Digitized by the Internet Archive in 2011 with funding from LYRASIS members and Sloan Foundation http://www.archive.org/details/tobaccosubstatioOOande L^-^^ I Bulletin 326 February, 1931 , Hl^^c TOBACCO SUBSTATION AT W^INDSOR REPORT FOR 1930 p. J. ANDERSON, T. R. SWANBACK, O. E. STREET AND OTHERS Qlnnnrrttrut Agnrultural iExpmitmtt ^tatinn bulletins of this station are mailed free to citizens of Connecticut )ply for them and to other applicants as far as the editions permit. CONNECTICUT AGRICULTURAL EXPERIMENT STATION BOARD OF CONTROL His Excellency, Governor Wilbur L. Cross, ex-officio. President Elijah Rogers, Vice-President • • ' Southington George A Hopson, Secretary Mount Carmel William L. Slate, Director and Treasurer Mew Haven Joseph w. Aisop ;;--;j,- . °" Edward C. Schneider Midd^etown Francis F. Lincoln .Cheshire S. McLean Buckingham Watertown STAFF E. H. Jenkins, Ph.D., Director Emeritus. Administration. William L. Slate, B.Sc, Director and Treasurer. Miss L. M. Brautlecht, Bookkeeper and Ltbrarian. Miss Dorothy Amrine, B.Litt , Editor. G. E. Graham, In Charge of Buildings and Grounds. Analytical E. M. Bailkv, Ph.D., Chemist in Charge. Chemistry. C. E. Shepard ^ Owen L. Nolan I . . ^, • Harry J. Fisher, A.B. >Asststaiit Chemists. W. T. Mathis David C. Walden, B.S. J Miss Harriet C. Yale, General Assistant. Frank C. Sheldon, Laboratory Assistant. V. L. Churchill, Sampling Agent. Mrs. A. B. Vosburgh, Secretary. Rinrhemistrv H B. ViCKERY, Ph.D., Biochemist in Charge. Biochemistry. george W. Pucher, Ph.D., Assistant Biochemist. Mrs. Helen Cannon Cronin, B.S., Dietitian. Q p. Clinton, Sc.D., Botanist in Charge. ^°'^"^- e' M. Stoddard, B.S., Pomologist. . m'iss Florence A. McCormick, Ph.D., Pathologist. A A. DuNLAP, Ph.D., Assistant Mycologist. A. D. McDonnell, General Assistant. Mrs. W. W. Kelsey, Secretary. E mology W. E. Britton, Ph.D., D.Sc, Entomologist in Charge, State " * Entomologist. B. H. Walden, B.Agr. 'j ^- ^- ^^^.l^^.«■^P« D ^Assistant Entomologists. Philip Garman, Jth.jj. i Roger B. Friend, Ph.D. J . ^, ^ _.^ ,, . „, . John T. Ashworth, Deputy in Charge of Gipsy Moth work. R C BoTSFORD, Deputy in Charge of Mosquito Elimination. J." P.' Johnson, B.S., Deputy in Charge of Asiatic and Japanese Beetle Quarantines. Mrs. Gladys Brooke, B.A., Secretary. Fnrpcjtrv WALTER O. FiLLEY, Forester in Charge. ^° ^" H W. HicocK, M.F., Assistant Forester. J. E. Riley, Jr., M.F., In Charge of Blister Rust Control. Miss Pauline A. Merchant, Secretary. Plant Breeding Donald F. Jones, Sc.D., Geneticist in Charge. W. R. Singleton, Sc.D., Assistant Geneticist. Lawrence C. Curtis, B.S., Assistant. Mrs. Catherine R. Miller, M.A., Secretary. Soils M. F. Morgan, M.S., Agronomist in Charge. H. G. M. Jacobson, M.S., Assistant Agronomist. Herbert A. Lunt, Ph.D., Assistant in Forest Soils. Dwight B. Downs, General Assistant. Tobacco Substation Paul J. Anderson, Ph.D., Pathologist in Charge. at Windsor. T. R. Swanback, M.S., Agronomist. O. E. Street, M.S., Plant Physiologist. Miss Dorothy Lknard, Secretary. the TUTTLE, MOREHOUSE & TAYLOR COMPANY, NEW HAVEN, CONN. CONTENTS Page Potash Fertilizer Experiments 357 Comparison of carriers 357 Sulfate, carbonate and nitrate of potash 357 Series of 1925 357 Series of 1927 361 Summary and discussion 363 Ground tobacco stems as a source of potasli 363 Chemical analyses of tobacco on stems plots 364 Cotton hull ashes as a source of potash 366 Quantity of potash 368 Series of 1926 368 Series of 1927 370 Effect of the quantity of potash applied in the fertihzer on the percentage of potash, Hme and magnesia in the leaf 371 Nitrogen Fertilizer Experiments 374 Urea as a source of nitrogen 374 Calurea as a source of nitrogen , 376 Nitrophoska tests ill Comparison of nitrate of soda and nitrate of hme as a mineral source of nitrogen 379 Burn tests 380 Fractional Application of Fertilizer 381 Manure as a Supplement to Commercial Fertilizer 384 Crop of 1930 384 Black rootrot on the manure plots 385 Increase in soil organic matter from use of manure 386 Increase in water-holding capacity of soil oW Effect of manure on the burn 387 Cover Crop Experiments 388 The Relation of Magnesia to the B,urning Qualities of Cigar Leaf Tobacco 391 354 Connecticut Experiment Station Bulletin 326 Page Effect of Topping and Suckering on Development of the Tobacco Plant 399 Influence of Plant Trimming on Weight of Seeds . 406 Curing Experiments 411 Tobacco Insect Studies in 1930 419 Survey of tobacco insects 419 Experiments with insecticides 421 Miscellaneous observations on tobacco insects 424 Flea beetle 424 Grasshoppers 427 Late infestation of wireworms 429 Stalk borer 430 Hornworms 430 Paris green injury to tobacco 431 Fertilizer Losses Through Leaching Measured by Lysimeter Experiments 432 The Use of Coke in Heating Tobacco Sheds 442 Fire-curing Tests on Stalk Tobacco 445 Appendix 447 TOBACCO SUBSTATION AT WINDSOR REPORT FOR 1930^ The season of 1930 was generally favorable for growth of tobacco, and satisfactory progress has been made on most of the projects under way. The temperature during the growing season was above average and the rainfall sufficient and well distributed up to the latter part of July. The accompanying table, which gives the rainfall by 10-day periods during the growing seasons of the last two years, shows how much more favorable were the distribu- tion and amount, than in 1929. Table 1. Distribution of Rainfall in Inches at the Tobacco Substation, Windsor, 1929 and 1930 By 10-day periods ^ear 1929 1930 1-10 1.44 0.43 — May- 11-20 1.56 2.01 21-31 1.79 2.04 1-10 0.08 2.38 — June ^ 11-20 21-30 0.03 1.56 0.73 0.62 By months 1-10 0.10 0.93 -July , 11-20 21-31 0.44 0.44 0.69 1.01 , — August — X 1-10 11-20 2.73 2.13 0.63 1.53 1929 1930 Mean' May 4.79 4.48 3.55 June 1.67 3.06 July 0.98 2.63 4.31 August 4.86 2.16 2.93 The weather in 1930 resulted in unusually rapid growth and development of the plants, and it was found necessary to start har- vesting before August 1, an unprecedented occurrence at the Sta- tion. During the latter half of July, however, the rain was insuf- ficient and the unusually hot days of low humidity caused tobacco on light lands to wilt badly. In many fields the lower leaves were spoiled by burning on the hot sands. Serious damage on the Station farm from this source was avoided by irrigating twice during the drought. The installation of this irrigation system, which consists of a fire hydrant and 700 feet of ordinary fire hose, was the most notable addition to our equipment during the year. Although the Connecticut Valley was visited by five destructive hail storms within the season, the crop on the experiment station farm was untouched. The loss from hail to the crop throughout the Valley approached but did not equal that of 1929. Fertilizer experiments on broadleaf tobacco were begun this year on the Farnham farm at East Windsor Hill, under the super- ' For bibliographical purposes all material should be credited to P. J. Anderson, T. R. Swanback and O. E. Street, unless otherwise indicated. '' Hartford records for the past 57 years. 356 Connecticut Experiment Station Bulletin 326 vision of J. S. Owens, Extension Agronomist of the Connecticut Agricultural College. The tobacco on these plots was destroyed by hail and therefore no results were recorded. Two assistants were added temporarily to the scientific staff during the summer. Donald S. Lacroix, under the direction of the Department of Entomology, spent three months in a survey and investigation of insects that attack tobacco. His summary of the work accomplished is published on page 419 of this bulletin. Dr. Th. Berthold, until recently tobacco expert of the province of Brandenberg, Germany, was on our staff throughout the summer under the direction of the Department of Plant Breeding. He continued the breeding work with shade tobacco and conducted investigations along other lines, some of which are reported on pages 319 and 406 of this bulletin. The purchase of a small field, the Mellon tract, adjacent to the Station farm furnished an opportunity for conducting experi- ments with phosphoric acid on land that had not grown tobacco previously, the only condition in which phosphatic fertilizers seem to be necessary. Leasing another field, the Pomeroy tract, directly across the road from the station farm also furnished an oppor- tunity for a comprehensive set of field plots on which the use of magnesian lime is under investigation. Dr. Vickery and Dr. Pucher of the Department of Biochemistry have continued their studies of the tobacco plant and have made valuable contributions to our knowledge of the organic acids of tobacco and of the changes in the nitrogenous compounds during curing. The work is published in Bulletins 323 and 324. Impor- tant investigations conducted by the Department of Soils on nutri- tion of tobacco have been published recently in Bulletin 320, "The Soils of Connecticut." A project on "Improvement of Havana Seed Strains," which has been inactive since 1926, has been resumed on a larger scale through the interest and financial support of three cigar manufac- turing corporations, Bayuk Cigars, Incorporated, the -General Cigar Company, and the Congress Cigar Company, under the direct supervision of W. H. Hatheway. We are cooperating with the Office of Tobacco Investigations, United States Department of Agriculture, in a project under the supervision of C. V. Kight- linger to find or develop strains of Havana Seed tobacco of satis- factory quality that are resistant to black rootrot. Fertilizer experiments with special reference to the use of manure for shade tobacco are being conducted in West Granby in cooperation with the growers F. S. Holcomb and Son and with the American Cigar Company. The following pages contain progress reports on projects of the Tobacco Substation that are sufficiently advanced to warrant pub- lication of results now obtained. POTASH FERTILIZER EXPERIMENTS The object of all the potash fertilizer experiments continued in 1930 is to determine, (1) which carriers of potash are best and (2) which is the optimum quantity to apply. Each involved the use of a different set of field plots and the two are discussed separately. o Comparison of Carriers Potash is ordinarily supplied in the tobacco fertilizer mixture as the sulfate, nitrate or carbonate of potash. Double sulfate of potash-magnesia has been used to some extent. The conclusion of a six-year experiment on this last salt was given in the Report of the Tobacco Substation for 1928. Carbonate of potash may also be supplied in cottonhuU ashes. Another source of potash is tobacco stems, originally applied in the unground condition separate from the fertilizer, but now supplied in a finely ground condition suitable for incorporation with the other ingredients. Sulfate, Carbonate and Nitrate of Potash In order to compare these carriers or combinations of them a series of ten plots was started in 1925 (designated in this report as "Series of 1925"). The experiment was enlarged in 1927 by the addition of 18 plots on a better type of soil (designated as "Series of 1927"). These two series are discussed separately before summarizing all results. Series of 1925.^ After the destructive hailstorm of August 1, 1929, the tobacco on the ten plots was harrowed in and the entire field sowed to oats, which made a luxuriant growth before winter, some plants even heading out. The location of these 10 one-fortieth acre plots on Field V was exactly the same as in previous years and they received the same fertilizer ration in 1930, applied on May 22.- Plants were set on May 27. After the heavy leaching rains of June 8 to 10, growth of the plants was somewhat checked, in a way which made us believe that the nitrogen had been leached out. This was also confirmed by results on the lysimeter. Therefore, on June 19, a side dressing was applied and worked in, consisting of 100 pounds nitrate of lime and 300 pounds cottonseed meal per acre. ^ In previous reports designated as "Old Qualitative Potash Series." ^ The composition of all fertilizer mixtures referred to in this bulletin is given in the appendix to which the reader is referred. See formulas Kl, K5, K7, K8, and K9. 358 Connecticut Experiment Station Bullciin 326 These plots are in duplicate. The five plots on one end of the field are identical in treatment with the five on the other end. As an irrigation experiment, the five plots on the east end were irri- gated three times (July 19, 23 and 27), when the weather became excessively dry and hot. The west end received no water. All the tobacco (except border rows) was harvested on July 30 and sorted in December. The only noticeable differences between the plots during the summer were the better growth on the east (irrigated) end and the Figure 18. Two-row setter, drawn by tractor on Station farm. All oper- ations on the farm are by tractor. poor growth on plots Kl-2 and K5 which, as mentioned in previous reports, were located in a corner apparently unfavorable for good growth. The sorting records of 1930 are presented in Table 2 and the- results for five years are summarized in Table 3. These show a slight advantage in favor of the triple source of potash. However, when it is recalled that the poor plots Kl-2 and K5 on the west end make this row of questionable value and we compare only the plots on the east end with each other (Table 4), it will be seen that the differences both in yield and grade index are quite small, not sufficiently large to be of significance. Potash Fertiliser Experiments 359 The results indicate that as far as yield and grading" are con- cerned it makes little difference which carrier or combination of carriers is used. Table 2. Comparison of Sulfate, Carbonate and Nitrate of Potash. Acre Yield and Grading of Crop of 1930. Series of 1925 , Percentage of grades ^ Grade L M LS SS LD DS F B indexi 3 10 9 14 28 21 15 0 .327 7 12 14 9 39 6 12 1 .395 2 3 14 10 32 18 18 3 .305 7 10 19 9 35 7 13 0 .403 1 3 21 12 31 15 14 3 .330 10 13 20 9 31 4 12 1 .409 2 6 21 11 34 12 12 2 .367 10 11 18 9 35 5 12 0 .438 9 13 16 10 32 9 10 1 .419 11 10 20 9 30 7 12 1 .434 ' In comparing the quality of tobacco grown on different plots it is very difficult to keep in mind the percentage of eight commercial grades of tobacco from one plot and compare it with a like number from another. To simplify these comparisons a grade index is used. The grade index is a single number expressing the grading of all the tobacco grown on a par- ticular plot. It is based on the percentage of carefully assorted commercial grades and the relative price value of the different grades. Although market prices vary from year to year, the ratios of prices between the different grades are fairly constant. These adopted price relationships for the different grades are as follows : (L) Light wrappers $1.00 (LD) Long darks (19" up).. $.30 (M) Medium wrappers 60 (DS) Dark stemming (17").. .20 (LS) Long sec. (19" up) 60 (F) Fillers 10 (SS) Short seconds (15" and (Br) Brokes 10 17") 30 The grade index of any plot is obtained by multiplying the percentage of each grade by the price in the above schedule, adding the products and dividing by 100. Carrier of potash Sulfate Plot No. Kl-2 Kl-3 Acre yield 1160 1431 Carbonate K5 KS-1 1194 1389 2/3 nitrate 1/3 carbonate K7 K7-1 1230 1398 1/2 carbonate 1/2 sulfate K8 K8-1 1305 1431 1/3 carbonate 1/3 sulfate 1/3 nitrate K9 K9-1 1470 1422 360 Connecticut E.vpcrinient Station B nil din 326 <■" O Q < 3 ,-t '^ ^ u^ CO CD a\ t^ CO o\ •* 00 CO O MD m T-H n CO ^ CO-* CT^ -^ ■* voo ooco 01 VO cooo \OCM COVD coco r.] o\ CO U-) ^ Tj- Tt •* -Tl- "* Qi 132 CM -* CO o t^ r^ (^1 1^ r-H C3 r^co CO"— I cvir^ T— (Lo CM '^ coco coco ^ CO o^ CO CO .— I CO 00 ^H cooo ^00 co-1- <>jco u-300 vooo coco coco coco coco vcf^i r^io r^ON >— (\o onJ^ >0(NJ ioOn S=^ 0(N1 0.-H ,— I ,— I ,-H (M ^H O yvoio^ u-ir-i oco c^jco (Ni-^ •<;cJt— I i-oON 00 r^i^i '^■— (CVJ coco coco CO-Tt IVi Tf mOOCO lOiO -t-lO OOf^ CO-* 01,— 110 <^1^ coCT\ lOON VOfM o\-!^io -^i-O -^^O ^-n- u-)io u ■I- . <>1 CO 1-^ r-^-— I.— I 1010 r^t^ 0000 OnOn HMh^ h^h-H h^^^ HHl^ h^HH i/l O C u Potash Fertiliser Experiments 361 Table 4. Comparison of Sulfate, Carbonate and Nitrate of Potash . Average Yields and Grade Index for Five Plots on East End for Five Years. Series of 1925 Average Average grade Carrier yield for 5 years index for S years Sulfate 1347 .403 Carbonate 1323 .420 2/5 titrate { ^3^^ ^^g>j 1/3 carbonate ) 1/2 carbonate [ I355 4^7 1/2 sulfate ) 1/3 sulfate ) 1/3 carbonate \ 1337 .410 1/3 nitrate ) Series of 1927^ This series of 18 one-fortieth acre plots is a replication in triplicate of the 1925 series, but is on Field I, w^here the soil is naturally more favorable for the production of larger tobacco. The fertilizer treatment was the same as for the 1925 series except for the addition of 400 pounds per acre of hydrated lime containing 29 per cent of MgO. The lime was added because this field was somewhat too acid and because the color of the ash needed improvement. The fertilizer was applied on May 22 and the plants were set on June 2. No second application of fertilizer was made on this field because the soil is not so "leachy" as Field V. It did not suffer so much from the dry, hot weather, but was irrigated once during the latter part of July. All the plots were harvested on August 5 and 6, one row only (center row of about 50 plants) being kept for the records. Growth was uniformly good on these plots and the yield heavy, being about the same as for 1923, 1925, and 1926 on this field. When sorted, there was found to be no pole sweat and very little yellow color but considerable white vein on most of the plots of this field. The yield and sorting records, presented in Table 5, do not show very large differences either in the yields or grade indexes for the different treatments when the replicates are averaged. ^ In previous reports designated as "Nev^r Qualitative Potash Series." 363 Connecticut Experiment Station Bulletin 326 Table 5. Comparison of Sulfate, Carbonate and Nitrate of Potash. Acre Yield and Grading of Crop of 1930. Series of 1927 Source of Plot potash No. Kl-4 Kl-S Sulfate Kl-6 Kl-7 Kl-9 , — Acre yield — Plot Average 1846 1884 1884 1910 202S 21 14 11 9 14 Percentage of grades M LS SS LD DS F 7 14 5 40 2 11 10 17 4 43 2 10 8 21 3 46 1 9 4 22 4 47 1 13 10 22 2 42 0 9 0 0 1 0 1 Gradf Plot . .486 .457 .443 .414 .474 : index leverage .455 KS-2 Carbonate KS-3 K5-4 1826 1838 1988 1884 16 19 12 7 7 10 16 18 23 5 3 4 44 43 43 2 0 0 10 10 8 0 0 0 .459 .488 .467 .471 2/3 nitrate K7-2 1/3 carbonate K7-3 K7-4 1900/ 1983 i 2114 1941 (?) 22 13 10 8 4 17 25 25 4 2 3 37 44 48 1 1 0 11 10 8 0 2 2 .506 .450 .437 .464 1/2 carbonate K8-2 1/2 sulfate K8-3 K8-4 1862^ 2010 J 2028 1936 (?) IS 14 8 8 3 9 27 21 23 1 3 5 40 48 43 1 1 2 8 9 10 0 1 0 .493 .449 .432 .458 1/3 carbonate K9-2 1/3 sulfate K9-3 1/3 nitrate K9-4 1762 1984 1944 1897 15 8 13 6 6 6 23 25 28 4 3 3 41 49 39 2 0 9 9 9 2 0 0 0 .472 .431 .471 .458 Records of three years on this series of plots are presented in Table 6. The only differences that may possibly be considered sufficiently large to be significant are the lower average yield and higher grade index on the carbonate plots. This tendency of the carbonate treatment has been noted in previous reports on both series. Table 6. Comparison of Sulfate, Carbonate and Nitrate of Potash. Summary of Three Years. Series of 1927 Source of potash Sulfate Plot , Acre yield \ No. 1927 1928 1930 Average Kl-4 1273 1410 1846 Kl-S 1250 1335 1884 Kl-6 1261 1476 1884 Kl-7 1258 1366 2025 Kl-8 1276 1356 (1910)' Kl-9 1320 1442 (1910) K5-2 1230 1320 1826 Carbonate K5-3 1246 1261 1838 KS-4 1307 1419 1988 2/3 nitrate K7-2 1271 1394 1900 1/3 carbonate K7-3 1250 1330 1983 K7-4 1318 1416 (1941) 1/2 sulfate K8-2 1319 1331 1862 1/2 carbonate K8-3 1280 1391 2010 K8-4 1345 1445 (1936) 1/3 carbonate K9-2 1292 1394 1762 1/3 nitrate K9-3 1319 1353 1984 1/3 sulfate K9-4 1284 1495 1944 , — Grade index — ^ 1927 1928 1930 Average .394 .455 .486 .389 .405 .457 1527 .372 .425 .443 .425 .413 .408 .414 .345 .511 (.455) .380 .420 .474 .410 .431 .459 1493 .394 .481 .488 .450 .421 .499 .467 .419 .416 .506 1534 .380 .507 .450 .426 .326 .392 .437 .396 .416 .493 1546 .369 .389 .449 .429 .463 .458 .432 .414 .385 .472 1536 .365 .441 .431 .426 .403 .452 .471 ^ Figures in parenthesis represent averages of other plots of the same treat- ment and year, and are substituted where the real records were accidentally lost or confused. Potash Fertiliser Experiments 363 Summary and Discussion Aside from these small but persistent differences in the car- bonate plots, it would seem, from all results up to date on both series, that it makes little difference which of the three or which combination of the three carriers is used to supply the potash requirements of a tobacco crop. The yield per acre and the grading" of the crop are essentially the same. This result is not surprising in view of the known behavior of potash salts when incorporated with soil. As soon as any one of these potash compounds goes into solution in the soil, the potassium cation unites with the large acetoid or colloidal exchange anion, displacing" hydrogen, calcium or magnesium, or possibly other cations which in turn unite with sulfate or nitrate or carbonate of the added fertilizer salt. Thus the potash compound, which is more or less stable in the soil and from which the tobacco roots must obtain the plant's ration of potash, is just the same, regard- less of the carrier in which potash was furnished in the fertilizer, and it would be anticipated that the results on the crop would be similar. When sulfate of potash is used, a difference in growth or quality due to increase in amount of sulfate in the soil solution might be anticipated, but such a difference is minimized by the limited capacity of the tobacco plant to accumulate or increase its sulfur content (see Report of the Tobacco Substation for 1929, page 213). In the formula where nitrate of potash is used, there is no increase in the total nitrate added because the nitrate of soda in the mixture is correspondingly reduced when nitrate of potash is added. When carbonate is used, it is decomposed or changed to a bicarbonate and if this anion is taken into the plant it could hardly produce a marked difference in a medium where carbonic acid from the air is ever present in sufficient quantity. Ground Tobacco Stems as a Source of Potash The three plots on which tobacco stems are used as the only source of potash are located on Field I, between the other Source- of-Potash plots. The fertilizer mixture used was essentially the same as in preceding years, composed as follows : 2,650 lbs. ground tobacco stems 1,529 lbs. cottonseed meal 260 lbs. nitrate of soda This formula furnishes 200 pounds of nitrogen and a like amount of potash. For the 1930 experiment, ground sterilized stems were substituted for the long stems used in the experiments of previous years. The stems were mixed with the other fertilizer ingredients and spread in a single application. Ground stems, which are a by-product of the extraction of nicotine sulfate (an 3(34 Connecticut Experiment Station Bulletin 326 insecticide), have certain advantages over long- stems: (1) Being sterilized by steaming in the process of nicotine extraction, they are free from danger of carrying such diseases as wildfire and mosaic to the young plants. (2) The moisture content is low and constant. (3) Their mechanical condition makes them easy to mix with the other fertilizer ingredients and also makes the plant food more quickly available. Dates of setting and harvesting were the same as for the adjacent plots. Table 7. Stems Plots. Acre Yield and Grading as Compared with Other Sources of Potash. Crop of 1930 -Acre yield- Potash Plot carrier No. Plot Average L K14 1958 20 Stems K14-1 1948 1977 17 K14-2 2025 10 1912 — Percentage of grades M LS SS LD DS F 9 15 5 16 8 22 4 42 0 9 5 46 1 9 6 46 0 7 Sulfate' Kl 1910 Carbonate' K5 1884 Nitrate' K7 1941 (average of 4 plots) (average of 3 plots) (average of 3 plots) — ^ Grade index B Plot Average 1 .492 1 .461 .466 1 .444 .455 .471 .463 .464 The yield and sorting records (Table 7) indicate that stems are a sufficient source of potash and a good one, since the yield was somewhat higher than that of plots where other potash carriers were used, and the grade index was about the same. Stems are a desirable fertilizer material because, being a residue of the tobacco plant, it is safe to assume that they contain, besides potash, a supply of all other elements that are needed by the grow- ing tobacco plants. In order to see whether this assumption is warranted, a pot experiment on very poor, worn-out soil was tried in the greenhouse. A heavy application of stems was used as the only fertilizer, on six out of 12 pots. No stems or other fertilizer were applied to the other six. Young tobacco plants in the untreated pots made practically no growth, while those in the stems pots grew rapidly and luxuriantly. One plant of each series is shown in Figure 19. Only in the last stages of growth did the plants grown on stems show some symptoms of shortage of nitrogen. Chemical Analysis of Tobacco on Stems Plots In order to see how much potash the crop actually absorbs from stems, as compared with other potash carriers, samples of the fermented crop of 1928 were analyzed.^ Since it was possible that ' Averages from Table 5. ^ The analyses were made by the Department of Analytical Chemistry, at New Haven. Potash I'crllliccr E.vpcriiiicnts 3G5 the other bases in the leaf might be affected, a determination of the percentage of CaO and MgO was also made. Results of the analyses are presented in Table 8. Do d!ern5 i Figure 19. Plant on right received tobacco stems only as a fertilizer. Plant on left had no fertihzer. Table 8. Analyses of Tobacco from Stems Plots' Source of potash in fertilizer Stems to supply 200 lbs. K2O 200 lbs. K.0 in nitrate, sulfate and carbonate 300 lbs. K.O in nitrate, sulfate and carbonate of potash K9-.S K9-5 K13-2 Kl 3-2 K13-3 K13-3 Plot No. Grade K14 D K14 S K14-1 D K14-1 S ■KoO- Percentage of ■CaO- -MgO- Plot Average Plot Average Plot Average 7.42 9.72 7.79 9.14 7.30 8.64 8..S4 9.14 8.19 8.83 8.52 7.97 8.68 8.57 5.50 5.83 6.22 5.Q7 6.52 5 57 6.05 6.43 6.53 6.25 5.98 .77 .73 .68 .82 .77 .77 .7.^ .78 .68 .77 .80 .74 Thus when 200 pounds of KoO per acre were supplied in stems, the crop absorbed at least as much as when the same amount of potash was supplied from mineral carriers. The percentage in the crop was almost as large as when 300 pounds of potash was supplied in mineral carriers. ' Figures are on air-dry basis ; water content, about seven per cent. 366 Connecticut Experiment Station Bulletin 326 In this connection, it is of interest to note that in the Poquonock experiments of 1892-96, Dr. Jenkins (Connecticut Agricultural Experiment Station Report for 1896, page 331) found the highest percentage of potash (nine per cent) in the tobacco on the plot that had been fertiHzed with stems. However, in these earlier experiments the stems applied to this plot contained more potash (486 pounds per acre) than the fertilizers applied to the other plots (340 pounds per acre). Cotton Hull Ashes as a Source of Potash Thirty to forty years ago, ash resulting from the burning of cotton seed hulls in the South was a commonly used source of potash for tobacco in New England. This ash contained a variable percentage (15 to 35 per cent) of potash in the form of car- bonate, carbonates of calcium and magnesium, and some phos- phoric acid (average about nine per cent), as well as small amounts of other elements, some of which may be of importance in plant growth.^ According to analyses made during the last year on 14 samples (Table 9), the chlorine content varies from .43 to 2.33 per cent. This is probably not sufficient to affect the burn seriously. ' The Station Report for 1897, page 144, gives the average analyses oC 185 samples of cotton hull ashes as follows : Ash of pure hulls Water Coal ) - , ( Sand and soil ( ^'^^ ) Silica 1.21 Oxide of iron and alumina 1.69 Lime 5.29 Magnesia 11.29 Potash 42.13 Soda 3.35 Sulfuric acid 3.04 Carbonic acid 20.10 Phosphoric acid 2.96 Chlorine 1.74 100.36 Deduct oxygen equivalent to chlorine .36 100.00 100.92 Commercial Average cotton composition hull ash (185 analyses) 5.95 9.00 1.86 ) 6.16 [ 1.00 ) 14.04 .54 2.07 15.68 8.85 3.47 9.97 36.76 23.40 1.18 2.58 1.92 2.56 22.22 3.10 9.08 1.40 1.60 101.24 .32 Potash Fertilizer Experiments 367 Table 9. Chemical Analyses of Samples of Cotton Hull Ashes Sold FOR Tobacco in 1930' Lab. No. of K2O P2O5 CaO MgO Cl Boron (B2O3) Water sample % % % % % % 3601 31.67 2.15 10.67 4.35 2.33 0.029 3667 23.35 2.90 12.52 4.38 1.34 0.024 3668 24.60 2.80 11.41 5.17 1.45 0.022 3669 24.79 2.78 11.47 4.99 1.47 0.035 3693 22.95 3.07 10.38 5.05 1.39 0.022 3694 23.64 2.86 10.15 4.48 1.59 0.024 3753 25.27 3.71 8.33 3.93 0.94 0.021 3786 29.97 2.72 8.50 3.97 1.97 0.029 3818 29.07 2.65 9.16 3.93 2.15 0.036 3826 23.03 2.33 15.35 5.64 0.87 0.008 3827 33.57 1.88 12.62 3.94 1.84 0.011 3949 20.97 3.18 19.94 8.14 0.43 0.011 3990 38.17 2.80 14.10 4.93 . 1.76 0.003 4026 25.31 5.00 12.73 5.35 0.83 0.009 average 26.88 2.92 11.95 4.87 1.45 0.020 App. 9.09-10.00 Later, the use of cotton hull ashes was discontinued, but within the last five years this material has again appeared on the market and has been used with success by many growers. In order to compare it with other sources of potash, five plots were added in 1929 to those described above as the Source-of- Potash plots on Field I. The formula used was as follows : 1765 lbs. cottonseed meal . 741 lbs. castor pomace . . . 260 lbs. nitrate of soda . . . 590 lbs. cotton hull ashes Total r Nitrogen — Nutrients Phosphoric acid per acre- Potash Magnesia 120 40 40 52.9 14.8 35.3 7.4 157.3 12.4 5.9 30.4 200 67.7 200.0 48.7 This supplied the same quantity of the elements as was applied to the adjacent plots. No results were computed in 1929 because the tobacco was destroyed by hail and the experiment was repeated in 1930 on the same plots. The records on- yield and grading given in Table 10 show a lower yield on all cotton hull ash plots as compared with the average of any of the other sources of potash. The grade index on two of the plots was about the same as for the other sources of potash. On the other plots it was lower. The results do not indicate that cotton hull ashes are a more favorable source of potash than the other carriers. Since the form in which the potash occurs here is mostly carbonate of potash, it is probable that any more favorable results that might be ' Analyses by the Department of Analytical Chemistry. Published by courtesy of Dr. E. M. Bailey. 368 Connecticut Experiment Station Bulletin 326 Table 10. Cotton Hull Ashes Compared with Other Sources of Potash. Acre Yield and Grading. Crop of 1930 Source Plot / — Acreyield — ^ , ■ — Percentage of grades ^ Grade index of potash No. Plot Average L M LS SS LD DS F B Plot Average K15-1 1809 11 4 26 6 47 1 4 1 .456 K15 1798 8 8 17 6 45 1 13 2 .400 Cotton hull ash K15-3 1562 1693 2 3 18 7 47 5 14 4 .327 .403 K15-2 1803 12 12 22 3 41 0 9 0 .466 K15-4 1494 6 5 21 5 31 12 14 6 .368 Sulfate 1910 .455 Carbonate 1884 Averages taken from .471 Nitrate 1941 Tables 4 and 6. .464 Stems 1977 .466 expected would come from the other elements in the ash. In view of our results with magnesia-, reported in another section of this bulletin, it seems likely that the quantity of this element in cotton hull ashes would have a favorable effect on the burn. Through the courtesy of one of the large cigar manufacturing corporations, the writers tested the smoking quality of cigars from tobacco raised on a cotton hull ash formula contrasted with an equal number of others raised on the same 12 fields in various parts of the Connecticut Valley, but with other sources of potash (standard commercial fertilizer mixtures). Neither the writers nor the experts from the manufacturing company could detect any consistently favorable influence on the burn, taste or aroma of the cotton hull ash cigars as compared with the others. Final conclusions as to the relative value of cotton hull ashes and other potash carriers must await the continuation of these experi- ments through a series of years. Nothing in the results at present indicates that this material is in any way superior to some of the others. Quantity of Potash In order to determine the optimum quantity of potash to apply in the fertilizer, a series of six one-fortieth acre plots was started on Field V in 1926 (Series of 1926^). In 1927 this experiment was enlarged by addition of 15 plots on Field I (Series of 1927"). These two series are discussed separately below. Series of 1926. These six plots received the same fertilizer treatment in 1930 as in the preceding four years.^ The fact that the crop of 1929 was harrowed into the soil and a cover crop of oats sowed early in August undoubtedly had an effect on the results in 1930. Omission of all special carriers of potash in 1930 did not have as serious effect in reducing yield and quality ' In previous reports designated as "Old Quantitative Series." ^ In previous reports designated as "New Quantitative Series." ^ See Appendix for formulas for plots K9, Kll, and K12. Potash Fertiliser Experiments 369 of the tobacco as might have been anticipated. The supply of available potash had been built up because none of it had been removed in the crop of the previous year. The tobacco plants in 1929 on the "no-potash" plots not only utilized the small amount that was present in the "organics" of the fertilizer ; they also were able to secure an additional supply from the natural but less avail- able soil potash compounds. This combined supply was returned to the soil in a perfectly available form and was next absorbed by the oats crop. This crop not only utilized the available supply from the tobacco, but also added some of the natural soil potash. When the oats crop was plowed under in the spring most of its potash was probably in a form available to the 1930 tobacco crop. Fertilizer was applied to these plots on May 22 and the plants were set on May 27. On June 19 a side dressing of 100 pounds nitrate of lime and 300 pounds cottonseed meal per acre were applied. All the plots except K9 were irrigated on July 19 and July 25. They were harvested July 31 (three rows of each plot) and sorted in December. Differences in growth as observed in the field were not so marked as they were in 1928, probably for reasons previously discussed. Sorting records presented in Table 11 show Table 11. Quantity of Potash. Acre Yield and Grading of Crop of 1930. Series of 1926 Lbs. potash Plot ,^ Acre yield— -, , Percentage of grades ^ Grade index per acre No. Plot Average L M LS SS LD DS F B Plot Average * Kll 1430 i,.Q 3 9 13 9 43 8 12 3 .349 ,,,, None Kll-1 1306 ^^^^ 0 4 12 6 39 14 15 10 .284 '^^^ L M 3 9 0 4 4 5 3 8 9 13 11 10 K12 1541 ,.„ 4 5 22 7 39 4 12 7 M7 ,.^ 100 K12-1 1304 ^^^^ 3 8 14 9 37 9 15 5 .338 -^^^ K9 1470 7^7" 9 13 16 10 32 9 10 1 .419 .^- 200 K9-1 1422 ^^^^ 11 10 20 9 30 7 12 1 .434 "^"^^ only a small reduction in yield due to shortage of potash. This would have been more pronounced if plot K9 had been irrigated as were the other plots. With the shortage of potash the reduc- tion in quality was more pronounced than the reduction in quantity, as shown by comparing the grade index of each. Table 12. Quantity of Potash. Summary of Yields and Grading for Three Years. Series of 1926 Lbs. potash Plot Acre yields by years ^— Index by years — s per acre No. 1927 1928 1930 Average 1927 1928 1930 Average Kll 1150 1107 1430 1^1, .281 .194 .349 ^-. None Kll-1 1137 1163 1306 ^^^^ .298 .233 .284 '^^^ K12 1247 1107 1541 .^.. .368 .321 .367 ,,- 100 K12-1 1141 1076 1304 ^^^^ .364 .324 .338 '^^^ K9 1152 1178 1470 ,,.. .411 .428 .419 ,„ 200 K9-1 1221 1422 ^^^^ .498 .434 "^^^ *No special carriers of potash were applied to any of the Kll plots but the cottonseed meal and castor pomace of the formula furnished 43 pounds K2O per acre. 370 Connecticut Experiment Station Bulletin 326 The sorting records for the three years 1927, 1928, and 1930 presented in. Table 12 show that the effects noted in 1930 are similar to those of the previous years. The notes taken at the time of sorting show as in previous years that without an adequate supply of potash, the tobacco becomes short, dry, yellow, non- elastic and lifeless. Although the leaves on these no-potash plots are much smaller than on the regular ration plots, the weight of tobacco per plot is not much less because the leaves are thick and heavy. Figure 20. Spearing tobacco on the Station farm, 1930. Series of 1927. Treatment on the replicate plots on Field I (Series of 1927) was the same as the Series of 1926 and the same notes as to treatment after the hail storm of 1929 appty here. Fertilizer was applied on May 22 and the plants set on June 2. No significant differences in growth in the field were observed except that the leaves appeared smaller on some of the no-potash plots. No side dressing of fertilizer was applied. The plots were irrigated once during the latter part of Jul3^ All plots were harvested August 5 and 6. The sorting records, presented in Table 13, lead to the same con- clusion as expressed for the other series, namely, that shortage of potash affects the grading more than the yield of tobacco. Potash Fertiliser Experiments 3?1 Table 13. Quantity of potash None 100 lbs. 200 lbs. 300 lbs. 300 Ibs.^ 200 Ibs.^ Quantity of Potash. Yield and Grading Records, Crop of 1930. Series of 1927 on Field I Plot No. K9-6 K9-7 -Acreyield- -Percentage of grades- Plot Average L M LS SS LD DS F B 8 6 24 4 46 2 10 3 4 16 5 49 4 19 Kll-5 1884 '°'^ 2 2 18 5 57 3 12 1 Kll-6 1758 7 4 18 11 45 3 12 Kll-2 1828 Kll-4 1782 1813 K12-2 1864 13 8 19 K12-3 2080 1945 11 3 17 K12-4 1890 5 3 30 K9-5 1932 12 5 25 1890 1886 14 8 22 1836 5 2 29 6 43 2 9 53 2 9 44 1 11 46 1 7 40 1 11 48 1 12 K13 1694 20 6 21 6 36 1 9 1 K13-1 1872 1834 9 4 29 5 45 8 K13-2 1935 17 4 24 4 41 1 9 K13-4 1755 11 7 17 11 42 12 K9-8 1682 14 4 21 7 41 3 10 Grade index Plot Average .436 .327 .345 .374 .452 .417 .427 .411 .459 .465 .403 .500 .446 .484 .425 .450 .442 .477 Effect of the Quantity of Potash Applied in the Fertilizer on the Percentage of Potash, Lime and Magnesia in the Leaf In previous reports it has been stated that with decreasing amounts of potash applied in the fertilizer, not only is the per- centage of potash absorbed by the leaf reduced, but also there is a corresponding increase in the other mineral bases. In order to see how far these changes had gone, samples of darks and seconds of the crop of 1928 from both series were analyzed for bases after the tobacco had been fermented and aged for a year. Results of these analyses are presented in Tables 14 and 15. Table 14. Percentage of Potash, Lime and Magnesia in Cured Leaves, Quantitative Potash Series of 1926. Crop of 1928, the Third on These Plots^ Lbs. potash per acre in Plot 'Potash (K2O) Per cent in the leaf ■ ^Lime (CaO)-^ Mag. . (MgO)' the fertilizer No. Grade Plot Average Plot Average Plot Average Kll D 3.63 8.12 .82 None Kll Kll-1 Kll-1 S D S 3.56 4.09 5.01 4.07 8.71 7.67 8.17 1.07 .96 .90 .94 K12 D 5.74 6.61 .64 100 K12 K12-1 S D 5.79 5.35 5.55 6.83 6.77 6.71 .60 .63 .62 K12-1 S' 5.33 6M .61 K9 D 6.27 5.83 .69 200 K9 K9-1 S D 7.01 6.70 6.69 5.74 5.78 5.87 .61 .74 .69 K9-1 S 6.76 6.11 .70 ^ These tvi^o plots were on a different part of the field from the others and therefore not included in averages. " Figures are on air-dry basis ; water content, about seven per cent. 373 Connecticut Experiment Station Bulletin 326 Table 15. Percentage of Potash, Lime and Magnesia in Cured Leaves. Quantitative Potash Series of 1927, Crop of 1928, THE Second on These Plots Lbs. potash per acre in the fertilizer None 100 200 300 Plot No. Kll-2 Kll-2 K12-2 K12-2 K9-5 K9-5 K13 K13 K13-2 K13-2 K13-3 K13-3 Grade D S D S D S D S' D S D S Potash (K2O) Plot Average t% 5.95 6.89 7.08 7.30 8.64 7.13 7.42 8.54 9.14 8.19 8.83 6.C 7.97 8.21 Per cent in the leaf - /•—Lime (CaO)-^ Mag (MgO) Plot Average Plot Average 7.19 7.93 7.56 .96 .90 .93 6.27 7.19 6.7Z .79 .79 .79 5.97 6.52 62S .82 .77 .80 5.87 .83 6.16 .67 5.57 6.05 6.00 .77 .7Z .74 5.87 .78 6.43 The potash analyses on the older (1926) series for three years are also shovi^n in Table 16. Table 16. Potash Content of Leaves of Three Crops. QuA] STTITATIVI Potash Series of 1926 Lbs. K2O f -Per cent of K2O in leaf ^ per acre Plot f 1926 1927 , ^ — 1928 , in fertilizer No. Grad i Plot Average Plot Average Plot Average Kll D 6.96 5.45 3.63 None Kll Kll-1 Kll-1 S D S 7.03 7.00 4.71 6.00 5.67 5.46 3.56 4.09 5.01 4.07 K12 D 7.20 7.19 5.74 100 K12 K12-1 K12-1 S D S 7.2>i 7.27 7.18 7.06 6.58 7.00 5.79 5.35 5.33 5.55 K9 D 7.97 7.69 6.27 200 K9 K9-1 K9-1 S D S 8.32 8.15 7.97 7.81 7.10 7.64 7.01 6.70 6.76 6.69 From an inspection of these tables we may drav^ the following conclusions : 1. Each increase in the amount of potash in the fertilizer up to 300 pounds KgO per acre increased the percentage of potash in the leaves. The increase was about one per cent for each 100 pounds of fertiHzer potash. 2. For three successive years the percentage of KjO in the leaves declined steadily when all special carriers of potash were omitted from the fertilizer. ^ Figures are on air-dry basis ; water content, about seven per cent. PotasJi Fertiliser Experiments 373 3. A steady decline, but less pronounced, followed when the KjO was reduced to 100 pounds per acre. 4. A still smaller decline in percentage of potash was noticed when 200 pounds K^O per acre were used. Whether this is seasonal may be determined by analyses of future crops. 5. The fact that in the third year of the "no-potash" plots, the percentage of KoO in the seconds was no higher than in the darks might be interpreted as meaning that either : ( 1 ) The lower leaves normally serve as storage places for surplus potash, or (2) when potash is short it is translocated from the older leaves to the new growth. 6. The percentage of CaO has varied inversely with KoO. This agrees with all our previous results. 7. The percentage of MgO was highest where no K^O was added to the fertilizer. The other rates of application, however, do not show any consistent diiTerences in this respect. This is probably due to the fact that 1928 was a very wet year with a num- ber of heavy leaching rains during the growing season, which resulted in a very low supply of MgO in the soil. If there had been an abundant supply of MgO in the soil probably we should have found the same relation to K^O and CaO as in previous years. NITROGEN FERTILIZER EXPERIMENTS Six sets of experiments that deal with the nitrogen ration of the crop are being conducted on the station farm. Progress reports on four of these are presented on the following pages. >)^* • .'* Figure 21. A two-row tractor cultivator at work in the Station tent. Urea as a Source of Nitrogen Progress on the urea experiments has been described in previous bulletins (Reports of the Tobacco Substation for 1925, page 12 ; 1926, page 33; 1927, page 55; 1928, page 190). The plots on Fields 11 and IX were continued in 1930 in the same location and with the same fertilizer ration'^ as in previous years. Fertilizer was applied on May 22 and the plants set on May 31. The tobacco was harvested on July 30. No significant differences in the appearance of the plants were recorded during the summer. Although this set of plots has now been continued through six years on the same soil, at no time has there been any significant difference in their appearance in the field, regardless of the fer- tilizer treatment. The crop of 1929 was destroyed, but if one could judge from field appearance, the results would not have dift'ered materially from those of the other years when the crop was sorted. Another comparison between the standard formula and one in which one-half the nitrogen is furnished in urea was started in 1927 on two small plots on Field II. These are designated in the tables as N 1-7 and N8-2. ^ See Appendix for composition of formulas for plots Nl, N8 and N9. Nitrogen Fertiliser Experiments 375 The yield and grading' records for 1930 presented in Table 17 indicate that a formula in which urea is used to furnish half of the nitrogen produces at least as much tobacco of as good quality as a standard formula in which none of the nitrogen is from urea. Table 17. Urea Plots. Yield and Grading of the Crop of 1930 Amount of urea Plot No. ,^Acre yield— > Plot Average l" Percentage M LS SS of grades LD DS F B Grade Plot A index verage None Nl-S Nl-6 1361 1545 1453 11 IS 6 13 17 20 13 6 Z6 31 4 4 13 11 .416 .478 .447 1/2 N in urea N8 N8-1 1460 1630 1545 14 12 8 1 20 18 8 6 32 41 5 4 13 12 .451 .431 .441 All N in urea N9 N9-1 1505 1577 1541 15 6 7 4 14 15 11 19 35 46 5 8 13 12 .437 .367 .402 None Nl-7 1701 9 9 27 4 Z7 2 12 .445 .445 1/2 N in urea N8-2 1610 13 11 20 6 37 -7 11 .460 .460 Tobacco from the plots where urea was the only source of nitro- gen has not graded out quite as well as the others. As in previous years, so in the 1930 crop, it was found that the leaves of all grades from the all-urea plots were a shade darker and the veins were more prominent and light colored. Since there are now records for five years on the plots of Field IX and for three years on the two small plots on Field II, we may summarize the entire series since its inception in 1925 in Table 18. One may draw the same conclusions from this table as from Table 17, namely, that a formula in which one half the nitrogen is in urea is as good as a standard formula in which there is no urea. Table 18. Urea Plots. Yield and Grading for Five Years Nitrogen Plot r Acre yield by years s, Grade index by years ^ treatment No. 1925 1926 1927 1928 1930 Average 1925 1926 1927 1928 1930 Average No urea Nl-5 1364 1501 1060 994 1361 ^^rn .268 .492 .337 .343 .416 ..^ Nl-6 1561 1711 1296 1105 1545 ^^^^ .411 .473 .411 .471 .478 ■'^^^ l/2Nin N8 1365 1488 1053 1109 1460 ..^^ .325 .545 .354 .430 .451 ..^ urea N8-1 1597 1695 1441 1170 1630 ^^^^ .303 .405 .446 .434 .431 "^^ AUNin N9 1347 1622 1060 1106 1505 ,^q^ .257 .489 .312 .319 .437 ,-. urea N9-1 1465 1810 1223 1206 1577 ^^^^ .352 .445 .428 .354 .367 '^^^ No urea Nl-7 1426 1157 1701 1428 .499 .451 .445 .465 1/2 N in urea N8-2 1386 1202 1610 1399 .445 .473 .460 .459 In this connection, it may be pointed out that the feeding of urea to tobacco is not a new practice in the Connecticut Valley. For two hundred years before the first carload of cottonseed meal for fertilizer was imported from the South, the tobacco farmers of Connecticut grew their tobacco on a fertilizer in which approxi- mately one half of the nitrogen was in urea, namely, stable manure. 376 Connecticut Experiment Station BuIIefin 326 The urea that contains 40 to 60 per cent of the nitrogen in manure is chemically identical with the product now produced synthetically ; the steps by which it breaks down in the soil and enters the growing tobacco plant and promotes growth are just the same, regardless of whether it came from the manure heap or the air nitrogen factory. The most obvious advantage in substituting synthetic urea for cottonseed meal is the reduction in the fertilizer bill. The present cost of nitrogen in urea is less than half that of nitrogen in cotton- seed meal. Some recent results obtained by the Soils Department (see page 435 of this Report) indicate that there is another possible advantage in using urea. After a leaching rain when all the available forms of nitrogen (nitrates mostly) have been washed away and the plant is in danger of being checked in its growth, urea seems to possess the ability of changing to an available form more quickly than cottonseed meal or other organics. In another set of experiments where single sources of nitrogen are being com- pared, it has been observed during every wet year that tobacco on the urea plots suffers but little from leaching, less so in fact than the cottonseed meal plot. Calurea as a Source of Nitrogen Calurea, a combination of urea and calcium nitrate in which four-fifths of the nitrogen is in the form of urea and one-fifth in calcium nitrate, has been found to be a cheap and favorable source of nitrogen in growing some crops. In order to determine its value in a tobacco fertilizer mixture, a series of 12 one-fortieth acre plots was started in 1928 on uniform soil on Field I. Three mixtures quite similar to the urea mixtures previously described were compared : ( 1 ) a standard mixture in which there was no Calurea, (2) one in which one-half of the nitrogen was in Calurea, and (3) one in which all the nitrogen was in Calurea.^ All treat- ments were in quadruplicate. Brief mention of this experiment was made in the Report for 1928. The crop was destroyed in 1929 but during the three years of the experiment, no differences in growth as between the three different treatments could be observed in the field. In 1930, the fertilizer was applied on May 22 and the plants set on June 2. All plots were harvested (the center row only kept for records) on August 5 and 6. The sorting records are shown in Table 19 and the results for two years are summarized in Table 20. The differences in yield for the three treatments are too small to be considered significant. The grade index is somewhat lower where Calurea is the only source of nitrogen. Altogether, the results parallel very closely those from the urea tests. ' For complete composition, see Appendix, formulas N25, N26, and N27. Nitrogen Fertilizer Experiments 377 Table 19. Calurea Series. Yield and Grading Records. Crop of 1930 Calurea used Plot No. ,— Acre yield- Plot Averag( -N , Percentage o 3 L M LS SS f grades— LD DS "1^ Grade index Plot Average None (standard N25 formula) N25-1 N25-2 N25-3 1959 1745 1675 1798 1794 8 3 19 9 15 12 19 11 25 4 22 7 18 6 19 7 50 2 32 3 35 1 35 1 8 8 13 8 .422 .507 .444 .506 .470 1/2 Calurea N26 N26-1 N26-2 N26-3 2002 1728 1720 1745 1799 14 9 17 10 12 8 12 11 19 4 20 6 21 5 19 6 42 1 2,7 2 42 1 42 1 11 8 11 9 .459 .491 .448 .455 .463 All Calurea N27 N27-1 N27-2 N27-3 1932 1760 1651 1861 1801 14 8 17 8 8 5 15 9 17 5 16 9 19 7 16 9 46 1 42 1 45 4 41 2 9 7 12 8 .454 .476 .400 .462 .448 Table 20. Calurea Series. Summary of T' wo Years Results Plol L No. t Acre yield 1930 -Gn A -X A Amount of Calurea 1928 Average 1928 iQC Inaex ^ 1930 Average None N25 N25-1 N25-2 N2S-3 1300 1140 1174 1273 1959 1745 1675 1798 1508 .387 .404 .419 .400 .422 .507 .444 .506 .436 244 lbs. to furn- ish 1/2 of the nitrogen N26 N26-1 N26-2 N26-3 1128 1160 1257 1092 2002 1728 1720 1745 1479 .398 .395 .394 .413 .459 .491 .448 .455 .432 488 lbs. to furn- ish all the nitrogen . . . N27 N27-1 N27-2 N27-3 1223 1201 1113 1298 1932 1760 1651 1861 1505 .386 .387 .400 .425 .454 .476 .400 .462 .424 Nitrophoska Tests Nitrophoska (No. 3) is a concentrated commercial mixture con- taining 16.3 per cent nitrogen, 16.3 per cent phosphorus and 20 per cent potash. It contains no chlorine, the potash being in the form of sulfate. Obviously if such a fertilizer were found to be suitable for growing tobacco it would be less expensive than our usual mixtures, both on account of its original cost and because it is less bulky and would thus involve less labor in handling. An experiment was begun in 1929 to determine its suitability, (1) as the principal source of the elements included and, (2) as the source of half of the elements. These treatments are compared with the standard formula on six one-fortieth acre plots on a part of Field I, which has always produced the best tobacco on the farm. These plots are in duplicate and the soil is uniform. The three formulas used are given in the appendix as N28, N29 and N30. The object of using a small quantity of urea in the N30 (all Nitrophoska) formula was to make the total quantity of nitrogen applied equal to the potash and thus correspond with 378 Connecticut Experiment Station Bulletin 326 our standard formula and customary practice. Carbonate of mag- nesia was also added to all three to prevent sand drown and to give a better burn. During 1929, observations in the field showed no difiference between growth on the standard formula and the half-Nitrophoska formula. On the all-Nitrophoska formula, how- ever, growth was not quite so large nor so uniform as on the others Figure 22. Arrangement of the Thumpsun aitaclimciits on the tractor for 'hilling" shade tobacco. at the time of the hailstorm that destroyed the crop on August 1. The same treatments, in duplicate, on the same plots were con- tinued in 1930. Fertilizer was spread on May 22 and the plants set on June 2. During the summer the growth was good and uni- form on all plots, no differences being apparent between the vari- ous treatments. All plots were harvested on August 5 (three rows of each plot, or about 150 plants, being kept for records). Table 21. Nitrophoska Series. Yield A.ND Grading Records. Crop of 1930 Plot Nitrophoska used No. ,^Acre yield— ^ ^— Plot Average L — Percentage of grades — M LS SS LD DS F — ^ Grade index B Plot Average None N28 (standard formula) N28-1 1884 .... IS 1829 ^^^^ 13 8 27 7 33 1 8 6 26 4 39 1 9 •491 477 2 .464 -^^^ 1/2 Nitrophoska N29 N29-1 1810 ,07^ 12 1934 ^^^^ 11 3 27 5 44 1 6 6 27 6 38 1 9 2 .457 „ 2 .453 -4^^ All Nitrophoska N30 N30-1 1915 ,.Qc 9 1875 ^^^^ 13 6 27 3 41 1 11 9 26 3 47 1 10 2 .435 ,^, 1 .473 -^^^ Sorting records shown in Table 21 indicate no significant dift'er- ences either in yield or grading. Neither do the observations taken Nitrogen Fertiliser Experiments 379 at time of sorting' show any real variations. All the tobacco on this field in 1930 had considerable white vein, but this apparently bore no relation to the fertilizer. It is planned to continue this experiment through a series of years and possibly enlarg'e it by the addition of further replica- tions. The only conclusion that may be drawn up to the present is that no falling off either in yield or quality occurred when Nitro- phoska was used for two years on naturally strong land. Comparison of Nitrate of Soda and Nitrate of Lime as a Mineral Source of Nitrogen The object of this experiment was to compare the merits of these two quickly available nitrogen compounds when used to supply a portion of the nitrogen in the fertilizer mixture. Neither of them is suitable as the only source of nitrogen in a mixture because of too rapid leaching. A preliminary statement of this test, which was started in 1927, was published in our Report for 1928, page 191. There are four different treatments,^ in which (1) they are compared when each furnishes one-fifth of the nitro- gen and (2) when each furnishes one-half of the nitrogen of the formula. Each treatment is in duplicate. The eight plots are located on Field III, where the soil is moderately uniform and crops are usually of only average size. During the wet years, 1927 and 1928, tobacco on the plots where half of the nitrogen was in nitrate form turned noticeably more yellow than on the other plots, thus indicating that there had been considerable leaching away of the nitrogen. However, significant differences between the two kinds of nitrate did not appear, as far as could be judged from the tobacco in the field. In 1930, the fertilizer was spread on May 22 and the plants set on June 2. No marked difi^erences in appearance were recorded for 1930. The tobacco was harvested on August 7. The yield and grading records for the 1930 crop are presented in Table 22 and a summary of three years in Table 23. Table 22. Comparison of Nitrate of Soda and Nitrate of Lime. Yield and Grading Records. Crop of 1930 Plot Carrier of nitrogen No. 1/5 N in Nitrate of soda Nl-8 (standard formula) Nl-9 ^— Acre yield- Plot Average I719 1788 17 9 —Percentage c M LS SS 13 17 6 10 18 5 fgra LD 35 42 des- DS 1 4 11 12 Grade Plot i» .486 .419 inde.x I verage .453 1/5 N in Nitrate of lime N18 N18-1 1795 1825 1810 20 9 14 10 13 16 5 9 39 43 4 9 9 .503 .419 .461 1/2 N in Nitrate of soda N2-3 N2-4 1403 1685 1544 7 17 9 10 22 17 5 5 35 29 7 6 15 16 .405 .462 .434 p . 1/2 N in Nitrate ; of lime N16 N16-1 1839 1674 1757 17 7 9 10 15 18 4 7 44 34 9 11 15 .469 .394 .432 'For composition of formulas see Nl, N2, N16, and N18 in the Appendix. 380 Connecticut Experiment Station Bulletin 326 Table 23. Nitrate of Lime Series. Summary of Yield and Grade Index Records for Three Years Source Plot r Acre yield ^ , Grade index- of nitrogen No. 1927 1928 1930 Average 1927 1928 1930 Average 1/5 in Nitrate Nl-8 1232 1162 1858 .^q. .386 .452 .486 ... of soda Nl-9 1239 1097 1719 ^^^^ .454 .463 .419 •^'^•^ 1/5 in Nitrate N18 1246 1153 1795 ..^.y .385 .446 .503 ... of lime N18-1 1265 1153 1814 ^^"^ .387 .446 .419 '^"^^ 1/2 in Nitrate N2-3 1145 1098 1403 .^.r .352 .422 .405 .n, of soda N2-4 1185 1080 1685 ^^^^ .344 .421 .462 '^^^ 1/2 in Nitrate N16 1348 1076 1839 .^.q .457 .408 .469 ,.r of lime N16-1 1128 1088 1674 '-^^^ .386 .376 .394 '^^^ These records show that there has been a marked reduction both in yield and grading where the larger quantities of nitrate of either kind were used. Such reduction has been greater where nitrate of soda was used than where nitrate of lime was applied. Burn Tests In order to see whether the kind or quantity of nitrate had any effect on the burn, both strip burn tests and cigar tests were made on the crops of 1927 and of 1928. One hundred and sixty strip burn tests on each of the four grades and for each year and each treatment failed to show any differences in the fire-holding capac- ity. All had an average fire-holding capacity of 54 to 60 seconds. No tobacco raised on the farm had a better burn. Some of the cigars had wrapper, binder and filler from the same plot ("clears"), while others had only the wrapper and binder. The fire-holding capacity of all was satisfactory, six to ten minutes. The color of the ash was a medium to dark gray, not light enough to be considered entirely satisfactory, but the different fertilizer treatments had no effect on this character. The coal band was narrow to medium, or even wide, and the coherence and evenness good. The taste and aroma were only fair. Therefore, from the standpoint of combustion characteristics, we may conclude, from the crops of 1927 and 1928, that there was no difference between these two kinds of nitrates or the quantities used. FRACTIONAL APPLICATION OF FERTILIZER On three one-twentieth acre plots on Merrimac coarse loamy sand on Field VII, the fertilizer was so divided that only a small part of the quickly available nitrogen was applied broadcast before setting, the remainder being divided and applied at the side of the growing plants on June 9 and June 27. On the adjacent check plots the same mixture in the same amount was broadcast on the field on May 23. The composition of the fertilizer for each application was as follows : Lbs. pel June 9 acre June 27 Total 400 400 1000 1400 100 100 200 100 50 60 110 90 117 400 Broadcast Carrier before setting Cottonseed meal 200 Castor pomace 1400 Nitrate of lime Calurea 100 Nitrate of potash Carbonate of potash 90 Sulfate of potash 117 Magnesia lime 400 This series of plots was started in 1926 and previous progress has been published in the Reports for 1927, page 57, and 1928, page 193. During 1926 and 1927 the results showed no advantage in fractional application. In the wet year of 1928, however, the benefit was quite apparent. (For details see Report for 1928.) In 1930 the first side dressing was applied during the heavy leach- ing rain of June 8 to 10 and again there was a distinct gain, both in yield and grading, as may be seen by reference to Table 24. That the poor grading of the check plots was due to leaching of the available nitrogen may be inferred from the lighter color of the leaves on these plots in the field and the dead, yellow condition of the seconds on the sorting bench. The sorting records of these plots for four years are presented in Table 24a. Table 24. Fractional Application Plots. Yleld and Grading as Com- pared WITH Adjacent Plots where Fertilizer was in One Application. Crop of 1930 Method of fertilizer Plot ^-Acre yield — ^ ,. Percentage of grades ^ Grade index application No. Plot Average L M LS SS LD DS F B Plot Average FS-1 1591 14 10 14 7 40 2 13 .442 Fractional F6 1235 1351 2 4 12 13 Z2 16 21 .304 .?>6?, F6-1 1228 6 9 10 9 29 19 18 .344 C3 1120 7 8 23 20 24 18 .217 Single C3-1 1260 hq. 1 2 16 3 31 9 21 16 .281 ^.q application C5 1157 ^^^^ 15 7 31 14 20 13 .265 '^^^ C5-1 1243 2 2 13 4 23 20 23 13 .267 382 Conncciicut Experiment Station ■ Bulletin 326 Table 24a. Fractional Application Plots. Yield and Grading Records FOR Four Years Method of Plot , Acre yield by years application No. 1926 1927 1928 1930 Average 1927 F5-1 1603 1413 1272 1591 Fractional F6 1300 1148 1092 1235 1322 F6-1 1550 1211 1225 1228 C3 1279 1062 880 1120 Single C3-1 1618 1343 1185 1260 application C5 1426 1221 805 1157 C5-1 1612 1291 1145 1243 1228 ^ — Grade index ^ 1927 1928 1930 Average .494 .446 .442 .375 .392 .304 .409 .433 .455 .344 .344 .280 .217 .435 .464 .281 .339 .419 .210 .265 .481 .399 .267 As a result of all fractional application experiments up to date on the station farm the following conclusions may be drawn : 1. During' dry years, that is, years without leaching rains that follow the first application of fertilizer, a single broadcast appli- cation is more profitable both in yield and grading than fractional application of the same fertilizer. 2. In years when heavy leaching rains occur in the grow- ing season, both yield and quality may be improved by reserving some of the quickly available nitrogen carriers for later applications. 3. Coarse sandy soils with porous subsoils subject to leaching, may be expected to show greater benefit from such later applications. 4. Timing such later applications is particularly important. Reference should be made not to the calendar nor to the time of hoeing, but to the time and character of the rains. After a heaw, leaching rain, the application should be made invmediately, without waiting to see whether the leaves will turn pale. From these conclusions, it is apparent that the better method of applying the fertilizer depends on whether it is a wet or a dry year. But it is obviously not possible to know beforehand what kind of a season is ahead. In order to be safe and to profit by either method, some growers follow this rule : Before setting, apply enough fertilizer to take care of all the nutrient requirements of the crop. Then, if leaching rains occur during the growing season, apply additional nitrogen at once. Obviously this increases the cost of the fertilizer throughout a series of years but probably the increase in yield and quality more than compensates for the extra cost. The heavy rains of June, 1930, presented an opportunity for testing this rule. As previously stated, 100 pounds of nitrate of lime and 300 pounds of cottonseed meal per acre were applied to Field V, but not to Field VII. These two fields are adjacent, and on the same type of soil and on the check plots they receive the same fertilizer application. The average acre yield of the four Fractional Application of Fertilizer 383 check plots on Field VII for the years 1926, 1927 and 1928 was 1,239 pounds, as compared with 1,140 pounds for the two check plots on Field V for the same years, a difference of 99 pounds per acre in favor of Field VII. In 1930, however, the average for Field V was 1,509, as compared with 1,195 for Field VII. A similar increase in grading is apparent by reference to Table 28. Such an increase more than pays for the additional cost of fertilizer. MANURE AS A SUPPLEMENT TO COMMERCIAL FERTILIZER The purpose of this experiment was to see what effect the addi- tion of manure to the regular commercial fertilizer ration has on the yield and quality of tobacco. The progress of the experiment, which was started in the autumn of 1925 on eight plots of Merri- mac loamy coarse sand, Field VII, and has now been continued for five years, was discussed in Reports^ for the years 1927, 1928 and 1929. One plot was treated annually with 20 loads (tons) per acre of mixed stable manure, another with 40 loads per acre, two others with about 30 loads of artificial (Adco) manure. Adjacent to each plot was a check plot of the same size and treated the same way throughout, except that it received no manure. In the dry year of 1926 the differences in yield and grading between the manure and no-manure plots were so small and incon- sistent it was apparent that the manure had been of little if any value. In the wet years of 1927 and 1928, tobacco on the manure plots seemed to suffer less from leaching and had a better appear- ance in the field. Both the yield and grade index were somewhat better on the manured plots in these wet years. See Table 26. During the dry year of 1929, however, the manured plots were extremely poor (as judged in the field, the crop being destroyed by hail) in contrast to the check plots. See Report for 1929. Crop of 1930 In preparation for the crop of 1930 the manure was harrowed into the soil the previous fall. Commercial fertilizer was applied on May 22 and the plants set on May 27. After the heavy rains of June 8 to 10 a side dressing of ^100 pounds nitrate of lime and 300 pounds of cottonseed meal per acre were applied to Field V on which the Adco manure and adjacent check plots were located. None, however, was applied to the other plots. All plots were irrigated twice during the latter part of July. All were harvested on July 31 and August 1. Early in the growing season it was apparent that the manure plots were not doing so well as the check plots. This difference became more pronounced as the season advanced. This part of the farm suffered severely from the dry and excessively hot weather of the latter half of July. The growth was not normal on the check plots, but on the manure plots the plants were so stunted that from a commercial standpoint they would not have been worth harvesting. They wilted so much that the lower leaves were badly burned on the hot sand. 'Tobacco Substation Bull. 10: 62-66. 1928. Conn. Agr. Exp. Station Bulletins 299 : 192. 1929, and 311 : 216-219. 1930. Mamire as a Supplement to Commercial Fertiliser 385 Table 25. Manure Plots. Yield and Grading for Crop of 1930 Plot ^-Acre yield-^ ^ Percentage of grades v Grade index Kind of manure No. Plot Average L M LS SS LD DS F B Plot Average Stable 20 loads Ml 989 q.. 16 54 30 .159 ^gy Stable 40 loads Ml-1 978 ^^^ 6 18 17 35 24 .235 -'^^ None C3 1120 7 8 23 20 24 18 .217 None H-H 1103 1210 14 17 34 25 10 .196 .231 None C3-1 1407 1 3 16 3 31 9 21 16 .281 '< Adco manure M2 1383 ...^ 7 4 11 11 35 11 21 .430 .37 M2-1 1236 ^^^" 7 8 8 12 31 16 18 .345 '^^^ None C14 1516 ...^ 15 16 12 8 29 7 13 .456 ... C14-1 1519 ^^^^ 11 12 18 8 32 4 13 2 .433 "^^^ The sorting records presented in Table 25 show a reduction in all cases of more than 200 pounds where manure was used. Also a corresponding- falling off in grading. The larger yields on the Adco plots and" their adjacent check plots are accounted for by the extra application of fertiHzer that was made to these plots after the heavy rains of June. Table 26. Manure Plots. Yields and Gradings for Four Years Plot f Acre yield ^ , Grade index , Kind of manure No. 1926- 1927 1928* 1930 Average 1926 1927 1928* 1930 Average Stable 40 loads Ml 1404 1375 1069 989 .^^7 .338 .480 .319 .159 ^.. Stable 20 loads MM 1489 1402 1106 978 ^^^^ .396 .391 .435 .235 -"^^^ None C3 1279 1062 880 1120 .^^7 .295 .344 .280 .217 ^^ C3-1 1618 1343 1185 1407 '■^^^ .379 .435 .464 .281 ""^^^ Adco 30 loads M2 1379 1259 1129 1383 .^.. .328 .442 .457 .430 ^.^ M2-1 1281 1300 980 1236 '■^^'^ .298 .482 .277 .345 -^'^^ None C14 1285 1161 912 1516 .^^^ .305 .350 .287 .456 ,„ C14-1 1217 979 1519 ^^"^ .274 .368 .351 .433 ^^^^ The yield and grading results of these plots for four years are presented in Table 26. The averages for the four years do not show any very striking differences, but averages do not tell the whole story. In the second and third years of the tests, manure undoubtedly improved both yield and grading. By the fourth year, 1929, however, the results were reversed and the manure plots were definitely inferior in appearance, and undoubtedly would have shown lower yields and grading if the hailstorm had not prevented taking records. In the fifth year the deterioration had progressed to such a point that the tobacco was worthless. Black Rootrot on the Manure Plots The field on which these tests are located has never been seri- ously affected by black rootrot, Thielavia basicola. In fact, rootrot is not commonly a factor on such coarse sandy soils. It has always * Crop of 1929 destroyed by hail on August 1. 386 Connecticut Experiment Station Bulletin 326 been possible, however, to find occasional lesions of this disease on the roots. This is true of practically all old tobacco fields of this state, but it has been observed in the last two years that the tobacco on the manure plots wilted more quickly during hot days than the tobacco on the other parts of the field. This, coupled with its dark green color, lead us to suspect the presence of black rootrot. After the removal of the 1930 crop a wheelbarrow load of roots from each of the manure plots and from the adjacent check plots was dug and washed. Completely rotted roots and lesions in all stages of development were present in abundance on root systems from the manure plots, while only occasional lesions were found on the check plots. The use of manure in such quantities on this land has thus greatly increased the prevalence of black rootrot and this is the most probable explanation of the decreasing yields. Bearing on this point also is an unpublished observation made during a series of experiments on brown rootrot at Poquonock from 1925 to 1928. Two manured plots in that experiment showed more black rootrot than any other plots on the field. Data to show how manure decreases acidity of the soil were published in our Report for 1929, page 217. Decreased acidity favors rootrot. It is also probable that increase in the humus con- tent of the soil would favor growth of the rootrot fungus. Increase in Soil Organic Matter from Use of Manure In order to see to what degree the application of manure increases the percentage of organic matter in this soil, analyses were made on two sets of samples, the first taken in the fall of 1927 and the second at about the same time in 1930. The results of these analyses^ presented in Table 27 show that there has been an average increase of about 50 per cent in the organic matter content as a result of adding manure for the five years. Average for 1930 Table 27. Percentage OF Organic Matter IN Soil of Mai Kind of manure Plot No. Percentage of organic matter 1927 1930 Difference None C3 C3-1 C14 C14-1 1.350 1.557 1.491 1.517 1.333 1.560 1.438 1.517 .023 decrease .003 increase .053 decrease no change Stable Ml Ml-1 1.672 1.815 2.057 2.477 .385 increase .662 increase Adco M2 M2-1 1.655 1.762 2.408 1.976 .753 increase .214 increase 1.462 2.229 ^ Made by Mr. Morgan and Dr. Lunt of the Soils Department according to the Parr-bomb method of determining carbon and calculating the organic matter by multiplying by the factor 1.724. Manure as a Supplement to Commercial Fertiliser 387 Increase in Water-holding Capacity o£ Soil Since the use of manure has increased the organic matter content of this soil it would be reasonable to expect that the capacity of the soil for retaining moisture would also be increased. Theoretically, such an increase should be beneficial to the tobacco crop by reduc- ing the ill effects of leaching on a wet year and of drought on a dry year. Determination of the water-holding capacity of the four manure plots and the four adjacent checks was made by the Soils Department in the fall of 1930. The average water-holding capacity of the four check plots was 27.94 per cent, as contrasted with 30.63 per cent for the manure plots, thus showing a gain of 2.69 per cent in the five years. The fact that the plants on these manure plots do not show any resistance to wilting, but rather are less resistant than those on the check plots is thus apparently due to reduction of the absorbing root surface through black rootrot. Effect of Manure on the Burn In strip burn tests on single leaves (see Report for 1929, page 218) it was found that leaves from the manure plots possessed as good fire-holding capacity as those from the check plots. More recently, cigars were made from this tobacco of the crop of 1928. On these cigars it was found that the burn was usually somewhat superior on the manure plots, in that the ash was lighter colored, and the coal band narrower. Also in most cases, the taste and aroma were judged to be better. These differences may be due to the increase in magnesia from the manure ; at least the effect is the same as was obtained in other experiments when the magnesia con- tent of the leaves was increased. COVER CROP EXPERIMENTS After the hailstorm of August 1, 1929, the tobacco was harrowed in and the cover crops sowed before August 10, this being one to two weeks earlier than the usual date for sowing them. Plenty of moisture was present and the fall weather was favorable for growth. Both the oats and the barley matured and much of it was headed out before freezing down in December. The vetch did not winter over very well. Most of it died before spring, but all others made an unusually large growth. The crops were plowed under early in May and the following fertilizer mixture (per acre) applied on May 23. 1,000 lbs. cottonseed meal 1,400 lbs. castor pomace 200 lbs. nitrate of lime 100 lbs. calurea 110 lbs. nitrate of potash 90 lbs. carbonate of potash 117 lbs. sulfate of potash 40O lbs. magnesian lime (29 per cent MgO) This mixture furnishes 200 pounds each of nitrogen and potash and 58 pounds of phosphoric acid. The plants were set on May 27. After the heavy rains in June a side dressmg of 100 pounds nitrate of lime and 300 pounds cottonseed meal per acre were applied to the plots on Field V, but not to those on Field VII. The tobacco on both fields was irrigated twice during the last ten days of July. This should have been started a week earlier for best results. All plots were harvested on July 31 and August 1. Observations during the growing season indicated that on Field VII the cover crop plots were doing better than the check plots. This difference was not apparent on Field V, probably due to the second application of fertilizer. The oats plot on the south end (C9) did not do so well, probably because the lysimeter had been built on part of it and the remainder was used as a roadway for the tractor and other machinery. The rye and timothy plots looked best on this field. The sorting records presented in Table 28 show a distinct benefit from the cover crops on Field VII, but none on Field V ; in fact, the check plots are slightly better on the latter. This difference in results on the two fields may be explained on the assumption that the benefit derived from a cover crop is due to conservation of nitrogen or other elements that might otherwise leach away to a smaller degree. On Field V, the second application of fertilizer after the only leaching rain of the summer was sufficient to prevent any shortage of nutrients on any of the plots. Growth during the later part of the season was limited by the low supply of moisture rather than the supply of nutrients on this field. Cover Crop Experiments 389 Table 28. Cover Crop Plots. Yield and Grading of Crop of 1930 Field VII Cover crop None Timothy Barley Rye Oats Vetch Plot No. C3 C3-1 CS C5-1 C5mk C6 C6-1 C7 C7-1 C8 C8-1 C9 C9-1 ClO-1 , — Acre yield — ^ Plot Average 1120 1260 1157 1243 1195 1326 1425 1243 1304 1276 1313 1679 1139 1471 1545 1326 1334 1290 1496 1305 1545 Percentage of grades M LS SS LD DS F 7 16 15 13 6 15 8 15 4 19 7 11 8 23 20 24 3 31 9 21 1 31 14 20 4 23 20 23 — ^ Grade index B Plot Average 18 .217 16 .281 13 .265 13 .267 4 14 4 34 14 18 10 .298 8 31 12 16 3 Z6 10 18 5 41 8 2,6 8 17 11 32 6 17 7 40 17 2 16 17 20 8 26 16 24 10 20 6 32 8 15 .349 .339 .340 .323 .390 .378 .284 .415 11 17 5 35 7 15 1 .408 .258 .298 .344 .332 .384 .350 .408 None Alfalfa Redtop Wheat C14 1516 C14-1 1501 C12 1519 C12-1 1466 C13 1524 C13-1 1367 C15 1427 C15-1 1525 1509 1492 1445 1476 Field V 15 16 12 11 12 18 14 11 14 10 12 12 13 14 15 11 15 9 11 14 11 11 14 17 8 29 8 32 9 ZZ 5 2>7 6 32 8 34 6 36 5 30 7 13 4 13 5 14 8 16 5 15 6 17 5 16 8 14 .456 2 .433 .440 .402 .443 .409 1 .413 1 .432 .445 .421 .426 .423 The low yield and grade index of plot C9 for 1930 are a reverse of the results on this plot in previous years and should be dis- regarded, since proper cultural conditions were disturbed by the installation of the lysimeter in 1929. An examination of Table 29 shows that in previous years this plot has been fully equal to C9-1 both in yield and in grading. Assuming the same yield in 1930 for C9 as for C9-1, the average for the four years for the oats plot is 1,418, about the same as for rye and vetch (Table 29). During the dry year of 1929 the rye plots were about the poorest on the field. Rye seems to be an excellent cover crop for a wet year, or a year at least when there are rains soon enough after the crop is turned under to rot it, as in 1930. During very dry years like 1929, however, decay does not occur at the proper time to help the growing tobacco crop. Oats seem more reliable throughout a series of years. Vetch also gave very good results, but it was observed in the sorting that the tobacco was a shade darker than the other plots. The same fault has been observed in previous years. 390 Connecticut Experiment Station Bulletin 326 Table 29. Cover Crop Plots. Yield and Grading Records for Four Years Field VII Cover crop Plot ,■ Acre yield ^Treatment, Grade Index .^Treatmc: No. 1926 1927 1928 1930 Average average 1927 1928 1930 Average averag. C3 1279 1062 880 1120 1085 .344 .280 .217 .280 None C3-1 1618 1343 1185 1260 1352 .^^^ .435 .464 .281 .393 „a C5 1426 1221 805 1157 1152 ^^^" .419 .210 .265 .298 -^^^ C5-1 1612 1291 1145 1243 1323 .481 .399 .267 .382 Muck C5mk 1073 1326 1200 .348 .298 .323 Timothy C6 1373 1203 1129 1425 1283 ..n. .450 .371 .349 .390 ,q, C6-1 1666 1278 1130 1243 1329 ^^^ .443 .342 .339 .375 -^^-^ Barley C7 1430 1296 1177 1304 1302 ,,^i .472 .447 .340 .419 ,^, C7-1 1507 1357 1057 1276 1299 ^^"^ .436 .396 .323 .385 ""^^^ Rye C8 1455 1387 1276 1313 1358 ..^ .480 .480 .390 .450 ^. C8-1 1717 1404 1225 1679 1506 ^^"^^ .488 .445 .378 .437 '^^ Oats C9 1687 1356 1310 1139 1371 . „-, .461 .521 .284 .491 ... C9-1 1621 1371 1070 1471 1383 ^"^^^ .478 .395 .415 .422 '^^^ Vetch CIO 1642 1430 1172 1415 ...q .392 .474 .433 .,q ClO-1 1479 1399 1258 1545 1420 ^^^^ .434 .554 .408 .465 '^^ 1224 ■ii'i 1.1 •;,, io\ .377 1241 Field V None C14 1285 1161 912 1516 1219 C14-1 1217 979 1501 1232 Alfalfa C12 1243 1198 1061 1519 1255 C12-1 1191 1238 1015 1466 1227 Redtop C13 1104 1212 1115 1524 1239 .^^y C13-1 1174 1240 1080 1367 1215 ^^-^ Wheat C15 1364 1261 951 1427 1251 .^-.. .377 .332 .413 .374 C15-1 1137 1210 970 1525 1211 ^^-^^ .448 .367 .432 .416 .350 .368 .287 .351 .456 .433 .371 .384 .435 .430 .344 .325 .440 .402 .406 .386 .399 .455* .447 .344 .443 .409 .430 .403 THE RELATION OF MAGNESIA TO THE BURNING QUALITIES OF CIGAR LEAF TOBACCO In previous publications of this Station concerning the effects of treating tobacco soils with lime or limestone (containing both cal- cium and magnesium), the following conclusions were drawn: 1. Liming reduces the fire-holding capacity of the leaves if this capacity is measured by the strip test (single leaf), but, 2. When the same leaves are rolled into a cigar, the fire-holding capacity of the cigar is not only as good, but usually somewhat better, than that of a cigar made from unlimed tobacco. 3. Liming the land improves the color of the ash, making it light gray or white in contrast to the dark gray to black ash of the unlimed cigar. 4. Too much lime, however, makes the ash "flaky" (therefore objectionable). 5. The burn is much closer (that is, the coal band very narrow) on the cigar made from limed tobacco. 6. The taste and aroma are improved by liming. 7. Chemical analyses of the leaves show that liming greatly increases the magnesia, but reduces the calcium and potassium con- tent of the tobacco. Since the decrease in calcium and potassium would not be ■expected to produce such changes in the character of combustion, and since such decrease is merely a result of the greater increase in magnesium, we are warranted in concluding that the improved combustion is due to absorption of larger amounts of magnesium by the plant. A more extensive test of the comparative burning qualities of imed and unlimed tobacco was conducted on the leaves from three plots of the 1928 crop. The three plots were fertilized as follows : Fertilizer material Cottonseed meal . . . 3astor pomace titrate of soda . . . . sulfate of ammonia Jrea Precipitated bone . . sulfate of potash . . Carbonate of potash PlotTl. — Lbs. per acre — Plot T2. Plot T3. Acid Alkaline General formula formula formula 1100 1100 1463 588 585 106 440 36 333 333 278 367 172 282 132 Each of these formulas furnished 200 pounds ammonia, 200 -rounds potash and 160 pounds phosphoric acid to an acre. One-half of each of the plots was limed each year from 1924 to ;927, inclusive, at the rate of one ton hydrated lime per acre. They [-i^ere not limed in 1928. Although magnesian lime was used, the 392 Connecticut Experiment Station Bulletin 326 actual percentage of MgO was not recorded. In the tables below the limed end of the plot is designated as a and the unlimed as b. After the tobacco had been fermented and aged for a year, the fire-holding capacity was first tested by the strip test on individual leaves. Then some of the leaves were made into cigars and the combustion tested. The ash from the cigars was then analyzed for magnesium, calcium and potassium. The Strip Test In making this test, the leaf is stretched between the hands and ignited by an incandescent electric filament. The number of seconds during which the fire, or glow, will continue to spread from the edge of the burned hole, after removal from the filament, is recorded as the fire-holding capacity. Twenty tests are made on each hand of tobacco and since four grades of leaves are included in each plot, the fire-holding capacity of any lot of tobacco is the average of 80 tests. Table 30 shows the average fire-hold- ing capacity of each of the plots for the crop of 1928. Since these same tests were conducted on the three preceding crops, the average fire-holding capacity for the four years is also presented. Table 30. Strip Burn Tests on Lime Plots Fire-holding capacity in seconds Lime , Average fire-holding capacity ^ Character of fertilizer treatment ^— Crop of 1928— ^ Cropsof 1925 to 1928 Acid(Tl) Lime No lime Alkaline (T2) Lime No lime Neutral (T3) Lime No lime Plot Treatment Plot Treatment 24 43 38 21 2,6 29 16 31 48 52 2,7 55 49 38 55 47 29 47 The data presented in Table 30 show that in every comparison, fire-holding capacity has been reduced by liming. This has been apparent during each of the four years of the test. The poor fire- holding capacity of the Tl plots as compared with the others is probably due to the large amount of sulfate of ammonia used in this fertilizer. The Cigar Test After sweating and re-sweating the tobacco from the above plots, leaves from the same hands as tested above were used to make clear cigars, that is, filler, binder and wrapper from the same hand. After ageing for a few weeks, these cigars were smoked and records were taken on length of time they would hold fire, color and coherence of ash, closeness of burn, taste and aroma. The results are recorded in Table 31. Magnesia and Burn of Leaves 893 ^3 O t-l o o -a o •s O o o s o o O XI ^ x: X 1 > 1-^ S:; Q u o > .ti bo ^0^-$ Q ^x ^^ be o S ^ bo 03 OJ t^X ^^ bf) o CJ jiR ;r^ r^ ^ o\ (U ^ ,;^^ Barium fluosilicate Full strength Diluted PViPf and calcium arsenate Full strength Diluted ] l^iicc Full strength Diluted June 21 1 7 3 6 39 June 23 5 8 8 15 "5 7 34 June 24 3 4 8 4 2 2 33 June 25 1 0 6 0 1 9 23 June 30 4 5 1 1 6 6 15 July 7 14 8 6 6 5 6 8 July 12 85 29 6 Shade grown 24 tobacco 2 13 27 June 21 1 3 0 0 9 June 23 3 3 2 1 '3 '2 12 June 24 0 0 1 0 0 0 9 Tune 25 0 0 1 1 1 1 8 June 30 0 0 1 0 0 0 4 July 7 1 2 3 7 5 8 7 July 12 15 16 6 21 3 4 178 Tuly 21 5 6 4 4 219 July 28 30 23 19 Z2, 145 increased in size a heavier dosage was necessary (15 to 18 pounds per acre) to get good coverage. In the experiments carried out at Windsor, repeated applications were made on account of the 422 Connecticut Experiment Station Bulletin 326 many rains in June with extensive burning as a result. The injury to the foHage was negligible when compared with the injury to the base of the stalk (Figure 30). It seems that the dust was washed off the leaves by successive rains and carried down the main stalk to its base at the soil surface. This accumulation of poison pro- duced a condition quite similar to that which is known as "sore- shin," the stalk becoming black and rotten. In some cases the stalk eventually broke off at this point. Figure 30. Paris green and calcium arsenate injury to base of stalk. This condition was found on shade grown as well as on Havana Seed treated with the same mixture. All of the shade grown tobacco dusted with this mixture at full strength suffered from this burning and about 50 per cent of the Havana Seed. \\'here the paris green-calcium arsenate mixture was used diluted with lime (one to five) the poison injury was slight. The Havana Seed tobacco that was dusted with full strength mixture was uniformly smaller than the check. Table 40 gives the comparisons in figures of the numbers of beetles found on 20 plants in each case. Tobacco Insect Studies 42o Cryolite This substance is a fluoride of aluminum and sodium, and under ordinary conditions is relatively insoluble, thus reducing the chances of foliage injury when used on plants. It is not toxic to man and animals, but is poisonous to insects. Cryolite was used at full strength and also diluted one to five with hydrated lime, and kept the flea beetle population down to a fairly low figure. See Table 40. Barium Fluosilicate This material has been used with some success against the flea beetle in the South, and was tried on several experimental plots this season at Windsor. It is also a substance of low solubility and therefore not dangerous to use on foliage. It is said to act some- what more slowly as a poison than does cryolite, but it kept the fiea beetle population down to a minimum. As with cryolite, no foliage injury from its use was observed on the experimental plots. While the differences in beetle population on treated and untreated plots were not great at first, there certainly was a great difference between treated and untreated shade grown tobacco at the time of the peak of emergence. For instance, on July 21 the treated shade grown tobacco had, in the case of cryolite, five and six beetles per 20 plants, while the untreated shade had a popula- tion of 219; on the same date, the untreated sun grown tobacco, just outside the tent, had 71 beetles per 20 plants. These figures are certainly of some significance. Cryolite and barium fluosilicate were also used on two plots of late-planted strain-test varieties to check the beetles that might migrate from the earlier tobacco as it was being harvested. The dusts were used at full strength and were applied at the rate of about 30 pounds per acre. At the time of application, July 28, there were 137 and 141 beetles per 20 plants on the barium fluosili- cate and cryolite plots respectively. Four days later, August 1, there were 97 and 92 respectively with 131 on the check. On August 9 there were 32 and 18 per 20 plants respectively and 51 on the check. The beetle population dropped off so rapidly on the untreated check plants as well as on the dusted ones after that date, that it was impossible to say that the scarcity of this pest was due to the effect of the poison or to a natural seasonal decline in abundance. The problem of residue on the foliage is one factor that stands in the way of the use of insecticides on Connecticut grown leaf. For this reason, the dustings were discontinued after July 16 on the Havana Seed. About a week later the beetles became very abundant on these plots, causing damage of commercial impor- tance. This means that the beetle must be completely suppressed in its first brood, or that some insecticide must be found that may be used up until harvesting time and that will leave no troublesome residue. 424 Connecticut Experiment Station Bulletin 326 Miscellaneous Observations on Tobacco Insects Flea Beetle, Epitrix cucumeris Harris In order to determine the nature of the activities of this insect on tobacco, daily observations were made on the plantation at the Tobacco Substation. Twenty plants were marked in each of five Fi.EA-BEETLE or TOBACCO WiND50R,C0hN. 1320. tv Figure 31. Flea beetle population on tobacco at Station plantation, 1930. parts of the field, and the beetles on each of the plants in these sections were counted every day. One of these sections, No. 5, was on the shade grown tobacco. Table 41 contains the figures showing the daily population secured in this manner. The flea beetle is more abundant on shade grown tobacco in the later part of the season, reaching its peak of abundance about ten Tobacco Insect Studies 435 days to two weeks earlier than the beetles on sun grown tobacco. A comparison of these facts is represented graphically in figures 31 and 32. Also the peak of abundance on shade grown tobacco coincides with or closely approximates the picking of the first leaves. There appears to be a smaller increase in numbers early in July, followed by a decline, which suggests that there are at least two broods. By the middle of August few beetles could be found on any plantation, and by August 20, they had practically disappeared. FLEA-BEtTLL TOBACCO WINDSOR, CONN.;':; n3o. 7— IT ? J J TT Figure 32. Comparison of flea beetle population on shade and sun grown tobacco at station. Although this pest may riddle the leaves of the young plants during June, its more serious form of damage is done from mid- July to late July, particularly on the shade grown foliage. At this period of its existence it feeds on both surfaces of the leaf, while earlier it confines its activities to the lower side only. After the lower leaves have been "primed" it migrates up to the next highest foliage and continues its feeding. The potato is the favorite food plant of this species, and under circumstances where potato and tobacco fields are planted side by side, the potatoes have been observed to suffer more than the tobacco. In one field, the tobacco next to heavily infested potatoes 4:26 Connecticut Experiment Station Bulletin 326 Table 41. Flea Beetle Population on Station Tobacco Taken on 100 Plants, 20 in Each of Five Parts of Field Date 1930 No. beetles No. beetles No. beetles No. beetles No. beetles Daily on Sec. 1 on Sec. 2 on Sec. 3 on Sec. 4 on Sec. 5 total June 19 5 6 11 31 3 56 June 20 8 15 16 27 8 74 June 21 18 16 30 39 9 112 June 23 49 38 27 34 12 160 June 24 34 27 31 iZ 9 134 June 25 36 15 24 23 8 106 June 26 35 9 31 24 12 111 June 27 25 14 20 20 10 89 June 28 24 20 13 17 5 79 June 30 33 20 15 15 4 87 July 1 30 16 12 6 5 69 July 2 13 14 13 10 10 60 July 7 29 12 3 8 7 59 July 8 22 10 8 16 14 70 July 9 31 20 5 13 53 122 July 10 23 16 7 23 93 162 July 11 13 23 7 26 166 235 July 12 9 21 6 27 178 241 Ju'y 14 23 73 18 87 418 619 July 15 29 45 17 71 313 475 July 16 24 37 18 39 368 486 July 17 18 48 27 49 393 535 July 19 37 73 21 53 264 448 July 21 75 80 67 71 219 512 July 23 74 176 89 85 293 717 July 24 83 134 58 57 221 553 July 25 66 248 58 78 234 684 July 28 101 85 37 59 145 427 July 29 90 63 31 47 170 401 Totals 1,057 1,374 720 1,088 3,644 7,883 was free from the flea beetle until the potato vines began to dry up. Then the tobacco next to the potatoes became infested. A tomato plant and a potato vine were placed in a row of tobacco that was heavily infested with beetles, and in a few days the insect was far more abundant on the potato than on either the tomato or the tobacco plant nearest. These facts suggest that it may be pos- sible to make up a poisoned bait, using- for an attractant the juice of potato vines or the substance that gives the potato its charac- teristic odor and flavor. Tobacco Insect Studies 427 Grasshoppers On July 18, 1930, our attention was called to a sudden infesta- tion of grasshoppers on 181-2 acres of shade grown tobacco in Avon. Most of the hoppers were young nymphs (probably Melanoplus atlantis Riley) apparently not more than a week old and had started toward the complete devastation of the crop. About 90 per cent of the plants showed injury of the nature indi- FiGURE 33. Grasshopper injury to leaves. cated in Figure 33, and individual plants had anywhere from one to six leaves with such holes in them. Badly infested plants had as many as ten holes to a leaf. The manager hired boys to hand- pick the hoppers, but judging from the size of the insects at that stage and their activity, it is doubtful that the boys were able to catch more than one in four. The immediate use of poisoned bran bait was advised, and the following standard mixture was made up : Bran 25 lbs. Paris Green 1 lb. Oranges J^ doz. Molasses 2 qts. Water to moisten This material was scattered on the ground between every other row. A visit to the plantation on July 21 showed that about 75 per cent of the grasshoppers had disappeared. At this time, the bait 428 Connecticut Experiment Station Bulletin 326 that was left over from the first appHcation was being scattered between the rows not treated before. On July 24 it was estimated that about 90 per cent of the hoppers were gone. A careful examination of the ground resulted in the recovery of some dead grasshoppers and many dead and dying ground beetles. The hoppers were hard to find because of their small size and because of the fact that they turn brown and shrivel up soon after death. All of these and some that were taken before the application of the poison were submitted to chemical tests for the presence of arsenic. The dead hoppers and dead ground beetles gave positive reactions to the tests, while the hoppers collected before the use of the poison reacted negatively. An estimate of the costs involved for the 18.5 acre field was made by the manager of the plantation. They are as follows : Materials Waste bran $22.00 Paris green 32.00 Molasses 4.40 Oranges (seconds) 15.00 Labor Mixing 7.50 Applying 10.00 Total $90.90 CAUTION : This mixture will cause serious burning if allowed to come in contact with or remain on the foliage. Domestic animals and fowl should be kept away. It is unusual to find grasshopper infestations under shade unless it be along the edges of the fields. In the above cited instance, the land had been in hay the year previous to planting tobacco, and in alfalfa for several years prior to that. As stated above, the princi- pal species in this infestation was thought to have been the lesser migratory locust, Melanoplus atlantis Riley. It was in a very- young nymphal stage. A few mature red-legged grasshoppers, M. femur-ruhrum DeG., were found, and an occasional Carolina locust, D. Carolina L, Another field (Broadleaf in this case) in Windsor Locks was being damaged by the red-legged grasshopper, and the Carolina grasshopper. It was a three-acre piece surrounded by hay land, and as soon as the hay had been cut, the hoppers migrated to the tobacco. While the edges of the field were attacked more seri- ously, the hoppers found their way to the middle. A single appli- cation of the grasshopper bait was sufficient to kill them all ofif in the center, and two applications in the immediate grass land and along the outside rows of tobacco completed the control measures very satisfactorily. Two days after the first application, dead hoppers could be picked up by the handful. Tobacco Insect Studies 429 Late Infestation of Wireworms A late infestation of wireworms was found at Burnside on August 16 in a field of Broadleaf. The larvae had apparently Figure 34. Wireworm injury to large tobacco stalks. been at work for some time as plants of all sizes could be found infested with them. Thirteen worms were taken from one plant. Figure 35. Specimen of wireworm on injured stalk. 430 Connecticut Experiment Station Bulletin 326 The injury (shown in Figures 34 and 35) resulted in malformed plants, dwarfed plants and chlorotic foliage. In some instances, half-grown plants had been bored and tunnelled so extensively that they broke off at the soil surface. Still others were connected with the root system only by a few rotting strands of plant tissue, so that a g'entle pull broke them off completely. Areas infested ranged in size from single plants up to a square rod or so. A late infesta- tion of this nature is unusual. Stalk Borer The common stalk borer, Papaipeina nitela Guen., (Figure 29) was found tunnelling tobacco stalks in a few fields. This insect usually works on the outside rows near weedy borders. The larva enters the stalk an inch or so above the soil surface (Figure 36) Figure 2i6. Stalk borer entrance in sucker. and bores out the pith from root to growing tip. Practically all summer is spent in the larval stage, and only one borer lives in a stalk. Infested plants transplanted in breeding cages tended to send out abnormal growth of suckers. In one case, the main stalk dried up and two suckers took its place, one of these being tunnelled by the original borer. Hornworms Both species of hornworm were found on tobacco this season. The first larvae were found in the latter part of July in Suffield and were anywhere from one-third to full grown at this time. On August 18, half-grown as well as full-grown worms were found in New Milford. The only larvae of the southern species taken Tobacco Insect Studies 431 this season were found on the Experiment Station plots in Windsor. Pupation in rearing" cages started in late July and continued through late August. No moths had emerged by August 23. A third species, the white lined sphinx, Celerio lineata Fab., was taken on tobacco as a larva July 18. It pupated July 20 and emero-ed on August 13. Paris Green Injury to Tobacco Our attention was called to foliage injury resulting from the promiscuous scattering" of cutworm bait on a field of young plants. FiGRUE 37. Paris green injury to shade leaves. A shade grown field had been treated with bran and paris green and the mixture had been allowed to lodge on the foliage of nearly every plant, with the result that most of the leaves were severely burned. Figure 37 shows how paris green spots the leaves. Some burning was also observed on the field treated for grass- hopper. The manager realized that burning would result from the poisoned bait lodging on the foliage and cautioned his men. How- ever, accidental spilling occurred and resulted in injury worse than that of the grasshoppers. FERTILIZER LOSSES THROUGH LEACHING AS MEASURED BY LYSIMETER EXPERIMENTS M. F, Morgan, O. E. Street and H. G. M. Jacobson As reported in Bulletin 311, a lysimeter equipment was installed at Windsor in the spring of 1929 for the purpose of studying the losses of plant food elements through leaching, as affected by soil type, character of the fertilizer and crop removal. It is common knowledge that the leaching of fertilizer during wet seasons is a serious problem in tobacco production, particularly on our sandy soils. Although less conspicuously in evidence, the fate of fer- tilizer ingredients that are applied in excess of crop removal in our heavy tobacco fertilizer practice is an important factor in the cumu- lative effect of continuous tobacco culture upon the productive power of the soil. Much light should be thrown upon these prob- lems by means of the experiments now in progress. Briefly, the plan of the investigations is as follows : Surface soil lysimeters. Cylindrical tanks 20 inches in diameter containing only the normal depth (seven inches) of surface soil are used in this experiment. Four different soils are compared : 1. An excessively sandy soil from the "plains," designated as Merrimac coarse sand; taken from a shade field operated by H. C. Griswold in the town of Windsor Locks. 2. A medium sandy loam of the type most generally used for tobacco, designated as Merrimac sandy loam; taken from the Tobacco Substation field. 3. A very fine sandy loam, characteristic of the lighter textured soils of the gently rolling low uplands that occur just east of the level terraces lying east of the Connecticut River, designated as Enfield very fine sandy loam; taken from a field owned by P. Chamberlain, near Broad Brook. 4. A reddish brown loam, typical of the heavier upland soils which are derived from red (Triassic) sandstone glacial till, designated as Wethersfield loam; taken from a field owned by Olds and Whipple, in the town of Sufiield. On each of these soils the following forms of fertilizer nitrogen are compared, each used in amounts equivalent to 200 pounds of nitrogen per acre per year : nitrate of soda, sulfate of ammonia, urea and cottonseed meal. Precipitated bone, carbonate of potash, sulfate of potash and sulfate of magnesia are added to each tank in amounts sufificient to supply 100 pounds of phosphoric acid, 200 pounds of potash and 25 pounds of magnesia per acre per year. The Merrimac sandy loam soil (No. 2) also includes tanks with- out nitrogen, but with other nutrients applied at the usual rate. All treatments are in duplicate. All of the surface soil lysimeters are uncropped, since this experiment is primarily designed to study the absolute amounts of nitrogen that become available and leach through the surface soil Lysimeter Measurements of Leaching 433 at different periods of the year from various sources of nitrogen on the different soils. Twenty-inch lysimeters. These are cylindrical tanks 20 inches in diameter and 20 inches deep, containing surface of normal depth (seven inches) placed over about 12 inches of subsoil. All the soil for this experiment is the Merrimac sandy loam from the Tobacco Substation field, identical with the one used in the surface soil lysimeters. In this experiment a comparison is made between the following sixteen nitrogenous fertilizers : Mineral and synthetic sources r nitrate of soda Nitrate nitrogen \ nitrate of potash ( nitrate of lime A™™ • -i S sulfate of ammonia Ammonia nitrogen ] ^^^phos "B" Urea Calurea (containing both urea and nitrate of lime) Cyanamid Vegetable organic sources Cottonseed meal Castor pomace Linseed meal Animal organic sources Dry ground fish mieal Hoof and horn meal Dried blood Animal tankage Cow manure In all cases precipitated bone, carbonate of potash, sulfate of potash and sulfate of magnesia are added in amounts sufficient to supply the soil in the tanks with a total of 100 pounds of phos- phoric acid, 200 pounds potash and 25 pounds of magnesia, except that the nitrate of potash necessary to furnish the required amount of nitrogen also supplied 677 pounds of potash, while the ammo- phos yielded 312 pounds of phosphoric acid.^ One tobacco plant is grown to normal harvest maturity on each of the 20-inch tanks. In 1929 the crop was destroyed by hail just before harvest time and the crop was chopped up and returned to the soil, as was the common field practice following the severe hail- storm that year. The crop of 1930 was harvested, dried, weighed and analyzed, in order to obtain data on crop removal of plant food elements. Whenever a sufficient amount of rain has fallen to produce leach- ing, the leachate is measured and sampled, and the nitrate nitrogen content of the leachate is determined. Complete chemical analyses ^ In 1929 errors in preliminary data on the nitrogen content of the fertilizer material caused the use of 227 pounds of nitrogen as cyanamid and 280.6 pounds of nitrogen as hoof and horn meal. 434 Connecticut Experiment Station Bulletin 326 of the composite samples of the leachings from each tank are made for two six-months periods for each year beginning about May 25. This is the normal date for the fertilizer applications that are made just before tobacco setting. The data collected during the year ending May 25, 1930, and for the six-months period ending November 25, 1930, will be briefly summarized in order to show the character of the results and for comparison with those of future years. It is to be remembered that soil leaching is dependent upon weather conditions, which fluctuate from season to season. These have been abnormal for practically the entire period since the beginning of these experiments. The rainfall since May 25, 1929, has been unusually low almost every month. The year ending May 25, 1930, showed a rainfall of only 32.75 inches, which is about 12 inches below normal. For the six-months period ending November 25, 1930, the rainfall showed a corresponding deficiency. In the tobacco growing season of 1929, no leaching occurred, except about 0.2 of an inch from the sandiest soil, until the heavy rain and hailstorm, which destroyed the crop on August 1. More than two inches of rain at this time saturated all the soils and pro- duced leaching, except in the 20-inch tanks. The amounts of leaching in terms of equivalent inches of rain- fall for the different soils during 1929-1930 were as follows: Surface soil tanks May 26- May 26- Nov. 26, 1929- Total Aug. 4, 1929 Nov. 25, 1929 May 25, 1930 for year Merrimac coarse sand 1.885 5.9 7.7 13.6 Merrimac sandy loam 1.255 4.6 6.4 11.0 Enfield V. f. s. 1 0.527 2.9 5.8 8.7 Wethersfield loam 0.760 3.8 6.7 10.5 Twenty-inch tanks Merrimac sandy loam 2.4 6.0 8.4 During the growing season of 1930, heavy rains at the end of May and in early June produced leaching of all the soils. No further rain of significant volume to produce leaching occurred until August 20, after the tobacco was harvested, and the season continued very dry until the heavy rains of November. Leachings during the first six months of the 1930-1931 lysimeter year, and for the rainy period of May 26 to June 11 are shown below, in equivalent inches of rainfall. Surface soil tanks Mav 26- Mav 26- June 11, 1930 Nov. 25, 1930 Merrimac coarse sand 2.019 5.268 Merrimac sandv loam 1.729 4 649 Enfield v. f. s. 1 1.208 3.324 Wethersfield loam 1.621 4.407 Twenty-inch tanks Ad^errimac sandy loam 1.166 2.470 Lysimeter Measurements of Leaching 435 It is thus seen that a distinct difference appeared in the distribu- tion of the rainfall in the growing seasons of 1929 and 1930. In 1929 leaching did not occur until at the end of the period, prac- tically all being collected on August 1 and August 4, more than two months after the fertilizer was applied. In 1930 all the leach- ing occurred in the first two weeks, before the nitrification of the non-nitrate forms of nitrogen had time to reach a maximum. This difference is clearly evident in a comparison of the amount of nitrate nitrogen and its maximum concentration during these periods of leaching. Table 42. Nitrate Nitrogen Leached from the Surface- Soil Lysimeters IN THE Growing Seasons of 1929 and 1930 ,. — Lbs. per acre — ^ Maximum concen- Soil and treatment 1929 1930 tration of leachate, in parts per million 1929 1930 Merrimac coarse sand Nitrate of soda 97.394 127.631 400 667 Sulfate of ammonia 36.013 17.344 108 67 Urea 70.153 33.413 198 150 Cottonseed meal 64.813 9.478 175 33 Merrimac sandy loam Nitrate of soda 113.401 102.914 540 630 Sulfate of ammonia 59.661 21.612 294 86 Urea 82.132 37.447 2>?>d, 180 Cottonseed meal 65.207 9.408 302 30 No nitrogen 25.701 10.684 155 38 Enfield v. f . s. 1. Nitrate of soda 33.299 36.917 358 288 Sulfate of ammonia 12.273 12.940 145 79 Urea 17.740 14.343 185 89 Cottonseed meal 13.907 9.119 164 40 Wethersfield loam Nitrate of soda 60.879 85.459 474 480 Sulfate of ammonia 21.975 13.434 213 62 Urea 37.851 48.971 319 255 Cottonseed meal 33.871 10.140 208 47 The greater volume of leachate in the growing season of 1930 explains the somewhat greater losses of nitrog'en from the nitrate of soda treatments for three of the four soils. It is evident that during the first two weeks after fertilizer application cottonseed meal has produced a nitrate nitrogen concentration much lower than sulfate of ammonia and that urea has produced a concentra- tion significantly higher. This indicates that nitrification of urea begins more promptly and it is believed that if it had continued to be wet during the latter half of June a serious loss of nitrogen from the urea-treated soils would have occurred. The heavy losses of nitrogen from the nitrate of soda tanks on all soils except the Enfield very fine sandy loam show that leaching 436 Connecticut Experiment Station Bulletin 326 at any time during the season will remove much of the available nitrogen from the surface soil. It is interesting to note that while the surface soil tanks of the Merrimac sandy loam in 1930 revealed a removal of more than half of the total application of nitrogen as nitrate of soda at the begin- ning of the growing season, the 20-inch tanks, with normal sub- soils under the same surface soils, showed that this nitrogen was not lost to the crop. The loss from none of the 20-inch tanks in the growing season exceeded 10 pounds per acre, while the crop of 1930 was able to take up more nitrogen from the tanks treated with nitrate of soda than from any other treatment. The data on crop removal on the 20-inch tanks in 1930 are shown in Table 43. Table 43. Nitrogen Removed by 1930 Crop of Tobacco from 20-Inch Lysimeter Tanks , Average of duplicate tanks ^ Lbs. per acre Percentage of nitrogen in dry har- vested tobacco^ plants Nitrate of soda Ammophos "B" Calurea Urea Nitrate of potash Nitrate of lime Sulfate of ammonia Cyanamid Cottonseed meal Castor pomace Linseed meal Dry ground fish Hoof and horn meal Dried blood Tankage Manure No nitrogen It is evident from the above data that in spite of the leaching in early June, more nitrogen was available in the soil from all the non-organic nitrogenous materials than from the organic sources. The only plants that showed evident symptoms of nitrogen starva- tion were on the tanks without fertilizer nitrogen or with manure as the only source of nitrogen. This agrees with field observa- tions, where nitrogen deficiency was noted only on soils similar to the Merrimac coarse sand, with nitrate of soda as the chief source of the element. By November 25, the end of the first six-months period, in both years practically the entire application of nitrate of soda was 109.481 3.06 107.107 3.14 102.983 2.99 99.308 2.90 98.560 3.20 91.045 3.07 89.585 3.15 80.015' 3.42 75.887 1.80 76.922 2.07 yemi 2.19 78.931 2.39 70.165 2.04 73.310 2.23 69.121 2.02 52.659 1.68 24255 1.27 ^Analyses by Dr. E. M. Bailey, head of the Department of Anal5rtical Chemistry. ' Two plants per pot on these tanks. Lysimeter Measurements of Leaching 437 leached from all the surface soils. The 20-inch lysimeters were not comparable in the two different years, since in 1929 the hail- destroyed crop was returned to the soil. The nitrogen losses dur- ing the second six months of the first lysimeter year were very low on all the surface soils, although the bulk of the nitrogen losses from the 20-inch tanks occurred during that period. Table 44. Nitrate Nitrogen Leached from Surface Soil Lysimeter Tanks in 1929 and 1930 Pounds per acre Soil and treatment May 26- Nov.25,1929 Merrimac coarse sand Nitrate of soda ..... 208.207 Sulfate of ammonia.. 117.982 Urea 148.637 Cottonseed meal .... 138.032 Merrimac sandy loam Nitrate of soda 250.328 Sulfate of ammonia. . 201.161 Urea 219.708 Cottonseed meal 174.612 No nitrogen 76.379 Enfield v. f . s. 1. Nitrate of soda 192.525 Sulfate of ammonia. . 146.581 Urea 142.732 Cottonseed meal 107.984 Wethersfield loam Nitrate of soda 167.216 Sulfate of ammonia. . 158.359 Urea 160.451 Cottonseed meal 99.714 Nov. 26, 1929- May25, 1930 6.727 6.050 5.963 6.594 12.056 11.451 10.413 12.421 11.768 14.127 24.831 16.291 20.625 55.887 31.597 50.633 35.395 May 26- Nov.25, 1930 218.608 72.939 90.509 80.324 233.918 172.383 169.164 119.132 58.771 210.662 164.265 153.350 93.597 220.442 153.336 189.629 91.680 Total to Nov. 25, 1930 423.542 196.971 245.109 224.950 496.302 384.996 389.285 306.165 146.918 417.314 335.677 312.373 222.206 443.545 343.292 400.713 226.789 From the above table it is evident that urea is leached from the soil at a distinctly more rapid rate than cottonseed meal, and that its rapid transformation into nitrate nitrogen proceeds until prac- tically all of the material is nitrified. Cottonseed meal has been more slowly and incompletely nitrified, and the 20-inch lysimeters show that this is true of all the organic ammoniates. Sulfate of ammonia begins to nitrify more slowly than urea, but the amount that becomes available as nitrates is as high or higher than for urea. This is true except on the Merrimac coarse sand, which is so sandy that it loses a considerable amount of ammonia nitrogen by leaching. The leaching of ammonia nitrogen during the first lysimeter year was as follows: 438 Connecticut Experiment Station Bulletin 326 Table 45. Nitrogen Leached as Ammonium Salts from Surface Soil Tanks. First Lysimeter Year (May 26, 1929, to May 25, 1930) Pounds per acre Nitrate Sulfate Urea Cottonseed of soda of ammonia meal Merrimac coarse sand 1.854 41.617 13.250 8.051 Merrimac sandy loam 1.308 11.189 2.197 2.156 Enfield V. f. s. 1 2.421 5.171 2.237 0.978 Wethersfield loam 0.702 1.165 0.700 1.117 The leaching of nitrate nitrogen from the 20-inch tanks has shown rather significant difi^erences in the absolute loss of nitrogen through the subsoil for the different sources of nitrogen. This is shown in Table 46, Table 46. Nitrate Nitrogen Leached from 20-Inch Lysimeter Tanks in 1929 and 1930 Pounds per acre Treatment May 26- Nov. 26, 1929- May 26- Total to Nov.2S,1929 May 25, 1930 Nov. 25,1930 Nov. 25,1930 Nitrate of soda 79.170 111.236 76.172> 266.579 Nitrate of potash 62.103 133.343 59.964 255.410 Nitrate of lime 66.796 122.909 52.159 241.864 Sulfate of ammonia .... 37.109 121.832 28.819 187.760 Ammophos 41.483 114727 40.398 196.608 Urea 43.750 98.378 38.569 180.697 Calurea 45.575 117.321 49.003 211.899 Cyanamid 19.459 58.501 44.659 122.619^ Cottonseed meal 43.476 71.023 34.644 149.143 Castor pomace 44.297 91.269 32.151 167.717 Linseed meal 38.039 80.354 28.848 147.241 Dry ground fish 35.481 78.898 36.591 150.970 Hoof and horn meal . . . 43.539 99.250 24.948 \67.72r Dried blood 41.111 117.980 32.026 181.117 Tankage 32.538 85.407 30.423 148.368 Manure 31.681 43.035 34.977 109.693 No nitrogen 13.088 43.272 15.296 71.656 The Effect of Various Nitrogenous Fertilizers Upon the Leaching of the Basic Constituents of the Soil Since there are great differences in the amounts of nitrate nitro- gen leached, which depend upon the form of nitrogen and the character of the soil, it is to be expected that there will be corre- sponding differences in the removal of the basic constituents, prin- cipally calcium, magnesium, potassium and sodium, from the soil by percolating water. Nitrate nitrogen is in the form of anions, ^ 227.5 pounds of nitrogen applied in 1929. * 280.6 pounds of nitrogen applied in 1929. Lyshneter Measurements of Leaching 439 which are tied up with these cations, as chemical compounds dis- solved in the leachate. The results of complete chemical analyses have verified this supposition. Table 47. The Leaching of Basic Constituents from the Surface Soil Lysimeter Tanks. The First Lysimeter Year (May 26. 1929, to May 25, 1930) Soil and treatment Merrimac coarse sand Nitrate of soda Sulfate of ammonia . . Urea Cottonseed meal Merrimac sandy loam Nitrate of soda Sulfate of ammonia . . Urea Cottonseed meal No nitrogen Enfield v. f. s. 1. Nitrate of soda Sulfate of ammonia . . Urea Cottonseed meal Wethersfield loam Nitrate of soda Sulfate of ammonia . . Urea Cottonseed meal On all soils, nitrate of soda shows a very high value for sodium leachings, with low figures for the other basic elements. On the two sandy soils, practically all the nitrate outgo can be accounted for as sodium nitrate. Sulfate of ammonia has very materially enhanced the leachings of calcium, magnesium and potassium Urea has produced materially greater calcium losses than has cottonseed meal, as would be expected with the more complete nitrification of the former material. Cottonseed meal and urea produced practically the same losses of magnesium, potassium and sodium. Similar data for the 20-inch tanks, shown in Table 48, have con- firmed these results, and show a close correlation between the cal- cium losses and the nitrate losses, except in the case of sodium nitrate, sulfate of ammonia and ammophos "B" (containing sulfate of ammonia). As in the shallow tanks, sodium nitrate has repressed the calcium outgo, while the sulfate of ammonia, furnishing the two soluble anions, has greatly enhanced it. 'ounds per acre Calcium Magnesium Potassium Sodium 68.700 153.123 112.891 90.125 11.168 14.415 11.056 17.167 50.388 62.767 49.203 62.294 299.298 13.567 10.027 13.131 148.445 376.153 246.959 204.239 122.769 24.850 48.599 37.820 32.710 19.098 43.203 71.133 61.144 63.633 59.583 263.591 24.725 21.688 17.399 9.469 102.991 187.375 165.861 113.733 25.265 34.090 35.031 28.436 51.617 64.407 51.342 55.725 164.851 32.257 24.699 25.647 162.803 341.436 238.594 184.112 24.314 34.328 25.647 26.324 11.999 13.321 12.888 13.441 186.533 40.183 48.833 49.333 Pounds per acre Calcium Magnesium Potassium Sodium 131.241 14.689 63.513 117.906 205.922 28.411 54.329 22.700 215.054 23.112 46.609 19.836 282.605 39.973 55.885 19.812 245.876 36.598 51.228 20.265 168.265 19.981 52.759 14.376 196.651 23.527 56.066 20.306 96.414 15.351 57.441 20.049 135.836 14.835 51.242 21.096 165.034 18.741 63.078' 15.827 148.840 20.084 60.701 17.569 141.744 14.227 51.136 11.164 173.348 22.470 58.441 10.828 172.380 31.810 56.171 26.741 143.866 16.124 50.096 27.848 90.393 12.038 47.379 20.952 76.485 18.773 52.138 20.785 440 Connecticut Experiment Station Bulletin 326 Table 48. The Leaching of Basic Constituents from the 20-Inch Tanks. The First Lysimeter Year (May 26, 1929, to May 25, 1930) Treatment Nitrate of soda Nitrate of potash Nitrate of lime Sulfate of ammonia .... Ammophos "B" Urea Calurea Cyanamid Cottonseed meal Castor pomace Linseed meal Dry ground fish Hoof and horn meal .... Dried blood Tankage Manure No nitrogen One very striking" point in these data is the fact that nitrate of potash, furnishing" 677 pounds of K^O per acre as compared with 200 pounds for all other apphcations, produced no increase in the amount of potassium in the leachate, although it caused almost as great a calcium outg"o as did nitrate of lime. It appears that while sodium caused very little base exchang"e, potassium was almost completely absorbed by the soil, its place being taken by calcium in the leachate. Other Important Constituents of the Leachate from Soils Treated With Various Nitrogenous Fertilizers Besides the nitrate, there are two other important anions in the leachates: the sulfate and the bicarbonate. Since all of the soils are acid and no liming material was used in this experiment, no carbonates were present, and the bicarbonates were very low on the two most acid soils, Merrimac coarse sand and Enfield very fine sandy loam. Tests for the chlorides showed a very low concentration in all the leachates, the outgo in no case exceeding one and one-half pounds per acre for the first year. No chlorides were used in the fertilizer. The concentration of phosphates was of about the same order of magnitude, and these two anions may be practically disregarded. Some of the surface soil tanks, particularly those treated with sulfate of ammonia and urea, showed a distinctly acid leachate. In such cases significant amounts of manganese and ahnninum were present. ^ One hundred and fifty pounds potash applied in 1929. Lysimeter Measurements of Leaching 441 Table 49. Leachings of Certain Constituents from the Surface Soil Tanks and their Relationship to the Reaction of the Leachate. First Lysimeter Year (May 26, 1929, to May 25, 1930) Soil and treatment Reaction , Lbs. per Acre v of leachate Bicarbonates (PH) (HCOs) Sulfur Manganese Aluminum Merrimac coarse sand Nitrate of soda 7.35 70.830 45.630 3.361 1.287 Sulfate of ammonia . . , 4.65 7.085 181.502 22.616 25.279 Urea 4.85 17.211 27.550 10.821 13.901 Cottonseed meal 6.60 33.346 26.826 5.812 5.099 Merrimac sandy loam Nirate of soda 7.25 69.731 61.858 0.435 1.382 Sulfate of ammonia .. . 6.90 37.^77 219.455 21.811 2.602 Urea 7.25 38.723 52.825 4.860 2.755 Cottonseed meal 7.10 76.204 53.661 1.331 1.751 No nitrogen 7.15 83.459 54.333 0.316 0.745 Enfield v. f. s. 1. Nitrate of soda 7.55 60.696 82.669 2.157 1.404 Sulfate of ammonia . . . 4.60 9.456 91.727 16.258 14.770 Urea 5.50 21.795 32.241 8.345 5.523 Cottonseed meal 6.85 27.793 39.699 4.442 4.630 Wethersfield loam Nitrate of soda 7.75 42.600 40.846 0.061 0.728 Sulfate of ammonia . . . 7.60 43.383 163.341 0.354 1.449 Urea 7.40 64.239 32.032 0.114 . 1.568 Cottonseed meal 7.65 57.202 36.138 0.000 0.450 The 20-inch tanks produced leachates that were slightly alkahne in nearly all cases, the reaction varying from 7.0 to 8.2 pH with no apparent correlation with fertilizer treatment. The manganese and aluminum concentration of all 20-inch leachates was almost negligible, which indicates that these elements, if leached from the surface soil of the Merrimac sandy loam, were absorbed in the subsoil. The bicarbonates in the 20-inch leachate. expressed as (HCOo) in the first lysimeter year ranged from 57 to 7?) pounds per acre, with no apparent correlation with treatment. The sulfate outgo, expressed as sulfur, for the same period from the same series of tanks ranged from 28 to 64 pounds per acre from all treatments except sulfate of ammonia and ammophos "B". These were 148.231 and 103.112 pounds, respectively, which indi- cates the rapid outgo through the subsoil of heavy applications of sulfate material. THE USE OF COKE IN HEATING TOBACCO SHEDS J. S. Owens' The use of artificial heat in curing tobacco must be determined by the effect on the quality of the product (including control of pole sweat) and the cost. Stalk growers in particular have con- FiGURE 38. Salamander and spreader designed for use in tobacco sheds. Note that 18-inch height permits spread of heat underneath lower tier of tobacco. sidered the use of charcoal too costly, except when extreme danger of sweat is imminent. Gas coke has been used in the Burley district. Coke made from bituminous coal is now distributed widely in this state. This project was undertaken to determine ^ Extension agronomist of the Connecticut Agricultural College. The project was conducted in cooperation with the Connecticut Coke Company and David M. Hadlow of that company assisted in developing the equipment used and in conducting the tests. Use of Coke in Heating Sheds 443 whether coke might be used with good results and whether its cost for both labor and fuel compared favorably with charcoal. In 1929, several types of salamanders or burners were developed and tried in empty sheds, at the Tobacco Substation, but loss of the crop by hail prevented using coke for curing trials there. Later in the season, coke fires were used in curing a small shed of tobacco on the farm of Lester Lloyd in Sufifield. Except for some difficulty in keeping the grates clean, the coke fires appeared satisfactory. On August 15, 1930, coke fires were started in shed No. 3 at the Station. This shed is 28 by 80 feet. Thirteen salamanders 12 inches in diameter and 16 inches high, and another 14 inches in diameter and 16 inches high, were distributed in two rows 12 feet apart. The salamanders were placed 13 feet apart in the rows and the rows were eight feet from the outside wall. The salamanders were made cylindrical, of sheet metal, fitted with a grate and covered with an inverted cone-shaped spreader of light metal, three feet in diameter. Thermometers located at varied elevations and distances from the burners showed variations within 4° F. The distribution of heat was excellent in spite of a strong, cool wind and large cracks in the sides of the shed. Thirty charcoal fires were started on the same day in shed No. 2, which is 28 by 93 feet. Fifteen fires were placed on each side and both sheds were heated 48 hours. Although an attempt was made to keep the sheds at approximately the same temperatures, the coke shed, with readings taken at the same time and in corresponding locations, averaged 7.2° F. higher than the shed heated with char-, coal, and 17.4° F. above the outside temperature. Observations on the tobacco both at time of curing and when sorted, showed no difference in quality. The coke consumed weighed 900 pounds, and the charcoal, 1,200. At prevailing prices of $13 and $20 per ton respectively, the fuel cost was $5.85 for the coke and $12.00 for the charcoal. Reduced to the basis of equal shed area the cost for charcoal would have been $10.34. On September 5, a shed 128 by 30 feet, filled with Broadleaf, on the farm of John M. Herr, East Hartford, was heated with coke. Twenty-four salamanders were used. Twenty-three were 12 inches in diameter and 16 inches tall, and one was 14 inches in diameter and 16 inches tall. The fires were run continuously for 40 hours and an average temperature of 81.2° F., or 12.9° above outside temperature was maintained. One ton of stove size coke and 60 pounds of charcoal (for starting fires) were used. Experience appears to warrant a few comparisons between coke and charcoal. In the test at the Tobacco Station more than 40 per cent of the fuel cost was saved by using coke. With 14 inch salamanders fitted with large mesh grates, attention was needed only every four to eight hours, while charcoal fires needed replen- 444 Connecticut Experiment Station Bulletin 326 ishing at periods ranging from 20 minutes to one hour. Coke gave a steady, practically odorless and smokeless heat, which with properly designed salamanders, and spreaders, was well distri- buted, and no trace of objectionable odor or injurious effect upon the tobacco appeared in any one of the three sheds so heated. Charcoal fires are more flexible. They are a little easier to start, and to increase or decrease than the coke. If continuous fires are needed, this factor would be insignificant. Coke requires special burners. A smaller number of fires are needed than with charcoal and the cost of the equipment might be covered by the saving in fuel during the first season. The equipment should last many years. FIRE-CURING TESTS ON STALK TOBACCO J. S. Owens Growers of shade tobacco consider artificial heat indispensible for hastening curing and for prevention of pole sweat. An occasional grower of stalk tobacco uses charcoal fires in emer- gencies to prevent pole sweat, and on a very few farms, fires are used to hasten curing. This project was undertaken to determine the practicability and the effect upon quality of curing stalk tobacco with methods similar to those that the shade growers believe essential. In 1928, one or more sheds were fired on each of nine farms. Four of the growers raised Havana and five Broadleaf. Imme- diately after filling, most of the sheds were heated for about forty- eight hours, or until the leaves became wilted. Open or covered charcoal fires were used and readings of 20° F. above outside tem- perature maintained, except during hot days. Every shed was fired from one to three times to dry off the leaves during the dampest spells. Because of the varying condition of the tobacco and weather, a uniform plan of heating and ventilation could not be followed. Only a small amount of pole sweat existed in any of the sheds fired, although the season was exceptionally favorable to such damage and heavy losses from sweat were general. Where sweat occurred in the fired sheds, either the firing had been delayed until weather conditions became bad or the fires were not maintained long enough to dry the leaves thoroughly. Injury from over- firing was not indicated, even in the sheds that were heated a total of 150 hours. The weather conditions were much more favorable for curing without artificial heat in 1929. Two growers of Broadleaf and four growers of Havana fired to hasten curing, almost regardless of weather conditions. Again the tobacco appeared to be of good quality. However, no comparisons with what would have resulted without firing were possible. In 1930 comparative tests were conducted on three farms. In each case tobacco was taken from the field and portions stored in two similar sheds on the same day. One of each pair of sheds was heated immediately and kept about 20° F. above outside tem- perature for approximately 48 hours, at which time the leaves were Note : The tests and observations on which this report is based were made with the assistance of R. G. Tryon and Dan Andrews, Glastonbury; R. E. Case, West Granby; S. R. Spencer, Suffield; Theodore Hurlbut, Somers; Horace Pease and DeCarli Brothers, Ellington; Harry N. Farnham and J. E. Shepard, East Windsor; and John M. Herr, East Hartford. The writer wishes to express appreciation for the splendid cooperation given by these men, 446 Connecticut Experiment Station Bulletin 326 well wilted. The same sheds were again heated a week later for a period of 12 to 24 hours, to remove the moisture during a damp spell. The sorting records for each of the three farms are as follows : John M. Herr, East Hartford (Broadleaf) Acre yield , Percentage of grades ^ Grade L M LS SS LD No.2S F B LT index Fired 2122 2.3 8.2 18.9 1.3 27.5 23.3 4.9 3.4 10.3 .372 Not fired 1908 ... 3.5 13.6 3.9 27.1 22.4 ... 12.5 17.1 .313 Harry N. Farnham, East Windsor (Broadleaf) Acre yield , Percentage of grades s Grade L M LS SS LD No.2S F B T index Fired 1656 2.4 8.1 19.1 0.4 31.7 20.5 9.8 ... 7.9 .371 Not fired 1790 9.6 19.8 13.5 5.1 24.9 8.1 9.4 ... 9.6 .444 S'. R. Spencer, Suffield (Havana Seed) Acre yield , M Percentage of grades LS SS LD No.DS F B T Grade index Fired 1412 17.3 Not fired 1412 20.1 8.5 12.2 27.8 ... 37.2 ... 9.2 24.2 ... 36.3 ... 7.2 ... .512 .536 Changes in the quality that can be credited to the firing are neither consistent nor large. The season was abnormally dry throughout. Some tobacco cured too rapidly without artificial heat, and pole sweat was almost unknown. It is hardly possible to deduce from the one season's experience what effect there would be upon quality in a season with an average range of humidity. The results did not seem to justify firing last year, but the fact that there was little, if any, injury from firing under the extremely dry conditions, would indicate that possible damage to the crop is not the most important factor in deciding whether firing should be done. APPENDIX Fertilizer Formulas Used in the Tests of 1930 Referred to in This Report Nl. Standard Formula; 200 Pounds Nitrogen f Carriers ^ ,- Nutrients per acre ^ Name Lbs. per acre N P2O5 K^O MgO Cottonseed meal 1765. 120 52.9 35.3 12.4 Castor pomace 741. 40 14.8 7.4 5.9 Nitrate of soda 260. 40 Sulfate of potash 163.8 78.6 Carbonate of potash 122.8 78.7 Magnesian lime 400. 100. Total 3452.6 200 67.7 200. 118.3 N2. One-Half of the Nitrogen in Nitrate of Soda Cottonseed meal 1103. 75 33.1 22.1 1 :! Castor pomace 463. 25 9.3 4.6 2i.l Nitrate of soda 649.4 100 Sulfate of potash 65.1 31.3 Carbonate of potash 122.8 78.6 Double sulfate 243.8 63.4 26 9 Total 2647.1 200 42.4 200. 38.3 N8. One-Half of the Nitrogen in Urea Cottonseed meal 1103. 75 33.1 22.1 1 :i Castor pomace 463. 25 9.3 4.6 3 7 Urea 217.4 100 Sulfate of potash 180.4 86.6 Carbonate of potash 135.4 867 Precipitated bone 71. 27. Magnesian lime 400. 100. Total 2570.2 200 69.4 200. 111.4 N9. All Nitrogen in Urea Urea 434. 200 Carbonate of potash 154. 100. Sulfate of potash 200. 100. Precipitated bone 180. 67.7 Magnesian lime 400. 100. Total 1368. 200 67.7 200. 100. N18. Standard Formula; Calcium Nitrate Substituted for Nitrate of Cottonseed meal 1764.7 120 52.9 35.3 12.4 Castor pomace 740.7 40 14.8 7.5 5.9 Nitrate of lime 260. 40 Sulfate of potash 163.8 78.6 Carbonate of potash 122.8 78.7 Magnesium carbonate .... 40. 20. Total 3092.0 200 67.7 200. 38.3 Soda 448 Connecticut Experiment Station Bulletin 326 N28. Standard Formula; 200 pounds Nitrogen , Carriers ^ , Nutrients per acre ^ Name Lbs. per acre N P2O5 K2O MgO CaO Cottonseed meal 1764.7 120 52.9 35.3 12.4 5.3 Castor pomace 740.7 40 14.8 7.4 5.9 5.9 Nitrate of lime 260. 40 72.8 Sulfate of potash 163.8 78.6 0.3 Carbonate of potash 122.8 78.7 Magnesium carbonate .... 36. 18. Precipitated bone 243. 92.3 109.4 Total 3331. 200 160. 200. 36. 193.7 N29. One-Hundred Pounds Nitrogen from Nitrophoska Cottonseed meal 1103. 75 33.1 22.1 7.7 3.3 Castor pomace 463. 25 9.3 4.6 3.7 3.7 Nitrophoska 645.2 100 100. 122.6 0.7 3.5 Precipitated bone 46.3 17.6 20.8 Carbonate of potash 79.2 50.7 Magnesium carbonate .... 50. 25. Total 2386.7 200 160. 200. 37.1 31.3 N30. Maximum Nitrogen from Nitrophoska Nitrophoska 1052.6 163.2 163.2 200. 1.0 5.8 Urea 80. 36.8 Magnesium carbonate .... 73. 36.5 Total 1205.6 200. 163.2 200. 37.5 5.8 Kl. Basal Ration with All Mineral Potash in Sulfate of Potash Cottonseed meal 1765. 120 52.9 35.3 12.4 Castor pomace 741. 40 14.8 7.4 5.9 Nitrate of soda 260. 40 Sulfate of potash 327. 157.3 Magnesian lime 400. 116. Total 3493. 200 67.7 200. 134.3 K5. All Mineral Potash in Carbonate Cottonseed meal 1765. 120 52.9 35.3 12.4 Castor pomace 741. 40 14.8 7.4 5.9 Nitrate of soda 260. 40 Carbonate of potash 245. 157.3 Magnesian lime 400. 116. Total 3411. 200 67.7 200. 134.3 Appendix 449 K7. Three-Fourths of the Mineral Potash in Nitrate; One-Third in Carbonate , Carriers s f Nutrients per acre \ Name Lbs. per acre N PaOs K2O MgO Cottonseed meal 1765. 120 52.9 35.3 12.4 Castor pomace 741. 40 14.8 7.4 5.9 Nitrate of potash 325. 40 143. Carbonate of potash 22. 14.3 Magnesian lime 400. 116. Total 3253. 200 67.7 200. 134.3 K8. Mineral Potash Divided Equally Between Sulfate and Carbonate of Potash Cottonseed meal 1765. 120 52.9 35.3 12.4 Castor pomace 741. 40 14.8 7.4 5.9 Nitrate of soda 260. 40 Sulfate of potash 163. 78.6 Carbonate of potash 123. 78.7 Magnesian lime 400. 116. Total 3452. 200 67.7 200. 134.3 K9. Mineral Potash Divided Equally Between Sulfate, Nitrate and Carbonate of Potash Cottonseed meal 1765. 120. 52.9 35.3 12.4 Castor pomace 741. 40. 14.8 7.4 5.9 Nitrate of soda 165. 25.4 Sulfate of potash 109. 52.4 Carbonate of potash 82. 52.4 Nitrate of potash 119. 14.6 52.5 Magnesian lime 400. 116. Total 3381. 200. 67.7 200. 134.3 Kll. No Mineral Potash Cottonseed meal 1765. 120 52.9 35.3 12.4 Castor pomace 741. 40 14.8 7.4 5.9 Nitrate of soda 260. 40 Magnesian lime 400, 116. Total 3166. 200 67.7 42.7 134.3 K12. Potash Reduced to 100 Pounds Cottonseed meal 1765. 120. 52.9 35.3 12.4 Castor pomace 741. 40. 14.8 7.4 5.9 Nitrate of soda 225. 34.7 Sulfate of potash 40. 19.1 Carbonate of potash 30. 19.1 Nitrate of potash 43. 5.3 19.1 Magnesian lime 400. 116. Total 3244. 200. 67.7 100. 134.3 450 Connecticut Experiment Station Bulletin 326 K13. Three Hundred Pounds Potash, Equally from Carbonate, Sulfate, and Nitrate / Carriers ^ , Nutrients per acre \ Name Lbs. per acre N P2O5 K2O MgO Cottonseed meal 1765. 120. 52.9 35.3 12.4 Castor pomace ;.. 741. 40. 14.8 7.4 5.9 Nitrate of soda 105. 16.1 Nitrate of potash 195. 23.9 85.7 Carbonate of potash 134. 85.8 Sulfate of potash 179. 85.8 Magnesian lime 400. 116. Total 3519. 200. 67.7 300. 134.3 K14. Two Hundred Pounds Potash, All in Stems Cottonseed meal 1529. 104 45.9 30.5 10.7 Nitrate of soda 260. 40 Stems 2650. 56 15.9 169.5 13.3 Magnesian lime 400. 116. Total 4839. 200 61.8 200. 140. K15. Two Hundred Pounds Potash in Cotton Hull Ashes Cottonseed meal ........ 1765. 120 52.9 35.3 12.4 Castor pomace 741. 40 14.8 7.4 5.9 Nitrate of soda 260. 40 CottonhuU ash 590. 17.7 157.3 30.4 Total 3356. 200 85.4 200. 48.7 v) u O O ■ • U 'J University of Connecticut Libraries o^ ioou2oo»d4ou