UNIVERSITY OF ILLINOIS Agricultural Experiment Station BULLETIN No. 196 THE USE OF COMMERCIAL FERTILIZERS IN GROWING ROSES BY F. W. MUNCIE URBANA, ILLINOIS, FEBRUARY, 1917 SUMMARY OF BULLETIN No. 196 1. Brown silt loam, the type of soil at the Illinois Station, will not pro- duce a maximum crop of roses without fertilization. Page 515 2. Grafted stock more than paid for the increase in initial cost by its larger production during the first year. Page 519 3. Dried blood in amounts exceeding 8 pounds per 100 square feet of bench space caused a decrease in production with own-root and grafted Brides and grafted Killarneys. Page 520 The effect of dried blood upon weekly production was found to be a decrease during fall months, no difference during winter, and an increase in the spring. Page 528 4. Acid phosphate gave a greatly increased production with all types used in the experiment except grafted Brides. Page 521 Further experiments showed that 20 pounds of acid phosphate per 100 square feet of bench space gave a profit of $176 per 1,000 plants; four times this quantity may be used without injury from overfeeding. Page 555 The beneficial effect was continuous thruout the year. Page 557 5. No benefit was obtained from the use of potassium sulfate under the con- ditions of the experiment. Pages 522-523 6. A definite relation was found to exist between the variation in hours of sunshine and the subsequent production of flowers. Page 526 7. A decrease in production resulted from mixing ground limestone with the soil, whether or not acid phosphate had been added, and this material is not rec- ommended for general use. Pages 561-562 8. CONCLUSIONS AND KECOMMENDATIONS. Page 562 THE USE OF COMMERCIAL FERTILIZERS IN GROWING ROSES BY F. W. MTJNCIE, ASSOCIATE IN FLORICULTURAL CHEMISTRY Since the fall of 1910, experiments have been carried on by the Horticultural Department of this station Avith regard to the use of commercial fertilizers in growing first-year roses. The commercial fertilizers used during 1910-13 were dried blood, acid phosphate, and potassium sulfate. During 1913-15, ammonium sulfate was used instead of dried blood. Assuming, as is ordinarily done, that the element deficient in the soil and so limiting plant growth is nitrogen, phosphorus, or potassium, the choice of fertilizers (dried blood or ammonium sulfate supplying nitrogen ; acid phosphate, phosphorus ; and potassium sulfate, potassium) is wide enough to establish the order of relative abundance of the elements in compounds available to the plants, and to form a basis for study of the effect of these fertilizers upon yearly and weekly production. 1 The experimental work reported in this bulletin deals with a gen- eral investigation (1910-13) with each of the three fertilizers men- tioned, alone and in various combinations, and a supplementary one upon the use of acid phosphate with and without lime (1913-15). DESCRIPTION OF THE EXPERIMENT, 1910-13 Bride and Killarney were the varieties grown, the first being a typical tea rose and the latter a hybrid tea; half the plants of each variety were own-root stock and half were grafted. Under the con- ditions of fertilizing prescribed by the plan of the experiment, com- parisons can also be made between the varieties used and between own-root and grafted stock. One greenhouse 28 feet by 105 feet, containing four benches 4 feet wide. 100 feet long, and 5 inches deep, with an area of 1,600 square feet, was used; the experiment begun in 1910-11 was repeated during 1911-12 and 1912-13. The roses propagated about December 1 of the previous year were successively potted into 2-inch and 4-inch pots, and set in the benches about July 10. The dates of setting the plants in the benches were as follows: July 6, 1910; July 13, 1911 ; July 10, 1912. 'In ordinary practice, the soil used in growing greenhouse crops is partially or entirely replaced each year. Also the gross returns with the average crop grown in a greenhouse are estimated to be as high as $40,000 per year per acre of enclosed space. T'nder these circumstances, the florist is concerned with pro- ducing the maximum crop upon the soil without consideration of soil depletion and with relatively little for the cost of the fertilizer used always a small item in comparison with the crop returns. 511 512 BULLETIN No. 196 [February, The soil used was of that type common to the part of the state in which the Experiment Station is located, the brown silt loam, the average composition of which is as follows: 1 Nitrogen 5,000 Ibs. Phosphorus 1,200 Ibs. Potassium 36,000 Ibs. per surface layer of 6% inches (2,000',000 Ibs.) The soil had been planted to corn for a number of years, had lain fallow since 1909, and had been twice plowed during the spring pre- vious to its use. Before being put in the benches, it was thoroly pulver- ized and a uniform mixture was secured by preparing the soil accord- ing to the methods described in the bulletin, ' ' The Use of Commercial Fertilizers in Growing Carnations. ' ' 2 The dried blood used contained approximately 14 percent nitro- gen, the acid phosphate 7 percent phosphorus, and the potassium sulfate 41 percent potassium, as shown by the analyses given below : TABLE 1. ANALYSES OP FERTILIZERS USED, 1910-13 Year Nitrogen in dried blood Phosphorus in acid phosphate Potassium in potassium sulfate 1910-11 1911-12 1912-13 percent 13.70 14.07 14.09 percent 7.03 7.03 6.62 percent 41.9 41.9 39.7 NOTE. These analyses were made by the Department of Agronomy. In order to make a comparison of the various kinds and pro- portions of fertilizers, the benches were laid off into sections 10 feet in length, giving an area of 40 square feet to each, with room for 32 plants. To each of these were added manure and the fertilizers in the amounts shown in Tables 3 and 4. Of the 32 plants in each section, 16 own-root and 16 grafted were placed in alternate ends of the sec- tions in alternate years, the own-root and grafted plants receiving the same quantities of fertilizer per section. On account of the larger root area of the grafted plants, however, they really received propor- tionately heavier applications. The arrangement of sections during 1910-11 and 1911-12 is shown in Fig. 1, and that during 1912-13 in Fig. 2. The fertilizers were applied and the plants set in the following manner: The benches were filled with the uniformly mixed soil, a straight-edge being used to secure even filling. To each section well- rotted manure (containing about 50 percent moisture) was applied at the rate of 115 pounds per 100 square feet of bench space. The fertilizers were then added in the amounts shown in Tables 3 and 4, only one-fourth of the total amount of dried blood, however, being Hopkins, Soil Fertility and Permanent. Agriculture, p. 82. 111. Agr. Exp. Sta. Bui. 176, p. 366. 1917] USE OF COMMERCIAL FERTILIZERS IN GROWING EOSES 513 / 3 4 5 6 7 e 9 10 Killamey II 12 13 Id 15 16 17 18 /9 go Killamey 7 e 9 10 1 a 3 4 S 6 Bride * 17 te 19 so ii it 13 14 /f Sride N 9 FIG. 1. ARRANGEMENT OF SECTIONS, 1910-11 AND 1911-12 20 19 ia 17 * 16 15 14 13 11 II Killamey 10 9 e 7 6 5 4 3 S / Killamey If 14 !} IZ II eo 19 18 17 16 Bride 6 f 4 3 Z 1 10 9 3 7 Briote FIG. 2. ARRANGEMENT OF SECTIONS, 1912-13 applied at this time. The remainder was added in three equal portions as a top-dressing at different times during the year, and was worked into the soil lightly with hand cultivators to a depth of one inch. At the time of planting, the fertilizers were well mixed with the soil by using hand trowels, after which the soil was tramped down into the benches, thoroly moistened with water, and left to stand overnight. The roses were set on the benches the next morning at distances twelve inches apart across the bench and sixteen inches the long way, and were arranged carefully so as to secure as nearly as possible a uniform distribution with respect to vigor thruout the house, with small and large plants alternating in each section. They were then planted and thoroly watered. The methods used during the next few days were those followed by every careful grower to pre- vent too rapid transpiration before the roses take hold of the soil. Thruout the year, watering, fumigating, disbudding, weeding, etc., were looked after as in a commercial greenhouse. All flowers were 514 BULLETIN No. 196 [February, cut back to the second or third leaf above the previous break, but no attempt was made to control the time of cropping by pinching the buds, since normal growth served better as a measure of the effects of the fertilizers. The temperature in the rose house was regulated as carefully as the heating system permitted ; at times of very cold weather, however, a rise of several degrees above the temperature of the outer end of the house occurred in the end next the cross house. The arrangement of sections was changed in 1912-13 to equalize this effect as far as possible. On cloudy days, when artificial heat was depended upon altogether, the temperature was kept as near as possible to 68 F., while on sunny days it was allowed to rise as high as 75 or over. From sundown it was lowered gradually until nine o'clock, when it reached 63, a temperature which was maintained until midnight. Between this time and morning it was allowed to drop two or three degrees lower, but very seldom below 60. During cold weather, tem- perature was regulated entirely by steam coils, without ventilation. Records were kept of the number of flowers produced and the length of stem of each flower during the seasons November to May inclusive (7 months), 1910-11, and October to May inclusive (8 months) the remaining two years. It was not known whether the number of flowers produced would be an accurate standard of meas- urement of the money value of the crop, since length of stem largely determines the price of roses. To overcome this difficulty, the "total stem length" produced by the plant, that is, the total number of inches of stem (found by adding the stem length of all flowers produced) was used as an additional standard of measurement. It is evident that this value is an approximate measure of the total amount of vegeta- tion produced by the plant, particularly with Killarney, which has no blind wood. Records were kept by the week, so that not only the total number of flowers and the total stem length produced during the season, but the production at various times during the season might be compared. EFFECT OF FERTILIZATION UPON YEARLY PRODUCTION, 1910-13 During the seasons of 1910-13, records were kept on 3,840 plants, one-fourth of which were own-root Bride, grafted Bride, own-root Killarney, and grafted Killarney, respectively. A complete summary of the number of plants grown, the number of flowers, and the total number of inches of stem length produced, is given in Table 2. The 3,840 plants produced a total of 95,013 roses, an average of about 25 flowers per plant per season. This may be considered a satisfactory yield for first-year plants, since a considerable number of them were injured by overfeeding as a result of the attempt to 1917] USE OF COMMERCIAL FERTILIZERS IN GROWING EOSES 515 determine the maximum amount of fertilizer that might safely be used. Of these 95,013 flowers, 45,135 were Killarneys and 49,878 Brides (the Brides producing about 5 percent more than half the total yield). Of the flowers produced by Killarney, 21,297 were pro- duced from own-root stock and 23,838 from grafted. Of the Brides, 23,019 were produced from own-root stock and 26,859 from grafted. Results indicate, then, that in production of flowers the plants rank as follows: grafted Bride, grafted Killarney, own-root Bride, own- root Killarney. A summary of the results obtained by applying commercial ferti- lizers to the) sections in differing amounts is given in Tables 3 and 4, showing the number of flowers produced during 1910-13 in the season when records were kept (Table 3), and the total stem length during the same time (Table 4). In these tables, which also give the amounts of the fertilizers applied to each section, it is shown that three sections (4, 10, 16) had no commercial fertilizer applied, while Sections 1, 7, and 13 had applied to each, 8 pounds of dried blood, 2 pounds of acid phosphate, and 2 pounds of potassium sulfate per 100 square feet of bench space (5 inches deep). The remaining sections received appli- cations of dried blood varying from 8 to 32 pounds, of acid phosphate from 2 to 8 pounds, and of potassium sulfate from 2 to 8 pounds per 100 square feet in various combinations. A comparison of the average results from the unfertilized sections with those from all fertilized sections is presented in Table 5. From these figures, in spite of the fact that the more heavily fertilized sections were somewhat injured by overfeeding, it is evident that the soil was not capable of producing a maximum crop without fertilization. COMPARATIVE EFFECTS UPON OWN-ROOT AND GRAFTED STOCK From Table 2 it is seen that the 960 plants of grafted Killarney produced 23,838 flowers, compared with 21,297 flowers from the same number of plants of own-root stock. Similarly, 960 grafted Bride plants produced 26,859 flowers, compared with 23,019 flowers from the same number of own-root Brides. The balance in favor of grafted Killarneys is 2,541 flowers (from 960 plants), and in favor of grafted Brides is 3,840 flowers. The average length of stem of own-root Killarneys was 10.6 inches and of grafted Killarneys 10.5 inches, while that of own-root Brides was 14.6 inches, compared with 14.9 inches, the average length of stem of grafted Brides. The quality of the flowers, measured in this manner, was about the same in the own-root and grafted Killarneys, while the quality of grafted Brides was some- what better than that of own-root Brides. The important question of whether it pays to grow grafted stock, with the larger initial invest- ment, can be answered within the conditions of the experiment by a study of these figures and the wholesale prices for flowers of such 516 BULLETIN No. 196 [February, 1 _. m co in co * 00 rH 3 ^ i' 1 " O OS b- b- co to co oo to O b- rH CO i 1 OO O OO 00 00 * cq cq cq oo eo rH os eo CO Cq in rH co os os in co co b- oo * * to * oq oq cq oq * ro O CO Oq rH rH rH p tJO rH rH-rH CO rH rH Oq tO to o b- co rH w (D CO CO CO OS rH O O3 rH 03 rH 11 rH I ** to b- o cq oco to * to eo I-H o rH OS b- b- o co cq o 2 {T* (Ti ' o o oq co rHi> q iq OS rHOQ rH s r~* ( i O rH^H 10 t* ^ rH CQ rH IOCS M ^< cq OS ^ CO CO CO CD CO cq OQ oq oq co 10 eo ^ 4S co os b- Cq rH rH o b- to eo cq oq oq oq .-* rH i-H i-H * CO CO tO b- i M rH i-H rH <* 1 0? 10 OS IO OS to os cq b- rH^b--* CO b- OS CO b- O tO O o * TJ c 0> o o o o ^ OS ^ b 00 C4 to CD to oco os co to b- co oq oq oq oq ^ T^ OS O3 M 3 CO CD CD OS CO CO b- O co oo eo b- O CO b- O CO CO CO CO rh rH rH rH IO OJ rH i-H IO if 'S cq cq cq co 2 H ti JJH a ^_^ Qj ^_^ B 09 9 3 1 00 I 2 1 os oq toco l| O t- O b- tq os_ to os oq' oq' os *' 0) s i to cq o o oq co' eo' co" 3 8 ffl os oq b- co b-; O OS oq to cq rH eo' W 03 rH r ^ ^ ^4 CO cq in co I-H & *H o os to m cq oq cq cq O .rH o os oo os 3 (-M o cocoa fr b- 00 * to co o eo r^ ^ co cq cq cq ^ bfl r rH rH rH <* fl " rt os oo oo co -J fl "*^ g O 13 i-H rH i-H in t> t3 ^i ^ OQ oo T3 ^ O2 m a at 01 oa a" Tl a "cS 4-3 Q^ to o o to to os o * 1 rH O CO O i l O> CO rH IO rH Tt< cq 00 to -2 (z; O O O oq cq cq co a to co co os oo ** cq in EH H OS 00 rj( cq oo co os to 0) hn t> b^OS 00 oq cq oq cq EH pq b- tO O OS IO OS os CD co os to to oo co u o oo o cq cq cq co oo TJI b- os tO b- 00 rH * oq os co co co eo eo co co ^ co cq cq cq oq oo ^ to os to eo * <* fe co co co o) CO IO b- O os o co co co co co co ^ b~ b* b* eo b- b- O in OT oq rH O i-H CO g rH rH rH CO a O OS tO * o o cq co rH 00 CO 00 OS OS CO * B o o o o oq oq ca to co co co os OS CO CO 00 rH Cq OS CO O CO rH 00 OS ^Jt rH "^ CO CO ^ "^ cq OS b- OS toco to * oq oq oq o oo" o co* cvi oo co co * S s* 8 ta J cococococococococooocococococoocooco m c oo in o O /"N S .^ PH O> fQ rj cd COCOCOOCOOOCOOCOCOOOCOCOCOCOC^COCO o * CD c 5 ti t- *IH -l-l 81 gi & S S O O5 CO* CO* Co' OJ O * Tji * CD Tj" l>* , ^ //Jt ^o ^ y \ } V * /; / i i * i % j \ 1 / 10< 1 / / ; i L >U fe i i 5 \ \ 1 / I / ^ V* ~)ki y >N X- ^ ^ ^ / _ - _- . * --- "'' 921 \ \ *0 -~-~ --- 1 Xi. \ 1 Li TQ ^ &vor\ = f/fotv- WJ- 9UOf- PXON- 2'PO- s o S \ S 8 sudMOijjo jaquunN 530 BULLETIN No. 15)6 [February, g 02 1917} USE OF COMMERCIAL FERTILIZERS IN GROWING ROSES 531 ^ ir> i!> -N--- "AS~ "b ^ - Q> Qj - 0-) C^ 8l" - ^ cx>- .> ^ J -C: _ V. ft) 55 cu J I. <^ \ <^ \ -} -> <0 <^g sjdMoyjoJdqiMnN 7 VMW- i 4 6 6 10 12 14 16 13 20 22 4 26 28 30 3d 34 Number of Weeks ,-- ?-- r *""" " -- .. '-. x ~y ^^* -J "- --. i //// y t V / 'A t ^ y 1 *C r S 5 C j 3 s j- ^ 5 J 5 s 1 2 x 5 5 i i \ \ ^ > (1 > > *^. ^ h. s A 7 VJt t" I V (j , / > ^. V S \ f \ j A *2S y- ! **+ d s K 1 R IfV / f < *: // f \ \ > i '/ ) OfO \ 9/7 t \ { / \ ^ ,-' ^ R ^ ...... -\ we I V- s. > \ i> n * 1917] USE OF COMMERCIAL FERTILIZERS IN GROWING ROSES 533 " V \ X I \ % o ^ A 534 BULLETIN No. 196 [February, V \ 1917] USE OF COMMERCIAL FERTILIZERS IN GROWING ROSES 535 -i^> j ft \ xfl 4 A \ 536 BULLETIN No. 196 [February, {0 S 1* :$ >" S* <^<>- >" V !O y o o> ^ ir> < LJ \ 02 ^ -C) OJ vo Is w -- g fe% S \ %* 03 N - ^ } > L H 7.0/7] USE OF COMMERCIAL FERTILIZERS IN ROWING ROSES 539 \ . -" * *- \ "-x^^ c\, x' ^ *! \ \ / o / \ ^5 ^ <> \ \ fy V / / 5P ^ )^s ^ a ^ s \ v ^ t > / * * U . / ^ - io rr> ^ "--s s ^ "i* | ^" \ *?fe fe! & X ^y Q c b . ^ in 1 / g &E + 7 I ^^: & P, i / a ^t & ; \ \ \ | y ^ c C\j ^ 1 / X "^> -4 \ t ^> w ) c Ci E \ \ X Xx S > i A / ^ k \ \, ^-^ -^ ^ c / ^.-* -- ^' / S X \ l> X \ "X s' < 1 C Q c V b D c ^ i t c r i si 1 \ 540 BULLETIN No. 196 [February, 1917] USE OP COMMERCIAL FERTILIZERS IN GROWING ROSES 541 DISCUSSION OF KESULTS A study of th'e effect of the application of these commercial fertilizers upon the weekly production of flowers shows that the appli- cation of dried blood was harmful in the fall, without result during th^ winter, and beneficial in the spring. Acid phosphate, on the other hand, gave a consistent increase in production thruout the year. This latter result is important because of the increased value of roses during the winter months. It points also to the advisability of mixing this fertilizer with the soil before filling the benches. DESCRIPTION OF THE EXPERIMENT, 1913-15 The experiments of 1910-13 showed acid phosphate to be, on the whole, the most profitable of the commercial fertilizers used, while a need for nitrogenous fertilizer amounting to about eight pounds of dried blood per 100 square feet of bench space, particularly in the spring of the year, was made evident also. With regard to acid phosphate, the maximum application that might be made with subse- quent profit and safety could not be determined from the data, since the largest amounts employed still gave increased returns over those from the use of smaller quantities. It was considered advisable in the experimental work of 1913-15 tc study the use of acid phosphate in more detail, varying quite widely the quantities applied to the different sections in order to reach the limit of practical application, and increasing the number of plants grown under each treatment by decreasing the number of treatments, in order that the averages of the data might afford a more reliable comparison. In addition, records were kept of the individual produc- tion of a number of moderately fertilized plants, for the purpose of calculating the probable error involved in comparing the production from the different sections. Certainty was given further by using a series of six treatments in which the quantity of acid phosphate was successively increased. To obviate that error due to unequal con- ditions of illumination, temperature, and humidity in different parts of the greenhouses, six repetitions of the treatments in each series were made progressively thru the house, and data for comparison were secured from the set of six thus differently located. Ammonium sul- fate (approximately 21 percent nitrogen) was used in the place of dried blood in these experiments. To half the sections used for each treatment finely ground limestone was applied as top-dressings, or mixed with the soil, in order to test the efficiency of this material in fertilizing roses 1 . MJpon acid foils (Eept. N. J. Sta. 1893 seq.) air-slaked lime has been found to benefit sweet peas, comet asters, poppies, and legumes, while dilute solutions of citric acid were found by Maxwell [Jour. Amer. Chem. Soc. 20 (1893) 103] to affect unfavorably the growth of some legumes and grasses, and all crucifers 542 BULLETIN No. 196 [February, The rose house used in this work measures 28 feet by 105 feet and contains 1,600 square feet of bench space -5 inches deep. It extends east and west and is a part of a range connected by a cross house at its west end, so that the east end is exposed. The temperature variations in different parts of the house proved rather complex under changing conditions of force and direction of wind, sunshine, outside temperature, heating arrangements, and methods of ventilation, but, as an average, the temperature of the inner end of the house was ap- proximately five degrees above that of the exposed end, during cold weather. Each bench was divided into eighteen 5-foot sections, separated from each other by a double partition with a 2-inch air space between, and with a "buffer" section about 3 feet long at each end growing roses upon which no records were kept. These end sections were not fertilized 1 and served as a sort of check on the experimental sections, altho conditions were hardly the same on account of the more rapid drying of the soil at the ends of the benches. Killarneys were grown in the two north benches, and Eichmonds in the south; each section, which held four rows of plants 12x15 inches apart, contained two rows of own-root and two of grafted stock, or eight plants of each type. The. plants were all first-year stock, potted into 4-inch pots on March 2-14 and into 5-inch pots on June 5-9. The plants were set in the benches on August 4, 1913, and on July 9, 1914. Each two benches, comprising thirty-six sections in all, were made up of six sets of six sections each, numbered in rotation from one bench to another. In numbering the sections the first (or unit) figure was carried from 1 to 6 and each represented a treatment. The second (or ten) figure, from to 5, indicated the location of the section in the house. The arrangement of the sections alternated progressively. The soil used during these years was of the same type and was prepared in the same manner as that described on page 512. Fertilizers were applied in 1913-14 in the amounts given in Table 15. The whole of the manure and the potassium sulfate was applied at the time of preparation of the soil (August 4, 1913). One-third of the amount of acid phosphate and ammonium sulfate was applied at this time, also, with another third on November 30, 1913, and the re- mainder on April 14, 1914. Limestone was used for top-dressings on August 9, 1913, and March 4, 1914. and clovers. On the other hand, azaleas, rhododendrons, begonias, lupines, some grasses, the heaths, gorse, broom, foxglove, vetches, etc. [Ibid and Abs. Jour. Chem. Soc. (Lond.) II (1909) 429] have been found by experiment or experience of gardeners to grow better in acid soils. Lime or limestone is often recommended for roses, carnations, and chrysanthemums, yet, so far as the author is aware, there is no extensive experimental work proving that it is a benefit to these crops. J In 1913-14 a small amount of manure was applied. 1917] USE OF COMMERCIAL FERTILIZERS IN GROWING ROSES 543 TABLE 15. APPLICATIONS OF FERTILIZERS TO ROSES, 1913-14 (Pounds per 100 square feet of bench space 1 ) Section Manure Ammonium sulfate A.cid phosphate Potassium sulfate Limestone* 1-1 1-2 1-3 1-4 1-5 1-6 115 115 115 115 115 115 15 15 15 15 15 15 10 20 40 80 160 2 2 2 2 2 2 10 10 10 10 10 10 ^Pounds per 100 square feet of bench space multiplied by 2 gives approximate pounds per 100 cubic feet, and when divided by 5 gives the application in tons per acre. 'Applied to alternate series only, viz., -0, -2-, -4-. In addition to obtaining the data regarding the production of flowers and their quality, which were secured during 1913-14, it was observed that the end sections, to which no fertilizer except manure had been applied, grew plants up to the middle of the winter with no signs of nitrogen starvation, which can be so easily detected by the unhealthy growth and the yellow color of the foliage. The second ap- plication of ammonium sulfate (November 30, 1913) was soon followed by a prolonged period of cloudy weather. While no signs of injury were noticed on the well-developed foliage, the young growth, which came on about January 1, showed marked chlorosis (whitening) of the leaflets, and in many cases drooping and blackening at the ends of the young shoots. This is to be distinguished from the dropping of leaves as a result of a disturbance of the root system, which has been observed to follow too deep cultivation, but which affects the oldest leaves first, while injury from overfeeding reaches the tender, most rapidly growing portions of the plant. As a result of overfeeding with ammonium sulfate at this time, no accurate records were secured during the midseason of the year. Tests made upon' the soil at the time of the second top-dressing of limestone (March 4, 1914) showed the soil to be quite acid at any depth greater than one-half inch below the surface. In the upper half-inch the soil seemed acid or neutral, depending on the evenness of the previous application of limestone. It was evident that top-dressings with limestone had not corrected acid- ity in that portion of the soil penetrated by the roots. The top-dress- ing with acid phosphate caused a wide-spread surface growth of roots, instead of a more desirable penetration of the entire soil by the root system. In fertilizing the soil for the experiment of 1914-15, the am- monium sulfate was applied on July 8, 1914 and April 27, 1915, omitting the midwinter application ; limestone was worked into the soil before setting in the plants, in addition to a top-dressing in the spring, while in Sections 2, 3, and 4 of each series, two-thirds of the acid 544 BULLETIN No. 196 [February, phosphate was worked into the soil before setting in the plants. No potassium sulfate was used. The schedule of fertilizers is given in Table 16. TABLE 16. APPLICATIONS OF FERTILIZERS TO KOSES, 1914-15 (Pounds per 100 square feet of bench space) Section Manure Ammonium &ulfate Acid phosphate 1 Limestone* 1-1 1-2 1-3 1-4 1-5 1-6 115 115 115 115 115 115 10 10 10 10 10 10 10 20 40 80 160 20 20 20 20 20 20 Applied July 8, December 21, and April 27 to Sections 1-5 and 1-6; on July 8 and April 27 to Sections 1-2, 1-3, and 1-4. 'Applied only to series -0-, -2-, -4-. The growth of the roses was satisfactory thruout the year except- ing that, judging from the color of the plants, the last application of ammonium sulfate should have been made about a week earlier. Records for both years (1913-14, 1914-15) were taken daily, excepting Sunday, upon the number of flowers produced in each sec- tion and the stem length of each flower. They were also grouped into classes according to stem length, firsts being those with a stem length over 18 inches; seconds, 12 to 18 inches; thirds, 6 to 12 inches; and fourths, under 6 inches. The roses in Sections 1 and 6 of each set were allowed to remain on the plant until fully open, and the size of each was measured, in order to test the effect of acid phosphate on the length of the petals of the flowers. ON THE ACCURACY OF THE RESULTS Record was kept of the number and stem length of the flowers pro- duced by each plant in Sections 123 and 134 (Richmond) and Sec- tions 124 and 133 (Killarney) during the approximate periods from November 1 to June 1, 1913-14, and from October 1 to June 1, 1914-15. The results are given in Table 17. Plants 1 to 8 were own-root and Plants 9 to 16 grafted stock in each case. The data in Table 17 were used for determining to what extent differences between the results from sections compared were due to the variation in production of individual plants and not to the influ- ence of the treatment. It is necessary to assume that the variation among plants in the moderately fertilized sections is repre- sentative of that in other sections, that is, that neither low nor high fertilizing affected the variability in production among the plants. The maximum standard deviation obtained from any set of 1917} USE OF COMMERCIAL FERTILIZERS IN GROWING SOSES 545 TABLE 17. INDIVIDUAL PRODUCTION BY EOSE PLANTS Killarney Eichmond Plant Number of flowers Plant Number of flowers No. Section 133 Section 124 No. Section 134 Section 123 1913-14 |1914-15 1913-14 1 1914-15 1913-14 1914-15 1913-14 11914-15 1 32 43 26 45 1 29 32 8 32 2 19 37 19 14 2 17 15 14 23 3 19 27 17 35 3 11 25 17 22 4 16 47 24 39 4 22 17 13 25 5 22 44 29 42 5 15 26 23 31 6 ' 20 27 24 37 6 13 22 15 15 7 15 25 14 29 7 23 15 18 27 8 23 60 19 28 8 16 32 10 27 Total Total (own- (own- root) 166 310 172 269 root) 146 184 118 202 9 38 35 29 39 9 40 30 21 49 10 18 42 25 33 10 24 37 25 26 11 36 32 18 45 11 19 34 20 34 12 31 40 34 28 12 41 34 27 30 13 34 30 46 51 13 30 43 29 40 14 20 38 30 41 14 42 38 30 29 15 27 32 18 27 15 26 25 19 26 16 19 39 24 47 16 26 46 34 29 Total Total (grafted) 223 288 224 311 (grafted) 248 287 205 263 NOTE. These sections were chosen because they received moderate applications of acid phosphate. 32 plants is only 1.09 higher than the standard deviation found by grouping the four sets of 32 each, and the minimum from any set of 32 plants is 3.9 lower than the standard deviation from the combined group. It would thus seem reasonable that in using 10.68 for a stand- ard deviation to predict a probable error for the entire experiment, the prediction would tend to be too large rather than too small. If resulting differences are found significant under a criterion which used values too large for the probable error, the differences would be all the more significant with a more accurate estimate. While the accuracy that might have been secured by an individual record of production for each plant is not to be had after these assumptions, the figure for probable error thus secured is an indication of the relia- bility of the results. The frequency . distribution for each variety of root stock is shown in Table 18, which contains also the arithmetical means and their probable errors, the standard deviations, and the coeffi- cients of variability 1 . The frequency distribution for Killarney and Richmond roses is shown graphically in Fig. 26. Calculations are made according to methods described in 111. Agr. Exp. Sta. Bui. 119 (1907), and Jpur. Agr. Sci. vol. 3 (1910), p. 417. 546 BULLETIN No. 196 [February, TABLE 18. FREQUENCY DISTRIBUTION OF BOSE PRODUCTION, 1913-1915 Bichmond Killarney Total No. of flowers Number of plants No. of flowers Number of plants Own-root | Grafted Both Own-root | Graf ted | Both 7-14 15-21 22-28 29-35 36-42 43-49 6 10 11 5 4 8 11 6 3 6 14 19 16 6 3 6-15 16-25 26-35 36-45 46-55 56-65 3 12 8 7 1 1 7 13 9 3 3 19 21 16 4 1 14 39 45 23 6 1 M 20.80 30.60 25.70 28.10 32.50 30.30 27.70 E .81 -K99 +.75 1.39 + 1.13 -K89 .63 D 6.88 8.36 8.93 11.77 9.16 10.68 10.68 G 33.10 27.30 33.70 41.90 28.10 35.30 38.60 E' 45.50 55.30 41.70 77.70 60.60 50.00 35.30 NOTE. The following abbreviations are used: M (Mean); E Error); D (Standard Deviation); and C (Coefficient of Variability). (Probable The production by any two treatments is nearly enough the same that a probable error (E') for use in a comparison of them may be secured by multiplying the corresponding E of Table 18 by V2 and dividing by the square root of the number of plants from which data were secured for the comparison. Since, in the succeeding tables, the production of all plants in a single treatment (48) is used as a basis Z4 n Killarney ~> $ ?> j? ^i ^ Richmond Number off/overs FIG. 26, FREQUENCY DISTRIBUTION OF PRODUCTION, 1913-1915 USE OF COMMERCIAL FERTILIZERS IN GROWING ROSES 547 Z80 \ 180 loo \ Kiltarney / Z 3 4 5 6 7 8 9 10 II IS 13 Length of Stems in Inches /s /6 17 /a t9 eo sr FIG. 27. FREQUENCY DISTRIBUTION OF STEM LENGTH IN KILLARNEY ROSES 6 10 14 16 18 20 Lengfh of Stems in /nches 26 36 30 JZ 34 FIG. 28.- FREQUENCY DISTRIBUTION OF STEM LENGTH IN RICHMOND ROSES 548 BULLETIN No. 196 [February, for comparison, the value for E' so secured is multiplied by 48 in order to obtain a probable error expressed in the same unit as the values to be compared. Thus, if the difference in production per 48 plants in two treatments of 48 plants each amounted to 45.5 flowers (own-root Richmond), the chances would be even that the difference resulted from the treatments given the series and not from variation in productivity of the plants chosen for the experiment. In the case of the combined data from both varieties and types of root stock, which were secured from 192 plants in each treatment, the probable error in comparison of average production per 48 plants amounts to 35.3 flowers. The frequency distribution of stem length of the flowers produced (Figs. 27 and 28) is given in Table 19, which also presents the mean stem lengths with their probable errors, the standard deviations, and the coefficients of variability. E' (season) and E' (year) are calcu- lated upon a basis of comparison between two sets of 400 flowers and 1,200 flowers respectively, these being the approximate productions for the season and the year. The results calculated for probable errors indicate that in com- paring two series of plants (48 plants each) which have received dif- ferent applications of fertilizer, the difference in average stem length of flowers must equal from. 117 to .258 inches, depending on the variety FIG. 29- -TYPES OF CURVES INDICATING THE EFFECT OF FERTILIZER APPLICATIONS UPON FLOWER PRODUCTION 1917] 549 TABLE 19. FREQUENCY STEM-LENGTH DISTRIBUTION OF ROSES Stem Number of flowers length Killarney Richmond Tntal (inches) Own-root Grafted Both Own-root | Grafted Both LOMU .6-1.5 . . . 1 1 1 1.6-2.5 1 "3 "4 2 8 10 14 2.6-3.5 7 13 20 19 29 48 68 3.6-4.5 14 23 37 30 51 81 118 4.6-5.5 26 32 58 67 78 145 203 5.6-6.5 60 63 123 58 66 124 247 6.6-7.5 71 80 151 47 81 128 279 7.6-8.5 79 95 174 48 79 127 301 8.6-9.5 95 135 230 38 60 98 328 9.6-10.5 107 130 237 46 70 116 353 10.6-11.5 86 126 212 44 68 112 324 11.6-12.5 94 95 189 54 78 132 321 12.6-13.5 66 94 160 25 67 92 252 13.6-14.5 69 66 135 37 75 112 247 14.6-15.5 44 37 81 26 59 85 166 15.6-16.5 28 17 45 22 43 65 110 16.6-17.5 21 15 36 18 28 46 82 17.6-18.5 13 9 22 16 21 37 59 18.6-19.5 5 4 9 15 20 35 44 19.6-20.5 8 2 10 9 9 18 28 20.6-21.5 6 2 8 8 11 19 27 21.6-22.5 3 2 5 13 4. 17 22 22.6-23.5 1 1 2 7 2 9 11 23.6-24.5 8 3 11 11 24.6-25.5 1 1 2 3 2 5 7 25.6-26.5 3 3 3 26.6-27.5 1 1 2 2 4 5 27.6-28.5 . . . 1 1 1 28.6-29.5 1 1 1 29.6-30.5 ... . . . 30.6-31.5 ... ... 1 1 1 Total 906 1045 1951 667 1016 1638 3634 M 10.7500 10.2100 10.4600 10.830 10.520 10.6500 10.5500 E .1130 -K0960 .0740 .199 .140 .1140 -K0660 D 3.6300 3.3300 3.4700 5.480 4.740 5.0100 4.2600 C 33.6000 32.8000 33.2000 50.600 45.100 47.0000 40.3000 E'( Season) .1730 .1590 .1170 .258 .227 -K1690 .1030 E' (Year) .0998 +.0918 +.0676 .151 .131 .0974 .0586 or type compared (see Table 19) before the chances become equal that the difference is not due to the error inherent in a comparison of two averages secured from the measurement of a limited number of flow- ers. In comparing the average stem length of a year's production from two treatments or in the combined data from the different vari- eties and types of root stock, the probable error is decreased because of the larger number of flowers used in making the comparison. It was thought likely that a simple relation could be secured be- tween the relative production of the sections and their positions in the house, on account of unequal conditions of temperature and humid- 550 BULLETIN No. 196 [February, ity at different places in the house. The study of these conditions proved more complex than a simple drop in temperature and rise in relative humidity toward the exposed end of the house, however, and no progressive rise or drop in production in the series could be ob- tained. In interpreting the results from a series of sections upon which successively increasing amounts of acid phosphate had been applied, a curve of one of five types would be expected, as shown in Fig. 29. A curve of the first type would indicate that neither benefit nor injury resulted from application of the fertilizer; of the second, injury; of the third, continued benefit up to the maximum applica- tion according to Mitscherlich 's Law of Diminishing Returns 1 ; of the fourth, benefit up to a certain maximum application, with neither increased yield nor decrease from injury with greater amounts; of the fifth, increased yield up to a maximum, with a decrease thereafter due to injury from overapplication of the fertilizer. EFFECT OF ACID PHOSPHATE ON WEEKLY AND YEARLY PRODUCTION A record of the data secured from the experiment, arranged in seasons, and totaled for the year, is given in Table 20 for 1913-14, and in Table 21 for 1914-15. In Table 20 the records of the second season are not included, except in the totals, since the roses were injured during the greater part of the season by overfeeding with ammonium sulfate. TABLE 20. NUMBER OF FLOWERS PRODUCED, 1913-14 (Average per 48 plants) Section Acid phosphate (Ibs.per 100 sq.ft.) Own-root Grafted Season I | Season III | Year Season I | Season III ) Year Killarney 1 358 493 1065 428 616 1308 2 10 390 619 1233 470 708 1460 3 20 369 620 1241 520 760 1599 4 40 382 650 1264 529 793 1594 5 80 405 645 1300 478 811 1584 6 160 385 612 1244 472 790 1549 Eichmond 1 467 479 1085 571 697 1450 2 10 487 565 1175 583 814 1581 3 20 472 605 1219 624 845 1674 4 40 466 631 1239 603 867 1722 5 80 484 650 1284 555 819 1567 6 160 470 553 1181 535 798 1549 Russell, Soil Conditions and Plant Growth (1912), page 24 (Longmans, Green, and Company, Monographs on Biochemistry). 1917] USE OF COMMERCIAL FERTILIZERS IN GROWING EOSES 551 TABLE 21. NUMBER AND QUALITY OF FLOWERS PRODUCED, 1914-15 Section Acid phosphate (Ibs. per 100 sq. ft.) Production of flowers 1 (per 48 plants) Aver- age size of flowers 2 Stem length Total number 1st - 2nd 3rd 4th Total Aver- age Own-root Killarney Season I 1 2 3 4 5 6 10 20 40 80 160 467 497 537 512 555 495 percent 0.8 0.8 1.2 1.5 1.8 1.0 percent 23.8 30.3 29.0 28.4 29.4 23.5 percent 63.1 60.1 64.0 63.6 59.9 62.1 percent 12.2 8.8 5.7 6.4 9.1 13.5 inches 4.53 4.51 inches 4476 ..5037 5595 5232 5729 4708 inches 9.58 10.13 10.41 10.22 10.32 9.51 Season II 1 2 3 4 5 6 10 20 40 80 160 318 370 382 380 391 326 2.5 4.0 4.7 3.4 4.1 1.8 37.8 41.9 43.2 38.7 39.0 35.0 57,0 52.7 51.4 54.5 53.8 56.0 2.8 1.3 .7 3.4 2.1 6.1 4.48 4.35 3637 4408 4498 4346 4548 3579 11.43 11.91 11.25 11.43 11.63 10.97 Season III 1 2 3 4 5 6 10 20 40 80 160 547 585 678 616 644 564 2.2 1.2 2.6 4.2 3.5 1.6 30.8 35.4 36.1 33.4 37.9 29.4 62.1 63.6 58.7 59.4 56.9 61.9 4.7 3.5 3.3 3.4 2.1 4.4 4.23 4.25 5906 6211 7606 6875 7323 5994 10.79 10.61 11.21 11.16 11.43 10.62 Year 1 2 3 4 5 6 10 20 40 80 160 1332 1452 1598 1508 1590 1385 1.8 1.8 2.7 3.1 3.0 1.4 30,0 35.0 35.4 33.1 36.4 28.6 61.2 60.5 57.8 59.0 56.5 64.5 6.9 4.8 3.5 4.4 4.5 8.8 4.39 4.37 14020 15656 17699 16454 17600 14282 10.52 10.78 11.07 10.91 11.07 10.31 Grafted Killarney Season I 1 2 3 4 5 6 10 20 40 80 160 600 608 604 579 598 639 0.3 0.1 0.9 0.3 0.1 0.0 25.0 22.9 24.4 20.6 18.4 13.0 65.1 67.1 64.5 70.8 70.0 73.2 9.3 9.8 10.2 8.6 11.7 13.9 4.47 4.50 5890 6012 5898 5544 5527 5533 9.81 9.88 9.76 9.58 9.24 8.66 Season II 1 378 1.0 32.9 63.5 2.7 4.38 4117 10.89 2 10 401 2.2 32.2 63.7 1.9 4452 11.10 3 20 448 2.7 39.0 55.4 2.9 5083 11.34 4 40 446 1.8 39.4 56.2 2.6 5040 11.30 5 80 455 .6 31.6 69.5 2.4 4971 10.95 6 160 383 .5 30.6 64.2 3.7 4.35 4063 10.61 J Flowers are grouped in classes as follows: 1st, length 18 inches and over; 2nd, 12 to 18 inches; 3rd, 6 to 12 inches; 4th, under 6 inches. 2 Size was determined of flowers from Sections 1 and 6 only, 552 BULLETIN No. 196 [February, TABLE 21. Continued Section Acid phosphate (Ibs. per 100 sq. ft.) Production of flowers (per 48 plants) Aver- age size of flowers Stem length Total number 1st 2nd 3rd 4th Total Aver- age Season III percent percent percent percent inches inches inches 1 668 1.5 29.6 64.2 4.4 4.21 6977 10.44 2 10 744 1.6 28.8 65.8 3.6 7840 10.53 3 20 736 1.7 32.8 61.9 3.3 8045 10.93 4 -40 732 1.3 36.6 58.1 3.8 8056 11.00 5 80 793 1.1 27.7 66.8 4.2 8332 10.50 G 160 784 .6 22.7 69.7 7.1 4.22 7849 10.01 Year 1 1646 1.0 28.7 64.6 5.8 4.34 16985 10.32 2 10 1753 1.3 27.5 65.8 5.4 18304 10.44 3 20 1788 1.7 31.5 61.3 5.6 19027 10.64 4 40 1757 1.1 31.9 61.8 5.1 18640 10.61 5 80 1846 .7 25.7 62.4 6.2 18831 10.20 6 160 1806 .4 21.0 69.9 8.8 4.35 17445 9.66 Own-root Richmond Season I 1 429 2.7 21.4 53.8 21.6 3.65 3951 9.21 2 10 431 7.6 21.5 50.9 19.9 4334 10.05 3 20 511 4.2 25.3 50.5 20.2 4983 9.75 4 40 501 6.3 22.2 49.5 22.0 4899 9.78 5 80 485 4.7 20.4 51.2 23.8 4532 9.34 6 160 518 5.2 24.7 50.5 19.8 3.65 4998 9.64 Season II 1 2 3 4 5 6 10 20 40 80 160 225 254 279 273 236 270 10.2 12.6 9.2 16.7 13.5 13.3 38.2 43.7 43.0 41.2 41.2 40.8 41.7 36.6 41.7 35.2 35.5 36.6 9.7 7.0 5.7 6.9 10.1 9.3 3.60 3.76 2645 3176 3376 3525 2906 3352 11.76 12.50 12.10 12.91 12.31 12.41 Season III 1 2 3 4 5 6 10 20 40 80 160 412 419 492 443 417 462 15.2 12.8 13.6 15.3 16.7 10.3 21.3 32.6 28.8 29.3 33.5 30.9 40.7 40.5 43.2 41.5 39.0 44.5 22.5 13.8 14.2 13.7 10.5 14.0 3.65 3.83 4633 4963 5808 5313 5215 5278 11.24 11.84 11.80 11.99 12.50 11.42 Year 1 2 3 4 5 6 10 20 40 80 160 1066 1104 1282 1217 1138 1250 9.4 10.8 9.7 12.0 11.0 8.9 24.8 30.9 30.4 29.1 29.5 30.5 46.1 43.7 45.8 43.5 43.4 45.3 19.1 14.7 14.8 15.6 16.1 15.4 3.64 3.83 11230 12474 14168 13738 12655 13628 10.53 11.29 11.05 11.20 11.12 10.90 Grafted Richmond Season I 1 2 3 4 5 6 10 20 40 80 160 490 603 630 617 603 590 7.7 7.1 7.4 5.5 5.8 2.8 30.6 35.4 28.0 27.4 27.2 18.1 47.8 47.3 51.2 49.2 51.5 54.6 14.3 10.2 13.5 17.8 15.5 24.4 3.71 3.59 5277 6758 6781 6257 6354 5272 10.77 11.20 10.76 10.12 10.36 8.93 NOTE. See footnotes 1 and 2, page 551. 1917] USE OF COMMERCIAL FERTILIZERS IN GROWING ROSES 553 TABLE 21. Concluded Section Acid phosphate (Ibs. per 100 sq. ft.) Production of flowers (per 48 plants) Aver- age size of flowers Stem length Total number 1st 2nd 3rd 4th Total Aver- age Season II percent percent percent percent inches inches inches 1 347 11.0 43.6 36.4 8.9 3.71 4326 12.46 2 10 375 10.4 46.5 37.8 5.3 4772 12.72 3 20 375 15.2 42.1 36.2 6.4 4828 12.87 4 40 402 8.7 48.2 35.0 7.9 4981 12.38 5 80 385 12.2 42.6 36.0 9.3 4740 12.31 6 160 360 6.1 36.3 44.1 13.3 3.71 3988 11.07 Season III 1 2 3 4 5 6 10 20 40 80 160 533 615 605 656 677 638 4.3 8.5 9.3 8.1 8.2 4.6 32.9 36.3 36.2 36.1 30.2 28.2 44.5 42.0 43.1 41.5 47.0 46.0 18.4 12.8 11.2 13.2 14.1 21.9 3.68 3.76 5451 7014 6935 7426 7486 6438 10.22 11.40 11.46 11.32 11.06 10.09 Year 1 2 3 4 5 6 10 20 40 80 160 1370 1593 1610 1675 1665 1588 7.2 8.7 9.9 7.3 8.5 4.6 34.7 38.4 34.4 35.8 32.1 26.3 43.6 42.7 43.7 43.2 46.4 49.8 14.5 10.1 11.0 13.7 13.5 20.3 3.70 3.68 15055 18545 18546 18665 18581 15699 10.98 11.64 11.51 11.74 11.16 9.88 NOTE. See footnotes 1 and 2, page 551. EFFECT OF ACID PHOSPHATE ON TOTAL YIELD OF FLOWERS The data upon the effect of acid phosphate on the yield of flowers arc arranged graphically in Figs. 30 and 31. The curves are consistent in showing an increased production as a result. of the smallest applica- tion of acid phosphate, the tendency toward a rise being the least pronounced in the first season. With succeeding applications the results are more or less definite in showing a smaller proportional increase up to the largest application, when a decrease is shown, as a rule. Since the curves are of the same type for both own-root and grafted stock and for each variety, it is permissible to use the data for both years and all types of plants in describing a curve from which to draw conclusions applicable equally to own-root and grafted stock, and to Killarney and Richmond varieties. The data averaged thus are arranged in Table 22 (production per 48 plants) and the grand average is described in Fig. 31. The excess in number of flowers produced in Section 3 over the yield from Section 1 (211 flowers per 48 plants) stands in the ratio 211 of - (=6) to the probable error obtained from the data given 35.3 554 BULLETIN No. 196 [February, 1700 660 1650 seo (600 760 /5SO 740 I4OO KO 1350 580 1300 54O IZ50 500 1200 ~46O 1150 420 HOO 300 1050 340 IOOO 300 ^ X I- ' ' ' x Zl Sea* Seas rear ml vnM /I / 1 / - V. // , .'' \ t 7 / ' \ ! / / 7 \ j / \ , \ / / \ i / ('-'" "-- 7^~ ./' \ . / J x \ . x^ ..' Ss / I*-- "" Owr Root Killo ney 6ro ted ," Illorn ey o n Rtx )tf?iC non d 123456 1 Z 3456 IZ '3456 Mumber of Treatment 1700 860 I6SO KO 1600 780 1550 740 1500 700 1450 660 1400 620 1350 560 1300 540 500 IZOO 460 1/50 42O 1100 360 1050 340 IOOO 30O ^\ ^ / \N 1 / V *"^ ./> ' \ ^^ // / Si \ , ^z~- ^*< \, / "s \ l t f ^ \ ^ '!! Grof ed R chmc nd {/ ft * ^ Wlon ffichn 9o/h^ ley. to ior>d 'ariefit ?tr/3t. y v/ts.c. verve, wfor^ year 1 Z 3 4 5 6 123456 Number of Treatment FIG. 30. NUMBER OF EOSES PRODUCED, 1913-14 1917] USE OF COMMERCIAL FERTILIZERS IN GROWING EOSES 555 in Table 18. A difference in results six times the probable error cor- responds to odds of 19,200 to I 1 that it is due to the difference in treatment of the sections, hence, considerable reliance may be placed in the data after making allowance for the approximate value of the probable error. TABLE 22. AVERAGE YEARLY PRODUCTION OP FLOWERS, 1913-15 Section Acid phosphate (Ibs. per 100 sq. ft.) Richmond Killarney Grand average 1913-14 1914-15 Average 1913-14 1914-15 Average 1 2 3 4 5 6 10 20 40 80 160 1267 1378 1446 1480 1425 1365 1218 1348 1446 1446 1401 1419 1242.5 1363.0 1446.0 1463.0 1413.0 1392.0 1186 1346 1420 1429 1442 1396 1489 1602 1693 1632 1718 1595 1337.5 1474.0 1556.5 1530.5 1580.0 1495.5 1290 1418 1501 1497 1496 1444 Two points are clear from the data. Applications of acid phos- phate up to 20 pounds per 100 square feet of bench space (40 pounds per 100 cubic feet) cause an increase in the number of roses produced. Applications up to four times this amount cause neither further in- crease nor decrease in production, hence, there is a wide difference between the minimum quantity of acid phosphate that should be applied to produce the maximum crop, and that quantity which will cause injury from overfeeding. This is especially important since with most commercial fertilizers great care must be taken not to ruin the crop by excessive applications. The significance of these results is perhaps not apparent at first glance. Calculated on the basis of 1,000 plants, an excess in produc- tion of 4,400 flowers is obtained by the use of 250 pounds of acid phosphate. This fertilizer costs at the present time about fifteen dol- lars per ton, while a conservative price for roses of the quality obtained (averaging about an 11-inch stem) is four dollars per hundred. The cost of fertilizer and of the labor to mix it with the soil are insignifi- cant compared with the additional profit of $176 per 1,000 plants obtained by its use. EFFECT OF ACID PHOSPHATE ON STEM LENGTH AND SIZE The results of fertilization with acid phosphate upon total stem length are similar to those upon the number of flowers produced (see Table 21), that is, an increase with all applications excepting the heaviest, which caused a slight decrease. The relative increase and decrease of the two characters are not quite the same, however, as seen '111. Agr. Exp. Sta. Bui. 119, p. 15. 556 BULLETIN No. 196 [February, 1900 800 jets 750 nso 700 1675 650 1600 600 I5Z5 550 1450 500 /J75 450 1300 400 &Z5 350 1150 300 1075 ^50 1000 ZOO Season I Season I Season M Year \ NX s Own Root Richmond Grafted imey \z Own Root Killamey eoo /8Z5 7SO 1750 700 1675 650 J600 600 I isis (^ 550 \ 1450 k 500 | /375 450 ^ I3OO 400 ISSS 35O //SO 300 1075 50 /OOO .---* \ / V / \ A ;-^- ^~~ V ^ *. / \ /ll / i ^ >-*-' s II y / > / x^ '/ / / -/r- / ' < y ' / Grand Average 19:3-15 lorney and Rich aftecf 1 <=,_, k vnf7oot/ Slock .X x * ^-, ^ V / N Gr nond ( Jrofte d Ric hmor id K - Kil/amey.botft s/ocfrs, average 1914- /S Richmond ' " " ' " ' --Both Varieties* - o eoo 23456 Number of Treat menr FIG. 31. NUMBER OF EOSES PRODUCED, 1914-15 in the figures for average length of stem 1 . These bring out the fact that the average as well as total length of stem increases when acid phosphate is used as a fertilizer, unless in excessive amounts, so that quality as well as production is benefited. No influence of fertilizing is seen upon the percentage of flowers in Classes 1, 2, 3, and 4, grouped according to the length of stem. While the average stem length may be increased slightly upon increas- 1 The differences, while small, are well beyond the experimental error calculated in Table 19 and so are considered significant. 1917] USE OF COMMERCIAL FERTILIZERS IN GROWING EOSES 557 ing the production by the use of acid phosphate, no marked increase in the number of long-stemmed flowers is to be expected. Attention may be called, at this point, to the larger percentage of long-stemmed flowers in the Richmond variety than in Killarneys, particularly of own-root stock. It is worthy of mention that own-root stock, while producing fewer flowers, does yield as long-stemmed and as long- petaled roses as grafted stock. No consistent relation is to be found between the difference in average size of the flowers from Sections 1 (no acid phosphate) and 6 (large amount of acid phosphate) and the difference in their ferti- lization. Since no figure for probable error has been obtained, it is impossible to attribute the existing differences to the treatment rather than difficulties of measurement and inaccuracy from insufficient num- ber of flowers for comparison. It can only be said that no difference in size* of roses great enough to be apparent in these records is obtained by fertilizing with acid phosphate. RELATION OF INCREASE FROM THE USE OF ACID PHOSPHATE TO WEEKLY PRODUCTION The relation of the increase in production obtained by the use of acid phosphate is shown in Figs. 32 to 35, in which the weekly pro-' duction of flowers upon Section 1-1, to which no acid phosphate was applied, and upon that section of each variety and type of root stock giving the greatest return (Section 1-3, 1-4, or 1-5), was used in describing the graphs. The application of acid phosphate upon the sections chosen for comparison with Section 1-1 varied from 20 to 80 pounds per 100 square feet of bench space in different cases. In the figures, numbers of flowers are used as ordinates and weeks of the sea- sons as abscissae. A curve representing the combined data of all types of plants used is shown in Fig. 36. It is unnecessary to comment in detail on these curves, since a glance at them shows that the advantage from the use of acid phosphate, while greatest in the spring as the soil is depleted of its original supply of phosphate, is evident thruout the year. Such evidence supports that found in the previous experi- ment (page 528) based on results from the use of smaller quantities of acid phosphate. EFFECT OF LIMESTONE ON PRODUCTION OF FLOWERS A comparison of the number of flowers produced from sections to which lime had been applied and those unlimed for the year 1914-15, is given in Table 23. During 1913-14, the limestone, applied as top- dressings, did not penetrate far enough beneath the surface to affect the condition of the soil materially, hence, no comparison is made for the production of that year. 558 BULLETIN No. 196 [February, tl "Is ii "f 1. *c 1 2 > \ 7 \ Q 10 eks 4 6 8 Number of We 8 6 1917] USE OF COMMERCIAL FERTILIZERS IN GROWING EOSES 559 ^) &} 5050 i 1 : t= 7 i K QO ^o 6 & ^ ** 1 ^j ^* rv 11 ,^> ^5 1 <- \ f ^ < 8 6 B A) O t>S 1 J 5i ^^ o If \rs ^ f- ? <9 560 BULLETIN No. 196 [February, \ \ K. 1917] USE OF COMMERCIAL FERTILIZERS IN GROWING EOSES 561 TABLE 23. EFFECT OF LIMESTONE ON PRODUCTION OF ROSES, 1914-15 (Total number of flowers) Lime No Lime Killarney Richmond Killarney Richmond Own-root Grafted Own-root Grafted Own-root Grafted Own-root Grafted 4324 5293 3465 4532 4541 5303 3592 4969 9617 7997 9844 8561 17614 18405 It is evident from the data that those sections receiving no lime- stone produced more flowers. The results by section (representing a treatment with acid phosphate) for all types are given in Table 24 and arranged graphically in Fig. 37. In comparison of yields per 96 plants, it would be necessary to have a difference exceeding 70 flowers (see page 548) in order to draw accurate conclusions. This difference is found with the first three 3250 3?00 3150 3100 3050 \ h 3000 2900 i* c *Z600 I 2750 2700 2650 &00 2550 t 7 \ 1234 56 Section and Treatment Number FIG. 37. EFFECT OF LIMESTONE ON FLOWER PRODUCTION, 1914-15 562 BULLETIN No. 196 [February, TABLE 24. EFFECT OF LIMESTONE ON PRODUCTION OF ROSES, 1914-15 Yield per 96 i lants Section (Pounds per 100 square feet) Lime No lime Difference due to lime 1 2 3 4 5 6 10 20 40 80 160 2557 2809 3043 3106 3050 3049 2857 3093 3235 3051 3189 2980 -300 -284 -192 + 55 -139 -f 69 treatments of the series. Beyond this, the results are not conclusive. The data are accurate, however, in showing a loss from the use of lime- stone with acid phosphate, up to the amounts which previous consider- ations have shown to be the maximum quantity it is advisable to apply. CONCLUSIONS AND RECOMMENDATIONS The soil used in the experiments described in the preceding pages was a brown silt loam. A description of the various soil types of Illinois, with their total content of the important fertilizing elements, is given in Bulletin 123 of this station, with the location of these types. Three facts are especially significant : The nitrogen content of the different soils suitable for rose growing varies from 1,870 to 8,900 pounds per acre (7 inches deep) and plans must be made by those florists who use soils of a type poorer in nitrogen than the brown silt loam to increase the content of nitrogen by use of green or barn- yard manures in the field, by heavier, applications of manure when mixing the soil for use in the greenhouse, and by more frequent appli- cations of manure or commercial nitrogenous fertilizer than have been used in this experimental work. The phosphorus content not only of brown silt loam but of all the types of soil is low compared to that of the other elements, and the results from the experimental work here described are applicable to all. Peaty and sandy soils are low in potassium content and are benefited by applications of this kind of fertilizer. It is probable, however, that the necessity for having a compact soil for successful rose growing would prevent the use of such soils for that purpose. KINDS OF FERTILIZER NEEDED Applications of phosphatic fertilizer give the most pronounced increase in the production of roses. Nitrogenous fertilizer also is needed, but applications of potassium sulfate not only give no increase but decrease the yield. (1) Nitrogenous Fertilizer. The need for nitrogenous fertilizer is particularly urgent after the turn of the year and makes itself ;. ( '7?J USE OP COMMERCIAL FERTILIZERS IN GROWING ROSES 563 apparent by the lightening of the color of the foliage that is associated bv every rose grower with lack of plant food. This characteristic is a better guide to the time for applying nitrogenous fertilizer than any rale. The florist will largely do away with danger of nitrogen starva- tion by enriching the soil before filling the benches by the use of green manures or farmyard manure up to twenty tons per acre in the field, or by mixing manure with the soil as it is put in the benches. If the need for nitrogen is apparent, it .may be supplied by liquid manuring, by mulching with rotted manure, or by applications of dried blood at the rate of 5 pounds per 100 square feet of bench space not oftener than six weeks apart. The nitrogen contained in such an application is equal to that in 130 pounds of average manure. Am- monium sulf ate and sodium nitrate, while satisfactory sources of nitro- gen, require too great care to prevent overfeeding to allow recom- mendation of them for general use. Applications of nitrogenous fer- tilizer should be lightest during periods of little sunshine arid when the plants are off crop. (2) PJiosphatic Fertilizer. Plants do not show marked signs of the need for phosphorus, and records of production alone can de- termine its need. Applications of acid phosphate up to 20 pounds per 100 square feet of bench space (40 pounds per 100 cubic feet of soil) have been found to give marked increases in production. The quantity of phosphorus contained in this application is equal to that contained in an application of 2,800 pounds of manure of average composition (50 percent moisture) to 100 square feet of bench space, or twice this amount mixed with 100 cubic feet of soil. Since manifestly it is impossible to use such a mixture, the need for phosphorus in the form of a commercial fertilizer is evident. Acid phosphate has proved to be a satisfactory source of phosphorus, but no comparison has been made in these experiments between acid phosphate and bone meal, basic slag, and other phosphate-containing fertilizers. - Since the benefit from the use of acid phosphate is continuous thru out the year, it should be mixed with the soil before the benches are filled. Top-dressings with it are not so satisfactory, since surface root groAvth is stimulated in this way, resulting in the roots having contact with the soil particles in only an upper layer of the soil in the bench. There is no danger of overfeeding with acid phosphate, for four times the quantity here recommended has been applied without injury. In this respect acid phosphate possesses an advantage over bone, which 'cannot be mixed v.ith soil or applied as top-dressings in excessive amounts without injuring the plants, as is true to a greater extent with high phosphate trnkage, and blood and bone. THE USE OF LIME With such a need fr phosphorus by rose plants, the use of lime or limestone in intimate contact with acid phosphate is to be dis- 564 BULLETIN No. 196 [February, couraged, since the solubility of the phosphate would be decreased by its use. The production from plants in soil with which limestone has been mixed is lower than from those on untreated soil, whether or not acid phosphate has also been used, hence, mixing lime or limestone with the soil, tho quite a common practice among growers, cannot be recommended. In case an application of lime is needed to prevent the growth of algse and moulds on the soil surface, finely ground lime- stone applied as a top-dressing at the rate of 10 pounds per 100 square feet of bench space and very lightly cultivated into the surface will accomplish this without being carried down into the soil further than an inch below the surface during the year. BENEFITS OF FERTILIZING The benefit from fertilizing is to be found in number of flowers produced and to a slight extent in the average stem length, tho not in percentage of long-stemmed flowers. No measureable change in length of petal follows fertilization with acid phosphate. KIND OF STOCK TO PLANT The experiments recorded in this bulletin with the varieties Kil- larney, Bride, and Richmond demonstrate the advisability of planting grafted stock, this conclusion being drawn from a record of production of grafted and own-root plants of first-year stock. Recommendations (1) Keep up the nitrogen content of the soil by turning under green or farm manure before use. If roses show signs of nitrogen starvation a lightening of the color of the foliage make up the need with applications of liquid manure, mulches of manure, or top-dress- ings of dried blood, the last in applications not exceeding 5 pounds per 100 square feet of bench space and applied not oftener than six weeks apart. Feed only during sunshiny weather and most generously during periods of heavy production. (2) Use generous quantities of acid phosphate in the soil. It may be added (a) at the rate of 4 to 8 tons per acre in the field, (b) in a compost with soil at the rate of 40 to 80 pounds per 100 cubic feet of soil, or (c) by'mixing it with the soil at the same rate just previous to filling the benches. (3) Do not mix lime or limestone with the soil. If needed for sweetening the soil and for preventing the growth of algge, make a top-dressing of finely ground limestone at the rate of 10 pounds per 100 square feet of bench space. tfl UNIVERSITY OF ILLINOIS-URBANA