Historic, archived document Do not assume content reflects current scientific knowledge, policies, or practices U. S. DEPARTMENT OF AGRICULTURE. DIVISION OP ENTOMOLOGY. Bulletin No. 2y. PWISION OF VEGETA^-^pbRT ON HIE BOLL WORM OF COTTON (Heliothis armiger Ilubn.j. MADE UNDER THE DIRECTION* OF THE ENTOMOLOGIST BY F. W. MALLY. PUBLISHED BY AUTHORITY OF THE SECRETARY OK AGRICULTURE. WASH I NGTON: GOVERNMENT PRINTING OFFICE, 1893. U.S. DEPARTMENT OF AGRICULTURE, division of entomology. Bulletin No. 29. DIVISION OF VEGE REPORT BOLL WORM OF COTTON (Reliothis armigcr Hlibn.). MADE UNDER THE DIRECTION OF THE ENTOMOLOGIST BY F. W. MALLY. PUBLISHED BY AUTHORITY OF THE SECRETARY OK AGRICULTURE. WASHINGTON: GOVERNMENT PRINTING OFFICE. 1893. LETTER OF TRANSMITTAL. U. S. Department of Agriculture, Division of Entomology, Washington, B. 2. Sir: I have the honor to transmit herewith, for publication as Bul- letin No. 29 of this Division, a report by Mr. F. W. Mally upon the Boll Worm of Cotton (Heliothis armiger Hiibn.), the first part cover- ing his observations upon the parasites and natural enemies of the Boll Worm while the second part is devoted to his bacteriological ex- periments with certain insect diseases affecting this larva. Respectfully, C. V. Riley, Entomologist. Hon. J. M. Rusk, Secretary, LETTER OF SUBMITTAL. Washington, D. C, May l, 1892. Sir : I submit herewith a report upon the remedies for, and the par- asites and natural enemies of, the Boll Worm (HeHothis armiger Hiibn.) covering the results of investigations carried on under your direction. Bespectfully yours, F. W. Mally, Assistant. Dr. C. V. Eiley, Entomologist. 5 CONTEXTS. Page. Acknowledgments 9 Habits and natural enemies of the Boll Worm 10 Destructiveness... 10 Food-plants other than cotton and corn 17 Tobacco 17 Tomatoes 18 Other food-plants 18 Characters and transformations 18 Larva 18 Pupa 21 Imago 22 Number of broods and hibernation 23 Parasites 24 Egg parasites «. 25 Parasites of the larva 2f> Other natural enemies 26 Ants 20 Wasps 28 Other insects 28 Insect ravages mistaken for those of the Boll Worm 29 Thecla poeas 29 Prodenia lineatella 30 Platynota rostrana 30 Agrolis yjwilon 30 Laphygma frugipeida ::<> Baris (erea Parajulus impressus 30 Calocoris rapidus ;'»1 Homalodis Simple cold-water decoction In Simple hot-water defections IT Summary of experiments 49 Advantages of the emulsified hot oil extracted from pyrethrnm. 5] 7 8 Page. Hand-picking of corn 52 Trap-corn experiments 52 Early and late cotton 58 Bacteriological experiments witli insect diseases 59 Introductory 59 General precautions 60 Culture media 60 Beef broth 61 Agar-agar 62 Special apparatus 62 Observations and experiments 64 A disease of Plusia brassicce 68 Discussion of results 69 ILLUSTRATIONS. Plate I. Diagram of cotton field, showing location of trap corn 56 II. Special apparatus for bacteriological work 62 REPORT ON THE BOLL WORM OF COTTON. ACKNOWLEDGMENTS. My first acknowledgments are due to Dr. 0. V. Eiley and to Mr. L. O. Howard, who have furnished valuable aid in the determination of specimens ; to Mr. E. A. Schwarz for identifying Coleoptera; to Mr. Theo. Pergande for naming ants, and also to Mr. Nathan Banks, who was assigned to me in my work at Shreveport, for determining spiders. The planters at Shreveport deserve great praise for their uniform courtesy, hearty cooperation, and the many sacrifices made in devoting time and labor to the promotion of field experiments. Those who as- signed portions of their plantations for the sole purpose of testing remedial measures, and who therefore deserve personal mention, are Messrs. J. H. Fnllilove, Daniel Nicholson, S. J. Ziegler, and John Caldwell, all of Shreveport; Mr. M. A. Curtis, of Curtis, La., and Mr. John Glassell, jr., of Rush Point, La. The wide geographical distribution of the Boll Worm, and the differ- ent natural conditions in the various regions where it occurs, made it quite impossible to cover the entire ground without the assistance and cooperation of intelligent persons throughout those regions. Accord- ingly arrangements were made with the several State weather services to have their observers give special attention to any facts of interest and value to the investigation. This was accomplished through the efforts of the directors of the several services. Mr. M. G. Wright, jr., of Shreveport, La.; Prof. K. B. Fulton, University, Miss.; Mr. F. II. Clark, Little Rock, Ark.; Dr. J. M. Cliue, Galveston, Tex.; Mr. George E. Hunt, Xew Orleans, La., and Prof. P. H. Mell, Auburn, Ala., merit special mention for their many favors. While on a trip through northern Texas for the purpose of making special observations, much depended upon the assistance of observers in that section. It is a source of much satisfaction to state that no de- lay was experienced at any point, and that the work was facilitated in every way possible. At Mesquite, Tex., Messrs. S. G. Lackey ami T. P. Worthington gave valuable information concerning their localities, 9 10 as did Mr. A. A. Pittuck, of The Texas Farm and Ranch, and Mr. F. Doreinus, of the Dallas Morning News. At Arlington, Tex!, sim- ilar courtesies were received from Dr. L. C. Page, Mr. 0. F. Mercer, Capt. M. J. Brinsan, Col. J. A. Ditto, and Hon. J. W. Hammack. Among those who by correspondence contributed freely of their ex- perience, Messrs. S. B. Mullen, of Harrisville, Miss. ; Jeff. Wclborn, of New Boston, Tex.; John C.Edgar, of Duval, Tex. ; Hon. George J. Twiley, of Holly Springs, Miss. ; Prof. H. A. Morgan, of Baton Rouge, La.; Prof. J. (x. Lee, of Calhoun, La., and many others, have my sincere thanks. Grateful acknowledgments are also due to Mr. Henry Hot- ter, Secretary of the New York Cotton Exchange ; Mr. Henry Hester, Secretary of the New Orleans Cotton Exchange, for many favors, and to Mr. A. B. Shepperson, of New York, for " Cotton Facts " and general statistical information. HABITS AND NATURAL ENEMIES OF THE BOLL WORM. DESTRUCTIVENESS. During May corn is practically the only abundant and available crop to be attacked by the Boll Worm. At that time a study of a number of corn fields on both bottom lands and uplands was made. The number of plants attacked was noted, as well as other data, as shown in Table I. Table I. — Ravages of Boll Worm on May corn. Field. Total. 1. 2. 3. 4. 5. 6. 7. 8. Plants examined Eaten 377 1G 361 3 2 1 296 7 289 4 1 3 1 472 12 460 1 1 720 21 699 5 1 1 2 4 1 422 4 418 496 6 490 511 16 495 3 1 2 2 368 15 353 8 ..... 7 3,662 97 3, 56.") 26 7 19 18 1 7 7 16 3 Color : Dark 3 3 1 3 1 1 1 1 1 1 1 ....„ 1 Mashed : 2 1 i It will be seen from this table that of the 3,602 plants examined, 97, or 2.6 per cent, showed injury, and 26, or 0.7 per cent, actually contained Boll Worms. Fields 5 and 6 were "hill country," and though plants were found apparently injured by Boll Worms, much of the damage done was due to Prodenia Mneatella. The other fields were Bed Eiver bottom lands. 11 June 1 a study was made of a small patch of sweet corn, which was then in good roasting ears. The results are given in Table II. Table II. — Ravages of Boll Worm on sweet corn roasting ears. Ear. Worms. Size. * Color. Grown. Hall grown. Very young. Dark. Green. Light green. Rose 1 2 3 4 5 d 7 8 9 10 11 12 13 14 15 16 .. 6 3 2 6 1 5 3 4 3 1 1 *12 5 6 3 2 3 1 c :; l 4 2 2 1 9 1 2 2 1 1 3 1 2 1 4 2 1 1 1 1 1 1 2 1 1 3 2 3 *9 3 . 3 5 5 3 1 2 17 18 .. 2 2 2 19 Total . . 1 1 1 i 64 6 22 36 28 2 13 2 * One dead. On the same date a field of crop corn just tasseling was similarly studied. Tlie number of plants examined is not given, but simply those upon which worms were found. The larvae in nearly every case were found in the freshly protruded or protruding tassel. The facts are pre- sented in Table III. Table III. — Worms found on tasseling corn. Plant. Worms. Size. Color. Grown. Half grown. Very : young. Dark. Green. Light green. Rose. 1 2 3 4 6 7 Total . . 1 3 1 2 2 1 2 1 \ 1 1 1 '-• 1 1 :::::::::: 1 2 1 1 1 I 1 :::::::::: :::::::::: 2 2 12 6 5 1 5 4 ' 3 1 Much error prevails among planters as to the causes of the shedding of the cotton crop, and that much of the blame has been misplaced is shown by the following tables. The data upon natural shedding (Table IV) and the natural or normal average number of bolls matured by a cotton plant under favorable conditions (Table V) were noted in fields entirely free from Boll Worm ravages. 12 Table IV. — Bolls in cotton plant — natural shedding. Plant. Bolls in plant. Shedding. Total. Natural. Other causes. 1 71 16 157 08 26 70 25 16 52 36 30 33 22 17 16 3 4 10 118 49 225 107 60 113 2..., 3 4 5 6 Total 408 68 192 32 72 12 672 112 Average per plant . . Table V. — Matured oolls in cotton plant. Plant. Bolls matured. Plant. ' Bolls matured. 1 25 16 18 26 25 22 22 17 9 12 17 22 31 33 2 10..! 3 11 4 12 13 5 G Total 7 286 22 8 Average per plant . . Table IV shows that when examined September 7 plant 3 had borne a total of 225 and plant 4 a total of 107 squares, forms, and bolls; Sep- tember 29 plant 3 bore only 98 bolls and forms, and plant 4, 59; Octo- ber 10 plant 3 had 96 and plant 4, 51; November 14, plant 3 bore 83 and plant 4, 44. At this time it was evident that the bolls yet remain- ing would mature and open if conditions continued favorable. Hence, by natural agencies the original numbers had been reduced to 83 and 44, respectively, by November 14. The shedding of plant 3 was, there- fore, 63.2 ; that of plant 4, 58.9 per cent. Even the figures given in Table IV on September 7 show that of a total of 672 bolls and squares originally borne by the six plants only 408 were then upon them, a shedding of 264 bolls, or 39.3 per cent. The plants examined were above the average in growth and vigor. These facts should impress the planter reporting damages supposed to be due to an insect with the necessity of making a careful examina- tion and discovering the real causes operating. If this be done, he will often find that much of his loss is'due to perfectly natural causes and not to insect depredations. The data contained in Table V are based upon average plants which had already matured, or at least had set their full crop. It must be noted in this connection that the average given applies to good cotton only, in such districts as northern Louisiana and Mississippi. In river bottoms, where cotton grows much more vigorously, the average per Xnant is proportionally greater. In some of the southern and central 13 portions of Texas, where cotton grows from 7 to 10 feel high, the aver- age may be twice that given in the table. These facts are recited to illustrate the necessity <>f making local studies for a given locality in order to arrive at anything like an accurate estimate of the injury and loss for that locality. Table VI. — Condition of cotton field at Mesqnite, Tex. Plant. Bored. 1 Other causes. Good bolls. Total. 1 15 2 4 3 5 4 ! 2 5 1 6 3 7 12 8 14 9 6 7 15 16 14 58 8 12 6 7 2 12 18 21 14 21 22 33 22 33 39 18 63 11 24 23 13 12 21 44 30 19 26 27 42 14 18 2 4 3 10 10 11 6 12 21 13 7 14 3 15 2 16 1 17 3 5 2 2 2 4 9 Total . . j 112 G9 286 467 Table VII.— Condition of cotton field at Arlington, Tex, Plant. I Bored. Other causes. Good bolls. Total. 1 7 2 1 3 1 4 1 5 1 6 12 7 8 3 9 9 10 8 11 1 12 7 13 •) 14 4 15 13 16 23 Total 9u 11 25 21 12 25 23 30 15 22 16 13 14 11 15 12 4:; 308 23 31 22 14 29 49 35 20 32 30 16 26 13 21 34 81 476 14 Table VIII. — Early and late cotton compared. LATE COTTON. Plant. Bored. Other causes. Good bolls. Total. 7 8 9 10 11 12 13 Total . . . 12 14 6 10 6 2.1 7 12 ] 12 18 21 24 23 13 12 21 44 30 3 3 5 2 76 13 78 167 EARLY COTTON. 14 3 2 14 19 15 2 3 21 26 16 1 4 22 27 17 Total . . . 0 9 33 42 6 18 90 114 Tables VI and VII exhibit data obtained while on a trip through the part of Texas which was worst infested by the Boll Worm. The figures for Table VI were noted at Mesquite, August 24, and for Table VII the facts were obtained at Arlington, August 27. In Table VI plants 1 to 6 inclusive were in the same field and stood consecutively in a row; plants 7 to 13 inclusive in a second field, consecutively as before ; while plants 14 to 17 inclusive were taken at random in a third field. Plants 7 to 13 were in a field of late cotton, still blooming profusely at the time of observation; 14 to 17 were in one of early cotton in which all the fruit had set and which, therefore, contained but few blossoms at that time. These data are compiled separately in Table VIII for the pur- pose of comparison. It presents some significant facts concerning the question of early and late cotton in Boll Worm districts. Thus, of a total of 167 bolls of the late cotton, 76, or 45.5 per cent, had been injured by the Boll Worm; of a total of 114 bolls of early cotton, only 6, or 5.2 per cent, were injured. Estimating the difference upon the basis of the normal average, we have the following result: The seven plants of late cotton averaged 11.14 bolls per plant, and the four plants of early cotton 22 per plant, 22 being the normal average arrived at m Table V. The late cotton therefore shows a loss of 50.6 per cent, while the early cotton shows no real loss. This may be taken as an extreme case, but the general principle remains that late cotton receives by far the greater portion of the Boll Worm attack, virtually protecting the early cotton fields about it. i: Table IX. — 8ked bolh found <>n ground. ; Table. Plants iuc .-lusive. Bored. Other causes. Total. 1-6 anrl 17 8 • 103 Ill VI 7-13 69 44 113 VI 14-16 10 42 52 VII 1-9 44 ■• VII 10-16 43 64 107 170 zr, 467 Table X. — Good bolls j>e>- plant. Source. Bored. Other Good bolls. Total. Table VI . Table VII Table IX . 112 17'i 69 75 o(l7 308 407 476 467 Total . . Per cent . . 375 441 594 20. 6 31. 3 42. 1 1.410 100 Table IX presents a study of the bolls and squares found shed and on the "round under the plants recorded in Tables VI and VII. The totals of Tables VI. VI L and IX are arranged for convenience in Table X. From the tacts thus presented it is found that 18 is the average number of good bolls per plant. The normal average has al- ready been given as 22. Hence the loss from injury is 18.1' per cent in- stead of 26.6, as found by tbe usual method. This difference is Largely due to having included the data of Table IX, which represents the shed bolls found on the ground under the plants examined. As has been shown by Table IV, many of these would have been shed by natural process, but were bored before having fallen. Hence, if in- cluded, they exaggerate the real damage. The actual damage should be estimated upon the basis of the average amount normally matured by the cotton plant in any given locality under favorable conditions. Omitting Table IX from Table X we have the following results: - lire. 15<>r»-.l. Other causes. TotaL Table VI Table VII .... Total ... Per cent 112 93 6'.» 75 308 4*7 20.') 144 J1.7 15.3 594 53 943 loo The above percentages are obtained upon the ba< of what was actually found upon the plant Augu>t 27, without reference to the num- ber of bolls normally matured per plant for that locality. The per cent of damage is shown to be 1*1.7. which, compared with the L&2 per cent obtained on the other basis, shows that this estimate is quite accurate and, for all practical purposes, satisfactory. 16 The damage detailed above was found only in the worse infested dis- tricts visited. In other localities the injury was much less, or none at all. Even in the infested districts some fields were found which had practically escaped injury. The estimated damage of 18.2 per cent applied more especially to a region approximately included by an imaginary line running from Paris, Texas, to Tyler, to Palestine, to Temple, to Greenville, to Paris. Other cotton-producing counties in Texas were much less affected, and for the entire State it will be safe to place the maximum limit of Boll Worm injury at 10 per cent, with probabilities that it is still less. Along the Red River and Mississippi valleys, and, in general, in the bottom lands along the smaller rivers and creeks, the injury is great- est. In the greater portion of Louisiana and Mississippi the damage is certainly not over 2 to 3 per cent. In Arkansas the damage along the rivers and in a belt across the State from Little Bock to Fort Smith was more serious, and for a considerable portion of the State ranged from 10 to 15 per cent. This is due in part to the greater acreage of corn in proportion to that of cotton. The reason for this lies in the fact that a greater number of individual ears are produced, and hence the probability of a greater number of worms reaching ma- turity. The relation existing between the acreage in corn and the acreage in cotton is no objection to the trap-corn method, to be subse- quently discussed, but rather makes it all the more advisable to use it at the proper time. In Alabama, Florida, Tennessee, Georgia, and the Carolinas the ravages are insignificant, and usually do not excite general attention. If the acreage and production of these States be included to ascertain the per cent of loss to the entire cotton crop from Boll Worm depredations, it is evident that the percentage will be reduced to a small figure. Those who have never spent a season among cotton-planters may consider this discussion of damage peculiar or even unnecessary. The fact is that the average observer, whether planter or newspaper re- porter, seldom comes to his conclusions upon a basis of what is found upon the plant, or after having considered natural causes of loss. He judges mostly by what he sees lying upon the ground, and to this, as has been shown, several causes contribute. Upon this basis (see Table IV) a damage of 39.2 per cent could be reported. Such reports are en- tirely misleading and erroneous, and have no foundation in fact. It is even more difficult to give an estimate of the damage to corn by Boll Worms. From Table I it is found that in May, 2.6 per cent of the young corn plants had been attacked. The plants were not ruined nor even checked in their growth, and ultimately produced sound ears of corn. The conditions presented in Table III are quite disgusting when viewing the ravaged tassels, but in the end the ear of corn is produced. Romantic discussions of these facts have been entirely mis- leading, and for corn it is safe to assert that no real damage is occa- 17 sioned, so far as the ultimate yi<*]pearance in numbers from about June 1. By the middle and latter part of June numerous young can be found. The second brood begins depositing about the latter part of July. After the first days in August the adults are not so abundant until the young begin maturing again. The male adults are easily at- 33 tracted to lights, while the female is rarely caught in fchia way. Of twenty-two specimens trapped twenty were males and two females. At a lamp experiment July n> nine specimens were captured^]ft] males. Mr. Banks often collected adults at random, and without regard to from cotton plants during the day. Eight specimens were taken oil one trip, four males and four females. July 15 fifteen were captured, fourteen being iemales and one mule. A third capture was found to contain six females and two females. This shows that females were plentiful in cotton fields at the time the lamp experiment had been made. but were not attracted. The damage to cotton by this species is due in great measure to the immature forms of the insect. It appears that during July and August cotton fields surrounded by poplar growths along the bayous suffer the greatest attack. This is to be expected, since during June the insect lives mostly upon these t: the young growth of which becomes too hard and tough later in the season. As has been stated, it is most numerous along river bottoms and bayous. Away from these regions this species is not at all com- mon in cotton fields. In the upland regions of Texas, where continuous observations for one week in August were made, not a single specimen was found upon cotton. 3Ir. Banks, who took an extensive trip through central and southwestern Texas during July, reports the rare occur- rence of this insect in those regions. Young poplar is probably their choice for food and egg deposition, but they are often found upon vari- ous kinds of weeds and miscellaneous plants. This being the case, the question of a remedy becomes a difficult one. The only recourse which seems at all practical is to control the number of young poplar trees along the bayous, keeping them at a minimum so as to obtain the maximum number of insects upon them. Then about the middle or latter part of June give them a thorough application of a strong solution of kerosene emulsion. This would kill many of the adults ami most of the young, which are abundant upon them at this time. Another nearly related species, Proconia inxl<(t<<, mostly found upon willow, is occasionally noted upon cotton. Whether its injury is similar to that of the Homalodisca has not been positively determined, but the facts already noted for the latter indicate that it may be. REMEDIES FOR THE BOLL WORM. LIGHTS FOB ATTRACTING THE MOTH. The experiments presented in Bulletin 24 (pp. 33-38] proved con- clusively that the ordinary lamps used by farmers and the methods of using them were inefficient. Until proven otherwise, the reasons .i^ Signed for such results were that the lights were not brilliant enough, together with being unprovided with extending wings as a background against which insects flying near by might strike and be trapped. A lamp was devised to meet all these requirements so that it could be ef- U935— No. 29 — -3 34 licient if the nature of the case permitted. The following is a descrip- tion of the lamp: A tin can, 6 inches in diameter, holding about a half gallon of oil, was provided with a No. 2 wick burner capable of receiv- ing and holding a large chimney. Around the tin can was fitted a movable tin band to which had been soldered four stout upright wires. To these wires were fastened sheets of tin a foot square, extending at right angles and from the top of the can. These wings, together with the 6-inch space between for the lamp, furnished a surface of 2£ feet toward any direction against which insects flying near might strike and drop into the large pan in which the lamp was placed. The lamp is not easily blown out of the pan if the precaution is taken to nave the wings extend out far enough to catch the rim of the pan. Though a chimney was always used, in no instance was the light blown out when a strong breeze was prevailing. By experiment this lamp was found to emit a brilliant light, which was not hindered in its transmis- sion by the extended wings. Experiments were begun as early as May 13, 1891, and repeated at intervals on through the season. The results were all so uniform that only a few experiments need be reported in detail. Experiment 2. May 15. — Lighted at 8 : 30 p. in. Sweet corn in a garden. Locality, upland in edge of timber. During the afteruoonof that day many recently deposited Heliothis eggs were found on the corn silks. 8:40. — Boll- worm moth flying along a row of corn next to the one in which the lamp is placed. When opposite the lamp, only 3 feet distant, it flew at right angles away from it. Lamp was on a level with the ears of corn on the plants and could he seen over the entire patch. 9:00. — Another moth flying as before came near, but flew away without showing any attraction to the light. 9 : 06, — One passed the lamp, turned, came near again, alighted upon the edge of the pan and sat there. In attempting to fly away it struck one of the tin wings and dropped into the pan. No more moths being observed, the experiment was closed at 9:40. Though the moths were not abundant several females were seen de- positing upon the fresh corn silks. To convey a general idea of the nature of the various trappings aside from the primary insect desired, experiments 3, 4, and 5 have been col- lated and are presented in Table XIII. 35 Table XIII. — Number and kind of insects caught. LEPIDOPl EB v J. , Date. Noctnidae. i ii ometridae. I'-, ralidae. TortricidflB. Tineidjf. Total 7 1 Spe- cies. Speci- niens. Sp< meus. •1 S|n-i 168. Sped mens. S,.e- <].(■( i mens. 3 1 ■■', 12 1 5S 2 9 Total Botia . . 15 Plutella 1 5 4 | 27 1 5 3 ltf 53 June 12 4 2 7 1 >\ 1" Plutella Total 2 7 1 4 4 25 36 Jnne 27 Total 5 •) 5 3 ■I 3 Botis.. 3 30 1 4 \ | 5 15 2 5 3 6 4 33 5 8 5 » 87 COLEQPTERA. 1 • . 1 || Date- CSb' II Bcmtaita. Wa v.- 2 - 45^ i. x " ai!2 x J; Cnccinel- lidae. „ ! T „ nn ( Lachnosterna. . 15 3 June9 30 \ Cvclocephala-35 , r „,„„ c c Pelidnota 1 4 June 12 v .' „-n 1.000 s Lachnosterna -.30 J 50 7 6 9 fc8 7 5 4 7 1. 104 , Small ,o ) Cyclocephala 40 5 t„„^-)- < Large 12 Cftn ( Lachnosterna. .55 . June., ..j SmaU15J, niu j Cyclocephala.. 75 i Megilla 2 1. 172 HETEKOPTERA. ExPeri" Pate ('v'1 Sto-" ment. ^ nidV ^ i LygaeidsB. ridse Acanthiidse. Coriside. 5 3 June 0 2 ... Consul W 62 1 June 12 5 June 27. . 30 \ 1 Melanocoryphns Ifl Melauucoryplius-.IOO G Tripbleps . .4 Corisa 15 Corisa ..l.ooo 13 1,133 1 IIOMMPTERA. Experi- ment. Date. Jassid;e. Fulgoridse. Memhracidae. 3 June 9 . . Tunc 12 . June 27 . Aulacisea Aulacises Deltocephalus 20 35 .15 .20 Ormenis . . .11 17 Ormenis - . . 10 37 4 13 36 Table XIII. -Number and kind of insects caught — Continued. DII'TERA. Experi- ment. Date. Mosquitoes. Tipulidae. Mycetoplrilidae. Totals. 3 1,000 30 200 15 12 11 13 100 1,026 4 5 June 12 . June 27 . ~>5 300 NEI'ROPTEEA. Experi- ment. 1 Date. Caddice flies. Chrysopa. May flies. Totals. 3 June 0 .. 20 5 7 10 G 30 4 i June 12 . June 27 . 11 5 1 g OKTHOPTERA. Experi- ment. Date. (Ecanthus. Nemobius. Platamodes. Totals. 3 4 5 June 9 3 3 4 4 1 5 Very few parasitic or beneficial Hymenoptera were trapped at any time and hence this order is omitted. Table XIII can best be reviewed by taking up the orders seriatim. Lepidoptera (Moths). — The only species of any considerable economic importance in the South is the Cabbage Plutella (Plutella cruciferarum). Experiments 3 and 4 were both located near a gardener's cabbage field. Its significance in this connection lies in the suggestion that gardeners growing cabbages extensively and troubled with this pest might resort to lam]) trapping with advantage. Coleoptera (Beetles). — Some of the large and well-known predaceous beetles were captured together with hundreds of many of the smaller species. Of the beneficial ladybird family a few specimens were trapped at various intervals. But this loss of beneficial insects is in part counterbalanced by the capture of several injurious species none of which, however, except the white grnb beetles, Lachnosterna longitar- sus and Cyclocephala immaculata were caught in great numbers. The last two species and a species of the wireworm beetles Monocrepidius vespertinus were caught by hundreds and may be considered a profit- able catch. Several species of injurious weevils and flea beetles were commonly trapped though not in great numbers. Following is a se- lected list of some of the beneficial and injurious beetles which were quite constantly trapped during the progress of the experiments. Xone of these were captured in great numbers. For the determina- tions of the species I am indebted to Mr. E. A. Schwarz of this Division. 37 HK.\ I I li I A I. Predaceous beetles: Loxandrua agilis. Badister micane. Chlwnius laticollis. CMamius penwtylranicna. Cratacanthus ditbius. Stenolophjis dissimilis. Brady evil ua nigricepa. Megilla maculate. Hippoda m ia <-,,„ rergens, Coeeinella 9-punciata. Coccivella oculata. Mysia pullata. Exochom us margin ipen n-is. Scymnus cervicalis. Flea-beetles : Sy8tetta elongate. Epitrix fuscula. dm turn, ikii pulicaria, Haiti ca i f/n Ha. Phyllotreta bipnstnlata. Wire-worm beetles: G lyp k on yx inqui n a tus. Monocrepiditi8 vespertin ns. Monocrepiditi8 lividus. INJTTJtlOl 8. Miscellaneous : Lachno8terna longitersus, Cyclocephala immaculate. Calandra oryzat. Typophorm canellus. Li n a script a. Diabroticd 1, '-punctata. Balaninus cargo*. Myochrous dentioollU. Colaspis flavida. Heteroptera True Bugs). — Only one species of known beneficial im- portance is noted. It is the small Triphleps insidio&us which punctures boll-worm eggs, hi other experiments not tabulated an occasional soldier-bug was caught usually Podisus spinosus. In some of the ex- periments an insect (Galacoris rapidus) which contributes much to what is popularly termed *• sharpshooter" damage was trapped in small num- bers. A probably injurious cotton insect which the planters often mis- took for the genuine Cotton Stumer (Dysdcrcus suturellus) is Melanocory- phus bicrucis. This insect was trapped by hundreds but subsequent stndy proved that fully DO per cent were males. Homoptera (Leaf-hoppers^ etc.). — Homtdotlisca coagulata was caught in great abundance. Subsequent study showed that about 90 per cent were males. In the three orders, Diptera, Xeuroptera, and Orthoptera, nothing worthy of consideration was captured except a tew specimens of the beneficial lace-wing tiies. Experiment 6. Arlington, Tex.. August .'7. — Lighted at 7 p. m. and placed between rows of cow- peas adjoining a cotton field. The rows of cowpeas were o t<> 8 feel apart ami had many Boll Worm moths flying about them feeding. The weather was warm ami pleasant, the nighl very dark. Being placed between the rows, a distance of only about 4 feel remained from which to attract the passing moths. For an hour the moths kept flying up and down the rows on either side of the lamp, fed freely, de- posited eggs, and paid no attention whatever to the Light. A volunteer pea vine w as near the center of the row having a tew branches extending \\ ell up projecting over the edge of the pan within 10 inches of the flaring light. Some fresh Mossoms upon them proved attractive, and a tew adventurous females visited them, sipped of 38 their sweets tor a time by lamplight and then flew away to continue their usual vocation. This act of defiance sent consternation to the hearts of some 15 or 20 planters who had been invited to attend the experiment and who during the day had insisted that if properly conducted, lights were effective agencies. All ad- mitted that the test had been made under the most auspicious circumstances and yielded their former position with commendable grace and sincerity. Their atten- tion was further called to a number of parasitic Hymenoptera which had been caught, some beneficial and predaceous beetles, soldier bugs, lace-wing flies, and many other species of little known economic importance such as have heretofore been given in detail and need not be repeated. To summarize briefly, it must be concluded that the use of lights for attracting and trapping the Boll Worm moth is entirely useless. The character and habits of the other insects caught, as shown by Table XIII and its discussion, are found to be pretty evenly divided between those which are beneficial and those considered injurious. Most of the insects noted as injurious are not of special economic importance throughout the cotton region, and hence their consideration in this connection may be justly omitted. The use of lights, so far as the cot- ton planter is concerned, results only in the destruction of beneficial insects and is, therefore, an absolute disadvantage. Such being the case money expended in this practice is an entire loss. As a protective agency lights are a failure and should be unhesitatingly discouraged and condemned. POISONED SWEETS. Much has been claimed for this method of destroying the moths and a number of experiments were made to test the value and importance of the remedy. The various mixtures were applied with a Woodasou liquid sprayer upon rows of cowpeas which had made a rank growth and were blooming profusely. They were freely visited by Heliothis from about 4 p. m. until 8 or 0 at night. All conditions for the experi- ments were favorable and furnished a good test of the poisons. The experiments were made upon Mr. C. F. Mercer's farm at Arlington, Tex., where Dr. L. C. Page, of that city, also rendered valuable assist ance. By direction Dr. Page prepared saturated aqueous solutions of the poisons, and mixtures of desirable strengths with vinegar or beer were made subsequently. Experiment 1. August 27 (4:15). — Beer, 8 ounces; saturated cold-water solution of arsenic, 4 ounces. August 28 (3 p. m.). — Leaves, blossoms, or young pods slightly or uncertainly in- jured. Experiment 2. August 27 (4:25). — Beer, 4 ounces, with 4 ounces of the same poison solution used in experiment 1. August 28 ( }). m.). — Foliage, blossoms, and very young pods badly scorched. 39 Experiment 3. August 27 (4:35). — Vinegar, 4 ounces; 3 ounces saturated arsenic solution. August 2S (p.m.). — Foliage, blossoms, and an occasional young pod badly scorched. August 27. — Dr. Page was directed to prepare the following solutions: 1 ) Saturated cold-water solution of commercial arsenic. (2) 1 ounce corrosive sublimate to 1 pint cold water. (3) 1 ounce potassium cyanide to 1 pint cold water. Samples of each solution were kept, taken to Sbreveport, and tested. They had been perfectly prepared, and the poisons were therefore actually in solution at the time of application. The following two mixtures were prepared and used to dilute the poisoned solu- tions in experiments 4 to 6. inclusive: (1) 3 pints beer to 1 pint molasses. (2) 3 pints vinegar to 1 pint molasses. Upon leaving Arlington, on the night of August 28, Mr. C. F. Mercer, of that city, was requested to make notes upon the damage done to the foliage by the several solutions in experiments 4 to 6, inclusive. These notes were submitted b> him in a letter September 1, and the facts contained are includedwith their respective experi- ments. Experiment 4. ttujnst 28 (4:15). — Beer, 8 ounces; cold-water solution commercial arsenic, 4 ounces. 4 >»ji'si 29. — Foliage scorched. Experiment 5. August 2S (4:45). — Beer. 4 ounces to 2 ounces potassium cyanide solution. August 29. — Foliage shows no signs of damage. August 30 — No damage to pea vines indicated yet. Experiment 6. August 28 (4:55). — Beer, 4 ounces to 2 ounces corrosive sublimate solution. August 29. — Foliage wilting. August 30. — Dead and badly damaged. Notes taken during the progress of the experiments show that re- cently issued females or those just beginning to deposit do, in fact must, meet with the poisoned liquid on the vines. Soon the moths began to alight upon the leaves or pea pods and sip of the drops of sweets to the practical neglect of the blossoms. After sipping the moths became somewhat uncertain in their flight and soon Hew away and hid. It was evident to anyone familiar with their flight that the moths were affected and it was only a question of a short time when death would occur. Intact the day following the firsl three experi- ments dead moths could be found here and there when the pea vines were raised from the ground. The specimens were not old or worn out individuals and their death was evidently attributable to tin' poisoned liquid which they had sipped from the vines the evening before. The practicability of this remedy is somewhat lessened by the fact that the poisoned mixture dries rather quickly. To attain the best results it must be applied each day for a time during the egg-laying 40 period. This objection is valid only to a certain extent as will be noted later. The remedy is certain to be effective if properly man- aged. Where Boll Worm ravages are very great the additional expense and application npon a minimum area of trap-planted peas becomes proportionately a matter of secondary consideration. The crop which can be most easily and successfully managed for this purpose is that of cowpeas planted in rows 6 or 8 feet apart as a trap bordering the cotton field. They should be planted late so as not to reach the height of their blooming period before the destructive August brood of moths appears. The area should be the minimum and will depend largely upon the size of the cotton field to be protected. The blooming pea vines attract the issuing moths for feeding purposes provided the cotton be early enough to have passed its attractive blooming period. It be- comes important, therefore, that the cotton be as early as possible. As will be seen from the experiments, the difficulty arises that even moderately weak solutions of the poisons scorch the pea vines if the weather be hot and sunshiny. This scorching at once brings to an ab- rupt end the utility of these plants as a trap crop. This result can be obviated by making the applications as weak as is advisable to insure death to the moths and then only applying it to portions of a row upon any one evening. This leaves unsprayed healthy portions for a series of evenings to follow. Applications should be made to only a portion of each row at any given time, since observation has shown that a moth once starting in a certain row, if undisturbed, is inclined to follow it up or down for some distance. The chances of poisoning are, there- fore, greater than were only certain of the rows sprayed and others not at all. In experiments 1 and 4 the same strength of the arsenical solutions was used. In the former the foliage was but slightly injured, in the latter, badly scorched. This is due to the arsenic for experiment 1 having been placed in cold water for about six hours before using, while in experiment 4 it was in cold water for twenty-four hours pre- vious. Hence a greater per cent of arsenic had been dissolved in the latter. A poisoned mixture of arsenic prepared as in experiment 1 and applied while fresh in the proportion of 12 parts of the vinegar solution to 4 of the poisoned liquid will be efficient and yet not injure the vines. From experiment (> it will be noted that the corrosive sublimate mix tare of the same strength as those of experiments 1 and 4 was less im- mediate in its effects. If the dilutions were carried to the same extent as just advised for the arsenic it could doubtless be used with safety and good results. The experiment with a preparation of potassium cyanide, designated as No. 5, shows that the solution did no appreciable injury to the plants. Since it is a swift poison for insects, its use is undoubtedly effective. There could be no hesitation in concluding from the experiments that preference should be given to the cyanide prepa- ration and its use in. the proportion given in the trial recommended were it not for the fact that it was lately determined that there was a 41 question as to the quality of the substance used. The tot of the prep- aration at Shreveport utter the experiments had been made proved beyond question that some eyanide was in solution, but no qualitative test could be made to determine the probable quality of the article used. There seems to be little, if any choice in the use of beer or vinegar with the molasses. Vinegar and molasses are probably more easily ob- tainable in the country districts, and hence are the cheapest. Fruit vinegar should be used, and a mixture of 4 parts to 1 of molasses is quite as effective as the ones used in the experiments. For the application a tine spray is not necessary, as it is preferable that the liquid should be formed in large drops on the plants. Any of the larger spraying' machines in use provided with a coarse nozzle can be used for the purpose. Plates of the poisoned liquids were left standing upon short pedestals among the pea vines, but the moths failed entirely to visit them. Stakes which had been set among the vines were sprayed to excess, but formed no attraction. In fact, anyone who has closely observed the feeding habits of the moth can have no hope for the efficiency of any remedy except an actual application upon the food plants themselves. The usual methods of utilizing poisoned sweets against this pest are evi- dently useless and involve expenditures of time and money which are practically an entire loss. This conclusion is based upon the behavior of the moths toward the sweets during the egg-laying period. That time over, many individuals may be caught, but then their capture has no real economic significance. Some advise cutting into halves numbers of ripened melons in patches adjoining cotton fields and saturating the cut surface with poisoned liquids such as have been mentioned. While at Arlington, Tex., a union patch was found between rows of pea vines and a large cotton field. During the day it was found that where melons had been broken open and left Lying during a hot day, Boll Worm moths visited them in the afternoon from about 3 o'clock. The moths unquestionably t\n\ upon the exudations; but the practice is objectionable, since during the day it had been noted that scores of the preying wasps constantly flying about cotton fields, honey bees, and some miscellaneous beneficial in- sects made visits to the broken melons. All of these would necessarily be poisoned and would be a direct loss. To a certain extent the same objection can be maintained against liquids applied to eowpeas. On these plants, however, the poisoned sweet is not applied until after the heat of the day, when beneficial insects are flying about less plentifully. Furthermore, the application dries the next day as soon as the dew of the night evaporates, which greatly lessens the danger of destroying desirable insects. The drying of the poisoned application is. therefore. in one sense an advantage, as it partially counterbalances the loss in efficacy of the application. 42 EXPERIMENTS WITH PYRETHRUM. Simple aqueous decoctions, as reported in Bulletin 24 (pp. 39-44), having proven a signal failure, it was thought advisable to experiment with some of the oils as agents for drawing out the insecticidal element. Headlight oil was selected, for the reason that the quality obtained from country dealers is much more constant and reliable and hence bet- ter for a series of experiments. Comparative tests of the power of ex- traction of the oil by various methods were made, as also of the oil combined in an emulsion with other than oil extracts. Asa check upon the pyrethrum emulsions the simple oil emulsion was used in several ex- periments, in order that the effect of the oil in the combination might be known and any additional advantage of the second factor rendered capable of more definite determination. Simple Emulsion. Method of preparation. — Oil 2 parts, water 1 part, and enough soap to emulsify well. Water heated and oil added while the water boiled. Churned until the mixture thickened. Prepared October8, and is yet in perfect condition November 10. Used in experiments 1, 2, 3, aud 4. As Boll Worms were scarce, the larvae of the Cotton Worm (Jletia xylina) were used in all pyrethrum experiments. Experiment 1. October 10 {12:35). — A 4 per cent water dilution was made and sprayed upon larvae on cotton plants in the held. The larva, seventeen in number, were taken from the sprayed plants and placed upon fresh unsprayed leaves in a box, later being placed upon fresh food in breeding cages. This method was followed in all subsequent ex- periments. The sprayed branches in the field were always appropriately marked, in order that the effect of the emulsions upon the foliage might be noted at any time. Date. Living. Dead. October 12 Total 2 half grown. Hi grown 15 2 Experiment 2. October 10 (12:30). — A 6 per cent dilution was sprayed upon 32 larvas. At 5 p. m. it was noticeable that the younger worms were somewhat affected, but the larger ones showed no uneasiness. Date. Living. Dead. October 12 Total 2 half grown. lu very young. 20 12 43 The living larvae less active than those in experiment 1. The foliage in experiments 1 and 2 was examined October 12 and November 28 and found uninjured. The emul- sion did not seem to render the foliage distasteful, for young larva- were subse- quently found feeding upon it with a relish. Experiment 3. October 24 (11: l'>). — A 13 per cent dilution sprayed upon 12 larva?; all nearly grown. At 4:43 1 seems slightly affected, others active. October 26. — All active and have fed freely; two have webbed. October 29. — Two larvae feeding vigorously; 1 webbed and .'! pupated. October SI. — Webbed larva} all pupated; 1 not perfectly formed. November 10. — Five imagos have issued. The imperfect pupa is dead, as also 4 others, which do not seem to have been normally formed, due probably to the effect of the emulsion by inducing premature pupation. Foliage slightly injured. Experiment 4. October 30 (4:40). — A 19 per cent dilution used upon 10 larvae. October 31 (9:30 a. m.). — Three larvae badly affected; rest active and feeding. November 2. — Six are badly affected and will probably die; others feeding. November 3. — Six are dead, 2 pupated normally, and 2 are attempting to do so. November 4. — Last two have pupated, but only about half the normal size. November 17. — Two pupae are dead; one imago has issued. December 16. — Remaining pupa produced an imago. Foliage examined November 10 and found badly scorched. Pyrethkum Emulsions. cold-water decoctions. Method of preparing first Emulsion. — To one pint of cold water one-fourth ounce of pyrethrum was added, well mixed and left to stand over night in a sealed Mason jar at a temperature of 66° F. This was done at 4 :30, October (3. Filtered on the morn- ing of October 7. Of the resulting filtrate one part was emulsified with two of head- light oil and soap as before and left to stand in a sealed Mason jar. This is the emul- sion used in experiments 5 and 6. It is worthy of note that on October 10 the simple water decoction which was perfectly clear when filtered had undergone some chemical change — fermentation probably. It became very turbid, offensive in smell, and e\ idently unlit for further use. On the other hand, the emulsion was still perfect a month later. Experiment 5. October W(1:S5). — Four per cent dilution. Number of larva' sprayed, 16. October 12. — Both large and small active and feeding. One, about half grown, dead. Experiment G. October 10 (1:20). — Seven per cent dilution. Number of larv», 21. Date. Living. Dead. October 12 1 papa !» half gron a. '.' ball grown Total 12 i) 44 Foliage in experiments 5 and 6 uninjured. The second emulsion was prepared as the first experiment, except that the propor- tions were 3 ounces of pyrethrum to 11 pints rain water. This is the emulsion nsed in experiments 7, 8, and 9. Experiment 7. October 10 (12 m.). — Four per cent dilution. Number of larvae, 16. Date. Living. Dead. October 12 Total 4 half grown. 2 Aery young. 10 6 Experiment S. October 10 (11:49 a. m.). — Six and one-half per cent dilution. Number of larvse, 19. Experiment 9. October 24 (11 :55 a. m.) — Thirteen per cent dilution. Number of larvae, 10. At 4 :37 p. m. 1 larva had webbed, but was badly affected. The other 9 were active and feeding. October 26. — One pupa, 8 active and feeding; 1 dead, half grown. October 27. — One more webbed. October 29. — One more pupa, 4 webbed, and 3 feeding. November 20. — All but one pupa which was imperfectly formed, have produced imagos. The imperfect pupa is dead. The foliage in experiments 7, 8, and 9 was uninjured. HOT-WATER DECOCTION. Three ounces of pyrethrum were added to \\ pints rain water, placed in a sealed Mason jar, and boiled for one hour. Filtered and emulsified a portion of the filtrate with headlight oil. This is the emulsion used in experiments 10, 11, and 12. Experiment 10. October 10 (It : 25). — Four per cent dilution. Number of larvae, 19. Date. Living. Dead. October 12 Total 2 half grown 2 very young 15 4 45 Experiment 11. October 10 {11:05). — Six per cent dilution. Number of l;irv;p, 10. Date. Living. October 12 Total m o Experiment 17. October 24 (12:05 p. m.). — Thirteen per cent dilution. Nnmbci of larva, 0. At 4 :•">:-; Larvae still active and apparently unaffected. October 26. — Have fed freely; 3 webbed. October 29. — One feeding; 5 webbed; 3pnpas. November 13. — Two pupated imperfectly and died; others have issued. COLD-OIL DECOCTION. One and one-half ounces pyrethrum added to one-half pint headlight oil placed in iled Mason jar and left over nighl at a temperature of 68 F. Filtered the next morning and emulsified the filtrate with half as much rain water. This emulsion was used in experiment- 13, 11. and 15. Exp< rimeni 13. 1 1 ■,■,.. io | f:05 p. n> . ). — Four per cent solution. O ■/ tber I.'. — Three nearly grown larvae lively: 5 dead, all about half grown. This bree ling cage, as also the one of experiment 14, was found to have cracks in, which had 1)^' !U nunotice 1. au I many of the larva' escaped. Expo') mi ni 11. October 10 1 12:55). — Seven per cent dilution. Late in the eveniug the larvse ap- peared somewhat uneasj . October 12. — Two half-grown ones may live; 11 half-grown ones are dead. Erperimi ni 15, October 24 (tt:25). — Thirteen per cent solution. Number of larvae, 10. At 1:25, 1 half-grown larva? are unable to crawl ; 2, about a third grown,in the Bame condition : 1 nearly grown ones can travel about, though their actions are not perfectly normal. October 26. — Two trying to web up; 1 larva feeding, and 7 dead. Of the dead, 5 are half grown, the other 2 yonnger. October 29. — A F> »11 Worm in the cage attacked and devoured one of the webbed- up larvae : the secon 1 one pupate I, and the third died in the attempt. November 17. — Pnpa has prodnced an imago. Inexptfrimoiitsl3a.nl 11 the foliage remaiued unimpaired, but in experiment 13 it w a> Blighl lv SCO! but faiutly recognizabh . Experiment 21 was made for the purpose of determining any difference in the effect of the changed or fermented decoction and the fresh nitrate. Ezpi riment 21. Ortoher 24 ( 12: ?o).— Full strength applied. At 5 p. m. all the larva1, 8 in number, lively. October 96. — All well and active. 2 having webbed. October 29. — One feeding vigorously, 4 webbed, and 3 pupa?. Evidently no results. and experiment closed. SIMPLE HOT-WATEB DECOCTIOIT. Three ounces pyrethrum to H pints rain water, boiled for one hour in a sealed Mason jar. After boiling, filtered and kept filtrate in sealed Mason jar. This decoc- tion was prepared October 8, and used in expei iments 22, 23, and 24. Experiment 22. October 8 (4:20). — Full strength sprayed npon 49 larvae. The very young began tumbling off in a few minutes. By 5 p. m. many of the newly hatched larvae were evidently dying. Pate. Living. l>ea .= = t>5 ^| fe T3 g 5 i M > fcf- x - P ? 3 r 4 per cent 6 per cent 13 per cent 19A per cent. . . 4 per cent 7 per cent 4 per cent fij, per cent 13 per cent. ... 4 per cent. . . ., 6 per cent 13 per cent. . . . 4 per cent 7 per cent 1;: per cent 4A per cent. . . . 13 per cent. . . . 21 per cent Full Full Full lllT 2 32 3 j 12 4 : io 5 j 16 6 1 21 7 i 16 8 ! 19 9 1 10 10 : 19 11 ' 40 12 9 13 (*) 14 (*) 15 10 16 ; 9 17 1 6 IS 10 19 37 ii 4 7 2 7 3 1 (5 8 4 8 7 3 -2 2 4 ii 2 2 io 15 20 2 15 12 10 12 8 15 31 7 2 i 4 1° 5 2 5 i 6 8 i Pyretlirum emulsion, cold-water de- \ 8 9 4 6 1 9 4 5 1 "2 "3 1 5 9 6 7 { 1 0 Fyrethrum emulsion, hot-water de- 'coction. { Fyrethrum emulsion, cold-oil ex- j 9 22 \ 4 6 4 ') 2 ft 2 :::: — .... 1 1 "2 3 7 ? 1 2 3 8 2 i tract. Fyrethrum emulsion, hot-oil extract < Cold-water decoction of pyretlirum ? Hot- water decoction of pyretlirum . 1 Check experiments < 0 2 ? 6 3 S "i 14 15' 8 18 10 99 21 8 t8 22 49 3 10 5 7 24 31 Full 24 I (*) 25 ! 16 Cold water Picked larvae - ^ 1 5 3 2 7 4 6 8 10 26 i 14 3 6 1 * Xot counted ; see record of experiments in the text, t See record of experiment in the text, 49 Table XV. — ExperimenU with different strengths of pyrcthrum* Totals. 6 per cent . - 6 per cent . . cenl - 7 per cent . . 7 per cent .. Totals Totals Totals. Survived. Dead. "-«*■ 5S u* 4 per cent - 4 )).-r cent . 4 per cent . 44 per cent 13 per cent 13 per cent 13 per « - e 1 1 1 13 per cent 13 per cent (*) 25 ') 32 4 11 40 8 - 19 <; 6 14 21 (*) •' : 2 ]_' 12 . 24 19} per cent 21 per cent 18 * Xot counted ; see record of experiment in tf this experiment not included in the totals. SUMMARY OF THE EXPERIMENTS. When studying the above tabulated results it must be constantly borne in mind that the larvae of Boll and Cotton Worms resist the ordi- nary liquid insecticides of such strengths as arc usually effective agaiust other insects, such as bugs or Leaf-hoppers. Another important fact to notice is that whatever effect was obtained from a certain solution or decoction is to be attributed solely to it, since the larvae were transferred to cages in the shade away from the direct sunlight. The assistance of direct sunlight in producing scorching effects with the oil emulsions is entirely eliminated, and explains why the larvae seem to have with- stood unusually strong solutions. For this reason the results obtained, though possibly less striking, have greater significance as to the real value as insecticides of the combinations made. The foliage in the field was injured less than might be expected with such strong solutions on account of the cool, dewy nights and moderate temperatures during the day at the time when the experiments were made. It is needless to dwell further upon these conditions, except to state thai the same strengths of emulsions if applied during the heat of day in midsummer would affect both larva- and foliage proportionately in a more decided and vigorous manner. This, however, has no direct bearing upon the primary purpose of the experiments, which was to 1 1935— No. 29 4 50 discover some easy and practical method of obtaining an extract of pyrethrum, which really added some insectieidal property to the remedy with which it was combined. For this reason in the oil experiments it was manifestly necessary to eliminate the factor of direct sunlight. In order, however, that this series might be complete in itself a few experiments with cold and hot water decoctions of pyrethrum were repeated. Their results are presented in experiments 19 to 24, inclu- sive. Comparing these with check experiments 25 and 26 it becomes evident that neither cold nor hot aqueous extracts have any value as remedies against the more mature larvae, and have but slight utility even against the youuger worms. This agrees with what has already been reported in Bulletin 24, p. 43. Eesults to be of great value in making comparative tests of the remedies should on the whole be ob- tained by experimenting with older individuals. In the experiments not already discussed considerable selection was exercised in this re- spect. The aqueous decoctions of the powder having proven of no value against the more mature larvae, we should expect to find that the re- sults of these experiments with the oil emulsions combined with these aqueous decoctions would not differ materially from those of the sim- ple oil emulsions of equal strengths. Inspecting Table XVI it is found that experiments 1 to 4, inclusive, were with simple oil emulsions; those of experiments 5 to 12, inclusive, were the same combined with cold and hot decoctions of pyrethrum. In Table XV equal strengths have been tabu- lated. Noting in this table the experiments just referred to, no appre- ciable difference is found in comparing experiments 1 with 5, 2 with 0, 8, or 11, 3 with 9 or 12. For a series of independent trials the vari- ation in results is but slight, and the combinations in question seem, therefore, to have no special advantage over the simple emulsion. Studying next the cold oil-extract emulsions by comjjaring experi- ment 15 with 3, 9, or 12, which latter are simple oil emulsions of equal strengths, some difference favorable to the oil extract is shown. The difference can not be fully discussed, since, by an accident, the records of two of the experiments are not complete. It was observed, however, that the activity of the larvae treated with the oil-extract emulsion was more excited and pronounced than that of those treated with the sim- ple emulsions. Coming now to the hot oil-extract emulsions, we find some remarka- ble results. For example, in experiment 16, where a 4£ per cent dilu- tion of this emulsion was used, it is found that grown larvae were affected to an extent almost equal to a 13 per cent solution of the simple emul- sion. Again, in experiments It) and 17, Table XV, it is found that when 13 per cent solutions of the hot and cold oil extracts were applied to grown larvae, results favorable to the emulsified hot oil extract followed, the latter killing every larva used in the experiment. The hot oil ex- tract having greatly increased the efficacy of the emulsion, it is to be expected that the cold oil will add to itself, in a less degree and more 51 slowly, a portion of the active principle of the pyrethrum. The slight advantage of the emulsified cold oil extract over the simple emulsion as already indicated is, therefore, corroborated by the decided advan- tage of the emulsified hot oil extract preparation. The effect of pyrethrum upon larva' is to throw them into convul- sions or paralyze the muscles so that they have no power to direct their movements. None of the emulsified extracts applied to the larvae pro- duced such effects until we come to the emulsified cold oil applications. In these, the characteristic effects are rather uncertainly indicated in the stronger applications. With the hot oil-extract emulsion such ac- tions wrere already manifested in the weaker 4£ per cent dilution, and very decidedly in the stronger applications. For example, in experi- ments 17 and 18, fifteen minutes after the application the full-grown larvae had utterly lost control of themselves, and it merely became a process of dying from that time. No chance for pupation, as in some of the other experiments. These facts show that there was really an additional insecticidal effect acquired by the hot oil decoction process, the extract of which was subsequently emulsified. ADVANTAGES OF THE EMULSIFIED HOT-OIL EXTRACT OF PYRETHRUM. The experiments above summarized again prove that the ordinary methods of extracting the active principle of pyrethrum are question- able, or at least unsatisfactory. The hot oil experiments show con- clusively that this method does to some extent draw out the insecticidal element of the powder, and retains it in the emulsion. However, its use upon host plants which are able to resist without injury an oil emulsion application of sufficient strength to destroy the insect is more expeditious and, perhaps, more economical than the use of the pyre- thrum emulsion. But plants which are injured by such an emulsion can be successfully treated with a weaker solution of the pyrethrum emul- sion, not injuring the foliage, and destroying the pest as effectually. This is shown by Table XIV, where, with a 4.] per cent pyrethrum emulsion in experiment 16, we have practically the same effect upon the growing larvae that a 13 per cent oil emulsion has in experiments 3, W or 12. The two latter can be regarded as purely oil emulsions, since it has been shown that the aqueous decoctions of the powder really contained no insecticidal properties. During high temperatures and bright sunshine it is well known that more or less danger of injury fo the plant is risked by the use of an oil emulsion when the strength which must be applied comes very near the maximum which the foliage will hear. This risk can be greatly lessened by using the pyrethrum emulsion, because the maximum strength which the plant will withstand need not be approached so closely. This ad- vantage should not be interpreted as a protective effect of the pyre thrum to the foliage, but as an additional insecticidal factor making the usual quantity of oil unnecessary. 52 HAND-PICKING OF CORN. Iii May, from the time when boll- worm injuries are first noticed in the buds of corn plants, the infested ones should be crushed in the hands so as to kill the worms found in them. To determine whether this could be successfully done, the method was tried while taking notes on the num- ber of worms and infested plants in a field during May and June. The result is given in Table I, and shows that of a total of 20 larvae, 23 (7 half grown and 16 very young) were crushed. It is therefore a sat- isfactory process. From the same table it is found that only 2.6 per cent of the plants showed injury. Hence but little time will be required to go over a large field in this manner. After an interval of two weeks, the process should be repeated. This will decrease the numbers of the later broods to such an extent that in many slightly-infested regions nothing further will be necessary, especially if infested ears of sweet corn be burned instead of simply thrown away. TRAP-CORN EXPERIMENTS. Experiment 1. A portion of a plantation owned by Mr. Dan. Nicholson was kindly set aside by him for a trap-corn experiment. The field was rich Red River bottom land, bordered on the east by a large forest, but surrounded on all other sides by cotton fields. Five rows were left vacant on the outer edge of the held, then eighteen rows of cot- ton planted, four more rows left vacant, then eighteen of cotton, and so on. The cotton was planted at the usual time. Two rows of each of the vacant strips were planted in corn April 4. May 7 this corn averaged about 1 foot in height. No boll worms were found in the buds of the plants, though in a held of corn some 300 yards away, which had been planted at the usual time, a few were collected. This field of early planted corn was near the garden and was surrounded on two sides by fences which were thickly grown over by dowering plants and dewberry vines. As no worms were found some distance from the edge of the field, it was evident that the first brood of moths had been somewhat attracted to the adjoining blossom- ing plants near the hedge and in the garden, and had confined their deposition to the outer edges of the field. This becomes an important factor when considering the feasibility of resort to killing the first brood of worms in the buds of corn by crush- ing. This does not apply to larger areas of corn whftre similar attractions are not near at hand. The trap corn was not so situated, but was in the midst of a large plantation, away from such early inducements. On July 3 a visit to the trap crop was made. It was, and probably for some days had been, silking profusely just as the second brood of moths was issuing. By July 6 the first planting had passed its prime in point of silking, though still in fit con- dition to receive the deposition of many eggs. At the time of the July 3 visit the following study of the number of larva? found in the young ears was made: Plant. Ears. Larvae. 1 2 6 2 1 1 3 •) 4 4 2 2 3 7 5 6 2 5 7 2 3 8 1 1 1 7' 0... 10 2 2 11 Total 1 2 18 41 53 The larvae at This time were Dearly all lesa than half grown, only two of fche num- ber being nearly gro^ n. These t\\ o w ere found alone in tin- ears of plants 2 and 8. In the ear of plant 9, which contained seven Larvae, all lesa than half grown, two were discovered being eaten by other-. By Julj 25, the second planting in the remaining t wo vacant rows was in its prime, but by August 1 had passed its best condition. The time of its greatest attractive- ness covered the period of the issnanee of tin- third brood. This brood deposited upon the fresh silks to such an extent as to product- an extremely crowded condition, for the larvse expected to find food upon snch a limited number of ears. Many ears were examined and all presented so nearly the same condition that only u lew counts were made. These were: Plant. Ears. Larvae. 1 2 3 10 8 •) :; Total (J 33 At the time of this examination. August 1. the larva' were still all very small. probably two-Thirds having never molted. In addition, the ears above noted for the larva- seldom bore leas than six to a dozen fresh eggs upon their silks, often ranging from a dozen to twenty. In the same Held, in ears in which nearly-grown larvae were found, only ;i few. if any, younger ones were present. This indicates that the crowded condition led the larger and stronger ones to prey upon the others, thna giving the victors more room and food. The ears of the first planting had now hardened, and no larva- were found in them and no fresh eggs were being deposited on their leaves and husks. Egga were still being deposited upon the plants of the second planting. The cotton between the rows of trap corn was carefully examined during the egg-laying period without finding eggs or bored bolls, even in the rows immediately adjoining those of the trap corn. Experiment :. Through the kindness of Mr. A. Curtis, of Curtis, La., a large and fertile tract of land, also in Red River bottom, was placed at my disposal for experimental pur- poaea. The cotton was planted at the usual time, one row for every fifteen being left implanted. This one row was planted in corn April i>. May 7 the rows of young corn were examined, but no larva- were found in the plants. A small held of crop corn, planted earlier and joining the experimental field on one Bide, had a tew worms in the buds of Borne of its plants. The Becond visit was made July :>. when the corn was found in splendid condition for egg-deposition. The following studies were made of infested ears: Plant. Ears, Lan a-. 1 •_> 5 1 1 1 1 t 10 l 1 •i (i 7 9 LO Total 1 43 54 A similar study, was made of the adjoining small field of rapidly maturing crop corn, with the following results : Plant. Ears. Larvae. 1 Bored* Bored Bored 1 o 1 1 1 Bored Bored 0 0 0 1 2 1 U 0 0 0 2 3 4 5 G . 7 8 1) 10 Total 11 4 *By bored is meant that a worm had been in the ear but had left, either for another ear or to pupate. Inasmuch as eggs were found quite plentiful upon the trap corn and none were found upon the other, it is apparent that the moths had chosen between the two. The number of plants and ears, such as the females would readily deposit upon, was counted. One row contained 148 plants with 267 ears. Each of the remaining rows was of the same length (about 10 or 15 rods) and contained approximately the same number of plants and ears. From the count of the number of worms in the ears of this trap corn, as above given, an average of 2.8 worms per ear is derived. There- fore the above row contained about 747.6 worms. For the eight rows of trap corn in this field, this makes 5,981 as the approximate number of worms trapped. This leaves out of consideration the unhatched eggs found in the silks at that time. May 23 a second trap planting was made, in a field immediately to the right of the first experimental field. By July 6 it had not yet tasseled, though it was badly in- fested with another species which was feeding in the buds, just as the Boll Worm does. Later, about the 1st of August, the second experimental field had silked and was well stocked with boll-worm eggs, many of which were parasitized. The larvae were plentiful in the ears, and as nothing of further interest could be attached to the experiment, Mr. Curtis cut the corn and fed it for forage. Experiment 3. [Mr. J. H. Fnllilove's plantation.] Corn was planted April 13. May 7 it was still small. No Boll Worms in the young plants. Two hundred yards away was a field of corn which had been planted much earlier. In this a few young Boll Worms were found. July 3 the ears of the trap corn were badly infested with Boll Worms and many unhatched eggs were upon the silks. The conditions in general were much the same as in the preceding, and need not be repeated in detail. Experiment 4. [Mr. S. J. Ziegler's Plantation.] One field was rather more upland and less favorable for a good growth of iate- planted corn. The first planting of corn was April 9. April 24 the corn was from 4 to 6 inches high, but contained no Boll Worms. It tasseled and silked subse- quently and the ears were badly infested. June 29 the second row was planted. July 28 the plants were 10 to 15 inches high, and had boll- worm eggs upon the leaves. The weather had been very dry during July, and the corn made an unsatisfactory growth, few plants producing ears with large flowing silks, 55 In another of Mr. Ziegler's fields corn was planted May 19. July 28 this was in line silk. By actual count the silk of a single car was found to" Lave twenty-five nnhatched boll- worm eggs. Most of the silks had only about a dozen eggs, with from three to six larvae in the ears. Late in .July notice was received from Mr. John Glassell, jr., a Leading planter at Rush Point, La., who had read the recommendations given by the Division upon the boll-worm question] and had prepared to test the suggestion. By his invitation the plantation was visited July 25, and a complete verification of our own experiments proved to he in waiting. Mr. Glassell had planted corn at the time of the second hoeing, when the cotton was ahout knee-high, or, as he informed me, about May 20. At the time of the visit the third hroodof moths was fairly issuing. The trap corn was in fine silk, and the record of a few of the many ears examined will suffice to indicate what they were accomplishing. One ear, 11 larvae, 7 eggs on silks; another ear, 6 larva?, 10 eggs on silks. The closest inspection of the cotton plants surround- ing this corn failed to reveal any traces of boll- worm injury. Various fields of corn near hy were examined hut no holl-worm eggs were found. The fresh silking corn was nearly in the center of a number of these fields and seemed to he receiving almost the entire egg deposition of the issuing brood in that immediate locality. Mr. Glassell enthusiastically accompanied your agent during all the observations, with a view of thoroughly informing himself of the facts and enabling himself to estimate the value of this method of protecting cotton. Subsequently he continued to make close observations and reported himself as being well satisfie-d with the remedy. In this connection it may also be stated that much valuable corroborative evidence was obtained from Mr. S. B. Mullen, of Harrisville, Miss., who had been advised of the trap-corn experiments. He arranged several small fields to make a test of the idea, and all of his reports by letter arc in entire accord with what has already been stated. The plantations thus far considered were bottom lands. The cotton in and about trap-planted fields was practically free from boll-worm injury. This could in a measure be said of other cotton fields in the valley, because the Boll Worm did not appear in destructive numbers during the season. This in reality does not affect the facts recorded for the corn experiments, and their significance relative to the moths which did appear remains the same. In the " hill country " of Louisiana and portions of Mississippi away from the river valleys, the Boll Worm is not noticed or feared much except during very destruc- tive years, when it spreads from the bottom lands. A small farm in the uplands west of Shreveport was prepared for experiment in much the same way as those in the valley. Corn was planted May 16. By dune 16 it was knee-high, but no worms were found. July 9 corn was tasseling and be- ginning to silk, but as no moths appeared in this locality, no eggs were found. To trap the first brood requires corn in silk from about May 15 to June 1. This is too early a date to be reached by the yellow or Dent corns. In its stead a sweet corn. commonly planted in the south for table use, meets the requirements. This corn had passed silking and was in good roasting ears before the first of June. Some of the studies made upon it are exhibited in Table II. which shows how badly if was and had been attacked. At the time of the count many nnhatched eggs were still to be fouud upon the silks. Care must be taken, however, not to estimate the abundance of the Boll Worm and the extent of its injuries from such examples. The Dent corns also make an unsatisfactory growth when planted late enough to bring silking about the first of August. In its stead the sweet corn again meets the conditions. The plan, therefore, to be recommended to the planter tor using the trap corn method of protecting his cotton against boll worm injury may be summed up as follows : When planting the cotton have vacant snips 56 of five rows for every twenty-five of cotton to be planted in corn. At the earliest possible time plant one row of this with an early maturing sweet corn. It should not be drilled in too thickly, since only a mini- mum number of plants and ears is desired. During the silking period of this corn frequent careful examinations must be made as to the num- ber of small white or brownish banded eggs, hardly larger than a pin head, found upon them, As soon as no more fresh white eggs are found each morning, the silks and ends of the ears should be cutaway and fed or burned in order to destroy the young worms and the eggs. A few eggs may be on the leaves of the plants, and since no more growth is to be made, they also should be cut and taken from the field. There is no reasonable objection to this method of handling the first planting, since the natural enemies are not yet numerous and the egg parasites appear iu greater numbers during the egg-laying period of the next brood. The next planting should be three rows of Dent corn, drilled in late enough to bring the silking period about the first of July or a little later. These rows catch immense numbers of eggs and larva?, but should be left to mature in order that the natural enemies which parasitize the eggs and prey upon the larvae may not be de- stroyed. Furthermore, the cannibalism previously discussed, which occurs in this corn under such crowded conditions, reduces the number of worms reaching maturity to a minimum, and these can well be al- lowed to escape if the natural enemies be saved thereby. To trap these escaped individuals, the fifth and last row of the vacant strips should be planted to sweet corn at a time calculated to make it reach full silk about August 1st, when the moths begin issuing again. This expedient allows the planter to save the second planting as a crop. The corn produced in this way is large enough in quantity to pay for the expense of cultivation and management and the sacrifice made in cropping the five rows with corn instead of cotton. However, it must be understood that this is immaterial so long as protection is afforded to the surrounding cotton. The last row of sweet corn should be care- fully watched. If it is found that a great many eggs are parasitized, a fact which is indicated by their uniform grayish or blackish color, it may be as well to allow it to mature as before and thus save the parasites. If this condition is not found, the corn should be cut and taken from the fields as soon as it shall appear that no more eggs are being deposited. If the first two plantings are well managed, the number of the earlier broods will be so reduced that the August brood will not be capable of inflicting great injury, and in less infested regions the third planting- may even become superfluous. It is not necessary or advisable to crop the entire plantation with corn and cotton as recommended. The end will be attained if five-acre strips of alternate corn and cotton be planted for every fifty acres of Bull 29, Division of Er.tomo'ogy, U. S. Dept. of Agricultur Plate I. CO _, cc ? Q £ < o o m o cO or ^ o CO z 3i It oO CD z o O 1— were to be made was not plentiful and the weather conditions were such as to ob- struct progress at every step. The laboratory was not complete enough for the most extended and exlhaustive researches, and the time at command was considerably divided in attending to other portions of the investigation. No noteworthy discoveries were made and no reliable ones could be rightfully expected in so short a time. So far as the strictly bacterio- logical work is concerned, it has just reached a satisfactory basis for exhaustive studies along the lines which the results of the investiga- tion indicate as the most promising. The studies were conducted as directed upon the practicability of ar- tificially utilizing the germs of insect diseases as remedial agents. Ac- cordingly the germs were isolated as pure cultures by the usual methods and artificial infection experiments made to ascertain the facts. The results as such are entirely satisfactory, though in no sense solving or setting at rest the problem under consideration. Yet. if properly in- terpreted, they contribute valuable suggestions relative to the basis upon which the problem should be considered, or a solution attempted. The results can not rightfully be takeu in a negative sense except in respect to the method and the basis upon which they were obtained. Practicability having been the object in these studies, only such experiments and observations are presented in this report as bear di- rectly upon that phase of the problem. The minutiae of some new methods of staining the germs, their specific descriptions, and like mat- ters, are entirely omitted, since, for the purposes of this report, they might be confusing and misleading. If this discussion contributes in any way towards freeing the minds of some from misleading and un bacteriological opinions concerning' the problem, or assists in putting future efforts on a more scientific basis, it will serve as great a purpose as our present knowledge of the specific organisms and the attending difficulties involved will permit. 60 GENERAL PRECAUTIONS, It will be unnecessary to enter into a detailed description of the laboratory and apparatus used, for both were such as are always re- quired for preliminary bacteriological studies. In general it can be stated that all the customary cleanliness and precautions Avere success- fully observed. The apparatus was thoroughly cleansed after using, and either disinfected or sterilized. Glassware requiring it was placed in sulphuric acid for a time, subsequently washed, rinsed in alcohol, and sterilized. Test tubes in which cultures had been made were first filled with water, again plugged, and boiled for a couple of hours, kill- ing the germs and lessening the danger of accidental infection from es- caping spores. After boiling, the tubes were washed quite clean in water and placed in sulphuric acid over night. The following day they were washed, rinsed in alcohol, and sterilized. When making trans- fers of cultures from old to fresh media, the needles were always first dipped in acid and sterilized, then in distilled water, and again ster- ilized. To some these may seem to be extreme precautions, but the fact that the sterilizing, filtering, and culture inoculating was all done in the same small room, fully justifies them. That cleanliness and thorough disinfections were constantly practiced, may be concluded from the fact that at no time were any stock media lost through acci- dental infection or faulty sterilization. At no time was a culture lost through accidental contamination. The incubator was provided with a thermostat, and the temperature controlled at will for any given purpose or set of conditions. The infection experiments were carried on in another portion of the city. Two large rooms were fitted up, thoroughly cleaned and fumi- gated. In one the experiments with the particular microbe under study would be carried on, in the other the check experiments. Six-inch flower pots, covered with netting, were used as cages. These were thoroughly washed with a disinfectant before being employed in any experiment. For each experiment a different pot was used, to avoid the danger of mixing the germs. After each experiment, the room was thoroughly fumigated before another was begun. CULTURE MEDIA. Many media could be profitably experimented with in the study of reducing the problem of insect diseases to a practical basis. When, however, immediate practical results are wrongly considered the pri- mary objects and experimentation is inaugurated upon that basis, it becomes impossible to use, at first, more than a few of the standard media. Those used in this work were beef broth, broth agar- agar, broth gelatine, and potato. The two most extensively used were beef broth and broth agar-agar, and for the purposes of this report it will 61 1)0 sufficient to consider only these two. The following is a brief ac- count of your agent's experience with, and methods of preparing, these two media: BEEF BROTH, The formula is the one most frequently used by Dr. S. A. Forbes and Prof. T. J. Burrill, of Champaign, 111.: One pound of round steak, free of fat. is chopped fine, placed in 1 quart of water, soft preferably, and allowed to stand over night. The next morning the meat is pressed dry. It is well to pour some of the liquid back on the meat, stir up thoroughly, let stand for half an hour, and press again. Strain the liquid through cheese-cloth, measure, and add enough to make the original quantity (1 quart). Pour into a liask. boil in steam, sterilize for an hour and a half. Strain through cheese-cloth or white flannel, niter, and allow to cool. Measure, and if necessary add enough dis- tilled water to make 1 quart. When about 60° C, neutralize with sodium carbonate (or if alkaline, with lactic acid). Cool to about 45° (_'.. and allow to stand for half an hour. Filter. Boil for au hour, cool to 60°, and filter through double thickness of best German filter paper. Sterilize for an hour, and let stand over night. If sediment forms, filter while cold. It is now safe to fill test tubes and proceed with three dis- continued sterilizations on as many successive days. In test tubes the sterilizations need not be continued for more thau twenty minutes. In large quantities an hour or more is required. The addition of the neutralizing agent often makes the liquid turbid. Added a little at a time and the liquid shaken, this cloudiness disap- pears. If so, it only indicates that the liquid is not yet neutral, ks the point of neutrality is reached the cloudiness disappears less per- fectly upon being shaken, and finally not at all, gradually forming a light, floeculent precipitate. The task of obtaining and retaining an absolute neutrality is a difficult one and the reaction just described, if carefully noted, will be of great assistance in making a delicate test. Some recommend the use of the white of an egg to assist in clarifying the broth. The method already detailed was so satisfactory that egg was used in only a few instances and then more as an experiment. It was found, if the broth was neutral or alkaline when the white was added, that it coagulated imperfectly when boiled and caused considerable difficulty. The broth had to be acidulated and then boiled to produce the proper coagulation. The filtrate was dear at first, but the process of neutralizing produced the same effect as to cloudiness and tine sedi- ment as already explained. The white of old eggs is somewhat more liquefied than that of fresh ones, and when used in a quantity of slightly acid broth it was difficult upon boiling to produce perfect coagulation. This merely emphasized the fact that only fresh eggs should be used in the work. 62 AGAR-AGAR. Ill the preparation of this, medium beef broth prepared as already described was used in every instance. For the most part, peptone did not seem to be required in the preliminary studies, and no time was spent in using it to determine additional differences in the growth ot the microbes studied. The agar-agar was finely cut before being placed in the broth to soak. It was found that the difficulty as to cloudiness in the media could be greatly lessened by soaking the agar-agar in water for a time and thoroughly washing before placing it in the broth. With these preliminary explanations, the following may be given as the formula, which is also the one used by Prof. Forbes and Prof. Bur- rill: One quart beef broth. Ten grams agar-agar. Five grams sugar (yellow clarified). Five grams salt (druggist's best). After shaking well, allow to staud and soak over night. The following morning boil for three, hours. Strain until clear; cool to about 60° C. Stir in the white of an egg and boil until well coagu- lated. Strain until clear; neutralize if necessary; keep hot without boiling, and allow to stand for fifteen minutes. Filter; sterilize for an hour. If sediment forms, filter again; sterilize for another hour, and let stand over night. If upon warming the next morning a sediment forms, filter again, after which it will be safe to fill the test tubes. These are then further sterilized the same as beef broth. After the last ster- ilization of the tubes the wire cage containing them should be laid on an inclined plane, so as to give a slanting and therefore greater surface in the tube for the growth of the germs. The agar agar medium sometimes looks slightly cloudy while yet hot or upon being heated, but, as in the broth, this disappears upon cool- ing. SPECIAL APPARATUS. The filtering of agar-agar and other solid media is often attended with great difficulties in winter, since the hot liquid cools and thickens so rapidly. For this reason an apparatus for hot filtering is necessary. A separate appliance requiring additional gas and burners is in com- mon use. Your agent was compelled to secure the necessary apparatus speedily and economically, and accordingly the following combination of the steam sterilizer and hot filter was devised (see Plate n, Figs. A, B, C, D, E). The lower portion (A) of the sterilizer was made as usual; this particular one 10 inches in diameter and 1 foot high, with three circles of tips («, «, a) on the inside, on which to lodge the perforated diaphragms (B) at various heights above the water. The legs (b) were high enough to allow an ordinary two-burner oil stove to Bui1. 29, Division of Entomology, U. S. Dept. of Ag Plate II. a-- a-- Special Apparatus for Bacteriological Work. 63 be placed under. The top or lid {<■) was made 8 inches high, of the same diameter as tlie lower portion, fitting into it tightly, so as to avoid the escape of steam and decrease of pressure as much as pos sible. Tli rough the center of this cover was fitted a collar (d) in the shape of an inverted cone, about '■> inches long. I inches in the larger and 2J in the smaller diameters. Fortius collar a tight fitting lid like that for a tin pail was provided, in order that the same top might be l'scd either for hot filtering or simply for sterilizing. To one side of the funnel collar, in the top of the lid a second small collar was fitted. for the reception of a thermometer. On the side a slender, slightly bent handle for lifting the top off and on was placed. When any hot filtering- is to be done, the flask containing the me- dium is placed in the sterilizer and brought to a boil. At the same time a second empty flask is put in on the diaphragm. The top is then placed on. The funnel is provided with the necessary filter paper and the whole inserted through the collar (d) in the top of the sterilizer. The steam arouud the funnel keeps it hot and that escaping through the neck moistens the filter paper. When the liquid to be filtered reaches the boiling point, the flask containing it is taken out. the hot funnel at once fitted through the collar, into the empty flask, inside the sterilizer. The liquid is then poured into the funnel and the filter- ing proceeds without further interruption or special care. The body of the funnel being inside the sterilizer the steam has full play upon it. keeping it and the liquid almost at boiling during the entire process. At the same time other flasks containing media can be placed in and sterilized while the filtering goes on. The flask receiving the filtrate being in the sterilizer, any danger of falling germs or spores in the air settling upon the liquid is avoided. The funnel collar should not extend more than a half inch above the level of the top, so as to -allow almost the whole of the funnel to be inside the sterilizer. For some media the pressure of the steam through the funnel checks the rapidity of the filtering. To avoid this a ruffled collar, instead of a perfectly circular one, can be made, thus allowing the escape of steam and relieving the internal pressure. The lid to the collar must be made to fit accordingly. If desirable a plain top(D) can be made for ordinary use in sterilizing, in which case the lid to the funnel collar in the other is not necessary. The respective dimensions must of course be adapted by each maker to his particular purpose. To prevent the radiation of heat as much as possible, the sterilizer is covered or bound in the usual manner with asbestos (g). Another piece of apparatus, which may be called an ••incubating cage," Fig. F. was also devised which in many respects materially facilitates work. The cages in which culture tubes are usually placed when transferred to the incubator are the well known wire cages, hold ing some twenty or thirty tubes. This entails a great inconvenience 64 when many cultures are in consideration, since neither the labels on the tubes nor the nature of the growth can be readily seen without taking out each individual tube. This difficulty is overcome in the new cage, the frame of which consists of wooden strips three-eighths or one-half inch thick, and about 1£ inches wide. The two upright ends (a, a) should be about 4 inches high, with grooves (b, b) cut along each side into which a plate of glass, c, can be slipped. The two up- rights are dovetailed into the horizontal piece (r7), pegged and firmly glued. This done, the two plates of glass are inserted, the bottom covered with cotton to the depth of half an inch, and the tubes placed in as shown in the figure. In doing so the slanting surface of solid media should be turned to the outside and the label placed on the same side. In this way no difficulty is experienced in speedily finding any tube desired and watching from the outside what progress any growth may be making. The cage should be wide enough to receive two rows of tubes, as then there is less danger of its falling over so easily. They can be made any desired length, and the uprights to any height de- manded for the best tubes in use. OBSERVATIONS AND EXPERIMENTS. June 11 a Boll Worm was placed in a cage to rear in confinement. It fed until June 13, when it entered the earth for pupation, but died in the attempt, June 15. The anterior part of the body began to decay and then darken. At the decaying portion a cut was made dorsally with the proper precautions, and a brownish golden-colored liquid is- sued. A drop of this was transferred to a tube of broth and a liquid culture made. From this in the usual manner pure cultures were ob- tained on solid media. The posterior portion of the body did not de- compose so rapidly, and though rather spongy, retained its natural color for some time. As the rotting proceeded, the color changed to a brownish or darker color. The germ which probably caused death changes beef broth to a de- cided white turbidity, with scant white deposit at first. As the growth advances the deposit becomes more abundant and the liquid begins turning greenish. Finally, the broth clears and is a beautiful deep green, with plenty of white sediment at the bottom. On agar-agar the growth is very thin and scant, beginning by numerous small, ir- regularly roundish, almost colorless colonies. They gradually spread a little, and if numerous enough form a thin, rather granular-appear- ing white film. The first pure cultures on solid agar media give the medium a faint, greenish tinge, but this power seems gradually to weaken with subsequent cultures. The germ was found to be quite sensitive to artificial cultivation, and doubtless loses much of its origi- nal power by such a process. In some respects the growths upon agar- agar and beef broth are quite similar to those of the cabbage- worm 65 (Pieris rapw) disease, but a microscopic examination shows the former to be a rather small bacillus. A similar observation was made June 10, when one of a lot of Boll Worms kept in a breeding cage for life-Jiistory purposes was found dead. A bacteriological study was made. The alimentary can al seemed to be the only portion of the body containing much liquefied matter, the fatty portions being rather slow to decay. Pressure of the decay- ing anterior third of the body forced out a drop of a rather golden- colored liquid, from which a broth culture was made. At the same time a pro-leg was snipped off with sterilized scissors, a platinum needle in- serted so as to miss the alimentary canal, and a second tube of broth inoculated. From each pure cultures upon agar-agar were isolated. In the beef broth the changes were the same as just described in the preceding study. Upon agar-agar a more profuse and vigorous growth was obtained, which was partly due to the fact that the tubes had been more recently prepared and were not so dry as in the first study. The film was smooth and white, with margins entire though irregular in some portions. This affection of Boll Worms is not very prevalent, though occasion- ally one is found in ears of corn dead or dying. From these in most cases the germ just considered can be isolated by the usual pure cul- ture methods. When affected, the larvae seem to lose their appetites, cease feeding, become rather sluggish, and appear somewhat disturbed. The color of the skin remains either partly or entirely normal, occa- sionally even for a time after death. At the same time, however, the tissues of the body are decaying and becoming watery, more especially along the alimentary tract. This condition at last imparts a grayish- brown or rose-tinted color to the body. Both cultures of this boll- worm bacillus were made from the pure ones on agar, and allowed to grow for eighteen days, when they were used in experiments 1, 2, and 3, which follow. Experiment 1. July 8 {5:30 p. m). — The husks of an ear of corn were torn aside and the silks and grains for a considerable space were well washed with the broth culture of the ba cillus. One nearly grown Boll Worm and one half grown, were placed within the husks, after which these were well closed down upon the ear. The ear was kept in a pot prepared as heretofore explained. The following day both larva- had fed freely upon grains of corn which had been drenched with the broth culture. No unfavorable symptoms. The second day the large worm had left the ear and entered tho earth for pupation. The small one was still feeding but showed no unhealthy symptoms. The third day the young larva molted. After this it continued to feed in the ear, pupating there and compl< its transformations by issuing as a moth July 27. The first pupa had hatched a few days earlier. l\xp> Hment P. During the same period of time four cabbage worms (Fieri* rapm) were fed upon a cabbage leaf which had previously been well drenched with a portion oi the broth 14935— So. 29 5 G6 culture used in experiment 1. One of the larvae was almost grown, one about half grown, and the others younger. The following day, July 9, the drenched leaf had been almost entirely eaten up. They were left to feed upon the remains until the second day, when a fresh leaf was placed iu. Up to July 14 no symptoms of disease appeared in any of the larvae, and on that day the last two pupated. July 15 the two oldest pupae died. One of those had heed noted as turning darker the previous day as if beginning to rot. To-day its wing-covers and head are entirely black, while the abdomen practically retains the normal color. The other dead pupa is entirely of a uniform dusky color. The two living pupae were lying just alongside the two dead ones, and were thoroughly ex- posed to infection, if any. Both, however, hatched, one on July 18, the other July 22. Experiment S. July S. — The culture liquid used was the same as in experiments 1 and 2. A small cabbage leaf was drenched and four Pieris rapce larvsB placed to feed. Two of (hem were nearly grown, the others about half grown. By July 10 the leaf had been en- tirely eaten, but no symptoms of disease were noted. Fresh leaves were placed in July 13. July 15 two pupae were found, one being imperfectly formed. The two remaining larvae fed freely, but did not seem to grow as rapidly as usual. At times their skin seems to be somewhat puckered and appears rather dusky. July 16 the ill-formed pupa is dead. July 18 the last larvae pupated. July 27, without any apparent outward changes to forewarn such a result, it was found that all the pupae had died. About the time of death, or soon after, the color becomes slightly brown- ish or dusky. The special attention due this experiment was frequently interrupted and fresh food was not provided the larvae as often, perhaps, as was conducive to their best development. This may have induced them to attempt pupation rather prematurely, or have weakened them so as not to be able to cope with the germ. Checks on experiments 2 and 3. The larvae in experiment 1 having completed their transformations without diffi- culty, a consideration of its check will not be necessary. For experiments 2 and 3 a number of Pieris rapes larvae were placed upon cabbage leaves in a separate jar to act as a cheok. July 9 the following was the condition of the larvae in the check : 1 pupated, 3 pupating, 4 grown, 1 half grown, and 3 younger. Up to July 16 the younger larvae had kept on feeding perfectly, and succeeded in maturing and pupating. Two adults issued on this day, and one pupa, which had been injured a few days before, was dead. July 17, 8 pupae remained. Two had become darker in color, as if begin- ning to decay internally. Later these 2 were found to be certainly dead, the one having turned quite blackish, the other more brownish gray. The other 6 hatched. One of the dead pupae of experiment 3 was taken for further study. The contents were a blackish liquid mass, from which a drop was taken with which to inoculate a tube of broth. From this other liquid cultures were made, and from these pure cultures upon agar-agar were obtained by the ordinary process. One of the dead pupa> was taken from the check for a similar study. Its contents were of the same nature as of the one just noted. In the same manner liquid cultures, and from these pure cultures upon agar-agar were obtained. A careful comparative study proved that the pure cultures obtained from the two pupae were identical, and a microscopic study developed the fact that both were cultures of the Micrococcus of the cabbage- worm disease. Accordingly the pupae in experiments 2 and 3 did not come to their deaths solely through the agency of the boll- worm disease, though the greater per cent of deaths in the experiments, as compared with that of the check, would indi- cate that the latter germ contributed in some manner to this end 67 Some diseased cabbage worms were received October 4 from Prof. C. P. Gillette, Ames, I«»wa. From one of these larva- a pure culture of the PieiHe rapce micrococcus was obtained. In this condition it was kept in a healthy growing state during the winter by frequent transfers to fresh media. In this manner the germ had been transferred eleven times, nine times on agar-agar and the la>t two in beef broth. The eleventh cultnre was used in the experiments August 20, alter having had about ten months of artificial cultivation. The culture was two days old when used in experiment 4. Experiment 4. A itf/Hst 20. — The culture liquid just spoken of was applied as follows : Two small bolls with involucres were well drenched in the liquid and two half-grown Heliothis larvae were placed on them. The larvae began sipping of the liquid, which insured their infection if possible. A small round cavity had been cut into the bolls and filled with the cultnre liquid. The worms decided to enter the bolls at these injured points, again exposing themselves to infection. Both continued healthy and hvK freely, bo much so that one fell a victim to the other through cannibalism. The survivor continued healthy to the last, pupated, and batched later as a robust, active moth. In isolating the cabbage-worm micrococcus from the diseased larva:' received from Profs. Osborn and Gillette, two other germs were isolated. On agar media the one produces a yellow growth, the other a beautiful pink one. In all the pre\ ions and subsequent studies the germ producing the pink growth was almost constantly ob- tained from diseased cabbage worms. It a\;is therefore thought advisable to give it a trial upon the Boll Worm. A broth culture was made and allowed to grow for two days, when it was applied as detailed in experiment 5. The germ had been carried over winter by artificial cultivations for a period of eight months and was t be tenth pure culture. On agar-agar the growth may be described as follows : At first small elevated round colonies having a translucent whitish appearance. These gradually spread and fuse, forming a continuous white growth. If it continues growing from the margins. these may be finely fringed, slightly branching or corrugated. As the growth be- comes thick, the surface becomes very much wrinkled or ridged. At this stage, and often earlier, i lie growth begins t inning to a pinkish color, finally becoming dial iiutly pink. The pink color appears in smooth growths or isolated colonies, as well. seemingly, developing as the germ ceases its most vigorous growth. The wrinkled scum seems rather to be evidence of a vigorous culture and the result of a s. i\ pro fuse growth. Experiment 5. August 15. — The husks of an ear of coin were torn away just enough to expose the silks and grains of corn. The culture liquid was then poured on the tip of the ear and allowed to soak in through t lie silks and run down t he length of t lie ear. One large Boll Worm and one half grown were in the ear. The liquid came into contact with both, and each was seen to sip of it. The following day the Larvae had eaten plentifully of the corn, including most of the grains w liieh bad been drenched w ith the charged liquid. Both larva' continued to feed, the larger one pupating and hatching later. The smaller one led for a time longer, during which no Unfavorable symptoms appeared, bu1 finally made good its escape from the pot while searching for a more desirable place to pupate. The disease ol Pieris rapce is found occasionally in most portions of Louisiana, but it is not of a virulent form in most cases, not causing deatb until the pupal stage is reached. June8 two dead pupae of this species were found upon cabbage plants in the field, A careful study 68 proved that they were not parasitized, and had not been injured. Pure cultures on solid media were isolated from the germs found in the liquid contents of the pupal skins. One of these germs proved to be the cabbage- worm micrococcus. In most portions of the South the dis- ease affects only a small percentage of the larvae, and as it is usually fully developed only in the pupal form, the contagion among cabbage- worms is reduced to a minimum. A DISEASE OF PLUSIA BRASSICJE. The first symptoms begin to appear about the region of the two white lateral patches just below the median line and over the first pair of pro-, legs. The patches look like whitish, cheese-like fatty bodies under the skin. From these the pale cream color of the body begins and spreads, the skin gradually becoming entirely of a lemon-yellow color. The pos- terior portion of the body shows these symptoms first, the anterior por- tion remaining quite natural in color until about the time of death. No fluids appear to issue from the mouth or vent during the course of the disease. When well affected by the progress of the disease, the larva ceases feeding, dying soon afterward. The entire body deliquesces very rapidly after death, producing a blackish, semifluid mass suspended in a bag of grayish skin, which finally bursts and allows its contents to escape. September 4, some living Plusia larvae were found on a cabbage leaf near a dead Plusia larva, which was already black and entirely de- liquesced. Two Plusia larvae and two of Pieris from the same plant were placed together in a collecting box, and later placed in the same breeding cage to rear. By September 7 the Plusia larvae had died and deliquesced. The Pieris larvae had certainly come in contact with the sick Plusia while crawling about and feeding upon the same cabbage leaves, and had thus been thoroughly exposed to infection. Both larvae, however, completed their transformations, and the butterflies showed no unfa- vorable symptoms. This experiment was repeated with a greater number of larvae of each species with exactly the same results, From this it becomes evident that the Plusia disease could not be very contagious so far as Pieris rapce was concerned; at the same time the disease acts very decidedly and rapidly among Plusia larva3. They often begin turning pale cream -colored, then yellowish, dying, and the body deliquescing, all within thirty to forty hours. This applies to nearly grown larvae, Those less than half grown succumb in half that time, In the usual manner pure cultures were obtained from the dead and deliquescing larvae, Three distinct germs, two of which were found almost constantly in the several specimens from which cultures were piade, were isolated by the usual process. On agar-agar one of these germs produces, at the beginning, numerous small, white roundish 69 colonies, which gradually spread and form a thin, white granular film. margins wavy or sometimes slightly corrugated. The growth has a slight tendency to liquefy at a certain period of its development. The second produces a pink growth, such as has been noted and described in considering experiment 5. The third is a profuse beautiful yellow growth, beginning at first by dense, thick round colonies, rather whitish at first, but soon turning yellow. When fused and the growth pretty nearly completed it is nearly always quite thick and deep yellow, with margins entire or wavy. The first and third of these are the ones which seem to be constantly associated with the disease. The one producing the thin, white film is the one which is parasitic, or at least partially so, in its relations to Plusia brassicce. Pure cultures of this germ were also received June 3 from Dr. J. 0. Neal, Lake City, Fla. Upon unsealing the tubes a small amount of gas escaped with a fizz, accompanied by a smell reminding one of rotten eggs. Fresh agar cultures were made, and from these, liquid cultures were prepared for use in experiments C and 7. The culture liquid used in these experiments was eighteen days old. Experiment 6. June 27. — A cabbage leaf was drenched with the charged liquid, and four Pieris larv* placed on to feed. The day following; all were feeding briskly. Two days later the leaf had all been eaten and fresh food was placed in. No indications of disease three days afterward. Later three larvae pupated, one of which was acciden- tally injured and died. The fourth larva died, but did not rot or turn dark. It dried up gradually, which indicated that the pot had been too strongly disinfected and that the larva had been poisoned from crawling about its walls. No cultures were subsequently obtained from it. The two living pupa} hatched in due time. Experiment 7. June 27. — Two bolls were prepared as described in experiment 4, but using the same culture liquid designated for experiment 6, namely, the Plusia disease genu. Two Boll Worms were placed on, and each was subsequently observed sipping of the liquid. One of the larvae was full grown, and had shortened some, preparatory to pupation; the other was also about full grown, but fed until mature. Both sub- sequently entered the earth and completed their transformation without any diffi- culty. Checks during the progress of the experiments recorded, and many others were carefully continued. Since no results were obtained from the experiments requiring it, the consideration of the checking will, in this report, be superfluous, save to remark that disease did not appear in them in any instance, except experiments 2 and o, which have already been included. DISCUSSION OF THE RESULTS. Prom the beginning complicating conditions were discovered. The most important one was that the species in question (Heliothis armiger) was subject to a disease which was probably as prevalent as the nature 70 of the case permitted. Great results had been anticipated by some from an introduction of the disease of the Cabbage Worm (Pieris rcvpce) as a remedy for destroying the Cotton Catterpillars (Aletia xylina) or the Boll Worm. Upon investigation it was found that this disease had already been introduced with its host through the natural dissemi nation of the disease from the locality of its first appearance. The third complication arose when it was observed that about 80 per cent of the larvse of Plusia brassicce, a very common cabbage insect in the South, were dying of disease. The importance and relation of these three conditions to each other will be better appreciated by bacteriol- ogists when it is explained that the system of small negro tenantry, which is customary in the greater portion of the cotton district, results in numerous small garden patches along the edges of, and often within, the centers of the fields. The one vegetable which can safely be pre- dicted to be present in nearly all of them is cabbage. These plants were always infested with either Pieris rapcv or Plusia brassicce, or both, and concerning both it was known that disease made its appear- ance. The Boll Worm and Cotton Catterpillar were therefore con- stantly exposed to the danger of infection. As a consequence, in the studies for the artificial infection of the Boll Worm, the following sources of error required elimination: First, infection through its own peculiar germ ; second, through that of Pieris rapce, and, third, through that of Plusia brassiew. This could be best guarded against by deter- mining, as much as possible, the relation of the three germs to the three insects involved. Before these points are discussed, it is advisable to dwell upon some other conditions of environment which will contribute to a better interpretation later. It is asked, why does not the disease of the Boll Worm itself spread more freely? The diseased Boll Worms, with few exceptions, were found in ears of coin. Here, as has been stated in the first part of this report, a struggle for food, due to crowded conditions, may and often does occur. This compels more or less traveling in search of suitable quarters. This in turn increases exposure to all sorts of unfavorable conditions, including the attack by parasites and natural enemies, as well as disease. Further, having probably fought for its freedom, the larva is forced to change at a time when it is weak and least fit to resist such conditions. It is under these conditions that the Boll Worm some- times falls a victim to disease, usually, however, getting under the cover of another ear before dying, thereby lessening the chances for its infecting others of its own species. In view of these facts, the trap-corn method recommended in another portion of this report becomes an im- portant factor, in that for this species it furnishes those conditions which are iavorable for the greatest propagation of disease. When the Boll Worm infests cotton, the chances for infection are even more diminished, in that each individual becomes a hermit in addition to feeding on the inside of the cotton boll. Upon cotton they seldom 71 come in emit act a\ ith each other, and then for a brief time only. There- fore, should a Boll Worm become diseased upon either com or cotton, the natural conditions and habits of the past are such thatthe chances for infecting- other individuals through it arc reduced to a minimum. This also explains the failure of the disease of Pieris rapce and Plusia brassicce to attack the Boll Worm, and spreading to it through the nat- ural processes of infection and dissemination. On this point, however, another consideration must be noted in the case of Pieris rapce. From the observations already recorded for this disease, it is found that, though present, it developed rather tardily in its host under the pre- vailing conditions. In addition, it appears to be less virulent and ap- parently has less power of contagion, since it does not seem to infest others of its own species so readily as in more northern districts, such as Illinois, Indiana, and Iowa. This seems to be due mostly to the dif- ferences in climatic conditions, the atmosphere being drier, much higher in temperature, and the hot summer season much more protracted in the South. The well-known devitalizing effect of hot, scorching sun-light under high temperatures upon many bacteriological organisms seems therefore to explain the lesser virulence of this germ in the locality where the investigation was prosecuted. Accordingly, the Pieris dis- ease is unpromising at present as an agent in destroying the Boll Worm in that section. The germ is doubtless becoming more acclimated and adapting itself to prevailing conditions, so that it may be expected to become more efficient in that region in the future. For the Plusia disease, however, the high temperature seems to be a necessary factor, and, so far as the writer's information goes, is less virulent in the cooler or northern districts. Experiment 1, in which a pure culture of the boll-worm germ obtained by artificial culture methods wasted to healthy Boll Worms, tailed again to produce the disease. The same genu was ted to larvae oi Pieris rapce as detailed in experiments 2 and -'5. In experiment 2, 50 per cent died. In experiment 3, all died. Subsequent studies of the dead pupse in these two experiments, as also the records of the checks upon them, together with microscopic examinations, proved that death could not be attributed to the boll-worm germ with any degree of cer- tainty. On the contrary, death seems to have been due to their own specific germs, as noted in the experiment. Reversing the trial, the germ of Pieris was used in experiment I. ami ted to the Boll Worms without producing disease. The Plusia germ was then led to Pieris larvae as in experiment 6, and to Boll Worms as in experiment 7, with- out bringing about diseased conditions. Furthermore, Pieris larvae, feeding upon the same plants and leaves along with diseased Plusias, did so w ith perfect immunity . What does it all mean 1 It is unsafe to hazard any positive state- ments and the discussion must be understood as being provisional. Granting that the germs in question are trulj parasitic upon then re 72 spective hosts, the first important fact indicated is that they are em- phatically specific as to the conditions required for their development. If this be so, the great differences in the life constitution and food of the three species of larvae under consideration would at once render mutual intercommunication of their respective diseases impossible. The theory held by some that a parasitic germ is readily transmissible from one species to another with power to produce disease, must be dismissed. Experience has shown that producing disease by artificial means in one species furnishes no guarantee that the same germ can in like maimer be used to produce disease in a nearly- related species and certainly not for those of distant relations. Actual experiment may prove it to be possible, which should therefore be done before any assertions are justifiable. The behavior of the germs in question, under the artificial culture conditions recorded in the experiments, indicates that they are faculta- tive rather than true parasites. This means that the germs can and do under certain conditions, develop as parasitic organisms, but under unfavorable conditions can undergo their development in other than living matter and thus tend toward saprophytism. Accordingly they may gradually adapt themselves to being more saprophytic or more par- asitic, whichever the prevailing environment may favor. This is quite certainly the nature of these organisms in relation to species of insects other than the one which for convenience may be called the natural host. Therefore the apparently negative results shown in the experi- ments are negative only as concerns the utility of the germ when used in the facultative condition in accordance with the usual method of procedure. The germs being facultative in their nature, cultivations on artificial culture-media begin at once to weaken their power to pro- duce disease. When a facultative organism, therefore, is used in the usual manner to produce artificial infection, failure is rat*her to be ex- pected, and it is manifestly erroneous to consider the results as having any direct bearing upon the practicability of parasitic organisms as remedial agents. The only interpretation which should be given the results recorded in the preceding experiments is that to the insect in question (Heliothis armiger) the germs cultivated and experimented with, bear only a facultative relation. This fact suggests the abandon- ment, as a primary method, of the generally accepted one for experi- menting with germs in the attainment of practical economic results. This consists in the simple isolation of an organism as a pure culture, feeding it to a given insect, and passing final judgment according to the results which follow. It further suggests that before the question of artificial infection can be satisfactorily solved, the germ used, whether really parasitic or only a facultative parasite, must first be studied in all its relations to environments which allow the organism to produce dis- ease. This done, the next step will be to determine how best to con- trol those conditions by artificial means, either in relation to the host 73 itself or for fixing upon the microbe a greater power for infection or a cooperation of both upon the same basis. Either cause would result in attaining the greatest infection. In preparing pure cultures the records show that in the process of isolating the desired germs from the dying host, at least one and some times two additional well-defined germs were obtained, which were either associated or coincident with the disease. It is an interesting and important study to determine what are the relations of these germs to each other, either preceding or during the progress of the disease. Primarily the results of the observations and experiments develop the following facts: (1) That the germs experimented with are only facultative in their relation to the Boll Worm. (2) That, as such, in the manner cultivated and in the condition applied, they fail to produce results which are of primary economic importance. (3) That such failure has no primary bearing upon the availability of strictly parasitic organisms to assist in producing infection by artificial methods and obtaining practical economic results. Secondarily they rather definitely suggest the following general prop- ositions : (1) The importance of giving the most exhaustive study to ascertain what environments, as to both the insect and the germ, are favorable or unfavorable to the infection and development of disease among in- sects. (2) The importance of first determining the biological character of the organism, /'. e., whether truly parasitic or only facultatively so. (3) The determination of the first and second specifies to a great ex- tent the further method of procedure, and the basis of experimenta- tion. (4) The importance of the three preceding considerations demon- strates the folly of attempting to obtain practical results by pure culti- vations and artificial disseminations, purely as such. (5) That making the attainment of practical results the primary ba- sis of such investigations is a mistake, and an obstacle to real progress in their final attainment. This suggests that — (6) The biological and physiological properties of the germs, together with their environments, should first be studied and determined upon a purely scientific basis, without regard primarily to the attainment of practical results,