larvae was unknown to the entomologists of that section where the insect had been found. This aroused the interest of the senior author, who had early recognized the serious nature of the pest. He accordingly collected pupae from infested cornstalks in the field during the month of July, 1917, from which the adults emerged early in August. Identifying the Species. To secure the identification of the species concerned. Dr. C. H. Fernald's extensive collection of both native and exotic moths was available at Amherst, Mass. An examination of his European collection revealed specimens of both male and female Pyralid moths, identical with those reared from the infested cornstalks in eastern Massachusetts. These European specimens had been determined by Mr. E. L. Ragonot, a French lepidopterist, as Pyrousta mtbilalis Hiibner. Specimens of the moths reared in Massachusetts were also submitted to Dr. H. G. Dyar of the United States National Museum at Washing- ton, D. C, who gave the same identification, stating that it was a com- mon and very destructive pest of various wild and cultivated plants in the Old World. A Previous Record in Massachusetts. Prior to 1917 this insect had never been reported as occurring in the United States, although the following supplementary facts should be recorded. During August, 1916, specimens of dahlia stems infested by lepidopterous larva^ were sent to the Massachusetts Agricultural Experi- ment Station from three localities near Boston, Mass. (Medford, Everett and Lynn). Adults were bred from this material, but their identity was not discovered nor their significance realized at the time. Later, however, the senior author determined these adults as being identical with the P. nubilalis bred from corn in 1917. Thus P. nubilalis was first bred in the United States in 1916, although its identity was not known until adults were bred from corn in 1917. Preliminary Investigations. As soon as this pest was found to be of foreign origin, and its potential menace to American agriculture realized, its presence became of more than local importance, and a survey was made in eastern Massachusetts during the latter part of September, 1917, to roughly determine its dis- tribution and any other pertinent facts bearing on the insect, and the results of this preliminary survey were published by the senior author (IS) in December, 1917. At this time it was found that the insect had estab- lished itself in an area covering approximately 100 square miles, im- mediately north and northwest of Boston, Mass., and that the towns at the mouth of the Mystic River were more generally infested than the others. In this section are several cordage factories which import hemp THE EUROPEAN CORN BORER AND ITS CONTROL. 7 (Cannabis saliva) from Europe. This fact, together with the knowledge that hemp is one of the favorite food plants of P. nuhilalis in Europe, at once suggested the possibility that this insect may have reached our shores through this medium. Early sweet corn grown in market gardens 10 to 12 miles inland had been seriously attacked by this pest for the past three or four years, and from this it is inferred that the species was im- ported about 1910, although this date is a mere conjecture. At this time (1917) sweet corn was found to be the only valuable commercial crop attacked by P. nuhilalis, the early crop being damaged to the extent of 10 to 20 per cent, while the loss to late plantings ranged as high as 75 to 80 per cent. Several weeds and grasses were also noted as food plants of P. nuhilalis. Observations made on the feeding habits of the species in the infested fields confirmed the original belief that the insect under consideration was possessed of characteristics that would render it a serious menace to the corn crop, and that it would be a very difficult pest to control. Burning, burying or feeding the plants containing over- wintering larvse were methods suggested for the control of the insect. It was pointed out that measures for insuring or compelling satisfactory handling of all infested material were very necessary, and that though these results might possibly be obtained by local organizations of farmers and gardeners instituting vigorous action, it seemed probable that the matter must be taken in hand by the State or Federal authorities if the insect was to be brought under control and its further spread prevented. Plans made for Further Investigations. Accordingly, Dr. H. T. Fernald, head of the Department of Entomology at the Massachusetts Agricultural Experiment Station, notified officials of the Bureau of Entomology at Washington, D. C, of the presence of the European corn borer in Massachusetts, and reviewed the facts already known as to the dangers existing from the presence of this pest. Plans were immediately made for co-operation between the Massachusetts Agi-icultural Experiment Station and the Bureau of Entomology, in a further investigation of the insect, in order to determine its biology and methods of possible control. Special attention was to be given to the food plants and distribution of the insect in the United States, with a view to recommending quarantine measures that would prevent the spread of the pest through avenues of commerce. Quarters were established at Arlington, Mass., in April, 1918, and the results of the investigations to Nov. 30, 1918, are presented in this bulletin. Control Measures during Spring of 1918. During the spring of 1918 a campaign was inaugurated by the Massa- chusetts State Board of Agriculture, which had for its object the burning of cornstalks and other infested plants within the infested towns. This work was under the direct supervision of Mr. Wilfrid Wheeler, Secretary 8 MASS. EXPERIMENT STATION BULLETIN 189. of the Board of Agriculture, and Mr. R. H. Allen, State Nursery In- spector. The infested towns were placarded with warning notices illus- trating the insect, and recommending the burning of all cornstalks re- maining from the previous year. This was supplemented by a detailed survey in each of the infested towns and the burning of cornstalks in instances where the owners failed to comply with the recommendations. The States Relations Service of the United States Department of Agri- culture, through the county agricultural advisers and other agents, aided in this campaign of publicity. Control Measures during Autumn of 1918. lu October, 1918, an extensive campaign was begun for the eradication of all cornstalks, weeds and crop remnants of the current season which contained the corn borer larva?. This was under a co-operative agreement between the Massachusetts Department of Agiiculture and the Bureau of Entomology, Section of Cereal and Forage Insect Investigations. Crews of men were placed in each of the infested towns, who, under the direction of competent foremen, burned infested material that had not been eliminated by property owners or their representatives. This was preceded by a similar campaign of publicity to that in force during the spring clean-up work, although on a larger scale. Town and State officials aided in this work in some instances by agreeing to destroy the infested plants growing on public projjerty under their jurisdiction, but, owing to the early approach of severe winter weather, it is probable that the clean- up of infested plants will not be completed until the early spring of 1919. Quarantine Measures enacted and their Origin. National Quarantine Measures. In late July, 1918, it was found that many sweet corn ears e.xposed for sale in the wholesale markets at Boston were infested by larvae and pupae of the European corn borer. This circumstance at once suggested the possibility that these infested products might be shipped outside the area already infested by the insect and become sources of new infestations. As a result of reporting these facts to the Federal Horticultural Board a public hearing was held at Washington, D. C, Aug. 27, 1918, to con- sider a proposed quarantine of that portion of Massachusetts known to be infested by the European corn borer. At this time, however, quar- antine action was deferred in order to await the results of the field con- ference scheduled to be held at Boston, Mass., Sept. 6, 1918, to consider ways and means of handling the problem. This conference was attended by entomologists and agricultural au- thorities from all of the New England States," New York and New Jersey, and by officials of the Bureau of Entomology, the Massachusetts Market Gardeners' Association, and the Boston Produce and Fruit Exchange. THE EUROPEAN CORN BORER AND ITS CONTROL. 9 A field meeting was held in the morning, during which those attending the conference were taken to a badly infested sweet corn field at West Med- ford, and the injury of the insect to corn and other plants observed. In the afternoon the present status of the insect was discussed and sug- gestions made for its control or possible extermination. The consensus of opinion inclined very strongly to the belief that vigorous quarantine and control measures were necessary if the destructive insect was to be confined within its present limits. This course of action was favored jointly by the entomologists and by the representatives of the market gardeners and produce dealers present. Accordingly, notice of quarantine No. 36, on account of the European Corn Borer, Pyrausta niibilalis, was issued by the secretary of agriculture tlirough the Federal Horticultural Board, and became effective on and after Oct. 1, 1918. This quarantine order applied to the towns which were known to be infested by the insect, and prohibited the interstate movement, to points outside the quarantined area, of all corn fodder or cornstalks, whether used for packing or otherwise, green sweet corn, roasting ears, corn on the cob and corn cobs. No restrictions were placed on the interstate movement of any of the enumerated articles that had originated outside of the quarantined area and were shipped through it on a through bill of lading. Further investigation -will probably show the necessity for amending this quarantine order to include additional territory and other articles, plants or plant products liable to contain the insect. State Quarantine Measures. The Hon. Elbert S. Brigham, Commissioner of Agriculture of Vermont, learning of the dangers existing from the presence of the pest in Massa- chusetts, immediately sent his assistant, Mr. H. L. Bailey, to investigate the situation in the infested fields near Boston, and as a result the State of Vermont issued a quarantine notice, on account of the European corn borer, which became effective on and after Aug. 26, 1918. This quar- antine prohibited the movement of all stalks or ears of the corn plant {Zea mays), either green or dried, from the State of Massachusetts into the State of Vermont, unless written permission be secured from the Commissioner of Agriculture of the State of Vermont. This restriction did not apply to ordinary commercial dried shelled corn used for feeding purposes, nor to any corn grown in other States and sent through Massa- chusetts in transit. A similar quarantine to that by Vermont was established by the State of Connecticut, effective Sept. 20, 1918. Permits to ship corn on the ear, stover or other parts of the corn plant (except the shelled dry kernels, or cooked or preserved products, or corn grown in other States passing through the State of Massachusetts in transit) must first be obtained from the Director of the Connecticut Agricultural Experiment Station, and accompany each shipment. 10 MASS. EXPERIMENT STATION BULLETIN 189. GEOGRAPHICAL DISTRIBUTION. In the Old World. The European Corn Borer, or Corn Pyralid, P. nuhilalis Hbn., is widely distributed in central and southern Europe, west central and northern" Asia, China, Japan and the Philippine Archipelago. Hiibner, in his original record of the species, gave the habitat as Europe, western Asia, the Himalayas and Assam (British India) . A closely allied, if not, indeed, the same, species is reported from the East Indies (8). In the United States. At the present time^ the European corn borer, so far as is known, is found in the United States only in the counties of Suffolk, Middlesex, «JWllV/ L, Map showing area in Massachusetts infested by European corn borer, November, 1918. Heavy black line denotes limit of distribution. Essex and Norfolk, in the State of Massachusetts. Thirty-four towns are infested, comprising an area of approximately 320 square miles, or about three times the area believed to be infested after the discovery and preliminary survey of the situation in 1917. This area is located imme- 1 Nov. 30, 1918. THE EUROPEAN CORN BORER AND ITS CONTROL. 11 diately west and north of the city of Boston, Mass., and has as its limits the towns of Beverly, Danvers, Topsfield, Peabody, North Reading, Reading, Woburn, Lexington, Waltham, Newton, Brookline and Boston. (See map.) All the towns within these limits are infested to a greater or less degree. Granting that the section near the mouth of the Mystic River was the original point of entrance, it will be noted that the European corn borer has shown a decided tendency to spread in a northerly and northeasterly direction. This characteristic has been exhibited by other insects intro- duced from Europe, notably the gj'psy moth {Porthetria dispar L.) and the brown-tail moth {Euprodis chrysorrhoea L.). An examination of the meteorological records shows that during the periods when the adults of P. 7nibilalis are in flight, the prevailing winds are from the south and southwest. This may be the decisive factor in influencing the direction of the spread of P. nubilalis, as it is thought to be in the case of the other insects mentioned. The area given above is believed to represent very accurately the limits of the district as 3'^et invaded by the European corn borer. During the past season several men were engaged in determining these limits. In addition to this, the surrounding and contiguous territory in the States of Massachusetts, New Hampshire and Maine was examined for possible isolated infestations. Some other sections of these States were also ex- amined because of the fact that their trade with infested sections near Boston might have led to the involuntary introduction of the pest in infested plant products. This was especially true of the sumaner hotel districts in Maine and New Hampshire, to which shipments of sweet corn were frequently made that had originated in the badly infested market-garden districts near Boston. Territory examined in Massachusetts. All of northeastern Massachusetts was examined to the New Hamp- shire line, and as far west 3S Tyngsborough, Westford, Acton, Sudbury, Wayland and Natick. On the south and east the territory was examined lo Dover, Westwood, Canton, Randolph, Holbrook and Wejonouth. Special attention was given the sections adjacent to the large cordage factories located at Andover and at Plymouth, with the idea that the pest may have been imported with hemp consigned to these factories. No infestation was found, however, outside the limits of the area pre- viously designated. Several reports were received during the season that the European corn borer was present in wadeh' separated localities throughout the State. Care was taken to investigate all of these reports, but 9side from those originating within the known area of infestation, it was found that insects other than P. nubilalis were responsible for the reported injury. 12 MASS. EXPERIMENT STATION BULLETIN 189. Territory examined in New Hampshire. The entire southeastern section of New Hampshire, in addition to the summer hotel districts, was examined for evidences of the European corn borer by Mr. F. H. Gates of the Bureau of Entomology. Particular attention was given the following localities, viz.: Portsmouth and surroundings, including New Castle; Greenland, Rye and Rye Beach; Hampton and Hampton Beach; Dover and vicinity; Rochester and vicinity; Farmington and vicinity; Concord and vicinity; Hookset; Manchester and vicinity, including Goffs Falls and Amoskeag; Derry and Londonderry; Nashua and vicinity; Pelham; Windham; Epping; and Thornton. No evidences of the insect were found anywhere in New Hampshire. Mr. W. A. Osgood, assistant to the deputy commissioner of agriculture of the State of New Hampshire, reports that, during October, 1918, he made a survey of the towns in the State bordering on Massachusetts, but did not find any indication of the European corn borer. Mr. Osgood had previously visited the infested fields near Boston, and had become famiUar with the appearance of the pest. Territory examined in Maine. The following locaUties were examined in the State of Maine by Mr. R. H. Van Zwaluwenburg of the Bureau of Entomology for the possible presence of the European corn borer: Portland, — city and suburbs, including South Portland, Deering, Woodfords, Falmouth Foreside, Peak's Island and Great Diamond Island; Kennebunkport and Kenne- bunk Beach; Kittery; Wells Beach and village; Yarmouth; South Poland Springs and eastward to Danville Junction; Bath, — city and suburbs, including Woolwich; Rockland, — town and suburbs; Camden and Crescent Beach; Bar Harbor, — town and vicinity south to New- port Mountain and north to within a mile of Hull's Cove; Bangor, — city and suburbs, north to Mount Hope, south to Hampden Highlands and on east bank of the river south through Brewer to North Orrington; Augusta, — town and suburbs within a radius of 2 miles north and west, on east bank of river north to Riverside, east to Togus and south to opposite Hallo well; Hallowell; Gardiner; Lewiston, — city and suburbs; Auburn; Minot; and Mechanic Falls. No evidences of the pest were found in the State of Maine. During the progress of this survey Mr. Van Zwaluwenburg learned that considerable quantities of early sweet corn, originating in Massa- chusetts, had been shipped into Kennebunkport, Me., during the past few seasons. One retailer stated that he had recently received sweet corn, grown near Boston, that was infested with worms of some kind. The merchant had sold this shipment along with his other corn, however, and could give no testimony as to its ultimate disposal. A very careful THE EUROPEAN CORN BORER AND ITS CONTROL. 13 examination of this section failed to reveal the presence of the European corn borer. This incident, however, demonstrates that the coastal region from Portland south to York, in the State of Maine, should be very care- fully watched for the appearance of the species. Mr. John A. Roberts, Commissioner of Agriculture of Maine, reported in August that his assistants had inspected sweet corn offered for sale in the stores at Augusta, Me., and were not able to find any evidence of the borer. • Similar reports were received from Mr. Dudley of the same office, and from Mr. Batchelor of the Maine Agricultural Experiment Station. These gentlemen had previously visited the infested fields near Boston, and were familiar with the appearance of the insect. Territory examined in Rhode Island and Connecticut. Reports were received concerning the possible presence of the European corn borer in corn at Providence, R. I., but an investigation proved that the injury was caused by Papaipeina nitela Gn. A similar report, received from Putnam, Conn., was investigated and also proved erroneous. FOOD PLANTS. In the Old Wokld. The principal food plants of the European corn borer in the Old World are corn, hemp, hops and millet. Corn (both field corn and fodder corn) and hop plants are recorded as being more severely injured by the pest than any of the other commercial crops grown in Europe. Foreign literature also contains references to a. great variety of minor food plants, including heather (14); artemesia (13); nettles (13); oak- galls (15); kidney-bean pods (15); grapevines (18); thistle (18); giant weed, Arundo donax (12); pigweed, Amaranthus retroflexus (18); fuller's teazel, Dipsacus fullonum (18); virgin's bower, Clematis vitalba (18); and several species of wild grasses and weeds. In the United States (Massachusetts). At the present time corn (sweet corn, field corn and fodder corn) is practically the only valuable commercial crop which is seriously attacked by the European corn borer in Massachusetts, although other commercial crops are attacked bj^ the insect to some extent. Corn is undoubtedly the favorite food plant of the pest. In the absence of corn, and in badly infested areas, the insect habitually attacks and enters a great variety of other wild and cultivated plants. Judging from observations made on the feeding habits of the species during the seasons of 1917 and 1918, it would not be surprising to find it present in almost 14 MASS. EXPERIMENT STATION BULLETIN 189. any plant possessing a moderatelj^ soft, fleshy stem or stalk, or bearing a soft seed head during its early growth. Along the outer edge of the infested region, and in areas only recently invaded, the insect is almost always found exclusively in corn. In badly infested fields the corn plants are frequently inhabited by so many feeding larvse that all of the desirable plant tissue is quickly con- sumed, and under these circumstances the laj-vae must leave their original host and enter other food plants growing in the vicinity in order to obtain food. Many of the eggs and smaller larvse are sometimes dislodged from their original location on the corn plant and fall to the ground or upon other species of plants growing underneath, or between the rows of the corn, to subsequently infest these other plants. This character- istic often accounts for the great variety of infested plants found in the vicinity of badly infested corn fields. The early season corn plants become dry and hard during July and August. Many of these plants contain belated F. nuhilalis larvae of the first generation, as well as small larvae of the second. The comparatively soft tissue of late season plants growing in the vicinity often attracts the corn borer larvae from their original food plant. Plants other than corn, growing in areas planted to corn during the preceding year, frequently have eggs laid upon them by moths resulting from the overwintering larvae in the crop remnants of the preceding year. In other instances the moths drift into areas where corn plants are absent, and deposit their eggs upon the most attractive food plant at hand. It is believed, however, that the moths prefer to deposit their eggs upon corn. Another factor which is of interest in connection with selection of food plants is that the larva? prefer large healthy plants, growing in well- fertiUzed land, to small plants of the same species, growing under less favorable conditions. List of Food Flants. The following table will show the list of food plants in which the Euro- pean corn borer has been found in Massachusetts to date. This list has been compiled by dissecting the larva from each plant mentioned. Adults were reared in instances where the identity of the larva was in doubt. The plants are arranged in order, with regard to their preference as food plants by the insect. THE EUROPEAN CORN BORER AND ITS CONTROL. 15 Table I. — Food Plants of the European Corn Borer in Massachusetts. CoMiMON Name. Scientific Name. Part of Plant attacked. -ed) Sweet corn, . Field corn, . Fodder corn, Barnyard grass, . Pigweed (redroot). Dock, . Ragweed (hogweed), Lamb's-quarters, . Dahlia, . Foxtail. Lady's-thumb (smart Burdock, Horsewoed, . Beggar-ticks (bur marigold) Purslane (pussley), Crab grass, . Scouring rush Panic grass, Timothy, Goldcnrod, Thistle, Apple of Peru, Gladiolus, Chrysanthemum, Celery, . Swiss chard. Beans, . Potatoes, Tomatoes, Beets, . Spinach, Oats, . Turnips, Zea mnus, .... Zea jnays, .... Zca mays, .... Echinochloa crus-gaUi Beauv., Aniaranthus retroflexuf: L., . Rumex crispus L. and R. obtus folia L. Ambrosia spp., Chenopodium album L., Setaria glauca Beauv., . Pcil!jgj7ium persicaria L., Arctium minus L., Erigeron canadensis L., Bidens frondosa 1j., Portulaca oleracea L., . Digitaria sanguinalis Scop , Equisetum spp Panirum dichotomifloruw Michx, Phleuni pratense L., Solidago sp. L Cirsiurn spp., A'icandra physaloides L., All except root. All except root. All except root. All except root. Stalk and seed head. All except root. Stalk and seed head. Stalk and seed head. Stalk and flower stems. Seed heads. Stalk. Stalk. Stalk. Stalk. Stalk. Stalk. Stalk. Stalk. Seed head. Stalk. Stalk. Stalk. Stalk. Stalk. Outside stems. Stalk and midrib of leaves. Pods, green beans and the vines. Vines. Vines. Tops (stem and midrib of leaves). Tops (stem and midrib of leaves). Stalks. Tops (feeding on exterior of leaf stems). 16 MASS. EXPERIMENT STATION BULLETIN 189. Emphasis should be placed upon the fact that the great variety of food plants attacked will undoubtedly prove a serious complication in the problem of controlling the insect. Several of these food plants or their products, notably sweet corn (green), field corn (on the cob), celery, beet tops, beans (string beans), Swiss chard, oat straw (used as pacldng material), dahlias, gladioli and chrysanthemums, are commonly transported through the regular chan- nels of trade, and may easily serve as agencies for carrjdng the insect into new localities. CHARACTER AND EXTENT OF INJURY. Corn, The following explanation, concerning the terms herein applied to different parts of the corn plant, maj' be of assistance. The corn plant is monoecius, bearing both staminate (male) and pistillate (female) flowers, separate, but both occur on the same plant. The corn tassel bears the male flowers and the corn silks are the female flowers. The cornstalk consists of nodes (joints) and internodes (intermediate spaces). A single leaf grows from each node. Each leaf is composed of three distinct parts, — the sheath, the Ugula and the blade. The sheath is the part of the leaf surroimding the stalk, and, beginnmg at a node, extends up- ward nearly to the next node, where it joins the long narrow blade of the leaf. Although the sheath surrounds the stalk, the edges merely overlap and are never grown together. The ligula is a thin, upward continuation of the sheath, above its junction wdth the blade, at the point where the sheath ends and the blade begins. The blade is the broad, flat portion of the leaf. The pedicel is that portion of the plant by which the ear is attached to the stalk. The pith is the soft, cork-Uke substance filling the interior of the stalk, between the internodes. Kinds of Corn injured. In Massachusetts the larvse of the European corn borer have been observed to attack sweet corn, field corn and fodder corn. In the area now infested by the insect, sweet corn is grown to a greater extent than either field corn or fodder corn, and most of the observations herein recorded were made on sweet corn. Wherever field corn has been found within the infested area the plants have been attacked by the insect with the same degree of severity as has been sweet corn, and, due to its longer period of growth, the damage to the ears is much greater than to the ears of sweet corn. Only one field of fodder corn was located within the infested area, and this was attacked by the insect to a slight extent. This infestation was on the edge of the infested area in the town of Topsfield, where only an occasional larva of the European corn borer was found. THE EUROPEAN CORN BORER AND ITS CONTROL. 17 Injury to the Tassel. The newly hatched larva of the European corn borer fii-st attacks the unopened staminate buds of the tassel. After entering and feeding upon the internal succulent parts of several staminate buds, it enters the stalk 2 or 3 inches above the lower branches of the tassel, and tunnels upward for 2 or 3 inches. It then returns to its original entrance and tunnels toward the base of the plant. Within a few days the larva completely consumes the central pith of the tassel stalk, soon causing a break at the point where it originally entered. The broken-over portion of the tassel still remains partlj^ attached to the plant, and in this condition its yellow-white color and broken-over position make it a very conspicuous object in a field of corn in contrast to the green color and upright position of tassels not infested. This type of injury indirectlj^ affects the formation of corn kernels on the cob by greatly reducing the amount of pollen. In the process of fertilization, pollen from the tassel must fertilize the corn silk in order that kernels may develop. It is apparent that if pollen is not present in large enough quantities the resulting ear of corn will show a lack of fully developed kernels. Field counts made in badly infested areas showed that as high as 61 per cent of the corn tassels had been broken over and were barren of pollen. This high percentage of injury was more common on late corn than on early corn, due, perhaps, to the greater number of larvse present. Out of a total of 3,810 tassels, counted in a field of late season, sweet corn at West Medford, Mass., 2,344 tassels, or 61 per cent, were infested and broken over. Many ears of corn from this field were noticeably small in size and with few kernels, even though not themselves directly injured by the insect. Much of this loss is believed to have been caused by the injury to the tassel, although this belief is contrary to the opinion of botanists consulted. It is apparent that botanists must reverse their opinion in this matter. Injury to the Stalk. In nearly all cases the terminal internode, bearing the tassel, furnishes sufficient food for the full development of a single lai-va. Other larvse, if present in the same tassel, are forced to leave and tunnel in the lower parts of the plant for food. Their operations are generally confined to the upper two-thirds of the stalk, but, if numerous, they may extend their tunneling to the very base of the stalk, or even into the upper part of the taproot. When several larvse are feeding in the same stalk the pith is nearly, if not entirely, consumed, and the interior of such a stalk is found to be practically hollow. There is a tendency for the lai-va^ to work in the internodes of the stalk, but, when necessary, they commonly pierce, and feed upon, the nodes. This latter observation is contrary to published records on the habits of the species by European writers. A total of 75 corn plants, growing in a badly infested field at West 18 MASS. EXPERIMENT STATION BULLETIN 189. Medford, Mass., were carefully dissected and counts made of the larvae found therein, in order to secure data concerning the number infesting single plants. A maximum of 117 larvae, and a minimum of 7, with an average of 46 larv© per plant, were found in these 75 plants. These plants composed a total of 17 hills taken at random in different parts of the field. A maximum number of 311 larvae, and a minimum of 151, with an average of 206 larvae per hill, were found in these 17 hills of corn. The 17 hills of corn composed of 75 plants contained a total of 3,503 larvae. The actual count of one-eighth of an acre in tliis field showed a total of 2,855 plants, or 22,840 plants to the acre. Each of these 2,855 plants was infested to a greater or lesser degree. An average infestation of 46 larvae per plant, as shown above, means a total of 1,050,640 larvae of the European corn borer per acre of corn. Natiu-ally, this extensive injury to the interior of the cornstalk, to- gether with the numerous entrance and exit holes of the larvae on the surface, weakens the plant to such an extent that it soon breaks over and lies prone upon the ground. The supply of nutriment to the ear is also cut off, causing a small or aborted ear of corn. Even when only a few larvae are present within the plant, the growth of the stalk and formation of the ear are greatly retarded. The tunnels left by the larvae of the European corn borer frequently serve as sources of infection by various rots and fungi, so that the interior of badly infested stalks is sometimes found to be a mass of putrif^ang matter, occupied by various scavenger insects that have gained ad- mittance to the plant by way of the entrance or exit holes of P. nxibilalis larvae. Injury to the Ear. The indirect injury to the ear by larvae of the European corn borer has already been mentioned. This is caused (1) by interference with proper poUenization resulting from larvae cutting off the tassel, and (2) by inter- nal injury to the stalk, which cuts oft" the normal supply of nutriment to the ear. The ear, however, is also directly injured by the external and internal feeding of the larvae. Frequently the moths of the first generation, and habitually those of the second generation, deposit their eggs directly upon the silk of the ear. The newly hatched larvae feed first upon the silk, thus contributing to improper fertilization, and later they work their way down into the ear, where they tunnel through all parts of the cob and also feed upon the newly formed kernels. Sometimes eggs are deposited upon the exterior, or husk, of the ear, and the newly hatched larva feeds for a time upon the exterior of the husk before entering the ear, either at its tip end, or between the edges of the leaves of the husk. The ear is frequently entered by partly grown larvae, which have left some other plant or another part of the same plant. These larvae may enter the ear at any point, — its tip end, along the sides, or through the side of the pedicel. In other instances they tunnel directly from the in- THE EUROPEAN CORN BORER AND ITS CONTROL. 19 terior of the stalk through the pedicel and into the ear; consequently, the infested ears may not show external indications of injury. A combination of these larval habits may result in the presence of several larva? within a single ear. In one instance a total of 15 were found feeding on the interior and exterior of one ear. Extensive feeding of this nature reduces the ear to a soft, decaying condition, totally unfit for market, and unsuitable, even, for feeding to stock. This deteriora- tion is hastened by the introduction of various rots and fungi, which gain entrance to the plant through the holes made by the borers. Even when only a single larva is present within the ear, its feeding renders the ear unfit for market, while its use for seed, or for storage in cribs, is abso- lutely prevented, owing to the softened condition of the kernels and their tendency to quick decay. The percentage of ears infested in any given field depends upon the degree of infestation. An actual field count, made in a one-eighth acre plot of sweet corn located at West Medford, Mass., showed that, out of a total of 3,311 ears present in this plot, the entire number were infested, to a greater or lesser degree, by larvae of the borer. This plot was typical of most of the fields and small garden patches of sweet corn found in the territory where the pest has become well established. It serves as a standard by which to judge the amount of damage to corn that may be expected if the pest is not brought under control. Injury to the Leaf. Newly hatched larvae of the European corn borer may feed upon the upper or lower epidermis of the leaf blade before they enter the buds of the tassel. This type of injury is of no economic importance, except that it offers a possibility for poisoning the young larvae by application of arsenicals. Partly grown larvae infrequently tunnel into the midrib of the leaf blade, and also feed between the leaf sheath and the stalk. Summary of Injury to Corn. The economic injury to corn may be summarized as follows: — 1. Injury to tassel which results in poor fertilization. 2. Injury to stalk which reduces vitality of plant. 3. Injury to stalk which causes breaking over of plant. 4. Injury to stalk which indirectly affects ear. 5. Injury to ear which directly affects the yield. 6. Injury to silk of ear which results in poor fertilization. Other Food Plants. Dock. In the absence of corn, dock is a common food plant of the first genera- tion of European corn borer larvae. The plant is represented by at least two different species in the area infested by P. nubilalis, and both species 20 MASS. EXPERIMENT STATION BULLETIN 189. are attacked by the insect. It grows plentifully as a weed in cultivated areas, and also in waste places, generally preferring rather moist soil. The newly hatched borer feeds first upon the tender seed head, or upon the epidermis of the tender leaves. As the larva develops it tunnels through the leaf petiole, and when about half grown enters the main stalk. It then usually tunnels downward, feeding through nodes and internodes, and consuming in its progress nearly all the interior of the stalk. This causes a weakening of the plant which soon breaks over at the point where the larva entered. The broken-over portion soon dies and turns brown in color, thus rendering it a very conspicuous object among plants not infested. A mass of conspicuous yellowish-white frass, extruded by the larva within, generally adheres to the point in the stalk where the larva entered. This serves to distinguish plants infested by P. nuhilalis, even in instances where the plants do not break over. The number of dock plants per acre is generally rather limited, so that all plants of this species in a given area are commonly infested, depending, of course, upon the degree of infestation. Economically, dock is important in that it serves as an early season host plant for the European corn borer in areas where corn is absent. The second generation adults emerging from dock deposit their eggs upon late corn and other commercial crops. Barnyard Grass. Barnyard grass is the most important and the most commonly infested weed among the uncultivated hosts of the European corn borer. All parts of the plant, except the root, are fed upon by the larv^a, including the seed head, the leaves and the stalk. Barnyard grass grows luxu- riantly in almost any waste area of ground, or in the spaces between economic plants in cultivated fields. It seems to prefer well-fertilized soil, and under favorable conditions may reach a height of 5 or 6 feet, with a diameter at the base of nearly half an inch. It is very abundant in all parts of the area infested by the European corn borer, and serves as a food plant for both generations of larvae. The newly hatched larvse feed for a short time upon the green buds of the seed head, and also upon the upper or lower epidermis of the leaves. They soon enter the main stalk of the plant, however, and tunnel upward or downward according to their individual preference. A dozen or more are sometimes found in each stalk, and as the stalks grow very thickly clustered t(>gether in clumps, a foot or more in diameter, the aggregate number of larvae infesting each clump of barnyard grass often equals the number normally found in a hill of badly infested corn. Many areas of vacant land, large or small in extent, throughout the infested region, are thicklj'^ covered by barnyard grass clumps of this description, which contain untold numbers of the depredating larvae. Owing to the small diameter of most barnyard grass stalks, the tun- neling of the larva leads to an early collapse of infested stalks, which THE EUROPEAN CORN BORER AND ITS CONTROL. 21 soon fall to the ground. This forms a mass of intertwined plants very difficult to remove or destroy during clean-up operations. The chief economic signincance of barnyard grass as a food plant of the European corn borer lies in the fact that it serves as a common host of the insect, and aids in its multiplication and distribution in areas where corn is absent. Pigweed. Pigweed, or redroot, is commonly found growing among cultivated crops, or closely adjacent thereto. It generally serves as a sort of over- flow host plant to accommodate the larger larvaj of the corn borer which have left their original host plant and are seeking other food. In rare instances newly hatched larvae are found feeding upon the green seed heads of this plant. This is generally caused by the dislodg- ment of these larva? from their original host. More commonly the plant is attacked by good-sized larvae which have partly completed their development in other food plants. The stalk is entered at any point along its surface, and the larva tunnels upward or downward in the same manner and with the same results as have been mentioned for other food plants. Pigweed is not generally infested by the European corn borer with the same degree of severity as are dock and barnyard grass, although it is important economically as an intermediate host of the insect, and may act as a host in the absence of more favored food plants. Ragweed and Lamb's-quarters. Ragweed, or hogweed, and lamb's-quarters serve as food plants for the European corn borer in the same manner and extent as has been described for pigweed. The larvse attack the green seed head and stalk of each of these plants. Lamb's-quarters sometimes grows to a height of 4 or 5 feet, and develops a tough, wood}/ stalk an inch or more in diameter. It is perhaps the hardest and toughest stalk in which the larvse of the European corn borer have been found. Both ragweed and lamb's-quarters are found widely distributed through- out the infested area, although the number of plants found in a given space is generally small. Dahlias. Larvse of the European corn borer tunnel through the main stalk and flower stems of dahlias during the late summer and fall. The percentage of dahlias in a given area, infested by the larvse, is generally very high. In Arlington, and other towns adjacent to Boston, almost every group of dahlia plants was found to be infested by P. nubilalis during the past summer. Small larvae are rarely found in dahlias, most of the damage being done by those which have hatched and fed for a time on other plants in the vicinity, and are about half grown when they enter the dahlia plants. Entrance may be effected at almost any place along the 22 MASS. EXPEKIMENT STATION BULLETIN 189. main stalk or flower stem, but the favorite point is at the junction of flower stems with the main stalk. The tunneling larva soon consumes the interior of the infested stalk or stem, and that portion hrst wilts and then breaks over in a dying condition. It is then verj^ conspicuous in contrast to the stems not infested, and ruins the appearance of dahha plantings. Half a dozen or more larvae hpve been cut from a single dahlia flower stem. The principal point to be considered in connection with the infestation of dahha plants by larvse of the borer is that the species may possibly be disseminated through the medium of cut flowers. Chrysanthemum and Gladiolus. The stalks of chrysanthemums and gladioU are tunneled by larvse of the European corn borer in a similar manner and with the same results as has been described for dahlias. Infested chrysanthemum stalks are commonly found in out-of-door gardens during the late summer and fall, and also in greenhouse plots. This characteristic renders chrysan- themums economically important because of the possibility that the pest may be accidentally spread by transporting recently infested plants which have not yet shown external effects of the larval injury. Infested gladiolus stalks are found in out-of-door gardens during the late summer, and though not as important economically as chr3^santhe- mums, this plant may also be a source of danger through the accidental transportation of infested plants to areas not yet inhabited by the pest. Timothy and Foxtail. Small larvse of the European corn borer have frequently been found feeding upon the seed heads of timothy and foxtail. This damage is not important economically, except that it affords a host for the larvae of the pest until they have reached a stage in their growth when they are large enough to attack other food plants. Larvse of the species have never been observed to feed within the stalks of these plants, and the plants are never noticeably injured. Miscellaneous Plants. The stalks of lady's-thumb, burdock, horseweed, beggar-ticks, purslane, crabgrass, mare's-tail, panicgrass, goldenrod, thistle and apple of Peru are often entered and tunneled by partly grown larv'ae of the European corn borer. These plants are rather numerous in restricted areas through the infested region, and serve as intermediate hosts of the borer, although the plants themselves are of no economic importance. Celery. Nearly full-grown larvae of the borer have been observed to enter and tunnel the outside stems of celery plants. This injury, however, has THE EUROPEAN CORN BORER AND ITS CONTROL. 23 been observed in only one field, and in this instance the celerj^ was growing adjacent to a very badly infested field of sweet corn. This corn was inhabited by so many larvae that the food supply was apparently exhausted, and the larvae were attracted to the green succulent stems of the celery plants. Several were commonly found in each of the outside stems, but none were found in the stems near the center of the plant. Similar circumstances to those which resulted in this infestation may be expected to occur from time to time, as celery is frequently grow^n adjacent to or between the rows of corn plantings. Celery may be considered an important economic food plant of the European corn borer because of the possibility that plants containing infested stems may be shipped outside the infested area. SicifiS Chard. The stalk and midrib of the leaves of Swiss chard plants were fre- quently found infested by the borer under the same circumstances and with the same result as has been recorded in the instance of celery. The injury to Swiss chard, however, was observed in a number of fields in widely separated localities. The green stalks and leaves of this plant are commonly shipped from town to town and must be considered as sources of danger. Beans. The pods, immature beans and interior of the vines of bean plants were found infested by larvse of the European corn borer in several fields. This generally occurred in instances where several crops were planted together, and the bean plants served to accommodate the overflow larvse from other food plants. The infestation was always found to be very light in character. Under exceptional circumstances the bean plant may become important economically as a host of the borer because of the possibility that larvse of the species might be transported within the immature pods of string beans. Potatoes and Tomatoes. In badly infested areas the larger larvse of the European corn borer may occasionally be found tunneling the stems of potatoes and tomatoes. Not more than a single larv'a has ever been observed within a plant, and the injury, so far as observed, is very slight and not at all important commercially. Beets and Spinach. Larvse of the European corn borer are infrequently found tunneling within the leaf stems of beets and spinach during the early fall. This type of injury may be of economic importance because of the possi- bility that infested plants may be transported for use as greens. 24 MASS. EXPERIMENT STATION BULLETIN 189. Oats. The stalks of volunteer oats were found infested by the larger larvse of the borer in one instance. The injury and its results were similar to that described for other plants with a like habit of growth (pigweed, etc.). Only a very small percentage of oat stalks present was infested. Oats may become important economically as a food plant of the borer because of the fact that oat straw is often used as packing material. Turnips. Large larvae of the European corn borer were observed feeding upon the outside surface of the tender leaf stems of the turnip. They were not found within the turnip plants, and it is believed that this plant is not at all important as a host of the borer. DESCRIPTIONS OF THE DIFFERENT STAGES. The Egg. Average length, .97 millimeter; average width, .74 milliineter; circu- lar ovate in shape, slightly convex on its upper surface, flat on its lower surface, or conforming to the shape of the object on which it is deposited. Exochorion sculptured with shallow pentagonal or polygonal pits. Endo- chorion apparently smooth. Color, when first deposited, opaque white, often strongly iridescent. In from eighteen to twenty-four hours after deposition a crescentiform clear space is formed m the center of the egg on its upper surface. About two days before hatching the egg assumes a yellowish tinge, and soon thereafter the developing larva becomes visible and imparts to the egg a yellow-black appearance. The eggs are commonly deposited in irregular-shaped masses, each egg overlapped by the adjacent ones in the manner of shingles. Each egg mass is composed of from 5 to 50 eggs. The Lakva. First Instar (see Plate I, Fig. 1). — Average measurements of 11 indi- viduals, newly hatched. Length, 1.6 millimeters; head width, .30 millimeter. Length of head and prothoracic shield, one-fourth total length of larva. Body subcylindrical, opaque white to yellowish green in color. Tubercles large, prominent, pale amber gray. Primary setae long, amber-colored. Anterior stigmatal tubercle on prothorax bisetose, the upper seta the shorter; sub ventral tubercle also bisetose, the anterior seta the shorter. Tubercles and setae iv and v are absent on meso- thorax and metathorax; coalescent on abdominal segments 1 to 8, inclu- sive; situated below the spiracle on segments 1 to 7; below and slightly anterior to spiracle on segment 8. Tubercles ia-i6 and iia-ii6 coalescent on mesothorax and metathorax. Seta? ia and iia are shorter than setae THE EUROPEAN CORN BORER AND ITS CONTROL. 25 i& and ii&. Seta iii is of medium length. On the dorsum of abdominal segments 1 to 7, tubercles i and ii form a trapezoidal figure, while on the dorsum of segment 8 they form a nearly rectangular figure. On the dorsum of abdominal segment 9 is a large irregular-shaped, nearly oblong, corneous tubercle bearing a long seta at ea^h of its posterior lateral angles, and a distinct pmicture on the median anterior border. The nearly elliptical preanal plate bears two short setae and one long seta along each of the posterior lateral angles, and one short seta centrally located on each side of the median line. Spiracles protruding, concolorous with tubercles. Head black or dark brown, declivous and flattened in the newly hatched larva, becoming more rounded as the larva develops. Adfrontal pieces not perceptible in this or succeeding instars until the fifth. Clypeus pale and distinct from frontal piece. Labrum pale, bilobate, with normal arrangement of seta?. Mandibles reddish brown, not protruding. Ocelli six in number, pale and protruding. Antennae with slight tinge of amber on distal segments. Prothoracic shield averages .25 millimeter, slightly lighter in color than the head, corneous, almost straight anteriorly, broadly rounded posteriorly. Each half of the shield bears three setse on the anterior border, two on the lateral posterior border, and one centrally located and near the median line. Bases of setse surrounded by a black ring. A perceptible indentation, but no division of shield along the middorsal line. Venter of prothoracic segment appears darker owing to presence of dark thoracic shield above. Thoracic legs, abdominal prolegs, preanal plate and anal prolegs amber. Circle of crotchets on abdominal prolegs broken externally. Thoracic feet and crotchets on abdommal and anal feet pale brown. Thoracic feet corneous. Second Instar (see Plate I, Fig. 2). — Average measurements of 13 individuals, just molted: Length, 2.625 millimeters; head width, .46 millimeter. Length of head and prothoracic shield, one-sixth total length of larva. Body subcjdindrical, amber-white to yellowish green in color. Tubercles large, prominent, pale amber, polished; iv and v present and coalescent on mesothorax and metathorax. Othei-wise the arrangements of tubercles and setse are similar to preceding instar, and remain fairly constant throughout the remaining larval stages. The relative length of the longer body setae diminishes in each succeeding instar. Spiracles pale amber at center, with black edges. Bases of tubercles and setae surrounded by a black ring. Head deflexed. Clypeus pale and distinct from frontal piece. Labrum pale brown. Mandibles reddish brown with black tips. Distal segments of antennae pale amber, otherwise colorless. Prothoracic shield averages .414 millimeter wide or nearly equal to that of head. Indentation along middorsal line more pronounced but no division. Venter of prothoracic segment darker. 26 MASS. EXPERIMENT STATION BULLETIN 189. Thoracic feet and crotchets on abdominal and anal feet dark brown. Preanal plate pale amber. Third Instor (see Plate I, Fig. 4). — Average measurements of 16 indi- viduals, just molted: Length, 4.75 millimeters; head width, .68 milli- meter. Body subcylindrical and darker than preceding instar. Ab- dominal segments, except 9 and 10, crossed transversely by shallow grooves. Anterior stigmatal and subventral tubercles of prothorax contiguous, nearly concolorous with head and somewhat corneous. Remaining body tubercles as before. Head deflexed, dark brown in color. Clypeus nearly concolorous with head, and not so distinct from frontal piece. Labrum pale brown. Prothoracic shield averages .71 millimeter wide. Line of division down middorsal line semi-distinct for one-half distance from anterior border. Venter of prothoracic segment dark. Thoracic legs, abdominal legs and preanal plate as before. Fourth Instar (see Plate I, Fig. 6). — Average measurements of 12 individuals, two or three days after molting: Length, 12.5 milhmeters; head width, 1.03 millimeters. Body cylindrical, varies in color from opaque white to pale or dark amber. Some individuals show, indistinct median and subdorsal longitudinal reddish brown or gray lines on the dorsum. Tubercles of medium size, arranged similar to second instar. Prothoracic tubercles not contiguous, slightly darker than remaining body tubercles. Spiracles nearly concolorous with tubercles. Head slightly paler than preceding instar. CljT^eus not distinctly marked off from front, concolorous with head, trapezoidal; average height, .12 millimeter, average width, .47 millimeter. Labrum dark brown. Mandibles dark brown, not protruding. Distal segments of antennse dark amber, otherwise nearly colorless. Prothoracic shield averages .98 millimeter wide, distinctly di\'ided along middorsal line, slightly paler in color than before, and often assuming a yellowish tinge on anterior border. Venter of prothoracic segment slightly darker than venters of remaining segments. Fifth Instar (see Plate II, Fig. 8). — Average measurements of 13 individuals, three daj'S after molting: Length, 14.46 millimeters; head width, 1.66 millimeters. Body cylindrical, varies in color from dusky opaque white to light pink, with distinct median and subdorsal longi- tudinal reddish brown, gray or pink lines on the dorsum. Tubercles medium and distinct, pale at center and surrounded by a dusky black ring which in turn is surrounded, on the abdominal segments, by a wider, pale amber-colored band. Tubercles on thorax uniformly dark amber, same arrangement as before. Spiracles nearly concolorous with tubercles. Head polished dark brown, not quite as high as wide. Clj'peus dis- tinctly marked off from front, central area paler than head; average height, .18 millimeter, average width, .66 millimeter. Adfrontal pieces distinct for first time and extend to the vertex. Labrum dark brown THE EUROPEAN CORN BORER AND ITS CONTROL. 27 at base, paler at free edge. IMandibles dark brown, protrude slightly. Distal segments of antennte amber, otherwise colorless. Prothoracic shield averages 1.72 millimeters wide, more distinctly- divided than preceding instar. General color pale brown to pale yellow, polished, with dark brown areas. Anterior border pale yellow. The median posterior margin bears a triangular area, and two large irregular areas are present in a shallow depression near the lateral corners of the shield. The posterior and lateral margins of the shield are dark browm. Bases of setse surrounded by a distinct black ring. Venter of prothoracic segment only slightly darker than venters of remaining segments. Prothoracic legs concolorous with head, mesothoracic legs dusky ex- ternally, metathoracic legs pale amber. Abdominal and anal legs as before. Sixth Instar (see Plate II, Fig. 10). — Average measurements of 9 individuals, four days after molting: Length, 19.95 millimeters; head width, 2.19 millimeters. Body cylindrical, abdominal segments, except 9 and 10, crossed transversely by deep grooves. General color darker than preceding instar, varying from dusky pale brown to dark brown or pink. Median line narrow, dark brown and very distinct; subdorsal line vague in outline, broad, pale brown or pink; lateral lines narrow, pale brown. Tubercles medium and darker than general color of body, more pronounced on thorax. Arrangement of prothoracic tubercles and setae as before. Tubercles ia-i6, iia-ii6 and iv-v are coalescent on meso- thorax and metathorax. Setse ia and iia are very much shorter than setse lb and iib. Seta v is very much shorter than seta iv, while setae iii and vii are of medium length. Tubercles iv-v are coalescent on ab- dominal segments 1 to 8, inclusive, situated below the spiracle on segments 1 to 7; below and slightly anterior to spiracle on segment 8. On dorsum of abdominal segments 1 to 7, tubercles i and ii form a trapezoidal figure, as before, while on dorsum of segment 8 these tubercles form nearly a rectangular figure. Large corneous tubercle on dorsum of segment 9, and preanal plate on dorsum of segment 10, with setal arrangement as before. The setse on lateral anterior borders of prothoracic shield, setse ib and ii6 of mesothorax and metathorax, the setae on dorsal tubercle of segment 9, and the long setae on preanal plate are nearly twice as long as any others present. Head polished brown, with pale brown areas on epicranial lobes. Clypeus as before; average height, .27 millimeter, average width, .84 millimeter. Adfrontal pieces more distinct. Labrum, mandibles and antennae as before. Prothoracic shield averages 2.34 millimeters wide. Colored areas on shield similar to preceding instar, with additional small pale brown depressions along each side of median line. The position and form of these colored areas are variable. Venter of prothorax concolorous with venters of remaining segments. Thoracic, abdominal and anal legs as before. 28 MASS. EXPERIMENT STATION BULLETIN 189. The Pupa. Average length of ^ , 13 to 14 millimeters; of 9 , 16 to 17 millimeters. Average width of ^ , 2 to 2.5 millimeters; of 9 , 3.5 to 4 millimeters. Color varies from light to dark brown, venter comparatively smooth, dorsum darker in color with fuie transverse wrinkles. Form elongate with peculiar "shouldered" appearance of the body, caused by the great width of the thorax as compared with width of the head. Appendages firmly cemented to the body. Wings, maxillae, antennae and mesothoracic legs, together with metathoracic legs which he beneath, are approximately equal in length and extend to the middle of the fourth abdominal segment. Prothoracic legs terminate midway between the head and the tip of the other appendages. Dorsum of thorax very dark, not shiny, with a distinct smooth, slightly elevated ridge extending along the dorso-median line. The fifth, sixth and seventh abdominal segments bear a ridge near the anterior border, which extends completely around the segment. On the dorsum of each of the fourth, fifth and sixth abdominal segments is a transverse line of four bicuspidate projections of the body wall. A pair of proleg scars are visible on the venter of the fifth and sixth abdominal segments. The last segment of the pupa terminates in a dark brown, or black, cremaster, which bears at its extremity eight small spines, ar- ranged transversely, which curve forward at their tips and serve to attach the pupa to its cocoon. Length of these curved spines about .19 milli- meter. Spiracles ellipsoidal, prominent and borne on abdominal seg- ments 2 to 7, inclusive. The pupa is always enveloped in a thin cocoon. The terminal segments of S and 9 pupae differ in shape and in arrange- ment of plates. The Adult. Alar expanse: male, 24-26 millimeters; female, 29-32 miUimeters. Length of body, 13-14 millimeters in both sexes. Head above covered with light yellowish brown scales, except adjacent to compound eyes, where scales are white; ventral surface, white. Labial palpi porrect; second segment covered with dense projectmg, cinnamon- brown to light brown scales attenuated to a point forward; terminal segment concealed; basal segment covered with white scales. An im- aginary line, passing through the axis of the body tangent to the lower edge of the compound eye, will divide the labial palpi into two portions according to their coloration, the upper portion being cinnamon-brown, the lower portion white. Maxillary palpi Hght brownish, erect, slightly dilated and converging at apex. Tip of labial palpi and maxillary palpi the same color in female, labial palpi somewhat darker in male. Pro- boscis long, with cream-colored scales, usually tightly coiled and almost completely hidden by labial palpi when viewed from the side. Antennae filiform, two-thirds the length of the front wing, with a longitudinal stripe of cream-colored scales on the posterior side; opposite side brownish. THE EUROPEAN CORN BORER AND ITS CONTROL. 29 Terminal half of antenna often curled in presented specimens. Ocelli present. Dorsum of thorax cinnamon-brown in male, light j'ellowish brown in female. Fore legs white exteriorly, fuscous internally. Ventral surface of thorax, mesothorax and metathoracic legs covered with white hairs and scales in both sexes. Inner spurs twice the length of outer ones. Fore wings as wide as hind wings, costal margin gently curved toward apex, anal angle rounded, inner margin straight. Fore wing of female dull yellow, the costa and inner two-thirds of wing more or less streaked with dull brown; a serrate brown line crosses the wing at about its outer third, followed externally by a narrow yellow band, the outer margin of which is also serrate; external to this is a brown band shot through with yellow toward the outer margin. Hind wing grayish brown, with a rather broad, pale band at the outer third, begin- ning a little behind the costa and extending nearly to the hinder margin. In some specimens the fore wing colors are dull yellow and cinnamon- brown, and the hind wings very pale brown with faint irregular streaks or shades of darker, instead of as described above; beneath, pale, with faint reproduction of the yellow band on the fore wing, its margins darker but not serrate. Male fore wing somewhat more reddish brown with a yellow discal spot, and a yellow serrate band at the outer third beginning a little behind the costa and often cut into outwardly by inward exten- sions from the darker color outside, tending to break it into a row of lunate spots; hind wings more gray, with the band of the female hind wing tending to disappear at its ends and become a large, elongate, rather oval area; beneath, dark, with a faint reproduction of the light band of the fore wing and a lighter shade corresponding to the oval area of the hind wing; also light along the inner margin over quite a width. Fore wing (Plate II, Fig. 12): la very weakly developed, bending slightly forward toward lb at the basal fourth of the latter; 4 and 5 fairly near at base, 5 arising considerably behind the middle of the outer end of the cell; cross vein closing end of cell nearly obsolete from 5 forward; 7 and 8 about as near each other at base as 4 and 5, 8-9 arising from the end of the cell, but almost in contact with 10, which it follows closely for some distance before diverging and forking, 8 extending ahnost exactly to the apex. Base of lb enlarged, bearing a tuft of long, forwardly directed hairs beneath. Hind wing (Plate II, Fig. 13) with three anal veins; veins 3, 4 and 5 arising close together; cross vein forming outer end of cell strongly re-entrant: vein 6 leaving the cross vein just before it unites with 7-8. Frenulum in male consists of one long, stout spine; in female (Plate II, Fig. 14), of two long spines and a shorter, more slender one. Ventral surface covered with whitish scales. Dorsum of male cinnamon-brown (excepting first two segments which are amber yellow); of female, amber yellow, the posterior border of each segment with a fringe of white. 30 MASS. EXPERIMENT STATION BULLETIN 189. LIFE HISTORY. First Generation. , Incubation Period. The eggs are deposited in masses of from 5 to 50 on the under surface of the upper blades of corn or other food plants. They hatch, on an average, in 7 days, with a maximum of 9 days and a minimum of 5 days (see Table II), the duration of the incubation period depending somewhat upon temperature conditions. Table II. — Duration of Incubation Period - - First Generation. Date of Depo- sition, 1918. Date of Hatch- ing. 1918. Incuba- tion Period (Days). Date of Depo- sition, 1918. Date of Hatch- ing. 1918. Incuba- tion Period (Days). May 24. May 25. May 26, May 26. May 26. May 27. May 28. May 28. May 28. May 28, May 28. May 29, liay 29. May 29. June 2. June 3, June 3, June 3, June 3, June 3, June 4. June 4, June 4. June 4, June 4, June 6, June 6, June 6, 9 9 8 8 8 7 7 7 7 7 8 8 8 May 29. May 31. June 1, June 1. June 2, June 2. June 3. June 3. June 4. June 5, June 6, June 7. June 8, June 9. June 5 June 6 June 7 June 6 June 8 June 8 June 10 June 11 June 11 June 12 June 15 June 15 June 16 June 18 8 9 Average length of incubation period. Maximum length of incubation period, Minimum length of incubation period, 7.43 days. 9 days. 5 days. Larval Period. In the course of their development the larva? feed upon, and within, various parts of their food plant, and pass tlirough from five to eight instars. Out of a total of 20 individuals reared from egg to pupa, in life-history cages 14 individuals required five instars to complete their larval growth, 3 required six instars, 2 required seven instars and 1 indi- vidual eight instars. It is probable that, under field conditions, there are normally five or six instars in this generation. In 20 life-history cages the average duration of the first instar was THE EUROPEAN CORN BORER AND ITS CONTROL. 31 7.25 days; second instar, 6 days; third iiistar, 5 days; fourth instar, 6.5 days; fifth instar, 13 days; sixth instar, 14 days; seventh instar, 8 days; and eighth instar, 13 days. The average duration of the total larval period was 44 days, with a maximum of 57 and a minimum of 35 days (see Table III). The duration of each instar and the total duration of the larval period depend upon temperature conditions. After reaching full growth the larva forms a cocoon within which it pupates. Table III. — Duration of Larval Instars — First Generation. Pigweed (Amaranthus). DcRATiov OF Larval Instars in Days. Date of Pupa- tion, 1918. Days in Larval Period. Date of Hatching, 191S. 2 I ■6 1 1 Sex. June 4, . June 10, . June 10, June 10, . 6 7 7 6 0 5 5 4 5 6 6 7 8 7 6 5 7 7 5 6 14 10 7 6 11 7 died 13 July 19 July 26 Aug. 4 45 46 55 9 9 9 Dock (Rumex). June 15, . 7 6 6 30 - - - Aug. 7 53 c? June 15, 7 6 9 19 - - - July 30 45 9 June 15, 9 6 5 18 - - July 31 46 9 June 15, 10 8 12 - - - July 23 38 d' June 15, 7 11 29 - - - Aug. 11 57 - June 15, 8 8 4 28 - - Aug. 11 56 9 June 15, 6 6 9 19 - - - July 30 45 d' June 16, 10 5 8 9 - - - July 22 36 9 June 16, 10 7 15 - - July 29 43 9 June 16, 10 6 13 ■ - - July 23 37 9 June 16, 10 7 8 - - - July 22 38 9 June 16, 9 5 13 - - - July 22 36 cf June 16, 9 5 12 - - - July 21 35 cf- June 16, 10 7 13 - - - July 28 42 9 June 10, 10 5 17 - - - July 29 43 cf June 16, 10 7 17 - - - July 29 43 9 Averag 3, 7.25 6.5 13 14 8 13 - 44 - Average duration of larval period 44 days. Maximum duration of larval period, .57 day.s. Minimum duration of larval period, 35 days. 32 MASS. EXPERIMENT STATION BULLETIN 189. Pupal Period. Pupation generally occurs within the tunnels made by the larva, although occasionally it occurs in masses of larval frass, or between closely attached leaves. The duration of the pupal period, in the instance of 49 individuals confined in life-history cages, averaged 8.5 days, with a maximum of 10 and a minimum of 7 days, depending upon temperature conditions (see Table IV). Table IV. — Duration of Pupal Period — First Generation. Number of Observa- tion. Date of — Num- ber of Days. Sex. Number of Observa- tion. Date of — Num- ber of Days. Pupa- tion. Emer- gence. Pupa- tion. Emer- gence. Sex. 1-120 July 15 July 24 9 d- 26-147 July 21 July 29 8 cf 2-121 July 15 July 25 10 <^ 27-149 July 22 July 29 7 9 3-122 July 16 July 24 8 cf 28-150 July 22 July 30 8 9 4-123 July 16 July 25 9 c? 29-151 July 22 July 30 8 9 5-124 July 16 July 25 9 c^ 30-152 July 22 July 30 8 9 6-125 July 16 July 25 9 cf 31-153 July 22 July 30 8 9 7-126 July 16 July 26 10 9 32-154 July 23 July 31 8 d> 8-129 July 17 July 25 8 9 33-155 July 23 July 31 8 & 9-130 July 17 July 25 8 9 34-156 July 23 July 31 8 9 10-131 July 18 July 26 8 9 35-157 July 23 July 31 8 cf 11-132 July 18 July 25 7 9 36-158 July 23 July 31 8 9 12-133 July 18 July 27 9 cf 37-159 July 23 Aug. 1 9 cf 13-134 July 18 July 25 7 9 38-160 July 23 July 30 7 9 14-135 July 19 July 29 10 cf 39-161 July 23 July 30 7 9 15-136 July 19 July 27 8 9 40-162 July 24 Aug. 2 9 cf 16-137 July 19 July 28 9 c? 41-163 July 23 Aug. 2 10 cf 17-138 July 18 July 27 9 9 42-164 July 24 Aug. 2 9 9 18-139 July 19 July 28 9 9 43-165 July 24 Aug. 3 10 9 19-140 July 19 July 28 9 9 44-166 July 25 Aug. 4 10 9 20-141 July 19 July 28 9 9 45-167 July 25 Aug. 4 10 9 21-142 July 20 July 28 8 9 46-168 July 26 Aug. 4 9 cf 22-143 July 20 July 27 7 9 47-169 July 27 Aug. 5 9 9 23-144 July 20 July 28 8 9 48-171 July 27 Aug. 6 10 cf 24-145 July 20 July 28 8 9 49-172 July 27 Aug. 5 9 9 25-146 July 20 July 29 9 d" -Average length of pupal stage, ........ 8.551 days. Maximum length of pupal stage, 10 days. Minimum length of pupal stage 7 days. THE EUROPEAN CORN BORER AND ITS CONTROL. 33 Adult Period. Soon after emerging from the pupa the female moth begins the ovi- position of second generation eggs. With 13 females, confined in indi- vidual life-historj^ cages, the average duration of the period, between emergence from the pupa and the first oviposition, was 3.2 days, with a maximum of 9 days and a minimum of 1 day (see Table V). Table V. Oviposition by Female Moths in Rearing Cages — First Generation. Sex. Date of — Number of Days — Number of Moths. d 9 Emer- gence, 1918. First Ovipo- sition. Last Ovipo- sition. Before Ovipo- sition. Of Ovipo- sition. From Emer- gence to last Ovi- position. Total Num- ber of Eggs. 2 July 25 July 29 Aug. 19 4 22 25 494 3 July 27 July 29 Aug. 14 2 17 18 590 3 July 27 July 29 Aug. 11 2 14 15 510 3 July 27 July 29 Aug. 3 2 6 7 415 2 July 29 Aug. 1 Aug. 11 3 11 13 594 2 July 29 July 30 Aug. 16 1 18 18 592 2 July 29 July 31 Aug. 10 2 11 12 132 2 July 29 July 30 Aug. 8 1 10 10 626 2 July 30 Aug. 3 Aug. 26 4 24 27 280 2 July 30 Aug. 8 Aug. 21 9 14 22 602 2 July 30 Aug. 3 Aug. 25 4 23 26 786 2 July 30 Aug. 2 Aug. 13 3 12 14 559 2 July 30 Aug. 4 Aug. 16 5 13 17 903 29 16 13 - - - - - - - Average 3.2 15 17.23 545 Maximum, 9.0 24 27.00 903 Minimum, 1.0 6 7.00 132 The duration of the oviposition period of these 13 females averaged 15 days, with a maximum of 24 and a minimum of 6 days (see Table V). The average length of life of 23 female moths, confhied in cages with male moths, approximating field conditions as nearly as possible, was 18 daj'S, with a maximum of 28 and a, minimum of 6 days. The average length of life of 27 male moths in these same cages was 14 days, with a maximum of 35 and a minimum of 3 days (see Table VI). 34 MASS. EXPERIMENT STATION BULLETIN 189. Table VI . — Length of Life of Male and Female Moths in Captivity — First Generation. Length of Life in Days. Number of Male Moths. Number of Female Moths. Length of Life in Days. Number of Male Moths. Number of Female Moths. 3 5 6 8 9 10 11 12 13. . . . 14 16, 2 2 2 1 2 3 1 1 2 1 2 2 17 18 19 23, . . . . 24 26 27 28 34.^ 35 Totals. . 2 2 3 1 1 1 2 2 27 23 Average length of life: male moths, 13.74 days; female moths, 18.26 days. Maximum length of life: male moths, 35 days; female moths, 28 days. Minimum length of life: male moths, 3 days; female moths, 6 days. It is believed that the duration of adult life, as well as the period before and during oAdposition, depends considerably upon the accessibility of the opposite sex, temperature conditions, and the facilities afforded for oviposition. Nevertheless, the data given above were secured under as near natural conditions as could be arranged in cages, and the averages are believed to represent very closely the actual duration of adult periods in the field. These figures are important, showing as they do the com- paratively long period during which the adults deposit their eggs. Life Cycle Summary. A complete life cycle is here considered to be the total period elapsing from the deposition of eggs of one generation to the time of deposition of eggs of the next generation. Therefore the average duration of the life cycle of the first generation of the European corn borer during 1918 was 63 days, with a maximum of 85 and a minimum of 48 days, as shown by the followuig table: — Table VII. — Life Cycle Summary of First Generation. Average. Maximum. Minimum. Incubation period in days Larval period in days ■ . Pupal period in days Adult preoviposition period in days. 7.43 44.05 8.51 3.20 9.00 57.00 10.00 9.00 5.00 35.00 7.00 1.00 Total 63.19 85.00 48.00 THE EUROPEAN CORN BORER AND ITS CONTROL. 35 Second Generation. Incubation Period. The eggs are deposited in masses on various parts of the food plant selected for oviposition. They hatch, on an average, in 6 days, with a maximum of 8 and a minimum of 4 days (see Table VIII). Duration of the incubation period depends upon temperature conditions. Table VIII. — Duration of Incubation Period — Second Generation. Observation Number. 200, 201, 202, 203, 204, 205, 206, 207, 208, 209, 210, 211, 212, 213, 214, 215, 216, 217, 218, 219, 220, 221, 222, 223, 224, 225, 226, 227, 228, Number of Eggs. 87 71 103 103 79 67 89 78 112 101 151 128 77 64 223 75 205 Deposi- Hatch- tion, 1918. ing, 1918. July July July July July July July July July July July July Aug. Aug. Aug. Aug. Aug. Aug. Aug. Aug. Aug. Aug. Aug. Aug. Aug. AuE. Aug. Aug. Aug. Aug. 2 Aug. 2 Aug. 3 Aug. 3 Aug. 5 Aug. 6 Aug. 6 Aug. 6 Aug. 6 Aug. 6 Aug. 6 Aug. 6 Aug. 6 Aug. 7 Aug. 7 Aug. 7 Aug. 8 Aug. 7 Aug. 8 Aug. 8 Aug. 9 Aug. 9 Aug. 12 Aug. 12 Aug. 14 Aug. 14 Aug. 15 Aug. 15 Aug. 16 Duration of Incuba- tion Period in Days. 36 MASS. EXPERIMENT STATION BULLETIN 189. Table VIII. — Duration of Incubation Period — Second Generation — Con. 229, 230, 231, 232, 2o3, 234. 235, 236, 237, 238, 239, 240, 241, 242, 243, 244, 245, 246, 247, 248, 249, 250, 251, 252. 253, 254, Observation Numbeb. Number ( Eggs. Deposi- Hatch- tion, 1918. ing, 1918. Aug. 11 Aug. 12 Aug. 13 Aug. 14 Aug. 15 Aug. 22 Aug. 23 Aug. 23 Aug. 23 Aug. 24 Aug. 24 Aug. 24 Aug. 25 Aug. 25 Aug. 25 Aug. 26 Aug. 26 Aug. 26 Aug. 26 Aug. 26 Aug. 27 Aug. 27 Aug. 27 Aug. 27 Aug. 27 Aug. 29 Aug. 17 Aug. 17 Aug. 19 Aug. 22 Aug. 23 Aug. 27 Aug. 29 Aug. 29 Aug. 29 Aug. 30 Aug. 31 Aug. 30 Aug. 30 Sept. 1 Sept. 1 Sept. 3 Sept. 3 Sept. 3 Sept. 3 Sept. 3 Sept. 4 Sept. 4 Sept. 4 Sept. 4 Sept. 4 Sept. 6 Duration of Incuba- tion Period in Days. Average duration of incubation period, 6.13 days. Maximum duration of incubation period, 8 days. Minimum duration of incubation period, 4 days. In the course of their development the larvse of the second generation feed in a manner similar to that described for the first generation. They pass through four or five instars before the advent of severe winter weather, which halts their activities and indefinitely prolongs the duration of the last instar or instars. According to data secured from 25 larvae, reared in life-history cages from eggs to the time when their activities ceased, the average duration of the first instar was 5.4 days; second instar, 4.2 days; third instar, 5 days; fourth instar, 9 days; and fifth instar, 10 days. The average duration of the total larval period was 26 days, with THE EUROPEAN CORN BORER AND ITS CONTROL. 37 a maximum of 32 days and a minimum of 20 days (see Table IX). The duration of each instar and the total duration of the larval period depend upon temperature conditions. Table IX. — Duration of Larval Instars and Activity to Nov. 30, 1918 — Second Generation. Barnyard Grass (Echinochloa crus-galli). Date of Hatching, 19 August 6, August 6, August 6, August 6, August 9, August 9, August 9, August 9, August 9, August 9, August 14, August 14, August 14, August 14, August 14, August 14, Duration of Larval In- stars IN Days. Date of Pupation 1918. Sept. 14 Days in Larval Period to Date. Activities to Nov. 30, 1918. Spun web Sept. 24. cf adult Oct. 14. Died Nov. 19. Died Sept. 6. Spun web Sept. 11. Died Nov. 15. Spun web Nov. 30. Died Nov. 3. Spun web Oct. 8. Spun web Oct. 10. Still feeding Nov. 30. Not emerged Nov. 30. Spun web Oct. 21. Still feeding Nov. 30. Still feeding Nov. 30. Spun web Oct. 16. Foxtail Grass (Setaria glauca). August 14, 6 4 3 Sept. 5 20 (f adult Nov. 4. August 14, 6 5 - - 24 Still feeding Nov. 30. Died Nov. 30. August 14, 7 3 - - 21 August 14, 6 3 13 - - 28 Spun web Oct. 14. August 14, 6 3 10 - 23 Died Nov. 30. August 14, 6 4 13 _ - 27 Spun web Nov. 8. August 14, 7 3 11 - - 25 Died Nov. 30. August 14, 6 3 ^ - 22 Died Nov. 18. August 14, 7 3 - - 23 Spun web Nov. 2. Average, 5.4 4.2 ■ 9.8 - 25.7 - Average duration of larval period to date, Maximvim duration of larval period to date. Minimum duration of larval period to date, 25.7 days. 32 days. 20 days. 38 MASS. EXPERIMENT STATION BULLETIN 189. Three of the larvse, confined in the life-history cages mentioned, formed pupse during September and October (see Table IX). This is believed to have been caused by the abnormal conditions which inevitably exist in confinement. No pupse of this generation were found in the field during the dissection of many hundreds of badly infested plants throughout the months of October, November and early December, 1918. The second generation larvse of the borer normally pass the winter within their host plants as full-grown, or nearly full-grown, larvae in the fifth and sixth instars. With the advent of warm weather in the spring the larvse begin feeding again, and pupate within a short period of time thereafter. Pupal Period. Pupation occurs in a similar manner to that described for the first generation. The duration of the pupal period for 35 individuals con- fined in life-history cages averaged 17 days, with a maximum of 20 and a minimum of 14 days (see Table X), depending upon weather conditions. Table X. — Duration of Pupal Period, Second Generation. Number " Date of — Num- ber of Days. Sex. Number of Observa- tion. Date of — Num- ber of Days. of Observa- tion. Pupa- tion. Emer- gence. Pupa- tion. Emer- gence. Sex. 1. . May 6 May 24 18 d 19, . . May 17 June 3 17 9 2 May 8 May 26 18 9 20, . May 18 June 3 16 d 3 May 10 May 24 14 9 21, . May 18 June 4 17 d 4 May 10 May 27 17 d 22, . . May 18 June 4 17 9 5 May 11 May 29 18 d 23, . May 18 June 4 17 9 6 May 11 May 28 17 9 24, . May 18 June 4 17 9 7 May 12 May 28 16 9 25. . . May 18 June 4 17 d 8 May 12 May 29 17 9 26, . . May 18 June 4 '' 9 9 May 12 June 1 20 d 27, . May 19 June 4 16 9 10 May 13 May 31 18 d 28, . May 20 June 6 17 9 11 May 13 May 31 18 9 29, . May 20 June 6 17 d 12 May 14 May 30 16 9 30, . May 21 June 7 17 9 13 May 14 June 2 19 d 31, . May 22 June 7 16 d 14 May 14 June 2 19 9 32, . May 25 June 12 IS c? 15 May 16 June 2 18 9 33, . . May 25 June 9 15 9 16 May 17 June 2 16 9 34, . May 31 June 17 18 9 17 May 17 June 3 17 d 35. . . June 2 June 19 17 9 18 May 17 June 3 17 9 A M 11 V [a li srage leng ximum le limuin le th of pup ngth of p ngth of p al stage, upal stag ipal stag e, e, . 17 . 20 . 14 11 days. days. days. THE EUEOPEAN CORN BORER AND ITS CONTROL. 39 Adult Period. The female moth begins the oviposition of first generation eggs within a few days after emerging from the pupa. With 15 females, confuied in individual life-history cages, the average duration of the period between emergence from the pupa and the first oviposition was 3.6 days, with a maximum of 7 daj^s and a minimum of 1 day (see Table XI). Table XI. — Oviposition by Female Moths in Rearing Cages, Second Generation. Sex. Date of — Number of Days — Number of Moths. cf 9 Emer- gence. First Ovipo- sition. Last Ovipo- sition. Before Ovipo- sition. Of Ovipo- sition. From Emer- gence to last Ovi- position. Total Num- ber of Eggs. 8 3 3 3 3 3 3 ' 3 2 3 3 3 4 2 2 2 2 2 2 1 2 2 2 May 21 May 24 May 24 May 24 May 24 May 25 May 25 May 26 May 28 May 29 June 1 June 1 May 24 May 29 May 28 May 31 May 28 May 28 May 28 May 29 May 29 June 2 June 3 June 8 June 10 June 13 June 23 June 6 June 3 June 17 June 16 June 3 June 9 June 16 June 10 June 24 3 2 7 18 16 7 7 7 21 20 6 12 14 7 16 20 20 10 13 10 23 22 8 12 18 8 22 1,261 389 190 157 348 727 713 223 107 586 210 137 40 25 15 - - - - - - ■ Average Maximum, ......... Minimum, ......... 3.66 7.00 1.00 13.66 21.00 6.00 16.4 23.0 8.0 336.53 727.00 107.00 The duration of the oviposition period of these 15 females averaged 14 days, with a maximum of 21 and a minimum of 6 days (see Table XI). The average length of life of 29 female moths, which were confined in cages with male moths, approximating field conditions as nearly as pos- sible, was 17 days, with a maximum of 29 and a minimum of 8 days. The average length of life of 40 male moths in these same cages was 13 days, with a maximum of 29 and a minimum of 6 days (see Table XII) . 40 MASS. EXPERIMENT STATION BULLETIN 189. Table XII. — Length of Life of Male and Female Moths in Captivity. Length of Life (Days). Number of Male Moths. Number of Female Moths. Length of Life (Days). Number of Male Moths. Number of Female Moths. 6 7 8 9 10 11 12 13 14 15 16 17 18 3 6 4 4 4 3 1 2 1 2 2 1 1 4 2 3 1 1 1 19 20 21 23, . . . . 24 25 26 27 28 29 Total. . . . 1 1 3 1 1 2 1 1 4 2 2 2 40 29 Average length of life of male moths, 13.1 days; of female moths, 16.86 days. Maximum length of life of male moths, 29 days; of female moths, 29 days. Minimum length of life of male moths, 6 days; of female moths, 8 days. These records were secured in the same manner as has been described in the instance of the first generation adults, and they have the same appHcation and quaUfication. Life Cycle Summary. It is rather difficult to give any accurate figures as to the duration of the life cycle of the second generation of the European corn borer, owing to the varying amount of time spent by the larva in an inactive condition during the late fall, winter and early spring. An attempt will be made, however, to approximate the correct figures by combining the results of the life-history studies as to the duration of the different periods of this generation during the early spring of 1918 and the summer and fall of 1918 up to November 30. According to these records the average duration of the second gene- ration of the European corn borer was 52.6 days, with a maximum of 67 and a minimum of 39 days, as shown by the following table: — THE EUROPEAN CORN BORER AND ITS CONTROL. 41 Table XIII. — Life Cycle Swrnnary of Second Generation. Average. Maximum. Minimum. Incubation period in days Larval period in days, i Pupal period in days, > . . . . Adult preoviposition period in days. 6.13 25.70 17.11 3.66 8.00 32.00 20.00 7.00 4.00 20.00 14.00 1.00 Total period in days, .... 52.60 67.00 39.00 I Excluding winter period of inactivity. SEASONAL HISTORY AND DEVELOPMENT. Number of Generations. There are two annual generations of the European corn borer in Massa- chusetts, a generation here being considered to begin with the egg and terminate with the moth of the same generation. Eggs of the first generation are deposited during late May or early June, and the resulting larvae pupate about the middle of July. The moths emerge during late July and early August to deposit eggs of the second generation. These eggs are deposited, therefore, during late July or early August, and the resulting larvae feed on, or within, their food plant until the advent of severe winter weather. Feeding is resumed with the coming of warm weather in the spring, and the larvae pupate about the middle of May. The second generation moths emerge during late May or early June, and deposit eggs for the first generation. A few moths of the second generation have emerged in life-history cages during September and October (see Table IX), but these indi- \'iduals died without depositing eggs. Under exceptional circumstances it is possible that moths emerging at this time may deposit eggs for a third generation, but this has not yet been observed. Seasonal History. The European corn borer passed the winter of 1917-18 as nearly full- grown larvae of the second generation within their tunnels in various food plants. The first pupa of the second generation was found in the field May 6, and the majority of the overwintering larvae pupated between May 15 and 20. The first moth of the second generation was observed in the field on May 16. Moths began to emerge from indoor cages May 18, and maxi- mum emergence occurred during the period from June 1 to 4. The last emergence of second generation moths was recorded on July 9, from labo- ratory cages (see Table XV). 42 MASS. EXPERIMENT STATION BULLETIN 189. Oviposition of second generation moths occurred within a few days after emergence, and extended over a period of about two weeks. Eggs of the first generation were first secured in life-histor}^ cages on Ma}^ 24 (see Table XI). Larvse of the first generation were first secured in life-history cages on June 2 (see Table II), and were observed for the first time in the field on June 13. The first pupa of the first generation was found in the field on July 11, and in life-history cages on July 15. Maximum pupation took place between July 19 and 23 (see Table IV). Emergence of first generation moths began about July 23 and reached its maximum between July 27 and August 4. The last emergence of first generation moths was recorded from indoor cages on August 29 (see Table XIV), and from corn in the field on September 6. On July 29 the first eggs of the second generation were secured in life- history cages (see Table V). Eggs of this generation were first observed in the field on August 13. Larvse of the second generation were first secured in life-history cages on August 2 (see Table VIII), and were observed for the first time in the field on August 13. On this date some of the larvse in the field were in the second and third instars. On September 2 many of the larvse in the field were in the fifth and sixth instars. When the last field observations were made, on November 30, most of the larvse were in the fifth and sixth instars, and in this stage of their development they probably will pass the winter of 1918-19. Seasonal Abundance. The larvse of the borer reach their greatest abundance and do the most damage to corn and other host plants during the late summer and fall. The damage to early corn by larvse of the first generation during June and July is much less than the damage to late corn by those of the second generation during August and September. The same is true for the other host plants infested by the insect. There is quite a heav>' mortality of overwintering larvse, due to natural causes, and this when added to the high percentage of overwintering larvse destroyed by control measures and cultural practices, greatly re- duces the numbers of the pest that remain to perpetuate the species in the spring. Consequently the first generation of larvae is much smaller in numbers each year than the second generation of the preceding year. HABITS OF LARViE. Hatching. About a day before hatching takes place, the black ej^e spots and red- dish mandible tips of the developing larva may be seen through the semi- transparent chorion of the egg. A few hours before hatching, the head THE EUKOPEAN CORN BORER AND ITS CONTROL. 43 and thoracic shield become black and are observed to occupy a central position in the egg. The body segmentation and the black spines on the body of the larva are also plainly discernible. At this time the develop- ing larva is curled up inside the egg with its mandibles resting upon the next to the last abdominal segment. These mandibles soon begin to move laterally, and the larva straightens itself out in such a manner that the mandibles are brought into contact with the eggshell. A slit in this is soon made, and the larva crawls forth. After hatching, the larva feeds, to some extent, upon the empty eggshell, but has not been observed to entirely devour it. Habits when attacking Corn. First Generation. The newly hatched larva crawls about over the surface of the corn blade on which it hatched, stopping here and there to eat a small area of the epidermis on either the upper or lower surface of the blade (see Plate I, Fig. 1). These small areas are usually bordered by veins on each side and are longer than wide. During its travels the larva gradually approaches the growing crown of the plant, and, upon reaching it, descends between the rolled leaf blades, or cone, composing the crown, and feeds internally upon the young and succulent epidermis of the unfolding leaf blades. If the tassel is present within the cone the first instar larvse may feed upon the epidermal cells composing the flower buds, but only rarely do larv^ae of this instar enter the buds. When ready to molt, the first instar larva spins a thin, silken molting cocoon in some protected location near its last feeding place, within which it molts to the second instar. Upon emerging from its molting cocoon this larva immediately attacks the staminate flower buds if the tassel is present within the crown. If the tassel is not present it feeds on the tightly rolled leaf blades composing the crown in a similar manner to that described for the first instar, except that larvse of the second instar are able to eat entirely through the blade, and do not confine their feeding to the epidermis. When the tassel is present within the crown the second instar larva bores a hole in the side of one of the staminate flower buds and feeds upon the internal succulent contents. Entrance to the bud maj^ be effected from the top, at the base or from the side, several buds are destroyed in turn by each larva. Dur- ing the process of feeding within the buds considerable frass is extruded, and this becomes webbed together with the silk spun by the larva in traveling from bud to bud, and forms a certain amount of protection for the larva. This webbing together of, frass for possible protection is char- acteristic of the second generation larva, as, although larvse of the first and later instars are capable of spinning a web, they do not use it for purposes of protection while feeding. When ready to molt, the second instar larva spins a molting cocoon, 44 MASS. EXPERIMENT STATION BULLETIN 189. within which it molts to the third instar. This molting cocoon may be located within a single, hoUowed-out flower bud, or may be situated be- neath the webbed-up frass from several flower buds. The third instar larva feeds at first within the staminate buds of the tassel in a similar manner to that described for the second instar lar\'a, but, when a little older, it may enter the terminal spike of the tassel, 1 or 2 inches above the last branch, and tunnel within this spike, and a small mass of frass collects at the point of entrance and renders the injury conspicuous. Instead of entering the tassel, many third instar larvae tunnel within the midrib of the leaf blade. These tunnels are never more than 1 or 2 inches in length, and closely resemble the injury to the tassel spike. Whether the third instar larva tunnels in the terminal spike, in the midrib of the leaf, or continues feeding in the ends, appears to be arbitrary with the individual. The third instar larva may molt to the fourth instar, either withm its tunnel or in some protected place outside. If molting occurs within its tunnel, a molting cocoon is not formed, but a silken partition is spun across the entrance hole. If molting occurs in some protected place out- side the tunnel, a t3Tpical molting cocoon is formed, and the larva molts to the fourth instar in a similar manner to that described for the preceding ones. After molting to the fourth instar the larva usually enters the main stalk of the tassel 1 or 2 inches from its base. Sometimes it enters the terminal internode at the point where the first leaf blade joins its sheath. Later the terminal internode of the corn plant grows so that this en- trance point, instead of being present at the junction of the leaf blade and the leaf sheath, is found 5 or 6 inches above that pomt. After cut- ting an entrance hole in the side of the stalk the larva tunnels out a small, spherical cell, which occupies nearly all the interior of the staUc at this point. From this it usually tunnels upward for 2 or 3 inches above the entrance hole, and then returns and tunnels downward. During this feeding a large amount of frass is pushed out of the entrance hole and is held there by means of small silken strands spun by the larva. This large mass of yellow-white frass is very conspicuous, and serves to identify infested tassels, even before they break over. Eventually the tassel becomes broken over at the point where the fourth instar larva entered the terminal internode. The fourth instar larva molts to the fifth instar within its tunnel, and only spins a silken partition across its entrance, thus using its tunnel for a molting cocoon. The fifth instar larva may complete its larval development within the terminal internode. The number of larval instars varies with different individuals, five being sufficient to complete the larval growth in some individuals, while six, or even seven or eight, instars are passed through in other cases. In the majority of instances, especially when an abundant supply of food is available, the fifth instar is the last and longest of the THE EUROPEAN CORN BORER AND ITS CONTROL. 45 larval instars. During this, or the succeeding instars, the larva; sometimes wander about and do their greatest amount of damage to the plant. Some individuals leave the terminal internode and tunnel through the lower parts of the stalk; some tunnel from the terminal internode down through the intervening nodes into the lower part of the stalk; while others enter the stalk at various places along its length and tunnel upward or down- ward according to their individual preference. The junction of the leaf sheath and node is a favorite point of entrance, although this is by no means universal. Frequently the larv^a enters a stalk and tunnels out a cavity, only to abandon it and enter the plant at a different point. The stalk may be tunneled by the larva? to its base, or even into the taproot, so that corn stubble is often infested and must be considered a source of danger in clean-up operations. During their wanderings the larger larvae very often descend the plant until they reach the side branch, or pedicel, on which the ear is borne. Here they may enter the main stalk or may enter the pedicel and tunnel into the ear. Some enter the ear directly by boring through the husk, later feeding on the immature kernels or tunneling through the cob. In other instances the ear is entered at the tip end, and the larvae tunnel through the kernels and the cob. Apparently the ear is very much favored as a food by the larvae. In instances when the fifth instar larvae molt into the sixth, seventh or eighth instars (see Table III), the molting process takes place in the same manner and location as has been described for the fourth to fifth instar molt. The habits of the larvae vary greatly with different individuals and under different environments. For this reason the preceding remarks are intended to give only an idea of their usual activities in this stage, and their habits when attacking corn. In general, it may be stated that they may attack all parts of the corn plant except the fibrous roots, and that this damage may occur in an indefinite number of ways by larvae of the different instars. Second Generation. When attacking corn the habits of the second generation larvae are essentially the same as have been described for those of the first generation. The only exception is that a large proportion of the larvae hatch from eggs which have been deposited directly upon the silk or husk of the immature ears. They feed first upon the tender leaves of the husk, and upon the silk, and then tunnel through all parts of the ear. This type of injury is of great economic importance, especially in sweet corn or that grown for seed. The amount of damage to corn by larvae of the second generation is, therefore, infinitelj^ greater than that caused by those of the first generation, due to the greater numbers of the second generation and their habit of attacking the ears directlj'. The nearly full-grown lai-vae winter over within their tunnels in the 46 MASS. EXPERIMENT STATION BULLETIN 189. stalk, in the ear or in the taproot. They do not generally spin any pro- tective cocoon, but remain quiescent during the cold weather. Feeding is resumed during the warm portions of pleasant days in early spring, but the larvse return temporarily to their quiescent state during cold nights or inclement and cold spring weather. The hardened condition of stalks and ears during the spring does not appear to present any diffi- culties to them, as they tunnel through all parts of the plant with the same apparent ease as when the plants were comparatively soft and green the preceding season. Cobs of seed corn, which had been stored on the cob all winter and were very hard and dry, contained living larvse of the borer in April, 1918. That they had been feeding during the warm periods of the early spring was evidenced by the mass of frass extruding from their tunnels. This occurrence will serve to illustrate the danger of disseminating the pest by the transportation of corn on the cob. Habits when attacking Dock. The first instar larva of the European corn borer feeds, to some extent, on the tender seed heads of the dock plant, and also upon the epidennis of the leaves, but soon works its way down between the main stalk and a leaf sheath. Here the first molt occurs, and the second instar larva feeds on the leaf sheath, the basal part of the leaf petiole, and on the small secondary' stalks which arise at the junction of the leaf and the stalk. When the leaf petiole is tender enough the second instar larva usually tunnels into it and molts into the third instar, either in this location or at the base of the leaf petiole when it has been unable to effect an en- trance. The third instar larva usually tunnels in the leaf petiole and molts to the fourth instar within its tunnel. Occasionally the third instar larva does not feed within the leaf petiole, but enters the main stalk at the junction of the petiole and stalk. Normally, the larva does not enter this main stalk until the fourth instar is "reached. After entering the stalk it usually tunnels downward through the nodes and internodes, practically consuming the interior of the stalk. The remaining instars are passed, and the larva becomes full grown and pupates, within this tunnel. A large quantity of frass is extruded by the larva through the entrance hole, and becomes webbed into the axial flowers situated between the main stalk and the petiole. This accumulation of frass makes in- fested dock plants very conspicuous, even before the upper portion of the plant breaks over at the entrance hole of the larva. By the 1st of August nearly all of the dock plants are dead, so the activities of the European corn borer in this plant are confined to the first generation. Habits when attacking Lady's-thumb. In this common host plant of the European corn borer the first instar larva tunnels directly into the main stalk at a point about 1 or 2 inches below the terminal leaves. Soon after the plant is attacked it may easily 'i THE EUROPEAN CORN BORER AND ITS CONTROL. 47 be distinguished from those not infested, as the terminal stalk withers and droops above the point where the small, first instar larva entered. After entering the stalk it tunnels downward, molting within the tunnel , as it develops in size. This tmmel is not continuous, owing to the fact that the larva emerges from the stalk at will, and enters again at a point nearer the base. It usually tunnels exclusivelj' in the internodes of this plant, very rarely passing through a node. In tliis particular the habits of the larva, when attacking lady's-thumb, are distinctive because the node is commonly tunneled in other plants. Habits whex attacking Barnyard Grass. The habits of the European corn borer larvje, when attacking barnyard grass, are verj^ similar to those that have been detailed in the injury to dock, except that the larger ones, instead of continuing to feed on their original host, often leave the stalks of barnyard grass, where they have partially completed their development, and enter others. Barnyard grass commonly serves as a host for the second generation larv'jB until the middle of October. At this time it becomes dry and hard, and many of the larvge desert it for more attractive food plants groM^ng in the vicinity, though a large percentage of the original number present continue feeding in the lower parts of the plant, and may be found inside the base of the stalk, below the level of the ground, as late as November 30. It is believed that the nearly full-grown larv^se pass the winter in this location, although complete data on this point will be lacking until observations are made in the spring of 1919. Superficial observers have frequently stated that barnyard grass is entirely deserted by larvae of the European corn borer during the late fall season, but close examination will reveal many at the bases of the stalk. In this position they are yerry difficult to destroy by ordinary clean-up methods. Molting. Wlien feeding on, or near, the surface of its food plant, especially during the early instars, the larva spins a molting cocoon within which it molts. This is formed of thin, silken strands, and is located in any protected place. WTien tunneling inside its food plant the larva does not form a molting cocoon, but merely closes the entrance to the tunnel with a thin, silken partition. It then molts inside this tunnel near its last feeding place. The process of molting varies in detail with the different instars, but in general is as follows. After all preparations to secure protection have been made, the larva enters a semi-quiescent state during which the head capsule becomes pushed forward uiitil a distinct non-contractile, white band appears between the head and the shield. After remaining in this condition approximately twelve to twenty-four hours the old larval skin splits longitudinally just back of the head capsule, and, as a result of 48 MASS. EXPERIMENT STATION BULLETIN 189. squirming movements from within, slips down and off the molting larva. When nearly free of the old larval skin the larva easily brushes off the old larval head mask, or remains of the head capsule. The newly molted larva is colorless, with an opaque, white head cap- sule and thoracic shield. In the course of two or three hours its body assumes the characteristic markings for the instar, while the head and thoracic shield darken and become fully pigmented. After completing its emergence and coloring the larva remains quiet until the body chitin becomes hard, and then resumes its activities. Length of Larval Life without Food. Newly hatched larvae of the European corn borer lived a maximum of two days in life-history cages without food or water. Nearly full-grown larvse, isolated in glass vial cages, without food or water, lived a maximum of thirty days during the active season. This latter characteristic is important with relation to the possible transportation of infested material to localities not infested by the insect. The long period of life without food would allow larvse to survive under very adverse conditions, and to start new colonies of the insect when opportunity afforded. Unusual Habits. Large larvse of the European corn borer will eat their way through an ordinary cork stopper and escape from confinement. They are unable, however, to make any impression upon a cotton plug, and are easily con- fined in glass vials when these are plugged with cotton. Larvse also eat through paper and pasteboard. On one occasion a full- grown larva, which had escaped from an indoor cage, tunneled through heavy pasteboard surrounding a bottle, and pupated between the bottle and its covering. Full-grown larvae have been observed crawling along the ground at some distance from any possible food plant. In cases of necessity these larvae could probably travel a considerable distance. Large larvse have been found underneath clods of earth and under- neath rubbish in badly infested fields, due possibly to some agency which forced them to leave their natural protection within the food plant. Infested cornstalks were buried in the soil to a depth of 6 inches during the spring of 1917. Within a few days the lai-vae deserted the buried corn- stalks and made their way to the surface. Although the larger larv^se normally feed within the plant, occasionally individuals are found feeding on its exterior. This is especially true of the full-grown larvse just before pupation. At this time they are fre- quently found feeding on the silk and on the outer husk of the ear. THE EUROPEAN CORN BORER AND ITS CONTROL. 49 PUPATION. Location of Pupa. Normally the pupa of the European corn borer is found inside the tunnel made by the larva, and not far from its last feeding place. A small per cent of the full-grown larvse, however, leave the interior of the plant when attacking corn and pupate in some protected place near by, such as the silk of the ear; between the husks of the ear; in a fold of the leaf blades; between two overlapping leaf blades; in the frass clinging to the tassel; in the frass at junction of leaf blade and leaf sheath; be- tween the leaf sheath and stalk; and on the surface of the ear in the hol- low made by the feeding larvse. Though in corn most of the larvae pupate within their tunnels in the stalk or in the pedicel of the ear, many pupae are found inside the cob and in the upper part of the taproot. Cocoon Formation. Most of the folio-wing remarks concerning cocoon formation apply only when the larva forms its cocoon and pupates within the larval tunnel. When the larva reaches full growth and is ready for pupation it cuts a circular exit hole to the surface of the plant. It then spins a silken partition across this exit hole from within, and this partition serves to block the outside entrance to its pupal chamber. It then retreats about 2 inches into its tunnel, and forms the base of its pupal chamber by packing the tunnel with a layer of frass about an inch thick. A silken partition is then spun on top of this protecting layer, and frequently another transverse partition of silk is spun about a quarter of an inch above this lower one. After thus closing both ends of the tunnel the larva proceeds to coat the walls of its pupal chamber with a very thin layer of silk, and then spins a single internal partition, also of silk, across the upper part of the pupal chamber and parallel to the exit hole. The larva then constructs two slanting partitions in the lower part of its pupal chamber, which intersect each other and form a partition resembling the letter "Y." After completing the bottom partitions of the pupal chamber the larva turns around and begins forming the upper ones. These are quite similar to the lower, but are usually more complicated and more substantial. They consist of a series of four or five intersecting partitions of silk which meet in the center to form a letter "Y", and make an angular roof over the head of the larva. The cocoon is then complete. About three or four days are usually required by the larva for its formation. After completing the upper partitions of its pupal chamber the larva attaches its anal legs firmly to the angle of the "Y" in the bottom par- tition, and then passes into a semi-quiescent state. 50 MASS. EXPERIMENT STATION BULLETIN 189. Changes undergone by the Lakva previous to Pupation. In the semi-quiescent state the larva is very sluggish, but is still capable of locomotion. Soon after entering this stage the head starts to bend downward, and the mouth parts become ventral instead of anterior. The second thoracic segment becomes swollen, and the third thoracic and first abdominal segments become compressed as a result of pressure exerted at the anterior and posterior ends of the larva. The second and third abdominal segments remain about normal, while the fourth to seventh become enlarged and swollen, and show distinctly the outlines of the pupal abdomen. At the termination of the semi-quiescent stage, which lasts for about twenty-four hom-s, the larval head is fully inflexed and the use of both thoracic and abdominal legs is lost. The larva then enters the true quiescent state. In this stage the larva is not capable of locomotion, but has the char- acteristic movements of a pupa. Soon after entering this stage the con- tents of the terminal segments of the lan^a shrink away from the larval body wall to form the terminal segments of the pupa. At this time the anal legs consist of only the external chitmous covering, with their crotch- ets firmly attached to the bottom silken partition. Wlien disturbed the larva twitches and turns with a movement resembling that of the pupa, while the empty anal legs remain attached to the silk and are often twisted around each other during the twisting movements of the larva. At this time the abdominal legs are flush with the venter, and the thoracic legs are folded close to the body. The quiescent stage requires from twelve to twenty-four hours for its completion, and then the larva begins the process of pupation. Process of Pupation. After a few straining movements forward, and as a result of pressure exerted from within, the larval skin suddenly splits along the dorsal line of the head and thoracic segments, and also down each side of the frontal head plate. After a few wriggling movements the larval skin slips down to the terminal segment, which then is liberated. As soon as it is freed from the larval skin the newly formed pupa turns around two or three times, thus firmly attaching its cremaster to the angle of the lower silken partition in the pupal chamber, at the point formerly occupied by the anal feet of the larva. A timed individual required two and one-half minutes to shed its larval skin, except the terminal segment, and the total time required to completely shed this skin and attach the cremaster was eight minutes. Changes undergone by the Pupa. The newly formed pupa is white in color, with a longitudinal pink line down the dorsum. Transverse pink lines extend across the center of the dorsum of each abdominal segment, but fade away laterally. The wing pads are yellow with a tinge of pink. The venter of the abdomen is THE EUROPEAN CORN BORER AND ITS CONTROL. 51 creamy white throughout. The cremaster and its spines, and also the chitinous braces arising from the last segment, are dark red. About one hour after pupation the transverse pink hues gradually widen and become darker in color, until the dorsum, except at the union of segments, is yellowish red. At this time the venter is almost pure white, but soon begins to turn pinkish yellow in the posterior half of each ab- dominal segment. This color then extends to include the entire venter of each abdominal segment. The terminal abdominal segment assumes its permanent color at this time. As permanent coloration proceeds, the dorsum of the thorax and abdomen, together with the wing pads, turn a darker red, and soon assume their permanent color. In approximately five or six hours after its formation the pupa is fully colored, and retains this coloration until about three or four days before the emergence of the moth. At this time it becomes very much darker and shows the adult markings. HABITS OF ADULTS. Emekgence of the Moths. After loosening its appendages the emerging moth pushes off the head cap of the pupal skin by exerting pressure from within, and frees itself until the head and eyes are visible. Here the moth rests for a few seconds before struggling completely out of the pupal skin. About two or three minutes are required for the moth to entirely free itself. At this time the wings of the moth are only partly developed, and are practically the size of the pupal wing pads. In this condition the moth escapes from the cocoon and crawls to the surface of the plant, providing pupation occurred within interior tunnels. After reaching the surface the moth obtains a foothold and assumes a perpendicular position. It is never found in a horizontal position at this time. The wings then lengthen and widen gradually, meanwhile being brought vertically over the body and held in this position until fully expanded. After reaching their full development and expansion the wings are lowered to their normal position of rest, and within a few hours the moth is ready to assume its adult activities. Maximmn adult emergence generally occurs very early in the morning, and the moths seldom emerge at any other time, unless the early morning hours are rather cold. In this event the moths are delayed in emerging until the early forenoon. A few, however, have been obsei*ved to emerge late in the afternoon. Copulation. Copulation occurs within twenty-four hours after the sexes emerge from the pupa, and at frequent intervals throughout the life of the adult, — thirteen to eighteen days' average (see Tables VI and XII). Late a'fter- noon or evening, when the adults are most active, is the usual time for copulation. The act is accomplished in a similar manner to that of other lepidopterous adults. Polygamy experiments were tried during the summer of 1918, but no 52 MASS. EXPERIMENT STATION BULLETIN 189. definite data were secured as to the number of females fertilized by each male. Bearing in mind the long period of adult Ufe, however, it is prob- able that each male will fertihze several females. Proportion of Sexes. First Generation. A total of 317 first generation pupse were collected from the field in July, 1918, and confined in individual cages. From these a total of 317 first generation adults emerged, of which 136, or 42.9 per cent, were males, and 181, or 57.1 per cent, were females (see Table XIV). Table XIV. — Proportion of Sexes and Time of Emergence of Moths, First Generation. Date of Emer- gence, 1918. July 23, July 24, July 25, July 26, July 27, July 28, July 29, July 30, July 31, August 1, August 2, August 3, August 4, August 5, August 6, August 7, August 8, August 9, August 10, August 11, Number of Males. Number of Females, Total Emer- gence. Date of Emer- gence, 1918. August 12, August 13, August 14, August 15, August 16, August 17, August 18, August 19, August 20, August 21, August 22, August 23, August 24, August 25, August 26, August 27, August 28, August 29, Total, Number of Males, Number of Total Emer- gence. Total emergence, 317 adults. Total males, 136, or 42.9 per cent. Total females, 181, or 57.1 per cent. A total of 49 first generation pupa were reared from full-grown, first generation larva) collected in the field during July, 1918, in order to secure data as to duration of the pupal period. From this material a total of 49 first generation adults emerged, of which 19, or 38.8 per cent, were males, and 30, or 61.2 per cent, were females (see Table IV). THE EUROPEAN CORN BORER AND ITS CONTROL. 53 On the night of August &-7, 191S, 17 first generation moths were cap- tured at a trap hght. Of these, 7, or 41.2 per cent, were males, and 10, or 58.8 per cent, were females. Thus out of a total of 383 first generation adults, 162, or 42.3 per cent, were males, and 221, or 57.7 per cent, were females. Second Generation. In April, 1918, two barrels of badly infested cornstalks were collected and placed in the laboratory in order to secure data as to adult emergence, proportion of sexes, etc. From these two cages 307 second generation adults emerged, of which 160, or 52.1 per cent, were males, and 147, or 47.9 per cent, were females (see Table XV). Table XV,. — Proportion of Sexes and Time of Emergence of Moths, Second Generation. Date of Emer- gence, 1918. Number of Males. Number of Females. Total Emer- gence. Date of Emer- gence, 1918. Number of Males. Number of Females. Total Emer- gence. May 18, . . 1 - 1 June 11, - - - May 19, 6 4 10 June 12, 1 5 May 20, 5 2 7 June 13, 2 2 May 21, 2 5 7 June 14, 1 2 May 22, 3 6 9 June 15, 2 3 May 23, 1 1 2 June 16, 1 3 May 24, 44 7 21 June 17, 3 - May 25, 7 2 9 June 18, 1 - May 26, 9 3 12 June 19, 1 2 May 27, 5 1 6 June 22, 1 1 May 28. 2 5 7 June 23, 1 - May 29, 8 7 15 June 24, 1 2 May 30, 2 - 2 June 25, - 2 May 31, 2 6 8 June 26, - 3 June 1, 12 8 20 June 27, - 2 June 2, 12 19 31 June 28, 1 - June 3, 17 16 33 June 29, 2 - June 4, 9 7 16 June 30, 1 1 June 5, 3 1 4 July 1, 1 - June 6, 6 3 9 July 2, - 3 June 7, 2 4 6 July 4, 1 - June 8, 4 2 6 July 5, - 1 June 9, 3 4 1 4 4 8 July 9, Total, - 1 June 10, 160 147 307 Total emergence, 307 adults. Total males, 160, or 52.1 per cent. Total females, 147, or 47.9 per cent. 54 MASS. EXPEEIMENT STATION BULLETIN 189. A total of 35 second generation pupse were reared from full-grown second generation larvae collected in the field during May, 1918, in order to secure data as to duration of the pupal period. From this material 35 second generation adults emerged, of which 13, or 37.3 per cent, were males, and 22, or 62.7 per cent, were females (see Table X). Thus out of 342 second generation adults, 173, or 50.5 per cent, were males, and 169, or 49.5 per cent, were females. It will be noted that, in the instance of the 725 adults of both genera- tions represented by these figures, the sexes were present in nearly equal proportions, there being 335 males and 390 females. Flight. Character of Flight. Both sexes of the European corn borer adults are capable of flight. They habitually fly very close to the ground, a tendency that is caused, perhaps, by the fact that the plants upon which the females deposit their eggs do not generally reach a height of more than 6 or 8 feet. When disturbed in their hiding places during the day the adults fly close to the ground, in a curious zigzag manner, for a distance of 10 or 20 feet, and then seek cover again under some object. It is rather difficult to observe the flight of the adults during the time of their maximum activity in the early evening. Such observations as were made, however, indicated that adults normally fly very low, even when seeking food plants upon which to deposit their eggs. The males apparently are more active than the females, and fly for greater distances and at higher altitudes. The character of their flight at this time is similar to that which has been described in the instance of moths dis- turbed from their hiding places during the day. Distances of Flight. Under most conditions the moths cover a very short distance in each flight, the maximum observed in any single flight being about 50 yards. The females make a series of short flights in search of food plants on which to deposit their eggs, so that the total distance covered by a female in a series of ffights may be considerable. The males make a similar series of flights in their search for the females. Effect of Wind on Flight of Moths. It is not beheved that the moths are carried any considerable distances by the wind, although the general direction in which the insect has spread, since its introduction into Massachusetts, has been with the prevailing winds. Meteorological records show that theee winds during May, June, July and August are from the south and the southwest. The fact that the insect has spread more rapidly toward the north and the north- THE EUROPEAN CORN BORER AND ITS CONTROL. 55 east than in any other direction would tend to indicate that the flight of the moths is influenced by the wind to some extent. The habit of the moth of flying close to the ground would seem to reduce the possibility of wind spread to a minimum, but future observations may show other influencing factors. Time of Maximum Activity. During the day the moths remain inactiVe. They may commonly be found hiding on the underside of the foliage of their food plant, or in strips of grassland and low weeds growing along the field borders and ditches of cultivated areas. They also remain inactive during cool periods, and also during high winds. They become active in the late afternoon, and reach their greatest period of activity about dusk. Attraction of Moths to Trap Lights. On the night of August 6-7, 1918, a trap light was placed midway, and 50 feet distant, from two areas of sweet corn which contained hundreds of first generation adults. These had recently emerged from early corn and were at the period of their greatest activity. The trap light was started at 8 p.m. At this time the moths were actively flying around among the corn plants. The first moth was caught at 8.45 p.m. Observa- tions were continued until 11.30 p.m., and the trap light was left burning until 8 A.M. the next morning. The total catch from this trap light experiment was 17 moths, of which 7 were males and 10 were females. Subsequent dissection showed that all of the females were gravid. The trap light used in the experiment was yellow in color. Examina- tion of blue arc lights along the streets in the vicinity of badly invested areas failed to show that the moths were attracted to the blue lights to any greater extent than has been detailed for the yellow light. OVIPOSITION. The females of the European corn borer begin ovipositing about three days after emerging from the pupa (see Tables V and XI). Oviposition generally occurs during the late afternoon or early evening. Details of Oviposition. The female assumes a position on the under surface of a leaf blade, and bends the end of the abdomen down, meanwhile extruding the ovipositor until its tip comes in contact with the leaf blade. The tip of the ovipositor is fleshy and circular. Around its periphery extends a circle of amber- colored hairs. After selecting the spot on which the egg is to be deposited the female stands still and vibrates the ovipositor until the spherical- shaped egg appears at its tip. The egg is then quickly pushed against the leaf and tamped down into place by the ovipositor, which at the same time flattens it. This act changes the egg from its original spherical 56 MASS. EXPERIMENT STATION BULLETIN 189. shape into a more flattened one. From 5 to 50 eggs are thus deposited in a flat egg-mass, each egg overlapping the adjoining one in the manner of shingles. The female rarely changes her position during the oviposition of an egg-mass, as the flexibility of the abdomen allows quite a radius of action. DlSTKIBUTION OF EgG MaSSES. During its period of fertility the female deposits a varying number of egg-masses, each mass being composed of from 5 to about 50 eggs. These are generally placed on the under sides of the leaves of several different plants, but in some instances all of the eggs may be deposited on the same plant. When selecting plants for egg deposition the female appar- ently prefers certain plants to the exclusion of others belonging to the same species. In life-history cages the daily rate of oviposition varied with different females and according to the temperature conditions. In some instances a single female deposited several egg-masses in twenty-four hours, while in other instances a period of several days elapsed between the deposition of successive egg-masses. Total Numbeb of Eggs deposited by Each Female. First Generation. In life-history cages 13 female moths of the first generation deposited an average of 545 eggs each. The maximum number of eggs deposited by a single female was 903, and the minimum, 132 (see Table V). Second Generation. In life-history cages 15 female moths of the second generation deposited an average of 337 eggs each. The maximum number of eggs deposited by a single female was 727, and the minimum, 107 (see Table XI). Duration of Fertility. The duration of fertility is here considered to be the period between the first and last deposition of eggs. First Generation. The duration of fertilitj^ of 13 female moths of the first generation that were confined in life-history cages during July and August, 1918, aver- aged fifteen days, with a maximum of twenty-four days and a minimum of six days (see Table V). Second Generation. The duration of fertility of 15 female moths of the second generation that were confined in life-history cages during May and June, 1918, averaged 13.66 days, with a maximum of twenty-one days and a minimum of six days (see Table XI). THE EUROPEAN CORN BORER AND ITS CONTROL. 57 The long period of fertility of the female moths in both generations of the European corn borer is important because it results in larvae of several' different instars being present in the same field, and often on the same plant at the same time. This may be an important consideration in any control measures that have for their object the destruction of the young larvae before they enter the plant. The long period of fertility also increases the chances that gravid females may start new infestations of the insect by being carried outside of the infested area. PARASITES. European Records of Parasites. European literature contains very few records of parasites bred from the European corn borer in any of its stages. Most of the literature on this species emphasizes the absence of any parasites. Robin and Laboulbene (11) mention the fact that one of their col- leagues, M. Jules Fallon, reared many specimens of P. nubilalis (Botys) from larvce to adults during several consecutive years prior to 1879, but secured no parasites, either hjanenopterous or dipterous, from any stage of the insect. Jablonowski (16) records breeding a parasite fly, Ceromasia interrupta Rdi., from the larva of P. mibilalis. The author states that "the insect is not much infested by parasites in Hungary." Kollar (6) mentions that some Ichneumonidae have been bred from the insect. Records of Parasites in Massachusetts. No parasites were bred from the egg of the European corn borer during the investigations in Massachusetts. Parasites of the Larva. In Massachusetts four different species of dipterous parasites belonging to the Tachinidse have been bred from larvse of the borer. These Tachinids were determined by Dr. J. M. Aldrich of the United States National Museum as Masicera myoidea Desv., Exorista pyste Walk., Exorista nigripalpis Tns., and Phorocera erecta Coq. No other parasites were bred from P. nubilalis larvae. In each of the species noted above the parasite larva emerged from its host larva just previous to normal pupation of the latter. All of these records were secured from host larvse collected in the field and kept under observation in cages. During the progress of dissecting infested plants in the field, occasional parasitic dipterous larvae and puparia were found in the tunnels of P. nubilalis. In these instances it was not possible to state definitely whether the parasite had emerged from P. nubilalis, or from some other larva which had wandered into the P. nubilalis tunnels. 58 MASS. EXPERIMENT STATION BULLETIN 189. For this reason these records are not included among the list of P. nubilalis parasites. Only a small per cent of P. nubilalis larvse were parasitized. During the entire season of 1918 a total of about twenty individual dipterous (Tachinid) parasites were bred, although several hundred larvse were under observation in life-history cages and in the process of securmg other biological data. The highest percentage of parasitism recorded was from a collection of 50 full-grown P. nubilalis larvse dissected from the stalks in a badly infested field in Revere, Mass., on Aug. 23, 1918. Two para- sitic larvse emerged from the total of 50 P. nubilalis larvse, a percentage of parasitism of 4. A fact worthy of recording here is that during July, 1918, the larvae of Papaipema nitela Gn. were very highly parasitized by Masicera myoidea Desv. The larvse of P. nitela were tunneling through the same plant, or plants in the same hill, as larvae of P. nubilalis, and the latter were only parasitized to a very small extent by the Tachinid. The statement has been made by foreign observers that one reason for the dearth of lar\^al parasitism in P. nubilalis is their protected mode of Uving within the plant, but in the instance recorded it would seem as though P. nubilalis should have been parasitized to as great an extent as P. nitela, which at this time was following the same mode of attacking its host plant. Parasites of the Pupa. In Massachusetts two different species of hymenopterous parasites have been bred from pupae of the European corn borer. These were determined by Mr. A. B. Gahan of the United States National Museum as (Pinipla) Epiurus pterophori Ashm., and {Ichneumon) Amblyteles brevicinctor Say. The hymenopterous larva of E. pterophori was found feeding on the internal juices of a P. nubilalis pupa which had been broken open. The full-grown parasite larva spun a brown silken cocoon and pupated within the remains of its host. Only two of these parasites were bred. The adult parasite A. brevicinctor emerged "from the fully formed pupa of P. nubilalis. Two of these parasites were bred durmg August, 1918. No other definite records of pupal parasitism were secured, although several hundred pupse were under observation in life-history cages and during the progress of securing other biological data. A single adult specimen of Agrypon sp. (det. Gahan) was found in a pasteboard box cage which contamed about a dozen discarded P. nubilalis pupae. The head cap of oiie of these had been forced off, so it is probable that the parasite emerged from this pupa. This cannot be considered a definite record of P. nubilalis parasitism, however, A single specimen of Macrocentrus sp. (det. Gahan) was bred from a hymenopterous cocoon found in the tunnels of P. nubilalis, near the remains of a P. nubilalis pupa; but this also cannot be considered a definite record of P. nubilalis parasitism. THE EUROPEAN CORN BORER AND ITS CONTROL. 59 Summarizing the records of parasites bred from the European corn borer it will be noted that there are four species of Diptera and two species of Hymenoptera represented. The number of different species attacking P. nuhilalis suggests the possibility that parasites may in the future have some influence in controlling the pest, but at the present time they cannot be rehed upon to accompUsh much. PREDATORS. Birds. Several species of birds, including woodpeckers, blackbirds and crows, have been observed to feed upon the larvae and pupse of the European corn borer. Blackbirds have been observed to pick them out of infested corn tassel-stalks, frequently breaking over the tassel-staUc to reach the insect within. On one occasion a flock of crows settled down in a:n infested patch of field corn and devoured nearly all of the P. nuhilalis larvae which were feeding on the ears. Incidentally they also devoured some of the corn. Insects. Larvae of the corn ear worm Chloridea obsoleta Fab. frequently kill and feed upon P. nuhilalis larv^ae which are feeding on the same ear of corn. A small beetle, 7ps fasciatus, is frequently found in P. nuhilalis tunnels but has not been observed to prey upon the larva of the pest. CONTROL. Destroying Plants containing Overwintering Larv^. Bearing in mind the life history and habits of the European corn borer, it is evident that any measures for controlling the insect must be pre- ventive rather than remedial. The most obvious method of preventing damage by the insect, or at least greatly reducing its numbers, is by the destruction of plants containing the overwintering larvae. This may be accomphshed any time during the period from the middle of October until the middle of the following May. Burning Infested Plants. Burning infested plants is undoubtedly the most practical and effective measure that can be adopted for the destruction of the overwintering larvae. At first thought this seems to be an easy method of handUng the problem, but when the great variety of food plants is considered, and also the extent of the infested area (320 square miles), it becomes one of great proportions. In order to destroy the larvae in any given area by this method, all parts of the different food plants within that area must be burned, including the roots or stubble of the plants. 60 MASS. EXPERIMENT STATION BULLETIN 189. In comparatively large areas occupied by weeds this result may be accomplished bj^ a running fire which, under favorable conditions, will effectively burn all plants to the surface of the ground, and kill any larvse that may be present in the roots. In the infinite number of small areas present throughout the infested region, and especially in the vicinity of buildings, it is not generally pos- sible to start or maintain a running fire, and, under these circumstances, it becomes necessary to remove the infested plants and burn them in piles or in some receptacle provided for the purpose. This method entails considerable labor and expense, and when appHed to the 320 square miles infested, presents a large problem. Cornstalks and other infested plants in cultivated areas may generally be cut very close to the ground and burned in piles. The stubble may then be plowed out, raked up and burned, if no better means for its de- struction are available. In small areas of corn it is sometimes more practicable to pull up and biu-n the entire plant than to remove and destroy the stubble. During the early fall of 1918 considerable difficulty was experienced in attempting to burn cornstalks and other infested plants, owing to the large amount of water still present in the stalks, some of these plants being still green in appearance and resisting all efforts to burn them, even when kerosene oil was applied. It is possible, therefore, that in some instances infested plants must be burned during the early spring or during mild periods of the winter. It is not necessary to entirely consume the infested plants in order to kill the larvse contained therein, but these plants should at least be given a thorough scorching or be exposed to considerable heat. While experimenting with methods for burning infested plants several different types of torches were used. None of these, however, gave any satisfaction during the fall of 1918. This result may have been due to the green condition of many infested plants on which the torches were used, and it is possible that this method may give better results during the %vinter and spring, when the infested plants are dead and dry. It is hoped that ultimately some tjT)e of a portable burning apparatus will be developed for use in burning large quantities of infested plants easily and at a low cost. Any method adopted for the burning of infested plants throughout the entire infested area will result in a considerable outlay of money. Never- theless, it is beheved that burning is the best method to use in clean-up operations. Figures, compiled from data concerning the towns in the area infested by the pest up to November, 1918, show that about 50,000 acres must be treated. Burying Infested Plants. Burying infested plants may destroy the contained larvse under some conditions. This method of eliminating infested material is especially desirable from an agricultural viewpoint, because the decaying plants THE EUROPEAN CORN BORER AND ITS CONTROL. 61 provide humus so necessary to the maintenance of fertility and texture in the soil. If this method is adopted, however, the infested plants must be buried at least a foot in the soil, and the surface packed, if possible. Experiments to date have indicated that this method of destroying in- fested plants cannot be relied upon unless undertaken with great care. In ordinary plowing operations infested plants are only partially turned under, and much of the plant remains are left on the surface of the ground. This is not an effective method for destroying mfested plants. During the month of May, 1918, infested cornstalks were buried in the soil to a depth of 6 inches, and in a manner resembhng the work of an ordinary plow. The second generation larvse contained in these buried stalks promptly made their way to the surface of the soil and entered plant remnants in the vicinity. Different results might possibly have been secured if the infested stalks had been buried in the fall and left in the soil through the winter, and experiments were started durmg the fall of 1918 to determine this point. Infested cornstalks, buried to a depth of 12 inches in October, 1918, were dug up five weeks later and found to contain h^dng larva?. These were still actively feeding, although the interior of each cornstalk was soft and had begun to decay. If a method could be developed for plowing under infested plants in order to destroy the larvse contained therein it would be very desirable but in the present state of our knowledge concerning the matter this practice cannot be recommended. Feeding of Infested Plants. The feeding of infested plants to hve stock is, from the economic view- point, the best possible means for destrojang the larvse of the European corn borer. The value of the stalks for fodder is not materially affected by the presence of the insects, and, if properly carried out, this method must result in the destruction of all insects within the infested plants. This is particularly true in the instance of infested corn fodder. Shredding the corn fodder, or cutting it into small sections before feeding, greatly reduces the chance that any of the contained larvse will survive. Live stock rehsh corn fodder when fed in this form, and will eat all parts of the plant. Ensilage, by ordinary methods, effectively destroys all larvse within the fodder, as the insects cannot survive the conditions existing in the silo. Composting Infested Plants. Whenever infested plants or parts of plants are placed in a compost or manure pile and covered deeply, the resulting decay and fermentation quickly result in the death of the larvse contained within the plants. It is a common practice on some farms to use corn fodder for bedding. This corn fodder ultimately becomes mixed with the manure, and any larvse present in the corn fodder do not survdve the treatment. 62 MASS. EXPERIMENT STATION BULLETIN 189. Application of Arsenicals to Plants. Although much of the literature dealing with the habits of the European corn borer emphasizes the fact that the larva feeds entirely within the plant, close observation of the habits of the insect has shown that a large proportion of the first and second instar larvae feed almost exclusively on the upper and lower leaf epidermis of some of their host plants. This circumstance at once suggests the possibility of control by the application of arsenical poisons, and experiments were attempted during the summer of 1918 in order to determine this point. Dusting with Lead Arsenate. An application of powdered lead arsenate was made on June 24, 1918, to 60 hills of sweet corn growing in the ex^ierimental plot at West Med- ford, Mass. At this time most of the corn borer larvse were feeding on the leaf epidermis or on the staminate flowers of the tassel. An attempt was made to get the poison into the unfolding tassel and around the bases of the corn blades, as well as to cover the surface of the leaf blades. This treatment did not noticeably curtail the activities of the larvse. When the ears developed they were infested in the same proportion as the check rows. Other Dusting Experiments. Calcium arsenate powder and equal parts of calcium arsenate powder and hydrated lime were applied in the same manner as arsenate of lead. The results were the same, although calcium arsenate appeared to be more effective than any of the other arsenical powders used. The check rows used in the calcium arsenate experiment were noticeably infested to a greater degree than the treated row. All the ears in the treated row were at least somewhat infested, however. Spraying with Lead Arsenate. Three applications of lead arsenate, at the rate of 1 ounce of the powder in 2 gallons of water, were made to 32 hills of sweet corn on Aug. 5, 13 and 22, 1918. Daily observations were made of these corn plants, and an effort was made to apply the poison at a time when it would be most effective in covering the surface areas of the plant that was being eaten by the larvse of the borer. At the time of apphcation the poison spray adhered to the foliage very well, and the excess liquid ran down the leaf blades and collected at the bases of the tassels and leaf blades, these points being the favorite feeding places of the young larvse. When the ears developed in this plot a close examination showed that 211 ears were present, of which the entire number were infested. Many THE EUROPEAN CORN BORER AND ITS CONTROL. 63 of these ears were only damaged to a slight degree, however, and in general' were in a much better condition than those in the check rows. About 52 per cent of the tassels were broken over in the sprayed plot while 61 per cent were broken over in the check rows. The stalks of the sprayed plants were all infested by the pest, but surface feeding had been entirely prevented. The sprayed plants had a much better (greener) color than the plants in the check rows. Late in October most of the plants in the check rows had fallen over as a result of P. nuhilalis attack, but only about 10 per cent of the sprayed plants had done this. The results of this experiment indicate that many of the European corn borer larvae can be killed by the application of arsenicals at the right time, but that the damage to the plants by the insect cannot be prevented to a paying degree. Corn grows very rapidly throughout the period when spraying is neces- sary, and the newly developed portions of the plant are the favorite points of attack, viz., bases of the leaf sheath, surface of the leaf blade, and the tassel. Tliis necessitates frequent sprayings in order to combat the larvae of the pest, which hatch over quite an extended period of time. The cost of sprajdng large areas would, therefore, be probably prohibitive. Spraying with Calcium Arsenate. Three applications of calcium arsenate, at the rate of one-haK ounce of the powder to 2 gallons of water, were made on the same date and in the same manner as have been detailed for lead arsenate. The results were practically the same, although calcium arsenate appeared to be more satisfactory in its prevention of injury than did lead arsenate. Cultural Practices to avoid Damage. Several observations made during the summer of 1918 seemed to suggest the possibihty that damage by the borer could be avoided by regulating the time of planting corn so that the plants would not be at a stage to attract the female moths of the insect during their time of activity. The female moths prefer to deposit their eggs upon some plant bearing a soft, green seed head. If corn plants bearing a tassel are not available the females habitually deposit their eggs upon some other species of host plant that bears a seed head in the desired stage of development. It was observed that adjoining corn fields, in different stages of develop- ment, were often infested in varying degrees by the insect. In one market garden at West Medford, Mass., a field of sweet corn, planted on April 1, 1918, was very severely infested by the borer. An adjoining field of sweet corn, planted about April 10, 1918, was only infested to a moderate degree. A third field of sweet corn, planted about April 30, 1918, was practically free from the pest, and an examination of the ears when harvested showed only a very small per cent of injury. 64 MASS. EXPERIMENT STATION BULLETIN 189. OTHER INSECTS FREQUENTLY MISTAKEN FOR THE EUROPEAN CORN BORER. The Stalk Borer. The stalk borer Papaipema nitela Gn. is frequently mistaken for the European corn borer. P. nitela attacks and tunnels in the stalks of a great variety of plants, including corn, tomatoes, potatoes and many other wild and cultivated plants. During the spring and early summer the larva is quite commonly found in the same field and often in the same plant with the European corn borer, but it may be distinguished from the latter during its early stages by the presence of a wide transverse brown band extending around the middle of the body. When nearly full grown the P. nitela larva more closely resembles P. nuhilalis, but may be easily distinguished from the latter at that time by the absence of the short stout spines which arise from the light-colored abdominal areas of P. nubilalis, and by the uniformly greater length and breadth of the P. nitela larva. Another point of difference between the two species is that P. nuhilalis pupates within its larval tunnels, while P. nitela leaves its host, when full grown, and pupates in the soil. In corn the larval tunnels of the two species are quite often similar, but the tunnels of P. nuhilalis are generally packed with a light colored frass, and in some instances contain the empty pupal skin, while the lai-val tunnels of P. nitela are generally free from frass, or, if present, the frass is much darker and composed of larger particles than that of P. nubilalis. Many reports of P. nubilalis injury have been found, upon investigation, to have for their basis the injury caused by P. nitela. The Corn Ear Worm. Larvae of the corn ear worm Chloridea obsoleta Fab. are sometimes mistaken for those of the European corn borer. The larvae of the first- named species are frequently found feeding on the same ear of corn "with larvae of P. nuhilalis, but may be easily distinguished from the latter by the presence of varicolored stripes running lengthwise of the body, and by theJact that larva? of the corn ear worm, as the name implies, confine their operations, when feeding on corn, almost exclusively to the kernels of the ear, and do not enter the cob or the stalk. They may generally be found feeding on the surface of immature ears. Cutworms. Several species of cutworms are occasionally found feeding on the ears of corn, but may be distinguished from larvae of the European corn borer by the same characteristics as have been mentioned in the instance of the corn ear worm. THE EUROPEAN CORN BORER AND ITS CONTROL. 65 SUMMARY. The European corn borer has recently become established in the eastern part of Massachusetts. This pest has long been recorded in Europe and Asia as one of the most serious insect enemies of corn, hemp, millet, hops and other crops. It was probably introduced into Massachusetts through the importation from Europe of raw hemp for use in cordage factories, about the year 1910. The insect was first discovered in Massachusetts in the summer of 1917. At that time it was causing severe damage to sweet corn and other plants. Preliminary investigations indicated that the insect had become established over an area of about 100 square miles immediately north and northeast of the city of Boston, and that the serious nature of the pest called for prompt and vigorous action by both State and Federal authorities if the corn crop of the country was to be safeguarded. During the season of 1918 the Massachusetts Agricultural Experiment Station and the United States Bureau of Entomology co-operated in a further investigation of the insect, in order to obtain detailed information concerning its distribution, habits and food plants, with a view to insti- tuting quarantine and control measures that would confine the pest to its present area and lead to its ultimate control. As a result of these investigations it was determined that up to Movem- ber, 1918, the European corn borer had estabhshed itself in an area of about 320 square miles, comprising 34 towns, located immediately west, north and northwest of the city of Boston. The insect attacks a great variety of both wild n CO CO I" ^o| Hs;^ ^ CO ^ s CO •* •« \n f^ O I-- „ in J ■* to ■* 00 o "" H •*j '"'' S « ^o1 1 1 ^ S % s s § s s 2 8 s s s s ' -.^ y I o m \n CO 00 Z ^o| CO "* " CO " " (.■J " S t^ s (n 1 CO CO ■* H ^ ^ ti s 00 CO lO lO u < « ^'Sl " o « « > "i ^ s on _ _, m o >o CO >o o o lO rt 1 0 0 ^ '" "* m « H M § ,n ^ ^ K ^al M t~ " " O T, 5 41 ■* ^ ss ^ "* s « 5 ^ CO ' s !5 . s 1 s 2 s s E= s - s g s s s 3 e2 ?^ ^ ■* ss § CO s S 2 o 2 S o s Trt Oi o § ?5 c. s 55 S ^ oo ^ s s ^ s g ^a° 2; t. ^ r-> 1 •^ -2 o S3 Si k^ S ?J g H Ul 3 -, it cT a 1 a ( 0 .2 > i 5 ^ s S 1 3 ^ 1 eq 0 1 X a ^ 1 1 1 i ^ 1 1 1 1 •< THE PROPAGATION OF APPLE TREES. 91 Grafting on Known Roots. Once trees are established on roots of known varieties it would seem a desii'able process to dig such trees and cut off the greater joart of the root system and replant them, that they may re-establish themselves on a renewed root system. Then the roots cut off may be used for grafting in the ordinaiy manner with scions of the same varietj^ as the root. By this method own-rooted trees should be secured without resorting to the seedhng nurse root, the subsequent removal of which is a severe check to the 3^oung tree, especially with those varieties that do not root freely. This method was tried out in 1915-16. Trees were dug in the fall and all roots suitable for whip grafting removed and the trees reset, the tops being severely cut back. All recovered and in time became vigorous trees. The roots were stored in moist sand and grafted in February and set in April. For some reason they failed to make a good stand, and those that did grow made less growth than adjoining trees grafted in a similar manner on seedling roots. The number of grafts planted, and the percentages growing in July after planting and also in July a j^ear later, are shown in Table 11. Seedling roots used in grafting are commonly one year old, while some of these roots were three or four years old, and this may have been responsible for the poor stand. The very fine sand in which the roots were stored was rather wet and compact, and this may have interfered with respiration and resulted in injury to the roots. It seems hardly reasonable to suppose that such poor results must necessarily follow grafting on the roots of known varieties. Table 11. — Grafts on Known Roots. • Number planted. Per Cent growing. July, 1916. July, 1917 Ben Davis, Bough, Northern Spy, . Red Astrachan, Wagener, . Wealthy, . 92 MASS. EXPERIMENT STATION BULLETIN 190. HISTOLOGY OF THE TWIG IN RELATION TO ROOT FORMATION. 1 Roots on the scion usually arise near a bud, either singly or in twos or threes. No case has been observed when roots arose at a node opposite the bud. Roots may also arise from the internode, but generally within a half inch of the node. Generally they arise above rather than below the bud. The first indication of the root is the falUng away of the axillary bud and the appearance of a swelling with two or three brownish white areas, — the growing points of the young roots. Free rooting varieties develop roots early in the season. An examination of Bough grafts in July showed that they were rooting freely. At the same time Red Astrachan, Ben Davis and Tompkins King showed incipient root formation in a few cases, while poor rooting varieties showed no signs of roots. An examination about the middle of October showed progress in aU these varieties, but the poorer rooting varieties showed hardly a tree with roots from the scion. Always, on digging, the poor rooting varieties have small roots (see Fig. 3) which have evidently formed the second season of growth. If we examine a cross section of a one-year-old twig we find between the bark and wood the cambium, consisting of a layer of eight to fourteen very small, thin-walled rectangular cells. Measurements of the thick- ness of the cambium layer were made and the number of cells noted on a number of the varieties used. Measurements of the thickness of the bark were also made. In choosing material, fresh twigs of the previous season's growth, from both bearing and nursery trees, were selected, and cross sections made usually at the fifth node back from the terminal bud. In the case of some immature tips it was necessary to go further back to secure a plump, mature bud. Sections were made with a sliding microtome and placed at once in 30 i^er cent alcohol for ten minutes. Then the alcohol was poured off and the sections stained for three to five minutes with Delafield's Hematoxylin, washed, mounted on the slide and measured at once. Measurements of the bark were to the wood, and included the cambium layer. They were made at a point one-fourth around the circumference of the twig from the bud when possible, and in aU cases care was taken to avoid the thickened bark near the bud. The limits of the bark as thus defined were clear, but more difficulty was experienced in measuring the cambium layer because of a less clear differentiation between it and the phloem. Often there are two or three cells that have no distinctive features of either cambium or phloem. In order to estab- lish a limit the phloem was considered as starting with the first cell, in which the cells were markedly larger and more rounding, with walls less 1 This discussion is based on work by Robt. P. Armstrong, graduate assistant, to whom the credit for it is due. THE PROPAGATION OF APPLE TREES. 93 deeply stained. In this way a fairly satisfactory criterion was established. (See Plates III and IV.) Four to thirteen twigs of each variety were ex- amined and five to ten measurements and counts of cambium cells made on each twig. No differences were detected between shoots from nurs- ery trees and from bearing trees. Table 12 gives the results of these measurements. Table 12. — Thickness of the Bark and Cambium. Per Cent rooting. Thickness of Bark in Millimeters. Thickness of Cambium in Microns. Number of Cambium Cells. Range of Number of Cambium Cells. Bough (Sweet), . Primate, Red Astrachan, . Tompkins King, Mcintosh, . Northern Spy, . Baldwin, . Yellow Transparent, Oldenburg, . Jewett, Tolman, 98 92 67 62 74 58 32 26 25 20 3 .513 .628 .633 .613 .525 .665 .611 .571 .743 .689 .592 80.0 80.5 78.0 75.0 56.0 69.8 75.0 67.2 58.0 10 10 10 9 6 9 9 9 7 9-11 9-10 9-13 9-10 8-10 8-9 -9 8-9 -9 7-8 It appears from this table that there is a difference in the thickness and number of cells in the cambium layer of the varieties examined, and that this is correlated with the ability of the variety to form roots from the scion. The only marked exception shown in the table is the Baldwin, which, having the fewest cells and the thinnest cambium layer of all, roots more freely than four of the other varieties studied. Further study of this question, including other varieties and extending through the growing season, should prove definitely whether we have here a significant reason for the variation in root formation among different varieties. DISCUSSION OF THE RESULTS. As a major result of the work here reported two facts are brought out : (1) varieties differ greatly in their readiness to form roots from the scion when propagated by the nurse-root method; (2) there is also great varia- tion within the variety in the number that form roots from the scion. Taking up first the varietal differences we find that a few varieties root in all, or nearly all, cases, while only one variety of Pyrus malus — Bethel — has failed entirely to yield trees rooted from the scion. Inas- much as this variety was grown in rather small numbers and under con- 94 MASS. EXPERIMENT STATION BULLETIN 190. ditions where other varieties gave low percentages of rooting trees, it is probable that Bethel would, under more favorable conditions, give at least a low percentage of rooted trees. Considering the number of va- rieties tested it seems safe to say that any variety of the common apple may be propagated on its own roots by the nurse-root method. There are fourteen varieties that have been propagated in consid- erable numbers in successive years and under different conditions, so that we may feel fairly certain that the percentage rooting is fairly rep- resentative for these varieties under the general conditions in which they have been grown. Arranged in order of percentage rooting they are as follows : — Bough (Sweet), 98 Rhode Island Greening, 30 Red Astrachan, 67 Oldenburg, .... 26 Northern Spy, 58 Yellow Transparent, . 26 Ben Davis, 51 Wealthy, .... 25 Wagener, . 45 Hubbardston, 21 Transcendent, 45 Jewett, . . . 20 Baldwin, 32 Tolman, .... 3 Coming now to the question of why certain of these varieties root better than others we find a rather difficult problem. We have made few investigations aimed directly at this question, but some discussions may be ventured. The property of rooting is not directly correlated with vigor. Tolman is fully as strong growing a variety in the nursery as Bough. Further- more, observations made on digging the trees fail to discover any noticeable correlation between vigor and rooting. It has seemed to the writer that a small, weak tree was as likely to be rooted from the scion as a strong one. Some varieties branch more freely than others. During the season of 1916 a block of yearling whips branched quite freely from the newly formed axillary buds. Notes taken at the time are as follows: No branches, Northern Spy; few, Baldwin, Bough, Oldenburg, Tolman; all. Trans- cendent (Crab). This gives no indication of any correlation between rooting from the scion and branch growth from axillary buds. A more reasonable expectation might be for a correlation between root formation and branching from adventitious buds on the stem. No exact record of branching from adventitious buds is available, but limited general ob- servation of the behavior of budded trees leads the writer to believe that such a correlation may exist, and that Bough and other free rooting varieties do send out shoots from adventitious buds more freely than Tolman and other varieties that root only sparingly. Further and more definite records may prove or disprove this belief. The relation of callus formation in cuttings has been referred to. (See page 75, Fig. 1.) Unfortunately no full notes of callus formation on the cuttings set was kept, but it is suggestive to point out that Yellow THE PROPAGATION OF APPLE TREES. 95 Transparent, which uniformly gave as large a callus as any variety, did not root as well as Wagener, which never gave any sign of callus forma- tion. Neither can we discover any relationship between rooting from the scion and season of maturity, either of fruit or wood, nor in size of leaves or density of foliage. Many woody plants are propagated from cuttings, and in general it is those with soft wood that grow most readily. There is considerable variation in hardness of wood among different varieties of apples, and we may inquire if those with softer wood are the ones that root most readily from the scion. No extended investigation of this question has been made at this station. Beach and Allen ^ made extensive tests of the hardness of wood of different varieties. They found considerable difference withm the variety, and a clear comparison of their results with rooting ability, as shown by their investigation, is difficult, but a general survey of their results leads to a belief that there is a general correlation. It is, however, subject to exceptions. Beach and AUen came to the con- clusion that there was a correlation between hardness of wood and resist- ance to winter cold, and here again there seems to be a rather loose cor- relation with rooting ability. Oldenburg and Wealthy are very hardy and root poorly, and Bough is tender and roots well. But Ben Davis is quite hardy and roots comparatively well, and Hubbardston and Tolman are less hardy than Wealthy and do not root so well. Wide variations in the rooting ability of different lots of the same variety are evident. Some of these are clearly seasonal. Such differences may be due to climatic conditions, to soil conditions, — for the soils used in different years are not all alike, — or they may be due to difference in the scions used. Any such difference would most likely trace back to the gi'owing conditions the previous season as affecting stored food and possibly structure. Slight differences in cultural treatment may have had an effect. Varying rainfall may have had an influence. It is im- possible from the evidence at hand to determine which of these possible factors have had an influence and to what extent. The general low percentages of Series 6 (Table 5) are striking, and the writer feels that they are due largely to poorly drained soil which prevailed over a considerable portion of the plot. While no direct comparisons are possible, careful observation indicated that rooting was better on the drier portions of the plot. A part of the plot on which Series 4 was grown was poorly drained, and may account for the rather low average of this series. 1 la. Expt. Station Res. Bui. 21 (1915). 96 MASS. EXPERIMENT STATION BULLETIN 190. SUMMARY. 1. Stem cuttings of the common apple grow only rarely; in the trials here reported none succeeded, though callus formation in some varieties was good. 2. Root cuttings grew well, especially when young roots were used, though growth was slow the first season. 3. Limited tests indicated that most varieties may be readily propa- gated by mound layers. 4. The best means of establishing trees on known roots is by the nurse- root method. The scion is wliip-grafted on a short piece of root and planted deeply; at the end of one or two seasons' growth the tree is dug, the seedling root removed and the tree replanted. Neither dwarf apple nor pear roots are of value as nurse roots. 5. Varieties vary greatly in the readiness with which they send out roots from the scion, the proportion varjang from none to practically all with different varieties. 6. There is also great variation within the variety in the numbers rooting from the scion. 7. Varietal differences may be loosely correlated with density of the wood, the softer the wood the higher the proportion rooting from the scion. 8. A fertile, well-drained, sandy loam probably offers the best conditions for securing a high percentage of rooting trees. 9. Once trees are estabhshed on known roots they may be propagated by root cuttings or by root grafting on known roots. 10. There seems to be a relation between the varietal ability to pro- duce roots from the scion and the thickness of the cambium layer at the dormant season. PLATE I. Fig. 1. — Green wood apple cuttings, showing callus formation. From left to right, Yellow Transparent, Fall Pippin, Red Astrachan, Bough, Ben Davis, Wagener. Fig. 2. — Matching cambium in root grafts: (a) one side only; (6) both sides only; (c) top only; (d) bottom only; (e) perfectly matched. Fig. 3. — Trees rooted from the scion after cutting off seedling nurse root; two-year-old trees cut back in spring of second year. Tolman at left, Bough at right, showing stronger roots of the latter. • • ■ 1 . 1 r 1 .=^ / r~«T J -i V > pfx_ \ y^ ^^€^^ V ] 1 r^-"^ \ -ri^_,__,„_ Fio. 4. — Own-rooted I{od A^tr^tchan two >tar^ altir cutting ofif seedling root. PLATE III. Fig. 5. — Section of Bough scion, showing origin of a young root. ^^K^^^K!SkJKKMSI^^^^BK^^tm\^. '3m^'''%L£M ; 1 "'■'•'■J ^*' . * ,^^Hfci<^rTfelrTi ^J^' Fig. 6. — Section of Bough, showing wlcin, ( iinliuun and plilc layer ha'^ nine or ten cells I PLATE IV. :^ M^'^.^^-'^ ■■■■■■p ^^^^^^ uJ^t' \. *'^/** % ^. . -^^ i^^ ^^£ "*» v\^^|^^ Cj^JBB|| sst-^^^^S t\^B I^^^^K V^vflL^' l^^^l J^WT^^^S' fmSH Xylem. P"iG. 7. — Section of Baldwin, --howins the thin cambium layer, averaging about five cells. Xylem. Fig. 8. — Section of Tolman, showing cambium layer, averaging about eight cells. 5 I MiimiiL