^ Q45 U. S. DEPARTMENT OF AGRICULTURE. BUREAU OF ENTOMOLOGY— BULLETIN No. 59. L. O. HOWARD, Entomologist and Chief of Bureau. PROLIFERATION AS A FACTOR IN THE NATURAL CONTROL OF THE MEXICAN COTTON BOLL WEEVIL. BY W. E. HINDS, Ph. D., In Charge of Cotto)i Boll Weevil Lahoratoi-y. Issued August 27, 1906. WASHINGTON: C40VERNMENT PRINTING OFFICE. 1906. Ikmognifib V, Bui 59, Bureau of Entomology, U, S Dept. of Agricultur Plate I. Proliferation from Weevil Feeding-punctures. Fit; 1 — KiKlit halt of square tilled with granular M,pi>eann,t; pniliforatioii. enlar,i,a'(l four diame- ters ¥)K 2— Interior of square, proliferation from feeding puncture dru'd and brown, enlarged two diameters. Fig. 3.— Seetii>n throu,gh feeding puncture trom whicli proliteration spread to tip of .Miuare. enlarged four liiaineters. Fig. I — Pn>liferation starting from teedmg punctures in tjoUs, enlarged two diameters. (Original ) U. S. DHPARTMFNT OF AGRICUITURE. BUREAU OF ENTOMOLOGY- BULLETIN No. 59. L. O. HOWARD, Entomologist and Chief of Bureau. PIIOLIFKKATION AS A FACTO Pt IN TflE NATliRAL CONTROL MEXICAN COTTON ]]OLL WEEA^IL. W. E. HINDS, Ph. D., 1)1 ('lt(irt/t' of Ciil/iiii J!(il/ ]]'ii'ri/ Lnhiinitiin/. IsfiUEU August 27, 190(i. WASH1N(tT()N: a o v e k n m k n t p 11 1 n t inc office, il»(M). ^^onograph LIHTRR OF TRANSMITTAL U. S. Department of Agriculture, Bureau of Entomology, WasJiington, D. C, April 10, 1006. Sir: I have the lionor to transmit herewith a manuscript prepared by Dr. W. E. Hinds, special field agent of this Bureau, engaged in work on the boll weevil. This manuscript is a study of the pro- liferation in the squares and bolls of cotton by means of which a certain percentage of weevil larvae are killed. It does not deal at length with the botanical aspects of the (luestion, Init is rather a practical statement of the effect of this formation of loose tissue cells upon the boll weevil, based upon a large number of observations made by agents of the Bureau of Entomology. The botanical side of this phenomenon has been fully considered by Mr. O. F. Cook, of the Bureau of Plant Industry, and this paper is therefore supple- mental to papers published by Mr. Cook on this subject. The preface is written by Mr. Hunter, and the conclusions in the paper have been revised by him. In addition to the general interest in the sub- ject, the information given will be undoid:)tedly of distinct advantage to those engaged in cotton-breeding work, and I therefore recommend that it be issued as Bulletin No. 59 of this Bureau. Respectfidly, C. L. Marl ATT, Acting Chief of Bureau. Hon. James Wilson, Secretary of Agriculture. 2 U. 01 y* PREFACE Aside from the hal)it of determinate growth, perhaps the most im]^ortant tendency of the cotton plant that has the effect of avoiding damage l)y the l^oll weevil is that of proliferation in the stpiares and bolls, which was first observed by Dr. W. E. Hinds in 1902 at Vic- toria, Tex. The present paper places on record a large number of observations and exj>eriments relating to this phenomenon, which have been carried on for several years by the boll-w(^evil investigation in Texas. The studies upon which the paper is based were planned primarily to determine the ways by which jn-oliferation actually affects the weevil. Other features of })roliferati()n have l)een dealt with fidly by Mr. (). F. (\iok, whose ])ublications are referred to more s])ecifically in the text. In addition to the general interest of the information given, nuu-h of it will undoubtedly be of special advan- tage to those who are engaged in cotton-breeding work. As will be seen in the following pages, it has been ascertained that the rate of mortality among weevils in s([uares of American u])land varieties of cotton is higher by about 13.5 per cent as a result of {)r()- liferation. This in itself is of no little significance, but it is to be noted that the greatest importance of proliferation is in connection with some of the foreign varieties of cotton, which seem to have this property develo])ed to a much greater extent than the American upland varieties. Consecpienth^, the discoveries of j\lr. O. F. Cook, relating to the Kekchi cotton of Guatemala and the possibility of utilizing this cotton in the Ignited States, are of great interest. An important didiculty which will be encountered in the work of breeding cottons which proliferate to a great extent, will he the capability for adaptation on the ])art of the boll weevil. That this insect has consideral^le ca]iability for ada])tation is shown in the great variation in the size of the adults, the result of conditions of food supply in the immature stages to which it has adapted itself, as well as in man}" other ways. As a matter of fact, the capacity of the weevil for ada})tation is probal)ly fully as great as is the natural adaptive capacity of the cotton plant. Nevertheless, the interference of man may throw the advantage greatly in favor of the plant. The work upon which this publication is based was ]>erformed under the general direction of the writer by Dr. W. E. Hinds. He was assisted hi various ways by practically all the agents of the boll- weevil investigations, but more ])articularly by Messrs. A. C. Morgan, W. W. Yothers, W. Dwight Pierce, A. W. Morrill, and F. C. Pratt. W. D. Hunter, //(- (liargc of Cotton Boll Weevil Investigations. ' 3 ro XT 1: NTS PilgO. liii rodiirl \nn. 7 Eiiiliest ohsoi'vaf ions. 7 Sc()[)t> oi" pn^scMil (lisciissii)ii. .... 7 Dcfinitioiis 8 Metliod of study !) Proliferation from fccdin"^ puncliirc-: in si|uai(' ^ K? Pfolifcralion from fccdin;:; |>unclui('s in hoils ... 11 Tiiflncncc of di(r<'i(Mil localit ics and s(>as<)ns 12 Ohsci'valiciis on s(|Uar('s 12 Ohscrvalions on Ix.lis II EflVcts of climatic conditions 15 l<]llcct on proliferation of fei'tilizin;'; 1 lie roil . 17 I'roliferation foilovvin Increased mortalit \' in s(|uarcs and ixills dxiv i o prolifcral ion. . 2() vSummary of results of ol)scr\ations 27 Formation of proliferation 27 Increased mortality of \vec\il sta^jes due to prolifcral ion 2S Stinuilation to proliferation \>v causes ol licr i liau wccxil at taci< 29 Proliferation slinudaled hv otiiei- insects 29 Prolifei'alion stinuilated by attacks of fmiiji 29 Artificial stinuilation to prolifcral ion. * 'AO Metliod of treatment '.iO Results with s(|uares 'M Results with holls :)2 Comparison of icsulls from simi)le needle punctures wilh ell'ects of chemical injections 31 (V)mparis()n of results from s(>alcd and unsealed j)vmctures 34 Conclusions 35 Maiuier in which proliferation causes death of wee\il staration 38 Practical application of conclusions from this st udy 39 Index 43 5 ILLUSTRATIONS PLATES. Page. Plate [. Fig. 1 . — Granular appearance of proliferation Frontispiece Fig. 2. — Proliferation among anthers following feeding puncture. . .Frontispiece Fig. 3. — Proliferation following weevil feeding puncture in scjuare. .Frontispiece Fig. 4. — Proliferation following weevil feeding puncture in boll Frontispiece ir. Fig. 5. — Bollworni full-grown, feeding on boll 2S Fig. (>. — Punctures of small i)ollworni in sf|uare, proliferation protruding. 28 Fig. 7. — Scpiarc-liiirer at work 28 III. Fig. S. — Boll fed on bv bugs, showing no external injury 28 Fig. '.I. -Insidr of carix'l sliown in lig. S, pi-oliferalion starting 28 Fig. 10. —Section of boll fed u]ion by Pcutatonia Jujatd, proliferation in seeds 28 Fig. 1 1 . — Sect ion of proliferated seed 28 Fig. 12. — Poll showing anthracno.se disease which may cause proliferation . 28 IV. Fig. V.i. — Proliferation in seeds from artiiicial injection of water 30 Fig. 14. —Proliferation in cai-jiel and seed from artiiicial injection of acetic acid 30 Fig. 1'). — Prolifcralion eii\ ('lo|)iiig lar\;i in l)oll 30 Fig. ](). — P>oli buist o])en l)y inteinal i)ressure of proliferation 30 V. Fig. 17. — Numerous proliferations starting on inner sides of carpels 36 Fig. IS. — Proliferation pushing into pupal cell, pressing upon pupa 36 Fig. U). — Adult weevil deformed by pressme of proliferation 36 Fig. 20. — Ma.ss of proliferous cells upon inside of carpel 36 VI. Fig. 21. — Ijocks completely filled with proliferation; a, original egg punc- ture ; h, larva surrounded and crushed to death 36 6 PROLIFERATION AS A FACTOR IN THE NATURAL CONTROL OF THE MEXICAN COTTON BOLL WEEVIL. INTRODUCTION. EARLIEST OBSERVATIONS. Soon after the beginning of the hiboratory work upon the cotton boll weevil at Victoria, Tex., in 1902, it was noticed that the attack of th(> weevil was frequently followed ])y a very decided change in the structure of the tissues near the point of attack in both l)uds and bolls. The significance of this change was not at that time fidly appreciated, and the observations made upon the weevil did not include records as to the occurrence and effect of this phenomenon. For this reason the early observations made before the autunm of 1903 have furnished comparatively little material which could be used in making tabular stAtements, such as have been made from more recent studies of the effect of proliferation upon the development of the boll weevil. When and by whom proliferation was first observed in cotton is not known to tlie writer, ])ut no publication relating to this phenomenon prior to that made in Bulletin No. 45 of the Bureau of Entomology, pages 96 and 97, has been found. The earliest notes upon the occur- rence of proliferation and its effect upon the weevil were made by the writer in September, 1902. Since that time there has been gradually accumulating in the notes of the agents of the Bureau of Entomology who have been studying the boll weevil, a large amount of data bearing ujion this subject. In the plans made for the .work of 1904, at the beginning of the season, definite provision was made for ol)servations upon this phe- nomenon in a munber of varieties of cotton and for testing the infiuenco of fertilizers in stimulating a greater manifestation of prolif- eration in the plants treated. Since that time the observations upon proliferation and its effect upon weevil development and injury liave been carried on continuously. SCOPE OF PRESENT DISCUSSION. The present paper does not pretend to l)e a study of proliferation in the botanical aspects of the question, Imt rather a practical statement of the large number of observations made by agents of the Bureau of Entomology primarily regarding the effect of this formation of loose 7 I'ROLIFERATION IN CONTROL OF KOLL WEEVIL. tissue cells upon the l)oll weevil. It is consequently of an entomolog- ical and not a botanical character. The botanical significance of the phenomenon has been very fully considered by Mr. O. F. Cook, of the Bureau of Plant Industry, to whose publications among those listed below" the r(>a(ler is referred for a discussion of that part of the subject. DEFINITIONS. In order that the statements here made may be readily understood by one who is not familiar with terms used in botany <»r entomology a few general definitions may be in order. In Bulletin No. 45 of the Bureau of Entomology, concerning the cotton boll weevil, the term "gehitinization" was used instead of ])roliferation, as it was believed that its significance woidd be b(>tter understood by the average reader, though it was realized that, strictly speaking, the term used expressed an incorrect idea, concerning the nature of the change^ to which it referred. The term "proliferation" is in general use both in botany and zoology, to denote a growth by the multiplication of (dementary parts. In the ])resent case we may define proliferation as being the development of numerous elementary cells from ]^arts of the bud or boll which are themselves normally th<^ ultimate product of combinations of iimch more highly sjiecialized cells. The resulting product is thus composed of comparatively large, thin-walled cells, which are placed so loosely together that the residting formation is of a soft texture, and has a granular appearance (PI 1, fig. 1) which may be ]^lainly seen with the unaided eye. The soft, ])uli)y nature of this growth led to the choice of the term" gelatinization" as being most appropriate to signify in a general way its appearance and texture. It appears that this forma- tion may originate from various causes in almost any part of the bud or boll (Pi. I, figs. 2-4). Whatever may have been the inciting cause, the character of the formation a])pears very much the same in any case. Ihidoubtedly, however, certain tissues proliferate more readily than do others. In squares, the outer layer of the column upon which the stamens are borne appears to be especially susceptible to stimuli which produce this reaction^ In bolls, the cells immediately adjoining the thin, hard layer lining the hulls or carpels are most frequently the o Bibliography of l-*iolil'eration: 1904. Hiintcf, W. D., and Hinds, W E.— Tlie Mexican Cotton Boll Weevil. Bui. No. 45, Biueau of Entoni()lop;y, I'. S. Dept. Agnc, pp. 9(5-97. 1904. Cook, () P.— Evolution of Weevil Resistance in Cotton. Science, Vol. XX, pp. 666-670. 1905. Hunter, W. I)., and Hinds, W. E.-~15ul. No. 51, Bureau of Entomology, U. .S. Dept. Agric, pp. 133-134. 1905. Cook, O. F.— Cotton Cidture in (Juateniala. YearlM)ok U. S. Dept. Agric, f. 1904, pp. 47r>-4cS8. 1906. Cook, (). F.— Weevil Resisting Adaptations of the Cotton Plant. Bui. No. 88, Bureau ot Phinl industry, V. S. Dept. Agric. METHOD OF STUDY. 9 point at which proHforatioii begins. In most cases the prohfcration appears to begin very near to the point of injury, but from that point it may spread tlu'oiigh an entire lock, or to all the inner parts of an injured bud. METHOD OF STUDY. As the significance of these observations came to be more fully appreciated it was believed that they contained at least a suggestion as to some very promising lines of work in the problem of controlling the weevil. Accordingly, it has been necessary to study carefidly the nature of the phenomenon, conditions of climate, soil, fertilizer, and variety of cotton which affected the occurrence of proliferation. Observations have, therefore, been made ujjon (piite a large number of varieties, and in locations ranging from Victoria to Dallas, Tex., upon various t^^pes of soil, and in connection with various exi)eriments with fertilizers and different conditions of cultivation. From a comparison of the results thus ()l)tained it was hoped that some factors might be found which coidd be used ])ractically in increasing proMferation, and thus rendering it more effective as a factor in controlling the weevil. In many cases the results of the work have been quite different from those anticipated, but enough has been learned to justify the assertion that at present proliferation is a more important factor in retarding the nmltiplication of the weevil than are the parasites which have thus far been found. Large munbers of squares and bolls have been carefidly examined in obtaining these records. In the examination of bolls, the lock has been made the unit rather than the l)oll. As a general rule, a larva confines its injur}^ almost, if not entirely, to the lock within which the egg was originally placed. Quite frequently two or more larvc^ occur within a lock, but even in such cases the injury does not often extend through the septum or partition which separates the locks. In making a comparison of varieties considerable care is recpiired in subdividing the classes of observations in order to render the influ- ential conditions sufiiciently imiform to make the observations fairly comparable and wherever possible to reduce the fundamental causes or stinndi producing variations in the proliferation to one essential factor. While the phenomenon in bolls is of a similar nature to that in scpiares, conditions in these -two cases are so ditferent that the results are not strictly comparable, and therefore separate tables have been made for squares and bolls. The effects of feeding and egg punctures also call for separate classification. Tliis treatment of the subject necessarily multiplies the number of tables, but we hope that it will render the results more easily intelligible. The personal ecpiation of the observer has been ecpialized by combining the records made by a nund)er of investigators. 31022— No. 59—06 ^2 10 prolifp:ration iisr control of boll wep]vil. PROLIFERATION FROM FEEDING PUNCTURES IN SQUARES. As the sfjuare precedes the boll in the natural process of develop- ment and the feedin^i; puncture precedes oviposition in the attack of the weevil, wc shall consider first the observations regardino; pro- liferation resulting from feeding punctures made in squares. These observations include nearly 25 varieties of cotton. They are grouped by years and localities in order to bring as closely together as is possi- ble those records which may be considered as most strictly compar- able. The totals and averages for so many seasons and localities should constitute a very fair average statement of the true condition. Table L — Proliferation resulting from feeding punctures in squares — comparison of varieties. Locality. • Variety. Total number of punc- tured squares exam- ined. Squares with proliferation. Squares with no proliferation. Date. Num- ber. Per cent of total. Num- ber. Per cent of total. 1902. Sept. 17 1903. July 8 to Several varieties King 16 470 83 102 101 79 74 82 78 53 10 21 70 33 47 45 69 25 39 36 IS 30 27 34 31 34 163 4 286 49 38 67 59 35 62 42 8 3 15 14 23 28 32 11 22 18 7 15 15 15 4 16 71 25.0 60.1 59.0 37.3 66.3 74.3 47.3 75.6 53.8 15.1 30.0 29. G 21.4 42.4 48.9 62.2 46.4 44.0 56.4 50.0 38.9 50.0 55.6 44.1 12.9 47.0 43.6 12 75.0 I do 184 1 39.9 Oct. 28 Aug. 10 to I do Parker Mascot 34 41.0 Oct. 19 Oct 25 1 do 64 34 20 39 20 36 45 62.7 Do do Dickson 33.7 Oct '^4 do Mit Aflfl .. 25.7 Oct '^2 do Ashmouni 52.7 Do do Jannovitch 24.4 Aug. 8 to do Native ... Egyptian 46.2 Oct. 28 1905. Nov. 9 1 Dallas, Tex do 84.9 Do Pachon 7 1 70.0 Do do Kekchi 15 1 71.4 Sept. 24 Do San Antonio, Tc.x Calvert Tex Shine . 55 1 78.6 King 19 j 57.6 Do do Shine 24 1 51.1 Do do Rowden 17 37 14 17 18 11 15 12 19 27 18 92 .37.8 Do ..do 53.6 Do do Triumph 56.0 Do do Tools 43.6 Do do Hawkins 50.0 Do . ...do Russell 61.1 Do do .Mien .50.0 Do do Bohemian 44.4 Do . ...do Truitt 55.9 Do do . .do Hetty 87.1 Do 53.0 Do... do Territory 56.4 ..als and averages Tol 1,870 965 "51.6 905 a 48.4 n Weighted average. The general average for all the various seasons and localities shows that in squares approximately one-half of all feeding punc- tures stimulate proliferation. The highest percentage shown is 75.6 per cent in "Jannovitch" (an Egyptian variety), at Victoria, Tex., on October 22, 1903, while the lowest percentage found was 12.9 PROLIFERATION FROM FEEDING rUNCTURErt IN BOLLS. 11 per cent for "Hetty," at ('alvert, Tex., Sopteiiil)er 24, 11)05. These figures show a wide rano;e. Five series of observations show prohf- eration in less than 30 per cent of the squares fed upon; three series show between 30 and 40 per cent; eight series between 40 and 50 per cent; six show between 50 and 60 per cent; three !)etween 60 and 70 per cent. It appears, therefore, that the range, while wide, is well balanced, the large majority of o})servations showing between 40 and 60 per cent. PROLIFERATION FROM FEEDING PUNCTURES IN BOLLS. Turning now to an examination of proliferation following feeding punctures made in bolls, records are presented of the observations made during 1905 only. These observations include IS varieties and 3 localities. Table IL — ProliJ) rot'ion nsii/iiiKifroiiiftcditit/ /iinirfii)(s in hal/.s. Date of I'Xiiini- nation. 1905. Sept. 25 Do.. Do.. Do.. Do . . Do . . Do . . Do.. Do.. Do.. Do.. Do.. Do... Do.. Sept. 27 . Nov. U . Do.. Do.. \'ariety. King Shine llowdon. . . Nicholson. Triumph . . Tools Hawkins. . Russell... Allen Bohemian. Tniitt Hetty Native Territory., Shine Mit Aflft. Pachon. . Korean . . Locality. Calvert, Te.x. . do do .do. .do. .do. .do. .do. do do do do ....do do San Antonio Tex. Dallas. Tex. . . . ....do ....do Totals and averages. Total bolls exam- ined. 80 01 03 83 57 109 110 98 83 90 94 97 94 055 150 Total locks in lot. 340 398 274 374 247 402 402 419 371 399 419 419 407 2, 830 050 244 7 3 Locks with feeding puncture; only. With prolifer- Without pro- ation. liferation. Num- ber. 123 159 94 195 14'i 239 302 254 180 173 187 248 122 1,251 104 04 5 3,908 Per cent of total. 30.2 40.0 34.3 52 . 2 .59. 1 51.7 (15. 4 07.8 48. 5 43.4 44.0 59.2 30. 44.0 25. 20.2 71.4 Num- lier. 284 05 Per cent of total. 9.4 5. 5 20.5 10.4 12.1 20.8 0.9 10.0 2.0 5,0 4.8 5.5 10.0 10.0 5.7 0.0 0.0 Per cent of punc- tured locks in which prolifer- ation fol- lowed. 79.4 87.9 02.7 83.3 08.0 81.0 70.9 89.7 83.0 95.6 90.0 92.5 84.7 81.5 71.0 82.0 100. 100.0 :81.3 a Weighted average. The bolls examined all showed distinct external signs of weevil injury. Among them, however, fully one-fourth of the total number of locks were found to have no noticeable internal injury, and ])rob- ably a majority of these locks would have matured had the bolls been allowed to remain upon the plants. As the ]>olls examined were selected especially for weevil injury, it appears that their con- dition would probably be worse than the average in fields where the weevil has done its worst damage. The figures are of interest, therefore, as indicating that even under the most severe conditions of weevil injury sufficient seed would still be produced to replant the 12 PROLIFKRATION IN CONTROL OF BOLL WEEVIL. field. While practically one-half of the squares attacked showed proliferation, a far greater proportion of the locks attacked by the weevil showed a similar formation. From these records it appears that 55 per cent of the nearl}^ 9,000 locks examined received feeding punctures only. Among the locks thus injured, an average of slightly over 81 per cent showed distinct evidence of proliferation. A comparison with Table I indicates that in bolls i^roliferation occurs from feeding punctures in a higher per- centage of cases than it does in squares. The records upon Pachon and Korean cottons were included in the table because of the special interest attached to these varieties, but the data regarding them are too meager to be reliable in drawing definite conclusions regarding proliferation in them, and they s.iiould be excepted in making a com- parison of varieties. It is to be regretted that the two varieties mentioned produced so little fruit at Dallas, Tex., that more extensive data regarding them could not be obtained, and the fruiting occurred so late in the season tluit no bolls could matiu'e. The range in the percentage of cases in which proliferation results from feeding punc- tures in bolls is not so great as it appears to have been in squares. This fact may ])ossil)ly be due to more iniiform climatic and cultural conditions, as nearly all the records for bolls were made from mate- rial collected in one locality at the same time. These records ajipear to the writer to show a remarkable uniformity, and to indicate that among the 15 varieties mentioned in the table which are most clearly comparable there is little difference in the natural tendency to ])roliferate in response to feeding injuries made by the weevil in bt)lls. INFLUENCE OF DIFFERENT LOCALITIES AND SEASONS. OBSERVATIONS ON SQUARES. This series of observations was made to determine, if possible, what influence difi^erent localities and seasons might have upon prolifera- tion in the same variety of cotton. While similar data have been secured for a number of varieties, the exhibit following is restricted to the two varieties on which the largest immber of observations was made, as the conclusions which may be drawn therefrom are consecjuently most reliable. In the case of King cotton, different seasons as well as localities are represented, while with Shine, difl'erent localities are represented at approximately the same time. In com- piling this table, both feeding and egg punctures have been included. It has seemed desirable also to present the figures showing the effect which the proliferation has had upon the weevil stages found. OBSERVATIONS ON SQUARES. 13 Table III. — Proliferation in King and. Shine squares — different seasons and localities. o3 Squares wi th proliferation. Squares without prolif- eration. ^9 ■3 A fe w g M OT C oj Variety and "5 > "3 > "3 ai ■a 1. 03 "Si 03 •^ -o o >4- Q> .5° locality. o o o 4J ^ ^ X3 t-i g R > > s^ % F Ol o 1 '? S| 3 3 ^ (2; 3 •A Ph 0) O) S 1- KING. 1904. Victoria, Tex. July to Oc- tober. 1905. 822 437 53.1 87 14 13.8 385 40.9 105 0.0 100.0 13.8 Calvert, Tex.. August and September. 218 124 50.2 04 24 27.3 94 43.8 01 2 3.2 92.3 24.1 Totals and averages . . 1,040 501 a54.0 151 38 a20.0 479 a40.0 220 2 n.9 "95.0 al9.1 SHINE. ; [ 1905. Calvert, Tex.. August 229 122 53.3 59 2S 32.2 107 40.7 ,52 5 8.H 84.8 23.4 San Antonio, Tex. September.. 443 212 47.9 152 51 25.1 231 52.1 178 18 9.7 73.9 15.4 Totals and averages .. 672 334 049.7 211 79 a27.2 338 "50.3 230 23 19.1 (i77.5 al8.1 General totals and av- verage 1,712 895 052.3 302 117 024.4 817 a47.7 450 25 a5.2 082.4 "19.2 a M^eighted average. Two rather striking contrasts arc showai l)y a study of the figures in this table. First, in the 1,040 King squares examined there were found 417 weevil stages, while m 672 Shine squares exammed there were found 543 stages. Stated in a way to make the contrast most evident, in King there was found an average of one weevil stage for each 2.5 squares; in Shine an average of one weevil stage for each 1.24 squares. That is, in Shine there were almost exactly twice as many weevil stages found, in pro]iortion to the number of squares examined, as in King. This is a factor, however, which would naturally vary widely with the degree of infestation found in the field and it is a well- established fact that weevils were much more numerous and injurious at San Antonio in 1905 than they were at Calvert, Tex. The second striking contrast is to be found in the percent>age of mortality. In King squares without proliferation only 0.9 per cent of the weevil stages found were dead, while in Shine squares without proliferation ten times as large a proportion, or 9.1 per cent, of the stages found were dead. Doubtless much of this dill'erence may have been due to seasonal rather than to varietal dilTerences, since it appears that in King squares at Calvert in 1905 the percentage of mortality was mucli greater than at Victoria in 1904. In other respects there is a most striking uniformity in the results shown. The percentage of squares showing proliferation varies only between 49.7 per cent for Shine and 54 per cent for King. The 14 PROLIFERATION IN CONTROL OF BOLL WEEVIL. av('rai!;(> increase in mortality apparently due to the proliferation varies only between 18.1 per cent for Shine and 19.1 per cent for Kin^. It appears that the "normal mortality," due to other causes than ])roliferation, varies much more widely in different localities and seasons than does the increase in mortality attributable to the presence of proliferation. OBSERVATIONS ON BOLLS. Before drawino; any general conclusion from Table III the similar records of examinations of bolls should be considered. The same varieties and localities are used as in Table III, the only change being the inclusion of the examination of King bolls made at Victoria, Tex. in 1903. Table IV. — Comparison of proJlferaflon in King and Shine bolls in different seasons and localities. Variety and locality. KING. Victoria, To.x.. Victoriii,, Tex. Do Cnlvert, Tex.. ■SHINE. Ciilvrrt, Tex Smm .Viitonio, Te.x. Totals and av- erages 1900. Oct. 14 1904. Sept. 5 Oct. 1 1905. Sept. 25 Sept. 25 Sept. 27 Total bolls ex- amined. G20 200 91 15(1 Total loclcs ex- amined. 865 843 398 {ioO 1,345 5,7G8 Locks with proliferation. 1.398 417 4(18 234 422 3,115 Per cent of total. 52.4 48.2 55.5 51.7 58.8 (14.3 Weevil stages jilive. 37 189 33G Weevil stages dead. Per cent of stages found dead. 13.3 22.2 30.2 26.7 Variety and locality. KING. Victoria, Tex . . . Calvert, Tex . .SHINE. Calvert, Tex San Antonio, Tex Totals and av- erages 1903. Oct. 14 1904. Sept. 5 Oct. 1 1905. Sept. 25 Sept. 25 Sept. 27 Locks without proliferation. Num- ber. Per cent of total 1,208 448 375 164 234 51.8 44.5 48.3 41.2 35.7 a 40.0 Weevil stages alive. Weevil stages dead. Per cent of stages found dead. 0.0 11.1 0.0 11.1 Percent of all dead stages in locks With with 1 prolif- prolif- eration, eration. Locks with feeding punctures only. 100.0 80.0 75.0 •J8.0 159 104 With- out prolif- eration. Percent with prolif- eration. 73.7 67.5 88.0 71.0 "77.4 " Weiglited average. EFFECTS OF CLIMATIC CONDITIONS. 15 In the case of bolls the conchisions indicated are c{uite simihir to those which have been stated for sciuares. The percentage of locks showing proliferation in consequence of weevil injury is remarka])ly uniform, varyint^; throuojh a rano;e of only 16 per cent in three years in three localities and with two varieties. Furthermore, the percenta*i;e is almost identical with that shown for squares. By far the hif^hest percentao:e of mortality amono; weevils in bolls was that found in Victoria in 1903, when an averalant growtli (hiiing this period, we beli(»ve they may safely l)e disregarded, assuming that the conditions inunediately prevailing would be most signilicant in their influence u))()n tlie growth of the ])lant, the development of the weevil stages in buds and bolls, and the formation of ])roliferati()iL The mean average tenq^erature at Victoria dnring S(^))tember, 1903, was 77.2° F., which was 3.7° below the normal. During the first thir- teen days of October the mean tenq>erature averaged 75. S°, which was about normal. Preci])itation during Septenfl)er was very , slight, amounting to only 0.54 inch, which was nearly 3 inches below the normal. During the first thirteen days of October the rainfall amounted to 1 .75 inches, which was 0.42 inch above the normal. 16 PROLIFERATION IN CONTROL OF BOLL WEEVIL. At Victoria in 1904, from July 1 to October 10, the mean tempera- ture averaged 80 . 6° F . , which was 1.17° below the normal . During the same period the total rainfall amounted to 8.50 inches, which was only 0.57 inch below the normal. In a general way this season might be described as slightly cooler than usual, with the humidity and rainfall practically normal . No records are available for Calvert, Tex., l)ut the reports from Hearne, which is only 8 miles fi-om Calvert, will serve to indicate the temperature and rainfall of the latter place with sufficient accuracy. During the months of July and August, 1905, the mean temperature averaged 82.85° F., which was 2.8° above the normal. No rain fell during ISeptember, and during August the rain amounted to only 0.63 of an inch. For these two months, therefore, the rainfall was 4.33 inches below the normal. The season may l)e characterized in general as exceptionally hot and dry. At wSan Antonio the mean tem]:)erature diu'ing these two months averaged 82.5° F., which was 1.6° above the normal. During this period the rainfall amounted to 2.31 inches, which was 3.35 inches below the normal. Here again the season was exce])tionally hot and dry. Considering these climatic conditions in relation to the figures given hi Table III, the following conclusions would seem to be indicated: (1) The percentage of squares which ])roliferate from attack by the weevil is not greatly afl'ected by varying conditions of temperature and moisture; (2) the increase in mortality due to jjroliferation is not greatly affected by the varying climatic conditions as showm for these localities; (3) the normal mortality of the weevil which may not be attributed to ])roliferation is decidedly greater during especially hot dry weather than it is imder cooler and more moist conditions. As for liolls, the range in formation of ]:)roliferation in locks fi'om 48.2 to 64.3 per cent is not unex])ectedly great. The most remarkable fact is that the maximum j^ercentage for locks and the minimum percent- age for squares occur at the same time, in the same locality, and with the same variety. It is ])lain, therefore, that climatic conditions can not be held resj^onsible for these contradictory results. The records concerning ]K'rcentages of mortality are also too inconsistent to point to any constant effect of the climatic conditions upon this particular point. The records for "normal mortality" also fail to show any con- sistent increase or decrease which may be attributed to exceptional conditions of heat or drought. The reason why the records for bolls fail to show as consistent conclusions as are indicated for squares may probably be found in the comparative difference in the length of the growing season for each and in the essential difference in the nature of the two sets of organs. Obviously the scpiare would be subject to cli- matic changes occurring within only a short ])eriod of time as com- EFFECT ON PRULIFERATION OF FERTILIZING SOIL. 17 pared with the boh, which woukl therefore exliil)it n more eoiii})osit(^. resuh of any influential conditions aft'ectinii; it. It is proba])ly true that the increased mortality in squares occur- ring din-ing hot dry wetither is more directly attrihiital)l(^ to the absolute maximiun temperature experienced than it is to the shghtly higher mean average temperature prevaiHng. The observatiors wliich have l)een here recorded refer only to conditions foinid in squares which have ])een attacked hy the weevil, but before tluy have fallen to the ground. After scpiares have fallen the influence of hot dry weather in largely raising the ])ercentage of mortality wherever the scpuires become directly exposed to the sun is unques- tionably a still more imjiortant factor in destroying the immatuni stages of the weevil. EFFECT ON PROLIFERATION OF FERTILIZING THE SOIL. In accordance with indiccitions shown by some of the examinations of bolls made in lOOo, it was expected that fertilization of cotton might ])roduce a considerable increase in the percentage of cases in which proliferation followed injury hy the weevil. A test of this point required a comparison of a consideral)le munber of varieties under similar cultural and soil conditions with check ])lats unfer- tilized for each variety. Tests of this nature were institiited at the beginning of the season of 1904. Favorable conditions for such observations were also furnished by the field experiment at Calvert, Tex., during the season of 1005. Fourteen varieties were there grown upon Brazos bottom lands. Ejich fertilized plat received an appli- cation of 400 pounds of acid phosphate per acre. Conditions in all plats were similar with the exception of fertilization. 31022— No. 5i)— 06 3 18 I'ROLIFERATION IN CONTROL OF BOLL WEEVIL. t* I .S ^<-' -'-. C3 C tn S f^ c ^ ^ '^Z>^ ' a O cj bOr; r^*- 5 t, -M o o5 ? -d o O OiCACO CI CS OO CC Ol '^ O) .— I C^l .— I OOOOOG0000500t--01000-^OOOC*Ot-- O t^ O O CC 00 o o o Q o c>o r- o :C' O O O GO t-- o 00 t^ O O OlO cc l-^OOOfCOiOOOOOO-^0»000000 1--I"-OOOOC^OC'Cl— tCC'000 0*OC;0 c^J CO o "^ c^I CO o o c^i c-J o ^^ i-H o o o ■-£ X o o o6 — ' c:' o '-c' ^ o o ii— l^o■*r-^oO'-^'-^ocoooooc^|^^oocof^^oo'^^"*oc:■'^ocoooc^O'-^o ; LO .-H -o r— lo o >— I lO c o * t-i-ic;GC--Hoo — C'l-H^OC')'— '.-'O'— I' ^ V. ^ S :?- o hog I- -^ c^ c; '^^ -^ >o t^ "* t-- o c; 00 — >— < O' oC' X) ■^ -xi »o a; c^i rc -^ T— ' o o o "O -^ r-- o CO lO CI coc^^^'^c■j'--*t^■n^'^cr5o6^-l"^»oo■-Hco^^TPcco'^"^■^'0 CI — < lo oi o -h cr- o f:^ co o:' ^ ^ o X o » >o -^ '^^ c; >o lo --H -T' -^ lO lo lo lO CO i-o ■^ -^ CO oi tc to o X >o -^ lO r-- '— lO !•-• -^ t- ■^ ic ' :^it-i— "C')t^rooro«o--0'-«icrooiOco"000'^'^0 — owJ^o^osioo'MOcoo i^"oo C-l O:) r-H c^) CO Cl X o 00 c^i ^D r-- .— I ^ ,— I CO r- '^ CO -^ 00 CO oi c; o o o o '^ CI r-* o '^ o»ooooo■^^oalt^loot^■OlO»oyr:■lOlO»ot--■^»ol0^^cot--^^'V'-| — CO ■ -Ooo ■ ■'^Q ■ 'lOC; OS Oi • • oi r~- O Ol • • CO -^ :• : :g IC ■ . lO ^ o ■ -o^o ■ -oo • • :t. <^j • *o^^ ' -oo ■ • ^-* '.V '.V '/T^ ' r^ ' t3 ' /t^ O CC • ' O CC * ■ 00 -o ■ -00 fa^fi.'hJ^i^I-'li^l-^l-'t^I^^l- ^'J!l;.j[S'^£,^[xiU'!-^^P^P£^fci^t^h3 ^ • ^ *-',*-'., "^ r/i t^ r/i II r/i _i r^, _j m -OO'- t^ C-t lO ^T' C* O ?C O O '* 'O 00 O (M t^Oi^ t^ t^ ^ <^i CO ^ c^ c^ 'iM C^C^C^r-t ^,-HC^^ •OlCCi— '-^OiOC^OOCiT-HOC'ItO-^fCiM'— lOOCC -t^-^»0'^r^Tj"c^ii-^ooiO'^'^^c;^t^t cc — t- ro :0 >0 O CI c: CO ■* ^ O] o CO O T) CO lO CO ^ :o "^ »o lO o C'l -^ -^ CO ■>; r- --o "-o d c-i r^- o r- "^ 1-* cc" i-H C^ ■ ■ (N C^l • • CM CS ■ • (M 1-H ^ T3 ,N -O _N 'C _N - &H^ti,;-j;jH;^ti,L^pM;^Pi,U'fH^iit^&H^tt-tj c3^ aJiS a ^ c3j^ ~.i 20 PKOLIFERATION IN CONTROL OF BOLL WEEVIL. At th(> end of Tal)le Y are ^iven totals and averag'e percentai^es for squares and for bolls, hut the differentiation of the result.^ for fer- tilized and unfertilized plats is more clearly shown in Tabic VI, whieli is ])r-M+3 ^ iz; "* ■^ '' Squares Po 1,604 1,571 50.5 49.5 544 400 182 159 53 35 25.0 23.8 10.2 6.6 14.8 17.2 18.9 Unfertilized 508 1 494 16.2 Bolls Fertilized 2, 094 2, 7,34 60.2 09.5 302 22 286 38 05 .52 2 3 17.7 15.4 8.3 7.3 9.4 8.1 17.1 Do Unfertilized and averages 14.5 Totals 8. 60.3 "01.9 1.040 1,010 4.58 93 021.4 08.4 " Weighted average. An examination of Table VI shows that proliferation follows weevil attack in api)r()xiniately two-thirds of the cases in bolls and in api)roximately c^ne-half of the cases in scpiares. As between squares on fertilized and unfertilized plats, tliere is found a difference of only 1 per cent, whicli is in favor of the fertilized plats. In the firiments, a]i]:»reciably a-ffected the percentage of instances in which ])rolifera- tion followed weevil attack. From a comi^arison of the mortalit}" percentages it a])]5ears that, although among the squares from fertilized })lats thei-e is a slightly larger ])ercentage of squares showing jn^oliferation following weevil attack, there is, on the contrary, a siualler difference in the average mortality which may be attributed to the ])resence of proliferation than is found among the squares from unfertilized plats. With squares on the unfertilized plats having a slightly smaller percentage showing ]:)roliferation there is a somewdiat higher average mortality apparently due to the presence of ])r()liferation. In a similar compari- s )n with bolls, among those grown on fertilized ground showing pro- liferation in 66.2 per cent of the injured locks, tliere was an average increase of 9.4 in the ]x^rcentage of mortality, wdiile in bolls growai on unfertilized ground showing ])roliferation in 69.5 per cent of injured locks the average excess of mortality apparently due to proliferation is onh^ 8.1 per cent. PROLIFERATION FOLLOWING OVIPOSITION IN SC^TTARES. 21 The second signiiicant feature of Table VI is that showing the effect of fertihzation upon the mortahty of the weevil ^\■itllo^!t regard to the presence or absence of prohferation. A coni):)arison of the percentages of niortahty shown in the last column of the table shows US that in the case of sciiiares there is a difference of 2.7 ])er cent and in the case of bolls of 2.(5 per cent in favor of the fertilized plats. These differences are so nearly ;dike in both scjiuires and bolls as to indicate ' that fertilization, as ])racticed in this case, would increase the general average mortality by a small j)ercentage, but that this increase was not due to any increase in thv proportion of cases showing proliferation. One general fact should be stated in regard to held conditions in connection with these observations. As has been stated, the experi- ment was located in the Brazos bottom. The a])])licati(»n of fertil- izer ]:»roduced little apparent difference in the size of ])lants, and the difference between varieties was Ijy no means as marked as is usually the case. It is possil)le that upon soil naturally k^ss fertile greater differences might have ]:)een produced both as regarding varietal char- acters and the effect of the application of fertilizer. From the data at hand, however, M'e woidd not venture to ])redict that such difl'er- ences would result in any greater iiicrease in the mortality o'f the weevil than has l)een found in the obs(U'vations here re])()rted. PROLIFERATION FOLLOWING OVIPOSITION IN SQUARES. The next series of observations to ])e ])resented will deal with a comparison of varieties in regard to the formation of ]n"oliferation following egg punctures in squares. The com])arison includes obser- vations made during three seasons and includes al)out 25 varieties. The table shows also the increase in mortality due, a])])arently, to proliferation. It is imj^ossible to obtain a close comparison of varie- ties in this examination, as the conditions of s«nl, cidtivation, and season were so diverse, and the influence of these varying conditions can not as yet l)e even closely estimated. In the following tabida- tion are included records where there were less than 100 observations in the series. It is noticeable that the greatest differences in the percentage of sc|uares examined which showed })roliferation occurs in cases where examination was made of only a small nund)er of scjiuares and late in the season. It is evident that the data in these cases are insufficient as a basis for reliable conclusions regarding those particular varieties, l)ut the figures may be included in the totals of examinations made without materiallvdisturbino- the freneral averasfes. 99. PROLIFERATION IN CONTROL OF BOLL WEEVIL. '^ s t^ fe « o CO C^l C^J O CC CO cs C C' C C O C' o t^-<}Hf^cioi^»oocooc-Jt-or^b- "-Ot^o0"^i-^'0 0:»oo;t^oox)00sci ci ci o oi (m' -^ -^ c 06 ci ci t^ C'i ^ o oi ^r M c^i 1-1 CI 01 r-i 01 ^ — CI CI -H o o c o ^ fern S cc C-l Oi — ^ ci o 1^ cv o — 00 CO c» CI »o CC' cc c: Pi! e • o 1, ?J 13 o o Micj c £ MS G £ oi 3 ^ St; c o i^ >o o s s r- lO cx T-'i C^ C<1 CC CC y~-> G '>i '>'i '^' ^^ O r-* "^ t-^ r-rH^coiooci^y^oooccor-oco . ■^ ■^ i^ »o 00 lo t-- o i- -0 00 ■^ lO o o o o o r- o - - ) '^ [M CO CO CM (M CC ^ CI •c c 10 .-H (N 10 1— I CC O '^ [M CO CO CM (M CC ' ,-tO"^OWC<)01GOOO'MCO'^I^iOC ■o r-i a; 1— I ■^icioiiOO'^O'-ioOc'rHO'l'^Q X. O Ci -^ O C: X 10 X r- r- •'^^ ?6 CC rH 1-H a: o; -^. 10 ■^ CO r-i -^ ^■; x' 'Oi t^ ro lO 00 o cc O "^ O r^rcoot^oooooooooooo ^ lli 2 d :r Q o o iOh-roa:rc ■^ t-- X c. o rC' — < c-i o o "^ ^ CO o o o lO -rJ 3: X o iM '^ r-^ o o -^ c-i oi o x lO C^? O O t-- I-^ C-l '^ O ro O O :0 O O O O r^ 'Cr: t^ X* --T X c- r- ^- o ci ■^ o o Q o o CC' M ^ o t^ cr. X o "^ « oi r^ o o o rrt-^Q^or^rc'XO-^^r^t^ociic »0 w w "O "^ -^ o »^ -^ lO --O TP CI oi ro (M C^» *M lO ;^ rC' TT -^ O O Ol ■n; CM X lO -^ r-( ^ lO t^ X X C^J "^ — C C — f 1 T I ^ — ' r 1 -H io '*^ -^ -M 55 ro ^t> lO lO Q ■^ ox: ^ . _ iS ^ ; .2 :r; • '5 ' • q; 5 • -^ C _ „ mQ oooooooccoo . oo w2;2 24 PROLIFERATIOlSr IN CONTROL OF BOLL WEEVIL. StiKlyins; Tal)le YII with a view to niakino; a com])arison of the varieties in regard to their tendency to proliferate in response to egg punctures and hirval injuries, it appears that, in cases where one hun- dred or more observations were made, the highest percentage showing proliferation was 67.7 per cent, found in Allen at Calvert, Tex., on September 11, 1905. A previous examination of this variety shows, however, only 50.4 per cent, wdiich is almost exactly the average per- centage found for all varieties. The average of all observations on Allen shows 59.3 per cent having proliferation. The lowest percentage, from observations which are closely comparable, was .33.9 per cent, found in Sunflower at Victoria, Tex., in 1904. Unfortunately there are no other observations upon Sunflower by which this residt may be checked to see whether it may be considered as a somewhat constant tendency in that variety. This being true, it would not be safe to conclude that Sunflower shows the least tendency to proliferate among the varieties examined. It should be noted that the average of the tliree Egyptian varieties grown at Victoria in 1904 is 52.3 per cent, which is slightly above the general average for all varieties examined. Considering all examinations for each of the four varieties having more than 500 observations each, we find for King 53.2 per cent, for Territory 52.8 per cent, for Shine 49.7 per cent, and for Native 45.7 per cent. Because of the larger number of observations made the average percentages shown for these four varieties are undoubtedly the most reliable of all those given in the table. It appears to the writer fi'om the small variation of 7.5 per cent that the tendency of difl'erent varieties to j)roliferate in resj^onse to weevil injury by ovi- position or l^y larval feeding is a remarkably imiform character. Much more extensive examinations woidd be required to determine the ]:)Ositive status of so many varieties in respect to this tendency to proliferate. SUMMARY OF RECORDS FOR FOUR VARIETIES. Examining more closely the portion of Table VII relating to mortality, we find that the percentage of mortality in scpiares with proliferation ranges from to 50 per cent. The latter figiu-e is found in Triumph at Calvert, August 25, 1905. A general average for the 4,121 squares examined is 22.3 })er cent. For the four varieties- Territory, King, Shine, and Native — a closer comparison can be made by presenting the figures in tal:)ular form. INCREASE OF MORTALITY WITH MORE SEV^ERE ATTACK. 25 Table VIII. — Comparison of four rarieties, each hari)t(j orer .'lOO observations in Table VII, shoiving average percentages of mortalitu and infuence of proliferation thereon. Variety. Squares with proliferation, t Squares witliout proliferation. Num- Ijer of squares exam- ined. Territory 1,568 King 1,525 Shine ! (i72 Native : 507 Num- ber of squa res with prolif- era- tion. 812 334 259 Niun- l)er of weevil stages found. 763 250 28S 172 Num- I ber of Per cent squares of stages with- found out dead. prolif- era- , tion. 22.4 Ig!? 27.1 17.5 Num- ber weevil stages found. Per cent of stages found dead. Aver- age num- ber of stages per 100 squares. 740 713 33S 30S 422 331 253 177 9.1 1.7 Average increase in rate of mortal- ity due to pro- lifera- tion. 18.7 16.1 IS.O 15.8 111 scjiiares haviiio- ])rolif oration the range in mortality varies between 16.7 per cent for Kini;' and 27.1 ]X'r cent for Shine. In s(|iiares without prohferation this range is between U.(3 ])er cent for King and 9.1 per cent for wShine. The most striking point in this com- parison is shown in the hist cohimn of the table giving the average increase in mortality due to j)roliferation in each variety. In spite of the variations of S.5 and 10. 4 in the preceding j^ercentage columns there is shown in the last column a variation of only 2.9 per cent. The unfavorable influence of proliferation appears, therefore, to be very nearly constant in different varieties, instead of varying widely, as early indications had led us to antici])ate that it might do. In respect to the rapidity of maturity these four varieties may be fairly considered as ranging from the very earliest to the late varieties. Rapid maturing or, in other words, "short season" cotton does not seem to increase especially either the formation of proliferation or the percentage of mortality occurring in the squares. INCREASE OF MORTALITY ACCOMPANYING MORE SEVERE ATTACK. While only injured scjuares were selected for these examinations there was a difference in the severity of the weevil attack in different fields. It was evident during the growing season that the field in which most of the data for Shine was obtained was being more severely injured than any other in which observations were made. The figures show some very interesting results of this condition, if, indeed, they do not indicate the ex]ilanation for the increased severity of the attack. The four varieties may be arranged in the order of the increasing j^roportion of weevil stages to number of squares examined. The figures for the number of weevil stages found and for the stages dead in scpiares without proliferation are reduced to the common basis of 1 .OOO squares for convenience in comparing. 26 PK()LIFERATI<>]Sr IN CONTROL OF BOLL WEEVIL. Increased mortal it>j lit squares irithout proliferation, accompany iiiy increased severity i7i weevil attack. Variety. Number of squares. Number of weevil stages found. Number Percent of stages of mor- dead. tality. King 1,000 1,000 1,000 1,000 380 620 7G0 810 6 17 91 1 6 Native 2 7 Territory 4 9 Shine 11 2 The comj^arison "[iven above shows very clearly the great increase in mortality accompanying; increased severity in the weevil attack. The (lata here given furnish a very interesting confirmation of the conclusions stated in Bulletin No. 51 of the Bureau of Entomology (p. 119). The statement most directly confirmed is here repeated. By this time the number uf weevils has become so great that the supply of squares is insufficient to meet their need for both feeding and oviposition. Selection of squares so that these two portions of their attack may l)e kept separate can no longer be exercised.- Female weevils are forced to deposit their eggs in squares which liave either received other eggs or been largely fed upon, and a much larger proportion of squai'es at this time shows that feeding punctures are made in squares having eggs or larva'. By these two factors the mortality among young larvie especially is greatly increased. An examination of the figures given shows that in Territory cotton were found twice as many weevil stages as in King, and among these there were 6 times as many dead. In Shine cotton having more than twice as many weevil stages as the King, 15 times as many stages were dead. INCREASED MORTALITY IN SQUARES AND BOLLS DUE TO PROLIFERATION. Next in order will be a s})ecial study of the increased mortality in S([uares and bolls which may be attributed directly to the formation of proliferation. The figures for squares and bolls together include more than 20,000 observations. In many cases the records are taken fi*om data which have been used in preceding tables. FORMATIOK OF I'ROLIFERATION. 27 Table IX. — Suininanj of obscrrationfi shoivuuj increased inortdlitij i n s pkolifp:ration in contkol of boll weevil. Arcw/c perceniaiieofliH'kx In iHi'ich jiroUffmUonfiillDirs trcevij punctures in hoUs as sliouui in srrcrdl pfrcaHiKj lahlis. Tiil)lc- II. Fee(lin,<^ punclurcs in Iwll.s. Prulilciiition fornicd in 44.S per cent of total locks and in SI.:-} jmt cent of locks actually fed upon. Table I\'. Kin<]!; and Shine bolls. Pioliferation foinied in 54 per cent of total locks and in 77.4 percentof locks actually fed upon. Tai)le \'. Fertilized and luifertilized l)()lls. Proliferation formed in (17. S per cent of total lucks. Thes? fiii'iin^s indicato that j^n'oliiVration is stimulated ])y weevil punctures in somewhat nioiv^ than 55 per cent of all locks in bolls attacked. The fio;ures in regard to feeding punctures only, show that jM'oliforation residts in nearly 80 jxr cent of the locks thus attacked. It should be noted here that in many cases the proliferation may have been stinuilated l)y secondary causes, such as the entrance of fimgi or by decay starting in the open feeding punctures. INCREASED MORTALITY OF WEEVIL STAGES DI^E TO PROLIFERATION. As would naturally be expected, a study of the increase of mortality attributable to proliferation will show a somewhat greater variation in the figures for various series of observations than has been found in the ]iercentages of instances in which proliferation occurs. Thus for scfiiares there is foimd a range of from 9 to 31.B per cent, the general average being only about 13.5 per cent. For bolls the range is not as great as fcH* squares, being only from 7.1 per cent to 15.6 per cent, while the general average increase in mortality in bolls was found to be only about 6.3 per cent. This increase is scarcely one-half as great as was the increase found in scpiares. In neither squares nor bolls is the ])ereentage of mortality sufficiently high to appreciably delay the time of maximum infestation by the weevil, since, if hibernated weevils survived in their usual numbers, the number of weevils developing would be abundantly able to totally infest a field by the time the weevils of the third generation had (lei)osited a majority of their eggs. However, the fact that prolifera- tion does ■ autliraenose, wliieh incited proliferatinn. Figs. 8-11 sliglitly enlarged; tig. IJ natural size. (Original, i PROLIFERATION FOLLOWING ATTACKS OF FUNGI. 29 STIMULATION TO PROLIFERATION BY CAUSES OTHER THAN WEEVIL ATTACK. PROLIFERATION .STIMULATED BY OTHER INSECTS. Since beginning this study of ])r()Iiferation it has Ijeen noticed frequently that it occurs commonly in localities where the weevil is not found and ii'om many other inciting causes. Some of the most abundant proliferations have been found in bolls and squares follow- ing the attack of young bollworms {Ileliotlvis ohsoleta Fab.). (See PI. II, figs. 5, 6.) In many cases small columns of purel}" proliferous material have b'een found projected from the punctures made by the ()ressure produced wTthin the square or boll. Similar cases resulting from the attacks of young square-borers — Uranotes melinus Ilbn. (PI. II, fig. 7) or other Thecla larva?, — are to be found in a probably larger proportion of the cases of attack than is generally true with weevils. Many species of bugs commonly produce proliferation of internal tissues in bolls, though no mark of their puncture can be seen in an external examination of the boll (PI. Ill, figs. 8, 9). In this way a Mexican bug (Pentatoma ligata Say) does great damage by inciting pro- liferation in the seeds (PI. Ill, figs. 10, 11) and preventing the opening of the boll. In a series of examinations covering 4,(300 punctures made liy this bug 34 per cent of the punctures were found to show distinct proliferation. A number of species of native bugs have ])een found to incite proliferation in a similar manner. Leptoglossus plvyllo'pus L. and Nezara hilaris Say have been studied especiall}^, and proliferation has ])een found in a large j^ercentage of punctures 'made by these species. In the fields the injury of several species is likely to be so similar in nature and effect as to make it impossible to separate the work of the various species concerned. Thus Euscliistus servus Say, Nezara hilaris wSay, and Thyanta custator Fab. commonly occur together. In an examination of bolls attacked by these three species ])roliferation was found in 52 per cent of the total number of locks examined. Other species of Leptoglossus, especially L. ojrpositus Say and L. zonatus Dall., have frequently l)een taken upon cotton, and undoubtedly they incite proliferatit)n exactly as L. pJiyllopus is known to do. Largus succinctus L. also feeds upon cotton bolls and in all |)robability incites proliferation, although specific instances have not l)een observed. PROLIFERATION STIMULATED BY ATTACKS OF FUNGI. Ill examining large series of bolls it was found that a small per- centage showed distinct and characteristic proliferation on the inner side of carpels, which were severely attacked externally by a fungous disease of cotton known as aiithracnose. Xo other cause for the pro- liferation could be seen, and the number of observations leaves 30 PROLTFERATION IN CONTROL OF BOLL WEEVIL. little doii])t that the anthracnose (PI. Ill, fig. 12) was the cause of the proliferation. In an examination of 1,800 bolls 71 locks showed pro- liferation from anthracnose. Undonbtedlv various species of fungi iind favorable ])laces for attack in the cavities formed by open feed- ing punctures, and these also appeared to incite proliferation in many cases, though it is possible that their attack accompanied rather than caused the proliferation. In still other cases decay seemed to l)e the inciting agent, but whether by chemical stimulus or in some other way is not known. ARTIFICIAL STIMULATION TO PROLIFERATION. In order to determine positively whether the formation of prolif- eration was connected specifically with weevil attack, a series of experiments was undertaken to see if it coidd be produced by artifi- cial stimulation without the intervention of any insect. The experi- ments, as originally planned, were much more extensive than is shown by the figures which follow^, but unfortunately a considerable portion of the records was destroyed in the field through the vandalism of some unknown person. The records secured are sufficient, however, to indicate reliable conclusions to be drawn from the work. METHOD OF TREATMENT. Punctures of two sizes were made in these tests, the smaller by a No. 12 needle, which is the smallest size that is commonly used. This needle is not as thick as the proboscis of a weevil, and it made a small puncture. The hypodermic syringe needle used woidd make a larger puncture than that ordinarily made l)y the weevil for ovi- positing, l)ut not as large as is often made in feeding. The needles were sterilized in a flame before starting a series of experiments, but not })etween the punctures made in the series. The "sealing" referred to in Table X, colunm 1, was accomplished by using a solution of shellac in alcohol. This was not supposed to resemble in nature the mucous secretion used by the weevil in sealing her egg punctures, but was simply expected to close the punctures a]>]^roximately as tightly as does the weevil. However, the shellac solution accomplished this object only partially, as in many cases it soon peeled away from the surface of the bud or boll. Only a single puncture was made in each square treated, but in more than one-third of these scpiares a weevil puncture was present also. The eflects of a few chemical solutions when injected into buds or bolls were tested also in these experiments (PL IV, figs. 13, 14). 59, Bureau of Entomology, U, S. Dept. of Agricuitu Plate IV. Proliferation from Artificial Stimulation. Fig 13— Hroliff ration in seeds followinK artilicial stimulation with iiijeelion of water, enlarpred three diameters. Fig. 14.— Proliferation in carpel and seed following artilieial treatment with aeetie acid. Fig. 1.5.— Prolifenition from carpel and septum enveloi>ing larva in boll. Fig. lt;._Loek of boll burst open by [)ressnre of proliferation formed within. Figs. 14-16 slightly enlarged. (Original.) ARTIFICIAL STIMULATION TO PROLIFERATION. 31 RESULTS WITH SQUARES. As ill previous tables the records for s([uares and bolls are kept separate. Table X.~RemUts s. Is M o „ 3 o is •" o P 0) p a 11 c o . +3 +J is OP, p4 o — MS +j o fe ftp, Puncture made with No. 12 sterilized nee- 5 5 4 7 3 10 4 5 5 5 4 7 3 10 4 5 1 1 2 2 5 2 3 12.8 13.4 14.0 12.7 14.0 13.4 12.5 10.2 1 1 2 2 4 1 3 1 5 2 5 2 • 3 5 4 5 3 5 2 2 100.0 Puncture made with No. 12 steriUzed nee- dle: .-ealcd with shellac 0.0 Puncture made with hyoodermic syringe 0.0 Pun'>ture made with hyDodermie syringe need'e, sterilized; sealed with shellac. . Puncture made with hypodermic syringe needle: one-half drop of 2 per cent solu- tion of formic acid injected: unsealed . . Puncture made with hypodermic syringe needle: one-half drop" of 2 pcrceni solu- tion of formic acid injected; sealed with shellac 28.6 0.0 50.0 Puncture made with hypodermic syringe needle; one-half drop of 2 per cent solu- tion of caustic potash injected; unsealed. Puncture made with hypodermic syringe needle; one-half drop of 2 per cent solu- tion of caustic potash injected; sealed with shellac 50.0 60.0 Totals and averages 43 i 43 16 13.0 1 14 1 1 1 17 26 a 40.0 a Weighted average. It must be remembered that in all cases these artificial punctures, thoutrh sealed, resemble feeding punctures of the weevil much more closely than they do egg punctures. It is impossible to imitate artificially the natural conditions following oviposition, the hatching of the egg, and the gradually increasing irritation accompanying the growth of the larva. Really the effect of a needle puncture upon the tissues penetrated is very different from the feeding punc- tures of the weevil. The needle simply crushes the cells, pushing them aside from its path, and leaving the cavity it makes more or less completely filled with sap and crushed cells. In the weevil puncture the work is far more neatly done than it can be in any bungling imitation. The sharply toothed mandibles at the tip of the weevil's snout cut away the tissues smoothly without crushing or injuring adjoining cells, and the material, being eaten, is entirely removed from the cavity, leaving it dry and clean with the adjoining tissues comparatively uninjured. The difference is really about as great as in a case of the accidental amputation of a limb under a railway train as compared with the work of a skilled sin-geon. The 32 PRULIFEKATION IN CONTROL OF BOLL WEEVIL. hoalini;- followinu- these two operations inijiiit also lie likened to the })roliferation following- artificial and weevil-made punctures. As the proba])ility would be greatly in favor of mortification following from the untreated railway accident, so would decay be more likely to follow a needle puncture in a boll than would the proper healing oi the wound by proliferation. The injection of chemical solutions by the hypodermic syringe would, in all probability, hinder rather than assist proliferation. Unfavoral^le as conditions for these tests seem, it appears that })roliferation was formed in a perceptible degree in 40 per cent of all cases in which artificial ]HUictures were made. The })ro])ortion in cases where punctures were left open is slightly greater than where punctures were sealed, being 43.7 per cent in the former case, and 37 per cent in the latter case. In all these cases the proliferation was entirely distinct from that caused by larva^ when such were ])resent. In many instances the proliferation was abu.ndant and very })lainly characteristic, in otlier cases there was only a slight formation with more or less of decay. The fact that the experiments were undertaken at very near the close of the growing season would account for the formation of less ])roliferation than might have resulted from similar experiments earlier in the season. RESULTS WITH BOLLS. More extensive exj^eriments were made with bolls than with squares in testing artificial stimulation to proliferation. Part of the experiments were performed at Dallas and part at Hidalgo, Tex. The same methods were used as with squares. The number of punc- tures per boll averaged about six. In the experiments at Hidalgo it was dillicrdt to find bolls which were wholly free from weevil attack. The fidl data from tlie experiments are given in the following table: RESULTS WITH BOLLS. 33 00 1- 1^ ■3" OC »0 "* '-D r- H C t^ « •noii'Biajiioid ira oi d -H d d d d -I d d d t^ c-i t-^ 10 d ^ ci d c-i d d d CO SuiMoqs sainjotind "^ e IBpgilJ'B JO ^nao jaj; "3 ranjdas ui 000000 — (O-H oooooco -^OCTO n .-( .h ^ N. 'i . mt^irat^ooo -s-OQO coootooo ^o^t^ co co 00 -h 00 ft •Jtooi ui M CO CJ -a" .-1 rt r-( r-1 CO •ladi'BO uj oi'OiOrtoo -a-i^oo o.-itoio«t^ (Not^— 1 -h 00 m ^ M i-H,-! rH .-l-9< rH -^ M ■>1< ■«■ .-( rt s? £ ^ M< LO (M M -0^0 10 -^ C-l 55 - - xo:r>p CO 00 t^ 00 ■no!:>Bjajnojd on ^^2S^ . ^ ^ g 11^ » cS 3 •noijBja 00^000 000 000000 0000 0000 -JHOid inomiAv •uoij 000000 0^0 oooi-irHo x-^cic-i «:) h- c^ 1-i ■w -tTjajijojci miAV ^rt-H • V« H« wi-t foil- -W -!3t^* •lajeuiB OJ -ip 'noq JO azTs sSvi3AY 1 c ■noi^oanoa 05. jnaun'Baij uiojj sXbq 0000 C'lo 1000 0000 t~o OfOC -t CO t^ CD -^ CO ^ i i i ' i ! ! ! ! I ! ^ --o CO • ■ ^ 0000c oco 000 • ■paj'Bajj uoq^\v § +j -a -a 13 -o -o 'O'3'S 'OT3'aiJ+j> ^-o-a-o -a -a -o 'd ^0 ■■■ooo O'- : : : : : : : : : : ;oo^ Z : : •juasajd ooccMOO o-H—i ooo-*r-io i^-*r^r^ oi 00 -o^ M C-l « r- rt rt ,-1 s sainjonnd ijAaaAv \v%ox, 1-H •XlfBIOIJIJIB Oi C5 10 ^ C; C 'J" O) !N (M C] 00 CO » -< O Ol c5 ap'Btn sajn'jound irjox i-H ^ CO'^COiO'OCJ ■^■^Cl C1C1"^00:01 C:COO)CO 00 iO 00 •paj^aJl siioq JO laciuiiiN ' rt CI C-l C-) rt r-i ■a c •a : !xj ft ■. ft ft ft ,+j CS 5 3 ii . 'r; fc^ t^ u. sh ■a ? -3 -a r) 'a . CS ^ uw >w <— — ■0 « ^ —, — « 3 3 0- t. It a ci oi . 03 C3~ ■a +j ..mi ft2 = TO ^ n"1 J J TO ' ' _£-! a Si X ■33" 3 C~3 C"' - "C ft o9£ •Sg| all 1 isiii'l 3 i" ^ ffi Sa:-=-c .ES.Sg.e.;.-.e-- •a 3 m . >» >» t>,<^ ;>.o s J; - 3 ii: — ij' 1 ? -o " "a ? ? -3 ^ 60 si) SEE ■a = .H - c S' f" E 0.0 CJ .0 .Ji, .=J^ '3 u 3 '5 a 'S 's" ■a O; 3 ii'a„ s c ^ 3Sgo >-..^t..s:,t-„t,'^ ill'l llllPlI Ills glSiaiai > '' •^ " ?oPo:SciS^ci ■a ft& 1 0' 'A 1 ,0 ■a ,0 ■a '3 ^' 3 ftB ■^'2 ■2.2 " ^ -a ^.s 2J '^ ■sf^ OJ c ■a ■a -a •a t3+i^+j.a+^'a+^ OS c^ |2 3 cj I2 S S =s OS c8 ci.c: osacjacjCcsC s B |se SHaE gS6&3&aS 0) 0^ (rf a; a; 0) a; ■a '^'i,? e 13 3 1 cSS oj ■3 p ..■a x' S "3 ^-.a S<5 p C3 « > S „ or; Eh p d c H H-<"mcomes unnecessary to assume that any irritating secretion is deposited by the weevil with the e^g. It has l)een shown conclusively that proliferation may occur entirely apart from weevil presence. Proliferation may result from a large number of causes, of which the following are noted in this bulletin: Weevil feeding punctures, weevil egg punctures, injury l)y the larva during its feeding period, bollworm punctures, square-l)orer ])unctures, feeding punctures of various bugs, fungous attacks in insect punctures, anthracnose, and artificial ])imctures of the bud or boll. MANNER IN WHICH PROLIFERATION CAUSES DEATH OF WEEVIL STAGES. REARING LARV^ ON PURELY PROLIFEROUS FOOD. In order to determine whether ])roliferation caused the death of larva^ l)y starving or poisoning them, numerous experiments have been performed. ITnhatched eggs and larvir just hatched have ])een placed in the midst of masses of ]:)iu-ely ])roliferous formation; and these have been kept in a tight, moist chamber and transferred to fresh masses of proliferation as frequently as necessary to preserve proper condi- tions for the larvae. In spite of these frec{uent transferences and the somewhat unnatural conditions necessary, it was found that but a very small proportion of the larv» died. In some cases the growth was completed in masses of ])rolife ration which were coin])letely decayed. This condition was allowed to exist in order to test the effect of what seemed the most unfavorable food conditions it was ])ossible to pro- duce so far as quality was concerned. In one series of ex]>eriments 8 very young larvas were ])laced in fresh ])roliferous masses. Of these, 1 died on the first day, but all others reached full growth and ])U])ated normally in from nine to twelve days, having fed on nothing ])ut pro- liferous material. In another series 8 larva> just hatched were started in locks in which dead fidl-grown larvte had l)een foimd entirely envelo])ed l)y the ])ro- liferation. Surely if ]>roliferation were poisonous, this test should prove it, since in each case a larva had l)een previously killed in the lock given each young larva for food. One larva was accidentally killed in examining the material on the fourth day. One larva com])leted its growth and ]>u])ated in the lock in which it was originally ])lace(l. Two died and the remaining 4 also ])U])at(Ml after being transferred to fresh locks from which dead larvaj were previously taken. ITnder tht^se most unpromising condi- 36 PROLIFERATION IK CONTROL OF BOLL WEEVIL. tions, therefore, 5 of the 7 larvse tested reached the pupal stage. Tliis experiment was performed between November 17 and December 12, 1905. The larval stage averaged about thirty days in duration. It does not follow from the length of this stage that the food conditions were even imfavorable, since at that season in bolls the length of the larval stage would undoubtedly exceed thirty days under normal field conditions. These experiments alone would be sufficient to prove that the mortality caused by proliferation is not due to insufficient nutrition or to poisonous qualities in the food material of the larvae affected. Furthermore, the examination of thousands of squares and bolls has shown that in a very great number of cases weevils reach maturity m the field on no other food than proliferous cells. MECHANICAL CRUSHING THE REAL METHOD. The real cause of death in the presence of abundant proliferation will become apparent to any one who will take the pains to examine care- fidly a few thousand bolls which have been injured by weevil attack. In one series of observations, covering 1,800 ])olls, 1,016 weevil stages were found. During this examination a partial record was kept of those cases in which the cause of death was unmistakably shown, with the following result: Three adults just alive, l)ut l>aclly deformed by pressure- Two pup« just alive, but l)adly deformed by pressure. Two pup£e unmistakably crushed to death. Twelve larvae unmistakably crushed to death. These 19 cases formed nearly 2 per cent of all the stages found. The record was not kept by all those engaged in the examination of this series of bolls and only the most unmistakable cases were recorded at all. It is certain, ther:fore, that this 2 per cent is but a small part of the true proportion of weevils which were killed in this way by the proliferation. Considering these facts in connection with the conclu- sions as to the food quality of proliferous cells, it seems safe to conclude that the great nuijority of deaths due to proliferation may be caused by the mechanical effect of the formation in first enveloping the larva so closely as to prevent its movement (PI. IV, fig. 15), and then the continued formation producing sufficient internal pressure (PI. IV, fig. 16) slowly but surely to crush to death the foe Vvdiose attack has called forth this effort at self-defenso on the part of the plant. Such an explanation alone fidly accords with the facts as we know them. These observations present to us in a very vivid way an illustration of the intensity of the struggle, continually going on between plant and insect life. It is a life and death struggle, and it is not always the insect that wins. 3ul. 59, Bureau of Entomology, U. S. Dept. of Agricultur Plate V. Origination and Effects of Proliferation in Bolls Fig. 17.— Proliferation beginning lUKicr farpd linintr. Fig. IS.— Proliferation p^■s^inl,' into pupal cell. Fig. 19.— .\ilnlt weevil dcinrmiMl by pres.'iure of proliferation. Fig. -'O.— Pro- liferous mass spreading over inner side of carpel. Figs. 17. is. 20 slightly enlarged: lig. 19 enlarged four diameters, i ( »rii;inal. i Jul 59, Bureau of Entomology, U. S, Dept. of Agriculture Plate VI. Locks Completely Filled by Proliferation. Fig. 21.— a, Point of original egg puncture and first proliferation: b, larva crushed l>v proliferatioi crowding upon it from all directions. Enlarged four diameters. ( Origiiial. ) CRUSHING OF BOLL WEEVIL BY PROLIFERATION. 37 EXPLANATION OF MECHANICAL ACTION. A brief explanation of some additional points regarding the forma- tion of proliferation may serve to show more clearly how it becomes possible for the plant to literally crush its irritating foe. The explana- tion will be given for bolls rather than for squares, though the real effect of proliferation is the same in squares as in bolls. Proliferation usually begins in the layer of cells adjoining the thin, tough lining within each section of the boll. By far the greater part of this formation projects through the rupture made by the weevil in the tough lining and forms a rather hemispherical mass protriuling from the inner side of the carpel (PI. V, fig. 1 7) and pressing into the lock. The formation sometimes, though not always, begins before the hatch- ing of the egg, which may be moved quite a distance, in some cases, by the pressure of the mass behind it. In other cases the egg becomes enveloped and the larva hatches into the proliferous mass. In such cases it may be destroyed early in life, though it will often make its way into the lock, eating its way as it goes. As it feeds the larva is continually injuring and irritating tissues capable of proliferation, which thus becomes started all around the larva and gradually pushes in u])on it from all directions (PI. V, fig. 18). It may happen in this way that the space which the larva has eaten out as it grew becomes filled by the masses of cells pushing in upon it and the larva can not possibly eat away the forming mass rapidly enough to preserve room for itself to move (PL V, fig. 20) . Though it may be nearly or quite full grown, it can not escape from its narrowing prison and soon becomes so closely enveloped as to be unable to move in any direction. It is then an easy victim for the relentless pressure of forming cells and is literally crushed to death in its prison (PI. VI, fig. 21). Very frequently, indeed, instances are to be found in which the plant gets a tardy vengeance on the pupa or the newly transformed adult (PI. V, fig. 1 9) . Whether death results within a short time or the victim is allowed to emerge with only some deformit}'' to tell of its narrow escape within the boll, depends largely upon the continiumce of the prolifera- tion. Deformed pupte and adults are by no means uncommon and in nearly all cases they are luidoubtedly the partial victims of this form of plant defense. Many of these specimens have been so deformed by pres- sure upon the pupa that the adult can not feed. These would be unable to make their escape from the boll did it not hapjjen sometimes that the maturing of the boll breaks open the prison cell of these victims and turns them out only to perish slowly by starvation. 38 PROLIFERATION IN CONTROL OF BOLL WEEVIL. PROLIFERATION IN PLANTS OTHER THAN COTTON. The most definite and abundant observations of proliferation in plants other than cotton have been made in two species of peppers in connection with the work of the pepper weevil {Anthonomus seneo- tinctus Champ.). Proliferation was very distinct in 93.5 per cent of the pods of sweet pepper which had been attacked by the pepper weevil. It was also found to liave formed in three-fourths of the feeding punctures. In pods of the chili pejjpers ])roliferation was found in about 38 per cent of those examined and in about 34 per cent of the cases of simple feeding punctures. Anu)ng the 300 pepper pods examined no trace of mortality residting from the proliferation was seen. Among other plants no special observations seem to have been made to determine the presence or absence of proliferation, but it may be allowable to state here that a similar formation, which has every appearance of being homologous with proliferation in cotton, has been observed by Mr. F. C. Pratt in the pods of garden peas, by Mr. C. R. Jones in the pods of cowpeas, by Mr. A. C. Morgan in the buds of CaJlirrTioe involucrata, and by Mr. J. C. Crawford in the seed pods of mesquite. It would appear probable that when six'cial inves- tigation shall be made of the occurrence of ])ro]iferation in other plants than cotton it will be found a not uncommon phenomenon in very widely separated species of ])lants. " Naturally, it may not 1)6 expected to occur in response to the great majority of insect injuries, since it depends upon a number of coincident favorable conditions, and the presence or absence of some other and entirely inirelated factor may prevent or obscure its formation even where some of the essential favorable conditions are present. CONCLUSIONS AS TO NATURE AND SIGNIFICANCE OF PROLIFERATION. In all cases, whatever the stinmlant, one factor is uniformly essen- tial. There nuist be a cell injury which is not sufficiently severe to overcome immediately the vital force of the injured organ or tissue. Proliferatiorf is simply the manifestation of a natural inherent tend- ency of plant cells to respond to an encountered irritation i)y nudti- plying or forming new cells. It is evidently a method of self-defense, and in the case of cotton the irritation apj^ears to be in nearly (1 The possible general occurrence of proliferation as the result of insect attack is shown by the followino; quotation lelating to Anthonoimis qnadrigihbns Say on apple, by Prof . C. S. Crandall: Many of the egg-cavities cut into were found to be more or less completely filled by intruding cell masses. These cell masses were quite firm in texture. Sometimes they invaded the cavity from tha bottom, but often grew as wart-like excrescences from small areas on the sides of the cavities. In several instances dead larvae were found pressed close to the cavity wall by these intruding cell masses. (Bui. 198, 111. Exp. Sta.; page 528.) CONCLUSIONS KEOARDING PROLIFERATION. 39 all cases strictly mechanical. The function of ]iroliferation in most cases is undoubtedly to repair an injury. From the numerous observations dealt with in detail in the preced- ing; pao;es a number of conclusions seem to be warranted. The phe- nomena considered are very complicated, and consequently only a few generalizations are made. (1) In a large nund)er of varieties of American u])land cotton proliferation has been found to oc-cur in 51 per cent of the cases of weevil attack upon squares and in .5.5 per cent of the cases of similar attack upon })olls. (2) P^liminating a certain jx'rcentage of mortality, which was found in cases where no ])r()lifera.tion occurred, the increased rate of mor- tality in all W(*vil stages apparently caused b}" proliferation was, in squares 13.5 per cent and iii l^olls 6.3 per cent. (3) Ordinary variations in climatic conditions seem to have com- paratively little effect upon the proportion of injuries })r()liferating, although hot, dry weather plainly increased the mortality occurring without regard to the jiresence of jiroliferation. (4) Contrary to a })revious tentative conclusion, ])ased u])on a much smaller mnnber of ol)servations,° the upland American varieties seem to l)e somewhat on a. j)arity so far as the tendency toward pro- liferation is concerned. (5) The application of different fertiliz(n-s to cotton has thus far failed to show any tendency toward increasing the proliferation. (6) Proliferous tissu{> is not toxic to weevils. Death results in most cases in a mechanical way from simple pressin-e. (7) Proliferation is caused by the attacks of a niunber of different insects, and is easily produced artificially hy needle ])inu'tures. Its stimulation appears to be from mechanical irritation, and, conse- quently, a secretion on the ])art of the insect does not seem to be essential. (8) Proliferation occurs commonly in plants other than cotton as the residt of hisect attack or from mechanical injury. It has ])een noticed in the seed ]>ods of several species of Leguminosa^ and in differ- ent species of Capsicum. PRACTICAL APPLICATION OF CONCLUSIONS FROM THIS STTTDY. The fuudamental purpose underlying all this study of proliferation, its causes and its effects, is to learn, if possible, facts which may be made of practical use in the effort to grow a profitable crop of cotton in the area infested ])y the boll weevil, either by so controlling the multiplication of the weevil as to reduce its injury to a sufferable ((uantity, or b}^ raising the crop so as to avoid the serious injury which the pest has shown itself capable of producing under the S3^8tem a Bui. Nq. 51, Bureau of Entomp|ogy, U. S. Dept, of Agric, p. 134. 40 PROLIFERATION IN CONTROL OF BOLL WEEVIL. of culture which lias been customarily employed. Many factors must be considered in any hopeful solution of this most serious problem. With insect pests ounces of prevention are worth many pounds of cure. The most promisini^' solution of the weevil problem is undoubt- edly found in a combination of the factors restricting" weevil develop- ment with those favorino- crop improvement. The facts learned from this study of proliferation may be utilized in the class of factors restricting weevil development. It appears that there is a small variation l)etween ditferent varieties of cotton in regard to the proportion of cases in which weevil punc- tures stinndate proliferation. It is evident that the presence of pro- liferation increases somewhat the percentage of mortality among the larvic and pupje in proliferating buds and bolls. The "plain conclusion is that the varieties whicii proliferate most freel}" will by that charac- teristic tend to restrain the rapid multiplication of the weevil. It is probable that varieties ma}^ be developed by repeated selections which will be more elective than any now known in restraining weevil development in this way; still, this factor alone will probably never be of more than secondary importance in reducing the number of weevils, as other considerations will inevita))ly be more important in determining the most desirable variety to plant. Although the obser- vations thus far made have failed to shoAV an}" conclusive etiect of fertili/ation of the soil upon proliferation, further investigations should l)e made upon this point. Much work would still l)e necessary to determine any constant relationshi}) l)etween the formation of pro- liferation and climatic conditions. Prohaldy little practical use could be made of a knowledge of such climatic relationships if ascertained, as the influential factors would always remain beyond the control of the cotton planter. The tendency to proliferate is by no means a recently acquired char- acteristic of cotton; therefore it should not l>e supposed that it is any more susceptible to such variation as will render it a still greater obstacle to weevil development than are many other characteristics wdiich ma}" l)e emphasized with equal or even greater advantage in the selection of new strains of cotton for growth in the weevil-infested area. Such selections require much time, and we may, therefore, feel somewhat encouraged to know that in the long tight yet to come we may expect this natural factor to accomplish no less than we have found it now doing toward the repression of the weevil. Of course complete reliance can never be placed in natural factors for a solution of the weevil prol)lein. Doubtless the capacity of the weevil for adap- tation to any new conditions which its food plant may present is just as great as is any adaptive capacity of the plant. Nevertheless, the interference of man is likely to throw the advantage greatly in favor CONCLUSION^; FROM THIS STITDY. 41 of the plant. Natural factors havino- a controllino- intiuence on the weevil will do their work no less etieetively if the intelligent assistance of the planter l)e given in the planting-, culture, and subsequent care of the crop. Repeated and widely varied experience has proven that the intelligent planter can, as a rule, so assist natural factors, l)y adopting certain changes in his methods of cultivation, as to produce a protital>le crop in tlu^ face of conditions which would otherwise have insured its failure. INDEX Page. Acid phosphate, in experiment with fertiUzers 17-21 Adaptation of boll weevil to new ctjnditions 3, 40 Allen cotton, observations recorded 10, 1 1 , 18, 22, 23, 24 Ant enemy of boll weevil. See Solenopttis (jcmlnata. A ntlionomus quadrigihhus, inciting proliferation in apple 38 Anthracnose, inciting proliferation 29-30 Apple, {iroliferation from attack of Anthonoihus quadriglhhns 38 Artificial stimulation, producing proliferation in scjuares 31-32 bolls 32-34 to i)rolifcrati(in 30-32 Ashmouni cotton, ol)Hervations recorded 10, 22 Bohemian cotton, observations recorded 10, 11, 10, 22, 23 r>()ll weevil, stages, num1)er in j)roportiou to squares, King cotton 13 Sliine cotton 13 Bolls, climatic conditions affecting proliferation 15-17 Bollworm. See Ili'Hiilhh olinolehi. Bugs, inciting proliferation 29 Callirrhfi' inrolnrrdfa, proliferation _ 38 Capsicum, proliferati( m 38 Carpel, proliferation 8-9, 29, 34, 37 Climatic conditions, influence on proliferation in bolls 16-17 scjuares 16-17 those most fatal to boll weevilstages 16,17 Cowpeas, proliferation 38 Crushing of boll weevil larvae by proliferation 36-37 Decay, following chemical injections 32, 34 in connection with i)roliferation 28, 34 Deformed boll weevils, caused by pressure of iiroliferation on pui»a 37 Dickson c< (tton, observations recorded 10, 22 Egyptian cotton, observations recorded 10, 23, 24, 33 Euscli iittus srrnis, probably inciting proliferatii >n 29 Feeding punctures in bolls, proliferation as result 11-12, 28 in squares, proliferation as result 10-1 1 , 27 stimulating proliferati( m 27-28 Fertilization of soil, effect on proliferation 17-21 Fungi, inciting proliferation 28, 29-30 Gelatinization, term applied to proliferation 8 Hawkins cotton, oI)servations recorded 1 0, 1 1 , 1 8, 22, 23 Ildlotlus olisoletd, inciting proliferation 29 Hetty cotton, observations recorded 10, 1 1, 19, 22, 23 Jannovitch cotton, observations recorded 1 0, 22 Kekchi cotton, oljservations recorded 10, 23 43 44 INDEX. Killer Imlls, liiji'li ))()11 wc't'vil luoi'talit y 15 Kin, IS, 2l', l.'4-L*(), 27--S, .">;> Korean cotton, observations recorded 11, li', L':] LaboratoTv work, l)eginnin_<; 7 Laryiis .snt-ciiu-his, ])roba1)ly i licit iny- ]irolil'eration 29 Lefynininos;i', ])roliferation 88 Lrploi/los.siis iipjHisiliix, ])ro))al)ly incitin.tj; pi'oiiferation 29 />li>l/l<)jifis, incitin.i:: ]iroHt'eration 29 zoiKitiiK, probalily inciting- proliferation 29 Lock, unit of obser\ation in bolls 9 Mascot cotton, ol)servations i-ecorded 10, 22 Maximnni infestation, not greatly delayed by j.roliferation 2S Mechanical action of proliferation explained 'M effect of proliferation .'>()-87 Mes(|uite, proliferation .SS Methods of stndyint; proliferation 9 Mit Atiti cotton, observations T-ecordcd 10, 1 1, 22, 38 Mortality, average incl•eas^' from jH'olifeialion, Sliine and King ];>-14, 25 from ]>roliferation, intlnenced little by climatic comlitioiis 16 somewhat constant facti ir 14, 16 increased by proliferation 14, lS-20, 2B-2S, 39, ublished record 7 function 38 general appearance 8 greater in Egyptian than in rjpland varieties 10,24 in l)oth plant and animal cells 8 in plants other than cotton 38 INDEX. 45 I'ase. Proliferiitiou, iiicthdil of sfmly „ 9 not connected especially with hyoil weeviF injnry 35, 39 not increased by fertilization of soil 20, 39 especially in "short season" varieties 25 occnrrence more frecpient in hulls than in s(juares 12,38-39 of similar ciiaracter in hnd and lioil 8 one of most important natnral factors in control 3, 28 points of originatii m 8-9 purpose of study 9 resnltinii in iiolls from artificial stinuilation 32-34 S(jnares from artilicial stinndation 31-32 secondary factor in choice of varieties 40 stinudated !iy feedimr punctures in l)olls 11-12 squares 10-11 fungi and decay 28, 29-30 tissues reacting most readily 8-9 Rearing larva:" on proliferous food 35-36 Red cotton, ol)servations recorded 33 Rowden cotton, ol;)servations recorded 10, 11 , 18, 22 Russell cotton, ol)servations recorded 10, 11, 18, 22, 23 Sealing of punctures, method 30 influence on proliferation 34 Peed, proliferation 29, 34 Septum, ]iroliferatiou 34 Shine cotton, o))servations recorded 10, 11, 12, 13-15, 18, 22, 23, 24-20, 27-28, 33 Soil types, range 21 Sob'uopsis gem iiiala, enemy ( if holl weevil 28 Scjuare 1)orer. See rV«»o/<>" nu/inux. S(|nares, climatic conditions affecting proliferation 16, 17 Starvation, not cause of death of holl weevils with pi-ol iteration 35-3t) Stinudi, causing cells to jiroliferate 8 Sun, influence on exposed fallen squares 17, 28 Suntli iwer cotton, observations recorded 22, 24 Territory cotton, observations recorded 10, 11, 19, 22, 23, 24-26 Texas wool cotton, oljservations recorded 33 Thecia larv;c, inciting proliferation , 29 Tln/aiild rnxhilur, prol)ably inciting proliferation 29 Tissues, <'hange following boll weevil attacks 37 Tools cotton, observations recorded 10, 11, 18, 22, 23 Triumi)h cotton, observations recorded 10, 11, 18, 22, 23, 24 Truitt cotton, observations recorded 10, 11, 19, 22, 23 rrdtioh's nicliiiits, inciting proliferation 29 Weather Bureau, use of records sliowing climatic t'onditions 15 o Lb Mr I \'^^ ."T^- ^^ /::x>^