ALBERT R. MANN LIBRARY New YorK STATE COLLEGES OF AGRICULTURE AND HOME ECONOMICS AT I | i CORNELL UNIVERSITY i C. H. Potter & Co., Inc. Book BINDERS 310 12TH STREET N. W. Washington, D. C. Cornell University Libra! SB 205.A4E92 study of the fertilization of alfalfa A STUDY OF THE / FERTILIZATION OF ALFALFA FLOWERS M. W. EVANS Office of Forage Crop Investigations Bureav of Plant Industry U. S. Department of Agriculture. C.F ~~ DO Ad Ee ad S2R 207 TABLE OF CONTENTS Page Introduction-~-~---------- 3 ne ene ree nnn 3 Description of the Alfalfa Flower ----------------= as Condition of the Flowers in Alfalfa Fields -------- 6 Description of the Experiments and Discussion of the Results -------------]8-e8=- y Comparative proportions of flowers producing seed under natural conditions when protected from insects and when artificially tripped---7 Effect of protecting the alfalfa flowers from sun and wind by a single thickness of mosquito netting, but with provision for ready access of insects------------------ 14 Comparative number of pods and seeds forming under natural conditions from flowers which have been tripped and from flowers which have not been tripped--~---~---~--------- 16 Effect of fertilization with pollen of different degrees of relationshin------------ 19 Pods formed when stigma and pollen do not come into contact with any object when the flower is tripped -----~--+--~----------------- 26 Pods and seed produced from automatically tripped flowers------------------------------ 28 Pods and seed developing from untripped flowers -~-- =e eee nee ene eee eee ee ee eee eee 40 Effect of tripping alfalfa flowers mechanically------------------------------+--- 40 Period during which alfalfa flowers may be pollinated------------------------------~- 42 Efficiency of Various Insects in Causing Fertilization of Alfalfa FPlowers----~------------- 44 Domestic Honeybees, bumblebees, and wild bees (Megachile spp.) ------------------------ AA (1) Variation in the Percentage of Flowers Producing Pods, and Number of Seeds Per Pod, on Different Plants ----------- 9-3-2 eee eee en eee ee 51 SUMMALY © - - mw me et ee ren rn nese nme en eee 53 Append SHH Ses eee eee ie 60 (2) INTRODUCTION. In 1907 a study of the factors controlling the pollination of alfalfa flowers was begun by the writer, who was then located at Pullman, Washington, in charge of the cooperative experimental work with forage crops being carried on there by the U. S. De- partment of Agriculture and the Washington State “x- periment Station. The investigation was begun in 1907 and was continued during the summers of 1908, 1909 and 1910. Most of the work was done on the experimental plots at Pullman, Washington, though considerable of the work done at Pullman was dupli- cated in alfalfa fields near Chinook, in northern Montana. Chinook is located in one of the most im- portant alfalfa seed producing districts in the North- west. The suggestion was made to the writer in July, 1907, by members of the Office of Forage Crop Investigations of the U. S. Department of Agriculture, that the relationship of the factors controlling the pollination of alfalfa flowers to the development of the alfalfa seed crop was a problem that had never been thoroughly investigated. At that time there was but a comparatively meager amount of information on this subject, and practically no investigations that have been recorded had been carried on in the United States. Since 1907 several other investi- gators, including three members of the Office of Forage Crop Investigations, have been making a study of the pollination of alfalfa flowers. This re- port, however, contains only a discussion of the re- sults which were obtained by the writer in the work conducted at Pullman, Washington, and at Thinook, Montana. DESCRIPTION OF THE ALFALFA FLOVER, Before the bud of the alfalfa flower has developed into a blossom the brosd standard petal is wrapped closely around the wing petals, which in turn tightly enclose the keel. As the flower develops, the standard unfolds and at the same time curves back- wards, while the two wing petals also unfold from over the keel. The two keel petals, however, remain joined together to about one-half the distance from the tip back toward their base. The stigma and the stamens are so situated that they are closely enveloped in the tip of the keel, with little chance for any pol- len from another flower to reach the stigma as long as the keel petals are united. If the tip of a pencil, toothpick, grass Sale stem, or some other similar object, be gently pushed down between the standard and the keel in such a way that the latter is pressed backwards, the strain will disunite the two petals of the keel. As soon as the restraining force of the united keel petals is removed the pistil and the staminal column instantly spring into the new position up against the standard. The flower is now "tripped". When the flowers are tripped by honey-gathering insects both anthers and stigma usvally come into contact with the body of the insect during the process. As the pollen adheres to different portions of the insect, there is abundant opportunity for the flowers to become cross-fertilized with pollen from the flowers of other plants. CONDITION OF FLOWERS IN THE ALFAL#s FIRLDS. Before any experiments had been planned or executed an examination was made of the condition of the alfalfa flowers in the fields near Pullman. It was found that a large proportion of the flowers had not been tripped and that the pistils of most of the tripped flowers were in a fresh, turgid condition; on the other hand, it was observed that many flowers had been tripped and that most of the tripped flowers were more or less wilted. During bright, warm days domestic honey bees and other honey gathering insects were more or less abundant in the fields. It was not known, however, whether it was these insects or some other agencies which had caused the flowers to become tripped. Furthermore, there was considerable difference of opinion in regard to the self-fertility of the alfalfa flower. DESCRIPTION OF EXPERIMENTS AUD DISCUSSION OF THE RESULTS. Comparative Proportions of Flowers Producing seed under Natural Conditions when Protected from Insects and when Artificially Trippred. In order to find out something in regard to the manner in which the alfalfa flowers become tripped, and what the effect of excluding insects from the plants is, an experirent was planned and carried out in 1907 and was repeated in 1908, 1909 and 1910. A considerable number of alfalfa plants were used in carrying out this experiment. The stems of each plant were divided into three groups each having an approximately equal nun- ber of flowers. The flowers on the first group of branches were left to develop under natural condi- tions. The other two groups of branches were en- closed in a small tent or cage made of a fine-meshed mosquito netting or tarlatan having about twenty-four meshes to the inch, which effectually excluded all honey gathering insects. The flowers on one of these groups of branches were left to develop with no other treatment than to exclude insects. On the third group of branches a number of flowers were artificially tripped by means of a toothpick, green alfalfa stem or ee some other similar object. In carrying out this experiment the flowers which were tripped artificially were not cross-pollinated. As a single toothpick or grass stem was ordinarily used to trip a number of flowers on one plant, many of the flowers might have been fertilized with pollen carried on the toothpick from one flower to another on the same plant. In carrying out the experiments in 1907 the raceme was taken as a wnit and no count was made of the actual number of flowers that were on each raceme. From 11 racemes which developed under natural condi- tions, 49 pods developed; from 36 racemes of flowers from which insects were excluded, 28 pods developed; and from 12 racemes on which the flowers were artifi- cially tripped, 25 pods developed. These results obtained in 1907 substantially agree with those ob- tained in the three succeeding years. The results obtained in 1908 are given in Table I. Table I. (Results of Alfalfa Tripping Experiments in 1908.) :Total:No. :Total:- of Tlow-:No. “HO. 2000S m6. <+67rs dével=-:seéds : Flow-:devel-:seeds:oping pods:ver :ers. :oved. : :pod. Outside netting: : conditions nat-: : : ural 869 BOY £.-BYa% Bose oert Inside netting; ‘ flowers not tripyved 2° OLE 26 32 fad Lees Enclosed in net-: ting; tripped artificially 576 148 205 Pie Lewd In 1909 the experiment was carried out in the same way at Pullman, ‘Jashington, and also at Chinook and Havre in northern Montana. are Piven in Pable 11. The results Table II. Pods and Seeds Forming Outside Wetting, Inside Netting, and from Flowers artificially Tripped. Outside Netting, Conditions Nsetural. Per cent of flow- No. of No.of ers de- seeds plants Wo. of No. of pods Yo.of veloping per used. racemes flowers formed seeds pods "Od, Pullman - 15 90 1468 480 1730 32.69 2660 Chinook - 10 76 944 143 420 15.14 Boo Havre - 9 41 566 51 15.95 34 207 2776 G74 F2050 24,25 “Seed Inside Netting, Insects Excluded. Pullman - 15 90 1500 131 opT 8.75 ale Chinook - 10 104 1186 138 oon 1It,65 2.04 Havre - 9 60 5OoD 30 5.60 O4 eb4 Beal 299 Goo. “Gseo “eect Enclosed in Netting, Flowers Artificially Tripped. Puliman = 15 LOL 1379 bo? 1783 45.43 Beat Chinook - 10 64 830 370 681 44.57 1.84 9 46 537 «+104 30.86 "sn “ET ean OYE Faded 2ney> eee Havre * umbers marked with * refer only to re- sults obteined at Pullman and Chinook. Bp We In 1910 the same experiment was reneated at Pullman, ‘Vashington. The results are given in Table III. fable LL. (Results of Alfalfa Tripping Experiment - 1910.) :Total:No. 2NO. os. of flow-:Fo.ma- “nO, < 81.87 = 1.82 The comparatively small cercentage of the flow- ers outside of the netting and of the flowers artificially tripped, which developed into pods, was appartently due to the droughty conditions existing at the time this experi- ment was conducted in 1910. The comparatively small num- - - per of seeds per ;od is probably due to the same cause. The results obtained in 1908, 1909 and in 1910 are combined in Tavrle IV. a sae Results of Alfalfa Tripping Experiments - 1708, 1909 1908 - 1909 - LOL = 1908 - 1909 - 1910 - 1908 - iyo: - 1910 cluded frost. Pullman Pullman Chinook Hevre Pullman Table IV. and 1910. Outside netting. Pullman Pullman Shinook Havre Pullman Per cent of flow- No. Totel Total ers de- seeds no. of No.of no.of veloping per flowers pods seeds pods pod. 869 noe 574 20380 eel 1468 480 L720 Deeb 5260 944 143 320 15 ...14 Bee 366 DL dee 228 a0 BG Les 8 pees Inside netting - not tripped. 812 26 Se Date deter 1500 ISL ABT Ba TZ Bu72 1186 138 ZS Lie 2204 Bod a0 5. 60 554 is 24 4.20 1460 4487 540 "695 Ge Ot 2 ek Inside netting = Pullman Pullman Chinook Havre Pullman 576 148 Leto BO? 830 370 oO7 104 e705 250 205 1783 681 Jee artificially tripred. 25.6 43.43 44.57 50.86 26.350 00,0" Lec Bat 1.84 2.0L. ee 22 The results obtained at Havre in 1909 are not in- for the reason that the seed did not mature before = 12 < The results of these experiments show that when honey gathering insects were excluded from the flowers the usual result vas to cause a decided de- crease in the number of pods and seeds formed. The only marked exception to this rule occurred at Chinook, Montana, in 1909, when excluding insects did not great- ly reduce the number of pods formed. The reasons why such a comparatively large proportion of the flowers, which were enclosed in netting, at Chinook produced pods and seeds will be discussed on page of this report. An examination of the preceding table shows that 23.50 per cent of the 3875 flowers which developed under natural conditions produced pods. Of the 4487 flowers which were inclosed in netting tents to exclude honey gathering insects only 7.57 per cent produced pods or seed. When the 4100 flowers from vhich insects were excluded were artificially tripped the percentage of flowers which produced seed was increased to 36.07 per cent. This increase in the number of pods and seeds formed was very marked in each year and at each lccality where the experiment was conducted. When the flowers were allowed to develop under natural conditions an average number of 3.09 well matured seeds per pod were formed. ‘Yhen the insects were ex- cluded the number of seeds per pod was reduced to 2.24; and when the flowers were artificially pollinated with a ae = no opportunity for pollen to be carried fror the flovers of one plant to those of another, the number of seeds per pod was 2.22,--about the seme as when the flowers under netting were not artificially tripped. These figures indicate that flowers fertilized with pollen from other plants produce more seed than if they are fertilized with pollen from the same plant. Another experiment, discussed on »age , brings out this point more clearly. Effect of Protecting the Alfalfs Flowers from Sum =nd ‘Yind by a single Thickness of Mosauito Netting, but with Provision for Ready Access of Insects. In carrying out the exneriments recorded above those plants or portions of the plants from which insects were excluded were protected by a single thickness of fine mosquito netting, arranged in tent-like form by plac- ing it over pieces of wood from four to six feet long. It was observed when the experiments were being carried out that on the different plants those flowers which were enclosed in the covering made by a single thickness of mosquito netting remained in bloom longer, and that the petals seemed to be larger, than on those flowers which were not enclosed. As it seemed possible that the ef- fect of the slight shade, or possibly the breaking of the force of the wind, might influence the development of seed, an experiment was carried out, in 1908, to ob- tain some information in regard to this point. Bae As Several plants were selected and a portion of eech plant was enclosed in a netting tent, as in the other experiment; the remaining portion of each plant Selected was not enclosed, but was protected on three sides and partially from above by one thickness of net- ting, which was between the plant and the sun, and also between the plant and the prevailing winds, yet did not prevent the access of bees and other insects. The results of this experiment are given in Table V. Table V. Effect of a Covering of One Thickness of Netting over the Flowers. Shaded by one thickness of netting, but not enclosed. Per cent No. of tlow- pods ers de- Plant Id. of devel- Wo. veloping Wo. seeds NO. flowers oped. seeds seeds. per pod. 3 160 55 53 2168 hives 03-a vu DA. a Boe feo 4l Lie ol 88 een pease Bg 39 Be 43 Baas 1.95 89-a ves 18 41 BDed Cent BLS dee 296 Coel bed Entirely enclosed in mosquito netting. 03 141 14 iv 9<9 1.25 Do-8 99 7 ig 7.0 Bw FL ‘4h Lee a 5 5-0 Le 2d 89 105 2 a Lee Led 89-a 43 0 0 0.0° 0.00 520 er 46 ool Ae a Te & As shovm in Table V, of those flowers not entirely enclosed in netting, but “hich were shaded by the netting, 24.7 per cent produced pods and seed. On other plants used in experiments in 1908, 2°.8 per cent of the flowers which were not enclosed °r shaded by netting produced pods and seed. According to these results there was no appreciable effect caused by the screen of one single thickness of mosquito net- ting on the development of seed on these plants. Comparative Number of Pods and Seeds Forming under WVatu- ral Conditions from Flowers which have been tripped and from #lowers which have not been tripped. The figures given in the preceding tables show that when alfalfa flowers are screened from the visits of honey-gathering insects, the effect is to greatly reduce the percentage of pods that are formed from the flowers. However, only a portion of the flowers used in the experiments described above, which were left to develop under netural conditions, were visited or tripped by honey-gathering insects. In order to determine what proportion of the flowers which are actually tripved produce seed, a num- ber of racemes of flowers on different plants were se- lected, in 1908, on which a portion of the flowers hed been tripped by natural causes. On the celyxes of those flowers which had been tripped a small quantity - 16 - of an insoluble drawing ink was applied. after the tripped flowers had been marked the racemes were en- Closed in netting to prevent the access of any more in- sects. Table VI shows the number of pods and seed which developed from tripped and untripped flowers. Table VI- snowing Number of Flowers Tripved under Natural Con- : ditions. NO-« uan= No.pods Total Totel no. nu flow- ber devel- NOs seeds ber Tun- ers pods oped seeds develoved flov- ber not devel- from on from Re- ers flow- tripped oped fls.not cluster fls.not cere in ers when from trimn.ed from tripoed num- ra- trip- in- tripned when tripped when ber ceme ped closed flowers enclosed flowers enclosed ze 11 4 ia o a 2 L1. 8 o 5 353 3 LS DA p 8: 55 4 8 6 S oS 2 5 Le TL 6 7 B3 La 1 6 10 Oe 5 4 14 7 18 10 8 2 2 8 g o O 8 Bo 9 a 6 5 5 at 10 Lo 8 5 a o cme 15 8 7 6 25 AZ 11 7 4 O E O Z Total L50 ee Sad 54 oS iy e Percentage of tripred flowers developing pods - - - ~ - 58.00 Percentage of flowers not tripred when enclosed devel- Oplne (Ots=.4 (== = soe eS Se Se Se eS ee 5.20 lueber Seeds per pod Trem tripoed Tlowers = = == + = = Ge25d Number seeds per pod from flowers not trivved when en- closed -----+--+-+-77-77-777424-2--- LeGO = 29 The figures given in Table VI show that 58 per cent of the flowers which had been tripped by in- sects or by some other natural agency produced seed; whereas only 5.2 percent of those flowers not tripred when enclosed in netting produced seeds. The flowers which were tripped when enclosed in the netting produced 5.53 seeds per pod, while the three flowers which were not tripped when enclosed in netting, which produced seeds, produced only one seed per pod. These results egree with those shown in Table IV. Table VI also indicates that most of the alfalfa seed produced at Pullman, ‘Jashington, in 1908, developed from flowers which had been tripped by honey-gathering insects; and that but a small proportion of the flowers which were not visited by honey-gathering insects produced seeds. ae AG oe Effect of Pertilization with Pollen of Different Degrees ot Relationsnin. As the results of previous exrerirents indi- cated that the manner in which a flower is fertilized has something to do with the percentage of pods and the number of seeds formed per pod, an experiment was planned and carried out in 1909 to obtain more definite information in regard to this problem. On six selected plants a portion of the flowers were tripped with a rough piece of string, each flower being tripped alter- nately with the flowers of one of the other plents. In this way there was opportunity to fertilize each flower with pollen from another plant. On three of these plants the remaining flowers that had been tagged were tripped with pieces of grass or alfalfa stems. As @ sevarate stem was used for each flower, each stigma received pol- len only from the same flower. On each of the other three plants, a portion of the flowers were tripped by pressure from the outside in such a way that no external object cane in contact with the stigma, thus preventing eross fertilization. The results are presented in tables VII and VIII. ee Table Vil. : Flowers tripped with separate Flowers fertilized with pol- : len from other plants. stems; fertilized with pollen from same flower. Plant Nos oF Togot [o.0L 406s Of Wo.of Po.ot number flowers sods seeds : flowers pods seeds De Bo 42 174 47 20 26 bo AT 56 LO 48 L2 25 55 Be BES 34: 48 7 Af Percentage of flowers : developing pods - - 59.09: Seeds per pod Sethe Percentage of flowers developing pods - - 22.16 peeds per pod Le oe Table Flowers fertilized with pol- len from other plants. : pressure; VII. Flowers tripped by outside fertilized with pollen from same flower. Plant No. of No.of llo.of : No. of No.of No.of number flowers pods seeds : flowers pods seeds 54 42 34 154 59 30 40 56 48 26 66 37 18 28 57 51. SS doe: oe 51 15 54 141 95 Se6 147 60 T0z Percentage of flowers developing pods - - 67.57: needs per pod Oe42: : Percentage of flowers developing pods - - 42.85 Seeds per pod Ladd In both tables the flowers fertilized with pollen from separate plants produced a larger percentage of pods and a larger number of seeds per pod than the flowers fertilized with their own pollen. The results of this experiment may are x perhaps be subject to the criticism that the flowers fertilized with their owm pollen should have been tripped with pieces of rough string,as the flowers fertilized with pollen from separate plants were, instead of with a smooth plant stem or by pressure; for the irritation of the stigma with the rough string may possibly have had some influence noon the development of the flowers into pods. In 1910 an experiment was conducted in which flowers which were fertilized with pollen from other plants, as in the experiment just described, were com- pared with flowers which were artificially fertilized with pollen from other flowers of the same plant. The flowers fertilized in each of these two ways were tripped in the same manner, by using a piece of alfalfa stem wound with thread. The results of this experiment are given in Table Ti. OL. Table 1X. Plants enclosed in netting; flowers tripped, with alfalfa stem wound with thread, alternately with flowers of another plant:- Plant Number Number Number Number Total number racemes flowers pods mature Humber seeds seeds 2 3 38 18 at 39 2 4 50 BL BL So 3 4 54 al 110 113 4 4 52 4 7 ? 5 4 48 24 5 49 6 4 45 10 3 15 q ioe ae. 20 18 58 27 338 118 eb 334 Percentage flowers producing pods--- 34.8 Number mature seeds per pod--~<- -<------ lee? Total number seeds per pod---------- 2485 (Including immature seeds) Flowers enclosed in netting; tripped with alfalfa stem wound with thread; fertilized with pollen from other flowers of same plant. Plant Number Number Number Bumber Total Number racemes flowers pods mature number seeds seeds i. 4 BS 5 "7 7 a 3 41 et 26 26 3 5 56 Lo 16 19 4 4 Te 14 ae 22 5 4. 56 30 at 73 6 2 23 2 S 5 7 5 52 9 6 10 re 357 84 109 Lé6e Percentage flowers producing pods ------ Zoe Number mature seed per podq----+-+ --e----- Led Total number seeds per pode~------eeeee= 1.93 The smaller percentage of flowers which produced pods and the smaller number of seeds per pod, from both cross fertilized and self fertilized flowers, in 1910 than in 1909, was apparently due to the drier weather conditions which existed in 1910. The percentage of flowers producing pods, the number of mature seeds per pod, and the total number of seeds per pod, is about 50 per cent more in each instance in 1910 where the flowers were fertilized with pollen from other flowers of the same plant. (23) The results of this experiment agree substantially with the results of the experiment conducted in 1909. In both instances the flowers which had been fertilized with pollen from other plants produced a larger proportion of pods and more seeds per pod than did the flowers which had been fertilized with their own pollen or with pollen from other flowers on the same plant. Another experiment was carried out in 1910 to obtain some information in regard to the relative effect of fertilizing with pollen from other flowers of the same plant compared with the effect of fertilizing with pollen from the same flower. The flowers that were fertilized with pollen from other flowers of the same plant were tripped with a piece of medium coarse wrapping twine, a single piece of twine being used to trip all of the flowes tripped on one plant, in order that pollen might be carried from one flower to another. The flowers fertilized with their own pollen were tripped by using a separate piece of twine for each flower. The hands were sterilized after tripping one flower, before tripping the next, in order to prevent pollen being carried from one flower to another. Both methods were used on flowers of each of several differ- ent plants. (24) Table X. Comparative effect of pollen from same flower and of pollen from other flower of the same plant, flowers enclosed in netting; fertilized with pollen and other flowers on the same plant. Number Total Plant Number Number Number mature number number racemes flowers pods seeds seeds warren BR wea G4 ee eee HQ eee een -------18 ween BG we eee 89 mew een HQ eee wee ] oe ----18 maemo Bee ene HBG eee eH HGH eee HH HG ee - ----90 wee eet 5 SSsSese 0 eeeteeky ssreerelS ereaeas] . 23 228 oe 63 149 Percentage of flowers developing pods---30.7 Number mature seeds per pod-------------- 9 Total number seeds per pod---~----------- Died Flowers enclosed in netting, fertilized with pollen from the same flowers. Number Total Plant Number Number Number Mature number number racemes flowers vpods Seeds seeds weccce So wwe ee 45 0 mene HQ Hee eee GQ eH HH-- 11 woreeee BS treet 29 wenn Q nme e ee OO ee H--= 2 meee ee G wr erem GL ame en4Q wenn nn e23 cnnn-= 55 eeseeee DO seen n= 50 ----89 -------24 ------ 34 eg | T83_ "ae "56 ee her ae Percentage flowers developing pods ---=-- 5943 Number mature seeds per pod------------- re Total number seeds per pod--~--~------=-= 2el9 The figures presented in Table X do not show a very mark- (25) ed.difference in the effect of the pollen from separate flowers of the same plant, compared with the effect of the pollen from the same flower. The average percentage of flowers producing pods is slightly less, while the number of mature seeds per pod is slightly more from the flowers fertii- ized with pollen from other flowers of the same plant, than from flowers of the same plant, than from flowers fertilized with their own pollen. The total number of seeds per pod developing from the two different methods of fertilization is almost exactly the same. Pods Formed When Stigma and Pollen Do Not Come Into - Contact With Any Object When The Flower is Tripned. Two experiments were performed in 1910, to find out what the effect would be when the flower is tripped, but when the stigma does not come into contact with the surface of the object tripping the flower or in contact with the standard of the flower. In the first experiment the standard was removed; then the flower was carefully tripped in such a way that the stigma and pollen did not come in contact with any object during the process. In the second experiment, the flowers were tripped by cutting the lower part of the anterior side of the keel with a sharp razor, so that the tip of the keel re- mained enclosing the anthers and stigma moving up with the staminal column when the flowers were tripped, thus prevent- ing the stigma and anthers from emerging from the inside of (26) the keel or from coming into contact with the standard or any other object. For the sake of comparison other flowers on the same plants were tripped by using a single alfalfa stem for all the flowers on one raceme. Other flowers which had been tripped naturally were also marked. A number of flowers which were untripped at the time when the flowers were inclosed in netting were also marked. Five plants were used in this experiment. The results are presented in Table x1. Table Xl. Pods developing from flowers tripped without pistil and stamens coming into contact with any object. / Percentage Total Number of flowers number pods developing flowers found pods. Standard removed before tripping =<<< <<“4+ss ese SesseHne> 113--<--+- 16 ---8--+- 1425 Tripped by cutting lower part of keel, stigma and pollem remain in tip of keele- 6>--<-~>= 10-222 ne 13.16 Flowers tripped with stem Bifalia. planio<+==-s++s+ssenee= 120 --- 2 == +55 - eee ee ALS Flowers tripped naturally----- L7B eee cr en OB ere eee 52.24 Plowera not tripped+<=<=-+--<- 3o8--sss+= 18 ------+-- 5.02 Table X1 shows that when the flowers were tripped with- out bringing the stigma and pollen into contact with any. (27) object, a comparatively small proportion of the flowers produced pods. In other words, this experiment indicates that when the staminal column only moves from its original position to the curved position it assumes when tripped, without rubbing the stigma and pollen against the object tripping the flower or the standard. The percentage of pods produced is much less than when the flowers are tripped in the ordinary way. Pods and Seed Produced From Automatically Tripped Flowers. The term automatically tripped is applied to an alfalfa flower which has become tripped without the aid of insects or any other external agency. The results of the experiments which have been recorded in this report indicate that the presence of honey gathering insects, which trip the flowers that they visit, under at least some conditions is essential in alfalfa fields where seed is to be produced. In the fields about Pullman, Wash. there are usually a considerable number of honey gathering insects flying about when the flowere are in bloom. On the other hand, there are certain localities in the Northwest, where alfalfa seed is produced, sometimes in very large quantities, though honeybees and bumblebees are quite rare. In the vicinity of Chinook, Montana, several car loads of alfalfa seed were produced in 1908, and also in some other years, though honey gathering insects were almost entirely (28) absent in most of the fields. This seemed to indicate that, under centain conditions, seed may develop abundantly ®@ith- out the presence of insects to trip the flowers. In the study of the fertilization of the alfalfa flowers conducted at Chinook and Havre, Mont., in 1909, it was espec- ially desired to find out whether alfalfa flowers sometimes do become automatically tripped, and develop seed without the presence of insects or other outside agency to trip the flowers, or whether the flowers might rot develop seed under some conditions without the flowers being tripped. This problem was rendered difficult to sclve, however, by the fact that tn 1909, in nearly all fields, but a very small proportion of flowers were developing pods, and these pods seemed generally to have been tripped by the limited number of insects which were present, or by some other external means. It was observed, however, that two of the plants enclosed in netting at Chinook, which had been producing a good crop of seed throughout the season, while most of the surrounding plants were producing but very few pods, were producing pods even from those racemes of flowers from which insects had been excluded. Accordingly, all wilted and all unopened flowers were removed from a number of racemes under the netting, leaving only opened, untripped flowers, which were closely observed during the following days. In a day or two, it was observed that several of the (29) flowers, from which insects were excluded, had become tripped. On one of these two plants,- plant numbered Mont- ana B,- in a number cof the flowers the keel petals were partially separating, evidently being forced apart by pressure of the pistil and stamens from beneath. While these flowers were being examined, one flower was seen to become self tripred. The pistil and stamens snapped up vigorously against the standard, scattering the pollen around, no object had come into contact with any portion of the flower. The calices of all tripped flowers were marked with carbon ink as soon as they were found to be tripped. The number of flowers that were tripped, and the pods that developed from tripped and untripped flowers, are shown in Table X1lll. (30) Table X111_ Pods and seeds from self-tripped flowers at Chinook, liont., 1909. Number Vumber Percent- Percent- pods pods are of “unber are of Total from from un- flowers Plant Raceme Number flowers flowers number tripped tripped forming tont. number number flowers tripped tripped nods flowers flowers pods " 2 34 Le 9 75.00 ¥ 7 0 Bo. 2S ” 2 35 La oS Bisset 3 3 0 Beet as S 6 7 G £9557 0 0 0 00,00 . o or At 7 65.62 5 S 0 45.45 2 5 38 10 9 90,00 5 5 0 50.00 " 2 39 6 3 50,00 i a 0 _16.66 57 oo by .a88 a, el 0 36.84 ont. 8 30 14 5 O5.471 3 % 0 21.42 . 8 al 6 2 26-55 0 0 0 00.00 . 8 oe 20 14 70,00 8 8 0 40.00 : 8 3 ck 2 Loek8 0 0 0 00.00 8 34 de LS 100,00 5 5 _O- 38.46 64 36 56.25 16 16 0 25.00 Percentage of all flowers tripped on pls. ont. 3 and Mont. 8--------- 57.02 ©ercentage of all flowers producing seed om pls. lont. 3 and Tiont. 8--30.57 Percentage of tripped flowers producing seed on pls. itiont. 3 and ont. 8---53.62 Percentage of untripped flowers producing seeds on pls. siont. 3 and !ont. 8---00.00 (31) The results shown in Table X111 show clearly that alfalfa flowers do sometimes become self-tripped. The figures in this table also show that. on these two plants a large percentage of the flowers that were self-tripped produced pods and seed. By refering to Table 1V, it will be observed that in each of the three years in which the pollination experi- ments which are summarized in this table were carried out, of those flowers which were not tripped when enclosed in mosquito netting, a small percentage produced pods every year. In 1910, nine alfalfa plants which were producing more pods and seed than most of the alfalfa plants in the aan garden at Pullman, wete inclosed in netting tents,. and were closely observed for several days to determine whether the small percentage of flowers that frequently produce pods when honey-gathering insects are excluded from the alfalfa plants, become self-tripped, or whether the flowers may produce pods without becoming tripped at all. The tents in which the plants were enclosed, were carefully covered with fine mesh mosquito-bar, so that there were no openings left for honey-gathering insects (32) to gain access to the flowers. The tents were made large enough and pains were taken so that no flowers which were being watched were in such a position that they could brush against. the sides or top of the tent. After the flowers from which insects were to be ex- cluded had been enclosed in the netting tents, they were examined every day or every second day until all of the flowers had become entirely wilted. Whenever any flower was found tripped, the calyx was marked with a mixture of carbon-black and water, in order that the pods that developed from tripped flowers might be distinguished from pods that might develop from flowers that had not peen observed to be tripped. Table X1V shows the number of flowers that became tripped and also the number of pods and seed that develop- ed from tripped and untrivped flowers. (33) Table X1V. Pods and seed from self tripped flowers at Pullman, Wash., 1910. Number mature. Mature Number seed Number seed from Number flowers Number from pods pods Plant Number flowers not marked marked not not number flowers tripped tripped vyods pods marked marked 8 128 9 119 i 0 0 0 9 104 Le 87 12 18 4 4 10 59 2 57 lt 0 1 0 J. 100 LS 85 4 3 0 0 22 62 4 58 0 0 0 0 13 91 i: 90 1 i. 0 0 18 79 3 76 2 4 0 0 19 tha 8 63 0 0 oO 0 20 81 ne 80 0 0 0 O 775 60 Tis aL 26 5 4 Percentage of flowers tripped--------------=-- tO Percentage of tripped flowers producing pods~-35.00 Percentage of flowers not tripped proqucing Tlowers=-s-=-+s-so>+--See sere eee eee -69 Mature seeds per pod from tripped flowers---- 1.23 Mature seeds per pod which developed from flowers not observed to be tripped----=- 80 (34) Table X1V shows that 7.7 percent of the flowers observed on the nine plants became tripped. In several of the flowers the two petals forming the keel were ob- served to gradually draw apart, and later the flowers were found to be tripped. As the tents were made so that the honey gathering insects did not have an oppor- tunity to have access to the flowers, as care was taken to prevent any other object from coming into contact with the flowers, it may presumably be assumed that nearly all of the flowers that were tripped were tripped automatically. The evidence obtained from this experiment and also from other experiments conducted at Pullman, Wash. indicate that in most seasons at Pullman, ona large proportion the alfalfa plants, a small percentage of the flowers become self-tripped; and that a consider- able proportion of the limited numbers of flowers which pecome self-tripped produce seed. In the latter part of August, 1910, alfalfa fields at Chinook, Mont., which were producing seed, were ex- amined, Two fields were found on which pods and seed were developing in umusual abundance. Both fields were profusely in blossom. On August 20 and 21, nearly all flowers that had been open for about one day or more, were tripped. A typical raceme, picked at random was found to have on it twenty opened flowers. Of this number (35) fifteen were tripped. The five not tripped were all some of the most recently opened flowers located at the tir cf the racere. An examination of a large number of plants in the field showed that during a period of about two weeks, or possibly for a somewhat longer period preceding August 20t2, the majority of the alfalfa flowers had been devel- oping into pods. Prior to this, there was a period during which but 2 small percentage of the flowers had developed into pods, as was ‘indicated by the stems of the older flower racemes, the majority of which were either without pods, or elise had sate @ small number of pods on then. On August 22, flowers which were just beginning to open were marked on ten different plants in these two fields and were watched in order to determine what was causing these flowers to become tripped. Table XV shows the number of flowers that were observed, and the number and percentage of flowers that had become tripped inside and outside of the netting during the four days after the flowers were first observed. Most of the flowers had become wilted by the latter date. (36) Table XV. Number of flowers becoming self-tripped in alfalfa fields at Chinook, Mont. 1910. Inside Netting. Number Plant Number Number flowers number racemes flowers tripped 9 5 29 O 10 5 48 0 Loe 6 41 0 i2 8 43 0 13 6 3 i te. 390 _ Percentage of flowers tripped=------ 00.25 Outside Netting. Number Plant Number Number flowers number racemes flowers tripped 9g 5 28 i 10 5 40 2 Lt 5 36 0 iz " 47 0 a 5 36 > 14 4 a5 0 15 6 39 0 16 5 32 0 L7 4 18 0 18 4 ae 2... 50 Doo 4 Percentage flowers tripped--------- Lee (37) As Fable XV indicates, very few flowers were tripped during the period from August 22 to August 26, either in- side or outside of the netting tents; whereas, in the few days preceding August 22, very nearly all flowers in these two fields became tripped after being open for a short time. Prior to August 22, the atmosphere was clear and fairly warm. By August 22, it became much colder; a hard frost occurred on the night of August 24th. During the three or four days following August 22, it was ob- served that when alfalfa flowers were artificially tripped the pistil and stamens did not snap up against the stand- ard as promptly and vigorously as in the case of those ‘flowers that were artificially tripped during the days preceding August 22. The only natural external agencies that could have caused the flowers to trip and produce pods so abundantly just prior to August 22 would have been honey-gathering insects, or possibly the action of wind. The latest date on which there was any appreciable amount of wind, prior to August 22, was on August 18. Flowers were watched, in another field during this wind, the wind was not found to trip any appreciable percentage of the open flowers. During the period of August 19 to 22 inclusive, when nearly all of the flowers which opened became tripped, there was no wind blowing sufficiently strong to cause the plants to sway enough to disturb the (38) alfalfa flowers. Wind, as a possible agency that might have caused the flowers to be tripped during this period, should, evidently be eliminated. During the period from August 22 to August 26, several hours were spent in the alfalfa fields each day; during this time not a single wild or domestic honeybee, or a bumblebee was seen. In the more favorable weather conditions that existed during the three or four days preceding about five or six wild bees of the genus Megechile were observed flying about the fields. No domestic honeybees were seen. The honey gathering insects that were present in the large fields of alfalfa, were present in such limited numbers that not more than a very small percentage of the flowers that were tripped during this period, could have been tripped by these insects. It was clearly demonstrated at Chinook, Montana, in 1909, and at Pullman, Wash., in 1910, that alfalfa flowers do become automatically tripped under certain conditions. The most plausible explanation for the fact that during the period immediately preceding August 22, nearly all of the alfalfa flowers in these two fields became tripped and developed into pods, while during an earlier period but few pods were produced, and during the period subsequent to August 22 practically no flowers became tripped, seems to be that just prior to August 22 condi- tions were favorable for alfalfa flowers to become (39) automatically tripped in large numbers, whereas during a period earlier in the season, as well as during a period following August 22, conditions were unfavorable for the flowers to become tripped in this manner. Pods and Seed Developing from Untripped Flowers. Table X1V shows that five of the 775 flowers observ- ed produced pods, when no evidence that these flowers had been tripped could be found. There is a chance for possible error, as some of these flowers mey have become tripped without being observed; or the carbon that may have been placed on the calyx may have been removed. However, two of these flowers were otserved, in which a rod had begun to develop, and where the tip of the young growing pod pushed through the tip of the keel, while the flower still remained untripped. These two flowers un- deubtedly produced pods without having been tripred. It is evidently only in rare instances, however, that a pod and seeds develop from an alfalfa flower which had not been tripped in some manner. Effect of Tripping Alfalfa Flowers Mechanically. In 1910, a number of racemes of opened alfalfa flowers on three different plants were clasped between the pages of a notebook. A record was made of the number of flowers that were tripped and which remained untripred, and of the pods and seeds that developed from them. The pods that (40) Pie Om ow developed from tripped flowers were distinguished by marking the calyxes of the tripped flowers with earbon- black. The object of this experiment was to illustrate the results that might be obtained by tripping alfalfa flowers in the field by any mechanical arrangement that might be devised to do this work. Table XV1l. Percentage of Flowers Tripped by Clasping Between the Pages of a Book. No. Number TO. mature mature No. flo. seed No. seeds Yo. flo. not Mark- from pods from pods race- No. tripp-trivp- ed marked not not mes flo. ed ed pods pod marked marked 4 SL 19 ie 14 14 aL 0 4 57 29 20 ee 29 i) 5 0 69 26 43 3 3 0 0 is 57 “4 BS 39 48 a 5 Percentage of flowers tripped---------------- 47.1 Percentage of marked flowers producing POdS — rn mr ne ner ren ne eee nee enenees= oes Fercentage of unmarked flowers producing pods ia CRA al RR a a Nal ag 4.8 (42 ) The results obtained show that 47.1 percent of the flowers were tripped by clasping tna paeeuee between the pages of a book, and that 52.7 percent of the tripred flowers produced pods; and that 4.8 percent of the flowers which were not tripped by this operation produced pods. The figures shown in the table indicate that any mechanical method that might be devised which would trip flowers in aifalfa fields, would, under conditions like those under which this experiment was conducted cause a considerable increase in the total vield of alfalfa seed. Period During Which Alfalfa Flowers May be Pollinated. An experiment :lanned to show at what stages of its development an alfalfa flower may be fertilized, and also to show during how long a period it remains capable of fertilization, was carried on at Pullman, Wash. in 1909. All opened and wilted flowers were removed from a number of racemes on five different plants. On the following day all unopened buds on these racemes were removed, leaving only those flowers which had opened during the preceding thirty-six hours. The flowers had been covered with netting tents in order to exclude insects. As the experiment was carried out in September, when the weather was comparatively cool, the flowers remained fresh and open for @ longer period than they would gave remained open in warmer weather. A mumber of the flowers on these (42) racemes were tripped each day up to the end of seven days when the tips pf some of the petals were beginning to wilt. The experiment wes discontinued at this time for the reason that there were no more flowers left to trip. The number of flowers that were tripped and the percentage of tripped flowers which produced pods are shown in Table X11. Table Xll. Showing period during which alfalfa flowers may be fertilized. Plant Plant Plant Plant Plant 54 58 59 60 61 : £i0s 82 > 10% ot & £16. 33 Co £16. =: Ee PG Tripped:No. flo. develop::No.flo. develop-:: Mo.flo. develop-:: No.flo. develop-:: TIo.flo. develos at the :tripp- ing titripp- ing :: tripp- ing :: tripp- ing it trippe dine end of : ed pods 1: ed nods ::led pods = ed nods nS ed pods L Gay + OL Ca.eo an 0 Ls 75.00 8 16. 66 2 le 30.00 2 days 14 21.42 LY 64.70 19 die te 19 5.20 3 days 26 7347 18 0 19 52.69 16 Le. 75 ask eoebe 4 days 16 6.25 16 Dt soe 15 26.66 17 58.82 5 days al 38.09 15 6.66 LG 50.00 15 Oe eas io 00.00 6 days 5 0 16 68.75 Le BG «do 20 65.00 7 days 10 30.00 8 0 9 88.88 14 23.67 16 6.20 a number of flowers were left untripped on these plants. Less than five percent of the un- tripped flowers produced pods. An exauination of Table X11 shows that there was practicully no dimination in the ability of the flowers to become fertilized up to the end of a period of seven days ofter they had opened, or up to about the time when the tins of the petals began to wilt FFFICIENCY OF VARIOUS INSECTS IN CAUSING ee ee en me mn ne me a ee eae ee ee neem GU FERTILIZATION OF ALFALVA FLOWERS. Domestic Honeybees, Bumblebees, and Wild Bees (Megachile_ Bpp.) t When alfalfa plants are in blossom at Pullman, Wash., particularly om warm bright days when there is not much wind blowing, honey gathering insects are quite numerous in the alfalfa fields. From observations, it was determ- ined that the most common insects which gather honey from the flowers of alfalfas are the domestic honeybees, (Apis mellifera L.); one or more of several species of bumble- bees (Bombus spp.)are usually present; and also a wild bee of the species Megachile. There are a few other species of insects which visit the alfalfa flowers to gather honey, but are present in the fields in only comparatively small numbers. In the vicinity of Chinook, in northern Montana where alfalfa seed is produced on several hundreds of acres annually, only a very limited number of insects are found in the alfalfa fields when the flowers are in blossom. The insects which are rost frequently found are, the domestic honeybee (Apis sp.), bumblebees of different species (Bombus spp.),and one or more species of Megachile. In 1909 and in 1910, insects were watched while at work getting honey from the alfalfa flowers; a count was (44) made of the number of flowers which were visited and of the number which were tripped by insects of different kinds. The results of these observations are given in Table XVil. Table XV1ll. Showing percentage of alfalfa flowers tripped by different honev gathering insects. Percent Total age number of Number ot Where flowers of flowers flowers Date Species observed visited tripped tripped 1909-Apis sp.-~ Pullman,Wash.---2318-------- L-----+- O46. 1910-Apis sp.-- Pullman Wash. o- +l BG enn eee ee Boer eee 1.58 190¢-Apis sp.- Chinook Heat ne -seoees Geren nee 4.76 1909-Bombus spp-Havre, Wont . ween 266 nnn nee 9-H He - 29 40 1909 -Mezachile"-Pullman ,Wash.--- ye el 90.38 1909-Megachile"-Chinook,Mont.--~--45------ BQerceen- 93.33 Table XV11 shows that on SO percent of the flowers visited by the wild bees of the Megachtle spp.were tripped. The bumblebees which were observed were somewhat less effieicnt than the Megachile spp., 29.4 percent of the flowers visited by the bumblebees being tripped. Some of the species of bumblebees appeared to be more efficient than others in tripping the flowers which they visited. Domestic honeybees do not trip more than a small percent- age of the flowers from which they gather honey. From the (45) figures shown in Table XV1l1, together with the results of additional observations made at Pullman, it seems evident that the domestic honeybees do not ordinarily trip more than about one percent, or less of the flowers that they visit. The reason why the Megachile species of wild bees trip so much larger proportion of the Piers. cual: they visit than the domestic bees, is apparently due chaifly to the difference in the way that they reach down to the honey secretion with their probosces. These two kinds of bees are about the same size, and have probosces of about the same length. The bees of the Megachile species reach down directly along the central portion of the standard, and have to push the standard and wing petals apart, to some extent, in order to get to the honey. The force which they exert when obtaining the honey seems to be sufficient to release the rechanism of the flower which holds the ristil and stamens in place, and the flower becomes tripped. The domestic honeybees extend their probosces down to the honey secretion from one side of the central part of the flower. The distance which they have to reach is less than the distance which the bees of the Megachile species have to reach; it is not necessary for the domestic honeybees to force the petals apart to a great extent, and consequently only a very small proportion of the flowers that they visit are tripped {46) Most species of humblebtees have a longer proboscis than either of the two kinds of honeybees which have teen described. They arproach the flower in such a way, how- ever, that a large proportion of the flowers are tripped. The comparatively heavy weight cof these insects,- part- icularly of some of the larger species, appears to have some influence in forcing the petals of the flowers apart. A count was made in 1907 of the number of flowers which several different bees of the species of Megachile common at Pullman, Wash., tripped within a definite period of time. One of these bees tripped 4 flowers within 30 seconds; another tripped 12 flowers in one minute and 10 seconds; while a third tripped 20 flowers in 2 minutes and 15 seconds. These three bees tripped the flowers at the rate of an average number of G.2 flowers per minute or 552 flowers per hour. A single one of these wild bees is capatle of fertilizing a very large number of flowers in one season, and if present in sufficient num- bers would be an important factor in fluencing the develop- ment of the seed crop. At Pullman, Wash., a large propor- tion of the alfalfa flowers which produced seed in the experimental plots, during the four seasons in which this study was carried on there, were fertilized by these insects. In the vicinity of Chinook, Montana, the numbers of these bees or of any other honey-gathering insects which are found there, and which trip the flowers that (47) they visit, are much too small to fertilize more than a small proportion of the flowers which produce seed. Thrips Thrips (Enthrips tritici ) were present in large numbers in the alfalfa flowers at Pullman,and in less numbers at Chinook, and Havre, Montana. At Pullman, 1119 thrips were found on 16 racemes of alfalfa flowers, or an average of 69.9 thrips on the flowers of each raceme. At Havre 48 thrips were found on 13 racemes. These minute insects do not trip the alfalfa flowers. They were abundant on practically all of the plants used in carrying out the experiments which have been described in this report. As only a very small percentage of the flowers which were not at Pullman produced seed, and as the flowers that were tripped produced seed abundantly, it may be assumed that the thrips are neither appreciatly peneficial or injurious in their influence upon the development of alfalfa seed. Night -flying Insects. An experiment was conducted at Pullman in 1909, to obtain information in regard to whether night-flying insects are of any importance in fertilizing alfalfa flowers. Seven plants were enclosed in fine meshed mosqu- ito netting. Five of these plants were left under the netting during the entire time of the experiment in order (48) to find out what proportion of the flowers become tripped when insects are entirely excluded. Two of the plants were kept enclosed in netting during the daytime, but were uncovered during the night. are given in Table XV111l. _Table XV1ll. The results obtained Showing efficiency of night flying insects. ‘Enclosed in netting during entire experiment. : : Percentage : Percentage:of flowers Percentage of flowers developing pods on same plants outside of the os ©8 +e oe we ew ee e oe Plant : of flowers :developing netting- open to number : tripped :pods insects Ll eer erene Q 2: reer ce ree L160 0 tere rene ee ee eee Seok 12 --9---+- O se eeeteee O et ewe eter e ere ne LG gees 14 A a ae 3292 SS A SSE A Se ee Cae EAS ey eee se eee ne ee oe 56.60 15 ---eee-- 3003 wore nnn Re0Q were w ween wenn ZO nas VJ enene eee 0 woe een91.36 ‘ Enclosed in netting in iidaytime ; open for night- flying insects Yemen eee 465 --nn eee by (Drees SSE eee Se =e Do= Log Saat ee 20 sesee = Bo Seeeld.S meee 1.25 {6 =seere 6G) SSeS See aes aes ao eHerlgse sores 2-55 09 Ses. Len Sees So Ss 224 <~----40.5 <----- 3.30 AG woenen- 92 --------- Q ----- QO ae-=- 0.0 --~--- 0.00 B39e-n---= BQ ween e eee 40 wane 200 ween 8,0 ------ 0.50 B3r.---n-- 60 waeneann Lo -ne-- OQ =n--- 1.6 --+=-- 0.00 40 --~-~- 143 --------- 4 wwen 40 anne eh seaees 1.00 40a---==- 133, -weee eH Qo nem Po ww eee 6467 ween Ox? 16 -n-nnnn- BQ annem n- QO w-2-4 QO wnn--0,0 ----ee 0.00 09 ------ 114 --eenn=-- Po --n- 16 ----- 6.1 -+++-- 2.30 AG ------ ol. sees ko wn---- Q ------ 1.2 ----- 0.00 i B50 were ern Jorn en-- GC. S855 20.0 9 --2-- 1.14 83 ween 68 tama me 18 ----~=- 23500 conn 26.4 ----- dee? 4Q ere 134. “Hee 5e5 SO ----<8 88) ----- 44.7 wwe ee 1.46 AQ0a -=89e"136 ---e HH B2ecrcc- 750 -ece- 38.2 eens 1.44 ee Boren 3 tree Bel tee 1.00 09 ----- 65. --seeeeo- Perec nn- 8 -e--- 10.7 9 -+--- 1.14 (52) Of the flowers observed on a large number of alfalfa plants at Pullman, Wash., and at Chinook and Havre, Mont., in 1908, 1909 and 1910, 23.5 nercent of those flowers which developed under natural conditions produced seed. When honey-gathering insects were excluded by means of fine-meshed mosquito netting, the number of flowers pro- ducing seed was reduced to 7.57 percent. Of those flowers from which insects were excluded, which were artificially tripped, 36.07 percent produced seed. The pods which developed from flowers which were open to the access of insects contained an average number of 3.09 seeds each; the pods developing from flowers from which insects were excluded contained 2.24 seeds each, and the flowers art- ificially tripped produced pods containing 2.22 seeds each. When alfalfa flowers were protected from the sun and wind by one thickness of mosquito netting but were access~- ible. to insects, the percentage of flowers which produced pods and seed was not arvreciably different than the pver- centage of flowers producing rods and seed when conditions were natural. These results indicate that the single thickness of mosquito netting which was placed on a portion or all of the plants in the various experiments (53) conducted, did not have any effect except by excluding insects, upon the development of the flowers under it. In 1908, flowers on a number of different alfalfa plants were enclosed in netting; a portion of the flowers had been naturally tripped before they were enclosed in netting, while other flowers had not been tripped. Of the flowers which were tripped before insects were excluded, 58 percent produced seed. Of the flowers which were not tripped at the time insects were excluded, 5.2 percent produced seed. The pods developing from the flowers which had been tripped contained 3.53 seeds each, while the pods developing from the flowers which were not tripped when enclosed in netting contained 1 seed each. The results of this experiment agree substantially with the results of other series of experiments, in showing that at Pullman, Wash., vods and seed ordinarily develop in large numbers only when the flowers have been tripped by insects or some other agency; but that even when insects have been excluded a comparatively small percentage of the flowers produce seed. The results of this experiment also indicate that a larger number of seeds per pod develop when the flowers have been tripped by insects, than when the flowers have been tripped under conditions where there is not an oppor- tunity for the flowers to become fertilized with pollen from other plants. (54) In 1909 an experiment was conducted for the purpose of determining the comparative effect of pollen obtained from the flowers of a different plant, from other flowers of the same plant, and of pollen from the same flower, in fertilizing flowers that were artificially tripped. In one experiment conducted in 1909, 59.09 percent of the flowers which had been fertilized with pollen from other plants, produced pods containing 3.71 seeds each; 32.16 percent of the flowers fertilized with pollen from the same flowers produced pods containing 1.56 seeds each. In another experiment carried out in the same year, 67.37 percent of the flowers fertilized with pollen from other plants produced pods containing 3.42 seeds each, and 42.85 percent of the flowers fertilized with pollen from the same flowers produced pods containing 1.61 seeds’ each. In 1910, 34.8 percent of the flowers. which were fertilized with pollen from other plants produced pods containing 1.87 mature seeds each; and 23.5 percent of the flowers which were fertilized with pollen of other flowers of the same plant produced pods containing 1.3 mature seeds each. The results of these three series of experiments show- ed that when alfalfa flowers were fertilized with pollen from other plants, a larger percentage of the flowers produced pods and a larger number of seeds per pod develop- ed than when the flowers were fertilized either with pollen ffom the same flower or from other flowers on the same (55) In an experiment conducted in 1910, there was not any marked difference in the results obtained by fertiliz- ing flowers with pollen from the same flowers or with polien from other flowers of the same plant. In another experiment carried out in 1910, the percentage of pods developing from flowers which were tripped without having the stigma or anthers come into contact with any object, was compared with the number of flowers producing pods under natural conditions. 52.24 percent of the flowers which were tripped naturally, provably by insects, rroduced pods. Of the flowers from which the standards were removed, and which were tripped so that the pistil and stamens did not come into contact with any object during the process, 14.15 percent produced pods. When the flower was tripped by cutting the pase of the keel petals in such a way that the pistil and stamens moved up to the standard, without emerging from the tip of the keel, 13.16 percent of the flowers tripped in this way produced pods. When the flowers were art- ificially tripped with a piece of alfalfa stem, using one piece of stem for all the flowers on one raceme, 41.73 percent of the flowers produced pods. The results of this experiment indicate that the mechanical action upon the surface of the stigma, which is brought about when it (56) strikes upon the object tripping the flower or against the standard has some influrnce in causing fertilization. This effect may be brought about either by bringing the pollen into contact with the surface of the stigma or by irritating the surface cells of the stigma, or to both of these causes. Alfalfa plants were observed at Pullman, Wash. and at Chinook, Mont., in 1909 and 1910, to determine whether pods and seed ever develop from flowers which have been automatically tripped. At Chinook, in 1909, two plants were found on which the flowers were becoming self-tripped in large numbers. On one of these plants 57.88 percent of the flowers and on the other plant 56.25 percent of the flowers observed were automatically tripped. At Pullman, Wash. in 1910, 7.7 percent of the flowers enclos- ed under netting were self tripped. At Chinook, Montana in 1910, nearly all of the flowers which opened during a period of two weeks or more early in August became tripped, when there were not enough insects in the fields to trip more than a small proportion of the flowers; in the few days succeeding this period, when climatic conditions were thanged, practically none of the flowers were tripped. The results of these observations indicate that usually a small percentage of the alfalfa flowers (57) become automatically tripped; and that under some condi- tions, the flowers on a portion or all of the alfalfa plants become self-tripped in large numbers. Observations meade at Chinook, Montana lead the writer to the conclusion that the large crops of alfalfa seed which are yroduced there in certain seasons, develop chiefly from flowers which have become automatically tripped. Observations made at Pullman, indicate that. in comparatively rare instances, alfalfa flowers which have not been tripped may produce pods. However, the numbers of pods and seed developing from untripped flowers evidently is very small. Alfalfa flowers on a number of racemes were tripped by clasping the racemes between the pages of a book. 47.1 percent of the flowers on these racemes were tripped in this way, and 52.7 percent of the flowers which were tripped produced pods. The results of this experiment indicate that the yield of alfalfa seed might be increased when insects are not present in the fields in sufficient numbers to trip the flowers, by any mechanical arrangement that might be devised to trip the flowers. An experiment was conducted at Pullman, Wash. in September 1909, to determine for how long a period alfalfa flowers remain susceptible to fertilization. It was (58) found that up to the time when the experiment ended, seven days after the flowers opened, when the tips of the petals were beginning to wilt, there was practically no diminution in the ability of the flowers tripped to become Tertitiged. The percentage of flowers visited which are tripped by different kinds of honey-gathering insects varies greatly. It was found that domestic honey bees usually trip only about one percent or less of the flowers. Different species of bumblebees vary in their efficiency in tripping flowers; at Havre, Mont., several different species which were observed tripped an average of 29.4 percent of the flowers that they visited for honey. Wild bees of the Megachile species tripped over 90 percent of the flowers which they were observed to visit. Observat- ions of the small thrips which were common on the alfalfa flowers in Washington and Montana, leads to the conculsion that these insects neither appreciably increase or de- crease their percentage of flowers producing pods and seed. An experiment conducted at Pullman, Wash. indicates that night-flying insects are not an important factor in influencing the development of the alfalfa seed cror. It was found that there is a considerable amount of variation in the percentage of flowers producing pods and in the number of seeds per pods, on different alfalfa (59) Appendix Historical Resume' (Verhandulungen Des Botanischen Vereins 1873, page 13 by Dr. I. Urban) The tripring mechanism of alfalfa flowers has been known to a limited number of biological investigators for a considerable time. The first reference that we have to the explosive mechanism in the flowers of Medicago sativa and of other leguminous plants is a statement by Urban, that the tripp- ing of these flowers was known in Linneus' time. Medicago Flowers Tripping Historical Resume' eee Vegetable Vol. 2 p 548 A.P. deCondolle 1832 In 1832 in his book on The Physiology of Plants, de Candolle gave a brief description of the tripping phenomna He said "Certain corolla contribute in an indirect way to fertilization, thus the pieces of the corolla of Indigo- fera and some lucernes are fixed one to the other by a kind a hock. When their development is accomplished these hooks are detached; the carina no longer fixed is warped elastically and stamps a blow upon the sheaf of stamens which determine the fall of the pollen." (60) Medicago Flowers Tripping Historical Resume' De Candolle: Pflanzen-Physiologie: p. 116 Aus dem Franz- Sischen ubersetzt von Johannes Roper, 1835, p. 116-111. Buch. Par. 6. In this translation into the German language Roper gives the same statement in regard to the tripping mechan- ism in Medicago that deCondolle originally stated in his original book on The Physiology of Plants written in French and published in 1832. Medicago Flowers Tripping Historical Resume' Journal of the Linnean Society of London 1867 p. 328 Rev. Geo. Henslow. In a paper read at a meeting of the Linnean Society in Nov. 1865 Henslow discussed the results of his obser- vations and experiments with flowers of alfalfa and other plants. Henslow was the first, according to any available record, who studied carefully the explosive mechanism of alfalfa flowers and attempted to explain its cause. He attributed the elasticity of the staminal column to the turgidity of the cells of the nine cohering filiments of the staminal tube. He found that the tenth or free stay- men appeared to have no such elastic property. When the pistil was removed from the tube it exhibited no elastic- -mcy to become curved. On the other hand (61) when the staminal tube had once become tripped he found that he could not straighten it to its original position without causing a transverse fracture. Henslow also observed hive bees sucking the nectar from the flowers of Medicago sativa. The flowers were frequently visited by the bees but im no instance did he find that they had power enough to set free the staminal tube. No opportunity of observing bumblebees presented itself. Medicago Flowers Tripping Resume! (Botanische Zeitung jahrg. 24 (1866) Von F. H. Hilde- brand P 74) In an article published in 1866 Hildebrand discusses the result of his studies of the flowers of Medicago and other plants. He gives a description of alfalfa flowers in their explosive arrangement. He quotes Triviranus' statement to the effect that flowers in Indigofera become tripped naturally. Hildebrand states that flowers in species of Medicago behave in the same way. According to a statement made in 1894 by Burkhill, Hildebrand asserted that the flowers of ™. sativa is fertile if allowed to remain unexploded. ( I.H. Burkhill Proceedings of the Cambridge Philosophical Society for 1894 p. 142) (62) Medicago Flowers Tripping Historical Resume! aware Zeitung jahrg. 25 (1867) Von F. H. Hildebrand P 283. In 1867 Hildebrand discusses the results of Delpino's study of the blossoms of different species of Medicago. Delpino described the alfalfa flower and its tripping mechanism but evidently made no important addition to what was already known on this subject. Medicago Flowers Tripping Historical Resume' Dr. I. Urban; Verhandlung des botanischen Veresns der Proving Prangenvure: Leva, bp. 1sit. Urban states that the peculiarity which the Medi- cago species show in their pollination, namely that the stamen is propelled out of the keel toward the standard when the keel is pressed on, or a needle inserted, was known in the time of Linneaeus, but its significance was not understood. De Candolle believes that this movement of the individual parts of the flower is for the purpose of bringing about direct self fertilization. "When the corolla barely reaches atove the calyx teeth a decided tension develops in the side of the stamen tube that is turned towards the keel." "Before pollination the standard lies close to the wing, so that it leaves only a narrow channel open which leads somewhat obliquely to the nectar chambers between the points on the wing just mentioned. These nectar chambers lie at the base pf the stamen tube on the inner (63) side. They are roll shaped projections, and can be reached from the base of the tube through two clefts, which are found on either side of the free stamens." "When the bloom reaches complete development the tension in the stamen tube ceases so that the stamens remain closed in the keel only in a few cases have I found fertilized ovules in such cases." According to my observations, which have been made frequently and at all times of the day, only bees bring about pollination, though butterflies enjoy the flowers just as well; but they can not push aside the points on the wings, and it is not necessary, since their more flex- ible proboscis can reach around them to the nectarium." "All these arrangements for pollination are not limited to the genus Medicago; besides the small flowered Medicagos, there are numerous Trigonella species, for example Fischerianna, Pinnatifida, Polycerata, etc., while others, for example Corniculata, Calliceras, etc., and Pocockia cretica have a pollination apparatus like that described by Hildebrand for Indigofera." Medicago Flowers Tripping Historical Resume! (The Fertilization of Flowers by Hermann Muller pars. 93 & 94 1873.) Muller gives a detailed description of the alfalfa flower and its explosive mechanism. He states that in Medicago sativa the elastic tension resides almost ex- clusively in the upper stamens. In Sarothamnus scoparius which also has explosive flowers, it is the style only which acts as a spring; in Genista finctoria the columm tends to spring upwards and the alae and carina together tend to spring downwards. Muller says : "self fertilization is undoubtedly efficient for Hildebrand has shown that flowers which wither unexploded when insects are excluded produce seed (64) by self fertilization." Muller states that : "although hive bees are by far the most numerous visitors of this flower, I have never seen them effect explosion; but I have often observed them. Though I have frequently watched this plant I have never succeeded in seeing explosion actually performed though the number of exploded flowers that we meet with shows that it mst take place very frequently. Butterflies visit the flowers in great numbers and it is doubtless by them that explosion and cross fertilization are effected; but they are generally too wild to permit close observation of their movements," Medicago Flowers Tripping Historical Resume' (Proceedings of the Cambridge Philosophical Society 1664 ps 142, 18, Burkhiil) At the meeting of the society named above on Feb. 12, 1894, Burkhill read a report of his quite extensive study. The subject of his paper was ; On the Fertilization of Some Species of Medicago in England. He first gave a historical resume' of the work that had already been done by different investigators and in- cluded a bibliographical reference to the published results of their work. Burkhill gives a detailed description of the alfalfa flower and its explosive mechanism. He states; "the ex- plosive action of the flower depends upon the uppermost (65) stamens of the fused nine : these by having the cells of their filiaments intensely turgid tend to make the whole staminal tube assume a curved form whereby the stigma and anthers are forced against the vexillum. This explosive force is resisted by the paired combining processes of the alae and carina and according to my observation not by the basal process of the alae." Burkhill covered a considerable number of alfalfa flowers with nets to prevent insect visits: no seeds are set he states in the unexploded flowers. Burkhill states that the stigma does not become receptive until rubbed or until its cells are injured in some manner. He says: "My proof is I think conclusive: Firstly, the stigma appears not to be moist, but when rubbed on glass leaves a sticky mark. Secondly, I have caused flowers to set seed though unexploded, (1) by pinching the stigma through the kell, (2) by perforating the keel with a needle and scratching the stigma, and (3) by cutting off the tip of the keel and rubbing the stigma with a stiff paint brush." Burkhill says that the hive bee does not explode the flower, but inserts its probosis obliquely over the basal processes and not between them. It is necessary to pro- duce explosion for an insect to insert its probosis be- tween these basal processes. (66) Burkhill found that it required an average weight of 1.68 grammes to explode a flower of Medicago sativa (Maximum and minumum 2.37 and .93) The worker of Apis he found to weigh about .096 and Bombus hortorum about .199 grammes. The mere weight of these two insects is there- fore according to Burkhill, insufficient to explode the flower. Burkhill says the flower is not always in the same degree of explosiveness; the hotter the weather the more explosive is the flower. He says shaking by the wind cannot explode the flowers Pieces of paper with a surface of 18 1/2 and 22 square inches were tied to stalks of this plant in order to give more power to the wind but no effect was observable from the shaking it produced. Medicago Flowers Tripping Historical Resume! (Contribution from the Entomological Laboratory; No. 65; Univ. of Kansas S.J. Hunter, 1898) Hunter published a bulletin in which he discussed alfalfa and honey bees. He describes the tripping mechan- ism of the flowers. He evidently assumed that the common domestic honey bee is capable of tripping the flowers but does not record any observations of his own. Hunter gathered a large number of representative ripened pods from an alfalfa field less than one half mile (67) away from a large apiary, and a like number from another field of much the same soil. and, vractically, under like conditions as the first field except that the second field was situated twenty-five miles away from a colony of bees. No bees were observed in the latter field. The pods from each field were carefully opened and the number of seeds in each counted. 87 pods taken half a mile from the bee hives contained 482 seeds or 5.58 per pod. 80 pods taken 25 miles from these produced 268 seeds or 3.35 per pod. Medicago Flowers Tripping Historical Resume' (Kirchner, Q. On the effect of self fertilization in Papilionaceae. Naturw. zeitschrift fur land- und forst- wirthschaf. January 1905. heft. 1-3, p. 45.) Kirchner states that the reports on the success of self fertilization of the violet blossoms of Medicago sativa which are provided with an explosive arrangement are contradictory. "The experiments made by us", he states, "in 1908, at Hohenheim, showed plainly that the plants are self stirile.. Fifty-four exposed clusters of blossoms on two plants with 432 blossoms produced Aug. 23, 208 pods which, though they were not perfectly ripe, showed plainly that they contained 636 well developed seeds. On the other hand 21 covered closters of blossoms on the same plant with 166 blossoms produced only 2 pods with 3 seeds." "Similar results were also obtained with the flowers (68) of Medicago falcata. Of one hundred and four exposed clusters of blossoms with 940 blossoms, 152 pods were produced with 177 seeds, whereas 206 covered clusters with 1882 blossoms produced no seeds at all. As the pistils are closely surrounded by the stamen,spontaneous self fertilization must take place, but without effect." Kirchner carried out experiments with the flowers of leguminous plants. during the years of 1900 to 1904. He classified them as either self fertile or self sterile. He classifies hoth Medicago sativa and M. falcata among the self sterile species. Medicago Flowers Tripping Historical Resume' After Hunters report on the relation of insects to alfalfa which was published in 1898 the next discussion of this subject, which is recorded in this country, is by Mr. J. M. Westgate, of the United States Department of Agriculture in a paper presented at the American Breeders' Association at Lincoln, Nebraska, Jan. 17 to 19, 1906, He presented a brief review of the work of Henslow, Urban, Kirchner and Burkhill. He pointed out the general dis- agreements between the results that have been obtained by different investigators. He suggested that a great deal of work needs to be done on the subject of the fertilizat- ion of alfalfa with the object of determining to what (69) extent the stigma is sterile to the pollen of its own flower, of the same cluster and of other flower clusters of the same plant. Medicago Flowers Tripping Historical Resume' (Bulletin No. 151, Kansas Experiment Station H. F. Roberts and G. F. Freeman, Dec. 1907.) Roberts and Freeman give the results of some experi- ments to determine the effect of close pollination. Plants were inclosed in wire netting tents. A portion of the flowers were pollinated by hand and a portion were left without pollination. On the plant on which this experi- ment was carried out 59 seeds were produced from flowers which were not hand pollinated and 118 seeds were pro- duced from the hand pollinated flowers. In another ex- periment with five different plants which were producing seed abundantly, wire netting cages were placed across. the middle of the plant thereby enclosing the half on which the flowers were to be hand pollinated within the cage and leaving the remaining half standing out side the cage freely accessible to insect visitors. Far more pods and even more pods coytaining seeds were produced on the hand pollinated series but a greater number of seeds per pod were produced in the insect pollinated series and as a matter of fact more seeds per plant on this account. Taking the green weight of the (70) plant as a basis of comparison, the insect pollinated series gave 29.7 seeds per each ten gramms of plant weight while the hand pollinated section produced slightly less, - 25.9 seeds per 10 gramms of weight of the plant. Seven different alfalfa plants were inclosed with- in wire netting cages; the number of seeds which developed while the plants were inclosed under netting were later counted. The results obtained are given in the following table. The flowers were artificially tripped. Plant No. Apparent seeding Total No. of seeds ability number per 10 gramms seeds green weight of plant. 69 Strong ey Lote 91 Weak d 0.14 89 Strong 214 20 g5% 90 Strong 257 21.24 91 Weak 23 0.735 92 Strong 59 Seok 70 Strong a0 2640 Roberts and Freeman describe a method of exploding flowers in large numbers by rolling the head carefully put firmly between the thumb and the first and second fingers. Medicago Flowers Tripping Historical Resume' (Cire. 24, B.Pels UsSs Depti of Agri, Cod, Brand & J.M. Westgate, March 5, 1909.) A very brief discussion of the relation of insects to the setting of alfalfa seeds is given in this bulletin. (Wi) The authors state that insect visits are essential to the proper pollination of the alfalfa flowers. They state that bumble bees are the most efficient of all insects in setting off the explosive mechanism and hence in bringing about pollination. Honey bees though not nearly so effect- ive as bumble bees should not be under rated in this conn- ection. Wild bees (Andrena spp. and Megachile spp.) and various butterflies are also valuable agents in pollinat- ing alfalfa flowers. A brief discussion of results obtained at Arlington Experimental Farm and at Chico, California by the U. S. Dept. of Agri. is given. The results show that when alfaifa flowers were artificially pollinated on the Arl- ington Farm Twenty-five and one half per cent more pods were produced than when the flowers were not artificially ‘pollinated. At Chico, Calif. a row of alfalfa plants artificially pollinated produced 129% more pods than the plants in the row not artificially pollinated. Medicago Flowers Tripping Historical Resume' (Report of American Breeders' Association, 1909) A report of questions submitted to men engaged in alfalfa breeding in different parts of the United States, is given by the committee in charge of breeding forage crops of which Prof. C. V. Piper was chairman. The opin- ion of the different alfalfa breeders, im regard to the (72) importance of insects and of other agencies which may trip the alfalfa flowers is given here. The opinion of two or three of these men in regard to whether alfalfa flowers may become self tripred or whether pods may develop without the flowers being tripped at all is also given. (73)