University of Illinois Library at Urbana-Champaign ACES yw*?,*^ UNIVERSITY OF ILLINOIS Agricultural Experiment Station BULLETIN No. 125 THIRTY YEARS OF CROP ROTATIONS ON THE COMMON PRAIRIE SOIL OF ILLINOIS BY CYRIL, G. HOPKINS, J. E. READHIMER AND WM. G. ECKHARDT UKBANA, ILLINOIS, MAY, 1908 SUMMARY OF BULLETIN No. 125 1. Authentic records show that our oldest soil experiment fields were established at the University of Illinois in 1879. Page 327 2. Under one system of farming- the yield of corn has decreased from 70 bushels to 27 bushels per acre within thirty years; while under another system the yield of corn has increased to 96 bushels per acre as an average . Page 324 3. The fertility of the soil can be maintained, or even increased,, by a proper system of grain farming with legumes in rotation. Page 325 4. A good system of live stock farming, which may or may not be more profitable than grain farming, will also maintain the fertility of the soil. Page 326 5. Farm manure has given a net profit in the first three crops of $1.30 per ton, or of $7.80 an acre when six tons have been applied. Page 341 6. As an average of 18 tests, covering three years, with a rotation of corn, oats, and clover, the use of 75 pounds of phosphorus has produced in- creases in crop yields worth $12.39. Page 342 7. L/arger crop yields have been secured (and phosphorus has been nearly twice as effective) in a three-year rotation of corn, oats, and clover, than in a two-year rotation of corn and oats. Page 346 8. While phosphorus is commonly the element that first limits the yields of crops on our common soils, nitrogen is lost from the soil so much more rapidly under poor systems of farming that nitrogen soon becomes the limit- ing element, after which phosphorus alone has no power to increase the crop yields. Page 353 9. This bulletin will be sent free of charge to anyone interested in Illinois agriculture, upon request to E. Davenport, Director Agricultural Ex- periment Station, Urbana, Illinois, and if so requested, the name of the applicant will be placed upon the permanent mailing list of the. Experiment Station, so that all subsequent bulletins will be sent to him as they are issued. THIRTY YEARS OF CROP U ROTATIONS ON THE COMMON PRAIRIE SOIL OF ILLINOIS BY CYRIL, G. HOPKINS, CHIEF IN AGRONOMY AND CHEMISTRY, J. E. READHIMER, SUPERINTENDENT OF SOIL EXPERIMENT FIELDS, AND WM. G. ECKHARDT, ASSISTANT IN Soil, FERTILITY Near the end of thirty years an average yield of 96 bushels of corn per acre on one field, and an average yield of 27 bushels of corn per acre on another field, must be accepted as the results of different systems of farming on land that was similar and uniform at the beginning. These results have been obtained in the heart of the corn belt and on typical Illinois prairie land, representing the most extensive and the most important type of soil in the state, land whose present market value ranges from $150 to $200 per acre, a value that may change rapidly in Illinois, as it has already changed in the older states. The 96 bushels is the average yield per acre for the years 1905, 1906, and 1907, in one system of farming; and the 27 bushels is the average yield for the same years in another system of farming on land originally the same. Between these extremes other results have been obtained from several other systems of farming. It is the purpose of this bulletin to report, especially to Illinois landowners and farmers, the valuable data that have been secured in these investigations. The results from these experiment fields are now beginning to influence the agricultural practice of the state, and they are destined to be of inestimable value to the com- monwealth. Before discussing the details of the work, a compre- hensive summary of the effects of the different systems of farming will be considered. SYSTEMS OF FARMING In Table i are given three-year averages of the yields of corn secured in recent years, including 1907, which is the 29th year of the oldest experiments and the i3th year of a newer and more ex- tensive series of experiments with crop rotations and soil treatment with special reference to two markedly different systems of farm- 323 324 BULLETIN No. 125. [May, ing, of which one is termed grain farming and the other live stock farming. (Some preliminary cropping and other variations from the systems indicated are reported in the detailed discussion and tabular statements in the following pages.) The crops in the 3Oth year of the older experiments may be seen by visiting these fields during the season of 1908. TABLE 1. LATEST CORN YIELDS PROM THE UNIVERSITY off ILLINOIS EX- PERIMENT FIELD AT URBANA: TYPICAL CORN BELT PRAIRIE SOIL (Three-year averages: Bushels per acre) Crop years. Crop system. 13-year experiments. 29-year experiments. 1905 -6 -7 Corn every year. .' 35 bu. 27 bu. 1903 -5 -7 62 " 46 " 1901,-4,-7 Corn, oats, clover 66 " 58 " Average of Three Corn Crops in Corn-Oats-Clover Rotation: 13-Year Experiments Crop years. Special treatment. Grain farming- (with legumes*) . Live stock farming- (with manuref). 1905 -6 -7 None. d't lin 81 bu 1905 -6 -7 Lime. 72 " 85 " 1905, -6, -7 Lime, phosphorus 90 " 93 " 1905, -6,-7 Lime, phosphorus, potassium. . . 94 " 96 " *Legume catch crops and crop residues. fManure applied in proportion to previous crop yields. As an average of the last three years where corn has been grown every year the yield has been 27 bushels in the 29-year ex- periments and 35 bushels in the 1 3-year experiments. The lesson of these experiments is that 12 years of cropping where corn fol- lows corn every year reduces the yield from more than 70 bushels to 35 bushels per acre, after which the decrease is much less rapid, amounting to only 8 bushels reduction during the next 16 years. Undoubtedly the rapid reduction during the first 12 years of con- tinuous corn growing is due in large part to the destruction of the more active decaying organic matter, resulting ultimately in insuf- ficient liberation of plant food within the feeding range of the corn roots. In addition to this, the development of corn insects in soil on which their favorite crop is grown every year is sometimes an important factor in reducing the yield. Where corn is followed by oats in a two-year rotation the av- erage yield of the last three crops of corn is 46 bushels in the 29- 1908.] THIRTY YEARS OF CROP ROTATIONS IN ILLINOIS. 325 year experiments, whereas in the 1 3-year experiments the average yield for the same three years is 62 bushels of corn per acre. In this case the destruction of humus is less rapid, and the develop- ment of the corn insects is discouraged by changing to oats every other year, so that the decrease in yield is less marked during the early years, although the reduction continues persistently with passing years. During the first n years the yield decreased from more than 70 bushels to 62 bushels, and during the next 16 years a further reduction of 16 bushels has occurred. With the three-year rotation corn is grown for one year, fol- lowed by oats with clover seeding the second year, and clover alone the third year. During the first 10 years under this system the yield of corn has decreased from more than 70 bushels to 66, and during the next 16 years the yield has further decreased to 58 bush- els, the average reduction being only one-half bushel a year. In this system the most marked reduction in crop yields has not yet appeared, although it must be expected in the future because the clover crop is already beginning to fail on the oldest field even in seasons when clover succeeds well on newer land under the same crop rotation. When clover fails we substitute cowpeas for that year on that field, which thus provides a legume crop and preserves the three-year rotation. Further time is required to determine how much the cowpeas will help to lessen the rate of decrease in yield of corn and oats. GRAIN FARMING In the lower part of Table I (third column) are recorded the average yields of corn for the last three years in a system of grain farming, in a three-year rotation of corn, oats, and clover. This system when fully under way provides that the corn shall be husked and the stalks disked down in preparation for the seeding of oats and clover the second year. In harvesting the oats as much straw as possible is left in the stubble, which may be mowed later in the summer to prevent the seeding of the clover or weeds. In the spring of the third year the clover is mowed once or twice be- fore the usual haying time and left lying on the land. The seed crop, if successful, is harvested with a hay buncher attached to the mower or in any other way to avoid raking, and afterward the threshed clover straw is returned to the land, all of this accumu- lated organic matter to be plowed under for the following corn crop, which begins the next rotation. In addition to this, catch crops of annual legumes, such as cowpeas, may be seeded in the corn at the time of the last cultivation and disked in the next spring 326 BULLETIN No. 125. [May, with the corn stalks. If biennial or perennial legumes are used as catch crops, the corn ground may be plowed for oats. The corn yields reported for this system in Table I were se- cured where the system was not fully under way, the legume catch crops being the only organic matter returned to the soil, aside from the residues necessarily left from the corn-oats-clover rotation. By using three different fields for this rotation, every crop may be grown every year, and the yields of corn reported are true three- year averages. With no special soil treatment aside from the use of legume catch crops, the yield of corn for 1905, 1906, and 1907 averaged 69 bushels. Where the equivalent of ^ ton per acre of ground limestone was applied five years ago the corn has yielded 72 bush- els per acre ; and, with phosphorus added for six years at the rate per annum of 25 pounds per acre of the element phosphorus (in 200 pounds of steamed bone meal) the average yield of corn has been 90 bushels per acre for the last three years. The yearly addi- tion of 42 pounds of potassium in 100 pounds of potassium sulfate has further increased the yield to 94 bushels. The cost per ton delivered is about $2 for the limestone, $25 for the steamed bone meal, and $50 for the potassium sulfate. LIVE STOCK FARMING Under the heading "Live Stock Farming" in Table i are re- corded the average yields of corn secured during the last three years where farm manure has been applied to the clover ground to be plowed under for corn. The plan of this system is to remove all crops from the land as usually harvested, including the corn and stover, oats and straw, and both first and second crops of clover. The amounts of manure applied to the different plots are deter- mined by the crop yields secured during the previous rotation. While the system of cropping followed during the past 13 years on these plots, and on those just described under "grain farming," has been approximately equivalent to a three-year rotation of corn, oats, and clover, the applications of manure have been made only for the three years, 1905, 1906, and 1907. If the average yields are decreasing on plots that receive only the amounts of manure that can be produced in practice from the crops grown, then the applications of manure must also be reduced on such land ; where- as if the crop yields are increasing where both manure and phos- phorus are applied, then the applications of manure for such plots may be increased in direct proportion. iooS.] THIRTY YEARS OF CROP ROTATIONS IN ILLINOIS. 327 Where manure alone has been used in this rotation the corn has averaged 81 bushels per acre for the last three years; with lime added the average is 85 bushels; with lime and phosphorus the manured land has averaged 93 bushels of corn, and this was in- creased to 96 bushels by adding potassium. While potassium has usually made some increase in crop yields on these fields it has not nearly paid its cost. The most profitable yields are the 9O-bushel average in the grain farming or the 93- bushel average in the live stock system. The effect of limestone has not yet been sufficiently uniform to recommend its use on this soil, but marked profit has resulted from the addition of phosphorus, which is applied in sufficient amount actually to enrich the land and not as a stimulant. HISTORY OF THE OLD EXPERIMENT FIELD It appears that the oldest soil experiment field in the United States with an authentic record of its origin and with a present continuation of the experiments originally inaugurated is on the campus of the University of Illinois, or rather it is surrounded by the University campus. In the biennial report for 1879 and 1880, on page 232, and under date of March 10, 1880, is the following: "The Farm Committee then submitted the following report: To the Hon. Board of Trustees of the Illinois Industrial University: "Your committee beg- leave to submit the following recommendations from the Professor of Agriculture, in regard to experiments for the coming season: "Fifth The formal commencement of what is designed to be a long con- tinued experiment to show the effect of rotation of crops, contrasted with continuous corn growing with and without manuring, and also the effect of clover and grass in a rotation. A commencement was made last year, and we are fortunate in having a piece of land more than usually well adapted for such a test. "The report was approved, and its recommendation concurred in." Thus, these oldest rotation experiments, begun, according to the official records, by Professor George E. Morrow, in 1879, wMl complete a record of thirty years in 1908. In Bulletin No. 13 of the Illinois Agricultural Experiment Sta- tion, published in 1901 and signed by Professor Morrow, the state- ment is made that from the beginning of these experiments plot No. 3 had "been in corn continuously," that plot No. 4 had been "in corn and oats alternately," and that plot No. 5 had "had this rotation: Corn, 2 years; oats, I year; meadow (clover, timothy, or both) 3 years." The records also state that these plots had re- ceived "no manure or commercial fertilizers of any kind." 328 BULLETIN No. 125. [May, k ^ j * z z' s:' X Z N. X O PLATE 1. DIAGRAM (DRAWN TO SCALE) OF THE URBANA SOIL EXPERIMENT FIELDS, SHOWING THE OLD FIELD SOUTH OF THE OBSERVATORY, AND ALSO THE SEVEN SERIES OF TEN PLOTS EACH LYING EAST OF MATHEWS AVENUE. The photographs shown on the opposite page were both taken from the same point. The lower view looking southwest shows the oldest plots on the right and series 100 on the left, with Mathews avenue between and farm buildings in the background. The upper vie.w looking southeast shows plainly series 200 to 600, with series 700 and the University forestry in the background. Note the plot labels (shown in full for series 400), also the oats harvesting in progress on series 500. None means no soil treatment (except the regular crop system); Le means grain farming with legumes (including catch crops) and crop residues returned to the land; M means farm manure; L means limestone; /'means phosphorus; A' means potassium (kalium) >and A^ means commercial nitrogen; $x means five times the standard applications. i 7 z Z z _J O 1 ^ 1 _/ ITl g J J J j M 3 u f 1 0) in Z _J 1 i ^0 _J i _! 1 i 1 ft O vO _J 1 "; a B c 1 8 10 J. o 8 &. _J i 8 5O9- AVL.P.fV 1 1 6 oi 1 s < i J O 1 o CL C i o J i 409-A\.L.PK. CL 1 O I _j A i 8 O c i 'ft: i i o u _J 1 (0 309-M.L.PK. 1 ft.; 1 O I i 6 li i i J 0" 1 ' _J (NJ I C X 1 o Csj 5 V 1 cs s; 8 j _J C o * r i vC o C i o * Ha U 3 gx'w ON ON ^ S g 3 S *'S CO o o n oo 10 "t O< ID ^SjS ^ 2 rXrX rH rH N cot- ON & a CO ON ON 10 oo CO * ga Pll ON t>- M 1> Sg i> C a J^ rH fS) 00 CO I 9S VO t> * N l> t- f} !fl S a 8-2 rj *d 00 5 S TT ^ && Ma Wri > O -M M J3 eu a ss; M >0 ^C t- rH VO ON N rH VC Tfr OB n h o> o 6j3 ON O ro t> i V U< ^8- rH 1C M rH rO rH t- * ON CO ON VO ^ rH -! ^1 a fe^^ c 8 a- Cd UH 1C ^ O X a SoS^ VC O ^ VO rH 1C GO * 10 MS fO tfl O 00 rt rH rH ON VO rO ON vo M rH . i.fe& a"J a ^ .2 '3 nn !j ra 6^ |l f H X a M M rH oo ON ri ^ rO ON iO 00 VO M * 0) s 00 rH fO M |. S a ^f rO rH VO t- rO ON VO M rH *vH J Si | - a ON rH rH t- t- iO UO * S rO CO VO rH VO CO iO 0) a s ) "O rH 5 0) 1> S s t> 00 iO rH 1C O vo i C1 ON r<5 M K 3 iO CO rj-t- rO oo rH >O ON t- M 00 VO M rH S5 O rH OOiO S S rl ro rH cot- i^ S a rHio vo S3 V S a * O ^j fsj (U 4> OJ a 3 rH CO rH 00 c^ I- rH O\ ON 00 oj co vo M rH be a ^-a 00 6^ bo 0) h4 rf> ro t- t- CO * ^f O ON C4t>- rO N ON co oo vo M rH o * cot- 1> * o fc vo rO 1> vo oo % a 3 bo bo :*;to !H > S ,2 O-H 5-00" 00 ON O rH ON ON O O 00 CO ON O\ i Partial soil 1902 -1 treatment ( be 0) p O 2*- - fc fl to" & fe^o uoo M ro TJ- p o o ON ON ON {Complete soil treatment be !fl a il* 5 X3^ v ; C tfl > S^o OOO iO vp t- ON ON ON O E o 0) 1 tfl cd M-i O a 2 "rt > a '5 S . S . it (A a N t> ON o ^ fk 1 a 3 ^ (^ '* VO >O X a O t- rl- OH ot- ro' to rH S rH m * o VO iO rH a fO t- '0 to O ^ V . ?s 7 T 4) a o o o r) 3 rH rH b P O O VO CO O 00 rt ON rH l " I VO Q ^ he V -* w a : i ; o H * J j3 ^ ! a 3 ?.g 2 d rt ^ rt \O D rO f-> O r^ JH .0 rH "U -M n JH ^H 4) a n HM (0 ^ C "rt l2 '(3 4> bi kl H O tfl -t- 1 W O 4) 338 BULLETIN No. 125. o r-i re co re ON r/l 3 C W> 5 D 0.2 fl J3 ) ON * r-l rH VO rH vo S * " S a S S f>l ON t^ ON r-l ^ VO vo re * re .5 CU M tf) rt Mo-'-' Is. rH ci f>i rN 00 >* &3 J " c'- g- J ^ S * [> vo ,? w S 1>VO M rH fN 0! SS re I> <* re ,2 O* o? -t M rt M o*; ta rH VO 'O fN CO * G CX t |3j2 J3 ac^ VO ro 00 * re '0 v. GO * r-l re ON rj- S 2g 8-S n-* 08 2| .-i Ji X 'S 00 rO r-l VC ?5 r-l r-l rH r-l OM> VO CO TJ- re ^3 a li rHt^ Q fN oo vT5 &j g.a &3g| J && vo VO CO * re >o tf> rH VO VO t^ VO fN rH ON I> *a9^ rt vo oo fkj vo re L/ime phosphorus VO re rH 00 re 00 rH rH O rH 00 vo Manure lime phosphorus * * I> * oo * 1 3 r\ 'O ^~ <* o o Leg-ume lime phosphorus l>- re rH VO re CO ^S8S^ a) S 2 is* 5>S a fag H X a I 1 ffi /.. vo re * ON * IONO l M 00 M rH ON t^ ivo'rn' 5** v4 a r-l re vo 1> Tf *a X : r*3 1^ I> re $ /- + 8 VO rH fN ^ *" re' rH ON I> t IJret- Ovo' ON* ^^re Legume lime rH 1> O * * CO r-l i-l 00 t- "* V 8 ^ 3 S bc* M 00 O rH re O O VC I> V h 3 * re ON VO f>1 N coi> vo re fc -g$ 8 ri a O vo t^ re 00 re '/? M vo co r- M re 4) s pi fH r-l O 00 fN CO 4 rH VO N 1> LO bn 0) a vo ci vo re n 3 re t^ ON ON ON 0) re Tt- O CO VO 00 0) O M 91 i- r^ fXl X vo rH re vo re r-i fN vo J> "* o g 1> l> VO M r-l S c a 3 r/l 333 .0.0.0 a c c rj^ '- '- O V O 1) omplete soil treatment beg 3 3 J= G uT 2- O O 1) (0 RJ O "o S o o o aao vo OI> ON ON ON oo oo oo rt ri rt rt OOOO 00 ON O rH ON ON O O 00 00 ON ON rH 8 ON o o 13 J2 oooo rN re <* vo SN ON ON ON U O rt oo <+- O 0) 2 "rt > 1 ON O vo vo r-l 9< >n j,J vc' nj #* | O S 3 re O 00 vo C l/l 1 T^ !> VO "S <& a) 3 O go C M a M t^ 10 a rH Ui a JC en 4-> CO 3 X u a O O O vo B 00 O ON vo 00 -1- a V & 3 rH r-l j3 m a ^ s 4- O * vo 8 3 cd rO 3 rH rH 2g a rH lanure M O t^ M Vo'rH VO 1.S rt -d &l .2) 2^ D o M S 3 S8 VO X^ U Wi be s* V O 0) J sS l>t to * (J CU > p LJ U i C .2 v> Wi c js a *j O O rt-J* 5 O TO O S -o ri vo W f. 0- O - Wi o > JJ fc B* -M C S 7-1 O cfl U O "a be t! # -T--M- OO <5 1908.} THIRTY YEARS OF CROP ROTATIONS IN ILLINOIS. 339 o VO > 3 g V O 3 P -c '35 ON ON C-l vc 'O vo ON >O v * "C S D 1> 2 o * I-H rO ro 00 M rO ON 1> ?>pl 3~! a s - oo i-H T-l n l^ M 3 l> O\ * VO a 2 S 4J O 3 S-C '35 ro M 00 vo 00 O ON Efl e U .. O 3 rt VO lO J o3 > X J t 1 O -M ^ 2 0.0. vo rO * aj*i 51^ o-2 fl & 2 P. Oc rO TH i-H 0^ M 00 * TH rj- eo 2 6 8 a; 0.5 3 n ** to * iO t i.M vo -t- ri 8 TH O VO ro O l> TH he C O. at w^gj a = S o rO O "* .H VO ONt^ I'll j ii i 1 1-1 1-1 8 t^ <* n ro >o H 3 c s .a o, Ug A a. 3 M ^S^ c E a rt ^ te ^ J O. O i-H i-H VC ^C r I- -t ro M t^ >O rO OM> * 10 P- M J3 O. M M * * tH vo ON r> flf H X O. 3 ,-H vo ffj iO VO * rO * s ro vo oo vo ri o IH D t-i m 3 g TH ro 00 ON 00 rO O> I> * UJ rt ri d ON o TH VO 00 1> Bj r5 S I-H O 1-H ON 'O 0> S S &S3 0 t^ C) 00 O * * ro rxo >o lO I> iO M S 00 00 ON 't > M ON CM CT\ * 0) S 3 O vo ro iO M O l> >O > D a IO t^ ro 't over- VO o 00 TH * rO O TH GO V 10 i i rO O ro O ON * ro 00 VO ro <* g rO Tj- o % g i o -d * ej c O K "o s 333 42 ,0 J2 c~ s" c _ _ u O O O OQO 1C vo t^ ON ON ON X 00 CO 1898 Cowpeas, (no recor 1899 Oats, (no record) . 1900 Cowpeas, (no recor 1901 Corn, bu ( Partial soil 1902 treatment ( begu at 52 O c o SH- 3 s - 1 - 1 fe' -' ' T > c M" > rt 5 o rt o OOOOO fN ro * 10 vo O O O> O O ON ON ON ON ON ( Complete soil 1907 -j treatment f begi 3 ,C r o i T I a LI u o Lj J3 to 3 o o _n "rt > 4-* < rt M _C 5 v Ja "O T-J rt ri o P O * g at 3 00 o Q) O Potassi * vo TH 1 1 = ^ ri O. n o 5 i^ O p t 0^0 a :assiu t- vo rO I- -t 1 > (U ^O 3 C Tj o ^ 1 ZC W ^ DH S c to 3 -j- O< u o O TH O VO vo ON X . rt fe. O. VO t> S c CO O 8 ^ .2 S o> o *CL, .~ti at 5 a; t^ 2 3 MH .t3 *^ 3 O SO VO 10 S3 S x 0. (M TH 10 .2 a, n o O rH *"^ ri S "^ 'O . U 1 o 4H 0) S TH O ON vo ? '1 a^ " 0) vo O 1 C - fe Cfl a 00 vo "rt 'O 5o 'ri a 00 r-i ^ g ri rt 'o t& S >s "^ n ri VO TH Izj a rti ^ *^^ hH ri ri ri 'O o >% ^ w ty ^ ,O T 1 j_j O o ^ O nH t 2 ^3 t> ri ^ o " O< D Ui ^ Ui M ri j. 0) JS "0 - 15 .2 r*> * -H -2 c 'S at j 2 S & E * T3 ri "Vg c" Li >> 4= ij ro 3 -C .S he a) > (1) n ri "a /: V-i C JS "0 CO a 'ri ^ S S3 ri ri <+H 'ri * 3 3 (L> LI H M +J O to O 1 o *o 0> iH 3 340 BULLETIN No. 125. PLATE 5. CORN AFTER CLOVER WITH NO SPECIAL SOIL TREATMENT: PLOT 301, URBANA SOIL EXPERIMENT FIELD, 1907: YIELD 80.5 BUSHELS PER ACRE. The complete grain system is first put under way for the 1908 corn crop, the equivalent of the clover grown in 1907 having been returned to the legume plots (102, 104, 106, and 108) to be plowed under for corn. The corn stalks will be disked down when seeding to oats and clover in the spring, and the effect of this on a subse- quent crop of corn on series 300 cannot be seen till 1913. Thus, while much valuable information has already been secured from these fields, as shown in this bulletin and in Circular 96 (published in 1905), it should be kept in mind that the full effect of these sys- tems of soil improvement will not be seen for several years. THREE- YEAR ROTATION In Tables 3, 4, and 5 are reported in detail the crop yields from series 100, 200, and 300, for the 13 years from 1895 to 1907. 1908.] THIRTY YEARS OF CROP ROTATIONS IN ILLINOIS. 341 PLATE 6. CORN AFTER CLOVER WITH LEGUME-LIME-PHOSPHORUS TREATMENT (GRAIN FARMING) : PLOT 306, URBANA SOIL EXPERIMENT FIELD, 1907 : YIELD 104.6 BUSHELS PER ACRE. With few exceptions all yields of grain and hay are recorded in the tables, and a financial statement is also given at the bottom of each table showing the effect of each addition in the soil treat- ment during the last three years, on each series of plots, counting corn at 35 cents a bushel, oats at 25 cents, and hay at $6 a ton. It will be seen that but little effect has been produced by sowing legume catch crops in the corn in the three-year rotation, where corn is followed by oats and clover. From 1902 to 1905-6-7, the legume catch crops produced but little benefit on other crops, as shown by comparing plots 2, 4, 6, and 8 with plots 3, 5, 7, and 9, except on the oat crop in 1904 and 1905. The effect of the farm manure on each of the three crops is shown only in Table 3 (series 100), the manure on series 200 and 300' having benefited only two crops and one crop, respectively. Six tons of manure per acre on series 100 have produced increases in 342 BULLETIN No. 125. [May, the three crops amounting to $9.60 per acre, or to $1.60 per ton of manure applied, leaving a profit of $1.30 a ton for the manure of the farm if it can be hauled and spread for 30 cents a ton. Of still greater interest will be the effect on the next three crops of the second application of manure ; (made for the 1908 corn crop on series 100), including any possible residual effect carried over from the previous application. Aside from the legume and manure, there is a duplicate test for each addition, and the effect is determined independent of the check plot (No. i). Thus the effect of lime is determined by com- paring plots 4 and 5, not with plot I, but with plots 2 and 3; and the effect of phosphorus is determined by comparing plots 6 and 7 with plots 4 and 5, respectively. On five plots out of six lime has shown some increase, and on three plots it has paid more than 50 cents per acre per annum, which is more than sufficient to pay for the applications actually made to these plots, but the results vary so greatly with the differ- ent crops and seasons that safe conclusions cannot be drawn until further data are secured. The apparently marked effects on plots 304 and 305 are evidently due in part at least to original soil dif- ferences,* indicated in the early records for those plots, as compared with 302 and 303. Aside from farm manure the only well established benefit is derived from the application of phosphorus, which has produced a marked increase on every plot and in five cases out of six has much *The large amount of data accumulated and reported in the preceding and following pages concerning these plots previous to the beginning of soil treat- ment has suggested, of course, that even small natural soil differences could be detected in the early yields secured under similar conditions of cropping, so that corrections might be made for such natural variations when considering subse- quent yields as affected by soil treatment A careful study of all the results shows, however, that it is very unsafe to try to make any such corrections. That plot A has produced larger yields than plot B for two or three years does not prove that it will continue to do so. In fact, under normal conditions, this relation may be reversed. The larger crops from plot A remove more plant food, and the later crops may be smaller in consequence ; whereas, if smaller crops are removed from plot B in the early years, the yields may be larger than on plot A in subsequent years. Several illustrations of this have actually occurred in these experiments. Thus for the three years, 1895 to 1897, plot 406 produced, as an average, 5 bush- els more corn per acre than plot 506, but as an average of the next three years plot 506 produced 2.9 bushels more corn per acre than plot 406. Likewise, a decision based upon the first three years would hold the natural productiveness of plot 407 to be 3.8 bushels higher than plot 507, but the records of the next three years would reverse the decision because for those years plot 407 was not 3.8 bushels better but 3.1 bushels poorer than plot 507, as an average. The aver- age yield for the entire six years differs by less than half a bushel for these two plots. (See also the last page of this bulletin.) Because of these facts it seems unsafe to try to make any corrections or to use "corrected" results. The data actually secured are correctly reported and if there is reason to believe that any results are untrustworthy they may be discarded. THIRTY YEARS OF CROP ROTATIONS IN ILLINOIS. 343 more than paid for its cost even in 200 pounds per annum of steamed bone meal, in which more phosphorus is applied than is removed from the soil in the crops grown. This, of course, gives promise of some residual effect in subsequent years, more especially for clover, because the land is growing richer year by year in phos- phorus and because clover should not be limited in yield by lack of nitrogen, while corn or oats may be, no matter how much phos- phorus is supplied. The very marked effect of phosphorus upon the yield of clover during the last three years, averaging more than 75 percent in- crease, is especially significant, and constitutes probably the most valuable information furnished from these particular experiment fields. The addition of potassium has produced . some increase on most plots, but never sufficient to pay for even one-half its 'cost. The effect of potassium indicates the need of more decaying organic matter, which has power to liberate sufficient potassium from the immense total supply in the soil. (Soil samples 59 to 421, whose analyses are reported on page 270 of Bulletin 123 were collected in 1901 from series 100 to 700, respectively, and samples 467 to 472 from plot 3 of the old experiment field.) TWO-YEAR ROTATION In Tables 6 and 7 are given data from the two-year rotation of corn and oats on series 400 and 500. It will be seen that corn was grown for six years on series 400 and 500 before the two-year rotation was begun. The subsequent substitution of oats for corn in alternate years lessens the liability to insect injury and somewhat reduces the draft upon the soil, so that larger corn yields are likely to be secured for a few years than were produced near the close of the six years of continuous corn, even with no increase in the annually available fertility. The legume catch crops, as cowpeas seeded in the corn at the last cultivation or clover seeded with the oats, have produced but little effect, and this effect has been a decrease as often as an in- crease in the regular crops. Plot i is apparently slightly better land than plots 2 and 3, as indicated by the crop yields secured previous to the beginning of soil treatment; but a comparison of plots 2, 4, 6, and 8 with plots 3, 5, 7, and 9, from 1902 to 1904-5 shows only a small average benefit on the oats. On the other hand, the legume treatment produced noticeably larger yields of corn in 1907 than were obtained from the plots to which light applications of farm manure had been applied. As a rule lime has produced some increase, but as an average it amounts to but little more than cost. 344 BULLETIN No. 125. [May, CO rH vo VO VO ON 1/3 3s il C J3 aj N O r-( rt eg o> o 2 s3 j: 'S H -1 O I> vo -1- rH -* CM 00 VO CN ON*^ vo vo vo t^ VO *O vo CO .5 0U i > i/i rt *l a o, vo vo ON vo <* CO .5 OH > ag| U& t^ 00 vo vo vo ON 00 ?, * * CO VO rH rH t> SS S.2 P -C vo ON vo la Bl rH 9 vo -* rH Q M CN o t^ O vo * * 1C VO vo cO S CU X i4 g * P VO 00 vo ^C * CO C w G- ) -| s -H& * 00 rH l> voi> ON 95 t^ rH N N O * ON 8 ^ 1151 J* * _ S 21 Ig^-a O 00 rH -t rH N ON VO i> rN d N cO vo VO CO SJe-a ^ te jS ^ is. 5 ^ Cu M * VO * Mfl ^ !& a CM * VO VO 00 10 s t^ rH CO l> TH> ON 3 V O S-c z vo 00 3 0) VH 1- d ON oo ON vo vc vo CM .5 OH Sjj a n M VO * s.S o, n) ~ ) fcrt O S ,c P< rH CO 00 1^ vo vo ON rO M 1 rH 00 1> t^ ON IO CO OM> rH CO O rH /} vo vo cO Legume lime phosphorus. * H ( l^ O a r-l O 00 vo VO 9 O 00 1> fN O ON cM e 6 aj 3 S M r-t VO u I, 3 - rH t> ON Tf r) _ o ON -H CO 00 > iO * co iO iO vo VO CO bog JF rH UJ ON VO * CO || 328S rH 8 r-t VO 00 ON ON O iH V 8 VO N rH V h 3 * vo rf 1^ VO I>1> 00 n oo vo IO rH VO CO VO vo i/> CO ^ ON O Tf- vo vo CO i S ON t^ ON vo vo vo oo |g -fr CO O VO t^ O vo V i 3 vo ON * o> ON * ON fe _ rH VO t> CO 00 VO CO 00 J> * vo iO vo CO So u I> C<1 00 vo * CO ho V HJ fN rf ON VO vo vo g rH * CN CM rH O 1> ON * -M- iO X & '4 333 ,0.0,0 ^J rt Pug E i 3 3 g XI > J3 *J -*-> CO C C C a c a n n) CC C 0) o CMC" rf o fc -*-> o s O O O 000 IT) VO I> ON ON ON 00 00 00 O O OOOO 00 ON O r-i CN ON Q Q 00 00 ON ON CH M g rt o rt ooo n co * ON ON ON O vo O ON O rt O OOO iO vo t> ON ON ON H- o CJ _3 "(3 > 5 3 '3 Cl U) ' vo VO o 1 PH sium SS 8 0) rt CO vo ? 00 . O 13 rH S E X O 3 a V* o S8 V > a ' vo -t o XI o ^ nl fe 'O ON D 3 rH .JH C ^ Q ~> Q - >s a VO vo ^ tiS c S o *s >% c qn DH ? (0 u> V Mg X M . . 3 (0 3 ^^ C c g J3 & ^ o c 1 'l^ r C A rt ^-C 3 ^^ C) S-T3 * o 5 fe o nJ ? D (U 4> 1) 3 ooo rH CM X CO -sl C 1 ^^ rH fl M-t .S^S n3 S * fl) *>^ 4^ a 3 be ^^ X VV r^t "^ -o > g - h > ' tr; "S *rl *^ KS V 28 "U T^ ^ ^ rH- |o ? O i_i | t : 2 f*% * rt "- ) ^ B C.O rH ^ O B-o g ^1x1 ^*^H s bC O "I ,j~ w * * ++ "rt tfl -l-> O * H- 1 ^ w o ON 00 3s ' *" 0) 3 fl J3 'S rO 00 rH M gX M a VO LO rj P_ i 4 tfi rt O -M S aa O C^rO VO I> VO vo LO ^43 ^! S o o It* 3% S O O rH CO T- t> vo LO &.i^i ^"o^ >-l ~ O a^ 00 00 rO rH VO vo -O vo M.S Ja rt 1) ^H o -M a *. 00 ON LO ro S cO ON <* rH rH ON CO to 3 gg 8-fl M ri M CO 09 hi 00 CO 0^ 3 * 00 * fN ONl> 00 rH O 00 iO t> vo VO .t3 a, a A a vo f3 rO rO VO I> vo vo ail * ! a rH rH vO * rH rO ^ * vo CO t> t> O f) vo N S 2 2- vo <* vo * 2 l CO 00 VO CO oo * * fN ONI> vo vo *f) to S.fe| J^^ a rH rO rH Tj- vo VO I> vo bcG O- ^^o M X a CO 00 VO CO 1-1 ro v? rH O <* rHOrJ-M- D S M rH ON rH u gfi 00 00 S i/) ON vo N M ONt^ if) vo iO LO ^4 l> M 00 00 10 10 * LO nj;3 S l> CO O CO S -t ON * I> rO -t O\ u S 3 S vo ON O 00 0) Ba; ? S ON rH ON CO 3 t^vo l- Cl 00 VO rH rH rH If) If) VO IO * hc^ a> ^ a * VO VO M iO * * LO M.H OJ ^ a 00 ON Tf N S 5 ON VO 00 M 00 VO Tt- rH rH * 10 VO IO * fe IO CO VO rH iO * * to cd S * ON LO M VO n 7 g * rH 1> O O rH O 0) a T|- M vo vo 0) a 3 ONTf w 10 r-( 00 ON rO 00 VO t> M * VO iO VO iO * be 0) h4 00 ro M to iO * iO iO bo U u rH ON LO CJ LO n ff: l-H O VO vo IO OM>iO D n VO to OtO ID i CO * * LO LO t^ oo ON CO 00 VO ON ^J" oo 00 to voi/) * Z I> M ON M iO * to r< S- sl ll ^ 3 a ,3 ^ tfl CSC G n" c" cd Q to" fl tc fl" ta ^ 6 g. "o s O O O OOO tO VO I> 0*1 ON ON oo oo oo o o o s OUOO 00 ON Q rH 00 00 O^ ON CH | O cd o rt OOOO n to * *o ^ c^ ^ ^ O o rt OO vot^ 3\ o^ H V 3 i5 > Potassium to o n- q iO to f to rt- 1 Potassium 00 Q VO O T CO 1 c Phosphorus 88 COLO c 1-1 1 "o ri 0) Phosphorus t^ Q S o HH "s 6 J LO O (N O M rH LO M "S CC 'o a 3 T rO Cl U 3 s Manure 00 O rH rH T ri Legume 88 ro 1 a o -5 o ri "S VO VO 00 " r-t a * .2 i- 10 00 ON VO O VO W co OMr^ rO CO CO vo vo vo Jl 3 42 PH vo vo vo O I> rO vo -* vo * 1 s ^ ^ PH oo t> rO rO ON TH l> ON TH 00 S .3 rO rO s t^~ ri ON M ON VO rO ON * LO vo vo 1 8 a 3 OH M * M ^ N rO vo * rO * 3 4 ri 3 a * cq rO CO oo TH VO r-( VO *!> go 3 c fl X O CO N 00 ** 3 L w 2 eJ3 ON O vo t- TH VO cNJ ON VO N 00 O Tf vo vo "> tfl H X a S ro n vo l> r*J VO iO * f<5 a i >4 o x (X T- rq *rO O O TH ON ON vo o SJQ VO O O^ ON ON : 2 S TH CO s O Tj- TH co ONI> 00 TH VO * vo vo .- OH a 8 ^ PH r<3 ci >! VO O rO VO * vo * PH "1 tfl ^ J a I> vo * CO CO M ON N r<5 * 1 ON 00 VO O\ VO c si VO TH * co co CO ON l^- cr\ vo o 10 vc vo S o a| a rO *!> M 00 rO 10 * vo * 3.1 a ro a a^ O\ ON VO * O a LO Tt O 10 vo vo .5 2 >-].t: a Ovo vo ON 00 e<) if) tt -*t <* "1 Ul t-J.t! 9 M M vo vo VO vo rf M O\ * ID s a CO vo TH fO V S * TH CO CO TH M COI> Tj- VO VO 4 r<5 * rO ON vo ro vo * - * 23 VO 00 O t> O * e ON vo ON M vo m S 00 TH vo O rH 1> CO 00 VO N Ot> iO VO vo H HJ TH M TH VO TH rO vo * Tj- * 5 CO1> CO rO VO vo CO 1> rO J> V a ON M (S O CO V c 00 VO i CJ 00 VO 8VO rO vo LO o fe VO r-t M VO CO M vo * rO rO 1 CO CO VO 00 W i> o O CO * a 3 t! bo S 0) d M.& C 4) ^ JJ- 3 3 3 ,0,0,0 3 3 3 J3 JO X5 s* O rt (A ON ON ON 00 00 CO 000 ooo ^^8 co oo o\ PL) TH c? O O O O O OOOOO TH f<) CO T(- VO ON CN wN ON ON O ou O t^* s * G\ 3 - tfl CO VO M 00 S co' * 10 "o S .5 VO VO ON X to TH C 1 ~t~ ri 1 I "o 1 1 iorus *) it-lf) r* a vo'vo CO vo to 1 'v m 3 JO aj 3 O VO ^t CO 00 VO 1> VO TH c~ LI ON O O CN VO CO I> 9 .t{ o ^ rt C V 0) o bo O M vo * ON MH E rH rj- f\l rj- ', C V a T- t> M CO vo ;/) ^ t_> ft) 3 a M vo M VO 0) "i CO * CO N be M i 1 -S nj rHC^L^ JO TH TH rH ^ "c "rt w O O O tfl .5.5.5 S "* u H > > > w 7905.] THIRTY YEARS OF CROP ROTATIONS IN ILLINOIS. 349 rr 10 10 VO rH |si sP*i iO >O tf) TH 10 VO * ^ 8K 1> CM iO VC VO iO 2g-E3 2|I 00 O TH M rO rO t >O 00 VO 3 J C "S. re S "o. ON 10 t^ VO g Mt~ t> 1> CO g6 ^ M"S Ml> VO VO >o * ?," 8,g ON 00 f^ VO M t- CO O VO iO 2^^ S4S \i VO O:00 O rO >O iO 1> O i re-r; g C -- ON t^ iO CO VO O vo 05 2 S a> J o 3 fl ^X'S M mo * 10 i cgfi 1) W 2 P s?*"3 O >O t> VO t^J>- oo r-i r> VO VO >O J?&2 S|fjS ex o e. VD ON Tj- O\ t*- rO VO iO l> iO .52^" >} G ta 1> W I> VO t: lOl^ VO TH TH M S 4J 0.3 c OC.G S iOI> VO TJ- iO tc s|l 1 1^1 CO CO t^ n oo oo i> VO CO 00 1- VO >O E P CU at ^ a x g p,a TH ON VO r-1 00 * o 10 oo >o .S 8 ft ^i! tc re " 5 o -a ex 2 a^ GO O l> vo s rH TH 00 TH CO 3g il SJ3 3 O\iO * *O TJ- to 3 g ^ 5 D O 3 S iB*S O O 1> vd O\ 00 J> >O ON O iO VO - *1 a js It O ON O O M 5. o 10 1> 10 .5 g.S > 3 II ON N 10 Tj- y? -t 00 rH xj- ON P |J ON * * VO CO 1 11 TH O ti ON t^ oo i> CO *O VO 3 ?-3 ^ is. M rO t> ON ON 13 11 * IO iO CO * CO VO iO t> ON n 0)0) = a 'i 00 fO * TH TH ID o> a a be - " o S CO Oi ti * ON ooi^ o o n CO VO VO i4.-b -| _a C3 ON -* ri ro rt iO >O l> VO 38 5 " .ti E. 'a Sg 3 oo ON O * rH S3 4 fO ON ON rH C<1 Th vo vo 10 a M c 5 B P ^^g S o-S. ^.^J c ~ 00 rO 00 <* 00 00 1^ iO a 0) l> S n bc .5 o-g. > IH 35 MM a ON CO CO * t^ vo fH g OMO CN o ooi^ 00 * O 2S Lime nitrogen phosphorus iO ro TH t^ TJ- iO M TJ- ro O rO vo >O t> O a (0 4) g a bo fe *" r^ FJ=: 2|1 C **" TJ-IO CO CO t>. VO ^- . s- O a; hC 1C ,-. 0) :!l 3 3 .0 > 3^3 X2 ^3 J2 -3-- .5 re 4-1 O S OO VOt^ ON ON CO 00 i-l rH o o o ooo 50 ON O OS ON O 30 OO ON rH rH rH PH rH g TH o o o o o UOOUO r-l N r5 * >O ON ON ON ON ON rH TH rH TH rH rH O OO ot^ o p ON ON rH TH 4- 10 D O W 4) vo ro j3 2 X re 1 5^ ro rn 9 CN V V QH -4-> JH n IH to O 3 -g il I -*0N p 5 "re N iO Cjj I> re "S *u ^^ ' o> ra IH f 6 3 > o Potassi TH TH TH I I . Q 1 ' | O ^, H to VO O 00 * ^2 'o re n-^THTH o J3 * ." D X jj; *-" fe 4 s O p S a .2 ^ vo >o o * a -= TH fH TH 5 a &. 2-5 * a g x E) co *o ^fr ON < -> s CO VO 'O H ^ Nl .S ^2 a, TH^ ' O O O o js a .v. increase 1? .v. increase 1* .v. increase 11 icrease for 19 re 3 fe * ^> > ~ ^ to d .M C o <\ << <^ I_H re a 350 BULLETIN No. 125. [May, In 1901 the nitrogen was applied in sodium nitrate, but since that year dried blood has been used as the nitrogen fertilizer. It is well to keep in mind that while nitrogen can be secured with profit from the inexhaustible supply of the air by means of clover and other legume crops, it costs at least 15 cents a pound in com- mercial form. The 250 pounds of nitrogen now being applied each year to all nitrogen plots on series 600 and 700 costs at least $37.50 per acre a year. It should be understood that the growing of clover produces several effects on the land and upon subsequent yields of other crops, only one of whch is due to the nitrogen secured from the air. In fact the chief effect of clover in increasing the yield of a subsequent crop of corn is not due to the nitrogen secured from bie air, but rather to the liberation of phosphorus by the decay of the clover residues in the soil. The physical improvement of the soil and subsoil are also important factors in some cases. PLATE 7. CORN EVERY YEAR WITH NO MANURE OR FERTILIZERS : PLOT 602, UR- BANA SOIL EXPERIMENT FIELD, 1907 : YIELD 34.6 BUSHELS PER ACRE. 1908.] THIRTY YEARS OF CROP ROTATIONS IN ILLINOIS. 351 But to determine the need and the effect of nitrogen itself we must apply nitrogen in some form and dried blood is the most sat- isfactory form to use. Sodium nitrate and ammonium sulfate are also common concentrated nitrogen fertilizers but they are soluble salts which may produce marked indirect effects, not infrequently of greater consequence than the effect due to the nitrogen itself. Phosphorus has been applied in steamed bone meal at the rate of 25 pounds of the element per acre per annum, the requirement of a hundred-bushel crop of corn being 23 pounds, and the annual loss in drainage water being about i pound per acre. The annual cost of the bone meal is $2.50 per acre. Potassium is applied at the rate of 42 pounds per acre a year in 100 pounds of potassium sulfate, also at a cost of $2.50. A hun- dred-bushel crop of corn contains 19 pounds of potassium in the grain and 52 pounds in the stalks. The application made is not PLATE 8. CORN EVERY YEAR WITH "EXTRA HEAVY" SOIL TREATMENT : PLOT 709, URBANA SOIL EXPERIMENT FIELD, 1907: YIELD 71.8 BUSHELS PER ACRE. 352 BULLETIN No. 125. [May, sufficient for the entire crop, but the natural supply of potassium in the soil is so exceedingly large that it is both unnecessary and un- profitable to return the potassium removed from the land in any ordinary systems of farming. On plots 709 and 710, beginning in 1906, the application of phosphorus was increased to five times the standard amount, and the application of manure was begun, using about five times as much manure as could easily be made from the crops grown, or 20 tons per acre each year. The addition of commercial potassium was discontinued on both plots and commercial nitrogen was also discontinued on 710, but is still used on plot 709, at the rate of 250 pounds per acre each year since 1906. The plots on series 600 and 700 are not quite so uniform in character and in natural productive power as on series 100 to 500, but since being tile-drained in 1904 the natural differences are re- duced. It may be seen that in 1896, for example, plots 703 to 706 were among the best yielding plots, while in the season of 1898 those four plots produced markedly smaller yields than the other plots in the series. This is explained in the foot note to Table 9. One other natural condition should be understood: A slight ridge crosses series 600 and 700, the crest passing through plots 606 and 607 and through 708. The lowest plots in the 600 series are the end plots, 60 1 and 610, while in the 700 series the lowest parts are plot 710 and the depression that covers part of plots 703, 704, 705, and 706. Because of the continuous cropping with a cultivated crop these two series are continuously exposed to sur- face washing, and there is some tendency for surface wash to be deposited on 601, 610, and 710 and to a less extent on the four plots affected by the depression mentioned. There is also a slight ten- dency for the dried blood applied to plots 605 and 606 to be carried north to plot 604 and south to 607. Because of these facts the most trustworthy data are secured from plots 602, 603, 605, 606, 608, and 609 and from plots 701, 702, 707, 708, and 709, and this is recognized in making the sum- mary given in the lower part of Tables 8 and 9. EFFECT OF LIME. The effect of lime on the yield of corn is chiefly due to the fact that calcium carbonate (limestone) encour- ages nitrification and the consequent liberation of the nitrogen and to a less extent of the phosphorus contained in the natural humus of the soil. This is best seen by comparing plots 603 and 602. The summary shows an increase from lime of 3.2 bushels in 1901-2, rising to 4.5 bushels in 1903-4, and falling to 2.5 bushels in 1907. Plot 604 shows larger effects, but the possible washing of dried blood from the adjoining plot may account for this difference. THIRTY YEARS OF CROP ROTATIONS IN ILLINOIS. Plot 709 compared with the average of 707 and 708 shows an average increase of 2 bushels per acre from lime from 1901 to 1905. From all of the investigations thus far conducted on the com- mon brown silt loam prairie soils of the Illinois corn belt, it is evident that the time is near for most of this soil, and already here for some of it, when ground limestone must be applied to give the most profitable results in crop yields. These soils were not rich in lime in the beginning, and lime is constantly being lost from the soil, both in crops removed and in the drainage waters, which per- colate through the soils and when collected in surface wells are al- ways recognized as "hard" waters because of the lime contained in them. EFFECT OF NITROGEN AND PHOSPHORUS. The effect of nitro- gen is determined* by comparing plots 605 and 606 with plot 603. These results are summarized in the lower part of Table 8. They show but very little effect from nitrogen in 1901-2, the average in- crease being about i l / 2 bushels. This was raised to about 5^ bush- els in 1903-4 to 13 bushels in 1905-6 and to 16 bushels in 1907. It is noteworthy, however, that this increasing difference is not due to an actual increase in yield on the nitrogen plots, but rather to a decrease in the yield of the plots not receiving nitrogen, so far as we can judge. By comparing plots 608 and 603 we have the most trustworthy data as to the effect of phosphorus with no addition of nitrogen or potassium. This shows an average increase of 5.4 bushels during the four years, 1901 to 1904, chiefly produced during the second and third years. The gain was 3.5 bushels for 1905-6; while the effect in 1907 was a loss of 5.1 bushels. The somewhat higher yields on plot 607 may be due to dried blood washed over from 606. In comparison with nitrogen, it will be seen that at the begin- ning of soil treatment in 1901 phosphorus was clearly the limiting element and produced a greater effect than nitrogen during the first four years, after which the effect of phosphorus soon decreased un- til it produced smaller yields than the untreated plots, while the ef- fect of nitrogen has been nearly to maintain the actual crop yields, with increasing differences above the untreated land. The lesson taught by these experiments is that on our ordinary corn belt prairie soils phosphorus is the most deficient element of plant food, because of which nitrogen applied alone produces little or no increase in crops ; but in continuous corn growing the supply of nitrogen is so rapidly depleted by removal in crops and in drain- age waters that within a few years nitrogen itself becomes the most *Plot 605 is probably the more trustworthy of the two nitrogen plots be- cause of possible wash from 607 to a part of 606. 354 BULLETIN No. 125. [May, deficient element, after which phosphorus alone has no power to increase the crop yields. Thus on plots 605 and 606, more than 60 bushels were grown in 1900, and, although nitrogen has been applied every year since, that yield has not been equaled; whereas on plots 607 and 608 phosphorus actually increased the yields for two or three years.* After that time the yields have been nearly maintained on the nitro- gen plots but have steadily decreased on the phosphorus plots, ni- trogen having become the limiting element. If, now, we first restore the nitrogen so as to remove the nitro- gen limit to crop yields, then phosphorus should again have power to increase the yield above that produced by nitrogen on plots 605 and 606, which averaged 53 bushels for 1906 and 1907. This has been done on plots 701 and 702, where both nitrogen and phos- phorus have been applied and where the average yield for 1906 and 1907 has been 69 bushels per acre. This is 16 bushels higher than on the nitrogen plots (605 and 606) and 30 bushels higher than on the phosphorus plot (608). This is shown in greater detail in the summary in the lower part of Table 9. The effect of nitrogen and phosphorus above that of phosphorus alone was 3.3 bushels in> 1901-2, 16.5 bushels in 1903-4, 25.4 bushels in 1905-6, and 31.9 bushels in 1907; where- as, the effect of phosphorus applied in addition to nitrogen was 7. i bushels the first two years, about 16 bushels the next four years, but only 10.8 bushels in 1907. Here again the apparent gain for nitrogen is not due to increasing yields on plots 701 and 702, which as a matter of fact were about the same in 1904-5 as they were in 1906-7, but the difference is due rather to the actual decrease on the phosphorus plot (608) where nitrogen has become the limit- ing element. It is very apparent that nitrogen alone cannot increase the yield of corn on this soil, and that the increase from phosphorus alone is very temporary, but that both nitrogen and phosphorus together have power to produce a decided increase. This is in harmony with the marked benefit of phosphorus on clover, which has power to secure nitrogen from the air when the soil nitrogen becomes de- ficient, as shown in the three-year rotation (Tables 3, 4, and 5). EFFECT OF POTASSIUM. The effect of potassium can be ascer- tained by comparing plot 609 with plot 603, also by comparing 707 and 708 with the average of 701 and 702. Under the most favor- able conditions, with both nitrogen and phosphorus present, the addition of potassium produced an average increase of 3 bushels in *The year 1901 must be overlooked because of the drouth. /90\ v -T . r ^Bt ^ -r\ /^**-ai..k/ /IK.^ ^-